Abstract Background Nonfatal myocardial infarction (MI) inadequately predicts cardiovascular and all-cause mortality in contemporary RCTs. The perceived efficacy of new treatments could be overstated if positivity in composite outcome is primarily attributed to reduction in nonfatal MI. Purpose This study seeks to assess pharmacological RCTs with positive results influenced by nonfatal MI reductions and scrutinize the associated clinical guideline recommendations. Methods Pharmacological RCTs focusing on mortality and nonfatal MI published from 2000 to 2023 were searched in PubMed and Web of Science, with subsequent citation tracking in Web of Science and Google Scholar for candidate RCTs referenced in English-language clinical guidelines. Main outcomes included: 1) relative risk reduction in the composite outcome excluding nonfatal MI; 2) relevant guideline recommendations based on the supporting trials. The impact of nonfatal MI on composite outcome was assessed by using the leave-one-out method. Meta-analysis utilized a random effects model. Results After reviewing 19,825 records, we identified 8 RCTs where positive composite outcomes were driven by reductions in non-fatal MI (Figure 1A). These were divided into three therapeutic groups: anti-thrombotic (3 trials), lipid-lowering (3 trials), and anti-inflammatory (2 trials). In guidelines citing these RCTs, lipid-lowering trials (IMPROVE-IT, FOURIER, CLEAR Outcomes; the latter has not been cited yet; a total of 60 recommendations across 17 guidelines; Figure 2) outperformed anti-thrombotic (27 recommendations across 15 guidelines) and anti-inflammatory trials (3 recommendations in 3 guidelines), with significantly higher Class I (45%) and IIa (33.3%) recommendations, and fewer Class IIb (21.7%) recommendations, compared to anti-thrombotic (0%, 22.2%, 66.7%) and anti-inflammatory (0%, 0%, 100%) trials (Figure 1B). In a meta-analysis of all trials on combined intensive lipid-lowering on top of maximally-tolerated statins (the ODYSSEY OUTCOMES, IMPROVE-IT, and FOURIER trials), the exclusion of nonfatal MI data sustained the negative findings (HR 0.94, 95% CI: 0.88 to 1.01). Conclusions Despite the overestimation of efficacy in certain anti-thrombotic, anti-inflammatory, and lipid-lowering RCTs due to the inclusion of nonfatal MI, lipid-lowering trials, particularly those supporting combined intensive therapy with statins, still receive disproportionately favorable guideline recommendations. This underscores the need to reassess the strength of evidence for these trials in guidelines, advocating for a more equitable and evidence-based approach to cardiovascular guideline recommendations.
Read full abstract