Repeat measurements on patient samples identifies unpredictable and poorly reproducible cardiac troponin results with a high-sensitivity cardiac troponin assay

There is much current interest in the use of low-normal high-sensitivity cardiac troponin (hsTn) concentrations, with or without minimal change, to rule out myocardial infarction (MI). Clifford-Mobley’s observations demonstrate that this a challenge even for platforms measuring hsTnT. ¹ Analytical imprecision may also affect algorithms that use hsTn change alone to rule in MI. For example, the imprecision observed with the Ortho hsTnI assay (Ortho Clinical Diagnostics, NJ, USA), using quality control or patient pools, has been found to exceed the European Society of Cardiology 0/1 h algorithm criterion. ² The Ortho hsTnI assay has also been shown to yield high and non-reproducible results (i.e. outliers), in addition to the problems with imprecision. ³ Outliers are often identified by repeat centrifugation and repeat testing. However, Ortho hsTnI results above the 99th percentile cut-offs may be discordant with respect to other cardiac troponin assays and the clinical diagnosis, even when imprecision from duplicate analysis is acceptable. ⁴ As part of a stability study assessing Ortho hsTnI concentrations in both EDTA plasma and lithium heparin plasma over 24 h, we have observed that this interference is random and not related to time on cells.

Briefly, samples were collected into paired lithium heparin and potassium EDTA vacutainers (Becton, Dickinson, and Company, NJ, USA), i.e. one lithium heparin and one EDTA tube were each collected at the same time from the same patient, and Ortho hsTnI was measured within 1 h of collection (baseline), and after various time intervals (4 h, 8 h, 24 h). Before measurement, the whole blood samples were mixed, aliquoted into new tubes and centrifuged at 2300 × g for 10 min. The plasma samples were stored on cells and at room temperature prior to measurement. Quality control was run in duplicate at the beginning and end of each day and met the laboratory precision goals for this assay. ³ , ⁴

Table 1 summarizes our findings. Reproducibility of duplicate measurements was unpredictable; unacceptable precision is highlighted in the table. Our study confirms that both plasma types can yield high and non-reproducible results, and that samples from some patients, e.g. patients 3 and 4, were more likely to yield unacceptable replicate results than others. The finding of acceptable replicate hsTnI results at one time point after collection did not guarantee reproducible results when measured at subsequent time points after collection; similarly, the finding of unacceptable replicate results tested at one time point did not reliably predict unacceptable results subsequently.

Table 1.

Acceptability of the duplicate Ortho hsTnI results (grey shaded area indicate poor duplicates with CVs > 10%) obtained from EDTA and lithium heparin plasma samples from five patients when stored at room temperature (15–30°C).

Room temperature	Baseline1st	Baseline2nd	Baselineaverage	CV	4h1st	4h2nd	4haverage	CV	8h1st	8h2nd	8haverage	CV	24h1st	24h2nd	24 haverage	CV	AbbotthsTnI [ng/L]	CRP [mg/L]	% Difference at 24h
Patient 1 EDTA	3.2	3.2	3.2	0%	2.7	3.1	2.9	10%	3.0	2.8	2.9	5%	3.1	3.2	3.2	2%			0%
Patient 1 lithium heparin	3.2	3.5	3.3	6%	4.1	3.9	4.0	4%	5.1	4.1	4.6	15%	3.8	No Result	3.8	.			15%
Patient 2 EDTA	99.4	No Result	99.4	.	102.1	104.3	103.2	2%	101.2	102.8	102.2	1%	110.7	104.4	107.6	4%			8%
Patient 2 lithium heparin	126.7	120.0	123.4	4%	125.9	121.5	123.7	3%	120.8	136.1	128.5	8%	144.2	134.7	139.5	5%			13%
Patient 3 EDTA	11.9	4.3	8.1	66%	8.1	16.7	12.4	49%	18.3	17.9	18.1	2%	15.3	16.6	16.0	6%			98%
Patient 3 lithium heparin	5.1	5.1	5.1	0%	3.6	9.3	6.4	63%	17.4	4.8	11.1	80%	6.7	7.3	7.0	6%			37%
Patient 4 EDTA	28.2	25.8	27.0	6%	70.0	55.2	62.6	17%	32.4	59.8	46.1	42%	38.6	53.6	46.1	23%	15		71%
Patient 4 lithium heparin	74.7	52.5	63.6	25%	46.1	49.3	47.7	5%	48.4	69.6	59.0	25%	62.9	81.2	72.0	18%		>90	13%
Patient 5 EDTA	20.4	20.5	20.5	0%	20.5	20.5	20.5	0%	19.5	19.2	19.4	1%	18.7	19.0	18.9	1%	33		-8%
Patient 5 lithium heparin	25.3	25.5	25.4	1%	28.5	27.8	28.1	2%	29.1	28.0	28.5	3%	31.8	33.0	32.4	3%			28%

Note: Patient 3 and patient 4 results are bolded with several poor duplicates with patient 4 results all above Ortho’s 99th percentile with Abbott hsTnI < 99th percentile in this patient.

Confirmation of myocardial injury with an independent method is warranted until hsTnI measurements with the Ortho assay can be shown to have acceptable reproducibility and predictability.