(P 288) ANGIOA: A SAFETY AND TOLERABILITY STUDY OF TRIPLE ANTI-ANGIOGENIC THERAPY COMBINED WITH PD-L1 BLOCKADE IN PEDIATRIC, ADOLESCENT AND YOUNG ADULT RECURRENT/REFRACTORY SOLID TUMORS

Objective: Children, adolescents, and young adults (AYA) with recurrent/refractory solid tumors face dismal outcomes. Attempts to improve treatment successes by intensifying therapy have largely failed, and new approaches are needed. Single-agent immune checkpoint inhibitors have largely been ineffective, but evidence supports combinatorial approaches with antiangiogenic agents targeting vascular endothelial growth factor (VEGF) signaling as a rational approach. VEGF is upregulated in a subset of pediatric and AYA sarcomas and has been determined to be a mechanism of resistance to standard chemotherapy protocols. Enhanced VEGF expression also correlates with an immunosuppressive microenvironment, a reduction of infiltrating T-cells, and an increase in PD-L1 expression. Collectively, these findings implicate VEGF signaling as a mechanism of therapy resistance. Therefore, the addition of VEGF signaling inhibition to checkpoint blockade may be useful for treating chemotherapy-refractory pediatric and AYA sarcomas.

ANGIOA is a two-part study. Part 1 sought to determine the safety of combining atezolizumab with low-dose cyclophosphamide, bevacizumab, and PK-guided sorafenib in pediatric AYA patients with recurrent, refractory solid tumors. Furthermore, it aimed to establish the feasibility of utilizing PK-guidance to optimize sorafenib (SOR) dosing and implementing an immune scoring system to gauge immune response to therapy. Here we report the findings from Part 1.

Methods: Patients between the ages of 1-30 years with biopsy-accessible, recurrent or refractory solid tumors with acceptable organ status were eligible for enrollment. In addition to standard labs and imaging studies, patients had a tumor biopsy performed prior to the initiation of study treatment and immunohistochemistry (IHC) was performed to quantify CD3+ and CD45RO+ cells. Patients received: atezolizumab (ATZ) IV day 1 (15 mg/kg (max 1200 mg)), bevacizumab (BEV) IV day 1 (15mg/kg), cyclophosphamide (CYCLO) PO days 1-21 (50mg/m2), and SOR PO days 1-21 (target area under the curve (AUC) target: 20-55 hr·µg/ml) of a 21 day cycle. After completion of two cycles, imaging was obtained, and the biopsy was repeated to calculate the ratio of CD45RO+/CD3+ of cells and the total number of CD3+ cells. Treatment was continued if the patient had at least stable disease and/or an increase in immune cell infiltrate with no safety concerns. CTCAE version 5 was used to grade toxicities and the dose-limiting toxicity (DLT) window was considered as the first 2 cycles of therapy. The recommended Phase 2 dose (RP2D) was defined as the dose/exposure level with ≤ 2/6 patients with a DLT and successful AUC target in > 60% of patients.

Results: Patient demographics and tumor characteristics are defined in Table 1. Of the 10 enrolled patients, 9 were evaluable for PK feasibility and 6 were eligible for toxicity and immune response. Reasons for being unevaluable for toxicity were (progression prior to completing 2 cycles (n=2) and physician decision (n=2)). The mean days on therapy was 104 (range 27-289) with 1 patient currently remaining on study. Grade 3 and 4 toxicities for the first two treatment cycles are defined in Table 2. The most common toxicities, including all grades, were leukopenia, lymphopenia, and treatment-associated palmar-plantar erythodesia syndrome. No study-defined DLTs were observed. Of the 43 SOR PK-targeting studies, 37 were within the target range for an 86% targeting success rate. 8/9 patients required a SOR dose adjustment to stay within target range (mean dose after adjustment: 65 mg/m2). Eight patients had evaluable tumor for immunology studies at baseline. The baseline mean absolute CD3+ infiltration and CD45RO+/CD3+ ratio was 49.1 per mm^2 viable tumor (range: 6.8-150) and .28 (0-0.43) respectively. Of patients who were able to have re-biopsy, 5/6 experienced an absolute increase in CD3+ cells and 4/6 had an increase in the CD45RO+/CD3+ ratio (Figure 1).

Conclusion: The combination of low dose cyclophosphamide, PK-guided sorafenib, bevacizumab, and atezolizumab was well tolerated with a recommended phase 2 dose of atezolizumab (day 1, IV, 15 mg/kg (max 1200 mg)), bevacizumab (day 1, IV, 15mg/kg), cyclophosphamide (days 1-21, PO, 50mg/m2), and sorafenib (days 1-21, PO, area under the curve (AUC) target: 20-55 hr·µg/ml). PK-targeting of sorafenib was found to be feasible and resulted in dose changes in most patients. Re-biopsy demonstrated increased immune infiltration with treatment in multiple patients. The phase-2 portion of the study is ongoing.