| 摘要: | 背景
兒童癌症是罕見的疾病，由於兒童癌症的罕見，使得此類研究是很困難的。因此，以人口為基礎的兒童癌症研究來探討發生率，死亡率及存活率是重要的。本研究利用台灣健保資料庫，估計1996-2002年出生兒童白血病，淋巴癌和中樞神經系統及腦瘤的小兒癌症的發生率及5年存活率。
方法
以人口為基礎的回顧性世代研究，利用台灣健保資料庫定義1996-2002年出生的兒童世代。考慮的變項包括性別、出生年、父母的保險金額及地理位置等，利用卜瓦松回歸分析估算各個相關因素發生比。並使用K-M估計方法描繪這些出生世代發生白血病，淋巴癌和中樞神經系統及腦瘤的小兒癌症個案之存活曲線，及使用log-rank test比較存活情形。最後，利用Cox之比例危險迴歸分析 (Cox’s proportional hazard regression analysis) 來評估各因子的死亡風險。
結果
1996-2002年出生的兒童世代在2008年前共有1003個兒童白血病發生個案，260個兒童淋巴瘤發生個案，578個兒童中樞神經系統癌及腦瘤發生個案。發生率分別為4.81、 2.44和2.77(每100,000人年)。整體來看，白血病的死亡率為4.48 (每100人年)，淋巴癌死亡率為2.03 (每100人年)，中樞神經系統癌及腦瘤死亡率為7.43 (每100人年)；以白血病來看，白血病亞型中的急性淋巴性白血病(ALL)和急性骨髓性白血病(AML)的存活曲線有顯著的差異，顯示其死亡風險小於其他種類的白血病。在出生年的部分則是2002年出生的小孩死亡風險會顯著的高於1996年出生的小孩。地理位置則是中部、南部和高屏及東部地區有較高的死亡風險；中樞神經系統癌及腦瘤則是在2002年出生及發病年齡為5-9歲的組別有顯著較高的死亡率。
結論
本研究顯示小兒三大癌症分別為白血病、淋巴瘤及中樞神經系統癌及腦瘤，此三大癌症的死亡率則以中樞神經系統癌及腦瘤為最高，白血病次之，淋巴瘤最低。本研究結果可以提供醫務政策制定者了解兒童癌症在流行病學方面的訊息，以發展對於兒童癌症預防和控制上的政策和策略。
Background
The estimation of childhood cancer using Taiwan nationwide data has never been reported. The objective of the current study was to estimate the incidence rates and 5-year survival rates of top three childhood cancers followed up to the 2008: leukemia, lymphoma, and central nervous system (CNS) and brain cancer from birth cohorts of 1996-2002 in Taiwan.
Methods
We conducted a population-based retrospective cohort study consisting of 1996-2002 birth cohorts for Taiwan between 1996 and 2002 and followed it up to 2008. The datasets of the study were from Taiwan National Health Insurance Research Database (NHIRD). The childhood cancer cases were identified from datasets of severe illness registry. We calculated incidence rates and 5-year survival rates of leukemia, lymphoma, and CNS and brain tumor. We estimated incidence rate ratios (IRRs) to evaluate the independent effects of gender, birth year, parents’ insurance premium and geographic area using Poisson regression analysis. We also explored the 5-year survival rate to find out prognosis factors using Cox’s proportional hazard model.
Results
A total of 1003 leukemia patients, 260 lymphoma patients, and 578 CNS and brain tumor patients were identified, and the incidences of these corresponding three majority childhood cancers were 4.81 per 105 person year, 2.44 per 105 person year and 2.77 per 105 person year, respectively. The multivariable-adjusted incidence rate of lymphoma was significantly higher in boys than girls (incidence rate ratio (IRR) =1.59, 95% CI, 1.24 to 2.05). For leukemia, the incidence rate of the 2002 birth cohort was significantly higher than that of the 1996 birth cohort. For CNS and brain tumor, the incidence rate of 2005 birth cohort was significantly higher than that of the 1996 birth cohort and Southern area was associated with significantly higher risk than Taipei area.
The adjusted HRs for leukemia subtype of ALL and AML, birth year of 2002, geographic areas of central area, southern area, Kao-ping area & eastern area and aboriginal area were significant. After adjusting for the other variables in the lymphoma model, none of them was significant. For CNS & brain tumor, the HRs of 2002 birth year and age of onset at 5-9 years old were significantly.
Conclusions
This study is the first to demonstrate leukemia, lymphoma, and CNS & brain tumor substantial incidence and survival difference from childhood cancer. It also shows that incidence varies according to birth year, gender, parents’ insurance premium and geographic area, and shows that 5-year survival rates varies according to birth year, gender, parents’ insurance premium, cancer subtype, age of onset and geographic area. Our findings are useful for prioritizing future childhood cancer research needs. |
