ZKSCAN3 expression in urothelial carcinoma of the upper urinary tract and its impact on patient outcomes

We recently found that a zinc-finger transcription factor, ZKSCAN3, could induce the growth of bladder cancer cells. In this study, we immunohistochemically stained for ZKSCAN3 in upper urinary tract urothelial carcinoma (UUTUC) specimens. ZKSCAN3 was positive in 52 [42.4%; 26 (26.3%) weak (1+), 13 (13.1%) moderate (2+), and 3 (3.0%) strong (3+)] of 99 UUTUCs, which was significantly (P<0.001) lower than in paired normal-appearing urothelial tissues [70 (86.4%) of 81; 17 (21.0%) weak, 36 (44.4%) moderate, and 17 (21.0%) strong]. There were no statistically significant associations between ZKSCAN3 expression pattern and tumor grade, pT/pN stage, or distant metastasis. However, the positive rate of ZKSCAN3 expression was significantly (P<0.001) higher in ureteral tumors (54.0%) than in renal pelvic tumors (26.7%). Kaplan-Meier and log-rank tests revealed that the levels of ZKSCAN expression did not considerably correlate with progressionfree or cancer-specific survival in the entire cohort of the patients. Meanwhile, strong (3+) ZKSCAN3 expression in muscle-invasive tumor was correlated with the risk of disease progression (P=0.001) and cancer-specific mortality (P=0.069), while only one patient had a ZKSCAN3(3+) muscle-invasive tumor. Thus, in contrast to the observations in bladder specimens, the decreased expression of ZKSCAN3 in UUTUC, compared with non-neoplastic urothelium, was seen. Moreover, the current results failed to provide conclusive evidence suggesting the prognostic value of ZKSCAN3 expression in patients with UUTUC.


Introduction
Upper urinary tract urothelial carcinoma (UUTUC) accounts for less than 10% of urothelial cancers but is often aggressive with poor prognosis compared with bladder cancer [1]. Presumably due to the scarcity of UUTUC, the pathophysiology of this disease is not well understood, while some of molecular or genetic factors are similar to those associated with bladder cancer [1,2]. Importantly, further validation is still required to determine the prognostic values of current molecular markers for UUTUC, other than histopathological features including tumor grade and stage, lymphovascular invasion, surgical margin status, and the presence of concomitant carcinoma in situ [2]. ZKSCAN3, a family member of the KRAB and SCAN domaincontaining zinc-finger transcription factors, has been shown to involve the outgrowth of malignancies [3][4][5][6]. It has indeed been documented that overexpression or gene amplification of ZKSCAN3 in tissue specimens correlates with colorectal cancer invasion [4] or prostate cancer metastasis [4], respectively. Recently, we demonstrated that ZKSCAN3 knockdown in bladder cancer lines resulted in considerable decreases in cell viability, along with increased apoptosis, as well as in cell migration and invasion [6]. Additionally, in xenograft-bearing mice, ZKSCAN3 knockdown significantly delayed tumor formation and subsequent tumor growth [6]. Thus, ZKSCAN3 activation appeared to contribute to the promotion of tumor progression in several types nephroureterectomy specimens. We obtained some results similar to previous observations in bladder cancer tissue samples and others dissimilar to them.
The Fisher's exact test was used to evaluate the association between categorized variables. The survival rates were determined using the Kaplan-Meier method and comparison was made by the log-rank test. Tumor progression was defined as the development of nonbladder lesions, including recurrence at the nephroureterectomy site and lymph node or visceral metastasis. P values less than 0.05 were considered statistically significant.

Results
We immunohistochemically stained for ZKSCAN3 in 99 UUTUC samples and corresponding 81 normal-appearing urothelial tissues. Positive signals for ZKSCAN3 were detected predominantly in the nucleus of benign and malignant epithelial cells (Figure 1). Next we analyzed the correlations between the status of ZKSCAN3 expression in UUTUCs and the clinicopathological profile available for our patient cohort ( Table 2). There were no strong associations between ZKSCAN3 expression pattern and the laterality of tumors, gender of the patients (data not shown), histopathological tumor grade, pT stage, pN stage, or distant metastasis (data not shown). However, the positive rate of ZKSCAN3 expression was significantly higher (P=0.012) in ureteral tumors compared with renal pelvic tumors.
We then performed Kaplan-Meier analysis coupled with the logrank test to assess the prognostic values of ZKSCAN3 expression in UUTUCs. There were no statistically significant differences in tumor progression (0 vs. 1+/2+/3+, P=0.361, Figure 2A Figure 2C) or cancer-specific mortality (P=0.307, Figure 2D). However, there was a link between strong (3+) ZKSCAN3 expression in a muscle-invasive UUTUC and the risk of tumor progression (P=0.001, Figure 2E) or cancer-specific mortality (P=0.069; Figure 2F).

Discussion
In the present study, we immunohistochemically determined the expression levels of ZKSCAN3 in a set of TMA consisting of    there was a significant decrease in the positive rate of ZKSCAN3 expression in renal pelvic tumors (26.7%), compared with ureteral tumors (54.0%), differences in tissue preservation among bladder, ureteral, and renal pelvic tumors due to their anatomic locations and/ or thickness of the specimens around the tumors (e.g. time to complete tissue fixation) might have had an impact on the immunoreactivity. Of note, delay to formalin fixation of surgical specimens is known to lead to false-negative results in immunohistochemistry of transcription factors (e.g. estrogen receptor or progesterone receptor expression in breast cancer [9]). Thus, the positive rate of ZKSCAN3 expression in UUTUC may have been underestimated in the current study. Indeed, a similar tendency was seen in the expression of androgen receptor (i.e. 42% in bladder tumors [10] vs. 28% in ureteral tumors [8] vs. 11% in renal pelvic tumors [8]) as well as another transcription factor GATA3 (i.e. 86% in bladder tumors [11] vs. 68% in ureteral tumors [12] vs. 36% in renal pelvic tumors [12]) we stained using the same protocol with identical antibodies. Meanwhile, we found no significant correlations of ZKSCAN3 expression with tumor grade or stage in both UUTUC and bladder cancer.
No prognostic significance of ZKSCAN3 expression was identified in the stains of the bladder TMA [6]. Similarly, ZKSCAN3 positivity was not strongly associated with the prognosis of the entire cohort of patients with UUTUC or that in a subgroup of patients with muscleinvasive UUTUC. Nonetheless, a single case of muscle-invasive UUTUC showing strong (3+) ZKSCAN3 positivity was found to have higher risks of disease progression and cancer-specific mortality. Thus, we failed to provide conclusive evidence suggesting that ZKSCAN3 expression serves as a prognosticator in UUTUC patients, while ZKSCAN3 signals have been shown to promote the progression of urothelial tumors in preclinical models [6].
In conclusion, we observed a significant decrease in the expression levels of ZKSCAN3 in UUTUC, compared with non-neoplastic urothelium included in the same surgical specimens, as well as in those in renal pelvic tumors, compared with ureteral tumors. The clinicopathological and prognostic significance of ZKSCAN3 expression in urothelial cancer needs to be further determined in larger patient cohorts of UUTUC and bladder cancer as well as completely normal urothelial tissues. Further assessments of ZKSCAN3 functions are also required to determine its biological significance in the development and progression of UUTUC.