Evaluation of a novel, stand-alone system to measure interleukin-6

To the Editor,

Early detection of sepsis and its subsequent treatment are crucial to prevent life-threatening situations in patient care. Interleukin-6 (IL-6) is a pro-inflammatory cytokine released in response to inflammation by a variety of cell types, including T lymphocytes and endothelial cells [1, 2]. IL-6 plasma concentrations are often used to assess acute inflammatory reactions due to trauma and/or infection, including severe COVID19 [3, 4], and especially in the early detection of neonatal sepsis [5]. In addition, IL-6 plasma levels correlate with the severity of sepsis and may predict outcome [6, 7]. Therefore, assessment of IL-6 plasma concentrations represents a clinically relevant diagnostic tool for monitoring critically ill individuals, including infants. Although well established, options for quantitative IL-6 measurements are often limited to large laboratories with expensive high-throughput devices. In addition, not all companies that offer automated laboratory systems have IL-6 in their own current parameter portfolio (e.g. Abbott, USA). Given the diagnostic and clinical importance of IL-6, however, more options for measuring its concentration in plasma and/or serum are necessary.

In this context, smaller stand-alone systems such as the novel iChroma 50 (Boditech Med Inc., Republic of Korea; distributed by WS Labor, Germany) might represent a useful alternative to the fully automated, large laboratory systems. In order to evaluate this, we compared the IL-6 assay using this novel system (iChroma 50) with the widely used, state-of-the-art method on the cobas e411 (Roche Diagnostics, Germany). Briefly, the iChroma 50 system is an automated, straightforward device for a test card-based fluorescent immunoassay (FIA), as compared to the well-known IL-6 electro-chemiluminescence immunoassay (ECLIA) on the cobas e411. Specific assay details can be found in Supplementary Table 1. In our comparison, 100 lithium heparinate plasma samples were identified based on their initial, medically justified IL-6 measurements on the cobas e411 in the central laboratory of the University Medical Center Goettingen (UMG), covering a wide concentration range (2.9–12,333.0 pg/mL). Samples were then stored at 4 °C before they were simultaneously measured on the iChroma 50 and cobas e411 for this comparative analysis.

The plasma used in this study was leftover material, and all patient samples were processed in an anonymized manner without consideration of further patient-related data. The study was conducted in accordance with the World Medical Declaration of Helsinki and approved by the UMG ethics committee (No.: 20/3/22). For statistical analysis and visualization, Microsoft Excel (Microsoft, USA) and RStudio (RStudio, USA) with the ggplot2 package were used.

IL-6 concentrations ranged from 2.0 to 13,193.6 pg/mL (median=30.9 pg/mL) on the iChroma 50, and 2.9–12,333.0 pg/mL (median= 43.9 pg/mL) on the cobas e411. Of note, ten samples had to be excluded from the following quantitative comparison as they showed negative results, i.e. below the limit of detection (LoD), on both devices. Passing-Bablok regression analysis of the remaining 90 samples revealed good comparability between both systems with a slope of 1.085 and a Y-intercept of 3.079 (Figure 1). The correlation was excellent with r>0.99 (Figure 1). Our data further indicate that IL-6 concentrations below ∼100 pg/mL are systematically measured slightly lower on the iChroma 50 vs. cobas e411, whereas this is the opposite for higher IL-6 concentrations (Figure 1). Agreement of qualitative results (positive/negative) - according to the proposed cut-off of 7 pg/mL - between both systems was very good as 95.0% of all samples (95/100) were equally interpreted using both systems, with a Cohen’s kappa of 0.86 (Supplementary Table 2).

Figure 1:

Passing-Bablok regression analysis. Passing-Bablok regression analysis between the iChroma 50 and the cobas e411 was performed from IL-6 measurements in 90 lithium heparinate plasma samples. The regression line is shown in red, the bisecting line is presented as a dashed black line. Regression equations with respective 95% confidence intervals (CI) and Pearson’s correlation coefficients (r) are presented in the respective figures. The magnified insert shows the regression line for IL-6 measurements up to 250 pg/mL.

Additionally, we assessed the imprecision according to the CLSI EP15 A2 protocol. This was found to be much higher on the iChroma 50 as compared to the cobas e411, with total imprecision coefficients of variation (CVs) ranging from 6.9 to 14.5% on the iChroma 50 vs. 2.7 and 2.4% on the cobas e411 (Table 1). A similar pattern could be observed for the within-run imprecision (CVs of 7.6 and 13.5% on the iChroma 50 vs. 2.8 and 2.6% on the cobas e411) (Table 1). According to our analysis, CVs were particularly high for the high control level (target value 876.81 pg/mL).

As mentioned above, IL-6 assessment is of great diagnostic importance in neonatal sepsis. Due to the limited amount of plasma in neonatal samples, the following algorithm is often used in laboratories to obtain as much information as possible from one single plasma sample containing ethylenediamine tetraacetic acid (EDTA) as anticoagulant: (i) blood count before centrifugation (this analysis requires EDTA as anticoagulant), (ii) centrifugation, (iii) measurements of the C-reactive protein (CRP) and IL-6 in the remaining EDTA plasma. Therefore, we tested whether IL-6 measurements on the iChroma 50 were also comparable with the cobas e411 in 30 EDTA samples (instead of lithium heparinate). Passing-Bablok analysis revealed moderate comparability with a slope of 1.234 and a Y-intercept of 16.122, indicating a relevant systematic bias (Supplementary Figure 1). In addition, this suggests that the type of anticoagulant may affect the results. This analysis had the limitation that EDTA samples were from adults rather than neonates due to the limited sample volume in neonatal specimens, which did not allow us to perform multiple measurements. This has to be evaluated separately in future studies.

Taken together, our data suggest that the iChroma 50 may serve as a suitable, stand-alone alternative to allow sufficient assessment of IL-6. This is particularly due to the good comparability with the state-of-the-art assay on the cobas e411 and the excellent correlation when using lithium heparinate samples. However, our data demonstrate relevant differences in comparability of both systems for EDTA plasma samples. Therefore, care should be taken when measuring IL-6 concentrations on the iChroma 50 in EDTA plasma, which is of particular clinical relevance with respect to neonatal samples. Another weakness is the relatively high imprecision, especially at high IL-6 concentrations, with CVs up to 14.5%. This is not in concordance with the precision data provided by the manufacturer (CV of 6.5% for an IL-6 concentration of 1,274 pg/mL) and should be investigated in further studies. It appears to be an assay-specific problem rather than a general issue with the iChroma 50, as a recent study demonstrated excellent precision for procalcitonin (PCT) measurements on the iChroma, even for high PCT concentrations [8]. The imprecision of IL-6 measurements clearly must be taken into account when using the iChroma 50 in regular patient care; false clinical conclusions could be made when patients are followed up longitudinally, e.g. a reduction in IL-6 could be interpreted as therapeutic success although it may simply be based on the imprecise measurements. This consequential problem needs to be addressed.