Use of the Nuclear Matrix Protein 22 BladderChek Test for the Detection of Primary and Recurrent Urothelial Carcinoma

Objective To evaluate the performance of the nuclear matrix protein 22 (NMP22) BladderChek test in urothelial carcinoma (UC). Methods We retrospectively analyzed 1318 patients who performed the NMP22 BladderChek tests. Of them, 103 were primary UC patients, 90 were surgical treatment UC patients, and 1125 were benign disease patients. The performance of the NMP22 BladderChek test for the diagnosis of primary and recurrent UC was evaluated. Moreover, the performance of urine cytology and the NMP22 BladderChek test for the diagnosis of primary UC was compared in 90 available subjects including 48 primary UC patients and 42 benign disease patients. Results The sensitivity and specificity of the NMP22 BladderChek test were 37.9% and 95.8%, respectively, for the diagnosis of primary UC (n = 1228). The corresponding parameters of the NMP22 BladderChek test were 31.0% and 88.5%, respectively, for the diagnosis of recurrent UC (n = 90). The sensitivity and specificity of urine cytology were 54.2% and 97.6%, respectively, for the diagnosis of primary UC (n = 90); the corresponding parameters of the NMP22 BladderChek test were 41.7% and 83.3%, respectively; the corresponding parameters of the two tests combination were 64.6% and 83.3%, respectively. There was a significant difference in the performance between the NMP22 BladderChek test and urine cytology or the combination of two tests (P = 0.017 and 0.001, respectively). Conclusions The NMP22 BladderChek test has a low sensitivity for detecting primary and recurrent UC. Urine cytology is superior to the NMP22 BladderChek test, and combined use of the two tests improves the sensitivity in the detection of primary UC.


Introduction
Urothelial carcinoma (UC) arises from the urothelium of the lower urinary tract (urethra and bladder) or the upper urinary tract (ureter and pyelocaliceal cavities). According to the National Cancer Institute of the United States, UC accounts for the vast majority (>90%) of bladder cancers. The estimated new cases and deaths from bladder cancer were 79030 and 16870 in the United States in 2017 [1]. Compared with the United States, the numbers were 80500 and 32900 in China in 2015 [2]. The current standard method for the detection of bladder cancer is an invasive cystoscopy. Urine cytology, the most accurate noninvasive test, is a secondary method after cystoscopy. Six urine tests for bladder cancer detection (UroVysion™, Immunocyt™, BTA stat, BTA TRAK, NMP22 ELISA, and NMP22 BladderChek) have received the US FDA approval [3], although none has sufficient accuracy to displace urine cytology [4]. Among these urine tests, only the NMP22 BladderChek test is approved by the Chinese FDA. The NMP22 BladderChek test is based on the detection of NMP22, a nuclear mitotic apparatus protein that is released from dead cells (e.g., apoptosis cells). In UC cells, NMP22 is elevated concordant with the structural and morphological change characteristic of malignant cell nuclei. So, the NMP22 BladderChek test can be used for the detection of UC. The NMP22 BladderChek test has been used in the clinic in China for several years. However, evaluation of the performance of this test is lacking. In the present study, we aimed to evaluate the performance of the NMP22 BladderChek test for the diagnosis of primary UC and compare its performance with urine cytology. In addition, the performance of the NMP22 BladderChek test for the detection of recurrent UC would be evaluated.

Patient Demographics and Clinical
Characteristics. The patient demographics are shown in Table 1. Simply, the major patients with primary or recurrent UC were older men. The distributions of primary and recurrent UC according to site, stage, and grade are shown in Table 2. UC mainly occurred in the bladder, mainly in high-grade, and its stages mainly were Ta, T1, and T2. Table 3 shows the diagnostic performance of the NMP22 BladderChek test, urine cytology, and the combination of two tests for the diagnosis of primary UC.

Diagnostic Performance of the Urine Tests for the Diagnosis of Primary UC.
3.3. The Sensitivity of Urine Tests for the Diagnosis of Primary UC Stratified by Tumor Site, Invasivity, and Grade. Table 4 shows the sensitivity of the NMP22 BladderChek test and urine cytology for the diagnosis of primary UC stratified by tumor site, invasivity, and grade.

Comparison of Diagnostic Performance among the NMP22
BladderChek Test, Urine Cytology, and the Combination of Two Tests. Table 5 shows the comparison of diagnostic performance for the detection of primary UC among the NMP22 BladderChek test, urine cytology, and the combination of two tests based on clinical diagnosis.

Diagnostic Performance of the NMP22
BladderChek Test for the Detection of Recurrent UC. Table 6 shows the diagnostic performance of the NMP22 BladderChek test for the detection of recurrent UC.
3.6. The False-Positive Rate of the NMP22 BladderChek Test. For benign disease patients, the false-positive rate of the NMP22 BladderChek test was 4.2% (47/1125). The falsepositive rates of the NMP22 BladderChek test were 20.0, 2.1, 6.9, 1.5, 2.7, and 31.8%, for the patients with urinary calculi, hematuria, urinary tract infections, benign prostate diseases, benign kidney diseases, and benign bladder diseases, respectively (P < 0:001).   3 Disease Markers and specificity of the NMP22 BladderChek test were from 11% to 85% and 69.6% to 100%, respectively. There were significant differences in the performance of these two urine tests in these studies. The reasons for those discrepancies may be due to different study populations and different study designs.
In the present study, the sensitivity and specificity of the NMP22 BladderChek test for the diagnosis of primary UC in 1228 patients were 37.9% and 95.8%, respectively; the corresponding data in 90 patients were 41.7% and 83.3%, respectively. The difference in the specificity for two study populations may be due to benign disease patient selection bias. For urine cytology, the sensitivity and specificity for the diagnosis of primary UC in 90 patients were 54.2% and 97.6%, respectively. The combination of urine cytology and the NMP22 BladderChek test can increase the sensitivity to 64.6% while decreasing the specificity to 83.3%. For the detection of recurrent UC in 90 patients with surgical treatment, the sensitivity and specificity of the NMP22 Blad-derChek test were 31.0% and 88.5%, respectively. The sensitivities of the NMP22 BladderChek test and urine cytology for the detection of primary UC in our study were similar to those of O'Sullivan et al.'s study (37.9% and 56.1%, respectively) [14]. Our study shows that the sensitivities of the NMP22 BladderChek test and urine cytology for the diagnosis of primary UC are low and new more sensitive biomarkers should be used in the detection of UC. It is reported that Cxbladder has excellent sensitivity in the detection of primary and recurrent UC [14,17].
Several studies have reported that the sensitivity of the NMP22 BladderChek test and urine cytology for the detection of UC increased when the stage or the grade rose [10,18]. In our study, the sensitivity of the NMP22 BladderChek test and urine cytology for different tumor site, invasivity, and grade was analyzed. Firstly, we found that the sensitivity of the NMP22 BladderChek test in kidney UC was higher than that in bladder UC or ureter UC (69.2% vs. 30.9% or 26.7%). So, the NMP22 BladderChek test may be more applicable to use for detecting kidney UC. The sensitivity of urine cytology has no significant difference among kidney UC, ureter UC, and bladder UC. Secondly, our study showed that the sensitivity of the NMP22 BladderChek test increased as tumor invasivity or grade rose. The sensitivity of urine cytology increased as tumor grade rather than tumor invasivity rose.
Previous studies have compared the performance of the NMP22 BladderChek test and urine cytology [5][6][7][8][9][10][11][12][13][14][15][16][17][18]. Our study showed the performance of urine cytology was superior to the NMP22 BladderChek test and the combination of two tests cannot improve the performance for detecting primary UC compared to urine cytology. However, the combination of two tests can increase the sensitivity in the diagnosis of primary UC.
Studies have reported that many factors can lead to falsepositive results for the NMP22 BladderChek test [19,20]. These factors included leukocytes, current use of blood pressure control drugs, urinary calculi, creatinine, recurrent urinary tract infections, and hematuria. In the present study, we found that benign bladder diseases and urinary calculi were the two most important factors for the false-positive results of the NMP22 BladderChek test. So, it is important to consider the influencing factors for interpreting the positive result of the NMP22 BladderChek test.
This study mainly evaluates the performance of the NMP22 BladderChek test for the diagnosis of primary UC with a large sample size. However, it has several limitations. First, it is a single-center and retrospective study. Second, not all patients performed urine cytology test. In the future, we will focus on resolving these issues by conducting a prospective multicenter study.

Conclusions
In conclusion, our study demonstrates that the NMP22 Blad-derChek test has a low sensitivity for the detection of primary and recurrent UC. Urine cytology is superior to the NMP22 BladderChek test, and the combination of two tests improves the sensitivity in the detection of primary UC. Moreover, new more sensitive biomarkers should be used in UC.

Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.