Prediction of disease specific and overall survival in men with prostate cancer using the Decipher assay

Objectives

This is a protocol for a Cochrane Review (prognosis). The objectives are as follows:

To determine the prognostic value of the Decipher gene assay in men with biopsy‐confirmed clinically localized prostate cancer before and after receiving treatment.

This objective can be framed in two questions following the population, index prognostic factor, comparator prognostic factor(s), outcome, timing, and setting (PICOTS) format as illustrated in Table 1.

Table 1. Review question in PICOTS format

Objective	Before treatment	After treatment
Population	Men with newly diagnosed, biopsy‐confirmed clinically localized prostate cancer	Men who have undergone radical prostatectomy for clinically localized prostate cancer
Index prognostic factor	Decipher gene assay
Comparator	Existing clinical prediction models
Outcome	Overall survival and prostate‐cancer survival
Timing	The outcome is to be predicted at any point after diagnosis
Setting	Before primary, local treatment with curative intent	Following primary surgical treatment with curative intent

Background

Description of the health condition and context

Prostate cancer is the second most commonly diagnosed cancer among men worldwide, especially in countries with a higher proportion of elderly men where this condition is more frequent (IACR France). In the USA, prostate cancer accounts annually for more than 260,000 new cases, approximately 27% of all new cancer cases, and the lifetime risk of prostate cancer is estimated to be approximately one in six (Siegel 2022). Similar numbers regarding the incidence of prostate cancer are found in Europe (Cancer Research UK; Netherlands Cancer Registry). The most prevailing risk factors for prostate cancer include family history, age, and ethnicity (Albright 2015; Hemminki 2012; Jansson 2012; Kamangar 2006). Modifiable risk factors may impact the risk of developing prostate cancer, but little evidence exists supporting any clear indication for primary prevention (Gandaglia 2021). Although the incidence of prostate cancer is high, the mortality of prostate cancer is relatively low (IACR France). Approximately 11% of all deaths due to cancer in men in the USA result from prostate cancer (Siegel 2022). An increase in survival rates has been found in recent years, attributable in large part to the widespread use of prostate‐specific antigen (PSA) testing, resulting in a more favorable stage distribution (De Angelis 2014; Etzioni 2013; Hayes 2014).

Population screening, defined as the systematic examination of men at risk of prostate cancer in the general population, is mainly based on the evaluation of the patient's PSA level. However, population screening for prostate cancer remains a controversial topic, given concerns about overdiagnosis and overtreatment (Ilic 2013; Ilic 2017). In this scenario, shared decision‐making is considered crucial in the decision to undergo screening (Carter 2013; Force 2018; Moyer 2012; Tikkinen 2017). Despite this controversy, PSA screening remains widespread and has resulted in the diagnosis of increased numbers of prostate cancer at earlier stages (Tikkinen 2018). Today, most men diagnosed with prostate cancer have non‐palpable disease detected due to an elevated PSA level. Many of these tumors are considered low or very low risk for progression with very low metastatic potential. It is unclear whether the increased use of multiparametric imaging and novel biomarkers will increase or reduce the issue of overdiagnosis and overtreatment (Carter 2004; Narayan 2017; Wilt 2017).

The diagnosis of prostate cancer is commonly established through tissue diagnosis of a prostate biopsy specimen. Pathological information obtained from the biopsy specimen, in particular on tumor grade (Gleason score) and volume, as well as the findings from the PSA levels and digital rectal examination (DRE) are critical to the determination of the likely prognosis to guide treatment decisions. When indicated, results of imaging tests (computed tomography (CT) scan, magnetic resonance imaging (MRI), or bone scan) can provide further information. Based on this prognostic assessment, men with prostate cancer are categorized into risk groups indicating the severity of the disease (Hanna 2019; Zelic 2020).

Description of the prognostic factor

Decipher is a commercially available gene‐expression panel (Decipher test by GenomeDx Biosciences, San Diego, CA, USA) that offers the promise of improved prediction of outcomes and, therefore, treatment selection choice tailored to a given patient’s risk. It is used as a prognostic tool for prostate cancer in two clinical settings: (i) at the time of biopsy‐confirmed diagnosis and ii) in the postprostatectomy setting.

Decipher assay evaluates messenger ribonucleic acid (mRNA) expression levels of 22 genes from both formalin‐fixed paraffin‐embedded biopsy tissue (i.e. Decipher biopsy) and prostate tissue after radical prostatectomy (i.e. Decipher postoperative). It provides a score ranging from 0 to 1, which categorizes patients into low‐, intermediate‐, and high‐risk categories based on the predicted risk of metastasis and prostate cancer‐specific mortality (Eggener 2020; Jairath 2021).

The Decipher test is one of several prostate cancer biomarkers of uncertain prognostic value that have become commercially available, mainly in the United States, are being aggressively marketed, and are being considered for integration into the standard of care (Narayan 2017). Some society guidelines acknowledge the selective use of tissue‐based genomic biomarkers when added risk stratification may alter clinical decision‐making. However, they do not endorse the routine use of tissue‐based genomic biomarkers for risk stratification or clinical decision‐making (Eastham 2022a; Eggener 2020). Additionally, genomic classifiers are not included in many current guidelines (NICE 2019; Parker 2020).

Health outcomes

The risk of disease progression and adverse oncologic outcomes of prostate cancer varies widely based on clinicopathologic characteristics such as tumor extension (T stage), histological grade, total PSA level, tumor volume, capsular invasion, and surgical margins (Buhmeida 2006). After the initial diagnosis of localized prostate cancer, risk stratification is vital to appropriately balance treatment‐specific risk and benefits, and consequently select the intensity of local treatment (Eastham 2022a; Eastham 2022b; Eastham 2022c). Main treatment therapeutic options in men assumed to have clinically localized prostate cancer include immediate treatment with curative intent, in the form of radical surgery or radiation therapy, versus deferred treatment, in the form of active surveillance (Knipper 2021).

Numerous methods have been described to determine the aggressiveness of a patient’s tumor, including grouping systems, nomograms, and scoring systems (D'Amico 1998; Eastham 2022a; Eastham 2022b; Eastham 2022c; Litwin 2017). One of the most extensively used systems is the National Comprehensive Cancer Network (NCCN) risk group classification (Spratt 2018). Its simplified 3‐tier grouping (low‐, intermediate‐, and high‐risk) is based on pretreatment PSA, Gleason score, and clinical T stage. Further subdivision of the low and high groups based on the percentage of positive core biopsies has been included in the updated 5‐tier classification. NCCN risk groups are routinely used in clinical trials reporting and provide a framework for guiding clinical decisions (Spratt 2018).

Men with prostate cancer who have elected to undergo radical surgery face the choice of undergoing additional treatment in the form of adjuvant or salvage radiation versus observation alone. This decision has traditionally been informed through risk stratification on the basis of pathological information obtained from the prostatectomy specimen (such as tumor grade, volume, presence of extracapsular extension, and positive margins), preoperative information (such as PSA level and DRE findings), as well as postprostatectomy PSA levels (Blute 2001; Mitchell 2015).

Why it is important to do this review

Given the heterogenous natural history of prostate cancer, prediction of risk for recurrence, metastatic spread, and disease‐specific survival is critical for guiding management decisions, and Decipher has the potential to make an important contribution (Spratt 2018). To date, several largely narrative reviews have painted a mostly positive picture of the value of this assay, fueling a rapid uptake in the United States (Lokeshwar 2022). A recent systematic review reported superior performance of Decipher when compared to standard risk prognostication (Jairath 2021). However, this review was not conducted according to the up‐to‐date methodological standards for prognostic reviews. Serious methodological limitations include lack of a formal risk of bias assessment and the absence of information regarding covariates used for adjusting in multivariate models of the included, largely retrospective studies. Therefore, there is an important need for a high‐quality systematic review performed in accordance with up‐to‐date Cochrane methods for prognostic reviews to better guide clinical decision‐making at the point of care, clinical practice guidelines, and health policy decision‐making.

Objectives

To determine the prognostic value of the Decipher gene assay in men with biopsy‐confirmed clinically localized prostate cancer before and after receiving treatment.

This objective can be framed in two questions following the population, index prognostic factor, comparator prognostic factor(s), outcome, timing, and setting (PICOTS) format as illustrated in Table 1.

Table 1. Review question in PICOTS format

Objective	Before treatment	After treatment
Population	Men with newly diagnosed, biopsy‐confirmed clinically localized prostate cancer	Men who have undergone radical prostatectomy for clinically localized prostate cancer
Index prognostic factor	Decipher gene assay
Comparator	Existing clinical prediction models
Outcome	Overall survival and prostate‐cancer survival
Timing	The outcome is to be predicted at any point after diagnosis
Setting	Before primary, local treatment with curative intent	Following primary surgical treatment with curative intent

Methods

Criteria for considering studies for this review

Types of studies

We will include longitudinal studies (prospective and retrospective). We will also include observational studies embedded in randomized controlled trials.

Targeted population

For our first objective, we will include men with a diagnosis of clinically organ‐confirmed (T1‐2, N0, M0) biopsy‐confirmed prostate cancer with a life expectancy of at least 10 years considering immediate (radical prostatectomy, radiation therapy) or deferred (active surveillance) primary treatment.

For our second objective, we will include men who underwent radical prostatectomy for clinically organ‐confirmed prostate cancer with a pathological diagnosis of T1‐3 without clinical evidence of regional or distant spread (N0, M0).

The preceding nomenclature is based on the current classification of prostate cancer (Brierley  2017).

• T1 Clinically inapparent tumor that is not palpable which can be an incidental histological finding in ≤ 5% (T1a) or > 5% (T1b) of the tissue resected or by needle biopsy (T1c).

• T2 Tumor that is palpable and confined within the prostate, involving one half of one lobe or less (T2a) more than half of one lobe (T2b) or both lobes (T2c).

• T3 Tumor extends through the prostatic capsule including microscopic bladder neck involvement (T3a) or seminal vesicle(s) (T3b).

• T4 Tumor is fixed or invades adjacent structures other than seminal vesicles.

• N0: no regional lymph node metastasis.

• M0: no distant metastasis.

Type of prognostic factor

We will include studies assessing the gene expression classifier test Decipher (Decipher Biosciences, San Diego, CA, USA). Decipher is a genomic classifier that uses a random forest algorithm based on the expression of 22 RNA biomarkers, applying a whole‐transcriptome microarray analysis related to androgen receptor signaling, cell proliferation, differentiation, motility, and immune modulation. Decipher has the broadest Centers for Medicare and Medicaid Services (CMS)‐approved indications for use in prostate cancer (Erho 2013).

Types of outcomes to be predicted

We will not exclude studies from the review solely because no outcomes of interest are reported. In cases where none of our outcomes of interest are reported in the included studies, we will report information about these studies in an additional table.

Primary outcomes

• Time to death from prostate cancer (time‐to‐event outcome)

• Time to death from any cause (time‐to‐event outcome)

Secondary outcomes

• Time to disease progression (time‐to‐event outcome) as determined by physical exam findings (e.g. increase in clinical tumor stage), PSA increase, or imaging studies (e.g. development of new metastases)

• Time to metastatic disease (time‐to‐event outcome) as established by imaging findings

If we are unable to retrieve the necessary information to analyze time‐to‐event outcomes, we plan to assess the incidence for dichotomized outcomes at five‐year intervals.

Main outcomes for summary of findings tables

We will present summary of findings tables reporting the following outcomes, listed according to priority. We will present one table for each objective (before treatment and after treatment).

• Time to death from prostate cancer

• Time to death from any cause

• Time to disease progression

• Time to metastatic disease

Search methods for identification of studies

We will conduct a comprehensive search with no restrictions on the language of publication or publication status. See Appendix 1 for the full search strategy for each database.

Electronic searches

We will search the following databases from inception.

• The Cochrane Library (Wiley)

• MEDLINE (Ovid)

• Embase (OVIDSP)

• Web of Science Core Collection (Clarivate)

• Latin American Caribbean Health Sciences Literature (LILACS) (Virtual Health Library)

• Scopus (Elsevier)

• Greynet (Grey Literature Network Service)

• OpenGrey (System for Information on Grey Literature in Europe)

• Database of Abstracts of Reviews of Effects (DARE) (Centre for Reviews and Dissemination (CRD) University of York)

• Health Technology Assessment (HTA) database (CRD University of York)

• INAHRA HTA database (The International Network of Agencies for Health Technology Assessment)

We will search the following trial registries.

• World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) search portal (apps.who.int/trialsearch/)

• ClinicalTrials.gov (www.clinicaltrials.gov/)

We will also search electronically available abstracts of the following national and international urology cancer meetings via Embase (OVIDSP) and Web of Science (Clarivate).

• Annual Meeting of the European Association of Urology (EAU)

• Annual Meeting of the American Society of Clinical Oncology (ASCO)

• American Urological Association Annual Meeting (AUA)

We will also search for health technology assessment‐related documents on the following organizations' web pages.

• CADTH ‐ Canadian Agency for Drugs and Technologies in Health (www.cadth.ca/node/8)

• NIHR HTA programme.(UK) (www.nihr.ac.uk/explore-nihr/funding-programmes/health-technology-assessment.htm)

• European Network for Health Technology Assessment (EUnetHTA) (www.eunethta.eu/)

• Australian Government Dept Health Medical Services Committee (MASC) (msac.gov.au/)

• USA Technology Assessment Program (www.ahrq.gov/research/findings/ta/index.html#tacomplete)

We will apply a MEDLINE email alert service to identify newly published trials using the search strategy as described for MEDLINE (Appendix 1). Should we identify new studies for inclusion, we will evaluate these, incorporate these studies into our review, and update this Cochrane Review.

Searching other resources

We will screen the reference lists of other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, reviews, and meta‐analyses. Additionally, we will contact the study authors of included trials to identify any studies that we may have missed.

Data collection and analysis

Selection of studies

We will use Covidence software to identify and remove potential duplicate records. Two review authors (JF, LG) will independently scan abstracts and titles to determine which studies should be assessed further. Two review authors will categorize all potentially relevant records as full‐text or map records to studies, and classify studies as included studies, excluded studies, studies awaiting classification, or ongoing studies, following the criteria in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021). We will resolve any disagreements between the two review authors through consensus or by recourse to a third review author (PD). If a resolution is not possible, we will designate the corresponding study as 'Awaiting Classification'. We will contact the study authors if we need clarification to determine the health status or diagnostic criteria of included patients. If we receive no response, clinical experts in our review group will classify the study, or we will list studies as 'Awaiting Classification'. We will document the reasons for the exclusion of studies in the 'Characteristics of excluded studies' table. We will present a PRISMA flow diagram showing the process of study selection (Page 2021).

Data extraction and management

We will develop a dedicated data extraction form based on the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS) checklist adapted for prognostic factor studies (Moons 2014; Riley 2019). We will pilot‐test the data extraction form ahead of time. For studies that fulfilled our inclusion criteria, two review authors (JF and LG) will independently abstract the following information, which we will present in the 'Characteristics of included studies' table.

• Study design (e.g. cohort, case‐control, randomized trial participants, or registry data)

• Study dates, settings, and country

• Participants: eligibility and recruitment method, and characteristics (e.g. age, baseline staging, baseline PSA at diagnosis, comorbidities, family history of prostate cancer)

  • For participants who underwent radical prostatectomy, we will also collect data on surgical decisions, pathology and postsurgical staging, postsurgical PSA, and hormonal treatment.

• Sample size: calculation, number of participants, outcomes, and events

• Prognostic factors: index and comparator prognostic factors: number and type of prognostic factors, definition and method for measurement of prognostic factors, the timing of prognostic factor measurement, handling of prognostic factors in the statistical modeling

• Outcome(s) to be predicted: definition and method for measurement of outcome(s), type of outcome(s), time of outcome(s), occurrence, or summary of duration of follow‐up

• Missing data: number of participants with any missing value, details of attrition (loss to follow‐up); for time‐to‐event outcomes: number of censored observations, and handling of missing data

• Analysis: modeling method, assumptions, unadjusted and adjusted prognostic effect estimates, adjustment factors used

• Results: interpretation of presented results, comparison with other studies, discussion of generalizability, strengths, and limitations

• Study funding sources

• Declarations of interest by primary investigators

We will resolve any disagreements by discussion, or if required by consultation with a third review author (PD).

We will provide information, including study identifiers, about potentially relevant ongoing studies in the 'Characteristics of ongoing studies' table.

We will contact the authors of included studies to obtain key missing data as needed.

Dealing with duplicate and companion publications

In the event of duplicate publications, companion documents or multiple reports relating to a primary study, we will maximize the yield of information by mapping all publications to unique studies and collating all available data. We will use the most complete data set aggregated across all known publications. In case of doubt, we will prioritize publications reporting the longest follow‐ups associated with our primary or secondary outcomes.

Assessment of risk of bias in included studies

We will use the Quality In Prognosis Studies (QUIPS) tool to assess the risk of bias of the included studies (Hayden 2013), including the following domains.

• Study participation

• Study attrition

• Prognostic factor measurement

• Outcome measurement

• Covariate adjustment

• Statistical analysis and reporting

We will judge each domain at low, moderate, high, or uncertain risk of bias according to the criteria set out in the tool, considering specificities of rating for observational studies. We will add an unclear risk option for judging the domains, because of the expected inconsistent reporting of the prognostic studies.

Measures of association to be extracted

We will determine for all of our outcomes (time to death from any cause, time to death from prostate cancer, time to disease progression, time to metastatic disease) the odds ratio or hazard ratio with 95% confidence intervals (CI), and P values. We will focus on association measures adjusted for other prognostic factors, but will also extract and present unadjusted estimates.

Dealing with missing data

We will contact the authors of the included studies for additional or updated information that is not available from the published articles. If measures of the association have not been reported, or details on precision are missing, we will estimate these from other information, including sample size, number of events, log‐rank P values and CIs, and Kaplan‐Meier curves (Altman 2003; Parmar 1998; Tierney 2007).

Assessment of heterogeneity

Heterogeneity between studies will be quantified using Tau². We will also assess between‐study heterogeneity using the I² statistic following the guidance of the Handbook (Higgins 2021), considering:

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity;

• 50% to 90%: may represent substantial heterogeneity;

• 75% to 100%: considerable heterogeneity.

Assessment of reporting bias

It is likely that reporting bias is an important problem in prognostic factor research; research in the cancer domain is known to include false‐positive studies, which would not have been published if their results were negative (Kyzas 2005; Kyzas 2007; Sauerbrei 2005). In addition, the prospective registration of prognostic factor studies is not common, making it impossible to check whether there were reporting deficiencies for all included studies (Peat 2014).

We will explore the presence of small‐study effects by creating a funnel plot with the measure of association and its standard error. We will perform a test for funnel plot asymmetry using the R package 'metamisc' (Debray 2019). We will interpret the results with caution, as these tests often have limited power to detect asymmetry, and many tests yield inadequate type‐I error rates (Debray 2018).

Data synthesis

Data synthesis and meta‐analysis approaches

If possible, we will perform a meta‐analysis for the adjusted hazard ratios for time to death from prostate cancer, time to death from any cause, time to disease progression, and time to metastatic disease. We will perform a meta‐analysis for each objective separately (before treatment and after treatment) and for each outcome separately. We will only conduct meta‐analysis when studies are sufficiently similar, and other covariates in the models are similar. We will focus our meta‐analysis on studies that adjust the analysis for at least age, and other prognostic factors, such as tumor extension (T stage), histological grade, total PSA level, tumoral volume, surgical margin, and comorbidities like chronic cardiovascular disease, chronic respiratory disease, and immunosuppression (Buhmeida 2006).

For meta‐analysis, we will use a random‐effects approach, using the restricted maximum likelihood estimation in the R package 'metafor'. In addition to calculating a confidence interval around the pooled estimate, we will calculate a 95% (approximate) prediction interval (Riley 2019).

Subgroup analysis and investigation of heterogeneity

If there is considerable heterogeneity, as judged by the prediction interval, we will consider performing a meta‐regression analysis to explore possible causes of this heterogeneity. We will consider the following parameters in this analysis: study design and population characteristics. If this is not possible, we will perform a narrative discussion of sources of clinical heterogeneity.

We will perform subgroup analysis considering these factors.

• Baseline staging (before treatment)/postsurgical staging (after treatment)

• PSA at diagnosis

• Family history of prostate cancer

Sensitivity analysis

For a sensitivity analysis, we will exclude studies with a 'high' risk of bias from the meta‐analysis. We will also conduct a sensitivity analysis for studies that reported complete information on hazard ratio and precision versus those for which we had to calculate this, based on other reported information as described in the section on 'dealing with missing data'.

Summary of findings table

We will present the prognostic value of the Decipher gene assay by expressing absolute risks by each defined outcome (time to death from prostate cancer, time to death from any cause, time to disease progression, and time to metastatic disease) in summary of findings tables. We will use the GRADE approach for prognostic reviews to assess the certainty of evidence of the listed outcomes (Foroutan 2020; Iorio 2015). The GRADE system classifies the certainty of evidence in one of four grades:

• High: we are very confident that the variation in risk associated with the prognostic factor (probability of future events in those with/without the prognostic factor) lies close to that of the estimate.

• Moderate: we are moderately confident that the variation in risk associated with the prognostic factor (probability of future events in those with/without the prognostic factor) is likely to be close to the estimate, but there is a possibility that it is substantially different.

• Low: our certainty in the estimate is limited: the variation in risk associated with the prognostic factor (probability of future events in those with/without the prognostic factor) may be substantially different from the estimate.

• Very low: we have very little certainty in the estimate: the variation in risk associated with the prognostic factor (probability of future events in those with/without the prognostic factor) is likely to be substantially different from the estimate.

The certainty of evidence can be downgraded by one (serious concern) or two levels (very serious concern) for the following reasons: risk of bias, inconsistency, indirectness, imprecision, and publication bias. Conversely, the certainty of the evidence can also be upgraded by one level due to a large summary effect.

Notes

The methods section of this protocol was developed based on another protocol by one of the authors (Lokuhetty 2020), and the background section incorporated some of the descriptive text from another review by one of the authors (Vernooij 2020).

Appendices

Appendix 1. Search Strategies

Ovid MEDLINE(R) ALL <1946 to 18 October 2021>

1 exp Prostatic Neoplasms/

2 (prostat* adj3 (cancer* or carcinoma* or neoplas* or adeno* or transitional or squamous or malignan*)).mp.

3 1 or 2

4 decipher*1.mp.

5 "22‐gene".mp.

6 genomic classifier*.mp.

7 genomic prognosticator*.mp.

8 or/4‐7

9 3 and 8

Embase <1974 to 20 October 2021>

1 exp Prostate tumor/

2 (prostat* adj3 (cancer* or carcinoma* or neoplas* or adeno* or transitional or squamous or malignan*)).mp.

3 1 or 2

4 decipher*1.mp.

5 "22‐gene".mp.

6 genomic classifier*.mp.

7 genomic prognosticator*.mp.

8 or/4‐7

9 3 and 8

Cochrane Library [Wiley]

Search Name: decipher gene assay

Date Run: 21/10/2021 14:29:51

ID Search

#1 MeSH descriptor: [Prostatic Neoplasms] explode all trees

#2 prostat*:ti,ab,kw NEAR/3 (cancer* or carcinoma* or neoplas* or adeno* or transitional or squamous or malignan*):ti,ab,kw

#3 #1 or #2

#4 decipher*:ti,ab,kw

#5 "22‐gene":ti,ab,kw

#6 "genomic classifier*":ti,ab,kw

#7 "genomic prognosticator*":ti,ab,kw

#8 #4 or #5 or #6 or #7

#9 #3 and #8

Scopus [Elsevier]

( TITLE‐ABS‐KEY ( ( prostat* ) W/3 ( cancer* OR carcinoma* OR neoplas* OR adeno* OR transitional OR squamous OR malignan* ) ) ) AND ( TITLE‐ABS‐KEY ( decipher* OR "genomic prognosticator*" OR "genomic classifier*" OR "22‐gene" ) )

Web of Science Core Collection (1900‐present) [Clarivate]

History

7 #1 and #6

6 #2 or #3 or #4 or #5

5 TS=("genomic prognosticator*")

4 TS=("genomic classifier*")

3 TS=("22‐gene")

2 TS=(decipher*)

1 TS=(prostat* NEAR/3 (cancer* or carcinoma* or neoplas* or adeno* or transitional or squamous or malignan*))

LILACS [BIREME]

(decipher or "22‐gene" or "genomic classifier*" or "genomic prognosticator*") [Words] and prostat* [Words]

Clinicaltrials.gov

Basic search including synonyms

1. Prostate cancer AND decipher

2. Prostate cancer AND “22‐gene”

3. Prostate cancer AND genomic classifier*

4. Prostate AND genomic prognostifier*

ICTRP search portal

Basic search including synonyms

(decipher OR "22 gene" OR genomic classifier* or genomic prognostifier*) AND prostate cancer

CRD [University of York]

Database of Abstracts of Reviews of Effects (DARE)

& HTA database

1. All fields Prostat* AND Decipher

2. All fields Prostat* and genomic*

3. All fields Prostat* and 22 gene

INAHTA HTA [database.inahta.org/]

Advanced search

1. All fields Prostat* AND Decipher

2. All fields Prostat* and genomic*

3. All fields Prostat* and 22 gene

Internet general search [any result screened for relevant studies to 10 pages]

allintext: decipher postate hta OR appraisal OR quality OR assessment

Specific websites searched using word ‘decipher’ in page/publication titles. [any result screened for relevant studies]

Canadian Agency for Drugs and Technologies in Health www.cadth.ca/node/8

NIHR HTA programme (UK) www.nihr.ac.uk/explore-nihr/funding-programmes/health-technology-assessment.htm

European Network for Health Technology Assessment (EUnetHTA) www.eunethta.eu/

Australian Government Dept Health Medical Services Committee (MASC) msac.gov.au/

USA Technology Assessment Program www.ahrq.gov/research/findings/ta/index.html#tacomplete

Contributions of authors

LG drafted the background and the methods section of the protocol.

JF drafted the methods section of the protocol.

SEM drafted the background section of the protocol.

RV revised the protocol and provided methodological input.

JHJ revised the protocol and provided clinical input.

ECH revised the protocol and provided clinical input.

JAD revised the protocol and provided methodological input.

CW revised the protocol and provided clinical input.

BC developed the search methods.

PD developed the review idea, revised the protocol, and oversaw the project.

Sources of support

Internal sources

• Heinrich Heine University Düsseldorf, Germany

  Provides in‐kind support to JVAF

• Instituto Universitario Hospital Italiano de Buenos Aires, Argentina

  Provides in‐kind support to LG.

External sources

• No sources of support provided

Declarations of interest

LG declares no conflicts of interest.

JF is a Contact Editor for Cochrane Urology; however, he was not involved in the editorial process of this protocol.

SEM declares no conflicts of interest.

RV is a Contact Editor for Cochrane Urology; however, he was not involved in the editorial process of this protocol.

JHJ is a Contact Editor for Cochrane Urology; however, he was not involved in the editorial process of this protocol.

ECH is a Contact Editor for Cochrane Urology; however, he was not involved in the editorial process of this protocol.

JAD declares no conflicts of interest.

CW declares no conflicts of interest.

BC declares no conflicts of interest.

PD is the Co‐ordinating Editor of Cochrane Urology; however, he was not involved in the editorial process of this protocol.

New