Does Kono-S Anastomosis Reduce Recurrence in Crohn’s Disease Compared with Conventional Ileocolonic Anastomosis? A Nationwide Propensity Score-matched Study from GETAID Chirurgie Group [KoCoRICCO Study]

Abstract

Background and Aims

Postoperative recurrence is a major concern in Crohn’s disease. The Kono-S anastomosis has been described to reduce the rate of recurrence. However, the level of evidence for its effectiveness remains low. The KoCoRICCO study aimed to compare outcomes between Kono-S anastomosis and conventional anastomosis in two nationwide, prospective cohorts.

Methods

Adult patients with Crohn’s disease, who underwent ileocolonic resection with Kono-S anastomosis, were prospectively included in seven referral centres between 2020 and 2022. Patients with conventional side-to-side anastomosis were enrolled from a previously published cohort. A propensity score analysis was performed to compare recurrence at first endoscopy in a matched 1:2 ratio population.

Results

A total of 433 patients with ileocolonic anastomosis were enrolled, of whom 155 had a Kono-S anastomosis. Before matching, both groups were unbalanced for preoperative, intraoperative, and postoperative characteristics. After matching patients with available endoscopic follow-up, endoscopic recurrence ≥i2 was found in 47.5% of the Kono-S group and 44.3% of the conventional side-to-side group [p = 0.6745].

Conclusions

The KoCoRICCO study suggests that Kono-S anastomosis does not reduce the risk of endoscopic recurrence in Crohn’s disease compared with conventional side-to-side anastomosis. Further research with a longer follow-up is necessary to determine whether there is a potential benefit on surgical recurrence.

1. Introduction

Crohn’s disease [CD] is a chronic condition that causes transmural inflammation in the bowel and can affect individuals of all ages and all parts of the digestive tract.¹ The most frequent localisations are the terminal ileum and the ileocolic region.^2,3 Whereas there is no curative treatment available for CD, immunosuppressors and biologic therapies are commonly used for induction and maintenance therapies. However, surgical treatment, specifically ileocolonic resection, is required when biotherapies fail, with 50% of CD patients operated on during their lifetime.

Despite the benefits of surgery, it exposes the patient to a risk of 57.6% of endoscopic recurrence at 12 months. Smoking, prior intestinal surgery, absence of prophylactic treatment, penetrating disease, perianal location, granulomas in resection specimen, and myenteric plexitis are considered as risk factors for recurrence after ileocolonic resection.⁴

Recent studies have highlighted faecal stasis as a trigger of disease recurrence, leading to proximal localisation of recurrence upstream of the anastomosis. Therefore, performing a wide ileocolic anastomosis is critical and currently recommended by ECCO Guidelines.⁵ However, a meta-analysis of 11 trials comparing different types of anastomosis showed that a stapled side-to-side anastomosis is associated with a lower risk of postoperative complications, without changing the rate of endoscopic recurrence.⁶

Based on anatomopathological observations, Kono et al. have described a new anastomosis technique that moves the anastomosis to the antimesenteric side of the ileum, strengthens it with a central supporting column, and ensures a wide lumen.^7,8 This technique has shown promising results, with a decrease in surgical but not endoscopic recurrence rates. A recent randomised monocentre study by Luglio et al. demonstrated a significant decrease in endoscopic and clinical recurrence rates at 6 and 18 months when using this new anastomosis technique.⁹ Following this publication, a meta-analysis of nine studies with variable statistical power showed similar results concerning morbidity and endoscopic recurrence.¹⁰

However, the sample sizes of these studies are low, and some are retrospective or not controlled. Therefore, the level of evidence remains insufficient to establish practice recommendations. The present KoCoRICCO study aimed to compare the outcomes of the Kono-S anastomosis technique [KA] with the conventional anastomosis [CA] in two large nationwide prospective cohorts.

2. Materials and Methods

2.1. Study population

This prospective, multicentre study included adult patients with Crohn’s disease [CD] requiring ileocolonic resection with either Kono-S anastomosis [KA group] or conventional side-to-side anastomosis [CA group]. Patients in the KA group were prospectively and consecutively included in seven French referral centres between February 2020 and October 2022; patients in the CA group were enrolled from a previously published nationwide cohort [RICCO study] operated on between September 2013 and September 2015.¹¹ Six of the seven referral centres participated in the RICCO study. Exclusion criteria were emergency surgery, anastomosis with defunctioning stoma, or lack of postoperative clinical, biologic, or endoscopic surveillance.

2.2. Data collection

Preoperative work-up, indications for surgery [disease behaviour/primary or iterative resection], and perioperative management were discussed during each institutional multidisciplinary inflammatory bowel disease [IBD] team meetings, based on the current European Crohn’s and Colitis Organisation [ECCO] guidelines. Risk factors for postoperative CD recurrence were collected, and CD medical prophylaxis was decided on a case-by-case basis after discussion during the multidisciplinary IBD team meeting, also based on the ECCO guidelines.

2.3. Surgery

All patients underwent ileocolonic resection, performed by open or laparoscopic approach. In the CA group, stapled mechanical or hand-sewn anastomosis was performed at the discretion of the surgical team. All patients followed a predefined management pathway involving preoperative nutritional assessment and perioperative enhanced recovery programme.

2.4. Endpoints

The primary endpoint was endoscopic recurrence assessed on an ileocolonoscopy performed according to the ECCO guidelines, using the recurrence score developed by Rutgeerts et al., defined by a ≥i2 score.¹² At the time of analysis, the follow-up of the KA group was too short to compare the two groups in terms of clinical and surgical recurrence.

The secondary endpoints were surgical and postoperative outcomes, including the rate of conversion to open surgery, intraoperative complications, duration of operation, postoperative morbidity, rate of anastomotic leakage, duration of hospital stay, reoperation, and readmission. Morbidity was graded according to the Dindo–Clavien classification.¹³

2.5. Population of analysis

The population of analysis was composed of all patients with endoscopic follow-up.

2.6. Statistical analysis

Continuous variables were described as median [interquartile range: IQR], and categorical variables as frequency with percentage. Medians and proportions were compared using Wilcoxon and Mann–Whitney, and χ2 tests [or Fisher’s exact test, if appropriate], respectively. Unconditional logistic regression modelling was used to estimate the odds ratios [ORs] and 95% confidence intervals [CIs] for factors associated with endoscopic recurrence, particularly for the type of anastomosis. A multivariate model, by entering all variables with a p-value <0.1 in univariate analysis and by forcing the type of anastomosis into the model, was produced. A backward selection until obtaining only p-values <0.1 was carried out to get the final multivariate model.

Then, a sensitivity analysis was done with a propensity score approach to deal with potential heterogeneity in characteristics between patients who had KA and CA. Two methods were used to address the potential confounding effect of the unbalanced factors: inverse probability of treatment weighting [IPTW] and propensity score matching. Propensity score construction was based on probability estimation to be in the KA group with a non-parsimonious, multivariable, logistic regression model including main parameters distributed unequally between the KA and CA groups. The propensity score matching technique, based on the caliper method with a ratio of 1:2 [two patients in the CA group matched with one from the KA group], was used to well balance the characteristics used in the propensity score construction.

All analyses were performed using SAS version 9.4 [SAS Institute, Cary, NC, USA]; p <0.050 was considered statistically significant, and all tests were two-tailed. No correction for multiple testing was applied. This study was performed according to the STROBE guidelines.¹⁴

2.7. Ethical statement

This study involving human participants was in accordance with the ethical standards of the institutional and national research committees and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

3. Results

3.1. Overall cohort

In total, 433 patients who underwent ileocolonic resection for CD were included in the study. Among them, 337 patients [78%] had available endoscopic follow-up data and were eligible for analysis. Differences were observed between patients with and without endoscopic follow-up regarding the type of anastomosis, smoking status, B1 and B2 behaviour, and previous CD treatment [Supplementary Table 1]. In all, 91 patients had Kono-S anastomosis [KA group] and 246 had conventional side-to-side anastomosis [CA group]. Statistically significant higher proportions of men, associated perianal Crohn’s disease, and disease behaviour B3 were observed in the KA group. Additionally, there were differences in American Society of Anesthesiologists [ASA] score, previous medical treatment, and postoperative medical treatment between the two groups [Table 1].

3.2. Factors associated with endoscopic recurrence

Factors associated with endoscopic recurrence in univariate and multivariate unconditional logistic regression are presented in Table 2. In multivariate analysis, B2 behaviour [OR = 1.754; 95% CI: 1.010-3.046, p = 0.046], and the absence of postoperative biotherapy prophylaxis [OR = 1.949; 95% CI: 1.177-3.228, p = 0.010] were independently associated with an increased risk of endoscopic recurrence. High body mass index [BMI] [OR = 0.932; 95% CI: 0.878-0.989, p = 0.019] and B3 behaviour [OR = 0.583; 95% CI: 0.341-0.999, p = 0.050] were independently associated with a decrease risk of endoscopic recurrence. Importantly, in both univariate analysis [OR = 1.096; 95% CI: 0.678-1.774 p = 0.708] and multivariate adjusted analysis [OR = 1.269; 95% CI: 0.723-2.227, p = 0.407], no significant association was observed between the type of anastomosis and endoscopic recurrence.

3.3. Propensity score construction

Bias due to unbalanced factors between KA and CA groups, which could distort the comparison of endoscopic recurrence, was minimised by using a propensity score approach. A non-parsimonious, unconditional, multivariate, logistic regression model to predict the probability to be KA was obtained by entering sex, smoking status, previous resection, penetrating behavior [B3 Montreal], associated perianal disease, previous anti-tumour necrosis factor [TNF] therapy and postoperative biologic therapy prophylaxis. This model exhibited good discrimination, with an area under the curves equal to 0.7369 [Supplementary Figure 1] and acceptable calibration [p = 0.091, Hosmer–Lemeshow goodness-of-fit test].

3.4. Inverse probability of treatment weighting method

Using inverse probability of treatment weighting [IPTW] analysis based on the propensity score, the type of anastomosis [Kono-S versus conventional side-to-side] was found to have no significant association with the risk of endoscopic recurrence [OR = 1.165; 95% CI: 0.855-1.589, p = 0.3335; N = 318 with 152 events].

3.5. Matching 1:2 analysis

Then, patients with KA were matched considering nearest neighbour for the propensity score with a caliper of 0.2 in a 1:2 ratio with patients with CA. Out of 155 patients in the KA group, 61 [39.3%] were successfully matched with 122 patients from the CA group [Table 3]. Upon analysis, the matched cohort exhibited no significant differences in the various patient characteristics used to establish the propensity score: sex [p = 0.752], smoking status [p = 0.456], previous resection [p = 0.797], penetrating behaviour [p = 0.915], associated perianal disease [p = 0.415], previous anti-TNF therapy [p = 0.596], and postoperative biologic therapy prophylaxis [p = 0.4624].

Surgical features and short-term operative outcomes are detailed in Table 4. In the KA group, the rate of overall morbidity was significantly lower than in the CA group [p = 0.025] but not the rate of severe morbidity [1.6% versus 7.4%; p = 0.168] and anastomotic leakage [1.6% versus 2.5%; p = 1].

In the matched cohorts, median delay between surgery and endoscopic assessment was 6.7 months [IQR, 5.8-8.3] and not statistically different between both groups. Endoscopic recurrence ≥i2 was found in 29 patients [47.5%] and 54 patients [44.3%] in the KA group and CA group, respectively [p = 0.6745] [Table 5]. In unconditional logistic regression, Kono-S anastomosis was not associated with a lower risk of endoscopic recurrence ≥i2 [OR = 1.141; 95% CI: 0.616-2.114, p = 0.5147]. However, when focusing on severe endoscopic recurrence  ≥i3, Kono-S anastomosis was identified with a favourable profile [OR = 0.595; 95% CI: 0.284-1.248, p = 0.1697]. Clinical recurrence was assessable in 106 patients but without any difference between both groups.

4. Discussion

This comparative study aimed to assess the impact of Kono-S anastomosis [KA] versus conventional anastomosis [CA] on postoperative endoscopic recurrence in patients undergoing surgery for ileal terminal Crohn’s disease [CD], in a large cohort of 433 patients. Our results indicate that KA does not significantly decrease endoscopic recurrence compared with CA. However, we observed a potential association between KA and a lower rate of overall postoperative morbidity.

Prior studies have reported inconsistent findings regarding endoscopic recurrence after KA. Our results contradict the only published, randomised, controlled trial, which demonstrated a significant decrease in endoscopic recurrence with KA compared with CA [22% versus 63%].⁹ Conversely, we found a similar lack of effectiveness of KA in reducing endoscopic recurrence, as reported in the first cohort study describing the outcomes of this anastomosis [83% versus 79%].⁸

One possible explanation for the lack of effectiveness in reducing endoscopic recurrence lies in the design of the anastomosis, particularly with regard to mesentery preservation. In CD, the mesentery undergoes macroscopic modifications characterised by fat thickening [creeping fat] that can cover over 50% of the bowel circumference. Microscopic studies have revealed shared anomalies between the mesentery and intestinal structure, such as hyperplasia of adipocytes and connective tissue.¹⁵ Given the initial localisation of ileal mucosal inflammation on the mesenteric side of the bowel and its correlation with mesenteric inflammation, it has been suggested that the mesentery may play a pathological role in the inflammatory process. Current surgical guidelines still recommend close bowel resection in CD. The innovative concept introduced by Kono-S anastomosis involves mesenteric exclusion, whereby the mesentery is removed from the anastomosis site, which is the main site of recurrence, while preserving the mesentery’s vascularisation and innervation. Another more radical technique, mesenteric excision resembling oncological resection, has shown promising results with a reduced surgical recurrence rate [3% versus 40%] over a 5-year follow-up period.¹⁶ Retrospective cohort studies have further supported the benefits of mesenteric resection on surgical recurrence [11% versus 30%] over a 3-year follow-up, with equivalent morbidity.¹⁷ However, the level of evidence remains limited, and ongoing, randomised, controlled trials aim to provide more definitive answers.¹⁸ Combining both techniques, along with wide anastomosis and reinforcing with a supporting column as initially designed by Kono, may offer potential synergistic advantages.¹⁹ The MEErKAT trial [NIHR131988; ISRCTN16900055] is a four-arm, randomised, controlled trial comparing traditional ileocolic anastomosis with Kono-S anastomosis with or without mesenteric resection; it will attempt to determine which surgical strategy is the most effective in reducing the risk of recurrence.

Another plausible explanation for the disparity in endoscopic recurrence rates may be associated with the potential for mechanical anastomotic staples to falsely resemble endoscopic recurrence, whereas hand-sewn anastomoses tend to heal without ulcerations. In the SUPREME trial, the Kono-S anastomosis was manually executed, whereas the side-to-side anastomosis was mechanically performed, possibly accounting for the substantial difference in endoscopic recurrence rates between the two groups.⁹ In our study, a majority of anastomoses in both groups were manually created. In a recent study, van der Does de Willebois et al. have shown that the presence of staples can result in an erroneous diagnosis of endoscopic recurrence when, in reality, it signifies the healing process at the anastomotic site.²⁰ Furthermore, a multicentre, French, randomised trial [NCT05974358] is scheduled to begin soon, aiming to compare the two types of anastomosis and conduct blinded evaluations of endoscopic recurrence.

Another factor to consider is the duration of follow-up, which plays a crucial role in evaluating surgical recurrence. Although our study did not allow for a robust assessment of surgical recurrence, it remains a significant concern as it can lead to terminal ileum Crohn’s lesions and increased morbidity.²¹ Notably, a study by Kono et al. reported a reduced surgical recurrence rate after 7 years of follow-up in the KA group [0% versus 15%, p = 0.0013].⁸

Our study confirms that KA is a safe and feasible technique, as evidenced by comparable operative time and intraoperative complication rates between the KA and CA groups. We observed a significant increase in overall postoperative complications in the CA group, although the rate of severe morbidity [Clavien–Dindo grade 3–4] was not statistically different. The overall complication rate was similar to that reported in the literature for CA.²¹ However, our KA cohort exhibited a higher overall complication rate compared with previously published KA cohorts. This difference can be partially explained by a comprehensive reporting of both medical and postoperative complications, whereas other studies focused solely on postoperative complications. For instance, in a large retrospective cohort study evaluating 187 KA postoperative outcomes, authors reported a cumulative postoperative complication rate of 13%, mainly consisting of surgical site infections and abdominal abscesses.²² Our prospective data collection allowed for better detection of complications, including medical complications. The observed benefit of KA in reducing overall complications was supported by a shorter length of hospitalisation. Nonetheless, given the comparison with retrospective data, this finding should be interpreted cautiously until validated by randomised studies.

The strengths of our study primarily lie in its design. First, it was conducted in seven French academic centres affiliated with the GETAID Chirurgie network, ensuring optimal medical and surgical management in accordance with ECCO guidelines, including IBD multidisciplinary consultations and specialised, minimally invasive surgery. Second, this is the largest multicentre prospective and controlled cohort study on KA, enabling robust statistical analyses. In addition, the application of a propensity score effectively mitigated disparities between the two study groups, encompassing risk factors for endoscopic recurrence. However, in comparison with a randomised trial, there is a potential for selection bias to emerge if important characteristics are omitted from the propensity score model or if data are missing.

In conclusion, the Kono-S anastomosis is a safe and feasible technique; however, based on this large, multicentre, prospective cohort study, it does not appear to reduce endoscopic recurrence . Continuing the follow-up of the KA cohort will allow for evaluation of the potential benefit on surgical recurrence.

Funding

This research was performed without any funding.

Conflict of Interest

The authors declare that they do not have anything to disclose regarding funding or conflict of interest relating to this publication.

Author contributions

All authors qualify for authorship and agree to the submission. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication. Study design: LV, DV, JHL, and ZL. Data collection: LA, LB, AF, GM, SB, PZ, JP, LM, AB, and GT. Data analysis: AF and DV. Data interpretation: LA, LB, AF, DV, JHL, and ZL. Writing: LA, LB, AF, DV, JHL, and ZL. Critical revision and final approval of the manuscript: all authors. The results of this study have been presented in part at the 18th Congress of ECCO in 2023 [DOP Session, Copenhagen, Denmark].

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Author notes