- 學位論文
選擇型COX-2抑制劑在食道癌化學預防的角色
Role of Chemoprevention by Selective Cyclooxygenase-2 Inhibitors in Esophageal Cancer Patients-A Population Base Study
摘要
背景：根據衛生署統計資料顯示，台灣地區食道癌的盛行率由民國84年的每十萬人中3.3人，增加至民國97年的每十萬人中8.6人，其中男性的盛行率由民國84年的每十萬人中5.9人，增加至民國97年的每十萬人中15.9人。近來國外研究顯示，長期使用阿司匹靈（aspirin）與非類固醇抗發炎藥物（NSAID），尤其是選擇型COX-2抑制劑（selective COX-2 inhibitor），可以有癌症化學預防的效果。國內目前針對這方面的食道癌罹患率影響之相關臨床研究較少。 研究目的：利用台灣全民健康保險資料庫的資料分析，來探討使用選擇型COX-2抑制劑與食道癌發生率的關係。 研究方法：研究設計為配對病例對照（matched case-control）實驗。利用全民健康保險資料庫，建構病例組及對照組；使用的資料庫包含健保資料庫特殊需求檔案（1997至2007）癌症患者追蹤的資料，及涵蓋1996至2006年之2005年承保抽樣歸人檔（Longitudinal Health Insurance Database 2005, LHID 2005）。統計分析以條件式邏輯斯廻歸模式（conditional logistic models），以病例組受試者的第一次確診的日期當作指標日（index date），對照組的指標日則與配對病例組受試者相同。主要分析變項是NSAID類藥物中的選擇型COX-2抑制劑的累計定義每日劑量（define daily dosage, DDD）、藥物持有率（medication possession ratio, MPR）其他NSAIDs類的藥物則以共變項（covariates）納入統計分析中；藥物使用累計的追蹤期次分為5個時間點，分別為5年、3年、1年、6個月、及三個月；選擇型COX-2抑制劑的MPRs則以10%進行切分。 結果：分析資料中共有2872位食道癌患者與28720位對照者，在index dates前五年，曾開立一次以上選擇型COX-2抑制劑處方籤的比率（p=0.1536）及平均DDD（p=0.3397）兩組並未達顯著性差異。選擇型COX-2抑制劑對不同追蹤期罹患食道癌的危險比在0.55至0.68之間，均未達統計顯著。其中，在老年族群，在追蹤期1年及三個月的呈現顯著較低的危險比分別為0.30（95% CI=0.09~0.99, p=0.0480）及0.29（95% CI=0.10~0.87, p=0.0269）。除了1年追蹤期MPR小於10%所估算出的危險比0.74（95% CI=0.56~1.00, p=0.0470）達統計顯著意義外，其他追蹤期及MPR切點間，危險比均未達統計上顯著性差異。次族群分析則顯示在老年次族群（65歲及大於65歲以上），在1年、3個月追蹤期，MPR為≥10%及≥20%時及6個月MPR為≥10%，對罹患食道癌具統計上顯著較低的危險比作用。 結論：全民健康保險資料庫分析後發現，選擇型COX-2抑制劑使用罹患食道危險機率，在不同追蹤期雖然勝算均小於1（保護作用），但因未達統計顯著，所以需更多資料來進一步分析其效益。次族群分析顯示選擇型COX-2抑制劑對老年族群（65歲及大於65歲以上）罹患食道癌的保護作用較年輕族群（小於65歲）為佳。
並列摘要
Background: In Taiwan, the incidence rate of esophageal cancer increased from 3.3/100,000 in 1995 to 8.6/100,000 in 2008. In the male population, the incidence rate of esophageal cancer increased from 5.9/100,000 in 1995 to 15.9/100,000 in 2008. Recent epidemiological studies have demonstrated long-term intake of aspirin and non-steroidal anti-inflammatory drugs (NSAID) could significantly reduce the risk for cancers incidences (e.g., colorectal). Domestic clinical studies on the effect of using selective COX-2 as chemoprevention on esophageal cancer incidences are limited and remain to be explored. Study objective: The effect of selective COX-2 inhibitors intake on esophageal cancer incidence was investigated using the Taiwan Health Insurance Research Database. Methods: Study design was a matched case-control study. The case and control groups were constructed based on two sets of data from the Taiwan Health Insurance Research Database (NHIRD). The databases used were the 11- year Longitudinal Database (1997-2007) of cancer patients and the Longitudinal Health Insurance Database 2005 (LHID2005, 1996-2006) from the National Health Research Institutes (NHRI) in Taiwan. Conditional logistic models were used for statistical analyses; the index date was the date of first diagnosis for the cancer patients and the same index dates were assigned to their matched controls. The major dependent variables were defined daily dosage (DDD) and medication possession ratio (MPR) of selective COX-2 inhibitors, and the usage of remaining NSAIDs was computed as covariates in the statistical analyses. Five follow-up periods for medications cumulatively used were 5-year, 3-year, 1-year, 6-month, and 3-month and the MPRs of selective COX-2 inhibitors were further categorized by every 10% as cut-off points. Results: A total of 2872 patients with esophageal cancer and 28720 controls were used in this study. In the prior 5 years of the index dates, the proportions of having at least one prescription for selective COX-2 inhibitors (p=0.1536) and DDD (p=0.3397) were similar between the two groups. The odds ratios of selective COX-2 inhibitor usages were similar among different follow-up periods, ranging from 0.55 to 0.68 with no statistical significances. Significantly lower ORs were seen in an older age group (age ≥ 65 yrs) at 1-year and 3-month follow-up periods with ORs of 0.3 (95%CI=0.09~0.99, p=0.0480) and 0.29 (95%CI=0.10~0.87, p=0.0269), respectively. Risk reduction at various MPR cut-off points at different follow-up periods showed no statistical significance expect for MPR<10% at 1-year follow-up (OR=0.74, 95%CI=0.56~1.00, p=0.0470). In subgroup analyses, selective COX-2 inhibitor usages for an older age group at MPR of 10% and 20% had significantly protective effects (all ORs<1 with p<0.05). Conclusion: Using Taiwan Health Insurance Database found non-significantly smaller odds ratios (all ORs<1) for selective COX-2 inhibitor usages on esophageal cancer incidence, and more data were needed to further confirm this beneficial effect. The results of the subgroup analyses showed that the protective effect of selective COX-2 inhibitor usages were more prominent in people aged ≥ 65 years compared to those < 65 years.