ProEx C as Diagnostic Marker for Detection of Urothelial Carcinoma in Urinary Samples: A Review.

ProEx C (BD Diagnostics-TriPath, Burlington, North Carolina) is an immunohistochemical cocktail containing antibodies against topoisomerase IIα (TOP2A) and minichromosome maintenance 2 (MCM2) proteins. TOP2A is a nuclear enzyme that controls and alters the state of DNA during transcription, occurring in processes such as chromosome condensation, chromatid separation and the relief of DNA torsional stress. This enzyme catalyzes the temporal breaking and rejoining of two strands of DNA, thus altering the DNA structure. The gene encoding TOP2A is the target for numerous anticancer agents; mutations in this gene have been related with the development of drug resistance 33 . MCM2 protein is a key component of the pre-replication complex and may be involved in the formation of replication forks and in the recruitment of other DNA replication related proteins 4 . Deregulation of MCM2 function has been suggested to contribute to tumorigenesis 6 . Both TOP2A and MCM2 are over-expressed in the cell nucleus during aberrant S-phase induction of human papillomavirus (HPV)-infected cells 7 , 8 , 34 - 36 . Several studies have confirmed the over-expression of ProEx C especially in cervical squamous dysplasia 9 , 37 - 41 . Guo and collaborators investigated on the efficacy of p16 and ProExC in detecting high-grade cervical intraepithelial neoplasia (CIN 2+) and cervical carcinoma 42 . The more severe was the cervical lesion, the more p16 and ProExC were positive. p16 immunostaining was more sensitive (79% for CIN 2+; 90% for CIN 3+) than ProExC (67% for CIN 2+; 84% for CIN 3+), whereas for CIN 3+ lesions, ProExC showed a specificity higher than p16. The highest specificity (100% for CIN 2+; 93% for CIN 3+) was found in samples with positivity of both biomarkers (p16+/ProExC+), suggesting that the combination these two biomarkers can be very useful to discriminate CIN 2/3 from its mimics. Furthermore, Siddiqui investigated on the utility of ProEx C for detecting CIN2+ lesions and compared it with high-risk HPV (hr-HPV) status in patients with ASC-US cytology: ProEx C showed a higher sensitivity compared to hr-HPV (98.04% and 82.35%, respectively), whereas the specificity was not statistically significant 9 , 34 . ProEx C immunostaining was also performed on formalin-fixed, paraffin-embedded histological sections for distinguishing HSIL from adenocarcinoma and from various non-neoplastic glandular lesions, such as microglandular hyperplasia, tubal metaplasia, cervical endometriosis, reactive endocervix and atrophy 43 , 44 . ProEx C showed a higher sensitivity, specificity, positive and negative predictive value in well-defined neoplastic lesions (high-grade squamous intraepithelial lesion/adenocarcinoma in situ (AIS)), compare to non-neoplastic lesions (squamous metaplasia/ reactive benign endocervix) 7 . The distribution of immunostaining for AIS was different from all benign mimics, but the intensity of staining for AIS overlapped with some mimics as it was not significantly different from endometriosis, microglandular hyperplasia and reactive endocervix 45 . These studies suggest that, although ProEx C is a valuable marker for distinguishing squamous and endocervical lesions of the cervix from reactive benign changes, caution should always be taken into account when using this marker in evaluating hyperchromatic crowded groups in Papanicolaou-stained gynecological smears. Walts and Bose proved positive staining (>50%) for ProEx C in Paget cells, in all cases of Paget's disease irrespective of tissue site (extramammary, mammary), albeit it appeared that the immunostaining could be unrelated to HPV 46 . ProEx C is a useful proliferation marker for high-grade vulvar intraepithelial neoplasia analogous to the staining patterns reported in high-grade cervical intraepithelial neoplasia, which is essentially limited to the basal and parabasal layers of the epithelium 37 . Similarly, ProEx C seems to be helpful in distinguishing melanoma from benign nevi, although ProEx C does not have prognostic significance in disease-specific survival in patients with primary melanoma 47 .

Urothelial carcinoma (UC) is one the most common malignancies derived from the urothelium of the lower urinary tract. Every year approximately 380000 new cases of UC occur in the world, with an estimated 15210 deaths from disease 1 . At initial diagnosis, most UCs are non-muscle invasive and the prognosis for these patients is generally good. Cancers will recur in 30-80% of cases, with a progression to muscle invasive disease of 1-45% within 5 year 2 , 3 . The accurate diagnosis is crucial for the appropriate management and routinary controls for UC are necessary once the diagnosis is made 4 - 7 . Urethro-cystoscopy, which is best for detecting low-grade urothelial carcinoma (LGUC), and urine cytology as supplement, often the test that recognizes high-grade urothelial carcinoma (HGUC) are the current approaches for the initial detection and follow-up of UCs. Imaging of the upper urinary tract is carried out as a further primary investigation and for follow-up in high risk cases 2 , as urothelial carcinoma can be also found in the renal pelvis or ureter. Both cystoscopy and biopsy are invasive and pricey procedures and therefore cytology is often used as first approach to investigate on a possible neoplasia, being a safe and cost-effective diagnostic modality of evaluation 8 , 9 . Routinary cytopathology can be very challenging in discerning neoplasia/dysplasia from reactive changes of urothelial cells: the morphology of the cells may overlap 10 and the diagnosis can be difficult when the tumor cells are few and/or degenerate 11 . It is even more difficult to diagnose a dysplasia in those cases in which urothelial atypia is observed in some cells but not all the morphological criteria are met for the cases to be classified as carcinoma 11 - 14 . Thus, the identification of atypical cell changes is of great importance for the correct management of these patients. However, urine cytology is operator dependent and the sensitivity for the detection of urothelial neoplasms is highly variable, as reported in the existing literature 15 , 16 . In patients with HGUC cytology has shown high sensitivity and specificity, whereas, in those patients with LGUC, the estimate of false-positives and false-negatives can be >10% 17 , 18 . Applying strict cytomorphological criteria to distinguish low grade lesions from reactive cells, the detection of false-negatives can be substantially improved 17 , 19 . The general opinion is that the only cytological approach is not sufficient for identifying the recurrence of the disease 20 . Multiple markers and urine based tests for UC have been developed 21 - 28 , which can help in the differential diagnosis 29 - 31 . Urine is in contact with the urothelium of the entire urinary tract and therefore a biomarker for detecting recurrence of the disease in urine samples would be preferable, especially if it could avoid the use of more invasive and expensive procedures 32 . This short review will focus on the employment of ProEx C marker as ancillary test to improve detection of UC in urine cytology specimens.

Minichromosome maintenance protein 2 (MCM2) and minichromosome maintenance protein 5 (MCM5) have been previously investigated as immunoassays in urothelial carcinoma 48 - 50 . A mini review by Stoeber and collaborators in 1999 reported on the possible use of MCM5 as a non-invasive, immunochemical method for the detection of UC on urinary cytological samples 50 . Numerous adjunct markers, such as CK20, p53, CD44, p16, thrombomodulin, Ki67, UroVysion and ImmunoCyt/uCyt have been evaluated as possible ancillary tests in cases of atypical urothelial cells (AUC) 31 , 48 , 51 - 53 . In particular, uCyt and UroVysion are the two ancillary tests most frequently used on exfoliative urothelial cytology. These two assays are not always processed and analyzed by cytotechnologists and/or cytopathologists, but their diagnostic evaluation requires trained and certified personnel. Besides, these tests are time-consuming and more costly to perform 54 . The first published study using ProEx C in urinary cytology 55 showed that ProEx C was very useful in stratifying patients with diagnosis of atypical urothelial cells into benign and malignant subsets. In a follow-up comparative study, ProEx C showed a high sensitivity in detecting HGUC (92%) (Figure 1 A) and a low sensitivity in LGUC (72%) (Figure 1 B). On histological sections of HGUC, ProEx C staining involves the whole thickness of the neoplastic epithelium (Figure 2 A) in contrast to LGUC where the reaction is only focal and closer to the basal layers (Figure 2 B). Therefore, the positive cells may not reach the surface for exfoliation into the urine. This observation may explain the lower sensitivity of ProEx C in LGUC 56 . Vergara-Lluri and colleagues in their study 57 demonstrated that the combination of ProEx C and uCyt ancillary tests greatly improved the sensitivity in detecting LGUCs (94%). In fact, cytology alone has a low sensitivity (5-18%) in LGUCs cases 58 . Moreover, they noted an remarkable sensitivity (92%) in detecting HGUCs using ProEx C alone or in combination with uCyt. To distinguish carcinoma in situ (CIS) from reactive atypia, McKenney and colleagues used a panel of three antigens: cytokeratin 20, p53, and CD44 10 . Their original work showed p53 positivity in 81%, CK20 over-expression in 57% and absence of CD44 reactivity in all cases of CIS. Furthermore, they showed that a diffuse, full-thickness staining with CD44 together with p53 over-expression and absence of CK20 would suggest more a reactive process. McKenney and his collaborators concluded that a combination of morphology and a panel of these 3 antibodies would be ideal to distinguish reactive from malignant urothelium. A study carried out by Yin and colleagues 59 showed that all cases of CIS of their series exhibited 100% reactivity with p16 immunostaining and a 71% in invasive UCs. Moatamed and his collaborators 56 , instead, found a 100% of ProEx C reactivity in both CIS and invasive UC. It appears that for the identification of urothelial lesions, ProEx C alone will provide good sensitivity and specificity, rather than a panel of markers. Burger and colleagues 48 used MCM2 antibody, one of the antibodies in the ProEx C cocktail, in histological samples to evaluate the risk of recurrence in bladder cancer. Its positive or negative reaction is more accurate than CK20, Ki-67 and histologic grade in the prediction of the recurrence of the disease. Liu and collaborators measured the expression of ProEx C in primary and metastatic UC, also comparing it with thrombomodulin immuhistochemical staining 60 . Both ProEx C and thrombomodulin had similar sensitivity for metastatic UC (84% vs. 77%), whereas ProEx C yielded a higher sensitivity for primary UC than thrombomodulin (93% and 72% respectively). This study demonstrated that ProEx C is useful for diagnosing primary UC but not helpful for detecting metastatic carcinoma, as it shows moderate to high expression in most of the common carcinomas such as colon, prostatic, renal cell, stomach, breast and lung carcinomas. Metastatic carcinomas can be seen in urine cytology specimens too, although very rarely 61 . Finally, Chang and colleagues, compared the utility of ProEx C and UroVysion in urine specimens 54 . They showed that ProEx C results were comparable to those previously published 55 , 57 : ProEx C displays a higher sensitivity than UroVysion for identifying UCs (88.9-55.6% respectively). In addition, positive predictive value (88.9%) and negative predicted value (77.8%) were much higher for ProEx C than those observed for UroVysion (64.3 and 30.8% respectively).

Urothelial carcinoma is an important health problem worldwide because of its silent clinical evolution, incidence and high recurrence rate. Conventional surveillance requires cystoscopy and urinary cytology. Unfortunately, cystoscopy is an invasive procedure for patients and very expensive for health care assistance. Urine cytology, although is a simple, safe and cost-effective diagnostic method of investigation, it is not highly sensitive for detection of LGUC. A wide range of non-invasive techniques have been evaluated that can improve early diagnosis, efficiency and costs of follow-up. Urinary biomarkers may also help to estimate and characterize bladder malignancies evolution. ProEx C seems to be a promising and simple adjunct device in urine cytology, especially in urine samples with either scant cellularity or with only a few atypical cells present, which can lead to cytological misinterpretation. ProEx C concretely differentiates high-grade lesions from benign reactive conditions; it helps to resolve queries regarding low-grade versus high grade UC, but it is not a useful marker in identifying metastatic UC, being also expressed in colon, stomach, breast and lung carcinomas. In conclusion, future investigations, using much larger series, will be necessary to further support and solidify these early promising findings.