ª Springer Science+Business Media New York 2015
Abdominal Imaging
Abdom Imaging (2015) DOI: 10.1007/s00261-015-0385-0
Quantification of liver, pancreas, kidney, and vertebral body MRI-PDFF in non-alcoholic fatty liver disease Ilkay S. Idilman,1 Ali Tuzun,2 Berna Savas,3 Atilla Halil Elhan,4 Azim Celik,5 Ramazan Idilman,2 Musturay Karcaaltincaba1 1
Department of Radiology, Faculty of Medicine, Hacettepe University, 06100 Ankara, Turkey Department of Gastroenterology, Faculty of Medicine, Ankara University, Ankara, Turkey 3 Department of Pathology, Faculty of Medicine, Ankara University, Ankara, Turkey 4 Department of Biostatistics, Faculty of Medicine, Ankara University, Ankara, Turkey 5 General Electric Healthcare, Istanbul, Turkey 2
Abstract Purpose: The purpose of the present study was to determine liver, pancreas, kidney, and vertebral fat deposition in NAFLD patients by proton density fat fraction (PDFF) using magnetic resonance imaging (MRI) and to evaluate the relationships among them. Methods: A total of 41 biopsy-proven NAFLD patients underwent MRI-PDFF with IDEAL-IQ. MRI protocol included T1-independent volumetric multi-echo gradientecho imaging with T2* correction and spectral fat modeling. The MR examinations were performed on a 1.5 HDx MRI system. MRI-PDFF measurements were obtained from liver, pancreas, renal cortex and sinus, and vertebral body. Liver biopsy specimens were retrieved from the archives and evaluated by one pathologist according to NASH CRN. Results: The median age of the patients was 47 years. The median interval between liver biopsy and MRI examination was 16 days. Mean liver, pancreas, renal cortex, renal sinus, T12 and L1 vertebral body MRI-PDFFs were 18.7%, 5.7%, 1.7%, 51%, 43.2%, and 43.5%, respectively. No correlation between either liver MRI-PDFF or histological steatosis, and other organ MRI-PDFFs was observed. A good correlation between pancreas and vertebral body MRI-PDFFs, and pancreas and renal sinus MRI-PDFFs was observed. Diabetic patients had higher average pancreas MRI-PDFF compared to nondiabetics (12.2%, vs., 4.8%; P = 0.028). Conclusions: Pancreas and vertebral body MRI-PDFF is well correlated in NAFLD patients and both of them are Correspondence to: Musturay Karcaaltincaba; email: musturayk@yahoo. com
higher in diabetic patients which may explain increased bone fractures in diabetics. MRI-PDFF can be used to demonstrate fat fractions of different organs and tissues and to understand fat metabolism. Key words: Non-alcoholic fatty liver disease—MRIPDFF—Pancreas—Kidney—Vertebral body
Non-alcoholic fatty liver disease (NAFLD) is the excessive accumulation of fat in hepatocytes and affects approximately 30% of general population with an increasing rate [1, 2]. Although frequently benign, there are several clinical manifestations of this disease related to liver injury like non-alcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma [1–4]. Beside hepatic manifestations, associations with obesity, insulin resistance (IR), type 2 diabetes mellitus, hypertension, hyperlipidemia, and cardiovascular disease suggest complex relationships among them. Liver biopsy remains the reference method for both confirming the diagnosis of NAFLD and grading the disease. Potential risks of liver biopsy like bleeding, perforation, and even mortality [5] motivated the investigators to find a non-invasive method for diagnosis and grading of NAFLD. Magnetic resonance spectroscopy (MRS) is the most accepted method for quantitative analysis of the liver fat [6–8]. However, it is not available for routine clinical practice because of technical difficulties in both performance and evaluation [7, 9]. A recent MR technique—MRIestimated proton density fat fraction (MRI-PDFF)—gives
I. S. Idilman et al.: Quantification of ectopic fat deposition in NAFLD
quantitative information of fat deposition in liver and has been shown to have a good correlation with both MRS [9– 13] and liver biopsy [14–16]. Apart from this, it has a potential for quantifying the fat fraction in other tissues. The relationship between NAFLD and its co-morbid diseases such as diabetes mellitus, hypertension, and osteoporosis remains unclear. A recent study found a correlation among pancreatic fat fraction, hepatic fat fraction, and liver biopsy determined hepatic steatosis [17]. Another study observed a correlation between insulin resistance and both pancreatic and liver fat content [18]. These findings documented a relationship between fat deposition of the organ and its disease. In the present study, we aimed to quantify deposition of fat in liver, pancreas, kidney, and vertebral bodies by MRI-PDFF and to assess the relationship among them.
Materials and methods Study population A total of 41 consecutive individuals diagnosed with NAFLD between June 2010 and April 2011 in our Liver Diseases Outpatient Clinic, who underwent PDFF calculation by IDEAL-IQ sequence for the quantification of hepatic steatosis were included in the study. This was a secondary analysis of a recent study [16]. In that study, PDFF was calculated from 10-mm-thick slices, but in the current study we only included subgroup of patients (n = 41) who also had PDFF maps with 5-mm slice thickness. Diagnosis of NAFLD in each case was based on biochemical, radiological, and histological criteria, and on exclusion of other forms of acute and chronic liver diseases. Criteria for inclusion were: (1) age >16 years; (2) evidence of absent or minimal alcohol consumption:
Abdominal Imaging
Abdom Imaging (2015) DOI: 10.1007/s00261-015-0385-0
Quantification of liver, pancreas, kidney, and vertebral body MRI-PDFF in non-alcoholic fatty liver disease Ilkay S. Idilman,1 Ali Tuzun,2 Berna Savas,3 Atilla Halil Elhan,4 Azim Celik,5 Ramazan Idilman,2 Musturay Karcaaltincaba1 1
Department of Radiology, Faculty of Medicine, Hacettepe University, 06100 Ankara, Turkey Department of Gastroenterology, Faculty of Medicine, Ankara University, Ankara, Turkey 3 Department of Pathology, Faculty of Medicine, Ankara University, Ankara, Turkey 4 Department of Biostatistics, Faculty of Medicine, Ankara University, Ankara, Turkey 5 General Electric Healthcare, Istanbul, Turkey 2
Abstract Purpose: The purpose of the present study was to determine liver, pancreas, kidney, and vertebral fat deposition in NAFLD patients by proton density fat fraction (PDFF) using magnetic resonance imaging (MRI) and to evaluate the relationships among them. Methods: A total of 41 biopsy-proven NAFLD patients underwent MRI-PDFF with IDEAL-IQ. MRI protocol included T1-independent volumetric multi-echo gradientecho imaging with T2* correction and spectral fat modeling. The MR examinations were performed on a 1.5 HDx MRI system. MRI-PDFF measurements were obtained from liver, pancreas, renal cortex and sinus, and vertebral body. Liver biopsy specimens were retrieved from the archives and evaluated by one pathologist according to NASH CRN. Results: The median age of the patients was 47 years. The median interval between liver biopsy and MRI examination was 16 days. Mean liver, pancreas, renal cortex, renal sinus, T12 and L1 vertebral body MRI-PDFFs were 18.7%, 5.7%, 1.7%, 51%, 43.2%, and 43.5%, respectively. No correlation between either liver MRI-PDFF or histological steatosis, and other organ MRI-PDFFs was observed. A good correlation between pancreas and vertebral body MRI-PDFFs, and pancreas and renal sinus MRI-PDFFs was observed. Diabetic patients had higher average pancreas MRI-PDFF compared to nondiabetics (12.2%, vs., 4.8%; P = 0.028). Conclusions: Pancreas and vertebral body MRI-PDFF is well correlated in NAFLD patients and both of them are Correspondence to: Musturay Karcaaltincaba; email: musturayk@yahoo. com
higher in diabetic patients which may explain increased bone fractures in diabetics. MRI-PDFF can be used to demonstrate fat fractions of different organs and tissues and to understand fat metabolism. Key words: Non-alcoholic fatty liver disease—MRIPDFF—Pancreas—Kidney—Vertebral body
Non-alcoholic fatty liver disease (NAFLD) is the excessive accumulation of fat in hepatocytes and affects approximately 30% of general population with an increasing rate [1, 2]. Although frequently benign, there are several clinical manifestations of this disease related to liver injury like non-alcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma [1–4]. Beside hepatic manifestations, associations with obesity, insulin resistance (IR), type 2 diabetes mellitus, hypertension, hyperlipidemia, and cardiovascular disease suggest complex relationships among them. Liver biopsy remains the reference method for both confirming the diagnosis of NAFLD and grading the disease. Potential risks of liver biopsy like bleeding, perforation, and even mortality [5] motivated the investigators to find a non-invasive method for diagnosis and grading of NAFLD. Magnetic resonance spectroscopy (MRS) is the most accepted method for quantitative analysis of the liver fat [6–8]. However, it is not available for routine clinical practice because of technical difficulties in both performance and evaluation [7, 9]. A recent MR technique—MRIestimated proton density fat fraction (MRI-PDFF)—gives
I. S. Idilman et al.: Quantification of ectopic fat deposition in NAFLD
quantitative information of fat deposition in liver and has been shown to have a good correlation with both MRS [9– 13] and liver biopsy [14–16]. Apart from this, it has a potential for quantifying the fat fraction in other tissues. The relationship between NAFLD and its co-morbid diseases such as diabetes mellitus, hypertension, and osteoporosis remains unclear. A recent study found a correlation among pancreatic fat fraction, hepatic fat fraction, and liver biopsy determined hepatic steatosis [17]. Another study observed a correlation between insulin resistance and both pancreatic and liver fat content [18]. These findings documented a relationship between fat deposition of the organ and its disease. In the present study, we aimed to quantify deposition of fat in liver, pancreas, kidney, and vertebral bodies by MRI-PDFF and to assess the relationship among them.
Materials and methods Study population A total of 41 consecutive individuals diagnosed with NAFLD between June 2010 and April 2011 in our Liver Diseases Outpatient Clinic, who underwent PDFF calculation by IDEAL-IQ sequence for the quantification of hepatic steatosis were included in the study. This was a secondary analysis of a recent study [16]. In that study, PDFF was calculated from 10-mm-thick slices, but in the current study we only included subgroup of patients (n = 41) who also had PDFF maps with 5-mm slice thickness. Diagnosis of NAFLD in each case was based on biochemical, radiological, and histological criteria, and on exclusion of other forms of acute and chronic liver diseases. Criteria for inclusion were: (1) age >16 years; (2) evidence of absent or minimal alcohol consumption:
