Letter: stool adalimumab detection in ulcerative colitis and Crohn's disease

Abstract

We read with great interest the excellent review article reported by Rosen et al.1 In this review, the authors explain the conceptual framework for hypothesised mechanisms of rapid clearance of therapeutic anti-tumour necrosis factor (TNF) therapies in acute severe ulcerative colitis (UC). In their discussion, they proposed that dashboard systems could incorporate baseline covariates into a PK model to reduce unexplained variability, and proposed a dosing regimen estimated to result in optimal drug exposure for a given patient. We also think that faecal anti-TNF levels in inflammatory bowel disease (IBD) patients can explain a drop in circulating anti-TNF, and so could be useful in acute severe ulcerative colitis. In a preliminary study including 11 patients with severe UC during infliximab (IFX) induction treatment, Brandse et al. found an association between faecal IFX levels, a drop in circulating IFX trough levels (TLI), and treatment nonresponse.2 To our knowledge, no studies have been carried out on faecal anti-adalimumab (ADA), in cases of loss of therapeutic response in UC or Crohn's disease (CD) patients. In a retrospective study, we included from a clinical database and from biological collection data, the first 36 IBD patients with faecal calprotectin levels above of 1800 μg/g stools. Faecal anti-TNF assays were conducted on all of these patients and compared with the results of six IBD patients with a faecal calprotectin level below 500 μg/g (below 500 μg/g in three cases and below 100 μg/g in the other three cases). At the same time, we analysed trough levels of anti-TNF and antibodies. All measurements were obtained just before infusion or injection of anti-TNF drugs. Forty-two samples were analysed (20 CD, 22 UC, 20 cases on ADA treatment and 22 under IFX therapy). The 36 patients with faecal calprotectin levels >1800 μg/g exhibited clinical activity. Stool anti-TNF trough levels (IFX for five cases and ADA for two cases) (>0.2 μg/g of stools; median: 0.3–1.2 μg/g of stools) was reported in seven cases. Anti-TNF was found to be present in the stools in five cases of UC (22.7%) and in two cases of colonic CD (10%) (P = NS). A positive anti-TNF threshold of stools was only isolated in cases where calprotectin was over 1800 μg/g (19.4%). In the seven cases showing faecal anti-TNF, an endoscopy detected ulcers in the colonic mucosa (100%) as compared with 5/29 showing ulcers in the absence of faecal anti-TNF (14%, P < 0.05). In conclusion, faecal anti-ADA can be detected in patients with CD as well as UC, irrespective of the anti-TNF used. The presence of colonic ulcers appears to be a pre-condition of intestinal leaks. These data could explain the low serum ADA concentrations observed after induction treatment with an anti-TNF in some patients with acute severe IBD. Declaration of personal and funding interests: None.