Background: Two-thirds to three-fourths of patients with either refractory luminal or fistulizing Crohn disease respond to infliximab treatment. The ability or inability to respond seems to persist over time. Biological characteristics and/or genetic background can influence the response to treatment. The aim was to assess the value of C-reactive protein and TNF-alpha serum levels before treatment as well as the TNF -308 gene polymorphism in the prediction of response to infliximab treatment in Crohn disease.
Methods: Two-hundred-and-twenty-six Crohn disease patients treated in the setting of an expanded access programme to infliximab in Belgium were studied. There were 136 refractory luminal diseases and 90 refractory fistulizing diseases. Luminal diseases were treated with one single infusion; fistulizing diseases with three infusions at weeks 0, 2 and 6. A clinical response to treatment was defined as either a Crohn disease activity index <150 (complete) or a drop of 70 points (partial) at week 4, for luminal disease, and as either complete fistula healing (complete) or a decrease of at least 50% of the number of draining fistulas on two consecutive visits between weeks 0 and 18, for fistulizing disease. CRP and serum TNF-alpha levels were measured at week 0 before treatment and were compared between responders and non-responders. Patients were genotyped for the -308 TNF gene polymorphism, and allelic as well as genotype frequencies were compared between responders and non-responders.
Results: There were 73.2% responders (46.4% complete and 26.8% partial) and 26.8% non-responders. Response rates were similar in luminal and fistulizing diseases. CRP level before treatment was significantly higher in responders than in non-responders (16.8 mg/l (5-160) versus 9.6 mg/l (5-143); P = 0.02). Furthermore, response rate was significantly higher in patients with elevated CRP (>5 mg/l) than in patients with a normal CRP value (<5 mg/l) before treatment (76% versus 46%; P=0.004; OR: 0.26 (0.11-0.63)). Allelic and genotype frequencies for -308 TNF gene polymorphism were not significantly different between responders and non-responders--with the exception of a slightly higher TNF2 frequency in non-responders in luminal disease (22.1 % versus 11.6%; P = 0.04). However, this was not associated with a significant difference in genotype frequencies.
Conclusion: A positive clinical response to infliximab was associated with a higher CRP level before treatment in our population of Crohn disease patients, but there was no relevant association with -308 TNF gene polymorphism. We therefore suggest that CRP level may help to identify better candidates for infliximab treatment.