Laparoscopic Abdominal Transanal Proctocolectomy with Coloanal Anastomosis Is a Good Surgical Option in Selective Patients with Low-Lying Rectal Cancer: A Retrospective Analysis Based on a Single Surgeon's Experience

Abstract

Purpose:

We intended to identify the oncological outcome for rectal cancer over the past 20 years and whether or not sphincter-preserving surgery is an appropriate approach for low-lying rectal cancer.

Materials and Methods:

The oncological outcomes from a total of 418 patients who electively underwent rectal cancer surgery for a lesion located within 8 cm of the anal verge by a single colorectal surgeon were reviewed retrospectively.

Results:

Of 418 patients, 175 patients underwent low anterior resection (LAR), 172 laparoscopic abdominal transanal proctocolectomy with coloanal anastomosis (LATA), and 71 abdominoperineal resection (APR). The factors related to the disease-free survival (DFS) were neoadjuvant chemoradiation (P = .016), pathologic stage (P < .001), circumferential margin involvement (P < .001), and initial (P = .016) and postoperative serum carcinoembryonic antigen level (P < .001). The factors related to the overall survival (OS) were similar with those related to DFS. Compared with DFS, OS, and local recurrence among three surgical techniques, APR was significantly poorer in DFS (P < .001), OS (P < .001), and local recurrence (P = .001), than was LAR or LATA.

Discussion:

The LATA procedure is a technically feasible and oncologically safe surgical option for low-lying rectal cancer. We suggest that LATA may be a good surgical option in selective low-lying rectal cancer patients.

Introduction

The treatment of rectal cancer has developed through the expertise of surgeons and a team composed of specialists from several disciplines (e.g., surgery, medical oncology, radiotherapy, endoscopy, and pathology). Of the various treatment modalities, curative surgery is a mainstay of treatment for rectal cancer. Miles demonstrated abdominoperineal resection (APR) for rectal cancer in 1908. He insisted that the perineal portion of the operation should be carried out as widely as possible so that the lateral and downward zones of spread may be effectively extirpated.¹ Until the 1980s, APR was the elective procedure for rectal cancer located less than 6 cm from the anal verge.² However, the introduction of total mesorectal excision, the concept of downsizing and downstaging from neoadjuvant chemoradiation therapy (nCRT), and the importance of circumferential radial margin rather than distal resection margin led to the increasing application of sphincter-preserving surgery.^3–5 Recently, as the ultimate evolution of sphincter-preserving surgery, the intersphincteric resection has been proposed for tumors located at the level of the anorectal ring or within the anal canal.^6,7

Laparoscopic surgery for low rectal cancer had had safety concerns, especially regarding long-term oncological outcome. However, recent trials and meta-analysis demonstrated that laparoscopic surgery for low rectal cancer is safe and feasible with regard to long-term oncological outcomes as well as short-term outcomes.^8–10

We started laparoscopic rectal cancer surgery in 1994, and first performed laparoscopic abdominal transanal proctocolectomy with coloanal anastomosis (LATA) in February 1996. Transanal abdominal transanal proctosigmoidectomy and coloanal anastomosis (TATA) was developed by Gerald Marks at the Thomas Jefferson University in 1984.¹¹ In this report, we refer to the laparoscopic TATA as “LATA.”

In this study, we intended to identify the oncological outcome for rectal cancer over the past 20 years, and determine whether laparoscopic sphincter-preserving surgery for rectal cancer is an appropriate approach for low-lying rectal cancer.

Materials and Methods

Patient enrollment

We retrospectively collected data for 695 patients who underwent curative resection for rectal cancer within 12 cm of the anal verge by a single colorectal surgeon between January 1994 and May 2011. Among these patients, patients who underwent transanal local excision for T1 rectal cancer, total or subtotal proctocolectomy, or emergency operation, including the Hartmann procedure, and patients with rectal cancer located between 8 and 12 cm from anal verge, were excluded from this study. A total of 418 patients who electively underwent rectal cancer surgery for a lesion located within 8 cm from the anal verge were enrolled in this study.

Ethics

After obtaining review board approval from The Catholic University of Korea, CMC Clinical Research Coordination Center (XC16RIMI00692), we analyzed the data and clinical information of these 418 patients.

Definition and procedures

The rectum may be conveniently divided into thirds, with its most proximal end at a variable level, usually several centimeters above the peritoneal reflection, and its most distal end near the dentate line. The lower rectum occupies a space between 0 and 5 to 6 cm proximal to the dentate line. The internal anal sphincter represents a thickened continuation of the rectal circular smooth muscle that is involuntary and ∼2.5 cm in length.¹² Therefore, we considered any lesion confirmed with adenocarcinoma as low-lying rectal cancer when it was within 8 cm of the anal verge. We measured the tumor location for all patients with rigid proctosigmoidoscopy at the time of initial diagnosis. For all enrolled patients, we performed a total mesorectal excision. During total mesorectal excision, the mesorectum was mobilized up to the level of the puborectalis muscle. If 1 cm or greater distal resection margin from the tumor was attainable by using a linear stapler, a double stapling technique was applied to perform a low anterior resection (LAR). However, if this was not possible, a laparoscopic abdominal procedure was performed, followed by a transanal approach for coloanal anastomosis, after changing the patient to the lithotomy position. In this technique, a circumferential incision was made immediately superior to the dentate line to extract the mobilized colorectum through the anus. After the proximal resection margin had been determined, it was resected, and a hand-sewn coloanal anastomosis was then performed. We defined the case of anastomosis done by transanal hand-sewn technique as the LATA (although LATA was included in TATA, we named this approach LATA because almost all TATAs, except two cases, were performed by the laparoscopic approach). A diverting ileostomy or a colostomy was performed for patients who had received preoperative radiotherapy and for those with an anastomotic line within 5 cm of the anal verge. Otherwise, APR was performed for rectal cancers located within 1 cm of the anal verge on preoperative proctosigmoidoscopy or with threatened circumferential resection margin (including uncertainty of tumor invasion to the sphincter muscle or pelvic floor) on preoperative imaging study or intraoperative finding. Our intention for radical rectal surgery was to secure a safe circumferential resection margin and a distal resection margin. We performed LATA only in patients who had been expected to have a secure grossly negative circumferential resection margin and more than 1 cm of distal resection margin in the operation room.

Staging work-up

In patients who had biopsy-proven adenocarcinoma in the low-lying rectum, colonoscopy to search for a synchronous lesion and rigid proctosigmoidoscopy to measure the length between the lesion and the anal verge were performed. For local staging, abdomen and pelvic computed tomography (CT), transanal ultrasound, and/or rectal magnetic resonance imaging (MRI) were used. In addition, abdomen and pelvic CT and chest CT or positron emission tomography-computed tomography (PET-CT) scans were obtained for staging work-up to search for distant metastatic lesions, and serum carcinoembryonic antigen (CEA) levels were assayed after the lesion was confirmed to be an adenocarcinoma. In patients who underwent nCRT, the aforementioned imaging studies for staging work-up were preoperatively repeated at 4 weeks after the end of nCRT.

Chemoradiation therapy

Although the decision whether patients were treated by radical surgery following nCRT or radical surgery alone was dependent on the surgeon, and most patients who had clinical T3–T4 or N+ rectal cancer received nCRT with conventional fractionation as follows: 1.8 Gy per day; five fractions per week; and a total dose of 50.4 Gy/28 fractions (45 Gy/25 fractions initially to the whole pelvis, followed by 5.4 Gy/3 fractions as a boost to the gross tumor). All of the patients received two cycles of concurrent chemotherapy with radiotherapy [5-fluorouracil (5-FU), 400 mg m² (IV) 1 h before radiotherapy and leucovorin, 20 mg m² (IV) immediately before each dose of 5-FU on days 1–5 and 29–33). Surgical treatment was performed within 6–8 weeks after the end of nCRT.

We omitted nCRT for some relatively young female patients who had a plan for child bearing and in cases that the surgeon was sure to secure the safe circumferential resection margin and distal resection margin without nCRT. The pathologic T3–T4 or N (+) tumors, which were not treated with nCRT, we tried to treat them with postoperative chemoradiation therapy.

Follow-up

For all patients, follow-up data were obtained during routine clinical practice. Using abdomen and pelvic CT and chest CT or plain chest X-ray, patients had been examined every 3 months during the first 2 years, and then every 6 months for the rest of the 3- to 5-year follow-up schedule. The cause and the date of death were obtained after the examination of medical records.

Primary endpoint

The primary endpoint was overall survival (OS), disease-free survival (DFS), and local recurrence rate according to three operative modalities.

Secondary endpoint

The short-term perioperative outcomes, including overall postoperative morbidity and recovery course after surgery were analyzed in the overall study population.

Data collection

For evaluating patients' preoperative condition, we analyzed sex, age, American Society of Anesthesiologist (ASA) score, body mass index, and serum preoperative CEA. Intraoperative parameters were analyzed with operative method (open or laparoscopy), operation time, intraoperative blood loss, and whether or not there were intraoperative complications. For postoperative short-term outcome, we compared the starting day of diet, postoperative hospital stay, and the severity of complications using the Clavien–Dindo classification among three surgical groups. We investigated the long-term oncological outcome with various factors, including pathologic findings.

Statistical analyses

Continuous variables were compared using Student's t-test and one-way ANOVA, and expressed as the mean ± standard deviation. Categorical variables were analyzed with the χ² test. The survival probability analysis was performed using the Kaplan–Meier method. The log-rank test was used to assess the difference of survival between strata. Significance was defined as a P value ≤.05. Multivariate analysis was applied with Cox's proportional hazard regression model. By using forward stepwise selection, independent factors were analyzed, and the statistically significant entry and staying values were set at 0.05. All statistical analyses were performed using the Statistical Package of the Social Sciences (SPSS) version 12.0 for Windows (SPSS, Inc., Chicago, IL).

Results

There were 231 men and 187 women in the present study. The patients' mean age was 60.3 ± 10.7 years. Table 1 shows the patients' demographic findings according to three surgical techniques. LAR was more frequently performed in female patients than in male patients (P < .001). The location of the tumor in the APR group was lower than those in the LAR or the LATA group (P < .001). Also, nCRT was more frequently performed in the LATA or the APR group than in the LAR group (P < .001).

Table 1.

Demographics of All Enrolled Patients

	LAR (N = 175)	LATA (N = 172)	APR (N = 71)	P
Age
<65 years, n (%)	113 (64.6)	109 (63.4)	44 (62.0)
≥65 years, n (%)	62 (35.4)	63 (36.6)	27 (38.0)	.925
Mean ± SD	59.9 ± 11.2	60.9 ± 10.6	59.9 ± 10.4	.623
Sex, n (%)
Male	72 (41.1)	118 (68.6)	41 (57.7)
Female	103 (58.9)	54 (31.4)	30 (42.3)	<.001
BMI
≤18.5 kg/m², n (%)	6 (3.5)	9 (5.2)	4 (5.8)
>18.5 kg/m²	167 (96.5)	163 (94.8)	65 (94.2)	.641
Mean ± SD	23.6 ± 3.3	23.8 ± 3.2	23.6 ± 3.2	.683
ASA, n (%)
Score 0	0	1 (0.6)	0
1	107 (64.5)	97 (57.1)	44 (68.8)
2	49 (29.5)	63 (37.1)	17 (26.6)
3	10 (6.0)	9 (5.3)	3 (4.7)	.535
Location from AV (cm) mean ± SD	6.61 ± 1.3	5.02 ± 1.5	3.40 ± 1.5	<.001
nCRT, n (%)
No	72 (41.1)	14 (8.1)	12 (16.9)
Yes	103 (58.9)	158 (91.9)	59 (83.1)	<.001
Initial CEA, n (%)
≤5 ng/mL	101 (66.4)	111 (70.3)	39 (61.9)
>5 ng/mL	51 (33.6)	47 (29.7)	24 (38.1)	.470

LAR, low anterior resection; LATA, laparoscopic abdominal transanal proctocolectomy with coloanal anastomosis; APR, abdominoperineal resection with permanent colostomy; BMI, body mass index; ASA, American Society of Anesthesiologists; AV, anal verge; nCRT, neoadjuvant chemoradiation therapy; CEA, carcinoembryonic antigen; SD, standard deviation.

We performed rectal cancer surgeries using the laparoscopic approach in most patients regardless of operation techniques. There was no statistically significant difference in intraoperative and postoperative complication rate, reoperation rate, or urination problems among the three surgical groups. Postoperative hospital stay was longer in the APR group than in the LAR and the LATA groups (20.23 ± 23.04 days versus 14.38 ± 7.52 and 14.23 ± 7.74 days, P = .001) (Table 2). Table 3 shows the list of early postoperative and late complications. The postoperative complications within 30 days after surgery occurred in 43 (24.6%) patients in the LAR group, 32 (18.6%) in the LATA group, and 18 (25.4%) in the APR group. Of these, the most common postoperative complication in all three operation groups was postoperative ileus in 19 (10.9%) patients in the LAR group, 21 (12.2%) in the LATA group, and 6 (13.0%) in the APR group. Especially, surgical site infections more frequently occurred in the APR (7 patients, 9.9%) group than in the LAR (3 patients, 1.7%) or the LATA (1 patient, 0.6%) group (P = .016). The late complications that occurred after postoperative 30 days were rectovaginal fistula (3 patients in the LAR group and 1 patient in the LATA group), anastomosis site stricture (4 patients each in the LAR and the LATA groups), and stoma problem (2 patients each in the LATA and the APR groups). In the present study, we always performed the diverting stoma during the LATA procedure. In all 152 patients in the LATA group, we were able to check if their diverting stomas were reversed or not. Seven (4.6%) of 152 patients did not undergo the operation for the reversal of their diverting stoma for following reasons: five patients refused the operation for the reversal of diverting stoma; one patient who underwent anastomosis site stricture refused any other procedure; and another patient committed suicide due to neuropsychiatric problems before the reversal of the diverting stoma.

Table 2.

Perioperative Short-Term Outcomes and Postoperative Findings

	LAR (N = 175)	LATA (N = 172)	APR (N = 71)	P
Approach, n (%)
Laparoscopy	170 (97.1)	165 (95.9)	66 (93.0)
Conversion	3 (1.7)	5 (2.9)	2 (2.8)	.407
Open	2 (1.1)	2 (1.2)	3 (4.2)
Intraoperative complication, n (%)
Yes	23 (14.2)	29 (17.3)	14 (22.6)
No	139 (85.8)	139 (82.7)	48 (77.4)	.318
Postoperative complication, n (%)
Yes	43 (24.7)	30 (17.6)	19 (26.8)
No	131 (75.3)	140 (82.4)	52 (73.2)	.171
Reoperation (+), n (%)	10 (5.7)	3 (1.7)	2 (2.8)	.129
Postoperative urinary problem, n (%)
Yes	9 (7.2)	12 (8.2)	7 (13.2)
No	116 (92.8)	135 (91.8)	46 (86.8)	.412
ICU stay (days)	1.80 ± 1.45	2.26 ± 1.91	2.65 ± 2.11	.002
Postoperative Hospital stay
Days	14.38 ± 7.52	14.23 ± 7.74	20.23 ± 23.04	.001
Range	7–66	7–81	9–90
Soft diet (POD)	6.76 ± 3.61	6.81 ± 3.69	6.55 ± 2.46	.869
Postop CEA, n (%)
≤5 ng/mL	157 (94.0)	157 (94.6)	50 (76.9)
>5 ng/mL	10 (6.0)	9 (5.4)	15 (23.1)	<.001

ICU, intensive care unit; POD, postoperative day.

Table 3.

Postoperative Complications in Detail

	LAR (N = 175), n (%)	LATA (N = 172), n (%)	APR (N = 71), n (%)	P
Perioperative complications
Postoperative ileus	19 (10.9)	21 (12.2)	6 (13.0)	—
Anastomosis leakage	5 (2.9)	2 (1.2)	0	—
Postoperative bleeding	4 (2.3)	1 (0.6)	0	—
Surgical site infection	3 (1.7)	1 (0.6)	7 (9.9)	.016
Chylous ascites	7 (4.0)	4 (2.3)	2 (2.8)	—
Other	5 (2.9)	3 (1.7)	3 (4.2)	—
C-D classification ≥3	12 (6.9)	5 (2.9)	1 (1.4)	.081
Late complications	7 (4.0)	7 (4.1)	2 (2.8)	.888
Rectovaginal fistula	3 (1.7)	1 (0.6)	0
Anastomosis stricture	4 (2.3)	4 (2.3)	0
Stoma problem	0	2 (1.2)	2 (2.8)

C-D, Clavien–Dindo.

We found that the distal resection margin was shorter in the LATA group than in the LAR or APR group (P < .001). There were 11 cases of R1 resection (they confirmed circumferential resection margin (CRM) involvement by pathological examination) and no R2 resection. Other pathological findings were not significant differences among the three groups (Table 4).

Table 4.

The Comparison of Pathological Result Among Three Surgical Techniques

	LAR (N = 175)	LATA (N = 172)	APR (N = 71)	P
Stage, n (%)
I	78 (48.8)	77 (47.0)	22 (33.3)
II	51 (31.9)	50 (30.5)	24 (36.4)
III	31 (19.4)	37 (22.6)	20 (30.3)	.233
T stage, n (%)
0–1	36 (22.5)	35 (21.3)	7 (10.6)
2	48 (30.0)	50 (30.5)	19 (28.8)
3	70 (43.8)	71 (43.3)	33 (50.0)
4	6 (3.8)	8 (4.9)	7 (10.6)	.214
N stage, n (%)
0	123 (78.3)	122 (77.2)	46 (70.8)
1 or 2	34 (21.7)	36 (22.8)	19 (29.2)	.465
PRM
<10 cm, n (%)	24 (15.7)	18 (11.3)	14 (22.6)
≥10 cm, n (%)	129 (84.3)	141 (88.7)	48 (77.4)	.103
Mean ± SD	16.93 ± 10.30	18.38 ± 8.03	15.10 ± 6.56	.041
DRM
<1 cm, n (%)	7 (4.1)	31 (18.5)	3 (7.3)
≥1 cm, n (%)	163 (95.9)	137 (81.5)	64 (95.5)	<.001
Mean ± SD	3.53 ± 2.04	2.28 ± 1.61	3.44 ± 1.77	<.001
CRM involvement, n (%)
(−)	174 (99.4)	165 (95.9)	67 (95.7)
(+)	1 (0.6)	7 (4.1)	3 (4.3)	.081
Differentiation, n (%)
Well	43 (28.3)	47 (29.9)	16 (24.2)
Moderately	102 (67.1)	103 (65.6)	45 (68.2)
Poorly	7 (4.6)	7 (4.5)	5 (7.6)	.811
Lymphatic invasion, n (%)
Yes	23 (15.9)	19 (13.0)	14 (25.0)
No	122 (84.1)	127 (87.0)	42 (75.0)	.116
Venous invasion, n (%)
Yes	6 (4.1)	4 (2.7)	5 (8.6)
No	140 (95.9)	142 (95.7)	53 (91.4)	.172
Perineural invasion, n (%)
yes	18 (12.4)	15 (10.3)	15 (25.4)
No	127 (87.6)	131 (89.7)	44 (74.6)	.014

PRM, proximal resection margin; DRM, distal resection margin; CRM, circumferential resection margin.

The mean DFS and OS for all enrolled patients in the present study were 105.85 ± 2.86 and 109.49 ± 2.82 months. The 3-year local recurrence rate for all enrolled patients was 5.3%. Table 5 shows the results of the oncological outcomes by univariate analysis according to various factors. The factors related to DFS were nCRT (P = .016), overall stage (P < .001), T stage (P < .001), node metastasis (P < .001), circumferential resection margin involvement (P < .001), lymphatic invasion (P < .001), vascular invasion (P = .049), perineural invasion (P < .001), initial serum CEA level (P = .016), and postoperative serum CEA level (P < .001). The factors related to OS were age (P = .024), ASA classification (P = .027), and proximal resection margin (P = .0310), as well as the factors related to DFS. Also, the factors related to local recurrence were overall stage (P < .001), T and N stage (P = .001 and P < .001), circumferential resection margin involvement (P < .001), lymphatic invasion (P < .001), perineural invasion (P < .001), and postoperative serum CEA level (P = .046). The surgical technique was a statistically significant factor in DFS (P < .001), OS (P < .001), and local recurrence (P = .001) (Fig. 1). In multivariate analysis, surgical technique (P = .043 and P = .011), circumferential resection margin involvement (P = .009 and P = .002), and postoperative serum CEA level (P = .003 and P < .001) were independent factors in DFS and OS; only circumferential resection margin involvement (P = .004) was an independent factor in local recurrence (Table 6).

FIG. 1.

This figure shows the oncological outcome according to three surgical techniques; (A) for DFS, (B) for OS, and (C) for local recurrence. (C) When we compared the local recurrence among each surgical groups, there was no significant difference in local recurrence between LAR group and LATA group (P = .156), but significantly better in local recurrence of LATA group than that of APR group (P = .0013). APR, abdominoperineal resection; DFS, disease-free survival; LATA, laparoscopic transabdominal transanal proctocolectomy with coloanal anastomosis; LAR, low anterior resection; OS, overall survival.

Table 5.

The Results of Oncological Outcomes Through the Univariate Analysis

	5 years DFS rate (%)	P	5 years OS rate (%)	P	3 years LR rate (%)	P
Sex
Male	71.2		79.9		5.5
Female	77.0	.186	78.7	.991	5.1	.760
Age
<65 years	72.7		82.5		5.3
≥65 years	75.5	.331	73.2	.024	5.3	.783
ASA
I	73.7		83.1		4.0
II	76.4		75.2		6.6
III	63.9	.375	63.0	.027	11.9	.572
nCRT
No	83.7		85.0		2.3
Yes	70.8	.016	77.5	.354	6.3	.434
Intraoperative complication
No	73.0		79.2		5.7
Yes	80.0	.286	78.8	.683	5.8	.872
Postoperative Complication
No	76.2		80.7		5.6
Yes	64.8	.055	73.7	.504	4.4	.823
Anastomosis leak
No	73.3		79.6		5.4
Yes	100	.202	85.7	.905	0	.544
Stage
0	100		100		0
I	88.1		86.5		2.2
II	69.8		78.7		5
III	50.1	<.001	64.3	.002	14.5	<.001
T stage
0–1	93.9		87.5		0
2	82.7		84.4		4.2
3	60.6		73.5		7.2
4	54.5	<.001	63.5	.036	25.4	.001
N stage
N0	79.6		83.0		3.5
N(+)	51.7	<.001	65.4	.003	14.1	<.001
PRM
<10 cm	77.4		72.0		5.3
≥10 cm	72.9	.902	80.7	.031	5.4	.739
DRM
<1 cm	76.2		79.8		12.3
≥1 cm	76.2	.975	79.1	.912	4.4	.126
CRM involvement
(+)	36.4		22.7		43.2
(−)	75.1	<.001	80.9	<.001	4.3	<.001
Differentiation
Well	75.5		81.8		3.6
Moderately	72.0		77.2		7.3
Poorly	62.3	.301	71.3	.631	0	.377
Lymphatic invasion
(+)	43.6		52.7		27.8
(−)	78.1	<.001	84.9	<.001	3.0	<.001
Venous invasion
(+)	32.2		54.5		0
(−)	73.2	.049	79.9	.014	6.5	.477
Perineural invasion
(+)	52.8		56.6		16.6
(−)	75.0	<.001	82.2	<.001	4.9	<.001
Initial CEA
≤5 ng/mL	78.9		85.3		3.9
>5 ng/mL	64.5	.016	70.0	.006	6.3	.155
Postop CEA
≤5 ng/mL	77.1		85.4		4.9
>5 ng/mL	39.1	<.001	32.6	<.001	9.8	.046
Surgical techniques
LAR	77.6		85.7		2.0
LATA	79.9		83.8		5.6
APR	49.3	<.001	54.0	<.001	11.2	.001

DFS, disease-free survival; OS, overall survival; LR, local recurrence.

Table 6.

The Multivariate Analysis for Oncological Outcomes

	DFS			OS			LR
	HR	P	95% CI	HR	P	95% CI	HR	P	95% CI
Age	0.805	.370	0.434–1.364	1.304	.253	0.759–2.849	0.430	.512	0.158–2.512
nCRT	1.387	.239	0.288–1.364	0.079	.779	0.401–1.985	0.179	.672	0.286–6.962
Surgical techniques	6.279	.043	1.187–4.141	6.414	.011	1.213–4.555	3.520	.313	0.546–6.600
TNM stage	0.802	.670	0.091–2.500	1.211	.271	0.066–2.150	0.707	.400	0.006–7.920
T stage	2.244	.134	0.721–11.584	2.488	.115	0.761–12.453	1.561	.211	0.022–22.467
N stage	0.032	.857	0.168–4.407	0.085	.770	0.026–7.455	0.041	.839	0.022–22.467
CRM	6.796	.009	0.052–0.657	9.429	.002	0.032–0.466	8.293	.004	0.005–0.368
Lymphatic invasion	1.219	.270	0.386–1.305	0.323	.570	0.393–1.672	3.247	.072	0.097–1.103
Vascular invasion	0.141	.708	0.277–2.391	0.833	.558	0.205–2.355	0.001	.975	0.000
Neural invasion	0.155	.694	0.588–2.220	0.387	.362	0.331–1.497	1.473	.225	0.129–1.618
Initial CEA	0.237	.626	0.654–2.026	0.387	.534	0.608–2.613	0.155	.694	0.345–4.957
Postop CEA	8.714	.003	0.129–0.662	12.241	<.001	0.072–0.476	0.599	.439	0.056–3.502

Bold indicates statistical significance.

HR, hazard ratio.

Discussion

Early efforts at sphincter preservation for rectal cancer were followed by an inordinately high incidence of local failure. Thus, the APR became the standard procedure for rectal cancer treatment from the time it was first described in the literature in 1908.¹ Although many surgeons had increased their concerns about sphincter preservation in rectal cancer treatment, most of them have performed APR for rectal cancer, especially low-lying rectal cancer, because of the serious consequences of the failed low anastomosis and unestablished safety for oncological outcome.¹¹ Steele et al.¹³ identified several major questions that had to be solved for acceptance of the sphincter-saving procedure for rectal cancer: spread pattern of lymph node, randomization, additional therapy to sterilize possible microscopic disease, and the role of salvage surgery. For solving these questions, they commented that the combination of adequate local surgery, external radiation, and chemotherapy would seem the most compelling. Since the preoperative radiation and sphincter-preserving surgical techniques were initiated in 1976 at the Thomas Jefferson University Hospital, they began the TATA procedure in 1984 with indications such as nonfixed tumors in the distal 3 cm of the rectum.¹¹ Recently, a multidisciplinary approach, including pre- or postoperative radiotherapy, chemotherapy, and surgery, has been improved and widely adopted for the treatment of rectal cancer, especially low-lying advanced rectal cancer.¹⁴ Advances in surgical technique have made the intersphincteric resection for tumors located in close proximity to the anorectal ring and sphincter complex a method to extend the opportunity for sphincter preservation to patients with very low rectal cancers.¹⁵

We reported our experiences for laparoscopic rectal cancer surgery in 2008.¹⁶ At that time, we reported that laparoscopic surgery did not adversely affect the long-term oncological outcome for patients with rectal cancer. Since 1994, most of the patients with rectal cancer underwent laparoscopic surgery by the corresponding author. Laparoscopic surgery enables the surgeon to visualize directly the narrow pelvic cavity and to perform accurate and sharp dissections, while the magnified vision clearly delineates the anatomy, thus permitting a more precise total mesorectal excision.¹⁷ This may be the reason that the laparoscopic surgery contributes to comparable or better oncological outcomes for rectal cancer.

In the present study, the overall intraoperative or postoperative morbidity was not different among the three groups. However, surgical site infection was more frequently observed in the APR group than in the LAR or the LATA group (P = .016). In our study, we had performed surgery mostly by laparoscopic approach, even in APR. One of obvious advantages of laparoscopic rectal surgery compared with open surgery is minimizing surgical site infection.¹⁸ Especially, only four or five small wounds within 1 cm on the abdomen are needed for trocar placement in the LATA procedure, in which the specimen is removed through the anus without any other wound for specimen extraction. On the other hand, a perineal wound is inevitable in APR. In the present study, all of the surgical site infections in the APR group were perineal wound infections. Because many patients in the APR group received nCRT before surgery, it seemed to be harder to treat perineal wound problems. Also, these wound problems made the postoperative hospital stays longer in the APR group.

The feasibility of sphincter preservation in low-lying rectal cancer depends on the ability to obtain an adequate distal resection margin above the sphincter, which accounts for the gross tumor as well as the area of potential microscopic distal spread that might lead to a positive distal resection margin on histopathology. In 1913, the 5-cm rule for rectal cancer surgery was first suggested to limit the intramural spread of rectal cancer.¹⁹ Since then, there have been some energetic debates about distal resection margins in rectal cancer surgery. In late 1980, one clinical trial demonstrated that margins of less than 2 cm, determined in vivo, appreciably increase the risk of local failure.²⁰ Recently, a 1-cm rule was used in patients with low-lying lesions when sphincter preservation is an issue. The 1-cm rule is based mainly on pathological data that demonstrate distal intramural spread greater than 1 cm in a substantial proportion of patients.²¹ Some authors even insisted that when total mesorectal excision is combined with radiotherapy, local control is not compromised in patients with distal resection margin of less than 1 cm in low-lying rectal cancer, as long as the pathological distal resection margin is negative.^22,23 Because of the confined and narrow pelvis, pelvic sidewall invasion by tumor cells may be increased as the rectum descends. Thus, it is very important to secure the negative circumferential resection margin in low-lying rectal cancer. Currently, many investigators have reported that the most potent risk factor for oncological outcome in rectal cancer patients is circumferential resection margin involvement.^5,24 We could draw similar results with the results from this study. In the LATA group, the distal resection margin was shorter than in the LAR or the APR group (2.28 ± 1.61 versus 3.53 ± 2.04 or 3.44 ± 1.77 cm, P < .001). The proportion of patients with a distal resection margin of less than 1 cm in the LATA group (18.5%) is larger than those in the LAR (4.1%) or the APR (7.3%) group (P < .001). However, these differences were not reflected in the oncological outcomes. The distal resection margin of less than 1 cm was not significantly related to DFS (P = .975), OS (P = .912), and local recurrence (P = .126). On the other hand, circumferential resection margin involvement was significantly related to DFS (P < .001), OS (P < .001), and local recurrence (P < .001) in univariate analysis (Table 5). Also, circumferential resection margin involvement was the most potent independent factor for DFS (Hazard Ratio [HR] 6.796, P = .009), OS (HR 9.429, P = .002), and local recurrence (HR 8.293, P = .004) in multivariate analysis (Table 6).

In our study, there was no significant difference in circumferential resection margin involvement between the LATA (4.1%) and the APR (4.3%) group. Only lower tumor location was significantly different between the LATA and the APR group. The LATA was more frequently performed in male than in female patients. That was because the rectal transection, keeping secure distal resection margin and CRM with a stapling device, such as a gastrointestinal anastomosis stapler, was more difficult in male patients than in female patients due to the narrower pelvis in male patients. Consequently, tumor location might be a bit higher in male patients of the LATA group than in female patients. This finding might be attributed to higher tumor location in the LATA group than in the APR group. Comparing with DFS, OS, and local recurrence among three surgical techniques (LAR, LATA, and APR), the patients in APR group than those in LAR or LATA groups had significantly poorer DFS (P < .001), OS (P < .001), and local recurrence (P = .001). Although the tumor location in itself may be a risk factor for oncological outcome in rectal cancer, it is an uncorrectable factor, regardless of whether they do a sphincter-preserving surgery or APR. Wibe et al.²⁵ proposed that poor outcome of the APR group may be the result of more patients with T4 stage tumors and predominantly low rectal cancer. Furthermore, lower tumor location and poor response to nCRT as well as a high proportion of tumor characterized by poor/undifferentiated histology may be one of the phenotypes of more aggressive tumor biology in the APR group.¹⁵ In the present study, surgical techniques, including LAR, LATA, and APR was the independent risk factor associated with DFS (HR 6.279, P = .043) and OS (HR 6.414, P = .011). Based on our results, the LATA procedure may be an oncologically safe procedure with better DFS and OS than APR.

Traditionally, most physicians had thought that sphincter preservation with the potential to avoid permanent stoma is preferable for the long-term quality of life for rectal cancer patients.²⁶ But, some authors insisted that individualization of care for rectal cancer patients is essential, but a policy of avoidance of APR cannot be justified on the grounds of quality of life alone.^27,28 All our authors agree with the comment that sphincter-preserving surgery may cause severe functional impairment after surgery. However, recently, several treatments with promising results have been applied to overcome the functional impairment after sphincter-preserving surgery. Therefore, sphincter-preserving surgery to avoid permanent stoma may give patients with low-lying rectal cancer one more chance to maintain their body shape and normal appearance as well as acceptable bowel function.

Our study has some limitations in that it was a retrospective study using the medical records of patients, and the follow-up observation could not be performed in all patients. There were missing data, which could make the study less reliable. Furthermore, we could not evaluate the effect of adjuvant chemotherapy on the oncological outcome. The results are obtained from surgery by a single surgeon. Although these data, as all single-surgeon reports do, suffer from issues of reproducibility, they represent a significant experience over a long period of time showing good outcomes of laparoscopic surgery in the low-lying rectum and accomplishing good oncological outcomes. Also, we did not evaluate functional outcome in the LATA group. Nevertheless, we think that these limitations do not change the importance of better oncological outcome in sphincter-preserving surgery for low-lying rectal cancer.

Based on these results, we suggest that LATA may be a good surgical option in selective low-lying rectal cancer patients who should undergo APR if they are treated by the previous 5-cm rule. For applying LATA to low-lying rectal cancer, surgeons have to do their best to achieve the negative circumferential resection margin. In addition, we think that research for the biological behavior of low-lying rectal cancer and quality of life after sphincter preservation should be performed before widespread adoption.

Footnotes

Acknowledgments

The authors thank DY Won, MK Kim, JK Lee, HJ Yang, and Gun Kim for data collection and KY Lee and KM Son for valuable suggestions and comments.

Disclosure Statement

No competing financial interests exist.

References

Miles

. A method of performing abdominoperineal excision for carcinoma of the rectum and of the terminal portion of the pelvic colon. Lancet, 1908; II:1812–1813.

Scala

, Niglio

, Pace

, Ruffolo

, Rega

, Delrio

. Laparoscopic intersphincteric resection: Indications and results. Updates Surg, 2016; 68:85–91.

Heald

, Husband

, Ryall

. The mesorectum in rectal cancer surgery: The clue to pelvic recurrence?. Br J Surg, 1982; 69:613–616.

Wagman

, Minsky

, Cohen

, Guillem

, Paty

. Sphincter preservation in rectal cancer with preoperative radiation therapy and coloanal anastomosis: Long term follow-up. Int J Radiat Oncol Biol Phys, 1998; 42:51–57.

Adam

, Mohamdee

, Martin

, Scott

, Finan

, Johnston

, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet, 1994; 344:707–711.

Rullier

, Sa Cunha

, Couderc

, Rullier

, Gontier

, Saric

. Laparoscopic intersphincteric resection with coloplasty and coloanal anastomosis for mid and low rectal cancer. Br J Surg, 2003; 90:445–451.

Schiessel

, Novi

, Holzer

, Rosen

, Renner

, Hőlbling

, et al. Technique and long-term results of intersphincteric resection for low rectal cancer. Dis Colon Rectum, 2005; 48:1858–1865.

Park

, Choi

, Jun

, Hasegawa

, Sakai

. Laparoscopic versus open intersphincteric resection and coloanal anastomosis for low rectal cancer intermediate-term oncologic outcomes. Ann Surg, 2011; 254:941–946.

Jeong

, Park

, Nam

, Kim

, Kang

, Lim

, et al. Open versus laparoscopic surgery for mid-rectal or low-rectal cancer after neoadjuvant chemoradiotherapy (COREAN trial): Survival outcomes of an open-label, non-inferiority, randomised controlled trial. Lancet Oncol, 2014; 15:767–774.

10.

Jiang

, Jiang

, Dai

, Wang

, Wu

, Wang

, et al. Laparoscopic versus open surgery for mid-low rectal cancer: A systematic review and meta-analysis on short- and long-term outcomes. J Gastrointest Surg, 2015; 19:1497–1512.

11.

Marks

, Mohiuddin

, Rakinic

. New hope and promise for sphincter preservation in the management of cancer of the rectum. Semin Oncol, 1991; 18:388–398.

12.

Yeatman

, Bland

. Sphincter-saving procedures for distal carcinoma of the rectum. Ann Surg, 1989; 209:1–18.

13.

Steele

Jr ., Hamilton

, Karr

. A rational next step in the treatment of some rectal adenocarcinomas. J Clin Oncology, 1989; 7:988–990.

14.

Kye

, Cho

. Overview of radiation therapy for treating rectal cancer. Ann Coloproctol, 2014; 30:165–174.

15.

Weiser

, Quah

, Shia

, Guillem

, Paty

, Temple

, et al. Sphincter preservation in low rectal cancer is facilitated by preoperative chemoradiation and intersphincteric dissection. Ann Surg, 2009; 249:236–242.

16.

Kim

, Heo

, Son

, Lee

, Suh

, et al. Impact of laparoscopic surgery on the long-term outcomes for patients with rectal cancer. ANZ J Surg, 2009; 79:817–823.

17.

Poulin

, Schlachta

, Gregoire

, Seshadri

, Cadeddu

, Mamazza

. Local recurrence and survival after laparoscopic mesorectal resection for rectal adenocarcinoma. Surg Endosc, 2002; 16:989–995.

18.

Kang

, Park

, Jeong

, Nam

, Choi

, Kim

, et al. Open versus laparoscopic surgery for mid or low rectal cancer after neoadjuvant chemoradiotherapy (COREAN trial): Short-term outcomes of an open-label randomised controlled trial. Lancet Oncol, 2010; 11:637–645.

19.

Cole

. The intramural spread of rectal carcinoma. Brt M J, 1913; 1:431–433.

20.

Wolmark

, Fisher

. An analysis of survival and treatment failure following abdominoperineal and sphincter-saving resection in Dukes' B and C rectal carcinoma. A report of the NSABP clinical trials. Ann Surg, 1986; 204:480–489.

21.

Nelson

, Petrelli

, Carlin

, Couture

, Fleshman

, Guillem

, et al. Guidelines 2000 for colon and rectal cancer surgery. J Natl Cancer Inst, 2001; 93:583–596.

22.

Bujko

, Rutkowski

, Chang

, Michalski

, Chmielik

, Kusnierz

. Is the 1-cm rule of distal bowel resection margin in rectal cancer based on clinical evidence? A systematic review. Ann Surg Oncol, 2012; 19:801–808.

23.

Fitzgerald

, Brinkley

, Zervos

. Pushing the envelope beyond a centimeter in rectal cancer: Oncologic implications of close, but negative margins. J Am Coll Surg, 2011; 213:589–595.

24.

Mukkai Krishnamurty

, Wise

. Importance of surgical margins in rectal cancer. Surg Oncol, 2016; 113:323–332.

25.

Wibe

, Syse

, Andersen

, Tretli

, Myrvold

, Soreide

. Oncological outcomes after total mesorectal excision for cure for cancer of the lower rectum: Anterior versus abdominoperineal resection. Dis Colon Rectum, 2004; 47:48–58.

26.

Williams

, Price

, Johnston

. The long-term effect of sphincter preserving operations for rectal carcinoma on the function of the anal sphincters in man. Br J Surg, 1980; 67:203–208.

27.

Grumann

, Noack

, Hoffmann

, Schlag

. Comparison of quality of life in patients undergoing abdominoperineal extirpation or anterior resection for rectal cancer. Ann Surg, 2001; 233:149–156.

28.

Cornish

, Tilney

, Heriot

, Lavery

, Fazio

, Tekkis

. A meta-analysis of quality of life for abdominoperineal excision of rectum versus anterior resection for rectal cancer. Ann Surg Oncol, 2007; 14:2056–2068.