应用99mTc=DTPA肾动态显像和双血浆法评估肾小球滤过率的一致性研究

Comparison and analysis of GFR measurements assessed from gates method of dynamic renal imaging and ^(99m)Tc-DTPA clerarance of double samples

目的:以双血浆法(dual GFR,dGFR)为标准,评价应用肾动态显像Gates法评估肾小球滤过率(GFR)的准确性和可靠性。方法:2009年6月至2010年6月可引起不同程度肾功能损害的各种常见病患者202例,男123例,女79例,平均年龄(49.5±3.9)岁,共398个肾。依据dGFR法并按Gates法分肾比值计算分肾GFR,将各肾分5级:normal(≥40)、4(30～39.9)、3(20～29.9)、2(10～19.9)、1(<10)[单位为mL/min·(1.73m2)]。分别采用Pearson相关分析和Bland-Altman检验比较dGFR法与Gates法的相关性和一致性。结果:normal级79个肾,4级74个肾,3级84个肾,2级87个肾,1级74个肾。(1)Gates法和dGFR法总体上有显著相关性(r=0.722,P=0.000),回归方程为:Gates=16.708+0.505dGFR;3、4、normal级呈显著弱相关性(分别为r=0.341,P=0.002;r=0.368,P=0.002与r=0.319,P=0.004),回归方程分别为Gates=-9.7+1.899 dGFR、Gates=31.671+0.060dGFR和Gates=17.225+0.879dGFR;1、2级无显著相关性(分别为r=0.038,P=0.747与r=0.198,P=0.077)。(2)Bland-Altman检验结果提示,Gates法和dGFR法的一致性各级患者均欠佳。结论:肾动态显像Gates法测量GFR的准确性尚可,但对其结果应按dGFR法回归计算以获得更准确的GFR,且对于1、2级患者仍应行dGFR法。

Objective The aim of this study was to assess clinical accuracy of GFR measurements assessed from gates methods of dynamic renal imaging by considering the 99mTc-DTPA-dual blood sample method as the reference in patients with different kidney diseases. Methods The study was performed in 202 patients between June 2009 and June 2010 with various degrees of renal function varied from normal to severely impaired [ 123 males and 79 females, age （49.5 ± 3.9） years old]. There were 398 kidneys. The GFR of each kidney was measured by gates and two-sample method. According to 99mTc-DTPA-dual blood sample method （dGFR）. These patients were divided into 5 groups as follows： normal（〉40）, 4（30 ± 39.9）, 3（20 ± 29.9）, 2（10 - 19.9）, 1 （ 〈 10） [mL.min-1. （1.73 mS）-1]. Pearson correlation analysis was used between GFR from gates and dGFR in all groups. Bland-Altman plotting was also performed to define the 95% limits of agreement, which was presented as bias ~ 1.96 s. Results There were 28 patients and 23 kidneys, 16 patients and 42 kidneys, 22 patients and 40 kidneys, 21 patients and 45 kidneys, 20 patients and 34 kidneys in normal, 4, 3, 2, 1 group respectively. （1） Pearson correlation analysis showed good correlation between gates and double-sample method （r = 0.722, P = 0.000） in total patients, the linear regression equation was gates = 16.708 ± 0.505 dGFR. The same correlations were poor in 3, 4 and normal group （r = 0.341, P= 0.002, r = 0.368, P= 0.002 and r = 0.319, P= 0.004, respectively）. The linear regression equation was respectively gates = -9.7± 1.899 dGFR, gates = 31.671 ± 0.060 dGFR and gates = 17.225 ±0.879 dGFR. There were no statically difference between group 1 and 2 （r = 0.038, P = 0.747 and r = 0.198, P = 0.077）. （2）The concordance study using the Bland-Altman method showed that there was poor consistency between gates and dGFR in all groups. Conclusions Dynamic renal imaging was acceptable for estimating GFR, but better correct GFR was assessed by linear regression equation according to dGFR. However, double-sample method should be recommended for accurate determination of the GFR in group 1 and 2.