Strauss, Sandra J.

11562 Background: Afami-cel is an autologous, HLA-A*02-restricted, specific peptide enhanced affinity receptor, T-cell therapy engineered to target MAGE-A4+ solid tumors. The pivotal, 2-cohort, single-arm, Phase 2, SPEARHEAD-1 trial (NCT04044768) with afami-cel met its primary endpoint based on Cohort 1 data. As of September 1, 2021, in 47 patients (pts) with metastatic synovial sarcoma (SyS) or myxoid/round cell liposarcoma (MRCLS), the overall response rate (ORR) per independent review was 34% with encouraging durability (Van Tine, et al. Paper 30: CTOS 2021; Virtual). To identify potential stratification factors for response and assess whether response is a proxy for progression-free survival (PFS), we present pooled analyses using data from the prior Phase 1 trial (NCT03132922) and Cohort 1 of the SPEARHEAD-1 trial. Methods: Eligible pts (16–75 years) were HLA-A*02+ with MAGE-A4+ tumors. Pts received afami-cel after lymphodepleting chemotherapy. The pooled analyses evaluated ORR per RECIST v1.1 by investigator review, stratified by 7 factors, and safety. Results: In the pooled data, 69 pts received afami-cel (2.12–9.99×109 transduced T-cells) and were evaluable for response (Phase 1, n = 18; Phase 2, n = 51); all expressed one eligible HLA-A*02 allele. Median (range) for: age was 42 years (19–76), number of prior lines of therapy was 2 (1–12), and tumor MAGE-A4 H-score was 230 (60–300). Median (range) H-score was higher in SyS (256 [60–300]) than in MRCLS (180 [112–230]). The pooled investigator-assessed ORR was 36.2% (40.7% in SyS; 10.0% in MRCLS). Responses occurred across a wide MAGE-A4 H-score range (134–300). Median (range) duration of response was 52 weeks (8.29–75.14). Response rate was higher in the 59 pts with SyS: with ≤2 vs ≥3 prior lines of therapy (55.2% vs 26.7%), baseline target lesion sum of longest diameters <10cm vs ≥10cm (53.1% vs 25.9%), MAGE-A4 H-score ≥200 vs <200 (46.3% vs 27.8%), without vs with bridging therapy (48.6% vs 29.2%), who were female vs male (46.4% vs 35.5%), aged ≥40 vs <40 years (45.7% vs 33.3%), and from North America vs Europe (42.6% vs 33.3%). In responders vs non-responders with SyS, respectively, median PFS was 58.3 vs 11. 0 weeks (log-rank p-value <0.0001); the probability of being progression-free at 24 weeks was 0.8 vs 0.2. The pooled benefit:risk profile of afami-cel was similar to that in the SPEARHEAD-1 trial (Van Tine, et al. Paper 30: CTOS 2021; Virtual.). Conclusions: We show that baseline tumor burden, prior systemic treatment history, and MAGE-A4 tumor expression levels are potential factors associated with response to afami-cel, although their true predictive value for response status awaits confirmation. Our findings will inform the ongoing clinical development of afami-cel in sarcoma, especially for prognostic studies with PFS or overall survival endpoints. Clinical trial information: NCT04044768, NCT03132922.

11504 Background: This phase 2, open-label trial (SPEARHEAD-1; NCT04044768) is designed to evaluate the efficacy, safety, and tolerability of afamitresgene autoleucel in 45 patients (pts) with advanced/metastatic synovial sarcoma or Myxoid/Round Cell Liposarcoma (MRCLS). Methods: Eligible pts are HLA-A*02 positive with MAGE-A4-expressing tumors. Pts undergo leukapheresis for collection of autologous T-cells for processing and manufacture into afamitresgene autoleucel cells. Pts were treated with afamitresgene autoleucel doses between 1–10 × 109 transduced T-cells after receiving lymphodepleting chemotherapy. The primary endpoint is overall response rate per RECIST v1.1 by independent review. An independent Data Safety Monitoring Board reviews ongoing safety and benefit: risk during the interventional phase. Results: As of Feb 4, 2021, 32 pts received afamitresgene autoleucel. Of these pts, 59% were male, 87.5% had synovial sarcoma, the median age was 43 yrs (range: 24–73), and they had a median of 3 prior systemic lines of therapy. The MAGE-A4 antigen expression level (histoscore) ranged from 112–300, and the transduced cell dose ranged from 2.7–9.9 x 109. At the data cutoff, 25 pts were evaluable for preliminary efficacy (23 with synovial sarcoma and 2 with MRCLS) and 7 pts (5 with synovial sarcoma and 2 with MRCLS) had insufficient follow-up (<8 weeks follow-up and/or awaiting first scan). Of the 25 evaluable pts, the investigator-assessed responses were: complete response (2 pts), partial response (8 pts), stable disease (11 pts), and progressive disease (4 pts). All responses were confirmed. Nine of the 10 responders had ongoing response at the data cutoff and 3 responders had MAGE-A4 antigen histoscores <200. The most common AEs of any grade (>30% pts) were neutropenia, lymphopenia, nausea, cytokine release syndrome, leukopenia, fatigue, pyrexia, and anemia. Cytokine release syndrome of any grade occurred in 19/32 pts; 95% of those events were ≤Grade 2. No immune effector cell-associated neurotoxicity syndrome (ICANS) has been reported to date. Cytopenia (≥G3) at 4 wks post-infusion was observed in 6 pts (anemia 3 pts, neutropenia 2 pts, and thrombocytopenia 1 pt). Conclusion: These preliminary data demonstrate afamitresgene autoleucel is efficacious and well tolerated in heavily pre-treated pts. Objective responses are reported across a wide range of MAGE-A4 antigen levels and deep responses have been observed. Initial durability data is encouraging. Preliminary response data in SPEARHEAD-1 is comparable to the findings of the prior Phase 1 trial [1]. To date, the safety profile of afamitresgene autoleucel has been favorable, with mainly low-grade cytokine release syndrome and tolerable/reversible hematologic toxicities. [1]. Van Tine BA, et al. CTOS; November 18-21, 2020; Virtual. Clinical trial information: NCT04044768.