Key result
Subsequent lots of hs-cTnT reagent did not change the prevalence of myocardial injury (≥14 ng/L) compared to the original lot, but increased values <5 ng/L.
Why the study?
Does the downward shift in subsequent lots of the hs-cTnT assay affect the prevalence of myocardial injury or the detection of low troponin concentrations in hospitalized patients?
Does the downward shift in subsequent lots of the hs-cTnT assay affect the prevalence of myocardial injury or the detection of low troponin concentrations in hospitalized patients?
Answers are grounded in this paper and cite their sources.
The downward shift in certain lots of the Roche hs-cTnT assay primarily affects concentrations below the 99th percentile, potentially impacting rapid rule-out algorithms but not standard myocardial infarction diagnosis.
To the Editor: Apple and Jaffe have clearly articulated possible clinical implications regarding an adjustment to the high-sensitivity cardiac troponin T (hs-cTnT) assay at concentrations near the limit of detection (1). They pose 5 thought-provoking questions and indicate their belief “that these findings have major implications for patient care” (1). The latter point is difficult to assess and is probably the most important concern for laboratories reporting hs-cTnT results and the clinicians who interpret these results. For hospital laboratories, the main intended use for the hs-cTnT assay is to identify myocardial injury in patients with acute conditions, such as acute coronary syndrome. For this clinical situation, a concentration above the 99th percentile is useful to aid in the clinical diagnosis of myocardial infarction (2). Roche's technical bulletin indicated that certain lots demonstrated a downward shift at the lower end of the analytical interval and that subsequent lots would be readjusted to conform to the original assay's specification. At our hospital site, we have been measuring hs-cTnT since 2010 in EDTA-containing plasma samples from a specific subgroup of hospitalized patients (approval obtained from the research ethics board). Over these 2.5 years we have used 4 different lots of hs-cTnT reagent (first lot, 153401; second lot, 157120; third lot, 160197; fourth lot, 163704). Roche has confirmed that only the first lot (153401) was not affected by the downward shift. Thus, we set out to retrospectively determine whether the prevalence of myocardial injury changed when the different lots were used for hs-cTnT assays in this homogeneous population. All hs-cTnT results reported since the introduction of the assay at our hospital (March 2, 2010) until September 20, 2012, were obtained from our laboratory information system. A total of 8835 hs-cTnT results were reported, with 10% of the hs-cTnT results having been obtained during use of the first lot. Plotting the hs-cTnT results per reagent lot indicated that there was a gradual loss in detectable concentrations (limit of detection for the hs-cTnT assay, 5 ng/L) (Fig. 1). For concentrations above the 99th percentile (14 ng/L), however, there did not appear to be a noticeable difference between reagent lots. In fact, we observed no significant change in the prevalence of myocardial injury between the first lot [152401: 27% (95% CI, 24%–30%) for hs-cTnT concentrations ≥14 ng/L] and subsequent lots [157120: 28% (95% CI, 26%–29%); 160197: 28% (95% CI, 26%–30%); 163704: 24% (95% CI, 22%–26%); P > 0.10, Fisher exact test]. The percentages of hs-cTnT concentrations below the limit of detection (<5 ng/L) were higher in lots 157120 (29%; 95% CI, 27%–30%; P = 0.032, one-tailed Fisher exact test), 160197 (33%; 95% CI, 31%–35%; P < 0.001, one-tailed Fisher exact test), and 163704 (45%; 95% CI, 43%–47%); P < 0.001, one-tailed Fisher exact test), compared with lot 153401 (25%; 95% CI, 22%–28%). These data lend support to Roche's claim that the loss in signal and concentrations mainly affected concentrations below the 99th percentile. This finding is important, because patient management that uses cutoffs at either the reported 99th percentile or higher concentrations may have been minimally affected; however, clinicians who use other algorithms for rapid rule-out of myocardial infarction, such as using the limit of detection for rule-out, could have missed individuals with these subsequent lots of hs-cTnT reagents (3). Apple and Jaffe have appropriately commented on the challenges faced by manufacturers to produce high-sensitivity cardiac troponin assays; however, these events have also highlighted the fact that clinical laboratories may need to monitor performance of high-sensitivity assays at concentrations below the 99th percentile. Presently, guidelines indicate that quality control be evaluated at the 99th percentile concentration to monitor precision (2); however, if lower cutoffs are to be used to direct care, clinical laboratories will have to use additional quality-assurance measures to ensure appropriate laboratory services. This consideration leads to the next major point regarding findings from published studies that used lots that have demonstrated lower recovery. Both Roche's technical bulletin and our retrospective review of hs-cTnT results suggest that the distributions of hs-cTnT results obtained with the affected lots were different below the 99th percentile (Fig. 1). Thus, findings from studies that assessed cutoffs or concentration intervals below the 99th percentile will need to be reviewed and possibly repeated to establish the correct cutoffs for risk stratification. We have recently reported on the role of the hs-cTnT assay (lot 153401) in risk stratification for myocardial infarction, heart failure, and cardiovascular death in a high-risk stable population (4). These data support hs-cTnT measurement as an important biomarker for predicting cardiovascular outcomes, with different concentration intervals yielding different risks (4). Importantly, the highest concentrations (>30 ng/L) identified patients at the highest risk (4), which is consistent with recent findings obtained with the fourth-generation cTnT assay indicating that concentrations >10 ng/L (i.e., ≥20 ng/L) identified patients at the highest risk of death by 30 days after noncardiac surgery (5). Additional and confirmatory studies that assess hs-cTnT concentrations for both short- and long-term cardiovascular outcomes will be required to support these latest data; however, concerns regarding published data that have assessed concentrations above the 99th percentile for the hs-cTnT assay may be diminished with these findings.
No takes yet. Share an insight, caveat, or question.
Kavsak et al. (2013) conducted a letter in Hospitalized patients (n=8,835). Subsequent lots of hs-cTnT reagent vs. Original lot (153401) was evaluated on Prevalence of myocardial injury (hs-cTnT concentrations ≥14 ng/L) (p=>0.10). Subsequent lots of hs-cTnT reagent did not change the prevalence of myocardial injury (≥14 ng/L) compared to the original lot, but increased values <5 ng/L.
Synapse has enriched 10 related papers. Consider them for comparative context: