231 Reproducibility study on invasiveness of pulmonary AdC with a modified classification Discussion In the present study, unanimous agreement was obtained using the current WHO classification criteria in about half of the small adenocarcinomas. In the other half discordant judgements for invasion “yes” and “no” were frequently present within a case. Moreover, invasive size measurements lead to major differences in pT staging. In round 2 we saw a major shift in the judgement for individual pathologists toward ‘in situ’ adenocarcinoma with a revised classification that utilized more precise and standardized criteria, which is in general justified by the clinical follow-up, and is accompanied by a major improvement of interobserver variability from 0.27 in round 1via 0.45 in round 2 to 0.62 in a ‘coached’ round 3. Diagnosing invasion in small pulmonary adenocarcinomas is remarkably inexact even among specialized pulmonary pathologists, with disagreement in around 50% of cases leading to differences in pT staging, which is similar to a recent study of the pathology panel of the IASLC317. Estimating invasive size in small pulmonary adenocarcinomas may show some overlap between two adjacent categories (e.g., pT1a/b or pT1b/c), especially for tumor sizes around the threshold of the categories. However, two other important aspects are i) the overlap between pT in situ and invasive adenocarcinoma, which improved from 46% of the small adenocarcinomas using unaltered WHO guidelines to 34% for the revised classification in round 3. Likewise, for differences in invasive size of more than one category the percentage decreases from 10% to 8,5% in the 3rd round. The heatmap based on the underlying largest invasive line, taken together with assigned growth patterns, reflects a detailed reflection of the pathologist’s interpretation. This provides a new opportunity for analysis of specific areas and possibly better understanding of the morphological features relevant for the individual pathologist in the assessment of invasion. On a personal level every pathologist probably has some uncertain feeling about the designation of invasion in about half of the small pulmonary adenocarcinomas. The cases with a high average uncertainty score clearly were also the cases with discordant scores between invasion “yes” and “no” across several samples and individual pathologists. This shared feeling of uncertainty on the same cases implies awareness of the difficulty about the decision of invasion. In applying the WHO classification of the lung for the assessment of invasion a low kappa score of 0.27 was obtained. Concerning the assessment of invasion, the data from this study match with data from a recent publication from a group of experts in pulmonary pathology. [see Supplementary Figure 2317]. In both studies, several pathologists participated from different areas around the globe. This demonstrates that it is a worldwide issue with marked scientific implications. Firstly, i) we (read pathologists) are currently not competent at making a similar decision about invasion in pulmonary adenocarcinomas across the globe. Taking into account the shared feeling of uncertainty about the decision of invasion, after combining the two studies the statement seems valid in terms of learning theory that “we are consciously incompetent 15
RkJQdWJsaXNoZXIy MTk4NDMw