| Literature DB >> 25455993 |
Anja C Roden1, Eunhee S Yi2, Sarah M Jenkins3, Janis L Donovan2, Stephen D Cassivi4, Yolanda I Garces5, Randolph S Marks6, Marie-Christine Aubry2.
Abstract
The prognostic importance of histologic classifications of thymic epithelial neoplasms is controversial. Evidence suggests that difficulties in reproducibility affect prognostic studies. Two thoracic pathologists independently classified 80 cases of type A or B3 thymoma and thymic carcinoma according to World Health Organization (WHO) classification. Ki-67 labeling index (LI) was used to identify cutoff points between WHO types. Recursive partitioning (Rpart) and ad hoc methods separated the data points. The pathologists agreed on type A (n = 31), type B3 (n = 21), and thymic carcinoma (n = 14). Ki-67 LI differed between types A and B3 (P < .001) and between thymic carcinoma and type A (P < .001) or type B3 (P = .001). Mitotic activity differed between thymic carcinoma and type A (P < .001) or type B3 (P < .001). Rpart revealed Ki-67 LI greater than 14.0% only in thymic carcinoma; cases with Ki-67 LI less than 5.1% did not represent thymic carcinoma. Ad hoc analysis showed Ki-67 LI greater than or equal to 13.5% represents thymic carcinoma; only type A had Ki-67 LI less than 2%. The pathologists disagreed on histologic type in 14 cases. In 11 of 14 cases with available Ki-67, the Rpart method predicted the WHO type; in 7 of 14 cases, the ad hoc method predicted the WHO type. In conclusion, Ki-67 LI is helpful in differentiating thymic epithelial neoplasms, with Ki-67 LI less than 2% and greater than or equal to 13.5% distinguishing type A thymoma and thymic carcinoma, respectively.Entities:
Keywords: Bcl-2; Ki-67 labeling index; Mitoses; Thymic carcinoma; Thymoma
Mesh:
Substances:
Year: 2014 PMID: 25455993 DOI: 10.1016/j.humpath.2014.10.001
Source DB: PubMed Journal: Hum Pathol ISSN: 0046-8177 Impact factor: 3.466