Francisco Pina1,2, Ana Ferro3,4, Francisco Botelho1, Margarida Manso1,5, Nuno Dias1,5, Gabriela Figueiredo6, Pedro Pereira7, Paulo Dinis1,5, Henrique Barros3,4, Nuno Lunet8,9. 1. S. João University and Hospital Center-Urology, Porto, Portugal. 2. Lapa Hospital-Urology, Porto, Portugal. 3. Departamento de Ciências da Saúde Pública e Forenses e Educação Médica, Faculdade de Medicina da Universidade do Porto, Alameda Prof. Hernâni Monteiro, 4200-319, Porto, Portugal. 4. EPIUnit, Instituto de Saúde Pública, Universidade do Porto, Porto, Portugal. 5. Department of Surgery and Physiology, University of Porto Medical School, Porto, Portugal. 6. S. João University and Hospital Center-Clinical Pathology, Porto, Portugal. 7. S. João University and Hospital Center-Pathology, Porto, Portugal. 8. Departamento de Ciências da Saúde Pública e Forenses e Educação Médica, Faculdade de Medicina da Universidade do Porto, Alameda Prof. Hernâni Monteiro, 4200-319, Porto, Portugal. nlunet@med.up.pt. 9. EPIUnit, Instituto de Saúde Pública, Universidade do Porto, Porto, Portugal. nlunet@med.up.pt.
Abstract
PURPOSE: New biomarkers may contribute to avoid unnecessary biopsies resulting from the suboptimal performance of prostate-specific antigen (PSA) testing. This study aimed to assess serum endoglin as a prostate cancer (PCa) diagnostic tool among biopsy candidates. METHODS: A total of 262 consecutive patients referred for prostate biopsy based on abnormal digital rectal examination and/or elevated total PSA (tPSA) who had serum endoglin assessed by solid-phase enzyme-linked immunosorbent assay were selected. Receiver operating characteristic curves were used to compare the predictive accuracy of different combinations of biomarkers to distinguish between PCa and benign prostatic conditions, and to identify cut-offs that maximize the ability of endoglin to rule out patients for biopsy (highest sensitivities). RESULTS: Serum endoglin levels were higher in patients with PCa (median: 7.86 vs. 5.88 pg/mL, P < 0.001). Among patients with baseline tPSA ≤ 10 ng/mL the area under the curve was 0.69 for endoglin. Approximately one-quarter of the patients had serum endoglin < 4.92 ng/mL (sensitivity: 90.3%; specificity: 32.8%), and the probability of PCa varied from 37.7% before testing to 15.2% among those with low endoglin levels [negative predictive value (NPV) = 84.8%]. When restricting the analyses to patients with free/total PSA ratio > 0.25, the probability of cancer was less than 5% among those with serum endoglin < 6.04 ng/mL (sensitivity: 93.8%; specificity: 56.1%), corresponding to a NPV of 95.8%; this could allow sparing approximately 40% of patients from biopsy. CONCLUSIONS: Serum endoglin may be useful in clinical practice to distinguish between PCa and non-cancer patients among prostatic biopsy candidates.
PURPOSE: New biomarkers may contribute to avoid unnecessary biopsies resulting from the suboptimal performance of prostate-specific antigen (PSA) testing. This study aimed to assess serum endoglin as a prostate cancer (PCa) diagnostic tool among biopsy candidates. METHODS: A total of 262 consecutive patients referred for prostate biopsy based on abnormal digital rectal examination and/or elevated total PSA (tPSA) who had serum endoglin assessed by solid-phase enzyme-linked immunosorbent assay were selected. Receiver operating characteristic curves were used to compare the predictive accuracy of different combinations of biomarkers to distinguish between PCa and benign prostatic conditions, and to identify cut-offs that maximize the ability of endoglin to rule out patients for biopsy (highest sensitivities). RESULTS: Serum endoglin levels were higher in patients with PCa (median: 7.86 vs. 5.88 pg/mL, P < 0.001). Among patients with baseline tPSA ≤ 10 ng/mL the area under the curve was 0.69 for endoglin. Approximately one-quarter of the patients had serum endoglin < 4.92 ng/mL (sensitivity: 90.3%; specificity: 32.8%), and the probability of PCa varied from 37.7% before testing to 15.2% among those with low endoglin levels [negative predictive value (NPV) = 84.8%]. When restricting the analyses to patients with free/total PSA ratio > 0.25, the probability of cancer was less than 5% among those with serum endoglin < 6.04 ng/mL (sensitivity: 93.8%; specificity: 56.1%), corresponding to a NPV of 95.8%; this could allow sparing approximately 40% of patients from biopsy. CONCLUSIONS: Serum endoglin may be useful in clinical practice to distinguish between PCa and non-cancer patients among prostatic biopsy candidates.