OncoSec Medical Inc (ONCSQ)

[hschlauch]

Now that the dust has settled somewhat, here are my personal thoughts on the updated combination data released yesterday.

These patients are not treatment naive - they have failed prior therapies. An unknown percentage of them have failed checkpoint inhibitor therapies including anti-PD-1. The question on some peoples' minds is whether or not the predicted nonresponder trial patients are indeed representative of an anti-PD-1 nonresponder population in advanced melanoma. They used a biomarker assay to select predicted nonresponders. While some, including myself, have questioned the reliability of the assay in accurately predicting nonresponders to anti-PD-1 therapy in advanced melanoma, a couple things come to mind.

The selected patients' exhausted phenotypes are well below the 30% CTLA-4/PD-1hi CD8+ threshold that is highly predictive of response to anti-PD-1 monotherapies. UCSF clinical researchers and others developed the assay thresholds based on observations in discovery and validation cohorts. The assay is based on a total of 40 patients. They "discovered" and "validated" that 100% of patients with greater than 30% of the exhausted phenotypes would respond; and 100% of those with less than 20% of the exhausted phenotypes would not respond. And those patients who fall somewhere in between the thresholds are split between responders and nonresponders. See https://www.jci.org/articles/view/87324.

At baseline, according to previously presented data, the patients enrolled in the combination trial that was just discussed yesterday by Algazi includes patients who are overwhelmingly below that 20% exhausted phenotype threshold. To see any responses out of these patients suggests to me that Immunopulse IL-12 is in fact elevating the exhausted phenotypes to the point where some of the patients land above the 30% threshold. Those who don't achieve responses, therefore, may be in that hit or miss zone or still below the 20% exhausted phenotype threshold. The question that needs to be answered is whether or not real anti-PD-1 nonresponders have something going on intratumorally that prevents them from observing significant elevations in their exhausted CD8-positive phenotype cells. Are their t-regs playing a role? Are CTLA-4 checkpoints on CD8 Tcells putting the breaks on immune responses?

Something to keep in mind for the PISCES trial: there will be enrolled patients who will be close to that 30% threshold for predictive response. The combination trial that was discussed yesterday by Algazi selected patients who were, again, way below 20%. The PISCES trial won't be as restrictive; there will be patients in that 20%-30% zone who would be more likely to achieve responses than patients who have very low percentages.

Another observation... I have never seen so many complete responses in either treatment naive or treatment experienced patients with stage III/IV metastatic melanoma. Those numbers are quite remarkable. You simply never see that level of response! I cannot express enough the significance of those data. Epacadostat plus pembrolizumab hasn't even come close to that percentage IN TREATMENT NAIVE PATIENTS!! I think these data alone will lead to some deal between Oncosec and Merck in a first line setting for metastatic melanoma and perhaps other indications as well. Really, no other combination has demonstrated anything close. This is the second time - the first was the retrospective analysis published last year that showed 50% complete responses in patients who went directly from Immunopulse IL-12 to an anti-PD-1 regimen - that have demonstrated such high levels of response.