RESUMEN
BACKGROUND: Non-small cell lung cancer (NSCLC) is one of the most aggressive tumors, with a very low overall survival rate. We investigated surgically resected squamous cell carcinoma (SCC) and adenocarcinoma (AC) to identify chromosomal imbalances and their value for individual prognostication. METHODS: A total of 80 cases, including 55 SCC and 25 AC, were retrospectively analyzed by comparative genomic hybridization. To model the sequential cytogenetic events, an oncogenetic tree model was applied based on maximum likelihood estimation. Clinicopathologic data and follow-up data were correlated with chromosomal imbalances. RESULTS: Fifty-one percent of patients were in stage I, 32% in stage II, and 17% in stage III, without statistically significant differences in staging distribution between SCC and AC. The average number of copy number imbalances was higher in SCC than in AC (9.4±1.2 vs 5.4±1.1; p=0.11). Frequency of chromosomal imbalances at -3p, +3q, -4q, +5q, -5q, +7q, and -13q were significantly different between SCC and AC. Subsequently, oncogenetic tree modeling identified different clusters of chromosomal imbalances for SCC and AC. Appearance of the -3p-cluster in SCC was associated with decreased overall survival independent of clinicopathologic parameters (mean, 42.8±7.5 months vs 80.1±9.1 months, log rank p=0.019), whereas in AC no prognostic value could be identified for specific clusters of chromosomal imbalances. CONCLUSIONS: Although, the limited number of analyzed cases allows a cautious statement on chromosomal imbalances, the oncogenetic tree modeling suggests distinct patterns of cytogenetic evolution for SCC and AC with implications for clinical outcome in SCC.
