Abstract Background Several observations suggest that immunomodulators (IMMs) and anti-TNFα agents can affect long-term clinical course of Crohn’s disease (CD). However, no convincing surgery-sparing direct effect of these medications has been demonstrated in large population-based studies. Failure to optimize these treatments, particularly a limited use in early disease course, may explain the underperformance of these strategies in the clinical practice setting. Early introduction of IMMs/anti-TNFα in CD after surgery, either as prevention of recurrence or as treatment of early endoscopic recurrence, has become a common clinical practice and may represent a surrogate model of early treatment. The aim of the present study was to evaluate the impact of IMMs/anti-TNFα on the long-term post-operative course after ileocaecal resection in CD patients. Methods All adult CD patients who underwent their first ileocaecal resection at three major referral Centers in Rome were retrospectively reviewed. Demographic and clinical characteristics were recorded: gender, age at the time of surgery, CD behaviour (penetrating or non-penetrating), previous treatment with IMMs/anti-TNFα, and time to prescription of IMMs/anti-TNFα after surgery. The primary endpoint was the need for second resection. Statistical analysis. The Kaplan–Meier survival method was used to estimate the cumulative probability of first prescription of IMMs/anti-TNFα after surgery and to estimate the post-operative course free from reoperation. Cox proportional hazards regression model was used to assess the association of clinical variables with time to reoperation. Results surgery: cohort 1 (1980–1998; n.198); cohort 2 (1999–2009; n.218); cohort 3 (2010–2020; n.241). The mean follow-up after surgery was 83 + 41 months. The cumulative probability of receiving IMMs/anti-TNFα after surgery was significantly different between the three cohorts but the probability of a postoperative course free from reoperation was not (figure). In the multivariate analysis, treatment with IMMs/anti-TNFα prior to the first resection (HR 9.15; 95%CI 2.77–30.21) and post-operative therapy with IMMs/anti-TNFα (HR: 0.24; 95%CI 0.07–0.74) were associated to the risk of reoperation. However, these associations had a time-varying effect and become not significant after 5 and 2 years after surgery respectively. Overall, 283 patients received IMMs/anti-TNFα at any time after surgery. Comparing early treatment (within 1, 2 or 3 years after surgery) with late treatment, no difference was observed in the cumulative risk of reoperation within 10 years Conclusion The increasing use of IMMs/anti-TNFα agents in the post-operative setting may have little impact on the long-term reoperation rate.

Neuroendocrine Neoplasms (NENs) are a heterogeneous group of tumours deriving from the diffuse endocrine system. NENs may occur almost everywhere in the body but are most common in the gastrointestinal tract, the pancreas, and the lungs, with gastroenteropancreatic (GEP) tumours representing 70% of all NENs. GEP-NENs have rarely been reported in association with inflammatory bowel diseases (IBDs) but no definitive relationship between these tumours and IBD has been established This was an ECCO COllaborative Network For Exceptionally Rare case reports project (ECCO-CONFER). We included cases of GEP-NENs diagnosed in patients with IBD that met the diagnostic criteria for NEN according to the European Neuroendocrine Tumour Society. Data were retrospectively collected in a standardized case report form and analysed for event association with patient’s and IBD-related factors GEP-NEN was diagnosed in 100 patients with IBD [61% female, 55% Crohn’s disease, median age 48 years (IQR 37–59)]. Overall the most common location was the appendix (39/100) followed by the colon (22/100). Complete IBD-related data was available for 50 individuals with a median follow-up of 30.5 months (IQR 11.2–70) following NEN diagnosis. At the last follow-up data, 47/50 patients were alive. Three deaths occurred, of which 2 were related to NEN. Median duration of IBD at NEN diagnosis was 84 months (IQR 10–151), and in 18% of cases NEN and IBD were diagnosed concomitantly. 20/50 of NENs were at stage I (T1N0M0) and 28/50 graded G1 (ki 67 ≤2 %) at diagnosis. Incidental diagnosis of NEN either during follow-up or during surgery as well as receiving diagnosis of NEN concomitantly with IBD was significantly associated with an earlier NEN stage (p< 0.01 and p<0.02, respectively). Exposure to immunomodulatory and/or biologic therapy was not associated with advanced NEN stage or grade. Interestingly, primary GEP-NEN sites significantly correlated to the segment affected by IBD (62% vs 38% p = 0.02) In the largest case series to date, prognosis of patients with concomitant GEP-NEN and IBD seems favorable. Incidental NEN diagnosis correlates with an earlier NEN stage and IBD-related therapies are independent of NEN stage and grade. The association of GEP-NEN location and the segment affected by IBD may suggest a possible role of inflammation in NEN tumorigenesis

Abstract Background Prevention of postoperative recurrence is a critical goal in Crohn’s disease (CD) management. Currently, postsurgical CD management and treatment are based on endoscopic monitoring performed within the first year after surgery. However, colonoscopy (CS) is an invasive and expensive procedure, unpleasant to patients. A non-invasive and patient friendly approach is required. Methods Consecutive CD patients who underwent ileo-cecal resection from July 2017 to January 2020 were prospectively enrolled in three Italian Centers and performed CS and bowel ultrasound (US) after six months from the surgery, in a blinded fashion. The patients also underwent complete clinical assessment and blood and stool samples were obtained for C-reactive protein (CRP), and fecal calprotectin (FC) measurements. The disease was considered clinically active if the Harvey–Bradshaw Index (HBI) was higher than 4. Uni- and multivariable analyses were used to assess the correlation between non-invasive parameters, including bowel US findings and FC values and endoscopic recurrence, defined by a Rutgeerts’s score (RS) > 2. Sensitivity, specificity, accuracy, PPV and NPV of bowel US parameters alone and in combination with FC in assessing endoscopic recurrence were calculated. Results Seventy patients were enrolled, 45 patients (64%) had an endoscopic recurrence (RS > 2) at 6 months. Thirteen out of 45 (29%) were symptomatic (HBI > 4). Bowel wall thickness (BWT), bowel wall flow (BWF, presence of vascular signals at color Doppler), the presence of mesenteric hypertrophy, the presence of limph-nodes and FC values significantly correlated with the endoscopic recurrence (p < 0.005). Independent predictors for endoscopic recurrence were BWT (for 1-mm increase: OR 2.63; 95% CI 1.136.12; p= 0.024), presence of lymph-nodes (OR 23.24; 95% CI 1.85291.15; p= 0.014) and FC > 50 µg/g (OR 11.86; 95% CI 2.60–54.09; p= 0.001). Sensitivity, specificity, accuracy, PPV and NPV of bowel US and/or FC are showed in Table 1. Table 1: Diagnostic accuracy of Bowel US and/or FC compared to CS in assessing endoscopic activity (CI 95%): per-patient analysis Conclusion Combined use of bowel US and FC is accurate in assessing endoscopic recurrence at 6 months in CD patients and represent a valid alternative to endoscopic assessment after surgery

Abstract Background Ustekinumab (UST) is increasingly used in Italy for the treatment of refractory Crohn’s disease (CD), however very few data concerning real-life experience has been reported. Therefore, the aim of this study was to assess the long-term effectiveness of UST in refractory CD patients treated in a large Italian cohort. Methods A retrospective study was conducted in 5 Italian tertiary centers. All adult CD patients who started UST because of anti-tumor necrosis factor (TNF) failure were included. The co-primary outcomes were steroid-free clinical remission (defined as Harvey Bradshaw Index, HBI ≤4) at weeks 26 and 52. Secondary outcomes were changes in HBI score, changes in C-reactive protein (CRP) values, normalization of CRP (≤0.5 mg/dl) at weeks 8, 26, and 52, and adverse events. Categorical variables were expressed as frequency and percentage. Unpaired t-test was used to compare variables. A p-value <0.05 indicated statistical significance. Continuous variables were expressed as mean and standard deviation (SD), and median with interquartile range (IQR). Results Between Nov 2018 and Feb 2020,140 patients (51.4% male; median age 45.0 years, IQR 36.3-54.0; median disease duration 16.0 years, IQR 8.0-22.0) were included. The majority of patients had ileocolonic disease (L1, 38.6%; L2, 11.4%; L3, 50.0%) and an inflammatory phenotype (B1, 50.7%; B2, 31.0%; B3, 18.3%). All patients had previously been exposed to at least one anti-TNF agent, 27.1% to 2 anti-TNF agents, and 20.0% to vedolizumab . At inclusion 15.7% of patients received corticosteroids and 8.6% immunomodulators. All patients received an intravenous dose of 6 mg/kg, followed by subcutaneous administration of 90 mg every 8 (90%) or 12 weeks (10%) according to clinical judgment. The proportion of patients achieving steroid-free clinical remission was 61.0% and 64.2% at weeks 26 and 52 respectively. A significant decrease in the mean HBI was reported from baseline to week 8 (6.8 ± 3.6 vs 4.5 ± 3.1; p <0.001), week 26 (3.5 ± 2.9; p <0.001), and week 52 (3.1 ± 2.4; p <0.001). The mean CRP values was also significantly decreased from baseline to week 8 (4.6 ± 7.3 vs 2.8 ± 7.1; p <0.001), week 26 (1.7 ± 3.8; p <0.001), and week 52 (1.1 ± 2.2; p<0.001). At baseline 93 of 119 patients had high CRP value: a normal CRP value was observed in 34.9%, 37.8%, and 49.3% of patients at weeks 8, 26, and 52 respectively. Overall, 11 patients (7.9%) discontinued UST within 1 year: primary failure (n=2), secondary failure (n=6), adverse events (n=3: 2 allergic reactions, and 1 arthralgia). Conclusion To our knowledge this is one of the largest Italian cohort followed up to 1 year, and the results confirm that UST is an effective and safe treatment in refractory CD patients.

Abstract Background The appropriate maintenance treatment for patients with acute severe ulcerative colitis (ASUC) responsive to intravenous steroids (IVS) is still a matter of debate. Although major Guidelines consider thiopurine maintenance an option in this setting, the long-term benefit of early immunomodulator (IMMs) initiation is not well established. The aim of our study was to explore the long-term outcome of patients with ASUC responsive to IVS who received different maintenance strategies Methods In a multicenter retrospective study, all patients with ASUC hospitalized between January 2005 and December 2017 in 14 Italian IBD referral centres were reviewed. Thiopurine and biologic-naïve patients experiencing their first acute severe attack and who responded to IVS were included in the study. Maintenance treatment was prescribed by attending physicians according to their clinical judgment. The main outcomes were recurrent flares requiring escalation of therapy, new hospitalization, and long-term colectomy rate. The Kaplan-Meier survival method was used to estimate the cumulative probability of a course without the main outcomes. Differences between curves were tested using the log-rank test. A propensity score matching analysis was performed to establish comparable groups of patients who received different maintenance treatment Results Overall 372 patients were reviewed. Of these, 141 met the inclusion criteria (males 61.7%, median age 34.5 (IQR 23–50). After response to IVS, 82 patients (58.1%) received maintenance treatment with aminosalicylates, 42 (29.8%) received IMMs and 17 (12.1%) were maintained with scheduled infliximab (IFX) + thiopurines. After a median follow-up of 48 (IQR 25–90) months, 94 patients (68.8%) experienced a flare requiring escalation of therapy, 51 (36.1%) required new hospitalization and 18 (12.8%) underwent colectomy. After 12, 36 and 60 months after the acute attack, the cumulative probability of a course without escalation of therapy was 59.6%, 33.3% and 23.1%; the cumulative probability of a hospitalization-free course was 83.9%, 67.4% and 59.5%; the cumulative probability of a colectomy-free course was 96.3%, 90.2%, and 88.9%. No differences were observed between patients receiving aminosalicylates, IMMs or IFX as maintenance treatment (log-rank test: p= 0.39; p = 0.41; p = 0.11 respectively). After a propensity score matching analysis, no significant difference in main outcomes was observed between patients maintained with aminosalicylates or IMMs/IFX Conclusion IMM-naïve ASUC patients responsive IVS remain at risk of relapse requiring escalation of therapy. Early IMMs introduction after the acute attack did not reduce the risk of escalation of therapy, hospitalization or colectomy

Prevention of postoperative recurrence is a critical goal in Crohn’s disease (CD) management. Currently, postsurgical CD management and treatment are based on endoscopic monitoring performed within the first year after surgery. However, colonoscopy (CS) is an invasive and expensive procedure, unpleasant to patients. A non-invasive and patient friendly approach is required. Consecutive CD patients who underwent ileo-cecal resection from July 2017 to January 2020 were prospectively enrolled in three Italian Centers and performed CS and bowel ultrasound (US) after six months from the surgery, in a blinded fashion. The patients also underwent complete clinical assessment and blood and stool samples were obtained for C-reactive protein (CRP), and fecal calprotectin (FC) measurements. The disease was considered clinically active if the Harvey–Bradshaw Index (HBI) was higher than 4. Uni- and multivariable analyses were used to assess the correlation between non-invasive parameters, including bowel US findings and FC values and endoscopic recurrence, defined by a Rutgeerts’s score (RS) > 2. Sensitivity, specificity, accuracy, PPV and NPV of bowel US parameters alone and in combination with FC in assessing endoscopic recurrence were calculated. Seventy patients were enrolled, 45 patients (64%) had an endoscopic recurrence (RS > 2) at 6 months. Thirteen out of 45 (29%) were symptomatic (HBI > 4). Bowel wall thickness (BWT), bowel wall flow (BWF, presence of vascular signals at color Doppler), the presence of mesenteric hypertrophy, the presence of limph-nodes and FC values significantly correlated with the endoscopic recurrence (p < 0.005). Independent predictors for endoscopic recurrence were BWT (for 1-mm increase: OR 2.63; 95% CI 1.136.12; p= 0.024), presence of lymph-nodes (OR 23.24; 95% CI 1.85291.15; p= 0.014) and FC > 50 µg/g (OR 11.86; 95% CI 2.60–54.09; p= 0.001). Sensitivity, specificity, accuracy, PPV and NPV of bowel US and/or FC are showed in Table 1. Table 1: Diagnostic accuracy of Bowel US and/or FC compared to CS in assessing endoscopic activity (CI 95%): per-patient analysis Combined use of bowel US and FC is accurate in assessing endoscopic recurrence at 6 months in CD patients and represent a valid alternative to endoscopic assessment after surgery

Abstract Background Inflammatory bowel disease (IBD) patient’s knowledge is a key factor for shared treatment decision making. Some studies have already investigated IBD knowledge but never focused on patient’s awareness about the major disease outcomes nor clinical course nor therapeutic aspects. The aim of the present study is to assess disease knowledge in IBD patients using a novel 11 items questionnaire. Methods An 11 items questionnaire was developed. The questionnaire explored different aspects of IBD course and prognosis: pattern of symptoms, disease complications, need for surgery, risk and prevention of colon-rectal cancer (CRC), extra-intestinal manifestations (EIMs), predictors of poor prognosis, benefit/risk of immunosuppressive and biological therapy and IBD in pregnancy. The questionnaire was handed out to consecutive Crohn’s disease (CD) and ulcerative colitis (UC) patients in regular follow-up at a referral centre. Clinical and socio-demographic characteristics were collected in a dedicate database. Descriptive analysis of quantitative variables was presented as mean or median. Fisher test and Student’s t-test were used for categorical and continuous variables. A p < 0.05 was considered statistically significant. Results Three hundred and four patients were enrolled and 290 completed the questionnaire (95%). Of these 53% UC and 47% CD. The mean age was 51 years (range 18–83); 56% were male. Most of the patients were in clinical remission (85%). The mean number of appropriate answer was 4 upon 11. Overall only 26% answered successfully more than half of the items. Looking at the single items, two third of patients were aware of current strategies of CRC prevention, chance of EIMs and woman capability to become pregnant. The questions about immunosuppression and biologic therapy were more correctly answered by those patients who had experienced them (p = 0.000001), whereas those about the risk of surgery were not influenced by the patient’s previous clinical history (UC p = 1; CD p = 0.067). Only about one third of women were aware of the most relevant risk factor of worst pregnancy outcome. Age <40 years old (p = 0.00003), high level of education (high school and degree, p = 0.0015) and current or previous biologic therapy (p = 0.0007) were associated with a higher level of IBD knowledge. Conclusion These results suggest that the overall disease knowledge in IBD patients is low and should be improved. Young age, education and biologic exposure are associated with an adequate awareness of disease course. This questionnaire should be validated and correlated to other patient-related measures such as health engagement and health literacy in prospective studies.

OBJECTIVE:The recurrence of Crohn's Disease after ileo-colonic resection is a crucial issue. Severe endoscopic lesions increase the risk of developing early symptoms. Prevention and treatment of post-operative Endoscopic Recurrence (ER) have been studied with conflicting results. We compare effi cacy of azathioprine (AZA) vs. high-dose 5-aminosalicylic acid (5-ASA) in preventing clinical recurrence and treating severe post-operative ER.PATIENTS AND METHODS:We performed a 1-year multicenter randomized double-blind double-dummy trial. Primary end-points were endoscopic improvement and therapeutic failure (clinical recurrence or drug discontinuation due to lack of efficacy or adverse events) 12 months after randomization. We also performed a post-trial analysis on symptomatic and endoscopic outcomes 10 years after the beginning of the trial, with a median follow-up of 60 months.RESULTS:Therapeutic failure occurred in 8 patients (17.4%) within 12 months from randomization, with no significant difference between patients treated with 5-ASA (20.8%, 5 patients) and those with AZA (13.6%, 3 patients). Therapeutic failure was due to clinical recurrence in the 5-ASA group and to adverse events in the AZA group. Endoscopic improvement at 12 months was observed in 8 patients, 2 (11.8%) in the 5-ASA group and 6 (30%) in the AZA group. No serious adverse event was recorded. At the post-trial analysis (median follow-up 60 months), 47.8% (22/46) of patients experienced clinical recurrence: 54.2% (13/24) in the 5-ASA group and 40.9% (9/22) in the AZA group, p=0.546. Patients treated with AZA had lower risk of drug escalation. Clinical recurrence was associated with smoking (p=0.031) and previous surgery (p=0.003).CONCLUSIONS:Our trial indicates that there was no difference in terms of treatment failure between 5-ASA and AZA in patients with severe ER. The main limit of AZA is its less favorable safety profile.

Background-Aims: The SOLE study was conducted on a large cohort of Italian patients with moderate-severe Crohn's disease (CD) to assess epidemiological and disease characteristics and their correlation with disease-related worries, treatment satisfaction and adherence, workability. Methods: The following tools were used over 12 months to assess: disease-related worries: Rating Form of Inflammatory Bowel Disease Patient Concerns, impact on workability: Work Productivity and Activity Impairment-CD, satisfaction: Treatment Satisfaction Questionnaire for Medication, adherence: Medication Adherence Rating Scale. Results were correlated with demographic and clinical variables with linear regression models. Results: 552 patients with active CD (51% men) were recruited. Higher worries were having an ostomy bag and undergoing surgery. Variables associated with a higher RFIPC score included female sex, higher disease activity, lower treatment adherence (p < 0.001), previous surgical treatments (p = 0.003). 60% of patients claimed difficulties with activities of daily living. Lower VAS scores were reported by patients with disease duration >6years; treatment satisfaction/adherence was higher with anti-TNF-alpha treatment. Decreased hospitalizations during follow-up and improved workability/daily activities occurred with adalimumab, infliximab, azathioprine (p < 0.001). Conclusion: Worries included having an ostomy bag, undergoing surgery, developing cancer: conditions significantly associated with worsened disease activity and low treatment adherence. Higher treatment adherence scores/greater workability improvements were observed in patients treated with anti-TNF-alpha agents. (C) 2019 The Authors. Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.l.

In a prospective, multi-centre, nested case–control study at 6 years (years), we aimed to characterise incident cases of cancer in inflammatory bowel disease (IBD). Secondary end point was to evaluate risk factors for cancerc in IBD. From 31 December 2011 to 31 December 2017, all incident cases of cancer in IBD patients referring to 16 IG-IBD Units (≥2 visits/year) were recorded. Each IBD patient with incident cancer was matched with 2 IBD patients with no cancer for: IBD type (Crohn’s disease, CD; ulcerative colitis, UC), gender, age (±5 years). Data expressed as median (range). Wilcoxon, χ2, Fisher exact test, multi-variate logistic regression analysis (OR [95% CI]). Incident cancer occurred in 403 IBD patients: 204 CD (CD-K),199 UC (UC-K). Overall, 1209 IBD patients were considered (403 IBD-K; 806 IBD-C). In IBD, cancer (n = 403) involved (n = [%]): digestive system (129 [32%]), skin (60 [14.9%]: 27 NMSC, 31 melanoma, 2 others),urinary tract (39 [9.7%]), lung (28 [6.9%]), breast (22 [5.5%]), genital tract (26 [6.5%]), thyroid (8 [1.98%]), lymphoma (11 [2.72%] all in CD), small bowel cancers (16 [3.9%];15 CD [7.3%], 1 UC ileal pouch [0.5%]), others (64[15.9%]). Cancer frequency was comparable between CD and UC considering (n = [%]):digestive system (61[30%] vs. 6 [34%]);skin (33[16%] vs. 27[13.5%]);lung (14[6.8%] vs. 14 [7.0%]);breast (22[10.7%] vs. 24[12.1%]);genital tract (15[7.3%] vs. 11 [5.5%];p>0.05). Colorectal and urinary tract cancers were more frequent in UC vs. CD (58[29%] vs. 35[17%],p < 0.005;26[13%] vs. 13[6.3%], p = 0.039). Extracolonic cancers were more frequent in CD vs. UC (35/204 [17%] vs. 58/199 [29%]; p < 0.005). Risk factors considered: age (<40 vs. ≥40 years), IBD duration (< 10 vs. ≥10 years), smoking (Yes/No), ISS and/or anti-TNFα (Y/N),IBD-related surgery, UC extent, CD pattern, perianal CD risk factors for any cancer identified in UC: UC-related surgery (4.63 [2.62–8.42]), extensive vs. distal UC (1.73 [1.10–2.75]). The other risk factors were not significant (OR 1.30 [0.74–2.39]; 0.92 [0.63–1.35]; 0.92 [0.55–1.52]; 0.84 [0.51–1.38]; 1.54 [0.95–2.51], respectively). In CD, perforating pattern was the only significant risk factor (OR 2.33 [1.33–4.11]) (other risk factors: OR 0.93 [0.59–1.48]; 0.98 [0.67–1.43]; 0.74 [0.51–1.07]; 1.31 [0.90–1.92]; 0.97 [0.62–1.51]; 1.25 [0.79–2.01]; 1.02 [0.65–1.60]).In CD, the frequency of B3 pattern was higher in CD-K vs. CD-C (26% [54/204] vs. 15% [63/408]; p = 0.0033).The frequency of extensive UC was higher in UC-K vs. UC-C (51% [101/199] vs. 38% [152/398]; p = 0.0045). In a prospective, multi-centre, nested-case–control study at 6 years, penetrating CD, extensive UC and UC-related surgery were significant risk factors for any incident cancer. Clinical characteristics of severity of IBD may increase the overall cancer risk. Lymphoma and SBC were associated with CD.

More than 70% of patients with Crohn's disease (CD) require surgery at least once during the course of their disease. Unfortunately endoscopic recurrence (ER) is up to 100% at 5 years with a risk of six-month severe ER (≥ i2) around 50% as showed in a previous Italian study. As well know symptomatic recurrence is strongly related to the severity of ER and ECCO guideline recommended prophylactic treatment after ileocolonic resection despite effectiveness of immunosuppressants remain debated. We performed a multi-centre randomised double-blind double-dummy trial to assess the role of azathioprine (AZA) vs. high dose of mesalamine (5-ASA) as treatment of early severe POR (Rutgeerts’ score ≥ i2) and as prophylaxis for clinical relapse (eligible patients and treatment allocation are showed in Table 1). Table 1 Primary outcomes: endoscopic improvement and clinical relapse after 12 months from randomisation. Post-trial analysis: data on clinical and endoscopic outcomes up to 10 years from T0. According to inclusion/exclusion criteria 46 patients were randomised (characteristics of screened patients are showed in Table 1): 65% males, overall median age at diagnosis and at surgery of 29.5 and 36.5 years, respectively. At the final analysis 17% of patients experienced a clinical relapse within 12 months from randomisation without differences between AZA and 5-ASA groups. Considering POR after 12 months of treatment no significant improvement were observed from T0 in both groups (p = ns). At the post-trial analysis, 53% of patients experienced a clinical relapse without differences between those previously treated with AZA or 5-ASA (p = ns). Smoking and previous surgery at T0 were risk factors for clinical relapse (p = 0.031 and 0.003). No significant AE were recorded. This multi-centre RCT does not show efficacy of AZA or 5-ASA in the treatment of severe POR or as prophylaxis for clinical relapse. In the post-trial analysis, in those with POR at 6-month from surgery, risk factors for severe CD (smoking and multiple surgery) could help to identify patients with worse prognosis to start biological therapy.

Acute severe ulcerative colitis (ASUC) is a potentially life-threatening event affecting up to 25% of patients during disease course. Intensive intravenous glucocorticoid treatment (IIVT) and early colectomy have reduced mortality to less than 2% in the last four decades. Rescue therapies -Infliximab (IFX) or Cyclosporin (CyA)- may reduce early colectomy in IIVT refractory patients but their impact in the long-term is unclear. Aim of the present study was to evaluate the long-term colectomy rate in patients escaping early colectomy after a severe attack From 2005 to 2016 all patients with ASUC meeting Truelove and Witts criteria modified by Chapman et al. referring to 14 Italian IBD referral centres were retrospectively reviewed. All patients received IIVT. IFX or CyA were used as rescue therapies. Primary outcome was long-term colectomy rate in patients escaping early colectomy (within 3 months). Secondary outcomes were overall need of escalation therapy (defined as need of anti-TNF agents or immunomodulators or steroids) or hospitalisation. Kaplan–Meier survival method was used to estimate the cumulative probability of a colectomy-free course and log-rank test to compare colectomy-free survival distributions in different subgroups. A stepwise regression model was used to look for predictive factors of long-term colectomy In total, 361 patients were enrolled. Of them, 15 (4.2%) underwent early colectomy and 346 avoided colectomy: due to of IIVT response (n = 223,64.5%) or rescue therapy response with IFX (n = 103, 29.7%) or CyA (n = 20, 5.8%). Clinical characteristics of patients. During a median follow-up of 43 months (range 1–156),67 patients (19.4%) required colectomy. The cumulative probability of a colectomy-free course was 92.7, 87, 81.9 and 79.7% after 12, 24, 36 and 60 months, respectively. Colectomy risk was similar in IIVT responders and in rescue therapy responders. During follow-up, 135 (39%) and 109 (31.5%) patients required at least one escalation of therapy and hospitalisation, respectively. At multi-variate analysis none of the covariates considered (age, gender, first or recurrent attack, disease extension, C-Reactive Protein levels, endoscopic severity, steroid/rescue therapy responsiveness, maintenance treatment) was associated to long-term colectomy risk The long-term colectomy risk after an acute severe attack is still relevant and do not seem to be influenced by the severity of the attack, resulting similar both in IIVT responders and in IIVT refractory patients responding to rescue therapies.