- Email: [email protected]
Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake?
CLB-08758; No. of pages: 6; 4C: Clinical Biochemistry xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Anne Tamigniau 1, Pierre Wallemacq, Diane Maisin ⁎ Laboratory of Clinical Chemistry, Saint-Luc University Hospital, Université Catholique de Louvain, Brussels, Belgium
a r t i c l e
i n f o
Article history: Received 14 March 2014 Received in revised form 9 May 2014 Accepted 13 May 2014 Available online xxxx Keywords: Carbohydrate-deficient transferrin Trisialotransferrin Cirrhosis Fibrosis Steatosis Heavy drinking Capillary electrophoresis
a b s t r a c t Objectives: Carbohydrate-deficient transferrin is a well-known biomarker widely used for detection of chronic excessive alcohol intake. However, under certain clinical conditions particularly frequently met amongst heavy drinkers (steatosis, fibrosis, cirrhosis…), it isn't a reliable biomarker. In this study, we tried to find additional biomarkers to CDT in order to improve detection of chronic excessive alcohol intake. Design and methods: We conducted a retrospective cohort study from December 2007 to December 2009. We focused mainly on three different groups: heavy drinking patients with active alcohol consumption (n = 243), cirrhotic patients (abstinent patients and non alcoholic cirrhosis, n = 44) and control group (n = 85). Results: In our study, CDT showed a poor sensitivity for diagnosis of heavy drinking patients (around 63%, and even lower) for cirrhotic patients and patients at advanced stage of fibrosis. Combination of CDT with trisialotransferrin enabled to improve significantly sensitivity and specificity (p-value AUC ROC b 0.001). When adding mean corpuscular volume and gamma-glutamyltransferase to this first combination, performances were even better (p-value b 0.001). This second cluster enabled to make a statistically significant difference between cirrhotic patients with active alcohol consumption compared to abstinent cirrhotic patients and to non alcoholic cirrhotic patients (p-value b 0.001). Conclusion: From our study, trisialotransferrin seems to be a useful additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Alcohol abuse remains a major health concern, with a prevalence of alcohol dependent patients accounting for up to 12.4% amongst Belgian hospitalized patients [1]. Former studies have already estimated that 20 to 30% of admissions in hospitals could be, directly or indirectly, related to alcohol disorders [2]. Various biochemical markers have been developed over the years to detect excessive alcohol consumption, notably for occupational medicine and forensic use. Carbohydrate-deficient transferrin (CDT) still remains one of the most widespread and easiest to use [3]. It has been demonstrated that a consumption of alcohol of 60 to 80 g per day over the past month could well correlate with an abnormal rise in CDT values [4]. However, its poor diagnostic sensitivity limits its use much more as a confirmatory and follow-up marker, than as a screening one [5]. Moreover, some studies have shown that CDT isn't a reliable marker under certain clinical conditions, such as steatosis, advanced fibrosis and cirrhosis [3,6]. Those clinical conditions are particularly frequently ⁎ Corresponding author. E-mail address: [email protected] (D. Maisin). 1 Present address: Hematology Laboratory, CHU UCL Mont-Godinne Dinant, Université Catholique de Louvain, Yvoir, Belgium.
met amongst heavy drinkers (N60 g/day or N210 g alcohol/week) [7], which precludes the use of carbohydrate-deficient transferrin to evaluate excessive alcohol consumption in those patients. In our study, we evaluated the performances of well-known biochemical markers of alcohol abuse such as AST–ALT ratio, MCV, GGT and CDT for heavy drinking patients [8–10]. We also especially focused on trisialotransferrin fraction which seemed, from our clinical practice, to increase with steatofibrosis and cirrhosis. One of the aims of this study was to determine if trisialotransferrin could whether or not be considered as an additional potential biomarker of chronic alcohol abuse in combination with others already used in clinical practice. We finally focused on patients suffering from cirrhosis, caused either by excessive alcohol consumption or not, and explored the potential use of trisialotransferrin and CDT to detect excessive alcohol drinkers (active and abstinent) amongst them.
Material and method Patients and study design We conducted a retrospective cohort study from December 2007 to December 2009. We selected three groups of patients: heavy drinking
http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
2
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
patients with active alcohol consumption, cirrhotic patients (abstinent patients and non alcoholic cirrhosis), and control patients. 272 alcoholdependent patients hospitalized in our Hepatology department for alcohol withdrawal treatment were selected in our heavy drinking group. Exclusion criteria were defined as follows: concomitant liver diseases, HIV patients, transferrin variants, immunosuppressive treatments, and uninterpretable electrophoretic profile [11]. From our initial group of 272 patients, 29 were excluded based on these criteria. The 243 patients included in this group were mostly men (66%), around 50 years old. 44 patients suffering from cirrhosis were recruited from our Transplant unit (before transplantation). Cause of cirrhosis and status of alcohol consumption were obtained from the pre-transplant visit. Exclusion criteria were the same as for the heavy drinking group, except for the concomitant liver diseases criteria. In this group, 21 patients were suffering from cirrhosis due to previous chronic alcohol consumption (abstinence at the time of the study), whereas 23 cirrhosis were due to other causes (viral hepatitis, cholangitis…). Those patients were mostly men (75%), aged around 55 years old. 85 control patients were recruited from our outpatient clinic. Inclusion criteria were defined as follows: no record in our Hepatology department concerning alcohol abuse, no evidence of liver injury and HIV, and no laboratory evidence of chronic alcohol abuse. Patient selection was conducted in order to reach similar proportions in terms of age and gender as the patients recruited in our study. This study was carried out in agreement with our local ethics committee. Alcohol consumption Alcohol consumption was evaluated at the entrance of the patient in our Alcohology department (start of abstinence program) for the heavy drinking group. From the Hepatology patients' files, we were able to extract the data relative to the self-reported alcohol consumption. Wines, beers, spirits and liqueurs were all considered as relevant in the evaluation of the daily consumption. Every patient included in our study had a self-reported consumption N 210 g alcohol/week and could thus be considered as heavy drinker [7].
Table 1 Characteristics of patients and controls included in the study. Heavy drinking group n = 243
Cirrhosis group n = 44
Control group n = 85
48.8 (0.7)
56.4 (1.5)
51.8 (1.7)
162 81 2
33 11 3
55 30 1.8
Fibrosis stage (n) Not assessed Failure F0–F1 F2 F3 F4
13 (5.4%) 11 (4.6%) 114 (47%) 34 (14%) 27 (11%) 44 (18%)
– – – – – –
– – – – – –
Liver injury (n) Not assessed Suboptimal examination No evidence of liver injury Steatosis Cirrhosis Others
2 (1%) 2 (1%) 34 (14%) 175 (72%) 28 (11%) 2 (1%)
– – – – 44 (100%) –
– – – – – –
Origin of cirrhosis (n) Chronic alcohol abuse Chronic viral infections Primary biliary cirrhosis Primary sclerosing cholangitis Cholestatic disease Others
28 – – – – –
21 6 5 1 1 10
– – – – – –
Biological markers, mean (SE) AST (IU/L) ALT (IU/L) AST/ALT ratio GGT (IU/L) MCV (fl) CDT (%)
75.6 (4.6) 56.3 (2.9) 1.4 (0.04) 248 (27) 95.7 (0.5) 4.3 (0.3)
55.8 (5.8) 40.5 (4.6) 1.5 (0.09) 117 (19) 90.3 (1.8) 0.73 (0.04)
22.4 (0.5) 19.8 (0.7) 1.2 (0.05) 24.6 (1.2) 86.6 (0.4) 0.76 (0.03)
Demographics Age, years, mean (SE) Gender Male (n) Female (n) Sex ratio (M/F)
Biological markers Aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyltransferase (GGT) were determined on heparinized plasma samples on DxC 800© from Beckman Coulter (Analis, Suarlée, Belgium). Prior to measurement, samples were centrifuged for 10 min at 3000 rpm. Reference values ranged from 7 to 50 IU/L for GGT, from 14 to 63 IU/L for ALT and from 6 to 33 IU/L for AST. Mean corpuscular volume (MCV) was determined on EDTA whole blood samples on Sysmex XE-2100 (Hoeilaart, Belgium). Reference values ranged from 85 to 95 fl for patients over 18 years old.
had to strictly determine if the result was pathological or not, results below 1.6% were considered negative [12]. Trisialotransferrin referred to isoforms with a pI at 5.6. For all patients, good isoform separation was obtained without bridging phenomenon.
CDT and trisialotransferrin Carbohydrate deficient transferrin (CDT) and trisialotransferrin measures were carried out on serum. The blood specimen were centrifuged at the reception in the laboratory, and the samples were then stored to − 20 °C until analysis (gentle thawing at 4 °C overnight). CDT and trisialotransferrin were measured by capillary zone electrophoresis on a Capillarys2 system from Sebia (Issy-les-Moulineaux, France) and data from electrophoretic profiles were extracted by the Phoresis 5.4.1 software. After complete iron saturation, transferrin isoforms were separated according to their isoelectric point (pI) and detection was achieved by UV determination at 200 nm. CDT fractions referred to diasialotransferrin (isoforms with a pI at 5.7). Asialotransferrin wasn't included in the determination of the CDT results. CDT results were considered pathological at or above 1.6%, negative below 1.3%, a gray zone existing between 1.3 and 1.6%. When we
Fig. 1. Number of cases of cirrhosis and steatosis according to fibrosis stage. Patients at early stage of fibrosis (FD–F1, F2) showed, most of the time, steatosis or no or minimal evidence of lesion. At advanced stage of fibrosis (F3, F4), patients were mostly suffering from steatosis and cirrhosis.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
3
Table 2 Sensitivity and efficiency of biomarkers for diagnosing heavy drinking pattern.
AST/ALT ratio MCV GGT CDT CDT CDT—fibrosis stage F0–F1 F2 F3 F4
Cutoff
Sensitivity (%)
CI95
Efficiency (%)
1.5 95 50 1.3 1.6
30.6 51.4 74.3 72.0 63.4
24.8–36.8 44.9–57.9 68.3–79.7 65.9–77.6 56.9–69.4
43.8 63.9 80.9 78.3 72.9
1.6 1.6 1.6 1.6
71.9 70.6 48.1 59.1
62.7–79.9 52.5–84.9 28.7–68.0 43.2–73.7
83.9 91.6 87.5 86.0
Liver fibrosis—METAVIR Scoring System Liver fibrosis was assessed by transient elastography—Fibroscan© from Echosens (Paris, France). This technique is an ultrasound-based method consisting of evaluating liver stiffness based on a shear wave velocity (kPa). Considering the shear wave velocity, patients were classified according to the METAVIR scoring system for fibrosis [13,14]. The results were only considered reliable when 10 valid measurements were achieved (success rate 100%) and the technique couldn't be performed with patients suffering from ascites [15,16].
Fig. 2. Receiver operating characteristic (ROC) curves of biomarkers for the detection of heavy drinking pattern (N210 g alcohol/week). The diagonal line represents an AUROC of 0.5 (null hyphothesis). The closer the curve is to the upper left corner, the better is the overall accuracy of the biomarker.
Evaluation of steatosis and cirrhosis by Doppler Ultrasound (US) Steatosis and cirrhosis were assessed by Doppler UltraSound (US) imaging performed on Acuson Sequoia 512© from Siemens (Munich, Germany). Ultrasound examinations revealed three main types of observations: 1) Normal echogenicity of the liver without any evidence of lesion, 2) increased echogenicity of the liver highly suggestive of steatosis, and 3) changes in liver morphology (nodular liver surface, coarse echopattern…) combined with signs of portal hypertension, highly suggestive of cirrhosis [17]. The main US signs of portal hypertension considered here were splenomegaly, alterations of portal vein blood flow velocity and alterations of resistive index of artery at porta hepatis and of intrahepatic branches [18,19]. Statistical analysis Normal distribution of variables was assessed by D'Agostino– Pearson test and outliers were detected with the Grubbs' test. T-test for independent samples was performed for parameters with a (log) normal distribution, whereas Mann–Whitney test was preferred for data that weren't normally distributed. P values were considered statistically significant below 0.05. Sensitivity and efficiency were determined for the different biomarkers in order to assess their potential performances for diagnosing heavy drinking pattern. Efficiency was defined as the ratio between the true determinations (true positive and true negative) and the total number of samples. Combination of biomarkers was achieved by logistic regression and was calculated as follows: logit(p) for cluster I = −10.05 + (5.46 × CDT) + (1.36 × trisialotransferrin) and logit(p) for
cluster II = − 30.28 + (4.03 × CDT) + (1.35 × trisialotransferrin) + (0.19 × MCV) + (0.11 × GGT). Logit(p) was back-transformed to p by the following formula: p = 1 / (1 + e−logit(p)). Analysis and comparison of receiver operating characteristic (ROC) curves was carried out according to Long et al. methodology [20]. We used MedCalc 12.7.0 (MedCalc software, Ostend, Belgium) and GraphPad Prism 6 (GraphPad Software Inc, La Jolla, USA) to perform all these statistical analysis. Results Cirrhosis and steatosis amongst heavy drinkers, according to fibrosis stage Patients were classified according to fibrosis stage (METAVIR scoring system) and to the results of their Doppler Ultrasound (Table 1). In Fig. 1, we represented the number of patients and their liver injury in regards to their corresponding fibrosis stage. We noticed that, at early stage of fibrosis (F0–F1, F2), patients were most of the time suffering from steatosis, or even, didn't show any evidence of liver injury. On the other side, at advanced stage of fibrosis (F3, F4), we observed an increased incidence of cirrhosis together with a decreased incidence of absence of liver injury. Diagnosis of heavy drinking Data relative to sensitivity and efficiency of biomarkers are summarized in Table 2. Sensitivity didn't exceed 75% for the different biomarkers tested in our study (range from 31 to 74%). CDT showed a quite poor sensitivity for diagnosis heavy drinking patients, around
Table 3 Receiver operating characteristics (ROC) curves analysis.
CDT Trisialotransferrin CDT–Trisialo CDT–Trisialo–MCV–GGT a b
AUC ROC
CI95
Cutoffb
Sensitivity
Specificity
p-valuea
0.90 0.79 0.96 0.99
0.86–0.93 0.73–0.82 0.93–0.98 0.97–0.99
1.0 3.9 0.73 0.80
79.4 72.0 86.8 92.9
89.4 74.1 94.1 98.8
– p = 0.0006 p b 0.0001 p b 0.0001
Comparison with CDT. Corresponding to the Youden Index.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
4
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Table 4 Carbohydrate-deficient transferrin and trisialotransferrin amongst heavy drinking patients.
Control group (n = 85) Fibrosis stage F0–F1 (n = 112) F2 (n = 34) F3 (n = 27) F4 (n = 44) a
CDT
p-valuea
Trisialotransferrin
p-valuea
Mean (CI95)
Mann–Whitney test
Mean (CI95)
t-test
0.76 (0.71–0.81)
–
3.2 (3.0–3.4)
–
5.5 (4.4–6.5) 5.6 (3.5–7.7) 3.1 (1.6––4.5) 2.8 (1.8–3.7)
p p p p
4.5 (4.2–4.7) 4.3 (3.9–4.8) 4.2 (3.8–4.6) 4.7 (4.4–5.1)
p p p p
b b b b
0.0001 0.0001 0.0001 0.0001
b 0.0001 b 0.0001 = 0.0001 b 0.0001
Comparison with control group.
63%. When considering fibrosis stage and liver injury, we noticed that sensitivity of CDT was even lower for cirrhotic patients at advanced stage of fibrosis. Combination of biomarkers was evaluated to explore if sensitivity could, whether or not, be improved for diagnosing heavy drinking patients. Two clusters of markers were studied: cluster I consisted of combination of CDT and trisialotransferrin, while cluster II consisted of combination of CDT, trisialotransferrin, GGT and MCV. All data relative to receiver operating characteristic curve are summarized in Table 3. Area under curve (AUC) of cluster I was significantly higher compared to AUC associated to CDT alone (p-value b0.0001). When considering cluster II, we showed that this combination was associated with a significantly higher AUC compared to cluster I and CDT alone (p-value b0.0001), reaching a sensitivity of 92.9% and a positive predictive value of 91.5% (Fig. 2). However, it should be kept it mind that such a result is directly dependent on the prevalence in our cohort, and could thus not be generalized to other groups. Trisialotransferrin and CDT amongst heavy drinkers Data relative to trisialotransferrin and CDT amongst heavy drinking patients are summarized in Table 4. We observed that CDT was statistically different from the control group at every stage of fibrosis (p-value b0.0001). However, we noticed that CDT was higher in early stage of fibrosis, F0–F1 and F2, compared to late stage of fibrosis, F3 and F4 (p-value b0.05) (Fig. 3 left). In patients at advanced stage of fibrosis, we noticed that some of them don't reach the manufacturer threshold of 1.6%, with sensitivity of CDT as low as 48% (Table 2). When focusing on heavy drinking patients at cirrhosis stage, we noticed that only 8 out of the 28 patients (b29%) had pathological values of CDT (≥1.6%). Trisialotransferrin was also statistically different from the control group at every stage of fibrosis. We didn't show any significant difference between the different fibrosis stages (Fig. 3 right).
Trisialotransferrin and CDT amongst cirrhotic patients Cirrhotic patients were divided into three main groups: alcoholic cirrhotic patients with active consumption, former alcoholic cirrhotic patients (abstinent) and non alcoholic cirrhotic patients. CDT in alcoholic cirrhotic patients with active consumption was significantly higher than in the control group and in the other two groups (p-values b0.001). No significant differences were observed between the control group and abstinent and non alcoholic cirrhosis (Fig. 4 left) (Table 5). Trisialotransferrin was significantly higher in the three groups compared to the control group. When comparing abstinent patients to active alcohol consumers, we noticed that trisialotansferrin was significantly higher in abstinent patients (Fig. 4 right). In order to assess if cirrhotic patients with active alcohol consumption could be distinguished from former alcoholic cirrhotic patients (abstinent), we tried to use our cluster II to see whether or not there was a difference between those two groups (Fig. 5). We observed that cluster II score was significantly higher in every group compared to the control group (p-values b 0.0001). Cirrhotic patients with active consumption seemed to have a significantly higher cluster II score than abstinent patients (p-value b0.05) and than non alcoholic cirrhotic patients (p-value b 0.001).
Discussion Our study confirmed some performance limitations of CDT to assess whether or not there is a chronic excessive intake of alcohol. As others had already described it [6], it is particularly true for cirrhotic patients and patients at advanced stage of fibrosis. Unlike others [12], we did show a significant difference in CDT levels between cirrhotic patients with active alcohol consumption compared to abstinent patients.
Fig. 3. Box-and-whisker plots of carbohydrate-deficient transferin (%) and trisialotransferin (%) determined by capillary zone electrophoresis amongst heavy drinking patients. The middle line in the box represents the median and the extreme parts of the box represent the 25th and 75th percentile. The extreme points represent the minimum and maximum values for all the data. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
5
Fig. 4. Box-and-whisker plots of carbohydrate-deficient transferin (%) and trisialotransferin (%) determined by capillary zone electrophoresis amongst heavy drinking patients. The middle line in the box represents the median and the extreme parts of the box represent the 25th and 75th percentile. The extreme points represent the minimum and maximum values for all the data. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
However, when strictly considering manufacturer threshold of 1.6%, we noticed that cirrhotic patients could have totally normal values of CDT. Those considerations should be kept in mind when taking CDT into account for the driving license reinstatement program for instance. Maenhout et al. [21] proposed a threshold of CDT at 2.3% for regranting driving license. If we consider this cutoff for our patient group, sensitivity would have been as low as 49%, hence leading to regranting driving license to non abstinent people. Those considerations should preclude the use of CDT alone to assess whether or not there is still an excessive intake of alcohol. Considering this, we tried to find an additional biomarker which could be used in combination with CDT in order to improve detection of heavy drinking pattern. We especially focus on trisialotransferrin which was already described in several studies, notably in chronic liver disease and excessive alcohol intake [22,23]. In our study, we showed that we had a significant rise in trisialotransferrin results at every stage of fibrosis amongst heavy drinkers, compared to the control group. We didn't have any significant differences between the different fibrosis stages, unlike Gressner [22]. In his study, he showed a “U” curve when representing trisialotransferrin fraction depending on the fibrosis stage amongst patients with chronic liver disease (hepatitis C and hepatocellular carcinoma). He even had significantly lower values at F1, F2 and F3 stages compared to the control group. When we especially focused on cirrhotic patients, we noticed that we had significantly higher results for cirrhotic patients compared to the control group, regardless of the origin of cirrhosis or the status of alcohol consumption. We did even have higher trisialotransferrin results for abstinent people compared to cirrhotic patients with active alcohol consumption, suggesting that trisialotransferrin rise is irreversible despite weaning. When combining trisialotransferrin and CDT, we noticed that we had better performances of this combination than for CDT alone. If we added GGT and MCV to trisialotransferrin and carbohydrate-deficient
transferrin (cluster II), sensitivity and specificity were even better. To our knowledge, this is the first report describing the clinical interest of combining these biomarkers together. Furthermore, when using cluster II parameters, we noticed a significant difference between alcoholic cirrhotic patients and abstinent people, and between alcoholic cirrhotic patients and non alcoholic cirrhotic patients. We could therefore discriminate cirrhosis of alcoholic origin from cirrhosis due to other causes. However, we have to highlight that trisialotransferrin shouldn't be used alone, since it isn't a specific biomarker of chronic alcohol consumption but a marker of liver injury which could rise in various conditions, such as chronic liver disease [22] and anorexia nervosa [24] and in patients with cancer due to co-migration of polysaccharides with trisialotransferrin [25]. This shift in glycosylation pattern has already been described and studied by several people [22,26–28]. Several hypotheses have been proposed to explain this shift, mainly concerning glycosyltransferases and sialidase activities. Ethanol seems to down regulate Galβ1, 4 GlcNAc and α2,6-sialyltransferase. This down regulation seems to be correlated with a defect in glycosylation. On the contrary, sialidase activity seems to rise up. It is also important to focus on the pre analytical and analytical variables, potentially influencing the results. Nowadays, there are several methods available to measure transferrin isoforms. But there are no consensus available on standardization and several questions are remaining, regarding the isoforms to measure, the technique that should be used, the reference material or the best threshold for interpretation [29,30]. There are also others challenges regarding measurements of transferrin fractions, such as di-tri bridging [31], transferrin variants, congenital disorders of glycosylation syndromes [32], … which prevent an accurate measurement of the different fractions [33]. Moreover, some conditions are now well known to give false positive results for CDT testing, such as pregnancy [34].
Table 5 Comparison of carbohydrate-deficient transferrin and trisialotransferrin between alcoholic cirrhotic patients, former alcohol cirrhotic patients (abstinent), and non alcoholic cirrhotic patients.
Alcoholic cirrhosis (n = 28) Alcoholic cirrhosis (abstinent patients) (n = 21) Non alcoholic cirrhosis (n = 23) a
CDT
p-valuea
Trisialotransferrin
p-valuea
Mean (CI95)
Mann–Whitney test
Mean (CI95)
t-test
1.6 (0.85–2.3) 0.7 (0.59–0.81) 0.75 (0.62–0.89)
– p = 0.0001 p = 0.0003
5.1 (4.5–5.7) 6.1 (5.4–6.9) 5.2 (4.5–5.9)
– p = 0.036 p = 0.881
Comparison with alcoholic cirrhosis group.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
6
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Fig. 5. Mean and 95% Cl of combination of carbohydrate-deficient transferin, trisialotransferin corpuscular volume and gammaglutamyl transferase amongst cirrhotic patients. Combination of those parameters was achieved by logistic regression. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
Conclusion In our study, we confirmed the poor performances of CDT alone to detect chronic excessive alcohol intake in patients with hepatic injury (cirrhosis, steatosis, fibrosis). We showed a potential benefit of using trisialotransferrin in addition to CDT, in order to improve diagnosis performances. All laboratories should keep in mind that a normal CDT value together with an increased trisialotransferrin level, should not exclude an excessive alcohol intake. Disclosure The authors declare no conflict of interest. References [1] Aertgeerts B, Buntinx F, Ansoms S, Fevery J. Questionnaires are better than laboratory tests to screen for current alcohol abuse or dependence in a male inpatient population. Acta Clin Belg 2002;57(5):241–9. [2] Lieber. Medical disorders of alcoholism. N Engl J Med 1995;19(333):1058–65. [3] Bortolotti F, De Paoli G, Tagliaro F. Carbohydrate-deficient transferrin (CDT) as a marker of alcohol abuse: a critical review of the literature 2001–2005. J Chromatogr B 2006;841:96–109. [4] Stibler H. Carbohydrate-deficient transferrin in serum: a new marker of potentially harmful alcohol consumption reviewed. Clin Chem 1991;37(12):2029–37. [5] Meerkerk G, Njoo K, Bongers I, Trienekens P, JA vO. Comparing the diagnostic accuracy of carbohydrate-deficient transferrin, gamma-glutamyltransferase, and mean cell volume in a general practice population. Alcohol Clin Exp Res 1999;23(6):1052–9. [6] Imbert-Bismut F, Naveau S, Morra R, Munteanu M, Ratziu V, Abella A, et al. The diagnostic value of combining carbohydrate-deficient transferrin, fibrosis, and steatosis biomarkers for the prediction of excessive alcohol consumption. Eur J Gastroenterol Hepatol 2009;21(1):18–27. [7] Glanz J, Grant B, Monteiro M, Tabakoof B, WHO/ISBRA obot. WHO/ISBRA study on state and trait markers of alcohol use and dependence: analysis of demographc, behavioral, physiologic, and drinking variables that contribute to dependence and seeking treatment. Alcohol Clin Exp Res 2002;26(7):1047–61.
[8] Alves PS, Camilo EA, Correia JP. The SGOT/SGPT ratio in alcoholic liver disease. Acta Med Port 1981;3:255–60. [9] Salaspuro M. Use of enzymes for the diagnosis of alcohol-related organ damage. Enzyme 1987;37(1–2):87–107. [10] Rinck D, Frieling H, Freitag A, Hillemacher T, Bayerlein K, Kornhuber J, et al. Combinations of carbohydrate-deficient transferrin, mean corpuscular eythrocyte volume, gamma-glutamyltransferase, homocysteine and folate increase the significance of biological markers in alcohol dependent patients. Drug Alcohol Depend 2007;89:60–5. [11] Fleming MF, Anton RF, Spies CD. A review of genetic, biological, pharmacological, and clinical factors that affect carbohydrate-deficient transferrin levels. Alcohol Clin Exp Res 2004;28(9):1347–55. [12] Gonzalo P, Pecquet M, Bon C, Gonzalo S, Radenne S, Augustin-Normand C, et al. Clinical performance of the carbohydrate-deficient transferrin (CDT) assay by the Sebia Capillarys2 system in case of cirrhosis. Interest of the Bio-Rad %CDT by HPLC test and Siemens N-Latex CDT kit as putative confirmatory methods. Clin Chim Acta 2012;413:712–8. [13] Nahon P, Kettaneh A, Tengher-Barna J, Ziol M, de Ledinghen V, Douvin C, et al. Assessment of liver fibrosis using transient elastography in patients with alcoholic liver disease. J Hepatol 2008;49(6):1062–8. [14] Foucher J, Chanteloup E, Vergniol J, Castéra L, Le Bail B, Adhoute X, et al. Diagnosis of cirrhosis by transient elastography (FibroScan): a prospective study. Gut 2006;55:403–8. [15] Bota S, Herkner H, Sporea I, Salzl P, Sirli R, Neghina AM, et al. Meta-analysis: ARFI elastography versus transient elastography for the evaluation of liver fibrosis. Liver Int 2013:1–10. [16] Wong GL-H. Transient elastography: kill two birds with one stone. World J Hepatol 2013;5(5):264–74. [17] Qayyum A, Chen DM, Breiman RS, Westphalen AC, Yeh BM, Jones KD, et al. Evaluation of diffuse liver steatosis by ultrasound, computed tomography, and magnetic resonance imaging: which modality is best? Clin Imaging 2009;33(2):110–5. [18] Berzigotti A, Ashkenazi E, Reverter E, Abraldes JG, Bosch J. Non-invasive diagnostic and prognostic evaluation of liver cirrhosis and portal hypertension. Dis Markers 2011;31:129–38. [19] Zekanovic D, Ljubicic N, Boban M, Nikolic M, Delic-Brkljacic D, Gacina P, et al. Doppler ultrasound of hepatic and system hemodynamics in patients with alcoholic liver cirrhosis. Dig Dis Sci 2010;55:458–66. [20] DeLong E. Comparing the areas under two or more correlated receiver operating characteristics curves: a nonparametric approach. Biometrics 1988;44(3):837–45. [21] Maenhout T, Baten G, De Buyzere M, Delanghe J. Carbohydrate deficient transferrin in a driver's license regranting program. Alcohol Alcohol 2012;47(3):253–60. [22] Gressner OA, Jafari S, Erkens M, Gao C, Stanzel S, Gressner AM. Evaluation of serum percent trisialotransferrin as potential predictive biomarker of hepatocellular dedifferentiation in chronic liver disease. Clin Chim Acta 2009;403:188–93. [23] Ramdani B. Analyte comigrating with trisialotransferrin during capillary zone electrophoresis of sera from patients with cancer. Clin Chem 2003;49(11):1854–64. [24] Arndt T, Erkens M, Holtkamp K, Keller T, Gressner AM. High prevalence of increased trisialotransferrin concentrations in patients with anorexia nervosa: implications for determination of carbohydrate-deficient transferrin. Clin Chim Acta 2007;379:150–3. [25] Ramdani B, Nuyens V, Codden T, Perpete G, Colicis J, Lenaerts A, et al. Analyte comigrating with trisialotransferrin during capillary zone electrophoresis of sera from patients with cancer. Clin Chem 2003;49(11):1854–64. [26] Waszkiewicz N, Szajda SD, Zalewska A, Szulc A, Kepka A, Minarowska A, et al. Alcohol abuse and glycoconjugate metabolism. Folia Histochem Cytobiol 2012;50(1):1–11. [27] Gong M, Castillo L, Redman RS, Garige M, Hirsch K, Azuine M, et al. Down regulation of liver Galb1, 4GlcNAc a2, 6-sialyltransferase gene by ethanol significantly correlates with alcoholic steatosis in humans. Metab Clin Exp 2008;57:1663–8. [28] Gong M, Garige M, Hirsch K, Lakshman MR. Liver Galb1, 4GlcNAc a2,6-sialyltransferase is down-regulated in human alcoholics: possible cause for the appearance of asialoconjugates. Metab Clin Exp 2007;56:1241–7. [29] Weykamp C, Wielders JP, Helander A, Anton RF, Bianchi V, Jeppsson J-O, et al. Toward standardization of carbohydrate-deficient tranfserrin (CDT) measurements: III. Performance of native serum and serum spiked with disialotransferrin proves that harmonization of CDT assays is possible. Clin Chem Lab Med 2012;51(5):991–6. [30] Bianchi V, Ivaldi A, Raspagni A, Arfini C, Vidali M. Use of carbohydrate-deficient transferrin (CDT) and combination of GGT and CDT (GGT–CDT) to assess heavy alcohol consumption in traffic medicine. Alcohol Alcohol 2010;45(3):247–51. [31] Arndt T, Van der Meijden B, Wielders J. Atypical serum transferrin isoform distribution in liver cirrhosis studied by HPLC, capillary electrophoresis and transferrin genotyping. Clin Chim Acta 2008;394(1–2):42–6. [32] Parente F, Mew NAh, Jaeken J, Gilfix BM. A new capillary zone electrophoresis method for the screening of congenital disorders of glycosylation (CDG). Clin Chim Acta 2010;411:64–6. [33] Caslavska J, Thormann W. Monitoring of alcohol markers by capillary electrophoresis. J Sep Sci 2013;36:75–95. [34] Kenan N, Larsson A, Axelsson O, Helander A. Changes in transferrin glycosylation during pregnancy may lead to false-positive carbohydrate-deficient transferrin (CDT) results in testing fo riskful alcohol consumption. Clin Chim Acta 2011;412(1–2):129–33.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Anne Tamigniau 1, Pierre Wallemacq, Diane Maisin ⁎ Laboratory of Clinical Chemistry, Saint-Luc University Hospital, Université Catholique de Louvain, Brussels, Belgium
a r t i c l e
i n f o
Article history: Received 14 March 2014 Received in revised form 9 May 2014 Accepted 13 May 2014 Available online xxxx Keywords: Carbohydrate-deficient transferrin Trisialotransferrin Cirrhosis Fibrosis Steatosis Heavy drinking Capillary electrophoresis
a b s t r a c t Objectives: Carbohydrate-deficient transferrin is a well-known biomarker widely used for detection of chronic excessive alcohol intake. However, under certain clinical conditions particularly frequently met amongst heavy drinkers (steatosis, fibrosis, cirrhosis…), it isn't a reliable biomarker. In this study, we tried to find additional biomarkers to CDT in order to improve detection of chronic excessive alcohol intake. Design and methods: We conducted a retrospective cohort study from December 2007 to December 2009. We focused mainly on three different groups: heavy drinking patients with active alcohol consumption (n = 243), cirrhotic patients (abstinent patients and non alcoholic cirrhosis, n = 44) and control group (n = 85). Results: In our study, CDT showed a poor sensitivity for diagnosis of heavy drinking patients (around 63%, and even lower) for cirrhotic patients and patients at advanced stage of fibrosis. Combination of CDT with trisialotransferrin enabled to improve significantly sensitivity and specificity (p-value AUC ROC b 0.001). When adding mean corpuscular volume and gamma-glutamyltransferase to this first combination, performances were even better (p-value b 0.001). This second cluster enabled to make a statistically significant difference between cirrhotic patients with active alcohol consumption compared to abstinent cirrhotic patients and to non alcoholic cirrhotic patients (p-value b 0.001). Conclusion: From our study, trisialotransferrin seems to be a useful additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction Alcohol abuse remains a major health concern, with a prevalence of alcohol dependent patients accounting for up to 12.4% amongst Belgian hospitalized patients [1]. Former studies have already estimated that 20 to 30% of admissions in hospitals could be, directly or indirectly, related to alcohol disorders [2]. Various biochemical markers have been developed over the years to detect excessive alcohol consumption, notably for occupational medicine and forensic use. Carbohydrate-deficient transferrin (CDT) still remains one of the most widespread and easiest to use [3]. It has been demonstrated that a consumption of alcohol of 60 to 80 g per day over the past month could well correlate with an abnormal rise in CDT values [4]. However, its poor diagnostic sensitivity limits its use much more as a confirmatory and follow-up marker, than as a screening one [5]. Moreover, some studies have shown that CDT isn't a reliable marker under certain clinical conditions, such as steatosis, advanced fibrosis and cirrhosis [3,6]. Those clinical conditions are particularly frequently ⁎ Corresponding author. E-mail address: [email protected] (D. Maisin). 1 Present address: Hematology Laboratory, CHU UCL Mont-Godinne Dinant, Université Catholique de Louvain, Yvoir, Belgium.
met amongst heavy drinkers (N60 g/day or N210 g alcohol/week) [7], which precludes the use of carbohydrate-deficient transferrin to evaluate excessive alcohol consumption in those patients. In our study, we evaluated the performances of well-known biochemical markers of alcohol abuse such as AST–ALT ratio, MCV, GGT and CDT for heavy drinking patients [8–10]. We also especially focused on trisialotransferrin fraction which seemed, from our clinical practice, to increase with steatofibrosis and cirrhosis. One of the aims of this study was to determine if trisialotransferrin could whether or not be considered as an additional potential biomarker of chronic alcohol abuse in combination with others already used in clinical practice. We finally focused on patients suffering from cirrhosis, caused either by excessive alcohol consumption or not, and explored the potential use of trisialotransferrin and CDT to detect excessive alcohol drinkers (active and abstinent) amongst them.
Material and method Patients and study design We conducted a retrospective cohort study from December 2007 to December 2009. We selected three groups of patients: heavy drinking
http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
2
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
patients with active alcohol consumption, cirrhotic patients (abstinent patients and non alcoholic cirrhosis), and control patients. 272 alcoholdependent patients hospitalized in our Hepatology department for alcohol withdrawal treatment were selected in our heavy drinking group. Exclusion criteria were defined as follows: concomitant liver diseases, HIV patients, transferrin variants, immunosuppressive treatments, and uninterpretable electrophoretic profile [11]. From our initial group of 272 patients, 29 were excluded based on these criteria. The 243 patients included in this group were mostly men (66%), around 50 years old. 44 patients suffering from cirrhosis were recruited from our Transplant unit (before transplantation). Cause of cirrhosis and status of alcohol consumption were obtained from the pre-transplant visit. Exclusion criteria were the same as for the heavy drinking group, except for the concomitant liver diseases criteria. In this group, 21 patients were suffering from cirrhosis due to previous chronic alcohol consumption (abstinence at the time of the study), whereas 23 cirrhosis were due to other causes (viral hepatitis, cholangitis…). Those patients were mostly men (75%), aged around 55 years old. 85 control patients were recruited from our outpatient clinic. Inclusion criteria were defined as follows: no record in our Hepatology department concerning alcohol abuse, no evidence of liver injury and HIV, and no laboratory evidence of chronic alcohol abuse. Patient selection was conducted in order to reach similar proportions in terms of age and gender as the patients recruited in our study. This study was carried out in agreement with our local ethics committee. Alcohol consumption Alcohol consumption was evaluated at the entrance of the patient in our Alcohology department (start of abstinence program) for the heavy drinking group. From the Hepatology patients' files, we were able to extract the data relative to the self-reported alcohol consumption. Wines, beers, spirits and liqueurs were all considered as relevant in the evaluation of the daily consumption. Every patient included in our study had a self-reported consumption N 210 g alcohol/week and could thus be considered as heavy drinker [7].
Table 1 Characteristics of patients and controls included in the study. Heavy drinking group n = 243
Cirrhosis group n = 44
Control group n = 85
48.8 (0.7)
56.4 (1.5)
51.8 (1.7)
162 81 2
33 11 3
55 30 1.8
Fibrosis stage (n) Not assessed Failure F0–F1 F2 F3 F4
13 (5.4%) 11 (4.6%) 114 (47%) 34 (14%) 27 (11%) 44 (18%)
– – – – – –
– – – – – –
Liver injury (n) Not assessed Suboptimal examination No evidence of liver injury Steatosis Cirrhosis Others
2 (1%) 2 (1%) 34 (14%) 175 (72%) 28 (11%) 2 (1%)
– – – – 44 (100%) –
– – – – – –
Origin of cirrhosis (n) Chronic alcohol abuse Chronic viral infections Primary biliary cirrhosis Primary sclerosing cholangitis Cholestatic disease Others
28 – – – – –
21 6 5 1 1 10
– – – – – –
Biological markers, mean (SE) AST (IU/L) ALT (IU/L) AST/ALT ratio GGT (IU/L) MCV (fl) CDT (%)
75.6 (4.6) 56.3 (2.9) 1.4 (0.04) 248 (27) 95.7 (0.5) 4.3 (0.3)
55.8 (5.8) 40.5 (4.6) 1.5 (0.09) 117 (19) 90.3 (1.8) 0.73 (0.04)
22.4 (0.5) 19.8 (0.7) 1.2 (0.05) 24.6 (1.2) 86.6 (0.4) 0.76 (0.03)
Demographics Age, years, mean (SE) Gender Male (n) Female (n) Sex ratio (M/F)
Biological markers Aspartate aminotransferase (AST), alanine aminotransferase (ALT) and gamma-glutamyltransferase (GGT) were determined on heparinized plasma samples on DxC 800© from Beckman Coulter (Analis, Suarlée, Belgium). Prior to measurement, samples were centrifuged for 10 min at 3000 rpm. Reference values ranged from 7 to 50 IU/L for GGT, from 14 to 63 IU/L for ALT and from 6 to 33 IU/L for AST. Mean corpuscular volume (MCV) was determined on EDTA whole blood samples on Sysmex XE-2100 (Hoeilaart, Belgium). Reference values ranged from 85 to 95 fl for patients over 18 years old.
had to strictly determine if the result was pathological or not, results below 1.6% were considered negative [12]. Trisialotransferrin referred to isoforms with a pI at 5.6. For all patients, good isoform separation was obtained without bridging phenomenon.
CDT and trisialotransferrin Carbohydrate deficient transferrin (CDT) and trisialotransferrin measures were carried out on serum. The blood specimen were centrifuged at the reception in the laboratory, and the samples were then stored to − 20 °C until analysis (gentle thawing at 4 °C overnight). CDT and trisialotransferrin were measured by capillary zone electrophoresis on a Capillarys2 system from Sebia (Issy-les-Moulineaux, France) and data from electrophoretic profiles were extracted by the Phoresis 5.4.1 software. After complete iron saturation, transferrin isoforms were separated according to their isoelectric point (pI) and detection was achieved by UV determination at 200 nm. CDT fractions referred to diasialotransferrin (isoforms with a pI at 5.7). Asialotransferrin wasn't included in the determination of the CDT results. CDT results were considered pathological at or above 1.6%, negative below 1.3%, a gray zone existing between 1.3 and 1.6%. When we
Fig. 1. Number of cases of cirrhosis and steatosis according to fibrosis stage. Patients at early stage of fibrosis (FD–F1, F2) showed, most of the time, steatosis or no or minimal evidence of lesion. At advanced stage of fibrosis (F3, F4), patients were mostly suffering from steatosis and cirrhosis.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
3
Table 2 Sensitivity and efficiency of biomarkers for diagnosing heavy drinking pattern.
AST/ALT ratio MCV GGT CDT CDT CDT—fibrosis stage F0–F1 F2 F3 F4
Cutoff
Sensitivity (%)
CI95
Efficiency (%)
1.5 95 50 1.3 1.6
30.6 51.4 74.3 72.0 63.4
24.8–36.8 44.9–57.9 68.3–79.7 65.9–77.6 56.9–69.4
43.8 63.9 80.9 78.3 72.9
1.6 1.6 1.6 1.6
71.9 70.6 48.1 59.1
62.7–79.9 52.5–84.9 28.7–68.0 43.2–73.7
83.9 91.6 87.5 86.0
Liver fibrosis—METAVIR Scoring System Liver fibrosis was assessed by transient elastography—Fibroscan© from Echosens (Paris, France). This technique is an ultrasound-based method consisting of evaluating liver stiffness based on a shear wave velocity (kPa). Considering the shear wave velocity, patients were classified according to the METAVIR scoring system for fibrosis [13,14]. The results were only considered reliable when 10 valid measurements were achieved (success rate 100%) and the technique couldn't be performed with patients suffering from ascites [15,16].
Fig. 2. Receiver operating characteristic (ROC) curves of biomarkers for the detection of heavy drinking pattern (N210 g alcohol/week). The diagonal line represents an AUROC of 0.5 (null hyphothesis). The closer the curve is to the upper left corner, the better is the overall accuracy of the biomarker.
Evaluation of steatosis and cirrhosis by Doppler Ultrasound (US) Steatosis and cirrhosis were assessed by Doppler UltraSound (US) imaging performed on Acuson Sequoia 512© from Siemens (Munich, Germany). Ultrasound examinations revealed three main types of observations: 1) Normal echogenicity of the liver without any evidence of lesion, 2) increased echogenicity of the liver highly suggestive of steatosis, and 3) changes in liver morphology (nodular liver surface, coarse echopattern…) combined with signs of portal hypertension, highly suggestive of cirrhosis [17]. The main US signs of portal hypertension considered here were splenomegaly, alterations of portal vein blood flow velocity and alterations of resistive index of artery at porta hepatis and of intrahepatic branches [18,19]. Statistical analysis Normal distribution of variables was assessed by D'Agostino– Pearson test and outliers were detected with the Grubbs' test. T-test for independent samples was performed for parameters with a (log) normal distribution, whereas Mann–Whitney test was preferred for data that weren't normally distributed. P values were considered statistically significant below 0.05. Sensitivity and efficiency were determined for the different biomarkers in order to assess their potential performances for diagnosing heavy drinking pattern. Efficiency was defined as the ratio between the true determinations (true positive and true negative) and the total number of samples. Combination of biomarkers was achieved by logistic regression and was calculated as follows: logit(p) for cluster I = −10.05 + (5.46 × CDT) + (1.36 × trisialotransferrin) and logit(p) for
cluster II = − 30.28 + (4.03 × CDT) + (1.35 × trisialotransferrin) + (0.19 × MCV) + (0.11 × GGT). Logit(p) was back-transformed to p by the following formula: p = 1 / (1 + e−logit(p)). Analysis and comparison of receiver operating characteristic (ROC) curves was carried out according to Long et al. methodology [20]. We used MedCalc 12.7.0 (MedCalc software, Ostend, Belgium) and GraphPad Prism 6 (GraphPad Software Inc, La Jolla, USA) to perform all these statistical analysis. Results Cirrhosis and steatosis amongst heavy drinkers, according to fibrosis stage Patients were classified according to fibrosis stage (METAVIR scoring system) and to the results of their Doppler Ultrasound (Table 1). In Fig. 1, we represented the number of patients and their liver injury in regards to their corresponding fibrosis stage. We noticed that, at early stage of fibrosis (F0–F1, F2), patients were most of the time suffering from steatosis, or even, didn't show any evidence of liver injury. On the other side, at advanced stage of fibrosis (F3, F4), we observed an increased incidence of cirrhosis together with a decreased incidence of absence of liver injury. Diagnosis of heavy drinking Data relative to sensitivity and efficiency of biomarkers are summarized in Table 2. Sensitivity didn't exceed 75% for the different biomarkers tested in our study (range from 31 to 74%). CDT showed a quite poor sensitivity for diagnosis heavy drinking patients, around
Table 3 Receiver operating characteristics (ROC) curves analysis.
CDT Trisialotransferrin CDT–Trisialo CDT–Trisialo–MCV–GGT a b
AUC ROC
CI95
Cutoffb
Sensitivity
Specificity
p-valuea
0.90 0.79 0.96 0.99
0.86–0.93 0.73–0.82 0.93–0.98 0.97–0.99
1.0 3.9 0.73 0.80
79.4 72.0 86.8 92.9
89.4 74.1 94.1 98.8
– p = 0.0006 p b 0.0001 p b 0.0001
Comparison with CDT. Corresponding to the Youden Index.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
4
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Table 4 Carbohydrate-deficient transferrin and trisialotransferrin amongst heavy drinking patients.
Control group (n = 85) Fibrosis stage F0–F1 (n = 112) F2 (n = 34) F3 (n = 27) F4 (n = 44) a
CDT
p-valuea
Trisialotransferrin
p-valuea
Mean (CI95)
Mann–Whitney test
Mean (CI95)
t-test
0.76 (0.71–0.81)
–
3.2 (3.0–3.4)
–
5.5 (4.4–6.5) 5.6 (3.5–7.7) 3.1 (1.6––4.5) 2.8 (1.8–3.7)
p p p p
4.5 (4.2–4.7) 4.3 (3.9–4.8) 4.2 (3.8–4.6) 4.7 (4.4–5.1)
p p p p
b b b b
0.0001 0.0001 0.0001 0.0001
b 0.0001 b 0.0001 = 0.0001 b 0.0001
Comparison with control group.
63%. When considering fibrosis stage and liver injury, we noticed that sensitivity of CDT was even lower for cirrhotic patients at advanced stage of fibrosis. Combination of biomarkers was evaluated to explore if sensitivity could, whether or not, be improved for diagnosing heavy drinking patients. Two clusters of markers were studied: cluster I consisted of combination of CDT and trisialotransferrin, while cluster II consisted of combination of CDT, trisialotransferrin, GGT and MCV. All data relative to receiver operating characteristic curve are summarized in Table 3. Area under curve (AUC) of cluster I was significantly higher compared to AUC associated to CDT alone (p-value b0.0001). When considering cluster II, we showed that this combination was associated with a significantly higher AUC compared to cluster I and CDT alone (p-value b0.0001), reaching a sensitivity of 92.9% and a positive predictive value of 91.5% (Fig. 2). However, it should be kept it mind that such a result is directly dependent on the prevalence in our cohort, and could thus not be generalized to other groups. Trisialotransferrin and CDT amongst heavy drinkers Data relative to trisialotransferrin and CDT amongst heavy drinking patients are summarized in Table 4. We observed that CDT was statistically different from the control group at every stage of fibrosis (p-value b0.0001). However, we noticed that CDT was higher in early stage of fibrosis, F0–F1 and F2, compared to late stage of fibrosis, F3 and F4 (p-value b0.05) (Fig. 3 left). In patients at advanced stage of fibrosis, we noticed that some of them don't reach the manufacturer threshold of 1.6%, with sensitivity of CDT as low as 48% (Table 2). When focusing on heavy drinking patients at cirrhosis stage, we noticed that only 8 out of the 28 patients (b29%) had pathological values of CDT (≥1.6%). Trisialotransferrin was also statistically different from the control group at every stage of fibrosis. We didn't show any significant difference between the different fibrosis stages (Fig. 3 right).
Trisialotransferrin and CDT amongst cirrhotic patients Cirrhotic patients were divided into three main groups: alcoholic cirrhotic patients with active consumption, former alcoholic cirrhotic patients (abstinent) and non alcoholic cirrhotic patients. CDT in alcoholic cirrhotic patients with active consumption was significantly higher than in the control group and in the other two groups (p-values b0.001). No significant differences were observed between the control group and abstinent and non alcoholic cirrhosis (Fig. 4 left) (Table 5). Trisialotransferrin was significantly higher in the three groups compared to the control group. When comparing abstinent patients to active alcohol consumers, we noticed that trisialotansferrin was significantly higher in abstinent patients (Fig. 4 right). In order to assess if cirrhotic patients with active alcohol consumption could be distinguished from former alcoholic cirrhotic patients (abstinent), we tried to use our cluster II to see whether or not there was a difference between those two groups (Fig. 5). We observed that cluster II score was significantly higher in every group compared to the control group (p-values b 0.0001). Cirrhotic patients with active consumption seemed to have a significantly higher cluster II score than abstinent patients (p-value b0.05) and than non alcoholic cirrhotic patients (p-value b 0.001).
Discussion Our study confirmed some performance limitations of CDT to assess whether or not there is a chronic excessive intake of alcohol. As others had already described it [6], it is particularly true for cirrhotic patients and patients at advanced stage of fibrosis. Unlike others [12], we did show a significant difference in CDT levels between cirrhotic patients with active alcohol consumption compared to abstinent patients.
Fig. 3. Box-and-whisker plots of carbohydrate-deficient transferin (%) and trisialotransferin (%) determined by capillary zone electrophoresis amongst heavy drinking patients. The middle line in the box represents the median and the extreme parts of the box represent the 25th and 75th percentile. The extreme points represent the minimum and maximum values for all the data. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
5
Fig. 4. Box-and-whisker plots of carbohydrate-deficient transferin (%) and trisialotransferin (%) determined by capillary zone electrophoresis amongst heavy drinking patients. The middle line in the box represents the median and the extreme parts of the box represent the 25th and 75th percentile. The extreme points represent the minimum and maximum values for all the data. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
However, when strictly considering manufacturer threshold of 1.6%, we noticed that cirrhotic patients could have totally normal values of CDT. Those considerations should be kept in mind when taking CDT into account for the driving license reinstatement program for instance. Maenhout et al. [21] proposed a threshold of CDT at 2.3% for regranting driving license. If we consider this cutoff for our patient group, sensitivity would have been as low as 49%, hence leading to regranting driving license to non abstinent people. Those considerations should preclude the use of CDT alone to assess whether or not there is still an excessive intake of alcohol. Considering this, we tried to find an additional biomarker which could be used in combination with CDT in order to improve detection of heavy drinking pattern. We especially focus on trisialotransferrin which was already described in several studies, notably in chronic liver disease and excessive alcohol intake [22,23]. In our study, we showed that we had a significant rise in trisialotransferrin results at every stage of fibrosis amongst heavy drinkers, compared to the control group. We didn't have any significant differences between the different fibrosis stages, unlike Gressner [22]. In his study, he showed a “U” curve when representing trisialotransferrin fraction depending on the fibrosis stage amongst patients with chronic liver disease (hepatitis C and hepatocellular carcinoma). He even had significantly lower values at F1, F2 and F3 stages compared to the control group. When we especially focused on cirrhotic patients, we noticed that we had significantly higher results for cirrhotic patients compared to the control group, regardless of the origin of cirrhosis or the status of alcohol consumption. We did even have higher trisialotransferrin results for abstinent people compared to cirrhotic patients with active alcohol consumption, suggesting that trisialotransferrin rise is irreversible despite weaning. When combining trisialotransferrin and CDT, we noticed that we had better performances of this combination than for CDT alone. If we added GGT and MCV to trisialotransferrin and carbohydrate-deficient
transferrin (cluster II), sensitivity and specificity were even better. To our knowledge, this is the first report describing the clinical interest of combining these biomarkers together. Furthermore, when using cluster II parameters, we noticed a significant difference between alcoholic cirrhotic patients and abstinent people, and between alcoholic cirrhotic patients and non alcoholic cirrhotic patients. We could therefore discriminate cirrhosis of alcoholic origin from cirrhosis due to other causes. However, we have to highlight that trisialotransferrin shouldn't be used alone, since it isn't a specific biomarker of chronic alcohol consumption but a marker of liver injury which could rise in various conditions, such as chronic liver disease [22] and anorexia nervosa [24] and in patients with cancer due to co-migration of polysaccharides with trisialotransferrin [25]. This shift in glycosylation pattern has already been described and studied by several people [22,26–28]. Several hypotheses have been proposed to explain this shift, mainly concerning glycosyltransferases and sialidase activities. Ethanol seems to down regulate Galβ1, 4 GlcNAc and α2,6-sialyltransferase. This down regulation seems to be correlated with a defect in glycosylation. On the contrary, sialidase activity seems to rise up. It is also important to focus on the pre analytical and analytical variables, potentially influencing the results. Nowadays, there are several methods available to measure transferrin isoforms. But there are no consensus available on standardization and several questions are remaining, regarding the isoforms to measure, the technique that should be used, the reference material or the best threshold for interpretation [29,30]. There are also others challenges regarding measurements of transferrin fractions, such as di-tri bridging [31], transferrin variants, congenital disorders of glycosylation syndromes [32], … which prevent an accurate measurement of the different fractions [33]. Moreover, some conditions are now well known to give false positive results for CDT testing, such as pregnancy [34].
Table 5 Comparison of carbohydrate-deficient transferrin and trisialotransferrin between alcoholic cirrhotic patients, former alcohol cirrhotic patients (abstinent), and non alcoholic cirrhotic patients.
Alcoholic cirrhosis (n = 28) Alcoholic cirrhosis (abstinent patients) (n = 21) Non alcoholic cirrhosis (n = 23) a
CDT
p-valuea
Trisialotransferrin
p-valuea
Mean (CI95)
Mann–Whitney test
Mean (CI95)
t-test
1.6 (0.85–2.3) 0.7 (0.59–0.81) 0.75 (0.62–0.89)
– p = 0.0001 p = 0.0003
5.1 (4.5–5.7) 6.1 (5.4–6.9) 5.2 (4.5–5.9)
– p = 0.036 p = 0.881
Comparison with alcoholic cirrhosis group.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052
6
A. Tamigniau et al. / Clinical Biochemistry xxx (2014) xxx–xxx
Fig. 5. Mean and 95% Cl of combination of carbohydrate-deficient transferin, trisialotransferin corpuscular volume and gammaglutamyl transferase amongst cirrhotic patients. Combination of those parameters was achieved by logistic regression. ⁎p-value b0.05, ⁎⁎p-value b0.001, and ⁎⁎⁎p-value b0.0001.
Conclusion In our study, we confirmed the poor performances of CDT alone to detect chronic excessive alcohol intake in patients with hepatic injury (cirrhosis, steatosis, fibrosis). We showed a potential benefit of using trisialotransferrin in addition to CDT, in order to improve diagnosis performances. All laboratories should keep in mind that a normal CDT value together with an increased trisialotransferrin level, should not exclude an excessive alcohol intake. Disclosure The authors declare no conflict of interest. References [1] Aertgeerts B, Buntinx F, Ansoms S, Fevery J. Questionnaires are better than laboratory tests to screen for current alcohol abuse or dependence in a male inpatient population. Acta Clin Belg 2002;57(5):241–9. [2] Lieber. Medical disorders of alcoholism. N Engl J Med 1995;19(333):1058–65. [3] Bortolotti F, De Paoli G, Tagliaro F. Carbohydrate-deficient transferrin (CDT) as a marker of alcohol abuse: a critical review of the literature 2001–2005. J Chromatogr B 2006;841:96–109. [4] Stibler H. Carbohydrate-deficient transferrin in serum: a new marker of potentially harmful alcohol consumption reviewed. Clin Chem 1991;37(12):2029–37. [5] Meerkerk G, Njoo K, Bongers I, Trienekens P, JA vO. Comparing the diagnostic accuracy of carbohydrate-deficient transferrin, gamma-glutamyltransferase, and mean cell volume in a general practice population. Alcohol Clin Exp Res 1999;23(6):1052–9. [6] Imbert-Bismut F, Naveau S, Morra R, Munteanu M, Ratziu V, Abella A, et al. The diagnostic value of combining carbohydrate-deficient transferrin, fibrosis, and steatosis biomarkers for the prediction of excessive alcohol consumption. Eur J Gastroenterol Hepatol 2009;21(1):18–27. [7] Glanz J, Grant B, Monteiro M, Tabakoof B, WHO/ISBRA obot. WHO/ISBRA study on state and trait markers of alcohol use and dependence: analysis of demographc, behavioral, physiologic, and drinking variables that contribute to dependence and seeking treatment. Alcohol Clin Exp Res 2002;26(7):1047–61.
[8] Alves PS, Camilo EA, Correia JP. The SGOT/SGPT ratio in alcoholic liver disease. Acta Med Port 1981;3:255–60. [9] Salaspuro M. Use of enzymes for the diagnosis of alcohol-related organ damage. Enzyme 1987;37(1–2):87–107. [10] Rinck D, Frieling H, Freitag A, Hillemacher T, Bayerlein K, Kornhuber J, et al. Combinations of carbohydrate-deficient transferrin, mean corpuscular eythrocyte volume, gamma-glutamyltransferase, homocysteine and folate increase the significance of biological markers in alcohol dependent patients. Drug Alcohol Depend 2007;89:60–5. [11] Fleming MF, Anton RF, Spies CD. A review of genetic, biological, pharmacological, and clinical factors that affect carbohydrate-deficient transferrin levels. Alcohol Clin Exp Res 2004;28(9):1347–55. [12] Gonzalo P, Pecquet M, Bon C, Gonzalo S, Radenne S, Augustin-Normand C, et al. Clinical performance of the carbohydrate-deficient transferrin (CDT) assay by the Sebia Capillarys2 system in case of cirrhosis. Interest of the Bio-Rad %CDT by HPLC test and Siemens N-Latex CDT kit as putative confirmatory methods. Clin Chim Acta 2012;413:712–8. [13] Nahon P, Kettaneh A, Tengher-Barna J, Ziol M, de Ledinghen V, Douvin C, et al. Assessment of liver fibrosis using transient elastography in patients with alcoholic liver disease. J Hepatol 2008;49(6):1062–8. [14] Foucher J, Chanteloup E, Vergniol J, Castéra L, Le Bail B, Adhoute X, et al. Diagnosis of cirrhosis by transient elastography (FibroScan): a prospective study. Gut 2006;55:403–8. [15] Bota S, Herkner H, Sporea I, Salzl P, Sirli R, Neghina AM, et al. Meta-analysis: ARFI elastography versus transient elastography for the evaluation of liver fibrosis. Liver Int 2013:1–10. [16] Wong GL-H. Transient elastography: kill two birds with one stone. World J Hepatol 2013;5(5):264–74. [17] Qayyum A, Chen DM, Breiman RS, Westphalen AC, Yeh BM, Jones KD, et al. Evaluation of diffuse liver steatosis by ultrasound, computed tomography, and magnetic resonance imaging: which modality is best? Clin Imaging 2009;33(2):110–5. [18] Berzigotti A, Ashkenazi E, Reverter E, Abraldes JG, Bosch J. Non-invasive diagnostic and prognostic evaluation of liver cirrhosis and portal hypertension. Dis Markers 2011;31:129–38. [19] Zekanovic D, Ljubicic N, Boban M, Nikolic M, Delic-Brkljacic D, Gacina P, et al. Doppler ultrasound of hepatic and system hemodynamics in patients with alcoholic liver cirrhosis. Dig Dis Sci 2010;55:458–66. [20] DeLong E. Comparing the areas under two or more correlated receiver operating characteristics curves: a nonparametric approach. Biometrics 1988;44(3):837–45. [21] Maenhout T, Baten G, De Buyzere M, Delanghe J. Carbohydrate deficient transferrin in a driver's license regranting program. Alcohol Alcohol 2012;47(3):253–60. [22] Gressner OA, Jafari S, Erkens M, Gao C, Stanzel S, Gressner AM. Evaluation of serum percent trisialotransferrin as potential predictive biomarker of hepatocellular dedifferentiation in chronic liver disease. Clin Chim Acta 2009;403:188–93. [23] Ramdani B. Analyte comigrating with trisialotransferrin during capillary zone electrophoresis of sera from patients with cancer. Clin Chem 2003;49(11):1854–64. [24] Arndt T, Erkens M, Holtkamp K, Keller T, Gressner AM. High prevalence of increased trisialotransferrin concentrations in patients with anorexia nervosa: implications for determination of carbohydrate-deficient transferrin. Clin Chim Acta 2007;379:150–3. [25] Ramdani B, Nuyens V, Codden T, Perpete G, Colicis J, Lenaerts A, et al. Analyte comigrating with trisialotransferrin during capillary zone electrophoresis of sera from patients with cancer. Clin Chem 2003;49(11):1854–64. [26] Waszkiewicz N, Szajda SD, Zalewska A, Szulc A, Kepka A, Minarowska A, et al. Alcohol abuse and glycoconjugate metabolism. Folia Histochem Cytobiol 2012;50(1):1–11. [27] Gong M, Castillo L, Redman RS, Garige M, Hirsch K, Azuine M, et al. Down regulation of liver Galb1, 4GlcNAc a2, 6-sialyltransferase gene by ethanol significantly correlates with alcoholic steatosis in humans. Metab Clin Exp 2008;57:1663–8. [28] Gong M, Garige M, Hirsch K, Lakshman MR. Liver Galb1, 4GlcNAc a2,6-sialyltransferase is down-regulated in human alcoholics: possible cause for the appearance of asialoconjugates. Metab Clin Exp 2007;56:1241–7. [29] Weykamp C, Wielders JP, Helander A, Anton RF, Bianchi V, Jeppsson J-O, et al. Toward standardization of carbohydrate-deficient tranfserrin (CDT) measurements: III. Performance of native serum and serum spiked with disialotransferrin proves that harmonization of CDT assays is possible. Clin Chem Lab Med 2012;51(5):991–6. [30] Bianchi V, Ivaldi A, Raspagni A, Arfini C, Vidali M. Use of carbohydrate-deficient transferrin (CDT) and combination of GGT and CDT (GGT–CDT) to assess heavy alcohol consumption in traffic medicine. Alcohol Alcohol 2010;45(3):247–51. [31] Arndt T, Van der Meijden B, Wielders J. Atypical serum transferrin isoform distribution in liver cirrhosis studied by HPLC, capillary electrophoresis and transferrin genotyping. Clin Chim Acta 2008;394(1–2):42–6. [32] Parente F, Mew NAh, Jaeken J, Gilfix BM. A new capillary zone electrophoresis method for the screening of congenital disorders of glycosylation (CDG). Clin Chim Acta 2010;411:64–6. [33] Caslavska J, Thormann W. Monitoring of alcohol markers by capillary electrophoresis. J Sep Sci 2013;36:75–95. [34] Kenan N, Larsson A, Axelsson O, Helander A. Changes in transferrin glycosylation during pregnancy may lead to false-positive carbohydrate-deficient transferrin (CDT) results in testing fo riskful alcohol consumption. Clin Chim Acta 2011;412(1–2):129–33.
Please cite this article as: Tamigniau A, et al, Could trisialotransferrin be used as an additional biomarker to CDT in order to improve detection of chronic excessive alcohol intake? Clin Biochem (2014), http://dx.doi.org/10.1016/j.clinbiochem.2014.05.052