藉由[氟-18]去氧葡萄糖正子電腦斷層掃描影像探討腹部脂肪體積與身體質量指數和腰圍之相關性

目的腹部脂肪，特別是臟器脂肪組織(Visceral Adipose Tissue, VAT)，已證實與代謝症候群和冠狀動脈疾病有關。最近已發表(J Am Coll Cardiol Img 2010;3:843-51.)的文章中，使用[氟-18]去氧葡萄糖估算臟器脂肪組織和皮下脂肪組織(Subcutaneous Adipose Tissue, SAT)的葡萄糖代謝。從正子斷層掃描中的低劑量衰減電腦斷層影像，藉由影像分割技術，分析皮下脂肪組織和臟器脂肪組織的體積以及肝臟、脂肪及肌肉的平均[氟-18]去氧葡萄糖攝取標準值(Standard Uptake Values, SUV)研究其相關性，並推導經由BMI值和腰圍簡易預估腹部脂肪體積的計算方程式。材料與方法在4年期間，有100個樣本於[氟-18]去氧葡萄糖正子電腦斷層掃描影像中，呈現無陽性反應，且在正子檢查結束後兩周內有血糖、血清膽固醇、三酸甘油脂等，相關血液生化值，可納入此次研究中。影像重組後，電腦斷層影像藉由Philips Syntegra®軟體設定脂肪組織介於-30和-190之間的韓森費爾德單位(Hounsfield Unit, HU)值，進行觀察者內和觀察者間影像分割處理，應用於第四腰椎及第五腰椎間的兩公分厚度，量測皮下脂肪組織和臟器脂肪組織的體積，並利用正子斷層掃描影像量測肝臟、脾臟、腰大肌、臟器脂肪組織和皮下脂肪組織的平均[氟-18]去氧葡萄糖攝取標準值。再進行全腹部脂肪(Total Abdominal Fat, TAF)體積、皮下脂肪體積、臟器脂肪體積、組織的平均[氟-18]去氧葡萄糖攝取標準值、身體質量指數(Body Mass Index, BMI)、腰圍和血清生化值皆進行相關性測試。結果在同一觀察者和不同觀察者測量的腹部脂肪相關性皆良好(r = 0.99)。身體質量指數和全腹部脂肪、臟器脂肪組織及皮下脂肪組織的相關性分別為(r = 0.79，0.69，0.62)；三者和腰圍的相關性分別為(r = 0.78，0.76，0.57)。組織之全部脂肪量測量的血液生化值及平均[氟-18]去氧葡萄糖攝取標準值相關性皆低(r < 0.5)。利用BMI值推估腹部脂肪的方程式為y = 6.538x + 426.53，腰圍推估腹部脂肪的方程式為y = 6.029x + 452.21。結論藉由正子電腦斷層掃描中的衰減校正電腦斷層影像量測腹部脂肪組織是可行且可重複的。影像量測結果和身體質量指數及腰圍相關性佳，但與血液生化值及肝臟、脂肪和肌肉的平均[氟-18]去氧葡萄糖攝取標準值無顯著相關性；利用BMI值和腰圍求得腹部脂肪的簡易方程式，可以提供臨床進行粗略估測脂肪的體積。

AimAbdominal fat, especially visceral adipose tissue (VAT), is proven to relate with metabolic syndrome and coronary artery disease. Using FDG to estimate VAT and subcutaneous adipose tissue (SAT) glucose metabolism had been recently reported (J Am Coll Cardiol Img 2010;3:843-851). The analysis of SAT and VAT volume by applying imaging segmentation on attenuation low dose CT images from PET/CT and correlated with the mean standard uptake values (SUV) of liver, fat and muscle were studied. The formula was derived from BMI and waist circumference to estimate the volume of abdominal fat.Materials and MethodsIn a 4-year period, 100 negative FDG PET/CT findings subjects who had blood sugar, serum cholesterol and triglycerol tests within two weeks were recruited for the study. After reconstruction, the CT image was display in Philips Syntegra® software for CT segmentation by setting the binary Hounsfield unit (HU) between -30 and -190 for fat tissue. The SAT and VAT volume measurement was applied between L4 and L5 spine level for 2 cm thickness. Intra- and inter-observer measurements of the SAT and VAT from CT were tested. The means of FDG SUVs of the liver, spleen, psoas muscle, VAT and SAT were measured from PET imaging. The correlation between total abdominal fat, SAT, and VAT volume, mean FDG SUVs of above mentioned tissue, body mass index (BMI), waist circumference, and serum biochemical analysis were tested.ResultsThe correlation between intra- and inter-observer measurements of the abdominal fat were all excellent (r = 0.99). The correlation between total abdominal fat (TAF), VAT, and SAT to BMI were (r = 0.79, 0.69, and 0.62), and those to the waist circumference were (r = 0.78, 0.76, and 0.57), respectively. The correlation between all the fat measurements to the blood biochemical tests and SUVs of above mentioned tissue were all low (r < 0.5). To estimate the abdominal fat from BMI, we could take the advantage of the formula y = 6.538x + 426.53. We could also estimate the abdominal fat from waist circumference through the formula y = 6.029x + 452.21.ConclusionsThe abdominal adipose tissue measurement was feasible and reproducible by using the attenuation CT imaging of the FDG PET/CT. The results of the measurements correlated well with the BMI and waist circumference but not with the blood biochemical tests and mean SUVs of liver, fat and muscle. The abdominal fat formula obtained from the BMI and waist circumference, it could provide for clinical to estimate the volume of fat.