We designed this multi-center prospective study with the following objectives: (
The National Cancer Institute Japan estimated about 23,000 new cases of bladder cancer in Japan in 2019[
Exosomes and microvesicles are membrane vesicles secreted at an elevated level in cancer patients[
In this prospective multi-center study, urinary EV mRNA markers were longitudinally evaluated before and after TURBT. Using the samples obtained before TURBT, we performed cross-sectional validation of the mRNA markers that we previously identified and analyzed[
We designed this multicenter prospective study with the following two objectives: (
Graph: Figure 1(A) Study design and sample collection schedule. Dark grey squares indicate mandatory procedures/sample collection, while light grey squares indicate optional procedures/sample collection. (B) Patient enrollment chart based on the study design. Numbers in parentheses indicate the numbers of corresponding patients/subjects. The two sample collection phases (CSV and LV) were indicated in perforated line boxes.
To achieve these objectives, we registered patients at our institutes prior to their scheduled TUR surgeries and collected urine samples for marker validation (CSV phase) (Fig. 1A). Once pathologically confirmed as their tumors are NMIBC (i.e., LGTa, HGTa, T1, any Tis by the NCCN risk categories), we enrolled and followed up these patients for recurrence monitoring by the EV mRNA markers (LV phase) (Fig. 1A).
After the initial TURBT, indications for a second TURBT, intravesical therapy (Bacillus Calmette–Guérin (BCG) or intravesical chemotherapy (IVC)) after TURBT, and surveillance frequency until recurrence were based on each institution and urologist's discretion according to the current clinical guideline[
Bladder tumor was diagnosed by cystoscopy, urine cytology and pathological diagnosis of resected tumors. Pathologic staging was reported according to the Union for International Cancer Control (UICC) staging system (7th edition)[
We registered 316 patients including control (42 subjects) between October 2017 and November 2021 (Fig. 1B). Non-UBC (benign tumor) (N = 8) and upper tract urothelial carcinoma (UTUC) (N = 4) were excluded due to out of study scope, and 26 UBC patients were excluded due to insufficient urine sampling, which resulted in 236 UBC patients. Following three groups of subjects were enrolled as control: (
For additional CSV, an independent validation cohort of 30 UBC patients was recruited prospectively in Central Europe between June 2021 and November 2021 through a commercial biospecimen procurement service (Discovery Life Sciences, IRB approval no. DLS-BB044-V.1). Clinical data including pathology and urine cytology results and urine samples were obtained as described above.
EV mRNA assay was performed as described previously[
UroVysion assay (Abbott, IL, USA) was performed by a commercial clinical laboratory (SRL, Tokyo, Japan) following its sample collection protocol: isolating urine sediments and shipping to the laboratory at 4 °C. NMP22 ELISA kit (Abbott, IL, USA) and BTA stat kit (Polymedoco, NY, USA) were purchased and the measurement was conducted following the manufacturers' protocols.
Data analysis was conducted by R unless otherwise noted[
This study was reviewed and approved by institution review boards at Hokkaido University Hospital (IRB approval no. 017-0036), Sapporo City General Hospital (IRB approval no. R02-059–704), Hokkaido Cancer Center (IRB approval no. 017-0036), Teine Keijinkai Hospital (IRB approval no. 2-017135-01), and Sapporo Keiyukai Hospital (IRB approval no. H29-26). Informed consent was obtained from all individual participants included in the study. All methods were performed in accordance with the relevant guidelines and regulations.
UBC patients (N = 236) enrolled in this study include LGTa (27%), HGTa (29%), T1 (22%), any Tis (14%), and MIBC (7%) by the NCCN risk categories and their clinical characteristics were summarized in Table 1.
Table 1 Patients' characteristics.
UBC Control Patient, n 236 42 Age, median year (IQR) 73 (66.5—79.5) 70 (63.3—76.8) Sex, n (%) Female 73 (31%) 15 (36%) Male 163 (69%) 27 (64%) Smoking (%) No 79 (33%) 6 (14%) Yes 129 (55%) 6 (14%) N/A 28 (12%) 30 (71%) Prior UBC history (%) No 168 (71%) 37 (88%) Yes 68 (29%) 5 (12%) 1 time 11 3 2 times 17 3 times 16 1 > 3 times 17 1 N/A 7 Tumor size, n (%) ≤ 3 cm 159 (67%) > 3 cm 58 (25%) N/A 19 (8%) Tumor number, n (%) 1 105 (44%) 2 to 7 94 (40%) ≥ 8 20 (8%) N/A 17 (7%) Pathological stage (%) pTa 151 (64%) pT1 61 (26%) pTis 7 (3%) pT2 or higher 17 (7%) with pTis 28 (12%) Tumor grade (%) LG 71 (30%) HG 165 (70%) Second TUR (%) No 170 (72%) Yes 66 (28%) Intravesical therapy No 116 (49%) Yes 120 (51%) BCG 59 IVC 41 BCG + IVC 16 N/A 4 NCCN risk category (%) LGTa 64 (27%) HGTa 69 (29%) T1 52 (22%) any Tis 34 (14%) MIBC 17 (7%)
IQR, interquartile range; UBC, urothelial bladder cancer; TUR, transurethral resection; BCG, Bacillus Calmette–Guérin; IVC, Intravesical chemotherapy; LG, low grade; HG, high grade; MIBC, muscle-invasive bladder cancer; N/A, not available.
After our guideline-based surveillance during this study period (median follow up: 19 months), their 1-, 2- and 3-year recurrence-free survival were 87.7%, 75.4% and 66.9%, respectively (Table 2), and the NCCN risk category, UBC history, tumor size, tumor number, sex, and smoking status were not significant predictors for recurrence-free survival (Fig. 2A, B, Suppl Fig. 1A to 1D). The breakdown of initial TUR pathology and recurrent pathology are shown in Suppl Table 2.
Table 2 Follow-up results.
Follow-up Patients, n 189 Follow-up period, median months (IQR) 19 (9–30) Recurrence-free survival, % (95% CI) 1 year 87.7% (82.6–93.2) 2 years 75.4% (68.3–83.2) 3 years 66.9% (57.6–77.8) Recurrence, n (%) No 151 (80%) Yes 38 (20%) Recurrence pathology results (NCCN category) LGTa 9 (24%) HGTa 10 (26%) T1 5 (13%) any Tis 7 (18%) MIBC 2 (5%) UTUC 2 (5%) Clinically diagnosed 3 (8%)
IQR, interquartile range; UBC, urothelial bladder cancer; CI, confidence interval; LG, low grade; HG, high grade; MIBC, muscle-invasive bladder cancer; UTUC, upper tract urothelial cancer.
Graph: Figure 2Recurrence-free survival curve analysis of the LV phase. A Kaplan–Meier survival curve analysis was conducted by NCCN risk categories (A) and bladder cancer history (B).
For the first study objective, we aimed for a CSV of the EV mRNA markers we identified in our previous study[
First, we investigated if the previously identified reference gene, ALDOB, is useful to normalize the EV mRNA markers in urine in this study cohort. ALDOB and other candidate reference genes including GAPDH and ACTB were analyzed by ANOVA in the CSV phase (Suppl Table 2). ALDOB was not only one of the most highly expressed genes among the tested genes but also no change of its expression levels was observed among absence and presence of NMIBC/MIBC or among the NCCN risk categories, which satisfy the requirements of a reference gene. On the other hand, GAPDH and ACTB, were differentially expressed in urinary EV mRNA although their expression levels in urine was high, therefore those conventional reference genes were not ideal for normalization in this urinary EV mRNA analysis. These data corroborated our previous study result and re-confirmed that ALDOB expression is high and stable in urine EV independent of bladder tumor status and can be used to normalize the EV mRNA marker expression levels as a reference gene.
Next, the EV mRNA marker expression, which was normalized by ALDOB, was analyzed in the CSV phase (Fig. 3). In this study, we focused on the three markers we identified in our previous study, KRT17, GPRC5A and SLC2A1[
Graph: Figure 3EV RNA expression in the CSV phase. Expression levels of EV RNA markers in the CSV phase are shown in boxplot graphs. Welch's t-test was performed to compare the significance of differences between diagnostic categories based on NCCN risk categories: p < 0.05*, p < 0.005**, p < 0.0005***. Perforated lines indicate the best diagnostic threshold of each gene to detect UBC, which was defined in the top left corner of each ROC curve (Table 3).
Table 3 Diagnostic performance in the CSV phase.
Markers AUC (95% CI) Sensitivity Specificity PPV NPV 0.760 (0.699–0.821) 0.633 0.786 0.94 0.289 0.730 (0.663–0.796) 0.62 0.81 0.945 0.288 Cytology 0.721 (0.614–0.829) 0.596 0.846 0.984 0.116 NMP22 0.692 (0.617–0.766) 0.56 0.824 0.945 0.257 0.674 (0.602–0.746) 0.534 0.81 0.937 0.248 Microhematuria 0.669 (0.562–0.777) 0.493 0.846 0.981 0.096 0.660 (0.581–0.738) 0.606 0.69 0.912 0.25 BTA 0.659 (0.581–0.737) 0.628 0.69 0.912 0.266 0.613 (0.531–0.695) 0.701 0.524 0.886 0.25
AUC, area under the curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value.
Diagnostic performance of these EV mRNA markers was estimated by ROC curve analysis and compared against that of urine cytology, microhematuria, and FDA cleared biomarkers (BTA and NMP22) (Table 3). Among the biomarkers, MDK showed the best AUC, 0.760, for the detection of UBC, followed by KRT17 with AUC 0.730, and both EV mRNA markers outperformed conventional markers such as cytology (AUC 0.721), NMP22 (AUC 0.692), microhematuria (AUC 0.669) and BTA (AUC 0.659).
To validate these EV mRNA markers further, we recruited a validation cohort of 30 subjects with prior UBC history in Central Europe. The patient characteristics of this cohort was summarized in Suppl Table 2. This cohort consists of 12 UBC subjects with LGHa (25%), HGTa (25%), T1 (42%) and MIBC (8%) as well as 18 control subjects who had previous UBC history and were at least six-month recurrence free. The EV mRNA markers as well as the other biomarkers were measured as described in the "Material and methods" section. In the Central Europe cohort, the EV mRNA markers were highly expressed in UBC patients compared to the control subjects (Suppl Fig. 3), which support the above CSV phase results (Fig. 3) as well as our previous study results[
Taken together, these two CSV suggested that the urinary EV mRNA markers, especially KRT17 and MDK, are promising biomarkers for the detection of UBC and may be superior to traditional biomarkers such as cytology, microhematuria and other FDA-cleared biomarkers although further validation is required.
For the second study objective, we aimed for a LV of the EV mRNA markers in a surveillance setting after TURBT, assuming such markers will help earlier detection of cancer recurrence and/or reduce a frequency of cystoscopy procedures eventually. In this LV phase, urine samples were collected for the EV mRNA marker measurement (median 5 samples per patient) while the enrolled patients were under our guideline-based surveillance for cancer recurrence (Fig. 1A). Time courses of these marker expression in the LV phase were analyzed by Loess trend-line analysis (Fig. 4). For those without cancer recurrence during the study period (N = 151, "No Rec"), the expression of the KRT17, GPRC5A, and MDK remained similar to the thresholds for UBC detection, which was predetermined in the CSV phase of the study (Table 3), almost for 12 months after the TURBT, and gradually decreased over time and went below the thresholds only after 12–24 months. On the other hand, the time courses of SLC2A1 and CXCR2 indicated disappointing results indicating those markers may not be helpful to monitor the cancer recurrence. SLC2A1 expression continuously increased over time. CXCR2 expression was at least fourfold higher than the threshold for UBC detection immediately after the TURBT and decreased over time and required 30 months to go below the threshold.
Graph: Figure 4EV RNA expression and other markers in the LV phase. (A) Loess trendline curve analysis of EV RNA marker expression levels in the LV phase. Red lines indicate patients with no recurrence (N = 151) and blue lines indicate patients with cancer recurrence (N = 38). Follow-up times indicate months after first TUR. Perforated lines indicate the best threshold of each gene to detect UBC in the CSV phase (Table 3). (B) Loess trendline curve analysis of cytology and microhematuria. Score 1 was assigned for test positives and score 0 for test negatives. Definitions of test positives and negatives are described in the "Material and methods" section. (C) Loess trendline curve analysis of the average scores of EV KRT17 and cytology/microhematuria. Regarding EV KRT17 , score 1 was assigned for test positives when the expression level was above the threshold (perforated line in Fig. 4A) and score 0 for test negatives when the expression level was below the threshold. Average scores were obtained by arithmetic means of each score for EV KRT17 , cytology, and/or microhematuria.
To understand why these EV mRNA markers were expressed at the levels near the best thresholds for UBC detection for more than 12 months after TUR when tumors were supposedly eradicated, "No Rec" group was split into sub-groups by TUR pathology results, tumor size, tumor number or post-TUR therapies and time courses of the EV mRNA expression was compared among the sub-groups (Suppl Fig. 4). Although overall trends are similar to each other, several distinct patterns were observed. For example, KRT17 and CXCR2 expression levels were elevated in those who had more aggressive tumors during the first 12 months after TUR, i.e., LGTa < HGTa = any Tis < T1. Similarly, the EV mRNA expression was elevated in those with larger or greater number of tumors than those with smaller or less (Suppl Fig. 4B, 4C). These observations are interesting because the EV mRNA expression seems to be affected by tumor status even after tumors were eradicated, and more aggressive tumors leave the EV mRNA elevated even after TUR. Lastly, second TUR and intravesical therapy, which are generally provided during the first 6 months after initial TUR, seem to increase the EV mRNA expression levels higher may be due to more aggressive tumors or intravesical immune response, though it is inconclusive whether these therapies directly affect the expression levels (Suppl Fig. 4D).
On the other hand, for those with cancer recurrence in the LV phase (N = 38, "Rec"), KRT17 expression was significantly elevated compared to not only the best threshold for UBC detection but also the expression level in the "No Rec" throughout the study period (p < 0.0001) (Fig. 4A). The difference of KRT17 expression between "Rec" and "No Rec" increased over time especially after about 12 months. In addition, MDK was also expressed significantly higher in the "Rec" group than in the "No Rec", time to time (p = 0.0224). For comparison, time courses of the cytology and microhematuria results in categorical scores (0 for negatives and 1 for positives) were analyzed (Fig. 4B). In the "No Rec" group, both cytology and microhematuria were close to score 0, or mostly negative, throughout the time, while BTA results were close to score 0.5, a mixture of positives and negatives, during the first 6 months after TUR and gradually decreased to score 0 over time. In the "Rec" group, on the other hand, cytology, microhematuria and BTA scores exceeded those in the "No Rec" group throughout the time.
Diagnostic performance of the EV mRNA markers for cancer recurrence detection was evaluated in the LV phase (Table 4A). KRT17 outperformed the other markers with AUC 0.653 and NPV 91.6%. Compared to the cytology (AUC 0.603 and NPV 86.6%), KRT17 could be a good alternative and may be useful to rule out those with no risk of cancer recurrence because of the relatively high NPV. Other EV mRNA such as CXCR2 and MDK performed better than cystoscopy with AUC 0.628 and 0.606, respectively, while BTA and UroVysion were less promising with AUC 0.565 and 0.488, respectively. Since the EV mRNA expression remained close to the threshold levels for 12–24 months after the TUR, diagnostic performance was evaluated in the LV phase excluding the first 6 or 12 months after the TUR (Table 4B, C). Diagnostic performance of KRT17 improved to AUC 0.667 and NPV 93.8% by excluding the first 6 months and AUC 0.700 and NPV 96.6% by excluding the first 12 months. Thus, KRT17 could be used to rule out those without cancer recurrence more accurately especially after 6 to 12 months after TUR.
Table 4 Diagnostic performance in the LV phase.
A. Follow-up period (≥ 0 month) Markers AUC (95% CI) Sensitivity Specificity PPV NPV 0.653 (0.601–0.706) 0.764 0.519 0.243 0.916 0.628 (0.572–0.683) 0.664 0.594 0.248 0.898 NMP22 0.625 (0.300–0.950) 0.5 0.75 0.5 0.75 0.606 (0.548–0.664) 0.555 0.64 0.237 0.877 Cytology 0.603 (0.555–0.651) 0.283 0.924 0.426 0.866 Microhematuria 0.583 (0.534–0.633) 0.337 0.83 0.287 0.861 BTA 0.565 (0.473–0.658) 0.548 0.583 0.397 0.721 0.561 (0.504–0.618) 0.582 0.558 0.21 0.869 0.531 (0.476–0.586) 0.536 0.521 0.184 0.848 UroVysion 0.488 (0.464–0.512) 1 0 0.109 NaN 0.697 (0.641–0.753) 0.8 0.536 0.25 0.933 0.671 (0.614–0.727) 0.719 0.59 0.257 0.914 0.655 (0.600–0.710) 0.723 0.565 0.249 0.911
Table 4 Diagnostic performance in the LV phase.
B. Follow-up period (≥ 6 months) Markers AUC (95% CI) Sensitivity Specificity PPV NPV 0.667 (0.607–0.726) 0.79 0.548 0.233 0.938 0.620 (0.555–0.685) 0.642 0.621 0.227 0.909 Cytology 0.600 (0.547–0.653) 0.264 0.936 0.422 0.878 0.595 (0.530–0.661) 0.531 0.642 0.205 0.888 Microhematuria 0.576 (0.520–0.633) 0.315 0.838 0.25 0.877 BTA 0.569 (0.448–0.690) 0.458 0.68 0.407 0.723 0.565 (0.500–0.630) 0.568 0.58 0.19 0.886 0.547 (0.483–0.612) 0.469 0.625 0.178 0.872 UroVysion 0.488 (0.464–0.512) 1 0 0.109 NaN 0.697 (0.634–0.760) 0.788 0.562 0.233 0.94 0.675 (0.612–0.738) 0.7 0.618 0.243 0.922 0.660 (0.598–0.722) 0.718 0.584 0.228 0.924
Table 4 Diagnostic performance in the LV phase.
C. Follow-up period (≥ 12 months) Markers AUC (95% CI) Sensitivity Specificity PPV NPV 0.700 (0.620–0.779) 0.825 0.573 0.182 0.966 0.628 (0.534–0.722) 0.575 0.712 0.187 0.936 Cytology 0.625 (0.549–0.700) 0.297 0.952 0.423 0.92 0.618 (0.527–0.709) 0.625 0.651 0.171 0.938 0.615 (0.524–0.706) 0.625 0.597 0.152 0.932 BTA 0.611 (0.413–0.810) 0.444 0.778 0.5 0.737 0.603 (0.516–0.690) 0.5 0.674 0.15 0.921 Microhematuria 0.530 (0.463–0.596) 0.189 0.87 0.143 0.904 UroVysion 0.486 (0.458–0.514) 1 0 0.103 NaN 0.718 (0.637–0.799) 0.8 0.603 0.19 0.963 0.714 (0.629–0.800) 0.722 0.659 0.198 0.953 0.667 (0.586–0.748) 0.722 0.625 0.179 0.952
AUC, area under the curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value.
To further improve the diagnostic performance, we employed a simple algorithmic analysis of KRT17 in combination with cytology and/or microhematuria results. For KRT17, we assigned score 1 (positive) when its expression level is above the threshold determined in the CSV phase and score 0 (negative) when the expression is below the threshold. Averaged scores of KRT17, cytology and/or microhematuria were analyzed, which represents the average number of positive markers out of each combination (Fig. 4C). Diagnostic performance was improved to AUC 0.697, 0.697 and 0.718 by adding cytology and hematuria results to KRT17 although NPV wasn't improved (Table 4).
Lastly, predictive value of KRT17 was investigated. First, KRT17 in urine obtained prior to TUR (CSV phase) was not predictive for the post-TUR recurrence-free survival (Fig. 5A). KRT17 during the first 6 months after TUR (LV phase) was not predictive either (log rank p > 0.05), however at least those with KRT17 positives indicated slightly poor prognosis compared to those with KRT17 negatives (Fig. 5B). On the other hand, the average score of KRT17, cytology and microhematuria during the first 6 months after TUR was clearly able to predict recurrence-free survival (Fig. 5C) while each marker alone did not (Fig. 5B, Suppl Fig. 5A, B). Therefore, KRT17 in combination with other conventional biomarkers may have a predictive value of cancer recurrence although further validations are necessary.
Graph: Figure 5Recurrence-free survival curve analysis of the LV phase. Patients were stratified by EV KRT17 in urine obtained before TUR (CSV phase) (A) or during the first 6 months after TUR (LV phase) (B) using the threshold defined in the CSV phase (Table 3), and recurrence-free survival was compared between EV KRT17 positives and negatives. Patients were also stratified by the average score of EV KRT17 , cytology, and microhematuria during the first 6 months after TUR (LV phase) using a cut-off of 0.5 (C). The significance of differences was assessed by the Log-rank test. Censored patients were indicated by crosses in the curves. (A) EV KRT17 , CSV phase (B) EV KRT17 , LV phase (<6 months) (C) Combination (EV KRT17 +Cytology+Microhematurial), LV phase (<6 months)
We designed this study with the two objectives: 1. CSV of the EV mRNA markers for UBC detection and 2. LV of the EV mRNA markers for NMIBC recurrence monitoring. For the first objective, we were able to confirm the EV mRNA markers were highly expressed in urine from UBC patients especially with higher stage/grade tumors compared to those without UBC in the CSV phase (N = 278 in total) and Central Europe cohorts (N = 30 in total) (Fig. 3, Suppl Fig. 1). We also confirmed that the EV mRNA markers especially MDK and KRT17 outperformed urine cytology and other conventional markers in both cohorts (Table 3, Suppl Table 4). Those data support the validity of the EV mRNA markers for UBC detection. For the second objective, we investigated the time courses of the EV mRNA markers prospectively while NMIBC patients (N = 189) were under our surveillance in the LV phase. One of the EV mRNA markers, KRT17, was confirmed to be expressed higher in those with cancer recurrence after TURBT until recurrence, while its expression gradually decreased over a time for those without cancer recurrence during the study period (Fig. 4A). Indeed, NPV of KRT17 obtained in this study was 91.6%, 93.8% and 96.6% for > 0 month, > 6 months and > 12 months after TURBT, respectively, thus it may be helpful to reduce surveillance frequency/duration using this marker although further validation studies are necessary (Table 4). These data suggest that the EV mRNA markers especially, KRT17, is a promising biomarker for UBC detection and recurrence monitoring.
It is intriguing that KRT17, and other mRNA expression remained high for more than 12 months after TURBT in those with no recurrence despite tumors were eradicated by TURBT (Fig. 4A), which may hamper the use of these markers immediately after TURBT. Our analysis indicated that those who had more aggressive tumors or those who received second TURBT tend to have high EV mRNA expression even after TURBT (Suppl Fig. 4). In those who had cancer recurrence, KRT17 expression was high already before cancer recurrence was detected by cystoscopy. Thus, the source of KRT17 and other markers to be further investigated in the future study. On the other hand, these data suggest KRT17 may be clinically more useful for selected patient populations such as those with less aggressive tumors.
KRT17 has been reported to be overexpressed in many types of cancer including UBC, breast cancer, colon cancer, non-small lung cancer, cervical cancer, oral cancer, esophagus cancer, pancreatic cancer, etc. KRT17 overexpression is associated with poor prognosis and cancer progression in non-small lung cancer[
Achieving the good performance for predicting recurrence is challenging, because the recurrent lesions under the follow-ups are usually very small to discharge EVs into urine. This is consistent with the result in which most of tumor size was less than 3 cm at the time of the recurrence. However, KRT17 can discriminate the tumor recurrence after 12 months after surgery. The discrimination ability was improved as the follow-up was prolonged. This discrimination ability might be due to the elimination of insignificant perioperative effect. Additionally, we created a simple algorithm combining KRT17 and conventional biomarkers such as urine cytology and microhematuria. We chose these two markers because of the high sensitivity of microscopic hematuria and the high specificity of urine cytology. In fact, this combination improved the diagnostic performance. Furthermore, with the recent advancement in this field, recurrence monitoring could be further improved in combination with other biomarkers such as urinary cells and circulating tumor DNA. Urinary cells could be analyzed from whole urine together with EV RNA, therefore new urinary cell RNA assay[
There are several limitations in this study. First, despite overall trend of the EV mRNA expression time course looks promising, the time course in each case looks noisy and difficult to take actual clinical actions based on the biomarker data yet. In the future, larger scale and long-term validation studies are necessary. In addition, new laboratory technologies, such as droplet digital PCR, which allow more accurate quantification, especially at very low concentrations, may overcome this problem, as mRNA expression levels in urinary EV are much lower than in tissue. In addition, the predictors with significant differences in the other reports were not necessarily significant factors in our study because a portion of patients were excluded due to radical cystectomy. Diagnostic performance was compared to UroVysion and NMP22, however sample sizes of these assays especially in the LV phase were limited (N = 46 and N = 12, respectively). EV ALDOB was selected for a reference gene to normalize EV mRNA marker expression levels because we confirmed EV ALDOB level was stable in our studies. However, ALDOB is highly expressed in kidney therefore EV ALDOB levels in urine may fluctuate depending on the patients' kidney functions, which we were not able to investigate in this study because of limited clinical data.
Taken together, these data suggest that the EV mRNA markers especially, KRT17, is a promising biomarker for UBC detection and recurrence monitoring though further validations especially in real-life clinical settings are necessary.
We thank for Dr. Hiroki Chiba (SGH), Dr. Masaya Miyazaki (SGH), Dr. Kiyoto Nagamori (SGH), Dr. Hiroshi Tanaka (SGH), Dr. Satoshi Nagamori (HCC), Dr. Takuya Moriguchi (HCC), Dr. Chihiro Kamijo (HCC), Dr. Naoto Miyajima (TKH), Dr. Toshiki Koyama (TKH), Dr. Kanta Hori (TKH), Dr. Kazushi Hirakawa (SKH) for collecting clinical data and samples and providing those to this study. We also thank for Ms. Yuka Tanaka (HUH) for preparing and shipping the clinical samples. We thank for Discovery Life Sciences team to support this study by recruiting patients and organizing their clinical specimen and clinical data. We thank for Showa Denko Material America co-workers, especially Ms. Mieko Ogura and Dr. Cindy Yamamoto for supplying the assay reagents and having thoughtful discussions and suggestions. We also thank for Dr. Melanie Oakes, University of California Irvine and Dr. Masato Mitsuhashi, Nanosomix, for providing thoughtful suggestions/advice.
T.M., T.O., and H.H. conceptualized and designed the study. T.O., S.M., K.M., T.H., A.K., and Y.S. prepared and obtained IRB approval. T.O., H.H., T.H., N.T., K.M., A.K., H.M., Y.S., R.M., H.K., T.A., S.M., and N.S. recruited patients and collected clinical data and urine samples. TM performed biomarker assays and analyzed study data. Y.I. advised and supervised statistical analyses. T.M. and T.O. discussed the data and wrote the manuscript. All the authors reviewed the manuscript.
All data generated or analysed during this study are included in this published article and its supplementary information files.
The authors of this manuscript have the following conflicts of interest to disclose. TM is an employee of Showa Denko Materials America, formerly known as Hitachi Chemical Co. America. The present study was funded by the Japan Society for the Promotion of Science (KAKENHI, Grant Number 22K09439) and Showa Denko Materials America. Parts of this study were presented in American Urological Association annual meetings in 2019 and 2020. Parts of this study were presented in American Urological Association annual meetings in 2019 and 2020.
Graph: Supplementary Table 1.
Graph: Supplementary Table 2.
Graph: Supplementary Table 3.
Graph: Supplementary Table 4.
Graph: Supplementary Table 5.
Graph: Supplementary Figure 1.
Graph: Supplementary Figure 2.
Graph: Supplementary Figure 3.
Graph: Supplementary Figure 4.
Graph: Supplementary Figure 5.
• EV
- Extracellular vesicles
• TUR
- Transurethral resection of bladder tumor
• AUC
- Area under the curve
• UBC
- Urothelial bladder cancer
• qPCR
- Quantitative polymerase chain reaction
The online version contains supplementary material available at https://doi.org/10.1038/s41598-024-55251-x.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author