Bronchopulmonary carcinoids 2012-2022 Manchester University NHS Foundation Trust, United Kingdom
Description
Retrospective analysis of surgically resected lung carcinoids to examine the significance of pathological and radiological variables in preoperative diagnosis. Particular focus was on Antigen Kiel 67 (Ki-67) proliferation index. Exploratory analyses was performed to study pre- and post-operative concordance of carcinoid histology, clinical and pathological staging, and post treatment outcomes of recurrence and survival. We found that high Ki-67 has reasonable accuracy (77.8%) in suggesting the final diagnosis of atypical carcinoid and is significantly associated with disease recurrence. Fluorodeoxyglucose-18 SUV (with cut-off of 5.0) showed modest sensitivity and specificity to distinguish typical and atypical subtype. Typical carcinoid showed better pre and post operative concordance than atypical carcinoids. More research is needed to identify atypical carcinoids more accurately. This study suggests that Ki-67 index should be calculated on all diagnostic samples before treatment of lung carcinoids, and may be incorporated as an essential criterion in lung carcinoid classification.
Files
Steps to reproduce
We searched our hospital’s pathology & thoracic surgery databases for patients who underwent surgical resection (during 2012-2022) and were confirmed to have a diagnosis of lung carcinoid on the resection specimen. Each pathological variable (mitotic rate, Ki-67, and necrosis) on the diagnostic specimen (attained by image guided lung biopsy or bronchoscopy guided) was studied for diagnostic accuracy using pre-defined classifications of each variable and the final pathological diagnosis on the resection sample. For the diagnostic accuracy of mitotic rate, a cut-off of <2/2mmsq. was taken to indicate a diagnosis of Typical Carcinoid(TC), 2-10/2mmsq. to indicate Atypical Carcinoid (AC) and >10/2mmsq. to indicate a high-grade poorly differentiated lung NET (Neuroendocrine Tumour). For Ki-67 index, a value of <5% was taken to indicate a diagnosis of TC, 5-30% to indicate an AC and >30% to indicate a high-grade poorly differentiated lung NET. The absence of necrosis was taken to indicate TC whilst the presence of necrosis was taken to indicate AC. The concordance of histological type between the biopsy and surgical specimen was also studied. Further analyses were performed to study the diagnostic accuracy of PET (Positron Emission Tomography) imaging & other radiological variables in differentiating TC versus AC and, the concordance of clinical and pathological staging in bronchopulmonary carcinoids. Exploratory analyses looked at the association between variables and post treatment outcomes of recurrence and mortality. The cohort was categorised based on the final diagnosis (TC vs AC based on full pathological analysis of the resected tumour). For interpreting diagnostic accuracy, sensitivity reflects the proportion of people with an atypical carcinoid who were correctly identified. Specificity is defined as proportion of people with a TC who were correctly identified. All tests were two-sided and conducted at the 5% significance level. A total of 26 tests were used. Bonferroni Correction was used to adjust the alpha level (threshold for significance) to preserve the 5% change of committing a type I error. The new threshold is 0.002, p-values less than this were considered significant.