Challenges with COVID-19 vaccine developmentand assessing impact of the pandemic on oncology trial outcomes

ABSTRACT

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes a respiratory disease called coronavirus disease (COVID)-19 in infected individuals. Since its first outbreak at the end of 2019, COVID-19 has rapidly evolved into a global pandemic and has infected more than 85 million (M) people worldwide and has resulted in over 1.85 M deaths. The pandemic has led to the race for finding therapeutics and vaccines, and to date, there have been two vaccines approved within the United States and the potential for more to follow in the coming months. In addition, this has also forced clinical researchers to look into the impact this may have on ongoing trials, given the risk of COVIDrelated complications confounding with the efficacy and safety outcomes and necessitating the collection of additional data and evaluating alternate ways of analyzing trial results. This is of particular concern in patients with cancer who are often immunosuppressed as a result of both their disease and the treatment they receive, which puts them at increased risk of severe complications from COVID-19. In this session, we will focus on two important but very distinct topics that have emerged due to the onset of the pandemic (1) the issues and challenges related to vaccine development including trial design, rapid execution, real-time regulatory feedback, and overall interpretation of trial results and (2) understanding potential impact of COVID-related deaths in Oncology trials, on time to event endpoints of overall survival and progression-free survival. Topic 1 Challenges with COVID-19 Vaccine development The SARS-CoV-2, first identified in December 2019, has caused a worldwide pandemic leading to widespread morbidity and mortality. There was no Food and Drug Administration-approved vaccine for the prevention of the coronavirus disease 2019 (COVID-19), the disease caused by SARS-CoV-2. The urgent need for safe and effective interventions to mitigate the global spread of SARS-CoV-2 has prompted international efforts to develop antivirals and vaccines. Numerous vaccine candidates based on traditional and new platforms are currently being evaluated including nucleic acid (DNA and RNA), viral vector (replicating and non-replicating), virus-like particles, peptide-based, recombinant protein, and live attenuated and inactivated virus modalities. COVID-19 vaccines are in various stages of clinical development, with several candidates in pivotal phase 3 clinical trials, including mRNA-based vaccines, of which two vaccines, one developed by Pfizer and the other by Moderna recently received the emergency use authorization from the Food and Drug Administration in December 2020. In this session, we will share experiences and challenges in the development of these two vaccines. We will share the practical and statistical challenges and considerations when designing COVID-19 vaccine studies, such as target population(s), endpoint selection and assessment, statistical analysis method, timing of interim analyses, and questions continuously to be answered after receiving the emergency use authorization and/or during the Biologics License Application review. We show that, when planning a vaccine pivotal study against a novel virus causing ongoing worldwide pandemic, special consideration needs to be given for the designing of interim analyses related to efficacy, so that a vaccine with favorable benefit-risk profile can be made available as early as possible;special attention also needs to be given for an independent data safety monitoring board for vaccine-associated enhanced respiratory disease and other safety signal monitoring to mitigate the risk for trial participants during an ongoing pandemic. Topic 2 Impact of COVID-19 on Oncology trial outcomes using overall survival or progression-free survival Simulations were conducted to assess (1) the impact of the COVID-19-related death and (2) missed RECIST visits on the statistical analysis of time-to-event outcomes in randomized phase 3 oncology trials and explore mitigation options for this r sk when COVID- 19 cohort is well-defined. Two simulated case studies of Phase 3 randomized controlled trials ongoing during the pandemic outbreak were used to evaluate five approaches (1, ITT approach;2, modified ITT excluding COVID-19-related deaths;3, censoring COVID-19- related deaths with target number of non-COVID-19- related deaths;4, censoring COVID-19-related deaths with original date-cut-off;5, Fine & Gray modeling of competing risk approach to treat COVID-19-related death and non-COVID-19-related death separately), at presence of COVID-19-related deaths with different pandemic onsets relative to timing of analysis and varying pandemic durations, the impact of COVID-19- related death due to pandemic, on the primary endpoints of overall survival and/or progression-free survival in terms of type 1 error, power, and hazard ratio estimates. It was found that COVID-19-related deaths would impact time-to-event analysis in terms of type 1 error and power for log rank test, and provide biased treatment effect estimation from Cox model if ITT approach is used;impact would be more severe if there was an imbalance in COVID-19-related deaths mainly in experimental arm. With same number of COVID-19- related deaths, the earlier the timing of the pandemic window, will lead to greater loss in power. Approaches censoring COVID-19-related deaths would minimize impact on power loss and bias in hazard ratio estimation, particularly if data cut-off was extended to mitigate for events loss due to censoring. The simulations conducted in this analysis provide a framework to help understand how to mitigate the risk to the randomized oncology trials in which COVID-19- related deaths are observed in the blinded assessment during the COVID-19 pandemic.