Supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and a grant from the Intractable Diseases, and Health and Labour Sciences Research Grants from the Ministry of Health, Labor and Welfare of Japan.
Previous studies have shown various risk factors for the initial and/or the second operation for Crohn's disease (CD). However, limited data are available with regard to the risk factors for a third operation. We aimed to clarify the risk factors for a third operation for CD.
A total of 200 CD patients who underwent a second intestinal surgery at 13 institutions were examined. We performed univariate and multivariate analyses to examine the influence of independent variables on the cumulative rate of needing a third operation.
A total of 95 patients underwent a third operation. The overall 5-year and 10-year cumulative rates for the third operation were 42.2% and 71.0%, respectively. In univariate analysis, the interval between the initial and the second operation (P = 0.0069), postoperative administration of infliximab (P = 0.0030), and the anatomical site of the disease (P = 0.0132) were significant risk factors for the third operation. In multivariate analysis, the interval between the initial and the second operation (P = 0.0287) and postoperative administration of infliximab (P = 0.0297) remained significant risk factors for the third operation. The cumulative 5-year third operation rate was significantly higher in patients with an interval of less than 5 years between the first and second operations than for those with an interval of 5 years or more (47.8% versus 35.2%, P = 0.0232).
An interval of less than 5 years between the first and the second operations is a significant risk factor for a third operation in patients with CD. (Inflamm Bowel Dis 2011;)
Approximately 50%–80% of patients with Crohn's disease (CD) require surgery at some point during their lifetime.1–3 Postoperative recurrence is common in CD, and after the initial operation some patients need a second and/or multiple operations. Reoperation rates for recurrence range from 48%–71% at 20 years after the initial surgery.4 Furthermore, the risk of needing a third operation reaches 40% at 10 years after the second operation.5, 6 Therefore, prevention of recurrence remains one of the major goals in the treatment of CD patients.
In order to prevent recurrence in CD, identification of patients at high risk for future recurrence is important because intensive therapy may be given to such patients to decrease recurrence needing surgical intervention. To identify such high-risk patients, previous studies evaluated various factors that potentially influenced the recurrence rates in CD patients, including age, gender, smoking, steroid use, duration of preoperative history, perforating disease, perianal disease, ileocolic disease, etc.1, 2, 7–13 However, these studies have focused on identifying risk factors for the initial or the second surgery. To date, few data have been generated with regard to the risk factors for the third operation except for one study with a comparatively small number of patients.14 Therefore, in the present study we aimed to evaluate risk factors for a third intestinal operation in a larger number of CD patients. We examined a total of 200 CD patients and showed that a shorter interval between the initial and the second operation was a significant risk factor for needing a third operation. To our knowledge, this is the first study that has shown that the interval between the initial operation and the second operation was a significant risk factor for a third operation. To the best of our knowledge, this is also the largest study of patients who underwent a second operation that has focused on the risk of a third intestinal operation for CD.
PATIENTS AND METHODS
Patients and Criteria for Diagnosis
A total of 200 CD patients who underwent initial and second intestinal surgeries were examined. Their onset of disease was between 1963–2003, and the diagnosis of CD was made according to the criteria provided by the Investigation and Research Committee for Intractable Inflammatory Bowel Disease organized by the Japanese Ministry of Public Welfare as described previously.15 The first and the second operation included intestinal surgery consisting of resection or strictureplasty. Surgeries for perianal disease or other minor surgical procedures without intestinal surgery were excluded from the initial and the second operations included in our study. This study was approved by the local Ethics Committee.
Data Management and Definitions
Case records were collected from 13 institutions which are participating in the Investigation and Research Committee for Intractable Inflammatory Bowel Disease organized by the Japanese Ministry of Public Welfare and scrutinized retrospectively. Data included the patient date of birth, date of onset of symptoms, date of diagnosis, disease localization at diagnosis, type of disease, type of surgery and date of initial/second surgery, and date of final follow-up, which were transferred to a data file (Microsoft Office Excel, Redmond, WA). The indications for surgery included acute abdominal pain, medical intractability, intestinal obstruction, palpable mass/abscess, internal fistulas, colonic dilatation, etc. The disease localization was established at the time of diagnosis and was classified into three groups: 1) small bowel disease (inflammation of the small bowel); 2) ileocolic disease (inflammation involving both the small bowel and the colon); 3) colorectal disease (inflammation confined to the colon or rectum or both). The type of disease was classified into perforating or nonperforating disease, as described previously.11 Perforating disease included patients who underwent their first operation due to perforating disease, whereas nonperforating disease patients were those who underwent the initial operation due to another cause, such as intestinal obstruction, medical intractability, hemorrhage, etc. Perforating disease was classified as perforating, regardless of the concomitant presence of additional nonperforating disease. The primary outcome measure of this study was the rate of patients needing a third intestinal resection or strictureplasty.
The statistical analysis was performed using the JMP software program (SAS Institute, Cary, NC). The cumulative third operation rate was calculated by the Kaplan–Meier method and compared by log-rank test. Univariate and multivariate analyses were performed by Cox proportional hazards regression models in order to examine the influence of independent variables on the cumulative probability of the third operation. Variables with P < 0.1 in univariate analysis were entered into each multivariate analysis. P ≤ 0.05 was considered statistically significant in all analyses. Probability values and confidence intervals were calculated at the 95% level.
Table 1 shows the characteristics of patients. In the 200 CD patients who underwent a first and second intestinal operation, 95 patients underwent a third intestinal surgery after a median of 3.5 years. The frequency of ileocolic disease or administration of infliximab was significantly higher in patients who underwent the third operation than for those who did not. The overall 5-year and 10-year cumulative rates of needing a third operation were 42.2% and 71.0%, respectively (Fig. 1).
Table 1. Patient Characteristics
Patients Who Received a 3rd operation (n = 95)
Patients Who Underwent Only 2nd operations (n = 105)
Total Number of Patients (n = 200)
Duration before surgery: duration between diagnosis and the first surgery.
Reoperation: reoperation after the first surgery.
Age at diagnosis
P = 0.439
P = 0.9417
P = 0.0006
Type of disease
P = 0.3032
P = 0.6132
P = 0.0170
P = 0.6175
Disease duration between onset and the first surgery (yrs)
P = 0.1494
Disease duration between onset and the second operation (yrs)
P = 0.0101
Interval between the first and the second operations (yrs)
P = 0.0897
Risk Factors for Reoperation and Cumulative Rate of Reoperation
The impact of possible risk factors that may have influenced the frequency of the third operation was evaluated by univariate and multivariate analyses (Table 2). In an analysis of duration of disease, we evaluated the following three different types of disease duration with respect to the risk of a third operation: first, the period between disease onset and the first operation; second, the period between disease onset and the second operation; and last, the interval between the first and the second operation. In a univariate analysis, significant risk factors for the third operation were the interval between the first and the second operation, the anatomical site of the disease, and postoperative administration of infliximab. The cumulative risk of the third operation was significantly higher in patients whose interval between the first and second operations was less than the median interval (4.7 years). We next examined whether the same trend could be observed when we divided patients according to the interval of either shorter or longer than 5 years between the surgeries. Patients whose interval between the initial and the second operation was 5 years or less also showed a higher risk of requiring a third operation (hazard ratio = 0.617 (95% confidence interval [CI], 0.401–0.935, P = 0.0226) compared to the patients whose interval was longer than 5 years. With regard to the anatomical site of the disease, patients with ileocolic disease showed significantly higher risk of needing a third operation than patients with either colorectal-only disease or small intestine disease. Other factors such as gender, age at diagnosis, preoperative duration of disease, and type of disease did not show any significant correlation with the third operation rate.
Table 2. Results of Univariate and Multivariate Analyses
Duration before surgery: duration between diagnosis and the first surgery.
Reoperation: reoperation after the first surgery.
Age at diagnosis
P = 0.4747
P = 0.2686
Small bowel, Colorectal
P = 0.0132
P = 0.0859
Type of disease
P = 0.1370
P = 0.3188
P = 0.0030
P = 0.0297
P = 0.6413
Disease interval between the first and the second operation
P = 0.0069
P = 0.0287
Disease duration before the first operation
P = 0.8263
Disease duration before the second operation
P = 0.0802
Next, we performed a multivariate analysis among the three risk factors that showed a significant impact on the rate of requiring a third operation by univariate analysis (Table 2). In multivariate analysis, the interval between the first and the second operation, and the use of infliximab remained significant risk factors.
Cumulative Rate of Patients Requiring a Third Operation
Cumulative 5-year and 10-year rates of the need for a third operation were significantly higher in patients whose interval between the first and the second operation was 4.7 years or less (P = 0.0069) (Fig. 2). Also patients whose interval was 5.0 years or less showed a higher third operation rate than those with interval more than 5 years (5-year third operation rate; 47.8% versus 35.2%, P = 0.0232). Cumulative 5-year and 10-year rates of the need for a third operation were also significantly higher in patients who received infliximab postoperatively than those who did not (P = 0.0015) (Fig. 3). With regard to the disease localization, patients with ileocolic disease showed a significantly higher rate of needing a third operation than those with small bowel or colonic disease (P = 0.0154) (Fig. 4).
The present study showed that the disease interval between the first and the second operation was a significant risk factor for CD patients to need a third operation. A shorter interval was significantly associated with a higher risk of needing a third operation. The 5-year cumulative rate of patients needing the third operation was 47.8% in those with an interval between the first and second surgeries of 5 years or less, while only 35.2% in patients whose interval was more than 5 years (P = 0.0232). These results suggest that intensive postoperative adjuvant therapy may be especially important for CD patients who have an interval of 5 years or less between the initial and second surgeries in order to avoid the need for a third operation.
Previous studies demonstrated various risk factors for intestinal operation in CD, however, most of these factors have been focused on predicting the risk of needing the first or second operation.1, 7–13 The risk factors for needing a third operation were unclear. To our knowledge, there has been only one study by Alves et al14 that evaluated risk factors for a third operation in CD. However, in Alves et al's study, the total number of patients was comparatively small. In their study, there were 28 CD patients who received a second intestinal operation, but since two cases were excluded because of missing data, they only examined a total of 26 CD patients. In the present study, we examined 200 CD patients who underwent a second operation, and to our knowledge, this is the largest number of patients among studies evaluating the risk of needing a third operation for CD. Another difference between Alves et al's study and the present study is the data source. Alves et al's study was based on the data from a single institution. However, a single institution-based study cannot rule out the possibility of patient selection biases. On the other hand, in the present study, we collected data from 13 institutions in Japan and we were able to analyze the risk factors for the third operation based on the multiinstitutional dataset.
Alves et al14 showed that the third intestinal resection rate was significantly lower in patients treated with immunosuppressive drugs (azathioprine and 6-mercaptopurine, or methotrexate) than in untreated patients (17% versus 58%, P < 0.02). However, with regard to the postoperative effect of immunosuppressants in preventing recurrence, previous studies have shown conflicting results.16–22 For example, Hanauer et al and D'Haens et al16, 18 showed that the postoperative recurrence rate was significantly lower in patients receiving immunosuppressants than in those receiving placebo. A meta-analysis also showed that immunosuppressants are more effective than placebo in preventing both clinical and endoscopic postoperative recurrence in CD.20 On the other hand, Ardizzone et al17 reported that there was no difference in the efficacy of immunosuppressants in preventing clinical and surgical relapses after conservative surgery. In the present study, administration of immunosuppressants was not a significant risk factor for needing a third operation. On the other hand, postoperative administration of infliximab was a significant risk factor for the patients needing a third operation in the present study. This is contrary to the results of recent studies, which showed that infliximab is effective for reducing the postoperative recurrence rate.23–27 In a recent randomized controlled study, Regueiro et al25 showed that endoscopic (9.1% versus 84.6%, P = 0.0006) and histologic (27.3% versus 84.6%, P = 0.01) recurrence rates were significantly lower in CD patients who received infliximab after intestinal resective surgery compared to patients who received placebo. One reason for the conflicting results between the present and other studies may be due to a selection bias of the patients who received infliximab. In the present study, patients received infliximab for therapy of recurrent disease. Therefore, there is a possibility that infliximab might have been administered preferably to higher-risk patients for a third operation, while lower-risk patients did not receive these treatments. This patient selection bias may have been responsible for the results indicating infliximab to be a risk factor in the present study.
The second reason may be a shorter follow-up period for patients who received infliximab. This is actually one limitation of the present study, because we were unable to evaluate the effect of infliximab with a long enough follow-up period because infliximab did not become available in Japan until 2002. In the present study, more than half of the patients (110 patients) underwent the third operation in or after 2002. Among these patients, the median follow-up period was only 2.8 years. We believe we need to follow patients for a longer period of time to evaluate the true effect of infliximab in the adjuvant setting.
The present study showed that an interval of less than 5 years between the first and the second operation for CD was a significant risk factor for needing a third operation. Previous studies have also shown that there is a correlation between the duration of the disease and a risk of surgery. A number of studies have shown a higher risk of surgery with a shorter history of disease.28–30 However, these studies examined the relationship between the disease duration before the first operation and this risk of a second operation. None of the previous studies examined the possible role of disease interval on the risk for needing a third operation. We have shown that patients who underwent a second surgery within 5 years of the first operation are at a higher risk of needing a third operation. With regard to the interval between operations, Greenstein et al11 examined 770 patients with CD and reported that third operations occurred faster after second operations than did second operations after the first. Also, in an analysis of CD patients who had undergone multiple operations, Greenstein et al31 showed that as patients undergo repeated surgical procedures, their postoperative recurrences develop faster after each successive operation. In the present study the mean interval between the second and the third operation (4.4 years) was significantly shorter than that between the first and the second operation (5.8 years) (P = 0.019). Our results were in accordance with Greenstein et al's observations.
To our knowledge, the present study was the largest study of CD patients after the second surgery showing the risk of needing a third surgery. These results suggest that patients with a short interval between the first and the second operations need intensive adjuvant therapy postoperatively, such as infliximab. On the other hand, in the present study the disease duration before the first operation was not a significant risk factor for needing a third operation.
Another risk factor for the third operation in the present study was the anatomical site of the disease. Patients who have colorectal involvement will often require a definitive resection with a permanent stoma. In fact, none of the patients with colonic-only disease underwent a third operation in our study. Therefore, we examined patients with small intestine disease and colonic-only disease together against patients with ileocolic disease. By univariate analysis we showed that ileocolic disease was a significantly higher risk factor for needing a third operation, although this did not remain significant by multivariate analysis.
Previously, a number of studies examined the impact of anatomical site of disease on the recurrence, and many studies have demonstrated that the risk of recurrence was highest for ileocolic disease and lowest for colonic-only disease.32–35 However, most of these studies examined the impact of the anatomical site on the first and/or the second surgery, and therefore, data concerning the need for a third operation with regard to the anatomical site involved is limited. In agreement with previous studies of initial and second surgeries, the present study indicated that there is a significantly higher risk of needing a third operation in patients with ileocolic disease. The present study shows ileocolic disease to be a risk factor, not only for the first or the second operation, but also for the third operation for CD.
Another unique factor that might affect the risk of needing surgery in Japanese CD patients is the use of the elemental diet (ED) therapy. In Japan, ED, rather than corticosteroid therapy, is considered to be effective in the primary remission-induction therapy for active intestinal inflammation.36–38 A Japanese randomized controlled trial showed that a “half elemental diet” therapy regimen, in which half of the daily calorie requirement is provided by an elemental diet and the remaining half by a free diet, is effective in reducing the relapse rate compared with patients eating purely a free diet (relapse rate; 34.6% versus 64.0%).36 However, in the present study ED was not a significant factor for needing a third operation.
One of the limitations of the present study is that we could not examine the association between the third operation and several well-established risk factors including smoking, steroid use, and perianal disease, since they were not available in retrospective review.7, 9, 10, 12, 13 Although these factors are known to be associated with the operation rate, it still remains unclear how these factors affect the risk of the third operation. We believe further studies are necessary to clarify this point. Another important issue is the endpoint of the present study. In the present study we included both intestinal resections and strictureplasties as an intestinal operation. Ideally these two procedures need to be analyzed separately. However, some patients receive both intestinal resections and strictureplasties at the same time, and furthermore the number of each procedure differs between each individual. Therefore, in the present study we did not divide patients according to each procedure. However, we believe that each procedure as well as the number of procedures needs to be evaluated separately. Lastly, although we examined multiple factors in association with the risk of third operation, the number of patients was comparatively small. Therefore, in order to clarify these issues we believe that a prospective study with a large number of patients is necessary.
In conclusion, to our knowledge, the present study is the first to show that a shorter interval between the first and the second operations is a significant risk factor for needing a third operation. Patients whose interval between initial and second surgeries is 5 years or less are at a higher risk of recurrence and, therefore, should receive postoperative adjuvant therapy to prevent the need for a third operation. However, to further confirm this we need to prospectively evaluate CD patients with a longer follow-up period. This is particularly important because the use of infliximab, which is generally thought to reduce disease symptoms and recurrence, was a risk factor for the third operation in our study. However, due to the retrospective nature of the present study, this was considered to be due to a bias, because these drugs might have been administered more frequently to higher-risk patients. Nevertheless, further studies are needed to confirm whether this is indeed the case.
The authors thank Dr. Toshinori Ito of Osaka University, Dr. Katsuyoshi Hatakeyama of Niigata University, Dr. Hiroki Ikeuchi of Hyogo Medical College, Dr. Masato Kusunoki of Mie University, Dr. Hisao Fujii of Nara Medical University, Dr. Masahiko Watanabe of Kitasato University, Dr. Shingo Kameoka of Tokyo Women's Medical University, Dr. Yuji Funayama of Tohoku Rosai Hospital, and Dr. Kazuhiko Yoshioka of Kansai Medical University for their cooperation on this study. The authors also thank Ms. Riyo Kakimoto for secretarial support.