1. ¹Department of Pathology, Federal University of Paraná, Curitiba, Parana, Brazil
2. ²Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY, USA
3. ³Division of Women's and Perinatal Pathology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA
4. ⁴Division of Cytology, Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA

Correspondence: Dr ES Cibas, MD, Department of Pathology, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA. E-mail: ecibas@partners.org

Received 8 April 2008; Revised 1 May 2008; Accepted 4 May 2008; Published online 13 June 2008.

Abstract

Topoisomerase II and minichromosome maintenance protein 2 are proteins associated with aberrant S-phase induction. The current study evaluated the performance of these biomarkers (ProEx™ C; TriPath Oncology, Burlington, NC) compared with p16^INK4A and MiB-1 in distinguishing high-grade squamous intraepithelial lesions (HSILs) from HSIL mimics. We collected archival cervical biopsy, cone, and curettage specimens from 96 cases in which the differential diagnosis of HSIL vs reactive epithelial changes was considered. Hematoxylin- and eosin-stained slides were reviewed independently by three pathologists and scored for the presence or absence of SIL. Immunostains for ProEx C, p16, and MiB-1 were available for 95, 96, and 59 samples, respectively, and classified blinded to histological interpretation. Strong nuclear and cytoplasmic staining for p16 and staining for MiB-1 and ProEx C that extended beyond the lower one-third of the epithelium were scored as positive. ²-tests and receiver operating characteristic analysis were conducted to statistically compare biomarker immunostaining performance against majority histological interpretation of SIL. Agreement between pathologists was also assessed by the -statistic. Inter-observer agreement ranged from fair to moderate (=0.37–0.57). All three biomarkers correlated strongly with the majority diagnosis of SIL (P<0.001). Positive staining for ProEx C, p16, and MiB-1 was observed in 87% (N=52/60), 84% (N=51/61), and 94% (34/36), respectively, of SIL and negative in 71% (N=25/35), 63% (N=22/35), and 52% (N=12/23), respectively, of majority diagnoses of NoSIL. The combination of p16/ProEx C predicted more SIL (92%, N=33/36) and NoSIL (61%, N=14/23) than p16 plus MiB-1 (94%, N=34/36 and 43%, N=10/23), although this difference was not statistically significant. ProEx C appears to provide an equivalent level of sensitivity and a higher level of specificity for HSIL alone or in conjunction with p16. Its principal value may be in providing a lower false positive rate for NoSIL relative to MiB-1.