The purpose of this study was to determine the effect of alcoholism on survival among patients with stage III and IV non-small cell lung cancer (NSCLC).
The records of 74 consecutive patients with clinicopathologic stage III/IV NSCLC receiving treatment at a VA Medical Center were analyzed. Twenty four alcoholic (Etoh) patients [defined by a DSM-IV diagnosis of dependence (303.90), abuse (305.00), or an alcohol consumption of > 60 ounces of alcohol per day] were compared to 50 non-alcoholic (Non-etoh) patients.
Overall survival was worse in the Etoh group compared to the Non-etoh group, 9.5 ± 2.0 vs. 13.8 ± 1.8 months, respectively, p<0.05. Progression free survival was 7.0 ± 1.6 and 12.5 ± 1.9 months for Etoh and Non-etoh patients, respectively, p=0.02, Figure 1. Survival differences were not explained by discrepancies in Eastern Cooperative Oncology Group (ECOG) performance status, stage of disease, Charlson comorbidity index, smoking history, pulmonary function, liver enzymes, or nutritional parameters. Alcoholics were more likely to refuse treatment, less likely to receive multimodality treatment, received a lower total dose of radiation, and were less likely to respond to chemotherapy than non-alcoholic patients. Toxicity of chemotherapy was similar between Etoh and Non-etoh groups, Table 1Table 1.
Clinical data in patients with stage III and IV non-small cell lung cancer (NSCLC). Alcoholic (Etoh) vs. Non-alcoholic (Non-etoh).EtohNon-etohP valueAlcohol consumption (ounces/day)137 ± 2326 ± 3.70.0041Divorce, jobless, homeless16 / 249 / 500.00012Stage of disease (III/IV)16 / 830 / 20NS2ECOG3 performance0.9 ± 0.11.2 ± 0.1NS1Charlson comorbidity2.2 ± 0.32.2 ± 0.2NS1Pack years smoking68 ± 6.960 ± 4.0NS1Bilirubin (mg/dl)0.6 ± 0.040.6 ± 0.03NS1Albumin (g/dl)3.8 ± 0.13.9 ± 0.1NS1Paclitaxel4 total dose (mg/M2)456 ± 101509 ± 59NS1Carboplatin4 total dose (mg/M2)1128 ± 209926 ± 113NS1Grade 2-44,5 Granulocytopenia6 / 138 / 29NS2Stage III patients refusing treatment67 / 161 / 300.0012Stage III patients receiving multimodality treatment68 / 1624 / 300.052Total radiation dose to primary tumor (cGy) in patients receiving radiation74145 ± 6115468 ± 2380.038Response to chemotherapy stage III/IV patients43 / 1317 / 290.0521
Two tailed unpaired student t test, p< 0.05 significant, NS=non-significant.2
Two tailed Fisher exact test, p<0.05 significant, NS=non-significant.3
Eastern Cooperative Oncology Group4
13 Etoh and 29 Non-etoh patients with stage III and IV NSCLC receiving chemotherapy.5
NCI Common Terminology Criteria for Adverse Events v. 3.0, Dec. 12, 2003.6
16 Etoh and 30 Non-etoh patients with stage III NSCLC.7
12 Etoh and 35 Non-etoh patients receiving radiation for stage IIIand IV NSCLC.8
Mann Whitney U test.. Alcohol abuse (B=0.59, p=0.03, hazard ratio=1.82) and ECOG status (B=0.57, p=0.02, hazard ratio=1.77) were the two best predictors of survival in a Cox proportionate hazards model. In multiple logistic regression, alcohol abuse predicted lack of response to chemotherapy (B=2.9, p=0.03, Odds ratio=18.3).
Alcoholic patients with stage III/IV NSCLC, even in the absence of cirrhosis, have a worse prognosis than non-alcoholic patients. This is due to non-compliance and a lower clinical response rate even when therapy is accepted.
Given the high prevalence of both alcohol abuse and lung cancer in the VA population, the results of this study, if confirmed, may have important prognostic and therapeutic implications.
D.E. Paull, None.