Serum sterols in patients with primary biliary cirrhosis and acute liver failure before and after liver transplantation

Journal of Hepatology 49 (2008) 936–945 www.elsevier.com/locate/jhep

Serum sterols in patients with primary biliary cirrhosis and acute liver failure before and after liver transplantationq Katriina Nikkila¨1, , Markku J. Nissinen2, , Helena Gylling3, Helena Isoniemi1, Tatu A. Miettinen4,* 1 Transplantation and Liver Surgery Clinic, Helsinki University Hospital, Helsinki, Finland Department of Medicine, Division of Gastroenterology, University of Helsinki, Helsinki, Finland 3 Department of Clinical Nutrition, University of Kuopio and Institute of Clinical Medicine, Internal Medicine, Kuopio University Hospital, Kuopio, Finland 4 Department of Medicine, Division of Internal Medicine, Biomedicum Helsinki, C422, P.O.B. 700, University of Helsinki, 00029 Helsinki, Finland 2

Background/Aims: Liver diseases modify sterol metabolism. Liver transplantation (LTX) provides a model to evaluate the impact of disease-affected liver on sterol metabolism. Methods: We studied serum sterol profiles and their relationships to other biochemical markers in consecutive cholestatic patients with acute liver failure (ALF, n = 39) and end-stage primary biliary cirrhosis (PBC, n = 67) before and 27d after LTX. Accordingly, we determined serum levels of sterols, bilirubin and prealbumin. Results: Due to weak cholesterol synthesis of ALF-patients before LTX, their serum levels of cholesterol, lathosterol/ cholesterol, cholestanol/cholesterol and lathosterol/campesterol were 18%–41% lower (P < 0.05 for each) than in PBC, but ratios of phytosterols to cholesterol were equal. In general, non-cholesterol sterol ratios reflected bilirubin and prealbumin concentrations. Interrelation of surrogate sterols showed that homeostasis of cholesterol metabolism prevailed in lowest cholestanol tertile of ALF-patients consistently, but not in PBC. After LTX, cholesterol levels and lathosterol ratios increased in both groups and phytosterol ratios decreased (P < 0.01). Cholestanol decreased profoundly in PBC, but remained 26% higher than in ALF (P < 0.05). Conclusions: Homeostasis of cholesterol metabolism was maintained only in ALF. Metabolism of phytosterols was equal in study groups. PBC- and ALF-patients have differential patterns in their serum sterols and cholesterol metabolism. Ó 2008 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Acute liver failure; Cholestasis; Liver transplantation; Non-cholesterol sterols; Phytosterols; Primary biliary cirrhosis

Received 10 February 2008; received in revised form 23 May 2008; accepted 14 July 2008; available online 26 September 2008 Associate Editor: P.-A. Clavien q The authors declare that they do not have anything to disclose regarding funding from industries or conflict of interest with respect to this manuscript. * Corresponding author. Tel.: +358 40 5111960; fax: +358 9 4717 1851. E-mail address: tatu.a.miettinen@helsinki.fi (T.A. Miettinen).   Authors have equally contributed to the manuscript. Abbreviations: LTX, liver transplantation; ALF, acute liver failure; PBC, primary biliary cirrhosis; ABC, adenosine triphosphate-binding cassette; ATP, adenosine triphosphate; NPC1L1, Niemann-Pick C1like 1 protein; LXR, liver X receptor; MARS, molecular adsorbent immunosorbent system; GLC, gas–liquid chromatography; ULN, upper limit of normal.

1. Introduction Endogenous cholesterol is synthesized predominantly by the liver through a complex enzymatic pathway consisting of several cholesterol precursor sterols [1]. Of the serum non-cholesterol sterols, lathosterol (a precursor of cholesterol synthesis), cholestanol (a metabolite of cholesterol), and plant sterols, e.g., campesterol and sitosterol, are associated with changes in cholesterol metabolism [2–5]. Normally, serum non-cholesterol sterol concentrations are only about 0.5% of the respective cholesterol values. In addition to the intestine, the liver

0168-8278/$34.00 Ó 2008 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2008.07.026

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

is crucially responsible for metabolism of both cholesterol and non-cholesterol sterols, and is – of the parenchymal organs – practically solely responsible for their excretion from the human body via bile. The hepatobiliary secretion of sterols is regulated by several recently discovered hepatic proteins including adenosine triphosphate (ATP)-binding cassette (ABC) transporters ABCG5 and ABCG8 [6], Niemann-Pick C1-like 1 protein (NPC1L1) [7] and liver X receptors (LXRs) [8]. On this theoretical background, liver diseases of various etiologies have the potential to disturb hepatic sterol metabolism differentially depending on which of the hepatic cells and structures are affected and the type of liver injury (e.g., parenchymal, cholestatic or mixed). This gives an intriguing option, first, to seek predictive non-cholesterol sterol markers for, e.g., severe liver failure and poor prognosis, and, second, to evaluate the role of non-cholesterol sterols in hepatic pathophysiology prior to and after liver transplantation (LTX). Studies on primary biliary cirrhosis (PBC), a chronic cholestatic autoimmune liver disease of unknown etiology and among the most common liver diseases leading to LTX [9], suggest that low serum ratios of lathosterol/ cholesterol and campesterol/sitosterol, but high ratios of cholestanol/cholesterol reflect advanced PBC [10,11]. Acute liver failure (ALF) is defined by the rapid development of hepatic synthetic dysfunction associated with significant coagulopathy and hepatic encephalopathy [12,13], of which the etiology is indeterminate in 17– 43% of cases [13,14]. Non-cholesterol sterols have not been measured in ALF, though. LTX provides a unique model to study the impact of the disease-affected liver on the complete regulation of sterol metabolism, and to compare it to circumstances dictated by the new genetic environment of the transplanted liver. Of the non-cholesterol sterols reflecting intestinal cholesterol absorption, i.e., campesterol, sitosterol and cholestanol, the first two are poorly absorbed by the intestine and, in the human body, are totally derived from dietary plants, whereas blood and tissue cholestanol in humans is predominantly synthesized from cholesterol in the liver. Thus, assessment of the latter provides a link between cholesterol metabolism and liver function [2–5]. Furthermore, serum cholestanol/cholesterol is even a more sensitive marker of cholestasis among early-stage PBC-patients than serum bilirubin [10,11]. Precursor sterols of cholesterol biosynthesis, e.g., lathosterol, serve as serum surrogate markers of cholesterol synthesis [4,5]. The aims of the present study were to evaluate preand post-LTX metabolism of cholesterol and non-cholesterol sterols among patients with ALF of unknown etiology leading to LTX, and to compare it to that of PBC-patients undergoing LTX. These liver diseases were chosen, because they both represent a cholestatic liver failure, and because serum sterol profiles and cho-

937

lesterol metabolism of ALF of unknown etiology have not been elucidated so far. In this respect, PBC-patients have been studied most widely [10,11,15,16], and PBC is one of the commonest indications for LTX in Finland; thus, they form a suitable reference group and study population of sufficient size. The following hypotheses were tested: (I) patients with ALF and PBC have differential patterns in their serum sterols and sterol metabolism before and after LTX [10,15], (II) certain profiles of serum sterols are related to poor liver function prior to LTX [15,16], and (III) ALF can be caused by abnormal sterol metabolism [17]. Accordingly, the surrogate sterol markers of cholesterol metabolism were determined and compared to serum bilirubin – a marker of cholestasis and severe liver disease – and to serum prealbumin – a biochemical marker of liver synthesis capacity – in patients with ALF and end-stage PBC undergoing LTX. 2. Patients and methods 2.1. Patients and study design Consecutively liver transplanted PBC- and ALF-patients between years 1987 and 2005 were examined for the purposes of the present study. Preliminary results of the PBC-patient group have been published earlier [16]. The ALF-patients had fulminant acute liver failure of unknown etiology. The ALF-patients were carefully examined prior to LTX. Diagnoses such as viral infections, autoimmune hepatitis, Wilson’s disease and other metabolic diseases, vascular diseases and liver diseases associated with pregnancy were excluded. Previous use of putatively toxic agents such as drugs, herbal products and other toxins like amanita etc. was also excluded. Some of the patients had fresh frozen plasma before LTX. In the ALF-group, 12 patients had preoperative treatment with molecular adsorbent immunosorbent system (MARS). After LTX, all patients received calcineurin-inhibitor-based initial immunosuppression. The majority received cyclosporine-based therapy, with azathioprine and methylprednisolone. Only some patients, participating in controlled clinical trials, had tacrolimus-based initial immunosuppression. The initial target of cyclosporine concentration was 200–250 ng/mL, decreasing with time to maintain a level of 70–150 ng/mL. For tacrolimus, these values were 15–20 ng/mL and 5–10 ng/mL, respectively. Mycophenolate mofetil was not used as initial immunosuppression for these patients but later it was added to some patients with calcineurin-inhibitor induced renal toxicity, and calcineurin-inhibitor doses were reduced or withdrawn. None of the patients were on statin or ezetimibe treatment two months prior to LTX or during the first postoperative month. Serum samples for the purposes of the present study were collected, immediately before LTX and after LTX, at discharge, once at both time points, on posttransplant day 27 for both groups of patients, and stored in 80 °C until sterol analyses. The discharge was considered to be the appropriate time to take serum samples, because most probably up to that point cholesterol metabolism had reached a steady-state. All patients gave informed written consent, and the study protocol had been approved by the Ethics Committee of the Transplantation and Liver Surgery Clinic, Helsinki University Central Hospital.

2.2. Determinations Serum bilirubin and prealbumin were determined by routine hospital methods. Also thromboplastin time and international normalized ratio were determined, but they were not used for data analysis because

938

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

some ALF- and PBC-patients had received fresh frozen plasma before LTX. Serum sterols, i.e., cholesterol, cholestanol, cholesterol precursor sterol lathosterol, and plant sterols (campesterol and sitosterol) were measured from nonsaponifiable material by gas–liquid chromatography (GLC), on a 50-m-long SE-30 nonpolar capillary column (Ultra 1 column, Hewlett-Packard, Wilmington, DE), with 5a-cholestane as internal standard [18,19]. Generally, in studies of non-cholesterol sterols in subjects with normal liver function, their serum values are expressed as ratios to the cholesterol concentration present, i.e., 102  mmol/mole of cholesterol (called ratios in the text) in order to exclude effects of varying serum lipoprotein concentration. However, among patients with severe liver disease (defined as low serum levels of biochemical markers of liver synthesis) with very low serum levels of lipoproteins it would be valuable to evaluate serum non-cholesterol sterols also in terms of absolute concentrations (lg/dl). Thus, in the present study, the data is ordinarily given as ratios, but in correlation analysis also absolute concentrations were used. The ratio of relative cholesterol synthesis sterol marker to that of absorption in serum, e.g., lathosterol/campesterol, lathosterol/sitosterol and lathosterol/ cholestanol, reflects cholesterol synthesis even better than a single lathosterol value. For subgroup analysis, the patient groups (both separately) were divided into tertiles according to pre-LTX cholestanol ratio, so that the tertiles illustrated patients with low to high absorption efficiency of cholesterol.

2.3. Statistical analysis Data analysis was performed using NCSS 2004 software for Windows (NCSS Statistical Software, Kaysville, UT). Data is expressed as median ±95% CI. Data was adjusted by BMI, age, and gender, and logarithmic transforms were used for ANOVA analysis. For comparison between the groups, the Mann–Whitney U test was used. The cholestanol tertiles were analyzed using ANOVA following Bonferroni’s adjustment. Correlations were analyzed with Spearman rank correlation test. A P value <0.05 was considered significant.

3. Results 3.1. Baseline characteristics The study consisted of 39 (15 men, 38%) ALF- and 67 (11 men, 16%) PBC-patients, who got through LTX (Table 1). Most of the transplanted PBC-patients had Child–Pugh B or C cirrhosis. In the ALF-patients, the histology of the explanted liver presented acute fulminant liver failure and had no signs of specific etiology. Median age and gender were equivalent in LTX of the ALF- and PBC- groups, while BMI was 8% lower in the PBC- than the ALF-group (P = 0.002) mainly due to low BMI of the PBC-women (Table 1). The ALFpatients had 18% (NS) lower preoperative median level of prealbumin than the PBC-patients, but postoperatively that was within normal ranges, and equal between the study groups (Table 1). Preoperatively, median serum bilirubin levels were 17and 8-times above the upper limit of normal (ULN) in the ALF- and PBC-groups, respectively (P < 0.001 for ALF vs. PBC) (Table 1). Postoperative bilirubin was normal in six ALF-patients, 26–100 lmol/l in 29 of them, and four had values above 100 lmol/l. Corresponding values of the PBC-group were 33, 30 and 4 (P < 0.001 ALF vs. PBC) (Table 1). Preoperatively, 12 ALF-patients had MARS-treatment, but their sterol ratios and levels of prealbumin and bilirubin were equal to those without MARS-

Table 1 Pre- and postoperative characteristics of study patients, biochemical marker of liver synthesis capacity (serum prealbumin) and serum bilirubin Variables

ALF (n = 39)

PBC (n = 67)

P

Preoperative age (yrs) Total F M P

50 (43–56) 52 (43–57) 44 (33–56) 0.242

54 (50–56) 54 (50–58) 50 (42–57) 0.143

0.052 0.237 0.350

Preoperative BMI (kg/m2) Total F M P

25 (24–27) 26 (24–28) 25 (23–28) 0.242

23 (22–25) 23 (22–24) 24 (20–26) 0.842

0.002 0.001 0.406

Prealbumin (mg/l) Preoperative Postoperative P D

66 (44–72) 260 (244–277) <0.001 +193 (+162 to +218)

80 (67–86) 259 (244–296) <0.001 +183 (+156 to +209)

0.090 0.827

Bilirubin (lmol/l) Preoperative Postoperative P D

430 (357–467) 39 (35–48) <0.001 360 ( 451 to 302)

192 (150–264) 26 (22–29) <0.001 +167 ( 241 to 124)

0.793 <0.001 <0.001 0.001

Values are medians ± 95%CI. P-values in the final column indicate the difference between ALF and PBC and P-values below age- and weight -sections indicate the difference between genders. P-values below prealbumin- and bilirubin –sections indicate difference between pre- and postoperative values. ALF, acute liver failure; PBC, primary biliary cirrhosis; F, female; M, male; D-values indicate difference between post- and preoperative values, BMI, body mass index.

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

treatment. An exception was the sitosterol ratio, which was almost doubled in treated vs. non-treated ALFpatients: 662 (378–1429) mmol/mol cholesterol vs. 361 (260–560) mmol/mol cholesterol, P = 0.032. 3.2. Cholesterol concentration The median preoperative cholesterol concentration in the ALF-group of patients was 41% lower (P = 0.004) but postoperatively 27% higher (P < 0.001) than that of the PBC-group. Median (95%CI) postoperative increases were 104% (72–189%) and 21% (4–53%) for the ALF and PBC groups, respectively (P < 0.001 for both) (Table 2). 3.3. Cholesterol precursor lathosterol ratios Median pre- and postoperative lathosterol ratios were 25% lower (P < 0.05 for both) in the ALF- than the PBC-group, and increased after LTX by 47% (2– 69%) (P = 0.004) and 31% (18–63%, P < 0.001), respectively; (NS for ALF vs. PBC) (Table 2). Median lathosterol/campesterol ratio was pre- and postoperatively 20% lower (P < 0.05) in the ALF- than PBC-group, but the ratio more than doubled equally in both groups (P < 0.01 for both) after LTX (Table 2). Consistently in the ALF-group and postoperatively in the PBC-group, median lathosterol/cholestanol ratio was within the same level as that of lathosterol/campesterol, but, preoperatively in the PBC-group, this ratio was very low due to high cholestanol (Table 2). Median ratio of the two absorption markers campesterol/cholestanol was preoperatively 70% and postoperatively 24% lower in the PBC- than in the ALF-group due to high serum cholestanol (Table 2).

939

tertiles (P < 0.05 for each). Preoperative lathosterol/ campesterol was the lowest in the middle tertile. Division into tertiles of the PBC-patients showed that postoperative cholestanol ratio and preoperative bilirubin concentration increased tertile-dependently (P < 0.001 for both) parallel to that of preoperative cholestanol. Preoperatively in the highest tertile, lathosterol/campesterol was almost three-times higher and campesterol/sitosterol 32% lower as compared to the other tertiles (P < 0.05 for each). 3.6. Correlations of serum sterols with each other (Table 3) The surrogate marker of cholesterol synthesis, the lathosterol ratio, was negatively associated with the respective absorption markers cholestanol (r-range 0.249 to 0.606, P-range 0.04 to <0.001), campesterol and sitosterol (r-range 0.407 to 0.515, P-range 0.010 to <0.001) in the ALF-group, but not in the PBC-group. Analysis according to cholestanol tertiles revealed that lathosterol and cholestanol ratios were interrelated only in the lowest tertile in the ALF-group pre- and postLTX (r-range 0.590 to 0.657, P < 0.01 for each). The surrogate marker of cholesterol absorption, the cholestanol ratio, was pre- and postoperatively inversely related to lathosterol/campesterol (r-range 0.327 to 0.663, P < 0.01 for each), except preoperatively in the PBC-group (r = +0.357, P = 0.009) (Table 3). Cholestanol ratio reflected preoperatively in the PBC-group inversely that of campesterol. Pre- and postoperative cholestanol concentration (not ratio) was positively related to cholesterol in PBC- and ALF-patients (rrange +0.470 to +0.796, P < 0.001 for each).

3.4. Absorption markers of cholesterol

3.7. Correlations of sterols to bilirubin and prealbumin

Median preoperative cholestanol ratio was about three-times higher in the PBC- than the ALF-group. Median reduction during the study period was significant 64% (P < 0.001) for the PBC-group only, resulting in 26% higher (P = 0.006) postoperative cholestanol ratio in the PBC- than the ALF-group. Preoperative ratios of the two plant sterols were equal in the two groups and the values were reduced during the study period by 30–40%, (P < 0.001 for each) (Table 2). Preoperative ratio of campesterol/sitosterol was lower in the PBC- than the ALF-group, P = 0.033), but increased by 26% (P < 0.001) postoperatively to comparable level as in the ALF-group.

Serum concentrations of cholesterol and lathosterol were positively associated with those of prealbumin, in both groups of patients prior to LTX (r-range +0.168 to +0.563, P-range 0.308 to <0.001) (Table 3). Pre-LTX lathosterol concentrations were inversely related to serum bilirubin concentrations in both groups of patients (rrange 0.382 to 0.396, P < 0.05 for both), but the same relationship held true for campesterol/sitosterol only in patients with PBC (Table 3). Lathosterol/bilirubin in patients with ALF and PBC was positively associated with prealbumin concentrations (Fig. 2), such that patients with lathosterol/bilirubin (10) above 7 lg lathosterol/lmol bilirubin had serum prealbumin concentrations above 36 mg/dl. The campesterol ratio preLTX was negatively associated with serum bilirubin concentrations in both groups of patients. Pre- and postoperative cholestanol ratios and the LTX-induced change were positively related to respective bilirubin concentrations, except preoperatively in the ALF-group (Fig. 3A and B). Preoperatively in

3.5. Tertiles by serum preoperative cholestanol ratio (Fig. 1A and B) Division into tertiles of the ALF-patients revealed preoperatively decreasing cholesterol and prealbumin levels, and lathosterol ratio with increasing cholestanol

940

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

Table 2 Comparison of pre- and postoperative values and their differences in patients with acute liver failure and primary biliary cirrhosis Variables

ALF (n = 39)

PBC (n = 67)

P

Cholesterol (mg/dl) Preoperative Postoperative P D

81 (+71 to +115) 208 (+179 to +226) <0.001 +117 (+83 to +133)

138 (+108 to +155) 164 (+149 to +176) 0.002 +22 (+6 to +64)

0.004 <0.001

Lathosterol (100  mmol/mol of cholesterol) Preoperative 64 (52–73) Postoperative 84 (69–101) P 0.004 D +28 (+1 to +41)

84 (79–92) 114(109–135) <0.001 +29 (+16 to +49)

<0.001 <0.001

Cholestanol (100  mmol/mol of cholesterol) Preoperative 360 (+317 to +411) Postoperative 288 (+238 to +344) P 0.054 D 61 ( 111 to 12)

972 (+879 to +1129) 363 (+330 to +441) <0.001 611 ( 744 to 530)

Campesterol (100  mmol/mol of cholesterol) Preoperative 477 (+347 to +568) Postoperative 290 (+219 to +393) P <0.001 D 169 ( 218 to 140)

489 (+365 to +555) 266 (+208 to +340) 0.001 169 ( 283 to 82)

Sitosterol (100  mmol/mol of cholesterol) Preoperative 413 (316–642) Postoperative 252 (198–384) P 0.006 D 186 ( 388 to 73)

542 (423–608) 221 (181–268) <0.001 297 ( 368 to 159)

Lathosterol/cholestanol (lg lathosterol/lg cholestanol) Preoperative 0.16 (0.14–0.21) Postoperative 0.28 (0.21–0.33) P 0.008 D +0.10 (+0.04 to +0.17)

0.01 (0.01–0.10) 0.31 (0.27–0.37) <0.001 +0.21 (+0.17 to +0.27)

Lathosterol/campesterol (lg lathosterol/lg campesterol) Preoperative 0.14 (0.10–0.21) Postoperative 0.30 (0.18–0.44) P 0.002 D +0.14 (+0.06 to +0.27)

0.17 (0.15–0.23) 0.39 (0.28–0.53) <0.001 +0.16 (+0.12 to +0.23)

Campesterol/cholestanol (lg campesterol/lg cholestanol) Preoperative 1.49 (1.07–1.63) Postoperative 0.97 (0.85–1.18) P 0.015 D 0.39 ( 0.61 to

<0.001 0.003

0.07)

0.45 (0.35–0.61) 0.74 (0.56–0.98) 0.003 +0.22 (+0.18 to +0.27)

Campesterol/sitosterol (lg campesterol/lg sitosterol) Preoperative 1.08 (0.91–1.31) Postoperative 1.14 (0.94–1.24) P 0.596 D +0.14 (+0.06 to +0.23)

0.95 (0.73–1.00) 1.20 (1.10–1.34) <0.001 +0.33 (+0.25 to +0.42)

0.033 0.373

<0.001

0.165 <0.001 0.006 <0.001 0.734 0.317 NS 0.846 0.160 0.380 <0.001 0.329 <0.001 0.044 0.007 0.844

<0.001

Values are medians ±95%CI. D-values indicate difference between postoperative and preoperative values. P-values in the final column indicate the difference between ALF and PBC and P-values below each section indicate the difference between preoperative and postoperative values. ALF, acute liver failure; PBC, primary biliary cirrhosis.

PBC, cholestanol concentration was positively related to prealbumin (r = +0.350, P = 0.004).

4. Discussion In general, at the pre-LTX stage, the patients in ALFand PBC-groups had as their common features deep cho-

lestasis and abnormally low serum cholesterol concentration. However, serum sterol profiles between the ALFand the PBC-groups differed considerably prior to LTX in that cholesterol concentration and lathosterol ratio were particularly low in the former group. This was most probably due to widespread hepatocellular necrosis and low liver synthesis capacity (to synthesize, e.g., cholesterol) as could be assessed by low level of serum

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

941

Fig. 1. Median preoperative serum levels of cholesterol, prealbumin, lathosterol/cholesterol, lathosterol/campesterol, campesterol/sitosterol and bilirubin and postoperative that of cholestanol/cholesterol in patients with acute liver failure (A) and primary biliary cirrhosis (B) according to preoperative cholestanol/cholesterol tertiles (n = 13 in ALF- and n = 22–23 in PBC-tertiles). Each triplet of columns represents the lowest, middle and highest tertiles (in that order) of serum cholestanol/cholesterol. Cut off points for cholestanol/cholesterol were 318 and 422 mmol/mol cholesterol 100 in the ALF-group, and 815 and 1199 mmol/mol cholesterol 100 in the PBC-group. *P < 0.05 compared to respective value of the lowest tertile, §P < 0.05 compared to respective value of the middle tertile.

prealbumin. Low serum lathosterol/bilirubin ratio preLTX indicated poor liver function. Median serum cholesterol concentration was relatively low in the PBC-group prior to LTX compared to that of PBC-patients in general [9]. This was most probably due to poor cholesterol synthesis at the end-stage PBC because of severe cirrhosis. The rationale to divide our study patients into tertiles according to serum cholestanol ratios arose from the fact that cholestanol reflects cholesterol absorption in general [3,5] and we have used it in that respect in our previous studies [16,20,21]. It is synthesized enzymatically from cholesterol in the liver, and its serum levels

are essentially not influenced by diet, because its levels in diet are very low and its intestinal absorption is only 3.3% [22]. Ratios of synthesis marker to absorption marker, e.g., lathosterol/campesterol, reflect positively cholesterol metabolism [4,5]. Thus, high ratios point to overwhelming synthesis, low ones predict high relative absorption. In liver injury this pattern has not been evaluated earlier. The present study revealed that high serum cholestanol ratio in the ALF-group was related to weak liver function (synthesis) as judged by preLTX serum levels of cholesterol, lathosterol, lathosterol/campesterol-ratio, lathosterol/bilirubin -ratio and

942

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

Table 3 Correlation matrices for serum non-cholesterol sterols, bilirubin and prealbumin in study patients before and after liver transplantation Variables

Cholestanol mmol/mol of cholesterol

Campesterol/sitosterol lg campesterol/lg sitosterol

Lathosterola

ALF preop ALF postop PBC preop PBC postop

0.606*** 0.411** 0.060 0.249*

Campesterola

ALF preop ALF postop PBC preop PBC postop

+0.220 +0.672*** 0.375** +0.194

Lathosterol/ campesterolb

ALF preop ALF postop PBC preop PBC postop

0.431** 0.663*** +0.357** 0.327**

+0.457** +0.262 0.543*** 0.103

Bilirubin lmol/l

ALF preop ALF postop PBC preop PBC postop

+0.195 +0.702*** +0.530*** +0.412***

0.068 0.415** 0.605*** 0.128

*

+0.303 +0.119 +0.252* +0.196 – – – –

Bilirubin lmol/l 0.396* 0.081 0.382** 0.193

Serum prealbumin mg/l +0.563*** +0.168 +0.336** +0.311*

0.370* +0.490** 0.545*** +0.187

+0.370* +0.028 +0.152 0.007

0.023 0.384* +0.319** 0.336**

0.152 +0.123 0.297* +0.170

– – – –

0.335* 0.238 0.099 0.082

P < 0.05, **P < 0.01,***P < 0.001. a mmol/mol of cholesterol for comparison with cholestanol and lg/dl for comparison with other variables. b lg lathosterol/lg campesterol.

prealbumin. This relationship was also clear in end-stage PBC, but not as consistent as in ALF, as high serum cholestanol (ratios) indicated low ratios of only lathosterol/bilirubin and campesterol/sitosterol. Accordingly, our results are in accordance with earlier studies among PBC-patients, which showed that hypercholestanolemia and accumulation of cholestanol in tissues develop gradually during worsening state of PBC [10,11,15,16]. Long-term recurrence rate of PBC in the transplanted liver is 18% [23]. Results of the present study

Fig. 2. Correlation between lathosterol/bilirubin -ratio and serum prealbumin in the two study groups. ALF = solid line, PBC = dashed line. r = Spearman rank correlations coefficient.

showed that despite lower pre- and postoperative serum bilirubin levels of the PBC- than the ALFpatients, the former ones were postoperatively more cholestanolemic. One putative explanation for this could be an ongoing pathogenesis of PBC early in the liver graft. However, most obviously post-LTX cholestanolemia was due to a slow diffusion of cholestanol from tissue deposits formed during several years before LTX. This explanation, however, is not self-evident as a plateau in serum cholestanol occurs by day 10 post-LTX, and cholestanol remains constantly slightly elevated in several PBC-patients after that [24]. A long-term post-LTX follow-up study in a group of 25 pediatric liver transplant recipients showed higher lathosterol, lower phytosterol, but equal cholestanol ratios as compared to controls [25]. However, these patients did not have PBC. Elevated post-LTX cholestanol ratio among PBC-patients should be further evaluated by a long-term follow-up study, which would include liver histology. Overall, although the most probable reason to hypercholestanolemia in PBC is cholestasis, analysis of biliary and fecal sterols in PBC-patients indicated that also hepatic cholestanol synthesis is modestly increased in PBC for an unknown reason [11]. Interestingly, our present results showed consistently a close positive relation of serum cholestanol ratios to bilirubin except pre-LTX in the ALFgroup, of which the latter finding is probably due to low hepatic synthesis of cholestanol among these patients. Proportion of women in our PBC-group was 84%, which is slightly lower than respective proportion (85–87%) generally in Finnish PBC-patients [26].

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

943

Fig. 3. Correlations between serum bilirubin concentration and cholestanol ratio in ALF- (A) and PBC-group (B). Pre-LTX = upper solid line, postLTX = lower solid line, D-value = change between pre- and post-LTX values = dashed line. r=Spearman rank correlations coefficient.

Normally, the homeostatic regulation of cholesterol metabolism indicates that low intestinal absorption of cholesterol up-regulates cholesterol synthesis, whereas an increase in the intestinal cholesterol flux to the liver suppresses cholesterol synthesis [27]. Thus, synthesis marker/absorption marker ratio is low in patients, if liver failure weakens synthesis of cholesterol. Whether

cholesterol homeostasis is maintained or not can be judged by interplay of surrogate sterol markers of cholesterol synthesis (e.g., lathosterol) and cholesterol absorption (e.g., cholestanol, campesterol and sitosterol) [2–5], which normally shows the best interrelation in the lowest cholestanol tertile (i.e., among subjects with low cholesterol absorption efficiency [28]). The

944

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

novel findings of the present study in this respect were, first, that this homeostasis is maintained in ALF before and after LTX suggesting that despite fulminant liver failure, the key elements with this respect, i.e., de novo cholesterol synthesis and LDL uptake by hepatocytes, maintain their sensitivity to detect serum cholesterol levels. Opposite to that, lack of interrelation of surrogate markers in PBC suggests either lacking homeostasis of cholesterol metabolism in end-stage PBC or changed metabolism of non-cholesterol sterols. Earlier studies on PBC-patients indicate that low serum ratios of lathosterol/campesterol and campesterol/sitosterol reflect advanced liver disease [10,15,16]. Comparison between the two groups in the present study revealed that low value of the former one reflected poor liver function equally in the two groups, whereas that of the latter one showed clinical applicability in this respect in the PBC-group only. This might indicate that PBC changes metabolism of plant sterols in a diseasespecific manner or that a change in their metabolism requires long duration of a liver disease. Mutations in ABCG5/G8-genes encoding sterol transporter proteins in hepatocytes and intestinal epithelium lead to, e.g., disease called sitosterolemia, in which high amounts of phytosterols, including campesterol and sitosterol, are accumulated in blood and tissues leading to premature atherosclerosis and early death [29]. Recently, a patient with an ABCG5/G8 mutation with liver failure and sitosterolemia leading to LTX has been reported suggesting that these abnormally functioning canalicular sterol transporters could be involved in liver pathogenesis [17]. Serum ratios of plant sterols were slightly above normal range prior to LTX equally in both groups, and none of the ALF-patients studied here had serum phytosterol ratios that of sitosterolemia. However, abnormally high campesterol values detected in some of our ALF-patients may be related to abnormalities in canalicular sterol transporters as hyperbilirubinemia did not explain high serum plant sterol ratios or concentrations. Results of the present study showed that serum sitosterol ratio was sensitive to MARS-treatment, which has to be considered when determining serum non-cholesterol sterols in these patients. In conclusion, non-cholesterol sterols depicted differences in cholesterol metabolism in two different types of cholestatic liver disease, and its changes before and after LTX. PBC- and ALF-patients had certain differences in their serum sterol profiles and cholesterol metabolism (hypothesis I). High cholestanol ratio, its close association with bilirubin and lathosterol ratio in serum, were typical of PBCpatients prior to and after LTX. Low serum cholesterol, lathosterol ratio and lathosterol/bilirubin predicted low liver synthesis capacity at pre-LTX stage in the ALF-group instead of similar serum prealbumin

levels compared to PBC. The ratio of markers of cholesterol synthesis to hepatic secretion capacity, i.e., lathosterol/bilirubin, was associated with prealbumin in both groups. Determination of serum sterols at pre-LTX stage described in this study does not challenge the specific criteria used in timing of LTX, but it may, for its part, help clinicians to conclude that patient’s liver failure has advanced to pre-LTX stage (hypothesis II). Metabolism of phytosterols was equal in the study groups. The present study did not show any direct evidence, that abnormalities in sterol metabolism would have been a causative factor in ALF (hypothesis III). Acknowledgements This study was supported by grants from the Paavo Nurmi Foundation and the Helsinki University Central Hospital. References [1] Goldstein JL, Brown MS. Regulation of the mevalonate pathway. Nature 1990;343:425–430. [2] Tilvis RS, Miettinen TA. Serum plant sterols and their relation to cholesterol absorption. Am J Clin Nutr 1986;43:92–97. [3] Miettinen TA, Tilvis RS, Kesa¨niemi YA. Serum cholestanol and plant sterol levels in relation to cholesterol metabolism in middleaged men. Metabolism 1989;38:136–140. [4] Miettinen TA, Tilvis RS, Kesa¨niemi YA. Serum plant sterols and cholesterol precursors reflect cholesterol absorption and synthesis in volunteers of a randomly selected male population. Am J Epidemiol 1990;131:20–31. [5] Nissinen MJ, Gylling H, Miettinen TA. Responses of surrogate markers of cholesterol absorption and synthesis to changes in cholesterol metabolism during various amounts of fat and cholesterol feeding among healthy men. Br J Nutr 2008;99:370–378. [6] Yu L, Hammer RE, Li-Hawkins K, Von Bergmann K, Lutjohann D, Cohen JC, et al. Disruption of Abcg5 and Abcg8 in mice reveals their crucial role in biliary cholesterol secretion. Proc Natl Acad Sci USA 2002;33:16237–16242. [7] Davis HR Jr HR, Zhu Li-Ji, Hoos LM, Tetzloff G, Maguire M, Liu J, et al. Niemann-Pick C1 Like 1 (NPC1L1) is the intestinal phytosterol and cholesterol transporter and a key modulator of whole-body cholesterol homeostasis. J Biol Chem 2004;32:33586–33592. [8] Repa JJ, Berge KE, Pomajzl C, Richardson JA, Hobbs H, Mangelsdorf DJ. Regulation of ATP-binding cassette sterol transporters ABCG5 and ABCG8 by the liver X receptors a and b. J Biol Chem 2002;277:18793–18800. [9] Kaplan MM, Gershwin ME. Primary biliary cirrhosis. N Engl J Med 2005;353:1261–1273. [10] Nikkila¨ K, Ho¨ckerstedt K, Miettinen TA. High cholestanol and low campesterol-to-sitosterol ratio in serum of patients with primary biliary cirrhosis before liver transplantatation. Hepatology 1991;13:663–669. [11] Gylling H, Vuoristo M, Fa¨rkkila¨ M, Miettinen TA. The metabolism of cholestanol in primary biliary cirrhosis. J Hepatol 1996;24:444–451. [12] Yee HF, Lidofsky SD. Acute liver failure. In: Feldman M, Friedman LS, Sleisenger MH, editors. Sleisenger and

K. Nikkila¨ et al. / Journal of Hepatology 49 (2008) 936–945

[13]

[14]

[15]

[16]

[17]

[18] [19]

[20]

Fordtrans’s gastrointestinal and liver disease. Saunders. p. 1567–1576. Ostapowicz G, Fontana RJ, Schiødt FV, Larson A, Davern TJ, Han SH, et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann Intern Med 2002;137:947–954. Brandsaeter B, Ho¨ckerstedt K, Friman S, Ericzon BG, Kirkegaard P, Isoniemi H, et al. Fulminant hepatic failure: outcome after listing for highly urgent liver transplantation – 12 years experience in the nordic countries. Liver Transpl 2002;8:1055–1062. Nikkila¨ K, Ho¨ckerstedt K, Miettinen TA. Serum and hepatic cholestanol, squalene and noncholesterol sterols in man: a study on liver transplantation. Hepatology 1992;15:863–870. Nikkila¨ K, Miettinen TA, Ho¨ckerstedt KVA, Isoniemi H. Sterol parameters as markers of liver function in primary biliary cirrhosis before and after liver transplantation. Transpl Int 2005;18: 221–225. Miettinen TA, Klett EL, Gylling H, Isoniemi H, Patel SB. Liver transplantation in a patient with sitosterolemia and cirrhosis. Gastroenterology 2006;130:542–547. Miettinen TA. Cholesterol metabolism during ketoconazole treatment in man. J Lipid Res 1988;29:43–51. Miettinen TA, Koivisto P. Non-cholesterol sterols and bile acid production in hypercholesterolaemic patients with ileal by-pass. In: Paumgartner G, Stiehl A, Gerok W, editors. Bile acids and cholesterol in health and disease. Lancaster, PA: MTP Press; 1983. p. 183–187. Miettinen TA, Gylling H, Strandberg T, Sarna S. Baseline serum cholestanol as predictor of recurrent coronary events in subgroup of Scandinavian simvastatin survival study. BMJ 1998; 316:1127–1130.

945

[21] Gylling H, Hallikainen M, Pihlajama¨ki J, Agren J, Laakso M, Rajaratnam RA, et al. Polymorphisms in the ABCG5 and ABCG8 genes associate with cholesterol absorption and insulin sensitivity. J Lipid Res 2004;45:1660–1665. [22] Bhattacharyya AK, Lin DS, Connor WE. Cholestanol metabolism in patients with cerebrotendinous xanthomatosis: absorption, turnover, and tissue deposition. J Lipid Res 2007;48: 185–192. [23] Gautam M, Cheruvattath R, Balan V. Recurrence of autoimmune liver disease after liver transplantation: a systematic review. Liver Transpl 2006;12:1813–1824. [24] Nikkila¨ K, Ho¨ckerstedt K, Miettinen TA. Liver transplantation modifies serum cholestanol, cholesterol precursor and plant sterol levels. Clin Chim Acta 1992;208:205–218. [25] Siirtola A, Ketoma¨ki A, Miettinen TA, Gylling H, Lehtima¨ki T, Holmberg C, et al. Cholesterol absorption and synthesis in pediatric kidney, liver, and heart transplant recipients. Transplantation 2006;81:327–334. [26] Rautiainen H, Salomaa V, Niemela¨ S, Karvonen AL, Nurmi H, Isoniemi H, et al. Prevalence and incidence of primary biliary cirrhosis are increasing in Finland. Scand J Gastroenterol 2007;42:1–7. [27] Brown MS, Goldstein JL. A receptor-mediated pathway for cholesterol homeostasis. Science 1986;232:34–47. [28] Miettinen TA, Gylling H, Viikari J, Lehtima¨ki T, Raitakari OT. Synthesis and absorption of cholesterol in Finnish boys by serum non-cholesterol sterols. The cardiovascular risk in young Finns study. Atherosclerosis 2008;200:177–183. [29] Bhattacharyya AK, Connor WE. b-Sitosterolemia and xanthomatosis A newly described lipid storage disease in two sisters. J Clin Invest 1974;53:1033–1043.