鼻咽癌患者經放射治療後缺血性腦中風的預防與預測

Abstract

背景：在台灣，鼻咽癌是第二常見的頭頸癌，佔所有癌症病例的5%以及頭頸癌的50%。放射治療是治療鼻咽癌最有效的方法，然而放射線治療照射部位包含顱底與頸部可能導致頸動脈狹窄，進而增加缺血性中風的風險。在現代，結合強度調控放射治療與全身化學治療，非遠端轉移的末期鼻咽癌患者的預後極佳。隨著長期存活者人數增加，預測與預防晚期放射治療相關併發症的重要性日益提升。 研究目的與方法：第一個研究目標（研究一）是評估在接受標準同步化學放射治療之晚期鼻咽癌患者中，使用他汀類藥物對缺血性中風風險的影響。在研究一中，我們使用了台灣癌症登記資料庫（TCRD）與台灣全民健康保險研究資料庫（NHIRD）連結的全國性資料。我們採用逆機率加權（IPTW）Cox 回歸分析，以探討在同步化學放射治療期間使用他汀類藥物與缺血性中風風險之間的關聯性。第二個研究目標（研究二）是建立一套風險預測模型，以辨識在接受放射治療後可能發展為中度以上內頸動脈狹窄的高風險鼻咽癌患者族群。研究二的資料來源為台大醫院整合醫療資料庫（NTUH-iMD）中的電子病歷資料。我們分析接受放射治療後鼻咽癌患者頸內動脈狹窄的風險因子，並利用逐步變數選擇法建立 Cox 比例風險模型的風險預測模型。此外，我們建立了接收者操作特徵曲線（ROC curve）與布萊爾分數（Brier score)，以評估該風險預測模型的判別能力與整體表現。第三個研究目標（屬於研究二）是釐清接受放射治療後鼻咽癌患者進行頸動脈超音波篩檢的最佳起始時機。為了闡明風險的時間分布，我們應用多變量廣義線性混合模型（Generalized Linear Mixed Model, GLMM），以判定放射治療後與發生中度以上內頸動脈狹窄最密切相關的年度時間點。 結果：在研究一中，相較於未使用他汀類藥物的族群，使用他汀類藥物患者的缺血性中風調整後風險比（aHR）為0.70（95%信賴區間：0.54–0.92；P 值為 0.0107）。累積規定日劑量（cDDD）分析顯示劑量反應關係，在 cDDD 較高的四分位數中觀察到較低的缺血性中風風險。此外，日定義劑量（DDD）大於 1 的患者缺血性中風風險降低，其 aHR 為 0.49（95%CI：0.31–0.63）；而DDD 小於或等於 1 的患者，其 aHR 為 0.59（95% CI：0.40–0.84）。在研究二中，根據我們建立的預測模型，風險分數達 5 分以上的高風險族群，其發生中度或以上內頸動脈狹窄的風險明顯增加，且可能在接受放射治療後的第四年即開始出現。此外，我們發現，相較於整體族群的平均風險，發生中度或以上內頸動脈狹窄的風險大約在放射治療後第七年左右開始顯著且持續上升。 結論：我們的研究一提供了證據，支持在接受標準同步化學放射治療的晚期鼻咽癌患者中，他汀類藥物於治療期間使用，能有效降低放射治療引起的缺血性中風風險。在研究二中，根據我們提出的風險預測模型，我們建議鼻咽癌患者於放射治療後第七年開始進行頸動脈超音波篩檢，而高風險族群患者則應提早開始於放數線治療後第四年即開始篩檢。

Introduction: In Taiwan, nasopharyngeal carcinoma (NPC) is the second most common type of head and neck cancer, accounting for approximately 5% of all cancer cases and representing nearly 50% of all head and neck malignancies. Radiotherapy (RT) serves as the most effective treatment for NPC but can induce carotid stenosis and increase the risk of ischemic stroke. In modern era of intensity modulated radiation therapy (IMRT) and effective systemic chemotherapy, prognosis of patients with localized NPC is excellent. With larger number of long-term survivors, the prediction and prevention of late RT-related complications has become important. Study Aims and Methods: The aim 1 of this doctoral dissertation (study 1) is to evaluate the impact of statin use on ischemic stroke risk in patients with advanced NPC undergoing standard concurrent chemoradiotherapy (CCRT). In the first study, we utilized nationwide data from the Taiwan Cancer Registry Database (TCRD), which was linked to the National Health Insurance Research Database (NHIRD) of Taiwan. In Study 1, We applied an inverse probability of treatment-weighted (IPTW) Cox-proportional hazards regression model to examine the association between statin use during CCRT and the risk of ischemic stroke. The aim 2 of this doctoral dissertation (Study 2) is to develop a risk prediction model to identify high-risk groups of NPC patients who are likely to develop moderate or greater internal carotid artery (ICA) stenosis after receiving RT. In the second study, the data source is from electronic medical record from National Taiwan University Hospital, integrative Medical Database (NTUH-iMD). We investigated the risk factors of ICA stenosis in post-RT NPC patients and developed risk prediction model by using stepwise variable selection in Cox proportional hazard model. The Receiver Operating Characteristic (ROC) curve and integrated Brier score were constructed to assess the discriminatory ability and overall performance of our risk prediction model. The aim 3 of this doctoral dissertation (study 2) is to determine the optimal timing for initiating carotid ultrasound screening in post RT NPC patients. To elucidate the temporal pattern of risk, a multivariable Generalized Linear Mixed Model (GLMM) was applied to determine the post-radiotherapy year most strongly associated with the onset of moderate or greater ICA stenosis. Results: In Study 1, the adjusted hazard ratios (aHR) for ischemic stroke in the statin group compared to the non-statin group was 0.70 (95% CI: 0.54–0.92). Analysis based on cumulative defined daily doses (cDDD) demonstrated a clear dose–response relationship, with a progressively lower risk of stroke observed in higher cDDD quartiles. Furthermore, patients with a daily defined dose (DDD) greater than 1 had a significantly reduced stroke risk (aHR: 0.49; 95% CI: 0.31–0.63), whereas those with a DDD of 1 or less also showed a protective effect (aHR: 0.59; 95% CI: 0.40–0.84). In study 2, according to our predictive model, patients in the high-risk group (risk score≥5) have significantly increased risk of moderate or greater ICA stenosis, which may begin to manifest as early as the fourth year after receiving radiation therapy. Additionally, we found that, compared to the average risk of the entire cohort, the risk of developing moderate-to-severe ICA stenosis began to increase significantly and persistently around the seventh year after radiotherapy. Conclusions: Study 1 provides compelling evidence supporting the beneficial effects of statin use during the CCRT period in reducing the risk of radiation-induced ischemic stroke among patients with advanced NPC undergoing definitive CCRT. In the study 2, according to our proposed risk prediction model, we recommend that carotid ultrasound screening begin in the seventh year following radiation therapy, while high-risk patients should start screening earlier, beginning in the fourth year.