(P 206) A SINGLE-ARM, PHASE II STUDY OF NEOADJUVANT PEMBROLIZUMAB COMBINED WITH RADIOTHERAPY IN THE PRIMARY TREATMENT OF SOFT TISSUE SARCOMAS.

Objective: Relapse after definitive surgery and radiotherapy (RT) is a common problem in STS. Past evidence has suggested potentially beneficial immunotherapeutic effects of RT. Immunotherapy with Programmed Death-1 receptor (PD-1) inhibitors has demonstrated preliminary or confirmed activity in a number of types of sarcomas, including undifferentiated pleomorphic sarcomas, cutaneous angiosarcomas, and alveolar soft part sarcomas. Combined therapy with PD-1 inhibitors and RT may yield enhanced efficacy versus either modality alone. Here, we investigated the feasibility and clinic-pathologic outcomes of concurrent therapy with the PD-1 inhibitor pembrolizumab (PEM) and neoadjuvant RT.

Methods: Patients ≥18 years were eligible if they had localized STS for which both surgical resection and neoadjuvant RT to 50-50.4 Gy (institutional standard) were determined to be feasible and appropriate by experienced sarcoma practitioners. Patients with FNCLCC grade 3 (>1.5 cm) or grade 2 (>3 cm) of any location were eligible. Bone sarcomas and STS for which chemotherapy is routinely administered at our institution were excluded. Patients with metastatic disease at baseline or contraindications to PD-1 inhibitor therapy were excluded.
All patients received 3 doses of PEM (200 mg IV q21 d) concurrently with planned neoadjuvant RT. 4-6 weeks after neoadjuvant therapy, patients underwent planned surgical resection.
The primary endpoint was the proportion of patients with complete pathologic necrosis, defined as ≥95% pathologic necrosis at the time of definitive surgical resection. Enrollment of 26 subjects has 94.4% power to detect an improvement in the complete necrosis rate from 15% to 40%. The observation of at least 7 subjects with ≥95% pathologic necrosis would be considered evidence to rule out a 15% complete necrosis rate.
Secondary endpoints included extent of surgical resection achieved; adverse events (AE) by CTCAE 5.0; local and distant progression-/relapse-free survival (PFS/RFS) from time-of-PEM initiation and from time-of-surgery. Prior to surgery, only confirmed development of distant metastases, local progression preventing surgical resection, or death from any cause were considered to represent events from the perspective of PFS/RFS. After surgery, any documented progression, or death from any cause, were considered events for purposes of RFS endpoints. Survival endpoints were analyzed by the Kaplan-Meier method. Disease response and progression were defined according to RECIST 1.1 criteria.

Results: 26 patients were evaluable; 21 (81%) had trunk/extremity and 5 (19%) had retroperitoneal/visceral disease. Histologies included undifferentiated pleomorphic sarcoma (UPS; 9/35%), dedifferentiated liposarcoma (8/31%), myxofibrosarcoma (2/8%), solitary fibrous tumor (2/8%), and other sarcoma (5/19%). 24 (92%) received all three planned PEM doses; 2 (8%) received only 2 doses (G2 rash, G2 fatigue). All completed planned RT.
Two UPS patients had objective partial responses at pre-operative imaging (ORR=8%). Progressive disease prior to surgery occurred in 3: 2 (8%) with local progression only and 1 (4%) with metastatic pulmonary disease. All three proceeded with surgery. All remaining patients had stable disease.
Median pathologic primary tumor size was 10 cm (range 2.1-26 cm). 14 (54%) patients had confirmed FNCLCC grade 3 tumors at baseline, while the remaining 12 (46%) were ≥grade 2. Negative margins were achieved in 23/26 (88%).
Median follow-up time at data cut-off (19 June 2023) was 18.8 months (m) from PEM initiation (PI) (range 2.0-59.2 m). At that time, 3 patients had no post-operative imaging and 4 patients had died: sarcoma progression (2), co-morbidity (1), ischemic colitis (1).
Median pathologic necrosis was 33% (range 0-99%; n=24), with 4 (17%) demonstrating ≥95% pathologic necrosis, the primary endpoint. Median percent viable cells was 25% (range 1-100; n=21), with 6 (29%) demonstrating less than 10% viable.
There were 11 PFS/RFS events in the evaluable population from PI and 10 RFS events from surgery. Median PFS/RFS of all evaluable PEM-treated patients from PI was 17.2 m (95% CI 7.6-NR m). Median RFS from surgery for those without metastatic disease pre-operatively and at least one post-operative imaging assessment was 14.6 m (95% CI 5.1-NR m). All relapses were distant.
There were no grade 5 AE’s judged to be related to study therapy. AE’s were manageable and consistent with the anticipated effects of PEM and RT.

Conclusion: Neoadjuvant therapy combining PEM and RT is feasible in the management of STS, including in retroperitoneal sarcomas. AEs were consistent with known PEM and RT effects, without evidence of synergism. This study did not achieve its primary endpoint, with only 4 patients demonstrating ≥95% pathologic necrosis. Objective responses were documented pre-operatively in two patients, both with UPS.

Updated outcomes analyses and preliminary results of correlative studies will be presented.