PREVENTING DRUG-ELUTING CORONARY STENT THROMBOSIS WITH ANTIPLATELET THERAPY IN THE SETTING OF A RECENT SEVERE HEMORRHAGE: A DECISION ANALYSIS

Purpose: We modeled the long-term risks and benefits of restarting antiplatelet therapy (APT) to prevent drug-eluting coronary stent thrombosis (DES-t) after APT was discontinued secondary to severe bleeding and there is a long-term increased risk of severe bleeding with APT.

Methods: Using published data, we built a Markov model comparing four treatment strategies: aspirin plus clopidogrel (dual APT); aspirin alone (ASA), clopidogrel alone; and neither aspirin nor clopidogrel (no APT).  The base case examined a 65-year old person treated with a DES and continuous dual APT for two years prior to a severe bleeding event (e.g., intracerebral hemorrhage, upper gastrointestinal bleeding) that resulted in APT being discontinued.  We considered risks of mortality from DES-t, myocardial infarction unassociated with DES-t, subsequent bleeding at the same site as the prior hemorrhage, and subsequent hemorrhage at another site.  We used a lifetime horizon and projected life expectancy (LE) without quality-adjustment.

Results: In the baseline analysis, either ASA or no APT was preferred (LE of 7.60 and 7.59 years, respectively).  In contrast, LE was only 7.39 years with dual APT, and 7.36 years with clopidogrel alone.  All parameters were varied over plausible ranges in sensitivity analyses.  We also performed sensitivity analyses on the time since stent placement (base case, 2 years) and on the duration of future treatment with clopidogrel (base case, life long).  Treatment with either ASA or no APT remained preferred over plausible ranges for the annualized probability of subsequent bleeding at the same site as the prior hemorrhage without APT and for the relative risk of bleeding with aspirin.  No APT became the preferred strategy when the efficacy of aspirin to prevent DES-t fell below 0.19 (baseline 0.30).  When the annualized probability of DES-t (baseline 0.0040) fell below 0.0025, no APT was the preferred strategy, and when it exceeded 0.067, dual APT was the preferred strategy.

Conclusions: The choice between ASA and dual APT was a relatively close call. Despite guidelines suggesting the use of thienopyridines, in patients with an increased risk of life threatening bleeding, the strategies containing clopidogrel are apparently not optimal beyond the first 24 months after stent placement unless the risk of DES-t is higher than that reported in long-term outcome studies.