# An Observational Study on Clinical Insights and Outcome Prediction of Lupus Enteritis in Indian Systemic Lupus Erythematosus Patients

## TL;DR

This study examines lupus enteritis in Indian SLE patients, analyzing clinical features and outcomes to identify predictors of recurrence and mortality.

## Contribution

The study provides new clinical insights and evaluates predictive models for LE recurrence and mortality in an Indian SLE cohort.

## Key findings

- 27.1% of patients experienced recurrent LE, with hydroureteronephrosis more common in these cases.
- Colon and bladder involvement were more frequent in recurrent LE cases.
- Anti-dsDNA antibodies, complement, SLEDAI, and albumin failed to predict recurrence or mortality.

## Abstract

Lupus enteritis (LE) is the most common serious gastrointestinal manifestation of SLE. Indian literature on LE is limited, while most of the data comes from China and Western series with recent efforts on developing prediction models for its occurrence & recurrence.

The objectives of the study were to analyse clinical and laboratory parameters of LE, to compare recurrent versus non-recurrent LE and to identify predictors of its recurrence and mortality.

Patients with LE who were admitted to the Rheumatology ward from 2018–2022 were considered cases. For cases, symptoms, abdominal CT findings, and outcome (recurrence or death) were recorded. Logistic regression was used for the prediction of recurrence and mortality.

Among 48 LE patients, 45 were female. The mean (± SD) age of the first enteritis episode was 27.04± 8.92 years. The most frequent extra-gastrointestinal manifestations were nephritis and arthritis (54% of cases). Mean disease duration, lupus nephritis, arthritis, and mean SLEDAI were higher in non-LE patients. Thirteen (27.1%) patients had recurrent LE & hydroureteronephrosis was more prevalent in them (p = 0.002). The logistic regression model using dsDNA complements and albumin failed to predict recurrence. Multiple repetitions of the confusion matrix for the mortality prediction model yielded variable ρ scores, suggesting insignificant accuracy prediction.

Colon and bladder involvement is more frequent in recurrent LE. Anti-dsDNA antibodies, complement, disease activity index, and serum albumin failed to predict recurrence and mortality in our cohort.