Background: Statistical testing in phase III clinical trials is subject to chance errors, which can lead to false conclusions with substantial clinical and economic consequences for patients and society. Methods: We collected summary data for the primary endpoints of overall survival (OS) and progression-related survival (PRS) (eg, time to other type of event) for industry-sponsored, randomized, phase III superiority oncology trials from 2008 through 2017. Using an empirical Bayes methodology, we estimated the number of false-positive and false-negative errors in these trials and the errors under alternative P value thresholds and/or sample sizes. Results: We analyzed 187 OS and 216 PRS endpoints from 362 trials. Among 56 OS endpoints that achieved statistical significance, the true efficacy of experimental therapies failed to reach the projected effect size in 33 cases (58.4% false-positives). Among 131 OS endpoints that did not achieve statistical significance, the true efficacy of experimental therapies reached the projected effect size in 1 case (0.9% false-negatives). For PRS endpoints, there were 34 (24.5%) false-positives and 3 (4.2%) false-negatives. Applying an alternative P value threshold and/or sample size could reduce false-positive errors and slightly increase false-negative errors. Conclusions: Current statistical approaches detect almost all truly effective oncologic therapies studied in phase III trials, but they generate many false-positives. Adjusting testing procedures in phase III trials is numerically favorable but practically infeasible. The root of the problem is the large number of ineffective therapies being studied in phase III trials. Innovative strategies are needed to efficiently identify which new therapies merit phase III testing.
Changyu Shen, Enrico G. Ferro, Huiping Xu, Daniel B. Kramer, Rushad Patell, and Dhruv S. Kazi
Sarju Ganatra, Sourbha S. Dani, Robert Redd, Kimberly Rieger-Christ, Rushin Patel, Rohan Parikh, Aarti Asnani, Vigyan Bang, Katherine Shreyder, Simarjeet S. Brar, Amitoj Singh, Dhruv S. Kazi, Avirup Guha, Salim S. Hayek, Ana Barac, Krishna S. Gunturu, Corrine Zarwan, Anne C. Mosenthal, Shakeeb A. Yunus, Amudha Kumar, Jaymin M. Patel, Richard D. Patten, David M. Venesy, Sachin P. Shah, Frederic S. Resnic, Anju Nohria, and Suzanne J. Baron
Background: Cancer and cardiovascular disease (CVD) are independently associated with adverse outcomes in patients with COVID-19. However, outcomes in patients with COVID-19 with both cancer and comorbid CVD are unknown. Methods: This retrospective study included 2,476 patients who tested positive for SARS-CoV-2 at 4 Massachusetts hospitals between March 11 and May 21, 2020. Patients were stratified by a history of either cancer (n=195) or CVD (n=414) and subsequently by the presence of both cancer and CVD (n=82). We compared outcomes between patients with and without cancer and patients with both cancer and CVD compared with patients with either condition alone. The primary endpoint was COVID-19–associated severe disease, defined as a composite of the need for mechanical ventilation, shock, or death. Secondary endpoints included death, shock, need for mechanical ventilation, need for supplemental oxygen, arrhythmia, venous thromboembolism, encephalopathy, abnormal troponin level, and length of stay. Results: Multivariable analysis identified cancer as an independent predictor of COVID-19–associated severe disease among all infected patients. Patients with cancer were more likely to develop COVID-19–associated severe disease than were those without cancer (hazard ratio [HR], 2.02; 95% CI, 1.53–2.68; P<.001). Furthermore, patients with both cancer and CVD had a higher likelihood of COVID-19–associated severe disease compared with those with either cancer (HR, 1.86; 95% CI, 1.11–3.10; P=.02) or CVD (HR, 1.79; 95% CI, 1.21–2.66; P=.004) alone. Patients died more frequently if they had both cancer and CVD compared with either cancer (35% vs 17%; P=.004) or CVD (35% vs 21%; P=.009) alone. Arrhythmias and encephalopathy were also more frequent in patients with both cancer and CVD compared with those with cancer alone. Conclusions: Patients with a history of both cancer and CVD are at significantly higher risk of experiencing COVID-19–associated adverse outcomes. Aggressive public health measures are needed to mitigate the risks of COVID-19 infection in this vulnerable patient population.