| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Centre de Recherche Public–Santé, Laboratoire National de Santé, Toxicologie, Université du Luxembourg, 162A, Ave de la Faïencerie, L-1511 Luxembourg;
aauthor for correspondence: fax 352-22-13-31, e-mail appenzel{at}cu.lu
Measurements of carbohydrate-deficient transferrin (CDT) are used to detect alcohol abuse (1)(2)(3)(4)(5)(6). Various patient factors(7)(8)(9)(10)(11) and comigration of analytes with transferrin (Tf) isoforms(12)(13) can affect the results of the measurements. Recent studies have questioned the influence of serum and blood storage conditions on CDT and yielded varying (and sometimes contradictory) findings(14)(15)(16)(17). Moreover, the stability of CDT may vary among serum samples(14)(16).
We studied CDT and Tf isoforms in serum specimens stored at 25 °C, 4 °C, and –20 °C. We used 100 discarded serum samples; 8 were from alcohol abusers in treatment (4 with increased disialo-Tf and the presence of asialo-Tf; 4 with increased disialo-Tf but without asialo-Tf), and 6 were from moderate drinkers (CDT within reference values). For sera from alcohol abusers, the alcohol abuse diagnosis was based on psychological evaluation; CDT, 
-glutamyltransferase, and transaminase (aspartate aminotransferase and alanine aminotransferase) concentrations; and erythrocyte mean corpuscular volume. Each of the 14 specimens was divided into 3 samples and stored in glass vials (4 mL; Supelco 27137) at room temperature (25 °C in the dark), in a refrigerator (4 °C), or in a freezer (–20 °C). Samples were then analyzed periodically for 28–81 days [mean (SD), 51.6 (14.7) days]. The main reason for discontinuing follow-up was the appearance of solid particles in the serum. For frozen samples (–20 °C), the time between thawing and refreezing was <5 min. Serum Tf was analyzed (13) by use of capillary electrophoresis (P/ACE 5500) and a reagent set (CEofix; ANALIS). Electropherograms displayed all major Tf isoforms: asialo-Tf (when present), disialo-Tf, trisialo-Tf, tetrasialo-Tf, pentasialo-Tf, and hexasialo-Tf. The signal was integrated in the valley-to-valley mode. Measurements were made in triplicate.
CDT was expressed as the percentage of total Tf:
 | (1) |
where 0sialo-Tf is asialo-Tf, 2sialo-Tf is disialo-Tf, 3sialo-Tf is trisialo-Tf, and so forth. Because CDT calculations included all Tf isoforms, relative error on the CDT calculation was expressed as 
CDT/CDT, derived from Eq. 1
by the logarithmic differential method, giving:
 | (2) |
 |
where Xsialo-Tf = |Xsialo-Tft0 – Xsialo-Tfti|.
Hexasialo-Tf was also expressed as percentage of total Tf, as was CDT in Eq. 1
. The hexasialo-Tf/pentasialo-Tf ratio was calculated similarly. Anti-human Tf (Q0327; Dako) was used for confirmatory identification of Tf isoforms. Serum was added to undiluted anti-Tf at ratios of 3:1 and 1:1 (by volume) and mixed by 5 simple pipette aspirations/expressions.
The storage of serum at room temperature markedly changed the pattern of Tf isoforms (Fig. 1
). In this sample, the decrease in asialo-Tf and disialo-Tf led to a 50% decrease in measured CDT (Table 1
; also see Fig. 1
in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue11). All samples behaved similarly, with lower CDT after storage at 25 °C (Table 1
). Sera from alcohol abusers displayed the most marked decrease in CDT, and in 3 samples (E, F, and H), CDT eventually decreased below the positive cutoff of 2%. In all sera stored at 4 °C and –20 °C (see Fig. 1
in the online Data Supplement), CDT was stable during the entire period (Table 1
). For frozen samples (–20 °C), repeated thawing and freezing had no apparent effect on CDT (see Fig. 1
in the online Data Supplement).
|
|
We used relative error (CDT/CDT), which takes into account the variation of all isoforms, to assess the validity of CDT calculation as a percentage of total Tf. CDT/CDT varied with storage temperature (see Fig. 2 in the online Data Supplement). At 4 °C and –20 °C, CDT/CDT was stable during the entire period of follow-up. This tendency was similar for all serum samples for which CDT/CDT was poorly correlated with time (Table 1
). For serum storage at 25 °C, CDT/CDT and time were significantly correlated (see Fig. 2 in the online Data Supplement). This tendency was observed for all serum samples studied here, but slopes varied widely among samples (Table 1
). Mean slopes of CDT/CDT vs time were higher for sera stored at room temperature [0.136 (0.070) for alcohol abusers and 0.080 (0.049) for moderate drinkers] than for sera stored at 4 °C [0.027 (0.042) for alcohol abusers and 0.013 (0.012) for moderate drinkers] or –20 °C [0.014 (0.011) for alcohol abusers and 0.012 (0.012) for moderate drinkers].
The CDT decrease occurring at 25 °C was accompanied by other changes in the Tf isoform pattern (Fig. 1
). All isoforms decreased with time except for pentasialo-Tf, which remained stable or decreased slightly, and hexasialo-Tf, which increased (Fig. 1
; also see Fig. 1
in the online Data Supplement). Serum treatment with anti-Tf confirmed that the apparent increase in hexasialo-Tf was not attributable to comigration of other substances and also demonstrated the presence of another anti-Tf–reactive substance visible just after hexasialo-Tf (probably heptasialo-Tf; Fig. 1
; see also Fig. 3 in the online Data Supplement). We observed signals in lines t = 2 days, t = 8 days, and t = 16 days (Fig. 1
) for all sera at time points between 6.4 and 6.6 min but not later. Although the peak distributions seemed to indicate Tf isoforms, they were neither cancelled by treating serum with anti-Tf nor correlated with an increase or decrease in any Tf isoforms. We attributed this signal to unidentified serum degradation products having no direct relationship to Tf. The increase in hexasialo-Tf in serum samples stored at 25 °C led to values ranging from 4.4% to >26% (see Table 1
in the online Data Supplement). In comparison, the mean (SD) value for hexasialo-Tf determined in 100 fresh sera was 3.4 (0.6)%. The increases in hexasialo-Tf concentration were significantly correlated with time for all sera investigated here, but slopes were highly variable among sera (see Table 1
in the online Data Supplement). Increases in hexasialo-Tf also correlated with decreases in CDT (see Fig. 4 in the online Data Supplement). Correlation was more significant for serum samples from alcohol abusers (see Table 1
in the online Data Supplement), but slopes varied markedly among samples. It is notable that the serum samples with the smallest increase in hexasialo-Tf (samples J and L) also displayed more stable CDT (Table 1
). The hexasialo-Tf/pentasialo-Tf ratio also correlated with both time and a decrease in CDT (see Fig. 4 and Table 2 in the online Data Supplement), and the correlation was more significant in serum samples from alcohol abusers. After storage at 25 °C, some samples displayed hexasialo-Tf/pentasialo-Tf ratios up to 1.5 (see Table 2 in the online Data Supplement). In comparison, the mean (SD) value observed in 100 fresh sera was 0.256 (0.046). For samples stored at 4 °C and –20 °C, as observed for CDT, the hexasialo-Tf concentration and the hexasialo-Tf/pentasialo-Tf ratio remained stable over the whole follow-up period (see Fig. 1
in the online Data Supplement).
Two mechanisms may explain the observed serum impairment: degradation of Tf molecules (attributable to endogenous enzymes and/or bacterial contamination), leading to a decrease in total Tf, and fixation of free sialic acids [naturally present in blood in considerable amounts (18)] on Tf isoforms having N-glycans but without terminal sialic acid, causing dominance of forms with a higher number of fixed sialic acid residues. The intensity of these mechanisms is dependent, however, on proper serum characteristics, because impairment kinetics were highly variable among sera. Because of this interserum variability, it is not practical to propose a time limit for room storage; therefore, validity of the CDT measurement may be assessed by use of indicators such as the hexasialo-Tf concentration and the hexasialo-Tf/pentasialo-Tf ratio, because both are correlated with a decrease in CDT. Because laboratories use various procedures and methods to quantify Tf in serum, definite values cannot be proposed for assessing serum impairment with these indicators, and adoption of a universal CDT cutoff value for alcohol abuse remains a point of contention. It may be more advantageous for laboratories to compare suspicious samples with reference values observed for hexasialo-Tf and the hexasialo-Tf/pentasialo-Tf ratio in fresh sera and to consider that a serum with abnormally increased indicator values (for example, 2-fold above reference values, corresponding to 
6% for 6sialo-Tf and 0.35 for the 6sialo-Tf/5sialo-Tf ratio, respectively, in our experiments) is inappropriate for CDT analysis. According to such criteria, all sera tested here that were stored at 25 °C were still fit for CDT analysis within 5 days, 21% were unfit after 7 days, 36% were unfit after 10 days, and 64% were unfit after 15 days. An analysis performed in parallel with the Bio-Rad %CDT immunologic test showed comparable decreases in CDT for sera stored at 25 °C (data not shown), confirming that serum degradation also interferes with immunologic tests. Nevertheless, evaluation of serum degradation with hexasialo-Tf and the penta/hexasialo-Tf ratio is possible only with techniques giving access to all Tf isoforms (e.g., capillary electrophoresis and HPLC) and hence is not applicable to CDT analysis performed with current immunoassay tests.
In summary, serum may be stored for several weeks at 4 °C or –20 °C before CDT analysis. Storage at 25 °C is acceptable for 5 days but not >7 days. After prolonged exposure to room temperature, Tf degradation may be assessed by measuring the hexasialo-Tf concentration and the hexasialo-Tf/pentasialo-Tf ratio.
Acknowledgments
This work was supported by Grant SAN-02-001 from the Luxembourg Ministry of Health.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
