原發性皮質醛酮症之心肌肥厚，纖維化及舒張功能 研究–自臨床到基礎

Abstract

背景及目的 原發性高醛固酮症(Primary aldosteronism) 為續發性高血壓的重要原因之一，具有較高的心血管疾病風險。左心室肥大為原發性高醛固酮症在心臟最早的變化之一，包括心臟質量、纖維化及舒張功能都會受到影響。目前並未有好的臨床指標可以預測這類病人是否已有左心室肥大。此外，造成心室肥大的機轉也並未完全清楚。而將醛固酮抑制劑用在舒張功能不良型的心衰竭的臨床試驗結果也不如預期。因此，對於醛固酮造成心室肥大及舒張功能不良的機轉值得進一步探討。本論文將針對上述要點從臨床到基礎進行一系列的研究。 原發性高醛固酮症病患常伴隨左心室不適當的肥大(相對於性別、體型及血壓而言)。目前並沒有好的臨床指標可以在原發性高醛固酮症病患中，預測是否有左心室肥大或是不適當的左心室質量(inappropriate left ventricular mass) 的存在。本論文的第一部分的假說為24小時尿液中的醛固酮因較能反映整體的醛固酮曝露量，因而更能預測是否有左心室肥大。我們將比較三種指標(24小時尿液中的醛固酮、血中腎素(renin) 活性及血中醛固酮對腎素活性比值)，何者較能預測病患是否具有左心室肥大或是不適當的左心室質量存在。 此外，膠原蛋白的代謝相關生物指標是否和醛固酮引起的舒張功能不良有關係目前並不清楚。原發性高醛固酮症病患已知常會有較明顯的舒張功能不良，但這舒張功能不良是否可逆目前也不清楚。目前也沒有好的臨床指標可以預測醛固酮引起的舒張功能不良。基質金屬蛋白脢-2 (matrix metalloproteinase-2)及基質金屬蛋白酶組織抑制因子-1 (tissue inhibitors of metalloproteinase -1) 為兩種膠原蛋白的代謝相關生物指標，已知和高血壓或收縮正常的心衰竭病人的舒張功能有關。本論文的第二部分的假說是，膠原蛋白的代謝相關生物指標，特別是基質金屬蛋白酶組織抑制因子-1，可能和心臟舒張功能有相關。我們設計臨床研究來探討，在原發性高醛固酮症病患中，基質金屬蛋白脢-2及基質金屬蛋白酶組織抑制因子-1 是否和這些病患的舒張功能有關；並且探討這兩個因子和舒張功能的可逆性是否有關。 醛固酮已知會造成心肌的纖維化。而基質金屬蛋白酶組織抑制因子-1在動物及人的壓力過荷的模式中，也顯示和纖維化有關。因此本研究的第三部分假說是醛固酮在心肌纖維母細胞中可以誘發基質金屬蛋白酶組織抑制因子-1的表現。我們設計細胞實驗來回答此問題。 研究對象與方法 在論文第一部分，我們招募106位原發性高醛固酮症病患(合併有高血壓)及31位本態性高血壓的病患作為研究對象，並測量其血中腎素、醛固酮濃度及24小時尿液中的醛固酮濃度。病患並接受心臟超音波檢查，測量其左心室質量指數 (left ventricular mass index)及左心室過度質量指數 (excessive left ventricular mass index)。 論文的第二部分，我們納入27位醛固酮分泌腺瘤引起的原發性高醛固酮症病患(病患皆合併有高血壓，且準備接受腺瘤切除手術)，及27位本態性高血壓的病患(對照組)作為研究對象。研究對象分別接受血中基質金屬蛋白脢-2及基質金屬蛋白酶組織抑制因子-1測定及心臟超音波組織都普勒影像檢查。在醛固酮分泌腺瘤的27位病患中，在手術後 1 年另外接受第二次心臟超音波檢查及血液測量。 論文的第三部分為基礎實驗。在細胞實驗中，先將醛固酮加入心臟纖維母細胞培養皿中。在24小時後，收集心臟纖維母細胞培養皿的上清液，來測量基質金屬蛋白酶組織抑制因子-1的蛋白濃度。基質金屬蛋白酶組織抑制因子-1 的 mRNA 表現量則使用定量的即時逆轉錄聚合酶鏈式反應進行測定。下游路徑則以化學抑制劑及干擾RNA 作檢驗。在動物實驗中，則利用醛固酮植入小鼠來驗證血中的基質金屬蛋白酶組織抑制因子-1是否因醛固酮給予而上升。 結果 在第一部分的研究當中，只有24小時尿液醛固酮量和左心室質量指數及左心室過度質量指數有顯著相關。在多變項分析中也證實此一關聯。比較 24小時尿液醛固酮量、血中醛固酮濃度及血中醛固酮對腎素活性比等三項指標，在診斷左心室肥大(ROC 曲線下面積分別為0.701, 0.568, 0.656) 及具有不適當的左心室質量(ROC曲線下面積分別為 0.61, 0.43, 0.493)的能力，發現以24小時尿液醛固酮量較優。 在第二部分的研究發現，基質金屬蛋白酶組織抑制因子-1、左心室質量指數及舒張功能不良指標(E/E’比值)在醛固酮分泌腺瘤病患中，都顯著較本態性高血壓患者要高。基質金屬蛋白酶組織抑制因子-1血中濃度和左心室質量指數、心室中膈厚度、左心房直徑及E/E’比值都有顯著相關。而基質金屬蛋白脢-2 和左心室結構的指標都無關(除了與心室中膈厚度有關)。病患在接受腺瘤切除手術後，心室質量指數及E/E’比值都有顯著改善。術後的基質金屬蛋白酶組織抑制因子-1也有顯著下降，但術後基質金屬蛋白脢-2 則未下降。基質金屬蛋白酶組織抑制因子-1在手術前後的變化和術後的 E/E’比值有顯著負相關 (校正年紀、性別、身體質量指數及平均血壓後: β coefficient = −3.6, p = 0.004)。 在第三部分的研究中，人類心臟纖維母細胞在接受醛固酮刺激後，基質金屬蛋白酶組織抑制因子-1的mRNA表現量及蛋白質分泌都有顯著增加。此增加在刺激後4小時出現，於8小時達到高峰。其下游的訊息傳遞係經由 PI3K/AKT (而不經 ERK 或p38)，最後經由細胞核因子NF-kB，而使基質金屬蛋白酶組織抑制因子-1 的表現增加。在小鼠的醛固酮植入模式中，也發現小鼠血液中的基質金屬蛋白酶組織抑制因子-1在植入後7天顯著上升，並持續到第21天。 結論 我們的研究發現24小時尿液醛固酮含量比其他血液指標更能預測原發性高醛固酮症病患是否有左心室肥大及不適當的左心室質量存在。在第二部分的研究中，我們證實過度的醛固酮和心臟的舒張功能不良有關；此舒張功能不良可以在腺瘤切除手術後得到改善。而血液中的基質金屬蛋白酶組織抑制因子-1和此舒張功能有關。而基質金屬蛋白酶-2則和舒張功能無關。在第三部分的研究進一步證實醛固酮可以在心臟纖維母細胞誘發基質金屬蛋白酶組織抑制因子-1表現。此一反應的細胞內途徑為 PI3K/AKT, NF-kB。醛固酮植入小鼠模式也證實在動物中額外給予醛固酮，會增加基質金屬蛋白酶組織抑制因子-1 在血中的濃度。 我們這一系列的研究找到新的心室肥大及舒張功能不良的指標，並證實基質金屬蛋白酶組織抑制因子-1 可能是醛固酮造成心臟纖維化的一個下游因子。

Background and objectives Primary aldosteronism (PA) is one important cause of resistant hypertension and is associated with a higher risk for cardiovascular complications. Left ventricular hypertrophy is an early manifestation of cardiac involvement. Primary aldosteronism is associated with a higher left ventricular mass, fibrosis and diastolic dysfunction. Currently, there are no ideal clinical parameters to predict the presence of left ventricular hypertrophy. Moreover, the underlying mechanism for how hyperaldosteronism relates to cardiac hypertrophy is also unknown. The use of mineralocorticoid antagonist in patients with diastolic heart failure yielded neutral results. Therefore, this topic is worthy of further investigation. This thesis will address these points in a series of studies. Primary aldosteronism is associated with inappropriate left ventricular hypertrophy in relation to gender, body size, and blood pressure. There is no ideal parameter to predict the presence of left ventricular hypertrophy or an inappropriate left ventricular mass in patients with primary aldosteronism. In the first part of this thesis, we hypothesized that the 24-hour urinary aldosterone level may reflect the total aldosterone exposure better than other plasma parameters. We designed a cross-sectional study to investigate the performance of 24-hour urinary aldosterone level, spot plasma renin activity and spot plasma aldosterone-to-renin ratio to predict the presence of left ventricular hypertrophy or inappropriate left ventricular mass. Furthermore, whether collagen markers are associated with aldosterone-induced diastolic dysfunction is still unknown. There are no known biomarkers associated with aldosterone-induced diastolic dysfunction. Two collagen-related biomarkers, tissue inhibitors of metalloproteinase -1 (TIMP-1) and matrix metalloproteinase-2 have been associated with diastolic dysfunction in patients with heart failure and preserved ejection fraction. In the second part of this thesis, we hypothesized that collagen markers, especially TIMP-1, are related to diastolic function among PA patients. We designed a follow-up study to test if TMP-1 or matrix metalloproteinase-2 were associated with diastolic dysfunction among patients with primary aldosteronism. We also tested if changes in TIMP-1 or matrix metalloproteinase-2 were associated with the reversibility of diastolic dysfunction. Aldosterone induces fibrosis in the myocardium. TIMP-1 has been shown to be associated with myocardial fibrosis among patients and animals with pressure overload. In the third part of this thesis, we hypothesized that aldosterone could induce TIMP-1 expression in cardiac fibroblasts and contribute to the fibrotic process. We designed a cellular study to demonstrate the aldosterone-induced TIMP-1 response. Materials and methods In the first part of this thesis, we performed echocardiography in 106 patients with primary aldosteronism and 31 subjects with essential hypertension. Plasma renin activity, aldosterone concentration and 24-hour urinary aldosterone levels were measured. In the second part of this thesis, we enrolled 27 patients with aldosterone producing adenomas preparing for adrenalectomy and 27 patients with essential hypertension. Plasma matrix metalloproteinase-2 and TIMP-1 levels were measured, and echocardiography, including tissue Doppler images, was performed for both groups as well as subsequent testing one year after receiving adrenalectomies for the aldosterone-producing adenoma group. In the third part of this thesis, human cardiac fibroblast cells were used in experiments to determine if aldosterone induced TIMP-1 expression. Cellular supernatant was collected to measure TIMP-1 levels. Pathways downstream of TIMP-1 were tested with chemical inhibitors and siRNAs. Aldosterone releasing pellets were implanted in eight-week-old wild-type mice (Institute for Cancer Research). The presence of TIMP-1 in the serum of these mice was measured at 0, 7 and 14 days after pellet implantation. Results In the first part of this thesis, only 24-hour urinary aldosterone correlated with left ventricular mass index and excessive left ventricular mass index among these parameters. The multivariate analysis revealed that 24-hour urinary aldosterone levels were an independent predictor for left ventricular mass index and excessive left ventricular mass index. The area under ROC curves for 24-hour urinary aldosterone, plasma aldosterone concentration and plasma aldosterone-to-renin ratio to identify the presence of left ventricular hypertrophy (area under ROC curve=0.701, 0.568, 0.656, respectively) and the presence of inappropriate LV mass index (area under ROC curve = 0.61, 0.43, 0.493, respectively) showed that 24-hour urinary aldosterone was the best predictor. In the second part of this thesis, the baseline plasma TIMP-1 levels, left ventricular mass index (LVMI) and E/E` ratio were higher in the aldosterone-producing adenoma group. The baseline TIMP-1 level was correlated with E/E` ratio, LVMI, and interventricular septum thickness. The matrix metalloproteinase-2 levels did not correlate with left ventricular structure parameters, except for interventricular septum thickness. After adrenalectomy, LVMI and E/E` ratios improved significantly. The post-adrenalectomy TIMP-1 levels, but not matrix metalloproteinase-2 levels, also decreased. The change of TIMP-1 was negatively associated with the post-adrenalectomy E/E` ratio after adjustment for age, gender, body mass index and mean blood pressure (β coefficient = −3.6, p = 0.004). In the third part of this thesis, TIMP-1 protein and mRNA expression increased after treating human cardiac fibroblasts with aldosterone. The downstream pathway of this increased expression was through the PI3K/AKT pathway, but not ERK or p38. Aldosterone acted via NF-kB to increase the expression of TIMP-1. This aldosterone-induced TIMP-1 expression was also noted in the mice serum. Conclusion In conclusion, 24-hour urinary aldosterone level performed better to predict the presence of left ventricular hypertrophy and inappropriate left ventricular mass index in patients with primary aldosteronism. Furthermore, we confirmed that an excess of aldosterone induced cardiac diastolic dysfunction, which was reversible by adrenalectomy. TIMP-1 was associated with aldosterone-induced diastolic dysfunction. Finally, aldosterone induced TIMP-1 expression via glucocorticoid receptors, PI3K and NF-kB pathways in cardiac fibroblasts, and resulted in enhanced collagen synthesis. Aldosterone increased the concentration of TIMP-1 in serum for the mouse model. Our thesis found clinical predictors for left ventricular hypertrophy and diastolic dysfunction, and established TIMP-1 as a new downstream mediator for aldosterone to induce myocardial fibrosis.