Response to Letter Regarding Article by Khan et al, “Prognostic Value of Troponin T and I Among Asymptomatic Patients With End-Stage Renal Disease: A Meta-Analysis”
We thank Drs Apple, Jaffe, and Herzog for their interest in our article and for drawing attention to the National Kidney Foundation (NKF) Guidelines on Evaluation of Cardiovascular disease.1 As stated in our paper, because the troponin studies used differing assays and variable cut points and reported conflicting results, direct comparison of the results was problematic, and a rigorous meta-analysis was required to clarify this association.2 The review from the NKF did not include all relevant studies (eg, results from 8 trials were not included in the NKF review), reported on results using differing cut points or outcomes, and did not evaluate for study heterogeneity and publication bias, given the significant possibility that negative observational studies may not be published3 (as was indeed found in our analysis). In addition, we respectfully submit that recommendations from professional organizations and regulatory bodies, although valuable, can be usefully informed by additional research—the existence of the NKF and FDA documents does not contradict our statement that the available literature has produced conflicting results.
Dr Apple and colleagues correctly point out that the cut point of 0.1 ng/mL is not a part of the ESC/ACC guidelines. We chose this cut point to improve between study comparisons and allow for mathematical synthesis, as the vast majority of troponin T studies evaluated prognosis at this level. Thus, from the evidence base, a cut point of 0.1 ng/mL was associated with reduced long-term survival among asymptomatic hemodialysis patients. The degree of imprecision at this cut point is also close to that recommended by the ESC/ACC. Most troponin assays do not currently meet this standard at the 99th percentile.4–5 However, as more studies accrue and precision of troponin assays at lower thresholds improves, a lower cut point may prove to have greater predictive power in this population.
We agree that troponin T elevations are more often seen in patients with chronic kidney disease compared with troponin I. It remains difficult to assess the ability of troponin I to predict mortality, given the variability of assays and assay cut points studied.
Thank you for identifying the typographical error in Table 2: The table column should read 99th percentile and not the 90th percentile.
Khan NA, Hemmelgarn BR, Tonelli M, Thompson CR, Levin A. Prognostic value of troponin T and I among asymptomatic patients with end-stage renal disease: a meta-analysis. Circulation. 2005; 112: 3088–3096.
Panteghini M, Pagani F, Yeo KT, Apple FS, Christenson RH, Dati F, Mair J, Ravkilde J, Wu AH; Committee on Standardization of Markers of Cardiac Damage of the IFCC. Evaluation of imprecision for cardiac troponin assays at low-range concentrations. Clin Chem. 2004; 50: 327–332.
Apple FS, Murakami MM. Serum 99th percentile reference cutoffs for seven cardiac troponin assays. Clin Chem. 2004; 50: 1477–1479.