Subbiah Addresses Controversy Surrounding Pembrolizumab Approval for TMB-High Tumors

Vivek Subbiah, MD

The FDA approval of pembrolizumab (Keytruda) for use in select adult and pediatric patients with unresectable or metastatic solid tumors that are tumor mutational burden (TMB)-high, has both generated a lot of excitement and been hotly debated, according to Vivek Subbiah, MD, who added that regardless, the decision will facilitate access to a therapy that can provide significant benefit for patients with no effective alternative treatment options.

The regulatory decision was based in part on data from a prospectively planned, retrospective analysis of 10 cohorts of patients with several previously treated, unresectable or metastatic, TMB-high solid tumors enrolled on the phase 2 KEYNOTE-158 trial. Results from the trial established a link between TMB-high status and improved overall response rate (ORR) with the immunotherapy in patients with various solid tumors.

The trial enrolled patients with anal, biliary, cervical, endometrial, salivary, thyroid, vulvar carcinoma, mesothelioma, neuroendocrine, or small cell lung cancer who also had an ECOG performance status of 0 or 1 and were relapsed/refractory to at least 1 prior line of therapy. The primary end points of the trial were ORR and duration of response (DOR) per RECIST v1.1 criteria, while key secondary end points included progression-free survival (PFS), overall survival (OS), and safety.

A total of 755 patients were determined to be evaluable for TMB; 102 of these patients had TMB-high tumors. Results showed that the ORR with pembrolizumab was 29% (95% CI, 21-39); 4% of patients achieved a complete response and 25% achieved a partial response. Although the median DOR had not been reached, 57% of patients experienced response durations of 12 months or longer and 50% of patients had response durations of 24 months or longer.

Furthermore, the 12-month PFS rates were 24.3% in the TMB-high cohort versus 14.0% in the TMB-low cohort, while the median OS was 11.1 months (95% CI, 8.1-16.1) versus 13.3 months (95% CI, 11.5-14.8), respectively. The 12-months OS rates were 48.0% versus 52.9%, respectively.

Regarding safety, common adverse effects (AEs) included fatigue, musculoskeletal pain, decreased appetite, pruritis, diarrhea, nausea, rash, pyrexia, cough, dyspnea, constipation, pain, and abdominal pain. There were also reports of immune-related AEs such as pneumonitis, colitis, hepatitis, endocrinopathies, nephritis, and adverse skin reactions.

“The FDA approval of pembrolizumab has been transformative for several reasons,” said Subbiah, an associate professor in the Investigational Cancer Therapeutics Department and center clinical medical director of the Clinical Center for Targeted Therapy, of the Cancer Medicine Division, at The University of Texas MD Anderson Cancer Center. “First, the magnitude and durability of responses mean that the approval will be life saving for a subset of patients with lethal cancers. Secondly, drug access and reimbursement will be enhanced for adult and pediatric patients with solid tumors, which is an enormous unmet need.”

He added that physicians, patients, and their families will be empowered to make decisions in the real world; immunotherapy access will be improved for underserved, minority, and economically disadvantaged populations who may not otherwise have access to this therapy; and the approval will encourage the adoption of modern, essential diagnostic genomic tests.

In a special episode of OncLive On Air^TM, we passed the mic to Subbiah, who discussed the FDA approval of pembrolizumab in TMB-high solid tumors and addressed the controversy surrounding the approval.

OncLive: The approval of pembrolizumab for the treatment of patients with TMB­-high solid tumors was based on findings from the KEYNOTE-158 trial. Could you expand on these data?

Subbiah: This approval was based on efficacy data from 10 refractory solid tumor cohorts which were enrolled in the multicenter, non-randomized, open-label, phase 2 KEYNOTE-158 trial. During this study, 102 patients, which comprised 30% of the patients enrolled, had TMB-high tumors, which was was defined as 10 or higher mutations/megabase.

Among these patients, the objective response rate was 29%. Overall, about half the responses were greater than 2 years with many patients obtaining durability of response, which was broadly observed in heavily pretreated metastatic cancers with treatment modalities other than immunotherapy.

Importantly, this was a tumor agnostic approval, for both adult and pediatric patients, which provided greater access to immunotherapy.

Do you believe the single-arm trial showed efficient evidence for this approval?

As it was based on a relatively small, single-arm study, the pembrolizumab approval has generated considerable debate within the oncology community. In my personal opinion, based on first-hand experience with immune checkpoint inhibitors in patients with TMB-high solid tumors, a subset of patients can go from a refractory state to a durable complete remission that can last many years. After reviewing available literature, I certainly support the tumor agnostic FDA approval of pembrolizumab for patients with TMB-high tumors.

I believe the approval will facilitate access to a therapy that can provide significant benefit for patients with no effective alternative treatment options. This approval will especially impact the underserved and minority populations who are less likely to have access to molecular profiling and off-label therapies. [The decision] will also benefit patients with rare cancer types that account for almost one-quarter of the cancer burden. This is far from the first FDA approval of pembrolizumab; rather, it’s an expansion of a drug label that is already known. The safety profile of the agent is also widely understood.

As of June 29, 2020, pembrolizumab received regulatory approval for over 15 indications and greater than 20 specific labels. As with all biologic and targeted cancer therapies, not all patients will achieve a durable response. Looking forward, biomarkers will be critical to optimize the use of this agent, as they will allow clinicians to identify the patients who are most likely to benefit.

Previously, only PDL-1 expression and microsatellite instability (MSI)-high deficient mismatch repair were recognized by FDA as predictive biomarkers of immunotherapy response.

However, the association of TMB-high with the response to anti–PD-1 and anti–PD-L1 therapy arose, not just from this small study, but from a multitude of retrospective analyses. For example, a recent meta-analysis of 117 clinical trials, which included more than 4450 patients treated with immunotherapy, revealed that TMB-high solid tumors were associated with enhanced objective responses to anti–PD-L1 therapy, anti–CTLA-4 therapy, and the combination of anti–PD-1, –PD-L1, and CTLA-4 therapy. Moreover, MSI-high also appears to be associated with immunotherapy response.

Would the conversation be different if the drug was not studied as well?

Yes, I believe that the conversation would be different in terms of the number of patients that are needed for such an approval; it would also depend on the rarity of the biomarker. As with many of the tumor agnostic approvals, the magnitude of response would be taken into account.

We currently have NTRK inhibitors, which were approved in a tumor agnostic fashion. We have pembrolizumab, of course, which was the first tumor agnostic approval for MSI-high cancers. Again, the magnitude of response was high. Based on this, I would say the magnitude of response or the number of patients for such an approval should require a great number of patients.

We know that the challenge is always the optimal cutoff for defining a biomarker, especially for TMB-high. This marker varies among cancer types. TMB greater than 10 thresholds arose from a consensus recommendation from the TMB Harmonization Project. Notably, this was a multi-stakeholder consortium and panel debate from academia, pharmaceutical, nonprofit cancer society representatives, many commercial sequencing companies, and nonprofit patient advocate groups with the recognition that it may not be optimal for all clinical scenarios. As with anything in oncology, understanding other biological variables that affect immune activation may be needed in order to better understand this biomarker.

How do you foresee the approval impacting molecular profiling?

We believe that lack of access to molecular profiling technology and off-label use of targeted immunotherapy is becoming an important contributor in terms of healthcare disparities. I’m hopeful that the FDA approval of pembrolizumab for patients with TMB-high tumors will help reduce disparities by expanding the reimbursement for tumor molecular profiling; this will facilitate access to a widely used therapy that can induce durable responses in this molecularly defined population. For instance, when testing a pediatric patient for high TMB, we may not get high TMB alone. The patient may have an NTRK fusion, for example, which is targetable. For that patient and their family, this is transformative.

Previously, access to these therapies and sequencing profiles were plausible only for patients who were referred to major cancer centers. With the advent of commercial sequencing companies, so many community oncologists are ordering next-generation (NGS) profiling. Some of the challenges are access to NGS profiling, reimbursement, and access to targeted therapies. Ultimately, the more we test patients, the more we find. If we don’t look for these aberrations, we don’t know if patients are eligible [to receive certain therapies].

If a patient is not eligible for a trial, they have to travel, potentially a long distance, to a major cancer center. The patient and the doctor must request off-label use, which could cause significant delays in treatment. These logistical considerations [reveal] disparities that impact patients of lower socioeconomic status the most. When looking at the expanded FDA indication for pembrolizumab, permitting broader access to profiling will help many patients.

There’s an idea that the FDA is the arbiter for what is best; however, this is not always the case. What are your thoughts?

One concern about this approval is that it will prevent patients from receiving other treatment options that are potentially better. It’s important to emphasize that the FDA does not dictate standard of care; it provides patients with options. The FDA approved pembrolizumab for patients with TMB 10 mutations/megabase or higher who have progressed on prior courses of therapy and have no satisfactory alternative options. A discussion is needed between doctors, patients, and their families.

Clearly, if there are better treatment options for patients with metastatic malignancies who have progressed after 4 or 5 lines of therapy, the oncologist could elect to use those options. Again, lack of access to molecular profiling and off-label use is becoming an important contributor of disparities. This approval and expanded indication permit broader access to the drug and molecular profiling for patients.

I believe that the FDA approval of pembrolizumab is based on compelling scientific rationale and compelling clinical data. This will provide patients with a therapeutic option that can confirm long-term durable remissions and, very importantly, the gift of time.

For anyone who is still on the fence with this approval, what is important for them to know?

There is no perfect biomarker in oncology. We know that PD-L1 expression and MSI-high are biomarkers, and now we have TMB as another biomarker of response to immunotherapy. It is likely that concentration of additional factors will enable the refinement of the predictive value of TMB. As with any complex biological process, antitumor immunity is associated with multiple biomarkers, and several are likely needed to increase the sensitivity and specificity of these predictive models over time. In the field of breast cancer, for example, the discovery and implementation of estrogen receptor, HER2, and BRCA have refined clinical treatment paradigms.

We continue to learn more about TMB. It is important to not let the immediate pursuit of perfection prevent progress. Undoubtedly, this approval will fuel further research and controversy, and will enhance our predictive capabilities. However, based on what we already know about the association of TMB with immunotherapy, these deliberations and debates will likely take many years, especially if the intent is to design prospective studies to determine the TMB threshold for each and every cancer type. As such, the prospective studies may not be feasible for many rare cancers.

In my opinion, the completion of such studies should not be used for access to a lifesaving drug. Based on my personal first-hand experience with immune checkpoint inhibitors, the available literature, and the FDA label, I strongly support the tumor agnostic regulatory approval of pembrolizumab for patients with TMB-high solid tumors.

Reference

Marabelle A, Fakih MG, Lopez J, et al. Association of tumour mutational burden with outcomes in patients with select advanced solid tumours treated with pembrolizumab in KEYNOTE-158. Ann Oncol. 2019;30(suppl 5):11920. doi:10.1093/annonc/mdz253.018