Re-analyzing randomized trials with incorporation of COVID-19 risk associated with cancer therapy

ABSTRACT

Background: Cancer therapy may put patients at risk of mortality from COVID-19. The impact of abbreviated treatment courses on outcomes in the setting of COVID-19 is unknown. We incorporated COVID-19-associated risks in re-analysis of practice-defining randomized trials in oncology that compared different radiation therapy (RT) regimens. Methods: We extracted individual patient level data (IPLD) from published survival curves from randomized trials in rectal cancer (Dutch TME, TROG 01.04), early stage breast cancer (CALGB 9343, OCOG hypofractionation trial, FAST-Forward, NSABP B-39), and localized prostate cancer (CHHiP, HYPO-RT-PC). Trials were simulated with incorporation of varying risk of SARS-CoV-2 infection and mortality associated with receipt of therapy. Results: IPLD from 14,170 patients were re-analyzed. In scenarios with low COVID-19-associated risks (0.5% infection risk per fraction [IRF], 5% case fatality rate [CFR]), fractionation did not significantly affect outcomes. In locally advanced rectal cancer, short-course RT appeared preferable to long-course chemoradiation (TROG 01.04) or RT omission (Dutch TME) in most settings. While moderate hypofractionation in early stage breast cancer (OCOG hypofractionation trial) and prostate cancer (CHHiP) was not associated with survival benefits in the setting of COVID-19, more aggressive hypofractionation (FAST-Forward, HYPO-RT-PC) and accelerated partial breast irradiation (NSABP B-39) were associated with improved survival in higher risk scenarios (≥5% IRF;≥ 20% CFR). In settings where RT can be omitted, such as favorable early stage breast cancer in the elderly (CALGB 9343), RT was associated with worse survival in higher risk pandemic scenarios (≥5% IRF, ≥ 20% CFR). Conclusions: Our framework, which can be adapted to dynamic changes in COVID-19 risk, provides a flexible, quantitative approach to assess the impact of treatment recommendations across oncology. The magnitude of potential benefit from abbreviated RT courses depends on the degree of hypofractionation and local COVID-19-associated risk. Abbreviated RT courses should be prioritized when possible and are increasingly beneficial in higher risk pandemic settings. With increased understanding and precautions against COVID-19 that can minimize risks for patients, our results support the continued use of evidence-based treatments for cancer patients in the COVID-19 era.