EDITOR:
Ridley and colleagues present retrospective data on the use of activated protein C (APC) in five UK hospitals in 2002–2005 [Reference Ridley, Lwin and Wyncoll1]. To identify patient groups that might benefit from APC the authors considered it more rational to use their cited approach rather than have further formal appraisal of the drug. The authors discuss the European Medicine Agency (EMEA) 2002 approval but, surprisingly, their crucial decision in early 2007 is not mentioned: the EMEA demanded a new randomized, placebo controlled trial of APC in severe sepsis to clarify the risk/benefit balance of APC [2] in the currently indicated high-risk population [3].
This EMEA opinion, on the most rational way to proceed, reflects results of the trials requested by the FDA. The PROWESS follow-up trial showed no significant difference in the number of patients discharged home between APC and placebo groups [Reference Laterre, Levy and Clermont4]. The authors suggest that ENHANCE confirmed drug efficacy and safety. However, this has been seriously questioned: as a non-randomized, uncontrolled trial it was not designed to assess efficacy; compared with PROWESS the serious bleeding rates were greatly increased in ENHANCE (a NNT of 16 for a serious bleed) [Reference Farmer5]. As the authors indicate, ADDRESS did not help APC and the drug’s failure to demonstrate efficacy and safety in paediatric sepsis [Reference Nadel, Goldstein and Williams6] raised further doubts around APC having any beneficial effect, even in adults [Reference Opal7].
It may be a particular concern that the authors’ data included a relatively high proportion of surgical patients (50%) vs. PROWESS (455/1690: 27%). Original Phase II and PROWESS trial data showed a higher mortality in the surgical patients randomized to APC [8]. A later retrospective reclassification of PROWESS patients may be difficult to interpret [Reference Barie, Williams and McCollam9]. With treated surgical patients in ADDRESS also having a higher mortality, there is a consistent trend towards poorer outcome, with APC administration, in almost 1500 surgical patients enrolled to all three placebo controlled trials in adults. It may be hazardous to imply that APC has an acceptable benefit/risk profile in this group of patients.
Finally, data from a small number of hospitals in 2002–2005 may not reflect current UK practice. Following recent trial results use may now be very low as in other countries: recent French data (2006) showed that APC was not used in 14 of the 15 ICUs surveyed [Reference Muller, Jaber, Raillard and Lefrant10]. Low usage will probably persist unless new trials demonstrate clear benefit in easily identifiable patient groups.
