Introduction
Patients with coeliac disease (CD) and unsatisfactory clinical response to a gluten-free diet (GFD) are identified as having non-responsive coeliac disease (NRCD), and may require investigations to exclude possible complications of CD. Refractory coeliac disease (RCD) is the most common complication of CD and can be divided into RCD type 1 and type 2 based on intraepithelial lymphocyte (IEL) immunophenotyping. Methods for the assessment of IEL phenotype in RCD include traditional immunohistochemistry (IHC), PCR-based clonality analysis of TCR-γ gene rearrangement and flow cytometry (FC). However, the role of monitoring IEL immunophenotype with FC in patients with NRCD and known RCD is poorly defined. Moreover, very little is known about the utility of FC for evaluating IEL phenotype in patients with non-coeliac enteropathies (NCEs), for which differential diagnosis with RCD is challenging.
Aims & Methods
We aimed to investigate the reproducibility and significance of monitoring IEL immunophenotype with FC in patients with RCD, NRCD, and NCEs under regular follow-up at a referral centre. Patients who underwent FC for IEL immunophenotyping between January-2012 and February-2023 were divided into two groups: group 1) patients with known RCD or patients with NRCD in whom complications of CD were suspected; group 2) patients with NCEs lacking clinical and histological response. Data was retrospectively collected on clinical features, IEL immunophenotyping with FC, IHC, analysis of γ-TCR clonality, and on mortality and progression to lymphoma. An aberrant IEL phenotype was defined as >20% CD103+,CD3-,CD8-,CD3ε+ at FC. Reproducibility of FC results and concordance of FC with IHC and γ-TCR clonality analysis were evaluated. Overall survival and lymphoma-free survival were compared among coeliac patients based on IEL phenotype at FC.
Results
52 patients (18 RCD, 21 NRCD, 13 NCEs; 38 F, mean age at FC testing 55±13 years) underwent 100 total FC IEL phenotype determinations and an aberrant IEL phenotype was found at FC in 9/52 (7RCD, 2NRCD). 22 patients had at least two FC determinations during follow-up. Interestingly, immunophenotype (both aberrant and normal) remained unchanged in all these 22 patients (Cohen’s κ=1.00). Concordance of FC with IHC was fair (Cohen’s κ=0.40) but only minimal with γ-TCR clonality (Cohen’s κ=0.22). Over a median follow-up of 33 months (IQR 3-59) after FC, 6/52 (11%) progressed to lymphoma and 13/52 (25%) died. Mortality was markedly increased in coeliac patients with aberrant IELs compared to those with a normal immunophenotype (89% vs 7% mortality, HR 5.5, 95% CI 1.8-17.3, p<0.01). None of the 13 patients with NCEs showed an aberrant IEL phenotype at FC, although 3/13 developed lymphoma and 4/13 died (three lymphoma, one pulmonary adenocarcinoma).
Conclusion
FC is an accurate and reproducible tool for evaluating IEL phenotype in patients with RCD and NRCD and provides valuable prognostic information. Further study is needed on the diagnostic and prognostic role of FC in NCEs.
