Y-ECCO Literature Review: Nathan Constantine-Cooke

Shahub and colleagues have previously developed a device which uses chemical impedance spectroscopy and target-specific antibodies to quantify concentrations of calprotectin, CRP and IL-6 in sweat. Having used a version of this device to monitor CRP and other cytokines in healthy populations [3, 4], the authors then turned their attention to monitoring IBD. They aimed to assess whether perspiration calprotectin, CRP and IL-6 differed between IBD patients with active and inactive disease. Linearity between perspiration and serum biomarkers, in addition to faecal calprotectin, was also explored.  

Methods

Shahub and colleagues recruited 33 IBD patients [12 ulcerative colitis (UC), 21 Crohn’s Disease (CD)] from a single centre in Chicago. Patients with immune-mediated comorbidities or recent exposure to methotrexate, anti-tumour necrosis factor or Janus kinase inhibitor therapies were excluded.

Disease activity was available for 24 subjects and was determined via serum levels and via the Mayo endoscopic score for UC and the simplified endoscopic score for CD.

The device was applied to the arms of subjects for 40–130 minutes, with perspiration CRP, calprotectin and IL-6 being recorded every minute. The device measured perspiration passively without actively stimulating the skin. Faecal calprotectin was collected for an unspecified subset of the cohort. 

The mean of the perspiration observations was calculated for each subject and used for comparison with serum data and faecal calprotectin. Coefficients of determination, R², were calculated using simple linear regression. Mann-Whitney U tests were used to assess whether perspiration, serum biomarkers and faecal calprotectin differed between subjects with endoscopically active and inactive disease.

Key findings

Whilst perspiration (p=0.03), serum (p=0.002) and faecal calprotectin (p=0.04) were found to differ significantly between the inactive and active disease groups, no difference was found with respect to perspiration or serum CRP. A significant difference in serum IL-6 (p=0.002) was reported between inactive and active disease, but the same behaviour was not observed for perspiration IL-6.

Close linear relationships were found between IL-6 and CRP in perspiration (R²=0.95), CRP and calprotectin in perspiration (R²=0.76) and calprotectin in serum and perspiration (R²=0.72).

Discussion

At a proof-of-concept level, the device appeared able to detect significant differences in calprotectin and IL-6 measured in sweat between subjects with active and inactive IBD. An obvious limitation of this study was the small sample size. Whilst 33 participants were recruited, only 13 of these had serum calprotectin available, reducing the effective sample size further for analyses involving serum calprotectin.

There is disagreement in the published literature regarding the utility of IL-6 for monitoring IBD activity. Whilst another small study previously reported serum IL-6 to differ significantly between IBD patients with inactive and active disease [5], a much larger study (n=378) found that serum IL-6 did not differ significantly between subjects with active and inactive disease although those with elevated IL-6 were more likely to have active disease [6]. In addition, a further study of 229 IBD patients suggested IL-6 to be clinically useful for CD but not for UC [7]. 

It should be noted that the longitudinal collection from the device was minimally utilised in this study, as the temporal observations were averaged. Whilst the next step is undoubtedly to conduct a larger-scale study, likely with wider inclusion criteria, later future work could explore the longitudinal nature of the data collected in more detail. For example, it may be possible that deep convolutional neural networks and/or signalling processing techniques and a classifier could be used to predict patients with endoscopically active and inactive disease. Similar approaches have previously been used to identify patients with atrial fibrillation from electrocardiogram data [8, 9].

Conclusion

The perspiration-based device assessed in this study could one day be used to allow disease activity to be non-invasively monitored from the comfort of patients’ homes. Moreover, the device’s ability to continuously monitor disease activity biomarkers presents unique potential for future research.

References

1. Plevris N, Lees CW. Disease monitoring in inflammatory bowel disease: evolving principles and possibilities. Gastroenterology 2022;162:1456–75.e1. doi: 10.1053/j.gastro.2022.01.024.
2. Mudter J, Neurath MF. IL-6 signaling in inflammatory bowel disease: pathophysiological role and clinical relevance. Inflamm. Bowel Dis 2007;13:1016–23. doi: 10.1002/ibd.20148.
3. Jagannath B, Pali M, Lin K-C, et al. Novel approach to track the lifecycle of inflammation from chemokine expression to inflammatory proteins in sweat using electrochemical biosensor. Adv Mater Technol 2022;7:2101356. doi: 10.1002/admt.202101356.
4. Jagannath B, Lin K-C, Pali M, Sankhala D, Muthukumar S, Prasad S. A sweat-based wearable enabling technology for real-time monitoring of il-1β and crp as potential markers for inflammatory bowel disease. Inflamm Bowel Dis 2020;26:1533–42. doi: 10.1093/ibd/izaa191.
5. Ogawa K, Matsumoto T, Esaki M, Torisu T, Iida M. Profiles of circulating cytokines in patients with Crohn’s disease under maintenance therapy with infliximab. J Crohns Colitis 2012;6:529–35. doi: 10.1016/j.crohns.2011.10.010.
6. Nikolaus S, Waetzig GH, Butzin S, et al. Evaluation of interleukin-6 and its soluble receptor components sIL-6R and sgp130 as markers of inflammation in inflammatory bowel diseases. Int J Colorectal Dis 2018;33:927–36. doi: 10.1007/s00384-018-3069-8.
7. Mavropoulou E, Mechie NC, Knoop R, et al. Association of serum interleukin-6 and soluble interleukin-2-receptor levels with disease activity status in patients with inflammatory bowel disease: a prospective observational study. PLoS One 2020;15:e0233811. doi: 10.1371/journal.pone.0233811.
8. Raghunath S, et al. Deep neural networks can predict new-onset atrial fibrillation from the 12-lead ECG and help identify those at risk of atrial fibrillation-related stroke. Circulation 2021;143:1287–98. doi: 10.1161/CIRCULATIONAHA.120.047829.
9. Yue Z, Jinjing Z. Atrial fibrillation detection based on EEMD and XGBoost. J Phys Conf Ser 2019;1229 (no. 1):012074. doi: 10.1088/1742-6596/1229/1/012074.

Nathan Constantine-Cooke - Short Biography

Nathan Constantine-Cooke is a postdoctoral research associate in Health Data Science in the Lees group at the University of Edinburgh. He is interested in characterising heterogeneity in Inflammatory Bowel Disease via longitudinal biomarker profiles and using these data to improve predictions for disease outcomes.