From the AGA Journals

Utility of three endoscopic IBD activity scores highlighted in two studies

View on the News

Cutoff values are still a moving target

The research field of measurements is fundamental for improving management strategies in inflammatory bowel diseases (IBDs) since the process can progress no faster than its key measured variables. In recent years, there has been a growing interest in mucosal healing as a goal of treatment in both CD and UC, since it seems to serve as a surrogate marker for long-term well-being and reduced complications (and this is our true ultimate goal of treatment).

However, the paucity of data on how to best measure mucosal healing limits our ability to treat patients to this target. Is mucosal healing best measured by macroscopic appearance of the mucosa? If so, how can we accurately quantify the degree of inflammation, and what cutoff scores are associated with our ultimate goal? Perhaps other markers of mucosal healing are superior, such as histology, calprotectin (which may be more sensitive for microscopic inflammation), or magnetic resonance (MR) enterography, which measures the entire thickness of the bowel wall and the entire length of the small bowel. Even after determining the best measurement tool and its cutoff values, change over time (i.e., responsiveness) and reliability must be determined before its implementation in treatment algorithms.

In the November issue of Gastroenterology, Ferrante et al. showed in the robust SONIC data set that a reduction of 50% in the CDEIS (CD Endoscopic Index of Severity) and the simplified SES-CD (Simple Endoscopic Score for CD), or mucosal healing by week 26, predicts to some extent 1-year clinical remission in CD. In the second study, Travis et al. elegantly report on the reasonable reliability and initial validity of the only validated endoscopic score in UC, the UCEIS (Ulcerative Colitis Endoscopic Index of Severity) score. Both studies present important data that advance our knowledge one step further toward the "treat to target" concept in IBD.

However, the two studies also highlight the shortcomings of our endoscopic definition of mucosal healing and their imperfect reliability, validity, and prediction power. Ferrante et al. should be applauded for reporting diagnostic utility statistics and not only P values in evaluating prediction power. Modest sensitivity/specificity of 74%-75%/45%-48% indicate that mucosal healing can be regarded as another important predictive variable in the decision-making process but not yet alone as the gold standard that mandates treatment escalation.

The studies facilitate significantly our evaluation of endoscopic mucosal healing, but as the authors of both studies rightly indicate, more studies are needed to further define our moving "target" before implementing it in IBD treatment algorithms.

Dr. Dan Turner is head of the pediatric gastroenterology and nutrition unit, Hebrew University of Jerusalem, Shaare Zedek Medical Center. He has no relevant conflicts of interest.


 

FROM GASTROENTEROLOGY

A reduction of at least 50% in either of two different Crohn’s disease endoscopic activity scores after 26 weeks of treatment predicted those patients likely to be off of corticosteroid therapy after about 1 year of treatment, with a sensitivity approaching 75%, reported Dr. Marc Ferrante and his coinvestigators.

The study provides evidence that this cutoff could be used as a reliable predictor of clinical response, but it needs to be studied further, according to the authors, who are members of the International Organization for the Study of Inflammatory Bowel Diseases. The study was published in the Novemberissue of Gastroenterology.

Another study, published in the same issue of Gastroenterology, evaluated the reliability of the Ulcerative Colitis Endoscopic Index of Severity (UCEIS), an instrument recently created to assess the endoscopic severity of UC. This study randomized 25 investigators to assess and score 28 endoscopic videos of patients with UC – without knowledge of clinical features in most cases. The results indicated that the UCEIS provides a "satisfactory" level of intrainvestigator and interinvestigator agreement and that it is "a reliable instrument for measuring the endoscopic disease activity of UC," according to Dr. Simon Travis, of the translational gastroenterology unit at John Radcliffe Hospital, Oxford, England, and his coinvestigators.

The version of the UCEIS that provides a score from 0 to 8, based on findings of vascular pattern, bleeding, and erosions/ulcers, is the "favored version" but needs to be validated further, they concluded (Gastroenterol. 2013 July 29 [doi: 10.1053/j.gastro.2013.07.024]).

In the Crohn’s disease (CD) study, which used data on 172 patients enrolled in SONIC (Study of Biologic and Immunomodulator Naive Patients in Crohn’s Disease), Dr. Ferrante, of the department of gastroenterology at University Hospitals Leuven, Belgium, and his colleagues identified a cutoff point in two endoscopic response scores after 26 weeks of treatment that was predictive of clinical response at 50 weeks.

At baseline, the patients had endoscopic lesions and CD Activity Index (CDAI) scores of 220-450 points (median, 277). Their median age was 34 years, and they had been diagnosed with CD for a median of 2.5 years. The patients were dependent on corticosteroids and were randomized to treatment with infliximab (Remicade) infusions and/or oral azathioprine, and were followed to 50 weeks. The primary endpoint of this study was corticosteroid-free clinical remission (CFREM) at week 26, and one of the secondary endpoints was complete healing of mucosal ulcerations.

The investigators evaluated different cutoff points of endoscopic responses for two methods used to score the severity of endoscopic lesions: SES-CD (Simple Endoscopic Score for CD) and CDEIS (CD Endoscopic Index of Severity).

Almost half of the patients achieved mucosal healing at the 26th week.

A decrease from baseline in the SES-CD of at least 50% at week 26 "appeared to be the best discriminative cutoff value" for predicting patients most likely to be in clinical remission (off steroids at week 50). This degree of SES-CD response was met by 112 (65%) patients, with 74% sensitivity and 48% specificity for predicting CFREM at week 50.

For the CDEIS values, the best cutoff was at least a 45% drop from baseline at week 26 for predicting patients most likely to achieve CFREM at week 50, with 75% sensitivity and 45% specificity.

Because of the "subtle" difference between the two measures (50% and 45% cutoffs), "we propose a definition of endoscopic response with a decrease in baseline in CDEIS of at least 50% at week 26 for both tests," the authors wrote. (For CDEIS, the 50% cutoff had a sensitivity of 73% and a specificity of 46%.)

"As such, this endoscopic endpoint could serve as a reliable predictor of the midterm clinical outcome of therapies," Dr. Ferrante and his coauthors concluded. A 50% reduction in endoscopic activity from baseline also has potential for use as an endpoint in studies evaluating pharmacotherapy or treatment strategies "to show healing capacity," they wrote.

These cutoff values should be validated in a prospective study, which should also look at correlations between the endoscopic response and "disease-modifying long-term outcomes," such as a sustained clinical response or surgery, they added.

In the study that evaluated the reliability of the UCEIS, 25 investigators from North America and Europe were randomized to assess 28 of 57 sigmoidoscopy videos of patients with UC, which included some duplicates to evaluate intraobserver reliability. Clinical details such as the number of stools per day were provided for only two of the videos. They were also trained.

The UCEIS provides a score of 0 to 8 based on the sum of three descriptors: vascular pattern (normal, patchy obliteration, or obliterated), bleeding (none, mucosal, luminal mild, luminal moderate, or severe), and erosions and ulcers (none, erosions, superficial ulcer, or deep ulcer). Almost 700 video evaluations were performed.

Pages

Next Article: