Original Title: Marcadores moleculares para la detección de cáncer de próstata significativo

Authors' objectives: The main objective of this Health Technology Assessment report is to evaluate the effectiveness, cost-effectiveness and ethical, legal, organizational, social and patient, and environmental aspects of the incorporation of diagnostic tests, based on biomarkers in samples of blood or urine in the identification of patients with a higher risk of suffering PCa, making it possible to reduce unnecessary procedures (overdiagnosis and overtreatment).

Authors' results and conclusions: RESULTS: EFECTIVENESS AND SAFETY: No RCTs were found that could be finally included according to the pre-established selection criteria. Therefore, diagnostic performance studies with an observational design were considered. Sixty-five studies were included with a total of 34,287 participants, conducted in Germany, Austria, China, Korea, the United States, Spain, France, the Netherlands, Ireland, Italy, Japan, Lithuania, Norway, the United Kingdom, Russia, Singapore, and Taiwan. Seven of the studies were conducted in multiple countries simultaneously. The overall quality of the evidence was considered very low. The sample size of the studies ranged from 50 to 1,544 participants. The mean age of the men included in the studies was around 64.45 years (SD: ±2.19). The diagnostic tests analyzed in the selected studies were as follows: PHI (37 studies), PHI density (5 studies), Progensa® PCA3 (12 studies), SelectMDx (5 studies), MyProstateScore (5 studies), 4Kscore® test (4 studies), TMPRSS2: ERG (2 studies), Stockholm3 (2 studies), ExoDx Prostate IntelliScore (1 study) and Proclarix® test (1 study). All studies included biopsy as a comparator. None of the studies considered evaluated the incorporation of the NADiA ProsVue and Prostate Core Mitomic Test blood tests, as well as the Post-DRE Urine Test and ProstarixTM urine-based tests. With respect to the urine tests, Progensa® PCA3, for the cut-off point of 15, yielded a sensitivity that was found between 93% and 99% and a specificity of 37%. The cut-off point of 20 showed a sensitivity of between 89% and 99% and a specificity of 51%. Finally, at the cut-off point of 35, the sensitivity ranged between 62% and 71%, while the specificity increased to 59-66%. In the case of the SelectMDx test, an average sensitivity and specificity of 84% (95% confidence interval (CI): 71% to 92%; I² = 79.7%; k = 5; n = 1,957) and 49% (95% CI: 26% to 72%; I² = 93.9%; k = 5; n = 1,957) were obtained, respectively. Regarding the MyProstateScore, this test showed a sensitivity of between 96.6% and 97.4% and a specificity of between 28.6% and 34.6%. No data related to the sensitivity and specificity of TMPRSS2: ERG (area under the curve, AUC: 0.64-0.75) were identified. Regarding the blood tests, for PHI, with a cut-off point between 15 and 20, the average sensitivity and specificity were 99% (95% CI: 97% to 100%; I² = 76.26%; k = 4; n = 2,994) and 14% (95% CI: 9% to 19%; I² = 87.03%; k = 4; n = 2,994), respectively. When the cut-off point was raised to the 20-25 range, an average sensitivity and specificity of 96% was obtained (95% CI: 94% to 98%; I² = 73.26%; k = 7; n = 6,698) and 24% (95% CI: 18% to 30%; I² = 95.57%; k = 7; n = 6,698), respectively. With a cut-off point of 25-30, the PHI showed a sensitivity for detecting the risk of suffering PCa of 95% (95% CI: 89% to 98%; I² = 85.72%; k = 10; n = 6,321) and a specificity of 33% (95% CI: 23% to 45%; I² = 98.27%; k = 10; n = 6,321). For the cut-off point between 30 and 35, the average sensitivity and specificity were 87% (95% CI: 81% to 91%; I² = 88.94%; k = 10; n = 5,964) and 49% (95% CI: 41% to 58%; I² = 95.53; k = 10; n = 5,964), respectively. For the cut-off point between 35 and 40, the average sensitivity and specificity were 79% (95% CI: 66% to 88%; I² = 83.13%; k = 5; n = 1,164) and 56% (95% CI: 48% to 64%; I² = 81.78; k = 5; n = 1,164). Finally, with a cut-off point of 55, the sensitivity was 42% (95% CI: 32% to 53%; I² = 73.61%; k = 3; n = 2,028), while the specificity was 87% (95% CI: 72% to 95%; I² = 98.28%; k = 3; n = 2,028). PHI shows a higher sensitivity and specificity in patients of Asian origin than in those patients of European origin for the 15-20 ranges (S = 100%; 95% CI: 99% to 100%; E = 13%; 95% CI: 4% to 22% vs. S = 98%; 95% CI: 96% to 100%; E = 13%; 95% CI: 8% to 19%) and 20-25 (S = 99%; 95% CI: 97% to 100%; E = 30%; 95% CI: 20% to 39% vs. S = 95%; 95% CI: 93% to 97%; E = 21%; 95% CI: 16% to 26%). The sensitivity in the 30-35 ranges was higher in patients of Asian origin (S = 91%; 95% CI: 84% to 97%) and African American origin (S = 91%; 95% CI: 78% to 100%) than in those patients of European origin (S = 85%; 95% CI: 79% to 91%), and the specificity was higher in patients of Asian origin (E = 58%; 95% CI: 46% to 70%) followed by those patients of European origin (E = 49%; 95% CI: 41% to 57%) and patients of African American origin (E = 26%; 95% CI: 11% to 41%). In the 35-40 range, no differences were observed between patients of Asian origin and those of European origin. No subgroup analysis could be performed by type of PCa risk definition. In relation to the 4Kscore® test and, assuming a risk of suffering csPCa equal to or greater than 7.5%, the sensitivity was 95.5% and the specificity was 32.1%; whereas when a risk of 12% was assumed, the sensitivity decreased to 90.1% and the specificity increased to 53.5%. Studies identified on the Stockholm3 test do not report sensitivity and specificity data for the Stockholm3 test. The AUC ranged from 0.77 to 0.86. Proclarix® had a sensitivity and specificity of 91% and 22%, respectively. COST-EFFECTIVENESS: Seven economic evaluations were included in the SR, five of them with a long-term horizon and two with a short-term horizon. The biomarkers evaluated were PHI, EPI, SelectMDx, and 4Kscore. One of the economic evaluations was carried out for Spain (financed by the industry) and consisted of a cost-utility analysis that compared SelectMDx before biopsy with biopsy for all subjects from the perspective of the health system. In general, all the studies are of medium methodological quality and conclude that biomarkers are cost-effective and even dominant, that is, more effective and less expensive than the usual strategy (biopsy). The results of the base case of the model developed for this report, for the short term and from the perspective of the SNS, place the alternative that includes the biomarker, specifically the SelectMDx, at €1,112.77 and 0.87 QALY, and at €862.20 and 0.84 QALY for the usual strategy. These values give rise to an ICER of €6,640.21 /QALY, which is well below the cost-effectiveness threshold recommended in Spain (€20,000 - €25,000 / QALY). Furthermore, this analysis allows us to conclude that the biomarker alternative generates a reduction in the number of performed biopsies, with an average of 414 avoided biopsies per 1,000 patients. For the long term, the biomarker becomes a dominant technology compared to the usual strategy. Probabilistic sensitivity analysis reinforces these results in both the short and long term. When analyzing the acceptability curve of the intervention to be evaluated, it is observed that, based on an availability to pay of €25,000/QALY, the biomarker would be cost-effective with a probability of 81% from the perspective of the NHS. Regarding the social perspective, the biomarker is cost-effective for both time horizons, presenting a lower ICER in the long term. The estimates obtained in the budget impact analysis show that the incorporation of biomarkers in the early detection of PCa in the NHS would mean an annual increase of approximately 15 million euros, for a population of 60,000 men and a unit price of the biomarker of €375. If the price fell to €100, there would be a saving of more than 1 million euros, with the same population size, and more than 2 million in the case of a population of 100,000 men. ETHICAL, LEGAL, ORGANIZATIONAL, SOCIAL AND PATIENT, AND ENVIRONMENTAL ASPECTS: Twenty articles were considered for the evaluation of these aspects. None of them address environmental aspects related to technology. For the rest of the aspects, the findings could be grouped into the following topics: Ethical challenges of molecular biomarkers Communication with patients and tools to aid decision-making Overdiagnosis, false negatives and new problems derived from biomarkers Problems for the clinical implementation of molecular biomarkers. CONSLUSIONS: Based on the SR on effectiveness and cost-effectiveness, the economic evaluation and the analysis of the ethical, legal, organizational, social, patient-related, and environmental aspects carried out for this ETS report, the following conclusions can be established on the incorporation of diagnostic tests based on biomarkers in blood or urine samples for the identification of patients with a higher risk of suffering csPCa: • The following conclusions can be drawn from the best available evidence for effectiveness, which comes from 65 studies with a total of 34,287 included patients that evaluate the diagnostic accuracy of a large number of different tests: [PROGENSA® PCA3] - With a cut-off point of 15, the sensitivity of the test is between 93% -99% and the specificity is 37%. - With a cut-off point of 20, the sensitivity of the test is between 89% -99% and the specificity is 51%. - With a cut-off point of 35, the sensitivity of the test is between 62% -71% and the specificity is between 59% -66%. [SELECTMDX] - The sensitivity and specificity of the test for the identification of patients with a higher risk of suffering PCa are between 54.5% -100% and 35.3% -78.6%, respectively (quality of evidence: very low ⊕⊖⊖⊖). [MYPROSTATECORE] - The available evidence on MyProstateScore is scarce (1 study). With a cut-off point of less than or equal to 10, the MyProstateScore shows a sensitivity of between 96.6% -97.4% and a specificity of between 28.6% -34.6%. When the cut-off point exceeds 10, the sensitivity ranges from 95.5% -96.7%, while the specificity remains between 29% -33.3%. [PHI] - With a cut-off point between 15 and 20, the sensitivity of the PHI diagnostic test is 99% and the specificity 14%. - With a cut-off point between 20 and 25, the sensitivity of the PHI diagnostic test is 96% and the specificity 24% (quality of evidence: low ⊕⊕⊖⊖). - With a cut-off point between 25 and 30, the sensitivity of the PHI diagnostic test is 95% and the specificity 33%. - With a cut-off point between 30 and 35, the sensitivity of the PHI diagnostic test is 87% and the specificity 49%. - With a cut-off point between 35 and 40, the sensitivity of the PHI diagnostic test is 79% and the specificity 56%. - With a cut-off point of 55, the sensitivity of the PHI diagnostic test is 42% and the specificity 87%. - The sensitivity and specificity of the PHI density for the identification of patients with a higher risk of suffering PCa are between 71.6%-97.9% and 29.1% -66.5%, respectively. For the identification of patients with PCa defined as a Gleason score ≥7, the range for sensitivity is 90% -97% and specificity is between 32% -39%. [4KSCORE® TEST] - According to the limited evidence available on the 4Kscore® Test (1 study), the sensitivity and specificity of it vary between 90.1% -95.5% and 32.1% -53.5%, respectively. [STOCKHOLM3 TEST] - No data related to the sensitivity and specificity of the Stockholm3 test were identified. The AUC of the Stockholm3 test ranges from 0.77 to 0.86. [PROCLARIX®] - According to the limited evidence available on Proclarix® (1 study), the sensitivity and specificity of the test is 91% and 22%, respectively. [OTHER TESTS] - The available evidence on Proclarix®, TMPRSS2: ERG, and ExoDx Prostate IntelliScore is still very scarce. [SUBGROUP ANALYSIS] - The analysis of subgroups by ethnic origin can be used to establish that: With a cut-off point between 15 and 20, the sensitivity of the PHI diagnostic test is slightly higher in patients of Asian origin (100%) than in those patients of European origin (98%). With a cut-off point between 20 and 25, both the sensitivity and the specificity are higher in patients of Asian origin (sensitivity=99%; specificity=30%) than in those patients of European origin (sensitivity=95%; specificity=21%). With a cut-off point between 25 and 30, the sensitivity of the PHI diagnostic test is higher in patients of Asian origin (sensitivity=98%) than in those patients of African-American origin (sensitivity=94%) and European (sensitivity=93%). The specificity of PHI is higher in patients of Asian origin (specificity=41%), followed by those patients of European origin (specificity=30%) and African-American (specificity=15%). With a cut-off point between 30 and 35, the sensitivity of the PHI diagnostic test is higher in patients of Asian origin (sensitivity=91%) and African-American (sensitivity=91%) than in patients of European origin (sensitivity=85%). Specificity is higher in patients of Asian origin (specificity=58%) followed by patients of European origin (specificity=49%) and African-American (specificity=26%). With a cut-off point between 35 and 40, no differences were observed between patients of Asian and European origin. • No studies were identified that met the pre-established selection criteria on the NADiA ProsVue and Prostate Core Mitomic Test blood tests, or on the Post-DRE Urine Test and ProstarixTM urine-based tests. • Published cost-effectiveness studies conclude that biomarkers are cost-effective in their respective countries in most circumstances. An economic evaluation of the SelectMDx biomarker before biopsy performed in Spain and funded by the industry concludes that it is a dominant strategy, that is, more effective and less expensive, compared to biopsy for all subjects. • The independent economic evaluation carried out for this report concludes that the strategy that includes performing the biomarker test before the prostate biopsy, and performing a biopsy only when the test indicates a probability of clinically significant cancer, can be a cost-effective strategy depending on its cost per determination and its sensitivity/specificity (or cut-off point chosen for the interpretation of results). The analyses carried out indicate that the SelectMDx biomarker is cost-effective from the NHS perspective and from the social perspective for a cost per determination of €375 (cost provided by the company). For the other biomarkers, only tentative estimates have been made due to the lack of information on their unit costs in Spain. • The adoption of the strategy consisting of performing the biomarker test before biopsy in the entire NHS would imply a net budgetary impact that can be positive or negative (savings) depending on the unit cost per determination. For a population of 60,000 men and a unit price of the biomarker of €375, the budget impact analysis would mean an annual increase of approximately 15 million euros. If the price fell to €100, there would be a saving between 1 and 2 million euros. • Studies are needed on patients' understanding of biomarkers, their expectations, attitudes, and values, as well as their concerns, to assist in future patient care and policy formulation. • Communication with patients is of utmost importance and requires closeness in the doctor/patient relationship: the decision to offer PCa screening tests needs to be individualized and needs to be supported by sources of information with proven solvency. • As regards fairness, older men should not be automatically excluded from molecular biomarker screening for PCa simply because of their age. • Regarding the implementation of biomarkers in clinical practice, the main problems are related to the design and validity of the studies. A facilitating factor is the dissatisfaction of professionals with the use of PSA, so that a better diagnostic test with clear guidelines is a powerful incentive for change.

Authors' recommendations: Due to the uncertainty associated with the estimates related to effectiveness at the time of preparing this report, with the available evidence on cost-effectiveness and the consideration of ethical, legal, organizational, social and patient, and environmental-related aspects, it is not possible to make a recommendation for the implementation of the SelectMDx biomarker for the identification of patients with a higher risk of suffering csPCa to guide decision making for first prostate biopsy, except in the strict context of research (Recommendation: CONDITIONAL). Due to the uncertainty associated with the estimates related to effectiveness at the time of preparing this report, with the available evidence on cost-effectiveness and the consideration of ethical, legal, organizational, social and patient, and environmental-related aspects, it is not possible to make a recommendation for the implementation of the PHI for the identification of patients with a higher risk of suffering csPCa to guide decision making for first prostate biopsy, except in the context of research (Recommendation: CONDITIONAL). According to the limited available evidence, it is not possible to make a recommendation on the implementation of diagnostic tests based on biomarkers in urine samples Progensa® PCA3, Mi-Prostate Score, TMPRSS2:ERG and ExoDx Prostate IntelliScore urine exosome assay and blood samples 4Kscore®, Stockholm3 and Proclarix®.

Authors' methods: EFFECTIVENESS: A systematic review (SR) of the literature published up to March 2021 was carried out in the following databases: Medline and EMBASE. A comprehensive search strategy was applied that included controlled vocabulary and free text terms. The search was completed with the manual examination of the bibliographic list of the included articles. Original studies published in English or Spanish were selected that evaluated the effectiveness of diagnostic tests based on biomarkers in blood or urine samples for the identification of patients with a higher risk of suffering PCa. Randomized clinical trials (RCTs) were selected. Failing that, it was proposed to consider cohort and case-control studies that performed an evaluation of the diagnostic validity of the markers in men with suspected PCa and clinical factors that suggested a higher risk of PCa. The outcome measures considered were: sensitivity, specificity, positive predictive value, negative predictive value and area under the ROC curve (AUC). The risk of bias of the included observational studies was assessed according to the QUADAS-2 scale. Quantitative synthesis of the results was performed by meta-analysis using the statistical program STATA in its version 14. The evaluation of the quality of the evidence and the grading of the strength of the recommendations was carried out following the methodology of the international working group Grading of Recommendations Assessment, Development and Evaluation (GRADE). COST-EFFECTIVENESS: An SR of cost-effectiveness studies was conducted from the search described above. Selection criteria included economic evaluations developed in parallel with primary studies and economic models. Outcome measures could be quality-adjusted life years (QALYs) or any of the other measures selected in the effectiveness review. The selection, critical reading, and data extraction was performed by economists. The methodological quality was assessed with the Critical Reading Cards of OSTEBA. A narrative synthesis of the results was carried out. A de novo economic evaluation was developed for Spain in which the performance of a biomarker test, before biopsy, was compared with only biopsy, with the aim of detecting PCa in suspected men in Spain. In the strategy to be evaluated, only men with a positive biomarker or higher than the cut-off point would undergo the biopsy. A decision tree was designed to model the short term (one year), and a Markov model, with annual cycles, for the long term (the entire life of the patient), in which both the perspective of the National Health System (NHS), including direct health costs, such as social, also adding productivity losses. Both perspectives were presented separately. Health outcomes were measured in terms of years of life gained (YLG) and QALY and, additionally, in the short term, avoided biopsies. The incremental cost-effectiveness ratio (ICER) is also estimated. The costs were expressed in 2021 euros. A deterministic sensitivity analysis and a probabilistic sensitivity analysis were carried out to quantify the uncertainty surrounding the decision to be made. As a complementary measure, a budget impact analysis was carried out to estimate the cost for the NHS of incorporating biomarker tests, before biopsy, in the population susceptible to receiving this technology. This analysis was carried out for a time horizon of one year. Similarly, several scenarios were evaluated, drawn from different unit costs of the biomarker and population sizes, through a sensitivity analysis on the budgetary impact. ETHICAL, LEGAL, ORGANIZATIONAL, SOCIAL AND PATIENT, AND ENVIRONMENTAL ASPECTS: The scope of the evaluation of these aspects was based on the same population, intervention and comparison described above for the evaluation of the effectiveness and cost effectiveness. The electronic databases MEDLINE, EMBASE and WOS were systematically searched for articles (search date: March 2021). A narrative synthesis of the results was performed, taking into account criteria of relevance and coherence of the results.

Project Status: Completed

Year Published: 2021

URL for published report: https://sescs.es/marcadores-moleculares-para-la-deteccion-de-cancer-de-prostata-significativo/

English language abstract: An English language summary is available

Publication Type: Full HTA

Country: Spain

Organisation Name: Canary Health Service

Contact Address: Dirección del Servicio. Servicio Canario de la Salud, Camino Candelaria 44, 1ª planta, 38109 El Rosario, Santa Cruz de Tenerife

Contact Name: sescs@sescs.es

Contact Email: sescs@sescs.es