Development and Validation of MyProstateScore 2.0 to Detect Clinically Significant Prostate Cancer
ABSTRACT Background The benefits of prostate cancer screening with serum prostate-specific antigen (PSA) have been largely offset by the high rate of negative prostate biopsies and overdiagnosis of indolent cancers. These outcomes result from the limited diagnostic accuracy of PSA for clinically significant prostate cancer and are considered a major obstacle to realizing the benefits of population-wide screening for prostate cancer.Methods We analyzed RNAseq data from a prostate cancer compendium to identify novel transcripts associated with cancer and high-grade cancer. Predefined nomination criteria were applied to 58,724 gene targets, yielding 54 differentially expressed transcripts. We designed a custom multiplex qPCR panel for non-invasive detection of candidate transcripts in urine. The panel was applied to a development cohort of men with elevated PSA (3 to 10 ng per milliliter) that underwent prospective, standardized urine collection and prostate biopsy at the University of Michigan. Elastic net modeling was used to derive the optimal model for clinically significant (grade group 2 or higher) prostate cancer, the 18-transcript MyProstateScore 2.0 (MPS2) test. The calibrated, locked MPS2 model was assessed in a blinded, external National Cancer Institute (NCI) – Early Detection Research Network (EDRN) validation cohort and compared to serum PSA, the Prostate Cancer Prevention Trial risk calculator (PCPTrc), and the MyProstateScore (MPS) test. The original MPS assay measures urinary expression of two cancer-associated markers (PCA3, TMPRSS2:ERG) and is endorsed by National Comprehensive Cancer Network guidelines for consideration prior to biopsy in the study population.Results We performed multiplex urinary testing of 1,623 clinical specimens in total, representing the largest such cohort to our knowledge. The prospective NCI-EDRN validation population included 743 men undergoing per-protocol urine collection and prostate biopsy. The median age was 62 years, median PSA was 5.6 ng per milliliter, and 151 men (20%) had clinically significant prostate cancer on biopsy. The area under the receiver-operating characteristic curve (AUC) for clinically significant prostate cancer was 0.597 (95% Confidence Interval [CI], 0.547 to 0.646) for PSA, 0.659 (95% CI, 0.611 to 0.707) for the PCPTrc, and 0.737 (95% CI, 0.694 to 0.780) for MPS, as compared to 0.818 (95% CI, 0.781 to 0.855) for the optimal MPS2 model (MPS2+). Under a clinically applicable testing approach providing 95% sensitivity for clinically significant cancer, the specificity (equivalent to the percentage of unnecessary biopsies avoided after pre-biopsy testing) was 11% for PSA, 20% for the PCPTrc, and 23% for MPS, as compared to 41% for MPS2+. In all sub-populations, MPS2 testing provided negative predictive value (NPV) of 95% to 99% for clinically significant cancer.Conclusions In a large, external validation population referred for prostate biopsy, the novel MPS2 assay provided exceptional sensitivity and NPV to rule out clinically significant prostate cancer. These data support the use of MPS2 as a highly accurate secondary test to reduce the harms associated with PSA screening and preserve its long-term benefits..
| Medienart: |
|
|---|
Preprint| Erscheinungsjahr: |
2023 |
|---|---|
| Erschienen: |
2023 |
| Enthalten in: |
bioRxiv.org - (2023) vom: 21. Apr. Zur Gesamtaufnahme - year:2023 |
|---|
| Sprache: |
Englisch |
|---|
| Beteiligte Personen: |
Tosoian, Jeffrey J. [VerfasserIn] |
|---|
| Links: |
Volltext [kostenfrei] |
|---|
| Themen: |
|---|
| doi: |
10.1101/2023.04.11.23288418 |
|---|
| funding: |
|
|---|---|
| Förderinstitution / Projekttitel: |
|
| PPN (Katalog-ID): |
XBI039295257 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | XBI039295257 | ||
| 003 | DE-627 | ||
| 005 | 20230422091331.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 230419s2023 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.1101/2023.04.11.23288418 |2 doi | |
| 035 | |a (DE-627)XBI039295257 | ||
| 035 | |a (biorXiv)10.1101/2023.04.11.23288418 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 100 | 1 | |a Tosoian, Jeffrey J. |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Development and Validation of MyProstateScore 2.0 to Detect Clinically Significant Prostate Cancer |
| 264 | 1 | |c 2023 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 520 | |a ABSTRACT Background The benefits of prostate cancer screening with serum prostate-specific antigen (PSA) have been largely offset by the high rate of negative prostate biopsies and overdiagnosis of indolent cancers. These outcomes result from the limited diagnostic accuracy of PSA for clinically significant prostate cancer and are considered a major obstacle to realizing the benefits of population-wide screening for prostate cancer.Methods We analyzed RNAseq data from a prostate cancer compendium to identify novel transcripts associated with cancer and high-grade cancer. Predefined nomination criteria were applied to 58,724 gene targets, yielding 54 differentially expressed transcripts. We designed a custom multiplex qPCR panel for non-invasive detection of candidate transcripts in urine. The panel was applied to a development cohort of men with elevated PSA (3 to 10 ng per milliliter) that underwent prospective, standardized urine collection and prostate biopsy at the University of Michigan. Elastic net modeling was used to derive the optimal model for clinically significant (grade group 2 or higher) prostate cancer, the 18-transcript MyProstateScore 2.0 (MPS2) test. The calibrated, locked MPS2 model was assessed in a blinded, external National Cancer Institute (NCI) – Early Detection Research Network (EDRN) validation cohort and compared to serum PSA, the Prostate Cancer Prevention Trial risk calculator (PCPTrc), and the MyProstateScore (MPS) test. The original MPS assay measures urinary expression of two cancer-associated markers (PCA3, TMPRSS2:ERG) and is endorsed by National Comprehensive Cancer Network guidelines for consideration prior to biopsy in the study population.Results We performed multiplex urinary testing of 1,623 clinical specimens in total, representing the largest such cohort to our knowledge. The prospective NCI-EDRN validation population included 743 men undergoing per-protocol urine collection and prostate biopsy. The median age was 62 years, median PSA was 5.6 ng per milliliter, and 151 men (20%) had clinically significant prostate cancer on biopsy. The area under the receiver-operating characteristic curve (AUC) for clinically significant prostate cancer was 0.597 (95% Confidence Interval [CI], 0.547 to 0.646) for PSA, 0.659 (95% CI, 0.611 to 0.707) for the PCPTrc, and 0.737 (95% CI, 0.694 to 0.780) for MPS, as compared to 0.818 (95% CI, 0.781 to 0.855) for the optimal MPS2 model (MPS2+). Under a clinically applicable testing approach providing 95% sensitivity for clinically significant cancer, the specificity (equivalent to the percentage of unnecessary biopsies avoided after pre-biopsy testing) was 11% for PSA, 20% for the PCPTrc, and 23% for MPS, as compared to 41% for MPS2+. In all sub-populations, MPS2 testing provided negative predictive value (NPV) of 95% to 99% for clinically significant cancer.Conclusions In a large, external validation population referred for prostate biopsy, the novel MPS2 assay provided exceptional sensitivity and NPV to rule out clinically significant prostate cancer. These data support the use of MPS2 as a highly accurate secondary test to reduce the harms associated with PSA screening and preserve its long-term benefits. | ||
| 650 | 4 | |a Biology |7 (dpeaa)DE-84 | |
| 650 | 4 | |a 570 |7 (dpeaa)DE-84 | |
| 700 | 1 | |a Zhang, Yuping |4 aut | |
| 700 | 1 | |a Xiao, Lanbo |4 aut | |
| 700 | 1 | |a Xie, Cassie |4 aut | |
| 700 | 1 | |a Samora, Nathan L. |4 aut | |
| 700 | 1 | |a Niknafs, Yashar S. |4 aut | |
| 700 | 1 | |a Chopra, Zoey |4 aut | |
| 700 | 1 | |a Siddiqui, Javed |4 aut | |
| 700 | 1 | |a Zheng, Heng |4 aut | |
| 700 | 1 | |a Herron, Grace |4 aut | |
| 700 | 1 | |a Vaishampayan, Neil |4 aut | |
| 700 | 1 | |a Arivoli, Kumaran |4 aut | |
| 700 | 1 | |a Trock, Bruce J. |4 aut | |
| 700 | 1 | |a Ross, Ashley E. |4 aut | |
| 700 | 1 | |a Morgan, Todd M. |4 aut | |
| 700 | 1 | |a Palapattu, Ganesh S. |4 aut | |
| 700 | 1 | |a Salami, Simpa S. |4 aut | |
| 700 | 1 | |a Kunju, Lakshmi P. |4 aut | |
| 700 | 1 | |a Zheng, Yingye |4 aut | |
| 700 | 1 | |a Wei, John T. |4 aut | |
| 700 | 1 | |a Chinnaiyan, Arul M. |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t bioRxiv.org |g (2023) vom: 21. Apr. |
| 773 | 1 | 8 | |g year:2023 |g day:21 |g month:04 |
| 856 | 4 | 0 | |u http://dx.doi.org/10.1101/2023.04.11.23288418 |z kostenfrei |3 Volltext |
| 912 | |a GBV_XBI | ||
| 951 | |a AR | ||
| 952 | |j 2023 |b 21 |c 04 | ||