Case-Matched Analysis Comparing Endoscopic Submucosal Dissection and Surgical Removal of Difficult Colorectal Polyps

Introduction

Endoscopic submucosal dissection (ESD) has been a well-established method for removing difficult colorectal polyp. The technique was described early in 1999. ¹ Nowadays, international guidelines recommended the use of ESD, regardless of the size, to remove lesions which cannot be resected easily or completely by simple endoscopic mucosal resection (EMR).^2,3 Traditionally, we have to rely on surgery, either colectomy or transanal excision, to deal with those polyps that are unable to be removed by EMR. The reports comparing ESD versus surgery for difficult polyps were scarce. This study aims to demonstrate how patients have been benefited from ESD.

Materials and Methods

During the period between January 2013 and December 2016, all patients who underwent second treatment in our center after failure of removal of colorectal polyps by EMR were reviewed. Outcomes between two groups of patients who underwent ESD and surgery were studied. The patients were matched by their age, sex, and polyp characteristics in a 2:1 ratio. Polyp size was recorded according to the pathology report.

Patients who had difficult colorectal polyps, which were not suitable for EMR, were subsequently managed by ESD or surgery. In the surgery group, transanal excisions were performed for lower rectal lesion. Other than that, the lesions were removed by laparoscopic colectomy. In the ESD group, the procedures were performed by four of our colorectal specialist surgeons. The procedures were performed in endoscopy unit under conscious sedation. Support from anesthetist was not required. ESD was performed in a classical way with initial submucosal injection of hyaluronic acid, followed by submucosal dissection by either needle type or scissors type ESD knife. All patients were admitted half day before surgery or ESD for preoperative workup.

Patient outcome, including length of stay, 30-day readmission, complication, and reintervention rate, as well as any local recurrence, was also reviewed.

With the assumption of nonparametric distribution, categorical data were analyzed by Fisher's exact test, while continuous data were analyzed by Mann–Whitney U test.

Result

After matching, we have made a review on 79 patients who underwent ESD and 38 patients who underwent surgery. Age, sex, and polyp characteristics were matched as shown in Table 1. Worthwhile to mention, there were 13 malignant polyps in the ESD group. Formal surgery was indicated in all of these cases because of submucosal invasion or inadequate margin. However, 4 of them did not have surgery because of patient refusal, high anesthetic risk, or the presence of incurable distant metastasis. In the 9 cases with surgery done, 1 of them had residual adenoma (11%), 2 of them had residual adenocarcinoma (22%), and 6 of them had no residual disease (67%). None of the resected specimens showed presence of lymph node metastasis.

Outcome comparing ESD versus surgery was shown in Table 2. A statistically significant difference in mean length of stay was observed. The length of stay was 4.2 days for ESD group and 8.7 day for surgery group (P < .001). There was significantly less than 30-day readmission in ESD group as well (0% versus 7.6%, P = .001).

Significantly more complications (7 out of 38 patients) were also observed in the surgery group (0% versus 18.4%, P < .001). Four patients suffered from minor complications, including wound infection and urinary tract infection. One of them was complicated with an intraabdominal abscess, which was managed conservatively. Two remaining patients required reintervention. One patient needed endoscopic examination for postoperative rectal bleeding, while the other one suffered from anastomotic leakage requiring reoperation. In this series, no complication was reported in the ESD group.

We did not make comparison on the local recurrence as we believed that this is inappropriate, because some of the lesions were removed by bowel resection. Fifty-nine patients in the ESD group have undergone surveillance colonoscopy. No local recurrence was found on surveillance. In the surgery group, three local recurrences were observed. All of them were rectal lesions, which were initially treated by transanal excision. Two of them were malignant polyps that patients refused our recommendation to have formal surgery. One of them was local recurrence of benign polyp.

Discussion

Since the introduction of colorectal screening program in our locality, we were encountering more and more advanced colorectal polyps, which could not be removed by simple EMR. ESD has been a fantastic technique in dealing with advanced colorectal polyps. Actually, ESD should also be considered for those lesions that are borderline cases for EMR, as systemic review and meta-analysis have already confirmed that ESD provides better en bloc resection rate and less recurrence.^4,5 One of the concerns of ESD would be the cost of those specialized equipments and extra time cost. A recent study indeed showed that selective use of ESD is cost-effective than EMR in suspicious polyps, as they prevented additional surgeries. ⁶ Studies comparing EMR and ESD were bound to be biased at the patients selection level. ⁷ Yet, the advantage of using ESD on difficult polyps is undeniable and is recommended by international guidelines.^2,3

Whether ESD is a better modality when compared with traditional surgery is not well proved. Study comparing ESD and minimal invasive surgery was scarce. So far, we could only find data from one Japanese center.^8,9 Their retrospective study showed that ESD is associated with less complication rate, which echoed the result of our series. They also reported a favorable en bloc resection rate and curative rate with ESD. Yet, we would like to cast doubt on this result. We believed that ESD should not give superior curative rate when compared with properly conducted formal colectomies. In our study, we demonstrated that ESD is associated with less hospital stay and readmission rate. This would be another extra merit for ESD. This did not only benefit our patients but also relieve the overwhelming hospital occupancy. Our patient was admitted in the afternoon before the procedure. The hospital stay is expected to be shorter if we adopt day admission later. Moreover, Japanese has suggested clinical pathway after ESD, which can further shorten length of stay. ¹⁰ We did not compare the cost between ESD and formal surgery. Yet, we can be pretty sure that the cost of ESD would be much lower than formal surgery. First, the procedure is performed in endoscopy unit under sedation instead of in operation theater. The input from anesthetist was not required. Second, the length of stay was much shorter than the surgery group. There was a study comparing EMR versus surgery for lateral spreading tumor, which has already demonstrated that the endoscopic treatment is more cost-effective than surgery. ¹¹ With all of the above discussed benefits, ESD should be a more preferred choice to deal with difficult colorectal polyps whenever technically feasible.

It is not reasonable to compare local recurrence rate between ESD and colectomy as bowels were resected in surgery group. A more meaningful comparison would be ESD versus transanal type of resection. Yet, we did not have enough cases in this series for comparison. A local series in 2011 has shown that ESD is associated with better short-term outcome in terms of morbidity and time to ambulation when compared with transanal local excision. ¹² Another small-scale study from Brazil reported similar en bloc resection rate and hospital stay comparing ESD and transanal endoscopic microsurgery (TEM). ¹³ Yet, the optimal choice for treatment of rectal lesions is still controversial. A meta-analysis concluded that TEM achieves a higher R0 resection rate when performed in full-thickness manner, significantly reducing the need for further abdominal treatment. The selection of rectal lesion should remain case by case. The size of the lesion, distance from anal verge, estimated chance of submucosal invasion by endoscopic assessment, and the availability of expertise are factors that affect our decision-making. ESD is an appropriate and safe modality for removing selected rectal lesions.

There were some concerns on the use of ESD to handle malignant colorectal polyps. Indeed, whether formal surgery is required can be determined by the presence of pathological risk factors. As most of the lesions requiring ESD are sessile, the risk of lymph node metastasis can be determined by the level of submucosal invasion according to the Kikuchi classification. ¹⁴ It classified lesions into sm1 (upper third of submucosa), sm2 (middle third of submucosa), and sm3 (lower third of submucosa) submucosal invasion. The chance of lymph node metastasis is relatively high for sm3 lesions which can be up to 25%. ¹⁵ Other pathological risk factors include poor differentiation, presence of lymphatic invasion, positive resection margin, and piecemeal removal.^15,16 For low-risk lesions without the above mentioned risk factors, ESD is oncologically safe.

There were few limitations in our study. Being retrospec tive in, there could be significant selection bias. Also, the matching incidentally selected out some of our ESD cases, which end up in complication. Hence, making our ESD series looked apparently flawless. There were complicated cases in the surgery group with much prolonged hospital stay. This may have significant influence on the statistical result as the sample size was small. We did not compare the procedure time as some of the ESD cases were done during our learning curve period. The procedure time was expected to be longer in those cases. Currently, we can complete two ESD cases in a 3-hour session most of the time. We will report the procedure time in our next review.

Conclusion

ESD is a well-established method to remove colorectal polyps that are not suitable for EMR. When compared with traditional surgery, ESD had the benefit of shortening hospital stay and less postprocedure complication.