TUG1 and H19 lncRNAs Can Predict Anti-TNF Unresponsiveness in Patients With Ulcerative Colitis: A Machine...

The present study aimed to explore the association of long noncoding RNAs (lncRNAs) with inflammation, disease activity, and predicting response to anti-tumor necrosis factor (TNF)-α therapy in patients with ulcerative colitis (UC). Whole blood samples and inflamed biopsies were collected from 42 UC patients at baseline (W0) and week 14 (W14) after receiving anti-TNF treatment as a discovery cohort. Colonoscopy images, histopathological, and clinical symptoms were used to monitor disease activity and response to treatment. LncRNA expression analysis showed increased expression of H19 in the active lesions of UC nonresponders (UCNs) compared to UC responders (UCRs) at baseline, whereas taurine upregulated gene 1 (TUG1) expression was lower. Higher expression of H19 in UCN compared to UCR was still observed at W14, whereas its expression was downregulated in UCR in the remission phase at W14 versus W0, suggesting it as a marker for monitoring disease activity. Moreover, colonic expression of H19 was positively correlated with erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). In blood, both H19 and TUG1 showed increased expression in UCN versus UCR at baseline and W14. Three-fold cross-validation-based machine learning approach and receiver operating characteristic (ROC) curve analysis showed that H19 and TUG1 had strong predictive performance for anti-TNF response, with accuracies of 90% and 93% in tissue and 85% and 60% in blood, respectively. In the validation cohort (12 UCR and 10 UCN), expression patterns were reproduced, and predictive performance remained high. H19 showed better accuracy in blood (85%) than tissue (70%), while TUG1 performed best in tissue (92%) and also remained highly accurate in blood (94%). The distinct expression of lncRNAs showed that they could play an important role in the response of UC patients to treatment. They can be a potential biomarker to monitor disease activity and predict response to anti-TNF treatment in UC patients.