P411 Early ultrasound assessment predicts therapy response: an easy tool for clinical decision making
CalabresePhD, E.(1)*;Castiglione, F.(2);De Cristofaro, E.(1);Rispo, A.(2);Capacchione , C.(1);Testa, A.(2);Guarino, A.D.(2);Caiazzo, A.(2);Monteleone, G.(1);Zorzi, F.(1);
(1)University of Rome Tor Vergata, Gastroenterology Unit- Department of Systems Medicine, Rome, Italy;(2)Federico II- School of Medicine- Naples- Italy, Gastroenterology- Department of Clinical Medicine and Surgery, Naples, Italy;
Background
Intestinal ultrasound (IUS) is an effective and easy-to-use tool in monitoring Crohn’s disease (CD) lesions during different biological therapies. Aim of our study was to evaluate whether lesion improvement during biological therapy could predict transmural healing.
Methods
We performed a prospective study enrolling CD patients (pts) with indication to biological therapies. IUS and doppler US parameters at baseline, at 3 and 12 months were: bowel wall thickening (BWT), lesion length, echopattern, complications, blood flow according to Limberg’ score. Transmural healing (TH) was defined as normalization of all parameters; IUS responders were defined as pts with improvement of BWT associated with decreased lesion length, Limberg’ score improvement, and no worsening of the other parameters. Delta (Δ) signified the difference between IUS parameters at baseline and at 3 months. Changes in parameters and in Δ values at baseline and after 3 months were analyzed using Mann Whitney test; the area under (a ROC) curve was calculated. Differences between combination of parameters were tested by Chi-square test.
Results
One hundred and fifteen CD pts were enrolled (63.5% males; median age 37 years; median disease duration 96 months). Forty-nine per cent of pts had L1, 9% had L2, 42% had L3 according to the Montreal criteria. TH rate at 12 months was 20% and IUS responder rate was 43%. At baseline, no statistical differences in terms of BWT were observed between pts achieving TH vs no TH. Similarly, no differences were observed in pts defined as IUS responders vs non responders. Patients achieving TH at 12 months had a higher ΔBWT than patient without TH (p=0.0004). On ROC curve, ΔBWT improvement of 1.25 mm showed sensitivity and specificity of 73% and 61%, respectively, in predicting pts who achieved TH. Similarly, IUS responder group at 12 months had a higher ΔBWT than patients IUS non-responder group (p<0.0001). On ROC curve, ΔBWT improvement of 1.25 mm showed sensitivity and specificity of 83% and 57%, respectively, in predicting IUS responders. A combination of a ΔBWT (>1 mm) and Limberg’ score improvement (1 point) was associated with a higher risk of TH at 12 months (p=0.0007; OR 5.1; 95% CI, 1.8-12.8). A combination of a ΔBWT (>1 mm) and Limberg’ score improvement (1 point) was associated with a higher risk of IUS responders at 12 months (p<0.0001; OR 15.7; 95% CI, 4.5-52).
Conclusion
An early ΔBWT, and a combination of a ΔBWT and Limberg’ score improvement showed high diagnostic accuracy in predicting pts who achieved TH and IUS responder at 12 months. This information may help clinical decision making in terms of prompt optimization or switching/swapping therapies in CD.