SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Background : Recurrences after surgery for Crohn's disease are frequent and unpredictable. To date, there is little agreement as to which factors increase a patient risk of early recurrence.

Aim : To assess whether the post-operative behaviour of diseased bowel walls, as determined by ultrasound, may be a useful predictor of relapse.

Methods : A total of 127 Crohn's disease patients were monitored after surgery by means of bowel ultrasound as well as by clinical and laboratory evaluations for a median follow-up of 41.0 months. Bowel wall thickness of diseased loops measured at ultrasound during follow-up was compared with the presurgery values. Multivariable survival analysis was performed to elucidate predictors of early post-operative recurrence. Receiver operating characteristic curves were also constructed taking into account bowel wall thickness for selecting Crohn's disease patients with high risk of clinical/surgical recurrence.

Results : The estimated 5 years survival probability of symptomatic Crohn's disease recurrence were 90% and 33%, respectively for unchanged/worsened bowel wall thickness vs. improved bowel wall thickness at 12 months from surgery. The hazard ratio for unchanged/worsened bowel wall thickness at 12 months was 8.9 (95% CI: 3.4–23.2). Receiver operating characteristic curve identified a bowel wall thickness > 6.0 mm at 12 months from surgery as directly associated with the risk of having a Crohn's disease recurrence (hazard ratio was 6.5, 95% CI: 2.8–15.4).

Conclusions : Systematic ultrasound follow-up of diseased bowel walls after conservative surgery allows the early identification of patients at high risk of clinical/surgical recurrence.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Crohn's disease (CD) is a chronic inflammatory gastrointestinal (GI) disorder with an unpredictable clinical course, characterized by phases of remission and frequent relapses which often lead to surgery. Indeed, up to 70% of adults with CD will eventually require at least one surgical treatment because of complications or refractoriness to medical treatment.1–3 Even the post-operative behaviour of CD is quite difficult to predict as some patients will experience a low morbidity whereas the majority will subsequently suffer recurrences of disease, often requiring further operative intervention. In particular, the rate of post-operative CD recurrence tends to steadily increase with time, reaching up to 50% at 20 years after surgery with no significant difference between conventional resective surgery and new bowel-sparing techniques.4–8

As ileocolonic CD may be better controlled by new pharmacological compounds,9 the main current issue in the management of post-operative CD is to identify strong predictors of negative outcome which allow the early identification of those patients who can theoretically benefit from a more aggressive medical prophylaxis after surgery.

Previous studies on risk indices of postsurgical recurrence in patients with CD have given conflicting results;10 the factors that have been more consistently linked to higher risk of relapse are cigarette smoking,11, 12 ileocolonic disease,6, 13 wide extent of bowel resection14, 15 and presence of perianal disease.15 Apart from colonoscopy, which can detect early luminal signs of recurrence after ileocolonic resection,16, 17 no other features of the bowel wall have provided so far useful information to differentiate between patients who relapse rapidly after surgery and those who do not.10

During recent years, bowel ultrasound (US) has been shown to be an accurate non-invasive method for evaluating bowel wall in patients with CD,18 being able to assess the presence, the location and the extension of transmural inflammatory changes as well as peri-intestinal abnormalities.19–21 To date, however, no large study has evaluated in the long run, the postsurgical behaviour of diseased bowel walls at US and if it may have any prognostic implication on subsequent clinical and surgical relapses.

The aim of the present study was to assess in a large series of CD patients who had undergone conservative surgery, whether the post-operative behaviour of the diseased bowel wall, as systematically evaluated by means of bowel US, is a useful predictor of symptomatic disease recurrence.

Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

L.Sacco University Hospital is a 573-bed tertiary care academic hospital located in Western Milan. It is a national referral Centre for inflammatory bowel disease (IBD), with the largest out-patient IBD clinic in the country and three consultant colorectal surgeons with specific interest in this field who perform all the operations on IBD patients (AMT, GMS, PGD). All records (including demographic and clinical data, laboratory, X-ray, endoscopic and surgical reports) of patients admitted to or seen as out-patients in our Centre, with a diagnosis of IBD, have been computerized since 1 January 1994 and are regularly updated.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

All consecutive patients with complicated or medical-therapy refractory CD who underwent surgery at our IBD centre from March 1994 to May 2002, were potentially eligible to enter this prospective study. At hospital admission their database was updated, if they were already under our care, otherwise a careful clinical history was taken, including all demographic data (age, gender, smoking habits, site and duration of disease, number of clinical recurrences, type and number of previous surgery, total length of previous bowel resections), modalities of current clinical presentation, indication for surgery, location of CD (see later on). Preoperative disease activity [CD activity index (CDAI)22] was also determined. The criteria for admission to the study also required that patients had adequate medical records of clinical, anatomical and therapeutic characteristics of disease, and had had a recent colonoscopy (or barium enema) and barium enteroclysis for disease restaging (within the last 6 months). Patients were excluded from the study if they had only anorectal, gastroduodenal or colonic localization of the disease or failed to obtain a preoperative bowel US. Patients who had to be submitted to surgical procedures other than conservative surgery (see later on) or for perianal disease only were also excluded from the study.

Study protocol

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Diagnosis of CD had been achieved in all patients using standard clinical, radiographic, endoscopic and pathological criteria.23

Preoperative work-up included chest X-ray, electrocardiography, routine blood tests, bowel US, and depending on the patients’ status at admission, endoscopic and radiological GI studies. The detailed indications and surgical techniques have been previously described.8, 24 Briefly, the conservative surgical approaches adopted were strictureplasty, minimal bowel resection or both. In the strictureplasty group, the Heineke-Mickulicz technique was used for short jejunoileal strictures (up to 10 cm), whereas side-to-side isoperistaltic plasty were used to treat long jejunoileal segments or multiple but close short strictures. A widening ileocolic strictureplasty was performed in case of short stenosis arising near the ileocaecal valve or a previous ileocolonostomy, and an ileocolic side-to-side isoperistaltic strictureplasty was done for longer stenoses in the same location. The minimal bowel resections were intestinal resections without macroscopically disease-free margins, with the removal of the intestinal segment with no residual lumen or a fistula arising from a stenotic segment or an abscess located close to a stenosis.

After surgery, patients were randomly assigned to a post-operative treatment with oral mesalazine (mesalamine) (3 g/day) or azathioprine (2 mg/kg/day) as part of another ongoing planned study of 24 months duration.

Clinical and laboratory evaluations (including a complete physical examination, routine biochemical tests and C-reactive protein assay), with determination of CDAI were done at 6-month intervals for the first year after surgery then once a year. Bowel US evaluations were performed before and at 6 and 12 months after surgery then once a year by two highly experienced operators who had an experience exceeding >4000 (FP) and >2000 (SG) sonographic examinations of the bowel, respectively. The sonographer was blind for the results of other diagnostic tests but aware of the diagnosis and previous bowel resection. Patients were also evaluated clinically at any time during the study period if a medical problem arose.

The study protocol was approved by the Ethics Committee of the L.Sacco University Hospital. Patients gave written informed consent before entering the study.

Bowel US

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Sonographic scan of the abdomen was carried out with a real-time ultrasonic apparatus (Hitachi EUB 525 or Aloka Prosound 5500 SV, Tokyo, Japan) initially using a convex 3.5–6 MHz transducer, and then, for a detailed examination of the bowel, a linear 7.5–10.0 MHz probe. Patients were examined after an overnight fast but without any special preparation. As previously reported,19 intestinal wall thickness ≥4 mm (measured from the edge of the outer wall to the inner echogenic mucosa-gas acoustic surface) was considered pathological provided that it could be measured both in longitudinal and transverse sections and it could be reproduced for at least 4 cm in length. The US parameters considered both preoperatively and after surgery were maximum bowel wall thickness (BWT), the length of BWT, and the bowel wall echopattern; the latter was classified into three type: hypoechoic echopattern characterized by the loss of normal wall stratification, stratified echopattern characterized by the persistence of echostratification and mixed echopattern characterized by the coexistence of intestinal tracts with and without echostratification. In the case of multiple disease locations with similar BWT, we considered the BWT of the longest diseased bowel segment.

As far as concerns the post-operative assessment of BWT, two conditions were considered: normalized (BWT: ≤4 mm) or improved BWT (reduction ≥ 40% of preoperative value and more than 2 mm) and unchanged or worsened BWT (reduction < 40% of preoperative value or increased BWT, of at least of 2 mm). A change of at least 2 mm was chosen in view of the good reproducibility obtained with repeated BWT measurement in the same individuals within this range of variability, as reported in a previous preliminary postsurgical experience of ours.25

Outcome measures

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

The primary outcome measure of our study was symptomatic recurrence of CD after surgery. Symptomatic recurrence was defined as the reappearance of symptoms related to CD, variably associated with radiological, endoscopic and laboratory findings, with a CDAI > 200, considered severe enough to require treatment with a systemic steroid at medium–high dose. We also considered the occurrence of surgical relapse which was defined as the presence of medically uncontrolled disease or CD-related complications requiring a new surgical procedures (i.e. strictures, intra-abdominal abscesses or high-flow fistulas).26

Patients who failed to attend the follow-up visit and bowel US within 4 weeks of the scheduled date were considered as protocol violations and excluded from the data analysis.

Statistical analysis

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Distribution of the individual characteristics was evaluated by simple descriptive statistics.

Survival probability of CD recurrences was estimated by the Kaplan–Meier method, stratified according to the behaviour of BWT. Time-to-event differences were tested using the log-rank test. To assess the relative excess risk of relapse, according to BWT behaviour at 6, 12 and 24 months, and to control for confounding factors the multivariate Cox proportional hazards models were fitted to obtain hazard ratios (HR) and the corresponding 95% confidence intervals (CI). The proportional assumption was examined by introducing time-dependent interaction terms into the model. To determine which characteristics (among age, gender, duration of disease, location of disease, smoking habit, history of perinal disease, number of previous surgery, length of bowel resection type of surgical procedures, post-operative treatment, preoperative BWT, echopattern, length of bowel involvement and CDAI at baseline) could influence the results, a backward stepwise selection analyses was performed using recurrence as the dependent variable. Variables excluded by the model were those whose corresponding chi-square test did not reach 95% significance.

Finally, receiver operating characteristic (ROC) curves were applied to find the best sensitivity and specificity cut off values of absolute BWT and its percentage of reduction, when compared with preoperative measurement, for selecting CD patients with high risk of clinical and/or surgical recurrence.

Patients

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Patient recruitment began in March 1994 and was completed in May 2002. A total of 169 were screened for the study. Of these, 26 patients were not enrolled due to violation of inclusion or applicability of exclusion criteria. Furthermore, 12 patients refused to take part in the study. Thus, 131 patients (77% of the eligible population) were enrolled. Four patients, did not attend the 6 month scheduled follow-up in time and were, therefore, not included in the final analysis. Eight patients (six on azathioprine and two on mesalazine) discontinued prophylactic treatment because of adverse events but continued clinical and US follow-up.

The baseline characteristics of the 127 patients who were included in the final study analysis are shown in Table 1.

Table 1.  Characteristics of the study population [n = 127 Crohn's disease (CD) patients]
Males/females83/44
Median age, years (range)32 (15–62)
Duration of CD, years (range)6 (0–23)
Smoking habit at surgery, n (%)
 Never smoker63 (49.6)
 Ever smoker64 (50.4)
Site of CD, n (%)
 Jejunum and Ileum22 (17.3)
 Ileum alone94 (74.0)
 Ileum and colon11 (8.7)
Perianal disease, n (%)
 No93 (73.2)
 Yes34 (26.8)
Previous surgery, n (%)
 Never64 (50.4)
 147 (37.0)
 213 (10.2)
 > 23 (2.4)
Total length of bowel resection (cm)
 < 50 or none97 (76.4%)
 ≥ 5030 (23.6%)
Indication for surgery, n (%)
 Strictures119 (93.7)
 Fistula/abscesses, other5 (3.9)
 Refractoriness to medical therapy3 (2.4)
Type of surgery, n (%)
 Minimal resection and strictureplasty57 (44.9)
 Strictureplasty70 (55.1)
Post-operative therapy, n (%)
 Mesalazine82 (64.6)
 Azathioprine45 (35.4)

Primary outcome parameters

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

About 29.9% (38 of 127) of the patients who had undergone conservative surgery eventually had a symptomatic CD relapse after a mean of 28.5 months. According to Kaplan–Meier function, the estimated survival probability was 66% after 5 years.

Predictors of relapse

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

From the backward stepwise selection analyses age (continuous term), history of perianal disease (yes/no), type of surgery (minimal resection and strictureplasty/strictureplasty) and history of previous surgery (yes/no), reached the significance and were therefore included in the multivariate model (Table 2). Among the candidate relapse predictors at baseline, only a history of perianal disease (HR = 16.9, 95% CI: 7.6–37.7) strongly increased the risk of relapses. This factor explained the apparent excess risk of having a recurrence for smoking habit (age-adjusted HR for ever vs. never smokers = 2.5, 95% CI: 1.2–5.1) and having a total length of bowel resection ≥ 50 cm (age-adjusted HR = 5.9, 95% CI: 3.1–11.4). Gender, duration and site of disease, post-operative prophylactic treatment, as well as preoperative BWT and echopattern, CDAI values at baseline, and length of bowel involvement had no significant influence on the probability of relapse (Tables 2 and 3).

Table 2.  Distribution of 127 consecutive Crohn's disease (CD) patients undergone conservative surgery by recurrence according to age, sex, and selected variables, and corresponding hazard ratios (HR) and 95% confidence intervals (CI)
 Number of subjects with recurrence Univariate (P-value) HRa (95% CI)
  1. aEstimated from proportional hazard regression models, including terms for age, history of perianal disease, number of previous surgery and type of surgery.

  2. bReference category.

Total38  
Age (years) 0.490 
 < 30181b
 30–3980.56 (0.24–1.32)
 ≥ 4060.41 (0.18–0.92)
Sex 0.635 
 Males261b
 Females121.42 (0.67–3.02)
Smoking habit at surgery 0.002 
 Never smoker111b
 Ever smoker271.15 (0.55–2.39)
Duration of CD (years) 0.124 
 < 7171b
 ≥ 7210.84 (0.37–1.89)
Site of CD 0.928 
 Jejunum and Ileum61b
 Ileum alone291.26 (0.50–3.20)
 Ileum and colon32.58 (0.58–11.53)
Perianal disease <0.001 
 No91b
 Yes2916.89 (7.56–37.73)
Previous surgery 0.006 
 No121b
 Yes262.40 (1.15–5.03)
Total length of bowel resection (cm) <0.001 
 < 50171b
 ≥ 50211.25 (0.35–4.47)
Type of surgery 0.114 
 Minimal resection  and strictureplasty131b
 Strictureplasty252.65 (1.30–5.43)
Post-operative therapy 0.851 
 Mesalazine251b
 Azathioprine131.44 (0.69–3.04)
Table 3.  Distribution of 127 consecutive Crohn's disease (CD) patients undergone conservative surgery by recurrence according to preoperative bowel wall thickness (BWT), echopattern, length of bowel involvement at US and CD activity index (CDAI) at surgery and corresponding hazard ratios (HR) and 95% confidence intervals (CI)
 Number of subjects with recurrence Univariate (P-value) HRa (95% CI)
  1. a Estimated from proportional hazard regression models, including terms for age, history of perianal disease, number of previous surgery and type of surgery.

  2. b Reference category.

Preoperative BWT (mm) 0.341 
 ≤7151b
 7.1–890.97 (0.42–2.24)
 > 8140.97 (0.44–2.11)
Echopattern 0.042 
 Mixed or  hypoechoic291b
 Stratified91.60 (0.72–3.58)
Length of bowel involvement (cm) 0.686 
 < 16151b
 16–25120.91 (0.40–2.07)
 ≥ 25111.30 (0.61–3.27)
Preoperative CDAI 0.341 
 ≤25081b
 251–350221.30 (0.55–3.07)
 > 35080.60 (0.20–1.80)

Among 121 patients whose information was available after 1 year from surgery, BWT improved with a reduction ≥ 40%, in 71 patients (58.7%) and remained unchanged or worsened (reduction < 40%) in 50 subjects (41.3%). In particular, it returned to normal in 33 of the 64 patients who had undergone minimal bowel resection combined with strictureplasty (57.9%), and in 38 of 64 who had undergone strictureplasty alone (59.4%).

The 10-year cumulative probabilities of recurrence as evaluated by the Kaplan–Meier time-to-event estimates for the patients with an ‘unchanged or worsened’ BWT at 12 months from surgery when compared to those with ‘improved’ BWT are reported in Figure 1. The survival functions were significantly different in the two strata of BWT behaviour (P < 0.0001). The estimated 5 years survival probability of clinical/surgical recurrence were 90% and 33% respectively for improved BWT and unchanged or worsened BWT at 12 months from surgery.

image

Figure 1. Kaplan–Meier recurrence-free survival curves demonstrating post-operative recurrence of Crohn's disease according to the bowel wall thickness (BWT) of diseased bowel loops at 12 months from surgery.

Download figure to PowerPoint

The multivariate Cox proportional hazard model showed that patients with an unchanged or worsened BWT at 12 months from surgery had a higher risk of recurrence (HR = 8.9, 95% CI: 3.4–23.3) which further increased at 24 months (HR = 16.4, 95% CI: 4.4–61.3) (Table 4). Even the presence of an hypoechoic or mixed echopattern in the context of a thickened bowel wall at 12 months was associated with an increased risk of relapse (HR = 4.16, 95% CI: 1.9–9.2) (Table 4). The combination of the two aforementioned parameters (i.e. unchanged or worsened BWT plus hypoechoic or mixed echopattern at 12 months) did not modified the risk of recurrence found for ‘unchanged and worsened’ BWT only (HR for the combination = 9.0, 95% CI: 3.2–25.2).

Table 4.  Distribution of 127 consecutive Crohn's disease (CD) patients undergone conservative surgery by recurrence according to various measures detected at different times, and corresponding hazard ratios (HR) and 95% confidence intervals (CI)
 6 months after surgery12 months after surgery24 months after surgery
Number of subjects with/ without recurrence HRa (95% CI)Number of subjects with/ without recurrenceb HRa (95% CI)Number of subjects with/ without recurrencec HRa (95% CI)
  1. BWT, bowel wall thickness.

  2. aEstimated from proportional hazard regression models, including terms for age, history of perianal disease, number of previous surgery and type of surgery. Reference category is ‘subjects without recurrence’.

  3. bBased on 121 subjects whose information was available 12 months after surgery.

  4. cBased on 97 subjects whose information was available 24 months after surgery.

Reduction of BWT at Xth month compared with preoperative BWT (%) 3.55 (1.33–9.48) 8.86 (3.37–23.25) 16.43 (4.41–61.31)
 ≥ 405/296/653/57
 < 4033/6026/2420/17
Pattern at xth month 1.97 (0.95–4.12) 4.16 (1.87–9.24) 7.10 (2.84–17.76)
 Stratified22/8517/879/71
 Mixed or hypoechoic16/415/214/3
BWT at xth month (mm) 1.96 (0.92–4.18) 6.52 (2.75–15.44) 5.90 (2.28–15.24)
 < 610/368/703/61
 ≥ 628/5324/1920/13

ROC curves

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

We used ROC curves to assess the percentage in BWT reduction and the absolute BWT value, when compared with the baseline measurement, with the best sensitivity and specificity for predicting clinical or surgical CD recurrence after operation. A BWT reduction < 40% of preoperative value [sensitivity = 0.81 (95% CI: 0.68–0.95); specificity = 0.73 (95% CI: 0.64–0.82)] and an absolute BWT ≥ 6.0 mm [sensitivity = 0.75 (95% CI: 0.60–0.90); specificity = 0.79 (95% CI: 0.70–0.90)] at 12 months from conservative surgery were identified as the two cut-off values strongly associated with the risk of having a CD recurrence in the long run (Figures 2 and 3).

image

Figure 2. Receiver operating characteristic curve showing the percentage reduction cut off of bowel wall thickness (BWT) at 12 months with the best sensitivity and specificity for a diagnosis of Crohn's disease recurrence.

Download figure to PowerPoint

image

Figure 3. Receiver operating characteristic curve showing the absolute cut off value of bowel wall thickness (BWT) at 12 months with the best sensitivity and specificity for a diagnosis of Crohn's disease recurrence.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Most CD patients will undergo at least one surgical procedure during the course of their illness because of complications or refractoriness to medical therapy.1–3 Although surgery is usually effective under these circumstances, symptomatic relapses occur frequently, requiring further medical treatment and, sometimes, further surgical interventions.26, 27

During the past two decades several retrospective and, more recently, prospective studies have dealt with the possible risk factors for CD recurrence after resective or conservative surgery, yielding conflicting results.10 These inconsistent findings are probably attributable to the heterogeneity of patient series, to differences in the study designs (i.e. retrospective vs. prospective studies) as well as to the wide diversities in preoperative and post-operative medical therapy. Factors that seem most consistently to be able to predict a higher risk of relapse after surgery are cigarette smoking, presence of perianal disease, ileocolonic localization and wide extent of bowel resection.11–15 However, most of these factors are rather ‘weak’ predictors, are often difficult to document (i.e. the total extent of bowel resection in patients who have undergone repeated operations), and may vary over time (i.e. cigarette smoking), making case selection for post-operative medical treatment quite difficult. A simple and non-subjective parameter, or a scoring system, which can be applied repeatedly after surgery and might easily predict the risk of relapse in an individual patient would be, therefore, highly desirable.

Only ileocolonoscopy has been successfully used so far as an imaging diagnostic tool to predict relapse after surgery; indeed, Rutgeerts et al.,17 in a series of 89 patients treated with ileal resection, showed that the endoscopic status of the terminal ileum examined at colonoscopy 1 year after surgery, was the most powerful variable in determining outcome. However, ileocolonoscopy is an invasive procedure, which cannot be easily performed in all patients after surgery; in addition, it may be used for this purpose only after ileocolonic resections, but not after sparing or resective procedures involving the ileum or the jejunum only, thanks to its limited capacity of small bowel exploration.

Bowel US has proved to be a reliable imaging tool in CD for disease localization, for detecting luminal and mesentery complications and, to some extent, for assessing disease activity, both at primary diagnosis and in case of postsurgical recurrence.18–21 The BWT and the echopattern of bowel walls are the most typical and constant US features of CD, and a good correlation between the degree of BWT, as measured at US, and the pathological transmural changes of the bowel wall has been documented.28, 29

The present prospective study, on a defined large cohort of CD patients undergoing conservative ileo-jejunal or ileocolonic surgery, demonstrates for the first time that the post-operative course of disease is best predicted by a serial US follow-up of bowel walls during the first year after surgery. Indeed, our findings show that patients with increased or unchanged BWT within 12 months of operation will get early clinical relapse and are more prone to suffer from CD complications, whereas those with BWT improvement or normalization have a greater chance of having an uneventful post-operative clinical course.

The behaviour of bowel walls during the first year after surgery remains a very strong predictive risk factor even after further adjustment for many preoperative covariates such as disease duration and location, CDAI values, and length of bowel involvement as well as several post-operative variables including type of surgical procedure and prophylactic treatment.

In addition, we documented that an hypoechoic or mixed echopattern in the context of a thickened bowel wall at 12 months from surgery was highly predictive of symptomatic CD recurrence; in this regards, there is a plausible explanation for the relationship between echopattern of diseased bowel wall and post-operative relapse. We and others have documented that the decreased echogenicity of diseased bowel wall is due to hyperaemia and neovascularization related to an increased inflammatory response;29, 30 it is therefore conceivable that an hypoechoic/mixed echopattern identify patients with local bowel inflammation, but subclinical disease who, however, are more prone to relapse in the near future. Indeed, in a preliminary study, looking at the vascularity of the diseased bowel wall by intravenous contrast-enhanced Power-Doppler sonography, it was shown the existence of a small group of CD patients with clinically inactive disease but enhanced Doppler signal in intramural vessels after contrast injection who relapsed within the subsequent 6–12 months of follow-up.31

We also found that a consistent proportion of patients had a normalized BWT after conservative surgery; the return to normal of BWT in 27% and 43% of patients at 12 and 24 months from operation, respectively, regardless of the surgical procedure and the post-operative medical therapy confirms and further reinforces previous data of ours, obtained in a preliminary postsurgical study of 6-month duration.25 Since BWT can be regarded as a surrogate index for transmural inflammation,29 our findings suggest the existence of a CD subpopulation whose disease may undergo morphological regression at the site of strictureplasty, as already shown in small series of patients by Poggioli et al., and Alexander-Williams.32, 33 Reasons for that are still unclear but it has been postulated that removal of bowel obstruction by strictureplasty may break a local vicious cycle of bacterial overgrowth, increased intraluminal pressure, with submucosal bacterial spread, ulcers and further scarring leading to a reduction of transmural inflammation (33).

Last but not least, our findings also have important implications in clinical practice as they suggest a possible strategy for an individualized medical prophylaxis of CD patients after conservative surgery. Indeed, all therapeutic efforts (including new pharmacological agents) should be directed towards those patients whose bowel wall behaviour within 1 year from operation is strongly suggestive of an impending recurrence, in order to prevent the evolution of US abnormalities to an overt symptomatic disease and its complications.

In conclusion, the results of the present study provide for the first time a practical and non-invasive follow-up scheme for CD patients undergone conservative surgery. The post-operative clinical course of CD may be well predicted by a serial US follow-up of the diseased bowel wall during the first year after surgery; this allows the early identification of those patients at high risk of clinical or surgical recurrence, thus making case selection for the most appropriate post-operative medical treatment much easier.

Acknowledgement

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References

Authors are indebted to Dr Perminder Phull, Consultant Gastroenterologist at the Aberdeen Royal Infirmary, UK, for his helpful suggestions and criticisms in preparing the manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Patients
  6. Study protocol
  7. Bowel US
  8. Outcome measures
  9. Statistical analysis
  10. Results
  11. Patients
  12. Primary outcome parameters
  13. Predictors of relapse
  14. ROC curves
  15. Discussion
  16. Acknowledgement
  17. References
  • 1
    Farmer RG, Whelan G, Fazio VW. Long-term follow-up of patients with Crohn's disease: relationship between the clinical pattern and prognosis. Gastroenterology 1985; 88: 181825.
  • 2
    Sales DJ, Kirsner JB. Prognosis of inflammatory bowel disease. Arch Intern Med 1983; 143: 2949.
  • 3
    Sachar DB. The problem of postoperative recurrence of Crohn's disease. Med Clin North Am 1990; 74: 1838.
  • 4
    Karesen R, Serch-Hanssen A, Thoresen BO, Hertzberg J. Crohn's disease: long-term results of surgical treatment. Scand J Gastroenterol 1981; 16: 5764.
  • 5
    Lock MR, Farmer RG, Fazio VW, Jagelman DG, Lavery IC, Weakley FL. Recurrence and reoperation for Crohn's disease: the role of disease location in prognosis. N Engl J Med 1981; 304: 15868.
  • 6
    Whealan G, Farmer RG, Fazio VW, Goormastic M. Recurrence after surgery in Crohn's disease: relationship to location of disease (clinical pattern) and surgical indication. Gastroenterology 1985; 88: 182633.
  • 7
    Ozuner G, Fazio VW, Lavery IC, et al. Reoperative rates for Crohn's disease following strictureplasty. Long-term analysis. Dis Colon Rectum 1996; 39: 1199203.
  • 8
    Sampietro GM, Cristaldi M, Parente F, et al. A prospective, longitudinal study of nonconventional strictureplasty in Crohn's disease. J Am Coll Surg 2004; 199: 820.
  • 9
    Rutgeerts P, Van Deventer S, Schreiber S. Review article: The expanding role of biological agents in the treatment of inflammatory bowel disease – focus on selective adhesion molecule inhibition. Aliment Pharmacol Ther 2003; 17: 143550.
  • 10
    Wolff BG. Factors determining recurrence following surgery for Crohn's disease. World J Surg 1998; 22: 3649.
  • 11
    Breuer-Katschinski BD, Hollander N, Goebell H. Effect of smoking on the course of Crohn's disease. Eur J Gastroenterol Hepatol 1996; 8: 2258.
  • 12
    Sutherland LR, Ramcharan S, Briant H, Fick G. Effect of cigarette smoking on recurrence of Crohn's disease. Gastroenterology 1990; 98: 11238.
  • 13
    Williams JG, Wong WD, Rothenberger DA, Goldberg SM. Recurrence of Crohn's disease after resection. Br J Surg 1991; 78: 109.
  • 14
    Hamilton SR, Boitnott JK, Morson BC. Relationships of disease extent and margin lengths to recurrence of Crohn's disease after ileocolonic anastomosis. Gastroenterology 1981; 80: 116672.
  • 15
    Bernell O, Lapidus A, Hellers G. Risk factors for surgery and recurrence in 907 patients with primary ileocaecal Crohn's disease. Br J Surg 2000; 87: 1697701.
  • 16
    Rutgeerts P, Geboes K, Vantrappen G, Kerremans R, Coenegrachts JL, Coremans G. Natural history of recurrent Crohn's disease at the ileocolonic anastomosis after curative surgery. Gut 1984; 25: 66572.
  • 17
    Rutgeerts P, Geboes K, Vantrappen G, Beyls J, Kerremans R, Hiele M. Predictability of the postoperative course of Crohn's disease. Gastroenterology 1990; 99: 95663.
  • 18
    Parente F, Greco S, Molteni M, Anderloni A, Maconi G, Bianchi Porro G. Modern imaging of Crohn's disease using bowel ultrasound. Inflamm Bowel Dis 2004; 10: 45261.
  • 19
    Parente F, Greco S, Molteni M, et al. Bowel ultrasound as primary imaging procedure in patients with suspected inflammatory bowel disorders. Aliment Pharmacol Ther 2003; 18: 100916.
  • 20
    Gasche C, Moser G, Turetschek K, et al. Transabdominal bowel sonography for the detection of intestinal complications in Crohn's disease. Gut 1999; 44: 1127.
  • 21
    Parente F, Maconi G, Bollani S, et al. Bowel ultrasound in assessment of Crohn's disease and detection of related small bowel strictures. A prospective comparative study versus X-ray and intraoperative findings. Gut 2002; 50: 4905.
  • 22
    Best WR, Becktel JM, Singleton JW. Rederived values of the eight coefficients of Crohn's disease activity index (CDAI). Gastroenterology 1979; 77: 8436.
  • 23
    Malchow H, Ewe K, Brandes JW, et al. European cooperative Crohn's disease study (ECCDS): results of drug treatment. Gastroenterology 1984; 86: 24966.
  • 24
    Cristaldi M, Sampietro GM, Danelli PG, et al. Long-term results and multivariate analysis of prognostic factors in 138 consecutive patients operated on for Crohn's disease using ‘bowel-sparing’ techniques. Am J Surg 2000; 179: 26670.
  • 25
    Maconi G, Sampietro GM, Cristaldi M, et al. Preoperative characteristics and postoperative behaviour of bowel wall on risk of recurrence after conservative surgery in Crohn's disease. A preliminary study. Ann Surg 2001; 233: 34552.
  • 26
    Lutenbach E, Berlin JA, Lichtenstein GR. Risk factor for early post-operative recurrence of Crohn's disease. Gastroenterology 1998; 115: 25967.
  • 27
    Dietz DW, Laureti S, Strong SA, et al. Safety and long-term efficacy of strictureplasty in 314 patients with obstructing small bowel Crohn's disease. J Am Coll Surg 2001; 192: 3307.
  • 28
    Kimmey MB, Wang KY, Haggitt RC, et al. Diagnosis of inflammatory bowel disease with ultrasound. An in vitro study. Invest Radiol 1990; 25: 108590.
  • 29
    Maconi G, Carsana L, Fociani P, et al. Small bowel stenosis in Crohn's disease: clinical, biochemical and ultrasonographic evaluation of histological features. Aliment Pharmacol Ther 2003; 18: 74956.
  • 30
    Stavros AT, Rapp CL, Thichman D. Sonography of inflammatory conditions. Ultrasound Q 1995; 13: 126.
  • 31
    Di Sabatino A, Fulle I, Ciccocioppo R, et al. Doppler-enhancement after intravenous levovist injection in Crohn's disease. Inflamm Bowel Dis 2002; 8: 2517.
  • 32
    Poggioli G, Stocchi L, Laureti S, et al. Conservative surgical management of terminal ileitis: side-to-side enterocolic anastomosis. Dis Colon Rectum 1997; 40: 2347.
  • 33
    Alexander-Williams J. The technique of intestinal strictureplasty. Int J Colorectal Surg 1986; 1: 547