Clinical Efficacy of Intersphincteric Resection for Low Rectal Cancer Compared With Abdominoperineal Resection: A Single-Center Retrospective Study

Background In recent years, intersphincteric resection (ISR) has been increasingly used to replace abdominoperineal resection (APR) in the surgical treatment of ultra-low rectal cancer. Aim This study was to compare the clinical efficacy of ISR and APR. Methods Between 2012 and 2018, 74 consecutive patients with ultra-low rectal cancer underwent ISR or APR in our medical center. A retrospective comparison of these 2 procedures was performed. Results A total of 43 patients underwent ISR and 31 underwent APR were included in the study. No significant differences were found between 2 groups in gender, age, BMI, and ASA score. Intersphincteric resection group showed shorter operative time (P = .02) and less blood loss (P = .001). Hospital stays, time to soft diet, and postoperative 30-day complications were not significantly different between the 2 groups. R0 resection achieved 100% in both the groups. As for the long-term outcomes, the survival and recurrence rate were similar between 2 groups. Moreover, the LARS and Wexner score showed that the postoperative anal function after ISR were satisfactory. Conclusion This study suggested that ISR was feasible and safe for selected patients with ultra-low rectal cancer, with clinically superior outcomes in select patients (small tumors/further from the anal verge) and similar oncological outcomes to APR, and the anal functional outcomes after ISR were acceptable.


Introduction
According to the global tumor statistical analysis, in both sexes combined, the incidence rate of colorectal cancer ranks fourth and the mortality rate ranks second. The rectal cancer, especially the low rectal cancer, which is located 5 cm or less from the anal verge, is more common, accounting for 70% to 80% of the total number of rectal cancers, which may be related to dietary habits, social environment, genetics, and other factors. 1,2 At present, the treatment of low rectal cancer mainly adopts the comprehensive treatment with surgery as the main and chemoradiotherapy as the auxiliary. Abdominoperineal resection (APR) is a standard operation for low rectal cancer, completely removes the distal colon, rectum, and anal sphincter complex to radically resect the tumor, resulting in a permanent colostomy ineluctably, which greatly influences the patient's quality of life. 3 In recent years, intersphincteric resection (ISR) has been performed at more and more specialized institutions to extend the opportunity for anus preservation for low rectal cancer. 4 However, there are also potential risks associated with ISR, including the tumor recurrence and postoperative anal function. To date, only a few studies have compared the long-term outcomes of ISR with APR and reported the postoperative anal function in patients after ISR.

Aim
This study was to compare the clinical efficacy of ISR and APR. In this study, the short-term and oncological outcomes, including disease-free survival and local recurrence, of ISR and APR for low rectal cancer were compared. Moreover, we conducted a preliminary follow-up and reported on the postoperative anal function of patients who underwent ISR.

Material and Methods Patients
From January 2012 to December 2018, a total of 464 patients with rectal cancer underwent curative surgical treatment in our medical center. All patients underwent routinely assessed preoperatively with digital anorectal examination, colonoscopy, pelvic magnetic resonance imaging (MRI), and/or endorectal ultrasonography, thoracoabdominal computed tomography (CT) scan, and sphincter manometry. Among them, 43 patients underwent ISR and 31 patients underwent APR who had invasive rectal cancer located within 5 cm from the anal verge were eligible and included in this study ( Figure 1). Inclusion criteria were histologically proven low rectal adenocarcinoma. Exclusion criteria included colon cancer or upper rectal cancer; tumor distant metastasis (clinical stage IV period); emergent setting, such as acute bowel obstruction or perforation from cancer 5 ; inflammatory bowel disease; and familial polyposis. What is more, the inclusion and exclusion criteria for ISR must be very strict. According to our clinical experiences, the inclusion criteria of ISR were T1-T3 tumor, well-moderately differentiated adenocarcinoma, and good anal function (AFN), regardless of age. The exclusion criteria of ISR were as follows: T4 tumor, fixed tumor, untreatable distant metastases, poorly differentiated adenocarcinoma, psychiatric disease, and poor AFN. Moreover, studies have shown that the oncological outcomes of the patients accepted neoadjuvant chemoradiotherapy (CRT) could be improved, 6,7 so the patients with T3-4 clinical stage or clinical lymph node metastasis were arranged neoadjuvant CRT; the neoadjuvant chemotherapy regimen was Capeox.
The study was approved by the Institutional Review Board of our hospital, and all patients provided written informed consent.

Surgical Procedure for ISR
Intersphincteric resection was performed on the basis of the theories described by Schiessel and his teams. 8 After successful general anesthesia, the lithotomy position was taken, the operative field was routinely disinfected, and the laparoscopic instruments were inserted into the abdomen.
Abdominal operation: the mesosigmoid was incised and the fusion fascia gap (Toldts gap) was extended, then ligation and dissection were performed at the root of the inferior mesenteric artery. Total mesorectal resection (TME) was performed to dissociate the rectum and its mesangium to the puborectalis muscle. 9 When necessary, to dissociate the splenic flexure to ensure the anastomosis of sigmoid colon and anal canal without tension. The rectum muscle tube was exposed at the pelvic floor and then separated downward along the gap between the internal and external sphincters. This procedure was performed laparoscopically and no cases were converted to open surgery. Perineal operation: the rectum and anal canal were irrigated and operating field was fully exposed. The lower incisional margin was determined at the distal 1-2 cm of the tumor, and the submucosal injection of 1:10,000 epinephrine was performed to reduce bleeding. Then the mucosa, submucosa, and internal sphincter were successively cut. Following closure of the anal orifice at the distal cut end, the rectum and mesangium were dissected in the sphincter plane. And then the position of the upper incisional margin was confirmed and the specimen was removed. Finally, a diverting ileostomy was established in all cases.

Data Collection and Definitions
Data on patients' characteristics and short-term outcomes were obtained from medical records and examination results. The basic diseases of patients in this study included hypertension, diabetes, heart disease, and chronic bronchitis. Postoperative 30-day complications were classified by Clavien-dindo scoring system. 10 Postoperative anal function was assessed among all patients who received closure of the ileostomy after ISR. The LARS and Wexner score were used to assess the anal function. 11,12 Low anterior resection syndrome score includes 5 items: exhaust control disorder (0∼7 points), loose feces (0∼3 points), defecation frequency (0∼5 points), redefecation within 1 hour after defecation (0∼11 points), and defecation urgency (0∼16 points). According to the score (0∼42 points), it was divided into 3°: 0∼20 points indicate "no low anterior resection syndrome"; scores ranging from 21 to 29 indicate "mild low pre-resection syndrome," while scores ranging from 30 to 42 indicate "severe low preresection syndrome". The Wexner score is a questionnaire that indicates the severity of fecal incontinence. It consists of 5 items, including the type and frequency of incontinence (solid, liquid, gas, and whether to use pads) and lifestyle changes. The total score is 20 points. A score of ≥10 is an indicator for stool incontinence. The higher the score, the worse the anal function. Patients were followed up at 6-month intervals for 2 years, and once annually thereafter. Follow-up examinations included physical examination, serum tumor marker assay, chest X-ray or CT, abdomino-pelvic CT or MRI, and colonoscopy. Recurrence was determined by imageological examinations and/or pathological confirmation. Local recurrence was defined as recurrent disease in the pelvis, including the anastomosis. Distant metastasis was defined as recurrent disease outside the pelvis. Follow-up and anal function data were obtained by telephone interviews or outpatient reexamination.

Statistical Analysis
All statistical analyses were conducted using software SPSS 16.0. Quantitative data were reported as mean, median, standard deviation (SD), and interquartile range (25%-75%). Student's t-test and chi-squared test (or Fisher's exact test) were used to compare normally distributed continuous variables and categorical variables, respectively. Survival curves were performed using the Kaplan-Meier method, and the difference between curves was assessed by the log-rank test. P values <.05 were considered statistically significant.

Patient Characteristics
The characteristics of the patients who underwent ISR or APR were listed in Table 1. No significant differences were observed between the 2 groups in age, sex, BMI, ASA score, basic diseases, and frequency of neoadjuvant CRT. R0 resection was performed in all patients.

Short-Term Outcomes
Operative time (mean 261.8 min ISR; 319.2 min APR; P = .021) and blood loss (mean 185.8 mL ISR; 375.0 mL APR; P = .001) were significantly lower in the ISR group compared with the APR group. The rate of 30-day postoperative complications (20.9% vs 25.8%) and Clav-ienDindo score were similar between the ISR and APR groups. As for the number of harvested lymph nodes, 2 groups showed no statistically significant difference (P = .219). In the ISR group, postoperative complications occurred in 9 patients, including 3 patients with anastomotic leakage, which was the most common complication in the ISR group. In the APR group, the most common complication was hemorrhage. One patient in the ISR group died in the hospital due to pulmonary infection, while there was no in-hospital mortality in APR group. The ISR group showed no difference from the APR group in time to anal or stoma exhaust (P = .953), time to soft diet (P = .248), postoperative hospital stays (P = .087), and 30-day reoperation or readmission (P = .814) ( Table 2).

Pathological and Oncological Outcomes
As shown in Table 3, the average distance from the tumor to the anal verge in the ISR group was 4.5 cm, while the APR group was 3.4 cm, showing a statistically significant difference (P = .0001). No difference was noted in tumor size in the 2 groups. Two groups were similar in tumor differentiation (P = .112). Similarly, pT stage and pN stage were comparable between the 2 groups. There was no statistical difference between the distal resection margin between the 2 groups (P = .244). Moreover, no patient showed positive circumferential resection margin; each patient achieved R0 resection, showing that ISR and APR both have a good effect on radical tumor resection.
The median follow-up period was 26 months (interquartile range 13-54 months) and 39 months (interquartile range 18-58 months) in the ISR group and APR group, respectively.
Three-and Five-year overall survival rates after ISR were 88% and 86% (95% confidence interval for the mean survival time was 64.5-79.4) and 77% and 57% (95% confidence interval for the mean survival time 49.9-70.4) in the APR group, respectively, showing no statistically significant difference in patients' overall survival when comparing both groups (P = .136, see Figure 2). The disease-free survival rates at 3 and 5 years were 79% and 72% in ISR and 63% and 63% in the APR group, respectively. There was no significant difference in patients' disease-free survival between the 2 groups (P = .303, see Figure 3). In terms of local recurrence, the 3-year local recurrence rate was 5% in the ISR group and 19% in the APR group, with no significant difference between the 2 groups (P = .309, see Figure 4).

Functional Results
We analyzed the functional data for all patients who had received closure of the ileostomy after ISR surgery. According to the low anterior resection syndrome score (LARS), of the 43 patients, 12 (28.0%) scored between 0 and 20, indicating no low anterior resection syndrome, and 31 (72.0%) scored between 21 and 29, indicating mild low anterior resection syndrome. None scored above 30. The mean LARS score was 19. According to the Wexner score, thirty-nine out of 43 patients had a score <10 (91%), representing good continence, and 4 patients had a Wexner score ≥10 (9%). The mean Wexner continence score was 5.9. Overall, the postoperative anal function of patients who underwent ISR was satisfactory.

Discussion
As laparoscopy and even robotics become more widely used, more and more patients with low rectal cancer can achieve tumor resection and anal preservation through sphincter-saving surgery. However, for ultra-low rectal cancers near the anus, which are located too close to the anus and have too little pelvic space, APR is still the standard procedure to achieve a negative resection margin, but the postoperative pain due to the permanent colostomy is indelible. Thus, it has always been the research direction to ensure the radical resection of tumor and maintaining the original gastrointestinal structure. Intersphincteric resection was first proposed by Professor Lyttle and Professor Parks in 1977 for the treatment of  patients with inflammatory bowel disease requiring colon and rectal resection, and it was first used in 1994 to treat rectal cancer. Intersphincteric resection dissected part or the entire internal sphincter by dissecting the sphincter space, and sutured the proximal colon and anal canal. It emphasized the correct separation of the inner and outer sphincter. The range of internal anal sphincter resection based on tumor location was divided into complete ISR (complete resection of internal sphincter and resection at the sphincter groove), subtotal ISR (subtotal resection of internal sphincter and resection between dentate line and sphincter groove), and partial ISR (partial removal of the internal sphincter and removal at the dentate line). 13 The proposal of ISR has brought hope to patients and the medical community, but also brought us a series of questions. This study compared a series of outcomes of ISR vs APR surgery for ultra-low rectal cancer patients in our center, including the short-term, oncological, and functional outcomes. Our data showed that ISR might provide a feasible and safety alternative to APR based on the optimized oncologic cure rate and satisfactory defecation function.
In our study, ISR group showed shorter operative time and less blood loss than APR group. Experienced surgeons could perform ISR combined with laparoscopic total mesorectal excision, thereby greatly reducing the damage. However, the perineal operation part of APR inevitably increased the surgical damage. We considered that the smaller surgical injuries were of great significance for the postoperative rapid recovery of patients. Furthermore, the mean length of postoperative hospital stay was shorter in ISR group in our study, although there were no statistical differences. In addition, 30-day complications and Clavien-Dindo score showed no statistical difference between the 2 groups; among them, anastomotic leakage was the main complication in ISR group, while hemorrhage was the main complication in APR. We considered that during the perineal operation of APR, the massive bleeding was mainly caused by the injury of the pre-sacral vein; when the visual field was poor or the vein was exposed, the electric knife should be used with caution. If bleeding occurs, hemostatic material compression and suture were preferred first. In addition, continued small amount bleeding in the perineal field could also cause serious complications, and the effective hemostasis for common bleeding sites such as the posterior urethra, prostate, seminal vesicles, and posterior vaginal wall must be paid attention to. What is more, we suggested that one-stage pelvic floor reconstruction and perineal suture should be performed as much as possible for patients with APR, which was of great significance for preventing incision infection, perineal hernia, and effusion. Clinically, as for the anastomotic leakage after ISR, we considered that the main causes included too-low anastomosis, poor blood supply or excessive tension in the anastomosis, and more opportunities for pelvic and intestinal contamination. Before ISR surgery, surgeons should fully understand the various complications, comprehensively evaluate patients before surgery, correctly grasp the surgical indications, and make preventive stoma when appropriate. Meanwhile, surgical skills and experience are also very important. These short outcomes demonstrated that ISR may provide a safe alternative to APR, with less surgical trauma and faster recovery.
In recent years, the survival results of ISR and APR have been reported in several studies. Rullier et al 14 found that the combination of preoperative radiotherapy with sphincter-saving procedures not only may facilitate good perioperative and functional outcomes but is also safe from the oncological point of view. Schiessel et al studied 121 rectal cancer patients who underwent ISR. After 16 years of follow-up, they found that there were no statistically significant differences in postoperative survival and recurrence rate between patients with ISR compared with low anterior resection (LAR) and APR. 15 Shunsuke Tsukamoto et al 16 followed up 285 patients (112 ISR and 173 APR) by propensity score matching and found the similar oncologic outcomes for ISR and APR without preoperative chemoradiotherapy in patients with low rectal cancer. In our study, the overall survival, disease-free survival, and local recurrence rate were similar between the 2 groups; R0 resection achieved 100% in both the groups, and the number of lymph nodes harvested and distal resection margin showed no difference, indicating that oncological outcomes and surgical efficacy were similar between ISR and APR. In addition, we found that postoperative local recurrence or metastasis in the ISR group mostly occurred in patients with T3 pathological stage or pathological lymph node metastasis or with a small number of lymph node detection, suggesting that ISR surgery should be carefully selected for patients. According to our experience, the indications for ISR include (1) the tumor is within 5 cm of the anal verge; (2) the tumor does not invade the external anal sphincter; (3) well-moderately differentiated; and (4) good anal function. Therefore, the inclusion criteria for ISR in this study are T1-T3, welldifferentiated tumors, no invasion of the external anal sphincter, and good preoperative anal function. According to the postoperative pathological results of the patients, ISR patients were more benign than the APR patients in terms of tumor location (further to AV, P = .0001), size, stage, and degree of differentiation, although the latter 3 results did not show significant statistical difference, which can explain the worse long-term survival outcomes of the patients in the APR group. According to a previously published metaanalysis that included 12 studies by our team, there was no statistically significant difference between the ISR and APR groups in 5-year OS, 5-year DFS, and 3-year LAR. 17 We also considered that preoperative evaluation was very important to ensure the radical effect of ISR, especially the pelvic MR, which was of great significance in judging the stage of rectal cancer and predicting whether the external sphincter was involved. For patients with insufficient lymph node detection or high pathological stage, postoperative follow-up interval should be shortened and close observation should be conducted. A number of studies have shown that the local recurrence rate of ISR is 4.9%-19.7%, [18][19][20][21]30 which is similar to APR. 22 Local recurrence usually occurs within 2 years after surgery. The possible reasons include advanced tumor stage, deep infiltration, incomplete lymph node dissection, incomplete resection of mesangium, and insufficient distal tumor margins. Yamada et al reported 107 patients with T2/T3 tumors with a distal margin of 2 cm and T1 tumors of 1 cm. The recurrence rate of these patients was only 2.5% at 5 years after surgery. 15 For patients with ultra-low rectal cancer who are less than 5 cm from the anal margin or less than 2 cm from the  dentate line, it is difficult for traditional surgical procedures to achieve a safe margin of 2 cm. The results of the above multiple studies show that under the premise of strictly grasping the indications, the postoperative oncological outcomes of ISR were satisfactory. [23][24][25] In our study, the mean length of the distal resection margin of ISR patients was 2.2 cm, and all patients achieved R0 resection, which further confirmed the superiority of ISR surgery. By removing part or the entire internal anal sphincter, more than 2 cm can be obtained; the resection margin may be able to avoid local recurrence.
In addition, whether patients could obtain good stool control after ISR due to partial or complete loss of the internal sphincter is always a concern. Most scholars believed that preserving all internal sphincter and mucosa was the key to maintaining good stool control function after the surgery. 26 When the distance between the lower edge of the tumor and the anorectal ring was less than 2 cm, most or even all internal sphincters needed to be removed to ensure the radical treatment effect, and the patients would suffer severe defecation control dysfunction and even anal incontinence. To avoid this outcome, APR was often preferred. Due to the removal of part or the entire internal sphincter in ISR, defecation control dysfunction would inevitably occur after a period of time, but the ability to control defecation could be recovered to varying degrees within 3 to 6 months after surgery. In our study, most of the patients after ISR presented with low anterior resection syndrome to varying degrees, but according to LARS and Wexner score, the anal function of the vast majority of patients after ISR was satisfactory. Yamada et al 27 also considered that postoperative anal function of ISR was generally satisfactory, and found no significant difference in postoperative defecation times between patients with partial internal sphincter resection and patients with subtotal resection, but anal function was relatively poor in patients with complete ISR. Therefore, it is of great significance to retain as much internal anal sphincter as possible to maintain good postoperative stool control. Motoi Koyama et al 28 suggested that younger patients with T1 or T2 rectal cancers who required no preoperative therapy are ideal candidates for ISR. With the help of the magnification effect of the laparoscope, the clearance and approach selection could be more accurate, and the perirectal tissue damage could be more effectively avoided. At the same time, the principle of total mesorectal excision should be strictly followed, and the peripheral incision edge should be kept smooth and intact, and the instinctual sense receptors located outside the intestinal wall should be preserved. This was of great significance for the recovery of postoperative anal function.
Currently, comprehensive treatment is emphasized for low rectal cancer. Neoadjuvant radiotherapy and chemical therapy are recommended for T3-4 or patients with positive lymph node metastasis, which can reduce the tumor stage, reduce the positive rate of circumferential resection edge, and improve the anal retention rate. However, many scholars considered that neoadjuvant chemoradiotherapy would increase the incidence of anastomotic fistula and impair postoperative anal function. 29 Ito et al 30 had performed postoperative anal function evaluation in 96 cases of patients with the ISR; single factor analysis showed that excessive internal sphincter resection range and preoperative radiotherapy and chemical drug treatment were the adverse factors affecting the anus function; multi-factor analysis showed that only the preoperative radiation therapy and chemical therapy were the independent factors affecting the anus function. Neoadjuvant chemoradiotherapy and sphincter-preserving total mesorectal excision give patients with locally advanced rectal cancer good local control and survival. However, the combination of the 2 approaches was associated with a variety of anorectal dysfunction, such as anal sphincter dysfunction, anorectal hypoesthesia, anorectal inhibitory reflex deficiency, rectal storage compliance, and volume reduction, which might permanently affect the patient's quality of life. Although neoadjuvant CRT had some toxic and side effects, we considered that preoperative radiotherapy had obvious effects of tumor downstaging, while chemotherapy, as a systemic treatment, could control tumor metastasis and spread, and had synergistic effect with radiotherapy. Therefore, neoadjuvant CRT had obvious effect on reducing local tumor recurrence and controlling distant metastasis. Nowadays, radical surgery combined with neoadjuvant CRT has become the standard of care for locally advanced rectal cancer (T3∼4 and/or N+). In the future, the application of neoadjuvant therapy will be more individualized, which requires a more comprehensive assessment of lesion location, stage, pathological characteristics, and individual risk factors before treatment to guide the selection of individualized programs. This puts forward higher requirements for imaging (including 3D endoscopic ultrasound and high-resolution MRI), pathology, molecular biology (such as cell proliferation marker Ki-67, radiosensitivity related genes, apoptosis index, B-K-RAS gene, etc.), and other auxiliary examination techniques.
The limitations of this study were retrospective and nonrandomized design of a single center and selection bias of patient. In addition, the sample size was also limited and the skill and experience of the surgeon might influence the outcomes. Therefore, it is necessary to expand the sample size and carry out multi-center research.