Back to 2018 Program and Abstracts

---

COMPARISON OF PERIOPERATIVE CHEMOTHERAPY VS. POSTOPERATIVE CHEMORADIATION THERAPY FOR ADENOCARCINOMA OF THE GASTRIC CARDIA: AN ANALYSIS OF THE NATIONAL CANCER DATABASE
Mihir M. Shah^*1, Rachel E. NeMoyer², Stephanie H. Greco¹, Yong Lin⁵, Miral S. Grandhi¹, Darren Carpizo¹, Parisa Javidian³, Salma K. Jabbour⁴, David A. August¹, H. Richard Alexander¹, Steven K. Libutti¹, Timothy J. Kennedy^1
1Surgical Oncology, Rutgers Cancer Institute of New Jersey, New Brunswick, NJ; ²General Surgery, Rutgers Robertwood Johnson University Hospital, New Brunswick, NJ; ³Pathology, Rutgers Robertwood Johnson University Hospital, New Brunswick, NJ; ⁴Radiation Oncology, Rutgers Cancer Institute of New Jersey, New Brunswick, NJ; ⁵Biostatistics, Rutgers School of Public Health, New Brunswick, NJ

Introduction
Perioperative chemotherapy (PEC) and postoperative chemoradiation therapy (POCR) are considered standard of care for patients with resectable gastric adenocarcinoma (GC) based on level 1 evidence. However, no significant evidence exists directly comparing PEC and POCR for gastric cardia adenocarcinoma. The purpose of this study was to compare these treatment modalities in patients with surgically resectable gastric cardia adenocarcinoma to determine the impact on overall survival (OS).

Methods
A retrospective review of patients undergoing definitive surgery for GC was performed using the National Cancer Database (2004-2014). Only patients with gastric cardia as the primary tumor site were included. The difference in OS in patients who received PEC or POCR was compared. A second analysis identified patients who received optimal therapy, defined as - negative surgical margins, radiation doses between 4500-5040cGy, radiation to the esophagus and stomach only, and chemotherapy started within 90 days of surgery. We utilized Cox proportional hazard model for our multivariable analysis.

Results
We identified 62,931 patients who underwent definitive surgery for GC. 1,932 patients with gastric cardia adenocarcinoma underwent PEC or POCR. No difference in OS was noted between the patients receiving PEC and POCR (p=0.486; HR 1.069) [Table 1]. The median OS was 35.2 months for all patients (PEC 38.47 months, POCR 34.04 months). No difference in OS was maintained, after selecting patients who received optimal therapy (n = 862) (p=0.7658; HR 1.035) [Table 2]. The median OS was 39.9 months for all patients (PEC 45.01 months, POCR 38.83 months). Age, grade, Charlson score and lymph node positive ratio were statistically significant predictors of OS.

Conclusion
After adjusting for covariates in patients who underwent definitive surgery for adenocarcinoma of the gastric cardia, there is no difference in OS in patients receiving PEC and POCR. Even after selecting patients who received optimal therapy, there is no difference in OS between PEC and POCR. The median OS is superior in patients receiving optimal therapy, regardless of the treatment type.

Table 1. Hazard Ratio for Each Factor

PARAMETER		p-value	Hazard Ratio	95% Hazard Ratio Confidence Limits
PEC	1	0.4860	1.069	0.886	1.289
SEX	1	0.8961	0.989	0.832	1.175
RACE	1	0.8584	0.960	0.614	1.501
RACE	2	0.2574	0.737	0.435	1.249
GRADE	1	0.0180	0.521	0.304	0.894
GRADE	2	0.0239	0.628	0.419	0.940
GRADE	3	0.0851	0.707	0.476	1.049
CHARLSON SCORE	0	0.0777	0.780	0.592	1.028
CHARLSON SCORE	1	0.2332	0.834	0.619	1.124
AGE		< 0.0001	1.019	1.012	1.025
TUMOR SIZE		0.3279	1.001	0.999	1.002
LYMPH NODE POSITIVE RATIO		< 0.0001	5.655	4.506	7.095

Table 2. Hazard Ratio for Each Factor

PARAMETER		p-value	Hazard Ratio	95% Hazard Ratio Confidence Limits
PEC	1	0.7658	1.035	0.823	1.303
SEX	1	0.7055	1.060	0.783	1.436
RACE	1	0.5056	0.787	0.388	1.594
RACE	2	0.1769	0.558	0.239	1.301
GRADE	1	0.0611	0.436	0.183	1.039
GRADE	2	0.0759	0.538	0.272	1.067
GRADE	3	0.1632	0.619	0.316	1.215
CHARLSON SCORE	0	0.9471	0.985	0.622	1.559
CHARLSON SCORE	1	0.9711	1.009	0.617	1.652
AGE		0.0050	1.015	1.004	1.025
TUMOR SIZE		0.5473	1.001	0.998	1.004
LYMPH NODE POSITIVE RATIO		< 0.0001	5.822	3.994	8.485

---

Back to 2018 Program and Abstracts