Cardiovascular risk stratification in diabetic patients following stress single-photon emission-computed tomography myocardial perfusion imaging: the impact of achieved exercise level.

BACKGROUND: Previous studies have demonstrated that diabetic patients undergoing exercise stress single-photon emission-computed tomography (SPECT) myocardial perfusion imaging (MPI) have significantly lower cardiac events when compared to the diabetic patients undergoing pharmacologic stress SPECT MPI across all perfusion categories. However, there are limited data on the level of exercise achieved during exercise SPECT MPI among diabetic patients and its impact on cardiovascular outcomes.
METHODS: We retrospectively analyzed 14,849 consecutive patients (3,654 diabetics and 11,195 non-diabetics) undergoing exercise stress, combined exercise and pharmacologic stress, and pharmacologic stress SPECT MPI from 1996 to 2005 at a single tertiary care center. Diabetic and non-diabetic patients were categorized into 3 groups based on the metabolic equivalents (METs) achieved: ≥5 METs, <5 METs, and pharmacologic stress groups. All studies were interpreted using the 17-segment ASNC model. The presence, extent, severity of perfusion defects were calculated using the summed stress score (SSS), and patients were classified into normal (SSS < 4), mildly abnormal (SSS 4-8), and moderate-severely abnormal (SSS > 8) categories. Annualized event rates (AER) for the composite end point of non-fatal myocardial infarction and cardiac death were calculated over a mean follow-up period of 2.4 ± 1.4 years with a maximum of 6 years.
RESULTS: In moderate-severe perfusion abnormality (SSS > 8) category, diabetic patients who were able to achieve ≥5 METs had significantly lower AER compared to diabetic patients who were unable to achieve ≥5 METs (3% vs 5.5%, P = .04), and non-diabetic patients unable to achieve ≥5 METs (3% vs 4.8%, P < .001). Diabetic patients who achieved a high workload of ≥10 METs had a very low AER of 0.9%. Diabetic patients, who attempted exercise but were unable to achieve ≥5 METs, still had significantly lower AER than diabetics undergoing pharmacologic stress MPI across all perfusion categories [1.5% vs 3.2%, P = .006 (SSS < 4); 2.5% vs 4.9%, P = .032 (SSS 4-8); 5.5% vs 10.3%, P = .003 (SSS > 8)]. After adjustment for cardiovascular risk factors, the percentage decrease in cardiac event rate for every 1-MET increment in exercise capacity was 10% in the overall cohort, 12% in diabetic group, and 8% in non-diabetic group.
CONCLUSIONS: Despite significant perfusion defects, diabetic patients who achieve ≥5 METs during stress SPECT MPI have significantly reduced risk for future cardiac events. Diabetic patients who achieve ≥10 METs have a very low annualized event rate. These findings support that exercise capacity obtained during SPECT MPI is a surrogate for outcomes among diabetic patients undergoing nuclear stress testing.