Abstract
Background
Calprotectin is an acute-phase protein used extensively in the assessment of gastrointestinal inflammation. It can readily be measured by enzyme-linked immunoassay (ELISA) and recently by point-of-care testing (POCT). We evaluated the Quantum Blue® POCT in this study and compared it with our existing ELISA method.
Methods
The method comparison study used faecal samples (n = 47) sent to the laboratory for routine calprotectin analysis. Linearity was assessed by serial dilution of extracted faeces (n = 4). Extraction efficiency was determined by repeat extraction of three different stools. The variation in results as a consequence of reading the POCT cartridges either side of the recommended 12 min was also assessed.
Results
The assay was linear across the range stated by the manufacturer. When multiple samples were taken from the same stool, results varied from −31.3% to +31.5%. For the clinical arm of our study, strictly applying the 50 μg/g cut-off recommended for both assays as positive for gastrointestinal inflammation, there were four patients where results fell a different side of the clinical cut-off; two patients had results higher by Quantum Blue® and two higher by ELISA.
Conclusions
In our hands, the Quantum Blue® method was a suitable screening test for excluding inflammatory bowel disease. It may be of value to laboratories wishing to offer calprotectin but who do not have sufficient numbers to warrant ELISA methodology or in ‘one stop’ gastrointestinal clinics where an immediate result is required.
Introduction
Calprotectin is an acute-phase protein which is stable to intestinal transit and can readily be measured in faeces. It has a wide role in gastroenterology as a non-invasive method for assessing intestinal inflammation. Calprotectin is of value in the diagnosis of organic bowel disease 1 and in differentiating irritable bowel syndrome (IBS) from inflammatory bowel disease (IBD). 2–5 It has been used to predict treatment response, 6 detect disease relapse 7 and monitor treatment in IBD sufferers. 4,8,9
Calprotectin can be measured by enzyme-linked immunoassay (ELISA) or more recently by point-of-care testing (POCT) using the Bühlmann Quantum Blue® Kit (Bühlmann, Schönenbuch, Switzerland), marketed by Alpha Laboratories, Hampshire, UK.
The ELISA method is a time-consuming technique requiring incubations of up to two hours and costs (including 4 h of staff time and 10% overheads) of £18.13 per sample if a full 96-well plate is used. Laboratories often run samples in batches to reduce costs at the expense of long turnaround times. At £29.70 per sample (including 20 min staff time and 10% overheads), POCT, although more expensive, offers the opportunity for better turnaround times and one-off analyses.
In this study, we evaluated the Quantum Blue® Test and compared it with our existing PhiCal® ELISA methodology (Calpro AS, San Diego, CA, USA).
Methods
The method comparison study used faecal samples, sent to the laboratory for routine calprotectin analysis (n = 47). Samples were stored at −20°C prior to analysis and after thawing, analysed using both the ELISA and Quantum Blue® POCT on the same day. The ELISA methodology has been published elsewhere.
10
In brief, between 40 and 120 mg of faeces was weighed and diluted 1 in 50 weight for volume with extraction buffer. After vortex mixing and centrifugation (10,000
Assay linearity was assessed using serial dilution of extracts (n = 4) in extraction buffer. The length of the acceptable time window on either side of the recommended 12-min time point for reading the test cartridge was assessed by reading the same extracts (n = 2) at 6, 8, 10, 12 14 and 16 min post sample application. To test the reproducibility of the extraction process, stools (n = 3) were extracted three times using the extractors provided with the kit and assayed by the POCT method. This was repeated using our existing extraction technique with a known quantity of weighed stool which was assayed by the ELISA method; results were compared. In each case, care was taken to avoid any undigested particulate material in the sample.
Statistical analysis was carried out using Microsoft Office Excel 2003. Pearson linear correlation and Passing–Bablok regression statistics were calculated for the linearity study. The Mann–Whitney U test was used to assess if there were any differences in results when different extraction methods were used.
Results
The assay was linear across the range 30–300 μg calprotectin per g faeces as stated by the manufacturer (r = 0.76).
Triplicate extractions of the same stool using the extraction devices supplied with the POCT kit produced results varying between −31.3 and +31.5% (median 0.31%) (n = 3). This compared with the weighing and measuring by ELISA of −21.4 to +18.4% (median 1.05%) (n = 3). There was no significant difference between these results (P > 0.05, Mann–Whitney U test).
The length of the acceptable time window for reading the POCT cartridge was −2 to +4 min where all values were within 5% of the 12-min reading, suggesting that a slight delay was acceptable.
Table 1 shows the results for the patient comparison study, available clinical details and scanning studies. Of the 47 samples used in the study, calprotectin was within the measurable range for 18 patients by either method. Where results were measurable by both methods (n = 12), differences ranged from −70 to +74% (median −4%) and the two methods showed poor agreement (R 2 = 0.32).
Summary of results for the 47 patients assessed in this study
IBD, inflammatory bowel disease; IBS, irritable bowel syndrome; CD, Crohn's disease; DVD, diverticular disease; MRI, magnetic resonance imaging; CRP, C-reactive protein
*/ Indicates no clinical details available
Both kit manufacturers recommend a <50 μg/g cut-off as normal. Although it is noted that in clinical practice other factors would be taken into account, further testing of patients falling outside this range would be performed in 19/47 patients by either method. Of these 19 patients, two (patients 31 and 41) were positive by ELISA but negative by POCT and two (patients 11 and 27) were negative by ELISA but positive by POCT. Patient 31 had a normal colonic biopsy and magnetic resonance imaging scan; patient 41 had large bowel colitis and a normal small bowel by biopsy. In both patients 11 and 27, there was a strong clinical suspicion of IBS and no further investigations were carried out.
Discussion
The Quantum Blue® method was simple, easy to use and able to produce more rapid results than the ELISA methodology. Each test, including extraction and incubation, took about 20 min. This was compared with four hours required for a result by ELISA, although it is appreciated that in this case, more samples can be run at one go. If a whole plate is used, this equates to 6.5 min per sample. An additional significant time saving could be made with the ELISA if faecal extraction devices were used. The use of such devices with the PhiCal® ELISA was not in practice at the time of this study but they were used with the POCT kit. We were able to show that there was no significant difference in results using weighing or the extraction devices, even if an additional variable of different kits was introduced. This supports the limited data available from Bühlmann comparing weighing and use of the Roche extraction devices. 11
The assay showed satisfactory linearity (r = 0.76, 95% confidence interval 0.29–0.97). We used Passing–Bablok regression analysis to compare slopes and intercepts with those reported by the manufacturer (Table 2). We found a much larger intercept than that reported by the manufacturer but with large confidence intervals; these were not quoted by the manufacturer. Due to cost limitations, we were not able to exactly duplicate the methodology of the manufacturer and run multiple applications of the same stool. This may have led to these discrepancies.
Comparison of the components of the regression line with those of the manufacturer
CI, confidence interval
/ Indicates CI from the manufacturer were not given
We found results −2 to +4 min on either side of the recommended 12-min reading point as the cartridge were all within 5%. Given that the analytical imprecision of the assay is quoted as 15%, 12 this was deemed acceptable.
In the patient comparison study, there was generally good agreement between the two methods, with only four patients showing results a different side of the 50 g/g cut-off recommended by the manufacturer. Our limited study suggested that the Quantum Blue® method may produce more false-positive results as in the two patients positive by this method but negative by ELISA, there was a strong clinical suspicion of IBS. It should be noted, however, that, firstly the lack of inflammation was not proven definitively by any invasive investigations as they were not deemed clinical appropriate. Secondly, our study only used a very limited number of patients and thirdly, as there is currently no gold standard for the measurement of calprotectin, it is difficult to assess which method gives the ‘correct’ result. Our existing ELISA has proven effectiveness in the literature 5,13 while there are currently no studies verifying the clinical utility of the Quantum Blue® method. There are a number of publications suggesting that an earlier semi-quantitative method 4,14,15 may be useful to rule out patients with calprotectins of <50 μg/g but none currently exist for the Quantum Blue® method.
Considering the absolute calprotectin values, there was poor agreement between the methods. This is hardly surprising given the different methodologies involved, the different antibody preparations (PhiCal® uses a polyclonal antibody, the Quantum Blue® method a monoclonal antibody) and the lack of a universal calprotectin reference standard. Given the wide biological variability in calprotectin in some patients, 16 this is probably of limited consequence. The key factor was that we were able to find little difference in the clinical performance of the assays.
Given the cost implications of using the Quantum Blue® method, it is doubtful that it will be used by laboratories which have sufficient numbers, equipment and time to warrant the ELISA methodology. The POCT device may have potential in the general practitioner community where the greatest cost-savings will be achieved. However, due to the lower prevalence of IBD in this population, the value of calprotectin in such a group remains to be proven. 17 The test may find value in a ‘one-stop’ gastrointestinal clinic where the result could be produced while the patient is waiting and so IBD could be ruled out while they were waiting.
DECLARATIONS