- 學位論文
以甲苯胺藍試劑篩檢口腔癌前病變之社區型臨床隨機實驗
Community-Based Screening for Oral Premalignant Lesions using Toluidine Blue dye staining – A Randomized Controlled Trial
摘要
[背景]： 依口腔癌之自然病史可知於惡性轉化成臨床症狀期之前，有相當長期無臨床症狀之「口腔黏膜癌前病變」存在；因此可以經由提供更加有效之篩檢工具，及時偵測出來予以適當介入，進而達到降低口腔癌發生率之終極目標。 [研究目的]： 本研究希望藉由甲苯胺藍試劑使口腔黏膜予以染色之輔助，證實能較目視篩檢偵測出更多之口腔病變。並分析利用此方法從事大規模篩檢是否可以偵測出更多之口腔癌前病變，進而降低口腔癌發生率並透過實證資料評估其成本效益。 [研究設計與方法]： 針對參加基隆市「闔家歡複合式篩檢」年齡大於15歲有「吸菸」或「嚼食檳榔」習慣者之民眾，排除先前已曾經被診斷、曾接受過口腔癌或癌前病變治療、拒絕/無法配合染劑漱口者。曾對甲苯胺藍試劑過敏與患有蠶豆症的個案亦拒絕其加入本研究。經由隨機化分配成兩組－「實驗組」經甲苯胺藍試劑漱口後由專業口腔外科醫師以目視篩檢口腔黏膜；「控制組」只以顏色、氣味相同之「安慰劑」染料漱口後，由不知隨機化分配組別之相同專業口腔外科醫師以目視篩檢口腔黏膜。兩組個案若篩檢結果為陽性則轉介至醫學中心接受複檢，以獲得確定診斷與安排口腔黏膜切片，得到病理組織報告。追蹤個案存活狀況及口腔癌發生至2004/12/31。 [研究結果]： 本研究共計8,101個案同意參與，經扣除資料不全及重覆篩檢者，共計剩餘7,975個案納入分析；實驗組共4,080人（佔51.16%），對照組共3,895人（佔48.84%），兩組於年齡分佈相當。篩檢陽性個案共有711例（8.9%），其中口腔黏膜癌前病變於實驗組共187 例（陽性偵測率4.58%），對照組共170 例（陽性偵測率4.36%）；兩組於口腔黏膜癌前病變之偵測率無統計上之差異（p值=0.64）。進一步分析發現實驗組（41/320, 偵測率12.81%）可比對照組（22/293, 偵測率7.51%）偵測更高比例之口腔黏膜下纖維化（p值=0.03）。對照組/實驗組之惡性轉移相對危險性比率為3.25（95%信賴區間為0.34~31.23）。實驗組比對照組口腔癌之發生率比率為0.80（95%信賴區間0.25~2.64）。以「甲苯胺藍試劑」輔助篩檢口腔黏膜癌前病變可以有效降低20%之口腔癌發生率。依隨機模式估計可得口腔黏膜下纖維化、均質性白斑及紅斑的發生率以均質性白斑約每千人一例最高。由均質性白斑進展至非均質性白斑平均約8.3年，由非均質性白斑進展至臨床症狀前期口腔癌平均約9.7年，由口腔黏膜下纖維化進展至臨床症狀前期口腔癌平均約21.5年，由臨床症狀前期口腔癌進展至臨床症狀期口腔癌平均約2個月。從本研究篩檢所得「口腔癌前病變」結果與電腦模擬估計追蹤十年後「口腔癌」發生個案數分別為實驗組10.67位與對照組16.04位，可推得若欲多減少一位口腔癌個案發生，需要多花費新台幣201,015元；每多拯救一個人年，平均需多支出新台幣8,740元。 [結論]： 本研究為首篇利用甲苯胺藍試劑篩檢口腔黏膜癌前病變之隨機試驗。雖然實驗組與對照組在整體在口腔黏膜癌前病變之偵測幾乎無差異，但實驗組比對照組有更高之口腔病變陽性偵測率與可偵測更高比例之口腔黏膜下纖維化。透過成本效益評估，此方法為符合成本效益之大規模篩檢輔助工具。
並列摘要
[Background] Oral squamous cell carcinoma is now the seventh highest type of malignancy in Taiwan. Incidence of oral cancer has continued to increase parallel to the consumption of betel quid since 1980s and with a similar but lesser extent of the consumption of tobacco. Long preclinical detectable phase in oral permalignant lesions (OPL), including oral submucous fibrosis, homogenous leukoplakia, non-homogenous leukoplakia, and erythroplakia, almost preceded the advent of clinical manifest oral cancer. Consequently, with proper screening, these OPLs could be detected early and proper intervention given, hence in hope of reducing incidence of oral cancer. [Study Purpose] (1) To assess whether oral screening with visual inspection with a supplement of Toulidine blue vital dye staining targeting at high risk group (betel quid chewing or smoking) can detect more oral premalignant lesions (2) To quantify the disease natural history for oral cancer and performed a simulation for the cost-effectiveness analysis [Designs and Methods] A community-based randomized controlled trial was designed and conducted in Keelung City in northern Taiwan during 1999. Subjects aged no less than 15 years old with habit of smoking or betel quids chewing were enrolled and randomly assigned into two arms. The experiment group was screened with Toluidine blue dye oral gargling following by visual inspection by experienced oral surgeons. The control group had only placebo dye gargling following by visual inspection by the same blinded oral surgeons. If suspicious oral lesions were identified, referral to tertiary referral medical center for clinical diagnosis and biopsy was arranged. The pathologic report was used as the gold standard for definite diagnosis. Linkage to death registry databank until December 31, 2004 was made to acquire data on occurrence of oral cancer, status of survival and cause of death. Analysis was done on an intention to treat basis. [Results] Of the 8,101 eligible participants recruited in this study, one hundred and twenty-six participants were excluded in the analysis owing to incomplete data collection. A total of 4,080 (51.16%) individuals were included in the experiment group, and 3,895 (48.84%) individuals in the control group. Three hundred and eighty-nine subjects (9.53%) in the experiment group yielded positive screening result while 322 subjects (8.27%) in the control group. After excluding false positive subjects, the positive OPL detection rate was not statistical significant (p=0.64) between the experiment group (4.58%, 187/4,080) and the control group (4.36%, 170/3,895). In subgroup analysis, the detection rate ratio of submucous fibrosis was 1.81 (95%CI 1.05~3.12) between the experiment (41/4,080, 1%) and the control (22/389, 0.56%) group. The relative malignant transformation rate or OPL was 1.29×10-3 in the experiment group and 4.20×10-3 in the control group. The ratio between two groups was 3.25 (95%CI 0.34~31.23) which mean the experiment group tended to detect OPL with higher malignant transformation potential. As for the occurrence oral cancer, it was 11.3×10-5 in the experiment group and 18.45×10-5 in the control group. [Conclusions] The current study is the first one to demonstrate that under the aid of Toluidine blue dye vital staining, mass screening for the high risk subjects could not detect statistically significantly more OPL than screening with visual inspection. The experiment group (41/4,080, 1%) tended to has higher detection rate on oral submucous fibrosis than the control group (22/3,895, 0.56%, p=0.03). In terms of economics, it was a convenient, safe and cost-effective adjunct in mass screening for oral premalignant lesions. Further research with larger cohort or longer follow-up period should be considered to delineate the results more in depth.