D0373 - Relationship Between Hospitalization, Surgery and Achievement of Clinical Remission or Clinical Response in Moderate to Severe Crohn’s Disease Patients: Results From the UNITI/IM-UNITI Trials

Introduction: Crohn’s disease (CD) is a chronic disease with high burden for patients and payers. Treatment aims include induction and maintenance of clinical remission (CRem). Understanding the relationship between rates of hospitalization and surgery and health states can help determine the economic burden. This analysis used results from the UNITI/IM-UNITI trials (CNTO1275CRD3001-3) of patients with moderate to severe CD, to determine the relationship between hospitalization and surgery rates and each health state over 1 year.

Methods: Hospitalization and surgery data from UNITI/IM-UNITI were pooled across arms and across induction and maintenance phases. Based on CRem and response achieved at each visit (weeks 0-52), patients’ exposure time were classified into three health states (HS) using the Crohn's Disease Activity Index score: CRem, clinical response without clinical remission (CResp) and active CD. Hospitalization and surgery events were assigned based on the HS at the visit prior to the event. The base case (BC) analysis included CD-related hospitalizations and surgeries only, a first sensitivity analysis was performed including both CD- and non-CD-related hospitalizations and surgeries, and a second sensitivity analysis was performed on patients randomized to maintenance only. Annualized rates were estimated by HS adjusting for the exposure time in each health state.

Results: The analysis included a total of 773 patients in CRem, 723 in CResp and 1,255 in active CD. BC results showed significantly fewer hospitalizations per 100 patient-years (PY) in CRem (6.50), and numerically fewer per 100 PY in CResp (14.37) compared to active CD (20.82). Similarly, patients in CRem had significantly fewer surgeries per 100 PY vs active CD (4.18 vs 15.57), and the difference was not statistically significant vs patients in CResp, who incurred 11.35 surgeries/100 PY. P-values are reported in Figure 1. CRem (9.11 days) and CResp (9.86 days) also had a numerically shorter mean duration of hospitalization compared to active CD (15.77 days). These results were consistent with those from both sensitivity analyses.

Discussion: Annualized hospitalization and surgery rates from the UNITI/IM-UNITI trials were significantly lower for patients in CRem and lower but non-significantly for patients in CResp, compared to patients in active CD. CRem and CResp had a numerically shorter mean duration of hospitalization compared to active CD.