PPS17. Predicting regional lymph node metastasis in goblet cell adenocarcinoma of the appendix using an internally validated model: can a right hemicolectomy be omitted?

Thursday, March 5, 2026

10:00 AM - 10:30 AM MST

Location: HUB Poster Park - West Hall 1-3

Poster Presenter(s)

Sara P. Ginzberg, MD, MS (she/her/hers)

Fellow, Complex General Surgical Oncology
MD Anderson Cancer Center
Houston, Texas, United States

Co-Author(s)

SE

Sacha El Khoury, MD

Research fellow
MD Anderson Cancer Center
Houston, Texas, United States
AY

Abdelrahman Yousef, MD

Resident
Department of Gastrointestinal Medical Oncology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
SL

Sun Mi Lee, MD

Associate Professor
Department of Anatomical Pathology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
WF

Wai Chin Foo, MD

Professor
Department of Anatomical Pathology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
MT

Melissa Wainwright Taggart, MD

Professor
Department of Anatomical Pathology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
PM

Paul F. Mansfield, MD

Professor
Department of Surgical Oncology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
KF

Keith F. Fournier, MD

Associate Professor
Department of Surgical Oncology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
Beth A. Helmink, MD, PhD (she/her/hers)

Assistant Professor
Department of Surgical Oncology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States
John P. Shen, MD (he/him/his)

Assistant Professor
Department of Gastrointestinal Medical Oncology, The University of Texas MD Anderson Cancer Center
Houston, Texas, United States
Yun Song, MD (she/her/hers)

Assistant Professor
Department of Surgical Oncology, University of Texas MD Anderson Cancer Center
Houston, Texas, United States

Introduction: Goblet cell adenocarcinoma (GCA) of the appendix is rare; thus, our understanding of its behavior and optimal management has been limited. The goal of this study was to evaluate the risk of regional lymph node (LN) metastasis (LNM) in GCA to inform surgical decision-making.

Methods: Using the National Cancer Database, we identified patients who were diagnosed with GCA (2010−2022) and underwent a right hemicolectomy (RHC) with ≥12 LNs evaluated. The primary outcome was the rate of LNM. Multivariable logistic regression was used to assess the associations between baseline clinicopathologic characteristics and LNM. The model was internally validated using a bootstrap method with 200 samples. From this model, a three-tier risk classification was created. The secondary outcome was overall survival (OS) according to LN status.

Results: Of 3,404 patients, 816 (24.0%) had LNM. Most patients had pT3 (1,907, 56.0%) or pT4 (1,307, 30.5%) disease. When histologic grade was available (2,348, 69.0%), 723 (30.8%) patients had G2 and 730 (31.0%) had G3 tumors. A RHC was performed in the setting of distant metastatic disease in 429 (12.6%) patients.
On multivariable analysis, age ³64 years (OR 1.34, p=0.005), Black race (OR 1.77, p< 0.001), Hispanic ethnicity (OR 1.89, p=0.033), increasing T stage (T3 OR 2.25, p=0.085; T4 OR 10.9, p< 0.001), increasing grade (G2 OR 1.31, p=0.100; G3 OR 3.98, p< 0.001), and presence of lymphovascular invasion (OR 5.82, p< 0.001) were significantly associated with LNM. The optimism-corrected area under the curve for the model was 0.856.
Based on the model, 196 (5.8%) patients were considered low risk (£5% LNM rate), 1,225 (36.0%) intermediate risk (5-10% LNM rate), and 1,983 (58.3%) high risk ( >10% LNM rate). The actual and predicted rates of LNM were: 2.0% and 2.0% (95% CI 1.9-2.0%) for low-risk, 6.2% and 6.5% (95% CI 6.2-6.8%) for intermediate-risk, and 37.1% and 36.9% (95% CI 35.8-38.1%) for high-risk patients.
Of 2,861 patients with follow-up data, LNM was associated with decreased 5-year OS rate (44.2% vs. 88.9%, log-rank p< 0.001), even in the presence of distant metastasis (16.2% vs 31.5%, log-rank p=0.002).

Conclusions: Regional LNM occurs in nearly one-quarter of patients with GCA, and only 6% of patients in a national cohort can be considered low risk. Accurate histopathologic assessment by an experienced gastrointestinal pathologist is crucial to defining risk. More evidence is needed before completion RHC can be omitted in a subset of patients with GCA.

Learning Objectives:

Understand the rate of lymph node metastasis in goblet cell adenocarcinoma of the appendix
Describe the risk factors for lymph node metastasis, as well as the association between regional disease and survival
Improve shared decision-making for patients with goblet cell adenocarcinoma of the appendix who are considering right hemicolectomy