J Ruiz-Aragón1, S Márquez-Peláez. 1. Agencia de Evaluación de Tecnología Sanitaria de Andalucía, Consejería de Salud, Junta de Andalucía, España. jesusm.ruiz.ext@juntadeandalucia.es
Abstract
OBJECTIVE: To assess the efficacy of diagnostic techniques based on PCA3 gene for early detection of prostate cancer. We carried out a systematic review of scientific literature and subsequent meta-analysis. MATERIAL AND METHODS: A literature search (2000-09) in MEDLINE, EMBASE, Cochrane Library, CRD, ECRI, Hayes databases and journals of Cancer and Urology. MESH terms used were "Prostatic Neoplasms", Prostate-Specific Antigen", "Antigens, Neoplasm", "Sensitivity and Specificity", "Predictive Value of Tests", and free terms "upm3", "PCA3", "dd3", "aptima PCA3" and "prostate cancer antigen 3". Patients were adults. The intervention was to determine the PCA3 gene, from urine samples for diagnosis of prostate cancer. The quality of the studies was checked according to QUADAS criteria. We calculated diagnostic accuracy rates and developed a meta-analysis to synthesize results. RESULTS: 14 studies of diagnostic tests were selected, with moderate-high quality. The sensitivity was between 46.9% and 82.3%, specificity ranged from 56.3% to 89%, positive predictive value had a range of 59.4-97.4% and negative predictive value 87.8-98%. The meta-analysis detected the existence of a threshold effect and heterogeneity between studies. Global sensitivity values was 0.85 [CI 0.84-0.87], specificity 0.96 [CI 0.96-0.97], positive likelihood ratio 22.21 [CI 15.12-32.63], and negative of 0.15 [CI 0.13-0.18]. CONCLUSIONS: Detection techniques have acceptable diagnostic accuracy rates for using in the diagnosis of prostate cancer.
OBJECTIVE: To assess the efficacy of diagnostic techniques based on PCA3 gene for early detection of prostate cancer. We carried out a systematic review of scientific literature and subsequent meta-analysis. MATERIAL AND METHODS: A literature search (2000-09) in MEDLINE, EMBASE, Cochrane Library, CRD, ECRI, Hayes databases and journals of Cancer and Urology. MESH terms used were "Prostatic Neoplasms", Prostate-Specific Antigen", "Antigens, Neoplasm", "Sensitivity and Specificity", "Predictive Value of Tests", and free terms "upm3", "PCA3", "dd3", "aptima PCA3" and "prostate cancer antigen 3". Patients were adults. The intervention was to determine the PCA3 gene, from urine samples for diagnosis of prostate cancer. The quality of the studies was checked according to QUADAS criteria. We calculated diagnostic accuracy rates and developed a meta-analysis to synthesize results. RESULTS: 14 studies of diagnostic tests were selected, with moderate-high quality. The sensitivity was between 46.9% and 82.3%, specificity ranged from 56.3% to 89%, positive predictive value had a range of 59.4-97.4% and negative predictive value 87.8-98%. The meta-analysis detected the existence of a threshold effect and heterogeneity between studies. Global sensitivity values was 0.85 [CI 0.84-0.87], specificity 0.96 [CI 0.96-0.97], positive likelihood ratio 22.21 [CI 15.12-32.63], and negative of 0.15 [CI 0.13-0.18]. CONCLUSIONS: Detection techniques have acceptable diagnostic accuracy rates for using in the diagnosis of prostate cancer.