Inclusion criteria:
"Patients must have histological or cytological diagnosis of melanoma with progressive locally advanced or metastatic disease that is not amenable to definitive local therapy with curative intent." When you have "progressed", you are refractory to treatment(s) or your disease is not stable without treatments. Non-responders would fall into this category.
However, the trial allows for "prior chemotherapy or immunotherapy or radiation therapy." This implies that they are enrolling patients who aren't just potential non-responders to anti-pd-1 monotherapy, but they are also potentially enrolling patients who are refractory to chemotherapies, other immunotherapies, or radiation therapies.
I am confident that the TIL phenotype exploratory biomarker developed by the UCSF is truly representative of patients that would be clinical non-responders to monotherapy pd-1 checkpoint inhibitors. But for this trial, they aren't actively recruiting patients who are determined non-responders to monotherapy pd-1 treatment. They have opened the door to patients who MAY be non-responders to immunotherapy in general, but they have also left the door open to non-responders to chemotherapy, non-responders to radiation therapy, and even to treatment-naive patients.
While the protocol calls for "<70% PD-1hiCD8+CTLA4+ in the CD45+ gate based on flow cytometry", they have further refined the requirement to only include patients with <25% PD-1hiCD8+CTLA4+ TIL; these details are left out of the clinical trials inclusion criteria, but it is still technically <70%.
Is this <25% requirement an accurate predictor of patients who would be non-responders on monotherapy pd-1 treatment? From what I can gather from the data presented in the Journal of Clinical Investigation, the two cohorts comprised of 20 patients each - "discovery" and "validation" - clearly shows that no one with a <20% PD-1hiCD8+CTLA4+ TIL profile had responded to anti-pd-1 monotherapy. In addition, everyone with a >30% PD-1hiCD8+CTLA4+ TIL profile did respond to anti-pd-1 monotherapy. Predictability is not significant for patients in the 20%-30%-positive range. This means patients in the 20%-25% (25% is the trial cutoff) range may or may not respond to monotherapy anti-pd-1. It will be interesting to see case studies on the phase II combination patients to know exactly what their biopsied TIL phenotypes show.
The biomarker assay is still very significant, because it informs the registration trial design. Trial investigators appear to be confident in the predictability of the biomarker, and if they see responses in the phase II combination trial, they are likely going to attribute those responses to the EP IL-12. This builds confidence in the product and the feasibility of moving forward with the registration trial.
