These studies of lung cancer suggest that a number of molecular mechanisms may be important in the pathogenesis of lung cancer, especially SCLC. An inherited predisposition to develop SCLC may correlate with a nonfunctional, recessive allele for a gene (McKusick #18228, McKusick 1986) that maps to chromosome region 3p(14-23). Individuals at risk would be heterozygous for this allele in their germ line, carrying one copy of a normal functional gene and one mutant, recessive allele. Exposure to carcinogens, in particular cigarette smoke, can produce somatic genetic changes such as chromosomal deletion or gene mutation in the functional allele of this gene, unmasking the nonfunctional allele. Loss of this normal gene may alter the regulation of cell growth, perhaps by allowing the deregulated expression of proto-oncogenes of the myc family, or autocrine growth factors such as GRP and/or its receptor. Alternatively, loss of this gene may result in the cell returning to a less differentiated developmental state where growth regulation is less stringent. Persons with this mutant gene should be at increased risk to develop SCLC, and further RFLP analysis of the 3p region in SCLC may allow identification of specific haplotypes with increased risk of developing lung cancer. If this notion is correct, one might expect to find an increased frequency of second tumors in lung cancer patients and the presence of similar chromosomal deletions in second tumors arising in SCLC patients. In this regard, cured lung cancer patients, including those with SCLC, have a tenfold increased risk of developing a second lung cancer (Fontana 1977; Cortese et al. 1983; Johnson et al. 1986b). In fact, a chromosome 3p deletion along with other chromosomal abnormalities was identified in acute erythroleukemia cells arising in a long-term survivor of SCLC (Bradley et al. 1982), implicating this same region in the pathogenesis of both tumors. Other predictions include the correction of at least a portion of the defect by introducing a normal chromosome 3 into SCLC cells. While c-myc is expressed in many fetal and adult tissues, high-level expression of N- and L-myc is very restricted as to tissue and stage in the developing mouse, with N-myc expressed in the fetal but not adult lung, whereas the lung was the only adult tissue where L-myc expression was detected (Zimmerman et al. 1986). Could these patterns provide a clue to the differential expression of c-, N-, and L-myc found in different lung cancers (Nau et al. 1986)?(ABSTRACT TRUNCATED AT 400 WORDS)