Diagnostic Biomarkers for Cancer Immunotherapy – Moving Beyond PD-L1
Santa Monica, CA, August 11, 2016 – Here at DeciBio we have been following the progression of the cancer immunotherapy, or immuno-oncology (I/O), landscape with great interest, curious how this burgeoning field will impact cancer diagnostics, particularly companion diagnostics (CDx). Personally, I believe that the collision of the impending tidal wave of cancer immunotherapy drugs with the surge of new and emerging diagnostic platforms (i.e., next-generation sequencing (NGS), single-cell technologies, multi-omic technologies, digital quantification technologies, multi-parameter/high-plex technologies) will lead to a transformation in oncology diagnostics over the next ten years.
While it is too early to tell exactly which tools and technologies will lead the charge in I/O diagnostics, it seems clear that the I/O biomarker landscape will explode in the coming years due to the variety of drugs and potential drug combinations coming through the pipeline, continued public interest in personalized medicine, and the ability of emerging diagnostic technologies to better interrogate the complexity of tumor-immune interactions than traditional technologies. In order to gather some insights into the future biomarker landscape for I/O, I did an analysis of clinical trials, focusing on open and active I/O-related trials that included “laboratory biomarker analysis” as a trial objective, and captured information (when available) about the types of biomarkers being explored, the purpose for biomarker analysis, sample types, and technologies utilized. This search yielded 97 different trials across all phases, and, though not an exhaustive picture of all I/O biomarker-related clinical activity, provided interesting insights about the future of I/O biomarkers and diagnostics.
See the interactive Tableau™ dashboard below to browse and filter the results (click on the buttons at the top, the biomarker circles, or the trial information in the table at the bottom to filter the data as desired; use Crtl + Click to make multiple selections):
From this analysis, as well as other discussions I have held with clinical and industry experts, I have identified a few trends and insights that I think are likely to shape the I/O biomarker landscape:
PD-L1 will likely remain the primary I/O biomarker in the near-mid term: Despite being perceived by many as a relatively poor biomarker (10 – 20% of PD-L1 negative NSCLC and melanoma tumors are responsive to Nivolumab and Pembrolizumab vs. 30 – 40% for PD-L1 positive tumors – source), PD-L1 was the single most-common biomarker mentioned in the trials analyzed. With regulatory approval for Dako’s and Ventana’s PD-L1 CDx tests for a handful of indications in both the U.S. and Europe, the rapid adoption of the current anti-PD-L1 checkpoint inhibitors, and the expected arrival of new PD-1/L1 drugs (e.g., Pfizer’s Avelumab, AstraZeneca / MedImmune’s Durvalumab) PD-L1 will pave the way for I/O CDx testing in the years to come, particularly as the labels for existing drugs and tests expand to first-line status for lung cancer and other solid tumors.
Multiplex assays will play a significant role: While PD-L1 and other single-marker tests may drive I/O diagnostics in the near term, many trials are also exploring multi-analyte markers. Common multiplex assays mentioned in the clinical trials include tumor mutation load profiling, T-cell receptor (TCR) repertoire and clonal expansion analysis, peripheral cytokine profiling, multi-parameter flow cytometry and IHC, and gene expression profile analysis. In addition to analyzing multiple markers of the same kind, many trials are exploring analysis of different types of biomarkers simultaneously, such as PD-L1 expression along with the number and types of different immune cells in the tumor microenvironment. Multiplex assays may provide better predictive capabilities than the single-marker tests used today. Many companies are currently pursuing multiplex I/O diagnostics; for example, Adaptive Biotechnologies and ImmunID are targeting TCR analysis using NGS and multiplex PCR, respectively, NanoString is offering I/O-related gene and protein expression assays, and Foundation Medicine is utilizing its FoundationOne assay to extrapolate tumor mutation burden. All these companies have announced clinical and/or commercial partnerships for I/O diagnostics development.
I/O drugs will likely be used primarily in combinations: The vast majority (~84%) of trials analyzed are exploring immunotherapies in combination with a variety of other types of cancer therapies. While the focus on combination therapies is not surprising, the relatively limited number of trials exploring biomarkers for predicting toxicity or adverse events (7 / 97) is surprising given the potential for synergistic toxicities with combination therapies. Additionally, the recent failure of BMY’s Nivolumab (source) to show improved response over chemotherapy as a first-line agent in NSCLC indicates that combination approaches may be required to achieve high-response rates in the earlier lines of treatment, at least in the case of non-biomarker selected patients.
Immune and therapy monitoring may represent additional diagnostic opportunities: In addition to exploring biomarkers for selecting patients for treatment or predicting likely responders, nearly half of the trials analyzed (46 / 97) are exploring markers to monitor immune response or drug efficacy over time. Common monitoring assays include assessing changes in peripheral immune cell populations and phenotypes, cytokine profiles, and TCR clonality over time compared to a pre-treatment baseline. Other trials are exploring the degree of infiltration of immune cells into tumor tissues or changes in PD-L1 expression following treatment. Such tissue-based monitoring, however, may prove challenging due to the need to collect repeat biopsy specimens of the original and/or metastatic tumor sites.
Liquid biopsies are gaining traction: Tissue remains the sample type of choice for biomarker analysis for predicting response to treatment (52 / 97 trials), however, approximately a third of all trials analyzed are exploring peripheral blood-based molecular and/or cellular biomarkers for therapy selection. Cytokine profiles, immune-cell population / phenotype analysis, and soluble or expressed PD-L1 are the most common biomarkers identified for therapy selection testing using peripheral blood samples. While a few companies are, or are planning to, offer liquid biopsy assays for oncology diagnostics (e.g., Roche’s FDA-approved EGFR test, Illumina spinout Grail Bio’s planned NGS screening assay, Pathway Genomics’ CancerIntercept), relatively few I/O-specific liquid biopsy tests are commercially available. To my knowledge, Biocept and Cynvenio Biosystems, both of which offer blood-based PD-L1/CTC assays, are the only companies to offer I/O-related liquid biopsy tests.
I will update and refine this trial analysis periodically, adding new trials and those that may have fallen outside of the scope of my initial search, so check back in if you are as interested in I/O biomarkers as I am. In the meantime, check out our other blogs or browse our market intelligence products to get our insights and analysis on other topics and markets.
Disclaimer: Some of the companies listed above may be DeciBio Consulting clients or customers.