A 2‐year retrospective study on pleural effusions: A cancer centre experience

Abstract

Background Cytopathological examination of pleural effusions is a fast and minimally invasive method for verification of the presence of neoplastic cells. We report our 2‐year experience using a categorised diagnostic system and reporting risks of malignancy (ROMs) for each defined category. Methods Cytological reports of patients between November 2016 and October 2018 were collected, with results primarily classified into a five‐tiered classification scheme. Immunohistochemistry markers used in cytology and their results were also recorded. Final agreement to histology and overall test performance was calculated for cases with available concomitant (up to 3 months) pleural biopsies. Results A total of 519 samples from 385 patients were collected, being 29 (5.6%) classified as non‐diagnostic, 291 (56%) as negative, 28 (5.4%) as atypical, 30 (5.8%) as suspicious and 141 (27.2%) as positive. Most requested markers were calretinin, TTF1, Ber‐EP4 and Gata‐3, being conclusive in 45 (76.3%) cases. Total cyto‐histological agreement was achieved in 49 (80.3%) specimens, with an overall sensitivity and specificity of 69.4% and 93.3%, respectively. Positive predictive value was 96.2% and negative predictive value was of 56%. ROM for each diagnostic category was 50% for non‐diagnostic, 44% for negative, 50% for atypical, 83.3% for suspicious and 96.2% for positive. Conclusions Our 2‐year retrospective study has shown a high specificity and positive predictive value for pleural cytology. The use of a five‐tiered system has also shown to be highly effective, with a concordantly progressive higher ROM for the assigned diagnostic categories.

Application of the International System for Reporting Serous Fluid Cytopathology with Cytohistological Correlation and Risk of Malignancy Assessment

“…Pericardial effusions only included three cases, and their distribution was 33.33% (one case) NFM and 66.67% (two cases) MAL. Our results are compatible with those of other studies [ 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ], some of which include only pleural effusions [ 25 , 26 , 27 , 30 ], while others include pleural and peritoneal effusions [ 24 , 28 , 31 , 32 ], and one includes only pericardial effusions [ 29 ]. In regard to pleural effusions, the rate of malignancy reported from oncologic centers is significantly higher [ 31 ], while a similar rate to ours is reported from other institutions [ 26 ], probably related to the “general” nature of our institution.…”

“…Meanwhile, for peritoneal effusions, the corresponding values were 80%, 99.3%, 98.5% and 89.6%. The accuracy parameter values are in line with those reported in other studies [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ]. Notably, oncologic centers report lower sensitivity [ 25 , 31 ].…”

The value of a tiered cytology diagnostic reporting system in assessing the risk of malignancy in indeterminate serous effusions

“…A meta‐analysis of 73 studies reported a mean ROM for the AUS and SFM categories of 65.9% and 81.8%, respectively 19 . Another recently published single‐institution study reported an ROM of 50% and 83.3% for AUS and SFM cases, respectively 20 . By comparison, the ROM for both the AUS and SFM categories in our study was lower (39% and 64%, respectively) than that in the published data.…”

Cytohistological Correlation in Pleural Effusions Based on the International System for Reporting Serous Fluid Cytopathology

“…In this same year, Valerio et al reported on a series of 65 pleural effusions with matching biopsies and found a ROM of 50% for ND, 44% for NFM, 50% for AUS, 83.3% for SFM, and 96.2% for MAL. They reported a sensitivity of 69.4%, a specificity of 93.3%, a PPV of 96.2%, and an NPV of 56% [ 24 ].…”