Back

Platinum Priority – Prostate Cancer
Editorial by Benedikta S. Haflidadóttir and Yvonne Ceder on pp. 42–43 of this issue

Exosomal miR-1290 and miR-375 as Prognostic Markers in Castration-resistant Prostate Cancer

By: Xiaoyi Huang a b , Tiezheng Yuan a , Meihua Liang c , Meijun Du a , Shu Xia a d , Rachel Dittmar a , Dian Wang e , William See f , Brian A. Costello g , Fernando Quevedo g , Winston Tan h , Debashis Nandy g , Graham H. Bevan i , Sherri Longenbach g , Zhifu Sun j , Yan Lu k , Tao Wang l , Stephen N. Thibodeau m , Lisa Boardman g , Manish Kohli g lowast and Liang Wang a lowast

European Urology, Volume 67 Issue 1, January 2015, Pages 33-41

Published online: 01 January 2015

Keywords: Exosome, microRNA, Extracellular RNA, RNA sequencing, Prostate cancer, Biomarker, Prognosis, Survival

Abstract Full Text Full Text PDF (2,6 MB) Patient Summary

Abstract

Background

Extracellular microRNAs (miRNAs) embedded in circulating exosomes may serves as prognostic biomarkers in cancer.

Objective

To identify and evaluate plasma exosomal miRNAs for prognosis in castration-resistant prostate cancer (CRPC).

Design, setting, and participants

RNA sequencing was performed to identify candidate exosomal miRNAs associated with overall survival in a screening cohort of 23 CRPC patients. Candidate miRNAs were further evaluated for prognosis using quantitative real-time polymerase chain reaction in a follow-up cohort of 100 CRPC patients.

Outcome measurements and statistical analysis

Cox regression and Kaplan-Meier survival analyses were used to evaluate survival association using candidate miRNAs along with clinical prognostic factors.

Results and limitations

RNA sequencing in screening cohort generated approximately 6.80 million mappable reads per patient. Of those with normalized read counts ≥5, 43% were mapped to miRNAs for a total of 375 known and 57 novel miRNAs. Cox regression analysis identified an association of miR-1290, -1246, and -375 with overall survival (false discover rate < 0.05). Of those, higher levels of miR-1290 and -375 were significantly associated with poor overall survival (p < 0.004) in the follow-up cohort. Incorporation of miR-1290/-375 into putative clinical prognostic factors-based models in CRPC stage significantly improved predictive performance with a time-dependent area under the curve increase from 0.66 to 0.73 (p = 6.57 × 10−6).

Conclusions

Plasma exosomal miR-1290 and miR-375 are promising prognostic biomarkers for CRPC patients. Prospective validation is needed for further evaluation of these candidate miRNAs.

Patient summary

In this study, we evaluated whether small RNAs circulating in blood could be used to predict clinical outcomes in late-stage prostate cancer patients. We identified two blood-based small RNAs whose levels showed significant association with survival. Our results warrant further investigation because the noninvasive blood-based test has great potential in the management of late-stage prostate cancer.

Take Home Message

We identified and verified two exosomal microRNAs (miRNAs) showing a significant association with overall survival in castration-resistant prostate cancer patients. The blood-based extracellular miRNA test is noninvasive and has great potential to serve as a prognostic biomarker for late-stage prostate cancer.

Keywords: Exosome, microRNA, Extracellular RNA, RNA sequencing, Prostate cancer, Biomarker, Prognosis, Survival.

1. Introduction

Prostate cancer (PCa) is the second leading cause of cancer-related mortality in men, with >240 000 newly diagnosed cases and 29 000 deaths annually in the United States [1] . In advanced metastatic stages, androgen deprivation therapy (ADT), a standard first-line therapy, leads to substantial cancer control and palliation that lasts for a median time of 20–30 mo[2] and [3]. After failure of ADT and emergence of a castration-resistant stage, circulating tumor cell counts (CTCs) are the only prognostic biomarker approved by the US Food and Drug Administration. This test is technically challenging and expensive with prognosis adversely affected for patients with >5 CTCs per 7.5 ml blood[4] and [5]. With the increased availability of therapeutic options for this stage, access and clinical utility of easy-to-use prognostic and predictive biomarkers has become clinically relevant. Cell-free nucleic acid profiling such as exosomal RNA in circulatory fluids offers an attractive potential for prognostic and predictive biomarker development [6] .

Exosomes are small (30–100 nm) membrane vesicles released into the extracellular environment on fusion of multivesicular bodies with the plasma membrane [7] . It is reported that exosomes contain RNAs selectively enriched from parent cells [8] . A number of plasma exosomal RNAs are reported as diagnostic, prognostic, or even therapeutic biomarkers in cancer patients [6] . miR-107, miR-574-3p, miR-375 and miR-141 are upregulated in metastatic PCa patients [9] . Circulating microRNA (miRNA)-375 and miRNA-141 are correlated with a high Gleason score or lymph-node positive status[10] and [11], implying great potential as a PCa biomarker. We recently developed a comprehensive pipeline for exosomal RNA profiling analysis including exosome isolation, RNA extraction, library preparation, sequencing, and bioinformatics analyses[12] and [13]. In this study, we applied this pipeline to examine exosomal RNA profiles based on the hypothesis that plasma exosomal RNAs can be used to predict survival in castration-resistant prostate cancer (CRPC) patients.

2. Material and methods

2.1. Patients and samples

Patients consented to participate in an advanced PCa biomarker registry prior to blood collection at the time of ADT failure. For plasma specimens, whole blood collection was first centrifuged at 2000 rpm for 10 min. The supernatant was then fractioned into multiple aliquots for storage at −80 °C. A screening cohort (n = 23) and a follow-up cohort (n = 100) were included. ADT failure was defined as the date of the first occurrence of either two serial prostate-specific antigen (PSA) rises during continuous ADT measured at least 1 wk apart with an absolute PSA level >2 ng/ml [14] or the appearance of new image-based metastasis during ADT or initiation of castration-resistant specific therapy to ongoing ADT and in the presence of subcastration testosterone levels (<50 ng/dl). Table 1 lists the demographic characteristics of the screening and follow-up cohorts.

Table 1 Clinical characteristics of castration-resistant prostate cancer patients in both cohorts

Screening cohort Follow-up cohort
Total number 23 100
Age at the time of CRPC and specimen draw, yr, median (range) 69.7 (53–86) 73.3 (43–92)
Gleason score at initial diagnosis
 5 0 3
 6 0 10
 7 10 32
 8 4 19
 9 9 26
 10 0 3
 Unknown 0 7
Clinical TNM staging at initial diagnosis
 T1 0 5
 T2 12 36
 T3 9 41
 T4 1 1
 Tx 1 17
 N0 7 34
 N1 7 15
 N2 1 4
 Nx 8 47
 M0 17 82
 M1 6 18
PSA at initial diagnosis, ng/ml, median (range) 10.8 (2.5–136.5) 21.4 (0.1–2324)
No. of patients treated with radical prostatectomy on initial diagnosis 12 49
No. of patients treated with radiation alone on initial diagnosis 3 25
No. of patients treated with radical prostatectomy and radiation on initial diagnosis 2 8
No. of patients who underwent salvage local treatments after primary prostate treatments 6 28
Time from initial treatments for localized stage disease to disease progression, mo, median * 12.94 26.47
Time from initial diagnosis to initiation of ADT for hormone-sensitive stage, mo, median (range) 26.89 (0–264) 12.2 (0–222.7)
Time from initiation of ADT for hormone-sensitive stage to progression for CRPC stage, mo, median (range) 19.08 (0.76–104) 19.4 (0.6–202.2)
Follow-up time from date of CRPC specimen collection to last follow-up, mo, median (range) 35.6 (7–47) 17.58 (1–48)
No. of patients dead during follow-up 10 37

* Disease progression in subjects diagnosed and treated for localized stage disease was defined by treating physician on the basis of either serial PSA rise on two separate occasions after achieving a posttreatment nadir PSA value or appearance of new radiologic disease or with the initiation of a new cancer-specific intervention, whichever came first.

ADT initiation for hormone-sensitive stage was defined as per treating physician and included either serial PSA rise on two separate occasions after achieving a nadir PSA value or appearance of new radiologic disease or with the initiation of a new cancer-specific intervention, whichever came first.

Progression to CRPC stage during ADT for hormone-sensitive stage was defined as per treating physician and included either serial PSA rise on two separate occasions after achieving a nadir PSA value during ongoing continuous ADT or appearance of new radiologic disease or with the initiation of a new cancer-specific intervention, whichever came first.

ADT = androgen-deprivation therapy; CRPC = castration-resistant prostate cancer; PSA = prostate-specific antigen.

For the follow-up cohort, patients were recruited between January 2010 and September 2012. At enrollment time, median PSA, hemoglobin levels, and serum lactate dehydrogenase (LDH) were 20.65 ng/ml (range: 0–2324), 12.7 g% (range: 9.2–15.6), and 192 IU/l (range: 92–500), respectively. Forty-eight patients received additional systemic therapies. A total of 31 patients underwent one systemic therapy regimen (29 patients with docetaxel chemotherapy alone and 2 patients receiving only abiraterone acetate and prednisone). Thirteen patients received two systemic treatments with nine receiving docetaxel followed by abiraterone acetate, three receiving cabazitaxel as second-line treatment after docetaxel, and one patient receiving mitoxantrone after docetaxel failure. Four patients received three systemic regimens including three patients with docetaxel, abiraterone acetate plus prednisone, and mitoxantrone and one patient with cabazitaxel instead of mitoxantrone as third-line therapy.

To determine the endogenous reference controls for quantitative reverse transcription polymerase chain reaction (qRT-PCR), we examined RNA sequencing profiles from plasma exosomal RNAs in 192 subjects including 50 healthy individuals, 100 colorectal cancer patients, 6 pancreatic cancer patients, and 36 PCa patients (23 were CRPC patients used in the study). Men and women were evenly distributed in these subjects except for PCa. This study was approved by the Mayo Clinic and the Medical College of Wisconsin institutional review board.

2.2. Exosomal RNA extraction and sequencing library preparation

Exosomal isolation, RNA extraction, and sequencing library construction were reported previously [12] . Small RNA libraries were constructed with 2 ng exosomal RNA. Index libraries were equally pooled for sequencing (Illumina HiSeq2000 platform).

2.3. Sequencing data analysis

A graphic user interface-based tool to process the sequencing data has been developed [13] . After initial data cleaning, sequences with a length ≥16 nt were aligned against miRBase (Release 19, 2043 entries) and human genome (Release 103). miRNA profiling was normalized using read counts per million mappable miRNA sequences. miRDeep2 [15] was used to identify novel miRNAs from sequencing data. Predicted miRNAs with scores ≥2 were considered significant.

2.4. Quantitative reverse transcription polymerase chain reaction

To validate miRNAs identified by RNA sequencing, TaqMan MicroRNA Assays (Life Technologies) and miScript SYBR Green PCR Kit (Qiagen, Valencia, CA, USA) were used to measure expression levels of miR-30a/e-5p, miR-99a-5p, let-7c, miR-1246, miR-1290, miR-16-5p, miR-125a-5p, and miR-375 in the plasma exosomal RNAs. Among those, miR-30a/e-5p, miR-99a-5p, miR-16-5p, miR-125a-5p, and let-7c were evaluated as candidates for normalization control. PCR reaction was performed in triplicate and analyzed using an ABI 7900HT system. Qiagen miScript II RT kit was also used for validation of miR-375, miR-30a/e-5p, and miR-16-5p.

2.5. Identification of endogenous references

In the screening stage, the normalized sequencing data from 192 individuals (23 with CRPC and an additional 169 healthy and nonhealthy subjects) were first screened for the most stably expressed miRNAs with an overall coefficient of variation (CV) <5%. Selected miRNAs were further processed by NormFinder [16] and BestKeeper [17] . Diagnostic power withp < 0.001 was considered as the reference candidate. In the follow-up cohort, raw CTvalues derived from real-time PCR were used to determine their stabilities.

2.6. Survival analysis

Overall survival was recorded from the date of plasma collection after ADT failure to death or last follow-up. Association between exosomal RNA levels and overall survival was assessed using Kaplan-Meier survival curves and Cox regression in Partek Genomics Suite 6. For the multivariate analysis, risk score was calculated by a linear combination of the ΔCTvalues from real-time PCR for selected miRNAs and/or clinical factors, weighted by their estimated regression coefficients. Time-dependent receiver operating characteristic (ROC) curve analysis was used to compare the predictive performance of the risk scores for the combined miRNAs and/or clinical variables. The area under the ROC curve (AUC) was used to evaluate survival prediction.

3. Results

3.1. Mapping of RNA sequencing

For the 23 RNA sequencing libraries, we received an average of 13 million raw reads per library, where reads with insert lengths >16 nt (query sequences) were approximately 10 million (76.60%). Of these, 6.7 million reads (51.42%) were mapped to known RNA species and the human genome. Among the mapped reads, mature miRNAs were the most common with an average of 2 784 360 reads (41.72%), followed by piwi-interacting RNA with 1 396 493 (20.92%), long noncoding RNA with 1 347 486 (20.19%), and messenger RNA with 434 955 (6.52%). Six other RNA species collectively constituted 10.64% of the mappable sequences ( Fig. 1 ; Supplementary Table 1). In this study, we reported our analysis on 375 common miRNAs and 57 putative miRNAs with >5 counts per million mappable reads (Supplementary Table 2). The raw sequencing data were deposited in the Gene Expression Omnibus database (accession number: GSE58410).

gr1

Fig. 1 RNA species and their distributions in plasma exosomes. Raw reads are the sequences detected by RNA sequencing. Query reads are those after trimming. Mappable reads are those mapped to known human RNA or genome. lncRNA = long noncoding RNA; miRNA = micro RNA; piRNA = piwi-interacting RNA; rRNA = ribosomal RNA; snRNA = small nuclear RNA; snoRNA = small nucleolar RNA; tRNA = transfer RNA.

3.2. Association of overall survival with exosomal microRNAs

To identify survival-related exosomal RNAs in CRPC patients, we further excluded those miRNAs with log2 transformed read counts <5. We performed Cox regression analysis using the remaining 248 known and 19 predicted miRNAs. We identified three miRNAs with apvalue <0.0008 (false discovery rate <0.05) including miR-1290 (hazard ratio [HR]: 1.79; 95% confidence interval [CI], 1.30–2.48), miR-1246 (HR: 8.04; 95% CI, 2.36–27.33), and miR-375 (HR: 2.69; 95% CI, 1.52–4.77). In addition, we also observed an association of other RNA types with survival (Supplementary Table 3). For miRNAs, we dichotomized these patients using miRNA mean expression values as a cut-off and performed Kaplan-Meier analysis. We confirmed the significant association of miR-1290, miR-1246, and miR-375 with overall survival (p < 0.05) (Supplementary Fig. 1).

3.3. Candidate microRNAs for endogenous normalization

To identify potential miRNA references for exosomal RNA quantification, we analyzed the 173 most abundant known miRNAs from the 192 exosomal RNA sequencing libraries with CV <5%. Based on NormFinder analysis, miR-30e-5p, miR-221-3p, miR-30a-5p, miR-99a-5p, and miR-181b-5p ranked as the top five most stable miRNAs in ungrouped samples ( Fig. 2 ; Supplementary Table 4). Under BestKeeper, most miRNAs identified by NormFinder were also among the best candidate list (p < 0.001), although rank of their diagnostic powers was slightly different (Supplementary Table 5). Based on their expression levels, stability, and diagnostic power, we selected miR-30a/e-5p, miR-99a-5p, let-7c, and miR-125a-5p for further validation. In addition, we also selected miR-16-5p and RNU6B because they have been used for endogenous control in other studies[18], [19], and [20].

gr2

Fig. 2 Evaluations of microRNA (miRNA) candidates for endogenous normalization. (A) NormFinder-determined intergroup and intragroup stabilities of 173 miRNA candidates; (B) intergroup and intragroup variations of six selected candidates; (C) normalized read counts of six selected candidates in 192 individuals.

3.4. miR-30a-5p and miR-30e-5p as best endogenous normalizers

To determine the best endogenous RNA references, we examined candidate miRNAs using the TaqMan qRT-PCR assay in the follow-up cohort. RNU6B, a commonly used endogenous control for cellular miRNA quantification, was not detectable by the TaqMan assay. miR-125a-5p was barely detectable, consistent with its relative low read counts in the RNA sequencing data. let-7c, although detectable, significantly varied across technical triplicates, possibly caused by its Ct value >35. We thus excluded RNU6B, miR-125-5p, and let-7c from endogenous candidates. For the remaining four miRNAs, the geometric mean of miR-30a and miR-30e-5p (GM-30a/e) was the best, followed by miR-30e-5p, miR-30a-5p, miR-99a-5p, and miR-16-5p, according to NormFinder ( Table 2 ; Supplementary Fig. 2). Defined by BestKeeper, miR-30a, miR-30e-5p, and GM-30a/e remained significant (cut-offp = 0.001). This result was further confirmed separately by the TaqMan assay using a smaller RT primer pool (miR-30a-5p, miR-30e-5p, and miR-16-5p) and Qiagen universal RT primer assay, where a more stable expression pattern was observed in all selected candidates ( Table 2 ; Supplementary Fig. 3). Of note, miR-30a and miR-30e-5p demonstrated better stability than miR-16-5p in these tests. Based on these analyses, miR-30a-5p and miR-30e-5p together with GM-30a/e emerged as the best endogenous references for follow-up validation.

Table 2 Performance of endogenous control candidates in follow-up cohort

Gene name NormFinder BestKeeper
Stability SE r p Power, x-fold
TaqMan qRT-PCR using 10 RT primer pool
 GM-30a/e * 0.251 0.048 0.766 0.001 1.64
 miR-30e-5p 0.311 0.046 0.73 0.001 1.65
 miR-30a-5p 0.361 0.046 0.696 0.001 1.63
 miR-99a-5p 0.743 0.061 0.592 0.001 1.99
 let-7c 0.764 0.062 0.735 0.001 2.68
 miR-16-5p 1.129 0.084 0.636 0.001 2.86
 Best candidate: GM-30a/e
TaqMan qRT-PCR using 3 RT primer pool
 GM-30a/e * 0.052 0.05 0.965 0.001 1.81
 miR-30e-5p 0.054 0.049 0.967 0.001 1.86
 miR-30a-5p 0.212 0.027 0.929 0.001 1.81
 miR-16-5p 0.617 0.046 0.913 0.001 2.59
 Best candidate: GM-30a/e
Qiagen universal qRT-PCR assay
 GM-30a/e * 0.075 0.034 0.98 0.001 1.95
 miR-30a-5p 0.183 0.023 0.956 0.001 1.84
 miR-30e-5p 0.215 0.023 0.957 0.001 2.08
 miR-16-5p 0.431 0.032 0.915 0.001 2.1
 Best candidate: GM-30a/e

* Geometric mean of miR-30a-5p and miR-30e-5p.

qRT-PCR = quantitative reverse transcription polymerase chain reaction; SE = standard error.

3.5. Association of miR-1290 and miR-375 with overall survival

Based on pretest quality on each assay, we selected the TaqMan assay for miR-1246 and miR-1290, and the Qiagen miScript for miR-375. We first normalized each qRT-PCR value to GM-30a/e to generate ΔCTand then performed Kaplan-Meier analysis. We observed significant associations of elevated levels of miR-375 and miR-1290 with shorter overall survival (p = 0.0028 and 0.0057, respectively) ( Fig. 3 ). No significant association was observed with miR-1246 (p = 0.13). The combined effect of higher levels in both significant miRs on survival was more significant than the individual miRs (p = 9.44 × 10−5). The patients with high levels in both miR-1290 and miR-375 (n = 21) had a mortality rate of approximately 80% at the 20-mo follow-up; lower levels in both miRs (n = 40) had a mortality rate of about 10% over the same follow-up time period. The median overall survival for patients with higher level miRs (7.23 mo) was significantly shorter than patients with lower level miRs (19.3 mo) (p = 0.0045).

gr3

Fig. 3 MicroRNAs (miRNAs)-based survival analysis in follow-up cohort (time in months). (A, C) Kaplan-Meier curves show that relative expression levels of miR-375 and miR-1290 are associated with overall survival; (B) miR-1246 shows a trend association but not statistically significance; (D) combination of miR-1290 and miR-375 expressions demonstrated a strong synergistic effect.

To evaluate performance of multivariate prediction models, we first estimated the association of individual clinical factors (age, PSA, LDH, and hemoglobin levels at the time of progression to CRPC stage, Gleason score at disease diagnosis, number of treatments after progression to CRPC, time from ADT initiation for hormone-sensitive stage to treatment failure) with overall survival. Individually, we observed a survival association with PSA levels at the time of CRPC stage (p = 0.0015) and ADT failure time from ADT initiation (p = 0.041) ( Fig. 4 ; Supplementary Fig. 4). We then computed the risk score for the combination of the selected miRNAs and clinical prognostic factors (multivariate model). We found a significant association of patients’ survival with the multivariate model that included miR-1290 and miR-375 levels, ADT failure time, and PSA levels at the time of CRPC stage (p = 0.00016). Based on this model, patients with a high risk score had a 2.58-fold higher risk of death than patients with a low-risk score (HR: 2.58; 95% CI, 1.51–4.41). In addition, a time-dependent ROC analysis using the multivariate model generated an AUC of approximately 0.73, which represented significant improvement over a clinical variable-based model (ADT failure time and PSA at CRPC stage) with an AUC of about 0.66 (p = 6.57 × 10−6; two-tailed Studentttest) ( Fig. 4 ).

gr4

Fig. 4 Multivariate-based survival analysis in follow-up cohort. Kaplan-Meier curves show significant survival association in androgen-deprivation therapy (ADT) failure time defined as from (A) initiation of ADT for hormone-sensitive stage to development of castration-resistant prostate cancer; (B) prostate-specific antigen (PSA) level; (C) multivariate model. (D) Time-dependent area under the curve (AUC) analysis shows significant improvement of multivariate model over clinical factors-only model.

4. Discussion

After failure of ADT and with the emergence of castration resistance, the survival period is limited but extremely variable between patients. Knowledge of prognosis for individual patients in this stage has become critical because several therapeutic choices are available to control progression. Such prognostic factors are lacking in advanced PCa therapeutics, barring the presence of CTC counts, which are technically challenging and have not been widely adopted. Clinical prognostic factors used in several nomograms at the CRPC stage have yet to be validated in prospective studies [21] . In this study, we observed a significant association of exosomal miRNA transcripts with overall survival and inculcated clinical prognostic factors using a hospital-based real-world registry. Our results suggest that circulating miRNAs may serve as a sensitive biomarker for the prognosis of CRPC patients.

There are several advantages for developing blood-based prognostic biomarkers in advanced stage cancer patients. Repeated blood measurements are minimally invasive and can provide critical information on a disease's molecular signature status over time, through the course of treatment and recurrence. While CTC sampling requires reference laboratory work-up and technically challenging equipment for assay performance, a blood-based PCR assay offers an attractive option. Real-time PCR assay is able to detect specific RNAs indicative of disease status from a simple sampling without the need for a biopsy or surgically resected tissue. However, this unmet need will have to take into account several challenges as per the Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK) [22] . In this initial attempt we attempted to adhere with these recommendations in the context of identifying candidate plasma-based biomarkers for prognosis.

As a well-known oncomiR (miRNA associated with cancer), circulating miR-375 has been implicated in a variety of cancers including PCa[10], [11], [23], and [24]. Upregulation of miR-375 has been correlated with PCa metastasis [9] and poor survival in esophageal squamous cell carcinoma patients [25] . Compared with miR-375, exosomal miR-1290 performed slightly better in the outcome prediction. A recent study shows that high levels of miR-1290 in serum discriminate low- from high-stage pancreatic cancer [19] . miR-1290 is involved in estrogen receptor α–positive breast cancer, probably due to targeted inhibition of N-acetyltransferase 1 (NAT1) and forkhead box A1 (FOXA1) [26] ; however a direct role for miR-1290 in PCa remains undetermined.

5. Conclusions

This study identified two miRNA candidates (miR-1290 and miR-375) for survival prognosis in CRPC after a systematic attempt for reducing biases from specimen processing and real-time PCR. These results provide strong preliminary evidence to support the notion that circulating exosomal RNAs contain various RNA species and changes in exosomal RNA contents are robust candidates as clinical biomarkers for advanced PCa. Prospective validation studies should be pursued in accordance with REMARK guidelines by the integration of extracellular RNAs and clinical prognostic factors in the future development of a prognostic algorithm for late-stage PCa patients.


Author contributions:Liang Wang and Manish Kohli had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design:Wang, Huang, Kohli.

Acquisition of data:Huang.

Analysis and interpretation of data:Yuan, Sun, Lu, Wang.

Drafting of the manuscript:Huang, Wang, Liang, Thibodeau, Boardman, Kohli.

Critical revision of the manuscript for important intellectual content:Du, Xia, Dittmar, Wang, See, Costello, Quevedo, Tan, Nandy, Bevan, Longenbach.

Statistical analysis:Yuan, Sun, Lu, Wang.

Obtaining funding:Wang, Kohli.

Administrative, technical, or material support:Dittmar.

Supervision:Wang, Kohli.

Other(specify): None.

Financial disclosures:Liang Wang certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.

Funding/Support and role of the sponsor:This study was supported by Advancing a Healthier Wisconsin (Project 5520227) and the National Institutes of Health (R01CA157881) to Liang Wang and by the Center for Individualized Medicine of the Mayo Clinic and John P. Vaile Development funds for prostate cancer research to Manish Kohil.

Appendix A. Supplementary data

mmc1

Supplementary Fig. 1 – Kaplan-Meier plots for overall survival in screening cohort (n = 23). Relative expression levels of miR-375, miR-1246, and miR-1290 are associated with overall survival.

mmc2

Supplementary Fig. 2 – Validation of normalization candidates using TaqMan quantitative real-time polymerase chain reaction method. With a pool of 10 primers in follow-up cohort, (a) expression variations and (b) stabilities of the five selected candidates are presented. When primer number of RT primer pool is reduced, the detected levels of the four candidates are increased (c) and their stabilities are significantly increased (d). All data represented in the scatter plots are sorted based on Ct value of miR-16-5p.

mmc3

Supplementary Fig. 3 – Validation of the normalization candidates using QIAGEN real-time qRT-PCR method. (a) Expression patterns and (b) stabilities of selected candidates re determined by QIAGEN qRT-PCR assay. All data represented in the scatter plots are sorted based on CTvalue of miR-16-5p.

mmc4

Supplementary Fig. 4 – Kaplan-Meier analysis of individual clinical factors. No survival association was found in (a) initial Gleason score, (b) age at the time of castration-resistant prostate cancer (CRPC) stage, (c) hemoglobin level at CRPC stage, and (d) the number of treatments after progression to CRPC stage.

References

  • [1] R. Siegel, D. Naishadham, A. Jemal. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11-30 Crossref
  • [2] C. Palmberg, P. Koivisto, T. Visakorpi, T.L. Tammela. PSA decline is an independent prognostic marker in hormonally treated prostate cancer. Eur Urol. 1999;36:191-196 Crossref
  • [3] E.D. Crawford, M.A. Eisenberger, D.G. McLeod, et al. A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med. 1989;321:419-424 Crossref
  • [4] D.C. Danila, G. Heller, G.A. Gignac, et al. Circulating tumor cell number and prognosis in progressive castration-resistant prostate cancer. Clin Cancer Res. 2007;13:7053-7058 Crossref
  • [5] J.S. de Bono, H.I. Scher, R.B. Montgomery, et al. Circulating tumor cells predict survival benefit from treatment in metastatic castration-resistant prostate cancer. Clin Cancer Res. 2008;14:6302-6309 Crossref
  • [6] X. Huang, M. Liang, R. Dittmar, L. Wang. Extracellular microRNAs in urologic malignancies: chances and challenges. Int J Mol Sci. 2013;14:14785-14799 Crossref
  • [7] G. van Niel, I. Porto-Carreiro, S. Simoes, G. Raposo. Exosomes: a common pathway for a specialized function. J Biochem. 2006;140:13-21 Crossref
  • [8] J. Skog, T. Wurdinger, S. van Rijn, et al. Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers. Nat Cell Biol. 2008;10:1470-1476 Crossref
  • [9] R.J. Bryant, T. Pawlowski, J.W. Catto, et al. Changes in circulating microRNA levels associated with prostate cancer. Br J Cancer. 2012;106:768-774 Crossref
  • [10] J.C. Brase, M. Johannes, T. Schlomm, et al. Circulating miRNAs are correlated with tumor progression in prostate cancer. Int J Cancer. 2011;128:608-616 Crossref
  • [11] H.C. Nguyen, W. Xie, M. Yang, et al. Expression differences of circulating microRNAs in metastatic castration resistant prostate cancer and low-risk, localized prostate cancer. Prostate. 2013;73:346-354 Crossref
  • [12] X. Huang, T. Yuan, M. Tschannen, et al. Characterization of human plasma-derived exosomal RNAs by deep sequencing. BMC Genomics. 2013;14:319
  • [13] T. Yuan, X. Huang, R.L. Dittmar, et al. eRNA: a graphic user interface-based tool optimized for large data analysis from high-throughput RNA sequencing. BMC Genomics. 2014;15:176 Crossref
  • [14] J.M. Fitzpatrick, J. Bellmunt, K. Fizazi, et al. Optimal management of metastatic castration-resistant prostate cancer: highlights from a European Expert Consensus Panel. Eur J Cancer. 2014;50:1617-1627 Crossref
  • [15] M.R. Friedlander, S.D. Mackowiak, N. Li, W. Chen, N. Rajewsky. miRDeep2 accurately identifies known and hundreds of novel microRNA genes in seven animal clades. Nucleic Acids Res. 2012;40:37-52 Crossref
  • [16] C.L. Andersen, J.L. Jensen, T.F. Orntoft. Normalization of real-time quantitative reverse transcription-PCR data: a model-based variance estimation approach to identify genes suited for normalization, applied to bladder and colon cancer data sets. Cancer Res. 2004;64:5245-5250 Crossref
  • [17] M.W. Pfaffl, A. Tichopad, C. Prgomet, T.P. Neuvians. Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper—Excel-based tool using pair-wise correlations. Biotechnol Lett. 2004;26:509-515 Crossref
  • [18] M. Redova, A. Poprach, J. Nekvindova, et al. Circulating miR-378 and miR-451 in serum are potential biomarkers for renal cell carcinoma. J Transl Med. 2012;10:55 Crossref
  • [19] A. Li, J. Yu, H. Kim, et al. MicroRNA array analysis finds elevated serum miR-1290 accurately distinguishes patients with low-stage pancreatic cancer from healthy and disease controls. Clin Cancer Res. 2013;19:3600-3610 Crossref
  • [20] J. Song, Z. Bai, W. Han, et al. Identification of suitable reference genes for qPCR analysis of serum microRNA in gastric cancer patients. Dig Dis Sci. 2012;57:897-904 Crossref
  • [21] O. Smaletz, H.I. Scher, E.J. Small, et al. Nomogram for overall survival of patients with progressive metastatic prostate cancer after castration. J Clin Oncol. 2002;20:3972-3982 Crossref
  • [22] L.M. McShane, D.G. Altman, W. Sauerbrei, S.E. Taube, M. Gion, G.M. Clark. Reporting recommendations for tumor marker prognostic studies (REMARK). J Natl Cancer Inst. 2005;97:1180-1184 Crossref
  • [23] L.A. Selth, S. Townley, J.L. Gillis, et al. Discovery of circulating microRNAs associated with human prostate cancer using a mouse model of disease. Int J Cancer. 2012;131:652-661 Crossref
  • [24] H.H. Cheng, P.S. Mitchell, E.M. Kroh, et al. Circulating microRNA profiling identifies a subset of metastatic prostate cancer patients with evidence of cancer-associated hypoxia. PloS One. 2013;8:e69239 Crossref
  • [25] S. Komatsu, D. Ichikawa, H. Takeshita, et al. Prognostic impact of circulating miR-21 and miR-375 in plasma of patients with esophageal squamous cell carcinoma. Expert Opin Biol Ther. 2012;12(Suppl 1):S53-S59 Crossref
  • [26] Y. Endo, T. Toyama, S. Takahashi, et al. miR-1290 and its potential targets are associated with characteristics of estrogen receptor alpha-positive breast cancer. Endocr Relat Cancer. 2013;20:91-102 Crossref

Footnotes

a Department of Pathology and Cancer Center, Medical College of Wisconsin, Milwaukee, WI, USA

b Biotherapy Center, Third Affiliated Hospital of Harbin Medical University, Harbin, China

c Department of Endocrinology, Second Affiliated Hospital of Harbin Medical University, Harbin, China

d Department of Oncology, Tongji Hospital of Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China

e Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, USA

f Department of Urology, Medical College of Wisconsin, Milwaukee, WI, USA

g Department of Oncology, Mayo Clinic, Rochester, MN, USA

h Department of Oncology, Mayo Clinic, Jacksonville, FL, USA

i University of Rochester Medical Center, Rochester, NY, USA

j Division of Biomedical Statistics and Informatics, Mayo Clinic, Rochester, MN, USA

k Department of Physiology, Medical College of Wisconsin, Milwaukee, WI, USA

l Division of Biostatistics, Medical College of Wisconsin, Milwaukee, WI, USA

m Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA

lowast Corresponding authors. Liang Wang: Department of Pathology and MCW Cancer Center, Medical College of Wisconsin, 8701 W. Watertown Plank Road, TRBC #C4678, Milwaukee, WI 53226, USA. Tel. +1 414 955 2574; Manish Kohli: Division of Medical Oncology, Department of Oncology, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905, USA. Tel. +1 507 284 3903.

Place a comment

Your comment *

max length: 5000