In this paper, the authors assess the impact of genetic deletion of the gut-homing CCR9 chemokine receptor on dextran sulfate sodium (DSS) induced colonic inflammation in mice. They find that, following DSS exposure for 7 days, CCR9-/- mice display impaired ability to recover from the DSS insult. We would like to bring the reader’s attention to what we consider the misrepresentation of existing information that is very relevant to their main conclusion.

Rather than summarizing the technical findings of their work, the authors choose instead to use the Conclusions section to warn the readers against “the use of CCR9-inhibitors in therapy of human IBD as they indicate a potential risk for patients with large intestinal inflammation”. The presumption on the part of the authors that their findings, obtained over the course of a couple weeks by administering detergent to genetically modified mice, might predict the results of pharmacological intervention in human patients is naïve, at best, but could be forgiven if it was one of many scientific observations all of which pointed in the same direction. Unfortunately, in this case their new observations are at odds with several existing preclinical and clinical studies with CCR9 antagonists. Some of these key existing studies are outlined below:

(1) Throughout 2009 and 2010, presentations were made at every major national and international gastroenterology meeting describing the results of the PROTECT-1 study, in which the CCR9 antagonist CCX282-B (Traficet-EN) was tested in a large trial (436 Crohn’s patients) for up to 1 year (e.g., Keshav et al. (2009) PROTECT-1 study demonstrated efficacy of the intestine-specific chemokine receptor antagonist CCX282-B (Traficet-EN) in treatment of patients with moderate-to-severe Crohn's disease. Gastroenterology 136(5): Suppl. 1; Keshav et al. (2009) PROTECT-1 study of intestine-specific chemokine receptor antagonist CCX282-B (TRAFICET-EN) in Crohn’s disease. Gut 58(II):A468; Keshav et al. (2010) Chemokine receptor antagonist CCX282-B (Traficet-EN) maintained remission of Crohn’s disease in PROTECT-1 study. Gastroenterology 138(5): Suppl. 1, S86). In fact, the last of these is cited by Wurbel et al. (Wurbel ref 8). Such abstract clearly states that the patients enrolled in the PROTECT-1 study had moderate to severe small bowel and/or colonic Crohn’s disease. The compound showed statistically significant improvement of both induction of response and maintenance of remission, with no evidence of disease worsening, as the Wurbel paper seems to warns us (2 years after the human study is completed!). A sizable fraction of the PROTECT-1 patients had colonic involvement and the study clearly demonstrated improvement across all patient types, regardless of disease localization. Wurbel et al. indicate several times in their paper that the PROTECT-1 study has shown promising activity for the CCR9 antagonist CCX282-B in Crohn’s patients; yet, somehow, the results of their mouse study must seem more relevant to the human condition than actual clinical data, given the warning expressed in their Conclusion.

(2) Genetic deletion of drug targets in mice often result in phenotypes that differ from those obtained with pharmacological inhibition of the same target in mature animals. Often, these discrepancies reflect compensatory changes suffered by the ko mouse during embryofetal development. CCR9 deletion appears to be one such case. For example, we recently published (Walters et al. (2010) JPET 335: 61-69) on the profound positive effects of CCX282-B in the mouse deltaARE-TNF ileitis model. In contrast to these robust and positive findings, earlier work from Apostolaki et al. (Gastroenterology (2008) 134: 2025) showed that genetic deletion of CCR9 in these mice had no effect on progression of the disease. Very recently, a new paper with the same deltaARE-TNF/CCR9-/- mice (Wermers et al., Gastroenterology (2011) Feb 4 [doi:10.1053/j.gastro.2011.01.044]) observed increased ileitis. The availability of the positive clinical data with pharmacological inhibition (i.e., CCX282-B dosing to humans) should be evidence enough that genetic manipulation in mice may produce results that do not recapitulate the human outcome. In fact, we have recently shown (Walters et al., unpublished) that pharmacological CCR9 inhibition in the mouse DSS model under the conditions used by Wurbel et al. has no impact on disease symptoms, in contrast to the CCR9-/- phenotype.

(3) Using other mouse models of colonic inflammation, more representative of human colitis than the DSS model, we have demonstrated that prophylactic as well as therapeutic intervention with various CCR9 antagonists is highly beneficial. For example, CCR9 inhibition with 2 separate CCR9 antagonists, including CCX282-B, results in robust protection in the mdr1a-/- spontaneous mouse model (Jaen et al., (2010) CCR9 inhibition in the treatment of colonic inflammation. UEGW 2010, Barcelona, Abstract # OP459). This disease, similarly to human ulcerative colitis, is T-cell driven and influenced by intestinal flora.

In summary, we believe that scientific advancement requires the dissemination of new observations, as long as the experiments are carefully designed and controlled, particularly when they draw conclusions that contradict earlier work. However, we think that the ignorance of such earlier work, combined with an over-interpretation of lab experiments that conflict with such earlier work, such as Wurbel et al. display in their manuscript, should not be tolerated. In this light, we hope that you will see fit to publish our letter in the journal, such that your readership is able to reach their own conclusions on this matter.

Respectfully yours,

Thomas Schall and Juan Jaen
ChemoCentryx, Inc. Mountain View, CA 94043