Is It Time to Resist Using RECIST as a Primary Endpoint for Rare Tumor Trials?

Epithelioid hemangioendothelioma is an ultra-rare cancer driven by YAP-CAMTA1 fusion. Based on the link of the fusion to the MEK pathway, SARC33 was performed. It is a phase 2 trial examining trametinib that missed its primary objective by RECIST but demonstrated patient-reported outcome benefits in improved pain. See related article by Schuetze et al., p. 4584.

A phase II study of pazopanib as front-line therapy in patients with non-resectable or metastatic soft-tissue sarcomas who are not candidates for chemotherapy.

BACKGROUND: Cytotoxic chemotherapy remains the standard of care first-line treatment for advanced and metastatic soft-tissue sarcomas (STSs). Certain patients may not be chemotherapy candidates based upon age or co-morbidities, leaving limited treatment options. Pazopanib is a multi-targeted tyrosine kinase inhibitor that is FDA-approved for metastatic STS after the first line. We proposed a phase II study evaluating pazopanib as a first-line agent in patients with advanced disease who are deemed not to be candidates for chemotherapy. METHODS: Eligible patients were at least 18 years old, not candidates for chemotherapyand were treatment naive. Pazopanib was titrated from 200 mg twice daily to a goal of 800 mg daily. The primary end point was the clinical benefit rate (CBR) (CBR = completed response + partial response + stable disease per Response Evaluation Criteria in Solid Tumours [RECIST 1.1]) at 16 weeks. The sample size of 56 evaluable patients was calculated to provide 80% power to test a hypothesised CBR of ≥35% against an unfavourable CBR of ≤20%. If ≥ 17 patients achieved benefit, the null CBR of 20% would be rejected at a nominal 5% alpha level. Secondary end points included progression-free survival (PFS), overall survival (OS), quality of life and serum biomarkers. FINDINGS: Fifty-six patients were enrolled from May 2015 to February 2019 and are included in the intention-to-treat analysis. Median PFS was 3.67 (2.62-7.25) months. Median OS was 14.16 (95% confidence interval [CI]: 8.4-NR) months, CBR = 39.29% (22/56) (CI = 0.265-0.533, p = 0.0007). No new or unexpected adverse events were seen. The most common grade I-II events were diarrhoea, nausea and fatigue. The most common grade III-IV events were hypertension and liver function test abnormalities. INTERPRETATION: These data suggest that there is a benefit to front-line pazopanib in patients with STS who are not candidates for cytotoxic chemotherapy.