Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P472. Head-to-head comparison of 5 fecal markers to predict response to induction and maintenance therapy with infliximab in ulcerative colitis patients; a prospective study

A.C. Frin1, S. Nancey2, J. Filippi1, G. Boschetti2, B. Flourié2, J. Drai3, P. Ferrari4, X. Hebuterne1, 1Archet 2 University Hospital, Gastro enterology, Nice, France, 2Hospices Civils de Lyon, Lyon-Sud hospital, Gasroenterology, Pierre Benite, France, 3Hospices Civils de Lyon, Lyon-Sud hospital, Biochemistry, Pierre-Benite, France, 4Pasteur University Hospital, Biochemistry, Nice, France

Background

The role of faecal markers in monitoring anti TNF alpha therapies has been insufficiently explored. This study aimed to determine the usefulness of five faecal proteins in the prediction of clinical response to Infliximab (IFX) in Ulcerative Colitis (UC): calprotectin (fCal), Lactoferrin (fLac), M2-PK (fM2-PK), neopterin (fNeo), and zonulin (fZo).

Methods

Thirty-one consecutive patients with an active UC, requiring IFX [5 mg/kg at week 0 (W0), 2, 6 and every 8 W] were prospectively studied. At W0, W2, W6 and W14, clinical activity was recorded and a stool sample collected. Clinical response to induction therapy was defined at W14 as a reduction of at least 3 points and 30% of the Mayo score. In 25 patients, endoscopies were performed at W0 and W12; an endoscopic Mayo subscore of 0 or 1 defined endoscopic remission. Clinical response to maintenance therapy was evaluated at W52 and optimization or discontinuation of IFX were considered as a failure.

Results

At W0 the median partial Mayo Score, the endoscopic Mayo and the UCEIS scores were 7/9 (2–9), 3/3 (2–3) and 8/11 (6–11) respectively. At W14, 19 patients (61%) were clinical responders and 13 (52%) experienced an endoscopic response. The median levels of fCal drop dramatically from W0 to W14 in responders [from 4260 µg/g (96–25051) to 128 µg/g (11–3782); p = 0.0001]. In contrast, it did not differ significantly in non-responders [from 9077 µg/g (215–50000) to 2781 µg/g (203–14149); p = 0.287]. Same trends were observed for fLac and fM2-PK levels. At W2, fLac and fM2-PK predicted accurately clinical response to IFX induction [area under the curve (AUC) = 0.82, 0.84 and 0.88 respectively]: cuts-offs of 800 µg/g for fCal, 20000 ng/g for fLac and 50 UI/mL for fM2-PK determined by ROC curves allowed to discriminate clinical responders from non responders to induction therapy, with good sensitivities (Se) (82%, 81% and 88%, respectively), and specificities (Sp) (69%, 70% and 80%, respectively). FLac measured at W2 were the more valuable marker to predict endoscopic remission at W12 (AUC=0.80, Se and Sp=72% with a cut-off of 32891 ng/g). At W14, the three previous markers were also reliable to predict clinical response at W52 (AUC=0.82, 0.86 and 0.75 respectively) with best cut-offs of 146 µg/g for fCal, 3457 ng/g for fLac and 2.25 UI/mL for fM2-PK. FCal, fLac and fM2-PK were well correlated with both the endoscopic Mayo subscore and the UCEIS. FNeo and fZo did not show any relevant result.

Conclusion

FCal, flac and fM2-PK predicted with a good accuracy the clinical response to induction and maintenance IFX therapy in UC. The measurement of one of these markers at W0 and at the end of induction might distinguish responders from non responders to IFX maintenance therapy within one year.