(P 152) FINAL RESULTS OF GEMMK: A PHASE I TRIAL OF GEMCITABINE AND PEMBROLIZUMAB IN PATIENTS WITH LEIOMYOSARCOMA AND UNDIFFERENTIATED PLEOMORPHIC SARCOMA

Objective: Patients with advanced leiomyosarcoma (LMS) and undifferentiated pleomorphic sarcoma (UPS) have poor outcomes. Single agent checkpoint inhibition has shown limited activity in sarcomas. The objective of this phase I trial (NCT03123276) was to determine the safety and maximum tolerated dose (MTD) of the combination of gemcitabine (G) and pembrolizumab (P) in advanced LMS and UPS patients (pts).

Methods: In Part A (dose escalation) a traditional 3+3 design with 3 G doses (800, 1000 and 1200 mg/m2) was employed: G and P (200 mg) were respectively administered on Days 1,8 (G) and Day 1 (P) every 3 weeks for 6–8 cycles. Pts with ongoing disease, disease stabilisation or response received maintenance with P on day 1 every 3 weeks to a maximum of 2 years. In part B (dose expansion) additional pts were enrolled at the MTD identified in part A. Pre- and on-treatment biopsies were mandated. The primary endpoint was safety (CTCAE v4.0). Secondary endpoints included radiological response (RECIST v1.1) at 9 weeks and median progression-free survival (mPFS) and are presented with 95% confidence intervals (CI). Stata v14 was used for all statistical analyses.

Results: In Part A, 13 pts (2 UPS, 11 LMS) were enrolled. One dose-limiting toxicity (DLT) was observed at G dose of 1000 mg/m2 (neutropaenia) but was not confirmed in the 1000 mg/m2 expansion group. There were no DLTs observed at G dose of 1200 mg/m2. This was selected as the MTD carried forward in the expansion phase. The MTD of G was not reached; recommended G dose for Part B was 1200 mg/m2. In Part B, 12 additional pts (4 UPS, 8 LMS) were enrolled. Pts’ characteristics in the whole cohort were: the median age was 56 years (range 39-78); 19 pts (76.0%) were female; 18 pts (72.0%) had an Eastern Cooperative Oncology Group performance status of 0; the median number of previous treatments was 1 (range 1–4).

Overall, 20/25 pts (80.0%) experienced at least one adverse event (AE) grade 3 or higher as per CTCAE v4.0. A total of 49 AEs grade 3 or higher were recorded, the most common being neutropaenia (9/49, 18.4%). Eighteen of the 49 (36.7%) AEs of grade 3 or higher were deemed unrelated or unlikely to be related to treatment. Three pts discontinued treatment due to AEs (2 immune-related AEs, 1 treatment-unrelated AE).
Twelve pts (48.0%) experienced at least one serious AEs (SAEs). A total of 31 SAEs were recorded, with fever being the most common (12/31, 38.7%). Febrile neutropaenia did not occur. Fourteen of the 31 (45.2%) SAEs were deemed unrelated or unlikely to be related to the treatment.
Common AEs of any grade reported in more than 35% of the pts included: rash (19/25, 76.0%); fatigue (18/25, 72%); nausea (13/25, 52%); neutropaenia (12/25, 48.0%); abdominal pain (11/25, 44.0%); hypothyroidism, vomiting, ALT increase (each 10/25, 40.0%); diarrhoea and fever (each 9/25, 36%).

At 9 weeks, objective radiological responses were seen in 1/19 (5%, 95%CI 1–33%) LMS pts and 4/6 (67%, 95%CI 22–93%) UPS pts. After a median follow up of 25.5 months, mPFS in the whole population was 7.0 months (95% CI 5.3-8.5).

Conclusion: The combination of G and P is a safe treatment for LMS and UPS pts. No unexpected safety signals were observed at the recommended dose of G 1200 mg/m2. Although the number of treated pts is relatively small, promising preliminary activity was seen in the UPS subgroup, warranting further evaluation in this subtype. Long-term follow-up and translational analyses are ongoing, including immunophenotyping of immune cells in pre- and post-treatment biopsies, evaluation of spatial distribution of immune cells relative to the microvasculature, and characterisation of PD-L1 expression.