Fujikawa K, Sasaki M, Itoh T, Arai Y, Ogawa O, Yoshida O
Department of Urology, Kobe City General Hospital, Japan.
Prostate. 1998 Oct 1;37(2):63-9. doi: 10.1002/(sici)1097-0045(19981001)37:2<63::aid-pros1>3.0.co;2-m.
Various criteria for patients with prostate cancer have been reported to be of prognostic value, and we have reported that estimates of volume-weighted mean nuclear volume (MNV), developed by Gundersen and Jensen based on a stereological technique, accurately predict the prognosis of prostate cancer. However, all of these studies were conducted on cases in a single institution, and it has remained unclear whether MNV calculations obtained at one institution apply to cases at another institution. In attempting to solve this problem, we made a prognostic index (P.I.) based on data from one hospital, and tested whether these data could be used to predict the prognosis of patients at another hospital.
A retrospective, multivariate prognostic study of 195 patients with prostate cancer, diagnosed at Kyoto University Hospital and treated conservatively, indicated that clinical stage, Gleason score, and MNV were all significantly correlated with the prognosis of patients with prostate cancer. From the relative strengths of these prognostic factors in a multivariate analysis, the following P.I. was constructed: P.I. = Clinical stage x 1.8040 + Gleason score x 1.5245 + MNV x 2.3162 (the constants correspond to the risk ratio estimated by Cox analysis). The P.I. was calculated for 104 patients with prostate cancer diagnosed at Shizuoka City Hospital and treated conservatively for analysis of disease-specific survival.
The prognostic index ranged from 3.841-16.142. Using the median value of 12.5 as a cutoff point, a clear separation of cases with poor and favorable prognosis was achieved (P < 0.0001, observation period: 1-167 months).
The results of this study suggest that estimates of MNV can be evaluated at multiple institutions with the use of P.I. calculation. Furthermore, combining estimates of MNV with Gleason score and clinical stage predicts most powerfully disease outcome of patients with prostate cancer.
已有报道称,多种前列腺癌患者标准具有预后价值,并且我们曾报道,贡德森和詹森基于体视学技术得出的体积加权平均核体积(MNV)估计值能够准确预测前列腺癌的预后。然而,所有这些研究均是在单一机构的病例中进行的,尚不清楚在一个机构获得的MNV计算结果是否适用于另一机构的病例。为解决这一问题,我们基于一家医院的数据制定了一个预后指数(P.I.),并测试这些数据是否可用于预测另一家医院患者的预后。
对195例在京都大学医院确诊并接受保守治疗的前列腺癌患者进行回顾性多因素预后研究,结果表明临床分期、Gleason评分和MNV均与前列腺癌患者的预后显著相关。根据多因素分析中这些预后因素的相对强度,构建了以下P.I.:P.I. = 临床分期×1.8040 + Gleason评分×1.5245 + MNV×2.3162(常数对应于Cox分析估计的风险比)。计算了静冈市医院确诊并接受保守治疗的104例前列腺癌患者的P.I.,以分析疾病特异性生存率。
预后指数范围为3.841 - 16.142。以中位数12.5作为分界点,实现了预后不良和良好病例的清晰区分(P < 0.0001,观察期:1 - 167个月)。
本研究结果表明,使用P.I.计算可在多个机构评估MNV估计值。此外,将MNV估计值与Gleason评分和临床分期相结合,对前列腺癌患者疾病转归的预测能力最强。