某北部醫學中心之表淺膀胱癌病患影響預後因子分析

Abstract

目的:我們利用北台灣某醫學中心的病患資料來分析原發性表淺膀胱癌病患的預後因子。 材料與方法:本研究採回溯性世代研究法，收錄115位1998年1月至2002年12月在國泰醫院就診之原發性表淺膀胱癌的病患。其中103位接受經尿道膀胱腫瘤切除術，七位接受部分膀胱切除手術，五位接受全膀胱切除手術。105位病患接受術後膀胱灌藥治療(包括bacille Calmette-Guérin、adriamycin或mitomycin C)。病患術後以膀胱鏡檢查追蹤，第一年每三個月追蹤一次，第二年每六個月追蹤一次，第三年起每年追蹤一次。我們使用Kaplan-Meier存活分析統計不同變相對於預後影響的差異性，並以Log-rank test檢定變相的差別是否顯著。並進一步擷取單變相分析中較顯著之變相來進行多變相Cox proportional-hazard回歸分析。 結果: 本研究病患平均年齡為69.5歲，平均追蹤期為61.2個月。五及十年無癌症復發比率分別為36.7% 和32.0%。五及十年癌症存活率分別為70.2% 和60.9%。根據單變相分析，和癌症復發有統計顯著之變相有腫瘤外型、腫瘤多發性、病理分期、腫瘤分化、肌酸酐高低、腫瘤大小及血紅素高低；和癌症存活有統計顯著之變相有年齡、腫瘤外型、腫瘤大小、病理分期、腫瘤分化、血紅素高低、肝功能及術後住院日數。我們進一步選擇單變相分析較顯著之因子進行多變相回歸分析後發現，最主要影響癌症復發的變相為腫瘤大小，其次為腫瘤外型及腫瘤多發性。另外,多變相回歸分析發現，最主要影響癌症存活的變相為腫瘤外型，其次為肝功能及腫瘤大小。 結論: 非乳突狀腫瘤及最大腫瘤大於或等於二公分皆為癌症復發及癌症死亡之危險因子。多發性腫瘤為癌症復發之危險因子。異常肝功能會影響癌症存活，此一新發現可能會對於評估台灣之原發表淺膀胱癌病患的預後有幫助。

Objectives: We analyzed the prognostic factors of recurrence and cancer-specific survival in patients with primary superficial bladder cancer. The present data are collected from a medical center in northern Taiwan. Material and Methods: This retrospective cohort study collected 115 newly diagnosed patients with primary superficial bladder cancers at the Cathay General Hospital from January 1998 to December 2002. Most (103, 89.6%) patients underwent transurethral resection of bladder tumors. 7 and 5 patients underwent partial and radical cystectomy, respectively. 105 patients received bladder instillation postoperatively with agents including bacille Calmette-Gu&eacute;rin (BCG), adriamycin, or mitomycin C. Patients were followed by cystoscopy every three months during the first year, every six months during the second year, and every twelve months subsequently. Kaplan-Meier survival analysis was used in order to evaluate the influence of different variables on prognosis. Log-rank test was utilized to verify significance of variables in this survival analysis. Multivariate Cox proportional-hazard regression analysis was performed to analyze the significant variables at 0.15 level from the univariate analyses. Results: In our study, the average age of patients was 69.5 years and the average follow-up period was 61.2 months. Recurrence-free rates in patients after five and ten years were 36.7 % and 32.0 %, respectively. Cancer-specific survival rates after five and ten years were 70.2 % and 60.9 %, respectively. Based on univariate analysis, recurrence of urinary bladder cancer was statistically significant for gross tumor shape, tumor multifocality, pathological tumor stage, pathological tumor grade, creatinine level, tumor size, and hemoglobin level; cancer-specific survival of urinary bladder cancer was statistically significant for age, gross tumor shape, tumor size, pathological tumor stage, pathologic tumor grade, hemoglobin level, liver function test, and postoperative hospital stay. After the univariate analysis, we selected above-mentioned significant factors to conduct multivariate Cox proportional-hazard regression model. The analysis showed that tumor size is the most significant determinant for cancer recurrence, followed by gross tumor shape and tumor multifocality. This model composed of three determinants yielded hazard ratios of: 3.2 for tumor size (≧2cm versus <2cm), 2.7 for gross tumor shape (non-papillary versus papillary), and 2.3 for tumor multifocality (unifocality versus multifocality). In addition, the multivariate analysis showed that gross tumor shape is the most significant determinant for cancer-specific survival, followed by liver function test and tumor size. This best model composed of three determinants yielded hazard ratios of: 6.8 for gross tumor shape (non-papillary versus papillary), 3.2 for liver function test (abnormal versus normal), and 2.8 for tumor size (≧2cm versus <2cm). Conclusions: Non-papillary tumor and the largest tumor greater than or equal to 2 cm were risk factors of recurrence and cancer-specific mortality. Tumor multifocality was a risk factor of recurrence. Abnormal liver function test influenced cancer-specific survival in our cohort patients; this new finding may be helpful for predicting the prognosis of primary superficial bladder cancer in Taiwanese patients.