Horwitz E M, Vicini F A, Ziaja E L, Dmuchowski C F, Stromberg J S, Martinez A A
Department of Radiation Oncology, William Beaumont Hospital, Royal Oak, Michigan 48073, USA.
Int J Radiat Oncol Biol Phys. 1998 May 1;41(2):267-72. doi: 10.1016/s0360-3016(98)00078-9.
We reviewed our institution's experience treating patients with external beam irradiation (RT) to determine if the ASTRO Consensus Panel definition of biochemical failure (BF) following radiation therapy correlates with clinical distant metastases free survival (DMFS), disease-free survival (DFS), cause-specific survival (CSS), and local control (LC).
Between 1/1/87 and 12/31/92, 568 patients with clinically localized prostate cancer received external beam irradiation (RT) using localized prostate fields at William Beaumont Hospital (median total dose 66.6 Gy; range: 60-70.4 Gy). Biochemical failure was defined as three consecutive increases in post-treatment prostate specific antigen (PSA) after achieving a nadir. Biochemical failure was recorded as the time midway between the nadir and the first rising PSA. Five-year actuarial rates of clinical DMFS, DFS, CSS, and LC were calculated for patients who were biochemically controlled (BC) versus those who failed biochemically. Median follow-up was 56 months (range: 24-118 months).
Five-year actuarial rates of DMFS, DFS, CSS, and LC were significantly greater in patients who were biochemically controlled versus those who were not (p < 0.001). In patients who were BC, the 5-year actuarial rates of DMFS, DFS, CSS, and LC were 99%, 99%, 98%, and 99% respectively. For patients who failed biochemically, the 5-year actuarial rates of DMFS, DFS, CSS, and LC were 74%, 64%, 89%, and 86% respectively. When stratifying by pretreatment PSA, Gleason score, and T stage these differences remained significant for DMFS, DFS, and CSS. The Cox proportional hazards model demonstrated that BC was the single most important predictor of clinical outcome for DMFS, DFS, CSS, and LC. Pretreatment PSA and Gleason score were also independent predictors of outcome for DMFS and DFS.
The ASTRO Consensus Panel definition of BF following radiation therapy correlates well with clinical DMFS, DFS, and CSS. These findings suggest that the Consensus Panel definition may be a surrogate for clinical progression and survival and should be considered a valid endpoint for separating successful versus unsuccessful treatment. Additional studies with longer follow-up will be needed to corroborate these findings.
我们回顾了本机构对外照射放疗（RT）患者的治疗经验，以确定美国放射肿瘤学会（ASTRO）共识小组对放疗后生化失败（BF）的定义是否与临床无远处转移生存期（DMFS）、无病生存期（DFS）、病因特异性生存期（CSS）和局部控制（LC）相关。
在1987年1月1日至1992年12月31日期间，568例临床局限性前列腺癌患者在威廉·博蒙特医院接受了使用局限性前列腺野的外照射放疗（RT）（中位总剂量66.6 Gy；范围：60 - 70.4 Gy）。生化失败定义为达到最低点后治疗后前列腺特异性抗原（PSA）连续三次升高。生化失败记录为最低点与首次升高的PSA之间的中间时间。计算了生化控制（BC）患者与生化失败患者的临床DMFS, DFS, CSS和LC的5年精算率。中位随访时间为56个月（范围：24 - 118个月）。
生化控制的患者的DMFS、DFS、CSS和LC的5年精算率显著高于未控制的患者（p < 0.001）。在生化控制的患者中，DMFS、DFS、CSS和LC的5年精算率分别为99%、99%、98%和99%。对于生化失败的患者，DMFS、DFS、CSS和LC的5年精算率分别为74%、64%、89%和86%。按治疗前PSA、Gleason评分和T分期分层时，这些差异在DMFS、DFS和CSS中仍然显著。Cox比例风险模型表明，BC是DMFS、DFS、CSS和LC临床结局的最重要单一预测因素。治疗前PSA和Gleason评分也是DMFS和DFS结局的独立预测因素。
ASTRO共识小组对放疗后BF的定义与临床DMFS、DFS和CSS密切相关。这些发现表明，共识小组的定义可能是临床进展和生存的替代指标，应被视为区分治疗成功与失败的有效终点。需要进行更长时间随访的进一步研究来证实这些发现。
