使用尿液胰蛋白酶原-2預測急性胰臟炎嚴重度 – 與床側急性胰臟炎嚴重度指標(BISAP)比較

Abstract

急性胰臟炎併發症及死亡率加總可以達到10%左右。準確的在第一時間預測急性胰臟炎的嚴重度對病患的預後及病情的告知是重要的。對於嚴重胰臟炎，及時的輸液治療可得到較佳的預後。床側急性胰臟炎嚴重度指標(Bedside Index for Severity of Acute Pancreatitis, BISAP)是由美國波士頓布萊根婦女醫院(Brigham and Women’s Hospital, Boston, USA)的臨床醫師收集臨床資料使用統計分析方法發展出來，可由常規使用的臨床數據，於到診的24小時內預測胰臟炎的嚴重度，較Ranson criteria及APACHE II快速簡單評分，準確度僅些微降低。目前亞洲已有印度、中國大陸、南韓等國對此指標作驗證，台灣仍無相關資料。本研究探討臨床嚴重胰臟炎指標的準確度，研究急性胰臟炎發病初期尿液中胰蛋白酶原-2與胰臟炎嚴重度及死亡率之關聯。在急性胰臟炎發生時，胰蛋白酶原-2(trypsinogen-2)的濃度會較胰蛋白酶原-1(trypsinogen-1)大幅增加，經血液被腎臟濾出而進入尿液中，目前已有尿液試紙檢測可快速檢驗。然而關於其尿液中濃度與急性胰臟炎重度的關聯性，目前僅有零星文獻顯示可能有相關。 我們預計在2015年9月至2016年5月間，施行急性胰臟炎的先期研究。所有至本院診斷患有急性胰臟炎且症狀發作於三天內的病患，若願意接受驗尿及後續疾病嚴重度追縱者納入本案。預計收案45人。入院後24小時內紀錄Glasgow coma scale，體溫，呼吸，心跳，CBC，血中尿素氮，胸部X光並取最不良值以記錄床側急性胰臟炎嚴重度指標(BISAP)。在24小時內也將留存尿液10ml以供研究，並冰存於-80°C。以urine trypsinogen-2 rapid strip test 檢驗尿液中胰蛋白酶原-2濃度是否大於50ng/ml及2000ng/ml(稀釋40倍)。於入院後48小時根據2012年修正版亞特蘭大分類來區分嚴重度。將以The Mann-Whitney U test、The Kruskal-Wallis test來判斷床側急性胰臟炎嚴重度指標及尿液中胰蛋白酶原-2濃度與急性胰臟炎嚴重度之間的關係。類別變項則使用皮爾森卡方檢驗及費雪精確性檢定來分析。包括床側急性胰臟炎嚴重度指標、尿中胰蛋白酶原-2濃度將使用接收者操作特徵曲線（receiver operating characteristic curve）來計算其預測胰臟炎嚴重度及死亡率的準確度。共收案42人，輕度急性胰臟炎與中重度及嚴重急性胰臟炎在胰蛋白酶原-2的濃度上，兩組並無統計上的顯著差別(P=0.0618)。接收者操作特徵曲線下面積(AUC)用以預測中重度及嚴重急性胰臟炎，在胰蛋白酶原-2下為0.68，BISAP score下為0.66，兩者間無顯著差別。針對胰蛋白酶原-2的接收者操作特徵曲線計算Youden index，並選擇500μg/l做為切點其預測中重度胰臟炎及嚴重胰臟炎的敏感度為100%，特異度50%，陽性診斷率50%，陰性診斷率100%，陽性相似比2.0，陰性相似比0.0。研究顯示胰蛋白酶原-2在預測中重度胰臟炎及嚴重胰臟炎的敏感度為100%，陰性診斷率100%，有負向指標潛力，待後續研究釐清。

The complication and mortality rate of acute pancreatitis are totally around 10%. The overall costs of acute panreatitis increased rapidly. It is important to accurately predict pancreatitis severity in order to decide treatments, to facilitate disposition, and to inform prognosis. Timely and adequate fluid infusion results in better prognosis in acute severe pancreatitis. The aim of this study is to preliminarily investigate the association between urine trypsinogen-2 level and acute pancreatitis severity and mortality. It will also investigate BISAP score accuracy in Taiwan. Bedside Index for Severity of Acute Pancreatitis (BISAP) is developed by clinicians from Brigham and Women’s Hospital (Boston, USA) using statistical analysis. Applying routine clinical data, BISAP can predict acute pancreatitis severity in 24 hours. BISAP is simple and quick scoring system comparing to Ranson criteria and APACHE II scores with only slightly inferior accuracy. BISAP score has been utilized and validated in Asia countries including India, China, South Korea; however, there is no Taiwanese Data. During acute pancreatitis, the serum concentration of trypsinogen-2 increases much more than trypsinogen-1 does. Trypsinogen-2 is than filtrated from serum into urine. Rapid urine strip test has been well developed for years as well as ELISA test. Until nowadays, few reports indicate that concentration of trypsinogen-2 in urine are associated with acute pancreatitis severity and mortality. From Septemper, 2015 to May, 2016, all patients with acute pancreatitis and with onset in 3 days are enrolled in this preliminary study after informed and consent. This study enrolled 42 patients. Total 10ml urine will be gathered within 24 hours of admission. The urine rapid strip test for trypsinogen-2 is done immediately. Dilute 40x urine strip test is done after positive result of trypsinogen-2 test. Within 24 hours of admission, we recorded BISAP scores, complete blood cell counts, LDH, calcium level, and triglyceride level. Within 48 hours, all patients underwent ultrasound. ELISA test for urine trypsinogen-2 quantification is done using Trypsinogen-2 IEMA test for urine. The acute pancreatitis severity is decided according to revised 2012 Atlanta criteria. The association between pancreatitis severity, mortality and urine trypsinogen-2 level is tested by the Mann-Whitney U test and the Kruskal-Wallis test. Receiver operating characteristic curve (ROC curve) are used to evaluate the prediction power of urine trypsinogen-2 concentration for severity and mortality of acute pancreatitis. Youden index was calculated for trypsinogen-2 cut-off value determination. Fisher’s exact test was performed for trypsinogen-2 and BISAP score cut-off value. There was no statistically significance of trypsinogen-2 level between mild acute pancreatitis group and moderate severe and acute severe pancreatitis group (p=0.0618). When receiver operating characteristic curve (ROC curve) was used to predict moderate severe and acute severe pancreatitis, there was no statistically significance between BISAP ROC curve and trypsinogen-2 ROC curve(0.66 v.s 0.68, p=0.8286 ). Using the trypsinogen-2 ROC curve, the Youden index was identified. The cutoff point of 500μg/l was chosen referring to Youden index and it showed sensitivity 100%, specificity 50%, positive predictive value 50%, negative predictive value 100%, positive likelihood ratio 2.0, and negative likelihood 0.0. It showed trypsinogen-2 may serve as negative predictor and future studies to confirm this finding are needed. The urine trypsinogen-2 rapid test potentially may be applied to clinical environment as rapid negative predictor.