Comparison between robotic-assisted and laparoscopic sphincter-preserving operations for ultra-low rectal cancer

Abstract

Aim

Sphincter-preserving operations for ultra-low rectal cancer include low anterior and intersphincteric resection. In low anterior resection, the distal rectum is divided by a transabdominal approach, which is technically demanding. In intersphincteric resection, a perineal approach is used. We aimed to evaluate whether robotic-assisted surgery is technically superior to laparoscopic surgery for ultra-low rectal cancer. We compared the frequency of low anterior resection in cases of sphincter-preserving operations.

Method

We investigated 183 patients who underwent sphincter-preserving robotic-assisted or laparoscopic surgery for ultra-low rectal cancer (lower border within 5 cm of the anal verge) between April 2010 and March 2020. The frequency of low anterior resection was compared between laparoscopic and robotic-assisted surgeries. The clinicopathological factors associated with an increase in performing low anterior resection were analyzed by multivariate analyses.

Results

Overall, 41 (22.4%) and 142 (77.6%) patients underwent laparoscopic and robotic-assisted surgery, respectively. Patient characteristics were similar between the groups. Low anterior resection was done significantly more frequently in robotic-assisted surgery (67.6%) than in laparoscopic surgery (48.8%) (P = 0.04). Multivariate analyses showed that tumor distance from the anal verge (P < 0.01) and robotic-assisted surgery (P = 0.02) were significantly associated with an increase in the performance of low anterior resection. The rate of postoperative complications or pathological results was similar between the groups.

Conclusion

Compared with laparoscopic surgery, robotic-assisted surgery significantly increased the frequency of low anterior resection in sphincter-preserving operations for ultra-low rectal cancer. Robotic-assisted surgery has technical superiority over laparoscopic surgery for ultra-low rectal cancer treatment.

BACKGROUND

Surgical outcomes of robotic-assisted surgery (RS) for rectal cancer have recently been reported.^1-3 Several studies have shown excellent outcomes with RS in terms of urinary function, sexual function, defecatory function, and local recurrence rate for rectal cancer compared to those achieved by laparoscopic surgery (LS) or open surgery (OS).^4-9

To date, differences in the types of RS and LS for low rectal cancer have not been fully investigated. The advanced technology of RS allows sphincter-preserving operations (SPOs) to be performed more often than LS or OS.^{10, 11} Further, RS enables patients with low rectal cancer to avoid abdominoperineal resection (APR) and the creation of a permanent stoma, suggesting that RS would be technically superior. Nevertheless, the indications for SPOs are determined not only by technical factors but also by anorectal function and lifestyle. SPOs markedly influence postoperative defecatory function.¹² Consequently, preoperative anorectal functions and the patient's acceptance of postoperative bowel dysfunction must always be considered when contemplating an SPO. Therefore, to evaluate the technical superiority of RS for rectal cancer in terms of selecting the type of operation, investigating only the frequency of SPOs would be insufficient.

For a more detailed investigation of the technical superiority of RS with regards to the operation type, we focused on two SPOs, low anterior resection (LAR) and intersphincteric resection (ISR), performed for ultra-low rectal cancer.¹³ Generally, LAR is performed when the distal rectum can be divided using linear staplers via a transabdominal approach, and ISR is performed when this is not possible. Dividing the distal rectum via the transabdominal approach is technically demanding, and the type of SPOs that can be performed depends mainly on technical factors. Therefore, the frequency of conducting LAR by a specific procedure can indicate the technical superiority of that procedure for ultra-low rectal cancer. Thus, in this study, we compared the frequency of LAR between RS and LS in patients who underwent SPOs for ultra-low rectal cancer to evaluate whether RS was technically superior to LS.

METHODS Patient selection

Data were collected from the institutional database at Shizuoka Cancer Center and analyzed retrospectively. Between April 2010 and March 2020, 322 patients with primary ultra-low rectal adenocarcinoma (lower edge of the tumor within 5 cm from the anal verge [AV]) underwent SPO with curative intent. SPOs included LAR and ISR, and all patients underwent transabdominal total mesorectal excision (TME).¹⁴ No patient underwent transanal TME (TaTME). During LAR, both TME and division of the distal rectum using a linear stapler were performed via a transabdominal approach. The coloanal anastomosis was performed using the double-stapling technique. ISR was defined as a surgical procedure that included both the transabdominal and transanal approaches. In ISR, TME was performed using the transabdominal approach, followed by transanal resection. This involved surgical dissection of the intersphincteric space and en bloc resection of the rectum, together with the internal anal sphincter within the anatomical anal duct. After removal of the specimen, a coloanal anastomosis was performed using transanal manual suturing in all patients.^{15, 16} Patients in whom the intersphincteric space could be entered and partial excision of the internal anal sphincter was performed via the transabdominal approach were included in the LAR group, as previously described.¹⁷

We excluded patients with tumors that extended into the anal canal, those who underwent OS, those who received preoperative treatment such as chemotherapy and/or radiotherapy, those with multivisceral resection, and those with a history of colectomy other than appendectomy (Figure 1). Consequently, 183 patients were enrolled in this study.

FIGURE 1. Flowchart of patient selection

Preoperative assessment included digital rectal examination, colonoscopy, computed tomography, contrast enema, and magnetic resonance imaging. If necessary, positron-emission tomography/computed tomography was performed. Tumor staging was performed using the tumor node metastasis classification.¹⁸ When the serosa of the rectum above the peritoneal reflection was invaded by the tumor, tumor depth was classified as T4a. Cases with T4b tumors were excluded from this study because multivisceral resection was required for these tumors. The indications for lateral lymph node dissection (LLD) were low rectal adenocarcinoma with cT3-4 or cT1-2 with metastasis to the lateral lymph nodes.^{19, 20} However, LLD was not performed if the patients without lateral lymph node metastasis on preoperative images were 75 years or older, had severe comorbidities, or received preoperative chemoradiation.¹⁹ Consistent with a previous report, we regarded lateral lymph nodes as regional lymph nodes.²¹ The study was approved by the institutional review board of Shizuoka Cancer Center (Approval Code: J2020-37-2020-1-3).

Indications for surgical approach

During the study period, we generally performed minimally invasive surgery (MIS), including LS and RS. However, tumors that required multivisceral resection with urinary diversion or reconstruction or total pelvic exenteration were treated using OS. Since there were significant differences in patient background between the OS and MIS groups, patients who underwent OS were excluded from this study. Furthermore, in the early to mid-stages of the study period, LLD was performed via OS or RS; LS was not performed for the cases of LLD. At our institution, RS was introduced in December 2011. Further, RS for rectal cancer was not covered under the national public health insurance plan in Japan until March 2018, making it more expensive than LS or OS until this period. After the patients provided informed consent, they indicated their preference for RS, LS, or OS, and the procedure was selected accordingly. All treatment strategies were approved in a multidisciplinary team conference.

Indications for SPO

Sphincter-preserving operation was indicated only for patients with the following characteristics: a clear resection margin, acceptable postoperative anal function, and absence of invasion into the levator ani muscles and external anal sphincter. The standard operative rule was total TME with autonomic nerve preservation.¹⁴ Japanese D2 or D3 lymph node dissection was performed.^{12, 15} LAR was indicated for patients in whom the transabdominal approach could divide the distal rectum using linear staplers. ISR was performed when the distal rectum could not be divided via the transabdominal approach, either because of technical difficulties or to ensure a sufficient distal resection margin. ISR was excluded for patients with poorly differentiated adenocarcinoma or with decreased anal sphincter tonus. APR was performed when (1) the tumor invaded the external anal sphincter or levator ani muscles, (2) the tumor was the diffuse infiltrative spreading type based on macroscopic appearance, (3) fecal continence was impaired preoperatively, or (4) patients could not tolerate postoperative defecatory dysfunction. All treatment strategies were approved in a multidisciplinary team conference.

Surgical procedure

The institutional standard surgical procedure has been described elsewhere.^22-24 The surgical procedure was almost the same in both LAR and ISR up until dividing the rectum. In brief, the inferior mesenteric artery and vein ligations and colon mobilization were performed via a medial-to-lateral approach. The rectum was mobilized down to the pelvic floor for TME. During LAR, the distal rectum was divided using linear staplers via the transabdominal approach. In RS, a robotic-assisted or a conventional laparoscopic linear stapler was used. In LS, only a conventional laparoscopic linear stapler was used. The specimen was extracted through a mini-laparotomy at the umbilical site. The standard double-stapling technique was utilized for end-to-end anastomosis. A diverting stoma was constructed if necessary. In patients who underwent ISR, intersphincteric dissection and specimen extraction were performed via the transanal approach.²⁴ The coloanal anastomosis was hand-sewn. A diverting stoma was constructed in all patients who underwent ISR. During the transanal part of the surgery, the surgical procedure was the same in both RS and LS.²⁰

Outcome variables

We collected data on patient characteristics as well as perioperative outcomes, including the frequency of LAR, postoperative complications, and pathological results. Postoperative complications within 30 days were categorized according to the Clavien-Dindo classification system.²⁵ The association between the clinicopathological factors and SPO type was analyzed. The radicality of each surgical procedure was defined as R0 (macroscopic complete resection with no microscopic residual tumor), R1 (macroscopic complete resection with microscopic residual tumor at the resection margin), or R2 (macroscopic residual tumor).

Statistical analyses

Parametric variables are expressed as median values. The Fisher's exact test, chi-squared test, and Mann-Whitney U test were appropriately used to assess the significance of between-group differences. Univariate and multivariate logistic regression analyses were performed to identify factors associated with the performance of LAR. A P-value of <0.05 was considered significant. All statistical analyses were performed using R software version 3.5.2 (The R Foundation for Statistical Computing).

RESULTS

A total of 183 consecutive patients who underwent SPOs for ultra-low rectal cancer were analyzed. The characteristics of the 41 (22.4%) patients who underwent LS and 142 (77.6%) patients who underwent RS are summarized in Table 1. Although the clinical stage in the RS group tended to be more advanced than that in the LS group, the difference was not statistically significant. In the RS group, the lower edge of the tumor was located just 2 cm from the AV in only one case. In the remaining 182 cases, the lower edge of the tumor was located within 3-5 cm of the AV.

TABLE 1 Patient characteristics

[Table omitted. See PDF]

Note:

Values are expressed as number (percentage) or median value (range).

Abbreviations: ASA-PS, American Society of Anesthesiologists-Physical Status; AV, anal verge; BMI, body mass index; LS, laparoscopic surgery; RS, robotic-assisted surgery.

Table 2 summarizes the perioperative outcomes. Compared to the LS group, LLD was performed significantly more frequently in the RS group. Operative times were significantly longer in the RS group than in the LS group. However, when the patients who underwent LLD were excluded, operative times in the RS group tended to be shorter than those in the LS group (Table S1). The frequency of LAR was significantly higher and blood loss was significantly lower in the RS group than in the LS group. The distal cut-end lines in ISR cases are summarized in Table S2.

TABLE 2 Perioperative outcomes

[Table omitted. See PDF]

Note:

Values are expressed as number (percentage) or median value (range).

Abbreviations: ISR, intersphincteric resection; LAR, low anterior resection; LS, laparoscopic surgery; RS, robotic-assisted surgery.

The associations between clinicopathological factors and the type of SPO are shown in Table 3. Univariate analyses indicated that tumor distance from the AV and RS were significantly associated with an increase in the frequency of LAR for ultra-low rectal cancer. Multivariate analyses revealed that tumor distance from the AV and RS was significantly associated with an increased LAR frequency.

TABLE 3 Univariate and multivariate analyses of factors linked to LAR performance frequency

[Table omitted. See PDF]

Note:

Values are expressed as number (percentage) or median value (range).

Abbreviations: AV, anal verge; BMI, body mass index; CI, confidence interval; ISR, intersphincteric resection; LAR, low anterior resection; LS, laparoscopic surgery; RS, robotic-assisted surgery.

Table 4 summarizes the postoperative complications. The rate of postoperative complications (≥Clavien-Dindo grade II) was similar between the groups. Postoperative mortality did not occur in either group.

TABLE 4 Postoperative complications

[Table omitted. See PDF]

Note:

Values are expressed as number (percentage) or median value (range).

Abbreviations: LS, laparoscopic surgery; RS, robotic-assisted surgery.

Table 5 shows the pathological results. Complete local resection (R0) rates and distal resection margins were similar between the LS and RS groups.

TABLE 5 Pathological outcomes

[Table omitted. See PDF]

Note:

Values are expressed as number (percentage) or median value (range).

Abbreviations: LS, laparoscopic surgery; RS, robotic-assisted surgery.

DISCUSSION

In this study, we evaluated whether RS has technical superiority over LS for ultra-low rectal cancer by comparing LAR frequency in patients who underwent SPOs. Compared to LS, RS was significantly associated with an increased frequency of LAR in SPOs for ultra-low rectal cancer. Our findings support the argument that RS has technical superiority over LS for ultra-low rectal cancer treatment.

Minimally invasive surgery for low rectal cancer requires highly advanced surgical skills. The lower the tumor location in the deep and narrow pelvis, the more technically challenging the procedure becomes. Previously, APR was the standard procedure for low rectal cancers within 5 cm of the AV.^{15, 26, 27} However, ISR is now widely recognized as an acceptable SPO procedure.²⁸ Moreover, recent advances in surgical techniques have made it possible to perform LAR, including partial excision of the internal anal sphincter, via a transabdominal approach for ultra-low rectal cancer.^{17, 29}

In other words, LAR for ultra-low rectal cancer requires mobilization of the rectum, even in the anal canal, by the transabdominal approach, which still requires highly advanced surgical techniques. Therefore, the technical superiority of RS over LS could be evaluated by comparing the frequency of LAR between these two types of SPOs for ultra-low rectal cancer. To the best of our knowledge, no previous study has attempted to compare the types of SPOs between LS and RS for ultra-low rectal cancer. In contrast to previous reports, we excluded APR, the indications of which are determined by the tumor and technical and non-technical factors, such as preoperative anorectal dysfunction and tolerance of postoperative defecatory dysfunction. As such, the types of SPOs utilized for the patients herein were determined largely based on technical factors, which is also a strong point of our study.

In this study, LAR frequency was significantly higher in the RS group (67.6%) than in the LS group (48.8%), and multivariate analyses showed that RS and the tumor distance from the AV were significantly associated with an increased frequency of LAR for ultra-low rectal cancer. Furthermore, there were no significant differences in the postoperative complications or the pathological results between the RS and LS groups. Although the operative time was longer in the RS group, this was attributed to the higher proportion of patients who underwent LLD in the RS group. Hence, compared to LS, RS allows transection of the distal rectum for ultra-low rectal cancer via the transabdominal approach without compromising surgical and pathological outcomes. These findings suggest that RS has technical superiority over LS for ultra-low rectal cancer. This superiority can be attributed to the technology involved in RS, such as digital suppression of hand tremors, free-moving multi-joint forceps, and high-quality three-dimensional visualization. Furthermore, the robotic stapler may be advantageous during stapling owing to its good maneuverability.^{30, 31}

In addition to the transabdominal approach, TaTME has been used for rectal cancer.³² It has been reported that the transanal approach could enhance access to the distal part of the rectum and enable better visualization, making a more accurate oncological dissection possible.^{33, 34} However, early adopters of TaTME underscored its technical difficulty. Several studies examining TaTME registry data have revealed that patients sustained visceral injuries during perineal dissection.^{35, 36} Additionally, Wasmuth et al³⁷ demonstrated that the oncological outcomes after TaTME were inferior to the national results. Further investigation is required to evaluate the long-term oncological results of TaTME. In contrast, RS makes it possible to perform surgery safely and precisely, even for ultra-low rectal cancer via the transabdominal approach, which is commonly used for TME. Since RS is performed in a familiar transabdominal field-of-view, it is easy to recognize the anatomy and the surgery can be performed safely, even in the deep pelvis. Furthermore, several studies have indicated good long-term results.^{2, 11, 38}

This study has several limitations. First, this was a retrospective study performed at a single institution. Second, long-term outcomes were not investigated. Randomized controlled trials are necessary to validate our findings. Third, although this study focused on the technical aspects, it was unclear whether LAR or ISR is better in terms of postoperative anorectal function. Previous studies have reported that patients who have undergone ISR have a higher risk of fecal incontinence and bowel dysfunction than those who have undergone LAR.^{39, 40} However, the characteristics of patients in previous studies were different from those in this study, and these results could not be extrapolated to the present study. Further studies are necessary to clarify the effects on postoperative anal function. Fourth, we excluded OS in this study. Therefore, we could not evaluate the superiority of RS over OS.

CONCLUSION

The findings of this present study revealed that RS significantly increased the frequency of transection of the rectum by the transabdominal approach in SPOs, supporting that RS has technical superiority over LS for the treatment of ultra-low rectal cancer.

DISCLOSURES

Funding: This manuscript did not receive sponsorship for publication.

Conflict of Interest: Authors declare no conflict of interest for this article.

Author Contributions: Tadahiro Kojima and Hitoshi Hino drafted the paper. Hitoshi Hino designed this study. Akio Shiomi, Hiroyasu Kagawa, Yusuke Yamaoka, Shoichi Manabe, Shunichiro Kato, and Marie Hanaoka obtained and analyzed data. All authors critically revised the report, commented on drafts of the manuscript, and approved the final report.

Ethical approval: The protocol for this research project has been approved by a suitably constituted Ethics Committee of the Shizuoka Cancer Center, and it conforms to the provisions of the Declaration of Helsinki. Committee of Shizuoka Cancer Center, Approval No. J2020-37-2020-1-3. All informed consent was obtained from the subject(s) and/or guardian(s).