OUTCOMES OF CRYOABLATION FOLLOWING ENDOSCOPIC SUBMUCOSAL DISSECTION OF GASTROESOPHAGEAL TUMORS IN SUBJECTS WHO INITIALLY REFUSE DEFINITIVE THERAPY AND ARE AT HIGH RISK FOR LYMPH NODE METASTASIS

Background:
With the increased detection of early gastroesophageal tumors (GET), the role of endoscopic submucosal dissection (ESD) has been expanding as an alternative to surgery. ESD can provide en bloc and ideally R0 specimens regardless of tumor size, as well as accurately stage the tumor pathology via examination of the entire resected tumor.
The risk of lymph node metastasis (LNM) has significant implications for treatment and is determined by factors that can be assessed after evaluation of the resected specimen. Factors increasing the risk of LNM include poorly differentiated grade, status of lymphovascular invasion (LVI), depth of tumor into the submucosa (T1b), and R1 resection status.² Minimal literature exists addressing outcomes and recurrence rates in resection of GET when ESD is followed by cryoablative (CYA) procedures. In this study we set out to compare the risk of recurrence for GET removed via ESD followed by CYA in subjects who have a high-risk of LNM when compared to a low-risk group.
Methods:
We included subjects who underwent ESD followed by CYA of GETs at an academic tertiary referral center in the United States. The subjects were analyzed in two groups: those with high-risk lesions (n=24) and those with low-risk lesions (n=19). A lesion was considered high risk if pathology noted LVI, poorly differentiated grade, depth of invasion was T1b, or tumor margins noted an R1 resection. Low-risk lesions had none of these factors. All high-risk subjects underwent discussion at a multidisciplinary tumor board and were offered surgery and chemo/radiation, but initially refused. We analyzed demographic, procedural, histopathologic characteristics and follow-up data including rate of recurrence using student's t-test, categorical variables, a one-tailed Fischer’s exact testing, and SPSS software.
Results:
Forty-three patients underwent ESD for GET. In the high-risk group, the average age was 65, and 71% were male; this was not statistically different compared with the low-risk group (Table 1). En bloc resection rate was 70%, and R0 resection rate was 50% in the high-risk group versus 100% in the low-risk group (p=0.01 and 0.0002). The mean follow-up for the high-risk group was 309 v. 399 days in the low-risk group (p=0.15).
Two patients in the high-risk group had evidence of recurrence on follow-up whereas zero patients had evidence of recurrence in the low-risk group (p =0.07). Two subjects in the high-risk group eventually underwent surgery. Adverse event rates were no different in the two groups (Table 2).
Conclusions:
ESD resection of high-risk versus low-risk GET who refused traditional therapy, followed by CYA has no difference in recurrence rates during the follow-up period. Though recurrence rates were not statistically different, larger sample sizes may be required to determine the effects of CYA after ESD for GET completely.