콘텐츠로 건너뛰기
Merck
  • ARID1A immunohistochemistry improves outcome prediction in invasive urothelial carcinoma of urinary bladder.

ARID1A immunohistochemistry improves outcome prediction in invasive urothelial carcinoma of urinary bladder.

Human pathology (2014-09-02)
Sheila F Faraj, Alcides Chaux, Nilda Gonzalez-Roibon, Enrico Munari, Carla Ellis, Tina Driscoll, Mark P Schoenberg, Trinity J Bivalacqua, Ie-Ming Shih, George J Netto
초록

AT-rich interactive domain 1A (ARID1A) is tumor suppressor gene that interacts with BRG1 adenosine triphosphatase to form a SWI/SNF chromatin remodeling protein complex. Inactivation of ARID1A has been described in several neoplasms, including epithelial ovarian and endometrial carcinomas, and has been correlated with prognosis. In the current study, ARID1A expression in urothelial carcinoma (UC) of the bladder and its association with clinicopathological parameters and outcome are addressed. Five tissue microarrays were constructed from 136 cystectomy specimens performed for UC at our institution. Nuclear ARID1A staining was evaluated using immunohistochemistry. An H-score was calculated as the sum of the products of intensity (0-3) multiplied by extent of expression (0%-100%). Average H-score per case was used for statistical analysis. ARID1A expression was categorized in low and high using Youden index to define the cut point. ARID1A expression significantly increased from normal to noninvasive UC to invasive UC. For both tumor progression and cancer death, Youden index yielded an H-score of 288 as the optimal cut point for ARID1A expression. Low ARID1A expression showed a tendency for lower risk of tumor progression and cancer mortality. Adding ARID1A expression to pathologic features offers a better model for predicting outcome than pathologic features alone. Low ARID1A expression was more frequently seen in earlier stage disease. There was a tendency for low ARID1A expression to predict better outcome. More importantly, the findings indicate that adding ARID1A expression to pathologic features increases the goodness of fit of the predictive model.