Insufficient endobronchial ultrasound-guided transbronchial needle aspiration specimens. When and why? The analysis of criteria and reasons behind the insufficient specimens.

BACKGROUND: The classification terminology systems for pulmonary cytology specimens have recently emerged. Inadequate samples, classified as "nondiagnostic," raise challenges in determining the threshold of cell numbers and the risk of malignancy (ROM). METHODS: The study retrospectively reviewed 248 endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) samples: 46 insufficient samples, 60 low cellularity samples, and 142 adequate samples. Characteristics as cellularity, number of benign and malignant cells, and background features were assessed. Receiver operating characteristic curve analysis was performed to establish cell sufficiency thresholds for the diagnosis. RESULTS: Out of the 248 samples analyzed, 108 were classified as benign, 94 as malignant, and 46 as insufficient. The study found that the cellularity thresholds for diagnosis in cell blocks and cytological samples were ≥50 cells and ≥100 cells, respectively. The thresholds for tumor cell counts were ≥1 - 10 cells for both types of cells, respectively. Considerably, some low cellularity samples were initially classified as insufficient despite meeting the diagnostic thresholds upon revision. The ROM varied across sample categories, with insufficient samples having a ROM of 10.9%, benign samples 15.7%, suspicious samples 92.0%, and malignant samples 100%. CONCLUSION: Insufficient EBUS-TBNA samples raise challenges in diagnosis and management. This study identified the root cause of insufficient samples, including factors related to humans, diagnostic methods, sampling, and laboratory processing. By understanding the root causes, diagnostic recommendations can be developed to improve the diagnostic process. The findings emphasize the importance of standardized classification and terminology systems for clear communication among healthcare professionals and institutions, ultimately improving patient care and enabling quality assurance measures.

Ectopic thyroid in EBUS: experience from a quality assurance programme.

A diagnostic challenge is presented: Distinguishing ectopic thyroid tissue from metastatic well-differentiated follicular carcinoma in cytological material. Two cases of thyroid tissue in mediastinal lymph nodes were sampled by endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA). Later, the cases were presented in Labquality's nongynecological external quality scheme rounds in the years 2017, 2019, and 2020. The same case was presented two times, both in the 2017 and in the 2020 rounds. The results of the three rounds and the discussion of diagnostic pitfalls of ectopic thyroid tissue are presented. A total of 112 individual laboratories worldwide participated in the external quality assurance rounds with whole-slide scanned images and digital still images of alcohol-fixed Papanicolaou-stained cytospin specimens in the years 2017, 2019, and 2020. Fifty-three laboratories participated in both the 2017 (53 of 70, 75.71%) and the 2020 (53 of 85, 62.35%) rounds. The given Pap classes between rounds were compared. Twelve (12 of 53, 22.6%) of the laboratories gave the same Pap class value, whereas 32 (32 of 53, 60.4%) were in the range of ±1 class difference (Cohen's kappa -0.035, p < 0.637). When comparing the diagnoses, 21 (21 of 53, 39.6%) laboratories gave the same diagnosis in 2017 and in 2020 (Cohen's kappa 0.039, p < 0.625). Thirty-two of the laboratories gave the same diagnosis both in 2017 and in 2020 (Cohen's kappa 0.004, p < 0.979). Ten (10 of 53, 18.9%) laboratories changed their diagnose from malignant to benign, and 11 (11 of 53, 20.8%) changed their diagnose from benign to malignant between the 2017 and the 2020 rounds. In conclusions, the expert reference diagnosis was thyroid tissue in mediastinal lymph node. Thyroid tissue in mediastinal lymph node may be either of ectopic or of neoplastic origin. The diagnostic work-up should include cytomorphological, immunohistochemical, laboratory, and imaging results. If a neoplastic change is excluded, the benign category is the most feasible one. The quality assurance rounds showed a large variability in the given Pap classes. Mirroring the problematic issue both inter- and intralaboratory of such cases both in routine diagnostics and in the classification terminologies is requiring multidisciplinary evaluation approach in the diagnostics.

EBUS-TBNA: A 2-Year Experience from a Tertiary Pathology Centre with Cyto-Histological Correlation.

INTRODUCTION: Endoscopic ultrasound-guided fine-needle aspiration and endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) methods are expanding due to their minimally invasive nature using real-time image guidance, a low incidence of complications, and a diagnostic rate ranging from 89% to 93.6%. The application of EBUS-TBNA was able to downgrade 40% of patients with suspicion of stage III lung cancer on PET scan. METHODS: In the present study, we analysed the diagnostic outcomes of EBUS cytology at the Laboratory of Pathology, Fimlab Laboratories, Tampere, Finland, during a 2-year period. In our analysis, 1176 EBUS samples from 486 patients that underwent EBUS-TBNA in Tampere University Hospital (TAYS) from January 2017 to December 2018 were reviewed. The mean age of the patients was 64 ± 14 (SD) years with the age range from 19 to 90 years. Cytospin cytology slides stained by Papanicolaou stain and cell blocks were performed in each sample. Further immunohistochemical stains were ordered by the pathologist on demand. RESULTS: The diagnostic yield was 89.5% and the diagnostic accuracy was 77.9% in the present study. The sensitivity for malignancy in EBUS-TBNA diagnosis was 95.69% and the negative predictive value 96.75%. Cyto-histopathological diagnosis correlation was 54.7% (95% confidence interval [CI]: 40.6%-66.4%, p < 0.001), and the histological accuracy compared to the exact original diagnosis was 43.5%. When considering the histological diagnoses that support the original diagnosis, the correlation was 61.3% (95% CI: 48.4%-71.5%, p < 0.001) and the accuracy was 72.2%. The overall inadequate rate of EBUS-TBNA samples in this study was 10.5%. CONCLUSIONS: The EBUS-TBNA diagnostic yield, diagnostic accuracy, and inadequate rate in the present study are in line with previously published studies.