| 摘要: | 目的：評估腎臟惡性腫瘤合併之慢性共病症與總生存率之間的關聯，並探討免疫基因譜對腎臟惡性腫瘤和糖尿病腎變病預後的影響。
背景：慢性共病症及免疫基因在腎臟惡性腫瘤患者的預後及存活率中扮演著重要的角色。不同的慢性共病症對預後及死亡率具有不同的影響，其中慢性腎臟疾病是腎臟惡性腫瘤最重要的慢性共病症之一。而糖尿病腎變病則是慢性腎臟疾病的主要原因，同時也會增加罹患腎臟惡性腫瘤的風險。儘管台灣地區慢性腎臟疾病的患病率很高，但遺傳變異在糖尿病腎病變中扮演的角色仍不清楚。腎臟惡性腫瘤是一種促血管生成和免疫原性腫瘤，富含豐沛的免疫細胞浸潤。然而，腫瘤內免疫細胞浸潤的表現與臨床分期和預後方面之相關性仍沒有一致的結論。
方法與材料: 首先，為了量化個別慢性共病症對腎臟惡性腫瘤患者的影響，收集了8964 名被診斷患有腎臟惡性腫瘤（ICD-9-CM 代碼：189.0）的病患。本研究分析了腎臟惡性腫瘤患者常見之慢性共病症並使用 Cox 回歸模型確立與死亡率有關的重要預測因子。依據這些與死亡率有顯著差異的重要預測因子，構建了預測模型和列線圖，以預測腎臟惡性腫瘤患者的存活率。其次，為了確立慢性共病症之一的糖尿病腎變病的遺傳變異角色，收集了275 名糖尿病腎變病患者來進行基因關聯研究。分析RAB38和CXCL10的單一核苷酸多型性與尿液全蛋白/肌酸酐比值（UPCR）的相關性。最後，為了探索腎臟惡性腫瘤微環境的免疫特徵，從癌症基因體圖譜中，收集了245名乳頭狀腎細胞癌(KIRP)病患之RNA-seq 數據和免疫相關基因的表現並透過應用單樣本基因集富集分析（ssGSEA）。分別使用線性回歸模型和Cox 回歸模型分析免疫元基因與腎臟惡性腫瘤之臨床分期及總體存活率之間的相關性。此外，從臺北醫學大學聯合人體生物資料庫收集了15名腎臟惡性腫瘤樣本，分析RAB38和CXCL10在不同分期腎臟惡性腫瘤及鄰近正常組織中表現量的相關性。
結果: 本研究分析了5090腎臟惡性腫瘤患者合併之慢性共病症，尤其以老年患者其發生率更高並確立與死亡率有關的重要預測因子，包括年齡，男性，慢性腎臟疾病，透析病患，腎結石，腦血管疾病和轉移性腫瘤。其預測模型之Harrell 的 C 統計值和ROC曲線下方的面積分別為 0.74 和 0.77。我們還分析了275名糖尿病腎變病患者，並顯示RAB38中的rs1027027，rs302647和rs302646與糖尿病腎病變患者的尿液全蛋白/肌酸酐比值（UPCR）有顯著相關，並且在男性族群中也有重要的遺傳影響。但是，在CXCL10基因型和UPCR之間沒有發現有顯著的關聯性。在腎臟腫瘤組織免疫圖譜研究中，自然殺傷（NK）細胞（CD56-），CD4+ T輔助2（Th2）細胞和活化樹突狀細胞（aDC）與腎臟惡性腫瘤之臨床分期及總體存活率呈顯著正相關，而樹突狀細胞（DC）則與總體存活率呈顯著負相關。此外，CXCL10的高度表現與晚期之腎臟惡性腫瘤相關。
結論: 腎臟惡性腫瘤患者的慢性共病症患病率很高，尤其是老年患者。慢性共病症及免疫基因在腎臟惡性腫瘤患者的預後及存活率中扮演著重要的角色。慢性共病症中的糖尿病腎病變患者RAB38變異體和性別與UPCR之高度相關性。 最後，免疫元基因和CXCL10的表現可用作潛在的生物標誌物，尤其是用於進一步研究腎臟惡性腫瘤的分子機轉以及腎臟腫瘤微環境。
Aim: To estimate the association between comorbid diseases and overall survival in kidney cancer and to determinate the impact of immune gene profiles on prognosis in kidney cancer and diabetic nephropathy.
Introduction: The effect of comorbid conditions and immune profiles of kidney cancer microenvironment have significant impact on prognosis and overall survival. Comorbidities have unique effects on mortality and one of the most important comorbidity of kidney cancer is chronic kidney disease. Diabetic nephropathy is a leading cause of chronic kidney disease and has a significantly higher risk for kidney cancer had been reported. Despite the high prevalence of diabetic kidney disease, the impact of genetic variants in diabetic nephropathy remain uncertain. Kidney cancer is a pro-angiogenic and immunogenic tumor and rich in immune infiltrates. However, the expression of intra-tumoral immune cell infiltrates demonstrated conflicting results on clinical outcomes and prognosis.
Method and Materials: Firstly, to quantify the effect of comorbidities of kidney cancer, 8964 participants diagnosed with kidney cancer (ICD-9-CM code: 189.0) were recruited. Demographic characteristics and comorbidities were examined. The Cox regression model was used and statistically significant variables were obtained. A prediction model and nomogram based on these significant prognostic factors were developed to predict overall survival of kidney cancer patients. Then, to determine the genetic role of comorbidity and diabetic nephropathy, 275 diabetic nephropathy patients were recruited to perform the genetic association study. The SNPs of RAB38 and CXCL10 were selected and the correlation with urinary protein-to-creatinine ratio (UPCR) was analyzed. Finally, to explore the immune profiles of the kidney cancer microenvironment, 245 RNA-seq data of renal papillary cell carcinoma (KIRP) and the expression of immune-related genes from The Cancer Genome Atlas (TCGA) database were analyzed. Each metagene was estimated from signature genes by applying the single-sample gene set enrichment analysis (ssGSEA). The correlation between immune metagenes and clinical characteristics and overall survival were determined using a linear regression model and cox-proportional hazard model respectively. Moreover, 15 kidney cancer samples were acquired from the Joint Biobank of Taipei Medical University and the correlation between the expression levels of RAB38 and CXCL10 in cancer and adjacent normal tissue of different stages was analyzed.
Results: 5090 kidney cancer patients were included and high prevalence of comorbidity especially elderly patients were identified. The significant predictors for mortality were metastasis tumor, renal stone, chronic kidney disease, cerebrovascular disease, dialysis requirement, age, and male gender were recognized. The Harrell's C- statistic and the integrated time-dependent AUC of our prediction nomogram were 0.74 and 0.77 respectively. We also recruited 275 diabetic nephropathy patients and demonstrated that rs302647, rs1027027, and rs302646 of RAB38 were correlated with urinary protein-to-creatinine ratio (UPCR) and also significant in male population. However, there is no significant association was identified between CXCL10 genotypes and UPCR. In the kidney tumor tissue profiles, the expression levels of CD4+ T-helper 2 cells, natural killer cells (CD56dim), and activated dendritic cells demonstrated a significant positive correlation with clinical stage and overall survival, while the expression level of dendritic cell showed a negative correlation with overall survival. Moreover, the higher expression level of CXCL10 was associated with advanced stage of kidney cancer.
Conclusions: Patients with kidney cancer exhibit high prevalence of comorbid disease, particularly older patients. The effect of comorbid conditions and immune profiles of kidney cancer microenvironment have significant impact on prognosis and overall survival. The variants of RAB38 and male gender have a significant association of UPCR in diabetic nephropathy patients. Finally, the expression levels of CXCL10 and immune metagenes in kidney cancer microenvironments can be utilized as potential biomarkers and prognostic factors, which may serve a critical role in tumor progression. |
