摘要
目的评价Gallina列线图预测前列腺癌精囊浸润（SVI）的准确性。方法2009年1月至2011年12月，89例前列腺癌患者接受耻骨后或腹腔镜下根治性前列腺切除术，术前资料具备血清PSA、临床分期、穿刺Gleason评分、前列腺穿刺活检阳性针数比例、盆腔MRI资料以及术后病理报告，并符合Gallina列线图与2001年版、2007年版Partintables的纳入标准。将其术前MRI及应用3个预测工具对前列腺癌SVI预测的结果分别与术后病理结果进行比较，计算MRI预测前列腺癌SVI的敏感性、特异性和准确性，再通过受试者工作特征曲线分析法检验Gallina列线图与两版Partintables预测前列腺癌SVI的准确性并进行比较。结果89例中器官局限性前列腺癌、包膜侵犯、SVI及淋巴结转移的发生率分别为58．4％、16．9％、22．5％、2．2％。应用MRI预测前列腺癌SVI的敏感性、特异性、准确性分别为45．0％（9／20）、71．O％（49／69）、65．2％（58／89）。应用2001年版、2007年版Partintables及Gallina列线图预测SVI的曲线下面积（AUC）分别为0．712、0．711、0．801。结论MRI预测前列腺癌SVI的敏感性较差，特异性与准确性一般；两版Partintables与Gallina列线图对国人前列腺癌术后SVI有预测价值（AUC〉0．7），Gallina列线图比两版Partintables显示出更佳的预测准确性。
Objective To evaluate the accuracy of Gallina nomogram in predicting seminal vesicle invasion （SVI） in prostate cancer. Methods From January 2009 to December 2011, 89 patients with prostate carcinoma underwent open retropubic or laparoscopic radical prostatectomy. Complete data of pre- operative serum prostate specific antigen （PSA） , clinical stage, biopsy Gleason score, percentage of positive biopsy cores, pelvic MRI and pathological report of prostatectomy specimen were collected, and all the pa- tients met the inclusion criteria of Gallina nomogram, 2001 Partin tables and 2007 Partin tables. Postopera- tive pathological results were respectively compared with MRI and the incidence of SVI predicted by the three tools, and the sensitivity, specificity and accuracy of MRI in predicting SVI were calculated. The receiver operating characteristics curves were performed to test the predictive accuracy of SVI of each tool. Results The incidences of organ-confined disease, capsule invasion, SVI and lymph node metastasis were 58.4% , 16.9% , 22.5% , and 2.2% , respectively. The sensitivity, specificity and accuracy of MRI in predicting SVI was 45.0% （9/20）,71.0% （49/69） and 65.2% （58/89）, respectively. The area under the curve （AUC） for SVI disease prediction of 2001 Partin tables, 2007 Partin tables and Gallina nomogram was 0. 712, 0. 711 and 0. 801, respectively. Conclusions The sensitivity of MRI in predicting SVI is poor,the specificity and accuracy are common. All the predictive tools have a reasonable value for SVI （ AUC 〉 0. 7） , and Gallina nomogram is superior to two versions of Partin tables in predicting SVI.
出处
《中华泌尿外科杂志》
CAS
CSCD
北大核心
2013年第5期369-373,共5页
Chinese Journal of Urology
基金
天津市科委抗癌重大专项攻关计划项目（12ZCDZSYl7200）
天津市应用基础及前言技术研究计划自然科学基金一般项目（12JCYBJC3140）
