27: The Prognostic Value of Pre- and Post-Neoadjuvant Chemotherapy Serum Carcinoembryonic Antigen Levels in Patients Having Resection of Colorectal Liver Metastases

Thursday, March 23, 2023

11:16 AM – 11:26 AM

Location: Ballroom C

Abstract Presenter(s)

Yi Song, MD, MS (she/her/hers)

Clinical research fellow
Memorial Sloan Kettering Cancer Center
Brooklyn, New York, United States

Disclosure(s): No financial relationships to disclose

First Author(s)

Yi Song, MD, MS (she/her/hers)

Clinical research fellow
Memorial Sloan Kettering Cancer Center
Brooklyn, New York, United States

Disclosure(s): No financial relationships to disclose

Author(s)

HK

Hannah Kalvin, MSPH

Assistant Research Biostatistician
Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
Alice C. Wei, MD MSc FRCSC FACS (she/her/hers)

Associate Attending Surgeon
Memorial Sloan Kettering Cancer Center
New York, New York, United States

Disclosure information not submitted.
KS

Kevin C. Soares, MD

Hepatobiliary Surgeon
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
TK

T. Peter Kingham, MD

Director of Global Cancer Disparities Initiatives
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
JD

Jeffrey A. Drebin, MD, PhD

Chair, Department of Surgery; Murray F. Brennan Chair in Surgery
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
WJ

William R. Jarnagin, MD

Chief, Hepatobiliary Service; Leslie Blumgart Chair in Surgery
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
MG

Mithat Gonen, PhD

Chief of the Biostatistics Service
Department of Biostatistics, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.
MD

Michael I. D’Angelica, MD

Enid A. Haupt Chair in Surgery
Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, United States

Disclosure information not submitted.

Disclosure(s):

Yi Song, MD, MS: No financial relationships to disclose

Introduction: Serum carcinoembryonic antigen (CEA) is a valuable pre-operative prognostic variable in patients with resectable CRLM. However, in neoadjuvant chemotherapy (NAC) recipients, the prognostic impact of CEA has not been well-studied. This study evaluated the roles of CEA before, after and during NAC in predicting survival outcomes after CRLM resection.

Methods: A retrospective review was performed of patients who received NAC followed by curative-intent resection of CRLM from 1991 to 2016. The prognostic values of three CEA variables, defined as pre-NAC CEA (+/- 30 days of CRLM diagnosis), post-NAC CEA (pre-op CEA), and ΔCEA (difference between pre- and post-NAC CEA), were studied. Univariate (UV) analysis and Classification and Regression Tree (CART) model were used to evaluate the association between the CEA variables and overall survival (OS). The optimal cut-off values of the significant variables were generated by the CART model. Survival estimates were derived using the Kaplan-Meier methodology.

Results: 529 patients were included in the cohort. 203 patients (38%) received Hepatic Arterial Infusion Pump (HAIP) adjuvant therapy. The median follow-up was 102 months (IQR: 67, 154), with 5-year OS of 52% (95% CI: 48%, 57%). UV analysis showed that OS was associated with pre-NAC CEA (p = 0.012), post-NAC CEA (p < 0.001), ΔCEA (p = 0.043), and modified clinical risk score (mCRS; CRS without pre-op CEA) (p < 0.001). When provided with all CEA variables and mCRS, the CART model demonstrated that post-NAC CEA, ΔCEA, and mCRS predicted OS. The three predictors and the associated cut-off values created four risk groups: (i) low risk: post-NAC CEA < 4.5 ng/ml, ΔCEA clearance ≥ 94%; (ii) intermediate-low risk: post-NAC CEA < 4.5 ng/ml, ΔCEA clearance < 94%, mCRS ≤ 2; (iii) intermediate-high risk: post-NAC CEA < 4.5 ng/ml, ΔCEA clearance by < 94%, mCRS ≥ 3; (iv) high risk: post-NAC CEA ≥ 4.5 ng/ml. The 5-year OS of the four groups were 88% (95% CI: 76%, 100%), 80% (95% CI: 72%, 90%), 50% (95% CI: 40%, 63%), and 42% (95% CI: 37%, 48%), respectively. When stratified by HAIP adjuvant therapy, the 5-year OS of the high-risk group were 49% (95% CI: 40%, 59%) and 38% (95% CI: 32%, 46%), with and without HAIP.

Conclusions: Post-NAC CEA and ΔCEA both predicted the OS after CRLM resection. ΔCEA clearance ≥ 94% during NAC with post-NAC normalization of CEA predicted the highest 5-year OS, while the inability to normalize the post-NAC CEA predicted the lowest 5-year OS. In patients without post-NAC CEA normalization, receiving HAIP adjuvant therapy was associated with higher OS.

Learning Objectives:

Upon completion, participants will be able to describe the significant CEA predictors (pre-NAC, post-NAC, or delta CEA) for overall survival in recipients of CRLM neoadjuvant chemotherapy.
Upon completion, participants will be able to describe the criteria that define the high-risk group and the low-risk group.
Upon completion, participants will be able to describe what treatment is associated with a better survival within the high-risk group.