Clinical Biochemistry 40 (2007) 278 – 281
Tissue-polypeptide-specific antigen levels in diabetic patients with normal and pathological biochemical profiles Javier Rodriguez a,b , Rubén Varela-Calviño a , Manuel Garrido Outeiriño a,b , Santiago Rodríguez-Segade a,b , Felix Camiña a,⁎ a
Department of Biochemistry and Molecular Biology, University of Santiago de Compostela, Spain b Clinical Biochemistry Division, University Clinical Hospital, Santiago de Compostela, Spain Received 18 May 2006; received in revised form 19 October 2006; accepted 5 November 2006 Available online 5 December 2006
Abstract Objectives: To identify causes for the raised TPS levels seen in diabetic patients. Design and methods: Relationships between TPS levels and biochemical markers for glycaemic control, hepatic dysfunction and renal dysfunction were investigated in 402 diabetic patients, none with evidence of cancer. Results: Median TPS level (range) was 34.6 (19–276) U/L in controls versus 40.5 (16–691) U/L in type 1 diabetes mellitus (T1DM) patients and 53 (6–1654) U/L in type 2 diabetes mellitus (T2DM) patients. TPS levels above the 95th percentile were observed in 26.1% diabetic patients and in 68.6% of these diabetic patients, raised TPS was associated with clinical complications or biochemical indicators of hepatic and/or renal dysfunction. Conclusions: The raised mean TPS levels seen in diabetic patients appear to be mainly due to the existence of hepatic or renal dysfunction. © 2006 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Tissue-polypeptide-specific antigen (TPS); Clinical complications; Diabetes mellitus; Proliferative marker
Introduction Tissue-polypeptide-specific antigen (TPS) is a soluble fragment derived from the carboxy-terminal end of cytokeratin 18 (CK-18) [1]. CK-18 is a protein with a molecular weight of 45 kDa that appears to be over-expressed by rapidly growing epithelial cells. Raised CK-18 levels are associated with epithelial cell proliferation and turnover [2]. Raised TPS levels are a marker of tumour activity, not necessarily tumour mass. TPS is not specific to any particular cancer, but is a general marker for proliferating epithelial cells [2,3]. In combination with tumour markers such as CA15-3, CA125, CA19-9, CEA, and PSA, TPS determinations may offer significant advantages for cancer characterization and management [4]. However, the ⁎ Corresponding author. Departamento de Bioquímica y Biología Molecular, Universidad de Santiago de Compostela, Campus Sur, s/n, 15782-Santiago de Compostela, Spain. Fax: +34 981 594912. E-mail address:
[email protected] (F. Camiña).
use of TPS for this monitoring is limited due to its nonspecificity because physiological conditions such as pregnancy [5], and physiopathological situations such as hepatic or renal injury [6,7], heart transplantation [8] or diabetes [9,10], all present with elevated TPS levels. Regarding diabetes, there are no data about the prevalence of raised TPS levels or the association of raised TPS levels with clinical and/or biochemical conditions. The aim of the present study was to identify factors that may explain the raised TPS levels seen in certain diabetic patients. Materials and methods Clinical status of patients Between January 2002 and January 2004, we enrolled 402 diabetic out-patients attending the diabetes clinic of the University Hospital Complex (Santiago de Compostela, Spain), all of whom had been prescribed insulin or oral anti-
0009-9120/$ - see front matter © 2006 The Canadian Society of Clinical Chemists. All rights reserved. doi:10.1016/j.clinbiochem.2006.11.006
J. Rodriguez et al. / Clinical Biochemistry 40 (2007) 278–281
diabetics in this clinic. Diabetic patients were classified into T1DM (88 patients) or T2DM (314 patients), in accordance with the criteria of the American Diabetes Association [11]; none of the patients showed evidence of neoplastic disease after review of their clinical records for the period between 6 months before and after blood sampling. Age, sex, time since onset of diabetes, and clinical complications (when present) were also recorded (Table 1). Assessment of clinical complications was done by specialized personnel as previously described [12]. The study was conducted in accordance with the Declaration of Helsinki and was approved by the hospital's Ethics Committee. All subjects gave prior written informed consent.
279
intervals are expressed as the 2.5th and 97.5th percentiles of the control group. For TPS and UAER, a cut-off value (80 U/L for TPS) was established at the 95th percentile of controls. The significance of differences between groups was assessed by Mann–Whitney tests (non-normal data). Associations between two variables were measured with Spearman's rho (two continuous variables) and point-biserial correlations (rpb) (dichotomous variable vs. continuous variable). Relationships between two dichotomous variables and odds ratio significances were assessed by Fisher's exact test. In two-sided tests, p values less than 0.05 were considered statistically significant. All analyses were performed with the statistics package SPSS for Windows (version 11.0, SPSS Inc, Chicago, IL, USA).
Analytical methods Results In whole blood samples from all subjects we determined fasting HbA1c by the ion exchange method (Hi-AUTOA1c HA8121, Menarini Diagnostic, Barcelona, Spain). In serum samples we determined fructosamine by the NBT method (Cobas Mira, Roche, Barcelona, Spain); glucose, total protein, albumin, creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), gamma-glutamyl transferase (γ-GT), alkaline phosphatase (AP), bilirubin, cholinesterase (Che), urea and uric acid by standard commercial methods (Boehringer Mannheim, Barcelona, Spain). Fasting serum TPS concentrations were determined by a sequential chemiluminescent enzyme-labeled immunometric assay using the specific monoclonal antibody M3 (IDL Biotech, previously Beki Diagnostic AB, Bromma, Sweden). In addition, the urine albumin excretion rate (UAER, μg/min) was estimated by analysis of a 24-h urine collection, with a monoclonal antibody in a BNA II nephelometer. Reference intervals for each laboratory parameter were previously determined using a group of 132 healthy subjects from the same geographic area and age range [13].
TPS levels in diabetic patients We determined serum TPS levels in 132 controls and 402 diabetic patients, specifically 88 T1DM patients (21.9%) and 314 T2DM patients (78.1%). Median TPS levels were 34.6 (19–276) U/L in controls versus 40.5 (16–691) U/L in T1DM patients and 53.0 (6–1654) U/L in T2DM patients (Table 1), but these differences were not statistically significant. TPS levels were higher in men than in women, but again this difference was not statistically significant. Of the 402 diabetic subjects, 26.1% showed TPS levels above the 80 U/L healthy-population 95% cut-off value. The proportion of subjects with TPS > 80 U/L was 14.8% in the T1DM and 29.3% in the T2DM patients. The p value for the two-tailed Fisher's exact test to compare these two proportions is 0.0059, indicating a statistically significant association between TPS and diabetes type, with an odds ratio (OR) of 0.418 (95% confidence interval, CI, 0.221–0.791). This OR indicates that TPS>80 U/L is twice as likely in the T2DM patients.
Statistical methods TPS and clinical complications of diabetes Data normality was checked by tests of skewness and kurtosis. Results are expressed as median (range). Reference Table 1 Clinical characteristics for the diabetic patients and controls Characteristic Sex (number) Men Women Age median (range) Duration of diabetes (year, mean ± SD) Nephropathy (%) Retinopathy (%) Neuropathy (%) Hypertension (%) Macroangiopathy (%) TPS median (U/L) (range)
All (N = 402)
T1DM (N = 88)
T2DM Control (N = 314) (N = 132)
186 216 61 (8–87) 8±7
45 43 34 (8–67) 6±7
141 173 72 (47–87) 9±8
70 62 51 (8–88) –
14.3 49.5 49.8 28.2 25.0 56.4 (6–1654)
11.9 37.6 35.1 9.2 7.3 40.5 (16–691)
16.6 60.0 60.1 39.3 33.7 53.0 (6–1654)
– – – – – 39.6 (19–276)
T1DM = type 1 diabetic patients; T2DM = type 2 diabetic patients.
The patients were grouped in accordance with clinical complications (presence or absence of each complication). TPS values in each group are shown in Fig. 1. Only patients with macro-angiopathy showed significantly higher serum TPS levels than patients with no complications (p = 0.042). Thus, only those diabetic patients with coronary heart disease showed an OR of 3.333 (95% CI 1.901–5.830) for TPS > 80 U/L, whereas in the remaining patients, without cardiovascular disease, no association with TPS levels was observed. A small, but statistically significant, point-biserial correlation between TPS level and hypertension was also observed (rbp = 0.101; CI = 0.001–0.197; p = 0.050). No significant correlations were observed between the other clinical diabetes complications and TPS levels. Glycaemic control and TPS levels We investigated whether serum levels of glucose and fructosamine and whole blood glycosylated haemoglobin were
280
J. Rodriguez et al. / Clinical Biochemistry 40 (2007) 278–281
Fig. 1. TPS levels in patients with (white bars) and without (black bars) clinical complications. CV, cardiovascular disease; HTA, hypertension; NP, neuropathy; RT, retinopathy; NeP, nephropathy. * indicates significant difference (p < 0.05).
related to TPS levels in our patient series. None of the three parameters differed significantly between T1DM and T2DM patients, although the difference in fructosamine serum levels were near-significant (p = 0.063) (data not shown). No statistically significant correlation was observed with TPS levels (Table 2). Hepatic dysfunction markers and TPS levels We analyzed standard markers of hepatic dysfunction in our diabetic patients. These markers fell within the reference intervals of the healthy control group in 89.9% (albumin), 90.2% (ALT), 94.6% (AST), 77.8% (γ-GT), 84.2% (Che), 88.3% (AP) and 94.4% (bilirubin) of patients. The only significant difference between the two diabetic populations was in γ-GT levels, which showed higher levels among the T2DM patients than among the T1DM patients (16 U/L (3–619) vs. 8 U/L (4–717), respectively; p = 0.031), although both values are within the reference intervals. We found positive correlations between the TPS levels and AST, ALT, γ-GT and AP levels (Table 2). Renal dysfunction markers and TPS level In both the T1DM and T2DM groups, renal dysfunction markers were within the reference intervals. None of them showed statistical differences between both diabetic groups except for UAER (7.8 μg/min (1.6–1148) in T1DM patients, vs. 9.6 μg/min, (0.2–5592) in T2DM patients; p = 0.048). In T1DM subjects, TPS levels were positively correlated with serum urea and uric acid levels, while in T2DM subjects the TPS levels were positively correlated with serum creatinine and uric acid levels and with UAER (Table 2). Discussion The present study evaluated the prevalence of raised TPS levels in diabetic patients and whether these raised mean TPS levels may be secondary to specific complications of the
disease, not to the diabetes itself. Our results show that the prevalence of raised TPS levels (> 80 U/L) is higher among both T1DM and T2DM patients than in the healthy population (14.8% and 29.3%, respectively). We found TPS levels >80 U/ L are twice as likely in the T2DM patients, and this could suggest TPS levels are an age-dependent factor, as T2DM patient group is much older. However, we suspect that this is related to the health condition of T2DM patients, since the percentage of patients with clinical complications, associated with elevated TPS levels, is higher in this group (see Table 1). Our results indicate that this higher prevalence of raised TPS levels is to a great extent related to the presence of various dysfunctions and clinical states (more prevalent in T2DM patients) that have previously been shown to influence TPS levels in non-diabetic individuals [7,10]. In particular, we found that indicators of hepatic and renal dysfunction correlated well with elevated TPS levels. Correlations with hepatic markers were significant in both groups of patients and are in accordance with previous reports [6]. This association is no surprise, since cytokeratin 18 is present in hepatocytes, and together with cytokeratin 8, is the only cytokeratin present in this cell type [2]; it is also expressed by the biliary epithelia [14], which would explain the association between raised TPS and γ-GT levels. Furthermore, other groups have shown that there is a correlation between raised TPS levels and γ-GT activity [6,14], because both TPS levels and γ-GT activity are raised in cellular stress situations provoked by toxic agents like alcohol. Raised TPS levels have also been documented in nonneoplastic patients with nephropathy [7]. We found positive associations of raised TPS levels with hypertension and with renal dysfunction markers, mainly in T2DM patients. The observed positive relationship with creatinine would suggest a negative relationship with glomerular filtration rate (GFR), as has been reported previously [7]. The positive relationship seen between TPS levels and UAER supports this interpretation Table 2 Correlations between TPS and biochemical parameters in two diabetic groups Parameter
N
T1DM rho*
p
N
T2DM rho*
p
Glucose Fructosamine HbA1c Albumin AST ALT γ-GT AP BIL Che Cre UAER Urea Uric acid GFR
82 82 82 82 82 82 82 74 74 82 82 33 74 74 50
0.150 − 0.310 − 0.131 0.011 0.326 0.332 0.235 0.395 − 0.250 0.130 0.087 − 0.100 0.305 0.246 0.146
0.893 0.784 0.909 0.920 0.003† 0.002† 0.033†
