zyxwvutsr P Magnetic Resonance Spectroscopy of the Liver: Correlation With Standardized Serum, Clinical, and Histological Changes in Diffuse Liver Disease 31 HESTERN. VAN zyxw HALL,' JEROEN VAN DER GROND,'JAN VAN HATTUM,'CAROLEKOOIJMAN,3 TJAARD u. HOOGENRLUD,~AND WILLEM P. TH. M. MALI' WASSENAER-VAN Many 31PMRS liver studies are assigned to determine TI values and absolute concentrations of hepatic metabolites or to test the diagnostic value in monitoring hepatic cirrhosis or hepatitis. Although most studies agree on relaxation time values, because these are independent of the method used, it is still not clear which hepatic metabolic processes are responsible for the observed spectroscopic changes in the liver of patients with hepatitis or fibrosis/cirrhosis. Meyerhoff et a12showed that the phosphomonoester (PME) concentration is increased in viral hepatitis but not in alcoholic hepatitis, whereas Angus et a16 showed that the PME concentration in alcoholic hepatitis was also increased and even was correlated with the severity of alcoholic hepatitis. Also, the description of spectroscopic changes in the liver of patients with liver fibrosis or cirrhosis is unclear: Angus et a16 suggested that hepatic PME concentration was not correlated to the severity of cirrhosis. This suggestion is supported by Rajanayagam et all and Meyerhoff et al,2 who both 31Pmagnetic resonance spectroscopy (MRS)has been showed that hepatic PME is not significantly increased shown t o be a useful method for studying metabolism in alcoholic cirrhosis. However, other studies"25demonin normal liver and in the liver with diffuse liver abnor- strated that the hepatic PME was increased in patients malities.'.' ' However, relatively little is known about with liver cirrhosis. These contradicting results in the potential of MRS the potential of MRS for diagnosing patients with diffuse liver disease into clinical and histological catego- in detecting hepatitis or cirrhosis may be caused by the ries. Because experimental conditions (field strength, fact that in a limited number of patients histological repetition time, acquisition technique, and quantifica- information was present. Because in these studies no, tion routine used) are often different between studies or only limited, information is provided about the reand patient inclusion criteria are not the same or are sults of biopsy or serum analysis, it cannot be excluded described poorly, it is difficult to compare individual that other hepatic abnormalities, or more specific hepatic abnormalities, may influence the PME concentrastudies with each other. tion. The goal of this study was to find possible correlations between MR spectroscopic results and standardAbbreviations: MRS, magnetic resonance spectroscopy: PME, phosphomonoester; AST, aspartate transaminase: ALT, alanine transaminase; PI, inorized serum, clinical, and biopsy values, to detect heganic phosphorus; VOI, volume of interest: PCr, phosphocreatine. patic abnormalities in patients with diffuse liver From the 'Department of Radiodiagnosis, 'Department of Gastroenterology, disease. ,"Department of Pathology, and 'Department of Neurology of the University The goal of this study was to analyze the possibilities of 31PMR spectroscopyto detect abnormal hepatic histological changes in patients with diffuse liver disease. 31P MR spectroscopy was performed, on a 1.5 T whole-body spectrometer using an image guided localization technique (ISIS), on 38 patients with various diffuse liver diseases, who all underwent histological and serum analysis, and 22 healthy volunteers. Phosphornonoester expressed as a fraction of total phosphorus (PMEP) showed a correlation with abnormal serum aspartate transaminase (AST), histological intralobular degeneratiodfocal necrosis, portal inflammation, and piecemeal necrosis. We found a lower correlation for PME/P with fibrosis. It was not possible to differentiate between fibrosis and cirrhosis. In summary, 31P MR spectroscopy is a technique to detect intralobular degeneration, inflammation and necrosis and to a less extent fibrosis. No diagnostic value was found with respect to steatosis and cholangitis. Furthermore,31PMR spectroscopy is a poor method for classifying patients into diagnostic categories. (HEPATOLOGY 1995;21:443-449.) zyxwvutsr zyx zyx zyxwvutsrqpo Hospital Iltrecht, the Netherlands. Received January 10, 1994; accepted September 14, 1994. Address reprint requests to: Jeroen van der Grond, MD, Academic Hospital Utrecht, Department of Radiology, Heidelberglaan 100, 3584 CX Utrecht, the Netherlands. Copyright h 1995 by the American Association for the Study of Liver Diseases. 0270-9139/95/2102-0027$3.00/0 SUBJECTS AND METHODS Subjects. 31PMR spectra of the liver were obtained from 22 healthy control subjects (21 to 65 years of age) and 38 patients (17 to 64 years) with diffuse liver disease (Table 1). Only patients of whom biopsies were performed within 1year before the spectroscopic examination were included in this 443 zyxwvut zyxwvutsrq zyxwvutsrqponm zyxwvutsr zyxwvuts zyxwvutsrqponmlkjih TABLE1. Individual Histological, Clinical, Serum, and Spectroscopic Patient Data Disease Normal Range AICAH B B B B B B B B C C C C C C C C C CCH GVH PBC PBC PBC PBC PSC PSC PSC W W W W W W W W W W W id f 0-4 0 pi 0-4 0 3 1 1 0 0 0 1 0 0 0 0 0 0 0 3 1 0 1 1 0 0 0 0 1 0 1 0 0 1 0 0 0 0 0 1 0 0 0 4 3 1 3 4 1 3 3 1 3 0 1 3 4 4 3 1 3 4 1 4 3 4 4 4 4 0 0 3 0 0 1 0 1 3 3 0 1 pbn 0-6 0 0-4 0 0 1 1 5 1 2 1 0 1 0 1 3 1 4 1 1 3 4 0 1 1 1 3 3 6 0 0 1 0 0 0 0 0 1 0 0 0 CP A-C aP ggt AST ALT pmol/L t100 IUL t40 pmol/L <30 pmollL a1 g L 0 bi pmol/L 3-17 <30 t35 - 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 0 1 1 1 0 1 0 0 0 0 0 0 0 0 0 7 14 10 6 11 11 12 5 7 9 5 9 10 8 8 14 8 7 34 69 7 11 208 54 11 20 7 12 6 14 10 8 11 7 16 5 6 6 50 73 46 110 86 48 108 60 45 65 60 103 28 67 61 114 53 130 128 787 200 409 1,642 510 66 430 122 85 90 65 102 81 49 343 134 370 126 90 19 39 24 70 73 21 54 50 18 15 24 32 18 32 62 161 10 62 90 2,460 90 61 526 285 157 160 712 15 63 15 26 57 34 19 72 34 21 55 15 59 63 47 34 27 77 68 25 31 15 70 38 84 48 96 41 79 554 66 35 53 223 98 23 125 61 12 18 10 25 35 19 26 62 24 16 15 11 174 110 130 32 58 88 43 23 58 4 80 55 180 69 250 55 70 41 3 15 55 106 284 106 26 150 226 13 17 10 46 106 27 55 90 46 21 7 41 42 38 35 41 41 32 47 38 37 43 37 38 40 36 40 33 35 39 32 35 33 27 30 29 35 32 40 32 15 39 32 37 36 41 40 40 40 A A A ch 0-2 0 st 0-2 0 0 1 1 0 4 1 3 0 1 0 0 4 4 1 3 4 0 4 1 0 0 1 1 3 4 3 0 1 4 0 3 0 0 1 4 3 1 0 0 2 0 0 0 0 0 0 0 0 0 0 1 1 2 1 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 1 0 0 0 1 0 B A A A A A A A A A A A A A A A B A A B B A A A A A A A A A A B A A A NTPP PMEP 0.07-0.19 Pim 0.21-0.29 PDEP 0.33-0.49 0.10-0.22 PH 7.08-7.34 0.14 0.18 0.15 0.16 0.15 0.10 0.22 0.13 0.13 0.18 0.16 0.15 0.15 0.17 0.22 0.20 0.12 0.18 0.09 0.08 0.15 0.08 0.10 0.28 0.15 0.18 0.11 0.14 0.17 0.10 0.09 0.23 0.11 0.14 0.25 0.18 0.11 0.10 0.19 0.13 0.27 0.27 0.21 0.23 0.22 0.18 0.25 0.24 0.20 0.21 0.17 0.23 0.22 0.21 0.23 0.22 0.22 0.22 0.23 0.18 0.14 0.28 0.25 0.16 0.23 0.30 0.22 0.21 0.28 0.06 0.26 0.25 0.26 0.22 0.21 0.20 0.45 0.54 0.40 0.40 0.50 0.47 0.33 0.48 0.45 0.37 0.46 0.48 0.42 0.37 0.40 0.41 0.43 0.40 0.52 0.49 0.47 0.58 0.64 0.28 0.39 0.49 0.49 0.39 0.43 0.46 0.47 0.51 0.50 0.43 0.29 0.42 0.50 0.50 0.22 0.15 0.19 0.17 0.13 0.19 0.23 0.21 0.17 0.21 0.17 0.15 0.26 0.23 0.16 0.18 0.22 0.20 0.17 0.20 0.15 0.15 0.12 0.17 0.22 0.18 0.18 0.17 0.18 0.23 0.16 0.19 0.13 0.18 0.21 0.18 0.18 0.21 7.12 7.23 7.34 7.29 7.19 7.04 7.34 7.44 7.36 7.28 7.26 7.21 7.09 7.43 7.37 7.34 7.02 7.16 7.19 7.29 7.10 7.23 7.42 7.29 7.14 7.09 7.20 7.16 7.05 7.31 7.32 7.17 7.14 7.26 7.10 7.38 7.36 7.07 Abbreviations: AICAH, auto immune chronic active hepatitis; B, hepatitis B; C, hepatitis C; CCH, cryptogene chronic hepatitis; GVH, graft versus host immune disease; PBC, primary biliar cirrhosis; PSC, primary sclerosing cholangitis; W, Wilson’s disease; pbn, periportal bridging necrosis; id, intralobular degeneration and focal necrosis; pi, portal inflammation; f, fibrosis; st, steatosis; ch, cholangitis; bi, bilirubin; ap, alkaline phosphatase; ggt, gamma-glutamyl transpeptidase; AST, aspartate transaminase; ALT, alanine transaminase; al, albumin; CP, Child-Pugh score; PME, phosphomonoester; Pi, inorganic phosphate; PDE, phosphodiester; NTP, 8-P NTP; P, total phosphorus. HEPATOLOGY Vol. 21, No. 2, 1995 zyx zyxwvut zyxwvutsr VAN WASSENAER-VAN HALL ET AL 445 TABLE2. Numerical Scoring of Liver Biopsy Specimens Periportal and Bridging Necrosis 0 None 1 Mild piecemeal necrosis 3 Moderate piecemeal necrosis (less than 50% of the circumference of most portal tracts) 4 Marked piecemeal necrosis (more than 50% of the circumference of most portal tracts) 5 Moderate piecemeal necrosis plus bridging necrosis 6 Marked piecemeal necrosis plus bridging necrosis 10 Multilobular necrosis Intralobular Degeneration and Focal Necrosis None Mild (acidophilic bodies, ballooning degeneration, or scattered foci of hepatonuclear necrosis in <: of lobule or nodules) Moderate involvement of <: of lobules or nodules Marked involvement of >: of lobules or nodules Portal Inflammation Fibrosis 0 None 1 Mild (sprinkling of inflammatory cells in <$of portal tracts) 0 None 1 Fibrous portal expansion 3 Moderate (increased inflammatory cells in <: of portal tracts) 3 Bridging fibrosis (portal-portal or portal-central linkage) 4 Marked dense packing of 4 Cirrhosis inflammatory cells in >: of portal tracts zyxwvutsr Reprinted with permission." study. This study was limited to patients having diffuse liver disease with a stable course. Therefore the kind of disease and severity obtained from the biopsy specimen is likely to be representative for the whole liver and also representative for at least on year. Patients showing abnormalities on MRI or biopsy specimens indicating malignancy were excluded from this study. Biopsy specimens were scored for periportal and bridging necrosis, intralobular degeneration and focal necrosis, portal inflammation, and fibrosis according to Knodell et a1" (Table 2). Histological steatosis was scored 0, 1, and 2 for absent, mild, and severe, respectively, and for cholangitis, we scored 0 or 1 (absent or present, respectively). Liver tests (aspartate transaminase [AST], alanine transaminase [ALT], alkaline phosphatase, gamma-glutamyltranspepsidase, bilirubin, albumin, and lactate dehydrogenase) were measured in all patients within 3 days of the study. We correlated biopsy scores, liver tests, and the ChildPugh13 classification with phosphomonoester expressed as a fraction of total phosphorus (PMEP), inorganic phosphate expressed as a fraction of total phosphorus (PUP), phosphodiester expressed as a fraction of total phosphorus (PDEP), and nucleoside triphosphate (NTP) expressed as a fraction of total phosphorus (p-P NTPP). The hepatic NTP signal mainly (280%) consists of adenosine triphosphate (ATP). Spectroscopic ratios are expressed as percentage metabolite of total phosphorus signal, in which total phosphorus was defined as the signal of PME + Pi + PDE + p-P NTP. Patients and volunteers participated in this study after their informed consent was obtained. All subjects fasted for at least 6 hours before MR spectroscopy examination. 31PMR Spectroscopy. MRS was performed on a 1.5 Tesla Philips whole-body system. A switchable ('H-31P) surface coil with a diameter of 15 cm was placed to the right side of the liver. The coil position was identified by coronal proton images, made with the body coil. The volume of interest (VOI) was selected on the basis of the anatomic information of the coronal as well as axial images. The VOI varied from 200 to 500 mL according to the size of the particular liver. Field homogeneity was optimized by shimming on the proton signal. Volume-selective "P spectra were recorded using the ISIS technique with a repetition time of 1,500 msec and 256 measurements. The total duration of the examination was 30 minutes: 7 to 8 minutes patient preparation and positioning of the surface coil, 7 to 8 minutes MR imaging, 7 minutes shimming, tuning/matching and RF optimization, and 7 minutes MR spectroscopy acquisition time. The averaged free induction decays were zero filled to 4,096 data points and processed with a convolution difference procedure (150 Hz) to remove broad signals from less mobile phospholipids. After exponential multiplication (4 Hz) and Fourier transformation, a linear phase correction was applied. Subsequent t o baseline correction, quantification of the metabolites was achieved by integrating peaks of interest. Phosphocreatine (PCr) was seen in almost all of the MR spectra. Although PCr is not present in the liver,3 it can sometimes be observed because of muscle contamination in the VOI in the ISIS experiment. pH was derived from the chemical shift of inorganic phosphate measured relative to the chemical shift of a-P NTP (referenced a t -7.50 ppm).14 Statistical Analysis. All spectroscopic data are presented a s mean ? SD. To analyze the differences in mean PMEP, Pi/ P PDEP, and p-P N T P P among groups (grouped according to serum value, biopsy score or Child-Pugh score), analysis of variance (ANOVA) was used. If the result of ANOVA was significant for PMEP, Pi/P PDE/P, or 0-P NTPF, we analyzed this metabolite ratio within the group using the unpaired Student's t-test, which was corrected for repeated measures. Furthermore, these ratios were compared with control values using the same test. We performed ANOVA in which patient groups were defined as follows: For the biopsy scores we used the numerical scoring of liver biopsy specimens (Table 2),12for steatosis we used 0, 1, and 2 for absent, mild, and marked, respectively, for cholangitis, we scored 0 or 1 (absent or present, respec- zyxwvu zyxwvutsrqpo 446 VAN WASSENAER-VAN HALL ET AL HEPATOLOGY February 1995 zyxwvutsr z zyxwvu zyxwvuts tively). The liver tests bilirubin and aIbumin we scored two categories; below or above the upper limit of the reference range for undiseased controls, for alkaline phosphatase, gamma-glutamyl transpepsidase, AST, and ALT we scored three categories; below the upper limit of the reference range for undiseased controls, below 3 times the upper limit of the reference range for undiseased controls, and above 3 times the upper limit of the reference range for undiseased controls. All normal values used are indicated in Table 1. The ChildPugh classification was scored in categories A, B, or C.13 RESULTS Diagnostic Categories Typical MR spectra of a healthy volunteer and of a patient showing increased P M E P ratio are shown in Fig. 1A and lB, respectively. The results of the biopsies and the serum analyses and corresponding metabolite ratios obtained with MR spectroscopy are summarized in Table 1. In this table, the spectroscopic results of the control subjects are expressed as a range extending 2 SD below and above the mean. Figure 2 shows the distribution of the MR data grouped to diagnostic categories. The shaded area represents the mean metabolite ratios of the control subjects -+ 2 x SD. There was no significant difference in metabolic ratios or pH between any of the diagnostic categories with control subjects or other diagnostic groups. the distribution of the P M E P ratio for all diagnostic categories grouped according to the biopsy score for periportal and bridging necrosis (Table 2). For periportal and bridging necrosis, we did not find a statistically significant difference in P M E P between patients with a biopsy score of 0 and 1(0.13 -+ 0.04, n = 18, and 0.15 t 0.04, n = 14, respectively). We found a significant difference in the P M E P ratio between patients with a biopsy score 0 and 3 and up (0.13 -+ 0.04 versus 0.18 -+ 0.05, n = 9, P < .01). Intralobular Degeneration and Focal Necrosis. Figure 3B shows the distribution of the P M E P ratio for all diagnostic categories grouped according to the biopsy score for intralobular degeneration and focal necrosis (Table 2). For intralobular degeneration and focal necrosis, we found a significant difference in the hepatic P M E P ratio between patients with a biopsy score of 0 and 1 (P < .0005, P M E P = 0.13 -+ 0.04, n = 26, and 0.19 ? 0.04, n = 10, respectively. Because of the low number of patients with a biopsy score of 2 (n = 2), we did not perform statistical tests with the spectroscopic data from this biopsy score ( P M E P = 0.18 ? 0.04). Portal Inflammation. Figure 3C shows the distribution of the P M E P ratio for all diagnostic categories grouped according to the biopsy score for portal inflammation (Table 2). For portal inflammation we did not find a statistically significant difference in the P M E P ratio between patients with a biopsy score of 0 and 1 in P M E P (0.11 ? 0.02, n = 7, and 0.13 t 0.04, n = 9, respectively), whereas the mean P M E P ratio for patients with biopsy scores 3 and 4 were significantly increased compared to biopsy score 0 (0.17 -+ 0.04, n = 12, P < .005, and 0.16 t- 0.05, n = 10, P < .05, respectively). Fibrosis. Figure 3D shows the distribution of the P M E P ratio for all diagnostic categories grouped according to the biopsy score for fibrosis (Table 2). For fibrosis we did not find a statistically significant difference in the hepatic P M E P ratio between patients with a biopsy score of 0 and 1(0.13 t 0.04, n = 13, and 0.12 ? 0.03, n = 11,respectively). Between patients with a biopsy score 0 and 3 we found a significant difference in the P M E P ratio (0.13 -+ 0.04 versus 0.18 +- 0.06, n = 6, P < ,051. We also found a significant difference in P M E P between patients with a biopsy score 0 and 4 (cirrhosis) (0.13 t 0.04 versus 0.18 -+ 0.06, n = 8, P < .05). zyxwvuts zyxwvu zyxwvutsrqp Biopsies and Serum Values We found a statistically significant difference (ANOVA) ( P < .05) among the means of the subcategories of periportal bridging necrosis, intralobular degeneration, portal inflammation, fibrosis, and AST with the P M E P ratio. This difference was not statistically significant for PUP, PDEP, or 0-P NTPP. No statistically significant difference (ANOVA) was found in any of the metabolite ratios among the means of the subcategories of steatosis, cholangitis, bilirubin, alkaline phosphatase, gamma-glutamyl transpepsidase, alanine transferase, albumin, and the ChildPugh score. Periportal and Bridging Necrosis. Figure 3A shows y-P m AST FIG. 1. 31P spectrum of the liver of a healthy volunteer (left) and a patient showing a n increased hepatic P M E P ratio. Peaks are assigned to phosphomonoester (PME, 6.5 ppm), inorganic phosphate (Pi, 4.9 pprn), phosphodiester (PDE, 2.6 ppm), phosphocreatine (PCr, 0 ppm), y-P NTP (-2.6 ppm), LY-P NTP (-8.0 ppm), and 0-P NTP (-16.5 ppm). For AST we found a significant difference in P M E P between category 1(n = 13) and 2 (n = 20) (1= normal value and 2 = 1 to 3 times normal value, 0.13 t 0.03 and 0.16 -+ 0.05, P < .02, respectively). Between category 1 and 3 (n = 5) (more than three times normal value P M E P = 0.17 t 0.081, we did not find a statistically significant difference. We only found a significant difference in the hepatic P M E P ratio ( P < .02) between control subjects and category 2. Figure 4 shows the distribution of the P M E P ratio for all diagnostic categories grouped according to the AST concentration. In zyxwvutsrqpo zy zyxwvuts zyxwvutsrqp HEPATOLOGY Vol. 21, NO.2, 1995 VAN WASSENAER-VAN HALL ET AL 447 7.40 7.20 7.00 0.30 0.20 0.10 0.00 0.80 0.80 0.40 0.20 0.40 0.30 0.20 0.10 0.00 0.30 0.20 zyxwvutsrq FIG. 2. The individual patient metabolite ratios grouped according to disease. The shaded area represents the control range ? 2 x SD (n = 22). 0.10 I I 0.00 this figure we excluded 2 AST values as outlyer who were far outside the normal range. The correlation coefficient between the PMEP ratio and serum AST was 0.45 ( P < .005). DISCUSSION Clinical. The most important finding of this report is that hepatic 31PMRS can detect pathologic processes such as beginning necrosis and moderate inflammation but to a less extent for detecting fibrosis. On the contrary, 31PMRS is a poor method for classifying patients into diagnostic categories. The increase in PMEP, which is in most literature attributed to hepatitis alone, is in our study for the most part associated with histological intralobular degeneratiodfocal necrosis, periportal and bridging necrosis, moderate portal inflammation, and abnormal serum AST. The finding of a correlation of PMEP with a combination of these three hepatic histological abnormalities may indicate a correlation with what they have in common, i.e., tissue damage. Increased serum levels of AST, a mitochondria1 and cytosolic enzyme distributed throughout the liver, can imply disturbed integrity of hepatocytic membranes (leakage) or cellular necrosis. Although generally ALT, a cytosolic enzyme predominantly in the periportal zone, is a better indicator of hepatitis than AST, we did not find a significant correlation between ALT and the PMEP ratio. However, in acute hepatocellular necrosis due to chemical or vascular injury, the elevation of AST is often greater than the elevation of ALT.15 We expected a correlation between AST and ALT, both serum markers for all forms of acute and chronic hepatitis, with histological portal inflammation. The absence of this correlation can be explained by the fact that biopsy and serum analysis was not performed at the same time. Furthermore, although we limited our study to diffuse liver disease and did not include focal liver disease, it should be realized that conclusions based on biopsy specimens only should be interpreted with care, because of the well-known possibility of sampling errors. Fig. 3A through 3D shows that the 31PMRS is a good method to detect pathological processes as periportal and bridging necrosis, intralobular degeneration and zyxw zyxwv z zyxwvutsrqpo zyx HEPATOLOGY February 1995 448 VAN WASSENAER-VAN HALL ET AL B 0.30 0.24 PME/P 1 + I 0.1 8 0.12 0.06 0.00 0.1 8 zyxwvutsrqpo zy * I 0.12 1 0.06 ** 0 1 2 *** 0.00 3 4 5 0 6 Periportal Bridging Necrosis C 0 A 0 1 2 3 4 Intralobular Degeneration D 0.30 I PME/P 0.24 A 0.24 0 Q 0 0.1 8 0.1 8 0.1 2 0.1 2 0.06 0.00 1 I' + * * 0.06 ** * 0.00 0 1 2 3 4 FIG. 3. The distribution of the P M E P ratios for all diagnostic categories grouped according to the biopsy score for periportal and bridging necrosis (A), intralobular degeneration and focal necrosis (B), portal inflammation (C), and fibrosis (D). The two horizontal lines represent the control range for the P M E P ratio 5 2 x SD (n = 22). A = autoimmune chronic active hepatitis, A = hepatitis B, 0 = hepatitis C, = cryptogene chronic hepatitis, 0 = graft versus host immune disease, = primary biliary cirrhosis, 0 = primary = Wilsclerosing cholangitis, son's disease. Statistical indices represent the significance between each histological subgroup and the normal control values; * = P < .05, ** = P < ,005, *** = P < ,0005. In A, the biopsy scores from category 2-6 were treated as one group. 0 Portal Inflammation focal necrosis, and portal inflammation, but less for detecting fibrosis. These findings are in agreement with earlier that suggested that an increased PME was mainly correlated to viral hepatitis and not to fibrosis. Cox et aI4 demonstrated that there was no change in the PME/ATP ratio in patients with liver cirrhosis only. However, when cirrhosis was accompanied by hepatitis, a significant increase in PMEfATP was observed. Nevertheless, other studies show that both hepatitis and cirrhosis contribute to an increased PME.3,5The differences for MR spectroscopy to diagnose liver abnormalities in these studies may be caused by the very broad patient inclusion criteria. Although we also found a statistical difference (P < .005) in PME/ P between hepatitis patients and control subjects (0.16 2 0.03 and 0.13 2 0.04, respectively), our study shows that hepatic 31PMRS is also a technique to detect differences between mild, moderate or severe inflammation, and also between differences between mild and severe fibrosis. Therefore, patient selection that is only based on the presence or absence of inflammation (hepatitis) may introduce a lot of patient bias. 1 2 Fibrosis 3 4 zy Technical. The PME peak is composed mainly from resonances that arise from phosphorylcholine (PC) and phosphorylethanolame (PE),which are both precursors in phospholipid biosynthesis. In small amounts, sugar phosphates and adenosine monophosphate contribute to the PME peak.16 Increased levels of phospholipid precursors in general point at an increased phospholipid biosynthesis, and the increase in PMEP in hepatic inflammation probably reflects hepatocyte regeneration, which is likely to occur in inflammation and necrosis. Although we realize that presentation of absolute metabolite concentrations is preferred, in this study metabolite ratios rather than absolute metabolic concentrations are measured. Measurements involving the calculation of absolute concentrations require individual TI measurements. To calculate saturation factors, the MRS experiments need to be repeated with at least one different repetition time. Unfortunately, because of the physical and mental condition of the patients, they would not stand such a long-lasting effort. However, especially because the TI values of liver metabo- zyxwvuts zy z z zyxwvutsrqp HEPATOLOGY Vol. 21, No. 2, 1995 VAN WASSENAER-VAN HALL ET AL REFERENCES 0.30 PME/P 0.00 449 1. Rajanayagam V, Lee RR, Ackerman Z, Bradley WG, Ross BD. Quantitative P-31 spectroscopy of the liver in alcoholic cirrhosis. J Magn Reson Imag 1992;2:183-190. 2. Meyerhoff DJ, Boska MD, Thomas AM, Weiner MW. Alcoholic liver disease: quantitative image-guided P-31 MR spectroscopy. Radiology 1989; 173:393-400. 3. Oberhaensli R, Rajagopalan B, Galloway GJ, Taylor DJ, Radda GK. Study of human liver disease with P-31 magnetic resonance spectroscopy. Gut 1990;31:463-467. 4. Cox IJ, Menon DK, Sargentoni J , Bryant DJ, Collins AG, Coutts GA, Iles RA, et al. P-31 magnetic resonance spectroscopy of the human liver using chemicalshift imaging techniques. J Hepatol 1992; 141265-275. 5. Munakata T, Griffiths RD, Martin PA, Jenkins SA, Shields R, Edwards RHT. An in uivo 31P MRS study of patients with liver cirrhosis: progress towards a non-invasive assessment of disease severity. NMR Biomed 1993;6:168-172. 6. Angus WA, Dixon RM, Rajagopalan B, Ryley NG, Simpson KJ, Peters TJ, Jewel1 DP, et al. A study of patients with alcoholic liver disease by 31-P nuclear magnetic resonance spectroscopy. Clin Sci 1990;78:33-38. 7. Cox IJ, Bryant DJ, Collins AG, George P, Harman RR, Hall AS, Hodgson HJF, et al. Four dimensional chemical shift MR imaging of phosphorus metabolites of normal and diseased human liver. J Comput Assist Tomogr 1988; 12(3):369-376. 8. Brinkmann G, Melchert UH. A study of T1-weighted 31P MR spectroscopy from patients with focal and diffuse liver disease. Magn Reson Imag 1992; 10:949-956. 9. Meyerhoff DJ, Karczmar GS, Matson GB, Boska MD, Weiner MW. Non invasive quantitation of human liver metabolites using image guided 31Pmagnetic resonance spectroscopy. NMR Biomeb1%0;3:17-22. 10. Cox IJ. Coutts. Gadian DG. Ghosh P. Sareentoni J. Youne IR. Saturation effects in P-31 'magnetic resonance spectra ofy the human liver. Magn Reson Med 1991; 17:53-61. 11. Buchthal SD, Thoma WJ, Taylor JS, Nelson S J , Brown TR. In viuo T1 values of phosphorus metabolites in human liver and muscle determined a t 1.5 T by chemical shift imaging. NMR Biomed 1989;5/6:298-304. 12. Knodell RG, Ishak KG, Black WC, Chen TS, Craig R, Kaplowitz N, Kiernan TW, et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. HEPATOLOGY 1981;5:431-435. 13. Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-649. 14. Taylor DJ, Bore PJ, Styles P, Gadian DG, Radda GK. Bioenergetics of intact human muscle: a 31P nuclear magnetic resonance study. Mol Biol Med 1983; 1:77-94. 15. Presig R, Tygstrup N, Price C. Assessment of liver function search tests. In Millward-Sadler GH, Wright R, Arthur MJP, eds. Wright's liver and biliary disease. Vol 1, Ed 3. London: Saunders, 1992. 16. Meyerhoff DJ, Karczmar GS, Weiner MW. Abnormalities of the liver evaluated by magnetic resonance spectroscopy. Invest Radio1 1989;24:980-984. 17. Evelhoch JL, Ewy CS, Siegfried BA, Ackerman JJH, Rice DW, Briggs RW. 31Pspin-lattice relaxation times and resonance linewidths of r a t tissue in vivo: dependence upon the static magnetic field strength. Magn Reson Med 1985;2:410-417. 18. Stark D, Mosely ME, Bacon BR, Moss AA, Goldberg HI, Bass NM, James TL. Magnetic resonance imaging and spectroscopy of hepatic iron overload. Radiology 1985;154:137-142. 19. Roth K, Hubesch B, Meyerhoff DJ, Naruse S, Gober JB, Lawry TJ, Boska MD, et al. Non-invasive quantitation of phosphorus metabolites in human tissue by NMR spectroscopy. J Magn Reson 1989;81:299-310. 20. Iles RA, Cox IJ, Bell JD, Dubowitz LMS, Cowan F, Bryant DJ. 3 1-P Magnetic resonance spectroscopy of the human paediatric liver. NMR Biomed 1990;3:90-94. zyxw zyxwvutsrqpo zyxwvutsrqp ' 0 I 100 Aspartate Transferase (microM/L) FIG.4. Cartesian plot showing the correlation between the PME/ P ratio and serum AST. The correlation coefficient between the PMEI P ratio and serum AST was 0.45 (P < ,005). A = autoimmune chronic active hepatitis, A = hepatitis B, 0 = hepatitis C, 0 = cryptogene chronic hepatitis, 0 = graft versus host immune disease, W = primary biliary cirrhosis, 0 = primary sclerosing cholangitis, = Wilson's disease. + lites are expected to be influenced by deposition of paramagnetic Fe or Cu,17-19 individual T1 relaxation measurements may be very useful. Munakata et a15 determined that there was a trend of a prolongation of T1 for all metabolites in cirrhotic patients compared with normal subjects. A prolonged T1 value, in general, causes a reduced signal intensity using the same repetition time. Using the data of Munukata e t al also a reduced P M E P ratio would exist in the cirrhotic liver compared with normal liver, based on the reduced T1 only. Therefore, the observed increase in P M E P with increased periportal and bridging necrosis, intralobular degeneration and focal necrosis, portal inflammation, fibrosis and AST is probably caused by an increased concentration of PME components, because a prolonged Tl only would decrease the P M E P ratio. Nevertheless, even if saturation factors are known, comparison between different studies would still be difficult. Differences in field strength, technical methods (Spectroscopic Imaging or ISIS), and quantification methods are of great influence on the final ratios. Furthermore, age-related changes in the P M E P ratio cannot be excluded,20although our adult control group did not show an age-related correlation with spectral data. CONCLUSION This study shows that 31PMRS of the liver is a poor method for classifying patients into diagnostic categories, but might be useful to detect markers of liver tissue damage, being abnormal serum AST, intralobular degeneration, piecemeal necrosis, and portal inflammation. __ zyxw