The cumulative incidence of HR was lower for malignant disease than for benign disease, although malignant disease was not an independent significant predictive factor for the incidence of reversal in the multivariate analysis. After reversal, SFS reduced significantly in the malignant disease group than in the benign disease group as the time from HR increased. HR was comparable in terms of functional outcomes in both groups.
HP remains an important surgical option due to the increasing incidence of colorectal cancer and diverticulosis of the left colon worldwide and in Asian countries, respectively [14]. The rate of stoma closure was reported to be lower after HP than after primary anastomosis [4]. In particular, the stoma closure rate after HP was different between benign and malignant diseases [5, 6]. Our findings showed that the cumulative incidence of HR was significantly different in the univariate analysis, but not in the multivariate analysis, between benign and malignant diseases.
Furthermore, our findings demonstrated that low CCI, urgent surgery, and home discharge were independently associated with a higher incidence of colostomy reversal. The CCI score captures the age and comorbidities of the individual and represents their background [11]. Royo-Aznar et al. [5] reported that patients with a low CCI had a higher rate of HR. In this study, the incidence of colostomy reversal was not significantly associated with ASA-PS at the time of HP, but was significantly associated with CCI. If patients with severe systemic disease at the time of HP (high ASA-PS) recovered, had a low CCI, and were discharged home, they had a good chance of HR. Urgent surgery meant that HP had to be performed and it was a lifesaving procedure. In such cases, the patient has a good chance of HR if they recover. Our data could help surgeons provide accurate information to patients and their families about the prospect of colostomy closure and a permanent stoma prior to obtaining informed consent.
HR is a difficult surgery with a high risk of anastomotic complications and mortality [7, 8]. However, with the introduction of minimally invasive techniques [15–18], the incidence of complications after HR has lowered [19–22]. Therefore, the indications of HR can now be expanded. Although there are many reports on short-term outcomes after HR [22, 25], only a few studies have examined its long-term outcomes [9, 10]. We speculate that the long-term follow-up of patients after HR may result in the identification of problems such as cancer recurrence in patients with malignant disease and anorectal disorders. These issues require further investigation.
Because patients with malignant disease undergoing HP often have colorectal obstruction or perforation, subsequent disease recurrence may occur even if the reversal is successful [24, 25]. In this study, stoma recreation was required in three cases of cancer recurrence in the pelvis. Furthermore, SFS reduced significantly in the malignant disease group compared to the benign disease group as the time from HR increased. For malignant diseases, an appropriate indication for HR, such as setting the adequate interval between surgeries, may avoid unnecessary HR. As mentioned earlier, in our center, HR is indicated when 1) there is no disease recurrence for at least 6 months post-HP and after adjuvant chemotherapy in patients who underwent curative HP, and 2) there is no disease progression in patients with distant metastasis controlled by systemic therapy. Although there was no significant difference, the interval from HP to HR in the three patients who underwent stoma recreation owing to cancer recurrence was shorter than that in the remaining 15 patients in the malignant disease group (9.7 months vs. 12.4 months, P = 0.250). A previous study reported HR intervals of 282 days (9.3 months) for malignant disease [6]. Determining the appropriate interval for HR could be the object of future research. Additionally, it is important to obtain informed consent before HR in patients with malignant disease owing to the risk of recurrence.
Few studies have examined anorectal function after HR. In their study of 64 patients with colostomy reversal, Sander et al. [10] reported that 15.6% and 17.2% of patients had minor LARS and major LARS, respectively. Caille et al. [9] stated that among 21 patients who underwent reversal after HP due to failure of the previous anastomosis, 33.3% reported minor LARS and 23.8% reported major LARS. In our study, 13.0% of 23 patients reported minor LARS and 17.4% reported major LARS. This result is comparable to those of previous studies. Furthermore, the present study revealed no significant differences in anorectal function between those with benign and malignant diseases. However, a short residual rectal stump could be associated with poor function. Although there were only two patients with residual stumps below the peritoneal reflection, they experienced major LARS. Interestingly, a previous study demonstrated that the length of the rectal stump did not differ significantly among patients with no, minor, and major LARS [10]. Further studies will be required to investigate the association between anorectal function and the length of residual stump. Based on our finding of a high incidence of HR, we suggest that interventions such as pelvic floor muscle exercises should be considered before HR for patients with a high likelihood of HR and a short residual stump.
This study had some limitations. First, we retrospectively collected data from the surgical database and medical records of a single center. Second, the sample size might not be sufficient to examine each variable related to the outcomes. Multicenter studies with large samples and minimal bias are required for more reliable statistical analyses. Third, the effect of time from HR on post-HR anorectal function varied from case to case, which may have influenced the functional assessment results. Fourth, because pre-HP anorectal function was retraced and assessed at the time of the survey, the actual function may not have been accurately represented. Further prospective studies are required to overcome these limitations.
Our findings indicated that the incidence of HR and SFS after HR could be worse in patients with malignant disease than in those with benign disease, although anorectal function after reversal was not significantly different. The postoperative long-term course of HR may vary between patients according to the indication for HP.