Questions Over Value of Bevacizumab in Many Cancers

Alexander M. Castellino, PhD

November 20, 2019

The antiangiogenic agent bevacizumab (Avastin, Roche/Genentech) is approved for use in six different types of tumor, but most of these approvals are based on improvements in progression-free survival (PFS), not overall survival (OS).

Bevacizumab has not provided OS benefit in metastatic pancreatic cancer, metastatic renal cancer, metastatic gastric cancer, glioblastoma, non–small cell lung cancer, and in five of seven colorectal cancer trials. (The two exceptions are metastatic cervical cancer and unresectable pleural mesothelioma, for which it has shown an OS benefit).

"On the basis of the OS gold standard, the angiogenesis-based VEGF [vascular endothelial growth factor] hypothesis is on life support," comment a pair of Australian oncologists.

Yet the product continues to be widely used. Since its initial approval in 2004, it has generated sales of more than $50 billion in drug costs alone, which represents a financial burden on healthcare systems worldwide, they point out.

The comments are made by Ian Haines, MBBS, of Monash University, Malvern, Australia, and George L. Gabor Miklos, PhD, Atomic Oncology, Sydney, Australia, in a letter to the editor published online November 7 in the Journal of Clinical Oncology.

The letter has prompted by a discussion of the latest clinical data on bevacizumab in ovarian cancer. Those data come from the GOG-0218 study, which showed that the addition of bevacizumab made no difference to median survival times for women with incompletely resected stage III or IV ovarian cancer treated with chemotherapy.

"These OS data [from GOG-0218] are congruent with three earlier phase III trials for this indication: ICON7, AURELIA, and OCEANS," they note.

"The high-probability conclusion is that bevacizumab-based targeting of vascular endothelial growth factor (VEGF) in ovarian cancer fails to increase median lifespan," Haines and Miklos continue.

Despite this lack of OS data, ovarian cancer experts say that bevacizumab is of value in this tumor type, and they are in agreement that bevacizumab should be incorporated into the first-line treatment of ovarian cancer.

In a published response to the Australian authors, Krishnansu S. Tewari, MD (lead author of GOG-0218), and colleagues argue that in ovarian cancer specifically, OS is an unrealistic endpoint, because post-progression survival is long, owing to the fact that the cancer remains chemosensitive. This, among other reasons, has been at the forefront of the US Food and Drug Administration's (FDA's) recognizing the value of PFS in ovarian cancer, and the FDA has approved use of the drug on this basis, they write.

Tewari and colleagues also note that nine phase 3 trials of newly diagnosed platinum-sensitive and platinum-resistant ovarian cancer that evaluated five different antiangiogenic agents have all met their primary endpoint of PFS.

Ovarian cancer expert Robert L. Coleman, MD, professor of gynecologic oncology and reproductive medicine at MD Anderson Cancer Center, Houston, Texas, agreed.

He noted that in the ovarian cancer trials cited in the correspondence, PFS (and not OS) was the primary endpoint of the studies. "In each of the ovarian cancer trials cited, OS was not an endpoint powered to do formal hypothesis-testing. Thus, the P values are meaningless because they have not been adjusted appropriately for confounders nor for multiple comparisons," Coleman told Medscape Medical News.

In addition, he pointed out that in order for GOG-0218 to have been powered for survival, it would have had to accrue more than 6000 patients. The PFS assessment was based on treating 1873 women. "The issue here relates to a fundamental lack of understanding of how statistical power operates relative to post-progression survival," Coleman said.

He explained that the FDA has been moving away from requiring blinded, independent radiology committee review [BICR] to determine PFS. "While some gain comfort in having BICR as a primary read for PFS, it actually ignores important patient-specific disease-related events that would be unknown to the radiologist reading the films and thus leads to censoring when a physician treating a patient declares progression based on clinical events but the radiologist declares no progression," Coleman told Medscape Medical News.

He further indicated that PFS is the most proximal event to the intervention and one that is not clouded by subsequent treatment interventions. "Since none of the cited trials actually prescribes what is and what is not allowed once a patient comes off trial, the impact that uncontrolled treatment interventions might have is impossible to discern," he said.

Support Not Overwhelming in Other Solid Tumors

In contrast to the support for bevacizumab in ovarian cancer, there is little support among experts who have used the drugs to treat several other cancer types.

Hedy L. Kindler, MD, of the University of Chicago, who led the CALGB 80303 study in pancreatic cancer, told Medscape Medical News that there is no role for bevacizumab in metastatic pancreatic cancer. "The trial was flat-out negative, the curves were superimposable," she said. Bevacizumab has not been approved for this indication.

George W. Sledge Jr, MD, told Medscape Medical News that he was heavily involved in the development of bevacizumab in breast cancer — his area of expertise.

"In the metastatic setting, multiple studies showed a PFS improvement, but no OS improvement," he said. He pointed out that the FDA initially approved bevacizumab for metastatic breast cancer, but then withdrew the approval because the risk-benefit was skewed toward risk, he said. However, the European authorities, unlike the FDA, never withdrew approval for bevacizumab for metastatic breast cancer.

Sledge commented that, across all the trials with bevacizumab, even when there has been evidence of a survival benefit, there has been no evidence of benefit in the adjuvant setting. "This suggests to me and others that antiangiogenic therapy just may not work without a tumor vasculature to inhibit, as probably occurs in small, early tumors," he said. "Antiangiogenic therapy was supposed to be the treatment 'resistant to resistance.' In fact, cancers are quite good at generating resistance to bevacizumab," he added.

Colorectal cancer expert Alan P. Venook, MD, the Madden Family Distinguished Professor of Medical Oncology and Translational Research at the University of San Francisco, in California, commented that whether or not to use bevacizumab in the absence of an OS benefit is a difficult issue. "It is not only a question of OS but also whether a patient shows a response or not," he said. "It is crazy to look only at OS," he told Medscape Medical News.

However, Venook acknowledged that the arguments extended by Haines and Miklos are not totally unjustified, but he added that "the same critique could apply to many drugs," he said.

Venook did agree that bevacizumab has been overused. In colorectal cancer, there is no evidence of OS benefit with the combination of FOLFOX and bevacizumab, and yet this is the default treatment of most oncologists (approximately 70%), he said. "There is a comfort level with long-standing use. It is difficult to change practice patterns, and patients do not like the side effects [acneiform rash] of cetuximab," he said.

However, he feels that it is a mistake to use FOLFOX with bevacizumab as first-line therapy for colorectal cancer. Instead, he uses FOLFIRI [5-fluorouracil, leucovorin, and irinotecan], but it is not the combination of choice of most physicians in the United States, he added.

Adverse Events With Bevacizumab

In their letter, Haines and Miklos also raise the issue of adverse events with bevacizumab, which include thromboembolism, gastrointestinal perforation, hypertension, and proteinuria. "These data have important implications for other more prevalent and incurable cancer types [than ovarian cancer]," they write.

Venook agrees. "With the combination of FOLFOX and bevacizumab, neuropathy develops after seven to eight doses, and bevacizumab has to be withdrawn. After this, disease may rapidly progress," he said. "With complications that develop, patients may lose the gains of initial therapy," he added.

However, Coleman observed that, other than hypertension, adverse events with bevacizumab are infrequently serious or unmanageable. "Since the drug has been available for some time, our experience in patient selection and management of adverse events has lowered patient treatment discontinuation rates for toxicity as opposed to progression," he said.

Bevacizumab also has an impact on symptomatic ascites, which frequently occurs in patients with newly diagnosed and some with recurrent ovarian carcinoma, according to Coleman. "The impact of bevacizumab on QoL [quality of life] from this debilitating effect is remarkable," he said. He noted that in the AURELIA trial, the improvement in QoL, despite the drug's greater toxicity, is a testament to the need to understand toxicity from patient-reported outcomes.

"One cannot assume the occurrence of one leads to the other. These are well documented in all the phase 3 trials and generally show no diminution in QoL despite differential rates of adverse events in bevacizumab-containing arms," he said.

"The major issue with bevacizumab is that we don't have a reliable biomarker to identify patients for whom the agent works best," Coleman said.

Other experts expressed similar sentiments. During the past 10 years, the search for a biomarker has proved futile, Venook said.

Elaborating on the observation that the adjuvant trials of bevacizumab in breast cancer were negative, Sledge said: "I think everyone who worked with this drug thought that there was a subset of metastatic patients who got real benefit, but we were never able to identify a biomarker to identify them.

"Imagine what would have happened to Herceptin (trastuzumab) without HER2 as a biomarker," he added.

Sledge has received clinical trial support, research grants, and travel accommodations from Eli Lilly and Company. He also receives research support from Pfizer, consults with Syndax, Symphogen, and Verseau Therapeutics, and is on the board of directors for Tessa Therapeutics. Kindler and Venook report no relevant financial relationships. Tewari reports receiving honoraria from TESARO and Clovis Oncology; consults with and/or is on the advisory board of Roche/Genentech, AstraZeneca, and Merck; receives research funding from AbbVie, Genentech, Roche, Morphotek, Merk, and Regeneron Pharmaceuticals; and receives travel and accommodations from Roche and Genentech. Coleman consults with and/or receives grants from AstraZeneca, Merck, Tesaro, Medivation, Clovis, Gamamab, Genmab, Roche/Genentech, Janssen, Agenus, Regeneron, OncoQuest, V-Foundation, and the Gateway Foundation.

J Clin Oncol. Published online November 7, 2019. Letter to the editor, Full text; Response, Full text

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.