We all want new treatments to “work” for breast cancer patients. For the National Breast Cancer Coalition (NBCC), that means the investigational therapy will significantly prolong overall survival (OS) without adverse impact on quality of life. We expect that the evidence of that benefit will be shown by well-designed, well-conducted prospective randomized clinical trials, sufficiently designed to support that result. Unfortunately, that has not been the case with TECENTRIQ® (atezolizumab) in first-line treatment of advanced or metastatic triple-negative breast cancer (TNBC).
In 2018, many in the medical oncology community celebrated a “practice changing” result from Impassion130, the Phase III randomized clinical trial of atezolizumab combined with nab-paclitaxel (Abraxane®) in patients with advanced or metastatic TNBC.[1],[2] That excitement was a reaction to data from a subgroup of patients with PD-L1+ tumors that suggested a beneficial trend in OS. However, according to the trial investigators’ own study protocol, that data could not be formally analyzed because the trial failed to show an OS improvement in the larger intention-to-treat (ITT) study population. Assessment of OS in the PD-L1+ subgroup in this case would have violated the analysis hierarchy implemented to control for a type 1 error (false positive result). However, the drug was approved under the US Food and Drug Administration (FDA) accelerated approval process based on the 2.6-month improvement in progression-free survival (PFS), and now costs around $13,450 per month.[3] The continued approval of atezolizumab was contingent on a confirmatory trial, Impassion131, which was already underway, to demonstrate clinical benefit (i.e., improved overall survival or improved quality of life).
NBCC takes issue with this trial as the confirmatory trial for two reasons. First, the trial used PFS as a primary endpoint with OS as a secondary endpoint, which was only to be analyzed for significance if PFS was significant in both the PD-L1+ subgroup and ITT population. Moreover, even after prior data had demonstrated that atezolizumab in this setting (i.e., advanced and metastatic disease TNBC) confers no benefit to patients who are not PD-L1+, the trial continued to enroll an “all-comer” population, guaranteeing that a large subset of the participating patients would be subject to harm with no likelihood of benefit.
On August 6, 2020, Genentech (a subsidiary of Roche) issued a press release[4] reporting key preliminary results from its Impassion131 trial. Similar to Impassion130, this trial tested atezolizumab or placebo combined with paclitaxel, instead of nab-paclitaxel, in first-line advanced or metastatic TNBC. Not only did Impassion131 fail to replicate the modest PFS benefit of the earlier Impassion130 trial, even in the PD-L1+ subgroup, disturbingly, data for the secondary endpoint of OS showed a negative trend for patients on the treatment arm compared to the control arm. Also troubling was the fact that patients in Impassion131 were randomized in a 2 to 1 fashion with twice as many patients in the atezolizumab arm.[5] The results of Impassion131 presented a complete reversal of findings reported in the New England Journal of Medicine in 2018,2 and subsequently in The Lancet in January 20201 from the Impassion130 trial. (It is interesting that the 7-month median overall survival benefit reported among PD-L1+ patients participating in the Impassion130 trial is described as “clinically meaningful,” and yet the 6.2-month loss in median survival among PD-L1+ patients in Impassion131 is not described as clinically meaningful in any way.)
Following the Genentech press release in August 2020, NBCC released a statement and called on the FDA to withdraw the accelerated approval of atezolizumab at that time. This was not a new position of NBCC. Indeed, for many years, NBCC has spoken out against accelerated approval of breast cancer drugs based on unvalidated surrogate endpoints.
We understand that the purpose of accelerated approval is to get drugs to patients with serious conditions earlier than might otherwise occur, where there is an unmet need, and the investigational drug may be helpful. What is the unmet need in breast cancer? We don’t know how to cure metastatic disease for one. You would assume then that accelerated approval is warranted if there is clinical trial evidence that a therapy confers clinical benefit over and above standard of care in that it at least extends metastatic breast cancer patients’ overall survival or significantly improves their quality of life. We already have many approved drugs in breast cancer that do neither. We would very much want to get one that does have these benefits to patients as soon as possible. But that is not atezolizumab. Accelerating the approval of a therapy of unknown meaningful clinical benefit and known harm, is not meeting an unmet need.
According to regulation, drugs granted accelerated approval based on surrogate endpoints must swiftly go on to demonstrate improvements in a clinically meaningful endpoint in a confirmatory trial or be subject to expedited withdrawal. But that is not how the FDA accelerated approval process has worked in practice.
In a 2-year follow-up to the FDA’s own published analysis[6] of all cancer drug indications receiving accelerated approval between 1992 and May 2017, Gyawali, et al. 2019[7] reported that only 19 of 93 (~20%) demonstrated improvements in overall survival in confirmatory trials. An additional 19 of 93 (20%) reported improvement in the same surrogate used in the preapproval trial, and 20 (21%) reported improvement in a different surrogate. Additionally, 24 confirmatory trials were either delayed, pending, or ongoing further highlighting problems with the intended timeliness of confirmatory trials.
When trials do not yield useful information about how well an intervention saves or prolongs lives, important questions are left unanswered. This means that further trials will be necessary – requiring more time, more money, and more patient participation that could be involved elsewhere. Meanwhile, drugs with no proven clinical benefit that have an accelerated approval status remain available to unsuspecting patients who may not understand the limited information available about the drug on outcomes that really matter. This can be avoided by selecting, at the outset, outcomes that provide the clearest indications of the true efficacy and harms of the intervention being examined.
In the specific case of atezolizumab, we continue to have no idea if this drug improves any clinically meaningful endpoints such as OS. We do know it confers harm. As such, under the rules guiding accelerated approval, this drug should be withdrawn from the market until such time that well-designed randomized clinical trials demonstrate clinically meaningful benefit.
We are all desperate to find interventions that will save lives or significantly prolong lives without reducing the quality of life. We should not accept less than real, high-level evidence that approved treatments actually do this. That would be progress. We should not redefine progress to include what is possible in the short term, rather than what is meaningful.
Breast cancer patients must be able to make health care decisions based upon the highest possible quality of evidence. To achieve this, we need well-designed clinical trials that use the most valid and meaningful outcome measures available. We recognize that selection of the ideal experimental endpoint is not always possible or practical. On the other hand, designing a trial with practical but inadequate outcome measures will only result in lives lost, wasted resources, and delays in identifying the most effective interventions.
[1] Schmid P, Rugo HS, Adams S, et al. Atezolizumab plus nab-paclitaxel as first-line treatment for unresectable, locally advanced or metastatic triple-negative breast cancer (IMpassion130): updated efficacy results from a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol. 2020;21(1):44-59. doi:10.1016/S1470-
[2] Schmid P, Adams S, Rugo HS, et al. Atezolizumab and Nab-Paclitaxel in Advanced Triple-Negative Breast Cancer. N Engl J Med. 2018;379(22):2108-2121. doi:10.1056/NEJMoa1809615 https://www.nejm.org/doi/full/10.1056/nejmoa1809615
[3] https://www.drugs.com/medical–answers/cost–tecentriq–3064818/
[4] https://www.businesswire.com/news/home/20200806005915/en/Genentech-Update-Phase-III-StudyTecentriqCombination
[5] https://cancerres.aacrjournals.org/content/78/4_Supplement/OT1–01–01.short
[6] Beaver JA, Howie LJ, Pelosof L, Kim T, Liu J, Goldberg KB, Sridhara R, Blumenthal GM, Farrell AT, Keegan P, Pazdur R, Kluetz PG. A 25-Year Experience of US Food and Drug Administration Accelerated Approval of Malignant Hematology and Oncology Drugs and Biologics: A Review. JAMA Oncol. 2018 Jun 1;4(6):849-856. doi: 10.1001/jamaoncol.2017.5618. PMID: 29494733.
[7] Gyawali B, Hey SP, Kesselheim AS. Assessment of the Clinical Benefit of Cancer Drugs Receiving Accelerated Approval. JAMA Intern Med. 2019 Jul 1;179(7):906-913. doi: 10.1001/jamainternmed.2019.0462. PMID: 31135808; PMCID: PMC6547118.