Viewpoint: paradoxical excess mortality in the PLATO trial should be independently verified

Thromb Haemost. 2011 May;105(5):752-9. doi: 10.1160/TH10-12-0807. Epub 2011 Mar 8.

Abstract

The PLATO trial revealed excess all-cause (4.5%) and vascular (4.0%) mortality after experimental pyrimidine, ticagrelor, and even higher death rates (5.9% and 5.1%, respectively) after clopidogrel, which have never been seen in any previous acute coronary syndrome (ACS) trial. The Food and Drug Administration (FDA) conducted, and recently released the ticagrelor review outlining some paradoxical mortality patterns in PLATO, including the existence of alive patient, who initially was reported dead. The drug was recently approved in Europe, but repeatedly delayed in the USA. The objective of this viewpoint article was to evaluate extremely high death rates in PLATO by scrutinising FDA-released evidence, and comparing mortality patterns in recent ACS trials. These data were first presented as the analytical report submitted to the FDA on October 26, 2010. The available evidence suggest that mortality rates in PLATO, so as death benefit of ticagrelor over clopidogrel are extreme, despite incomplete follow-up, short duration of the trial, frequent preloading with clopidogrel, and gross mismatch between conventional average myocardial infarction rates but disproportionally frequent vascular fatalities, and heavily imbalanced sepsis-related deaths. In contrast to the overall PLATO results, the deaths rates in the USA were much lower (3.2% vs. 3.8%) not only favouring clopidogrel, but more importanly matching very well with identical rates in TRITON (3.2%), and one-year ACUITY (3.6%-3.9%) fatalities. Since the «play of chance» cannot explain these discrepancies due to excess death rates in both PLATO arms, and considering that study sponsor self-monitored sites in most countries, but not in the USA, the mortality data are questionable, and should be independently virified. It was concluded that excess mortality rates and delayed timing of the benefit onset in PLATO do not match with any recent ACS trial, and do not look natural. Reevaluation of the survival, especially driven from the several high-volume sponsor monitored sites in Eastern Europe may reveal discrepancies between those reported in PLATO and actual vital records. Future practice of self monitoring in pivotal indication-seeking clinical trials should be completely banned.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Acute Coronary Syndrome / drug therapy*
  • Acute Coronary Syndrome / epidemiology*
  • Acute Coronary Syndrome / mortality*
  • Adenosine / analogs & derivatives*
  • Adenosine / therapeutic use
  • Bias
  • Clinical Trials as Topic
  • Clopidogrel
  • Europe / epidemiology
  • Humans
  • Mortality
  • Survival Rate
  • Ticagrelor
  • Ticlopidine / analogs & derivatives*
  • Ticlopidine / therapeutic use
  • Treatment Outcome
  • United States / epidemiology
  • United States Food and Drug Administration

Substances

  • Clopidogrel
  • Ticagrelor
  • Adenosine
  • Ticlopidine