Impact of Procedure Time of Preceding Endoscopic Submucosal Dissection on the Difﬁculty of Laparoscopic Rectal Surgery

Objetive: In this study, we addressed which factors, including endoscopic submucosal dissection (ESD)–related parameters, affect the difﬁculty of laparoscopic rectal surgery. Summary of background data: Endoscopic treatment for gastrointestinal cancer can cause inﬂammation, edema, and ﬁbrosis formation in the surrounding tissue. Recently, we reported that preceding endoscopic treatment increased the volume of intraoperative blood loss and slightly prolonged the operative time of laparoscopic surgery for rectal cancer. Methods: We retrospectively reviewed 24 consecutive patients who underwent ESD followed by laparoscopic surgery for rectal cancer in our hospital. Short-term surgical outcomes were evaluated by intraoperative blood loss and operative time for laparoscopic surgery. The correlations between the surgical outcomes and preoperative parameter were analyzed by multiple linear regression analyses. Results: The patient cohort comprised 12 men and 12 women. The median distance between primary cancer and anal verge was 7 cm. The median procedure time of ESD was 120 minutes (21 available cases). Based on multiple linear regression analyses, the ESD procedure time ( P ¼ 0.007) and tumor location from the anal verge ( P ¼ 0.046) were independently predictive of intraoperative blood loss. On the other hand, only tumor location was found to be an independent predictor of surgical time ( P ¼ 0.014). Conclusions: A long session of ESD for rectal cancer may make subsequent laparoscopic surgery difﬁcult based on intraoperative blood loss.

E ndoscopic treatments, such as endoscopic mucosal resection and endoscopic submucosal dissection (ESD), have been developed as minimally invasive techniques for the en bloc removal of early cancer of the gastrointestinal tract, including the colon and rectum. [1][2][3][4] However, some patients with early colorectal cancer are referred for additional surgery after endoscopic treatment because of histologic findings suggestive of possible metastasis in regional lymph nodes. 5 We recently demonstrated that preceding endoscopic treatment increased the volume of intraoperative blood loss and slightly prolonged the operative time for laparoscopic surgery for early rectal cancer. 6 Electrocoagulation injury to the rectal wall induced by endoscopic treatment may lead to transmural burn, causing inflammation of the mesorectum and subsequent fibrosis in the surrounding tissue, which is considered a possible explanation for the results. ESD usually takes longer than endoscopic mucosal resection 4,7 and is therefore more likely to cause deep mural injury, leading to post-ESD electrocoagulation syndrome. 8 Moreover, a long procedure time was reported to be a risk factor for perforation during colorectal ESD. 9 Collectively we hypothesized that the ESD procedure time is associated with the extent of inflammation and increases the volume of blood loss and/or operative time during additional laparoscopic rectal surgery.
In this study, we investigated which factors affect the level of difficulty of laparoscopic surgery after ESD for early rectal cancer, particularly focusing on ESD-related parameters.

Patients and Methods
We retrospectively reviewed consecutive patients with early rectal cancer who underwent ESD and subsequent surgery via a laparoscopic approach between November 2012 and June 2020 at the Department of Surgical Oncology, The University of Tokyo Hospital. Patients in whom ESD was discontinued for technical reasons during the procedure were included, whereas patients treated by abdominoperineal resection, those who underwent combined resection of multiple colorectal cancers, and those who needed conversion to open laparotomy were excluded. The study was approved by the ethics board of the University of Tokyo (3252-10). Written informed consent was obtained from all patients.
All endoscopic procedures were performed by experienced endoscopists in our hospital and affiliated hospitals. The indications and procedure of ESD for rectal lesions were described previously. 2,6 Pathologic findings of ESD specimens indicating noncurative, namely, oncologically insufficient treatment, included cancer-positive margin, extensive submucosal invasion (!1 mm; pT1b), lymphatic and venous infiltration, dedifferentiated component, and grade 2 or 3 budding based on the Japanese Society for Cancer of the Colon and Rectum guidelines. 5 All surgical procedures were performed as described previously. 6 Briefly, low tie of the inferior mesenteric artery was performed preserving the left colic artery via the medial-to-lateral approach. The splenic flexure was taken down when required. Then, sharp dissection in front of the prehypogastric nerve fascia and behind Denonvilliers fascia was performed based on total mesore ct a l e x ci s i o n . I n a n t e r i o r re s e c t i o n , a defunctioning stoma was created at the surgeon's discretion considering the patient's comorbidities and risk of anastomotic leakage, such as an incomplete doughnut after stapled anastomosis and positive findings on air insufflation. A diverting stoma was created in all patients undergoing intersphicteric resection.
Baseline variables age, sex, body mass index, ESD procedure time, interval between ESD and surgery, tumor location from the anal verge, tumor size, serum hemoglobin level, prothrombin time-international normalized ratio (PT-INR), and activated partial thromboplastin time (APTT) before surgery were retrieved from medical charts. The normal ranges of PT-INR and APTT are 0.85 to 1.15 and 25.5 to 36.1 seconds, respectively, in our hospital. The main outcomes of interest in this study were operative time and estimated volume of blood loss during laparoscopic surgery. Other clinical and pathologic outcomes, such as perioperative blood transfusion, postoperative complications graded according to the Clavien-Dindo classification, 10 postoperative hospital stay, pathologic depth (pT), and lymph node metastasis (pN), were reviewed.
To investigate associations of preoperative variables with the surgical outcomes, multiple linear regression analysis was performed with a stepwise (forward selection/backward elimination) method (significance level to enter ¼ 0.25, significance level to stay ¼ 0.1). All statistical analyses were conducted using the JMP 15.1 software (SAS Institute Inc, Cary, North Carolina). A P value below 0.05 was considered to indicate significance.

Results
Twenty-eight patients underwent salvage surgery via a laparoscopic approach after ESD in the study period. Becauase 2 patients underwent abdominoperineal resection, 1 patient had synchronous sigmoid colon cancer, and 1 patient needed conversion to open laparotomy, 24 patients were analyzed in the current study.
The characteristics of the patients (12 men and 12 women; median age: 63 years) are summarized in Table 1. The ESD procedure time was available for 21 patients, which ranged from 28 to 720 minutes with a median of 120 minutes. The primary tumor was located 7 cm (median) from the anal verge. Two patients had chronic antithrombotic therapies that were stopped before surgery. The preoperative PT-INR and APTT were normal in all patients. Additional surgery was performed a median of 62 days after ESD.
The intraoperative and postoperative outcomes are summarized in Table 2. The median blood loss was 75 mL, and the median surgical time was 341 minutes. A stoma was created in 16 patients. Seven patients developed grade 2 or higher postoperative complications that were treated conservatively, except for leg compartment syndrome. There was no perforation in the rectal wall during surgery and no anastomotic leakage. Patients were discharged around 2 weeks after surgery. Two patients in whom ESD was discontinued and 1 with a positive vertical margin received a diagnosis of pT2 tumors. Regional lymph node metastases were detected in 5 patients (21%). During the follow-up period (medi-an, 35 months), only 1 patient developed lung metastases after rectal surgery, which were removed by video-assisted thoracoscopic surgery. There were no deaths.
To address relationships between clinicopathologic factors and surgical outcomes, multiple linear regression analyses were performed. Based on stepwise linear regression analysis, the ESD procedure time and tumor location from the anal verge (estimate, 0.41 mL/min, P ¼ 0.007; and estimate, À12.66 mL/cm, P ¼ 0.046, respectively) were independent predictors for intraoperative blood loss (Table 3), whereas age, sex, body mass index, hemoglobin level, and treatment interval were not. By a similar analysis, the optimal model to predict surgical time included only tumor location (estimate: À10.14 min/cm, P ¼ 0.014; Table 4). In these analyses, it was confirmed that the residuals followed a normal distribution (data not shown).

Discussion
Our recent study shed light on the adverse effects of preceding endoscopic therapies on laparoscopic rectal surgery in terms of intraoperative bleeding. 6 In that study, the median volume of blood loss was larger in patients treated by ESD than by endoscopic mucosal resection (80 versus 20 mL), 6 suggesting a significant impact of cautery duration during endoscopic treatment on the extent of inflammation in the mesorectum and beyond. These findings were an impetus to the current study, and we found a positive correlation between the ESD procedure time and estimated volume of blood loss during laparoscopic rectal surgery. The ESD procedure time reflects its technical difficulty. In general, the ESD procedure time increases according to tumor size. In a study of specialized hospitals in Japan, the average ESD procedure time was 129 minutes for tumors !40 mm, whereas it was 66 minutes for tumors of 20 to 29 mm. 11 In addition, a long ESD procedure time is often required for tumors with severe submucosal fibrosis, 12 which may be associated with massive invasion into the submucosa or other histologic characteristics, for example, desmoplastic reaction. 13 The median ESD procedure time for our patients who were referred for salvage surgery (120 minutes) was longer than that reported by a systematic review (75 minutes), although it ranged widely in both our cohort and the review (28-720 minutes and 5-600 minutes, respectively). 14 We consider it important to not only distinguish tumors with out-of-indication features from those of Tis or those invading the submucosal layer less than 1 mm as pre-ESD diagnosis, but also to pay attention to findings of possible massive submucosal invasion, such as muscle-retracting sign during ESD, considering the correlation between a long session of ESD and the difficulty of subsequent laparoscopic rectal surgery demonstrated in the current study.
After neovascularization that occurs as a pathologic response to inflammation, fibrosis generally develops at a delayed stage in the healing process. 15 Thus, we initially hypothesized that the long interval between ESD and surgery will affect short-term surgical outcomes. However, we observed no significant effects of the treatment interval on intraoperative blood loss or surgical time. This suggested that the timing of salvage surgery does not matter in terms of the difficulty of laparoscopic rectal surgery, although our findings may contain type II errors.
In addition to the ESD procedure time, the distance between the tumor and the anal verge was significantly associated with increased blood loss during laparoscopic rectal surgery after ESD in the current study. It was also correlated with a longer time for laparoscopic rectal surgery. Previous studies in tertiary hospitals also reported that a low tumor location is associated with difficult laparoscopic surgery when evaluated by these outcome measures in rectal cancer patients who did not undergo preceding ESD. 16,17 We recognize that the current study has several limitations. First, this was a retrospective study conducted at a single hospital with a limited number of surgical patients. Another important limitation of the study is that different endoscopists performed ESD; laparoscopic surgery was also performed by many doctors, although they formed board-certified expert surgeon teams. Moreover, we were unable to assess the extent of mesorectal inflammation in other patients who underwent ESD for rectal cancer without additional surgery. We did not evaluate the severity of inflammation in the surgically resected specimens histologically because regional lymph nodes were removed from the mesorectum as routine procedure for pathologic examination in Japan. Lastly, the findings obtained from Japanese patients may not necessarily reflect those in patients in Western countries because of differences in patient background, for example, body mass index distribution.
To conclude, the current study revealed that the procedure time of preceding ESD for rectal cancer is associated with increased blood loss, one of the surrogate markers of the difficulty of laparoscopic surgery. Doctors should be aware of the impact of ESD on salvage surgery when ESD results in noncurative treatment. It is therefore important to communicate well among endoscopists, surgeons, and pathologists before initiating treatment, and to accurately diagnose the tumor depth even during ESD.