Long-term outcome after chronic anastomotic leakage following surgery for low rectal cancer

Colorectal cancer represents the third most common malignant tumor worldwide [1]. After surgical treatment, acute anastomotic leakage (AL) is the most feared postoperative complication, with an incidence of up to 20% in rectal cancer, increasing patients’ morbidity and mortality by leading to worse oncological and functional outcomes and formation of chronic anastomotic leakage (CAL) or a persisting presacral sinus, which is associated with further complications [2‐6].

Surgical treatment of low rectal cancer and its early postoperative complications has become relatively standardized and is already thoroughly reported. Still, only limited literature is available for late-onset postoperative morbidity management, such as CAL. CAL represents a complication with complex treatment strategies and a severe impact on patients’ quality of life, some even requiring a permanent colostomy [7, 8]. These treatment strategies might also be impacted by reduced healing tendency, through increased inflammation in the affected area, in patients who received neoadjuvant radio- or chemotherapy [9].

AL and CAL may lead to systemic infections and impaired intestinal continuity, while CAL may especially result in the formation of a recurrent fistula or a persisting presacral sinus [10, 11]. Secondary complications may include involvement of the surrounding tissue, leading to osteomyelitis, necrotizing fasciitis, and periureteric fibrosis with ureteral stenosis and hydronephrosis [11]. Additionally, permanent inflammation in the region of the chronic sinus, together with the epithelialization of the surface, promotes increased cell proliferation and cell turnover rate, which increase the chance for further malignant transformations. However, since such de novo carcinomas are a rarity, extensive long-term observations would be necessary to demonstrate a connection between chronic sinus and malignant tumor development [12].

Accordingly, the literature regarding treatment strategies in patients with CAL is very heterogeneous.

In general, the treatment ranges from conservative over interventional/endoscopic to surgical management, with various individual pathways in each category. Because there is no thoroughly established treatment algorithm, some patients might also receive unnecessary multimodal treatment after primary therapy failure (e.g., permanent terminal stoma), resulting in further physical, psychological, and financial strain for these patients in some health care systems [3, 13, 14].

Furthermore, in hindsight of rising budgetary pressure on public health care systems, it seems only logical to implement a treatment algorithm for CAL with a persisting presacral sinus to reduce costs for long-term complication treatment and preventable treatment modalities.

This study aims to analyze the prevalence of CAL, suspected risk factors (e.g., chemo-/radiotherapy or surgical technique) for its development, the treatment modalities (conservative, interventional, or surgical) in patients who received surgery for low rectal cancer, and the possible influence of stoma reversal on postoperative pain and bowel continuity in patients with presacral sinus.

This study retrospectively analyzed the prevalence, therapy, and outcome of chronic anastomotic dehiscence following resection of low rectal cancer. In the study population, 71 patients (17.9%) developed any type of anastomotic leakage after LAR. According to recent literature, this rate is in the upper range, which might be attributed to patients with late-onset AL being added, and most studies only reporting on AL during the hospital stay [2, 3]. Moreover, patients primarily treated in other centers increased the rates of patients with CAL attended at our reference hospital. A total of 39 patients treated for CAL with persisting presacral sinus were further analyzed. This represents 54.9% of all considered anastomotic leakages and 9.8% of the total cohort and is therefore in accordance with the cross-sectional study by Borstlap et al. which reports rates of 48% and 9.5% [18].

Initial treatment modalities (conservative, interventional, or surgical) for CAL showed no influence on the rate of patients with chronic pain or the rate of reconstruction of bowel continuity. However, surgical therapy showed a significant advantage in patients’ survival compared to the other initial treatment modalities, despite a higher comprehensive complication index and Clavien–Dindo grade, at the time of low anterior resection, in the surgical group. Nevertheless, this survival benefit might be due to the small patient cohort and a selection bias, which was not analyzable through a retrospective study. The rate of surgical revisions and multitherapy between the initial interventional and surgical groups was comparable. This data should be seen as a trigger for future studies analyzing if primary surgical management should be considered the first option for CAL in patients presenting as fit for surgery because the limited sample size of this study does not allow for a therapy suggestion.

Factors influencing the outcome, such as multitherapy, secondary surgical revision, or stoma reversal, were evaluated in the further therapy course. A statistically significant negative relationship between multitherapy and bowel continuity was observed, wherein patients without multiple therapies had a higher rate of natural bowel continuity. Secondary surgical revision also correlated significantly with therapeutic success. Patients with secondary reoperation showed an increased rate of pain symptoms, while patients requiring no revision required a temporary or permanent stoma less often. These results again emphasize the possible association of primary surgical management to reduce the rate of unsuccessful treatment modalities. Subsequently, the influence of postoperative reversal of a protective loop ileostomy was examined, showing a significant coherence between stoma reversal and natural bowel continuity. Patients without reversal had an increased risk of staying with a permanent stoma. Although stoma reversal is naturally directly associated with bowel continuity, four (19%) patients nevertheless required a discontinuity resection to create a permanent stoma, and three (14%) patients required a temporary secondary stoma. In addition, there was a significant correlation between stoma reversal and post-therapeutic pain symptoms, showing that patients without reversal suffered from a higher risk of developing post-therapeutical pain. Patients with reversal also had a statistically significant overall 10-year survival advantage in this small cohort. According to this data, the influence of protective ileostomy reversal despite persistent CAL should be further assessed in larger cohorts, which might also be achieved in this specific cohort through multicentricity. One possible explanation for this suggested effect is that stoma reversal counteracts dehydration and electrolyte deficiency, thereby supporting granulation and epithelialization at the leakage and increasing its healing tendency [19]. The reversal of a protective stoma thus has a positive association with the therapy success regarding bowel continuity and symptomatic complaints, and survival. However, misinterpretation as a result of the small sample size and selection bias, due to preoperative assessment, is possible. The proposed selection bias is a possible confounder as some patients might not be fit enough for stoma reversal surgery or suffer from more severe comorbidities inhibiting stoma reversal.

Moreover, various clinical risk factors were analyzed. Hereby, neoadjuvant radiochemotherapy and neoadjuvant chemotherapy were identified as factors increasing the possibility of developing chronic leakages as already described in the literature.

This study was performed in a retrospective, non-randomized setting and had a small study population, especially in the subgroup analysis on stoma reversal. Accordingly, there might be a higher occurrence of type II errors. In addition, the analyzed period spans 25 years, during which different treatment concepts without constant therapeutic strategies as well as new surgical methods have been established. Furthermore, patients from peripheral hospitals influenced the study population through inconsistent therapy concepts and pre-existing complex courses. As a result, adjustment for confounding was not performed due to the distinct and limited sample size. Results should be interpreted accordingly. Although rectovesical and/or rectovaginal fistulas can be a possible clinical characteristics of CAL, these patients were excluded because they require different treatment modalities with involvement of additional departments. For future studies, patient-reported outcome should be considered an indicator for therapy success, as this allows a more detailed interpretation.

Summarizing our results for therapeutical management, our data suggest that patients, which require a surgical revision or any kind of other additional therapy after their primary treatment, might suffer from reduced rates of bowel continuity. Nevertheless, before a specific treatment recommendation is possible, these results have to be validated in larger cohorts to examine if in patients with a CAL after LAR, primary surgical treatment should be preferred. In our cohort primary surgical treatment showed the same therapy result in the sense of bowel continuity and absence of pain. Furthermore, our data suggests that it may reduce the rate of patients receiving multiple therapies due to unsuccessful conservative or interventional treatment attempts.