Final Progress Report

Proposal No. IBD-0211
Principal Investigator:  Lee A. Denson, M.D.
Applicant Organization:  Cincinnati Children's Hospital Medical Center (Ohio, U.S.A.)
Project Title:  GM-CSF bioactivity and IBD phenotype
Period of Award:  October 1, 2007 - September 30, 2009

Project Aims and Significant Results

Aim 1. Characterize anti-GM-CSF antibody regulation of monocyte/neutrophil function and IBD phenotype. Hypothesis: Endogenous anti-GM-CSF antibodies regulate monocyte/neutrophil function and IBD phenotype. The serum concentration of neutralizing anti-GM-CSF antibodies will be determined and related to alterations in monocyte/neutrophil function and IBD phenotype in patients with both pediatric and adult onset disease. The CARD15 genotype and titers of ASCA, pANCA, OmpC, I2, and CBir1 will be determined and related to anti-GM-CSF levels and IBD phenotype and behavior. These data will determine whether neutralizing anti-GM-CSF antibodies regulate innate immunity and IBD phenotype, relative to current genetic and serologic markers.

We made significant progress in enrolling pediatric and adult onset patients for the studies in this Aim, and in performing all of the studies needed to test our hypotheses. We enrolled 555 pediatric onset IBD patients, and, now in collaboration with Dr. Judy Cho at Yale, 385 adult onset IBD patients, and determined their serum anti-GM-CSF concentration and IBD phenotype. We found that the serum anti-GM-CSF concentration is elevated in both pediatric and adult onset CD patients with ileal or ileo-colonic involvement (L1_L3), relative to those with colon-only CD (L2), UC, or healthy controls. After controlling for age of onset and small bowel versus colon-only involvement, we found that the rate of stricturing/penetrating disease behavior is increased two-fold with elevated serum anti-GM-CSF, and exceeds 80% in adult-onset CD. We determined results for the other IBD serologic assays in the pediatric cohort. We found that CD patients with elevated serum anti-GM-CSF exhibit a high frequency of sero-positivity for both ASCA IgA and IgG (approximately 65% positive if also anti-GM-CSF positive, versus approximately 15% positive if anti-GM-CSF negative), but not CBir1, OmpC, or I2. We measured serum anti-GM-CSF in parents and affected children with CD (n=115), and to date have found that the serum anti-GM-CSF level correlates between affected probands with CD and their unaffected parents. We have tested for a relationship between elevated serum anti-GM-CSF and monocyte and neutrophil function in a subset of pediatric-onset patients. We found that elevated levels of anti-GM-CSF neutralize GM-CSF in a sensitive neutrophil CD11B stimulation assay. We previously reported that neutrophil phagocytosis is quite sensitive to inhibition by anti-GM-CSF. Consistent with this, we found that neutrophil phagocytic capacity is reduced in CD patients with elevated anti-GM-CSF. Moreover, we found that neutrophil phagocytic capacity is inversely related to serum anti-GM-CSF in both healthy controls (r=-0.71, p=0.001) and CD patients (r=-0.46, p=0.04). By comparison, we observed no difference in neutrophil oxidative burst between CD patients with high versus low serum anti-GM-CSF. These results have been published in Gastroenterology and Blood this year, and replication of these results in the larger adult cohort with Dr. Cho has been submitted to the 2010 DDW meeting.

Aim 2. Determine the effect of anti-GM-CSF antibodies and CARD15 deficiency upon mucosal barrier function. Hypothesis: Anti-GM-CSF antibodies promote increased intestinal permeability in CD.   Intestinal permeability will be determined using the lactulose:mannitol test in CD patients in remission with elevated and normal serum anti-GM-CSF. CARD15 genotype and fecal S100A12 will be determined and controlled for in the analysis. Antibodies and immune responses (LBP and sCD14) to endotoxin will be determined and related to anti-GM-CSF level and intestinal permeability.

We enrolled 30 subjects with pediatric-onset CD for the studies in this Aim, 15 with elevated serum anti-GM-CSF, and 15 with low serum anti-GM-CSF, as well as 15 healthy pediatric controls with no familiy history of IBD. All subjects tolerated the L:M solution well, and provided urine samples which yielded results. We found that the mean (SD) urinary L:M ratio is increased from 0.038 (0.02) in CD with low serum anti-GM-CSF, to 0.07 (0.03) in CD with high serum anti-GM-CSF. The L:M ratio in healthy controls was also lower than that observed in CD with high serum anti-GM-CSF, and did not differ from CD with low serum anti-GM-CSF. Intestinal inflammation, as measured by fecal S100A12, did not vary between the two CD groups, and was significantly higher than the control group. The fecal S100A12 was measured by Dr. Dirk Foell, who Dr. Denson first met at the annual BMRP meeting. This has led to an active ongoing collaboration. Antibody responses to endotoxin were also increased in the CD group with high serum anti-GM-CSF, compared to the CD group with low serum anti-GM-CSF, and healthy controls. Taken together, these exciting results haven confirmed a specific increase in intestinal permeability and endotoxin exposure in pediatric CD patients with high serum anti-GM-CSF. These data have been submitted to the 2010 DDW meeting, and a manuscript is being written.

Lay Summary 

New therapies which block specific pro-inflammatory proteins in the body are more effective in healing the intestine in IBD than the current first line therapies such as steroids. However, recent studies have suggested that at least some patients with Crohn’s disease have a primary problem with an immune system that is too weak, and can not control bacteria in the intestine. We have sought to define fundamental features of the causes of IBD amenable to targeted therapies in specific subsets of patients. Our studies have now defined a novel basis for dysfunction of white cells in the body needed to control bacteria, involving naturally occuring proteins, antibodies, which block the action of the Granulocyte Macrophage Colony Stimulating Factor (GM-CSF) protein.  GM-CSF is needed for the white cells to controll bacteria in the intestine. We have discovered that IBD patients who have these antibodies which block GM-CSF are at increased risk for early surgery. Our recent studies supported by the Broad Medical Research Program have: 1) defined this increased risk for surgery in hundreds of children and adults with IBD, 2) determined how white cell function is reduced in patients with these antibodies against GM-CSF, and 3) discovered that the intestine is more leaky in patients with these antibodies against GM-CSF, allowing more bacteria through. We have currently developed a new diagnostic test to measure the antibodies against GM-CSF. In the near future, this test may provide a strategy which will define whether an individual patient may benefit from receiving GM-CSF itself as a medical therapy, to boost the anti-bacterial function of their white cells and seal their leaky intestine. Collectively, these studies have provided a significant advance towards achieving targeted therapy for IBD patients.