Trial of Immunotherapy With Avelumab and Pepinemab As Second Line For Patients With Metastatic Pancreatic Adenocarcinoma

Pancreatic adenocarcinoma (PDAC) is the third leading cause of cancer related deaths with an incidence expected to increase over the next decade.3 Despite modest advances in conventional chemotherapy, five-year survival remains dismal at 5-10%.4 Therefore, the development of effective therapies for treating PDAC represents a significant unmet medical need. The near 95% 5 year mortality observed in PDAC is in part due to advanced disease stage at presentation, and a propensity for recurrence following curative intent resection. As a result, 5FU regimens (including FOLFIRINOX) or Gemcitabine based cytotoxic chemotherapies are standard of care for patients who present with both resectable and unresectable disease.5,6 However, despite advances in cytotoxic chemotherapy, a majority of patients fail to respond to first line therapy, and only around 10-15% of patients experience treatment response in the second line setting, while a third of patients experience grade 2-3 toxicities from second-line regimens.7,8 Thus, novel second-line treatments with improved toxicity profiles compared to cytotoxic chemotherapy are a dire medical need and strongly supported by patient advocates of this dismal disease.

A body of clinical and preclinical work has shown that the profoundly immunosuppressive tumor microenvironment (TME) in PDAC remains a significant barrier to effective cytotoxic and immune based therapies.9-11 Somewhat unique to PDAC, a dense fibrotic stroma is infiltrated with bone-marrow derived myeloid cells of monocytic and granulocytic origin including tumor associated macrophages (TAM) neutrophils (TAN) and immature suppressive myeloid progenitors (MDSCs). Tumor-infiltrating myeloid cells are highly immunosuppressive and the most prevalent immune cells found in the TME of PDAC.12,13 These cells produce immunosuppressive cytokines (Il6, Il10, Il4), secrete considerable amounts of arginase, and produce reactive oxygen species, essentially shielding the tumor from effective T cell immunity.14 Thus, PDAC is an immunologically "cold" tumor with minimal T cell adaptive response, which in part explains the failure of immune-checkpoint blockade therapy to demonstrate efficacy in PDAC.15-17 Previous work has shown that small molecule chemokine inhibitors targeted at monocytic myeloid cells (TAM precursors) can effectively prevent mobilization and trafficking of these cells to the TME, and improve the standard of care chemotherapy in both preclinical and clinical studies (NCT01413022, NCT02732938, NCT02345408).18 However, these successes in reprogramming of the tumor microenvironment are accompanied by a compensatory increase in immunosuppressive granulocytes (TANs and pmn-MDSCs), resulting in treatment resistance to single subtype myeloid blockade.18 Based on this finding of myeloid compensation by TANs in the absence of TAMs, and the potential therapeutic value in dual myeloid inhibition, new therapies targeted at both myeloid subtypes represent a unique opportunity for exposing PDAC to effective immune surveillance.13,19 Human data has previously correlated both SEMA4D and Plexin B1 expression with nodal metastasis, invasiveness, and worse overall survival in patients who underwent resection for PDAC.25 Analysis of resected tumor specimens demonstrate an avid infiltrate of Semaphorin 4D cells (Figure 3B). Concomitantly, myeloid suppressor cells within the PDAC TME demonstrate avid expression of Semaphorin 4D receptors, Plexin B1 and Plexin B2 (Figure 3B). Thus, Semaphorin 4D blockade represents a novel immunotherapeutic strategy within the PDAC TME.

The trial is carried out in two stages. In Phase 1b, a maximum number of up to 44 patients is accrued, but as few as 16 patients are needed and 1000 simulations of this model anticipates between 24-29 patients. Phase 1b will continue until 16 evaluable patients accrue in a dose. Once 16 patients have accrued to a particular dose based on the BOIN design, the MTD dose is established as the dose yielding 30% or lower DLT rate. If there is 1 or less objective responses among these 16 patients, the study will be stopped early for efficacy futility according to Simon's two-stage design.1,2 If after evaluation of futility the trial is meeting early markers of efficacy (at least 2/16 responses), an expansion cohort will begin. Dose escalation will cease and all patients will subsequently enter the second stage of the trial at the MTD dose defined by Phase 1b. In this expansion cohort, Phase 2, up to 24 patients will be enrolled. These patients will be followed for objective response rates by RECIST 1.1 criteria and iRECIST. When combined with the 16 patients from Phase 1b, the Phase 2 cohort will total an evaluable sample size of 40. If there are 5 or more responses among these 40 patients, we reject the null hypothesis and claim that the treatment is promising. An objective response rate of 23% will afford a 10% type 1 error with 80% power to detect a difference. The null hypothesis is that the true response rate is 0.10 or less, and the alternative hypothesis is that the true response rate is 0.23 or greater. The design controls the type I error rate at 0.1 and yields the power of 0.8.