Comparison of laparoscopic lateral lymph node dissection for rectal cancer with and without routine resection of the visceral branches of internal iliac artery

This study aimed to explore the safety and feasibility of the modified lateral lymph node dissection (LLND) with routine resection of the visceral branches of internal iliac vessels (IIVs) for mid‐low‐lying rectal cancer.


| INTRODUCTION
The existence of the lateral drainage pathway is the main cause of lateral lymph node metastasis (LLNM) in rectal cancer. 1 Approximately 15%-25% incidence of LLNM was observed in middle to lowlying rectal cancer patients. 2LLNM is the main cause of lateral-pelvic recurrence and predicts poor prognosis if left untreated. 3After the application of total mesorectal excision (TME) and neoadjuvant chemoradiotherapy (neoCRT) for locally advanced rectal cancer, the local recurrence rate has greatly decreased to 5%-7%. 4 In the current practice, lateral pelvic recurrence has replaced central-pelvic recurrence to be the main pattern of local recurrence. 5r a long time, Japan and Western countries adopted different perspectives about and treatment regimens for LLNM. 6In Western countries, LLNM was regarded as distant metastasis and managed with neoCRT.However, in Japan and a few Asian centers, LLNM is considered as regional disease and treated with upfront lateral lymph node dissection (LLND) or in combination with neoCRT. 7Previous multicenter studies have demonstrated the insufficiency of neoCRT alone in the control of lateral pelvic recurrence for patients with swollen lateral lymph nodes, 3 with only 5% of swollen lateral lymph nodes disappearing after neoCRT.Meanwhile, LLND was shown to provide improved local control for lateral pelvic recurrence in both prospective and retrospective studies. 3,8wever, LLND without neoCRT has been associated with a high rate of local recurrence in patients with pathological LLNM, ranging from 22.2% to 56.8%, for which lateral-pelvic recurrence accounted for 30%-50%. 9Up to now, the causes of lateral-pelvic recurrence after LLND have not been well understood.Considering that neoCRT could achieve similar local control, as well as control of lateral local recurrence, compared with prophylactic LLND plus TME, 10 one of the reasons is that neoCRT could sterilize microscopic lateral node particles.Another reason is assumed to be the residual of small nodes in the distal part of the lateral compartment.The most common practice of prophylactic LLND does not include routine resection of pelvic visceral branches, which have highly variable anatomy.
Therefore, preservation of these visceral arteries and veins might increase the risk of incomplete resection.This assumption, though without direct supportive evidence, is also indicated by the finding that the existence of the middle rectal artery has been associated higher rate of LLNM.
Here we reported the short-term outcomes of routine resection of this modified technique and compared it to our previous series when this technique was not adopted.All surgeries were performed by the same colorectal surgeon with more than 25-year experience (Ziqiang Wang).This study was approved by the Institutional Review Board of the West China Hospital, Sichuan University.
(5) Internal iliac and obturator nodes were dissected at least.

| Exclusion criteria
(1) Other kinds of rectal tumors including neuroendocrine tumors, stromal tumors, malignant melanoma, and squamous cell carcinomas of the anal canal.

| Indication and technique for LLND
LLND was only indicated for patients with enlarged lymph nodes (≥5 mm on the short axis) around IIVs and the obturator region.LLND was also occasionally performed for patients with lateral lymph nodes 4-5 mm but with malignant features on imaging.In both groups, LLND would be performed if the size of lateral lymph node increased after neoCRT and it would be omitted while the lateral lymph nodes disappeared after treatment.
In the RVR group, the LLND was performed according to the modified technique which was detailed in our previous report. 9Briefly, visceral branches, including inferior vesical vessels, vaginal vessels, and middle rectal vessels, were routinely ligated; whereas umbilical vessels, superior vesical vessels, and uterine vessels were selectively preserved according to the location of lateral lymph nodes and surgeon's discretion during surgery.The posterior trunk of the IIVs (superior gluteal vessels) was routinely preserved and the anterior trunk (internal pudendal vessels, inferior gluteal vessels) was selectively preserved or resected according to its relation to lateral lymph nodes and the branching level of visceral vessels.Usually, the posterior arterial trunk was resected when the branching-off of the inferior or vaginal artery occurred more distally.The venous trunk was usually preserved.In the NRVR group, the LLND was performed generally according to Matsumoto's description. 11The visceral branches of the IIVs were routinely preserved during the procedure and only selectively resected when invaded by positive lateral lymph nodes.

| Survival outcomes
Overall survival, disease-free survival, and the location of recurrence.

| Patient demographics
A total of 75 patients were prospectively enrolled in the RVR group, whereas 57 patients were retrospectively included in the NRVR group (Figure 1).The baseline of included patients is shown in Table 1.No significant difference was observed between the two groups except for lower tumor location (p = 0.044) and higher neoCRT rate in the RVR group.

| Surgical information
The intraoperative details are shown in Table 2.No significant difference was observed between the two groups for the choice of surgical procedure.However, significantly more patients in the RVR group received temporary stomas among those with low anterior resection or intersphincteric resection (p < 0.001).
F I G U R E 1 CONSORT diagram outlining the trial.NRVR, nonroutine visceral branches of the internal iliac vessels; RVR, routine visceral branches of the IIVs resection.
However, a higher proportion of bilateral LLND was observed in the NRVR group, which was not significantly different (p = 0.054).
Besides, 15 (20%) and six (10.5%) patients received surgery beyond TME for suspected tumor invasion in the RVR and NRVR groups, respectively (p = 0.140).The multiple tissue/organs resected beyond TME showed no significant difference between the two groups.The median operative time for the total procedure was shorter in the RVR group (median 270 min vs. 300 min, p = 0.020), whereas it was comparable for the procedure of LLND (median 75 min vs. 78 min, p = 0.464).The median of intraoperative blood loss was comparable between the two groups (80 mL vs. 80 mL, p = 0.708).One patient in the RVR group received intraoperative blood transfusion because of bleeding.

| Postoperative recovery and complications
Patients in the RVR group showed a significantly faster recovery, with shorter postoperative stays (p < 0.001) and earlier flatus (p < 0.001), fluid diet (p < 0.001), and catheter removal (p < 0.001).
No significant difference was observed between the two groups for the overall, major, or minor incidence of postoperative complications.
Besides, the incidence of reoperation and readmission was not significantly different between the two groups (Table 3).LLN was also comparable between the two groups (Table 4).In the subgroup analysis according to the compartment of LLND and the application of neoCRT, a higher incidence of LLNM and more harvested lateral lymph nodes were observed in the RVR group, but it was not significantly different (Table 5).Eleven patients died of tumor progression.Comparing the DFS, the Kaplan-Meier analysis resulted in an estimated average DFS of 36 months (range, 1-96 months) for all of the patients in the NRVR group and 12 months (range, 1-21 months) for the patients with RVR group (log-rank p = 0.097, Figure 2).Regarding the OS of all of the patients with RVR group and NRVR group, Figure 2 shows a Kaplan-Meier analysis resulting in an estimated average survival of 14 months (range, 1-27 months) for patients with RVR group and 40 months (range, 1-96 months) for the 94 patients with NRVR group (log-rank p = 0.790) (Figure 3).The lateral drainage pathway of rectal cancer is significantly associated with the blood flow of the visceral branches of the IIVs, including the inferior vesicle artery and middle rectal artery. 12aditionally, LLND was performed along the surface of the vascular and the main trunk, while visceral branches of IIVs were usually isolated and preserved. 13,14Vascular ligation was only performed when tumor or node invasion was suspected. 15However, vascular preservation was technically demanding and added the risk of unexpected bleeding.Besides, preserving visceral branches might lead to residual nodal diseases around the distal part of IIVs.Recently, Akiyoshi 16 indicated that lymph node metastasis distal to the infrapiriformis foramen accounted for 44% of all LLNM and was predictive of better disease-free survival (hazard ratio: 0.412; 95% confidence interval: 0.159-0.958,p = 0.039), as compared with metastasis to the more proximal lateral compartment.Furthermore, as the most of nodes in the distal lateral compartment are close to the inferior vesical artery, the routine resection of visceral branches will certainly facilitate the complete removal of lymphatic and adipose tissue in this area. 17erative time and blood loss are two important parameters to demonstrate the quality of operation, especially for difficult procedures including LLND. 18Compared with traditional procedures, LLND with routine resection of the visceral branches could simplify the procedure of isolating vessels.Thus, the total operative time was significantly reduced.Besides, the blood loss for LLND was also significantly less in our study than in previous reports. 19The favorable result in operation time and bleeding in RVR can't be simply attributed to gain in experience over time but also indicates the efficacy and safety of this modified technique.
Fortunately, in our study, we did not observe an increase in intraoperative and postoperative complications in the RVR group.
Urinary and sexual dysfunction was another concern after resection of visceral branches.The incidence could range from 30% to 50% as reported by the measurement of bladder residual volume. 20wever, most patients did not need any intervention and their urinary function would recover.In this study, 9.3% and 5.3% reported that the resection of the bilateral inferior vesicle artery in LLND would significantly harm the urinary function, including increasing the residual volume and the need for catheter reinsertion, compared with unilateral or no resection of the inferior vesicle artery. 21In our practice, we also noticed this problem.Thus, in the later phase, for bilateral LLND, we would try to preserve at least one inferior vesicle artery as possible to reduce the incidence of urinary retention after surgery. 22Besides, the latest data of JCOG0212 study demonstrated patients with the clinical Ⅱ stage could not benefit from prophylactic bilateral LLND. 23Thus, the resection and preservation of the inferior vesicle artery must be carefully planned when bilateral dissection is needed.
The clinical stage of the primary tumors was significantly more advanced in the RVR group than in the NRVR group.However, no significant difference was observed between the two groups for the pathological stage.We thought this might be because over 85% of patients in the RVR group received long-course neoCRT, whereas that was 40% in the NRVR group.Although no significant difference was detected between the two groups in terms of the incidence of LLNM and the number of lateral lymph nodes harvested, the trend indicated that routine resection of visceral branches of IIVs is promising for improving the completeness of LLND.
However, the limitation of this study should also be indicated.The effect of RVR on the recovery of sexual function and whether patients could benefit from long-term survival outcomes are still unclear, and it needs a larger sample size and a longer time of follow-up to illustrate.
In conclusion, LLND with routine resection of visceral branches of IIVs is safe and feasible.It could help improve the lateral lymph node clearance and would not bring additional complications compared to traditional procedures.However, the effect on sexual function and survival outcomes still needs to be further investigated with a longer time of follow-up.

2. 1 |
Patients This study was single-center, case-control designed.Patients who received routine vascular resection of visceral branches of the internal iliac vessels (IIVs) were enrolled in the routine vascular resection group (RVR group), which was prospectively registered from October 2018 to March 2021 (NCT03826862).Patients without routine vascular resection (NRVR group) were respectively reviewed from January 2012 to September 2018, in the prospective colorectal cancer database of the Colorectal Cancer Center, Department of General Surgery, West China Hospital, Sichuan University.
Continuous variables were expressed as the median (range) or mean (SD) and analyzed using a nonparametric Mann-Whitney U test or an independent-sample t test.Ranked data were also analyzed by a nonparametric test.Categorical variables were shown as a number and were analyzed using χ 2 or Fisher's exact tests.A logistic regression analysis was performed to identify the factors associated with postoperative complications.Kaplan-Meier survival curves were used to analyze the cancer-specific outcomes and Cox regression was performed as a multivariable analysis.p < 0.05 were considered to indicate statistical significance.All statistical analyses were performed using the SPSS 22.0 software program.

T A B L E 1
Patient demographics.

T A B L E 3
Postoperative recovery and complications.RVR (n = 75) NRVR (n = 57) p Time to first flatus, day 1
incidence of urinary retention were observed in the RVR and NRVR groups (p = 0.587), respectively.This demonstrated that the routine resection of visceral branches, at least when performed on one side would not bring additional urinary dysfunction.One previous study F I G U R E 3 Overall survival for patients with and patients without with and without routine resection of the visceral branches of internal iliac artery.cens, censored; NRVR, nonroutine visceral branches of the internal iliac vessel resection; RVR, routine visceral branches of the internal iliac vessel resection.
RVR (n = 75) NRVR (n = 57) p Note: Bold values represent that the results are statistically significant.Abbreviations: BMI body mass index; CEA, carcinoembryonic antigen; neoCRT neoadjuvant chemoradiotherapy; neoCT neoadjuvant chemotherapy; NRVR, nonroutine visceral branches of the internal iliac vessel resection; PAS, previous abdominal surgery; RVR, routine visceral branches of the internal iliac vessel resection; SRT, short-course radiotherapy.T A B L E 2 Surgical information.RVR (n = 75) NRVR (n = 57) pAs for pathological outcomes, no significant difference was observed between the two groups in terms of pT stage, pN stage, or pTNM stage.Thirteen (17.3%) and nine (15.8%) patients achieved pathological complete response after neoadjuvant treatment in RVR and NRVR groups, respectively.More neural invasions were observed in the RVR group than that in the NRVR group (p = 0.002).The two groups had comparable incidences of lymphovascular invasion, positive circumferential resection margin, LLNM, and mesorectal node metastasis.The harvested number of lateral lymph nodes and mesorectal nodes were similar between the two groups as well.The number of positive