P311 Measurement of functional blocade of TNF-alpha by anti-TNF agents is a stronger predictor than trough levels and anti-drug antibodies: 2-year prospective clinical data
P. Juillerat*1, P. Andrew2, J. Macpherson2, J. Cahenzli2, N. Patuto2, E. Slack2, F. Seibold1, K.D. McCoy2, A.J. Macpherson1, 2
1Clinic for Visceral Surgery and Medicine, Gastroenterology, Bern, Switzerland, 2University Bern, Maurice E Müller Laboratories, Universitätsklinik für Viszerale Chirurgie und Medizin, Inselspital, Bern, Switzerland
In case of loss of response (LOR) to anti-TNF agents in inflammatory bowel disease (IBD) patients, interventions, such as dose increase and shorten the interval, lead only to a transient improvement and a majority patients will eventually loss response. Patients could also be successfully switch within class following antibodies development. However, the LOR in some patients is due to a non anti-TNF driven pathway of inflammation. A functional in vitro test (CD-62L shedding) measuring TNF functional blockade should help us identified those specific situations.
An in vitro test was used to predict the response to the drug: the shedding of the L-selectin (CD62L) quantified by flow cytometry on the surface of granulocytes before and after the anti-TNF agent administration. In a subgroup of patients trough level of the drug (TL) and antibodies against the drug (ADA) have been performed in order to compare both tests. The treatment strategy during the 2 years of the study was blinded to the Results of the CD62L shedding, TL and ADA and followed clinical symptoms-based interventions or switch by IBD specialists.
From June 2012 to October 2014, 33 IBD treated with anti-TNF agents at Bern University Hospital were followed prospectively (clinicians blinded) to correlated clinical outcome with their response profile tested at baseline. The 22 responders (R) and 11 non responders (NR) had similar clinical characteristics. During a median follow up of 22 months (range 7- 26; 53 patient-years), 17 medico- surgical events occurred (3 adverse events, 1 CMV colitis, 9 flares, 3 intestinal resections and 1 de novo fistula) 5 in R and 12 in NR, which means 11% vs. 60% (p< 0.001) of the patient-year follow up. The delta of calprotectin between year 1 and year 2 of follow up was 261 for NR and 1.3 for R. ADA and TL measurement could be performed in 15 patients (45%; 9 R and 6 NR). Only 2 patients developed ADA (one in each group). There was no significant difference in trough levels between R and NR (2.8 vs. 4.8; p=0.4) and 62% had a therapeutic level ( >1.5).
Patients stable without need for intervention were 16/20 (84%) in R vs. 1/11 in NR (p<0.001). In the NR group all the dose optimisation failed, whereas in the responders group, interventions that would have been suggested on the basis of TL and ADA have not been performed (clinicians blinded), but were finally not required, based on the favourable clinical outcome.
Testing the in vitro functional blockade of TNF alpha (CD62L shedding) in anti-TNF treated IBD patients seem to be a better long term predictor than trough levels and antibodies against the drug measurements.
This could certainly minimize interventions and therefore reduce costs and risk of adverse events.