Core classification of lung cancer: Correlating nuclear size and mitoses with ploidy and clinicopathological parameters

Abstract We attempted to establish a microscopy based tumour characterization providing insight into the genetics of cancer cells and in particular their DNA ploidy. The core classification defined semi-quantitative criteria for scoring the nuclear size ranging from small (core score 1) to giant nuclei (core score 4). By listing all nuclear sizes according to their relative frequencies it provided a measure for core size variability as a correlate of nuclear pleomorphism. Additionally, the mitosis size, their variability and the presence/abundance of tripolar and tetrapolar mitoses were determined. This classification was applied to 155 lung cancer samples from all major histologic types and the results were correlated with the analysis by DNA image cytometry and patient survival. The morphological assessments correlated highly significantly with the DNA ploidy parameters, e.g. small cell lung carcinomas showed the smallest values for nuclear size (mean core score of 1.18) and DNA content (DNA index mean of 2.08c) being highly significantly different from adenocarcinomas (1.95/3.10c), large cell lung carcinoma (2.00/3.26c) and squamous cell carcinoma (2.20/3.42c). In non-small cell lung carcinoma (NSCLC) in general and adenocarcinoma in particular, the core size variability correlated significantly with grading and survival. Furthermore, parameters indicative for chromosomal variability, i.e. 2c deviation index and 5c exceeding rate, were predictors of poor survival in NSCLC patients. As a complement to histologic tumour diagnosis the core classification should help to better stratify cancer subtypes.