Validation and Update of the Lémann Index to Measure Cumulative Structural Bowel Damage in Crohn's Disease

Abstract

Background & aims: The Lémann Index is a tool measuring cumulative structural bowel damage in Crohn's disease (CD). We reported on its validation and updating.

Methods: This was an international, multicenter, prospective, cross-sectional observational study. At each center, 10 inclusions, stratified by CD duration and location, were planned. For each patient, the digestive tract was divided into 4 organs, upper tract, small bowel, colon/rectum, anus, and subsequently into segments, explored systematically by magnetic resonance imaging and by endoscopies in relation to disease location. For each segment, investigators retrieved information on previous surgical procedures, identified predefined strictures and penetrating lesions of maximal severity (grades 1-3) at each organ investigational method (gastroenterologist and radiologist for magnetic resonance imaging), provided segmental damage evaluation ranging from 0.0 to 10.0 (complete resection). Organ resection-free cumulative damage evaluation was then calculated from the sum of segmental damages. Then investigators provided a 0-10 global damage evaluation from the 4-organ standardized cumulative damage evaluations. Simple linear regressions of investigator damage evaluations on their corresponding Lémann Index were studied, as well as calibration plots. Finally, updated Lémann Index was derived through multiple linear mixed models applied to combined development and validation samples.

Results: In 15 centers, 134 patients were included. Correlation coefficients between investigator damage evaluations and Lémann Indexes were >0.80. When analyzing data in 272 patients from both samples and 27 centers, the unbiased correlation estimates were 0.89, 0,97, 0,94, 0.81, and 0.91 for the 4 organs and globally, and stable when applied to one sample or the other.

Conclusions: The updated Lémann Index is a well-established index to assess cumulative bowel damage in CD that can be used in epidemiological studies and disease modification trials.

Keywords: Bowel Damage; Crohn’s Disease; Lémann Index; Validation.

Figure 1.

Scatterplot of the investigator organ resection–free cumulative damage evaluation as a function of the organ Lémann Index in the validation sample, upper tract (A), small bowel (B), colon/rectum (C), anus (D), and of the investigator global damage evaluation as a function of the global Lémann Index in the validation sample (E). The area of each circle is related to the number of coincident data plotted there.

Figure 2.

Scatterplot of the updated organ predicted index vs the organ resection–free cumulative damage evaluation given by the investigator for (A) upper tract, (B) small bowel, (C) colon and rectum, (D) anus, and of the updated global predicted index vs the global damage evaluation given by the investigator (E). Results obtained on the development sample in blue and validation sample in red when analyzed together. For small bowel, 1 patient with a high level of damage (investigator damage evaluation of 77 was not shown in [B]). Predicted indexes were estimated from the multiple linear mixed model. The area of each circle is proportional to the number of coincident data plotted there. “r” is the unbiased estimate of the correlation coefficient between predicted organ (and global) damage indexes and investigator organ (and global) damage evaluations, derived from the bootstrap method.

Figure 3.

(A) Excel file allowing calculation of the Lémann Index in 1 patient. (B) User’s guide describing the 4 steps necessary to calculate the Lémann Index. In case of previous surgical resection, the investigator should type 100 if complete resection of the segment and the percentage of resected segment in case of partial resection. After identifying stricturing and penetrating lesions and their grade of severity at each mandatory examination according to CD location, the investigator should type the grade of maximal severity (1, 2, or 3) of each lesion present in a segment at each examination. At the end, the most severe lesion in each segment will be automatically selected. The Lémann Index can range between 0 (no damage) and 115 (maximal damage).