Takashi M, Wakai K, Hattori T, Ono Y, Ohshima S
Department of Urology, Hekinan Municipal Hospital, Aichi, Japan.
Int Urol Nephrol. 2002;34(3):329-34. doi: 10.1023/a:1024431519652.
To evaluate factors affecting recurrence after intravesical bacillus Calmette-Guérin (BCG) therapy (Tokyo 172 strain), we reviewed data for 101 patients with superficial bladder cancer (pTa [n = 80] and pT1 [n = 21]) treated between 1985 and 1999. The median follow-up period was 58.9 months. Factors affecting the first tumour recurrence were evaluated using Cox's proportional hazards model and those affecting multiple recurrence with Andersen-Gill's model. The 5-year recurrence-free rate was 63% for all 101 patients. The recurrence frequency, defined as times per 100 patient-months of follow-up, greatly decreased from 7.3 +/- 9.6 (SD) before the instillation to 2.6 +/- 5.6 after the therapy (p < 0.0001). Patients with pT1 tumours tended to have earlier recurrence than those with pTa tumours (p = 0.06). Multivariate analysis using Cox's proportional hazards model revealed that a history of bladder cancer and pathological stage were independent factors affecting the first tumour recurrence after the BCG therapy. When multiple endpoints of recurrence were evaluated using the Andersen-Gill's model, number of tumours as well as a history of bladder cancer and pathological stage demonstrated significant links to tumour recurrence after the BCG therapy. The 5-year progression-free and 5-year survival rates were 89.3% and 85.3% for all the 101 patients, respectively. Because intravesical recurrence may involve multiple events during the clinical course of patients with bladder cancer, the Andersen-Gill's model appears useful for evaluation of risk factors.
为评估影响膀胱内卡介苗(BCG,东京172株)治疗后复发的因素,我们回顾了1985年至1999年间接受治疗的101例浅表性膀胱癌患者(pTa [n = 80]和pT1 [n = 21])的数据。中位随访期为58.9个月。使用Cox比例风险模型评估影响首次肿瘤复发的因素,使用Andersen-Gill模型评估影响多次复发的因素。101例患者的5年无复发生存率为63%。复发频率定义为每100患者-月随访时间内的复发次数,从灌注前的7.3±9.6(标准差)大幅降至治疗后的2.6±5.6(p < 0.0001)。pT1肿瘤患者的复发往往比pTa肿瘤患者更早(p = 0.06)。使用Cox比例风险模型进行多变量分析显示,膀胱癌病史和病理分期是影响BCG治疗后首次肿瘤复发的独立因素。当使用Andersen-Gill模型评估多个复发终点时,肿瘤数量以及膀胱癌病史和病理分期均显示与BCG治疗后的肿瘤复发有显著关联。101例患者的5年无进展生存率和5年生存率分别为89.3%和85.3%。由于膀胱内复发可能在膀胱癌患者的临床过程中涉及多个事件,因此Andersen-Gill模型似乎有助于评估风险因素。