Framingham risk score and novel cardiovascular risk factors underpredict major adverse cardiac events in kidney transplant recipients.

BACKGROUND

Framingham Risk Score (FRS) is an insufficient cardiovascular event predictor in unselected kidney transplant recipients. Its role in different risk subgroups and the value of adding novel risk factor candidates to FRS is unknown.

METHODS

We reviewed patients who underwent transplantation from 1998 to 2008 with minimum 3 months graft function, determining FRS-ascertained 10-year risk at 3 months along with relevant clinical and laboratory information. Major adverse cardiac events (MACE) (myocardial infarction, coronary artery revascularization, or cardiac death) 3 months posttransplant were captured. Time-to-MACE multivariate Cox modeling with FRS and novel risk factors (C-reactive protein, uric acid, urine albumin-to-creatinine ratio) as independent variables was performed.

RESULTS

Of 956 patients, 89 experienced MACE (2.17 events/100 patient-years). FRS-predicted 10-year risk was 14.7% ± 10.0% in males with and 9.2% ± 8.2% in those without subsequent MACE (P < 0.0001), although FRS substantially underestimated MACE (actual-to-predicted event ratio 1.2-8.4 in different subgroups, all P < 0.0001). Although patients with MACE had a higher C-reactive protein (5.4 ± 6.0 vs. 3.8 ± 2.5 mg/L, P = 0.026) and uric acid (417 ± 109 vs. 386 ± 101 μmol/L, P = 0.012) level as well as lower 3-month estimated glomerular filtration rate (50.1 ± 20.1 vs. 54.8 ± 18.3 mL/min/1.73 m(2), P = 0.022), only FRS more than or equal to 10% (hazard ratio 2.313, 95% confidence interval 1.49-3.58, P = 0.0002) and estimated glomerular filtration rate less than 50 mL/min/1.73 m(2) (hazard ratio 2.291, 95% confidence interval 1.06-4.94, P = 0.034) predicted MACE in multivariate analysis. Adding novel risk factors to FRS did not improve FRS prediction.

CONCLUSION

FRS substantially underpredicts MACE in kidney transplant recipients among all risk subgroups. Commonly available novel risk factors do not improve FRS predictive value.