Adequate Margins for Anorectal Cancer Can Be Achieved by Single-Site Laparoscopy

Introduction

Despite initial trepidation that a laparoscopic approach toward the treatment of colon cancer would yield unfavorable oncologic results compared with open surgery, it has been demonstrated beyond contention that when laparoscopic techniques are appropriately applied by experienced practitioners, minimally invasive colon cancer surgery can routinely attain adequate margins of resection with comparable cancer-specific survival compared with open colon surgery.^1,2 Furthermore, these comparable outcomes can be achieved while affording a clinical advantage as measured in shorter lengths of hospital stay and an accelerated return to baseline levels of activity.^1,2 Many of these findings have been demonstrated with the strength of multicenter, prospective randomized trials.^2,3 In contradistinction, the oncologic adequacy of laparoscopic rectal cancer surgery remains a more controversial subject.^4–8 Surgeons who frequently operate on rectal cancer are well aware of the greater technical challenges attendant to a dissection in the restricted confines of the pelvis compared with the wider territory sustained in the abdomen. Additionally, the difficulty encountered in the application of stapling devices in the distal pelvis raises concerns about whether laparoscopy can consistently replicate the margins of resection and the prudent tissue handling customarily achieved in open rectal resections. ⁷ Although there are previous studies that have provided evidence that minimally invasive surgery can be applied to the treatment of rectal cancer with good results, the sum of the evidence advocating for this approach as comparable to open rectal surgery is less established than the case for laparoscopy for colon cancer. ⁵ There are ongoing randomized trials comparing the oncologic outcomes of patients with rectal cancer who undergo laparoscopic versus open resections, with the results still pending.

Introducing an additional wrinkle in the debate over the proper role of laparoscopy in rectal cancer surgery is the recent development of several variants of standard laparoscopy (SL), one of which involves single-site laparoscopy (SSL). ⁹ There are few data on the use of SSL for any particular disease process, and thus almost no data dedicated to the subject of SSL for anal or rectal cancers are available.^10–14 The same challenges present for SL rectal cancer surgery, such as achieving adequate distal margins of resection, attaining an intact mesorectal excision, and the hindrance of using endostaplers in the deep pelvis, are persistent but magnified all the more by restricting the surgical approach for rectal cancer to a single incision.

If SSL is to have any role in rectal cancer surgery, then at minimum it should consistently produce surgical specimens that meet the gross and microscopic marks commensurate with an adequate resection. This issue is preeminent in assessing any novel surgical approach for malignancies of the anorectum, considering the unforgiving nature and the frequently incurable condition of recurrent anorectal cancer. The purpose of this study was to describe the pathologic assessments of consecutive patients undergoing SSL low anterior resection (LAR) and abdominoperineal resection (APR) for malignancies of the mid- and distal rectum and anal canal. The rationale for the study was founded on the concept that the quality of rectal resection as objectively measured by the pathologist is principal to rectal cancer surgery, and any future studies regarding feasibility or clinical outcomes with SSL would be unwarranted if the quality of the resection is not first demonstrated.

Subjects and Methods

This was a retrospective analysis of an Institutional Review Board–approved, prospectively maintained database, linked to an Institutional Review Board–approved study of SSL for large intestinal malignancies. Consecutive patients from October 2011 until February 2012 who would have otherwise been offered a laparoscopic resection for rectal and anal cancers using a standard laparoscopic approach, which for the authors had traditionally involved the use of two 5-mm bladeless trocars and a single 12-mm bladeless trocar, underwent resection of their cancer using the Ethicon SSL Access System (Ethicon Endosurgery, Cincinnati, OH). All surgeries were performed by board-certified colorectal surgeons. All patients who did not meet the previously established exclusion criteria for elective laparoscopic surgery (more than three previous laparotomies and/or a body mass index of ≥50 kg/m²) underwent a single-site LAR or APR (excluded: n=1). No other form of laparoscopy besides SSL was used during this time period, so that every laparoscopic surgery was at least initiated as SSL on consecutively encountered patients.

Information collected during the study included demographic information such as age and gender, body mass index, American Society of Anesthesiologists score, and location of the cancer from the anal verge as measured by rigid proctoscopy. Operative details included the type of resection performed (LAR or APR), mean operative times, estimated blood loss, location of the single-site trocar, the need to convert to a multiple-trocar surgery, which was defined as the insertion of any additional trocar(s), the need to convert to a laparotomy, and whether a conversion was preemptive (because of the anticipated need for conversion to avoid an intraoperative complication) or reactive (because of the committal of an intraoperative complication). Postoperative outcomes included time to tolerating a regular diet, time to return of flatus, mean lymph node yield, length of hospital stay, 30-day complication rates, and 30-day mortality rates. Each patient was evaluated personally by the authors 30 days after surgery, to ensure that all postoperative complications were captured.

Results

Patient demographics are provided in Table 1. Median patient age was 66 years (range, 36–87 years), with 58% of the study population being male. The overwhelming majority of patients were white (91%), with a median body mass index of 28 kg/m². All patient American Society of Anesthesiologists scores were 3, with approximately half of the patients having a diagnosis of systemic hypertension or coronary artery disease. Six (50%) of the patients had undergone previous surgeries, all of which were laparotomies, with 4 (67%) of these patients having undergone more than one prior surgery. Eleven (91%) of the patients underwent surgery for adenocarcinoma of the rectum, whereas 1 patient underwent an SSL APR for an anal canal melanoma.

Table 2 provides a summary of operative details. The median operative time when considering all patients was 200 minutes (mean, 210±44 minutes); median operative times for SSL APRs were longer than for LARs but did not represent a statistically significant difference (LAR, 185 minutes; APR, 240 minutes; P=.2). Median estimated blood loss for all subjects was 100 mL (mean, 109±53 mL), with APRs demonstrating a higher median estimated blood loss (LAR, 88 mL; APR, 150 mL; P=0.3). For each APR, the SSL trocar was placed in a left-sided colostomy site; for LARs, 2 patients had their trocar placed in the umbilicus, whereas the remaining patients had their trocar placed in a potential stoma site. All patients undergoing an LAR who underwent neoadjuvant chemoradiation therapy also had a protecting loop ileostomy constructed (n=2). Mean incision length was 3.1 cm.

Table 3 provides cancer characteristics for each patient. Seven (58%) patients underwent an LAR, with the remaining patients undergoing an APR. The median distance of the inferior border of each cancer from the anal verge as measured via rigid proctoscopy performed by the surgeon prior to neoadjuvant therapy was 5 cm (mean, 6±4 cm). None of the patients who underwent neoadjuvant chemoradiation (n=6) developed a change to the location of the inferior border of their cancer beyond 1 cm as measured by a post-therapy/preoperative proctoscopy. Median distal margins of resection for LARs were 3 cm (mean, 3±0.7 cm), and for APRs they were 1.9 cm (mean, 2±1.3 cm). Median radial margins for LARs were 15 mm (mean, 10.8±7.7 mm), whereas median radial margins for APRs were 2 mm (mean, 5.3±5.7 mm).

Based on pathologic staging, a plurality of the cancers were T3 (41%), with an equal number of T1 (25%) and T2 (25%) lesions. One rectal cancer was a T4 lesion that required an en bloc ileocecectomy with an ileocolic anastomosis in addition to an LAR and colorectal anastomosis. Pathologic nodal staging demonstrated 58% of the cancers as being N0, with 33% being N2 and 9% being N1. Median node yield for all patients was 18 (mean, 19.7±12.5); for patients who underwent an LAR, median node yield was 16.5 (mean, 23±16.3), with similar results (P=.2) for APRs with a median node harvest of 18 (mean, 15.8±4.7). At least 12 nodes were harvested from all patients who did not receive neoadjuvant therapy. Two of the patients who received neoadjuvant therapy had fewer than 12 nodes (8 nodes and 10 nodes, respectively). Regarding the number of nodes involved with cancer, for all patients the median number was 0 (mean, 1.6±2.7), with LARs having a median of 2 nodes (mean, 2.8±3.2) and APRs having a median of 0 (mean, 0.2±0.4), a finding that did not represent a statistically significant difference (P=.33). All mesorectal excisions were complete (Grade 3) according to criteria provided by Nagtegaal et al. ¹⁵

Table 4 provides a summary of salient 30-day outcomes for the study group. There were no conversions from SSL to either open surgery or to SL. All patients were able to tolerate a solid food diet on the day of surgery, with median time to return of flatus of 1 day (LAR, median of 1 day; APR, median of 3 days; P=.5). Median length of hospital stay for all patients was 2 days (mean, 3.1±1.6 days); patients undergoing an LAR had a shorter median length of hospital stay than those undergoing an APR (LAR, 2 days; APR, 5 days; P=.03). Only 1 subject required re-admission to the hospital within 30 days of surgery, which involved a patient who underwent an LAR and a diverting loop ileostomy and who required 23-hour observation for a food bolus obstruction. No patients required blood transfusions during the perioperative period. There were no intraoperative complications or 30-day surgical site infections, anastomotic leaks, hollow viscus injuries, ureteral injuries, venous thromboembolic events, unplanned re-operations, or deaths. Ten (91%) patients were able to have their bladder catheter removed within 24 hours of surgery, without urinary retention.

Discussion

The present investigation represents a preliminary study on the application of SSL for rectal cancer. Despite the small number of patients, the issue raised in the present study is vitally important because the ability to achieve adequate margins of resection with SSL is the logical starting point prior to forays into issues related to clinical outcomes and cancer-related survival with this new technique.^12–14 The current investigation demonstrates that an intact mesorectal excision can be performed safely using an SSL approach without compromising oncologic principles and without adding significant morbidity to an already morbid procedure.

To date, there is minimal Level 1 and Level 2 evidence supporting the laparoscopic approach for rectal cancer as oncologically satisfactory compared with open resections, and this despite the increasing acceptance of a laparoscopic approach for rectal resections. Although there are several randomized studies either that were not designed to evaluate rectal cancer, per se, or that have small study populations that limit the strength of their conclusions, there is no large-scale, multicenter, randomized controlled trial validating minimally invasive rectal cancer surgery.^4,5,7,16 The CLASICC trial included 85 laparoscopic total mesorectal excisions for cancer, with a conversion rate of 57% and with a rate of involved circumferential resection margins of 12%.^1,17,18 Although the high conversion rate might be explained by a more verdant experience with the approach, this finding was of particular concern because this subgroup of study patients experienced a higher rate of postoperative complications compared with patients whose surgery was completed laparoscopically. In a similar manner, the differences between circumferential resection margin status among laparoscopic and open surgeries was thought to reflect the greater technical difficulties associated with laparoscopic rectal cancer surgery and the lack of experience among study surgeons. In the current study of unselected patients undergoing SSL rectal resections, there were no conversions to open surgery or SL, with an associated minimal blood loss and with no intraoperative complications. The single postoperative complication was a re-admission for a food bolus obstruction. Although SSL was technically challenging in some of the study subjects, each mesorectal excision specimen was complete with adequate distal and radial margins.

Although the laparoscopic approach for pelvic surgery is evolving to include robotic-assisted multiport resections, the technology more than outpaces the data regarding oncologic safety with all of these techniques.^19–22 Sphincter-sparing total mesorectal excision should provide the patient an overall locoregional recurrence of less than 5%. Heald ²³ and Heald and Ryall ²⁴ had reported a locoregional recurrence of 2% for Stage I disease, with local recurrence rates of 4% for Stage II disease and with recurrence rates of 7.5% for Stage III cancers, following open total mesorectal excision. Similar results should be reported for SSL before the initiation of a randomized trial and certainly prior to the widespread use of SSL for rectal cancer. The current pilot study demonstrates that R0 resections can be achieved with SSL resections, even for distal rectal cancers, despite the technical challenges.

If adequate margins of resection cannot be achieved during a total mesorectal excision, long-term studies comparing oncologic outcomes for SSL and SL are superfluous. The seminal data presented here indicate that SSL cannot be given a wholesale dismissal due to conceptual reservations that single-port access prevents an intact mesorectal excision and an adequate distal margin of resection. Taking into consideration that laparoscopic approaches can now be approached through robotic assistance, by SSL, or via a standard multiport approach, future randomized trials should be designed to control for these various laparoscopic techniques in terms of assessing which technique will most consistently achieve the goals of rectal cancer surgery.

The present study has several limitations. This is a pilot study, with small numbers of patients, evaluating early postoperative outcomes and pathology indices for rectal and anal cancer resections. Larger retrospective studies, followed by well-designed prospective studies, will be required before SSL is considered equivalent to SL for use with anorectal malignancies. The present study does not offer comparative analysis to another surgical technique, with the present work being principally focused on whether adequate margins of resection and an acceptable mesorectal excision can be achieved using SSL. An additional factor to be considered is that the authors had previously used a three-trocar technique (two 5-mm trocars and a 12-mm trocar) for virtually all of their laparoscopic surgeries, including more complex surgeries such as LARs, APRs, and ileal-pouch constructions. Transitioning from this approach to SSL did not, therefore, represent as significant a transition as would have been the case with changes from a different form of SL. Therefore, the results in the present manuscript may not be reproducible if SSL is applied by surgeons with a different previous laparoscopic experience. Certainly, great care should be shown by the surgeon prior to adopting a new technique that could potentially alter the outcome of a patient, whether oncologic or otherwise.

One additional issue related to SSL is how the additional challenge introduced by this approach will impact the education of surgical residents and fellows and whether this will be to the detriment of their future success as independent surgeons. The authors believe that there is an often-overlooked benefit to educating surgeons in training with an SSL platform. It is the authors' opinion that one of the key obstacles toward the trainee becoming an efficient laparoscopist is related to the ineffectual use of their nondominant hand (the retracting instrument). Improficient habits and techniques in this regard are a hurdle with any laparoscopic approach, although it is the authors' opinion that standard, multiple-trocar laparoscopy provides for a greater margin of leniency than does SSL. In essence, SSL makes a greater issue of those techniques that limit the surgeon's efficiency than SL because the latter allows for the choosing of multiple ports to address issues with exposure. Although SSL may not be an appropriate introductory technique to the youngest surgeon, the authors feel that SSL has a useful pedagogical role in educating trainees to be efficient, two-handed surgeons.

Single-site laparoscopic resections for rectal and anal cancers can achieve adequate radial and distal margins of resection, even when operating in the distal pelvis and after preoperative radiotherapy. Although SSL may potentially develop into a regularly used alternative to SL for rectal and anal cancers, larger prospective studies are needed to validate the oncologic outcomes for this technique.