Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P350. Serum CRP is a better early marker for response to infliximab induction therapy than fecal calprotectin in patients with moderate to severe ulcerative colitis

J.F. Brandse1, J.M. Jansen2, P.A. Baars3, M. Löwenberg1, C.Y. Ponsioen1, G.R. van den Brink1, G. D'Haens1, 1Academic Medical Center, Inflammatory Bowel Disease Centre, Amsterdam, Netherlands, 2Onze Lieve Vrouwe Gasthuis, Gastroenterology and Hepatology, Amsterdam, Netherlands, 3Academic Medical Center, Clinical Immunology, Amsterdam, Netherlands

Background

Serum CRP and fecal calprotectin are established makers for response to infliximab (IFX) maintenance therapy in Ulcerative Colitis (UC). However early measurement and its predictive value for response to IFX induction in UC remain uncertain. We aimed to define the optimal timing of measurement and compare both markers for response to therapy.

Methods

In this multicenter prospective observational study UC patients starting on IFX were included. Serum CRP and albumin and fecal calprotectin (Buhlmann ELISA) were measured at days 0, 1, 4, 7, 11, 14, 18, 21, 28, 42 during the first 6 weeks of induction therapy and Clinical Colitis Activity Index was monitored weekly. Absence of response was defined as the need for higher dose infusion during induction or colectomy within 3 months. Endoscopic response was defined as improvement at week 6–8 endoscopy.

Results

Fifteen UC patients (5/15 left-sided, 10/15 pancolitis, 14/15 Mayo endoscopic score 3) were included. All but one patients received IFX according to the regular induction regime of 5 mg/kg at week 0, 2, 6 and 7/15 with concomitant thiopurines. Median (IQR) baseline serum CRP was 36 (3–96) mg/L, albumin 38 (30–41) g/L and fecal calprotectin 1800 ug/g (647–1800). Median (IQR) serum CRP at day 4 was 59 (30–96) mg/l for patient with absence of response (n = 3) compared to 3.8 (1.3–11.3) for responders (n = 12), P = 0.01. Median (IQR) CRP at day 7 was 15.3 (2–35) mg/l for endoscopic non-responders (n = 8) compared to 1.6 (0.8–3.4) mg/l for endoscopic responders (n = 7), P = 0.06. CRP of >25 mg/l at day 4 seemed to be an optimal cut-off OR:175 (95% CI:2.9–10520, P < 0.01) for predicting absence of response. A cut-off for serum CRP of >5 mg/l at day 7 had an OR: 23 (95% CI:0.99–556, P = 0.02) for predicting lack of endoscopic response.

Both serum albumin and fecal calprotectin discriminated less or later between clinical and endoscopic response.

Analysis of fecal samples was however limited by the upper detection limit of the test (1800 ug/g).

Figure: Serum CRP during IFX induction therapy.

Table (abstract P350): Markers at various timepoints that significantly discriminate between absence of response and response or endoscopic response and non-response
Absence of response (n = 3) Median (IQR)Responders (n = 12) Median (IQR)P valuePredictive value cut-off
Baseline CRP (mg/l), Cut-off 75 mg/l124 (95–128)9 (2–41)P = 0.04OR: 54 (95% CI 1.76–1637, P < 0.01)
Day 1 CRP (mg/l), Cut-off 75 mg/l93 (80–123)7 (2–29)P = 0.04OR: 54 (95% CI 1.76–1637, P < 0.01)
Day 4 CRP (mg/l), Cut-off 25 mg/l59 (30–96)4 (1–11)P = 0.01OR: 175 (95% CI 2.9–10520, P < 0.01)
Day 4 Albumin (g/l), Cut-off 33 g/l29 (29–32)40 (37–46)P = 0.02OR: 54 (95% CI 1.76–1637, P < 0.01)
Endoscopic Non-Responders (n = 8)Endoscopic responders (n = 7)
Day 7 CRP (mg/l), Cut-off 5 mg/l15 (2–35)2 (1–3)P = 0.06OR: 23 (95% CI 0.99–556, P = 0.02)
Day 7 calprotectin (ug/g), cut-off 1300 ug/g1562 (1800–1800)465 (200–1410)P = 0.02OR: 55 (95% CI 1.86–1624, P < 0.01)
Day 14 calprotectin (ug/g), cut-off 750 ug/g1800 (1421–1800)222 (130–837)P < 0.01OR: 55 (95% CI 1.86–1624, P < 0.01)

Conclusion

Serum CRP is a better early marker for response to IFX therapy compared to fecal calprotectin or serum albumin in patients with UC. Optimal timing for measuring serum CRP seems to be day 4 and day 7 in order to predict absence of clinical or endoscopic response, with cut-off values of 25 mg/l and 5 mg/ml respectively. Larger numbers of patients are needed to confirm these findings.