SPINK1 Expression in Bladder Urothelial Carcinoma In Situ: Important Clinical Implications
Francesca Khani, Kyung Park, Ya-Lin Chiu, Brian D Robinson, Mark A Rubin, Juan Miguel Mosquera. Weill Cornell Medical College, New York
Background: Diagnosis of urothelial carcinoma in situ (CIS) can be challenging at times, particularly on small bladder biopsies. p53 overexpression, aberrant CK20 expression and reduced CD44 expression have been used as diagnostic aids in CIS with variable results. Based on our preliminary observations in tissue microarrays, we assessed the utility of SPINK1 immunohistochemistry (IHC) in the diagnosis of CIS and urothelial dysplasia in clinical specimens.
Design: 159 bladder specimens including 106 biopsies, 32 cystectomies, and 21 transurethral resections from 108 men and 51 women, age 39 to 94 years old were studied. IHC was performed using a monoclonal antibody for SPINK1 (Abnova). Two pathologists performed semi-quantitative IHC evaluation: more than 20% staining extent and 2-3 intensity (scale 0 to 3) was considered positive. Expression in CIS (n=82) was compared to urothelial dysplasia (n=13) and urothelial atypia (n=18). Benign urothelial samples (n=46) were used as controls, including cases of papillary cystitis, radiation cystitis, and cystitis cystica. SPINK1 results were compared with p53 expression in a subset of CIS (n=33), urothelial dysplasia (n=7), urothelial atypia (n=6) and benign controls (n=4).
Results: SPINK1 overexpression was seen in 74/82 CIS cases (90%), 7/13 dysplasia cases (54%), 3/18 urothelial atypia cases (17%), and 3/46 benign control cases (6.5%). In the subset group, p53 overexpression was seen in 23/33 CIS cases (70%), 5/7 dysplasia cases (71%), 5/6 urothelial atypia cases (83%), and 1/4 benign control cases (25%). SPINK1 staining was observed only in the umbrella cells in 40% of the benign controls, including in the cases of papillary cystitis, radiation cystitis, and cystitis cystica. For detection of CIS or urothelial dysplasia, SPINK1 IHC showed a sensitivity of 85% and specificity of 91%. In contrast, p53 IHC had a sensitivity of 70% and specificity of 40%. Used in combination, these two biomarkers have a sensitivity and specificity of 88% and 85%, respectively, for the detection of CIS and urothelial dysplasia.
Conclusions: We found SPINK1 to be a highly sensitive and specific marker for urothelial CIS and urothelial dysplasia. While it shows a similar sensitivity to p53, our study reveals that it is more specific. Positive SPINK1 expression may be helpful to distinguish CIS and urothelial dysplasia from atypical lesions, particularly in small bladder biopsies. We are currently planning validation studies incorporating either SPINK1 alone, or in combination with p53, as an additional diagnostic tool in routine practice.
Category: Genitourinary (including renal tumors)
Wednesday, March 21, 2012 1:00 PM
Poster Session VI # 150, Wednesday Afternoon