Joel Thunberg1, Olle Björkqvist2, Michael Eberhardsson3, Daniel Bergemalm1, Carl Eriksson1 and Jonas Halfvarson1

Affiliates

[1] Department of Gastroenterology, Faculty of Medicine and Health, Örebro University, Örebro, Sweden. [2] Department of Laboratory Medicine, Clinical Microbiology, Faculty of Medicine and Health, Örebro University, Örebro, Sweden. [3] Department of Medicine Solna, Karolinska Institutet, Stockholm, Sweden.

Abstract

Background
Point-of-care test (POCT) devices for measuring concentrations of anti-tumour necrosis (TNF) agents, such as infliximab (IFX), have recently been developed to provide rapid and user-friendly measurements. We aimed to compare the agreement between a POCT IFX assay (ProciseDx, San Diego, CA, USA) and the conventional in-house ELISA at Karolinska University Hospital, Sweden.

Methods
Adult patients with inflammatory bowel disease treated with infliximab at Örebro University Hospital were prospectively recruited between June and December 2021. IFX levels were consecutively measured as part of the clinical routine using the in-house ELISA. After obtaining written informed consent, additional blood samples were collected, and the serum was separated and stored as aliquots at -80°C. After the inclusion of all patients, IFX levels were measured in a single batch using the POCT. Agreement between the two assays was visualised on a Bland-Altman plot and a scatter diagram. On the Bland–Altman plot, a good agreement is reflected by a horizontal line near the mean difference of zero. A Deming regression analysis was also performed, and the Pearson correlation coefficient was calculated. Values below the lower limit of detection (LOD), i.e. <0.5 µg/mL for the in-house ELISA and <1.7 µg/mL for the POCT, were substituted with LOD/√2.

Results
Sixty-one serum samples were collected and analysed. Using the POCT, six measurements were below the LOD, and three of these were also below the LOD when the in-house ELISA was used. A significant correlation in IFX levels was observed when the POCT and the ELISA were compared (r=0.95, p<0.001) (Figure 1). A Deming regression analysis resulted in a slope of 1.16 and -0.73 for the intercept. A Bland-Altman plot of all measurements is shown in Figure 2 and resulted in a bias of -0.77.

Conclusion
The POCT showed no clinically relevant bias compared to the conventional ELISA but seemed to generate slightly higher IFX concentrations. However, the upper and lower limits of agreement in the Bland-Altman plot seemed clinically acceptable.