Optimizing Endoscopic Respiratory Diagnostics with Cytology: An Update on Touch Imprints with a Comparative...

Background: Major diagnostic and therapeutic changes led to the implementation of the 'lung cancer diagnosis in small biopsies and cytology specimens' concept in the WHO Classification of Tumours of the Lung, Pleura, Thymus and Heart in 2015. Touch imprints are an established technique in cytology that provides a rapid and minimally invasive method for morphological diagnoses of clinical specimens, accurate subtyping, and molecular characterizations of malignancies. The extended diagnostic requirements from the increasingly limited material provided by minimally invasive biopsy techniques pose major challenges for pathology. Recognizing the relevant features and potential pitfalls is essential for cytologists to avoid misinterpretation. Materials and Methods: A retrospective analysis of endoscopic and surgical biopsy diagnostics was performed on 717 patients (303 women and 414 men; average age of 66.9 years) with clinically suspicious tumor findings at the LungenClinic Grosshansdorf in 2020. A total of 1363 cytological samples were obtained pre-therapeutically from 986 distinct biopsies covering 330 surgically and 656 endoscopically examined pulmonary, mediastinal, and bronchial regions. These samples were then compared with the histological diagnoses that were critical for determining the final therapy. Results: Out of a total of 656 endoscopically examined cases, 322 were classified as malignant, 308 as benign, and 26 as undetermined. While touch imprints and histological analysis separately achieved specificity values of 95.4% and 98.8%, both methods showed sensitivity values of 82.1% and 86.5%, respectively. In contrast, combining the two methods improved the sensitivity by 8 percentage points to 94.6%. Out of 330 cases of surgically examined samples, 137 were malignant, 190 were benign, and 3 were undetermined. The specificity of the morphological examinations for these samples was comparably high at 97.9% and 100%, respectively. In this surgical setting, touch imprints alone achieved a sensitivity of 75.9% (n = 104/137 cases), with a specificity of 97.9% (n = 186/190 cases). The outcome of the histological approach alone and in combination with touch imprints reached a sensitivity of 96.4% (n = 132/137 cases). Conclusions: Cytology and histology achieved comparably high sensitivity and specificity values on small biopsies. Under optimal conditions for morphological analysis in a surgical setting, the sensitivity of cytology for detecting malignant tumors was only 6 percentage points lower compared with the clinical endoscopic setting. A combined approach of cytologic-histologic evaluation for endoscopically examined specimens significantly increased the sensitivity by approximately 8% compared with the surgical setting (p < 0.003).