Using Interpretable Machine Learning to Identify Baseline Predictive Factors of Remission and Drug Durability in Crohn's Disease Patients on Ustekinumab.

Ustekinumab has shown efficacy in Crohn's Disease (CD) patients. To identify patient profiles of those who benefit the most from this treatment would help to position this drug in the therapeutic paradigm of CD and generate hypotheses for future trials. The objective of this analysis was to determine whether baseline patient characteristics are predictive of remission and the drug durability of ustekinumab, and whether its positioning with respect to prior use of biologics has a significant effect after correcting for disease severity and phenotype at baseline using interpretable machine learning. Patients' data from SUSTAIN, a retrospective multicenter single-arm cohort study, were used. Disease phenotype, baseline laboratory data, and prior treatment characteristics were documented. Clinical remission was defined as the Harvey Bradshaw Index ≤ 4 and was tracked longitudinally. Drug durability was defined as the time until a patient discontinued treatment. A total of 439 participants from 60 centers were included and a total of 20 baseline covariates considered. Less exposure to previous biologics had a positive effect on remission, even after controlling for baseline disease severity using a non-linear, additive, multivariable model. Additionally, age, body mass index, and fecal calprotectin at baseline were found to be statistically significant as independent negative risk factors for both remission and drug survival, with further risk factors identified for remission.

Long-Term Real-World Effectiveness and Safety of Ustekinumab in Crohn's Disease Patients: The SUSTAIN Study.

BACKGROUND: Large real-world-evidence studies are required to confirm the durability of response, effectiveness, and safety of ustekinumab in Crohn's disease (CD) patients in real-world clinical practice. METHODS: A retrospective, multicentre study was conducted in Spain in patients with active CD who had received ≥1 intravenous dose of ustekinumab for ≥6 months. Primary outcome was ustekinumab retention rate; secondary outcomes were to identify predictive factors for drug retention, short-term remission (week 16), loss of response and predictive factors for short-term efficacy and loss of response, and ustekinumab safety. RESULTS: A total of 463 patients were included. Mean baseline Harvey-Bradshaw Index was 8.4. A total of 447 (96.5%) patients had received prior biologic therapy, 141 (30.5%) of whom had received ≥3 agents. In addition, 35.2% received concomitant immunosuppressants, and 47.1% had ≥1 abdominal surgery. At week 16, 56% had remission, 70% had response, and 26.1% required dose escalation or intensification; of these, 24.8% did not subsequently reduce dose. After a median follow-up of 15 months, 356 (77%) patients continued treatment. The incidence rate of ustekinumab discontinuation was 18% per patient-year of follow-up. Previous intestinal surgery and concomitant steroid treatment were associated with higher risk of ustekinumab discontinuation, while a maintenance schedule every 12 weeks had a lower risk; neither concomitant immunosuppressants nor the number of previous biologics were associated with ustekinumab discontinuation risk. Fifty adverse events were reported in 39 (8.4%) patients; 4 of them were severe (2 infections, 1 malignancy, and 1 fever). CONCLUSIONS: Ustekinumab is effective and safe as short- and long-term treatment in a refractory cohort of CD patients in real-world clinical practice.

This large retrospective study demonstrated the short- and long-term effectiveness and safety of ustekinumab in patients with Crohn's disease in real-world clinical practice, including those with refractory disease.

Prognostic factors of liver cirrhosis mortality after a first episode of spontaneous bacterial peritonitis. A multicenter study.

INTRODUCTION: Spontaneous bacterial peritonitis is an infectious complication with a negative impact on survival of patients with cirrhosis. OBJECTIVE: To analyze the short- and long-term survival after a first episode of bacterial peritonitis and the associated prognostic factors. PATIENTS AND METHODS: This was a retrospective, multicenter study of patients admitted to hospital for spontaneous bacterial peritonitis between 2008 and 2013. Independent variables related to mortality were analyzed by logistic regression. The prognostic power of the Child Pugh Score, the Model for End-Stage Liver Disease (MELD) and the Charlson index was analyzed by ROC curve. RESULTS: A total of 159 patients were enrolled, 72% were males with a mean age of 63.5 years and a mean MELD score of 19 (SD ± 9.5). Mortality at 30 and 90 days and one and two years was 21%, 31%, 55% and 69%, respectively. Hepatic encephalopathy (p = 0.008, OR 3.5, 95% CI 1.4-8.8) and kidney function (p = 0.026, OR 2.7, 95% CI 1.13-16.7) were independent factors for short- and long-term mortality. MELD was a good marker of short- and long-term survival (area under the curve [AUC] 0.7: 95% CI 1.02-1.4). The Charlson index was related to long-term mortality (AUC 0.68: 95% CI 0.6-0.77). CONCLUSIONS: Short- and long-term mortality of spontaneous bacterial peritonitis is still high. The main prognostic factors for mortality are impairment of liver and kidney function. MELD and the Charlson index are good markers of survival.