Identifying hospitalized patients with acute severe ulcerative colitis (ASUC) who will be refractory to corticosteroid therapy and require rescue therapy remains difficult. Hypoalbuminemia worsens with time during hospitalization and is associated with rapid clearance of and reduced response to infliximab (IFX) rescue. Early use of rescue therapy may therefore be more effective. Simple clinical and laboratory predictors of corticosteroid responsiveness would facilitate earlier use of rescue therapy.
Retrospective study of a prospectively maintained database of 3600 patients attending a single center was conducted. Patients with histologically confirmed ulcerative colitis admitted with ASUC over a 5-year period from January 2010 to December 2014 were identified. All patients initially received intravenous corticosteroids. Patient demographics were collected; C-reactive protein (CRP) and albumin levels were recorded at baseline and during admission. Receiver operating characteristic statistics were used to determine the optimal stool frequency, CRP, albumin, and CRP/albumin ratio (CAR) to predict steroid response.
A total of 124 ASUC patients were admitted during a 5-year period. Median follow-up was 2.3 years. A total of 62 patients (50%) were steroid responsive, 55 patients (44%) received rescue IFX, 22 patients (18%) required colectomy within 30 days of admission, whereas a further 14 (11%) required colectomy during follow-up. By receiver operating characteristic statistics, day 3 CAR was a more accurate marker of steroid responsiveness than day 3 CRP or day 3 albumin alone [area under curve=0.75 (P<0.001)]. The optimal CAR to predict response to steroids on day 3 was 0.85 (sensitivity 70%, specificity 76%). When combined with D3 stool frequency, specificity improved to 83%. If at day 3, CAR was >0.85 and stool frequency was >3, the relative risk of steroid nonresponse was significantly raised at 3.9 (95% confidence interval, 2.1-7.2).
Raised D3 CAR is an early predictor of steroid-refractory ASUC. When combined with D3 stool frequency, its predictive ability improves. In patients with predicted steroid nonresponse, early introduction of rescue IFX at this stage may be more effective, before serum albumin falls profoundly.
*Centre for Colorectal Disease, St Vincent’s University Hospital
†School of Medicine, University College Dublin, Dublin, Ireland
D.J.G.: is acting as submission’s guarantor. D.J.G., G.C., S.T.M., and G.A.D.: were involved in concept and design. D.J.G., K.H., J.D., J.N., K.B., D.K., M.B., J.S., and GH: were involved in patient recruitment and data collection. D.J.G., G.A.D., and G.C.: were involved in drafting of manuscript. D.J.G. and H.E.M.: were involved in data analysis.
The authors declare that they have nothing to disclose.
Address correspondence to: David J. Gibson, MB, ChB, BAO, MD, Centre for Colorectal Disease, St Vincent’s University Hospital, Elm Park, Dublin 4, Ireland (e-mail: firstname.lastname@example.org).
Received November 11, 2016
Accepted May 19, 2017