Scandinavian Journal of Gastroenterology. 2014; 49: 617–624
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Impact of sequential proton density fat fraction for quantification of hepatic steatosis in nonalcoholic fatty liver disease
ILKAY S. IDILMAN1, ONUR KESKIN2, ATILLA HALIL ELHAN3, RAMAZAN IDILMAN2* & MUSTURAY KARCAALTINCABA1* 1
Liver Imaging Team, Hacettepe University, School of Medicine, Department of Radiology, Ankara University, Faculty of Medicine, Ankara, Turkey, 2Department of Gastroenterology, Ankara, Turkey, and 3Department of Biostatistics, Ankara, Turkey
Abstract Objective. To determine the utility of sequential MRI-estimated proton density fat fraction (MRI-PDFF) for quantification of the longitudinal changes in liver fat content in individuals with nonalcoholic fatty liver disease (NAFLD). Methods. A total of 18 consecutive individuals (M/F: 10/8, mean age: 47.7 ± 9.8 years) diagnosed with NAFLD, who underwent sequential PDFF calculations for the quantification of hepatic steatosis at two different time points, were included in the study. All patients underwent T1-independent volumetric multi-echo gradient-echo imaging with T2* correction and spectral fat modeling. Results. A close correlation for quantification of hepatic steatosis between the initial MRI-PDFF and liver biopsy was observed (rs = 0.758, p < 0.001). The median interval between two sequential MRI-PDFF measurements was 184 days. From baseline to the end of the follow-up period, serum GGT level and homeostasis model assessment score were significantly improved (p = 0.015, p = 0.006, respectively), whereas BMI, serum AST, and ALT levels were slightly decreased. MRI-PDFFs were significantly improved (p = 0.004). A good correlation between two sequential MRI-PDFF calculations was observed (rs = 0.714, p = 0.001). With linear regression analyses, only delta serum ALT levels had a significant effect on delta MRI-PDFF calculations (r2 = 38.6%, p = 0.006). At least 5.9% improvement in MRI-PDFF is needed to achieve a normalized abnormal ALT level. The improvement of MRI-PDFF score was associated with the improvement of biochemical parameters in patients who had improvement in delta MRI-PDFF (p < 0.05). Conclusions. MRI-PDFF can be used for the quantification of the longitudinal changes of hepatic steatosis. The changes in serum ALT levels significantly reflected changes in MRI-PDFF in patients with NAFLD.
Key Words: follow-up, hepatic steatosis, MRI, nonalcoholic fatty liver disease, PDFF
Introduction Nonalcoholic fatty liver disease (NAFLD) is one of the most common causes of chronic liver disease with an increasing prevalence especially in developed countries [1–3]. At present, specific biochemical or serological tests for the diagnosis and follow-up of patients with NAFLD are not available. Liver biopsy remains the reference method for establishing the diagnosis and also evaluating the quantification of steatosis during follow-up period. However,
invasiveness of this procedure and risks for bleeding and perforation are undesirable factors [4]. In addition, sampling error and intra-inter-observer diagnostic variability have also been reported [4,5]. Therefore, a new noninvasive diagnostic modality is needed to evaluate and quantify hepatic steatosis as an alternative to liver biopsy. MRI modalities have been used for determination of hepatic fat content for several years [6,7]. Chemical shift-based water–fat separation methods include the approaches known as “Dixon” water–fat separation
Correspondence: Musturay Karcaaltıncaba, MD, Hacettepe University, Faculty of Medicine, Department of Radiology, Ankara, 06100, Turkey. E-mail: [email protected] *Musturay Karcaaltincaba has been supported by the Turkish Academy of Sciences (TUBA), in the framework of the Young Scientist Award Program (EA-TUBA-GEBIP/2011). Ramazan Idilman is an associate member of the TUBA.
(Received 17 December 2013; revised 1 February 2014; accepted 7 February 2014) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2014.894118
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
618
I. S. Idilman et al.
and advanced methods that rely on the phase shifts created by fat-water resonance frequency differences [8–10]. MRI-estimated proton density fat fraction (MRI-PDFF) is a recently described chemical shiftbased water–fat separation technique, which eliminates all major biases seen in conventional MRI methods and provides PDFF maps of the whole liver [11–15]. This technique has been shown to provide accurate quantification of hepatic fat content as compared to other MRI modalities [11–15] and liver biopsy [16–18]. However, there are very limited data regarding the usefulness of MRI-PDFF for the evaluation of the hepatic steatosis at the two different points, from baseline to the end of the follow-up period, in patients with NAFLD [19]. The aim of the present study was to determine the utility of sequential MRI-PDFF for quantification of the longitudinal changes in liver fat content in individuals with NAFLD. Methods Patients This is a longitudinal follow-up study. Between July 2010 and January 2011, a total of 18 individuals diagnosed with NAFLD, who underwent sequential PDFF measurements to quantify hepatic steatosis and evaluate changes in liver fat content, were included in the study. The diagnosis of NAFLD was based on a combination of biochemical, radiological and histological criteria, and on exclusion of other forms of acute and chronic liver diseases as previously published [18]. BMI was calculated as weight in kilograms divided by height in meters squared. Obesity was defined based on the World Health Organization criteria with BMI of 25–29.9 kg/m2 defining overweight and BMI ‡30 kg/m2 defining obesity [20]. Insulin resistance (IR) was calculated on the basis of fasting plasma glucose and insulin values using the homeostasis model assessment-insulin resistance method (HOMA-IR: plasma glucose (mg/dl) insulin (mu/ml)/405) [21]. Histological evaluation. All patients underwent liver biopsy for once at the initial diagnosis. All liver biopsy specimens were retrieved from the archives of the department of pathology. Biopsy specimens were evaluated by one experienced pathologist blinded to the clinical and biochemical data. Biopsies were scored by using the nonalcoholic steatohepatitis (NASH) Clinical Research Network NAFLD activity score and fibrosis score [22]. Significant fibrosis was defined as fibrosis stage 2–4.
MRI examinations. The MRI examinations were performed on a 1.5-T HDxt MRI system (GE Healthcare, Milwaukee, WI, USA). The subjects were examined in the supine position. A three-plane gradient echo localizer sequence was performed at the beginning of the examination. An eight-channel phased array body coil was used for acquisition. The MRI protocol included IDEAL-IQ (Iterative Decomposition of water and fat with Echo asymmetry and Least Square Estimation) sequence. This sequence is a three-dimensional (3D) volumetric imaging sequence used to create T2* and PDFF maps from a single breath hold acquisition noninvasively. The technique was used to estimate R2* (1/T2*) and PDFF (water-triglyceride fat separation) in the liver simultaneously in a single acquisition. Afterward, a correction was applied to resulting PDFF maps to correct for T2* effects. Six gradient echoes were applied to reconstruct water and triglyceride fat images, relative triglyceride fat fractions, and R2* maps. IDEAL-IQ sequence uses a novel “complex field map” to incorporate the T2* and field inhomogeneity effects into the overall multi-echo acquisition signal model. It has been shown by Yu et al. [23] that acquiring a six-echo image and estimating complex field map using an iterative least-square estimation algorithm, it is possible to achieve fat–water separation and T2* estimation in a single breath-hold acquisition. Imaging parameters. The parameters of this sequence were time of repetition (TR) = 12.9 ms, field of view (FOV) = 35–40 cm, Matrix = 224 160, 125 kHz Bandwidth, flip angle: 5 , and slice thickness 10 mm, and a single 3D slab with 22–28 slices was acquired. We acquired data sets with six different echoes ranging from 1.6 to 9.8 ms. A 2D selfcalibrated parallel-imaging technique called auto calibrating reconstruction of Cartesian sampling was used with an acceleration factor of 2. The images were processed using the software provided by the manufacturer to create water, fat, IP, OP, R2*, and fat fraction maps.
Image processing. By using a workstation (AW 4.4, GE Healthcare), a radiologist who was unaware of clinical data and biopsy results placed an elliptic region of interest of approximately 4 cm2 on the PDFF maps, in Couinaud segments 5–6, avoiding blood vessels, bile ducts, and artifacts in the baseline and approximately the same region in follow-up evaluation. Follow-up. After the diagnosis of NAFLD was confirmed, the management was focused in the following
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Sequential MRI-PDFF in NAFLD areas: establishment of an appropriate diet including a conventional diet of 25 kcal/kg ideal body weight (kg) with three meals per day containing 60% carbohydrate, 25% fat, and 15% protein and an exercise program including walking (initially as 300 steps per day for 3 days, thereafter adding 500 steps at 3-day intervals until a level of 10,000 steps was attained) and/or jogging (20 min twice a day) [24,25]; improvement in associated conditions such as diabetes mellitus, IR, and moderate/severe hyperlipidemia; and discontinuation of potentially hepatotoxic drugs such as herbal medicine. The dietitian interviewed all subjects, and the same diet and exercise program was suggested. The dietitian and one of the authors monitored patient compliance with the diet and exercise program during the follow-up period via verbal communication. All subjects were seen at the fourth week and at three- or six-month intervals thereafter in the outpatient clinic. All patients underwent one sequential MRI-PDFF in the follow-up period. Laboratory tests were performed during the follow-up period. The variability of repeat MRI-PDFF in the same patient was found to be less than 1% in a previous study [26]. Therefore, 1% or more change in MRI-PDFF in the follow-up MRI examination was determined as alteration in lipid fraction of the liver in our study [19]. Statistical analysis. Wilcoxon signed ranks test was used to compare the variables from baseline to the end of the follow-up period. Degree of association between variables was evaluated by Spearman’s correlation coefficient. Differences between two groups for baseline clinical and laboratory characteristics were evaluated by Mann–Whitney U test. A multiple regression analysis was performed to determine the independent variables that significantly predict the outcome variable. p Value less than 0.05 was considered significant. Results At the time of the initial imaging, the mean age was 47.7 ± 9.8 years (range 27–66 years), the median body weight and BMI were 80.6 kg (range: 55.1–102.2 kg) and 28.5 kg/m2 (range: 23.3–34.8 kg/m2), respectively; 33.3% of the patients (6/18) were obese. Sixty-one percent of the patients (11/18) had hyperlipidemia, 5.6% (1/18) hypertension, 5.6% (1/18) diabetes mellitus, and 78% (14/18) IR (HOMA score ‡ 2.7). Median serum AST, ALT, and GGT levels were 34 U/L (range: 20–76 U/L), 53.5 U/L (range: 20–102 U/L), and 60 (range: 17–113 U/L), respectively. Median HOMA score was 3.2 (range: 2–
619
12.1). According to liver biopsy examination, 38.9% (7/18) of the biopsy samples revealed grade 1 hepatic steatosis, 27.8% (5/18) grade 2 hepatic steatosis, and 33.3% (6/18) grade 3 hepatic steatosis. Necroinflammation, either lobular or portal, was graded as follows: 5.6% of the patients (1/18) had no necroinflammation, 88.9% (16/18) had grade 1, and 5.6% (1/18) had grade 2 necroinflammation. Of note, 33.3% of the samples (6/18) revealed the presence of fibrosis, and 5.6% of those (1/18) showed significant fibrosis (‡F2) (Table I). Mean initial MRI-PDFF was 18.8 ± 9.7% (median: 22.1%, range: 5.3–32.4%), and the mean percentage of histological steatosis was 47.2% ± 25.9% (median: 50%, range: 10–80%). The median interval between the liver biopsy and initial MRI-PDFF measurements was 3.5 days (range: 1–61 days). A close correlation in Table I. Baseline characteristics of 18 patients with NAFLD. All study patients Age 47.7 ± 9.8 (49.5) BMI 29 ± 3.4 (28.5) Obesity (%) 6 of 18 (33.3%) Diabetes mellitus (%) 1 of 18 (5.6%) Insulin resistance (%) 14 of 18 (78.9%) Hypertension (%) 1 of 18 (5.6%) Hyperlipidemia (%) 11 of 18 (61.1%) Fasting glucose (mg/dl) 83.7 ± 11 (80) Cholesterol (mg/dl) 206.2 ± 33.6 (200) Triglyceride (mg/dl) 179.7 ± 82.8 (179.5) HDL (mg/dl) 40.9 ± 9 (38) Serum AST level (U/L) 37.8 ± 14.8 (34) Serum ALT level (U/L) 54.8 ± 22.6 (53.5) Serum GGT level (U/L) 58.1 ± 27.8 (60) MRI-PDFF (%) 18.8 ± 9.7 (22.1) Percentage of hepatic steatosis (%) 47.2 ± 25.9 (50) Steatosis grade Grade 0 (66%) 6 of 18 (33.3%) Inflammation 0 (None) 1 of 18 (5.6%) 1 (4 foci) 0 Fibrosis F0 (None) 12 of 18 (66.7%) F1 (Perisinusoidal or periportal) 5 of 18 (27.8%) F2 (Perisinusoidal and portal/periportal) 1 of 18 (5.6%) F3 (Bridging fibrosis) 0 Abbreviations: ALT = Alanine aminotransferase, AST = Aspartate aminotransferase, GGT = Gamma glutamyl transpeptidase, HDL = High-density lipoprotein. Data are mean ± standard deviation, with the median in parentheses. Lobular inflammation was evaluated at 200 optical filed as 0, no inflammation; 1, less than 2 foci; 2, 2–4 foci; and 3, more than 4 foci of inflammation. Fibrosis was defined as 0, none; 1, perisinusoidal or periportal; 2, perisinusoidal and portal/periportal; 3, bridging fibrosis.
620
I. S. Idilman et al. 35 30
MRI-PDFF (%)
25 20 15 rs = 0.758 p < 0.001
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
10 5 0 0
10
20
30
40
50
60
70
80
90
Steatotic hepatocytes (%) Figure 1. Correlation between baseline MRI-PDFF and liver biopsy.
terms of the quantification of the percentage of steatotic hepatocytes was observed between the initial MRI-PDFF calculation and liver biopsy (rs = 0.758, p < 0.001) (Figure 1). The correlation was better in patients without hepatic fibrosis (n = 12, rs = 0.825, p = 0.001). From baseline to the end of the follow-up period, 11 patients (61.1%) had weight loss, 6 patients (33.3%) had weight gain, and 1 patient (5.6%) had a stable weight. Median body weight and BMI were slightly decreased (80.6 kg vs. 78.7 kg and 28.5 kg/m2 vs. 27.8 kg/m2, respectively, p = 0.50). Serum GGT level and HOMA score were significantly improved (median 60 U/L vs. 42.5 U/L, p = 0.015 and median: 3.2 vs. 2.1, p = 0.006, respectively), whereas serum AST and ALT levels were slightly decreased (median 34 U/L vs. 27 U/L, p = 0.256 and median 53.5 U/L vs. 40 U/L, p = 0.266, respectively) (Table II). The median interval between two sequential PDFF measurements was 184 days (range: 115–
274 days). Mean MRI-PDFF at the follow-up was 12.8% ± 10.5% (median: 8.9%, range: 1.6–35.5%). There was a good correlation between two sequential MRI-PDFF calculations (rs = 0.714, p = 0.001) (Figure 2). From baseline to the end of the followup period, MRI-PDFFs were significantly improved (median 22.1% vs. 8.9%, p = 0.004). Thirteen patients (72.2%) had a decrease in liver steatosis, 2 (11.1%) had an increase, and 3 (16.6%) had no alteration in MRI-PDFF (Figure 3). With linear regression analyses, only delta serum ALT levels had a significant correlation on delta PDFF calculations (r2 = 38.6%, p = 0.006) (Table III). Patients were divided into two distinct groups according to the alteration in the sequential MRIPDFF. Group 1 (n = 13) was defined as patients with decreased MRI-PDFF from baseline to the end of the follow-up period and group 2 as patients with stable and increased MRI-PDFF. In group 1 patients, at the time of the sequential imaging, nine patients had an
Table II. Comparison of baseline and follow-up patient characteristic of the whole cohort. Baseline BMI HOMA score Fasting glucose (N:75–115 mg/dl) Cholesterol (N:120–200 mg/dl) Triglyceride (N:40–165 mg/dl) HDL (N:40–60 mg/dl) Serum AST level (N:10–37 U/L) Serum ALT level (N:10–37 U/L) Serum GGT level (N:0–55 U/L) MRI-PDFF (%)
29 4.3 83.7 206.2 179.7 40.9 37.8 54.8 58.1 18.8
± ± ± ± ± ± ± ± ± ±
3.4 (28.5) 2.6 (3.2) 11 (80) 33.6 (200) 82.8 (179.5) 9 (38) 14.8 (34) 22.6 (53.5) 27.8 (60) 9.7 (22.1)
Data are mean ± standard deviation, with the median in parenthesis.
Follow up 29.2 2.5 80.7 200.7 161.8 41.8 32.2 46.7 43.2 12.8
± ± ± ± ± ± ± ± ± ±
3.7 (28.2) 0.9 (2.1) 8.5 (80.5) 36.5 (193.5) 101.7 (124) 6.9 (42) 17 (27) 28.7 (40) 28.1 (42.5) 10.5 (8.9)
p 0.433 0.006 0.368 0.740 0.136 0.470 0.256 0.266 0.015 0.004
Sequential MRI-PDFF in NAFLD
621
40
MRI-PDFF on follow-up (%)
35 30 25 20 rs = 0.714 p = 0.001
15 10
0 0
5
10
15
20
25
30
35
40
MRI-PDFF at baseline (%) Figure 2. Correlation between two sequential MRI-PDFFs.
improvement in body weight. The improvement of MRI-PDFF score was associated with the improvement of clinical and biochemical parameters; BMI (p = 0.075), HOMA scores (p = 0.003), serum AST (p = 0.023), ALT (p = 0.016), and GGT (p = 0.019) levels were improved. Median serum ALT levels were decreased from 55 to 34 U/L. Ten patients had an improvement in their abnormal serum ALT levels; six patients’ ALT levels were normalized with a median MRI-PDFF improvement of 11.9% (range 2.1– 23.6%). With linear regression analyses, at least 5.9% improvement in MRI-PDFF is needed to
23%
On 120th day of follow-up BMI: 25.7
12%
In-phase
PDFF
At baseline BMI: 27.3
achieve a normalized abnormal ALT level. Of note, serum ALT level was increased in only one patient in Group 1. Five of the 13 patients had a decreased MRI-PDFF more than 50%. These patients had lost a median weight of 4.7 kg (range: 2.4–6.6 kg): four had normal serum ALT levels. Of note, five had a decreased MRI-PDFF below 5%, with a weight loss of a median 4.8 kg (range: 0–8.6 kg) in group 1. There was also a good correlation in terms of the delta MRI-PDFF differences and delta serum AST (rs = 0.691, p = 0.002) and ALT levels (rs = 0.608, p = 0.007) in group 1 patients. We also compared the
Opposed-phase
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
5
Figure 3. MRI-PDFF calculations at days 0 and 120 of a patient who had also a decrease in BMI show reduction of hepatic steatosis from 23% to 12%. The decrease in fat fraction can also be realized on in- and opposed-phase MR images.
622
I. S. Idilman et al.
Table III. Association between delta MRI-PDFF and delta changes with linear regression analyses. p
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
DBMI DHOMA score DSerum AST level DSerum ALT level DSerum GGT level
0.644 0.190 0.622 0.006 0.964
changes in clinical and biochemical parameters and MRI-PDFF between Group 1 and Group 2, and there were significant differences in delta serum AST (p = 0.004), ALT (0.004), triglyceride (p = 0.046) levels, and delta MRI-PDFF (p < 0.001) (Table IV). Discussion In the present study, we evaluated the utility of MRIPDFF measurements to quantify the liver fat content cross-sectionally and longitudinally in patients with NAFLD. At baseline, MRI-PDFF measurement had good diagnostic accuracy for quantifying liver fat content compared with liver biopsy results (rs = 0.758, p < 0.001) as previously demonstrated by several investigators [16–18]. The correlation was better in patients without hepatic fibrosis (rs = 0.825, p = 0.001). Median interval between two sequential MRI-PDFF measurements was 184 days. From baseline to the end of the follow-up period, MRI-PDFFs were significantly improved. A close correlation between two sequential MRI-PDFF calculations was also observed (rs = 0.714, p = 0.001). The findings are consistent with previous study, demonstrating that liver MRI-estimated PDFF strongly correlated with magnetic resonance spectroscopy (MRS)-measured liver fat content both at baseline and at week 24 in patients with NASH treated with colesevelam [19]. MRS has been shown to accurately measure hepatic
lipid fraction in previous studies [27,28]. However, it has limited clinical application, because it is not widely available on all clinical MR systems and analyses require specific software. MRI-PDFF, a chemical shift-based water–fat separation method, is a novel MRI technique, which addresses confounding factors for accurate lipid fraction quantification. MRI-PDFF permits quantification of lipid fraction of the whole liver, which is a limitation for MRS, as well as liver biopsy. These results indicate that MRI-PDFF had good diagnostic accuracy for quantifying liver fat content at both cross-sectional and also longitudinal evaluation. The changes in clinical and biochemical parameters from baseline to the end of the follow-up period after an appropriate diet, and an exercise program were also evaluated in the present study. There is no proven beneficial therapy in NAFLD cases. Several investigators have evaluated whether or not diet and lifestyle modification with/without insulin sensitizers or UDCA or antioxidant therapy can improve the metabolic, biochemical, and histological abnormalities in such individuals [29–34]. An appropriate diet and exercise is currently the most accepted first-line therapy for patients with NAFLD and the goal of this approach is to lose 7–10% of their body weight in 6– 12 months. From baseline to the end of the follow-up period, 61% of the patients had weight loss, 33% had weight gain, and 6% had a stable weight in the present study. Sequential MRI-PDFFs were significantly improved (p = 0.004); 72% of the patients had a decrease in liver steatosis, 11% had an increase, and 17% had no change. The investigators recently reported that liver MRI-estimated PDFF measured changes in liver fat content correlated with changes in body weight, serum aminotransferases, and GGT levels in NASH patients [19]. In the present study, in contrast to previous study [19], with linear regression analysis, the changes in serum ALT levels
Table IV. The comparison of the delta differences in clinical, laboratory parameters, and MRI-PDFF between subgroups according to MRIPDFF alterations in the follow up. Group 1 (Decreased MRI-PDFF, n = 13) DBMI DHOMA score DFasting Glucose DCholesterol DTriglyceride DHDL DAST DALT DGGT DMRI-PDFF (%)
–0.5 –1.1 –4.6 –7.6 –31.1 2 –12.2 –19.4 –21 –8.8
(1) (1) (13.2) (46.7) (72.5) (6.6) (22.6) (25.6) (34.1) (7.1)
Data are mean with the standard deviation in parenthesis.
Group 2 (Stable or increased MRI-PDFF, n = 5) 1.6 –3.4 1 0 16.4 –2 11.6 21.4 0.8 1.5
(4.3) (4.8) (5.5) (31.8) (37.8) (11.5) (14.6) (19.4) (13.1) (2.3)
p 0.208 0.566 0.443 0.633 0.046 0.503 0.004 0.004 0.019 < 0.001
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Sequential MRI-PDFF in NAFLD significantly reflected changes in MRI-PDFF calculations (r2 = 38.6%, p = 0.006) in NAFLD patients. Serum aminotransferase levels have been investigated at diagnosis and also follow-up evaluation of NAFLD. The changes in serum aminotransferases levels in the follow-up evaluation are confounding. A study, which investigated the follow-up changes in liver biopsy, reported a decrease in serum aminotransferase levels paralleling improvement in steatosis and inflammation but not changes in fibrosis stage [35]; however, another study reported disease progression despite a reduction in ALT levels more than 25% [36]. In the subgroup analysis, from baseline to the end of the follow-up period, among 13 patients (Group 1), 9 had an improvement in body weight and 10 had an improvement in abnormal serum ALT levels. The improvement of MRI-PDFF score was associated with the improvement of biochemical parameters including HOMA scores (p = 0.003), serum AST (p = 0.023), ALT (p = 0.016), and GGT (p = 0.019) levels (p < 0.05). Changes in MRI-PDFF correlated with changes in serum AST and ALT levels (p = 0.002, p = 0.007, respectively). There were significant differences in delta serum AST, ALT, triglyceride levels, and MRI-PDFF between Group 1 and Group 2. In conclusion, based on the result of the present longitudinal study, MRI-PDFF can be used for quantification of the longitudinal changes of hepatic steatosis. The changes in serum ALT levels reflected changes in MRI-PDFF in patients with NAFLD. Acknowledgments None of the authors have relevant conflicts of interest to disclose. All authors agree with the manuscripts submission and its content. ISI collected data, conceived and carried out experiments, OK collected data, AHE analyzed data, RI and MK conceived and carried out experiments. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 2002;346:1221–31. [2] Charlton M. Nonalcoholic fatty liver disease: a review of current understanding and future impact. Clin Gastroenterol Hepatol 2004;2:1048–58. [3] Vuppalanchi R, Chalasani N. Nonalcoholic fatty liver disease andnonalcoholicsteatohepatitis:selectedpracticalissuesintheir evaluation and management. Hepatology 2009;49:306–17.
623
[4] Bravo A, Sheth S, Chopra S. Liver biopsy. N Engl J Med 2001;344:495–500. [5] Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, et al. Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology 2005; 128:1898–906. [6] Qayyum A, Goh JS, Kakar S, Yeh BM, Merriman RB, Coakley FV. Accuracy of liver fat quantification at MR imaging: comparison of out-of-phase gradient echo and fat-saturated fast spin-echo techniques – initial experience. Radiology 2005;237:507–11. [7] Pilleul F, Chave G, Dumortier J, Scoazec JY, Valette PJ. Fatty infiltration of the liver: detection and grading using dual T1 gradient echo sequences on clinical MR system. Gastroenterol Clin Biol 2005;29:1143–7. [8] Dixon WT. Simple proton spectroscopic imaging. Radiology 1984;153:189–94. [9] Glover GH. Multipoint Dixon technique for water and fat proton and susceptibility imaging. J Magn Reson Imaging 1991;5:521–30. [10] Bley TA, Wieben O, François CJ, Brittain JH, Reeder SB. Fat and water magnetic resonance imaging. J Magn Reson Imaging 2010;31:4–18. [11] Reeder SB, Sirlin CB. Quantification of liver fat with magnetic resonance imaging. Magn Reson Imaging Clin N Am 2010;18:337–57. [12] Reeder SB, Cruite I, Hamilton G, Sirlin CB. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J Magn Reson Imaging 2011;34:729–49. [13] Reeder SB, Robson PM, Yu H, Schimakawa A, Hines CD, McKenzie CA, et al. Quantification of hepatic steatosis with MRI: the effects of accurate fat spectral modeling. J Magn Reson Imaging 2009;29:1332–9. [14] Yokoo T, Bydder M, Hamilton G, Middelton MS, Gamst AC, Wolfson T, et al. Nonalcoholic fatty liver disease: diagnostic fat-grading accuracy of low-flip-angle multiecho gradient-recalled-echo MR imaging at 1.5 T. Radiology 2009;251:67–76. [15] Kim H, Taksali SE, Dufour S, Befroy D, Goodman TR, Petersen KF, et al. Comparative MR study of hepatic fat quantification using single voxel proton spectroscopy, two point Dixon and three-point IDEAL. Magn Reson Med 2008;59:521–7. [16] Permutt Z, Le TA, Peterson MR, Seki E, Brenner DA, Sirlin C, et al. Correlation between liver histology and novel magnetic resonance imaging in adult patients with nonalcoholic fatty liver disease – MRI accurately quantifies hepatic steatosis in NAFLD. Aliment Pharmacol Ther 2012;36:22–9. [17] Tang A, Tan J, Sun M, Hamilton G, Bydder M, Wolfson T, et al. Nonalcoholic fatty liver disease: MR imaging of liver proton density fat fraction to assess hepatic steatosis. Radiology 2013;267:422–31. [18] Idilman IS, Aniktar H, Idilman R, Kabacam G, Savas B, Elhan A, et al. Hepatic steatosis: quantification by proton density fat fraction with MR imaging versus liver biopsy. Radiology 2013;267:767–75. [19] Noureddin M, Lam J, Peterson MR, Middleton M, Hamilton G, Le TA, et al. Longitudinal comparison between MRI, MRS and histology-determined steatosis in NAFLD patients at two-time points in a randomized trial. Hepatology 2013;58:1930–40. [20] World Health Organization. Report of a WHO consultation: definition of metabolic syndrome in definition, diagnosis and classification of diabetes mellitus. Geneva: World Health
624
[21]
[22]
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
[23]
[24]
[25]
[26]
[27]
[28]
I. S. Idilman et al. organization, Department of Noncommunicable Disease Surveillance; 1999. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–19. Kleiner DE, Brunt EM, Van Natta ML, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41:1313–21. Yu H, McKenzie CA, Shimakawa A, Vu AT, Brau AC, Beattu BJ, et al. Multiecho reconstruction for simultaneous water-fat decomposition and T2* estimation. J Magn Reson Imaging 2007;26:1153–61. Ueno T, Sugawara H, Sujaku K, Hashimoto O, Tsuji R, Tamaki S, et al. Therapeutic effects of restricted diet and exercise in obese patients with fatty liver. J Hepatol 1997;27: 103–7. Kukkonen K, Rauramaa R, Siitonen O, Hänninen O. Physical training of obese middle-aged persons. Ann Clin Res 1982;34:80–5. Bydder M, Shiehmorteza M, Yokoo T, Sugay S, Middleton MS, Girard O, et al. Assessment of liver fat quantification in the presence of iron. Magn Reson Imaging 2010;28:767–76. Longo R, Pollesello P, Ricci C, Masutti F, Kvam BJ, Bercich L, et al. Proton MR spectroscopy in quantitative in vivo determination of fat content in human liver steatosis. J Magn Reson Imaging 1995;5:281–5. Cho SG, Kim MY, Kim HJ, Kim YS, Choi W, Shin SH, et al. Chronic hepatitis: in vivo proton MR spectroscopic evaluation of the liver and correlation with histopathologic findings. Radiology 2001;221:740–6.
[29] Neuschwander-Tetri BA, Brunt EM, Wehmeier KR, Oliver D, Bacon BR. Improved non-alcoholic steatohepatitis after 48 weeks of treatment with the PPAR-gamma ligand rosiglitazone. Hepatology 2003;38:1008–17. [30] Belfort R, Harrison SA, Brown K, Darland C, Finch J, Hardies J, et al. A placebo-controlled trial of pioglitazone in subjects with non-alcoholic steatohepatitis. N Engl J Med 2006;355:2297–307. [31] Uygun A, Kadayifci A, Isik AT, Ozgurtas T, Deveci S, Tuzun A, et al. Metformin in the treatment of patients with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2004;19:537–44. [32] Idilman R, Mizrak D, Corapcioglu D, Bektas M, Doganay B, Sayki M, et al. Clinical trial: insulin-sensitizing agents may reduce consequences of insulin resistance in individuals with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2008;28:200–8. [33] Nobili V, Manco M, Devito R, Di Ciommo V, Comparcola D, Sartorelli MR, et al. Lifestyle intervention and antioxidant therapy in children with nonalcoholic fatty liver disease: a randomized, controlled trial. Hepatology 2008;48:119–28. [34] Promrat K, Kleiner DE, Niemeier HM, Jackvony E, Kearns M, Wands JR, et al. Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis. Hepatology 2010;51:121–9. [35] Adams LA, Sanderson S, Lindor KD, Angulo P. The histological course of nonalcoholic fatty liver disease: a longitudinal study of 103 patients with sequential liver biopsies. J Hepatol 2005;42:132–8. [36] Pais R, Charlotte F, Fedchuk L, Bedossa P, Lebray P, Poynard T, et al. A systematic review of follow-up biopsies reveals disease progression in patients with non-alcoholic fatty liver. J Hepatol 2013;59:550–6.
ORIGINAL ARTICLE
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Impact of sequential proton density fat fraction for quantification of hepatic steatosis in nonalcoholic fatty liver disease
ILKAY S. IDILMAN1, ONUR KESKIN2, ATILLA HALIL ELHAN3, RAMAZAN IDILMAN2* & MUSTURAY KARCAALTINCABA1* 1
Liver Imaging Team, Hacettepe University, School of Medicine, Department of Radiology, Ankara University, Faculty of Medicine, Ankara, Turkey, 2Department of Gastroenterology, Ankara, Turkey, and 3Department of Biostatistics, Ankara, Turkey
Abstract Objective. To determine the utility of sequential MRI-estimated proton density fat fraction (MRI-PDFF) for quantification of the longitudinal changes in liver fat content in individuals with nonalcoholic fatty liver disease (NAFLD). Methods. A total of 18 consecutive individuals (M/F: 10/8, mean age: 47.7 ± 9.8 years) diagnosed with NAFLD, who underwent sequential PDFF calculations for the quantification of hepatic steatosis at two different time points, were included in the study. All patients underwent T1-independent volumetric multi-echo gradient-echo imaging with T2* correction and spectral fat modeling. Results. A close correlation for quantification of hepatic steatosis between the initial MRI-PDFF and liver biopsy was observed (rs = 0.758, p < 0.001). The median interval between two sequential MRI-PDFF measurements was 184 days. From baseline to the end of the follow-up period, serum GGT level and homeostasis model assessment score were significantly improved (p = 0.015, p = 0.006, respectively), whereas BMI, serum AST, and ALT levels were slightly decreased. MRI-PDFFs were significantly improved (p = 0.004). A good correlation between two sequential MRI-PDFF calculations was observed (rs = 0.714, p = 0.001). With linear regression analyses, only delta serum ALT levels had a significant effect on delta MRI-PDFF calculations (r2 = 38.6%, p = 0.006). At least 5.9% improvement in MRI-PDFF is needed to achieve a normalized abnormal ALT level. The improvement of MRI-PDFF score was associated with the improvement of biochemical parameters in patients who had improvement in delta MRI-PDFF (p < 0.05). Conclusions. MRI-PDFF can be used for the quantification of the longitudinal changes of hepatic steatosis. The changes in serum ALT levels significantly reflected changes in MRI-PDFF in patients with NAFLD.
Key Words: follow-up, hepatic steatosis, MRI, nonalcoholic fatty liver disease, PDFF
Introduction Nonalcoholic fatty liver disease (NAFLD) is one of the most common causes of chronic liver disease with an increasing prevalence especially in developed countries [1–3]. At present, specific biochemical or serological tests for the diagnosis and follow-up of patients with NAFLD are not available. Liver biopsy remains the reference method for establishing the diagnosis and also evaluating the quantification of steatosis during follow-up period. However,
invasiveness of this procedure and risks for bleeding and perforation are undesirable factors [4]. In addition, sampling error and intra-inter-observer diagnostic variability have also been reported [4,5]. Therefore, a new noninvasive diagnostic modality is needed to evaluate and quantify hepatic steatosis as an alternative to liver biopsy. MRI modalities have been used for determination of hepatic fat content for several years [6,7]. Chemical shift-based water–fat separation methods include the approaches known as “Dixon” water–fat separation
Correspondence: Musturay Karcaaltıncaba, MD, Hacettepe University, Faculty of Medicine, Department of Radiology, Ankara, 06100, Turkey. E-mail: [email protected] *Musturay Karcaaltincaba has been supported by the Turkish Academy of Sciences (TUBA), in the framework of the Young Scientist Award Program (EA-TUBA-GEBIP/2011). Ramazan Idilman is an associate member of the TUBA.
(Received 17 December 2013; revised 1 February 2014; accepted 7 February 2014) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2014.894118
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
618
I. S. Idilman et al.
and advanced methods that rely on the phase shifts created by fat-water resonance frequency differences [8–10]. MRI-estimated proton density fat fraction (MRI-PDFF) is a recently described chemical shiftbased water–fat separation technique, which eliminates all major biases seen in conventional MRI methods and provides PDFF maps of the whole liver [11–15]. This technique has been shown to provide accurate quantification of hepatic fat content as compared to other MRI modalities [11–15] and liver biopsy [16–18]. However, there are very limited data regarding the usefulness of MRI-PDFF for the evaluation of the hepatic steatosis at the two different points, from baseline to the end of the follow-up period, in patients with NAFLD [19]. The aim of the present study was to determine the utility of sequential MRI-PDFF for quantification of the longitudinal changes in liver fat content in individuals with NAFLD. Methods Patients This is a longitudinal follow-up study. Between July 2010 and January 2011, a total of 18 individuals diagnosed with NAFLD, who underwent sequential PDFF measurements to quantify hepatic steatosis and evaluate changes in liver fat content, were included in the study. The diagnosis of NAFLD was based on a combination of biochemical, radiological and histological criteria, and on exclusion of other forms of acute and chronic liver diseases as previously published [18]. BMI was calculated as weight in kilograms divided by height in meters squared. Obesity was defined based on the World Health Organization criteria with BMI of 25–29.9 kg/m2 defining overweight and BMI ‡30 kg/m2 defining obesity [20]. Insulin resistance (IR) was calculated on the basis of fasting plasma glucose and insulin values using the homeostasis model assessment-insulin resistance method (HOMA-IR: plasma glucose (mg/dl) insulin (mu/ml)/405) [21]. Histological evaluation. All patients underwent liver biopsy for once at the initial diagnosis. All liver biopsy specimens were retrieved from the archives of the department of pathology. Biopsy specimens were evaluated by one experienced pathologist blinded to the clinical and biochemical data. Biopsies were scored by using the nonalcoholic steatohepatitis (NASH) Clinical Research Network NAFLD activity score and fibrosis score [22]. Significant fibrosis was defined as fibrosis stage 2–4.
MRI examinations. The MRI examinations were performed on a 1.5-T HDxt MRI system (GE Healthcare, Milwaukee, WI, USA). The subjects were examined in the supine position. A three-plane gradient echo localizer sequence was performed at the beginning of the examination. An eight-channel phased array body coil was used for acquisition. The MRI protocol included IDEAL-IQ (Iterative Decomposition of water and fat with Echo asymmetry and Least Square Estimation) sequence. This sequence is a three-dimensional (3D) volumetric imaging sequence used to create T2* and PDFF maps from a single breath hold acquisition noninvasively. The technique was used to estimate R2* (1/T2*) and PDFF (water-triglyceride fat separation) in the liver simultaneously in a single acquisition. Afterward, a correction was applied to resulting PDFF maps to correct for T2* effects. Six gradient echoes were applied to reconstruct water and triglyceride fat images, relative triglyceride fat fractions, and R2* maps. IDEAL-IQ sequence uses a novel “complex field map” to incorporate the T2* and field inhomogeneity effects into the overall multi-echo acquisition signal model. It has been shown by Yu et al. [23] that acquiring a six-echo image and estimating complex field map using an iterative least-square estimation algorithm, it is possible to achieve fat–water separation and T2* estimation in a single breath-hold acquisition. Imaging parameters. The parameters of this sequence were time of repetition (TR) = 12.9 ms, field of view (FOV) = 35–40 cm, Matrix = 224 160, 125 kHz Bandwidth, flip angle: 5 , and slice thickness 10 mm, and a single 3D slab with 22–28 slices was acquired. We acquired data sets with six different echoes ranging from 1.6 to 9.8 ms. A 2D selfcalibrated parallel-imaging technique called auto calibrating reconstruction of Cartesian sampling was used with an acceleration factor of 2. The images were processed using the software provided by the manufacturer to create water, fat, IP, OP, R2*, and fat fraction maps.
Image processing. By using a workstation (AW 4.4, GE Healthcare), a radiologist who was unaware of clinical data and biopsy results placed an elliptic region of interest of approximately 4 cm2 on the PDFF maps, in Couinaud segments 5–6, avoiding blood vessels, bile ducts, and artifacts in the baseline and approximately the same region in follow-up evaluation. Follow-up. After the diagnosis of NAFLD was confirmed, the management was focused in the following
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Sequential MRI-PDFF in NAFLD areas: establishment of an appropriate diet including a conventional diet of 25 kcal/kg ideal body weight (kg) with three meals per day containing 60% carbohydrate, 25% fat, and 15% protein and an exercise program including walking (initially as 300 steps per day for 3 days, thereafter adding 500 steps at 3-day intervals until a level of 10,000 steps was attained) and/or jogging (20 min twice a day) [24,25]; improvement in associated conditions such as diabetes mellitus, IR, and moderate/severe hyperlipidemia; and discontinuation of potentially hepatotoxic drugs such as herbal medicine. The dietitian interviewed all subjects, and the same diet and exercise program was suggested. The dietitian and one of the authors monitored patient compliance with the diet and exercise program during the follow-up period via verbal communication. All subjects were seen at the fourth week and at three- or six-month intervals thereafter in the outpatient clinic. All patients underwent one sequential MRI-PDFF in the follow-up period. Laboratory tests were performed during the follow-up period. The variability of repeat MRI-PDFF in the same patient was found to be less than 1% in a previous study [26]. Therefore, 1% or more change in MRI-PDFF in the follow-up MRI examination was determined as alteration in lipid fraction of the liver in our study [19]. Statistical analysis. Wilcoxon signed ranks test was used to compare the variables from baseline to the end of the follow-up period. Degree of association between variables was evaluated by Spearman’s correlation coefficient. Differences between two groups for baseline clinical and laboratory characteristics were evaluated by Mann–Whitney U test. A multiple regression analysis was performed to determine the independent variables that significantly predict the outcome variable. p Value less than 0.05 was considered significant. Results At the time of the initial imaging, the mean age was 47.7 ± 9.8 years (range 27–66 years), the median body weight and BMI were 80.6 kg (range: 55.1–102.2 kg) and 28.5 kg/m2 (range: 23.3–34.8 kg/m2), respectively; 33.3% of the patients (6/18) were obese. Sixty-one percent of the patients (11/18) had hyperlipidemia, 5.6% (1/18) hypertension, 5.6% (1/18) diabetes mellitus, and 78% (14/18) IR (HOMA score ‡ 2.7). Median serum AST, ALT, and GGT levels were 34 U/L (range: 20–76 U/L), 53.5 U/L (range: 20–102 U/L), and 60 (range: 17–113 U/L), respectively. Median HOMA score was 3.2 (range: 2–
619
12.1). According to liver biopsy examination, 38.9% (7/18) of the biopsy samples revealed grade 1 hepatic steatosis, 27.8% (5/18) grade 2 hepatic steatosis, and 33.3% (6/18) grade 3 hepatic steatosis. Necroinflammation, either lobular or portal, was graded as follows: 5.6% of the patients (1/18) had no necroinflammation, 88.9% (16/18) had grade 1, and 5.6% (1/18) had grade 2 necroinflammation. Of note, 33.3% of the samples (6/18) revealed the presence of fibrosis, and 5.6% of those (1/18) showed significant fibrosis (‡F2) (Table I). Mean initial MRI-PDFF was 18.8 ± 9.7% (median: 22.1%, range: 5.3–32.4%), and the mean percentage of histological steatosis was 47.2% ± 25.9% (median: 50%, range: 10–80%). The median interval between the liver biopsy and initial MRI-PDFF measurements was 3.5 days (range: 1–61 days). A close correlation in Table I. Baseline characteristics of 18 patients with NAFLD. All study patients Age 47.7 ± 9.8 (49.5) BMI 29 ± 3.4 (28.5) Obesity (%) 6 of 18 (33.3%) Diabetes mellitus (%) 1 of 18 (5.6%) Insulin resistance (%) 14 of 18 (78.9%) Hypertension (%) 1 of 18 (5.6%) Hyperlipidemia (%) 11 of 18 (61.1%) Fasting glucose (mg/dl) 83.7 ± 11 (80) Cholesterol (mg/dl) 206.2 ± 33.6 (200) Triglyceride (mg/dl) 179.7 ± 82.8 (179.5) HDL (mg/dl) 40.9 ± 9 (38) Serum AST level (U/L) 37.8 ± 14.8 (34) Serum ALT level (U/L) 54.8 ± 22.6 (53.5) Serum GGT level (U/L) 58.1 ± 27.8 (60) MRI-PDFF (%) 18.8 ± 9.7 (22.1) Percentage of hepatic steatosis (%) 47.2 ± 25.9 (50) Steatosis grade Grade 0 (66%) 6 of 18 (33.3%) Inflammation 0 (None) 1 of 18 (5.6%) 1 (4 foci) 0 Fibrosis F0 (None) 12 of 18 (66.7%) F1 (Perisinusoidal or periportal) 5 of 18 (27.8%) F2 (Perisinusoidal and portal/periportal) 1 of 18 (5.6%) F3 (Bridging fibrosis) 0 Abbreviations: ALT = Alanine aminotransferase, AST = Aspartate aminotransferase, GGT = Gamma glutamyl transpeptidase, HDL = High-density lipoprotein. Data are mean ± standard deviation, with the median in parentheses. Lobular inflammation was evaluated at 200 optical filed as 0, no inflammation; 1, less than 2 foci; 2, 2–4 foci; and 3, more than 4 foci of inflammation. Fibrosis was defined as 0, none; 1, perisinusoidal or periportal; 2, perisinusoidal and portal/periportal; 3, bridging fibrosis.
620
I. S. Idilman et al. 35 30
MRI-PDFF (%)
25 20 15 rs = 0.758 p < 0.001
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
10 5 0 0
10
20
30
40
50
60
70
80
90
Steatotic hepatocytes (%) Figure 1. Correlation between baseline MRI-PDFF and liver biopsy.
terms of the quantification of the percentage of steatotic hepatocytes was observed between the initial MRI-PDFF calculation and liver biopsy (rs = 0.758, p < 0.001) (Figure 1). The correlation was better in patients without hepatic fibrosis (n = 12, rs = 0.825, p = 0.001). From baseline to the end of the follow-up period, 11 patients (61.1%) had weight loss, 6 patients (33.3%) had weight gain, and 1 patient (5.6%) had a stable weight. Median body weight and BMI were slightly decreased (80.6 kg vs. 78.7 kg and 28.5 kg/m2 vs. 27.8 kg/m2, respectively, p = 0.50). Serum GGT level and HOMA score were significantly improved (median 60 U/L vs. 42.5 U/L, p = 0.015 and median: 3.2 vs. 2.1, p = 0.006, respectively), whereas serum AST and ALT levels were slightly decreased (median 34 U/L vs. 27 U/L, p = 0.256 and median 53.5 U/L vs. 40 U/L, p = 0.266, respectively) (Table II). The median interval between two sequential PDFF measurements was 184 days (range: 115–
274 days). Mean MRI-PDFF at the follow-up was 12.8% ± 10.5% (median: 8.9%, range: 1.6–35.5%). There was a good correlation between two sequential MRI-PDFF calculations (rs = 0.714, p = 0.001) (Figure 2). From baseline to the end of the followup period, MRI-PDFFs were significantly improved (median 22.1% vs. 8.9%, p = 0.004). Thirteen patients (72.2%) had a decrease in liver steatosis, 2 (11.1%) had an increase, and 3 (16.6%) had no alteration in MRI-PDFF (Figure 3). With linear regression analyses, only delta serum ALT levels had a significant correlation on delta PDFF calculations (r2 = 38.6%, p = 0.006) (Table III). Patients were divided into two distinct groups according to the alteration in the sequential MRIPDFF. Group 1 (n = 13) was defined as patients with decreased MRI-PDFF from baseline to the end of the follow-up period and group 2 as patients with stable and increased MRI-PDFF. In group 1 patients, at the time of the sequential imaging, nine patients had an
Table II. Comparison of baseline and follow-up patient characteristic of the whole cohort. Baseline BMI HOMA score Fasting glucose (N:75–115 mg/dl) Cholesterol (N:120–200 mg/dl) Triglyceride (N:40–165 mg/dl) HDL (N:40–60 mg/dl) Serum AST level (N:10–37 U/L) Serum ALT level (N:10–37 U/L) Serum GGT level (N:0–55 U/L) MRI-PDFF (%)
29 4.3 83.7 206.2 179.7 40.9 37.8 54.8 58.1 18.8
± ± ± ± ± ± ± ± ± ±
3.4 (28.5) 2.6 (3.2) 11 (80) 33.6 (200) 82.8 (179.5) 9 (38) 14.8 (34) 22.6 (53.5) 27.8 (60) 9.7 (22.1)
Data are mean ± standard deviation, with the median in parenthesis.
Follow up 29.2 2.5 80.7 200.7 161.8 41.8 32.2 46.7 43.2 12.8
± ± ± ± ± ± ± ± ± ±
3.7 (28.2) 0.9 (2.1) 8.5 (80.5) 36.5 (193.5) 101.7 (124) 6.9 (42) 17 (27) 28.7 (40) 28.1 (42.5) 10.5 (8.9)
p 0.433 0.006 0.368 0.740 0.136 0.470 0.256 0.266 0.015 0.004
Sequential MRI-PDFF in NAFLD
621
40
MRI-PDFF on follow-up (%)
35 30 25 20 rs = 0.714 p = 0.001
15 10
0 0
5
10
15
20
25
30
35
40
MRI-PDFF at baseline (%) Figure 2. Correlation between two sequential MRI-PDFFs.
improvement in body weight. The improvement of MRI-PDFF score was associated with the improvement of clinical and biochemical parameters; BMI (p = 0.075), HOMA scores (p = 0.003), serum AST (p = 0.023), ALT (p = 0.016), and GGT (p = 0.019) levels were improved. Median serum ALT levels were decreased from 55 to 34 U/L. Ten patients had an improvement in their abnormal serum ALT levels; six patients’ ALT levels were normalized with a median MRI-PDFF improvement of 11.9% (range 2.1– 23.6%). With linear regression analyses, at least 5.9% improvement in MRI-PDFF is needed to
23%
On 120th day of follow-up BMI: 25.7
12%
In-phase
PDFF
At baseline BMI: 27.3
achieve a normalized abnormal ALT level. Of note, serum ALT level was increased in only one patient in Group 1. Five of the 13 patients had a decreased MRI-PDFF more than 50%. These patients had lost a median weight of 4.7 kg (range: 2.4–6.6 kg): four had normal serum ALT levels. Of note, five had a decreased MRI-PDFF below 5%, with a weight loss of a median 4.8 kg (range: 0–8.6 kg) in group 1. There was also a good correlation in terms of the delta MRI-PDFF differences and delta serum AST (rs = 0.691, p = 0.002) and ALT levels (rs = 0.608, p = 0.007) in group 1 patients. We also compared the
Opposed-phase
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
5
Figure 3. MRI-PDFF calculations at days 0 and 120 of a patient who had also a decrease in BMI show reduction of hepatic steatosis from 23% to 12%. The decrease in fat fraction can also be realized on in- and opposed-phase MR images.
622
I. S. Idilman et al.
Table III. Association between delta MRI-PDFF and delta changes with linear regression analyses. p
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
DBMI DHOMA score DSerum AST level DSerum ALT level DSerum GGT level
0.644 0.190 0.622 0.006 0.964
changes in clinical and biochemical parameters and MRI-PDFF between Group 1 and Group 2, and there were significant differences in delta serum AST (p = 0.004), ALT (0.004), triglyceride (p = 0.046) levels, and delta MRI-PDFF (p < 0.001) (Table IV). Discussion In the present study, we evaluated the utility of MRIPDFF measurements to quantify the liver fat content cross-sectionally and longitudinally in patients with NAFLD. At baseline, MRI-PDFF measurement had good diagnostic accuracy for quantifying liver fat content compared with liver biopsy results (rs = 0.758, p < 0.001) as previously demonstrated by several investigators [16–18]. The correlation was better in patients without hepatic fibrosis (rs = 0.825, p = 0.001). Median interval between two sequential MRI-PDFF measurements was 184 days. From baseline to the end of the follow-up period, MRI-PDFFs were significantly improved. A close correlation between two sequential MRI-PDFF calculations was also observed (rs = 0.714, p = 0.001). The findings are consistent with previous study, demonstrating that liver MRI-estimated PDFF strongly correlated with magnetic resonance spectroscopy (MRS)-measured liver fat content both at baseline and at week 24 in patients with NASH treated with colesevelam [19]. MRS has been shown to accurately measure hepatic
lipid fraction in previous studies [27,28]. However, it has limited clinical application, because it is not widely available on all clinical MR systems and analyses require specific software. MRI-PDFF, a chemical shift-based water–fat separation method, is a novel MRI technique, which addresses confounding factors for accurate lipid fraction quantification. MRI-PDFF permits quantification of lipid fraction of the whole liver, which is a limitation for MRS, as well as liver biopsy. These results indicate that MRI-PDFF had good diagnostic accuracy for quantifying liver fat content at both cross-sectional and also longitudinal evaluation. The changes in clinical and biochemical parameters from baseline to the end of the follow-up period after an appropriate diet, and an exercise program were also evaluated in the present study. There is no proven beneficial therapy in NAFLD cases. Several investigators have evaluated whether or not diet and lifestyle modification with/without insulin sensitizers or UDCA or antioxidant therapy can improve the metabolic, biochemical, and histological abnormalities in such individuals [29–34]. An appropriate diet and exercise is currently the most accepted first-line therapy for patients with NAFLD and the goal of this approach is to lose 7–10% of their body weight in 6– 12 months. From baseline to the end of the follow-up period, 61% of the patients had weight loss, 33% had weight gain, and 6% had a stable weight in the present study. Sequential MRI-PDFFs were significantly improved (p = 0.004); 72% of the patients had a decrease in liver steatosis, 11% had an increase, and 17% had no change. The investigators recently reported that liver MRI-estimated PDFF measured changes in liver fat content correlated with changes in body weight, serum aminotransferases, and GGT levels in NASH patients [19]. In the present study, in contrast to previous study [19], with linear regression analysis, the changes in serum ALT levels
Table IV. The comparison of the delta differences in clinical, laboratory parameters, and MRI-PDFF between subgroups according to MRIPDFF alterations in the follow up. Group 1 (Decreased MRI-PDFF, n = 13) DBMI DHOMA score DFasting Glucose DCholesterol DTriglyceride DHDL DAST DALT DGGT DMRI-PDFF (%)
–0.5 –1.1 –4.6 –7.6 –31.1 2 –12.2 –19.4 –21 –8.8
(1) (1) (13.2) (46.7) (72.5) (6.6) (22.6) (25.6) (34.1) (7.1)
Data are mean with the standard deviation in parenthesis.
Group 2 (Stable or increased MRI-PDFF, n = 5) 1.6 –3.4 1 0 16.4 –2 11.6 21.4 0.8 1.5
(4.3) (4.8) (5.5) (31.8) (37.8) (11.5) (14.6) (19.4) (13.1) (2.3)
p 0.208 0.566 0.443 0.633 0.046 0.503 0.004 0.004 0.019 < 0.001
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
Sequential MRI-PDFF in NAFLD significantly reflected changes in MRI-PDFF calculations (r2 = 38.6%, p = 0.006) in NAFLD patients. Serum aminotransferase levels have been investigated at diagnosis and also follow-up evaluation of NAFLD. The changes in serum aminotransferases levels in the follow-up evaluation are confounding. A study, which investigated the follow-up changes in liver biopsy, reported a decrease in serum aminotransferase levels paralleling improvement in steatosis and inflammation but not changes in fibrosis stage [35]; however, another study reported disease progression despite a reduction in ALT levels more than 25% [36]. In the subgroup analysis, from baseline to the end of the follow-up period, among 13 patients (Group 1), 9 had an improvement in body weight and 10 had an improvement in abnormal serum ALT levels. The improvement of MRI-PDFF score was associated with the improvement of biochemical parameters including HOMA scores (p = 0.003), serum AST (p = 0.023), ALT (p = 0.016), and GGT (p = 0.019) levels (p < 0.05). Changes in MRI-PDFF correlated with changes in serum AST and ALT levels (p = 0.002, p = 0.007, respectively). There were significant differences in delta serum AST, ALT, triglyceride levels, and MRI-PDFF between Group 1 and Group 2. In conclusion, based on the result of the present longitudinal study, MRI-PDFF can be used for quantification of the longitudinal changes of hepatic steatosis. The changes in serum ALT levels reflected changes in MRI-PDFF in patients with NAFLD. Acknowledgments None of the authors have relevant conflicts of interest to disclose. All authors agree with the manuscripts submission and its content. ISI collected data, conceived and carried out experiments, OK collected data, AHE analyzed data, RI and MK conceived and carried out experiments. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References [1] Angulo P. Nonalcoholic fatty liver disease. N Engl J Med 2002;346:1221–31. [2] Charlton M. Nonalcoholic fatty liver disease: a review of current understanding and future impact. Clin Gastroenterol Hepatol 2004;2:1048–58. [3] Vuppalanchi R, Chalasani N. Nonalcoholic fatty liver disease andnonalcoholicsteatohepatitis:selectedpracticalissuesintheir evaluation and management. Hepatology 2009;49:306–17.
623
[4] Bravo A, Sheth S, Chopra S. Liver biopsy. N Engl J Med 2001;344:495–500. [5] Ratziu V, Charlotte F, Heurtier A, Gombert S, Giral P, Bruckert E, et al. Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology 2005; 128:1898–906. [6] Qayyum A, Goh JS, Kakar S, Yeh BM, Merriman RB, Coakley FV. Accuracy of liver fat quantification at MR imaging: comparison of out-of-phase gradient echo and fat-saturated fast spin-echo techniques – initial experience. Radiology 2005;237:507–11. [7] Pilleul F, Chave G, Dumortier J, Scoazec JY, Valette PJ. Fatty infiltration of the liver: detection and grading using dual T1 gradient echo sequences on clinical MR system. Gastroenterol Clin Biol 2005;29:1143–7. [8] Dixon WT. Simple proton spectroscopic imaging. Radiology 1984;153:189–94. [9] Glover GH. Multipoint Dixon technique for water and fat proton and susceptibility imaging. J Magn Reson Imaging 1991;5:521–30. [10] Bley TA, Wieben O, François CJ, Brittain JH, Reeder SB. Fat and water magnetic resonance imaging. J Magn Reson Imaging 2010;31:4–18. [11] Reeder SB, Sirlin CB. Quantification of liver fat with magnetic resonance imaging. Magn Reson Imaging Clin N Am 2010;18:337–57. [12] Reeder SB, Cruite I, Hamilton G, Sirlin CB. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J Magn Reson Imaging 2011;34:729–49. [13] Reeder SB, Robson PM, Yu H, Schimakawa A, Hines CD, McKenzie CA, et al. Quantification of hepatic steatosis with MRI: the effects of accurate fat spectral modeling. J Magn Reson Imaging 2009;29:1332–9. [14] Yokoo T, Bydder M, Hamilton G, Middelton MS, Gamst AC, Wolfson T, et al. Nonalcoholic fatty liver disease: diagnostic fat-grading accuracy of low-flip-angle multiecho gradient-recalled-echo MR imaging at 1.5 T. Radiology 2009;251:67–76. [15] Kim H, Taksali SE, Dufour S, Befroy D, Goodman TR, Petersen KF, et al. Comparative MR study of hepatic fat quantification using single voxel proton spectroscopy, two point Dixon and three-point IDEAL. Magn Reson Med 2008;59:521–7. [16] Permutt Z, Le TA, Peterson MR, Seki E, Brenner DA, Sirlin C, et al. Correlation between liver histology and novel magnetic resonance imaging in adult patients with nonalcoholic fatty liver disease – MRI accurately quantifies hepatic steatosis in NAFLD. Aliment Pharmacol Ther 2012;36:22–9. [17] Tang A, Tan J, Sun M, Hamilton G, Bydder M, Wolfson T, et al. Nonalcoholic fatty liver disease: MR imaging of liver proton density fat fraction to assess hepatic steatosis. Radiology 2013;267:422–31. [18] Idilman IS, Aniktar H, Idilman R, Kabacam G, Savas B, Elhan A, et al. Hepatic steatosis: quantification by proton density fat fraction with MR imaging versus liver biopsy. Radiology 2013;267:767–75. [19] Noureddin M, Lam J, Peterson MR, Middleton M, Hamilton G, Le TA, et al. Longitudinal comparison between MRI, MRS and histology-determined steatosis in NAFLD patients at two-time points in a randomized trial. Hepatology 2013;58:1930–40. [20] World Health Organization. Report of a WHO consultation: definition of metabolic syndrome in definition, diagnosis and classification of diabetes mellitus. Geneva: World Health
624
[21]
[22]
Scand J Gastroenterol Downloaded from informahealthcare.com by Nyu Medical Center on 05/08/15 For personal use only.
[23]
[24]
[25]
[26]
[27]
[28]
I. S. Idilman et al. organization, Department of Noncommunicable Disease Surveillance; 1999. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985;28:412–19. Kleiner DE, Brunt EM, Van Natta ML, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41:1313–21. Yu H, McKenzie CA, Shimakawa A, Vu AT, Brau AC, Beattu BJ, et al. Multiecho reconstruction for simultaneous water-fat decomposition and T2* estimation. J Magn Reson Imaging 2007;26:1153–61. Ueno T, Sugawara H, Sujaku K, Hashimoto O, Tsuji R, Tamaki S, et al. Therapeutic effects of restricted diet and exercise in obese patients with fatty liver. J Hepatol 1997;27: 103–7. Kukkonen K, Rauramaa R, Siitonen O, Hänninen O. Physical training of obese middle-aged persons. Ann Clin Res 1982;34:80–5. Bydder M, Shiehmorteza M, Yokoo T, Sugay S, Middleton MS, Girard O, et al. Assessment of liver fat quantification in the presence of iron. Magn Reson Imaging 2010;28:767–76. Longo R, Pollesello P, Ricci C, Masutti F, Kvam BJ, Bercich L, et al. Proton MR spectroscopy in quantitative in vivo determination of fat content in human liver steatosis. J Magn Reson Imaging 1995;5:281–5. Cho SG, Kim MY, Kim HJ, Kim YS, Choi W, Shin SH, et al. Chronic hepatitis: in vivo proton MR spectroscopic evaluation of the liver and correlation with histopathologic findings. Radiology 2001;221:740–6.
[29] Neuschwander-Tetri BA, Brunt EM, Wehmeier KR, Oliver D, Bacon BR. Improved non-alcoholic steatohepatitis after 48 weeks of treatment with the PPAR-gamma ligand rosiglitazone. Hepatology 2003;38:1008–17. [30] Belfort R, Harrison SA, Brown K, Darland C, Finch J, Hardies J, et al. A placebo-controlled trial of pioglitazone in subjects with non-alcoholic steatohepatitis. N Engl J Med 2006;355:2297–307. [31] Uygun A, Kadayifci A, Isik AT, Ozgurtas T, Deveci S, Tuzun A, et al. Metformin in the treatment of patients with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2004;19:537–44. [32] Idilman R, Mizrak D, Corapcioglu D, Bektas M, Doganay B, Sayki M, et al. Clinical trial: insulin-sensitizing agents may reduce consequences of insulin resistance in individuals with non-alcoholic steatohepatitis. Aliment Pharmacol Ther 2008;28:200–8. [33] Nobili V, Manco M, Devito R, Di Ciommo V, Comparcola D, Sartorelli MR, et al. Lifestyle intervention and antioxidant therapy in children with nonalcoholic fatty liver disease: a randomized, controlled trial. Hepatology 2008;48:119–28. [34] Promrat K, Kleiner DE, Niemeier HM, Jackvony E, Kearns M, Wands JR, et al. Randomized controlled trial testing the effects of weight loss on nonalcoholic steatohepatitis. Hepatology 2010;51:121–9. [35] Adams LA, Sanderson S, Lindor KD, Angulo P. The histological course of nonalcoholic fatty liver disease: a longitudinal study of 103 patients with sequential liver biopsies. J Hepatol 2005;42:132–8. [36] Pais R, Charlotte F, Fedchuk L, Bedossa P, Lebray P, Poynard T, et al. A systematic review of follow-up biopsies reveals disease progression in patients with non-alcoholic fatty liver. J Hepatol 2013;59:550–6.
