頭頸癌病患之健康生活品質研究

The Investigation of Health Related Quality of Life for Head-and-Neck Cancer Patients

方富民(Fu-Min Fang)

指導教授 ： 郭文烈 ； 邱亨嘉

高雄醫學大學/醫學院/醫學研究所/博士(2003年)

摘要

健康生活品質研究是目前頭頸癌研究之重要課題之一。探討頭頸癌病患之健康生活品質可幫助了解病患之生理、心理、社會等功能及相關症狀之自我表現外，臨床上並可當作治療結果之參考標準及治療後病患存活之預測因子。放射治療﹝簡稱放療﹞為治療頭頸癌之主要方法之一，其副作用常造成頭頸部器官之不可逆破壞。本研究目的為探討(1)放射副作用如何影響頭頸癌病患之健康生活品質，其影響模式為何；(2)放療期間健康生活品質對頭頸癌病患存活之預測模式為何。利用中文版SF-36量表，以及中文版EORTC-QLQ-C30與QLQ-H&N35量表，依臨床上頭頸癌病患接受放療之不同特性，分為三個研究子題。子題一，為一橫向性設計，以鼻咽癌存活者為例，探討放射副作用對健康生活品質之影響模式。子題二，為一縱貫性設計，以口腔咽喉癌接受手術後放療為例，探討放射副作用對健康生活品質之影響模式。子題三，為一前瞻性設計，以晚期頭頸癌接受放療為例，建立健康生活品質當作病患存活之預測模式。研究結果如下：
一、 不同之放射副作用對頭頸癌病患之健康生活品質有不同程度之影響。(1).以鼻咽癌存活者為例，經線性迴歸分析後發現影響整體生理功能之主要因子為：教育程度﹝迴歸係數(β)=0.247, p=0.000﹞、其他慢性病數目﹝β=-0.247, p=0.000﹞、唾液分泌功能﹝β=-0.151, p=0.026﹞、聽力功能﹝β=-0.157, p=0.015﹞、吞嚥功能﹝β=-0.140, p=0.029﹞；此模式之解釋程度R2值為0.51。而影響整體心理功能統計上有意義之因子只有經濟程度﹝β=0.235, p=0.001﹞；各種頭頸部功能障礙中只有頸部僵硬程度﹝β=-0.137, p=0.059﹞、發聲功能﹝β=-0.152, p=0.056﹞二者統計上接近有意義；此模式之解釋程度R2值為0.25。(2).以頭頸癌病患接受手術後放療為例，經Logistic迴歸分析後，我們發現臨床變項中，癌症分期是影響整體生活品質改變最有意義之變項，第三期病患比第四期病患於放療一年後整體生活品質變差之相對危險率(RR)為3.85﹝95%信賴區間( C.I.): 1.08-13.68, p=0.04﹞；而QLQ-H&N35頭頸部症狀變項中，吞嚥問題改變、言語問題改變是影響整體生活品質改變最有意義之變項。 吞嚥問題變差者其整體生活品質變差之RR為3.78﹝95% C.I.: 1.06-13.51, p=0.02﹞；言語問題變差者其整體生活品質變差之RR為5.07﹝95% C.I.: 1.36-18.98, p=0.0001﹞；此影響模式之解釋程度(sensitivity)為83.2%。
二、 放療前之健康生活品質為預測晚期頭頸癌病患存活之獨立因子，尤其是疲累分數。經Cox model分析發現，將非生活品質變項預測模式加入疲累分數後，其預測能力(Log Likelihood)由23.0增加至31.9﹝p<0.001﹞。疲累分數每增加10分，存活將減少約17%﹝95% C.I.: 8%-27%, p<0.001﹞。
本研究有系統建立放射副作用對頭頸癌病患健康生活品質之影響模式，有其臨床價值。其次，我們發現放療前病患疲累分數之高低是影響存活之重要因子，此意謂臨床醫師在實施放療前可參考病患疲累分數，設計出適合個別病人之治療策略。

關鍵字

頭頸癌 ； 健康生活品質 ； 生活品質 ； 放射治療 ； 放射副作用 ； 預測模式

並列摘要

Health-related quality of life (HRQL) has been one of the most important issues in recent clinical trials for head-and-neck cancer (HNC) patients. The investigation of HRQL could assist clinicians explore the patients’ self-reported physical and psychosocial functions and symptom problems, and compare potential therapeutic difference between rival treatment modalities. Besides, the HRQL levels could also be used as a survival predictor. Radiotherapy (RT) is one of the main treatment methods for HNC patients. The irreversible complications of RT usually cause organs at the head-and-neck area permanently dysfunction. The purposes of this research were to investigate how the RT complications affect the HRQL for HNC patients and whether HRQL could be used as a survival predictor for HNC patients treated with RT. In this research, we used the instruments of Taiwan Chinese versions of Short Form-36 (SF-36), European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30) and the Head and Neck Cancer Module (QLQ-H&N35) to investigate the HRQL issues. Three topics were designed based on the research endpoints and the patient groups with different RT intentions. The first was a cross-sectional design to establish the model of the effect of RT complications on HRQL for nasopharyngeal cancer (NPC) survivors. The second was longitudinally to establish the model of the effect of RT complications on HRQL for HNC patients receiving postoperative RT. The third was to test the survival predictability of HRQL for advanced HNC patients treated with RT. The results were as follows:
First, we confirmed that RT complications had different effects on HRQL for HNC patients. Take NPC survivors for example, analyzed by the linear regression model, we found the education level (regression coefficient:β=0.247, p=0.000), comorbidity number (β=-0.247, p=0.000), salivary dysfunction (β=-0.151, p=0.026), hearing dysfunction (β=-0.157, p=0.015), and swallowing dysfunction (β=-0.140, p=0.029) were the independent factors significantly correlated with patients’ physical component score (PCS) detected by SF-36 (R2 of the model: 0.51). On the other hand, economic level (β=0.235, p=0.001) was the only factor significantly correlated with mental component score (MCS). Among the head-and-neck functional impairments, only the neck stiffness(β=-0.137, p=0.059) and speech dysfunction (β=-0.152, p=0.056) approach marginal significance correlated with MCS (R2 of the model: 0.25).
Second, in the longitudinal assessment of HNC patients treated with postoperative RT, we found cancer stage was the clinical variable significantly correlated with the change of overall quality of life detected by EORTC QLQ-C30. The relative risk (RR) to have worse change of overall quality of life for patients with stage IV versus stage III was 3.85 (95% confidence interval (CI): 1.08-13.68, p=0.04). Among the 13 symptom scales of EORTC QLQ-H&N35, the change of swallowing problems (RR:3.78, 95% CI:1.06-13.51, p=0.02) and speech problems (RR:5.07, 95% CI:1.36-18.98, p=0.0001) were the variables most significantly correlated with the change of overall quality of life after postoperative RT (sensitivity of the model: 83.2%).
Third, the fatigue level before RT was an independent survival predictor for advanced HNC patients. The log likelihood significantly increased from 23.0 to 31.9 after adding the fatigue variable to the non-QoL model. An increase in fatigue score of 10 points corresponded to a 17% reduction of survival in the likelihood of survival (95% CI:8%-27%, p<0.001).
In conclusion, this research was the first approach to systemically establish the models of RT complications on HRQL for HNC patients. These models could provide useful reference in clinical practice. Also, we found the pre-treatment fatigue score was an independent survival predictor, indicating this parameter could be used to adjust the modality in treating the individual HNC patients.

並列關鍵字

Head and neck cancer ； Health related quality of life ； Quality of life ； Radiotherapy ； Radiation complication ； Predictive model

參考文獻

1. 劉文良，盧瑞芬，吳淑瓊. 長期照護需求量表的建立─SF-36中文版，行政院衛生署八十七年度科技發展計劃研究報告.

2. Allison PJ. Health-related quality of life comparisons in French and English-speaking populations. Community Dental Health 2001; 18(4): 214-8.

3. Allison PJ. Alcohol consumption is associated with improved health-related quality of life in head and neck cancer patients. Oral Oncol 2002; 38(1):81-6.

4. American Cancer Society Workshop Conference on Methodology in Behavioral and Psychosocial Research. Cancer 1984; 53 (Suppl): 2217-384.

5. Annual report 1999, Formosa Cancer Foundation.