| 摘要: | 研究目的
肺癌是一種世界性的致命癌症。有許多的風險因子與肺癌有關，如：支氣管炎、氣喘、肺結核病、抽菸、空氣汙染...等。這些風險因子同時與一種被認為是良性的疾病有關，那就是自發性氣胸。而且，自發性氣胸也可以是肺癌最先出現的臨床症狀。自發性氣胸與肺癌間有何關聯？大膽假設，自發性氣胸患者有較高的風險罹患肺癌。本研究目的為探討其間的關聯及證實可能的假設。
研究方法
計畫使用全人口資料庫─台灣全民健保資料庫，來進行回溯性世代研究。這個資料庫涵蓋超過99%的台灣人口。在時間區間2000-2009年間，我們建立了一個27405人的自發性氣胸人口族群，及一個109620人的無氣胸對照族群來進行回溯性追蹤。氣胸首次診斷日為觀察起始日，觀察起始日後6個月為參考日，在參考日前有惡性腫瘤者將由研究族群排除。對照組以性別及年齡配對，以1:4比例由無氣胸患者群中依觀察起始日，隨機配對，同時必須符合排除條件。由回溯性世代研究來釐清氣胸與肺癌間的關係。劑量效應的研究由年發作氣胸次數來進行分層並計算相對風險。
研究結果
在描述性統計資料中，可以發現許多共病症在兩個族群間呈現有意義的差異，如高血壓、糖尿病、支氣管炎、氣喘、慢性阻塞性肺病，及肺結核病及肺氣腫。在多變項分析中，自發性氣胸患者族群確實有較高的風險罹患肺癌。整體而言，在校正了共病症、年齡及性別的影響後，相對風險(Hazard ratio)比率為2.09(95%信賴區間; 1.69-2.58)。各年齡層分析，在35-49歲族群有最高的相對風險。在有肺部共病症存在的情形下，氣胸的影響不具統計學上的差異。氣胸發生的頻率與肺癌發生的風險研究中，可以發現氣胸年發生率越高，罹患肺癌的相對風險就越高，具有劑量效應。年平均發作次數分為小於1次的輕度，1-2次的中度及大於2次的重度患者，可以發現風險逐步上升。相對風險 1.51、7.76及37.09，均具有統計學上的意義。
研究結論
自發性氣胸患者確實有較高的風險罹患肺癌，相對風險在35-49歲族群中是最高的。發作次數越多，未來罹患肺癌的風險就越高，在年平均發作次數大於2次的患者，相對風險超過30倍。當治療自發性氣胸患者時，如果年平均發作次數大於2次的病人，需要注意肺癌發生的風險。
Purpose: Lung cancer was a lethal malignancy world-widely. There were many risk factors such as bronchitis, asthma, tuberculosis, smoking, air pollution...etc. Those risk factors were also related with a benign disease, spontaneous pneumothorax, and it can be the first sign of lung cancer. It was interesting the relationship between spontaneous peumothorax and lung cancer. We hypothesized that patients with spontaneous pneumothorax have more risk to develop lung cancer in the future. We designed a population-based retrospective cohort study to evaluate the relationship between spontaneous pneumothorax and lung cancer.
Material and methods: We used the population-based Taiwan Health Insurance Research Database to perform the population-based retrospective cohort study. The database included more than 99% people of Taiwan. We established a 27405 pneumothorax cohort and a 109620 comparison cohort from 2000-2009 to evaluate the relationship between spontaneous pneumothorax and lung cancer. The whole population was separated to two groups, spontaneous pneumothorax and non-spontaneous pntumothorax people. The day of first pneumothorax episode was the starting observation day. There was index day which was 6 months after starting observation day. The patients of spontaneous pneumothorax would be excluded if they got any malignancy before the index day. Comparison group was matched by the relative starting observation day and 4 folds from non-pneumothorax patients randomly. The same excluded criteria was performed in comparison cohort. Multi-variant analysis would be performed to evaluate the relationship in different aspects. Dose effect was evaluated by stratified patients with average attacking number per year. The hazard ratios were calculated in different stratified and adjusted by age, gender, and those comorbidities.
Result: In demographic analysis, there was many comorbidities different between the two cohorts significantly such as: hypertension, diabetes mellitus, bronchitis, asthma, chronic obstructive disease, tuberculosis and emphysema. The multi-variant analysis showed that patients with spontaneous pneumothorax have more relative risk to develop lung cancer. The overall hazard ratio(HR) was 2.09(95% CI; 1.69-2.58) adjusted by age, gender, and those comorbidities. In age stratified, the highest relative risk was at the group of 35-49 years old. Over those patients with pulmonary comorbidities, the risk of pneumothorax was not significant. The dose effect was confirmed, the more frequency of spontaneous pneumothorax attack the more relative risk to develop lung cancer. Separated patient to 3 groups, less severity( average attacking number per year, less than 1), median severity( average attacking number per year, 1-2) and severe severity(average attacking number, greater than 2). Among the three groups, the hazard ratios were 1.51, 7.76 and 37.09 significantly.
Conclusion: Patients of spontaneous pneumothorax had higher risk to develop lung cancer, especially among those patients with age 35-49 years old. It was significant in multi-variant analysis after correct of gender, age, and some comobidity. It was found that the more frequency of spontaneous pneumothorax attack the higher relative risk to develop lung cancer. As the frequency more than 2 times per year, the relative risk to develop lung cancer would be more than 30 folds. As treating patients with spontaneous pneumothorax with the frequency more than 2 times per year, the risk to develop lung cancer should be considered. |
