Angiogenesis inhibitors for the treatment of epithelial ovarian cancer
Abstract
Background
Many women, and other females, with epithelial ovarian cancer (EOC) develop resistance to conventional chemotherapy drugs. Drugs that inhibit angiogenesis (development of new blood vessels), essential for tumour growth, control cancer growth by denying blood supply to tumour nodules.
Objectives
To compare the effectiveness and toxicities of angiogenesis inhibitors for treatment of epithelial ovarian cancer (EOC).
Search methods
We identified randomised controlled trials (RCTs) by searching CENTRAL, MEDLINE and Embase (from 1990 to 30 September 2022). We searched clinical trials registers and contacted investigators of completed and ongoing trials for further information.
Selection criteria
RCTs comparing angiogenesis inhibitors with standard chemotherapy, other types of anti‐cancer treatment, other angiogenesis inhibitors with or without other treatments, or placebo/no treatment in a maintenance setting, in women with EOC.
Data collection and analysis
We used standard methodological procedures expected by Cochrane. Our outcomes were overall survival (OS), progression‐free survival (PFS), quality of life (QoL), adverse events (grade 3 and above) and hypertension (grade 2 and above).
Main results
We identified 50 studies (14,836 participants) for inclusion (including five studies from the previous version of this review): 13 solely in females with newly‐diagnosed EOC and 37 in females with recurrent EOC (nine studies in platinum‐sensitive EOC; 19 in platinum‐resistant EOC; nine with studies with mixed or unclear platinum sensitivity). The main results are presented below.
Newly‐diagnosed EOC
Bevacizumab, a monoclonal antibody that binds vascular endothelial growth factor (VEGF), given with chemotherapy and continued as maintenance, likely results in little to no difference in OS compared to chemotherapy alone (hazard ratio (HR) 0.97, 95% confidence interval (CI) 0.88 to 1.07; 2 studies, 2776 participants; moderate‐certainty evidence). Evidence is very uncertain for PFS (HR 0.82, 95% CI 0.64 to 1.05; 2 studies, 2746 participants; very low‐certainty evidence), although the combination results in a slight reduction in global QoL (mean difference (MD) ‐6.4, 95% CI ‐8.86 to ‐3.94; 1 study, 890 participants; high‐certainty evidence). The combination likely increases any adverse event (grade ≥ 3) (risk ratio (RR) 1.16, 95% CI 1.07 to 1.26; 1 study, 1485 participants; moderate‐certainty evidence) and may result in a large increase in hypertension (grade ≥ 2) (RR 4.27, 95% CI 3.25 to 5.60; 2 studies, 2707 participants; low‐certainty evidence).
Tyrosine kinase inhibitors (TKIs) to block VEGF receptors (VEGF‐R), given with chemotherapy and continued as maintenance, likely result in little to no difference in OS (HR 0.99, 95% CI 0.84 to 1.17; 2 studies, 1451 participants; moderate‐certainty evidence) and likely increase PFS slightly (HR 0.88, 95% CI 0.77 to 1.00; 2 studies, 2466 participants; moderate‐certainty evidence). The combination likely reduces QoL slightly (MD ‐1.86, 95% CI ‐3.46 to ‐0.26; 1 study, 1340 participants; moderate‐certainty evidence), but it increases any adverse event (grade ≥ 3) slightly (RR 1.31, 95% CI 1.11 to 1.55; 1 study, 188 participants; moderate‐certainty evidence) and may result in a large increase in hypertension (grade ≥ 3) (RR 6.49, 95% CI 2.02 to 20.87; 1 study, 1352 participants; low‐certainty evidence).
Recurrent EOC (platinum‐sensitive)
Moderate‐certainty evidence from three studies (with 1564 participants) indicates that bevacizumab with chemotherapy, and continued as maintenance, likely results in little to no difference in OS (HR 0.90, 95% CI 0.79 to 1.02), but likely improves PFS (HR 0.56, 95% CI 0.50 to 0.63) compared to chemotherapy alone. The combination may result in little to no difference in QoL (MD 0.8, 95% CI ‐2.11 to 3.71; 1 study, 486 participants; low‐certainty evidence), but it increases the rate of any adverse event (grade ≥ 3) slightly (RR 1.11, 1.07 to 1.16; 3 studies, 1538 participants; high‐certainty evidence). Hypertension (grade ≥ 3) was more common in arms with bevacizumab (RR 5.82, 95% CI 3.84 to 8.83; 3 studies, 1538 participants).
TKIs with chemotherapy may result in little to no difference in OS (HR 0.86, 95% CI 0.67 to 1.11; 1 study, 282 participants; low‐certainty evidence), likely increase PFS (HR 0.56, 95% CI 0.44 to 0.72; 1 study, 282 participants; moderate‐certainty evidence), and may have little to no effect on QoL (MD 6.1, 95% CI ‐0.96 to 13.16; 1 study, 146 participants; low‐certainty evidence). Hypertension (grade ≥ 3) was more common with TKIs (RR 3.32, 95% CI 1.21 to 9.10).
Recurrent EOC (platinum‐resistant)
Bevacizumab with chemotherapy and continued as maintenance increases OS (HR 0.73, 95% CI 0.61 to 0.88; 5 studies, 778 participants; high‐certainty evidence) and likely results in a large increase in PFS (HR 0.49, 95% CI 0.42 to 0.58; 5 studies, 778 participants; moderate‐certainty evidence). The combination may result in a large increase in hypertension (grade ≥ 2) (RR 3.11, 95% CI 1.83 to 5.27; 2 studies, 436 participants; low‐certainty evidence). The rate of bowel fistula/perforation (grade ≥ 2) may be slightly higher with bevacizumab (RR 6.89, 95% CI 0.86 to 55.09; 2 studies, 436 participants).
Evidence from eight studies suggest TKIs with chemotherapy likely result in little to no difference in OS (HR 0.85, 95% CI 0.68 to 1.08; 940 participants; moderate‐certainty evidence), with low‐certainty evidence that it may increase PFS (HR 0.70, 95% CI 0.55 to 0.89; 940 participants), and may result in little to no meaningful difference in QoL (MD ranged from ‐0.19 at 6 weeks to ‐3.40 at 4 months). The combination increases any adverse event (grade ≥ 3) slightly (RR 1.23, 95% CI 1.02 to 1.49; 3 studies, 402 participants; high‐certainty evidence). The effect on bowel fistula/perforation rates is uncertain (RR 2.74, 95% CI 0.77 to 9.75; 5 studies, 557 participants; very low‐certainty evidence).
Authors' conclusions
Bevacizumab likely improves both OS and PFS in platinum‐resistant relapsed EOC. In platinum‐sensitive relapsed disease, bevacizumab and TKIs probably improve PFS, but may or may not improve OS. The results for TKIs in platinum‐resistant relapsed EOC are similar. The effects on OS or PFS in newly‐diagnosed EOC are less certain, with a decrease in QoL and increase in adverse events. Overall adverse events and QoL data were more variably reported than were PFS data.
There appears to be a role for anti‐angiogenesis treatment, but given the additional treatment burden and economic costs of maintenance treatments, benefits and risks of anti‐angiogenesis treatments should be carefully considered.
PICOs
Plain language summary
Do medicines that restrict new blood vessel growth (angiogenesis inhibitors) help women with epithelial ovarian cancer?
What did we want to find out?
We wanted to find out if treatments that prevent new blood vessel formation (angiogenesis) improve outcomes for women with epithelial ovarian cancer (EOC).
Ovarian cancer is the eighth most common cancer in women (and other females) worldwide, with an annual mortality rate of 4.2 per 100,000 women. EOC originates from the surface layers of ovaries or fallopian tubes and represents 90% of all ovarian cancers.
Treatment of EOC involves surgery to remove cancer deposits and platinum‐based chemotherapy (medicines that kill fast‐growing cells). However, despite good initial response, many with advanced disease eventually require further treatment.
Cancers need new blood vessels to supply oxygen and nutrients for growth; inhibiting angiogenesis may slow or stop cancer growth. Angiogenesis can be blocked either by smothering the angiogenesis hormone (called VEGF) with a monoclonal antibody (an antibody that recognises a single target) or by interfering with cell responses to VEGF binding with its receptor (VEGF‐R), by inhibiting enzymes (tyrosine kinases (TK)) associated with VEGF‐R (tyrosine kinase inhibitor (TKI)).
What did we do?
We collected and analysed all relevant studies in women with EOC. Studies compared angiogenesis inhibitors with or without conventional chemotherapy, or different biological agents against treatment with placebo (a dummy medicine), no treatment or different biological agents. We investigated whether these medicines improved how long women with EOC lived after treatment (overall survival (OS)), if medicines delayed disease re‐growth (progression‐free survival (PFS)), what were the harms (adverse events), and whether they impacted on quality of life. How well EOC responds to subsequent chemotherapy depends on previous chemotherapy treatment and time from last platinum‐based chemotherapy, so we analysed the results by whether people had newly‐diagnosed or recurrent EOC, and by platinum‐sensitivity.
What did we find?
We found 50 studies with 14,836 women.
Main results
Newly‐diagnosed EOC
Monoclonal antibody treatment (called bevacizumab or Avastin) given with chemotherapy, and continued as maintenance, probably has little effect on survival following an initial diagnosis of EOC. The evidence for delaying progression is very uncertain. Treatment increases serious side effects and slightly reduces quality of life.
TKIs given with chemotherapy and continued as maintenance, probably have little effect on survival following an initial diagnosis of EOC, but may delay disease progression. Treatment causes a slight reduction in quality of life, and a slight increase in the risk of serious side effects, with a big increase in the risk of needing treatment for high blood pressure (hypertension).
Recurrent EOC (platinum‐sensitive; relapse over a year after last platinum chemotherapy)
For women with platinum‐sensitive recurrent EOC, bevacizumab given with chemotherapy and continued as maintenance may have little effect on survival, but may delay progression. There may be little impact on quality of life, but treatment slightly increases the risk of serious side effects. All studies found that treatment increased rates of hypertension.
In this same group of women, TKIs given with chemotherapy and continued as maintenance probably have little effect on survival after relapse, likely delays progression, and may have little to no effect on quality of life. We were not able to estimate the effect on overall serious side effects, although serious hypertension was more common with treatment.
Recurrent EOC (platinum‐resistant; relapse within six months of last platinum chemotherapy)
For women with platinum‐resistant recurrent EOC, bevacizumab increased survival and probably results in a large delay in progression. However, treatment causes significant risk of hypertension and may increase the risk of bowel perforation. Other serious side effects were inconsistently reported, as were quality of life outcomes.
The addition of TKIs to chemotherapy in this group probably doesn't affect survival, but may delay progression, with little meaningful difference in quality of life. However, TKIs increase the risk of serious side effects slightly. The effect of treatment on bowel perforation rates and hypertension is very uncertain, largely due to small studies and different TKI drugs used in different studies.
What are the limitations of the evidence?
This is a rapidly moving field and evidence may change with further studies and longer follow‐up of studies.
How up to date is this evidence?
This review updates our previous review of 2011 and is up to date to September 2022.
Key messages
Newly‐diagnosed epithelial ovarian cancer (EOC)
The effects of bevacizumab and TKI anti‐angiogenesis treatment in women with newly diagnosed EOC are uncertain.
These treatments may have a minimal effect on how long women survive or disease re‐growth (progression), with a decrease in quality of life and an increase in serious side effects.
Platinum‐sensitive EOC
Bevacizumab and TKIs probably delay progression, but may or may not improve how long women live.
Platinum‐resistant EOC
Bevacizumab probably improves how long women live and probably results in a large delay in progression.
TKIs probably delay disease progression, but may or may not improve how long women live.
There appears to be a role for anti‐angiogenesis treatment, but additional treatment burden and financial costs of maintenance treatment of anti‐angiogenesis treatments should be carefully considered.
Authors' conclusions
Summary of findings
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.97 | 2776 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely results in little to no difference in overall survival. | |
| 590 per 1000 | 599 per 1000 | |||||
| Progression‐free survival (PFS) | Averaged | HR 0.82 | 2746 | ⨁◯◯◯ | The evidence is very uncertain about the effect of chemotherapy with bevacizumab on progression‐free survival. | |
| 550 per 1000 | 612 per 1000 | |||||
| Quality of life (QoL) | The mean global quality of life score was 76.1 | MD 6.4 score lower | ‐ | 890 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab results in a slight reduction in global quality of life. |
| Any adverse event grade ≥ 3 | 566 per 1000 | 657 per 1000 | RR 1.16 | 1485 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely increases any adverse event (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) | 44 per 1000 | 224 per 1000 | RR 4.27 | 2707 | ⨁⨁◯◯ | Chemotherapy with bevacizumab may result in a large increase in hypertension (grade ≥ 2). |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 trials (chemotherapy alone arms) | ||||||
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.99 | 1451 | ⨁⨁⨁◯ | Chemotherapy with TKI likely results in little to no difference in OS. | |
| 590 per 1000 | 593 per 1000 | |||||
| Progression‐free survival (PSF) | Averagec | HR 0.88 | 1451 | ⨁⨁⨁◯ | Chemotherapy with TKI likely increases PFS slightly. | |
| 550 per 1000 | 591 per 1000 | |||||
| Quality of life (QoL) Follow‐up: not specified | The mean quality of life score was 70.68 | MD 1.86 score lower | ‐ | 1340 | ⨁⨁⨁◯ | Chemotherapy with TKI likely reduces QoL slightly, although this may not be clinically significant. |
| Any adverse event grade ≥ 3 | 703 per 1000 | 921 per 1000 (780 to 1000) | RR 1.31 | 188 | ⨁⨁⨁◯ | Chemotherapy with TKI likely increases any adverse event (grade ≥ 3) slightly. |
| Hypertension grade ≥ 3 | 7 per 1000 | 43 per 1000 | RR 6.49 | 1352 | ⨁⨁◯◯ | Chemotherapy with TKI may result in a large increase in hypertension grade ≥3. |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 (chemotherapy alone arms) | ||||||
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI [peptide‐Fc fusion protein] and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.99 | 1015 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in overall survival. | |
| 590 per 1000 | 593 per 1000 (517 to 659) | |||||
| Progression‐free survival (PSF) | Averaged | HR 0.93 | 1015 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in progression‐free survival. | |
| 550 per 1000 | 574 per 1000 (521 to 624) | |||||
| Quality of life (QoL) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Any adverse event grade ≥ 3 | 661 per 1000 | Ranged from 727 to 1000 | RR ranged from 1.10 (grade 3) to 9.96 (grade 5) | 1011 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely increases any adverse event grade ≥ 3 slightly. |
| Hypertension grade ≥ 3 | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐sensitive EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.90 | 1564 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in overall survival. | |
| 490 per 1000 | 526 per 1000 | |||||
| Progression‐free survival (PFS) | Averagec | HR 0.56 | 1564 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely increases progression free‐survival. | |
| 230 per 1000 | 439 per 1000 | |||||
| Quality of life (QoL) Assessed with: TOI‐FACT‐OC questionnaire Scale from: 0 to 152 | The mean quality of life was 77 | MD 0.8 higher | ‐ | 486 | ⨁⨁◯◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in quality of life. |
| Any adverse event (grade ≥3) | 804 per 1000 | 892 per 1000 | RR 1.11 | 1538 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab followed by maintenance bevacizumab increases any adverse event (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) Assessed with: CTCAE versions 3.0‐4.0 where reported | ‐ | ‐ | ‐ | ‐ | ‐ | All three trials included in this comparison reported only hypertension grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 3.0‐4.0 where reported | ‐ | ‐ | ‐ | ‐ | ‐ | Two trials included in this comparison (MITO‐16b 2021 and GOG‐0213 2017) reported only gastrointestinal perforations of any grade. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON6 2021, GOG‐0213 2017 and OCEANS 2015 trials (chemotherapy arms only). | ||||||
| Patient or population: recurrent platinum‐sensitive EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI and as maintenance | |||||
| Overall survival (OS) Assessed with: survival rate | Average | HR 0.86 | 282 | ⨁⨁◯◯ | Chemotherapy with TKI followed by maintenance with TKI likely results in little to no difference in overall survival. | |
| 490 per 1000a,b | 541 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.0 criteria | Average | HR 0.56 | 282 | ⨁⨁⨁◯ | Chemotherapy with TKI followed by maintenance with TKI likely increases progression‐free survival. | |
| 230 per 1000 | 439 per 1000 | |||||
| Quality of life (QoL) Follow‐up: 12 months | The mean quality of lIfe was 62.6 | MD 6.1 higher | ‐ | 146 | ⨁⨁◯◯ | Chemotherapy with TKI followed by maintenance with TKI may result in little to no difference in quality of life. |
| Any adverse events (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Hypertension (grade ≥ 2) | ‐ | ‐ | ‐ | ‐ | ‐ | A single trial included in this comparison reported only events of grade ≥ 3 (ICON6 2021). |
| Bowel fistula/perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON6 2021, OCEANS 2015 and GOG‐0213 2017 trials (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐resistant EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.73 | 778 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab increases overall survival. | |
| 10 per 1000 | 35 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.0‐1.1 where reported | Averageb | HR 0.49 | 778 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely results in a large increase in progression‐free survival. | |
| 40 per 1000 | 207 per 1000 | |||||
| Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Any adverse event grade ≥ 3 | 460 per 1000 | 773 per 1000 (350 to 1000) | RR 1.68 (0.76 to 3.69) | 101 (1 RCT) | ⨁⨁◯◯ | Chemotherapy with bevacizumab may increase any adverse events (grade > 3) slightly. |
| Hypertension (grade ≥ 2) Assessed with: CTCAE version 3.0 | 73 per 1000 | 228 per 1000 | RR 3.11 | 436 | ⨁⨁◯◯ | Chemotherapy with bevacizumab may result in a large increase in hypertension (grade ≥ 2). |
| Bowel fistula / perforation (grade ≥ 2) | 4 per 1000g | 28 per 1000 (3 to 220) | RR 6.89 (0.86 to 55.09) | 436 (2 RCTs) | ⨁⨁◯◯ | Chemotherapy with bevacizumab may increase rates of bowel fistula / perforation (grade ≥ 2) slightly. Two studies included in this comparison although one reported only gastrointestinal perforations (grade ≥ 2) (AURELIA 2014). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in AURELIA 2014, MITO‐11 2015 and TRIAS 2018 trials (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐resistant EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.85 | 940 | ⨁⨁⨁◯ | Chemotherapy with TKI likely results in little to no difference in overall survival. | |
| 10 per 1000 | 16 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.1 criteria where specified | Averagec | HR 0.70 | 940 | ⨁⨁◯◯ | Chemotherapy with TKI may increase progression‐free survival. | |
| 40 per 1000 | 87 per 1000 | |||||
| Quality of life (QoL) assessed with: Global Quality of Life and EORTC core QoL questionnaire (QLQ‐C30) Scale from: 0 to 100 | MD in Quality of Life score ranged from ‐0.19 (95%CI ‐9.77 to 9.39) at 6 weeks (METRO‐BIBF 2020) to ‐3.40 (95%CI ‐13.22 to 6.42) at 4 months (TAPAZ 2022) | 164 | ⨁⨁◯◯ | Chemotherapy with TKI may result in little to no difference in quality of life. | ||
| Any adverse events (grade ≥3) Assessed with: CTCAE versions 3.0‐4.1 | 581 per 1000 | 657 per 1000 | RR 1.23 | 548 | ⨁⨁◯◯ | Chemotherapy with TKI may increase any adverse events (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) | ‐ | ‐ |
|
|
| Trials included in this comparison reported only on events of grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 4.0‐4.1 | 4 per 1000 | 11 per 1000 | RR 2.74 | 557 | ⨁◯◯◯ | The evidence is very uncertain about the effect of chemotherapy with TKI on bowel fistula/perforation (grade ≥ 3). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in AURELIA 2014, MITO‐11 2015 and TRIAS 2018 (chemotherapy alone arms). | ||||||
| Patient or population: recurrent EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI [peptide‐Fc fusion protein] | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.92 | 1250 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in overall survival. | |
| 60 per 1000 | 75 per 1000 (51 to 105) | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST version 1.1 criteria | Averaged | HR 0.73 | 1250 | ⨁⨁⨁⨁ | Chemotherapy with TKI [peptide‐Fc fusion protein] increases progression‐free survival. | |
| 110 per 1000 | 200 per 1000 (164 to 238) | |||||
| Quality of life (QoL) Assessed with: TOI‐FACT‐OC questionnaire Scale from: 0 to 152 Follow‐up: 25 weeks | The mean change from baseline QoL was ‐1.6 | MD 0.8 lower | ‐ | 315 | ⨁⨁◯◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] may result in little to no difference in quality of life. |
| Any adverse events (grade ≥3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Hypertension (grade ≥ 2) | ‐ | ‐ | ‐ | ‐ | ‐ | All three trials included in this comparison reported events of grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 3.0 | 18 per 1000 | 6 per 1000 | RR 0.35 | 108 | ⨁⨁◯◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] may result in little to no difference in bowel perforation/fistula G3+. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in TAPAZ 2022, TRINOVA‐1 2016 and TRINOVA‐2 2017 trials (chemotherapy alone arms). | ||||||
Background
Description of the condition
This is an update of the review, originally published in 2011 (Gaitskell 2011).
Each year, worldwide, over 300,000 women and people with ovaries are diagnosed with epithelial ovarian cancer (EOC) and over 200,000 die, corresponding to an annual age‐standardised incidence of 6.6 cases per 100,000 females, an annual mortality rate of 4.2 deaths per 100,000, and a cumulative lifetime risk of 0.73% for incidence and 0.49% for mortality (GLOBOCAN 2020). In terms of both incidence and mortality, it is the eighth most common cancer in females. The onset is often insidious, as abnormal cells developing on the surface of the ovaries or lining of the fallopian tubes have ready access to spread throughout the abdominal cavity, with no effective screening tests. Approximately 60% of women with EOC in the USA are diagnosed when the disease has spread (stage III or IV) and the five‐year survival rate is 20% to 30% (Berek 2018; Cancer Research UK 2022; Siegel 2021). EOC accounts for 90% of all ovarian cancers and typically presents in post‐menopausal women, with a peak incidence when women are in their early sixties, although it does occur in younger women, often associated with genetic predispositions (Quinn 2001).
EOC may be divided into several different histological types, on the basis of different microscopic appearances and characteristic molecular features. There is growing evidence that these different histological types may have different origins (Kurman 2011; Prat 2012; Prat 2018; WHO 2020). In particular, many cases of high‐grade serous ovarian carcinoma (the most common type, accounting for about 70% of EOC (Prat 2018)) are thought to arise from precursor lesions within the fallopian tubes (serous tubal intraepithelial carcinomas) which share the same p53 mutations characteristics of high‐grade serous carcinoma (Ahmed 2010; Kurman 2013; Labidi‐Galy 2017; Shih 2021). Some cases of endometrioid and clear cell ovarian cancer are thought to arise from endometriosis (Kurman 2011).
Description of the intervention
Management of advanced ovarian cancer consists of a combination of debulking surgery, either before or during chemotherapy, and platinum‐based chemotherapy, with or without the addition of a taxane (Coleridge 2021; Stewart 1999). However, in women presenting with advanced disease, the five‐year survival rate for stage III to IV of the disease remains poor (Engel 2002; Cancer Research UK 2022; Siegel 2021). Despite good initial response to platinum‐based chemotherapy, the majority of women with advanced disease at presentation will relapse, require further treatment with chemotherapy, and eventually develop resistance to conventional chemotherapeutic agents.
Conventional chemotherapeutic agents have activity on all rapidly dividing cells; hence, the common side effects, such as hair loss, bone marrow suppression, and mucositis (inflammation and ulceration of the mucous membranes lining the digestive tract). Increasing knowledge of the genetic basis for cancer has led to the development of novel reagents, which target cancer‐specific pathways. It is hoped that these reagents will spare normal cells and reduce the toxic side effects of chemotherapy, in addition to having an enhanced therapeutic effect.
Since the 2011 version of this review, bevacizumab is now approved for use in both first‐line and second‐line settings in clinical practice in several regions, including the USA (NCCN 2022), Europe (Colombo 2019; Ledermann 2013) and the UK (BGCS 2017). Bevacizumab is sometimes used as part of first‐line therapy in advanced disease, where it may be given both concurrently with conventional first‐line cytotoxic chemotherapy, and then continued alone as a maintenance treatment. Bevacizumab is also sometimes given alongside chemotherapy in the second‐line (relapsed/resistant) setting. In some cases, bevacizumab may also be given alongside another type of novel therapy known as poly(ADP‐ribose) polymerase (PARP) inhibitors (NCCN 2022).
How the intervention might work
Angiogenesis and ovarian cancer
Angiogenesis is the development of new blood vessels. Once a tumour deposit is larger than 1 mm in diameter, it cannot receive adequate nutrients or oxygen from surrounding tissues by diffusion alone, and it must then stimulate new blood vessel formation to support further growth. Angiogenesis is a vital part of embryo development, but is tightly controlled in adults and normally occurs during wound healing and as part of ovulation. Abnormal angiogenesis can occur in a variety of illnesses, either stimulated by low oxygen levels in tissues (e.g. diabetes and metastatic cancer), or in inflammatory conditions, such as rheumatoid arthritis (Fidler 1994; Folkman 1990). In contrast to the ordered formation of new blood vessels during embryonic angiogenesis, tumour angiogenesis is disordered and results in abnormal and leaky blood vessels (McDonald 2002). Blocking this process may prevent growth of small tumour deposits and improve survival of people with cancer.
Angiogenesis requires signalling between tumour cells and nearby endothelial (lining)cells of normal blood vessels, stimulating them to sprout, multiply and invade the growing tumour. The process involves release of agents by cancer cells, stimulated by low oxygen levels or low pH. These agents bind to receptors on endothelial cells, which then trigger downstream intracellular signalling, leading to growth and migration of endothelial cells. This process can be inhibited at each of these stages. Because angiogenesis is normally inactive in adults, its inhibition is an attractive candidate for selective anti‐tumour therapies. Another advantage is that tumour endothelial cells are not themselves malignant and so, unlike cancer cells themselves, do not have pre‐existing mutations that favour the development of further mutations, which could lead to drug resistance. In addition, anti‐angiogenic agents may work synergistically with conventional chemotherapeutic agents or other novel systemic agents, due to their different mechanisms of action.
Vascular endothelial growth factor (VEGF) is one of the key elements in the stimulation of angiogenesis. VEGF is released by cancer cells and binds to a receptor on endothelial cells (VEGF‐R) (Figure 1 A‐B). VEGF binding stimulates tyrosine kinase activity in the VEGF‐R (Figure 1 B), which in turn stimulates downstream signalling and activation of endothelial cells (Figure 1 C). VEGF over‐expression is associated with ascites formation (build up of fluid within the abdominal cavity) and poorer prognosis (Oehler 2000). There are different forms of VEGF: VEGF‐A has an important role in the formation of new blood vessels; VEGF‐B is involved with the maintenance of newly formed blood vessels (Zhang 2009).
(A) The VEGF‐R is a transmembrane protein, found on cells, which line blood vessels (endothelial cells). (B) Following binding to its ligand, VEGF, the VEGF‐R is stimulated and develops tyrosine kinase activity. (C) Tyrosine kinase activity sets off a sequence of downstream events that lead to stimulation of cell growth and new vessels grow in, to supply the growing tumour. (D) VEGF‐R activity can be blocked by antibodies, which bind to VEGF, and so stop it binding to the receptor, or using chemicals, which inhibit the tyrosine kinase enzyme activity of the VEGF‐R.
VEGF signalling can be blocked at several levels (Figure 1 D). First, anti‐VEGF antibodies or soluble VEGF‐R molecules mop up excess VEGF and prevent binding to, and stimulation of, cellular VEGF‐R. Second, antibodies have been developed that bind to VEGF‐R and block binding and activation by VEGF. Third, VEGF‐R signalling may also be inhibited by small molecules which specifically inhibit the intracellular tyrosine kinase activity of VEGF‐R following stimulation by angiogenic factors.
VEGF‐R inhibitors
Small molecule inhibitors of VEGF‐R tyrosine kinase (tyrosine kinase inhibitors (TKIs)) have been developed and investigated in clinical trials. One advantage of these compounds is that many are orally active. For details of the mechanism of action, see Table 1.
| Study ID | Number of references | Intervention/s (N) | Control (N) | Number randomised | Randomisation ratio | Type of anti‐angiogenesis agent | Newly‐diagnosed or relapsed/recurrent EOC | Population in relation to platinum‐sensitivity* | Percentage (%) stage IV (newly‐diagnosed EOC only) | Prior treatment |
|---|---|---|---|---|---|---|---|---|---|---|
| Newly‐diagnosed EOC | ||||||||||
| 6 | Chemotherapy + nintedanib (911) | Chemotherapy | 1366 | 2:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Newly‐diagnosed | PS 100% | 24% in intervention arm; 24% in control arm); overall 24% | N/A | |
| 13 | Pazopanib | Placebo | 940 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Newly‐diagnosed | PS 100% | 16.3% in intervention group; 16.9% in control group; overall 16.6% | N/A | |
| 6 | Chemotherapy + bevacizumab | Chemotherapy | 95 | 2:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 26% in intervention group; 35% in control group; overall 30% | N/A | |
| 6 | Chemotherapy + nintedanib (124) | Chemotherapy | 188 | 2:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Newly‐diagnosed | PS 100% | N/A | N/A | |
| 3 | Chemotherapy + bevacizumab | Chemotherapy | 68 | 1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 34% in intervention arm and 33% in control arm; 33.8% overall | N/A | |
| 20 | Chemotherapy + bevacizumab (625) | Chemotherapy | 1873 | 1:1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 26.2% intervention in initiation only arm; 26.5% in initiation and maintenance arm; 24.5% in control arm; 25.7% overall | N/A | |
| 4 | Chemotherapy + sorafenib (43) | Chemotherapy | 85 | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Newly‐diagnosed | PS 100% | 33% in control arm; 19% in intervention arm; 25.9% overall | N/A | |
| 3 | Sorafenib | Placebo | 246 | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Newly‐diagnosed | PS 100% | Stage at diagnosis not provided but all stage III/V; 8.1% suboptimally debulked at primary surgery in each arm | N/A | |
| 16 | Chemotherapy + bevacizumab | Chemotherapy | 1528 | 1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 12% in control arm; 13% in intervention arm; 13.2% overall | N/A | |
| 2 | Chemotherapy + celecoxib (97) | Chemotherapy (99) | 196 | 1:1 | Celecoxib: COX‐2 inhibitor | Newly‐diagnosed | PS 100% | 25.3% in control arm; 22.7% in intervention arm; 23.7% overall | N/A | |
| 2 | Chemotherapy + trebananib | Chemotherapy | 1015 | 2:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Newly‐diagnosed | PS 100% | 24% in control arm; 27% in intervention arm; 26.2% overall | N/A | |
| Platinum‐sensitive recurrence | ||||||||||
| 5 | Niraparib + bevacizumab | Niraparib | 97 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS (66%) PPS (34%) | N/A | Platinum‐based chemotherapy. Overall previous lines of treatment: one = 49.5%; two = 44.3%; three or more = 6% | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 164 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based chemotherapy | |
| 3 | Chemotherapy + bevacizumab (377) | Chemotherapy (337) | 674 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based chemotherapy | |
| 8 | Chemotherapy + cediranib + | Chemotherapy | 486+ (456) | 2:3:3 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PS 67% | N/A | Platinum‐based chemotherapy, 89% with paclitaxel. Overall 5% had had previous bevacizumab treatment. | |
| 1 | Chemotherapy with paclitaxel and carboplatin + bevacizumab (34) | Chemotherapy with paclitaxel and carboplatin (34) | 68 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS 100% | N/A | Platinum‐based chemotherapy at least (presumed) | |
| 5 | Olaparib + cediranib (44) | Olaparib (46) | 90 | 1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PS 100% | N/A | Platinum‐based chemotherapy and max 1 non‐platinum therapy in recurrent setting | |
| 4 | Cediranib + olaparib | Chemotherapy | 565 | 1:1:1 | TKI with PARPi | Recurrent | PS
| N/A | Platinum and non‐platinum based chemotherapy (65% only 1 prior line of chemotherapy) | |
| 2 | Chemotherapy + bevacizumab | Chemotherapy | 406 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS 100% | N/A | First‐line platinum‐based treatment, including bevacizumab | |
| 12 | Chemotherapy + bevacizumab (242) | Chemotherapy + placebo (242) | 484 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based front‐line chemotherapy | |
| Platinum‐resistant recurrence | ||||||||||
| 5 | Olaparib + cediranib (16) | Olaparib + durvalumab (14) | 30 for relevant comparison [3 other arms, N = 70 in total] | 1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR | N/A | At least 2 prior lines of anticancer therapy | |
| 3 | Chemotherapy + apatinib (78) | Chemotherapy | 150 | 1:1 | Apatinib: TKI targeting VEGF‐R2 | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy | |
| 16 | Chemotherapy + bevacizumab | Chemotherapy | 361 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy (max 2) | |
| 2 | Cediranib‐olaparib combination (continuous n= 41) (intermittent n= 41) | Weekly paclitaxel (n= 41) | 123 | 1:1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR (100%) | N/A | 39.8% up to 2 previous lines of chemotherapy; 60.2% ≥3 previous lines of chemotherapy. 53.7% prior anti‐angiogenic treatment | |
| 2 | Bevacizumab (33) atezolizumab + cevacizumab + placebo Atezolizumab + bevacizumab + acetylsalicylic acid | Atezolizumab + placebo | 122 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy (max of 2 non‐platinum regimens) | |
| 5 | Aflibercept (29) | Placebo (26) | 55 | 1:1 | Aflibercept: fusion protein targeting VEGF‐A and VEGF‐B | Recurrent | PR | N/A | At least 2 lines of previous chemotherapy, one platinum‐based | |
| 1 | Bevacizimab + albumin‐binding paclitaxel | Albumin‐binding paclitaxel | 70 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR | N/A | Unclear as English language abstract only | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 86 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy | |
| 1 | Chemotherapy + bevacizumab (38) | Chemotherapy (38) | 76 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR | N/A | Platinum‐based chemotherapy; platinum‐free interval 4.3 months ±0.6 months control group and 4.8 ±0.8 months in bevacizumab group (P=0.06) | |
| 3 | Chemotherapy + olaratumab (62) | Chemotherapy (61) | 123 | 1:1 | Olaratumab: monoclonal antibody targeting PDGFR‐α | Recurrent | PR | N/A | Platinum‐based chemotherapy | |
| 3 | Cyclophosphamide + nintedanib | Cyclophosphamide | 117 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PR or intolerant 100% | N/A | Two or more lines of chemotherapy | |
| 3 | Chemotherapy + pazopanib (37) | Chemotherapy (37) | 74 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PR | N/A | Prevous chemotherapy lines: one = 43.8%; two = 47.9% three or more = 8.2% | |
| 2 | Nintedanib | Chemotherapy | 91 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy. participants had clear cell carcinoma of EOC or endometrial origin. 91 participants with EOC. | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 103 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | previously treated with ≥3 cycles of bevacizumab + platinum chemotherapy; progression occurred <6 months after completion of platinum treatment | |
| 3 | Olaparib + cediranib | Olaparib (46) Chemotherapy (46)
| 139 | 1:1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR 100% | N/A | Prior PARPi therapy (22%) | |
| 3 | Ixabepilone + bevacizumab | Ixabepilone | 76 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR or refractory 100% | N/A | Not reported | |
| 2 | Etoposide + cyclophosphamide + pazopanib | Etoposide + cyclophosphamide | 75 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R, c‐kit, and FGF‐R | Recurrent | PR 51%; | N/A | Prior treatment with at least 2 chemotherapy regimens in advanced tumor | |
| 3 | Chemotherapy + vandetanib (63) | Chemotherapy (66) | 129 | 1:1 | Vandetanib: TKI targeting VEGF‐R, EGF‐R, and RET | Recurrent | All patients were considered platinum resistant or refractory | N/A | Platinum‐based front‐line chemotherapy +/‐ up to 3 chemotherapy regimens in current setting +/‐ primary anti‐angiogenic therapy | |
| 3 | Topotecan + sorafenib | Topotecan + placebo | 174 (172) | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Recurrent | PR or refractory 100% | N/A | No more than two prior treatment regimens for recurrent EOC | |
| Mixed platinum‐sensitive and platinum‐resistant recurrence | ||||||||||
| 3 | Chemotherapy + pazopanib (75) | Chemotherapy (76) | 148 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PS (40%) PR (60%) | N/A | Chemotherapy (max 3) | |
| 2 | Cyclophosphamide + celecoxib | Cyclophosphamide | 52 | 1:1 | Celecoxib: COX‐2 inhibitor | Recurrent | PS 38.5% | N/A | No limit on prior lines of therapy | |
| 9 | Chemotherapy + lower‐dose trebananib (AMG386) (53) Chemotherapy + higher‐dose trebananib (AMG386) (53) | Chemotherapy + placebo (55) | 161 | 1:1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PS (52%) PR (47%) | N/A | Platinum and non‐platinum based chemotherapy (max 3 in total) | |
| 3 | Nintedanib (BIBF 1120) (43) | Placebo (41) | 84 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PS (59%) PR (41%) | N/A | Chemotherapy (2 or more rounds) | |
| 2 | Cabozantinib (57) | Chemotherapy (54) | 111 | 1:1 | Cabozantinib: TKI targeting VEGF‐R2, c‐MET, c‐kit, Tie2, FLT‐3, and RET | Recurrent | PR (50%) PS (50%) | N/A | Platinum‐based chemotherapy +/‐ non‐platinum based regimens (max 3 in total) | |
| 3 | Chemotherapy + pazopanib (54) | Chemotherapy + placebo (52) | 106 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PR (51%) PS (49%) | N/A | Platinum‐based chemotherapy +/‐ non‐platinum based chemotherapy (max 3 in total) | |
| 3 | Paclitaxel + pazopanib (79) | Paclitaxel | 116 | 2:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PPS 70.7%; PR 29.3% | N/A | Not reported | |
| 8 | Chemotherapy + trebananib (461) | Chemotherapy + placebo (458) | 919 | 1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PR (53%) PS (47%) | N/A | Platinum‐based chemotherapy +/‐ up to 2 other chemotherapy regimens +/‐ anti‐angiogenic therapy | |
| 3 | Chemotherapy + trebananib (114) | Chemotherapy + placebo (109) | 223 | 1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PR (59%) PS (41%) | N/A | Platinum‐based chemotherapy +/‐ up to 2 other chemotherapy regimens +/‐ anti‐angiogenic therapy | |
| Other | ||||||||||
| 3 | Chemotherapy (two different regimes) + bevacizumab | Chemotherapy (two different regimes) | 50 | 1:1:1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed & recurrent. Mucinous EOC only | N/A | N/A | No previous chemotherapy | |
| 3 | Intraperitoneal chemotherapy + bevacizumab (31) | Intraperitoneal chemotherapy (27) | 58 | 1:1 | Bevacizumab: antibody against VEGF | Unclear | Unclear | 77.4% in intervention group; 77.8% in control group; overall 77.6% | Unclear | |
PPS: partially‐platinum sensitive; ALT: alanine transaminase; AMG386: trebananib; Ang1: angiopoietin 1; Ang2: angiopoietin 2; AST: aspartate aminotransferase; AUC: area under the curve; BIBF : BIBF 1120 = nintedanib; BRCA: breast cancer gene; CA125: cancer antigen 125; COX‐2: cyclo‐oxygenase‐2; CT:computed tomography; CTCAE: Common Terminology Criteria for Adverse Events; ECOG: Eastern Cooperative Oncology Group; EOC: epithelial ovarian cancer; EORTC: European Organisation for Research and Treatment of Cancer; FACT/GOG NTX: Functional Assessment of Cancer Therapy/Gynecologic Oncology Group – Neurotoxicity; FACT‐O (TOI): Functional Assessment of Cancer Therapy‐Ovarian (Trial Outcome Index); FGF‐R : fibroblast growth factor receptor; FIGO: International Federation of Gynecology and Obstetrics; FLT‐3: Fms‐like receptor tyrosine kinase‐3; FOSI: Functional Assessment of Cancer Therapy (FACT)/National Comprehensive Cancer Network (NCCN) Ovarian Symptom Index; GCIG: Gynecological Cancer InterGroup; GFR: glomerular filtration rate; GOG: Gynecologic Oncology Group; HRD: homologous recombination deficiency; ICON: International Collaborative Ovarian Neoplasm study; IDS: interval debulking surgery; IQR: interquartile range; ITT: intention‐to‐treat; IV: intravenous(ly); IVRS/IWRS: Interactive Voice Response System/ Interactive Web Response System; KGOG: Korean Gynecologic Oncology Group; KPS: Karnofsky Performance Status; MET: mesenchymal epithelial transition; MRI: magnetic resonance imaging; NCI: National Cancer Institute; ORR: objective response rate; OS: overall survival; PARP: poly(ADP‐ribose) polymerase; PARPi: poly(ADP‐ribose) polymerase inhibitor; PDGF‐R: platelet‐derived growth factor receptor; PDGFR‐α: alpha subunit of PDGF‐R; PD‐L1: Programmed death‐ligand 1; PLB: placebo; PLD: pegylated liposomal doxorubicin; PFS: progression‐free survival; PPS: partially‐platinum sensitive; PR: platinum‐resistant; PS: platinum‐sensitive; QoL: quality of life; RAF: Rapidly Accelerated Fibrosarcoma; RECIST: Response Evaluation Criteria in Solid Tumors; RET:REarranged during Transfection; SWOG: Southwest Oncology Group; Tie2: angiopoetin‐1 receptor; TKI: tyrosine kinase inhibitor; TNM: tumour nodes metastases; ULN: upper limit of normal; USS: ultrasound scan; VEGF: vascular endothelial growth factor; VEGF‐R: vascular endothelial growth factor receptor; WHO: World Health Organization
+ICON6 2021 486 randomised overall, of which 30 randomised to initial 30 mg dose of cediranib and excluded because of increased toxic effects, leaving 456.
-
Cediranib (AZD2171 or Recentin AstraZeneca) is a small molecule inhibitor of VEGF‐R and also inhibits the c‐kit proto‐oncogene receptor tyrosine kinase, and has some weak activity against platelet‐derived growth factor receptor (PDGF‐R) (Wedge 2005; Brave 2011; NCI‐DCTD‐Cediranib 2022; ICON6 2021).
-
Pazopanib is a potent selective receptor tyrosine kinase inhibitor of VEGF‐R, PDGF‐R and c‐kit that blocks tumour growth and inhibits angiogenesis (Sonpavde 2007; Richardson 2018).
-
Nintedanib (BIBF 1120) is an oral, small molecule, triple angiokinase inhibitor, targeting VEGF‐R, fibroblast growth factor receptor (FGF‐R) and PDGF‐R (Erber 2004; Hilberg 2008; Ledermann 2011).
-
Brivanib is a small molecule inhibitor targeting the VEGF‐R and FGF‐R tyrosine kinase receptor families (Bhide 2010; Cai 2008).
-
Cabozantinib is a small molecule inhibitor targeting multiple tyrosine kinase receptors, primarily c‐MET (mesenchymal epithelial transition) and VEGF‐R2, but also c‐kit, Tie‐2 (an angiopoietin receptor), FLT‐3 (Fms‐like receptor tyrosine kinase‐3) and RET (REarranged during Transfection ‐ a proto‐oncogene) (Yakes 2011; Matulonis 2019).
-
Vandetanib (ZD6474) is a small molecule inhibitor of VEGF‐R, EGF‐R and RET(Carlomagno 2002; Wedge 2002; SWOG‐S0904 2014).
-
Sorafenib (BAY 43‐9006/Nexavar) is a small molecule tyrosine kinase inhibitor that directly inhibits VEGF‐R in addition to other angiogenic and growth stimulatory pathways, including via PDGF‐R and Raf kinase inhibition (Mross 2007; Siu 2006).
-
Sunitinib (SU11248) is a small molecule inhibitor of the VEGF‐R, c‐kit and FLT3 tyrosine kinase receptors (Abrams 2003; Mendel 2003).
-
Apatinib is a small molecule inhibitor targeting several tyrosine kinase receptors, including VEGF‐R2 (Ding 2019; Tian 2011).
VEGF blockade
Monoclonal antibodies are antibodies that have a specific target pattern to which they bind. Bevacizumab (Avastin) is a humanised monoclonal antibody that binds VEGF, prevents it binding to VEGF‐R, and so inhibits VEGF‐R activation and angiogenesis (Ferrara 2004). Bevacizumab has been shown to have activity in phase II trials in women who had platinum‐resistant relapsed ovarian cancer (13% to 16% partial response rates and 25% to 55% stable disease), although complete responses, in this group of pre‐treated patients, were low (0% to 5%) (Burger 2007; Cannistra 2007). Side effects encountered were different to those seen with conventional chemotherapy, in line with its alternative mode of action. They included hypertension, bleeding episodes, thromboembolism and bowel perforation.
On the basis of success from these studies, phase III trials have been performed combining bevacizumab with carboplatin and taxol chemotherapy in postoperative patients with ovarian cancer in the GOG 218 (GOG‐0218 2019) and the ICON 7 (ICON7 2015) studies. These trials are also assessing the role of bevacizumab in the maintenance treatment of these patients.
Aflibercept is a recombinant fusion protein that binds to VEGF‐A and VEGF‐B, acting as a soluble decoy receptor (Ciombor 2014; NICE 2022).
VEGF inhibitors combined with PARP inhibitors
PARP inhibitors are a novel type of cancer treatment that targets the DNA damage repair pathways in cancer cells. They work particularly well in people who have already inherited genetic mutations affecting other DNA damage pathways (e.g. the BRCA breast and ovarian cancer susceptibility gene variants). There have been several studies investigating PARP inhibitors as a possible treatment for ovarian cancer; in some cases, alongside bevacizumab (Tattersall 2022). Part of the rationale for trying the combination of PARP inhibitors and angiogenesis inhibitors is that there is some early evidence that some angiogenesis agents may also affect DNA damage repair pathways, while PARP inhibitors may also affect angiogenesis, and thus the combination of these two types of treatment might be even more effective at targeting both DNA damage repair and angiogenesis (Alvarez Secord 2021).
Agents targeting other aspects of angiogenesis
Trebananib (AMG 386) inhibits angiogenesis via a different mechanism. Trebananib is a peptide‐Fc fusion protein (composed of the Fc portion of human immunoglobulin IgG1 fused to a peptide of interest) which inhibits angiogenesis by binding to the angiopoietins Ang1 and Ang2, and so preventing them binding to the receptor Tie2 (Neal 2010; Oliner 2004; TRINOVA‐1 2016).
Celecoxib is an inhibitor of the enzyme cyclo‐oxygenase‐2 (COX‐2) and is a non‐steroidal anti‐inflammatory drug. There is also some evidence that it can act to inhibit angiogenesis (Gupta 2019; Masferrer 1999; Masferrer 2000).
Olaratumab (IMC‐3G3) is a monoclonal antibody targeting platelet‐derived growth factor alpha (PDGFRα), a transmembrane receptor tyrosine kinase which is involved in the maturation of new blood vessels. Olaratumab inhibits angiogenesis by binding to PDGFRα and so inhibiting these pro‐angiogenic downstream signalling pathways (Choi 2015; McGuire 2018).
Why it is important to do this review
Novel treatment strategies working in different ways to conventional chemotherapy have been developed. It is important to establish whether the addition of these new drugs to conventional chemotherapy regimens has additional benefit, in terms of survival, and if so, at what cost, in terms of additional harmful effects. Furthermore, since these compounds may be less toxic compared to conventional chemotherapeutic agents, these newer treatments are used increasingly in people who are not currently taking chemotherapy (so‐called maintenance treatment), to reduce the chance of, or delay, the recurrence of their ovarian cancer. Ensuring that maintenance treatment gives additional benefit, with improvement in overall survival without significant degradation of quality of life, is extremely important in what, for many, will be a life‐limiting illness. Since the previous version of the review was published (Gaitskell 2011), there have been significant developments in this area and this represents a significant update in the field.
Objectives
To compare the effectiveness and toxicities of angiogenesis inhibitors for treatment of epithelial ovarian cancer.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials (RCTs) of angiogenesis inhibitors plus conventional chemotherapy versus conventional chemotherapy alone, and angiogenesis inhibitors versus no treatment. We only included results from trials with a minimum of 10 participants.
Types of participants
We included adult women (and other females), aged 18 and over, with histologically‐proven epithelial ovarian cancer (EOC) (including high grade serous tubal and primary peritoneal malignancies). We excluded women with other concurrent malignancies.
In this review update, we planned to stratify by clinically relevant treatment setting; that is, to analyse separately those with newly‐diagnosed EOC and relapsed EOC. We further subdivided recurrent EOC by platinum‐sensitivity (platinum‐sensitive, platinum‐resistant/refractory and studies with a mixed or unclear platinum sensitivity) since these have different biology and response rates to chemotherapy treatment.
We had not previously explicitly specified that the review was limited to EOC, although all previously included studies were limited to those with EOC.
Types of interventions
There have been a number of developments in this field since the publication of the original protocol and previous version of the review (Gaitskell 2011). Angiogenesis inhibitors are used as standard of care for some people in some settings. We therefore planned to include studies that contained the following comparisons in this update of the review:
-
angiogenesis inhibitors plus conventional chemotherapy versus conventional chemotherapy (including studies where the angiogenesis inhibitor is continued as maintenance after chemotherapy);
-
angiogenesis inhibitors versus no treatment (e.g. in a maintenance setting);
-
angiogenesis inhibitor 1 versus angiogenesis inhibitor 2, with either chemotherapy in each arm or no other treatment;
-
chemotherapy plus angiogenesis inhibitor 1 versus chemotherapy plus angiogenesis inhibitor 1 plus angiogenesis inhibitor 2;
-
angiogenesis inhibitor versus an alternative chemotherapy treatment.
Types of outcome measures
Primary outcomes
Overall survival (OS): survival until death from any cause
Secondary outcomes
-
Progression‐free survival (PFS)
-
Quality of life (QoL), measured by a validated scale
-
Toxicity. Based on the outcomes reported by studies included in the 2011 version of this review, we identified the following adverse event outcomes by Common Terminology Criteria for Adverse Events (CTCAE) criteria:
-
any severe adverse event (grade 3 or higher (≥ G3));
-
hypertension (≥ G2);
-
proteinuria (≥ G2);
-
pain (≥ G2);
-
abdominal pain (≥ G2);
-
neutropenia (≥ G3);
-
-
febrile neutropenia (any grade);
-
venous thromboembolic event (any grade);
-
arterial thromboembolic event (any grade);
-
non‐central nervous system bleeding (≥ G3);
-
gastrointestinal adverse events (≥ G2);
-
bowel fistula or perforation (≥ G3).
Please see academy.myeloma.org.uk/wp-content/uploads/2015/04/CTCAE_v5.pdf for details regarding CTCAE criteria.
Search methods for identification of studies
We sought papers in all languages and obtained translations when necessary.
Electronic searches
See Cochrane Gynaecological Cancer Group methods used in reviews.
We searched the following electronic databases from 1990 to 30 September 2022:
-
Cochrane Gynaecological Cancer Review Group's trials register;
-
Cochrane Central Register of Controlled Trials (CENTRAL; 2022, Issue 9) in the Cochrane Library;
-
MEDLINE Ovid;
-
Embase Ovid.
We present the CENTRAL, MEDLINE and Embase search strategies in Appendix 1, Appendix 2 and Appendix 3, respectively.
We searched databases from 1990 to October 2010 for the 2011 version of this review, and from October 2010 to 30 September 2022 for this update. The novel agents we focus on in this review have been developed relatively recently: trials published before 1990 would not have been relevant.
We also used the ‘related articles’ feature on PubMed, and searched the reference lists of included studies in this review and systematic reviews on this topic published between 2020 and September 2022. We limited our search to systematic reviews published in the last two years, as many of the primary studies included in this review were only published in the last few years, and thus older systematic reviews would not be comparable.
Searching other resources
We searched the Physician Data Query (the US National Cancer Institute's comprehensive cancer database), three clinical trials registers (the ISRCTN registry: www.isrctn.com; ClinicalTrials.gov: www.clinicaltrials.gov; and the National Cancer Institute clinical trials portal, www.cancer.gov/about-cancer/treatment/clinical-trials) and the National Research Register (NRR) for ongoing trials on 29 September 2022. We also sought details of ongoing or unpublished trials from the FDA (Food and Drug Administration, the regulatory body for medicines within the USA, www.fda.gov), the EMEA (European Medicines Agency, the drug regulatory body within Europe, www.emea.europa.eu) and from pharmaceutical company sources. We contacted the main investigators of the relevant completed and ongoing trials for further information.
Correspondence
We contacted authors of relevant trials to ask if they knew of further data which may or may not have been published.
Data collection and analysis
Selection of studies
For this update of the review, we downloaded all titles and abstracts retrieved by electronic searching to Covidence (Covidence), and removed duplicates. Two review authors (a combination of KG, SP, YC, MAEA and JM) independently examined the remaining references. We excluded those references which clearly did not meet the inclusion criteria and obtained copies of the full texts of potentially relevant references. Two review authors (a combination of KG, SP, YC, MAEA and JM) independently assessed the eligibility of retrieved papers. We resolved disagreements by discussion between the two review authors and, when necessary, by involving a third review author (ER and/or JM). We listed studies that initially appeared to meet the inclusion criteria but that we later excluded in the Characteristics of excluded studies table, with the reasons for exclusion.
Data extraction and management
For included studies, we extracted the following data.
-
Author, year of publication and journal citation (including language)
-
Country
-
Setting
-
Inclusion and exclusion criteria
-
Study design, methodology
-
Study population
-
Total number enrolled
-
Participant characteristics
-
Age
-
Comorbidities
-
Previous treatment
-
-
Total study duration
-
Total number of intervention groups
-
Ovarian cancer details at diagnosis
-
FIGO stage (FIGO = Fédération Internationale de Gynécologie et d'Obstétrique or International Federation of Gynaecology and Obstetrics), derived from FIGO's five‐stage typology of cancer (stages 0 to 4)
-
Histological cell type
-
Tumour grade
-
Extent of disease
-
-
Intervention details
-
Type of angiogenesis inhibitor
-
Dose
-
Duration of treatment
-
Consolidation treatment or treatment of active disease
-
-
Comparison details
-
Type of control: conventional chemotherapy or no treatment
-
Dose (if appropriate)
-
Duration (if appropriate)
-
-
Deviations from protocol
-
Risk of bias in study (see Assessment of risk of bias in included studies below)
-
Duration of follow‐up
-
Outcomes: OS, PFS, QoL, toxicity.
-
For each outcome: outcome definition (with diagnostic criteria if relevant).
-
Unit of measurement (if relevant).
-
For scales: upper and lower limits, and whether high or low score is good.
-
Results: number of participants allocated to each intervention group.
-
For each outcome of interest: sample size; missing participants.
-
We extracted data on outcomes as follows.
-
For time‐to‐event data (OS and PFS), we extracted the log of the hazard ratio (HR) (log(HR)) and its standard error (SE) from trial reports. If these were not reported, we attempted to estimate them from other reported statistics using the methods of Tierney 2007.
-
For dichotomous outcomes (e.g. toxicity), we extracted the number of participants in each treatment arm who experienced the outcome of interest and the number of participants assessed at an endpoint, in order to estimate a risk ratio (RR).
-
For continuous outcomes (e.g. QoL measures), we abstracted the mean difference (MD) and standard deviation (SD) between the final value of the outcome measure in each treatment arm at the end of follow‐up. If SDs of final values were not available, we used change scores if their SDs were available. If no SDs were available, we omitted these trials from the analyses.
When reported, we extracted both unadjusted and adjusted statistics. Where we extracted adjusted results, we recorded the variables that were adjusted for. Where possible, all data that we extracted were those relevant to an intention‐to‐treat (ITT) analysis, in which participants were analysed in the groups to which they were assigned. We noted the time points at which outcomes were collected and reported. In the case of repeated reporting of outcome measurement, we used the data with the longest follow‐up. Where time‐to‐event outcomes were assessed by more than one method (e.g. independent radiology review, investigator assessment or independent oncology review), we used the independent radiology review data. For toxicity, we recorded whether the outcomes were reported as any adverse event or drug‐related adverse event. If both were given, we used any adverse events. Where a trial evaluated the same drug in two or more different doses, we extracted all the combined data, but in the data synthesis we used only the estimated individual data for the most efficacious dose/regimen versus the comparator.
For this update of the review, two review authors (a combination of KG, SP, YC, MAEA, AT, ER and JM) extracted data onto a data extraction form specially designed for the review. The review authors resolved differences by discussion or by appeal to a third review author (JM and/or ER) when necessary.
Assessment of risk of bias in included studies
We assessed the risk of bias in included RCTs using Cochrane's risk of bias tool (Higgins 2011). This included assessment of the following domains:
-
random sequence generation (selection bias);
-
allocation concealment (selection bias);
-
blinding of participants and personnel (performance bias);
-
blinding of outcome assessment (detection bias);
-
incomplete outcome data (attrition bias). We recorded the proportion of participants whose outcomes were not reported at the end of the study; we noted whether loss to follow‐up was not reported. We coded a satisfactory level of loss to follow‐up for each outcome as:
-
'low risk', if fewer than 20% of participants were lost to follow‐up and reasons for loss to follow‐up were similar in both treatment arms;
-
'high risk', if more than 20% of participants were lost to follow‐up or reasons for loss to follow‐up differed between treatment arms; and
-
'unclear risk' if loss to follow‐up was not reported.
-
-
selective reporting of outcomes (reporting bias);
-
other possible sources of bias.
Two review authors (a combination of KG, SP, YC, MAEA, AT, ER and JM) independently applied the risk of bias tool and resolved differences by discussion or by appeal to a third review author (JM or ER). We have presented results in both a risk of bias graph and a risk of bias summary. We interpreted the results of meta‐analyses in light of the findings with respect to risk of bias.
Measures of treatment effect
We used hazard ratios (HR) with 95% confidence intervals (CIs) for time‐to‐event data, risk ratios (RR) with 95% CIs for dichotomous outcomes, and mean differences (MD) with 95% CIs for continuous outcomes.
Unit of analysis issues
The unit of analysis was the individual participant. As some of the eligible trials included women with recurrent EOC regardless of their platinum sensitivity status (sensitive and resistant), we decided to incorporate these trials into the analyses along with trials including solely women with recurrent platinum‐resistant EOC.
Dealing with missing data
We did not impute missing outcome data for any of the outcomes.
Assessment of heterogeneity
We assessed heterogeneity between trials by visually inspecting forest plots, by estimating the percentage of heterogeneity (I² statistic) between trials that could not be ascribed to sampling variation (Higgins 2003), and by performing a formal statistical test of the significance of the heterogeneity, the Chi² test (Deeks 2001). We regarded statistical heterogeneity as substantial if the I² statistic was greater than 50% and either the t2 test (a measure of between‐study variance) was greater than zero, or the P value of the Chi² test was less than 0.10. If there was evidence of substantial heterogeneity, we investigated the possible reasons for this and reported it.
Assessment of reporting biases
There was an insufficient number of included studies per analysis to adequately evaluate the potential for small study effects, such as publication bias, using funnel plots.
Data synthesis
Where deemed clinically and methodologically appropriate, we meta‐analysed trial data. Our main approach was to pool data in a two‐stage, fixed‐effect, inverse‐variance meta‐analysis based on the assumption that all studies included in a given comparison were conducted under sufficiently similar conditions and in similar populations. We applied a random‐effects, inverse‐variance model in comparisons with platinum‐resistant EOC where we included data from studies that evaluated the effect of treatment options in a population with recurrent EOC regardless of platinum‐sensitivity status. If the outcome was rare (few events), we used the Mantel‐Haenszel, fixed‐effect model.
Dealing with multi‐arm trials
The GOG‐0218 2019 trial had multiple treatment groups (three‐arm trial). We divided the control group between the treatment groups and treated comparisons between each treatment group and a split control group as independent comparisons for all adverse event outcomes. This was not necessary for overall survival, as we obtained HR estimates from a Cox regression model.
Dealing with non‐proportional hazards
If studies identified non‐proportional hazards, we used the reported hazard ratios as a measure of the effect, if reported. However, we indicated the detection of non‐proportionality, reported value of the log‐rank test and alternative measure of the effect (e.g. restricted mean survival times) if available.
Subgroup analysis and investigation of heterogeneity
For populations with newly diagnosed EOC, we added a subgroup analysis by risk of disease progression. Although this was a post hoc subgroup analysis, it is in line with the current use of bevacizumab in clinical practice.
We stratified all tyrosine kinase inhibitors (TKI) analyses by the inhibitor type, with data pooled only for the TKIs with a similar mechanism of action.
Sensitivity analysis
We performed sensitivity analysis for survival outcomes (overall survival, progression‐free survival) for one comparison (the combination of TKI with chemotherapy in recurrent platinum‐resistant EOC) to explore the influence of our decision to incorporate in the analysis data from studies in recurrent EOC that recruited mixed populations (i.e. participants with platinum‐sensitive EOC, platinum‐resistant EOC, or unclear platinum sensitivity).
Summary of findings and assessment of the certainty of the evidence
We generated summary of findings tables for the most clinically relevant comparisons (chemotherapy with bevacizumab versus chemotherapy alone, and chemotherapy with TKI versus chemotherapy alone) (Schünemann 2017a), for the following outcomes:
-
overall survival (OS);
-
progression‐free survival (PFS);
-
quality of life (QoL);
-
adverse events: overall severe adverse events (Grade 3+);
-
adverse events: hypertension (Grade 2+);
-
adverse events: bowel fistula/perforation (Grade 3+).
For each assumed risk cited in the tables, we provided a rationale and used the GRADE system to rank the certainty of the evidence (Schünemann 2017b). If the following limitations were present, we downgraded the evidence certainty by one or two levels, according to the seriousness of the limitation: study design limitations, inconsistency, imprecision, indirectness and publication bias. Where the evidence was based on single studies, or where there was no evidence for a specific outcome, we included the outcome in the summary of findings tables and graded or explained accordingly. We reported and interpreted results based on the interactive GRADEpro summary of findings table guidance (Schünemann 2019).
Results
Description of studies
For details of included, excluded, 'awaiting classification' and ongoing studies, see: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies, respectively.
Results of the search
This review was first published in 2011 (Gaitskell 2011), when we identified five relevant randomised controlled trials (with 20 references) from an initial de‐duplicated yield of 4248 references. These studies had at least published preliminary results in the form of conference abstracts.
Updated searches (starting in November 2010, with the most recent conducted on 30 September 2022) yielded the following:
-
CENTRAL: 631 references;
-
MEDLINE: 2617 references;
-
Embase: 2638 references.
Following preliminary de‐duplication across the databases, the combined total yield was 5339 references. We found an additional five published articles from five included studies by handsearching after 30 September 2022. We retrieved trial protocols (e.g. from the MetaRegister of Controlled Trials, or the ClinicalTrials.gov website) for 60 studies (across all four categories: Included, Excluded, Ongoing, or Awaiting classification studies). See Figure 2 for details of the results of this search update. In total, we sifted 5406 references following de‐duplication for this update of the review, and we excluded 5124 by screening titles and abstracts.
PRISMA flow diagram of studies considered for this review update. Please see previous version of review for further details of previous search (Gaitskell 2011).
We identified 95 studies (282 references) as potentially eligible for this review through title and abstract screening by two independent review authors (two of KG, SP, YC, MAEA and JM). We excluded 33 studies (55 references) after obtaining the full texts, for the reasons described in the Excluded studies section. We identified 50 studies (227 references) with at least preliminary results data published, which were suitable for inclusion, including the five studies (20 references) already included in the previous version of this review. Of these, 36 contributed data to the meta‐analysis. We summarise the characteristics of these studies in Table 1.
From searching online registers of clinical trials, we identified 15 ongoing studies (18 references) which are likely to be relevant for this review when completed (see Characteristics of ongoing studies). We identified two further studies which may or may not be relevant to this review when they are completed or we are able to obtain full texts (see Studies awaiting classification).
Included studies
We included 50 RCTs with a total of 14,836 participants (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; AMBITION 2022; ANTHALYA 2017; APPROVE 2022; AURELIA 2014; AVANOVA2 2019; BAROCCO 2022; CHIVA 2019; Cong 2019; Duska 2020; EORTC‐1508 2021; GEICO‐1205 2019; GOG‐0213 2017; GOG‐0218 2019; GOG‐0241 2019; Gotlieb 2012; Gupta 2019; Hainsworth 2015; Herzog 2013; ICON6 2021; ICON7 2015; Karlan 2012; Ledermann 2011; Li 2019; Li 2021; Liu 2019a; Liu 2019b; Liu 2021a; Liu 2022; Matulonis 2019; McGuire 2018; METRO‐BIBF 2020; MITO‐11 2015; MITO‐16b 2021; NICCC 2020; Nishikawa 2020; OCEANS 2015; OCTOVA 2021; Reyners 2012; Richardson 2018; Roque 2022; Sharma 2021; SWOG‐S0904 2014; TAPAZ 2022; TRIAS 2018; TRINOVA‐1 2016; TRINOVA‐2 2017; TRINOVA‐3 2019; Zhao 2015).
Five of these studies were included in the previous version of the review, using information from conference abstracts and presentations (GOG‐0218 2019; Gotlieb 2012; ICON7 2015; Karlan 2012; Ledermann 2011).
Eleven studies ‐ now included in the review ‐ were identified in the previous version of this review as ongoing studies (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; AURELIA 2014; GOG‐0213 2017; Hainsworth 2015; Herzog 2013; ICON6 2021; McGuire 2018; OCEANS 2015; TRIAS 2018; TRINOVA‐1 2016). In some cases, they were listed under their ClinicalTrials.gov reference IDs.
We grouped the included studies by the main clinically relevant settings that we used for analyses, looking separately at newly‐diagnosed EOC and recurrent EOC, which we further divided into platinum‐sensitive EOC, platinum‐resistant EOC, and those with mixed populations (i.e. participants with platinum‐sensitive EOC, platinum‐resistant EOC, or unclear platinum‐sensitivity). Details of included studies are summarised by category in Table 1.
Newly‐diagnosed epithelial ovarian cancer (EOC)
We included 13 studies evaluating the effect of angiogenesis inhibitors in 7708 participants with newly‐diagnosed EOC (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; ANTHALYA 2017; CHIVA 2019; GEICO‐1205 2019; GOG‐0218 2019; GOG‐0241 2019; Hainsworth 2015; Herzog 2013; ICON7 2015; Reyners 2012; TRINOVA‐3 2019; Zhao 2015).
GOG‐0241 2019 was a study in mucinous epithelial ovarian cancer and contained participants with both newly‐diagnosed and recurrent disease. Since the original review protocol, it has become clear that this disease differs from other types of EOC. Thus, we did not include this study in meta‐analysis.
Bevacizumab
Six trials evaluated use of bevacizumab in newly‐diagnosed EOC (ANTHALYA 2017; GEICO‐1205 2019; GOG‐0218 2019; GOG‐0241 2019; ICON7 2015; Zhao 2015).
ANTHALYA 2017 was a multi‐centre, open‐label, phase II trial, evaluating the effect of adding bevacizumab (three cycles) to neoadjuvant chemotherapy (four cycles of carboplatin and paclitaxel) in 95 women with initially unresectable FIGO stage IIIC/IV ovarian cancer. After interval debulking surgery, all women received adjuvant chemotherapy (four cycles of carboplatin‐paclitaxel), with bevacizumab added in the sixth cycle and continued as maintenance therapy (up to 26 cycles). The primary outcome was the complete resection rate; the secondary outcome was safety/adverse events.
GEICO‐1205 2019 was an open‐label, phase II trial evaluating the effect of adding bevacizumab (three cycles) to neoadjuvant chemotherapy (four cycles of carboplatin and paclitaxel) in 68 women with newly‐diagnosed stage III/IV high‐grade serous or endometrioid EOC. After interval debulking surgery, all women received adjuvant chemotherapy (three cycles), with bevacizumab alongside and continued as maintenance therapy for up to 15 months. The primary outcome was the complete macroscopic response rate at interval debulking surgery; secondary outcomes included safety, surgical operability, optimal cytoreduction, response rate, and progression‐free survival.
GOG‐0218 2019 was a double‐blind, phase III trial in 1873 women with newly‐diagnosed, incompletely resected, stage III/IV EOC. The study compared three regimens of treatment: chemotherapy alone (six cycles of carboplatin and paclitaxel) versus chemotherapy plus concurrent bevacizumab (in cycles 2 to 6) versus chemotherapy plus concurrent and maintenance bevacizumab (cycles 2 to 22). The primary outcome was progression‐free survival (changed from overall survival during the trial); secondary outcomes included overall survival, safety, and quality of life (as measured by the Functional Assessment of Cancer Therapy‐Ovary Trial Outcome Index (FACT‐O TOI)).
GOG‐0241 2019 was an open‐label, multi‐centre, phase III factorial trial in 50 women with mucinous EOC of FIGO stage II‐IV or recurrent stage I. The study randomised participants to four different treatment arms, evaluating the effect of adding bevacizumab to either chemotherapy with carboplatin and paclitaxel, or chemotherapy with oxaliplatin and capecitabine. The primary outcome was intended to be overall survival; secondary outcomes were progression‐free survival, response rate, toxicity, and quality of life. The trial was stopped early by the data monitoring committee due to difficulty recruiting in this rare histological type of EOC.
ICON7 2015 was a multi‐centre, open‐label, phase III trial in 1528 women with newly‐diagnosed EOC that was either high‐risk early‐stage disease (high‐grade stage I‐IIa) or more advanced disease (FIGO IIb‐IV). The study compared chemotherapy (six cycles of carboplatin and paclitaxel) to chemotherapy plus bevacizumab (given concurrently with chemotherapy and then as maintenance therapy for up to 12 further cycles). The primary outcome was progression‐free survival; secondary outcomes were overall survival and safety; exploratory outcomes included quality of life, health economics, and translational research.
Zhao 2015 was a phase III trial evaluating the effect of adding bevacizumab to cisplatin (both delivered intraperitoneally) in 58 women with EOC and malignant ascites (i.e. an accumulation of fluid in the abdomen due to the ovarian cancer). The primary outcome was the objective response rate of partial or complete remission of the ascites; secondary outcomes included safety, the number of required peritoneal drainages during the trial, the speed of peritoneal drainage, and quality of life (measured by Karnofsky Performance Status). The trial population was somewhat unclear, but appeared to be mainly women with newly‐diagnosed EOC (as prior anti‐tumour treatment was an exclusion criterion).
Tyrosine kinase inhibitors (TKIs)
Six trials evaluated the use of tyrosine kinase inhibitors (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; CHIVA 2019; Hainsworth 2015; Herzog 2013; TRINOVA‐3 2019).
AGO‐OVAR 12 2020 was a double‐blind, placebo‐controlled, phase III trial evaluating the effect of adding nintedanib to chemotherapy (carboplatin and paclitaxel), in 1366 women with newly‐diagnosed FIGO stage IIB‐IV EOC. The primary outcome was progression‐free survival; secondary endpoints included overall survival, time to tumour marker progression, safety and quality of life (as measured by the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire ovarian cancer module (EORTC QLQ‐OV 28) and EORTC core quality of life (QLQ‐C30) questionnaires).
AGO‐OVAR 16 2019 was a multi‐centre, double‐blind, placebo‐controlled, phase III trial evaluating the effect of adding pazopanib maintenance therapy for up to 24 months after first‐line chemotherapy (with platinum and a taxane) in 940 women with newly diagnosed FIGO stage II‐IV EOC. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and quality of life.
CHIVA 2019 was a multi‐centre, placebo‐controlled, phase II trial evaluating the effect of adding nintedanib to neoadjuvant chemotherapy (three to four cycles of carboplatin and paclitaxel) prior to interval debulking surgery, and continuing nintedanib as maintenance therapy for up to two years, in 188 women with FIGO stage IIIC‐IV EOC. The primary outcome was progression‐free survival; the secondary outcome was response rate.
Hainsworth 2015 was a multi‐centre, open‐label, phase II trial evaluating the effect of adding sorafenib to chemotherapy (up to six cycles of carboplatin and paclitaxel), and continuing sorafenib as maintenance therapy for up to a total of 12 months, in 85 women with FIGO stage III/IV EOC and residual measurable disease or elevated serum cancer antigen 125 (CA125 ‐ an ovarian tumour marker) levels after maximal surgical cytoreduction. The primary outcome was progression‐free survival; secondary outcomes included overall survival and safety.
Herzog 2013 was a multi‐centre, double‐blind, placebo‐controlled, phase II trial evaluating the effect of adding sorafenib as maintenance therapy in 246 women with FIGO stage III/IV EOC with clinical complete response following tumour debulking surgery and chemotherapy (platinum and a taxane). The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and quality of life.
TRINOVA‐3 2019 was a multi‐centre, double‐blind, phase III trial evaluating the effect of adding trebananib to chemotherapy (carboplatin and paclitaxel), and continuing trebananib as maintenance therapy for up to 18 months afterwards, in 1015 women with FIGO stage III/IV EOC. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety, pharmacokinetics and patient‐reported outcomes.
Other anti‐angiogenic agents
Reyners 2012 was an open‐label, phase II trial evaluating the effect of adding the selective cyclo‐oxygenase‐2 (COX‐2) inhibitor, celecoxib, to first‐line chemotherapy (docetaxel and carboplatin) in 196 women with newly‐diagnosed ovarian cancer. The primary outcomes were response rate and progression‐free survival; secondary outcomes included safety/toxicity and overall survival.
Recurrent epithelial ovarian cancer
We included 37 studies evaluating the effect of angiogenesis inhibitors in 7128 participants with solely recurrent EOC: nine studies in 3034 women with recurrent platinum‐sensitive disease (AVANOVA2 2019; Cong 2019; GOG‐0213 2017; ICON6 2021; Li 2019; Liu 2019b; Liu 2022; MITO‐16b 2021; OCEANS 2015), 19 in 2174 women with recurrent platinum‐resistant disease (AMBITION 2022; APPROVE 2022; AURELIA 2014; BAROCCO 2022; EORTC‐1508 2021; Gotlieb 2012; Li 2021; Liu 2019a; Liu 2021a; McGuire 2018; METRO‐BIBF 2020; MITO‐11 2015; NICCC 2020; Nishikawa 2020; OCTOVA 2021; Roque 2022; Sharma 2021; SWOG‐S0904 2014; TRIAS 2018), and nine studies in 1920 participants recruited regardless of their platinum sensitivity status (Duska 2020; Gupta 2019; Karlan 2012; Ledermann 2011; Matulonis 2019; Richardson 2018; TAPAZ 2022; TRINOVA‐1 2016; TRINOVA‐2 2017).
Platinum‐sensitive recurrent EOC
Bevacizumab
Five studies evaluated the effect of adding bevacizumab to chemotherapy in women with platinum‐sensitive recurrent ovarian cancer (Cong 2019; GOG‐0213 2017; Li 2019; MITO‐16b 2021; OCEANS 2015). Cong 2019 evaluated the effect of adding bevacizumab to chemotherapy (carboplatin and paclitaxel) in 164 women with platinum‐sensitive recurrent ovarian cancer. The primary outcome was the objective response rate; secondary outcomes included progression‐free survival, overall survival, safety and quality of life. GOG‐0213 2017 was a multi‐centre, open‐label, phase III trial evaluating the effect of adding bevacizumab to chemotherapy (carboplatin and paclitaxel) in 674 women with platinum‐sensitive recurrent ovarian cancer. The primary outcome was overall survival; secondary outcomes included progression‐free survival, safety and quality of life. Li 2019 evaluated the effect of adding bevacizumab to chemotherapy (carboplatin and paclitaxel) in 68 women with platinum‐sensitive recurrent ovarian cancer. Outcomes included safety and "clinical efficacy". MITO‐16b 2021 was a multi‐centre, open‐label, phase III trial evaluating the effect of adding bevacizumab to chemotherapy (carboplatin plus one of paclitaxel, gemcitabine, or pegylated liposomal doxorubicin) in 406 women with platinum‐sensitive recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, objective response rate, safety and biomarkers. OCEANS 2015 was a multi‐centre, blinded, placebo‐controlled, phase III trial evaluating the effect of adding bevacizumab to chemotherapy (carboplatin and gemcitabine) in 484 women with platinum‐sensitive recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included objective response rate, duration of response, overall survival and safety. AVANOVA2 2019 was an open‐label, phase II trial evaluating the effect of adding bevacizumab to the PARP inhibitor niraparib in 97 women with platinum‐sensitive and partially platinum‐sensitive recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included the objective response rate, patient‐reported outcomes and safety/tolerability; exploratory outcomes included overall survival.
Tyrosine kinase inhibitors
One study evaluated the effect of adding a tyrosine kinase inhibitor to chemotherapy in women with platinum‐sensitive ovarian cancer (ICON6 2021). ICON6 2021 was a multi‐centre, double‐blind, placebo‐controlled, phase III trial evaluating the effect of adding cediranib alongside chemotherapy (carboplatin with either paclitaxel or gemcitabine), or cediranib alongside chemotherapy followed by cediranib maintenance therapy, compared to chemotherapy alone (three arms), in 486 women with platinum‐sensitive recurrent ovarian cancer. The primary outcome was progression‐free survival (changed during the trial); secondary outcomes included overall survival, safety and quality of life.
Two studies evaluated a tyrosine kinase inhibitor in combination with a PARP inhibitor in women with platinum‐sensitive recurrent ovarian cancer (Liu 2019b; Liu 2022). Liu 2022 was an open‐label, three‐arm, phase III trial evaluating the effect of adding cediranib to the PARP inhibitor olaparib, compared to olaparib alone, or to chemotherapy (carboplatin with either paclitaxel, gemcitabine or liposomal doxorubicin), in 565 women with recurrent platinum‐sensitive ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and patient‐reported outcomes. Liu 2019b was a multi‐centre, open‐label, phase II trial evaluating the effect of adding cediranib to olaparib in 90 women with recurrent platinum‐sensitive ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included the objective response rate, safety/toxicity and overall survival.
Platinum‐resistant recurrent EOC
Nineteen studies evaluated the use of angiogenesis inhibitors in platinum‐resistant recurrent ovarian cancer (AMBITION 2022; APPROVE 2022; AURELIA 2014; BAROCCO 2022; EORTC‐1508 2021; Gotlieb 2012; Li 2021; Liu 2019a; Liu 2021a; McGuire 2018; METRO‐BIBF 2020; MITO‐11 2015; NICCC 2020; Nishikawa 2020; OCTOVA 2021; Roque 2022; Sharma 2021; SWOG‐S0904 2014; TRIAS 2018).
Bevacizumab
Six studies evaluated the effect of adding bevacizumab to chemotherapy in women with platinum‐resistant recurrent EOC (AURELIA 2014; Li 2021; Liu 2019a; Liu 2021a; Nishikawa 2020; Roque 2022). AURELIA 2014 was an open‐label, phase III trial evaluating the effect of adding bevacizumab to single‐agent chemotherapy (investigator choice of pegylated liposomal doxorubicin, weekly paclitaxel or topotecan) in 361 women with platinum‐resistant EOC. The primary outcome was progression‐free survival; secondary outcomes included objective response rate, overall survival, safety, tolerability and quality of life. Li 2021 was a study with 70 participants, which appears to have compared albumin‐binding paclitaxel monotherapy (days 1, 8 and 15) with or without bevacizumab. Treatment was for six cycles and there does not appear to have been a maintenance phase following chemotherapy. The details available are limited to an English‐language abstract and we have not been able to access a copy of the full‐text (Chinese) paper to date. Details are therefore unclear, including risk of bias in all categories and whether the study was blinded or open label. Outcomes included progression‐free survival, overall survival, objective response rate, adverse reactions, quality of life and immune function. As this study is at such high risk of bias, we have not included its results in the meta‐analyses, and will not do so until the full text paper can be interrogated. Liu 2019a evaluated the effect of adding bevacizumab to chemotherapy (albumin‐bound paclitaxel) in 86 women with platinum‐resistant EOC. The primary outcome was the objective response rate; secondary outcomes included progression‐free survival, overall survival and safety/ toxicity. Liu 2021a evaluated the effect of adding bevacizumab to chemotherapy (liposomal doxorubicin) in 76 women with platinum‐resistant recurrent EOC. Primary outcomes included the objective response rate and disease control rate; secondary outcomes included progression‐free survival, overall survival and safety. The study was open‐label and follow‐up ranged from 3.2 to 30.0 months. Insufficient data were provided to allow us to extract hazard ratio (HR) data for overall survival and progression‐free survival. Adverse outcomes were presented as total number of adverse events, not by participant, and were not categorised by grade of adverse event. We were therefore not able to include these data in the meta‐analyses. Nishikawa 2020 was an open‐label, phase II trial evaluating the effect of adding bevacizumab to single‐agent chemotherapy in 103 women with platinum‐resistant EOC. All participants had previously been treated with at least three cycles of bevacizumab in previous lines of chemotherapy. The primary outcome was progression‐free survival; secondary outcomes included objective response rate, safety and tolerability. Roque 2022 was a multi‐centre, phase II trial evaluating the effect of adding bevacizumab to chemotherapy (ixabepilone, a microtubule‐stabilising agent) in 76 women with platinum‐resistant/refractory ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and response rates.
One other study evaluated bevacizumab in platinum‐resistant recurrent ovarian cancer (EORTC‐1508 2021). EORTC‐1508 2021 was a multi‐centre, phase II trial that evaluated the effect of adding bevacizumab to atezolizumab (an immunotherapy drug), and also the effect of adding bevacizumab to atezolizumab and acetylsalicylic acid (aspirin, an inhibitor of the cyclo‐oxygenase enzymes COX 1/2), in 122 women with platinum‐resistant recurrent EOC. The primary outcome was progression‐free survival at six months; secondary outcomes included safety, progression‐free survival, response rate and time to first subsequent therapy.
A further study evaluated a different antibody targeting angiogenesis in platinum‐resistant ovarian cancer (Gotlieb 2012). Gotlieb 2012 was a double‐blind, placebo‐controlled, phase II study evaluating the effect of giving aflibercept in 55 women with chemotherapy‐resistant ovarian cancer and recurrent symptomatic malignant ascites (accumulated fluid in the abdomen due to the ovarian cancer). The primary outcome was the time to repeat paracentesis (draining of the abdominal fluid); secondary outcomes included the frequency of paracentesis, safety and patient‐reported outcomes.
Olaratumab
Olaratumab (IMC‐3G3) is a monoclonal antibody targeting platelet‐derived growth factor alpha (PDGFRα). McGuire 2018 was a randomised, open‐label, phase II study evaluating olaratumab plus liposomal doxorubicin compared with liposomal doxorubicin alone in 123 participants with platinum‐resistant and platinum‐refractory recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, objective response rate, duration of response and safety.
Tyrosine kinase inhibitors
Seven studies evaluated the effect of adding a tyrosine kinase inhibitor to chemotherapy in women with platinum‐resistant recurrent EOC (APPROVE 2022; METRO‐BIBF 2020; MITO‐11 2015; NICCC 2020; Sharma 2021; SWOG‐S0904 2014; TRIAS 2018). APPROVE 2022 was a multi‐centre, open‐label, phase II trial evaluating the effect of adding apatinib to chemotherapy (pegylated liposomal doxorubicin) in 150 women with platinum‐resistant recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, objective response rate, disease control rate and safety. METRO‐BIBF 2020 was a double‐blind, placebo‐controlled, phase II trial evaluating the effect of adding nintedanib to chemotherapy (low dose metronomic cyclophosphamide) in 117 women with platinum‐resistant recurrent ovarian cancer. The primary outcome was overall survival; secondary outcomes included progression‐free survival, response rate, toxicity and quality of life. MITO‐11 2015 was an open‐label, phase II trial evaluating the effect of adding pazopanib to chemotherapy (weekly paclitaxel) in 74 women with platinum‐resistant or platinum‐refractory ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and objective response rate. NICCC 2020 was an open‐label, phase II trial comparing nintedanib 200 mg or chemotherapy (paclitaxel, pegylated liposomal doxorubicin, or topotecan) in 91 participants with relapsed clear cell carcinoma. The primary outcome was progression‐free survival in participants with clear cell EOC. Secondary outcomes included overall survival, response rate, disease control rate and patient reported outcomes. Sharma 2021 was an open‐label, phase II trial evaluating the effect of adding pazopanib to chemotherapy (etoposide and cyclophosphamide) in 75 women with platinum resistant/refractory EOC. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety and quality of life. SWOG‐S0904 2014 was a phase II trial evaluating the effect of adding vandetanib to chemotherapy (docetaxel) in 129 women with platinum‐resistant, refractory, persistent or recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included response to treatment, overall survival and translational studies. TRIAS 2018 was a multi‐centre, double‐blind, placebo‐controlled, phase II trial evaluating the effect of adding sorafenib to chemotherapy (topotecan) in 174 women with platinum‐resistant/refractory recurrent ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, objective response rate, patient‐reported outcomes and safety/tolerability.
Three other studies evaluated the use of tyrosine kinase inhibitors in platinum‐resistant recurrent ovarian cancer. AMBITION 2022 was an umbrella study of biomarker‐driven targeted therapy. One of the comparisons evaluated the effect of adding cediranib to the PARP inhibitor olaparib, compared to adding the immunotherapy drug durvalumab to olaparib, in 30 women with homologous recombination‐deficient platinum‐resistant ovarian cancer. The primary outcome was the response rate; secondary outcomes included progression‐free survival, overall survival, safety, immune‐related response criteria and duration of response. BAROCCO 2022 was a phase II trial evaluating the effect of adding cediranib to the PARP inhibitor olaparib (in a continuous or intermittent schedule), compared to chemotherapy (paclitaxel) in 123 women with recurrent platinum‐resistant high‐grade EOC. The primary outcome was progression‐free survival; secondary outcomes included treatment compliance, reasons for discontinuation and treatment modification, objective response rate, partial response, overall survival and quality of life. OCTOVA 2021 was an open‐label, three‐arm, phase II trial evaluating the effect of adding cediranib to the PARP inhibitor olaparib, compared to olaparib alone, or to chemotherapy (paclitaxel), in 139 participants. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety/tolerability, objective response rate and quality of life.
Mixed recurrent EOC (platinum‐sensitive, platinum‐resistant and unclear)
Ten studies included mixed populations of women with platinum‐sensitive and platinum‐resistant recurrent ovarian cancer (Duska 2020; Gupta 2019; Karlan 2012; Ledermann 2011; Matulonis 2019; NICCC 2020; Richardson 2018; TAPAZ 2022; TRINOVA‐1 2016; TRINOVA‐2 2017).
Six studies evaluated the addition of a tyrosine kinase inhibitor to conventional therapy in women with recurrent ovarian cancer regardless of platinum‐sensitivity status (Duska 2020; Karlan 2012; Richardson 2018; TAPAZ 2022; TRINOVA‐1 2016; TRINOVA‐2 2017). Duska 2020 was a multi‐centre, open‐label, phase II trial evaluating the effect of adding pazopanib to chemotherapy (weekly gemcitabine) in 148 women with persistent or recurrent EOC. The trial population included both platinum‐resistant and platinum‐sensitive participants. The primary outcome was progression‐free survival; secondary outcomes included overall survival and safety/toxicity. Karlan 2012 was a double‐blind, placebo‐controlled, three‐arm phase II trial evaluating the effect of adding trebananib (AMG 386) (at either a higher or a lower dose, in two separate trial arms) to chemotherapy (weekly paclitaxel) in 161 women with recurrent ovarian cancer. The trial population included platinum‐refractory, platinum‐resistant and platinum‐sensitive participants. The primary outcome was progression‐free survival; secondary outcomes included overall survival and safety. Richardson 2018 was a multi‐centre, double‐blind, placebo‐controlled phase II trial evaluating the effect of adding pazopanib to chemotherapy (paclitaxel) in 106 women with persistent or recurrent ovarian cancer. The trial population included platinum‐resistant and platinum‐sensitive participants. The primary outcome was progression‐free survival; secondary outcomes included safety, overall survival, proportion responding and duration of response. TAPAZ 2022 was a phase II trial evaluating the effect of adding pazopanib to weekly paclitaxel in 116 women with ovarian cancer who relapsed during bevacizumab maintenance therapy. The trial population included platinum‐resistant and platinum‐sensitive participants. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety, pharmacokinetics and quality of life. TRINOVA‐1 2016 was a multi‐centre, double‐blind, placebo‐controlled, phase III trial evaluating the effect of adding trebananib to chemotherapy (weekly paclitaxel) in 919 women with recurrent EOC (the trial included a mixture of women with platinum‐resistant and potentially platinum‐sensitive disease). The primary outcome was progression‐free survival; secondary outcomes included overall survival, objective response rate, safety and patient‐reported outcomes. TRINOVA‐2 2017 was a double‐blind, placebo‐controlled, phase III trial evaluating the effect of adding trebananib to chemotherapy (pegylated liposomal doxorubicin) in 223 women with recurrent partially platinum‐sensitive or platinum‐resistant ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, safety/toxicity, objective response rate and duration of response.
Four studies evaluated a tyrosine kinase inhibitor plus other agents or a tyrosine kinase inhibitor on its own, in a mixed population of women with platinum‐sensitive and platinum‐resistant recurrent EOC (Gupta 2019; Ledermann 2011; Matulonis 2019; NICCC 2020). Gupta 2019 was a phase II trial evaluating the effect of adding celecoxib to chemotherapy (oral cyclophosphamide) in 52 women with recurrent or persistent EOC. The trial population included platinum‐refractory, platinum‐resistant and platinum‐sensitive participants. The primary outcome was response rate; secondary outcomes included safety/toxicity, time to treatment failure and overall survival. Ledermann 2011 was a double‐blind, placebo‐controlled, phase II trial evaluating the effect of giving nintedanib (BIBF 1120) maintenance therapy for up to 36 weeks in 84 women who had recently completed chemotherapy for recurrent ovarian cancer. Matulonis 2019 was an open‐label, phase II trial evaluating cabozantinib versus chemotherapy (weekly paclitaxel) in 111 women with persistent or recurrent EOC. The trial population included platinum‐resistant and platinum‐sensitive participants. The primary outcome was progression‐free survival; secondary outcomes included overall survival and safety/toxicity. NICCC 2020 was a multi‐centre, open‐label, phase II trial evaluating nintedanib versus chemotherapy (paclitaxel, pegylated liposomal doxorubicin or topotecan) in 91 women with recurrent clear cell ovarian cancer. The primary outcome was progression‐free survival; secondary outcomes included overall survival, response rate, disease control rate and patient‐reported outcomes. The investigators did not attempt to restrict the study population by platinum sensitivity, as clear cell ovarian cancer often shows poor response to platinum chemotherapy.
Excluded studies
In line with Cochrane guidelines, we describe here only those studies assessed and excluded at the full‐text screening phase, which might plausibly have been included in the review (Lefebvre 2022, Section 4.6.5). In the previous version of the review, we excluded 12 studies (14 references); some of these do not now make the updated criteria for listing in the Excluded studies section. We have added additional references identified in this search update to some previously excluded studies. For this review update, we identified an additional 33 potentially relevant studies (55 references). Thus, there are now a total of 45 excluded studies (69 references). The reasons for study exclusion are as follows:
-
ineligible study design: 18 studies (Azad 2008; Campos 2013; Colombo 2012; Hagemann 2013; Harter 2013; Ikeda 2013; Jones 2019; Krasner 2019; Ma 2022; Nasu 2022; NCT01972516; OCTAVIA 2014; Schilder 2013; Tillmans 2012; Tillmans 2013; Vergote 2017; Verschraegen 2012; Wenham 2013);
-
ineligible comparator: 13 studies (Baumann 2012; NCT00017303; Ojeda 2011; PACOVAR‐trial 2011; PAZOFOS 2020; Pfisterer 2021; Ray‐Coquard 2019; Schwandt 2014; STAC 2011; Tew 2014; Tew 2018; Tredan 2022; Zhang 2020).
-
ineligible intervention (e.g. alternative drug randomised and angiogenesis inhibitor in both arms): six studies (BOOST 2011; Chan 2016; DUO‐O 2018; ENGOT‐ov65 2021; GOG‐3018 2020; Heiss 2010).
-
ineligible population: two studies (ALIENOR/ENGOT‐ov7 2020; Brown 2014).
-
not clinical trials (e.g. review articles or systematic reviews): four studies (Burger 2010; Markman 2009; Osterweil 2010; Sennino 2010; Trillsch 2021).
-
one study compared adding two cycles of intraperitoneal bevacizumab to carboplatin‐paclitaxel chemotherapy prior to primary debulking surgery and looked at short‐term outcomes (Tao 2022).
Risk of bias in included studies
For overall risk of bias and assessment of the risk of bias items for individual studies, see Figure 3 and Figure 4.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
Risk of bias summary: review authors' judgements about each risk of bias item for each included study
Allocation
Newly‐diagnosed epithelial ovarian cancer (EOC)
Six of 13 studies in newly‐diagnosed EOC specified the method of randomisation; we judged these to be at low risk of bias (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; GOG‐0218 2019; GOG‐0241 2019; ICON7 2015; TRINOVA‐3 2019). In the remaining seven studies, we judged the method of randomisation to be at unclear risk of bias (ANTHALYA 2017; CHIVA 2019; GEICO‐1205 2019; Hainsworth 2015; Herzog 2013; Reyners 2012; Zhao 2015).
Five of 13 studies in newly‐diagnosed EOC specified the method of allocation concealment; we judged these to be at low risk of bias (AGO‐OVAR 12 2020; GOG‐0218 2019; GOG‐0241 2019; ICON7 2015; TRINOVA‐3 2019). In the remaining eight studies, we judged allocation concealment to be at unclear risk of bias, typically due to limited information (AGO‐OVAR 16 2019; ANTHALYA 2017; CHIVA 2019; GEICO‐1205 2019; Hainsworth 2015; Herzog 2013; Reyners 2012; Zhao 2015).
Recurrent EOC
Nineteen of 37 studies in recurrent EOC specified the method of randomisation; we judged these to be at low risk of bias. In 14 studies, we judged the method of randomisation to be at unclear risk of bias, typically due to limited information (AMBITION 2022; Cong 2019; EORTC‐1508 2021; Gupta 2019; Karlan 2012; Ledermann 2011; McGuire 2018; Matulonis 2019; METRO‐BIBF 2020; NICCC 2020; Nishikawa 2020; OCEANS 2015; OCTOVA 2021; TRINOVA‐2 2017). We judged the remaining four studies to be at unclear risk of bias due to the methods used (e.g. random number table) (Li 2019; Li 2021; Liu 2019a; Liu 2021a).
Twenty of 37 studies in recurrent EOC specified the method of allocation concealment; we judged these to be at low risk of bias. In the remaining 17 studies, we judged the method of allocation concealment to be at unclear risk of bias, typically due to limited information (AMBITION 2022; APPROVE 2022; BAROCCO 2022; Cong 2019; EORTC‐1508 2021; Gupta 2019; Li 2019; Li 2021; Liu 2019a; Liu 2022; Liu 2021a; NICCC 2020; Nishikawa 2020; Matulonis 2019; Roque 2022; TAPAZ 2022; TRINOVA‐2 2017).
Blinding
Newly‐diagnosed EOC
In six of 13 studies in newly diagnosed EOC (ANTHALYA 2017; GEICO‐1205 2019; GOG‐0241 2019; Hainsworth 2015; ICON7 2015; Reyners 2012), interventions were not concealed (open‐label design), and hence participants and investigators were aware of the allocated treatment. As the key outcomes included progression‐free survival and adverse events, which have an element of subjectivity in their assessment, the open‐label design led to a potentially high risk of performance bias and detection bias. However, estimates for overall survival should be at low risk of bias. We judged the risk of performance and detection bias as low in six studies due to a 'double‐blind' study design (in which both participants and investigators/assessors are masked as to intervention) (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; CHIVA 2019; GOG‐0218 2019; Herzog 2013; TRINOVA‐3 2019), and as unclear in one study (Zhao 2015), as the study did not specify whether it was open‐label.
Recurrent EOC
In 22 of 37 studies in recurrent EOC, interventions were not concealed (open‐label design) (AMBITION 2022; APPROVE 2022; AURELIA 2014; AVANOVA2 2019; BAROCCO 2022; Duska 2020; GOG‐0213 2017; Gupta 2019; Liu 2019b; Liu 2021a; Liu 2022; McGuire 2018; Matulonis 2019; MITO‐11 2015; MITO‐16b 2021; NICCC 2020; Nishikawa 2020; OCTOVA 2021; Roque 2022; Sharma 2021; SWOG‐S0904 2014; TAPAZ 2022). Thus, we judged assessment of outcomes such as PFS and adverse events to be at high risk of performance bias and detection bias, though estimates for overall survival should be at low risk of bias. One multi‐arm study was described as 'double‐blind' (EORTC‐1508 2021), but appeared to be open‐label for the comparison relevant to this review (i.e. with versus without bevacizumab). Thus, we considered it to be at high risk of bias for outcomes such as PFS and adverse events. We judged four studies to be at unclear risk of bias, typically because they were unclear about whether the trial was open‐label (Cong 2019; Li 2019; Li 2021; Liu 2019a).
Incomplete outcome data
Newly‐diagnosed EOC
We did not consider incomplete outcome data to be an issue in most studies, and judged them to at low risk of bias for this domain. The exceptions were ANTHALYA 2017, where it was difficult to judge this domain, as the analysis used a modified intention‐to‐treat approach, and CHIVA 2019, where there was insufficient information to judge this domain; we judged these to be at unclear risk of bias.
Recurrent EOC
We did not consider incomplete outcome data to be an issue in most studies, and judged them to be at low risk of bias for this domain. The exceptions were five studies in which we judged the risk of bias as unclear (BAROCCO 2022; Cong 2019; Li 2021; Liu 2021a; NICCC 2020), mostly because relevant published outcome information was only available from conference abstracts. We judged BAROCCO 2022 to be at unclear risk of attrition bias, as 12 of the 17 participants who did not complete four weeks of treatment were in the control arm of a three‐arm study.
Selective reporting
Newly‐diagnosed EOC
We did not consider selective reporting of outcomes to be an issue in most studies, and judged them to be at low risk of bias for this domain. The exceptions were CHIVA 2019 and Reyners 2012, where there was insufficient information to judge this domain.
Recurrent EOC
We did not consider selective reporting of outcomes to be an issue in the majority of studies, and judged them to be at low risk of bias for this domain. However, in 11 of 37 studies, there was insufficient information to judge this domain (Cong 2019; EORTC‐1508 2021; Gupta 2019; Li 2019; Li 2021; Liu 2019a; Liu 2021a; Matulonis 2019; NICCC 2020; Nishikawa 2020; OCTOVA 2021), in some cases due to limited information from conference abstracts, and we assessed these as having an unclear risk of bias.
Other potential sources of bias
We judged 43 of 50 included studies (in both newly‐diagnosed and recurrent settings) to have an unclear risk of bias in this domain because they were either fully or partly industry‐sponsored, with at least some authors from each study disclosing a financial conflict of interest. The other studies appeared to have non‐industry funding and no declared conflicts of interest.
Effects of interventions
See: Summary of findings 1 Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone in newly‐diagnosed EOC; Summary of findings 2 Chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone in newly‐diagnosed EOC; Summary of findings 3 Chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone in newly‐diagnosed EOC; Summary of findings 4 Chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC; Summary of findings 5 Chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC ; Summary of findings 6 Chemotherapy with bevacizumab compared to chemotherapy alone in recurrent platinum‐resistant EOC; Summary of findings 7 Chemotherapy with TKI compared to chemotherapy alone in recurrent platinum‐resistant EOC; Summary of findings 8 Chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone in recurrent EOC
This review assesses the treatment options at different time points in the treatment of EOC where it may be appropriate to use angiogenesis inhibitors, each of which is a separate clinical question. These time points are: newly‐diagnosed EOC; and recurrent EOC, where disease may be further divided by platinum‐sensitivity (platinum‐sensitive recurrence; platinum‐resistant recurrence; mixed/unclear platinum sensitivity). We divided studies between these different clinical scenarios; the studies are summarised in Table 1. We present summary of findings tables for the most clinically relevant comparisons.
Newly‐diagnosed epithelial ovarian cancer (EOC)
Thirteen included studies evaluated the effect of angiogenesis inhibitors in newly‐diagnosed EOC (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; ANTHALYA 2017; CHIVA 2019; GEICO‐1205 2019; GOG‐0218 2019; GOG‐0241 2019; Hainsworth 2015; Herzog 2013; ICON7 2015; Reyners 2012; TRINOVA‐3 2019; Zhao 2015). The GOG‐0241 2019 and Zhao 2015 trials did not contribute data to quantitative synthesis due to their reporting of outcome data. Of the remaining trials, four evaluated the use of bevacizumab (ANTHALYA 2017; GEICO‐1205 2019; GOG‐0218 2019; ICON7 2015), and six evaluated the use of tyrosine kinase inhibitors (TKIs) (AGO‐OVAR 12 2020; AGO‐OVAR 16 2019; CHIVA 2019; Hainsworth 2015; Herzog 2013; TRINOVA‐3 2019), in addition to or after standard chemotherapy. One study evaluated the use of celecoxib (Reyners 2012).
We graded the certainty of the evidence of the three most clinically relevant comparisons:
-
chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone (summary of findings Table 1);
-
chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone (summary of findings Table 2);
-
chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone (summary of findings Table 3).
1. Chemotherapy with bevacizumab compared to chemotherapy alone (placebo for all in the maintenance phase)
One included study compared chemotherapy with bevacizumab to chemotherapy alone for newly‐diagnosed EOC (GOG‐0218 2019). This three‐armed study enrolled 1873 participants in total. For this comparison, 1250 women were randomised to one of two study arms (625 women to each arm).
Overall survival (OS)
Chemotherapy with bevacizumab likely results in little to no difference in OS compared to chemotherapy alone (hazard ratio (HR) 1.06, 95% confidence interval (CI) 0.94 to 1.20; 1250 participants; Analysis 1.1).
Progression‐free survival (PFS)
Chemotherapy with bevacizumab likely results in little to no difference in PFS compared to chemotherapy alone (HR 0.91, 95% CI 0.79 to 1.04; 1250 participants; Analysis 1.2).
Quality of life (QoL)
Chemotherapy with bevacizumab likely results in little to no difference in QoL at six months of follow‐up (measured using the Trial Outcome Index score of the Functional Assessment of Cancer Therapy ‐ Ovarian Cancer questionnaire) compared to chemotherapy alone (mean difference (MD) 1.80, 95% CI ‐0.32 to 3.92; 709 participants; Analysis 1.3).
Adverse events
The GOG‐0218 2019 trial did not report the effect of chemotherapy with bevacizumab on any severe adverse events (grade ≥ 3), proteinuria (grade ≥ 2), abdominal pain (grade ≥ 2), neutropenia (grade ≥3), or bowel fistula or perforation, which were the prespecified outcomes for this review. However, many studies detailed adverse events by different category groupings (e.g. grade 1‐2 and grade ≥ 3). We judged it appropriate to present the data as reported, rather than discard the data. Chemotherapy with bevacizumab compared with chemotherapy alone likely results in a large increase in hypertension (grade ≥ 2) (RR 2.33, 95% CI 1.55 to 3.26; 1208 participants; Analysis 1.4); likely results in little to no difference in proteinuria (grade ≥ 3) (RR 0.99, 95% CI 0.25 to 3.94; 1 study, 1208 participants; Analysis 1.5) and pain (grade ≥ 2) (RR 1.00, 95% CI 0.88 to 1.14; 1208 participants; Analysis 1.6); slight increases in neutropenia (grade ≥ 4) (RR 1.16, 95% CI 1.06 to 1.26; 1208 participants; Analysis 1.7); likely results in little to no difference in febrile neutropenia (any grade) (RR 1.41, 95% CI 0.82 to 2.44; 1208 participants; Analysis 1.8), venous thromboembolic events (any grade) (RR 1.02, 95% CI 0.65 to 1.60; 1208 participants; Analysis 1.9), arterial thromboembolic event (any grade) (RR 0.79, 95% CI 0.21 to 2.94; 1208 participants; Analysis 1.10), non‐central nervous system (non‐CNS) bleeding (grade ≥ 3) (RR 1.58, 95% CI 0.52 to 4.81; 1208 participants; Analysis 1.11) and gastrointestinal adverse events (grade ≥2) (RR 1.88, 95% CI 0.88 to 4.01; 1208 participants; Analysis 1.12).
2. Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone in newly‐diagnosed EOC
Two included studies (2776 participants) compared chemotherapy with bevacizumab followed by maintenance bevacizumab versus chemotherapy alone (GOG‐0218 2019; ICON7 2015). See summary of findings Table 1.
Overall survival (OS)
Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in OS compared to chemotherapy alone (HR 0.97, 95% CI 0.88 to 1.07; 2776 participants; moderate‐certainty evidence; Analysis 2.1). There was evidence of subgroup differences in the risk of disease (P = 0.007) (Analysis 2.2), which suggests that those at higher risk of disease progression may have more benefit (HR 0.86, 95% CI 0.76 to 0.98; 1316 participants) compared to those at lower risk (HR 1.13, 95% CI 0.97 to 1.31; 1460 participants). However, these data are based on retrospective subgroup analysis within the studies and should be interpreted with caution.
Progression‐free survival (PFS)
The evidence is very uncertain about the effect of chemotherapy with bevacizumab followed by maintenance bevacizumab on PFS (HR 0.82, 95%CI 0.64 to 1.05; 2746 participants; very low‐certainty evidence; Analysis 2.3).
Quality of life (QoL)
Studies differed in their reporting of quality of life measures. Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone reduces QoL at 54 weeks, measured using the European Organisation for Research and Treatment of Cancer (EORTC) QLQ‐C30 questionnaire (ICON7 2015) (MD ‐6.40, 95% CI ‐8.86 to ‐3.94; 1 study, 890 participants; high‐certainty evidence; Analysis 2.4), and likely results in little to no difference in QoL at six months of follow‐up, measured using Trial Outcome Index score of the Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire (GOG‐0218 2019) (MD 2.00, 95% CI ‐0.12 to 4.12; 1 study; 709 participants; Analysis 2.4).
Adverse events
The two included studies for this comparison did not report the effect of chemotherapy with bevacizumab followed by maintenance bevacizumab by our prespecified outcomes of hypertension (grade ≥ 2), proteinuria (grade ≥ 2), abdominal pain (grade ≥ 2), venous thromboembolic events, arterial thromboembolic events, non‐CNS bleeding, gastrointestinal adverse events or bowel fistula or perforation. For reasons described above, presented grades of hypertension, pain and abdominal pain are different from the prespecified outcomes and grade of pain data are limited to only grade 3 (not grade ≥ 3 as prespecified).
Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone likely results in little to no difference in any adverse events (grade ≥ 3) (RR 1.16, 95% CI 1.07 to 1.26; 1 study, 1485 participants; moderate‐certainty evidence; Analysis 2.5). Chemotherapy with bevacizumab followed by maintenance bevacizumab may result in a large increase in hypertension (grade ≥ 2) compared to chemotherapy alone (RR 4.27, 95% CI 3.25 to 5.60; 2 studies, 2707 participants; low‐certainty evidence; Analysis 2.6).
Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone likely increases proteinuria (grade ≥ 3) (RR 1.95, 95% CI 1.18 to 3.20; 2 studies, 2707 participants; Analysis 2.7); likely results in little to no difference in pain (grade ≥ 2) (RR 1.13, 95% CI 0.99 to 1.28; 1 study, 1209 participants; Analysis 2.8), neutropenia (grade ≥ 3: RR 1.09, 95% CI 0.86 to 1.38; 1 study, 2707 participants; grade ≥ 4: RR 1.10, 95% CI 1.00 to 1.20; 1 study, 1209 participants; Analysis 2.9), and febrile neutropenia (any grade) (RR 1.33, 95% CI 0.87 to 2.03; 2 studies, 2707 participants; Analysis 2.10). Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone likely increases the rate of venous thromboembolic events (any grade) (RR 1.39, 95% CI 1.02 to 1.89; 2 studies, 2707 participants; Analysis 2.11) and arterial thromboembolic events (any grade) (RR 1.93, 95% CI 1.05 to 3.57; 2 studies, 2707 participants; Analysis 2.12); likely results in little to no difference in non‐CNS bleeding (grade ≥ 3) (RR 2.10, 95% CI 0.85 to 5.21; 2 studies, 2707 participants; Analysis 2.13); likely results in a large increase in gastrointestinal adverse events (grade ≥ 2) (RR 2.17, 95% CI 1.04 to 4.55; 1 study, 1209 participants; Analysis 2.14); and may result in a large increase in gastrointestinal perforation (grade ≥ 3) (RR 3.71, 95% CI 1.04 to 13.23; 1 study, 1498 participants; Analysis 2.14).
3. Chemotherapy with tyrosine kinase inhibitors (TKIs) followed by TKI maintenance compared to chemotherapy alone in newly‐diagnosed EOC
Three studies (1639 participants) compared chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone in newly‐diagnosed EOC (AGO‐OVAR 12 2020; CHIVA 2019; Hainsworth 2015). See summary of findings Table 2. AGO‐OVAR 12 2020 (1366 participants) evaluated the addition of pazopanib, CHIVA 2019 (188 participants) evaluated the addition of nintedanib, whereas Hainsworth 2015 (85 participants) evaluated the addition of sorafenib.
Overall survival (OS)
Chemotherapy with TKI followed by maintenance with TKI likely results in little to no difference in OS compared to chemotherapy alone (HR 0.99, 95% CI 0.84 to 1.17; 2 studies, 1451 participants; moderate‐certainty evidence; Analysis 3.1).
Progression‐free survival (PFS)
Chemotherapy with TKI followed by maintenance with TKI likely results in little to no difference in PFS compared to chemotherapy alone (HR 0.88, 95% CI 0.77 to 1.00; 1451 participants; moderate‐certainty evidence; Analysis 3.2).
Quality of life (QoL)
Chemotherapy with TKI followed by maintenance with TKI (pazopanib) compared to chemotherapy alone reduces the mean global health status and QoL score over the treatment period, measured using the European Organisation for Research and Treatment of Cancer QLQ‐C30 questionnaire in one study (AGO‐OVAR 12 2020) (MD ‐1.86 95% CI ‐3.46 to ‐0.26; 1 study, 1340 participants; moderate‐certainty evidence; Analysis 3.3).
Adverse events
The two included studies for this comparison did not report the effect of chemotherapy with TKI followed by maintenance with TKI on our prespecified outcomes of hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), venous thromboembolic events, arterial thromboembolic events, non‐CNS bleeding, febrile neutropenia, gastrointestinal adverse events or bowel fistula or perforation. For reasons described above, presented grades of hypertension and abdominal pain are different from our prespecified outcomes.
Chemotherapy with TKI followed by maintenance with TKI compared to chemotherapy alone increases adverse events (grade ≥ 3) (RR 1.31, 95% CI 1.11 to 1.55; 1 study, 188 participants; moderate‐certainty evidence; Analysis 3.4) and may result in a large increase in hypertension (grade ≥ 3) (RR 6.49, 95% CI 2.02 to 20.87; 1 study, 1352 participants; low‐certainty evidence; Analysis 3.5). Chemotherapy with TKI followed by maintenance with TKI compared to chemotherapy alone likely results in little to no difference in abdominal pain (grade ≥ 3) (RR 1.54, 95% CI 0.81 to 2.92; 1 study, 1352 participants; Analysis 3.6). Chemotherapy with TKI followed by maintenance with TKI compared to chemotherapy alone likely results in little to no difference in neutropenia (grade ≥ 3) (RR 1.11, 95% CI 0.95 to 1.30; 1 study, 1352 participants; Analysis 3.7).
4. Chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone in newly‐diagnosed EOC
One included study (1015 participants) evaluated the effect of the TKI agent trebananib, a peptide‐Fc fusion protein (TRINOVA‐3 2019) (see summary of findings Table 3).
Overall survival (OS)
Chemotherapy with TKI (trebananib) followed by maintenance with TKI likely results in little to no difference in OS compared to chemotherapy alone (HR 0.99, 95% CI 0.79 to 1.25; moderate‐certainty evidence; Analysis 4.1).
Progression‐free survival (PFS)
Chemotherapy with TKI (trebananib) followed by maintenance with TKI likely results in little or no difference in PFS compared to chemotherapy alone (HR 0.93, 95% CI 0.79 to 1.09; moderate‐certainty evidence; Analysis 4.2).
Quality of life (QoL)
TRINOVA‐3 2019 did not report this outcome.
Adverse events
TRINOVA‐3 2019 did not report the effect of chemotherapy with TKI (trebananib) followed by maintenance with TKI on hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), venous thromboembolic events, arterial thromboembolic events, non‐CNS bleeding, gastrointestinal adverse events or bowel fistula or perforation. For reasons described above, presented grades of pain and abdominal pain are different from those we prespecified and grade of pain data are limited to only grade 3 (not grade ≥ 3 as prespecified).
Chemotherapy with TKI (trebananib) followed by maintenance with TKI compared to chemotherapy alone likely increases any adverse events (grade ≥ 3) (RR ranged from 1.10 (grade 3) to 9.96 (grade 5); moderate‐certainty evidence; Analysis 4.3).
Chemotherapy with TKI (trebananib) followed by maintenance with TKI compared to chemotherapy alone may result in little to no difference in pain (grade 3) (RR 1.00, 95% CI 0.09 to 10.94; Analysis 4.4) and likely results in little to no difference in abdominal pain (grade ≥ 3) (RR 1.22, 95% CI 0.61 to 2.43; Analysis 4.5).
Chemotherapy with TKI (trebananib) followed by maintenance with TKI compared to chemotherapy alone likely results in little to no difference in neutropenia (grade ≥ 3) (RR 0.95, 95% CI 0.77 to 1.18; Analysis 4.6 (grade 3)). The evidence is very uncertain about the effect of TKI (trebananib) followed by maintenance with TKI on febrile neutropenia (RR ranged from 0.50 (grade 1 to 2) to 3.49 (grade 4); Analysis 4.7).
5. Maintenance with TKI versus placebo after first‐line chemotherapy in newly‐diagnosed EOC
Two studies (1186 participants) compared maintenance with TKI versus placebo after first‐line chemotherapy in newly‐diagnosed EOC (AGO‐OVAR 16 2019; Herzog 2013). AGO‐OVAR 16 2019 (940 participants) evaluated pazopanib versus placebo, whereas Herzog 2013 (246 participants) evaluated sorafenib versus placebo.
Overall survival (OS)
After standard chemotherapy, maintenance with TKI compared to placebo likely results in little to no difference in OS (HR 0.98, 95% CI 0.83 to 1.16; Analysis 5.1).
Progression‐free survival (PFS)
After standard chemotherapy, maintenance with TKI compared to placebo likely results in little to no difference in PFS (HR 0.87, 95%CI 0.63 to 1.20; Analysis 5.2).
Quality of life (QoL)
After standard chemotherapy, maintenance with TKI compared to placebo likely results in little to no difference in QoL at the end of the maintenance phase, measured using the Functional Assessment of Cancer Therapy (FACT)/National Comprehensive Cancer Network (NCCN) Ovarian Symptom Index (FOSI) in one study (Herzog 2013)(MD 0.48 95% CI ‐0.70 to 1.66; 156 participants; Analysis 5.3).
Adverse events
The two included studies for this comparison did not report on any severe adverse events (grade ≥ 3), hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), neutropenia (grade ≥ 3), febrile neutropenia, venous thromboembolic events, arterial thromboembolic events, non‐CNS bleeding, gastrointestinal adverse events, or bowel fistula or perforation. For reasons described above, presented grades of hypertension, proteinuria, abdominal pain and neutropenia are different from those we prespecified.
After standard chemotherapy, maintenance with TKI compared to placebo likely results in a large increase in hypertension (grade ≥ 3) (RR 5.59, 95% CI 3.78 to 8.25; 2 studies, 1184 participants; Analysis 5.4); and likely results in little to no difference in proteinuria (grade 3 or 4) (RR 2.90, 95%CI 0.59 to 14.29; 1 study, 938 participants; Analysis 5.5) and abdominal pain (grade ≥ 3) (RR 1.46, 95% CI 0.52 to 4.08; 1 study, 1184 participants; Analysis 5.6). After standard chemotherapy, maintenance with TKI compared to placebo likely results in a large increase in neutropenia (grade 3 or 4) (RR 6.49, 95% CI 2.96 to 14.21; 1 study, 938 participants; Analysis 5.7).
6. Neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all in newly‐diagnosed EOC
Two studies (163 participants) assessed the effect of bevacizumab in participants who had neoadjuvant chemotherapy prior to interval debulking surgery and compared giving chemotherapy alone before surgery versus chemotherapy plus bevacizumab in combination, with all participants receiving chemotherapy and bevacizumab in combination following surgery (ANTHALYA 2017; GEICO‐1205 2019).
Overall survival (OS)
Neither of the two included studies examining this comparison reported this outcome (ANTHALYA 2017; GEICO‐1205 2019).
Progression‐free survival (PFS)
Neoadjuvant chemotherapy with bevacizumab likely results in little or no difference in PFS compared to chemotherapy alone (HR 1.13, 95% CI 0.66 to 1.93; 1 study, 68 participants; Analysis 6.1).
Quality of life (QoL)
Neither of the two included studies examining this comparison reported this outcome (ANTHALYA 2017; GEICO‐1205 2019).
Adverse events
The two included studies did not report the effect of neoadjuvant chemotherapy with bevacizumab on hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), febrile neutropenia, venous thromboembolic events, arterial thromboembolic events, non‐CNS bleeding (grade ≥ 3), gastrointestinal adverse events (grade ≥ 2), and bowel fistula or perforation. For reasons described above, presented grades of hypertension, abdominal pain and gastrointestinal adverse events are different from those we prespecified.
The evidence is very uncertain about the effect of neoadjuvant chemotherapy with bevacizumab on any adverse events (grade ≥ 3) (RR 0.83, 95% CI 0.58 to 1.19; 2 studies, 163 participants; Analysis 6.2), hypertension (grade ≥ 3) (RR 0.94, 95% CI 0.06 to 14.47; 1 study, 68 participants; Analysis 6.3) and neutropenia (grade ≥ 3) (RR 1.89, 95% CI 0.37 to 9.62; 1 study, 68 participants; Analysis 6.5). Neoadjuvant chemotherapy with bevacizumab may result in little to no difference in abdominal pain (grade ≥ 3) (RR 0.19, 95% CI 0.01 to 3.79; 1 study, 68 participants; Analysis 6.4) and gastrointestinal adverse events (grade unclear) (RR 0.52, 95% CI 0.18 to 1.52; 1 study, 95 participants; Analysis 6.6), compared to chemotherapy alone.
7. Chemotherapy with celecoxib versus chemotherapy alone in newly‐diagnosed EOC
One included study evaluated the addition of celecoxib to primary chemotherapy in women with newly‐diagnosed ovarian cancer (Reyners 2012).
Overall survival (OS)
There may be little to no difference in OS (HR 1.16, 95% CI 0.86 to 1.57; 1 study, 196 participants; Analysis 7.1).
Progression‐free survival (PFS)
There may be little to no difference in PFS (HR 1.07, 95% CI 0.85 to 1.34; 1 study, 196 participants; Analysis 7.2).
Quality of life (QoL)
Reyners 2012 did not report this outcome.
Adverse events
There may be little to no difference in febrile neutropenia (grade ≥ 3) (RR 0.94, 95% CI 0.45 to 1.96; 1 study, 196 participants; Analysis 7.3) or gastrointestinal adverse events (grade ≥ 3) (RR 1.15, 95% CI 0.46 to 2.85; 1 study, 196 participants; Analysis 7.4). There was evidence that participants in the group that received celecoxib may be more likely to have skin rash (11.2% versus 0%; P < 0.001) and changes in liver function tests (7.2% versus 1%; P = 0.034), but these were not outcomes included in our key outcomes. Overall, this was a small study, and our certainty in the survival results is lowered further by discontinuation of celecoxib for over six months in the study. This was due to wider safety concerns about another similar drug (rofecoxib) for which approval was withdrawn by the Food and Drug Administration (FDA) due to safety concerns regarding cardiovascular events. The study was re‐started seven months later after participants were informed about the potential for increased cardiovascular toxicity. In the Reyners 2012 study, 24% (23/97) of participants in the chemotherapy plus celecoxib group discontinued celecoxib during chemotherapy and 27% (17/63) of those who started maintenance treatment discontinued treatment, largely due to adverse reactions.
Recurrent epithelial ovarian cancer (EOC)
Thirty‐seven included studies evaluated the effect of angiogenesis inhibitors in recurrent EOC. Of these, there were nine studies in women with recurrent platinum‐sensitive disease (AVANOVA2 2019; Cong 2019; GOG‐0213 2017; ICON6 2021; Li 2019; Liu 2019b; Liu 2022; MITO‐16b 2021; OCEANS 2015), 19 studies in women with recurrent platinum‐resistant disease (AMBITION 2022; APPROVE 2022; AURELIA 2014; BAROCCO 2022; EORTC‐1508 2021; Gotlieb 2012; Li 2021; Liu 2019a; Liu 2021a; McGuire 2018; METRO‐BIBF 2020; MITO‐11 2015; NICCC 2020; Nishikawa 2020; OCTOVA 2021; Roque 2022; Sharma 2021; SWOG‐S0904 2014; TRIAS 2018), and nine studies who recruited participants regardless of platinum‐sensitivity status (Duska 2020; Gupta 2019; Karlan 2012; Ledermann 2011; Matulonis 2019; Richardson 2018; TAPAZ 2022; TRINOVA‐1 2016; TRINOVA‐2 2017).
Eleven studies evaluated the addition of bevacizumab to conventional chemotherapy (AURELIA 2014; Cong 2019; GOG‐0213 2017; Li 2019; Li 2021; Liu 2019a; Liu 2021a; MITO‐16b 2021; Nishikawa 2020; OCEANS 2015; Roque 2022). Thirteen studies evaluated the addition of TKIs to conventional therapy (APPROVE 2022; Duska 2020; ICON6 2021; Karlan 2012; METRO‐BIBF 2020; MITO‐11 2015; Richardson 2018; Sharma 2021; SWOG‐S0904 2014; TAPAZ 2022; TRIAS 2018; TRINOVA‐1 2016; TRINOVA‐2 2017). One evaluated the addition of olaratumab to conventional chemotherapy (McGuire 2018). Eleven studies evaluated a combination of TKIs with other agents or on their own (AMBITION 2022; BAROCCO 2022; Ledermann 2011; EORTC‐1508 2021; Gotlieb 2012; Gupta 2019; Liu 2019b; Liu 2022; NICCC 2020; Matulonis 2019; OCTOVA 2021). The final study compared the addition of bevacizumab to a PARP inhibitor (niraparib) (AVANOVA2 2019).
We graded the certainty of the evidence of the five most clinically relevant comparisons:
-
chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone (summary of findings Table 4);
-
chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone (summary of findings Table 5);
-
chemotherapy with bevacizumab compared to chemotherapy alone (summary of findings Table 6);
-
chemotherapy with TKI compared to chemotherapy alone (summary of findings Table 7); and
-
chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone (summary of findings Table 8).
A. Platinum‐sensitive EOC
8. Chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC
Three included studies (1564 participants) compared chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC (GOG‐0213 2017; MITO‐16b 2021; OCEANS 2015). See summary of findings Table 4.
Overall survival (OS)
Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in OS compared to chemotherapy alone (HR 0.90, 95% CI 0.79 to 1.02; moderate‐certainty evidence; Analysis 8.1).
Progression‐free survival (PFS)
Chemotherapy with bevacizumab followed by maintenance bevacizumab likely increases PFS compared to chemotherapy alone (HR 0.56, 95% CI 0.50 to 0.63; moderate‐certainty evidence; Analysis 8.2).
Quality of life (QoL)
Chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone likely results in little to no difference in QoL at 12 months after the first cycle, measured using the Trial Outcome Index score of the Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire in one study (GOG‐0213 2017) (MD 0.80, 95% CI ‐2.11 to 3.71; 1 study, 486 participants; low‐certainty evidence; Analysis 8.3).
Adverse events
The included studies did not report the effect of chemotherapy with bevacizumab on hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), venous thromboembolic events (any grade), non‐CNS bleeding (grade ≥ 3), gastrointestinal adverse events (grade ≥ 2), or bowel fistula or perforation (grade ≥ 3). For reasons described above, presented grades of hypertension, proteinuria, pain, abdominal pain, venous thromboembolic events, non‐CNS bleeding, and gastrointestinal perforations are different from those we prespecified.
Chemotherapy with bevacizumab followed by maintenance bevacizumab slightly increases any adverse event (grade ≥3) compared to chemotherapy alone (RR 1.11, 95% CI 1.07 to 1.16; 3 studies, 1538 participants; high‐certainty evidence; Analysis 8.4).
The evidence is very uncertain about the effect of chemotherapy with bevacizumab followed by maintenance bevacizumab compared to chemotherapy alone on hypertension (grade ≥ 3) (RR 5.82, 95% CI 3.84 to 8.83; 3 studies, 1538 participants; Analysis 8.5), abdominal pain (grade ≥ 3) (RR 16.88, 95% CI 4.72 to 60.34; 2 studies, 1058 participants; Analysis 8.8), neutropenia (grade ≥ 3) (RR 1.04, 95% CI 0.83 to 1.31; 2 studies, 1058 participants; Analysis 8.9), febrile neutropenia (any grade) (RR 1.20, 95% CI 0.70 to 2.06; 3 studies, 1538 participants; Analysis 8.10), and gastrointestinal perforations (RR 4.96, 95% 0.86 to 28.51; 2 studies, 1058 participants; Analysis 8.14).
Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in venous thromboembolic events (grade ≥ 3) compared to chemotherapy alone (RR 1.73, 95% 0.65 to 4.60; 1 study, 480 participants; Analysis 8.11).
Chemotherapy with bevacizumab followed by maintenance bevacizumab may result in a large increase in proteinuria (grade ≥ 3) (RR 20.27, 95% CI 6.42 to 64.00; 3 studies, 1538 participants; Analysis 8.6), arterial thromboembolic events (any grade) (RR 3.63, 95% 1.49 to 8.84; 1 study, 657 participants; Analysis 8.12), and non‐CNS bleeding (RR 3.77, 95% 2.70 to 5.26; 1 study, 657 participants; Analysis 8.13).
Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in a large increase in pain (grade ≥ 3) (RR 3.09, 95% CI 1.81 to 5.28; 2 studies, 1058 participants; Analysis 8.7).
9. Chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC
One study with 282 participants compared chemotherapy with TKI (cediranib) followed by TKI maintenance compared to chemotherapy alone in recurrent platinum‐sensitive EOC (ICON6 2021). See summary of findings Table 5.
Overall survival (OS)
Chemotherapy with TKI followed by maintenance TKI likely results in little to no difference in OS compared to chemotherapy alone (HR 0.86, 95% CI 0.67 to 1.11; low‐certainty evidence; Analysis 9.1).
Progression‐free survival (PFS)
Chemotherapy with TKI followed by maintenance TKI likely increases PFS compared to chemotherapy alone (HR 0.56, 95% CI 0.44 to 0.72; moderate‐certainty evidence; Analysis 9.2).
Quality of life (QoL)
Chemotherapy with TKI followed by maintenance TKI compared to chemotherapy alone may result in little to no difference in QoL measured at 12 months using the Global Quality of Life European Organisation for Research and Treatment of Cancer Questionnaire QLQ‐C30 (MD 6.10, 95% CI ‐0.96 to 13.16; 146 participants; low‐certainty evidence; Analysis 9.3).
Adverse events
The one included study for this comparison did not report the effect of chemotherapy with TKI followed by maintenance TKI on any adverse events (grade ≥ 3), hypertension (grade ≥ 2), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), febrile neutropenia (any grade), venous thromboembolic events (any grade), arterial thromboembolic events (any grade), non‐CNS bleeding (grade ≥ 3), gastrointestinal adverse events (grade ≥ 2), and bowel fistula or perforation (grade ≥ 3). For reasons described above, presented grades of hypertension, proteinuria, neutropenia, and febrile neutropenia are different from those we prespecified.
Chemotherapy with TKI followed by maintenance TKI likely results in a large increase in hypertension (grade ≥ 3) (RR 3.32, 95% CI 1.21 to 9.10; 444 participants; Analysis 9.4) and febrile neutropenia (grade ≥ 3) (RR 1.92, 95% CI 0.68 to 5.46; 444 participants; Analysis 9.7) compared to chemotherapy alone.
Chemotherapy with TKI followed by maintenance TKI may increase proteinuria (grade ≥ 3) (RR 1.76, 95% CI 0.09 to 36.34; 444 participants; Analysis 9.5) and increases neutropenia (grade ≥ 3) (RR 1.10, 95% CI 0.75 to 1.60; 444 participants; Analysis 9.6) compared to chemotherapy alone.
B. Platinum‐resistant EOC
10. Chemotherapy with bevacizumab compared to chemotherapy alone in recurrent platinum‐resistant EOC
Five studies (778 participants) compared chemotherapy with bevacizumab compared to chemotherapy alone in recurrent platinum‐resistant EOC (APPROVE 2022; AURELIA 2014; Liu 2019a; Nishikawa 2020; Roque 2022). See summary of findings Table 6.
An additional included study (Liu 2021a) also compared chemotherapy with bevacizumab compared to chemotherapy alone in recurrent platinum‐resistant EOC but data presented were insufficient for inclusion in the meta‐analyses
Overall survival (OS)
Chemotherapy with bevacizumab increases OS compared to chemotherapy alone (HR 0.73, 95% CI 0.61 to 0.86; high‐certainty evidence; Analysis 10.1).
There were insufficient data to extract HR data for OS from Liu 2021a. Median OS was 17.2 months in the bevacizumab group and 14.1 months in the control group (P = 0.015).
Progression‐free survival (PFS)
Chemotherapy with bevacizumab likely results in a large increase in PFS compared to chemotherapy alone (HR 0.49, 95% CI 0.42 to 0.58; moderate‐certainty evidence; Analysis 10.2).
There were insufficient data to extract HR for PFS from Liu 2021a. Median PFS was 10.9 months in the bevacizumab group and 7.8 months in the control group (P = 0.007).
Quality of life (QoL)
None of the included studies for this comparison reported this outcome.
Adverse events
Most of the included studies did not report the effect of chemotherapy with bevacizumab on any adverse events (grade ≥ 3), proteinuria (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2), febrile neutropenia (any grade), venous thromboembolic events (any grade), arterial thromboembolic events (any grade), non‐CNS bleeding (grade ≥ 3), gastrointestinal adverse events (grade ≥ 2), and bowel fistula or perforation (grade ≥ 3). For reasons described above, presented grades of proteinuria, febrile neutropenia, venous thromboembolic events, arterial thromboembolic events and gastrointestinal perforations are different from those we prespecified.
Chemotherapy with bevacizumab may increase any adverse events (grade ≥ 3) slightly (RR 1.68, 95% CI 0.76 to 3.69; 1 study, 101 participants; low‐certainty evidence).
Chemotherapy with bevacizumab compared to chemotherapy alone may result in a large increase in hypertension (grade ≥ 2: RR 3.11, 95% CI 1.83 to 5.27; 2 studies, 436 participants; low‐certainty evidence; Analysis 10.4; grade ≥ 3: RR 3.24, 95% CI 1.46 to 7.19; 4 studies, 623 participants; Analysis 10.4).
The evidence is very uncertain about the effect of chemotherapy with bevacizumab compared to chemotherapy alone on proteinuria (grade ≥ 3) (RR 6.26, 95% CI 1.13 to 34.70; 4 studies, 683 participants; Analysis 10.5), febrile neutropenia (grade ≥ 3) (RR 0.33, 95% CI 0.04 to 3.04; 1 study, 101 participants; Analysis 10.7); arterial thromboembolic evens (grade ≥ 3) (RR 9.10, 95% CI 0.49 to 167.79; 1 study, 360 participants; Analysis 10.9). Bevacizumab may increase the risk of gastrointestinal perforations (grade ≥ 2) slightly (RR 6.89, 95% CI 0.86 to 55.09; 2 studies, 436 participants; Analysis 10.10).
Chemotherapy with bevacizumab compared to chemotherapy alone may increase the risk of neutropenia (grade ≥ 3) (RR 1.35, 95% CI 1.01 to 1.81; 3 studies, 308 participants; Analysis 10.6). The effect on rates of venous thromboembolic events (grade ≥ 3) is very uncertain (RR 0.58, 95% CI 0.21 to 1.63; 2 studies, 436 participants; Analysis 10.8).
11. Chemotherapy with TKI compared to chemotherapy alone in recurrent platinum‐resistant EOC
Nine studies evaluated the effect of TKI agents with similar mechanisms of action, as follows: one study each in apatinib (APPROVE 2022), nintedanib (METRO‐BIBF 2020), sorafenib (TRIAS 2018), vandetanib (SWOG‐S0904 2014); and five studies in pazopanib (Duska 2020; MITO‐11 2015; Richardson 2018; Sharma 2021; TAPAZ 2022). See summary of findings Table 7.
Overall survival (OS)
Chemotherapy with TKI likely results in little to no difference in OS compared to chemotherapy alone (HR 0.85, 95% CI 0.68 to 1.08; 8 studies, 940 participants, moderate‐certainty evidence; Analysis 11.1). This is based on evidence from trials in apatinib, nintedanib, pazopanib, sorafenib and vandetanib, with no strong evidence of subgroup differences depending on the type of TKI agent (P = 0.05). Sensitivity analysis limited to studies with only platinum‐resistant EOC showed a significant difference in OS between chemotherapy with TKI and chemotherapy alone (HR 0.77, 95% CI 0.63 to 0.93; 5 studies, 589 participants; fixed‐effect model).
Progression‐free survival (PFS)
Chemotherapy with TKI may increase PFS compared to chemotherapy alone (HR 0.70, 95% CI 0.55 to 0.89; 8 studies, 940 participants; low‐certainty evidence; Analysis 11.2), based on evidence from trials in apatinib, nintedanib, pazopanib, sorafenib and vandetanib, with evidence of subgroup difference depending on the type of TKI agent (P = 0.009). Findings from a sensitivity analysis limited to studies with only platinum‐resistant EOC were consistent with the main analysis (HR 0.67, 95% CI 0.57 to 0.79; 6 studies, 695 participants; fixed‐effect model).
Quality of life (QoL)
Chemotherapy with TKI may result in little to no difference in QoL compared to chemotherapy alone, based on evidence from three studies (METRO‐BIBF 2020; Sharma 2021; TAPAZ 2022). All studies measured global QoL using the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire QLQ‐C30 at three different time points: at six weeks (METRO‐BIBF 2020), after six cycles (Sharma 2021), and after four months (TAPAZ 2022). The mean difference (MD) ranged from ‐3.40 (95% CI ‐13.22 to 6.42) to 17.50 (95% CI 1.11 to 33.89; low‐certainty evidence; Analysis 11.3). The evidence was limited to trials in nintedanib and pazopanib.
Adverse events
Included trials did not report the effect of chemotherapy with bevacizumab on hypertension (grade ≥ 2), venous thromboembolic events (any grade), arterial thromboembolic events (any grade), and gastrointestinal adverse events (grade ≥ 2). Presented grades of hypertension and gastrointestinal adverse events are different from the prespecified.
Chemotherapy with TKI may increase any adverse events (grade ≥ 3) slightly compared to chemotherapy alone (RR 1.23, 95% 1.02 to 1.49; 4 studies, 548 participants; low‐certainty evidence; Analysis 11.4), based on evidence from trials in apatinib, nintedanib, pazopanib and sorafenib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.06).
Chemotherapy with TKI may result in a large increase in hypertension (grade ≥ 3) compared to chemotherapy alone (RR 4.20, 95% CI 1.58 to 11.14; 7 studies, 844 participants; Analysis 11.5), based on evidence from trials in apatinib, nintedanib, pazopanib, sorafenib and vandetanib, with evidence of subgroup difference depending on the type of TKI agent (P = 0.02).
The evidence is very uncertain about the effect of chemotherapy with TKI on proteinuria (grade ≥ 2) compared to chemotherapy alone (RR 4.00, 95% CI 0.49 to 32.86; 2 studies, 262 participants; Analysis 11.6), based on evidence from trials in apatinib and vandetanib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.86).
Chemotherapy with TKI may result in little to no difference in pain (grade ≥ 2) compared to chemotherapy alone (RR 0.97, 95% CI 0.44 to 2.15; 3 studies, 361 participants; Analysis 11.7), based on evidence from trials in pazopanib and sorafenib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.72).
Chemotherapy with TKI may result in little to no difference in abdominal pain (grade ≥ 2) compared to chemotherapy alone (RR 0.78, 95% CI 0.20 to 3.09; 1 study, 116 participants; Analysis 11.8), based on evidence from a trial in pazopanib.
Chemotherapy with TKI may increase neutropenia (grade ≥ 3) compared to chemotherapy alone (RR 1.73, 95% CI 1.15 to 2.61; 9 studies, 1069 participants; Analysis 11.9), based on evidence from trials in apatinib, nintedanib, pazopanib, sorafenib and vandetanib, with evidence of subgroup difference depending on the type of TKI agent (P = 0.008).
Chemotherapy with TKI may result in little to no difference in febrile neutropenia (any grade) compared to chemotherapy alone (RR 1.49, 95% CI 0.68 to 3.30; 6 studies, 748 participants; Analysis 11.10), based on evidence from trials in nintedanib, pazopanib, sorafenib and vandetanib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.94).
Chemotherapy with TKI may result in little to no difference in non‐CNS bleeding (grade ≥ 3) compared to chemotherapy alone (RR 1.07, 95% CI 0.07 to 17.44; 1 study, 172 participants; Analysis 11.11), based on evidence from one trial in sorafenib.
Chemotherapy with TKI likely results in little to no difference in gastrointestinal adverse events (grade ≥ 3) compared to chemotherapy alone (RR 1.08, 95% CI 0.46 to 2.53; 3 studies, 386 participants; Analysis 11.12), based on evidence from trials in nintedanib, pazopanib and sorafenib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.09).
The evidence is very uncertain about the effect of chemotherapy with TKI on bowel fistula or perforation (grade ≥ 3) compared to chemotherapy alone (RR 2.74, 95% CI 0.77 to 9.75; 5 studies, 557 participants; very low‐certainty evidence; Analysis 11.13), based on evidence from trials in nintedanib and pazopanib, with no evidence of subgroup difference depending on the type of TKI agent (P = 0.99).
12. Chemotherapy with olaratumab compared to chemotherapy alone in recurrent platinum‐resistant EOC
One trial (with 123 participants) evaluated the effect of the platelet‐derived growth factor alpha (PDGFRα)‐targeting monoclonal antibody olaratumab in addition to chemotherapy (McGuire 2018).
Overall survival (OS)
Chemotherapy with olaratumab may result in little to no difference in OS (HR 1.10, 95% CI 0.71 to 1.71; Analysis 12.1).
Progression‐free survival (PFS)
Chemotherapy with olaratumab may result in little to no difference in PFS (HR 1.04, 95% CI 0.70 to 1.56; Analysis 12.2).
Quality of life (QoL)
McGuire 2018 did not report on QoL.
Adverse events
The evidence from the McGuire 2018 study is very uncertain about the effect of chemotherapy with olaratumab on incidence of pain (grade ≥ 3) (RR 0.33, 95% CI 0.01 to 7.90; Analysis 12.4), abdominal pain (grade ≥ 3) (RR 0.25, 95% CI 0.05 to 1.11; Analysis 12.5), or neutropenia (grade ≥ 3) (RR 1.57, 95% CI 0.55 to 4.54; Analysis 12.6).
McGuire 2018 did not formally report on incidence of hypertension, febrile neutropenia, venous thromboembolism, arterial thromboembolism, non‐central nervous system bleeding (grade ≥ 3), or gastrointestinal perforation (grade ≥ 3) by trial arm. There were no cases of proteinuria (grade ≥ 3) (Analysis 12.3).
Other comparisons in platinum‐resistant recurrent EOC
The following included studies were not included in meta‐analyses.
In the recurrent platinum‐resistant setting, two studies had an experimental design and compared a combination of anti‐programmed death‐ligand 1 (PD‐L1) antibody and aspirin with and without bevacizumab (EORTC‐1508 2021), and TKI with placebo (Gotlieb 2012). Another two studies compared a combination of PARP inhibitor with TKI to chemotherapy or to PARP inhibitor alone (OCTOVA 2021), and to PARP inhibitor with durvalumab (immunotherapy agent) (AMBITION 2022). None of the studies found evidence of a beneficial effect of the evaluated interventions on survival outcomes.
NICCC 2020 enrolled participants with clear cell cancer of either EOC or endometrial origin. It was a phase II study powered to detect an improvement in PFS from three to five months (HR 0.6), with greater than 90% power, with single agent nintedanib, to determine whether a phase III study was warranted. There was no significant difference in either OS (HR 0.77, 95% CI 0.46 to 1.28) or PFS (HR 0.79, 95% CI 0.5 to 1.125), although there was evidence of non‐proportionality of hazards for OS. The study authors concluded that there was insufficient evidence of activity with nintedanib alone, but that combination treatment was worth further examination.
C. Mixed platinum‐sensitive, platinum‐resistant and unclear recurrent EOC
13. Chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone in recurrent mixed platinum‐resistance EOC
Three trials evaluated the effect of the TKI agent trebananib, a peptide‐Fc fusion protein (Karlan 2012; TRINOVA‐1 2016; TRINOVA‐2 2017). In the synthesis, we used the 10 mg dose of trebananib data from the Karlan 2012 study. See summary of findings Table 8.
Overall survival (OS)
Chemotherapy with TKI (trebananib) likely results in little to no difference in OS (HR 0.92, 95% CI 0.80 to 1.06; 3 studies, 1250 participants; moderate‐certainty evidence; Analysis 13.1).
Progression‐free survival (PFS)
Chemotherapy with TKI (trebananib) increases PFS (HR 0.73, 95% CI 0.65 to 0.82; 3 studies, 1250 participants; high‐certainty evidence; Analysis 13.2).
Quality of life (QoL)
Chemotherapy with TKI (trebananib) may result in little to no difference in QoL measured at 25 weeks using the Functional Assessment of Cancer Therapy Ovarian Cancer questionnaire (MD ‐0.80, 95% CI ‐4.31 to 2.71; 1 study, 315 participants; low‐certainty evidence; Analysis 13.3) compared to chemotherapy alone.
Adverse events
Included trials did not report the effect of chemotherapy with bevacizumab on any adverse events (grade ≥ 3), hypertension (grade ≥ 2), pain (grade ≥ 2), abdominal pain (grade ≥ 2) and gastrointestinal adverse events (grade ≥ 2). Presented grades of hypertension, pain and abdominal pain are different from those prespecified.
Compared to chemotherapy alone, chemotherapy with TKI (trebananib) may result in little to no difference in hypertension (grade ≥ 3) (RR 2.92, 95% CI 0.70 to 12.18; 3 studies, 1242 participants; Analysis 13.4), febrile neutropenia (any grade) (RR 0.49, 95% CI 0.04 to 5.39; 1 study, 913 participants; Analysis 13.9) and gastrointestinal perforation (grade ≥ 3) (RR 0.35, 95% CI 0.01 to 8.30; 1 study, 108 participants; low‐certainty evidence; Analysis 13.13).
Compared to chemotherapy alone, chemotherapy with TKI (trebananib) likely results in little to no difference in abdominal pain (grade ≥ 3) (RR 0.99, 95% CI 0.60 to 1.65; 3 studies, 1242 participants; Analysis 13.7) and venous thromboembolism event (any grade) (RR 0.68, 95% CI 0.25 to 1.85; 2 studies, 1021 participants; Analysis 13.10).
Chemotherapy with TKI (trebananib) increases neutropenia (grade ≥ 3) compared to chemotherapy alone (RR 0.60, 95% CI 0.40 to 0.89; 2 studies, 1134 participants; Analysis 13.8).
The evidence is very uncertain about the effect of chemotherapy with TKI (trebananib) compared to chemotherapy alone on proteinuria (grade ≥ 3) (RR 6.86, 95% CI 0.36 to 132.5; 1 study, 913 participants; Analysis 13.5), pain (grade ≥ 3) (RR 2.94, 95% CI 0.12 to 72.02; 1 study, 913 participants; Analysis 13.6), arterial thromboembolic event (any grade) (RR 3.11, 95% CI 0.13 to 74.72; 1 study, 108 participants; Analysis 13.11) and non‐CNS bleeding (grade ≥ 3) (RR 3.11, 95% CI 0.13 to 74.72; 1 study, 108 participants; Analysis 13.12).
Other comparisons in mixed platinum‐sensitive, platinum‐resistant and unclear recurrent EOC
The following included studies were not included in meta‐analyses because of the diversity of the comparisons: Gupta 2019; Ledermann 2011; Matulonis 2019.
We included three studies in recurrent EOC that recruited participants with mixed platinum‐sensitivity status: Matulonis 2019 compared a TKI with chemotherapy; Ledermann 2011 compared a TKI with placebo; and Gupta 2019 compared a combination of chemotherapy with celecoxib with chemotherapy alone. Only one study reported an effect on survival outcomes (overall survival) that was statistically significant (HR 2.27, 95% CI 1.17 to 4.41), favouring the comparator (chemotherapy) (Matulonis 2019).
Discussion
Summary of main results
The five studies which were included in the previous version of this review (Gaitskell 2011), using data from conference abstracts, have now all been published in more detail as full papers (GOG‐0218 2019; Karlan 2012; Ledermann 2011; ICON7 2015; Gotlieb 2012). Additionally, we identified 45 new studies published within the last ten years.
Our systematic review identified 50 randomised trials with 14,836 individuals diagnosed with epithelial ovarian cancer (EOC): 13 trials (7708 participants) in newly‐diagnosed EOC and 37 trials (7128 participants) in recurrent disease. The studies examined the effects of various angiogenesis inhibitors (e.g. bevacizumab, sorafenib, trebananib) in a range of clinical scenarios. Where possible, we grouped and synthesised the evidence by the type of population (newly‐diagnosed EOC, recurrent platinum‐sensitive EOC, recurrent platinum‐resistant EOC) and according to the angiogenesis inhibitor mechanism of action (bevacizumab, tyrosine kinase inhibitors (TKIs) and the TKI agent trebananib, a peptide‐Fc fusion protein). We performed quantitative synthesis for thirteen comparisons (as specified below). We graded the certainty of the evidence for the eight most clinically relevant comparisons (indicated with a * in the list below) and the six most critically important outcomes (overall survival (OS), progression‐free survival (PFS), quality of life (QoL), any adverse events grade ≥ 3, hypertension grade ≥ 2, and bowel fistula/perforation grade ≥ 3).
Newly‐diagnosed EOC
-
Chemotherapy with bevacizumab versus chemotherapy alone with placebo for all in the maintenance phase (one study, 1250 participants)
-
Chemotherapy with bevacizumab followed by bevacizumab as maintenance versus chemotherapy alone (two studies, 2776 participants)*
-
Chemotherapy with TKI followed by TKI as maintenance versus chemotherapy alone (three studies, 2639 participants)*
-
Chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI as maintenance versus chemotherapy alone (one study, 1015 participants)*
-
Maintenance with TKI versus placebo after first‐line chemotherapy (two studies, 1186 participants)
-
Neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all (two studies, 167 participants)
-
Chemotherapy with celecoxib versus chemotherapy alone (one study; 196 participants)
Recurrent platinum‐sensitive EOC
-
Chemotherapy with bevacizumab followed by maintenance bevacizumab versus chemotherapy alone (seven studies, 1893 participants)*
-
Chemotherapy with TKI followed by TKI maintenance versus chemotherapy alone (one study, 486 participants)*
Recurrent platinum‐resistant EOC
-
Chemotherapy with bevacizumab versus chemotherapy alone (seven studies, 894 participants)*
-
Chemotherapy with TKI versus chemotherapy alone (seven studies, 810 participants)*
-
Chemotherapy with olaratumab versus chemotherapy alone (one study, 123 participants)
Recurrent EOC
-
Chemotherapy with TKI (peptide‐Fc fusion protein) versus chemotherapy alone (three studies, 1250 participants)*
Eighteen included trials (1850 participants) evaluated the effect of angiogenesis inhibitors in combination with or in comparison to other treatments, such as poly(ADP‐ribose) polymerase inhibitors (PARPi), anti PD‐L1 antibody, aspirin and celecoxib, in cohorts of participants with mixed populations (newly‐diagnosed and/or platinum‐sensitive and platinum‐resistant recurrent disease). We only briefly summarised the findings of these studies due to their experimental and proof‐of‐concept nature. Future updates will include an analysis of PARPi‐anti‐angiogenesis combination treatment.
Newly‐diagnosed EOC
Based on available evidence, bevacizumab given with chemotherapy and then continued as a maintenance treatment results in little to no difference in OS and a slight reduction in QoL compared to chemotherapy alone. The combination likely results in little to no difference in any adverse events (grade ≥ 3). The evidence on the effect on PFS and hypertension (grade ≥ 2) is very uncertain.
Equally, TKIs (nintedanib, pazopanib, sorafenib) given with chemotherapy and continued as maintenance treatment likely result in little to no difference in OS compared to chemotherapy alone. However, the combination likely slightly increases PFS and slightly reduces global QoL compared to chemotherapy alone. The combination increases any adverse events (grade ≥ 3) and may result in a large increase in hypertension (grade ≥ 3).
Chemotherapy with a peptide‐Fc fusion protein (trebananib) given with chemotherapy and continued as maintenance treatment likely results in little to no difference in OS and PFS compared to chemotherapy alone. The combination increases any adverse events (grade ≥ 3).
Celecoxib plus chemotherapy versus chemotherapy alone may result in little to no difference in OS or PFS in newly‐diagnosed ovarian cancer. There may be little to no effect on most adverse events of grade 3 or higher, but 40 of the 97 participants discontinued treatment, mainly due to adverse events.
Recurrent platinum‐sensitive EOC
Bevacizumab given with chemotherapy and then continued as a maintenance treatment may have little to no effect on OS; however, the combination may improve PFS. The combination may result in little to no difference in QoL with a slight increase in the rate of any adverse events (grade ≥3). Included trials reported only rate of events of hypertension of grade 3 or above, which were higher in arms with bevacizumab.
The combination of TKIs (cediranib) with chemotherapy likely results in little to no difference in OS; however, it likely increases PFS and may have little to no effect on QoL. Included trials reported only rate of events of hypertension of grade 3 or above, which were higher in arms with TKIs.
Recurrent platinum‐resistant EOC
Bevacizumab given with chemotherapy compared to chemotherapy alone, and then continued as a maintenance treatment, increases OS and likely increases PFS. However, the combination may result in a large increase in hypertension (grade ≥ 2).
The combination of TKIs (apatinib, nintedanib, pazopanib, sorafenib, vandetanib) with chemotherapy compared to chemotherapy alone likely results in little to no difference in OS and may result in little to no meaningful difference in QoL; however, it may increase PFS. The combination slightly increases the rate of any adverse events (grade ≥ 3). The effect on bowel fistula/perforation rates is uncertain as is the effect on hypertension (grade ≥ 2), largely due to the small study size and heterogeneity in the effect between different TKIs.
Chemotherapy with a peptide‐Fc fusion protein (trebananib) likely results in little to no difference in OS, although it increased PFS. The combination may result in little to no difference in QoL. The only safety data available were for bowel perforation/fistula (grade ≥ 3), suggesting that trebananib may result in little to no effect on this outcome.
Olaratumab plus liposomal doxorubicin versus liposomal doxorubicin alone did not improve PFS or OS in platinum‐resistant or platinum‐refractory recurrent ovarian cancer. There was little to no effect on any adverse events of grade 3 or higher.
In summary, bevacizumab in combination with chemotherapy seems most beneficial for individuals with the most advanced diseases. The evidence supporting the use of TKI with chemotherapy for the treatment of EOC was not available, except for its use in combination with chemotherapy for recurrent platinum‐resistant EOC.
Overall completeness and applicability of evidence
In terms of applicability, the women in the included studies tended to be younger and fitter than the general cohort of women with ovarian cancer. The increased risk of severe adverse events and effects on the quality of life of long‐term maintenance treatment may therefore be different in the wider population of women with ovarian cancer. Furthermore, the included studies were generally powered for PFS rather than OS or other patient‐reported outcomes, including adverse events, quality of life and time at home. This limits the certainty and full breadth of information available to women, and their caregivers, needed to make fully informed decisions about treatment.
We are relatively confident that we have captured the majority of studies assessing anti‐angiogenesis inhibitors in ovarian cancer, having identified ongoing studies in a previous review update and compared included studies to recent systematic reviews. In addition, this review has been performed alongside other reviews of biological agents and pegylated liposomal doxorubicin (PLD) in ovarian cancer (Morrison 2018; Newhouse 2023; Tattersall 2022). Studies have been shared between review teams, where there was found to be overlap, so the search net has effectively been wider than the search strategy of each individual review. However, this is a fast‐moving field with multiple new studies, drugs and drug combinations included in this update, as well as several ongoing studies; we acknowledge that relevant studies may have been missed.
Certainty of the evidence
This is a comprehensive review of literature on the effect of angiogenesis inhibitors in the treatment of epithelial ovarian cancer. Nevertheless, the quality of the evidence is not always satisfactory. The main difficulty in the assessment of the study quality was suboptimal or inadequate reporting of important features of trial design, such as randomisation procedure (in 23 studies, we assessed this domain as unclear) or allocation concealment (in 25 studies, we assessed this domain as unclear). Another indicator of the quality of the evidence, a clinical trial registration number, was unavailable for several studies.
Over half of the included studies (28 out of 50) had an open‐label design, which put the studies at high risk of bias for blinding for all outcomes except OS (Figure 4). Overall, we deemed only two studies to be at low risk of bias in all evaluated domains (ICON6 2021; Richardson 2018), and five studies to be at low risk in six domains (AGO‐OVAR 12 2020; GOG‐0218 2019; TRIAS 2018; TRINOVA‐1 2016; TRINOVA‐3 2019).
For an informative analysis of the primary outcome of this review (overall survival (OS)), the most commonly used statistical model relies on the assumption of proportional hazards (Schemper 2009), in order to yield an informative effect estimate (hazard ratio). Unfortunately, in a number of key trials in this area (e.g. Duska 2020; ICON6 2021; ICON7 2015; TRINOVA‐2 2017), there was evidence of non‐proportionality. However, in the final summary of the evidence, we did not downgrade certainty due to this finding. A formal assessment of non‐proportionality was infrequently reported, and we perceived it more as a mark of the study’s quality rather than its weakness.
Potential biases in the review process
We aimed to reduce any potential biases in the review process by adhering to Cochrane methodology as much as possible. We performed a comprehensive search, including a thorough search of the grey literature. At least two review authors sifted and independently extracted data for all studies. However, we recognise that some studies ‐ published in non‐indexed journals or in less accessible languages (e.g. Chinese) ‐ could have been missed. However, studies in Chinese were identified by the search (as they had English language abstracts) and we were able to obtain and translate full text articles with assistance where necessary (see Acknowledgements) (e.g. Liu 2021a).
None of the review authors have any links to drug companies, any financial interest in the prescription of chemotherapeutic agents, nor were we involved in the conduct of the included studies.
Agreements and disagreements with other studies or reviews
We conducted a systematic search for other recent systematic reviews on randomised controlled trials of angiogenesis inhibitors in ovarian cancer, published since January 2020 (see Appendix 4 for search strategy). This identified 15 references, with an additional reference identified in a further search to October 2022. We limited our search to systematic reviews published in the last two years, as many of the primary studies included in this review were only published in the last few years, and thus older systematic reviews would not be comparable.
We excluded seven of these references on the basis of the abstract, and two on the basis of a full‐text review. We considered six systematic reviews to be at least partially relevant, and have summarised these below.
Helali 2022 is a comprehensive systematic review and network meta‐analysis of anti‐angiogenic agents in advanced epithelial ovarian cancer. The authors identified 23 relevant randomised controlled trials (RCTs) and looked separately at newly diagnosed, recurrent platinum‐sensitive, and recurrent platinum‐resistant ovarian cancer, focusing on the outcomes of overall survival (OS; primary outcome) and progression‐free survival (PFS; secondary outcome). The authors concluded that the best interventions for improving overall survival were likely to be: chemotherapy with concurrent bevacizumab followed by maintenance bevacizumab for high‐risk (defined as FIGO stage IV or inoperable/suboptimally‐resected stage III) newly‐diagnosed (chemotherapy‐naive) advanced ovarian cancer; and pazopanib combined with chemotherapy for platinum‐resistant recurrent ovarian cancer. They concluded that the evidence was less convincing for a benefit in OS with angiogenesis inhibitors in a setting of non‐high‐risk, newly‐diagnosed, or platinum‐sensitive recurrent, ovarian cancer. One author reported being employed in industry; the other authors reported no conflicts of interest.
Hirte 2021 is a systematic review of consolidation or maintenance systemic therapy for newly‐diagnosed stage II‐IV ovarian cancer. This included a review of studies of VEGF‐R TKIs (pazopanib and sorafenib, four studies), anti‐VEGF monoclonal antibodies (bevacizumab, two studies), triple angiokinase inhibitors (nintedanib, two studies), and angiopoietin inhibitors (trebananib, one study), in addition to other agents. The authors described the results of the trials but did not conduct a meta‐analysis. Their overall conclusion was that, compared with placebo, maintenance therapy with bevacizumab improves PFS for certain patients with newly‐diagnosed stage III‐IV EOC, but that there is thus far no evidence of a benefit in OS. The authors declared no conflicts of interest.
Chilimoniuk 2022 is a systematic review covering RCTs for a variety of new therapies in ovarian cancer, including 15 studies of angiogenesis inhibitors. The authors concluded that bevacizumab was beneficial for the treatment of recurrent ovarian cancer; that there was some promising evidence for cediranib, apatinib, ramucirumab and nintedanib, but that further studies were needed; and that cabozantinib and motesanib could not be recommended for treatment of ovarian cancer because of toxicity. The authors declared no conflicts of interest.
Broekman 2021 is a systematic review of licenced systemic therapies for ovarian cancer, which aimed to assess their benefits according to the European Society of Medical Oncology Magnitude of Clinical Benefit Scale. This review included the angiogenesis inhibitor, bevacizumab, which has been licensed. The authors concluded that the addition of bevacizumab to chemotherapy in the platinum‐resistant setting was one of only three treatments assessed which showed a substantial benefit. One of the authors declared some financial links to two relevant companies; the other authors declared no conflicts of interest.
Liu 2021b is a systematic review of phase II‐III clinical trials of bevacizumab in advanced ovarian cancer. The review included 35 studies, of which eight were included in a quantitative synthesis. This review included single‐arm phase II studies, as well as randomised phase II and III studies, and included studies in which the comparison was not chemotherapy with, versus without, bevacizumab. The authors found that patients with newly‐diagnosed ovarian cancer who were treated with bevacizumab combined with chemotherapy, compared to chemotherapy alone, had improved PFS, but no significant difference in OS. They found that patients with recurrent ovarian cancer treated with regimes including bevacizumab had both improved PFS and improved OS, compared to treatment regimes without bevacizumab. The authors declared no conflicts of interest.
Trillsch 2021 is a meta‐analysis of three studies which included participants with platinum‐resistant ovarian cancer. One of the included studies involved anti‐angiogenesis inhibitors (TRIAS 2018); the two other studies randomised participants to two different topotecan schedules (Sehouli 2011), or oral versus intravenous treosulphan (Sehouli 2017). Datasets were provided by authors of the original studies, who were included as authors of the meta‐analysis. They compared prognoses in participants with platinum‐resistance developed after first‐line chemotherapy (primary platinum resistance (PPR)) versus those who developed platinum resistance after subsequent lines of chemotherapy (secondary platinum resistance (SPR)). They found that PPR had a negative prognostic impact compared with SPR on PFS, although the clinical significance was minimal (3.9 months versus 3.1 months), and the difference in overall survival was not statistically significant. Retrospective subgroup analysis of the TRIAS 2018 study suggested that sorafenib was more effective in those with PPR, where statistically significant improvements in OS (PPR median survival 13.2 months (sorafenib) versus 8.6 months (placebo) (HR 0.52, 95% CI 0.33 to 0.85) compared to those with SPR (median survival 18.6 months (sorafenib) versus 13.5 months (placebo); HR 0.82, 95% CI 0.48 to 1.41) were observed. Similar results were observed for PFS (PPR = 6.9 months (sorafenib) versus 3.8 months (placebo); HR 0.40, 95% CI 0.26 to 0.64); SPR = 5.8 months (sorafenib) versus 4.9 months (placebo) HR 0.83 (95% CI 0.51 to 1.36)). They recognised the need for effective treatments in those with PR disease, especially PPR disease, but also the need to consider reducing treatment burden and introducing palliative care in this cohort, given their poorer prognosis.
Qi 2021 is a systematic review and meta‐analysis focused on evaluating the safety and efficacy of apatinib combined with chemotherapy for treatment of advanced ovarian cancer. This review included 12 studies, all conducted in China. The authors reported that patients treated with apatinib combined with chemotherapy, compared to chemotherapy alone, had higher risk of proteinuria, but did not find a significant difference in risk of other adverse events. This review reported on disease control rates and objective response rates, but did not report on progression‐free survival or overall survival outcomes. The authors declared no conflicts of interest.
These recent systematic reviews were mostly narrower in scope than the current review, and included fewer studies. Where the scopes overlapped, the conclusions of these other reviews were broadly in keeping with the findings of this review. For example, the finding that bevacizumab was beneficial in the setting of recurrent platinum‐resistant ovarian cancer was fairly consistent between reviews, while the evidence of benefit from other agents and in other settings was more variable.
(A) The VEGF‐R is a transmembrane protein, found on cells, which line blood vessels (endothelial cells). (B) Following binding to its ligand, VEGF, the VEGF‐R is stimulated and develops tyrosine kinase activity. (C) Tyrosine kinase activity sets off a sequence of downstream events that lead to stimulation of cell growth and new vessels grow in, to supply the growing tumour. (D) VEGF‐R activity can be blocked by antibodies, which bind to VEGF, and so stop it binding to the receptor, or using chemicals, which inhibit the tyrosine kinase enzyme activity of the VEGF‐R.
PRISMA flow diagram of studies considered for this review update. Please see previous version of review for further details of previous search (Gaitskell 2011).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies
Risk of bias summary: review authors' judgements about each risk of bias item for each included study
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 1: Overall survival
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 2: Progression‐free survival
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 3: Quality of life ‐ Trial Outcome Index score of Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 4: Hypertension (grade ≥ 2)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 5: Proteinuria (grade ≥ 3)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 6: Pain (grade ≥ 2)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 7: Neutropenia (grade ≥ 4)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 8: Febrile neutropenia (any grade)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 9: Venous thromboembolic event (any grade)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 10: Arterial thromboembolic event (any grade)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 11: Non‐central nervous system bleeding (grade ≥ 3)
Comparison 1: Newly‐diagnosed EOC: chemotherapy with bevacizumab versus chemotherapy alone (placebo for all in the maintenance phase), Outcome 12: Gastrointestinal adverse events (grade ≥ 2)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 2: Overall survival by risk status
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 3: Progression‐free survival
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 4: Quality of life
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 5: Any adverse event (grade ≥ 3)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 6: Hypertension (grade ≥ 2)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 7: Proteinuria (grade ≥ 3)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 8: Pain (grade ≥ 2)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 9: Neutropenia (grade ≥ 3)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 10: Febrile neutropenia (any grade)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 11: Venous thromboembolic event (any grade)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 12: Arterial thromboembolic event (any grade)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 13: Non‐central nervous system bleeding (grade ≥ 3)
Comparison 2: Newly‐diagnosed EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 14: Severe gastrointestinal adverse events
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 3: Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 4: Any adverse event (grade ≥ 3)
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 5: Hypertension (grade ≥ 3)
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 6: Abdominal pain (grade ≥ 3)
Comparison 3: Newly‐diagnosed EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 7: Neutropenia (grade ≥ 3)
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 3: Any adverse event (grade ≥ 3)
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 4: Pain (grade 3)
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 5: Abdominal pain (grade ≥ 3)
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 6: Neutropenia (grade ≥ 3)
Comparison 4: Newly‐diagnosed EOC: chemotherapy with TKI (peptide‐Fc fusion protein) followed by TKI maintenance compared to chemotherapy alone, Outcome 7: Febrile neutropenia (any grade)
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 1: Overall survival
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 2: Progression‐free survival
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 3: Quality of life ‐ Functional Assessment of Cancer Therapy (FACT)/National Cancer Center Network (NCCN) Ovarian Symptom Index (FOSI) score
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 4: Hypertension (grade ≥3)
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 5: Proteinuria (grade 3 or 4)
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 6: Abdominal pain (grade ≥3)
Comparison 5: Newly‐diagnosed EOC: maintenance with TKI versus placebo after first‐line chemotherapy, Outcome 7: Neutropenia (grade 3 or 4)
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 1: Progression‐free survival
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 2: Any adverse event (grade ≥ 3)
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 3: Hypertension (grade ≥ 3)
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 4: Abdominal pain (grade ≥ 3)
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 5: Neutropenia (grade ≥ 3)
Comparison 6: Newly‐diagnosed EOC: neoadjuvant chemotherapy with bevacizumab versus chemotherapy alone followed by adjuvant chemotherapy with bevacizumab and bevacizumab maintenance for all, Outcome 6: Gastrointestinal disorders
Comparison 7: Newly‐diagnosed EOC: chemotherapy with celecoxib versus chemotherapy alone, Outcome 1: Overall survival
Comparison 7: Newly‐diagnosed EOC: chemotherapy with celecoxib versus chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 7: Newly‐diagnosed EOC: chemotherapy with celecoxib versus chemotherapy alone, Outcome 3: Febrile neutropenia (grade ≥ 3)
Comparison 7: Newly‐diagnosed EOC: chemotherapy with celecoxib versus chemotherapy alone, Outcome 4: Gastrointestinal adverse events (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 3: Quality of life ‐ Trial Outcome Index score of Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 4: Any adverse event (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 5: Hypertension (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 6: Proteinuria (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 7: Pain (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 8: Abdominal pain (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 9: Neutropenia (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 10: Febrile neutropenia (any grade)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 11: Venous thromboembolic event (grade ≥ 3)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 12: Arterial thromboembolic event (any grade)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 13: Non‐central nervous system bleeding (any grade)
Comparison 8: Recurrent platinum‐sensitive EOC: chemotherapy with bevacizumab followed by bevacizumab maintenance compared to chemotherapy alone, Outcome 14: Gastrointestinal perforations (any grade)
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 3: Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 4: Hypertension (grade 3)
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 5: Proteinuria (grade ≥ 3)
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 6: Neutropenia (grade ≥ 3)
Comparison 9: Recurrent platinum‐sensitive EOC: chemotherapy with TKI followed by TKI maintenance compared to chemotherapy alone, Outcome 7: Febrile neutropenia (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 3: Any adverse event (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 4: Hypertension (grade ≥ 2 & grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 5: Proteinuria (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 6: Neutropenia (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 7: Febrile neutropenia (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 8: Venous thromboembolic event (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 9: Arterial thromboembolic event (grade ≥ 3)
Comparison 10: Recurrent platinum‐resistant EOC: chemotherapy with bevacizumab compared to chemotherapy alone, Outcome 10: Gastrointestinal perforations (grade ≥ 2)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 3: Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 4: Any adverse event (grade ≥ 3)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 5: Hypertension (grade ≥ 3)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 6: Proteinuria (grade ≥ 2)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 7: Pain (grade ≥ 2)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 8: Abdominal pain (grade ≥ 2)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 9: Neutropenia (grade ≥ 3)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 10: Febrile neutropenia (any grade)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 11: Non‐central nervous system bleeding (grade ≥ 3)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 12: Gastrointestinal adverse events (grade ≥ 3)
Comparison 11: Recurrent platinum‐resistant EOC: chemotherapy with TKI compared to chemotherapy alone, Outcome 13: Bowel fistula or perforation (grade ≥ 3)
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 3: Proteinuria (grade ≥ 3)
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 4: Pain (grade ≥ 3)
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 5: Abdominal pain (grade ≥ 3)
Comparison 12: Recurrent platinum‐resistant EOC: chemotherapy with olaratumab compared to chemotherapy alone, Outcome 6: Neutropenia (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 1: Overall survival
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 2: Progression‐free survival
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 3: Quality of life ‐ Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 4: Hypertension (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 5: Proteinuria (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 6: Pain (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 7: Abdominal pain (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 8: Neutropenia (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 9: Febrile neutropenia (any grade)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 10: Venous thromboembolic event (any grade)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 11: Arterial thromboembolic event (any grade)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 12: Non‐central nervous system bleeding (grade ≥ 3)
Comparison 13: Recurrent EOC: chemotherapy with TKI (peptide‐Fc fusion protein) compared to chemotherapy alone, Outcome 13: Gastrointestinal perforation (grade ≥ 3)
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.97 | 2776 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely results in little to no difference in overall survival. | |
| 590 per 1000 | 599 per 1000 | |||||
| Progression‐free survival (PFS) | Averaged | HR 0.82 | 2746 | ⨁◯◯◯ | The evidence is very uncertain about the effect of chemotherapy with bevacizumab on progression‐free survival. | |
| 550 per 1000 | 612 per 1000 | |||||
| Quality of life (QoL) | The mean global quality of life score was 76.1 | MD 6.4 score lower | ‐ | 890 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab results in a slight reduction in global quality of life. |
| Any adverse event grade ≥ 3 | 566 per 1000 | 657 per 1000 | RR 1.16 | 1485 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely increases any adverse event (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) | 44 per 1000 | 224 per 1000 | RR 4.27 | 2707 | ⨁⨁◯◯ | Chemotherapy with bevacizumab may result in a large increase in hypertension (grade ≥ 2). |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 trials (chemotherapy alone arms) | ||||||
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.99 | 1451 | ⨁⨁⨁◯ | Chemotherapy with TKI likely results in little to no difference in OS. | |
| 590 per 1000 | 593 per 1000 | |||||
| Progression‐free survival (PSF) | Averagec | HR 0.88 | 1451 | ⨁⨁⨁◯ | Chemotherapy with TKI likely increases PFS slightly. | |
| 550 per 1000 | 591 per 1000 | |||||
| Quality of life (QoL) Follow‐up: not specified | The mean quality of life score was 70.68 | MD 1.86 score lower | ‐ | 1340 | ⨁⨁⨁◯ | Chemotherapy with TKI likely reduces QoL slightly, although this may not be clinically significant. |
| Any adverse event grade ≥ 3 | 703 per 1000 | 921 per 1000 (780 to 1000) | RR 1.31 | 188 | ⨁⨁⨁◯ | Chemotherapy with TKI likely increases any adverse event (grade ≥ 3) slightly. |
| Hypertension grade ≥ 3 | 7 per 1000 | 43 per 1000 | RR 6.49 | 1352 | ⨁⨁◯◯ | Chemotherapy with TKI may result in a large increase in hypertension grade ≥3. |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 (chemotherapy alone arms) | ||||||
| Patient or population: newly‐diagnosed EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI [peptide‐Fc fusion protein] and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.99 | 1015 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in overall survival. | |
| 590 per 1000 | 593 per 1000 (517 to 659) | |||||
| Progression‐free survival (PSF) | Averaged | HR 0.93 | 1015 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in progression‐free survival. | |
| 550 per 1000 | 574 per 1000 (521 to 624) | |||||
| Quality of life (QoL) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Any adverse event grade ≥ 3 | 661 per 1000 | Ranged from 727 to 1000 | RR ranged from 1.10 (grade 3) to 9.96 (grade 5) | 1011 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely increases any adverse event grade ≥ 3 slightly. |
| Hypertension grade ≥ 3 | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Bowel fistula / perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON7 2015, GOG‐0218 2019 and AGO‐OVAR 12 2020 (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐sensitive EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab and as maintenance | |||||
| Overall survival (OS) | Averagea | HR 0.90 | 1564 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in overall survival. | |
| 490 per 1000 | 526 per 1000 | |||||
| Progression‐free survival (PFS) | Averagec | HR 0.56 | 1564 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely increases progression free‐survival. | |
| 230 per 1000 | 439 per 1000 | |||||
| Quality of life (QoL) Assessed with: TOI‐FACT‐OC questionnaire Scale from: 0 to 152 | The mean quality of life was 77 | MD 0.8 higher | ‐ | 486 | ⨁⨁◯◯ | Chemotherapy with bevacizumab followed by maintenance bevacizumab likely results in little to no difference in quality of life. |
| Any adverse event (grade ≥3) | 804 per 1000 | 892 per 1000 | RR 1.11 | 1538 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab followed by maintenance bevacizumab increases any adverse event (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) Assessed with: CTCAE versions 3.0‐4.0 where reported | ‐ | ‐ | ‐ | ‐ | ‐ | All three trials included in this comparison reported only hypertension grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 3.0‐4.0 where reported | ‐ | ‐ | ‐ | ‐ | ‐ | Two trials included in this comparison (MITO‐16b 2021 and GOG‐0213 2017) reported only gastrointestinal perforations of any grade. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON6 2021, GOG‐0213 2017 and OCEANS 2015 trials (chemotherapy arms only). | ||||||
| Patient or population: recurrent platinum‐sensitive EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI and as maintenance | |||||
| Overall survival (OS) Assessed with: survival rate | Average | HR 0.86 | 282 | ⨁⨁◯◯ | Chemotherapy with TKI followed by maintenance with TKI likely results in little to no difference in overall survival. | |
| 490 per 1000a,b | 541 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.0 criteria | Average | HR 0.56 | 282 | ⨁⨁⨁◯ | Chemotherapy with TKI followed by maintenance with TKI likely increases progression‐free survival. | |
| 230 per 1000 | 439 per 1000 | |||||
| Quality of life (QoL) Follow‐up: 12 months | The mean quality of lIfe was 62.6 | MD 6.1 higher | ‐ | 146 | ⨁⨁◯◯ | Chemotherapy with TKI followed by maintenance with TKI may result in little to no difference in quality of life. |
| Any adverse events (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Hypertension (grade ≥ 2) | ‐ | ‐ | ‐ | ‐ | ‐ | A single trial included in this comparison reported only events of grade ≥ 3 (ICON6 2021). |
| Bowel fistula/perforation (grade ≥ 3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in ICON6 2021, OCEANS 2015 and GOG‐0213 2017 trials (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐resistant EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with bevacizumab | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.73 | 778 | ⨁⨁⨁⨁ | Chemotherapy with bevacizumab increases overall survival. | |
| 10 per 1000 | 35 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.0‐1.1 where reported | Averageb | HR 0.49 | 778 | ⨁⨁⨁◯ | Chemotherapy with bevacizumab likely results in a large increase in progression‐free survival. | |
| 40 per 1000 | 207 per 1000 | |||||
| Quality of life | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Any adverse event grade ≥ 3 | 460 per 1000 | 773 per 1000 (350 to 1000) | RR 1.68 (0.76 to 3.69) | 101 (1 RCT) | ⨁⨁◯◯ | Chemotherapy with bevacizumab may increase any adverse events (grade > 3) slightly. |
| Hypertension (grade ≥ 2) Assessed with: CTCAE version 3.0 | 73 per 1000 | 228 per 1000 | RR 3.11 | 436 | ⨁⨁◯◯ | Chemotherapy with bevacizumab may result in a large increase in hypertension (grade ≥ 2). |
| Bowel fistula / perforation (grade ≥ 2) | 4 per 1000g | 28 per 1000 (3 to 220) | RR 6.89 (0.86 to 55.09) | 436 (2 RCTs) | ⨁⨁◯◯ | Chemotherapy with bevacizumab may increase rates of bowel fistula / perforation (grade ≥ 2) slightly. Two studies included in this comparison although one reported only gastrointestinal perforations (grade ≥ 2) (AURELIA 2014). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in AURELIA 2014, MITO‐11 2015 and TRIAS 2018 trials (chemotherapy alone arms). | ||||||
| Patient or population: recurrent platinum‐resistant EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.85 | 940 | ⨁⨁⨁◯ | Chemotherapy with TKI likely results in little to no difference in overall survival. | |
| 10 per 1000 | 16 per 1000 | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST 1.1 criteria where specified | Averagec | HR 0.70 | 940 | ⨁⨁◯◯ | Chemotherapy with TKI may increase progression‐free survival. | |
| 40 per 1000 | 87 per 1000 | |||||
| Quality of life (QoL) assessed with: Global Quality of Life and EORTC core QoL questionnaire (QLQ‐C30) Scale from: 0 to 100 | MD in Quality of Life score ranged from ‐0.19 (95%CI ‐9.77 to 9.39) at 6 weeks (METRO‐BIBF 2020) to ‐3.40 (95%CI ‐13.22 to 6.42) at 4 months (TAPAZ 2022) | 164 | ⨁⨁◯◯ | Chemotherapy with TKI may result in little to no difference in quality of life. | ||
| Any adverse events (grade ≥3) Assessed with: CTCAE versions 3.0‐4.1 | 581 per 1000 | 657 per 1000 | RR 1.23 | 548 | ⨁⨁◯◯ | Chemotherapy with TKI may increase any adverse events (grade ≥ 3) slightly. |
| Hypertension (grade ≥ 2) | ‐ | ‐ |
|
|
| Trials included in this comparison reported only on events of grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 4.0‐4.1 | 4 per 1000 | 11 per 1000 | RR 2.74 | 557 | ⨁◯◯◯ | The evidence is very uncertain about the effect of chemotherapy with TKI on bowel fistula/perforation (grade ≥ 3). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in AURELIA 2014, MITO‐11 2015 and TRIAS 2018 (chemotherapy alone arms). | ||||||
| Patient or population: recurrent EOC | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with chemotherapy alone | Risk with chemotherapy with TKI [peptide‐Fc fusion protein] | |||||
| Overall survival (OS) Assessed with: survival rate | Averagea | HR 0.92 | 1250 | ⨁⨁⨁◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] likely results in little to no difference in overall survival. | |
| 60 per 1000 | 75 per 1000 (51 to 105) | |||||
| Progression‐free survival (PFS) Assessed with: progression‐free rate according to RECIST version 1.1 criteria | Averaged | HR 0.73 | 1250 | ⨁⨁⨁⨁ | Chemotherapy with TKI [peptide‐Fc fusion protein] increases progression‐free survival. | |
| 110 per 1000 | 200 per 1000 (164 to 238) | |||||
| Quality of life (QoL) Assessed with: TOI‐FACT‐OC questionnaire Scale from: 0 to 152 Follow‐up: 25 weeks | The mean change from baseline QoL was ‐1.6 | MD 0.8 lower | ‐ | 315 | ⨁⨁◯◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] may result in little to no difference in quality of life. |
| Any adverse events (grade ≥3) | ‐ | ‐ | ‐ | ‐ | ‐ | Outcome not reported |
| Hypertension (grade ≥ 2) | ‐ | ‐ | ‐ | ‐ | ‐ | All three trials included in this comparison reported events of grade ≥ 3. |
| Bowel fistula / perforation (grade ≥ 3) Assessed with: CTCAE versions 3.0 | 18 per 1000 | 6 per 1000 | RR 0.35 | 108 | ⨁⨁◯◯ | Chemotherapy with TKI [peptide‐Fc fusion protein] may result in little to no difference in bowel perforation/fistula G3+. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
GRADE Working Group grades of evidence | ||||||
| aThe control risk is an average number of participants reported alive at 36 months in TAPAZ 2022, TRINOVA‐1 2016 and TRINOVA‐2 2017 trials (chemotherapy alone arms). | ||||||
| Study ID | Number of references | Intervention/s (N) | Control (N) | Number randomised | Randomisation ratio | Type of anti‐angiogenesis agent | Newly‐diagnosed or relapsed/recurrent EOC | Population in relation to platinum‐sensitivity* | Percentage (%) stage IV (newly‐diagnosed EOC only) | Prior treatment |
|---|---|---|---|---|---|---|---|---|---|---|
| Newly‐diagnosed EOC | ||||||||||
| 6 | Chemotherapy + nintedanib (911) | Chemotherapy | 1366 | 2:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Newly‐diagnosed | PS 100% | 24% in intervention arm; 24% in control arm); overall 24% | N/A | |
| 13 | Pazopanib | Placebo | 940 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Newly‐diagnosed | PS 100% | 16.3% in intervention group; 16.9% in control group; overall 16.6% | N/A | |
| 6 | Chemotherapy + bevacizumab | Chemotherapy | 95 | 2:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 26% in intervention group; 35% in control group; overall 30% | N/A | |
| 6 | Chemotherapy + nintedanib (124) | Chemotherapy | 188 | 2:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Newly‐diagnosed | PS 100% | N/A | N/A | |
| 3 | Chemotherapy + bevacizumab | Chemotherapy | 68 | 1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 34% in intervention arm and 33% in control arm; 33.8% overall | N/A | |
| 20 | Chemotherapy + bevacizumab (625) | Chemotherapy | 1873 | 1:1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 26.2% intervention in initiation only arm; 26.5% in initiation and maintenance arm; 24.5% in control arm; 25.7% overall | N/A | |
| 4 | Chemotherapy + sorafenib (43) | Chemotherapy | 85 | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Newly‐diagnosed | PS 100% | 33% in control arm; 19% in intervention arm; 25.9% overall | N/A | |
| 3 | Sorafenib | Placebo | 246 | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Newly‐diagnosed | PS 100% | Stage at diagnosis not provided but all stage III/V; 8.1% suboptimally debulked at primary surgery in each arm | N/A | |
| 16 | Chemotherapy + bevacizumab | Chemotherapy | 1528 | 1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed | PS 100% | 12% in control arm; 13% in intervention arm; 13.2% overall | N/A | |
| 2 | Chemotherapy + celecoxib (97) | Chemotherapy (99) | 196 | 1:1 | Celecoxib: COX‐2 inhibitor | Newly‐diagnosed | PS 100% | 25.3% in control arm; 22.7% in intervention arm; 23.7% overall | N/A | |
| 2 | Chemotherapy + trebananib | Chemotherapy | 1015 | 2:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Newly‐diagnosed | PS 100% | 24% in control arm; 27% in intervention arm; 26.2% overall | N/A | |
| Platinum‐sensitive recurrence | ||||||||||
| 5 | Niraparib + bevacizumab | Niraparib | 97 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS (66%) PPS (34%) | N/A | Platinum‐based chemotherapy. Overall previous lines of treatment: one = 49.5%; two = 44.3%; three or more = 6% | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 164 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based chemotherapy | |
| 3 | Chemotherapy + bevacizumab (377) | Chemotherapy (337) | 674 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based chemotherapy | |
| 8 | Chemotherapy + cediranib + | Chemotherapy | 486+ (456) | 2:3:3 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PS 67% | N/A | Platinum‐based chemotherapy, 89% with paclitaxel. Overall 5% had had previous bevacizumab treatment. | |
| 1 | Chemotherapy with paclitaxel and carboplatin + bevacizumab (34) | Chemotherapy with paclitaxel and carboplatin (34) | 68 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS 100% | N/A | Platinum‐based chemotherapy at least (presumed) | |
| 5 | Olaparib + cediranib (44) | Olaparib (46) | 90 | 1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PS 100% | N/A | Platinum‐based chemotherapy and max 1 non‐platinum therapy in recurrent setting | |
| 4 | Cediranib + olaparib | Chemotherapy | 565 | 1:1:1 | TKI with PARPi | Recurrent | PS
| N/A | Platinum and non‐platinum based chemotherapy (65% only 1 prior line of chemotherapy) | |
| 2 | Chemotherapy + bevacizumab | Chemotherapy | 406 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS 100% | N/A | First‐line platinum‐based treatment, including bevacizumab | |
| 12 | Chemotherapy + bevacizumab (242) | Chemotherapy + placebo (242) | 484 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PS | N/A | Platinum‐based front‐line chemotherapy | |
| Platinum‐resistant recurrence | ||||||||||
| 5 | Olaparib + cediranib (16) | Olaparib + durvalumab (14) | 30 for relevant comparison [3 other arms, N = 70 in total] | 1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR | N/A | At least 2 prior lines of anticancer therapy | |
| 3 | Chemotherapy + apatinib (78) | Chemotherapy | 150 | 1:1 | Apatinib: TKI targeting VEGF‐R2 | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy | |
| 16 | Chemotherapy + bevacizumab | Chemotherapy | 361 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy (max 2) | |
| 2 | Cediranib‐olaparib combination (continuous n= 41) (intermittent n= 41) | Weekly paclitaxel (n= 41) | 123 | 1:1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR (100%) | N/A | 39.8% up to 2 previous lines of chemotherapy; 60.2% ≥3 previous lines of chemotherapy. 53.7% prior anti‐angiogenic treatment | |
| 2 | Bevacizumab (33) atezolizumab + cevacizumab + placebo Atezolizumab + bevacizumab + acetylsalicylic acid | Atezolizumab + placebo | 122 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy (max of 2 non‐platinum regimens) | |
| 5 | Aflibercept (29) | Placebo (26) | 55 | 1:1 | Aflibercept: fusion protein targeting VEGF‐A and VEGF‐B | Recurrent | PR | N/A | At least 2 lines of previous chemotherapy, one platinum‐based | |
| 1 | Bevacizimab + albumin‐binding paclitaxel | Albumin‐binding paclitaxel | 70 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR | N/A | Unclear as English language abstract only | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 86 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy | |
| 1 | Chemotherapy + bevacizumab (38) | Chemotherapy (38) | 76 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR | N/A | Platinum‐based chemotherapy; platinum‐free interval 4.3 months ±0.6 months control group and 4.8 ±0.8 months in bevacizumab group (P=0.06) | |
| 3 | Chemotherapy + olaratumab (62) | Chemotherapy (61) | 123 | 1:1 | Olaratumab: monoclonal antibody targeting PDGFR‐α | Recurrent | PR | N/A | Platinum‐based chemotherapy | |
| 3 | Cyclophosphamide + nintedanib | Cyclophosphamide | 117 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PR or intolerant 100% | N/A | Two or more lines of chemotherapy | |
| 3 | Chemotherapy + pazopanib (37) | Chemotherapy (37) | 74 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PR | N/A | Prevous chemotherapy lines: one = 43.8%; two = 47.9% three or more = 8.2% | |
| 2 | Nintedanib | Chemotherapy | 91 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PR 100% | N/A | Platinum‐based chemotherapy. participants had clear cell carcinoma of EOC or endometrial origin. 91 participants with EOC. | |
| 1 | Chemotherapy + bevacizumab | Chemotherapy | 103 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR 100% | N/A | previously treated with ≥3 cycles of bevacizumab + platinum chemotherapy; progression occurred <6 months after completion of platinum treatment | |
| 3 | Olaparib + cediranib | Olaparib (46) Chemotherapy (46)
| 139 | 1:1:1 | Cediranib: TKI targeting VEGF‐R, PDGF‐R, and c‐kit | Recurrent | PR 100% | N/A | Prior PARPi therapy (22%) | |
| 3 | Ixabepilone + bevacizumab | Ixabepilone | 76 | 1:1 | Bevacizumab: antibody against VEGF | Recurrent | PR or refractory 100% | N/A | Not reported | |
| 2 | Etoposide + cyclophosphamide + pazopanib | Etoposide + cyclophosphamide | 75 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R, c‐kit, and FGF‐R | Recurrent | PR 51%; | N/A | Prior treatment with at least 2 chemotherapy regimens in advanced tumor | |
| 3 | Chemotherapy + vandetanib (63) | Chemotherapy (66) | 129 | 1:1 | Vandetanib: TKI targeting VEGF‐R, EGF‐R, and RET | Recurrent | All patients were considered platinum resistant or refractory | N/A | Platinum‐based front‐line chemotherapy +/‐ up to 3 chemotherapy regimens in current setting +/‐ primary anti‐angiogenic therapy | |
| 3 | Topotecan + sorafenib | Topotecan + placebo | 174 (172) | 1:1 | Sorafenib: TKI targeting VEGF‐R, PDGF‐R, and RAF kinases | Recurrent | PR or refractory 100% | N/A | No more than two prior treatment regimens for recurrent EOC | |
| Mixed platinum‐sensitive and platinum‐resistant recurrence | ||||||||||
| 3 | Chemotherapy + pazopanib (75) | Chemotherapy (76) | 148 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PS (40%) PR (60%) | N/A | Chemotherapy (max 3) | |
| 2 | Cyclophosphamide + celecoxib | Cyclophosphamide | 52 | 1:1 | Celecoxib: COX‐2 inhibitor | Recurrent | PS 38.5% | N/A | No limit on prior lines of therapy | |
| 9 | Chemotherapy + lower‐dose trebananib (AMG386) (53) Chemotherapy + higher‐dose trebananib (AMG386) (53) | Chemotherapy + placebo (55) | 161 | 1:1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PS (52%) PR (47%) | N/A | Platinum and non‐platinum based chemotherapy (max 3 in total) | |
| 3 | Nintedanib (BIBF 1120) (43) | Placebo (41) | 84 | 1:1 | Nintedanib: TKI targeting VEGF‐R, PDGF‐R and FGF‐R | Recurrent | PS (59%) PR (41%) | N/A | Chemotherapy (2 or more rounds) | |
| 2 | Cabozantinib (57) | Chemotherapy (54) | 111 | 1:1 | Cabozantinib: TKI targeting VEGF‐R2, c‐MET, c‐kit, Tie2, FLT‐3, and RET | Recurrent | PR (50%) PS (50%) | N/A | Platinum‐based chemotherapy +/‐ non‐platinum based regimens (max 3 in total) | |
| 3 | Chemotherapy + pazopanib (54) | Chemotherapy + placebo (52) | 106 | 1:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PR (51%) PS (49%) | N/A | Platinum‐based chemotherapy +/‐ non‐platinum based chemotherapy (max 3 in total) | |
| 3 | Paclitaxel + pazopanib (79) | Paclitaxel | 116 | 2:1 | Pazopanib: TKI targeting VEGF‐R, PDGF‐R and c‐kit | Recurrent | PPS 70.7%; PR 29.3% | N/A | Not reported | |
| 8 | Chemotherapy + trebananib (461) | Chemotherapy + placebo (458) | 919 | 1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PR (53%) PS (47%) | N/A | Platinum‐based chemotherapy +/‐ up to 2 other chemotherapy regimens +/‐ anti‐angiogenic therapy | |
| 3 | Chemotherapy + trebananib (114) | Chemotherapy + placebo (109) | 223 | 1:1 | Trebananib: TKI targeting Ang1 and Ang2 (angiopoietins) | Recurrent | PR (59%) PS (41%) | N/A | Platinum‐based chemotherapy +/‐ up to 2 other chemotherapy regimens +/‐ anti‐angiogenic therapy | |
| Other | ||||||||||
| 3 | Chemotherapy (two different regimes) + bevacizumab | Chemotherapy (two different regimes) | 50 | 1:1:1:1 | Bevacizumab: antibody against VEGF | Newly‐diagnosed & recurrent. Mucinous EOC only | N/A | N/A | No previous chemotherapy | |
| 3 | Intraperitoneal chemotherapy + bevacizumab (31) | Intraperitoneal chemotherapy (27) | 58 | 1:1 | Bevacizumab: antibody against VEGF | Unclear | Unclear | 77.4% in intervention group; 77.8% in control group; overall 77.6% | Unclear | |
| PPS: partially‐platinum sensitive; ALT: alanine transaminase; AMG386: trebananib; Ang1: angiopoietin 1; Ang2: angiopoietin 2; AST: aspartate aminotransferase; AUC: area under the curve; BIBF : BIBF 1120 = nintedanib; BRCA: breast cancer gene; CA125: cancer antigen 125; COX‐2: cyclo‐oxygenase‐2; CT:computed tomography; CTCAE: Common Terminology Criteria for Adverse Events; ECOG: Eastern Cooperative Oncology Group; EOC: epithelial ovarian cancer; EORTC: European Organisation for Research and Treatment of Cancer; FACT/GOG NTX: Functional Assessment of Cancer Therapy/Gynecologic Oncology Group – Neurotoxicity; FACT‐O (TOI): Functional Assessment of Cancer Therapy‐Ovarian (Trial Outcome Index); FGF‐R : fibroblast growth factor receptor; FIGO: International Federation of Gynecology and Obstetrics; FLT‐3: Fms‐like receptor tyrosine kinase‐3; FOSI: Functional Assessment of Cancer Therapy (FACT)/National Comprehensive Cancer Network (NCCN) Ovarian Symptom Index; GCIG: Gynecological Cancer InterGroup; GFR: glomerular filtration rate; GOG: Gynecologic Oncology Group; HRD: homologous recombination deficiency; ICON: International Collaborative Ovarian Neoplasm study; IDS: interval debulking surgery; IQR: interquartile range; ITT: intention‐to‐treat; IV: intravenous(ly); IVRS/IWRS: Interactive Voice Response System/ Interactive Web Response System; KGOG: Korean Gynecologic Oncology Group; KPS: Karnofsky Performance Status; MET: mesenchymal epithelial transition; MRI: magnetic resonance imaging; NCI: National Cancer Institute; ORR: objective response rate; OS: overall survival; PARP: poly(ADP‐ribose) polymerase; PARPi: poly(ADP‐ribose) polymerase inhibitor; PDGF‐R: platelet‐derived growth factor receptor; PDGFR‐α: alpha subunit of PDGF‐R; PD‐L1: Programmed death‐ligand 1; PLB: placebo; PLD: pegylated liposomal doxorubicin; PFS: progression‐free survival; PPS: partially‐platinum sensitive; PR: platinum‐resistant; PS: platinum‐sensitive; QoL: quality of life; RAF: Rapidly Accelerated Fibrosarcoma; RECIST: Response Evaluation Criteria in Solid Tumors; RET:REarranged during Transfection; SWOG: Southwest Oncology Group; Tie2: angiopoetin‐1 receptor; TKI: tyrosine kinase inhibitor; TNM: tumour nodes metastases; ULN: upper limit of normal; USS: ultrasound scan; VEGF: vascular endothelial growth factor; VEGF‐R: vascular endothelial growth factor receptor; WHO: World Health Organization +ICON6 2021 486 randomised overall, of which 30 randomised to initial 30 mg dose of cediranib and excluded because of increased toxic effects, leaving 456. | ||||||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Overall survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.2 Progression‐free survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.3 Quality of life ‐ Trial Outcome Index score of Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 1.4 Hypertension (grade ≥ 2) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.5 Proteinuria (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.6 Pain (grade ≥ 2) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.7 Neutropenia (grade ≥ 4) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.8 Febrile neutropenia (any grade) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.9 Venous thromboembolic event (any grade) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 1.10 Arterial thromboembolic event (any grade) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.11 Non‐central nervous system bleeding (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.12 Gastrointestinal adverse events (grade ≥ 2) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Overall survival Show forest plot | 2 | 2776 | Hazard Ratio (IV, Fixed, 95% CI) | 0.97 [0.88, 1.07] |
| 2.2 Overall survival by risk status Show forest plot | 2 | Hazard Ratio (IV, Fixed, 95% CI) | Subtotals only | |
| 2.2.1 Women at high risk for disease progression | 2 | 1316 | Hazard Ratio (IV, Fixed, 95% CI) | 0.86 [0.76, 0.98] |
| 2.2.2 Women at lower risk for disease progression | 2 | 1460 | Hazard Ratio (IV, Fixed, 95% CI) | 1.13 [0.97, 1.31] |
| 2.3 Progression‐free survival Show forest plot | 2 | 2746 | Hazard Ratio (IV, Random, 95% CI) | 0.82 [0.64, 1.05] |
| 2.4 Quality of life Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.4.1 Trial Outcome Index score of Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.4.2 Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30 | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 2.5 Any adverse event (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 2.6 Hypertension (grade ≥ 2) Show forest plot | 2 | 2707 | Risk Ratio (IV, Fixed, 95% CI) | 4.27 [3.25, 5.60] |
| 2.7 Proteinuria (grade ≥ 3) Show forest plot | 2 | 2707 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.97 [1.20, 3.23] |
| 2.8 Pain (grade ≥ 2) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 2.9 Neutropenia (grade ≥ 3) Show forest plot | 2 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 2.9.1 Grade ≥ 3 | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 2.9.2 Grade ≥ 4 | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 2.10 Febrile neutropenia (any grade) Show forest plot | 2 | 2707 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.33 [0.87, 2.04] |
| 2.11 Venous thromboembolic event (any grade) Show forest plot | 2 | 2707 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.39 [1.03, 1.89] |
| 2.12 Arterial thromboembolic event (any grade) Show forest plot | 2 | 2707 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.95 [1.07, 3.54] |
| 2.13 Non‐central nervous system bleeding (grade ≥ 3) Show forest plot | 2 | 2707 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.13 [0.87, 5.20] |
| 2.14 Severe gastrointestinal adverse events Show forest plot | 2 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 2.14.1 Grade ≥ 2 GI events | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 2.14.2 Grade ≥ 3 GI perforation | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 3.1 Overall survival Show forest plot | 2 | 1451 | Hazard Ratio (IV, Fixed, 95% CI) | 0.99 [0.84, 1.17] |
| 3.1.1 Chemo with nintedanib [nintedanib maintenance] | 1 | 1366 | Hazard Ratio (IV, Fixed, 95% CI) | 0.99 [0.83, 1.18] |
| 3.1.2 Chemo with sorafenib [sorafenib maintenance] | 1 | 85 | Hazard Ratio (IV, Fixed, 95% CI) | 1.03 [0.34, 3.12] |
| 3.2 Progression‐free survival Show forest plot | 2 | 1451 | Hazard Ratio (IV, Fixed, 95% CI) | 0.88 [0.77, 1.00] |
| 3.2.1 Chemo with nintedanib [nintedanib maintenance] | 1 | 1366 | Hazard Ratio (IV, Fixed, 95% CI) | 0.86 [0.75, 0.98] |
| 3.2.2 Chemo with sorafenib [sorafenib maintenance] | 1 | 85 | Hazard Ratio (IV, Fixed, 95% CI) | 1.21 [0.74, 1.97] |
| 3.3 Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30 Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3.4 Any adverse event (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 3.4.1 Chemo with nintedanib [nintedanib maintenance] | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 3.5 Hypertension (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 3.5.1 Chemo with nintedanib [nintedanib maintenance] | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 3.6 Abdominal pain (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 3.6.1 Chemo with nintedanib [nintedanib maintenance] | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 3.7 Neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 3.7.1 Chemo with nintedanib [nintedanib maintenance] | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 4.1 Overall survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.1.1 Chemo with trebananib [trebananib maintenance] | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.2 Progression‐free survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.2.1 Chemo with trebananib [trebananib maintenance] | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.3 Any adverse event (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.3.1 Chemo with trebananib [trebananib maintenance] | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.4 Pain (grade 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.4.1 Chemo with trebananib [trebananib maintenance] | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.5 Abdominal pain (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.5.1 Chemo with trebananib [trebananib maintenance] | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.6 Neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.6.1 Chemo with trebananib [trebananib maintenance] | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 4.7 Febrile neutropenia (any grade) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.7.1 Chemo with trebananib [trebananib maintenance] | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 5.1 Overall survival Show forest plot | 2 | 1186 | Hazard Ratio (IV, Fixed, 95% CI) | 0.98 [0.83, 1.16] |
| 5.1.1 Pazopanib | 1 | 940 | Hazard Ratio (IV, Fixed, 95% CI) | 0.96 [0.80, 1.14] |
| 5.1.2 Sorafenib | 1 | 246 | Hazard Ratio (IV, Fixed, 95% CI) | 1.48 [0.68, 3.20] |
| 5.2 Progression‐free survival Show forest plot | 2 | 1186 | Hazard Ratio (IV, Fixed, 95% CI) | 0.81 [0.69, 0.95] |
| 5.2.1 Pazopanib | 1 | 940 | Hazard Ratio (IV, Fixed, 95% CI) | 0.77 [0.65, 0.91] |
| 5.2.2 Sorafenib | 1 | 246 | Hazard Ratio (IV, Fixed, 95% CI) | 1.09 [0.72, 1.64] |
| 5.3 Quality of life ‐ Functional Assessment of Cancer Therapy (FACT)/National Cancer Center Network (NCCN) Ovarian Symptom Index (FOSI) score Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 5.3.1 Sorafenib | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 5.4 Hypertension (grade ≥3) Show forest plot | 2 | 1184 | Risk Ratio (M‐H, Fixed, 95% CI) | 5.63 [3.81, 8.31] |
| 5.4.1 Pazopanib | 1 | 938 | Risk Ratio (M‐H, Fixed, 95% CI) | 5.46 [3.67, 8.13] |
| 5.4.2 Sorafenib | 1 | 246 | Risk Ratio (M‐H, Fixed, 95% CI) | 10.00 [1.30, 76.94] |
| 5.5 Proteinuria (grade 3 or 4) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 5.5.1 Pazopanib | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 5.6 Abdominal pain (grade ≥3) Show forest plot | 2 | 1184 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.46 [0.52, 4.07] |
| 5.6.1 Pazopanib | 1 | 938 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.55 [0.51, 4.69] |
| 5.6.2 Sorafenib | 1 | 246 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.00 [0.06, 15.81] |
| 5.7 Neutropenia (grade 3 or 4) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 5.7.1 Pazopanib | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 6.1 Progression‐free survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 6.2 Any adverse event (grade ≥ 3) Show forest plot | 2 | 163 | Risk Ratio (IV, Fixed, 95% CI) | 0.84 [0.66, 1.06] |
| 6.3 Hypertension (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.4 Abdominal pain (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.5 Neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6.6 Gastrointestinal disorders Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 7.1 Overall survival Show forest plot | 1 | 196 | Hazard Ratio (IV, Fixed, 95% CI) | 1.16 [0.86, 1.57] |
| 7.2 Progression‐free survival Show forest plot | 1 | 196 | Hazard Ratio (IV, Fixed, 95% CI) | 1.07 [0.85, 1.34] |
| 7.3 Febrile neutropenia (grade ≥ 3) Show forest plot | 1 | 196 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.94 [0.45, 1.96] |
| 7.4 Gastrointestinal adverse events (grade ≥ 3) Show forest plot | 1 | 196 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.15 [0.46, 2.85] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 8.1 Overall survival Show forest plot | 3 | 1564 | Hazard Ratio (IV, Fixed, 95% CI) | 0.90 [0.79, 1.02] |
| 8.2 Progression‐free survival Show forest plot | 3 | 1564 | Hazard Ratio (IV, Fixed, 95% CI) | 0.56 [0.50, 0.63] |
| 8.3 Quality of life ‐ Trial Outcome Index score of Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8.4 Any adverse event (grade ≥ 3) Show forest plot | 3 | 1538 | Risk Ratio (IV, Fixed, 95% CI) | 1.11 [1.07, 1.16] |
| 8.5 Hypertension (grade ≥ 3) Show forest plot | 3 | 1538 | Risk Ratio (M‐H, Fixed, 95% CI) | 5.82 [3.84, 8.83] |
| 8.6 Proteinuria (grade ≥ 3) Show forest plot | 3 | 1538 | Risk Ratio (M‐H, Fixed, 95% CI) | 20.27 [6.42, 64.00] |
| 8.7 Pain (grade ≥ 3) Show forest plot | 2 | 1058 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.09 [1.81, 5.28] |
| 8.8 Abdominal pain (grade ≥ 3) Show forest plot | 2 | 1058 | Risk Ratio (M‐H, Fixed, 95% CI) | 16.88 [4.72, 60.34] |
| 8.9 Neutropenia (grade ≥ 3) Show forest plot | 2 | 1058 | Risk Ratio (IV, Fixed, 95% CI) | 1.04 [0.83, 1.31] |
| 8.10 Febrile neutropenia (any grade) Show forest plot | 3 | 1538 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.20 [0.70, 2.06] |
| 8.11 Venous thromboembolic event (grade ≥ 3) Show forest plot | 2 | 1137 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.73 [0.65, 4.60] |
| 8.12 Arterial thromboembolic event (any grade) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 8.13 Non‐central nervous system bleeding (any grade) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 8.14 Gastrointestinal perforations (any grade) Show forest plot | 2 | 1058 | Risk Ratio (M‐H, Fixed, 95% CI) | 4.96 [0.86, 28.51] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 9.1 Overall survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 9.2 Progression‐free survival Show forest plot | 1 | Hazard Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 9.3 Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30 Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 9.4 Hypertension (grade 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 9.5 Proteinuria (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 9.6 Neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 9.7 Febrile neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 10.1 Overall survival Show forest plot | 5 | 778 | Hazard Ratio (IV, Fixed, 95% CI) | 0.73 [0.61, 0.88] |
| 10.2 Progression‐free survival Show forest plot | 5 | 778 | Hazard Ratio (IV, Fixed, 95% CI) | 0.49 [0.42, 0.58] |
| 10.3 Any adverse event (grade ≥ 3) Show forest plot | 1 | 101 | Risk Ratio (M‐H, Random, 95% CI) | 1.28 [0.88, 1.87] |
| 10.4 Hypertension (grade ≥ 2 & grade ≥ 3) Show forest plot | 5 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 10.4.1 Grade ≥ 2 | 2 | 436 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.11 [1.83, 5.27] |
| 10.4.2 Grade ≥ 3 | 5 | 769 | Risk Ratio (M‐H, Fixed, 95% CI) | 3.83 [1.79, 8.20] |
| 10.5 Proteinuria (grade ≥ 3) Show forest plot | 4 | 683 | Risk Ratio (M‐H, Fixed, 95% CI) | 6.26 [1.13, 34.70] |
| 10.6 Neutropenia (grade ≥ 3) Show forest plot | 3 | 308 | Risk Ratio (IV, Fixed, 95% CI) | 1.35 [1.01, 1.80] |
| 10.7 Febrile neutropenia (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 10.8 Venous thromboembolic event (grade ≥ 3) Show forest plot | 2 | 436 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.58 [0.21, 1.63] |
| 10.9 Arterial thromboembolic event (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 10.10 Gastrointestinal perforations (grade ≥ 2) Show forest plot | 2 | 436 | Risk Ratio (M‐H, Fixed, 95% CI) | 6.89 [0.86, 55.09] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 11.1 Overall survival Show forest plot | 8 | 940 | Hazard Ratio (IV, Random, 95% CI) | 0.85 [0.68, 1.08] |
| 11.1.1 Chemo with apatinib | 1 | 152 | Hazard Ratio (IV, Random, 95% CI) | 0.66 [0.30, 1.44] |
| 11.1.2 Chemo with nintedanib | 1 | 117 | Hazard Ratio (IV, Random, 95% CI) | 1.03 [0.69, 1.54] |
| 11.1.3 Chemo with pazopanib | 4 | 370 | Hazard Ratio (IV, Random, 95% CI) | 0.79 [0.57, 1.09] |
| 11.1.4 Chemo with sorafenib | 1 | 172 | Hazard Ratio (IV, Random, 95% CI) | 0.65 [0.45, 0.93] |
| 11.1.5 Chemo with vandetanib | 1 | 129 | Hazard Ratio (IV, Random, 95% CI) | 1.25 [0.93, 1.68] |
| 11.2 Progression‐free survival Show forest plot | 8 | 940 | Hazard Ratio (IV, Random, 95% CI) | 0.70 [0.55, 0.89] |
| 11.2.1 Chemo with apatinib | 1 | 152 | Hazard Ratio (IV, Random, 95% CI) | 0.44 [0.28, 0.70] |
| 11.2.2 Chemo with nintedanib | 1 | 117 | Hazard Ratio (IV, Random, 95% CI) | 0.91 [0.62, 1.33] |
| 11.2.3 Chemo with pazopanib | 4 | 370 | Hazard Ratio (IV, Random, 95% CI) | 0.68 [0.47, 0.98] |
| 11.2.4 Chemo with sorafenib | 1 | 172 | Hazard Ratio (IV, Random, 95% CI) | 0.60 [0.43, 0.83] |
| 11.2.5 Chemo with vandetanib | 1 | 129 | Hazard Ratio (IV, Random, 95% CI) | 0.99 [0.78, 1.25] |
| 11.3 Quality of life ‐ Global Quality of Life European Organization for Research and Treatment of Cancer Questionnaire QLQ‐C30 Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 11.4 Any adverse event (grade ≥ 3) Show forest plot | 4 | 548 | Risk Ratio (IV, Random, 95% CI) | 1.23 [1.02, 1.49] |
| 11.4.1 Chemo with apatinib | 1 | 146 | Risk Ratio (IV, Random, 95% CI) | 2.22 [1.30, 3.81] |
| 11.4.2 Chemo with nintedanib | 1 | 114 | Risk Ratio (IV, Random, 95% CI) | 1.18 [0.87, 1.60] |
| 11.4.3 Chemo with pazopanib | 1 | 116 | Risk Ratio (IV, Random, 95% CI) | 1.24 [0.99, 1.56] |
| 11.4.4 Chemo with sorafenib | 1 | 172 | Risk Ratio (IV, Random, 95% CI) | 1.09 [0.98, 1.20] |
| 11.5 Hypertension (grade ≥ 3) Show forest plot | 9 | 1075 | Risk Ratio (M‐H, Random, 95% CI) | 4.20 [1.58, 11.14] |
| 11.5.1 Chemo with apatinib | 1 | 146 | Risk Ratio (M‐H, Random, 95% CI) | 12.65 [0.73, 220.58] |
| 11.5.2 Chemo with nintedanib | 1 | 114 | Risk Ratio (M‐H, Random, 95% CI) | 1.24 [0.29, 5.30] |
| 11.5.3 Chemo with pazopanib | 5 | 518 | Risk Ratio (M‐H, Random, 95% CI) | 7.64 [3.17, 18.41] |
| 11.5.4 Chemo with sorafenib | 1 | 172 | Risk Ratio (M‐H, Random, 95% CI) | 1.07 [0.22, 5.16] |
| 11.5.5 Chemo with vandetanib | 1 | 0 | Risk Ratio (M‐H, Random, 95% CI) | Not estimable |
| 11.6 Proteinuria (grade ≥ 2) Show forest plot | 3 | 387 | Risk Ratio (M‐H, Random, 95% CI) | 4.00 [0.49, 32.86] |
| 11.6.1 Chemo with apatinib | 1 | 146 | Risk Ratio (M‐H, Random, 95% CI) | 4.87 [0.24, 99.65] |
| 11.6.2 Chemo with pazopanib | 1 | 116 | Risk Ratio (M‐H, Random, 95% CI) | 3.32 [0.18, 62.76] |
| 11.6.3 Chemo with vandetanib | 1 | 0 | Risk Ratio (M‐H, Random, 95% CI) | Not estimable |
| 11.7 Pain (grade ≥ 2) Show forest plot | 3 | 361 | Risk Ratio (IV, Random, 95% CI) | 0.97 [0.44, 2.15] |
| 11.7.1 Chemo with pazopanib | 2 | 189 | Risk Ratio (IV, Random, 95% CI) | 0.98 [0.20, 4.88] |
| 11.7.2 Chemo with sorafenib | 1 | 172 | Risk Ratio (IV, Random, 95% CI) | 1.07 [0.45, 2.57] |
| 11.8 Abdominal pain (grade ≥ 2) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 11.8.1 Chemo with pazopanib | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 11.9 Neutropenia (grade ≥ 3) Show forest plot | 9 | 1069 | Risk Ratio (IV, Random, 95% CI) | 1.73 [1.15, 2.61] |
| 11.9.1 Chemo with apatinib | 1 | 146 | Risk Ratio (IV, Random, 95% CI) | 1.78 [0.70, 4.57] |
| 11.9.2 Chemo with nintedanib | 1 | 114 | Risk Ratio (IV, Random, 95% CI) | 14.00 [0.82, 239.49] |
| 11.9.3 Chemo with pazopanib | 5 | 512 | Risk Ratio (IV, Random, 95% CI) | 2.35 [1.42, 3.90] |
| 11.9.4 Chemo with sorafenib | 1 | 172 | Risk Ratio (IV, Random, 95% CI) | 1.03 [0.78, 1.35] |
| 11.9.5 Chemo with vandetanib | 1 | 125 | Risk Ratio (IV, Random, 95% CI) | 0.92 [0.64, 1.32] |
| 11.10 Febrile neutropenia (any grade) Show forest plot | 6 | 748 | Risk Ratio (M‐H, Random, 95% CI) | 1.49 [0.68, 3.30] |
| 11.10.1 Chemo with nintedanib | 1 | 114 | Risk Ratio (M‐H, Random, 95% CI) | 2.80 [0.12, 67.32] |
| 11.10.2 Chemo with pazopanib | 3 | 337 | Risk Ratio (M‐H, Random, 95% CI) | 1.44 [0.41, 5.06] |
| 11.10.3 Chemo with sorafenib | 1 | 172 | Risk Ratio (M‐H, Random, 95% CI) | 1.29 [0.41, 4.06] |
| 11.10.4 Chemo with vandetanib | 1 | 125 | Risk Ratio (M‐H, Random, 95% CI) | 3.15 [0.13, 75.76] |
| 11.11 Non‐central nervous system bleeding (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 11.11.1 Chemo with sorafenib | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 11.12 Gastrointestinal adverse events (grade ≥ 3) Show forest plot | 3 | 386 | Risk Ratio (IV, Random, 95% CI) | 1.08 [0.46, 2.53] |
| 11.12.1 Chemo with nintedanib | 1 | 114 | Risk Ratio (IV, Random, 95% CI) | 0.93 [0.06, 14.54] |
| 11.12.2 Chemo with pazopanib | 1 | 100 | Risk Ratio (IV, Random, 95% CI) | 2.00 [0.88, 4.53] |
| 11.12.3 Chemo with sorafenib | 1 | 172 | Risk Ratio (IV, Random, 95% CI) | 0.69 [0.43, 1.11] |
| 11.13 Bowel fistula or perforation (grade ≥ 3) Show forest plot | 5 | 557 | Risk Ratio (M‐H, Random, 95% CI) | 2.74 [0.77, 9.75] |
| 11.13.1 Chemo with nintedanib | 1 | 114 | Risk Ratio (M‐H, Random, 95% CI) | 2.80 [0.12, 67.32] |
| 11.13.2 Chemo with pazopanib | 4 | 443 | Risk Ratio (M‐H, Random, 95% CI) | 2.73 [0.68, 10.90] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 12.1 Overall survival Show forest plot | 1 | 123 | Hazard Ratio (IV, Fixed, 95% CI) | 1.10 [0.71, 1.71] |
| 12.2 Progression‐free survival Show forest plot | 1 | 123 | Hazard Ratio (IV, Fixed, 95% CI) | 1.04 [0.70, 1.56] |
| 12.3 Proteinuria (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 12.4 Pain (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 12.5 Abdominal pain (grade ≥ 3) Show forest plot | 1 | 123 | Risk Ratio (IV, Fixed, 95% CI) | 0.25 [0.05, 1.11] |
| 12.6 Neutropenia (grade ≥ 3) Show forest plot | 1 | 123 | Risk Ratio (IV, Fixed, 95% CI) | 1.57 [0.55, 4.54] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 13.1 Overall survival Show forest plot | 3 | 1250 | Hazard Ratio (IV, Fixed, 95% CI) | 0.92 [0.80, 1.06] |
| 13.2 Progression‐free survival Show forest plot | 3 | 1250 | Hazard Ratio (IV, Fixed, 95% CI) | 0.73 [0.65, 0.82] |
| 13.3 Quality of life ‐ Functional Assessment of Cancer Therapy—Ovarian Cancer questionnaire Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 13.4 Hypertension (grade ≥ 3) Show forest plot | 3 | 1242 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.92 [0.70, 12.18] |
| 13.5 Proteinuria (grade ≥ 3) Show forest plot | 2 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 13.6 Pain (grade ≥ 3) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 13.7 Abdominal pain (grade ≥ 3) Show forest plot | 3 | 1242 | Risk Ratio (IV, Fixed, 95% CI) | 0.99 [0.60, 1.65] |
| 13.8 Neutropenia (grade ≥ 3) Show forest plot | 2 | 1134 | Risk Ratio (IV, Fixed, 95% CI) | 0.60 [0.40, 0.89] |
| 13.9 Febrile neutropenia (any grade) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 13.10 Venous thromboembolic event (any grade) Show forest plot | 2 | 1021 | Risk Ratio (IV, Fixed, 95% CI) | 0.68 [0.25, 1.85] |
| 13.11 Arterial thromboembolic event (any grade) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 13.12 Non‐central nervous system bleeding (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
| 13.13 Gastrointestinal perforation (grade ≥ 3) Show forest plot | 1 | Risk Ratio (IV, Fixed, 95% CI) | Totals not selected | |
