Prediction of severe cardiovascular events by VE/VCO2 slope versus peak VO2 in systolic heart failure: A meta-analysis of the published literature

Background Peak VO2 has traditionally been used for prognostic evaluation in systolic heart failure. However, in the past years, VE/VCO2 slope has been shown to be similar or even superior in many studies. We performed a systematic review and a meta-analysis of diagnostic studies of VE/VCO2 slope to assess its ability to predict cardiovascular events in systolic heart failure. Methods We searched the published literature in PubMed and ISI Web of Science for VE/VCO2 slope in heart failure, and performed a systematic review and a meta-analysis of diagnostic studies in articles fulfilling previously established selection criteria. End points were serious cardiovascular events defined as death or the combined end point of death, ventricular assist device implantation, or heart transplant. A sub-analysis was also performed with those articles providing enough data to compare VE/VCO2 slope prognostic ability to that of peak VO2. Results Four hundred ninety-one articles that are potentially relevant were identified, and 12 studies were selected based on our predefined criteria. No heterogeneity or evidence of publication bias was found. The 12 studies included a total of 2,628 patients with a mean follow-up of 31 months (95% CI 16-46 months). The combined event rate at 1 year was 11.2% (95% CI 7.8%-14.6%). Diagnostic odds ratio and area under the curve for serious cardiovascular events were 5.02 (95% CI 4.06-6.21) and 0.75 (95% CI 0.72-0.78), respectively. Six studies provided sufficient data for VE/VCO2 slope and peak VO2 comparison. Both variables showed similar performance, although VE/VCO2 did present a trend to superiority. Conclusions In this meta-analysis, VE/VCO2 slope represents a reasonable ability to predict serious cardiovascular events in systolic heart failure, and is at least as effective as peak VO2. (Am Heart J 2010;160:1004-14.)