人類乳突瘤病毒感染在台灣女性肺癌形成之角色研究

自民國七十五年開始，肺癌一直是台灣女性和男性癌症的第一和第二大癌症死亡原因。已知抽菸是導致肺癌的最重要原因，歐美國家的肺癌有90 % 以上可用抽菸來解釋。但是台灣地區則有一半的肺癌無法以抽菸來解釋，尤其台灣女性肺癌患者有90 % 以上是不抽菸者。由本研究室先前研究發現，HPV 16/18 感染和不抽菸台灣女性肺癌形成有關 (Cheng et al., 2001)。最近又發現，在肺癌病患肺腫瘤組織切片上的血球抹片上可測得HPV DNA。本研究擬探討台灣女性肺癌患者HPV的感染路徑，以及評估血液循環中之血球中HPV DNA存在，是否可做為罹患肺癌之危險指標。因此本研究選取149位肺癌患者及175位非癌症的成人，以巢疊式聚合酶連鎖反應(nested-PCR)分析血球中HPV的感染率，並比較同時具有組織檢體與血液檢體之70例肺癌病患，其HPV表現在肺腫瘤組織中與患者血液檢體中之相關性。我們發現肺癌患者血球中HPV 16/18感染率均高於非癌症者 (HPV 16, 47.7% vs 12.6%, P < 0.0001; HPV 18, 30.9% vs 5.2%, P < 0.0001)。HPV 16的感染頻率於女性肺癌患者高於男性肺癌患者(57.6% vs 41.1% , P=0.048)，第三期與第四期肺癌患者高於第一期與第二期患者（54.6% vs 29.3%, P=0.006）。由多變項統計分析的結果顯示，參與肺癌之危險因子中，在HPV 16或HPV 18單獨感染方面，感染者比未感染者分別有6.5倍 (P < 0.0001)及9.2倍(P < 0.0001)罹患肺癌之危險性，而同時感染HPV 16/18者比未感染者罹患肺癌之危險性高達75.7倍(P < 0.0001)。由此顯示血球中有HPV 16/18 DNA的感染可做為罹患肺癌的危險指標。 由先前的研究可推論HPV可能與女性肺癌的形成有關，但致癌的機轉仍不清楚，其他研究指出，p16INK4a 基因的去活化在肺癌中發生的頻率很高。p16INK4a基因的高度甲基化與抽菸有關。而台灣女性肺癌患者大多為不抽菸者，因此假設不抽菸之女性肺癌之p16 基因的去活化機制上，可能與男性肺癌患者不同。為了查驗環境暴露對p16INK4a 高度甲基化的影響，包括抽菸、環境致癌物的暴露和HPV感染，我們以methylation-specific PCR分析了162個肺癌患者腫瘤組織，包括67位抽菸男性肺癌，41位不抽菸男性肺癌和58位不抽菸女性肺癌。結果顯示在67位抽菸男性肺癌，59.7 ﹪（40/67位）可偵測到p16INK4a 高度甲基化，不抽菸男性肺癌36.6 ﹪（15/41位），不抽菸女性肺癌60.3 ﹪（35/58位）。這個結果似乎透露性別和抽菸對p16INK4a 高度甲基化有相同的影響。因此，我們推測HPV感染可能促進不抽菸女性的肺癌p16INK4a 高度甲基化。由我們的結果知道，不抽菸女性肺癌腫瘤有HPV感染，其p16INK4a高度甲基化的頻率高達70 ﹪（30/43位），相較於無HPV感染的僅33 ﹪（5/15位）。有HPV感染和p16INK4a 高度甲基化僅於不抽菸女性肺腫瘤有相關性（P＝0.017），於抽菸男性或不抽菸男性肺癌腫瘤則無相關性。更進一步分析，p16INK4a 免疫染色和p16INK4a 促進子高度甲基化的負相關性僅見於不抽菸女性肺癌。這些結果高度建議著HPV感染的台灣女性肺癌，HPV感染可能有部分經由促進子甲基化涉及p16INK4a基因的不活化。我們的結果可以提供環境-超越基因（environment-epigenetic）的相互作用在人類腫瘤發展因地域的變化扮演一個角色的證據。

Lung cancer is the leading cause of cancer death in Taiwan, and the paucity of dependable risk markers has impeded the early management of lung cancer. An association of human papillomavirus 16/18 infection with lung cancer among non-smoking Taiwanese women was revealed in our previous study. Nested PCR was employed to detect HPV 16/18 DNA in blood circulation of 149 lung cancer patients and 174 non-cancer controls. Additionally, correlation of prevalence of HPV DNA between the blood circulation and lung tumor tissue was compared from 70 sets of paired tumor tissues and peripheral blood samples available. The results showed that the prevalence rate of HPV 16/18 in the blood circulation of lung cancer cases was significantly higher than that of non-cancer controls (47.7% vs. 12.6% for HPV 16, P<0.0001; 30.9% vs. 5.2% for HPV 18, P<0.0001). A significantly higher HPV 16 prevalence was detected in female lung cancer patients than that of male (57.6% vs. 41.1%, P=0.048), as well as in cases with tumor stages III/IV than those with tumor stages I /II (54.6% vs. 29.3%, P=0.006). After adjusting the effects of age, gender, and smoking status, a 6.5-fold greater risk of lung cancer was demonstrated for those subjects with HPV type 16 positive (95% CI 3.7-11.3, P<0.0001), a 9.2-fold for HPV type 18 positive (95% CI 4.2-20.2, P<0.0001), and a 75.7-fold greatest risk for those with both HPV type 16 and 18 positive (95% CI 9.8-582.1, P<0.0001). These results suggested that the presence of HPV DNA in the blood circulation may serve as a feasible risk marker of lung cancer. Inactivation of p16 INK4a gene through promoter hypermethylation has been frequently observed in non small cell lung cancer; however, various studies have shown a controversial correlation between p16 INK4a hypermethylation and cigarette smoking. Our recent report showed that human papillomarvirus (HPV) 16/18 infections were associated with the development of nonsmoking female lung cancer in Taiwan and we further speculated that HPV infection may be linked with p16INK4a hypermethylation. To verify the influence of environmental exposure, including cigarette smoking, environmental carcinogen exposure and HPV infections on p16 INK4a hypermethylation, tumors from 162 lung patients, including 67 smoking males, 41 nonsmoking males and 58 nonsmoking females, were subjected to p16INK4a hypermethylation analysis by methylation-specific PCR. As the results showed, p16 INK4a hypermethylation was detected in 40 (59.7%) of 67 smoking male, 15 (36.6%) of 41 nonsmoking male and 35 (60.3%) of 58 nonsmoking female lung tumors. This result seemed to reveal that gender and cigarette smoking both possess an equal influence on p16INK4a hypermethylation. This result also led to a speculation that HPV infection may promote p16 INK4a hypermethylation in nonsmoking female lung cancer patients. From our data, p16 INK4a hypermethylation frequency in nonsmoking female lung tumors with HPV infection was as high as 70% (30 of 43) compared to those without HPV infection (33%; 5 of 15). In fact, the correlation between HPV infection and p16 INK4a hypermethylation was only observed in nonsmoking female lung tumors (p = 0.017), but not in smoking male or nonsmoking male lung tumors. Moreover, the reverse correlation between p16INK4a immunostaining and p16INK4a promoter hypermethylation was also only observed in nonsmoking female lung tumors. These results strongly suggested that the involvement of HPV infection in lung tumorigenesis of nonsmoking female cancer patients in Taiwan may be mediated at least in part through the increase of hypermethylation to cause p16 INK4a inactivation.