Elevated Ratio of Urinary Metabolites of Thromboxane and Prostacyclin Is Associated with Adverse Cardiovascular Events in ADAPT

Thomas J Montine,Ginger Milne,John C S Breitner,Joshua A Sonnen,Laura D Baker,Mark A Smith

doi:10.1371/journal.pone.0009340

Abstract

Results from prevention trials, including the Alzheimer's Disease Anti-inflammatory Prevention Trial (ADAPT), have fueled discussion about the cardiovascular (CV) risks associated with non-steroidal anti-inflammatory drugs (NSAIDs). We tested the hypotheses that (i) adverse CV events reported among ADAPT participants (aged 70 years and older) are associated with increased ratio of urine 11-dehydrothromboxane B2 (Tx-M) to 2′3-donor–6-keto-PGF1 (PGI-M) attributable to NSAID treatments; (ii) coincident use of aspirin (ASA) would attenuate NSAID-induced changes in Tx-M/PGI-M ratio; and (iii) use of NSAIDs and/or ASA would not alter urine or plasma concentrations of F2-isoprostanes (IsoPs), in vivo biomarkers of free radical damage. We quantified urine Tx-M and PGI-M, and urine and plasma F2-IsoPs from 315 ADAPT participants using stable isotope dilution assays with gas chromatography/mass spectrometry, and analyzed these data by randomized drug assignment and self-report compliance as well as ASA use. Adverse CV events were significantly associated with higher urine Tx-M/PGI-M ratio, which seemed to derive mainly from lowered PGI-M. Participants taking ASA alone had reduced urine Tx-M/PGI-M compared to no ASA or NSAID; however, participants taking NSAIDs plus ASA did not have reduced urine Tx-M/PGI-M ratio compared to NSAIDs alone. Neither NSAID nor ASA use altered plasma or urine F2-IsoPs. These data suggest a possible mechanism for the increased risk of CV events reported in ADAPT participants assigned to NSAIDs, and suggest that the changes in the Tx-M/PGI-M ratio was not substantively mitigated by coincident use of ASA in individuals 70 years or older.

Highlights

Prostaglandin (PG) H synthase- or cyclooxygenase (COX) 2 inhibitors, called coxibs, were developed as an alternative for treatment of arthritis in patients who could not tolerate traditional (t) non-steroidal anti-inflammatory drugs (NSAIDs)
Production of TxA2 is closely linked to COX-1 activity while production of PGI2 is more closely linked to COX-2 activity, with aspirin preferentially inhibiting COX-1 activity and t-NSAIDs inhibiting both COX isozymes without bioactivation [7,8,9,10]
We sought to test three hypotheses: (i) that CV adverse events in Alzheimer’s Disease Anti-inflammatory Prevention Trial (ADAPT) participants were associated with an increase in the Tx-M/PGIM ratio, (ii) that use of ASA might ameliorate NSAID-induced changes in urine Tx-M/PGI-M ratio, and (iii) that use of NSAIDs and/or ASA would alter urine or plasma F2-IsoP concentrations

Summary

Introduction

Prostaglandin (PG) H synthase- or cyclooxygenase (COX) 2 inhibitors, called coxibs, were developed as an alternative for treatment of arthritis in patients who could not tolerate traditional (t) non-steroidal anti-inflammatory drugs (NSAIDs). It appears that exposure to coxibs, perhaps even for even as little as two weeks [6], provokes a small but significantly increased risk for adverse CV events that is related in part to the presence of baseline CV risk factors [1,2]. This coxib effect is probably multifactorial, abundant data from experimental models suggest that it derives at in part from suppressed production of prostacyclin (PGI2), especially relative to that of thromboxane (Tx) A2. Production of TxA2 is closely linked to COX-1 activity while production of PGI2 is more closely linked to COX-2 activity, with aspirin preferentially inhibiting COX-1 activity and t-NSAIDs inhibiting both COX isozymes without bioactivation [7,8,9,10]

Methods

Results

Conclusion