You are here
Decreasing trends in hospitalizations during anti-TNF therapy are associated with time to anti-TNF therapy: Results from two referral centres
Digestive and Liver Disease, Volume 46, Issue 11, November 2014, Pages 985–990
Hospitalization is an important outcome measure and a major driver of costs in patients with inflammatory bowel disease. We analysed medical and surgical hospitalization rates and predictors of hospitalization before and during anti-TNF therapy.
Data from 194 consecutive patients were analysed retrospectively (males, 45.4%, median age at diagnosis, 24.0 years, infliximab/adalimumab: 144/50) in whom anti-TNF therapy was started after January 1, 2008. Total follow-up was 1874 patient-years and 474 patient-years with anti-TNF exposure.
Hospitalization rates hospitalization decreased only in Crohn's disease (odds ratio: 0.59, 95% confidence interval: 0.51–0.70, median 2-years’ anti-TNF exposure) with a same trend for surgical interventions (p = 0.07), but not in ulcerative colitis. Need for hospitalization decreased in Crohn's disease with early (within 3-years from diagnosis,p = 0.016 by McNemar test), but not late anti-TNF exposure. At logistic regression analysis complicated disease behaviour (p = 0.03), concomitant azathioprine (p = 0.02) use, but not anti-TNF type, gender, perianal disease or previous surgeries were associated with the risk of hospitalization during anti-TNF therapy.
Hospitalization rate decreased significantly in patients with Crohn's disease but not ulcerative colitis after the introduction of anti-TNF therapy and was associated with time to therapy. Complicated disease phenotype and concomitant azathioprine use were additional factors defining the risk of hospitalization.
Keywords: Adalimumab, Anti-TNF infliximab, CD, Hospitalization, Risk, Surgery, UC.
Inflammatory bowel diseases (IBD) are lifelong, progressive diseases. Both Crohn's disease (CD) and ulcerative colitis (UC) have a variable disease course; the majority of patients experiences a relapsing disease course, with 20–25% of patients experiencing chronic continuous symptoms and eventually developing complications necessitating hospitalization or surgery.
Relatively limited data are available on hospitalization rates in IBD. In a pre-biologics multicentre European prospective follow-up study, the cumulative risk of all-cause hospitalization was 52.7% 10 years after diagnosis, but with considerable differences among countries. Of note, hospitalization rates declined significantly from the second year after diagnosis  . In UC, hospitalization reflects the failure of medical therapy and disease severity, and is associated with the need for colectomy and mortality. In the Oxford cohort, the need for hospitalization for acute severe colitis was the most important predictor for colectomy  . Similar findings were also reported from North America with stable hospitalization rates in UC and , with colectomy rates of around 10–15% in a referral severe UC cohort from Hungary 10-years after the diagnosis  . In addition, surgical-resection rates in CD vary widely over time among published studies, ranging between 25% and 61% in the first five years  . Until recently, there was little evidence that outcomes for Crohn's disease had changed over the past decades. However, a recent meta-analysis from the IOIBD Epidemiology Task Force reported that the risk of surgery in CD in prebiologic population-based cohorts has been decreasing during the last decade  , at least partly due to changes in patient monitoring and more aggressive therapeutic strategies  .
Therefore, the requirement for hospitalization and surgery may be regarded as markers of IBD severity  . Furthermore medical and surgical hospitalizations are associated with high costs and contribute significantly to the economic burden of IBD, being responsible for about half of all direct medical costs  reducing these events has become an important therapeutic goal  .
Lower hospitalization and surgery rates were reported for both the anti-TNFs infliximab and adalimumab- in randomized clinical trials, , and and in some observational studies. More recently, a systematic review and meta-analysis  reported that infliximab reduced hospitalization risk in both CD and UC. The magnitude of risk reduction was comparable in randomized clinical trials (OR: 0.51) and observational case-control studies (OR: 0.29) without a difference between CD and UC. However, the observational period was limited to one-year of infliximab exposure in the majority of the trials, and most of the observational studies included assessed economical outcomes and costs but did not target the systematic capture of hospitalization events.
Since limited data are available on the long-term hospitalization rates before and after anti-TNF therapy, especially in patients treated with adalimumab and no data are available from Eastern Europe, our aim was to prospectively analyse the long-term hospitalization rates and predictors of hospitalization before and after anti-TNF therapy prospectively in a large consecutive referral IBD cohort treated with either adalimumab or infliximab.
2. Patients and methods
One hundred ninety-four consecutive IBD patients in whom anti-TNF therapy was started after January 1, 2008 were included in this study. Total follow-up was 1874 patient-years. Both in- and out-patient records were collected prospectively and comprehensively reviewed. The analysis was retrospective. The clinical characteristics of the patients are presented in Table 1 . Patients with a primary non-response to anti-TNFs induction therapy were excluded. We included patients receiving at least one maintenance anti-TNF therapy, with a clinical response to anti-TNF induction therapy at week 12 (>70 point decrease in CDAI were included, or >3 point decrease in MAYO score).
|Males||66 (43.4%)||22 (52.3%)|
|Age at presentation (years)*||23 (19–28)||30.5 (22.5–48)|
|Total follow-up (years)*||8 (5–13)||8.5 (5–11.2)|
|Total follow-up (patient-years)||1523||351|
|Anti-TNF type (infliximab/adalimumab)||102 (67.1%)
|Previous anti-TNF use in patients with adalimumab (%)||53 (34%)||–|
|Time to anti-TNF therapy (years)*||5 (2–10)||6 (3–9)|
|Anti-TNF exposure (years)*||2 (1–4)||1 (1–2)|
|Anti-TNF exposure (patient-years)||415||59|
|Familial IBD||11 (7.2%)||3 (7.1%)|
|L4 only||9 (5.9%)|
|Perianal disease||72 (47.4%)||–|
|Steroid use (ever/during antiTNF)||140 (92.1%)/74 (48.7%)||42 (100%)/27 (64.3%)|
|Azathioprine use (ever/during antiTNF)||145 (95.4%)/112 (73.7%)||40 (95.2%)/23 (54.8%)|
|Smoking habits during anti-TNF|
|No||99 (65.2%)||29 (69.1%)|
|Yes||35 (23.0%)||3 (7.1%)|
|Ex||18 (11.8%)||10 (23.8%)|
CD, Crohn's disease; UC, ulcerative colitis; IBD, inflammatory bowel disease; L1, ileal; L2, colon; L3, ileocolonic; L4, upper gastrointestinal; B1, inflammatory; B2, stenosing; B3, penetrating; PSC, primary sclerosing cholangitis.
Results are presented asn(%) or *median (IQR).
Diagnosis was based on the Lennard–Jones Criteria  . The disease phenotype (age at onset, duration, location, and behaviour) was determined according to the Montreal Classification  . Medical records, including data regarding the presence of major extra-intestinal manifestations (EIM), previous frequency of flare-ups (frequent flare-up: >1 clinical relapse/year  ), previous surgical procedures (resections or perianal procedures), the presence of familial IBD, smoking habits, and perianal involvement, were determined by a thorough review of the patients’ medical charts, which had been collected in a uniform format. Previous and concomitant medical therapy (steroid, immunosuppressive, or previous biological therapy) was registered.
Continuous biological therapy (infliximab or adalimumab) became routinely available in 2008 in the outpatient setting for the treatment of CD patients with luminal and additional fistulizing disease with a CDAI >300 despite adequate previous conventional therapy (including appropriate use of steroids and immunosuppressants for at least three months; or failure to respond or intolerance to, either steroid or immunosuppressive therapy). Due to health authority regulations, patient follow-up appointments are mandatory at least every three months (regular or extraordinary visits). These visits include clinical assessment, review of patient diaries, CDAI  and PDAI determination, laboratory assessment (including CRP), chest X-ray every six-months, and in fistulizing patients, objective determination of response, by MRI, EUA or, rectal ultrasound every 12 months. The indication for anti-TNF therapy in UC in Hungary was steroid-refractory acute severe colitis (MAYO ≥9, with an endoscopy subscore ≥2 despite five-day full-dose intravenous steroid therapy), moderate-to-severe disease activity despite three months’ adequate dose oral steroid and immunosuppressive therapy, or steroid dependency (>10 mg prednisolone for more than three months) in patients receiving or intolerant-to azathioprine therapy. Consequently, patient selection and follow-up was standardized and uniform in all specialized centres from this date on. Of note, the Hungarian Insurance Bureau (OEP) monitored all centres for quality of care and regulation compliance in June 2011 and 2013. The definition of current smoking consisted of smoking ≥7 cigarettes per week for at least six months  .
All-cause and IBD-related hospitalizations (including medical/surgical complications) were registered throughout the entire disease course. Clinical, laboratory data, medication, and hospitalizations – including surgeries – were captured prospectively in all patients from the time of the diagnosis. In referral patients, disease phenotype, course, and hospitalizations were captured retrospectively at the time of the referral visit and prospectively thereafter. Major surgical events were defined as major oesophageal, gastric, duodenal, intestinal, and rectal procedures, including laparotomy, open surgery, and gastrostomy and colostomy, while minor surgery was defined as perianal fistula drainage, fistulectomy, and abscess drainage. Biological therapy was monitored exclusively in the aforementioned IBD centres. Hospitalization rates were compared in the two years preceding and during the continuous anti-TNF therapy.
Central coordination and database management was completed at the 1st Department of Internal Medicine, Semmelweis University (by PLL). The study was approved by the Semmelweis University Regional and Institutional Committee of Science and Research Ethics and University of Szeged Institutional Committee of Science and Research Ethics.
2.3. Statistical methods
Variables were tested for normality by Shapiro Wilk'sW-test. Theχ2-test andχ2-test with Yates correction, McNemar test and logistic-regression analysis were used to assess the association between categorical variables. Odds ratios (OR) were calculated. Variables with ap-value <0.2 in univariate analysis were included in the multivariate testing with anti-TNF type and gender added arbitrarily. Results for continuous variables are expressed as median (interquartile range, IQR) unless otherwise stated. Peter L. Lakatos performed all statistical analysis. For statistical analysis, SPSS®200 (SPSS Inc., Chicago, IL) was used. Ap-value of <0.05 was considered significant.
3.1. Patient phenotype, previous disease course and exposure to medications
One hundred ninety-four consecutive IBD patients (152 CD, 42 UC, male/female: 88/106, median age at diagnosis: 24.0, IQR: 19–30 years, disease duration: 8 years, IQR: 8–12.5 years) in whom anti-TNF therapy was started after January 1, 2008 were included in this study. Patient clinical characteristics are summarized in Table 1 . In CD 60% of patients had extensive disease and two-thirds of patients had complicated disease behaviour with perianal involvement in 47.4%. Median disease duration until the start of anti-TNF therapy was five and six years from diagnosis in CD and UC, respectively. Concomitant steroid and AZA use was common in both CD and UC. Of note, 84.2% of CD patients received concomitant AZA, while 40.1% of CD patients had previous major abdominal surgery.
3.2. Hospitalization rates before and during anti-TNF therapy: the importance of early therapy
The IBD-related hospitalization rate in the two years preceding anti-TNF therapy was significantly higher compared to the hospitalization rate during anti-TNF therapy (61.6/100 patient-years versus 43.2/100 anti-TNF exposed patient-years, OR: 0.69, 95% CI 0.60–0.79,p < 0.001 byχ2-test with Yates correction, and for the eventsp = 0.014 by McNemar test) in the total cohort with similar outcomes for all-cause hospitalizations (data not shown). There was no difference in hospitalization rates during anti TNF therapy in the first versus subsequent years (first year: 42.8/100 anti-TNF exposed patient-years, versus after the first year: 38.6/100 anti-TNF exposed patient-years).
The risk for IBD-related hospitalization decreased only in CD (65.6/100 patient-years versus 41.2/100 anti-TNF exposed patient-years, OR: 0.59, 95% CI: 0.51–0.70,p < 0.001 byχ2-test with Yates correction and for the eventsp = 0.02 by McNemar test), but not UC (48.8/100 patient-years versus 54.3/100 anti-TNF patient-years,p = NS) ( Fig. 1 ). There was no difference in hospitalization rates during anti-TNF therapy in the first and subsequent years (first year: 42.1/100 anti-TNF exposed patient-years, versus after the first year: 38.5/100 anti-TNF patient-years) in CD. In addition, there were no significant differences in the hospitalization pattern before and after anti-TNFs (pIFX = 0.001, OR: 0.49 andpADA = 0.05, OR: 0.78, by McNemar test) according to the anti-TNF type in CD.
In CD, hospitalizations due to all surgeries (major or minor: 18.8/100 patient-years in the two years prior to anti-TNF therapy versus 12.5/100 patient-years in the first year during anti-TNFs,p = 0.07), but not major surgeries (7.9/100 patient-years vs. 7.2/100 patient-years during anti-TNFs), were numerically lower after the start of the anti-TNF therapy. Interestingly, 88.9% of patients requiring surgery during the first year of anti-TNF therapy were done in patients without a previous surgery. In contrast, only two patients (4.9%) requiring surgery in the two years prior to anti-TNF therapy required a subsequent surgical intervention in the first year of anti-TNF therapy and six patients (14.6%) required subsequent surgery during all anti-TNF exposed years.
In addition, there was an association between the time to initiation of anti-TNF therapy and the risk of hospitalization in CD. The need for hospitalization decreased in patients with early- (within three years of diagnosis, hospitalization rate: 81.7/100 patient-years versus 48.6/100 anti-TNF exposed patient-years, OR: 0.60, 95% CI 0.48–0.75,p < 0.001, byχ2-test with Yates correction and need for hospitalizationp = 0.017 by McNemar test) but not with late-anti-TNF exposure (hospitalization rate: 53.4/100 patient-years vs. 35.3/100 anti-TNF exposed patient-years, OR: 0.48, 95% CI 0.27–0.85,p = 0.015 byχ2-test with Yates correction for hospitalization rates, butp = 0.38 by McNemar test). There was no difference in the hospitalization rates during anti TNF therapy in the first and subsequent years of therapy (first year: 47.6/100 patient-years versus after the first year: 49.6/100 patient-years). In a sensitivity analysis, results remained unchanged if the upper limit of early anti-TNF use was defined as four years after diagnosis (p < 0.001 byχ2-test with Yates correction andp = 0.07 by McNemar test), but not for later exposure (p = 0.022 byχ2-test with Yates correction,p = 0.19 by McNemar test).
3.3. Predictors of hospitalization
The association between clinical phenotype and risk of hospitalization was also analysed. In CD, the risk of hospitalization in the two years prior to anti-TNF therapy was associated with female gender (p = 0.04, OR: 1.92, 95% CI: 1.03–3.82) and complicated disease phenotype (p = 0.07), while the risk of hospitalization during anti-TNF therapy was associated with complicated disease behaviour (p = 0.01, OR: 2.60, 95% CI: 1.25–5.39), concomitant AZA use (p = 0.01, OR: 0.40; 95% CI: 0.20–0.81) and previous surgeries (p = 0.05, OR 1.92) in univariate analysis ( Table 2 ). In a logistic regression analysis, complicated disease behaviour (p = 0.03) concomitant AZA (p = 0.02) use but not anti-TNF type, gender, perianal disease or previous surgeries were associated with the risk of hospitalization during anti-TNF therapy. None of the assessed parameters were associated with the need for hospitalization in UC.
|Univariate analysis||Multivariate analysis|
|Female gender||p = 0.17||p = 0.43|
|Duration (per year)||p = 0.19||p = 0.15|
|Location||p = 0.037||p = 0.13|
|Complicated disease behaviour||p = 0.01, OR: 2.60,
95% CI: 1.25–5.39
|p = 0.03, OR: 2.80,
95% CI: 1.12–7.03
|Perianal disease||p = 0.31||–|
|Anti-TNF type (infliximab/adalimumab)||p = 0.19||p = 0.54|
|Concommitant azathioprine use||p = 0.01, OR: 0.40, 95% CI: 0.20–0.81||p = 0.02, OR: 0.40, 95% CI: 0.18–0.87|
|Concomittant steroid use||p = 0.30||–|
|Previous surgery||p = 0.05, OR 1.92, 95% CI: 1.00–3.71||p = 0.13|
|Frequent relapses (>1/year)||p = 0.21||–|
|Smoking||p = 0.16||p = 0.63|
The major finding of our study was that long-term IBD-related hospitalization rates and risk of hospitalization after anti-TNF therapy decreased by 40% in CD patients with early clinical response to anti-TNFs. In contrast, hospitalization rates did not decrease in patients with UC in this cohort. In addition, the risk for hospitalization decreased only in patients with early- but not late-anti-TNF exposure in this moderate-to-severe referral IBD cohort.
The benefit of anti-TNF therapy was first demonstrated in randomized clinical trials. The clinical efficacy of infliximab and adalimumab in CD and UC were reported in the ACCENT, ACT, Classic and Charm randomized clinical trials, , , , and . In addition to clinical efficacy, anti-TNF therapy was reported to be associated with a reduced risk of surgery and hospitalizations, , and . In the ACCENT I trial  , Rutgeerts et al. reported that the risk of hospitalizations (p = 0.014) and need for surgeries (p = 0.01) was lower in luminal CD patients with scheduled infliximab therapy strategy patients compared to patients receiving intermittent therapy only after 12-months of anti-TNF therapy. Parallel steroid and immunosuppressive therapy, respectively, were given in approximately 50% and 30% of patients, while average disease duration at the start of anti-TNF therapy was 7.9 years. Disease-related hospitalization rates were 23 and 24 per 100 patients in the scheduled treatment group while this was 38 per 100 patients in the episodic treatment group (p = 0.047 andp = 0.023) with a similar pattern for surgeries (surgery rates: 14/188 episodic and 11/385 for the combined scheduled group,p = 0.01). Similar data were reported by Lichtenstein et al. in patients with fistulising disease  . In the 195 randomized responder patients with a disease duration of 12.4 years, patients on infliximab maintenance had lower hospitalization rates (11 versus 31 per 100 patients or 18.2% versus 7.3%;p < 0.05) and fewer hospitalization days (0.5 versus 2.5 days;p < 0.05), total surgeries and procedures (65 versus 126;p < 0.05), inpatient surgeries and procedures (7 versus 41;p < 0.01), and major surgeries (2 versus 11;p < 0.05) compared to placebo. Similarly, in ACT  , moderate to severe UC patients (Mayo score 6–12 points and endoscopic subscore ≥2) on infliximab had fewer IBD-related hospitalizations and surgeries/procedures (20 versus 40 per 100 patient-years,p = 0.003 and 21 versus 34 per 100 patient-years,p = 0.03) compared to placebo, respectively. Parallel steroid and immunosuppressive therapy was given in 49.6–59.8% and 41.7–54.5% of the patients, respectively, while disease duration at the start of anti-TNF therapy was 5.9–8.4 years. In addition, in a real-life setting, persistence with infliximab  , most probably identifying patients with a continued clinical benefit, was associated with reduced hospitalization rates.
Reduced hospitalization rates were reported in moderate-to-severe anti-TNF-naïve or -exposed CD patients treated with adalimumab in the CHARM and ADHERE trials and . Previous anti-TNF exposure and parallel immunosuppressive use was reported in approximately half of patients, and disease duration of approximately eight years. The 12-month risk of CD-related hospitalization were 9.8% and 8.4% for the adalimumab every-other-week and combined groups compared with 15.5% for placebo (allp < 0.02 versus placebo). The hazard ratios for CD-related hospitalization were 0.50, 0.34, and 0.42 for the adalimumab every-other-week, weekly and combined group. Relatively few data are available beyond one-year. A continued benefit was reported in ADHERE, the open-label extension of the CHARM study, in the subsequent 12-months of adalimuamb therapy. The week-116 cumulative estimated rates for CD-related hospitalization were 23%, 19%, and 25% for the every-other-week, weekly, and placebo treatment groups corresponding a hazard rate of 0.80 (95% CI: 0.69–0.93). Interestingly, the majority of patients requiring a hospitalization in ADHERE were new patients not hospitalized during the first year of adalimumab therapy. Similarly, hospitalization requirements decreased in UC patients treated with adalimumab in the ULTRA (ulcerative colitis long-term remission and maintenance with adalimumab) trial  .
Concordantly, we report a significantly reduced hospitalization rate and risk in CD patients treated with either infliximab or adalimumab beyond one-year, compared to the rates before anti-TNF use in patients presenting with complicated disease. Of note, the indication for anti-TNF use was more stringent in the current cohort, with a higher disease activity (CDAI >300, and Mayo Clinical score ≥9, with an endoscopy subscore ≥2), a high percentage of perianal complications in CD, while the median time-to-anti-TNFs was lower, which may at least partly explain the numerically higher hospitalization rates reported in the present study. In contrast, hospitalization rates did not decrease in UC. Of note, patients enrolled in clinical trials, especially for CD, do not necessarily represent the patient population encountered in clinical practice  .
The real-life setting differs in many regards from clinical trials. One of the landmark observational studies that reported the long-term clinical efficacy of biological therapy in a referral IBD cohort was published by the Leuven group  . However, the majority of clinical observational studies reporting hospitalization rates primarily assessed healthcare utilization; their primary aim was to assess cost-utility rather than rigorously investigating clinical outcomes. A good example is a multicentre retrospective observational study from the UK  , that reported a significant reduction in the mean annualized per-patient rates of elective (0.18–0.11;p = 0.0035) and non-elective (0.46–0.29;p < 0.0001) hospitalizations, and the number of consultations with gastroenterologists, gastrointestinal surgeons, and radiologists (4.0–3.5, 0.7–0.5, and 0.5–0.2, respectively; allp < 0.001) in CD patients in the 24 months post-infliximab compared to the 12-months before anti-TNF therapy. This corresponds to an approximate 30–40% risk reduction, similar to the present study. However, the median disease duration until infliximab was approximately eight years, longer than in the present study. In addition, we did not find a difference according to the type of anti-TNF, and only limited data are available for adalimumab from routine clinical practice. A small single-centre study from Norway  reported a benefit of anti-TNF therapy on hospitalization and surgery rates compared to rates before anti-TNFs, in patients treated with either infliximab or adalimumab.
Similar findings were reported in a propensity-score matched case–control study from Canada  , where the authors reported a significantly lower incidence of unadjusted disease-related major surgery (HR: 0.64; 95% CI: 0.51–0.81), hospitalization (HR: 0.73; 95% CI: 0.63–0.85) risk and lower rates of hospitalized days (rate ratio: 0.69; 95% CI: 0.49–0.97) in 338 CD patients exposed to infliximab in a healthcare provider database. After adjustment, subjects who received infliximab had a significantly lower risk of hospitalization at any time during follow-up (HR: 0.69, 95% CI: 0.59–0.81,p < 0.001). Cumulative hospitalization rates were 32.1%, 44.4%, 56.1%, and 76.2% versus 24.7%, 38.1%, 51.3%, and 56.0% at 1, 2, 3, and 5 years for infliximab non-users versus infliximab users (p < 0.005, for each comparison), respectively. In this study, almost two-thirds of the patients received parallel azathioprine therapy; disease duration at the start of anti-TNF therapy was 3.5 years, yet disease phenotype was not available for comparison.
Another UK study  , published somewhat earlier assessed the medical and surgical hospitalization requirements and length as well as outpatient, imaging, and laboratory costs 12 months pre- and post-infliximab. There were 90 CD-related acute hospital admissions in the 12 months prior to infliximab. In the 12 months after the start of infliximab therapy, there were 98 admissions, but only 22 of these were CD-related acute medical or surgical admissions. The majority of admissions (76 of 98) were for the scheduled administration of infliximab. Other studies also assessed health-care utilization in adherent versus non-adherent patients  , however adherence in the present paper was not systematically assessed.
Finally, a small Canadian study  reported health resource use and costs including hospitalization outcome before and after infliximab in a small cohort of CD (n = 66) patients. Disease duration at the start of the infliximab was more than ten years and rates of parallel immunosuppressive therapy were not given. In the year following initiation of infliximab therapy, there were significant decreases in healthcare use, including in total hospital events and days (from 45 to 25,p = 0.06 and from 495 to 155,p < 0.05) and major surgeries (from 10 to 2,p < 0.05).
However, from population-based studies it is clear that reoperation rates are lower compared to first surgery rates and this does not necessarily reflect the efficacy of the postoperative medical therapy  . Consequently, surgical rates in patients with previous surgery may introduce a significant bias to the analysis. Interestingly, in the present study there was a trend for lower surgical rates during anti-TNF therapy in CD (p = 0.07), however most patients requiring a surgical intervention during anti-TNF therapy were new cases, confirming that reoperation rates were low. In contrast, the rate of major surgeries was unchanged.
Significantly, none of the above studies assessed the association between the time to anti-TNFs and the disease outcomes. One of our most striking findings was the association between the decrease in hospitalization rates in patients with short disease duration before anti-TNFs. Of note, we defined early drug exposure based on the report from Munkholm et al.  , predicting the long-term disease outcome based on the first three years of the disease course. In addition, 65% of the CD patients in the present cohort already had a complicated disease phenotype at the start of anti-TNFs, which by definition could not be interpreted as early disease. Thus, our results confirm a benefit of early anti-TNF therapy even in patients presenting with complicated-disease.
Similar to previous reports  , female gender, complicated disease phenotype, concomitant AZA use and previous surgeries were identified as factors associated with the need for hospitalization in CD. Furthermore, crude hospitalization rates decreased in patients receiving combination AZA treated with both infliximab (from 62.6 to 27.3/100 patient-years,p = 0.007 by McNemar test) and adalimumab (59.4 to 29.2/100 patient-years,p = 0.04 by McNemar test), but the anti-TNF monotherapy group was too small to enable an adequate analysis.
In UC, the need for medical hospitalizations was reported to be associated with the need for colectomy  , however in this cohort none of the parameters assessed were predictive of hospitalization need while the cumulative colectomy rate during or after anti-TNF therapy was relatively low (19.1%, 8/42). Of note, factors other than disease activity may be associated with the need for hospitalization also in UC, including geographic location, gender, reimbursement factors or therapy persistence  .
The authors are aware of possible limitations of the present study. First, the findings from a referral cohort are not necessarily directly generalizable to general gastroenterology practices (e.g., to CD cohorts with less complicated-disease) and non-IBD specialized settings. The surgery and reoperation rates in patients with previous surgery are known to be lower, so the comparison of surgery rates before and during anti-TNFs are suboptimal. The relative small cohort size in UC may have prevented a robust analysis, as well as the requirement that treatment be ceased after 12 months in UC may have influenced the findings. Finally, factors other than disease activity may influence the actual hospitalization rates, including geographical location and ethnicity. However, these factors are similar in the present patient cohort before and during anti-TNF therapy. In contrast, among others, one of the strengths of the present study is that a cohort of well-characterized CD patients representing the nationwide clinical practice in Hungary with standardized indication for anti-TNF use and harmonized follow-up, according to the regulations of the Hungarian Insurance Bureau (OEP) was studied. Moreover, centres were monitored for quality of care and compliance with regulations by the OEP in June 2011 and June 2013.
In conclusion, in the present study we found that hospitalization rate decreased significantly in CD but not UC patients after the introduction of anti-TNF therapy in this referral IBD cohort and the decrease in hospitalization rates was associated to the time to anti-TNF therapy. There was a trend for lower surgical hospitalizations during anti-TNF therapy. Complicated disease phenotype and concomitant AZA use were additional factors defining the risk of hospitalization in CD.
Conflict of interest
Peter L. Lakatos has acted as speaker for AbbVie and MSD. In the past two years he has consulted for AbbVie and MSD Hungary and has received unrestricted research grants from AbbVie and MSD Hungary. Tamas Molnar has acted as speaker for AbbVie and MSD Hungary. In the past two years he has consulted for AbbVie and MSD Hungary.
-  S. Odes, H. Vardi, M. Friger, et al. Cost analysis and cost determinants in a European inflammatory bowel disease inception cohort with 10 years of follow-up evaluation. Gastroenterology. 2006;131:719-728 Crossref
-  L.C. Dinesen, A.J. Walsh, M.N. Protic, et al. The pattern and outcome of acute severe colitis. Journal of Crohn's and Colitis. 2010;4:431-437 Crossref
-  A.N. Ananthakrishnan, M. Issa, D.B. Beaulieu, et al. History of medical hospitalization predicts future need for colectomy in patients with ulcerative colitis. Inflammatory Bowel Diseases. 2009;15:176-181 Crossref
-  C.N. Bernstein, A. Nabalamba. Hospitalization, surgery, and readmission rates of IBD in Canada: a population-based study. American Journal of Gastroenterology. 2006;101:110-118 Crossref
-  T. Molnár, K. Farkas, T. Nyári, et al. Response to first intravenous steroid therapy determines the subsequent risk of colectomy in ulcerative colitis patients. Journal of Gastrointestinal and Liver Diseases. 2011;20:359-363
-  F.L. Wolters, M.G.V.M. Russel, R.W. Stockbrugger. Systematic review: has disease outcome in Crohn's disease changed during the last four decades. Alimentary Pharmacology and Therapeutics. 2004;20:483-496 Crossref
-  C.N. Bernstein, E.V. Loftus Jr., S.C. Ng, et al. Epidemiology and natural history task force of the international organization for the study of inflammatory bowel disease (IOIBD). Hospitalisations and surgery in Crohn's disease. Gut. 2012;61:622-629 Crossref
-  P.L. Lakatos, P.A. Golovics, G. David, et al. Has there been a change in the natural history of Crohn's disease? Surgical rates and medical management in a population-based inception cohort from Western Hungary between 1977–2009. American Journal of Gastroenterology. 2012;107:579-588 Crossref
-  J. Burisch, T. Jess, M. Martinato, et al. The burden of inflammatory bowel disease in Europe. Journal of Crohn's and Colitis. 2013;7:322-337 Crossref
-  G. Van Assche, S. Vermeire, P. Rutgeerts. The potential for disease modification in Crohn's disease. Nature Reviews Gastroenterology and Hepatology. 2010;7:79-85 Crossref
-  P. Rutgeerts, B.G. Feagan, G.R. Lichtenstein, et al. Comparison of scheduled and episodic treatment strategies of infliximab in Crohn's disease. Gastroenterology. 2004;126:402-413 Crossref
-  G.R. Lichtenstein, S. Yan, M. Bala, et al. Infliximab maintenance treatment reduces hospitalizations, surgeries, and procedures in fistulizing Crohn's disease. Gastroenterology. 2005;128:862-869 Crossref
-  W.J. Sandborn, P. Rutgeerts, B.G. Feagan, et al. Colectomy rate comparison after treatment of ulcerative colitis with placebo or infliximab. Gastroenterology. 2009;137:1250-1260 Crossref
-  J. Costa, F. Magro, D. Caldeira, et al. Infliximab reduces hospitalizations and surgery interventions in patients with inflammatory bowel disease: a systematic review and meta-analysis. Inflammatory Bowel Diseases. 2013;19:2098-2110 Crossref
-  J.E. Lennard-Jones. Classification of inflammatory bowel disease. Scandinavian Journal of Gastroenterology. 1989;24(Suppl. 170):2-6 Crossref
-  M.S. Silverberg, J. Satsangi, T. Ahmad, et al. Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: report of a Working Party of the 2005 Montreal World Congress of Gastroenterology. Canadian Journal of Gastroenterology. 2005;19(Suppl. A):5-36
-  E.F. Stange, S.P.L. Travis, S. Vermeire, et al. European evidence based consensus on the diagnosis and management of Crohn's disease: definitions and diagnosis. Gut. 2006;55(Suppl. I):i1-i15 Crossref
-  W.R. Best, J.M. Becktel, J.W. Singleton, et al. Development of a Crohn's disease activity index: National Cooperative Crohn's Disease Study. Gastroenterology. 1976;70:439-444
-  T. Szamosi, J. Banai, L. Lakatos, et al. Early azathioprine/biological therapy is associated with decreased risk for first surgery and delays time to surgery but not reoperation in both smokers and non-smokers with Crohn's disease, while smoking prevents colectomy in UC. European Journal of Gastroenterology and Hepatology. 2010;22:872-879 Crossref
-  S.B. Hanauer, B.G. Feagan, G.R. Lichtenstein, et al. Maintenance infliximab for Crohn's disease: the ACCENT I randomised trial. Lancet. 2002;359:1541-1549 Crossref
-  B.E. Sands, F.H. Anderson, C.N. Bernstein, et al. Infliximab maintenance therapy for fistulizing Crohn's disease. New England Journal of Medicine. 2004;350:876-885 Crossref
-  P. Rutgeerts, W.J. Sandborn, B.G. Feagan, et al. Infliximab for induction and maintenance therapy for ulcerative colitis. New England Journal of Medicine. 2005;353:2462-2476 Crossref
-  W.J. Sandborn, P. Rutgeerts, R. Enns, et al. Adalimumab induction therapy for Crohn disease previously treated with infliximab. Annals of Internal Medicine. 2007;146:829-838 Crossref
-  J.F. Colombel, W.J. Sandborn, P. Rutgeerts, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn's disease: the CHARM trial. Gastroenterology. 2007;132:52-65 Crossref
-  C.T. Carter, H. Leher, P. Smith, et al. Impact of persistence with infliximab on hospitalizations in ulcerative colitis. American Journal of Managed Care. 2011;17:385-392
-  B.G. Feagan, R. Panaccione, W.J. Sandborn, et al. Effects of adalimumab therapy on incidence of hospitalization and surgery in Crohn's disease: results from the CHARM study. Gastroenterology. 2008;135:1493-1499 Crossref
-  R. Panaccione, J.F. Colombel, W.J. Sandborn, et al. Adalimumab sustains clinical remission and overall clinical benefit after 2 years of therapy for Crohn's disease. Alimentary Pharmacology and Therapeutics. 2010;31:1296-1309 Crossref
-  B.G. Feagan, W.J. Sandborn, A. Lazar, et al. Adalimumab therapy is associated with reduced risk of hospitalization in patients with ulcerative colitis. Gastroenterology. 2014;146:110-118
-  C. Ha, T.A. Ullman, C.A. Siegel, et al. Patients enrolled in randomized controlled trials do not represent the inflammatory bowel disease patient population. Clinical Gastroenterology and Hepatology. 2012;10:1002-1007 Crossref
-  F. Schnitzler, H. Fidder, M. Ferrante, et al. Long-term outcome of treatment with infliximab in 614 patients with Crohn's disease: results from a single-centre cohort. Gut. 2009;58:492-500 Crossref
-  J.O. Lindsay, R. Chipperfield, A. Giles, et al. A UK retrospective observational study of clinical outcomes and healthcare resource utilisation of infliximab treatment in Crohn's disease. Alimentary Pharmacology and Therapeutics. 2013;38:52-61 Crossref
-  A. Riis, T.C. Martinsen, H.L. Waldum, et al. Clinical experience with infliximab and adalimumab in a single-center cohort of patients with Crohn's disease. Scandinavian Journal of Gastroenterology. 2012;47:649-657 Crossref
-  J.P. Leombruno, G.C. Nguyen, P. Grootendorst, et al. Hospitalization and surgical rates in patients with Crohn's disease treated with infliximab: a matched analysis. Pharmacoepidemiology and Drug Safety. 2011;20:838-848 Crossref
-  M.B. Sprakes, A.C. Ford, N.C. Suares, et al. Costs of care for Crohn's disease following the introduction of infliximab: a single-centre UK experience. Alimentary Pharmacology and Therapeutics. 2010;32:1357-1363 Crossref
-  C.T. Carter, H.C. Waters, D.B. Smith. Impact of infliximab adherence on Crohn's disease-related healthcare utilization and inpatient costs. Advances in Therapy. 2011;28:671-683 Crossref
-  D.E. Loomes, C. Teshima, P. Jacobs, et al. Health care resource use and costs in Crohn's disease before and after infliximab therapy. Canadian Journal of Gastroenterology. 2011;25:497-502
-  L. Lakatos, L.S. Kiss, G. David, et al. Incidence, disease phenotype at diagnosis, and early disease course in inflammatory bowel diseases in Western Hungary, 2002–2006. Inflammatory Bowel Diseases. 2011;17:2558-2565 Crossref
-  P. Munkholm, E. Langholz, M. Davidsen, et al. Disease activity courses in a regional cohort of Crohn's disease patients. Scandinavian Journal of Gastroenterology. 1995;30:699-770
a 1st Department of Medicine, Semmelweis University, Budapest, Hungary
b 1st Department of Medicine, University of Szeged, Szeged, Hungary
© 2014 Editrice Gastroenterologica Italiana S.r.l., Published by Elsevier B.V.