Clinical Characteristics, Image Findings, and Treatment Modalities Associate with Outcomes of Patients with Glioblastoma Multiforme

臨床表現、影像特徵及治療方式對多形性神經膠芽細胞瘤病人之預後影響

梁祥光(Hsiang-Kuang Liang) ； 黃昭源(Chao-Yuan Huang) ； 王駿瑋(Chun-Wei Wang) ； 陳育瑄(Yu-Hsuan Chen) ； 郭頌鑫(Sung-Hsin Kuo) ； 成佳憲(Jason Chia-Hsien Cheng)

《放射治療與腫瘤學》 19卷2期 (2012/06) Pp. 93-104

https://doi.org/10.6316/TRO/201219(2)93

摘要

目的：在此研究中，我們評估臨床表現、影像特徵及治療方式是否影響多形性神經膠芽細胞瘤病人的臨床預後。材料與方法：從2001年3月至2008年10月，我們評估114位在單一醫院經組織學診斷為多形性神經膠芽細胞瘤的病人。所有的病人都接受腫瘤全切除或局部切除。多數病人則接受術後放射治療或併用temozolomide。病人之臨床表現、影像特徵、治療方式與整體存活（overall survival, OS）都被檢視與分析。結果：共35位病人接受腫瘤全切除或局部切除與術後放射治療併同步及輔助藥物temozolomide，55位病人接受腫瘤切除與術後放射治療，而24位病人僅接受腫瘤切除。在平均中位數追蹤時間40.6個月後（範圍從0.3至61.8個月），病人的中位數整體存活時間為12.5個月，而l年及2年的整體存活率則分別為50%與17.5%。接受手術切除與術後放射治療併temozolomide使用的病人，比起接受腫瘤切除與術後放射治療的病人，有較好的中位數整體存活時間（分別為19.4及10.1個月，p = 0.01）。在單變數分析中，Karnofsky氏體能表現狀態不佳者（Kamofsky performance status, KPS）（≤ 70，p = 0.036）、多重神經學障礙（p = 0.044）、腦部多發病灶或胼胝體侵犯（p = 0.003）、腫瘤局部切除（p = 0.001），未接受放射治療（p＜0.001）或未使用temozolomide（p = 0.001）為預測病人整體存活不良的重要預後因子。在多變數分析中，多重神經學障礙（p = 0.008）、腦部多發病灶或胼胝體侵犯（p = 0.007）、腫瘤局部切除（p = 0.005），未接受放射治療（p＜0 .001）或未使用temozolomide（p = 0.007）仍為預測病人整體存活不良的重要預後因子。結論：我們的研究結果顯示：多重神經學障礙、影像特徵，尤其是腦部多發病灶或胼胝體侵犯以及非整合性治療為預測整體存活不良的重要預後因子。

關鍵字

多形性神經膠芽細胞瘤 ； 預後 ； 治療結果 ； 台灣

並列摘要

Purpose: In this study, we evaluated the clinical outcomes of glioblastoma multilorme (GBM) patients with their prognostic factors.Materials and Method: From March 2001 through October 2008, we evaluated 114 consecutive patients with histologically confirmed GBM in a single institution. Most patients received tumor resection followed by radiotherapy (RT) with or without temozolomide use, while some received tumor resection alone. Clinical characteristics, image findings, treatment modalities, and overall survival (OS) of these patients were reviewed and analyzed.Results: Thirty-live patients were treated with total or subtotal tumor resection followed by RT with concurrent and adjuvant temozolomide, 55 treated with tumor resection and adjuvant RT, and 24 treated with resection alone. At a median follow-up time of 40.6 months (ranging from 0.3 to 61 .8 months), the median OS for all the patients was 12.5 months, and the one-year and two-year OS rates were 50% and 17.5%, respectively. Patients receiving combined tumor resection with RT and temozolomide had a better median OS than those without temozolomide used (19.4 vs. 10.1 months, p = 0.01). In univariate analysis, poor Karnofsky performance status (KPS) (≤ 70, P = 0.036), multiple neurologic deficits (p = 0.044), multiple brain lesions or corpus callosum invasion (p = 0.003) , subtotal resection (p = 0.001) , no RT (p ＜ 0.001) , and no temozolomide (p = 0.001) were significant prognostic factors for predicting poor OS of these patients. In multivariate analysis, multiple neurologic deficits (p = 0.008), multiple brain lesions or corpus callosum invasion (p = 0.007), subtotal resection (p = 0.005), no RT (p ＜ 0.001), and no temozolomide (p = 0.007) remained independent worse prognostic factors for OS 01 these patients.Conclusion: Our results indicated that multiple neurologic deficits, image findings of corpus callosum invasion or multiple brain lesions, and single modality treatment were important factors for predicting poor prognosis.

並列關鍵字

Glioblastoma multilorme ； Prognosis ； Treatment outcome ； Taiwan