Prognostic Value of Cardiac Troponin in Patients With Chronic Kidney Disease Without Suspected Acute Coronary Syndrome

review (n = 3492) Excluded (n = 3317) Article review (n = 1182) Excluded (n = 1052)* No original data: 28 Meeting abstract: 265 Did not include patients with CKD or end-stage renal disease: 72 Did not evaluate troponin I or troponin T: 383 Troponin and CKD results not presented separately: 257 No comparison of interest: 34 No outcome of interest: 168 Not applicable: 83 Other reason: 8 Included studies (n = 98 [105 publications]) Addressed other objectives (n = 25) CKD chronic kidney disease. * Articles could be excluded for 1 reason. This online-first version will be replaced with a final version when it is included in the issue. The final version may differ in small ways. Review Prognostic Value of Troponin in Patients With CKD Without Suspected ACS www.annals.org Annals of Internal Medicine 3 Downloaded From: http://annals.org/ on 10/06/2014 Patients With CKD Receiving Dialysis Troponin T Level and All-Cause Mortality Forty-three cohort studies (in 49 publications) examined the association between troponin T level and all-cause mortality (12, 20, 21, 27, 29–32, 36, 38, 40–42, 44, 47, 50, 51, 53, 59, 61, 65, 67–69, 71, 73, 77, 79, 81, 82, 85, 88, 89, 91–97, 99, 100, 102–104, 109, 111, 116, 119). We excluded 10 of these studies from our meta-analyses because they had insufficient data or did not present results separately for patients receiving dialysis (31, 36, 38, 40, 85, 102, 103, 109, 111, 116). We pooled HRs from 11 studies that adjusted for age and CAD or a risk equivalent (Figure 3) and found a 3-fold increased risk (HR, 3.0 [95% CI, 2.4 to 4.3]); heterogeneity was not significant. Similar risk was noted in meta-analyses of HRs that were adjusted for age but not CAD or were unadjusted. We also pooled ORs from 24 studies (Supplement 4, available at www.annals.org) and found a nearly 5-fold increased risk (OR, 4.7 [CI, 3.6 to 6.5]) with significant heterogeneity (I 53%). One study (31) tested high-sensitivity troponin T level on a continuous scale rather than by cut point. The ageadjusted risk increased 1.4-fold (CI, 1.0to 2.0-fold) for every 2.72-ng/L increase in troponin T level. Troponin I Level and All-Cause Mortality We identified 30 studies (in 31 publications) on the association between troponin I level and all-cause mortality (3, 19–21, 25, 30, 32, 36, 43, 47, 48, 51, 56, 67, 68, 72–74, 81, 86, 88, 89, 93, 94, 102–105, 113, 118, 120). We excluded 7 of them from our meta-analyses because they reported insufficient data, presented troponin levels as continuous variables rather than cut points, or did not present results separately for patients receiving dialysis (30, 36, 89, 102, 103, 105, 113). We included 7 studies in a meta-analysis of HRs adjusted for age and CAD or a risk equivalent (Figure 4). The overall pooled HR was 2.7 (CI, 1.9 to 4.6); heterogeneity was not significant. Similar levels of risk were seen in studies that presented HRs that were adjusted for age but not CAD or were unadjusted. In a meta-analysis of 19 studies, we found a pooled OR of 2.6 (CI, 1.9 to 3.6) (Supplement 5, available at www.annals.org). One study (113) evaluated the risk associated with each 10-ng/L increase in high-sensitivity troponin I level and did not find a statistically significant association, but the study was underpowered for this outcome. Troponin T Level and Cardiovascular Mortality Twenty studies (of 16 unique cohorts) addressed the association between troponin T level and cardiovascular mortality (12, 44, 50, 53, 59, 66–68, 77, 79, 91–95, 99, 107, 108, 111, 114). We excluded 2 of these studies from our meta-analyses because they had insufficient data (111, 114). Five studies were included in a meta-analysis of pooled HRs adjusted for age and CAD or a risk equivalent (Supplement 6, available at www.annals.org). We found a 3-fold increased risk (HR, 3.3 [CI, 1.8 to 5.4]) with significant heterogeneity (I 66%). We included 9 studies in the meta-analysis of ORs (Supplement 7, available at www.annals.org) and found a 4-fold increased risk (OR, 4.3 [CI, 3.0 to 6.4]). One study (114) used a high-sensitivity troponin T assay but presented results per 100-U increase in troponin T level as a continuous variable (OR, 1.5 [CI, 1.2 to 1.9]). Troponin I Level and Cardiovascular Mortality Thirteen studies addressed the association of troponin I level with cardiovascular mortality (3, 19, 25, 43, 58, 67, 68, 74, 90, 93, 94, 107, 118). One study was excluded Figure 2. Pooled HRs from studies that examined the association of elevated troponin level with outcomes among patients who were receiving dialysis and those who were not. Patients receiving dialysis All-cause mortality All-cause mortality CVD mortality