273 General discussion and future perspectives AIS with 100% 5-year disease free survival may be limited in two ways. On the one hand resected non-mucinous adenocarcinomas may lead to cure. The presence of invasive characteristics without metastases is not excluded by the disease-free survival rate. In any managerial classification, thus also the revised adenocarcinomas classification, will have cases that do not fit into the designed categories. Although no system will ever be perfect, a possible improvement in the kappa score from poor reproducibility (0.27) according to the WHO to 0.62 for the revised classification is a major stimulus for further exploration. Borderline cases will always exist and downgrading seems an appropriate approach, also with the Hippocratic oath in mind: “primum non nocere” (“first, do no harm”) Suggestions for future studies As two independent studies (the reproducibility study in chapter 15 and the study of the pathology panel of the IASLC317) the data show that for the assessment of invasion the pathologists when using the WHO classification in terms of the Dunning-Kruger effect are consciously incompetent339. Artificial intelligence is a topic of many studies today. A future analysis of the cases of chapter 14 with an unsupervised AI approach is interesting364. The dream of AI is creating the hope for the ideal solution. In a study of the previous century AI was applied in the differential diagnosis between small cell lung cancer and NSCLC397, what is a problem area in only 5%. To this end the tumor cell nuclei were digitized with scanning stage microscope (resolution 0.42 micron). AI decreased the fraction of problem cases by 50%. In other words, AI may be supportive in reduction of this problem area, but not a perfect solution: drawing lines in a biological system will always have boundary cases. Although expectations may be high, the AI researchers presenting at the last USCAP were modest. Time will tell. Conclusions Basic histopathology remains a critical aspect of diagnosing and treating lung cancer. Evaluating the extent of pathologic response in resected specimens after neoadjuvant therapy can be immensely valuable in predicting prognosis. Removing “Spread through air spaces” (STAS) from the WHO classification could potentially alleviate some of the confusion surrounding it since its assumptions of invasion are unproven and we believe it to be an artifact. Identifying a collapsed adenocarcinoma in situ pattern with the help of additional cytokeratin 7 and elastin staining can improve the classification and staging of resected non-mucinous adenocarcinomas. Providing further training on the proposed revised classification may prove advantageous for patients with pulmonary adenocarcinomas worldwide. 17
RkJQdWJsaXNoZXIy MTk4NDMw