Back to Annual Meeting Program

Introduction: While minimally-invasive techniques have an integral role in foregut surgery, their optimal use in the resection of gastric neoplasms awaits validation in clinical practice. We hypothesized that conversion of operations to open could be predicted by specific anatomical and pathological factors.
Methods: A retrospective analysis was conducted on patients with attempted laparoscopic resection (n=69) and open resection (n=25) of submucosal neoplasms of the stomach from October 2002 through October 2011. Nonparametric statistical tests were used for comparisons between groups.
Results: Patients were 63 ± 14 years old, 52% male, and had a BMI of 29.5 ± 7.3 kg/m2. Diagnostic endoscopic ultrasound use was greater in the laparoscopic group than the open (86% vs. 64%, p<0.05). Lesions in the laparoscopic and open cohorts included gastrointestinal stromal tumors (71% vs. 88%), leiomyomas (9% vs. 12%), schwannomas (9% vs. 0), and other (11% vs. 0). There were 7 (10%) conversions to open in the laparoscopic group. Posterior location was a predictor of conversion (71% vs. 32%, p<0.01) and selection for an open technique (67%, p<0.01). There were no other predictors of conversion by location of the mass along the stomach. Conversions (29%) and open resections (40%) were more likely to have multivisceral involvement than the laparoscopically-treated patients (2%, p<0.0001). Combined laparoendoscopic approaches were used in 10% of laparoscopic procedures (n=62), which included wedge or sleeve resection in 47 (76%), transgastric wedge resection in 5 (8%), submucosal resection in 3 (5%), midbody gastrectomy in 2 (4%), antrectomy in 1 (2%), and other in 4 (6%). Patients selected for open resection were more likely to require a gastroenteric anastomosis in comparison to patients initially selected for a laparoscopic approach (40% vs. 4%, p<0.0001). Significant differences in tumor size, operative time, morbidity, and length of stay were noted between groups (refer to Table, p<0.001).
Conclusions: Selection for laparoscopic versus open resection appears to be influenced by factors including tumor size, multivisceral involvement, and the need for gastric reconstruction. Conversion to open is also more likely with tumors in a posterior location. Laparoscopic gastric resection has decreased morbidity, operative time, and hospital length of stay and is appropriate in well-selected patients with gastric submucosal neoplasms.

	Laparoscopic (n=62)	Conversion (n=7)	Open (n=25)	p
Symptoms: - - - - -	14 (24%) 23 (39%) 8 (14%) 0 14 (24%)	1 (14%) 2 (29%) 1 (14%) 0 3 (43%)	1 (4%) 8 (32%) 2 (8%) 1 (4%) 13 (52%)	0.01
Neoadjuvant Imatinib	2 (3%)	2 (29%)	6 (24%)	0.005
Preoperative Size (cm)	3.5 ± 1.8	7.7 ± 5.4	9.5 ± 7.1	<0.0001
Pathological Size (cm)	4.0 ± 2.2	7.6 ± 5.0	8.5 ± 6.2	0.0004
Operative Time (min)	145 ± 60	246 ± 84	231 ± 96	<0.0001
Estimated blood loss (mL)	35 ± 70	376 ± 422	373 ± 280	<0.0001
Accordian Severity Score 0 - no complications 1 - minor (wound, foley) 2 - minor (PRBCs, Abx, TPN) 3 - Endoscopic / IR Intervention 4 - Operative Intervention 5 - MSOF 6 - Death	51 (82%) 3 (5%) 4 (7%) 1 (2%) 2 (3%) 0 1 (2%)	3 (43%) 1 (14%) 3 (43%) 0 0 0 0	9 (36%) 4 (16%) 7 (28%) 2 (8%) 1 (4%) 1 (4%) 1 (4%)	0.0003
Hospital length of stay (days)	3.5 ± 3.9	6.4 ± 1.1	8.7 ± 5.8	<0.0001