摘要目的 定量评价地中海贫血患者心脏、肝脏铁沉积程度,探讨两者关系及临床用肝铁浓度＞ 15 mg/g干重预测心铁沉积的准确性.方法 连续搜集2010年9月至2011年6月期间103例5岁以上总输血(全血)剂量＞10 U(1 U=200 ml)地中海贫血患者资料,行心脏、肝脏MR检查并测量T2*值,采用Spearman秩相关分析其相关性.以肝脏T2*＜0.96 ms(相当于肝铁浓度＞15 mg/g干重)为标准,将患者分为两组,采用秩和检验比较各组心肌T2*值的差异.以心脏T2*＜10 ms、10 ms≤T2*＜20 ms和T2*≥20 ms为标准,将患者分为3组并采用秩和检验比较各组肝脏T2*值有无差异.绘制用肝铁浓度＞ 15 mg/g干重预测患者心脏铁沉积的ROC曲线.结果 103例地中海贫血患者的心脏T2*值中位数为24.00ms(4.70～51.10 ms),肝脏T2*值中位数为1.16 ms(0.68～14.80 ms),两者呈低度相关(r=0.453,P=0.000),且未见规律性.肝脏T2*＜0.96 ms组25例患者的心脏T2*值中位数为12.10 ms(4.70～41.70 ms),T2*≥0.96 ms组78例患者的心脏T2*值中位数为26.10 ms(4.80～51.10 ms),两组之间差异有统计学意义(Z=-3.566,P=0.000).心脏T2*＜10 ms组20例患者的肝脏T2*值中位数为1.06ms(0.68 ～3.83 ms),T2*≥20 ms组58例患者中位数为1.76 ms(0.74～14.80 ms),差异有统计学意义(Z=-3.553,P=0.000)；10 ms≤T2*＜20 ms组25例患者的肝脏T2*值中位数为0.99 ms(0.69 ～ 13.59 ms),与T2*＞20 ms组间差异有统计学意义(Z=-3.951,P=0.000)；心脏T2*＜10 ms组与10 ms≤T2*＜20 ms组相比,两组患者的肝脏T2*值之间差异无统计学意义(Z=-0.046,P=0.964).以肝铁浓度＞15 mg/g干重预测心脏铁沉积的ROC曲线下面积为0.771,敏感度为42.2％,特异度为89.7％.结论 多次输血地中海贫血患者的心、肝铁水平之间呈低度相关.肝铁浓度＞ 15 mg/g干重的患者心铁沉积发生率相对较高,用其预测心脏铁沉积的准确度为中等偏低.

abstractsObjective To quantify the heart and liver iron overload in thalassemia patients and discuss the relationship of iron deposition between them,and to evaluate the accuracy of using hepatic iron concentration ＞ 15 mg/g dry tissue as an index to predict heart iron deposition as used in clinical practice.Methods One hundred and three transfusion-dependent patients with thalassemia,who were older than 5 years,underwent MRI heart and liver measurement to obtain T2 * values.The Spearman rank correlation was employed to analyze the relationship between cardiac T2 * and liver T2 * values.By using liver T2 * =0.96 ms as standard setting,patients were divided into two groups,and the differences of cardiac T2 * values between the two groups were compared by Wilcoxon rank sum test.Then by using cardiac T2 * =10,20 ms as standard setting,patients were divided into 3 groups,and the differences of liver T2 * values among the 3 groups were compared by Wilcoxon rank sum test.The ROC curves were drawn to predict the possibility of using hepatic iron concentration ＞ 15 mg/g dry tissue as an index of cardiac iron deposition.Results The cardiac and liver T2 * values of the 103 thalassemia patients showed low correlation(r =0.453,P =0.000).With the liver T2 * value reduced,the cardiac T2* value did not decline proportionally.The cardiac T2 * value range and median of 25 patients' group whose liver T2 * ＜ 0.96 ms were 4.70 to 41.70 ms and 12.10 ms,respectively.The cardiac T2 * value range and the median of 78 patients' group whose liver T2 * ＞ 0.96 ms were 4.80 to 51.10 ms and 26.10 ms,respectively.There was statistically significant difference between those of the two groups(Z =-3.566,P =0.000).The liver T2 * value range and the median of 20 patients'group whose cardiac T2 * ＜ 10 ms was 0.68 to 3.83 ms and 1.06 ms,respectively.The liver T2 * value range and the median of 58 patients' group whose cardiac T2 * ≥20 ms were 0.74 to 14.80 ms and 1.76 ms,respectively.There was statistical difference between those of the two groups(Z =-3.553,P =0.000).The liver T2 * value range and the median of 25 patients' group with cardiac 10 ms≤T2 * ＜20 ms were 0.69 to 13.59 ms and 0.99 ms,respectively.The values were significantly different from that of T2* ≥20 ms group(Z =-3.951,P =0.000).The liver T2 * values of cardiac T2* ＜ 10 ms group was not statistically different from that of 10 ms≤T2* ＜20 ms group(Z =-0.046,P =0.964).To predict cardiac iron deposition with the index of hepatic iron concentration ＞ 15 mg/g dry tissue,the area under the ROC curve was 0.771.The sensibility was 42.2％,the specificity was 89.7％.Conclusions There is low correlation between heart and liver iron level in thalassemia patients with long-term transfusions.Patients with hepatic iron concentration ＞ 15 mg/g dry tissue have a higher incidence of heart iron deposition,but the accuracy of using hepatic iron concentration as an index to predict myocardial iron deposition is low to moderate.