Sequencing of chemotherapy and radiation therapy for early breast cancer
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To determine the effects of the sequencing of radiation therapy and chemotherapy for surgically treated early breast cancer.
Background
After conservative surgery or mastectomy for localised breast cancer, adjuvant radiotherapy improves local control and has an impact on breast cancer specific survival (Ragaz 2005). In patients at risk of harbouring micrometastatic disease, adjuvant chemotherapy improves 10 year survival by 7‐11% in women aged <50 and by 2‐3% in women aged 50‐69 years ( Lancet 1998).
It is not clear what the optimum sequence of administration of adjuvant chemotherapy and radiotherapy for early stage breast cancer. Current practices include: administering chemotherapy prior to radiation therapy, administering CT synchronously with RT, or "sandwiching" RT in the middle of the chemotherapy course. The sequence of administration of therapeutic modalities may affect outcome ( Recht 1996), but this is not clear. A delay in initiating radiation therapy has been shown to increase the risk of local recurrence, OR 2.28 (95% CI 1.45‐3.57) ( Huang 2003). However, delay in commencing CT may have a detrimental effect on survival. A published randomised trial, demonstrated a non‐significant improvement in overall survival with the initial administration of chemotherapy ( Recht 1996). However, with longer follow‐up the same trial showed no difference in the rates of local or distal recurrence or death between the two arms ( Bellon 2001).
In many parts of the world, waiting lists for radiation therapy are a fact of life( Ash 2000; Kenny 2004; Mackillop 1994; Mackillop 1995). Delivery of CT before RT allows patients to continue their anti cancer therapy while they waiting for RT to start and may make waiting list management easier for RT staff (Kenny 2004).
If a systematic review confirms that sequencing of CT and RT is unimportant in terms of cancer related outcomes such as survival and local recurrence, then choosing to give the CT first may be preferable for both logistic reasons and women's preference. If a systematic review does not confirm this approach, and indicates that RT should be given first, it will be important evidence to support expansion of RT facilities and resources to allow equity of access for patients to RT.
Determining appropriate sequencing for adjuvant CT and RT is therefore an ideal question for a systematic review.
Objectives
To determine the effects of the sequencing of radiation therapy and chemotherapy for surgically treated early breast cancer.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials evaluating different ways of sequencing chemotherapy and radiation therapy. Trials incorporating the use of monoclonal antibodies or hormonal manipulation are eligible, if equally applied in both arms. Both published and unpublished studies to be included. Each trial is to be un‐confounded (i.e. differ only by the sequencing of the two adjuvant treatment modalities).
Types of participants
Inclusion criteria:
Women with surgically treated,histologically confirmed early breast cancer (UICC stage T1‐3N0‐1M0) who require both adjuvant chemotherapy and radiotherapy. Surgery may include mastectomy, lumpectomy, wide local excision or quadrantectomy, plus or minus axillary dissection, axillary sampling or sentinal node biopsy.
Exclusion criteria:
Patients with prior therapy for breast cancer will be excluded.
Types of interventions
These may include trials where;
1) Adjuvant radiation therapy, followed by adjuvant chemotherapy, is compared to adjuvant chemotherapy followed by adjuvant radiation therapy
2) trials where adjuvant radiation therapy, followed by adjuvant chemotherapy is compared with a "sandwich technique"( where initial chemotherapy is followed by radiation therapy, then further chemotherapy)
3) trials where adjuvant radiation therapy, followed by adjuvant chemotherapy is compared to concurrent adjuvant chemotherapy and radiation therapy.
Chemotherapy regimens will include those delivered at standard doses (i.e. not high dose), and may include cyclophosphamide, 5 Fluorouracil, anthracyclines, taxanes or other agents.
Radiation therapy delivered to breast or chest wall plus or minus supra‐clavicular fossa and axilla in standard fractionation (1.8‐2.7 Gy per fraction), to 40‐61 Gy at Reference point, and may include a boost (using electrons, interstitial therapy or external beam) or new techniques.
Types of outcome measures
Primary endpoints: local control (ipsilateral), cause‐specific survival
Secondary endpoints:overall survival, distant metastases (in isolation or with concurrent local failure), relapse‐free survival, mastectomy rate, harms including acute and late effects of radiation therapy, chemotherapy related toxicity, ability to deliver prescribed chemotherapy dose, costs, quality of life and consumer preference.
Search methods for identification of studies
The specialised register maintained by the Secretariat of the Cochrane Breast Cancer Group will be searched. Details of the search strategy applied by the group to create the register, and the procedure used to code references are described in the Group's module on the Cochrane Library. Studies coded as "early' and "chemotherapy" and "radiotherapy" on the specialised register will be extracted for consideration.
In addition a comprehensive search of the following electronic databases will be conducted: Current Contents, Science Citation Index, CINAHL, Psych lit. All databases will be searched from the beginning of electronic records.
The following sources will also be searched: registers of controlled trials in progress, registers of unpublished studies ( SIGLE, National Research Register and HSRProj), conference proceedings in related fields including the Proceedings of American Society of Therapeutic Radiation Oncologists, cited references in retrieved papers, contact with key authors in the field, relevant professional associations and colleges (oncology, radiology, radiotherapy), publications of Health Technology Assessment Agencies.
No language restrictions will be employed.
Data collection and analysis
Study selection: The title and abstracts identified from the databases will be checked by all 3 reviewers. Three reviewers (BH, TD and ML) will independently assess the full text of all studies of possible relevance and discussion will be used to resolve any discrepancies. Trials will be assessed with the results section (and any other section where results appear) masked. Where possible additional information will be sought from the principal investigator of the trial concerned where information is lacking.
Data extraction will be performed by two reviewers (BH and ML), and any discrepancies resolved by discussion. For unpublished trials or published trials with unpublished data, the trialists will be contacted to obtain further information. Data will be checked and entered into RevMan 4.2.
Data extracted (where possible) will include tumour stage, nodal status, margin status, receptor status, hormonal manipulation,treatment allocation and surgery performed. Outcome data will be extracted for local control, distant metastases,survival (cause specific and overall), time from randomisation to commencement of radiation therapy, and chemotherapy, distant metastases, treatment‐related toxicity (including that related to acute and late effects of radiation therapy and chemotherapy), duration of chemotherapy, duration of radiation therapy, radiation dose and dose per fraction, costs of treatment, consumer preference and quality of life.
Time to event analysis will be used (if possible) for time to local recurrence and time to death (Parmar 1998 ). Dichotomous results will be presented as both relative risk (RR) and odds ratio (OR), using 95% confidence intervals (95% CI) (Deeks 2003). Mantel‐Haenszel methods will be used to calculate (Greenland 1985, Mantel 1959). For continuous variables, e.g. quality of life, a judgement about combination of data will be made if and how the data is collected. A mean difference method will be used, unless different scales are employed in different trials, in which case a standard mean difference will be used to summarize data (Deeks 2003). Heterogeneity will be tested for (both visually and statistically, using the Chi‐squared test of heterogeneity)( Altman 1992, Walker 1988). If heterogeneity is found, the reasons for it will be explored, and an attempt made to explain it, recognising the need to be cautious in doing so.
Analysis will be by intention to treat. A weighted treatment effect will be determined, using the fixed effects model to combine results (Mantel 1959), using the Cochrane statistical package RevMan 4.2, if meta‐analysis is possible and appropriate.
Sub‐group analysis may be used to investigate the possible effects of nodal status, margin status, receptor status, hormonal manipulation and tumour stage, if so, this will be done in collaboration with a statistician.
