[612] Identifying Low-Risk and High-Risk T1 Esophageal Adenocarcinoma Based on Pathologic Risk Factors.

Jon Davison, James Luketitch, Michael Gibson, Rodney Landreneau, Katie Nason. UPMC, Pittsburgh, PA

Background: Pathologic evaluation of endoscopic resection (EMR) specimens is a critical aspect of the pre-operative staging evaluation for superficially invasive adenocarcinoma of the esophagus and GE junction (stage T1-EAC), helping to direct appropriate treatment. The purpose of this study is to quantify the risk of lymph node metastasis (LNM) associated with known pathologic risk factors in a series of surgically resected T1-EAC.
Design: We reviewed pathology reports and clinical records from a consecutive series of 232 patients who underwent esophagectomy without induction therapy for T1-EAC at the UPMC between 1996 and 2009.
Results: Tumor size (Tsize), angiolymphatic invasion (ALI), submucosal invasion (stage T1b) and histologic grade (poorly differentiated, G3) were all significant risk factors for LNM on univariate analysis. When controlling for all four covariates, Tsize, ALI and grade were independent predictors of lymph node metastasis, but T stage was not (Table 1).

Table 1

Pathologic Risk Factors		LN Metastasis *
		N/Total (%)	Adjusted OR (95% CI)
T stage	T1b	39/151 (25.8)	1.5 (0.44, 5.1)
	T1a	5/80 (6.3)
Tumor size, cm (median, IQR)	(2.0, 1.0-3.0)		1.4 (1.1, 1.8)
ALI	+	28/44 (63.6)	10.5 (4.3, 26)
	-	15/181 (8.3)
Grade	G3	23/65 (35.4)	3.0 (1.2, 7.2)
	G1 or G2	22/163 (13.5)

* The median number of LN evaluated per patient was 18.5.

Using the cumulative number of risk factors with tumor size of 1.0 cm as a cutoff, a simple, unweighted model shows that as the number of pathologic risk factors increases, there is a progressively higher incidence of regional metastasis (Table 2).

Table 2

Number of Risk Factors	LN Metastasis *
(T1b, ALI, G3, Tsize>1cm)	N/Total (%)
4	18/20 (90.0)
3	14/52 (26.9)
2	8/72 (11.1)
1	4/55 (7.3)
0	1/33 (3.0)

* P<0.001, Chi-squared test


Conclusions: We demonstrate that T stage, when controlling for ALI, grade and tumor size, is not an independent predictor of nodal metastasis and should not be used by itself to determine whether or not a patient can safely be treated with endoscopic resection and ablation therapy. When one accounts for the cumulative number of pathologic risk factors, one can stratify T1-EAC into low-risk and high-risk for LNM. These features can be evaluated pre-operatively and, in combination, be used to select the most appropriate treatment approach (endoscopic vs esophagectomy) and potentially identify patients who will benefit from systemic chemotherapy.
Category: Gastrointestinal

Tuesday, March 1, 2011 8:45 AM

Platform Session: Section E, Tuesday Morning