目的：本研究目的針對攝護腺特定抗原（prostate-specific antigen,t PSA）為4 ～ 10 奈公克／毫公升（ng/mL）且經肛門指診檢查（Digital Rectal Examination, DRE） 正常 的攝護腺癌（prostate cancer）病患，找出較好的診斷指標。方法：我們利用回顧性評估， 從2008 年1 月至2013 年6 月，490 例患者在我們醫院接受了經直腸超聲波引導下攝 護腺穿刺（transrectal ultrasound-guided prostate biopsy）。排除肛門指診（DRE）異常 和攝護腺特定抗原（PSA）>10 奈公克／毫公升（ng/mL）的病人。最後收集了238 名男性，以此數據來做統計分析。另外使用接受者操作特性曲線（receiver operating characteristic，ROC）分析，我們評估t PSA、f/tPSA、攝護腺特定抗原密度（Prostatespecific antigen density, PSAD），來比較診斷前列腺癌的準確度。結果：238 人中的47 人（19.7%）有攝護腺癌。攝護腺特定抗原密度（PSAD）有顯著差異（平均0.306 奈 公克／毫公升平方（ng/mL2）和0.156 奈公克／毫公升平方（ng/mL2），P <0.001）。 ROC 曲線下對tPSA、f/tPSA、PSAD 分別為59.8%，36.6% 和77.4%。當PSAD 在0.16 奈公克／毫公升平方（ng/mL2）的數值，針對攝護腺癌的診斷有較高的靈敏度和特異 性（分別為72% 和61%）。ROC 曲線分析表明，PSAD 是一個更好的預測前列腺癌 相對於f/tPSA。結論：結果表明，PSAD 在護腺特定抗原為4 ～ 10 奈公克／毫公升（ng/ mL）且經肛門指診（DRE）檢查正常的病患中，有提高診斷攝護腺癌的準確率。

Purpose: This study was conducted to identify the optimal predictor of prostate cancer among several prostate-specific antigen (tPSA) derivatives in prostate biopsies of patients with PSA levels of 4–10 ng/mL and normal digital rectal examination (DRE) findings. Materials and methods: We retrospectively assessed prostate biopsy specimens of 490 patients who underwent transrectal ultrasound-guided prostate biopsy in our hospital from January 2008 to June 2013. We excluded patients with abnormal DRE findings and serum tPSA levels of >10 ng/mL. Finally, 238 men with serum tPSA levels of 4–10 ng/mL and normal DRE findings were included in the statistical analysis. We performed receiver operating characteristic (ROC) curve analysis and assessed the predictive power of tPSA, percentage of free PSA (f/tPSA), and PSA density (PSAD) for detecting prostate cancer. Results: Overall, 47 (19.7%) of the 238 prostate biopsy specimens were positive for prostate cancer. Patients with positive and negative biopsy results had significant differences in the mean PSAD (0.306 and 0.156 ng/mL2, respectively, p < 0.001). The areas under the ROC curves for tPSA, f/tPSA, and PSAD were 59.8%, 36.6%, and 77.4%, respectively. Moreover, when PSAD was considered as a predictor at an optimal cutoff value of 0.16 ng/mL2, the detection of prostate cancer was highly sensitive (72%) and specific (61%). The ROC curve analysis indicated that PSAD was a more accurate predictor of prostate cancer than was f/tPSA. Conclusion: The results suggest that PSAD improves the specificity of cancer detection in men with normal DRE findings and an intermediate PSA level.