2536 Background: Cell-surface protein PSCA is upregulated in many solid tumors and correlates with disease stage. BPX-601, an autologous T-cell product expressing a PSCA-CD3ζ CAR and a rimiducid (Rim)-inducible MyD88/CD40 co-activation switch to augment T-cell proliferation and persistence, is designed to have enhanced efficacy in solid tumors vs traditional CARs. This ongoing first-in-human study assesses safety, biologic, and clinical activity of BPX-601+Rim in PSCA+ cancers. Updated results, including those from patients (pts) who underwent LD with Flu/Cy, are presented. Methods: BP-012 is a 2-part, open-label trial. Part 1 is a 3+3 dose escalation of BPX-601 (1.25–5.0x106 cells/kg; Day [D] 0) given prior to a single, fixed Rim dose (0.4 mg/kg; D7) in pts with previously treated PSCA+ metastatic pancreatic, gastric, or prostate cancers with measurable disease. Results: As of Jan-22-2019, 15 pts have received BPX-601±Rim. Two pts at the highest cell dose received Flu/Cy for LD on D−5 to D−3 before BPX-601; LD after Flu/Cy was 96.6% and 84.3%. Thirteen pts received Cy alone on D−3; in these pts, LD ranged from 0–68.6%. Rapid cell expansion by D4 was observed in all pts with peak vector copy number 8.3-fold higher with Flu/Cy (n = 2) vs Cy LD (n = 13). Serum IP-10, IL-6 and TNFα increased > 2-fold from baseline in ≥1 pt in all Rim cohorts, with 3- to 20-fold Rim-dependent cell expansion in 6 pts. No CRS or DLTs were reported. After Rim, one Flu/Cy pt experienced a serious Grade 2 AE (encephalopathy) related to BPX-601+Rim that resolved with IV steroids; despite time-matched nonserious Grade 1 pyrexia, the pt had no other CRS symptoms. After BPX-601+Rim and ≥1 scan, best responses were 8 SD and 3 PD (1 non-evaluable). With a median follow-up of 9.8 wks, time to next treatment (tx) after BPX-601 ranged from 2.7–22.1 wks (n = 8) and ongoing tx-free intervals range from 9.1–30.1 wks (n = 4). Conclusions: BPX-601+Rim was well-tolerated with manageable safety and early evidence of enhanced CAR T-cell expansion and prolonged persistence after Flu/Cy vs Cy. Additional pts will undergo Flu/Cy LD prior to BPX-601 with single- and repeat-dose Rim. Clinical trial information: NCT02744287.
144 Background: The combination of anti-PD-1 and anti-CTLA-4 is a promising strategy that is being clinically explored to treat a variety of cancer types. Some patients display primary resistance to this combination treatment, while others relapse after treatment. Although some cancer patients have long-term durable responses to the combination therapy of anti-PD-1 and anti-CTLA-4, physicians have been looking for biomarkers that predict response. Methods: Using multiple immunocompetent syngeneic and K-rasLA1/+p53R172H?g/+ spontaneous animal models of lung cancer, we have explored the mechanisms of resistance to anti-PD-1/CTLA-4 combination therapy by evaluating the molecular and cellular immune profiles of the tumor microenvironment. Antibody-mediated cell depletion assays were also conducted to validate the mechanisms. Lastly, to determine the applicability to patients, we analyzed 791 lung cancer patients’ specimens with immunohistochemistry staining and mined many immune markers in multiple large independent patient databases (~1900 tumors). Results: We observed that tumor bearing mice treated with combined PD-1 and CTLA-4 blocking antibodies developed resistance through the up-regulation of CD38, and that targeting CD38 abolished the resistance in a manner dependent on B7-costimulation. Additionally, combined PD-1 and CTLA-4 blockade eradicates CD38-deficient tumors in mice. Further study revealed that control of tumors by triple blockade of CD38, PD-1, and CTLA-4 depended on improved functionality of CD4+/CD8+ TILs, which were reactivated by enriched CD103+ dendritic cells in the tumor microenvironment. Pathologic analysis revealed positive immunohistochemical staining for CD38 on tumor cells in 15-23% of cases and bioinformatic analyses revealed a strong correlation between CD38 expression and an immune inflammatory signature. Conclusions: Targeting CD38 improves the efficacy of anti-PD-1/CTLA-4 combination therapy in lung cancer. CD38 on tumor cells could potentially serve as a novel biomarker of resistance for immune checkpoint inhibition.