Synergism between anti-angiogenic and immune checkpoint inhibitor drugs: A hypothesis
Introduction
The combination of bevacizumab (an anti-angiogenic drug) with atezolizumab (an immune checkpoint inhibitor) scored major success in two trials, IMbrave150 and IMmotion151, against hepatocellular (HCC) [1] and renal cell (RCC) cancers [2] respectively. These successes are particularly noteworthy as liver and kidney cancers are particularly resistant to conventional chemotherapy. Classical cytotoxic drugs work by interfering with the machinery in the nucleus that controls cell division. As organs specialized for waste disposal, the liver and kidney are superbly evolved to deal with toxins, including chemotherapy drugs. Cancer cells derived from them inherit similar properties. These drugs are thus eliminated efficiently before they can cause major damage. Oncologists have acknowledged the impenetrable nature of HCC and RCC to conventional chemotherapy after multiple failed trials. Their focus then shifted to targeting the cells in the surrounding stroma (such as vascular and immune cells) – and have achieved limited success. Anti-angiogenic (AA) drugs that cut off blood supply improve survival by a few months. These are either monoclonal antibodies such as bevacizumab, aflibercept and ramucirumab, or a multitude of small molecule tyrosine kinase inhibitors (TKIs) such as sorafenib, sunitinib, lenvatinib, axitinib, pazopanib and cabozantinib. (It should be noted that most TKIs are not pure AA drugs but are multikinases, targeting other receptors as well, such as RAF and PDGFR). Further success followed by use of immunotherapy drugs that stimulate host immunity against cancer cells, such as immune checkpoint inhibitors (ICIs) like pembrolizumab, nivolumab and atezolizumab. These work indirectly by activating the host cytotoxic T lymphocytes against the cancer cells and are thus pro-drugs. Recent trials suggest that the combination of an AA drug (bevacizumab) with an ICI (atezolizumab) is much more than the sum of its parts (Table 1). Interestingly, the combination also appears to work well in another hard-to-treat cancer – liver metastases (such as from lung cancer [15]).
There are AA TKI plus ICI combinations approved for treatment in RCC [16], but as mentioned earlier, TKIs have broad-spectrum activities and I shall mainly concentrate on bevacizumab as the prototype AA drug.
The rationale behind the combination, as pointed out in the trial publications [1], [2], is that the bevacizumab reverses VEGF-mediated immunosuppression, and by normalizing tumour vasculature and improving perfusion, can convert the immunosuppressive tumour micro-environment (TME) to an immune stimulatory one. Three key mechanisms related to VEGF–mediated immunosuppression are inhibition of dendritic cell maturation, reduction of T-cell tumor infiltration, and promotion of inhibitory cells in the TME [17]. Bevacizumab, by virtue of being an anti-VEGF antibody, is believed to reverse these mechanisms.
The conventional explanation of the mechanism of action of bevacizumab is that it shuts off blood supply to the tumours and causes cell death by hypoxia. However, alternative theories have been proposed. Jain [18] hypothesised that under specific circumstances, bevacizumab can do the opposite, that is, “normalise” tumour vasculature, improve perfusion and reduce interstitial pressure, resulting in better drug delivery to the tumour. Normalizing circulation converts the immune suppressive TME into an immune stimulatory one (see below).
Incidentally, it is known that bevacizumab has limited single agent activity and its efficacy improves when given in combination with chemotherapy drugs. Blagosklonny [19] has suggested that the chemotherapy drugs have anti-angiogenic properties, that resistance to these drugs occurs by elevation of VEGF, and bevacizumab works by blocking this increase. Bocci [20], however, accepts the conventional view point but suggests that accompanying chemotherapy drugs may actually contribute to anti-angiogenesis resistance.
Jain [18] acknowledges that normalizing vasculature is both dose and schedule dependant, and in other situations, bevacizumab leads to pruning of vasculature and tumour hypoxia, causing cell death. When the tumour vasculature is excessively pruned, the resultant hypoxia can aggravate an already immunosuppressive TME. This is characterized by increase in immunosuppressive regulatory T cells (Tregs), M2 tumour associated macrophages (TAM), and myeloid-derived suppressor cells (MDSCs), and reduced anti-tumour cytotoxic T lymphocytes (CTLs), M1 TAMs and mature dendritic cells (DCs). Hypoxia can also upregulate multiple immune-suppressive growth factors and cytokines (e.g., VEGF and TGF-β). Intratumoral hypoxia also compromises antigen presenting cell (APC) functionality and the stimulation of T cell responses. (Summarized in [21]). Despite the attractiveness of Jain’s hypothesis, augmented tumour hypoxia and immune suppression appears to be the more likely outcome of bevacizumab activity. Hypothesis: It is hypothesized that bevacizumab increases the efficacy of ICIs by
1. Increasing immunogenicity of the tumours by aggravating hypoxia; this causes stress-induced mutagenesis and increased mutation load.
2. Starving rapidly proliferative cancer cells during the hyperprogression precipitated by ICIs.
Generation of hypermutator profile by stress-induced mutagenesis (SIM) caused by hypoxia:
ICIs are more effective against “hot” immunogenic tumours, ie, tumours displaying an excess of “neoantigens”. What defines an immunogenic tumour is still a matter of investigation, but these are generally identified by predictive biomarkers such as percentage of cells positive for PD-L1 (or its variants like tumour proportion score (TPS) or combined positive score (CPS)), by microsatellite instability (MSI), or by the Tumour Mutational Burden (TMB). TMB [22] (measured as mutations per megabase (mutations/Mb) by next generation sequencing (NGS) or by actual number of mutations by whole exome sequencing (WES)) is a biomarker that indicates the number of tumor-specific somatic mutations in the tumor’s exome. High TMB tumors are “hot” tumours with more surface neoantigens, producing an enhanced immune response [23]. The KEYNOTE-158 trial showed that pembrolizumab, an ICI, is effective against tumours with high TMB irrespective of tissue of origin [24], and has been FDA-approved in this indication.
It is possible that bevacizumab can convert “cold” cancers into an immunogenic (“hot”) cancer. ICIs work by blocking the PD-1/PD-L1 interaction, and increasing PD-L1 level is one possibility. This has been shown in ovarian cancer, where bevacizumab-responsive cisplatin resistant ovarian cancer had higher levels of PD-L1 than the non-responsive tumour (58.7% vs. 39.3%) [25]. Similarly, sorafenib (an AA TKI) induced hypoxia with activation of the SDF1-α/CXCR4 pathway, and increased expression of PD-L1 in the cancer cells [26]. In addition, hypoxia has been shown to directly up-regulate – via HIF1α activation – the expression of PD-L1 by MDSCs, DCs and cancer cells to aid immune-suppression and evasion [27].
However, PDL1 is not used as a predictor for immunotherapy responses in HCC and RCC. The other predictor, TMB, is relatively low in RCC and does not seem to be predictive of survival, as shown in the MSKCC IMPACT cohort [28]. Data of HCC was not reported in this cohort, but in another study of 755 patients with advanced HCC [29], median TMB was low at 4 mutations/Mb. Only 6 tumors (0.8%) were TMB-high (conventionally accepted as >10 mutations/Mb).
Both HCC and RCC are highly vascular tumours. HCC’s vascularity relates to upregulation of HIF1a and multiple angiogenic factors such as VEGF, BMP4, PAI-1 and SCF [30]. The predominant clear cell subtype of RCC (ccRCC) arises from the inactivation of vHL gene with upregulation of HIF-1a; the resultant “pseudo-hypoxic” signal causes increased vascularity [31].
Paradoxically, despite their vascularity, HCC and RCC are highly hypoxic. Most mammalian tissues are maintained at 2–9% oxygen (O2) (on average 5.5%, 40 mm Hg). Tissue hypoxia is defined as reduced oxygen partial pressure ≤ 2% (≤15 mm Hg) [32]. The critical pO2 in tumours is proposed to be 8–10 mmHg [33]. Below this level, detrimental changes associated with reduced oxygen consumption have been observed (ATP depletion, intracellular acidosis, glycolysis and apoptosis). HCC is one of the most hypoxic tumors with median oxygen level as low as 0.8% [34]. Again, although ccRCC is a vascular tumour, mean pO2 value was 9.6 mmHg [35].
Despite Jain’s intriguing concept [18], there are several lines of evidence that AA drugs do, as predicted, inhibit angiogenesis and induce hypoxia. For instance, sustained sorafenib treatment has been shown to reduce microvessel density and promote intratumoral hypoxia; the resultant HIF-mediated cellular responses favor the selection of resistant cells adapted to the hypoxic microenvironment and leads to sorafenib resistance [36].
Similarly, multiple studies have documented that bevacizumab primarily induce hypoxic stress. Treatment of mice bearing colon cancer xenografts with bevacizumab resulted in depletion of tumor microvasculature, upregulation of HIF-1α, and increased pimonidazole staining, consistent with an anti-angiogenic effect and induction of hypoxia [37]. Another study with colon cancer xenografts showed that intratumoral hypoxia induced by bevacizumab treatment resulted in activation of HIF-1α protein and upregulated stanniocalcin 2 (STC2) [38]. AA treatment (sunitinib and bevacizumab) of breast cancer xenografts induce hypoxic conditions and increase in cancer stem cells, primarily mediated by HIF-1α (EMT induction) [39]. Other studies have consistently demonstrated reductions in microvessel density in HCC [40] and RCC [41].
Aggravation of hypoxia alters expression of multiple genes, both random and specific (the latter for the upregulation of adaptive resistance pathways) and also results in increased tumor cell invasiveness and selection of more invasive metastatic clones [42], [43], [44]. The increased aggressiveness post AA therapy is a possible explanation for the lack overall survival (OS) benefit despite the increased progression free survival (PFS) seen with AA agents [43]. For instance, two randomized clinical trials with bevacizumab and taxanes for the treatment of metastatic breast cancer (the E2100 study [45] and the AVADO [46] study) demonstrated small benefits in PFS but not an OS benefit. Similar induction of aggressive behavior has been noted in glioblastoma [47] and colorectal cancer [48].
Uedo et al. [49] demonstrated that in metastatic breast cancer treated with a combination of paclitaxel and bevacizumab, non-responders exhibited more severe hypoxia than responders (despite higher angiogenesis). The authors confirm that addition of bevacizumab results in acute hypoxia, and increased cytokine secretion associated with cancer progression, but the exact genetic mechanism was not studied.
The paradoxical effects of enhanced tumour aggressiveness and induction of an invasive phenotype with the use of bevacizumab have been reviewed [50]; the basis appears to be upregulation of resistance mechanisms by additional mutations; specific causes cited include increased epithelial-to-mesenchymal transition, and activation of MET receptor. The significance of these studies is that cancer cells do not “roll over and die” when stressed with hypoxia, but survive by acquiring additional mutations. The additional mutations contribute to increased aggressiveness.
As Uedo et al. [49] notes, the behavior of bevacizumab is a double-edged sword; it induces reoxygenation if there is proper vascular remodeling, otherwise it can lead to severe hypoxia because of destruction of the tumor vasculature.
In summary, it would appear that use of AA drugs leads to increased hypoxia, and this hypoxic challenge can result in additional tumour cell mutations and evolution of an aggressive phenotype.
Not all mutations are random. When bacteria and yeast are exposed to stress (such as by hypoxia or nutrient deprivation), they respond by actively inducing survival mutations. This is by a process called adaptive or stress-induced mutagenesis (SIM). It is now known that cancer cells when stressed, like bacteria, can also undergo SIM. Hypoxic stress activates adaptive mutagenesis, partly by downregulation of mismatch repair proteins; it induces genomic instability and a hypermutator profile (reviewed by Fitzgerald et al. [51]); optimum conditions for checkpoint inhibitors to be effective.
It is hypothesized that hypoxic stress caused by AA therapy leads to SIM and creates a hypermutator profile, and the induction of neoantigens and a higher mutation burden is a major factor in the improved efficacy of ICIs in combination with AA drugs.
There is limited data on mutational changes after exposure to AA therapy. Metastatic colorectal cancer monitored by liquid biopsy while on treatment with bevacizumab (and chemotherapy), developed mutations of CREBBP and FBXW7; these genes are associated with tolerance to hypoxia [52]. A report from the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium quantified hypoxia in 1188 tumours spanning 27 cancer types. It showed that elevated hypoxia is associated with increased mutational load across cancer types, irrespective of underlying mutational class [53].
A recent study [54] on cell lines showed that hypoxic stress resulted in transcriptional downregulation of DNA repair genes. Hypoxic conditions also increased the mutational burden which was characterized by an increase in frameshift insertions and deletions, and increase in the formation of potential neoantigens.
The downregulation of DNA repair genes on exposure to hypoxia have been reported earlier [55], [56]; this helps in the propagation of the new mutations. Hypoxic cells can thus acquire a mutator phenotype that consists of decreased DNA repair, increased mutation rate, and increased chromosomal instability [57].
Conforming to the same principle, gliomas treated with temozolomide (an alkylating agent) develop hypermutator profile [58]; similar changes induced in a case of neuroendocrine carcinoma was exploited to treat the patient successfully with an ICI, pembrolizumab [59].
In summary, it is hypothesized that use of bevacizumab increases hypoxic stress in tumours, ushers in stress-induced mutagenesis, which increases the mutational burden of the tumour, and converts a “cold” non-immunogenic tumour into a “hot” tumour, thus increasing the efficacy of ICIs.
Section snippets
Starving hyperprogressing tumours
The “metabolic war” against cancer has been on for several decades with the basic principle of “attacking the supply wagons to starve cancer cells to death” [60].
Cancer cells have unique metabolic requirements and differ from normal cells by genetic and epigenetic programming for adaptation and survival. These differences present tempting targets for cancer treatment. There has been a long standing belief that the addiction of cancer cells to certain metabolites may prove to be its Achilles
Consequences of the hypothesis
The advantages of combining AA with ICIs have been extensively reviewed [81], [82], [83] and combinations are being expanded to cover other tumour types [84]. The efficacy of ICIs depends on two pre-conditions – the availability of cytotoxic immune cells that can be activated against the cancer cells, and the presence of targetable antigens on the latter. While ICIs activate the cytotoxic cells, it is hypothesized that bevacizumab increases tumour antigen expression by aggravating tumor
Funding
Nil.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (84)
- et al.
Atezolizumab plus bevacizumab versus sunitinib in patients with previously untreated metastatic renal cell carcinoma (IMmotion151): a multicentre, open-label, phase 3, randomised controlled trial
Lancet
(2019) - et al.
Atezolizumab plus bevacizumab and chemotherapy in non-small-cell lung cancer (IMpower150): key subgroup analyses of patients with EGFR mutations or baseline liver metastases in a randomised, open-label phase 3 trial
Lancet Respir Med
(2019) Antiangiogenesis strategies revisited: from starving tumors to alleviating hypoxia
Cancer Cell
(2014)- et al.
Atezolizumab and bevacizumab attenuate cisplatin resistant ovarian cancer cells progression synergistically via suppressing epithelial-mesenchymal transition
Front Immunol
(2019) - et al.
Renal cell carcinoma for the nephrologist
Kidney Int
(2018) - et al.
Antitumor effect of the angiogenesis inhibitor bevacizumab is dependent on susceptibility of tumors to hypoxia-induced apoptosis
Biochem Pharmacol
(2008) - et al.
Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis
Cancer Cell
(2009) - et al.
Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis
Cancer Cell
(2009) - et al.
Co-repression of mismatch repair gene expression by hypoxia in cancer cells: role of the Myc/Max network
Cancer Lett
(2007) - et al.
IMbrave150 Investigators. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma
N Engl J Med
(2020)
Phase II trial evaluating the clinical and biologic effects of bevacizumab in unresectable hepatocellular carcinoma
J Clin Oncol
Pembrolizumab as second-line therapy in patients with advanced hepatocellular carcinoma in KEYNOTE-240: a randomized, double-blind, phase III trial
J Clin Oncol
Safety and clinical activity of durvalumab monotherapy in patients with hepatocellular carcinoma (HCC)
J Clin Oncol
Randomized phase II study of erlotinib combined with bevacizumab compared with bevacizumab alone in metastatic renal cell cancer
J Clin Oncol
Nivolumab for metastatic renal cell carcinoma: results of a randomized phase ii trial
J Clin Oncol
Nivolumab plus ipilimumab versus sunitinib in advanced renal-cell carcinoma
N Engl J Med
First-line pembrolizumab (pembro) monotherapy in advanced clear cell renal cell carcinoma (ccRCC): updated results for KEYNOTE-427 cohort A
J Clin Oncol
Avelumab monotherapy as first-line or second-line treatment in patients with metastatic renal cell carcinoma: phase Ib results from the JAVELIN Solid Tumor trial
J Immunother Cancer
Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma
N Engl J Med
Avelumab plus axitinib versus sunitinib for advanced renal-cell carcinoma
N Engl J Med
Combinations of bevacizumab with cancer immunotherapy
Cancer J
How Avastin potentiates chemotherapeutic drugs: action and reaction in antiangiogenic therapy
Cancer Biol Ther.
Hypoxia-inducible factors in regulation of immune responses in tumour microenvironment
Immunology
Tumor mutational burden as a predictive biomarker for response to immune checkpoint inhibitors: a review of current evidence
Oncologist
Genomics of response to immune checkpoint therapies for cancer: implications for precision medicine
Genome Med
Efficacy of pembrolizumab in patients with noncolorectal high microsatellite instability/mismatch repair-deficient cancer: results from the phase ii Keynote-158 study
J Clin Oncol
CXCR4 inhibition in tumor microenvironment facilitates anti-PD1 immunotherapy in sorafenib-treated hepatocellular carcinoma in mice
Hepatology
PD-L1 is a novel direct target of HIF-1α, and its blockade under hypoxia enhanced MDSC mediated T cell activation
J Exp Med
Tumor mutational load predicts survival after immunotherapy across multiple cancer types
Nat Genet
Prevalence of established and emerging biomarkers of immune checkpoint inhibitor response in advanced hepatocellular carcinoma
Oncotarget
Targeting the vasculature in hepatocellular carcinoma treatment: starving versus normalizing blood supply
Clin Transl Gastroenterol
Why is the partial oxygen pressure of human tissues a crucial parameter? Small molecules and hypoxia
J Cell Mol Med
Tumor hypoxia: definitions and current clinical, biologic, and molecular aspects
J Natl Cancer Inst
Hypoxia inducible factors in hepatocellular carcinoma
Oncotarget
Assessing regional hypoxia in human renal tumours using 18F-fluoromisonidazole positron emission tomography
BJU Int
Sorafenib resistance in hepatocarcinoma: role of hypoxia-inducible factors
Exp Mol Med
Anti-VEGF antibody therapy induces tumor hypoxia and stanniocalcin 2 expression and potentiates growth of human colon cancer xenografts
Int J Cancer
Antiangiogenic agents increase breast cancer stem cells via the generation of tumor hypoxia
Proc Natl Acad Sci USA
Cited by (4)
Tumor Mutational Burden for Predicting Prognosis and Therapy Outcome of Hepatocellular Carcinoma
2023, International Journal of Molecular SciencesHepatocellular Carcinoma: Pick the Winner - Tyrosine Kinase Inhibitor Versus Immuno-oncology Agent-Based Combinations
2022, Journal of Clinical OncologyPolypharmacy in the elderly on immunotherapy: Problem or opportunity?
2021, Cancer Research, Statistics, and Treatment