Spatial distribution of EGFR and KRAS mutation frequencies correlates with histological growth patterns of lung adenocarcinomas

Multiregional analysis provided first indications for morphological and molecular heterogeneity in lung adenocarcinomas (ADCs), but comprehensive morpho-molecular comparisons are still lacking. The purpose of our study was to investigate the spatial distribution of EGFR and KRAS alterations systematically throughout whole tumor cross-sections in correlation with the tumor cell content and the histopathological patterns. Central sections of 19 ADCs were subdivided into 467 segments of 5 mm x 5 mm. We determined the predominant histological growth pattern and the allele frequencies of driver gene mutations by digital PCR in every segment. We further quantified the absolute cell counts and proportions of tumor and nonneoplastic cells in all segments to normalize the mutant allele frequencies. Driver gene mutations could be detected in >99% of the tumor containing segments, with high levels of inter- and intratumor heterogeneity regarding the mutant allele frequency (range: 0.04-19.36). Different patterns for the distribution of the variant allele frequency within a tumor were recognizable. While some cases showed ubiquitously low or high levels, others revealed regions with focally elevated frequencies. Differences between KRAS and EGFR alterations were not significant. The great majority of the analyzed tumor sections (16/19) exhibited two or more morphological growth patterns. Mutant allele frequencies were significantly higher in segments with a predominant solid pattern compared to all other histologies (p < 0.01). Our data indicate that driver gene mutations are present with high levels of inter- and intratumor heterogeneity throughout the whole tumor, with a correlation between the allele frequencies and histological growth patterns.