Lane P H
Department of Pediatrics, St. Louis University School of Medicine, Missouri, USA.
Lab Invest. 1995 Jun;72(6):765-70.
Mean glomerular volume (VG) determined by Cavalieri, maximal profile area (MPA), and Weibel-Gomez methods correlated in paraffin-embedded needle biopsies; the Disector method did not correlate, probably because of sampling limitations. The relationship of these methods in other types of specimens has not been established.
Sections of whole kidneys from 19 rats, 10 treated with furosemide to increase VG, were immersion-fixed, embedded in glycol methacrylate, serially sectioned at 5 microns through 150 microns of tissue, and stained with Jones' silver stain. VG was determined by the Cavalieri, MPA, Disector, and Weibel-Gomez methods with three samples from each animal to assess the correlation, agreement, and repeatability of each of these methods. Methods were considered to correlate if p < 0.01 on linear regression analysis and to agree if the line of identity (x = y) fell within the 95% confidence interval for the slope of the regression curve. Repeatability was assessed with the coefficient of variation for each technique.
Other methods correlated with the Cavalieri method (MPA r = 0.86, Disector r = 0.66, Weibel-Gomez r = 0.87), although none of these methods agreed with Cavalieri VG. These results were similar when treated animals, untreated animals, and all animals were analyzed. Cavalieri and MPA methods showed 7% coefficient of variation with repeated samples; the Disector method showed higher variation (12%), and Weibel-Gomez showed significantly higher variation when compared with the other methods (25%).
MPA, Weibel-Gomez, and Disector methods all provide estimates of VG that correlate with Cavalieri VG; however, none gives the same value. Reasons for these discrepancies may include assumptions regarding glomerular shape and size distribution inherent in the various methods and the experimental conditions. Tissue availability will often dictate VG method, and consistency of method within an experiment is critical.
在石蜡包埋的针吸活检中，通过卡瓦列里法、最大轮廓面积法（MPA）和韦贝尔 - 戈麦斯法测定的平均肾小球体积（VG）具有相关性；而分割法不具有相关性，可能是由于取样限制。这些方法在其他类型标本中的关系尚未确定。
选取19只大鼠的全肾切片，其中10只经速尿处理以增加VG，将肾脏进行浸入固定，包埋于甲基丙烯酸乙二醇酯中，沿组织连续切成5微米厚、共150微米厚的切片，并用琼斯银染法染色。通过卡瓦列里法、MPA法、分割法和韦贝尔 - 戈麦斯法对每只动物的三个样本进行VG测定，以评估这些方法各自的相关性、一致性和可重复性。如果线性回归分析中p < 0.01，则认为方法具有相关性；如果恒等线（x = y）落在回归曲线斜率的95%置信区间内，则认为方法具有一致性。用每种技术的变异系数评估可重复性。
其他方法与卡瓦列里法具有相关性（MPA，r = 0.86；分割法，r = 0.66；韦贝尔 - 戈麦斯法，r = 0.87），尽管这些方法中没有一种与卡瓦列里法测定的VG一致。对处理过的动物、未处理的动物以及所有动物进行分析时，结果相似。卡瓦列里法和MPA法在重复样本时变异系数为7%；分割法变异系数较高（12%），与其他方法相比，韦贝尔 - 戈麦斯法变异系数显著更高（25%）。
MPA法、韦贝尔 - 戈麦斯法和分割法都能提供与卡瓦列里法测定的VG相关的VG估计值；然而，没有一种方法得到的值相同。这些差异的原因可能包括各种方法中关于肾小球形状和大小分布的假设以及实验条件。组织的可获取性通常决定VG测定方法的选择，并且在一个实验中方法的一致性至关重要。
