Adjuvant chemoradiotherapy versus radical surgery after transanal endoscopic microsurgery for intermediate pathological risk early rectal cancer: A single-center experience with long-term surveillance.

BACKGROUND
The choice of subsequent treatment for intermediate-risk rectal tumors after transanal endoscopic microsurgery between adjuvant chemoradiotherapy and total mesorectal excision is controversial. The present study aimed to compare survival and functional outcome between these 2 strategies.


METHODS
This retrospective study included intermediate-risk patients with early rectal cancer after transanal endoscopic microsurgery in our center between 2010 and 2017. Patients were divided into adjuvant treatment and total mesorectal excision groups. Intermediate risk was defined as pT1 with lymphovascular invasion, poor differentiation, or large diameter (3-5 cm) or pT2 with small diameter (<3 cm). The study was based on follow-up data on survival and results from distributed validated scales for functional outcome.


RESULTS
Postoperative overall survival and disease-free survival were comparable between the groups (P = .619 and P = .712, respectively). Pathological T stage was an independent risk factor for disease-free survival (hazard ratio 3.09, 95% confidence interval 1.66-4.18, P = .044). Anorectal symptoms, such as buttock pain, were significantly prevalent in the total mesorectal excision group (P = .030). In addition, the total mesorectal excision group presented with poorer bowel function, including stool urgency (P < .001), bowel frequency (P = .016), severity of low anterior resection syndrome (P = .039) and total low anterior resection syndrome score (P = .040). Except for a lower score of vaginal lubrication in the total mesorectal excision versus the adjuvant treatment group, sexual function was similar between the groups.


CONCLUSION
Similar to total mesorectal excision, adjuvant chemoradiotherapy is an alternative option for intermediate-risk early rectal cancer after transanal endoscopic microsurgery and is associated with similar survival outcomes and better bowel function.

0.001), bowel frequency (p = 0.016), severity of LARS (p = 0.039) and total LARS score (p = 0.040). Except for a lower score of vaginal lubrication in TME group than the adjuvant treatment group, sexual function was similar between the groups.

Conclusion
Like TME, adjuvant chemoradiotherapy is an alternative option for intermediate risk early rectal cancer after TEM and is associated with similar survival outcomes and better bowel function.

Background
Because of the widespread popularization of colonoscopy and other screening methods, the overall incidence of rectal cancer has exhibited a dramatically increase in recent years [1,[2][3][4]. Total mesorectal excision (TME) has remained the gold-standard for the treatment of rectal cancer without metastases since it was proposed by professor Heald in 1982 [5]. Radical TME surgery, primarily including low anterior resection (LAR) and abdominoperineal resection (APR), was adopted for over 30 years [6]. Investigations of oncologists validated that early stage rectal cancer, especially T1 rectal cancer without risk factors for recurrence and metastases (i.e., low differentiated and lymphatic or vascular invasion) may be treated via local excision with the absence of radical TME [7][8][9]. Transanal endoscopic microsurgery (TEM), which was invented by Dr. Buess in 1983, was put into constant use for the treatment of rectal carcinoma over years of practice [8][9][10]. The full-thickness resection makes it an advisable approach for low-risk early rectal cancer, according to several guidelines [9,[11][12][13][14]. Despite of the advantages, the de ciency of resecting lymph nodes in TEM resection remains highly controversial.
Patients with poor histological characteristics hinder its application in clinical practice. Patients with pT1 or pT2 stage rectal cancer with high risk factors after TEM are highly recommended to undergo radical resection to lower the recurrence risk and improve survival outcome [15][16]. However, patients with intermediate risk of local recurrence and lymph node metastases after TEM occupy over half of locally resectable rectal cancer patients, which create a dilemma for treatment options between adjuvant chemoradiotherapy (CRT) and radical surgery [17]. TME may include lymph node dissection, but it is accompanied with a higher morbidity rate. CRT is bene cial for health-related quality of life (HRQoL) because of organ preservation, but it is inferior to radical surgery because of the absence of lymph node dissection. An increasing number of studies reported high morbidity and HRQoL problems caused by radical surgeries [16,[18][19][20].
The present study primarily compared oncological and functional outcomes between TME and adjuvant CRT in patients who were diagnosed with early rectal cancer with intermediate risk after primary TEM resection at our hospital.

Methods
Patients and study design A retrospective study was performed based on the constructed database at Peking Union Medical College Hospital (PUMCH), Department of General Surgery. TEM has been adopted for treating early rectal cancer for one decade in PUMCH. Surgeons in our unit have accumulated abundant clinical experience, and patients' intact pathological and clinical information were available in the speci c database created in our clinical center.
The present study involved patients who underwent TEM in our clinical center and were diagnosed with intermediate risk early rectal cancer via pathological examination, in accordance with relevant guidelines, between January 2010 and December 2017. Intermediate risk early rectal cancer was de ned as a patient with one of the following properties: (1) pT1: Diameter range from 3 to 5 cm or < 3 cm but accompanied with at least one risk factor, including lymphatic and/or venous invasion and/or neural invasion and/or poor differentiation; or (2) pT2: Diameter less than 3 cm with the absence of any risk factors. For large benign lesions with small invasive component, tumor size was considered to be the maximum diameter of malignant tissue during pathological examination. Before the TEM procedure, all patients underwent pelvic MRI to exclude nodal metastases and preoperative colonoscopy to obtain biopsy results to satisfy the indications for surgery. Patients who met the following criteria were excluded from the study: (1) unable to give consent due to dementia; (2) barrier in comprehension of the questionnaire provided; (3) failed to return questionnaires by the cutoff date; and (4) diagnosed with malignant rectal neoplasms other than adenocarcinomas. All patients in our clinical center diagnosed with intermediate risk pT1 or pT2 rectal cancer were recommended to receive the following treatment, including adjuvant CRT or radical surgery. Patients bearing intermediate risk rectal cancer who underwent adjuvant therapy after TEM were in a follow-up that included more intense postoperative examinations.
Patients meeting the selection criteria were grouped into the TME group (including LAR and APR) or adjuvant treatment group. Oncological outcome was acquired from telephone interviews and outpatient follow-up records. Functional outcome was primarily obtained from distributed paper questionnaires, online llings and telephone follow-up. The European Organization for Research and Treatment of Cancer Quality of Life colorectal cancer speci c module (EORTC-QLQ-CR29) was used to obtain an overall impression of a patient's HRQoL. Low anterior resection syndrome (LARS) score and Wexner score were used as scales to assess patients' defecatory or anorectal function. The Female Sexual Function Index (FSFI) and ve items version of the International Index of Erectile Function (IIEF-5) were used to evaluate female and male postoperative sexual function, respectively. To avoid bias from potential confounding variables, we used the Propensity Score match (PSM) method (match tolerance = 0.05). Enrolled patients had an average 42.63 months of follow-up (range 12 -73 months).

Radical surgery
Specimens resected during TEM were reviewed by experienced pathologists and diagnosed as intermediate risk pT1 or pT2 rectal cancer. Patients underwent radical surgeries 4 -8 weeks after primary TEM. The choice of surgical technique between LAR and APR were primarily determined by the distance from the anal verge. All patients were informed of the possibility and inconvenience created by a permanent stoma. The anastomosis method involved stapling and hand sewing. Surgeons assessed the suitable time for diverse stoma if it was suitable. Patients received a contrast enema and abdominal CT scan before the reversal to certify the continuity and liquidity of the distal digestive tract and exclude probable distant metastasis. Patients who had lymph node metastasis on pathological examination after TME were recommended for subsequent adjuvant therapy.

Adjuvant treatment
Postoperative CRT included concurrent chemoradiotherapy (CCRT) and radiotherapy alone. Patients who received CCRT had a long course of radiotherapy with the chemotherapeutics. Chemotherapy regimen primarily included single-agent capecitabine, oxaliplatin plus intravenous 5-Fu/LV (FOLFOX) or oxaliplatin combined with capecitabine (XELOX). The choice of regimen was determined based on the oncologist's judgment.

Oncological outcome
Patients' oncological prognoses were assessed using overall survival (OS) and disease-free survival (DFS). The interval of OS was the time between the primary surgery and the date of death. The DFS interval was calculated as the interval between the primary surgery and recurrence date. Recurrence was primarily con rmed via biopsy during follow-up colonoscopy, and death information was certi ed via telephone interviews.

Functional outcomes
The following questionnaires were distributed after radical surgery or the eventual course of adjuvant treatment. For patients who received LAR with diverting ileostomy, the questionnaires were distributed after the reversal of stoma. For patients who underwent APR or LAR without diverting ileostomy, the questionnaires were sent after primary radical surgeries. Since patients in the adjuvant therapy group did not have stoma, the items related to stoma care was not listed in the distributed modules.

EORTC QLQ-CR29
EORTC QLQ-CR29 is a speci c module of conventional EORTC QLQ-C30 for colorectal cancer patients, and it consists roughly of 4 main scales and 18 diverse single items [21]. All items are divided into functioning scales and symptom scales. Answers were graded as the extent of severity from 1 (not at all) to 4 (very much) and from 1 (very poor) to 7 (excellent) for items about HRQoL. The initially calculated scores were transformed into scales of 0 -100 according to a previously reported standard procedure [22].
LARS score and Wexner score These two scales were used primarily to evaluate defecation problems. The LARS score has 5 questions on bowel movement-related disability, such as control of flatus, feces and liquid stool. The total LARS score ranges from 0 to 42. Patient scores of 0 to 20, 21 to 29 and 30 to 42 were categorized as no LARS, minor LARS and major LARS, respectively [23]. The Wexner score includes 5 questions on the frequency of solid and liquid stool, gas incontinence and the frequency of using a pad due to fecal incontinence. The overall Wexner score ranged from 0 to 20, and a higher score indicates greater severity of fecal incontinence [24].

FSFI and IIEF-5
HRQoL problems related to sexual dysfunction were evaluated using the FSFI for female patients and the IIEF-5 for male patients. FSFI is a self-report scale that includes 6 domains (desire, arousal, lubrication, orgasm, satisfaction and pain). The separate scores for each domain are calculated as a total score [25].

Statistical analysis
Continuous variables are presented as the means (range) or proportions as necessary. Differences in continuous variables, such as median LARS score, were examined using the Mann-Whitney U test.
Categorical variables in this study were assessed using Pearson's chi-squared test or Fisher's exact test (if applicable). Criteria for signi cant differences was a p value < 0.050. A p value < 0.001 was de ned as an extremely distinct difference. The Kaplan-Meier method was used to compare OS and DFS. The log-rank test was used to distinguish differences in oncological outcome. Statistical analyses were performed using SPSS version 24.0 for Windows (SPSS Inc., Chicago, IL, USA).

Patients' Characteristics
Of the 153 patients who met the inclusive criteria, 10 patients in the adjuvant treatment group and 13 patients in the TME group were excluded due to various factors, including barrier in the comprehension of the scales provided (n = 12), dementia or other diseases prevented consent (n = 3), and functional outcome scales feedback missing (n = 8). The remaining 130 patients were enrolled and matched one-toone using the nearest neighbor method. Tumor diameter, tumor height, cT classi cation, pT classi cation and age were considered. The proper math tolerance was calculated as 0.05 in propensity score matching, which resulted in 110 patients included in the study (n = 55 for each group). The entire inclusion and exclusion ow chart is shown in Figure 1. Patients' demographic, tumor characteristics and treatment details are listed in Table 1.

Oncological outcome
Of the 110 enrolled patients, 12 patients (10.91%) developed recurrence after initial TEM surgery, including 8 patients (7.27%) of local recurrence and 4 patients (3.64%) of distant recurrence ( Table 2). The local recurrence rate of patients undergoing adjuvant treatment was higher than patients who underwent TME (9.09% vs. 5.45%), but this difference was not signi cant (p = 0.463). Four patients suffered distant metastases, with 2 patients in each group. The average interval from primary TEM surgery was 34.25 (24 -79) months. The distant metastasis sites primarily included liver, lung, bone and distant lymph nodes. The Kaplan-Meier survival curve showed no signi cant differences in 5-year OS and DFS between the two treatments after primary TEM for intermediate risk early rectal cancer patients (Figure 2a and 2b).
The univariable analysis revealed that pT2, large tumor diameter, positive lymphovascular invasion and poor differentiation were the risk factors predicting recurrence ( Table 3). The subsequent multivariable analysis indicated that only pT2 was an independent risk factor for DFS (HR 3.09, 95% CI 1.66-4.18, p = 0.044) (Figure 2c).

Functional outcomes
The study reclaimed 105 effective questionnaires (excluding patients lost to follow up and death), including 52 (94.5%) from the adjuvant treatment group and 53 (96.4%) from the TME group.

EORTC QLQ-CR29
Compared with patients in the adjuvant treatment group, patients who underwent TME suffered from buttock pain more seriously (17.99 vs. 4.67, p = 0.030). The adjuvant treatment group complained more about taste problems (7.91 vs. 1.73, p < 0.001). Other functional and symptoms scales, including body image and sexual interest, were similar between the two groups without signi cant differences (Table 4).

LARS score and Wexner score
Patients' LARS and Wexner scores are shown in Table 5. The anorectal function in both groups was comparable before TEM. The TME group had a higher frequency of bowel urgency (p < 0.001). Categorization of the severity of LARS in the two groups and median total LARS scores also showed signi cant differences (p = 0.039 and 0.040, respectively). However, there was no signi cant difference between the 4 other domains of the LARS score related to atus or liquid stool control between the two groups (p > 0.05). The results of Wexner scores between the two groups were not signi cantly different either (p = 0.173).

FSFI and IIEF-5
Based on the preoperative data, the sexual function of all enrolled patients was not different between the two groups. Lubrication disorder of the vaginal canal was signi cantly severe in female patients who underwent TME surgery (7.52 vs. 2.03, p < 0.001), but other indexes of FSFI and total FSFI score were comparable to the adjuvant treatment group patients (Figure 3). Postoperative erectile function evaluation revealed that males in the radical TME group experienced a higher proportion of moderate and severe ED (20.69% vs. 11.76%), but without a signi cant difference. Patients in both groups had a similar overall IIEF-5 score, which re ected parallel erectile function (Table 6).

Discussion
According to previous reports, approximately 75% of the patients who suffered recurrence after local excision of rectal cancer were intermediate risk pT1 or pT2 stage [27][28]. The choice between postoperative radical surgery and adjuvant treatment was always a barrier for surgeons and oncologists to implement the most appropriate treatment option. TEM is a relatively novel local excision surgical method to perform en-bloc resection through a speci c working tunnel [10,12,14]. We performed this study to compare the oncological and functional outcomes in a series of intermediate-risk early rectal cancer patients who received adjuvant treatment or TME after TEM. Overall, each option had its pros and cons in different aspects.
Nearly 11% of the enrolled patients developed local or distant recurrence after initial TEM resection. The overall recurrence rate was consistent with previous reports [29]. The present study found no signi cant differences in 5-year OS and DFS between the different post-TEM treatment methods. Pathological T stage was a predisposing factor for recurrence. The present study found that although several pathological characteristics, such as tumor diameter and lymphovascular invasion, were high risk factors for a low DFS rate, only pT stage was an independent predicting factor for low DFS rate in multivariable analysis. This result suggests that patients who were diagnosed as pT2 after TEM should be in a closer out-patient follow-up regardless of post-TEM therapy. The results were in great accordance with a systematic review done by Cutting et al., which maintained that local excision plus adjuvant therapy should not be recommended for pT2 patients [30].
We clearly observed that patients who underwent TME suffered from a greater proportion of buttock pain, which is consistent with some related work on comparisons of local excision and TME. Taste complaint was more obvious in the adjuvant treatment group patients, but this nding may be somewhat explained by the chemotherapy module bias after TEM. No other symptom or functional scales showed signi cant differences. There is no denying that adjuvant treatment may produce a series of side effects, such as perforation, diarrhea or proctitis, especially in patients who choose radiotherapy [31][32]. Borstlap et al. and Smart et al. reported the complication rate ranged from approximately 6-52% [17,31]. The 2 patients who underwent CCRT in our study reported no radiotherapy-related side effects. However, this result may not represent the low risk of side effects of adjuvant treatment due to the small sample size.
TEM is an advisable method for the treatment of early rectal cancer partially because of its unique transanal working pattern. Several studies demonstrated that TEM had little effect on patients' HRQoL after surgery. Biviano et al. enrolled 66 rectal cancer patients and concluded that TEM had no signi cant effect on anal function [33]. Perioperative damage to the anal sphincter caused by TME is consistently a concern for surgeons and patients. Anorectal function was evaluated using 2 validated questionnaires, and great disparity occurred between the groups. Patients in the TME group demonstrated worse bowel function, particularly bowel urgency and stool control. More patients in the TME group reported major LARS, which indicates a higher level of bowel dysfunction. These results corroborate the ndings of many previous works. Chen et al. found that nearly half of the patients who underwent TME suffered from major LARS [34]. Bohlok et al. (2019) also showed that a large proportion (72.1%) of patients after TME complained of LARS to different extents [35]. A meta-analysis done by Kidane et al. also proposed the viewpoint that TEM could be adopted as a reliable device which provides similar oncological control but lower mortality rate [36]. One unexpected nding was the extent to which the overall Wexner score between the 2 groups was not signi cantly different. This result may be partially due to the nonspeci city of the Wexner scale and distortion of the included questions, especially after lling out the LARS questionnaire. Cheong et al. performed a prospective observational study and revealed severe bowel dysfunction caused by anterior resection for rectal cancer. The dysfunction partially originated from the procedure of coloanal anastomosis [37]. Sclafani et al. concluded that CRT did not increase the signi cant detrimental effect on HRQoL and bowel function in rectal cancer patients [38]. Therefore, the potential bene t of adjuvant therapy in the lower rate of LARS should not be underestimated. The followup time was another vital factor for postoperative LARS based on some published reports. Therefore, our data should be interpreted with caution because the follow-up months for each patient were diverse [33,39].
Latent damage to the pelvic nerve during TME is a controversial side-effect of postoperative sexual function preservation. Published reports on the differences in sexual function between adjuvant CRT and radical surgery were limited [40]. We employed different gender-speci c scales for patients' sexual functions and clearly found no obvious differences between the treatment of options. Except for a higher score of vaginal lubrication in the adjuvant treatment group, no other scales in the FSFI were different between the two groups. Male patients' erectile function is a big concern for surgeons, but there was no signi cant difference on median overall IIEF-5 score between the treatments. The ratio of moderate and severe ED patients was higher in the TME group than the adjuvant treatment group (20.69% vs. 11.76%), but no signi cant difference was found. These results may be biased to some extent due to the selfreported nature of the scales and the small number of patients who underwent adjuvant therapy when diagnosed with N2 stage after TME.
Our study has several limitations, such as its retrospective design and relatively short following-up period. Except for these points, the self-reported questionnaires we used in this study were more or less subjective. Some examinations, such as anal manometry, should be used to increase the objective evidence of bowel function. The same surgeon performed most of the TEM in our unit after standard training. However, the learning curve may also cause bias of the pathological results for the earliest cases.

Conclusion
All in all, compared with TME, adjuvant therapy poses a similar oncological outcome with no inferiority in postoperative overall or disease-free survival for intermediate risk early rectal cancer patients after primary TEM resection. Pathological T stage may be an independent risk factor for recurrence regardless of treatment. TEM with adjuvant therapy has less impact on patients' HRQoL with better bowel function and similar levels of sexual function.

Declarations
Ethics approval and consent to participate: Our protocol has been approved and got ethics approval and consent from the IRB of Peking Union Medical College Hospital (S-K1237).

Consent for publication:
This article doesn't include any individual person's data.

Availability of data and materials:
The data that support the ndings of this study are available on request from the corresponding author. The data are not publicly available due to their containing information such as sex function that could compromise the privacy of research participants.
Competing interests: All the authors declare that they have no con ict of interest.    LARS, low anterior syndrome; Scores 0 to 20 (no LARS), 21 to 29 (minor LARS), 30 to 42 (major LARS)