Misreported Cancer Deaths in PLATO Trial

Abstract

The potential link between antiplatelet agents and anticoagulants with excess cancer deaths (CD) was reported first for prasugrel (TRITON, DAPT), clopidogrel (DAPT), vorapaxar (TRACER), apixaban (APPRAISE-2), and later ticagrelor (PEGASUS). However, verified CD in the ticagrelor indication-seeking PLATO were not public. We obtained the complete list of deaths and their primary causes in PLATO, matched that dataset against local site records, and analyzed the patterns of CD reporting. The FDA-issued spreadsheet contains 31 precisely detailed CD (PLATO code 12-3). We obtained local site evidence for four CD and matched them with FDA-reported. We also assessed the patterns of how CD were reported among non-vascular death database column “S” by scrolling the FDA Excel file down among 938 PLATO entries. Clopidogrel CD (n = 17) were reported exclusively by sponsor, while independent CRO’s reported only ticagrelor CD (3 out of 14 PLATO total). Among four matched verified outcomes, one ticagrelor CD was correct, second ticagrelor CD was unreported, and two (ticagrelor and clopidogrel) CD were reported inaccurately. Of the remaining 16 clopidogrel CD six were reported as three separate next in line paired entries in Denmark (236–237), Poland (597–598), Romania (679–680), and as two more fatalities in South Africa (786) and Spain (789), while patients 787 and 788 received ticagrelor out of 938 records suggesting possible late addition of incorrect clopidogrel CD reports. We conclude that some CD were misreported in PLATO, favoring ticagrelor. Such mismatch may require reevaluation of this critical outcome in the trial focusing on the exact death cause reported by site investigators.