Clinical Insights

How should we evaluate the benefit of immunotherapy combinations?



Every medical oncologist who has described a combination chemotherapy regimen to a patient with advanced cancer has likely been asked whether the benefits of tumor shrinkage, disease-free survival (DFS), and overall survival are worth the risks of adverse events (AEs).

Dr. Alan P. Lyss, now retired, was a community-based medical oncologist and clinical researcher for more than 35 years, practicing in St. Louis.

Dr. Alan P. Lyss

Single-agent immunotherapy and, more recently, combinations of immunotherapy drugs have been approved for a variety of metastatic tumors. In general, combination immunotherapy regimens have more AEs and a higher frequency of premature treatment discontinuation for toxicity.

Michael Postow, MD, of Memorial Sloan Kettering Cancer Center in New York, reflected on new ways to evaluate the benefits and risks of immunotherapy combinations during a plenary session on novel combinations at the American Association for Cancer Research’s Virtual Special Conference on Tumor Immunology and Immunotherapy.

Potential targets

As with chemotherapy drugs, immunotherapy combinations make the most sense when drugs targeting independent processes are employed.

As described in a paper published in Nature in 2011, the process for recruiting the immune system to combat cancer is as follows:

  • Dendritic cells must sample antigens derived from the tumor.
  • The dendritic cells must receive an activation signal so they promote immunity rather than tolerance.
  • The tumor antigen–loaded dendritic cells need to generate protective T-cell responses, instead of T-regulatory responses, in lymphoid tissues.
  • Cancer antigen–specific T cells must enter tumor tissues.
  • Tumor-derived mechanisms for promoting immunosuppression need to be circumvented.

Since each step in the cascade is a potential therapeutic target, there are large numbers of potential drug combinations.

Measuring impact

Conventional measurements of tumor response may not be adequately sensitive to the impact from immunotherapy drugs. A case in point is sipuleucel-T, which is approved to treat advanced prostate cancer.

In the pivotal phase 3 trial, only 1 of 341 patients receiving sipuleucel-T achieved a partial response by RECIST criteria. Only 2.6% of patients had a 50% reduction in prostate-specific antigen levels. Nonetheless, a 4.1-month improvement in median overall survival was achieved. These results were published in the New England Journal of Medicine.

The discrepancy between tumor shrinkage and survival benefit for immunotherapy is not unexpected. As many as 10% of patients treated with ipilimumab (ipi) for stage IV malignant melanoma have progressive disease by tumor size but experience prolongation of survival, according to guidelines published in Clinical Cancer Research.

Accurate assessment of the ultimate efficacy of immunotherapy over time would benefit patients and clinicians since immune checkpoint inhibitors are often administered for several years, are financially costly, and treatment-associated AEs emerge unpredictably at any time.

Curtailing the duration of ineffective treatment could be valuable from many perspectives.

Immunotherapy combinations in metastatic melanoma

In the CheckMate 067 study, there was an improvement in response, progression-free survival (PFS), and overall survival for nivolumab (nivo) plus ipi or nivo alone, in comparison with ipi alone, in patients with advanced melanoma. Initial results from this trial were published in the New England Journal of Medicine in 2017.

At a minimum follow-up of 60 months, the 5-year overall survival was 52% for the nivo/ipi regimen, 44% for nivo alone, and 26% for ipi alone. These results were published in the New England Journal of Medicine in 2019.

The trial was not statistically powered to conclude whether the overall survival for the combination was superior to that of single-agent nivo alone, but both nivo regimens were superior to ipi alone.

Unfortunately, the combination also produced the highest treatment-related AE rates – 59% with nivo/ipi, 23% with nivo, and 28% with ipi in 2019. In the 2017 report, the combination regimen had more than twice as many premature treatment discontinuations as the other two study arms.

Is there a better way to quantify the risk-benefit ratio and explain it to patients?

Alternative strategies for assessing benefit: Treatment-free survival

Researchers have proposed treatment-free survival (TFS) as a potential new metric to characterize not only antitumor activity but also toxicity experienced after the cessation of therapy and before initiation of subsequent systemic therapy or death.

TFS is defined as the area between Kaplan-Meier curves from immunotherapy cessation until the reinitiation of systemic therapy or death. All patients who began immunotherapy are included – not just those achieving response or concluding a predefined number of cycles of treatment.

The curves can be partitioned into states with and without toxicity to establish a unique endpoint: time to cessation of both immunotherapy and toxicity.

Researchers conducted a pooled analysis of 3-year follow-up data from the 1,077 patients who participated in CheckMate 069, testing nivo/ipi versus nivo alone, and CheckMate 067, comparing nivo/ipi, nivo alone, and ipi alone. The results were published in the Journal of Clinical Oncology.

The TFS without grade 3 or higher AEs was 28% for nivo/ipi, 11% for nivo alone, and 23% for ipi alone. The restricted mean time without either treatment or grade 3 or greater AEs was 10.1 months, 4.1 months, and 8.5 months, respectively.

TFS incentivizes the use of regimens that have:

  • A short duration of treatment
  • Prolonged time to subsequent therapy or death
  • Only mild AEs of brief duration.

A higher TFS corresponds with the goals that patients and their providers would have for a treatment regimen.

Adaptive models provide clues about benefit from extended therapy

In contrast to cytotoxic chemotherapy and molecularly targeted agents, benefit from immune-targeted therapy can deepen and persist after treatment discontinuation.

In advanced melanoma, researchers observed that overall survival was similar for patients who discontinued nivo/ipi because of AEs during the induction phase of treatment and those who did not. These results were published in the Journal of Clinical Oncology.

This observation has led to an individualized, adaptive approach to de-escalating combination immunotherapy, described in Clinical Cancer Research. The approach is dubbed “SMART,” which stands for sequential multiple assignment randomized trial designs.

With the SMART approach, each stage of a trial corresponds to an important treatment decision point. The goal is to define the population of patients who can safely discontinue treatment based on response, rather than doing so after the development of AEs.

In the Adapt-IT prospective study, 60 patients with advanced melanoma with poor prognostic features were given two doses of nivo/ipi followed by a CT scan at week 6. They were triaged to stopping ipi and proceeding with maintenance therapy with nivo alone or continuing the combination for an additional two cycles of treatment. Results from this trial were presented at ASCO 2020 (abstract 10003).

The investigators found that 68% of patients had no tumor burden increase at week 6 and could discontinue ipi. For those patients, their response rate of 57% approached the expected results from a full course of ipi.

At median follow-up of 22.3 months, median response duration, PFS, and overall survival had not been reached for the responders who received an abbreviated course of the combination regimen.

There were two observations that suggested the first two cycles of treatment drove not only toxicity but also tumor control:

  • The rate of grade 3-4 toxicity from only two cycles was high (57%).
  • Of the 19 patients (32% of the original 60 patients) who had progressive disease after two cycles of nivo/ipi, there were no responders with continued therapy.

Dr. Postow commented that, in correlative studies conducted as part of Adapt-IT, the Ki-67 of CD8-positive T cells increased after the initial dose of nivo/ipi. However, proliferation did not continue with subsequent cycles (that is, Ki-67 did not continue to rise).

When they examined markers of T-cell stimulation such as inducible costimulator of CD8-positive T cells, the researchers observed the same effect. The “immune boost” occurred with cycle one but not after subsequent doses of the nivo/ipi combination.

Although unproven in clinical trials at this time, these data suggest that response and risks of toxicity may not support giving patients more than one cycle of combination treatment.

More nuanced ways of assessing tumor growth

Dr. Postow noted that judgment about treatment effects over time are often made by displaying spider plots of changes from baseline tumor size from “time zero” – the time at which combination therapy is commenced.

He speculated that it might be worthwhile to give a dose or two of immune-targeted monotherapy (such as a PD-1 or PD-L1 inhibitor alone) before time zero, measure tumor growth prior to and after the single agent, and reserve using combination immunotherapy only for those patients who do not experience a dampening of the growth curve.

Patients whose tumor growth kinetics are improved with single-agent treatment could be spared the additional toxicity (and uncertain additive benefit) from the second agent.

Treatment optimization: More than ‘messaging’

Oncology practice has passed through a long era of “more is better,” an era that gave rise to intensive cytotoxic chemotherapy for hematologic and solid tumors in the metastatic and adjuvant settings. In some cases, that approach proved to be curative, but not in all.

More recently, because of better staging, improved outcomes with newer technology and treatments, and concern about immediate- and late-onset health risks, there has been an effort to deintensify therapy when it can be done safely.

Once a treatment regimen and treatment duration become established, however, patients and their physicians are reluctant to deintensity therapy.

Dr. Postow’s presentation demonstrated that, with regard to immunotherapy combinations – as in other realms of medical practice – science can lead the way to treatment optimization for individual patients.

We have the potential to reassure patients that treatment de-escalation is a rational and personalized component of treatment optimization through the combination of:

  • Identifying new endpoints to quantify treatment benefits and risks.
  • SMART trial designs.
  • Innovative ways to assess tumor response during each phase of a treatment course.

Precision assessment of immunotherapy effect in individual patients can be a key part of precision medicine.

Dr. Postow disclosed relationships with Aduro, Array BioPharma, Bristol Myers Squibb, Eisai, Incyte, Infinity, Merck, NewLink Genetics, Novartis, and RGenix.

Dr. Lyss was a community-based medical oncologist and clinical researcher for more than 35 years before his recent retirement. His clinical and research interests were focused on breast and lung cancers, as well as expanding clinical trial access to medically underserved populations. He is based in St. Louis. He has no conflicts of interest.

Next Article: