BACKGROUND The p53 tumor suppressor gene (also known as TP53) is often mutated in a wide variety of cancers, including angiosarcoma of the liver (ASL). Anti-p53 antibodies have been detected in the sera of patients with leukemia, childhood lymphoma, or cancers such as those of the breast, lung, colon, esophagus, and liver (hepatocellular carcinoma). OBJECTIVE The objective of this study was to determine the prevalence and time of appearance of serum anti-p53 antibodies during the pathogenesis of ASL associated with occupational exposure to vinyl chloride. METHODS Enzyme-linked immunoassay (EIA) was used to detect anti-p53 antibodies in 148 serum samples from 92 individuals occupationally exposed (in France or in Kentucky) to vinyl chloride; 15 of these individuals (six from France and nine from Kentucky) had ASL. A subset of coded EIA-positive and EIA-negative sera was further analyzed for anti-p53 antibodies by immunoblotting and immunoprecipitation. Nucleotide sequence analysis of exons 5-8 of the p53 gene was conducted on ASL DNA from six patients. We tested sera from 31 men who had no occupational exposure to vinyl chloride; they made up the control group. Statistical analyses were done using the Kruskal-Wallis chi-squared approximation and the Wilcoxon two-sample test for normal approximation. All P values result from two-sided tests. RESULTS Fourteen serum samples (from nine individuals) were positive in the EIA. Five of the 15 individuals with ASL were positive for anti-p53 antibodies by EIA, immunoblotting, and immunoprecipitation: one individual at 11.3 and 10.8 years before diagnosis, another at 4 months before and shortly after diagnosis, and three when diagnosed or shortly thereafter. Four of the 77 vinyl chloride-exposed workers without diagnosed ASL were positive for anti-p53 antibodies; two of the four had symptoms related to vinyl chloride toxicity. Tumors from three of the six vinyl chloride-exposed workers from which sufficient DNA for analysis was obtained had A:T to T:A missense mutations of the p53 gene. Anti-p53 antibodies were detected in two of these individuals. Among the control group, two of 15 serum samples from 15 lung cancer patients and zero of 15 serum samples from control subjects without cancer had anti-p53 antibodies as substantially lower levels than the nine (10%) of 92 vinyl chloride-exposed workers who were positive for anti-p53 antibodies. CONCLUSIONS Serum anti-p53 antibodies can predate clinical diagnosis of certain tumors, such as ASL, and may be useful in identifying individuals at high cancer risk, such as workers with occupational exposure to vinyl chloride.