Abstract 2830: Effect of tobacco and alcohol intake on the future development of liver metastasis

Abstract

Abstract While cigarette smoking and drinking of alcohol have been implicated in carcinogenesis, their role in the genesis of liver metastasis (LM) is undefined. Animal studies support a role for alcohol consumption but it appears that alcoholic liver cirrhosis may be protective. Given this conundrum, we prospectively evaluated a VA cohort of patients at high risk of cancer and therefore LM taking into account self- reported and serum-based substance abuse at baseline, documenting future development of cancer and LM correlated with tumor markers. Methods: Patients who completed a questionnaire to assess colorectal cancer (CRC) risk factors and scored positively were enrolled into the study. Data were compared to those in the computerized record and objective clinical data such as serum alcohol and CEA levels were recorded. Cumulative cancer diagnosis was noted on follow-up and LM had to be confirmed by biopsy or imaging. Tumor markers were serum CEA and stool p87 ELISA. Results: 21 patients developed LM (Group A), 529 cancer patients (Group B), and 588 remained cancer free (Group C). Follow up time was similar (A 9.7±5.4; B 9.4±5.4 and C 10.1±6.9 years). Mortality was 83, 42 and 38% in groups A,B,C respectively. In group A the LM were: 3 of each lung, CRC, breast, metastatic hepatocellular cancer and unknown primary; two of each: pancreatic cancer and bladder; and 1 of each B-cell lymphoma, prostate and renal cell carcinoma. These categories were represented in B by 88% of all cancer types. Smoking percentage was highest in the A (63% ;p=0,002 compared to cancer group), 46% in C (p=0.0001 compared to B), and 27% in B. Serum alcohol levels were highest in A 30.7±75.2 vs B 26.6±89.8 and C 25.9±85.8 mg/dL). Percent of active drinkers were lowest in A 12% and highest in C with B intermediate with no significant differences. Serum CEA levels were highest in A (12.24±13.66) vs B (3.59±4.81: p=0.063);B vs C (2.56±2.15 ng/ml: p=0.026). A direct correlation was found between serum alcohol and CEA levels (r=0.996;p=0.059). p87 Stool tests (81, 50 and 47% in A,B,C respectively) correlated with CEA (r=0.999;p<0.011) and ethanol levels (r=0.993,p=0.06). Conclusions: Only 1.8% of this sample developed LM but smoking would appear to be a factor as 62% of the LM were from cancers associated with smoking-risk. The combination of both factors was accordingly highest in A (38%) but significance was found only between groups B and C (24 vs 35% respectively, p=0.004) perhaps reflecting the habits of the entire VA population rather than cancer-associated. In an animal model for LM resistance we also confirmed a direct correlation between CEA and p87 in tissue extracts. Alcohol intake is more difficult to quantify than smoking and the ability to correlate the 2 serum analytes suggests a relevant assumption of the basis proof. CEA has long been recognized as being pro-metastatic in LM and the direct correlation of alcohol with CEA suggests that alcohol is a promoter for LM. Citation Format: Martin Tobi, Fadi Antaki, Mary Ann Rambus, Edi Levi, Harvinder Talwar, Michael Lawson, Benita McVicker. Effect of tobacco and alcohol intake on the future development of liver metastasis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2830.