Endoscopic full-thickness resection using an over-the-scope device for treatment of recurrent / residual colorectal neoplasia: a single-center case series

Between 01/2016 and 11/2018 we performed 30 endoscopic full thickness resections using the FTRD® device (Ovesco, Tübingen, Germany) in 29 consecutive patients (19 male / 10 female; median age 72,7 years / range 21,5-81,6; Table 1). All lesions were < 20 mm in size and were judged difficult to resect because of non-lifting due to scar formation after previous endoscopic resection. Localization of the lesions was in the right colon (n = 18), left colon (n = 9) and rectum (n = 3). The procedure time was measured from the initial marking of the lesion until the final endoscopic control of the clip closure. All procedures were performed by a single endoscopist (F.L.D.) who had received team training with the FTRD® device on isolated pig colon provided by the manufacturer. Informed consent for the procedure was obtained from all patients. Data on endoscopic procedures were prospectively recorded (Clinic WinData, E&L, Erlangen, Germany). Retrospective analysis was done by review of the digital patient charts. The study conforms to the provisions of the Declaration of Helsinki. It was approved by institutional board review of the Gemeinschaftskrankenhaus Bonn and - in accordance with federal regulations (§ 15 of the professional code / ‘Berufsordnung für die nordrheinischen Ärztinnen und Ärzte’) - specific consent to participate in this retrospective case series was waived.

EFTR was carried out under conscious sedation with propofol, fentanyl and prophylactic single shot antibiosis (cefuroxime). After endoscopic evaluation and marking of the lesion (Fig. 1a, b), the FTRD® system was mounted and the endoscope re-advanced to the lesion. Full thickness resection with the integrated snare was accomplished after pulling the tissue into the cap and release of the clip (Fig. 1c, d). The settings of the Erbe Vio 3 electrosurgical unit (Erbe Elektromedizin, Tübingen, Germany) were forced coagulation 1.0 for initial marking of the target lesion and endocut Q 1.0 for cutting the lesion. After retrieval of the specimen and removal of the FTRD® system a repeat endoscopy was performed to control the positioning of the clip (Fig. 1e) and to rule out any possible injury that might have happened during the passage of the FTRD® device to the target lesion.

The specimens were gently stretched and then fixed on corkboard. Histopathology was performed with specific attention to the resection margins to confirm a complete resection (Fig. 1f). Patients with early cancer were presented in the interdisciplinary tumor board.

Patients had a liquid diet for the first 6 h after the procedure and were allowed to consume a soft diet the day after. We performed clinical visits and laboratory controls on the evening after the procedure as well as on the first post-interventional days and in any case of suspected adverse events. The majority of patients could be discharged on the second post-interventional day. We recommended follow up endoscopies in accordance with the current German S3 guideline [3].

We defined ‘technical feasibility’ as reaching the target lesion and ‘R0 resection’ if both vertical and lateral margins were free of adenoma or cancer. A ‘perforation’ was diagnosed if a complete transmural lesion was identified during the procedure or if clinical and radiological signs of perforation were demonstrated after the procedure. A ‘major bleeding’ was defined as a loss of ≥3 hemoglobin units with signs of gastrointestinal bleeding after the completion of the intervention.

Group comparisons were done with Fisher’s exact test for categorical (https://​www.​socscistatistics​.​com) and students t-test (Microsoft Excel for Mac 2011) for continuous variables.

EFTR was technically feasible in 28/30 patients (93,3%) with a median procedure time (marking to full thickness resection) of 34,5 min (range 11–120). In two patients the FTRD device could not be advanced to the lesion due to sigmoid (inflammatory) strictures. Both full thickness and R0 resection rates were of 80% (24/30) on intent to treat analysis and 85,7% (24/28) on per protocol analysis (Fig. 2, Table 2).

In four EFTR interventions we could not achieve R0 resection. The endoscopic resection was judged complete in two of them, but histology showed positive resection margins. Follow up has been done for one patient and did not show residual neoplasia. In two other patients we had experienced difficulties in mobilizing the colonic wall into the resection cap due to fibrosis. The resection was diagnosed as incomplete at the final endoscopic control of the resection area. One patient had additional surgery due to high grade intraepithelial neoplasia; follow up data are not available for the other (Fig. 2).

The final histology showed advanced lesions in 12 patients including four carcinomas that had been diagnosed beforehand for possible curative treatment (Table 3). Two of these were evaluated as low risk lesions and - according to the German S3 guideline [3] – the patients were left without additional surgery. The two other lesions had been classified as potentially deep submucosal invasive according to the NICE classification [20] and the patients had been informed about the possibility of a non-curative resection. Histopathology showed pT2, G2, Pn0, L0, V0 - R0 in both cases and surgery was recommended. No lymph node metastasis was found after surgical resection in one patient; the other refused additional treatment and is without recurrence or metastasis after 14 months of follow-up.

Endoscopic follow-up data are available for 12 patients with one adenoma recurrence despite previous R0 EFTR resection after intial piecemeal EMR of a large rectal lesion (Fig. 2).

We observed one major complication, a delayed perforation the day after EFTR necessitating emergency surgery. In addition, three patients reported minor, self-limiting hematochezia and one patient developed fever without abdominal pain after EFTR. The 30-day mortality rate was 0% (Table 2).