中文摘要：

目的使用MR定量重型地中海贫血患者心脏铁沉积，探讨其与血清铁蛋白（SF）、肝铁浓度（LIC）的相关性。方法对58例10岁以上长期接受输血治疗的β-重型地中海贫血患者行心脏MR检查并测量T2＊值，采用Spearman相关分析探讨心脏T2＊值和SF、LIC之间关系。分别以SF〉2500μg／L和LIC〉15mg／g干重为界值将患者分组。用完全随机设计两样本秩和检验比较2组患者的心脏T2＊值。以T2＊〈20ms为诊断心脏铁沉积参考标准，计算以SF〉2500μg/L或LIC〉15mg／g干重预测心脏铁沉积的敏感度、特异度，分别绘制用SF和LIC预测心脏铁沉积的ROC曲线。结果58例患者的心脏T2＊值范围4．7～51．1ms，中位数14．0ms；SF范围1345～23640μg/L，中位数5741灿g／L；LIC范围9．0～〉43．0mg／g干重，中位数41．4mg／g干重。所有患者的心脏T2＊值与sF尚不能认为有线性相关关系（r=-0．240，P=0．070）；心脏T2值和LIC之间呈低度负相关（r=-0．402，P=0．002）。SF≤2500μg／L组7例，心脏T2＊值范围6．1-47．6ms，中位数23．7ms；SF〉2500μg／L组51例，心脏T2＊。值范围4．7-51．1ms，中位数13．5ms，2组心脏T2＊值之间差异无统计学意义（Z=-0．489，P=0．625）。LIC≤15mg／g干重组5例，心脏T2＊值范围24．4-51．1ms，中位数44．8ms；LIC〉15mg／g干重组53例，心脏T2＊值范围4．7-45．5ms，中位数13．2ms，2组心脏T2＊值之间差异有统计学意义（Z=-2．895，P=0．004）。以SF〉2500μg／L和LIC〉15mg／g干重预测心脏铁沉积的敏感度分别为90．9％（30／33）、100．0％（33／33），特异度分别为16．0％（4／25）、20．0％（5／25）。以SF和LIC预测心脏铁沉积的ROC曲线下面积分别为0．652、0．775。结论MRI—T2＊可以直接定量重型地中海贫血患者心脏铁沉积，心脏铁浓度与SF无相关，与LIC低度相关。临床用SF或LIC预测心脏铁沉积不可靠，对诊断及?

英文摘要：

Objective Using MRI-T2 ＊ method to quantify the cardiac iron overload in patients with β-thalassemia major and to evaluate the correlation between cardiac T2 ＊ values and serum ferritin （ SF ） , liver iron concentration（LIC）. Methods Fifty-eight over 10 years old transfusion-dependent patients with β-thalassemia major were underwent MRI heart measurement to obtain T2 ＊ values. Spearman rank correlation was used to analyze the relationship between cardiac T2 ＊ , SF, and LIC. Patients were divided into two groups based on standard setting （ SF 〉2500 p,g/L or LIC 〉 15 mg/g of dry tissue）. Differences of cardiac T2 ＊ values between two groups were evaluated by Wilcoxon rank sum test with cardiac T2 ＊ 〈 20 ms as diagnosis standard. The sensibilities and specificities of prediction for cardiac iron deposition with the index of SF 〉 2500 μg/L or LIC 〉 15 mg/g dry tissue were calculated, and receiver operating characteristic （ROC） curve analysis was performed. Results The range （median） of cardiac T2 ＊ values, SF and LIC in58 patients were 4. 7-51.1 ms （14.0 ms）, 1345-23 640 μg/L （5741 μg/L）, 9.0-〉43.0 mg/g dry tissue （41.4 mg,/g）, respectively. There was no linear correlation between cardiac T2 ＊ values and SF （r = - 0. 240, P = 0. 070）. Cardiac T2 ＊ values and LIC was weakly correlated （ r = - 0. 420, P = 0. 002）. The range （median） of cardiac T2 ＊ values was 6. 1-47. 6 ms （23. 7 ms） in 7 patients of SF≤2500 μg/L group. The range （median） of cardiac T2 ＊ values was 4. 7-51.1 ms（ 13.5 ms） in 51 patients of SF 〉 2500 μg/L group. There was no statistically significant difference between two groups （ Z = - 0. 489, P = 0. 625 ）. The range （median） of cardiac T2 ＊ values was 24.4-51.1 ms （44. 8 ms） in 5 patients of the LIC ≤15 mg/g dry tissue group. The range （median） of cardiac T2 ＊ values was 4. 7-45.5 ms （ 13.2 ms） in 53 patients of HIC 〉 15 mg/g dry tissue group. There was significant differen