Home > Focus Areas > I-O Connect > Post
  • Saved

1w
PD-1/PD-L1 checkpoint inhibition: predictive biomarkers, combination strategies, and overcoming resistance across tumor types

PD-1/PD-L1 checkpoint inhibitors have produced durable responses across multiple solid tumor types, yet primary and acquired resistance remains a substantial challenge. Predictive biomarker identification beyond PD-L1 tumor proportion score (TPS) and tumor mutational burden (TMB) continues to evolve.

PD-L1 TPS shows variable predictive utility across histologies, and TMB-high status does not uniformly confer benefit across tumor types. Emerging multiparameter approaches, incorporating tumor microenvironment profiling, T-cell inflamed gene expression, and circulating tumor DNA dynamics, are under prospective evaluation. Combination strategies with anti-CTLA-4, anti-LAG-3, anti-TIGIT, and VEGF/VEGFR inhibitors have expanded across indications; LAG-3 co-inhibition demonstrated improved progression-free survival in advanced melanoma versus PD-1 blockade alone. Immune-related adverse events, including colitis, pneumonitis, and endocrinopathies, require multidisciplinary management and careful rechallenge decision-making.

Medical oncologists, tumor immunologists, and immuno-oncology specialists managing diverse solid tumors will benefit from discussion on biomarker-driven patient selection, rational combination design, irAE management, and resistance strategies.

How do you incorporate multiparameter biomarker data, beyond PD-L1 TPS, into your checkpoint inhibitor selection and combination strategy decisions across tumor types? What is your approach to checkpoint inhibitor rechallenge following a significant immune-related adverse event, and what factors most influence this decision?

  • 1w
    Beyond PD-L1 TPS, the two other FDA-validated, clinically actionable biomarkers are MSI-H/dMMR and TMB-high (≥10 mut/Mb), both of which support tumor-agnostic pembrolizumab and dostarlimab (dMMR only) use, but all three are only modest predictors with important limitations, so the practical approach is to integrate them as complementary not interchangeable signals whose weighting differs by tumor type.
    A key reason to test all three rather than relying on any single marker is that they identify largely non-overlapping populations: across 11,348 tumors, few cases are triple-positive and the overlap pattern varies dramatically by histology.
  • 2w
    I use PD-L1, TMB, tumor type, molecular profile, and clinical factors together rather than relying on a single biomarker, with combination strategies such as CTLA-4 or VEGF/VEGFR inhibition considered when supported by disease-specific data. For rechallenge after a significant immune-related adverse event, I consider the severity and organ involved, resolution to baseline, need for steroids or immunosuppression, availability of alternative therapies, and the expected benefit of rechallenge; severe or life-threatening toxicities generally make rechallenge inappropriate.
  • 2w
    Selection criteria are getting more complicated all the time usually it's a multidisciplinary team approach.
    Would be very comfortable with rechallenge depending on the risk ratio.
  • 3w
    uSe PD-L1 status alongside other clinical and molecular factors—such as tumor type, genomic findings, disease burden, and emerging biomarkers when available—to guide checkpoint inhibitor selection. I consider combination immunotherapy for patients who are likely to benefit based on tumor characteristics and overall health, while balancing the higher risk of immune-related side effects. After a significant immune-related adverse event, I individualize the decision to rechallenge based on the severity and type of toxicity, how well it resolved, the patient’s cancer response, available alternative treatments, and current clinical guidelines, often involving a multidisciplinary team.
  • 3w
    I think we are all familiar on what patients will respond to PD-LT (TPS, CPS, Pole, MSI H etc). The newer biomarkers can help us decided who make need more than single agent IO (STK11, KEAP 1, SMARCA4) . This data is evolving.

    Rechallenge of IO: IF patient was on a doublet, I prefer to rechallenge single agent based on the grade. Grade 4 rechallenge is the difficult decision and I try to avoid this if at all possible, but have rechallenged in shared decision making scenarios

Show More Comments