- Email: [email protected]
Effect of anticoagulants and storage temperatures on stability of plasma and serum hormones1
Clinical Biochemistry 34 (2001) 107–112
Effect of anticoagulants and storage temperatures on stability of plasma and serum hormones Margaret J. Evans*, John H. Livesey, M. Jane Ellis, Timothy G. Yandle Endolab, Department of Endocrinology, Christchurch Hospital, PO Box 4710, Christchurch, New Zealand Received 25 October 2000; received in revised form 1 February 2001; accepted 6 February 2001
Abstract Objectives: To determine the effect of different anticoagulants and storage conditions on the stability of hormones in plasma and serum. Design and methods: Human blood samples were collected from volunteers into EDTA, lithium heparin, sodium fluoride/potassium oxalate, or tubes without anticoagulant, plasma and serum left at ⫺20°C, 4°C or 30°C for 24 and 120 hours then assayed for ACTH, aldosterone, ␣-subunit, AVP, CRH, C-peptide, estradiol, FSH, glucagon, GH, IGF-1, IGFBP-3, insulin, leptin, LH, PPP, PTH, prolactin and VIP, or at room temperature for 0 to 72 hours (BNP, NT-BNP)(n ⫽ 6 per condition). Results: The anticoagulant altered the measured concentrations for 9 hormones when compared to EDTA. All hormones except ACTH were stable for ⬎120 hours in EDTA or fluoride at 4°C, but only 13 hormones were stable in all anticoagulants. At 30°C, 8 hormones were stable for ⬎120 hours in EDTA, and 3 hormones in all anticoagulants. BNP and NT-BNP were stable for ⬍24 hours when stored in EDTA or heparin at room temperature. Discussion: Storage of samples in EDTA plasma at 4°C is suitable for most hormones (except ACTH) for up to 120 hours. © 2001 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Plasma; Storage; Hormones; Anticoagulant; Serum; Temperature; Stability
1. Introduction With the development of referral centres specialising in the measurement of hormones, there has been an increase in the number of samples that require shipping from the place of collection to the laboratory where the analysis takes place. In our own laboratory, the most frequent enquiries from collection centres relate to the suitability of particular anticoagulants, and whether or not the sample needs special transport conditions. The latter, in particular, impacts on the convenience and cost of using a specialist laboratory outside the area from which the sample is collected. Previous studies have looked at the effect of different storage temperatures on a smaller number of analytes measured in either EDTA [1], or serum [2]. In the present study, we extended these observations to 22 hormones, and undertook to examine the effect of three of the most commonly used anticoagulants and serum, as well as the
* Corresponding author. Tel.: ⫹64-33640848; fax: ⫹64-33640818. E-mail address: [email protected] (M.J. Evans).
effect with each of these collection methods of storing samples frozen, at 4°C or at 30°C for 24 and 120 hours (1 and 5 days). These temperatures and time points were selected to differentiate between analytes which require immediate freezing, analytes which could be sent on ice (4°C) to arrive either overnight or within 5 days, or analytes which were stable either overnight or for 5 days at ambient temperature (likely maximum 30°C). Samples (whole blood, serum and plasma) from heart failure subjects for natriuretic hormone measurements were stored at room temperature at intervals from 0 to 72 hours to reproduce collection procedures in a general practitioner’s surgery and transport to the laboratory before or after separation.
2. Methods 2.1. Collection of samples Blood samples (approximately 260 ml) were collected from non-fasting volunteers between 0800 and 1000 hours at the Outpatient Bleeding Service, Department of Endocri-
0009-9120/01/$ – see front matter © 2001 The Canadian Society of Clinical Chemists. All rights reserved. PII: S 0 0 0 9 - 9 1 2 0 ( 0 1 ) 0 0 1 9 6 - 5
108
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
nology. Volunteers gave informed consent, and the ethical committee for our institution approved the study. Sequential samples were collected into each of the following four blood collection tubes: (a) EDTA (ethylenediaminetetra-acetic acid); (Vacuette威, Greiner Labortechnik, Austria); (b) Serum (Vacutainer威, Becton Dickinson Vacutainer System USA, Rutherford, NJ, USA); (c) Lithium heparin (Vacutainer威), and (d) Sodium fluoride/potassium oxalate (Vacutainer威). 2.2. Storage of samples After collection, samples were centrifuged at 4°C immediately (plasma) or after 10 min (serum) and the plasma and serum were separated. Five aliquots of each of the 4 samples from each volunteer were prepared on ice/water and immediately placed in the following storage conditions: a One aliquot at ⫺20°C* (reference temperature) b One aliquot at 4°C for 24 hours (1 day) c One aliquot at 4°C for 120 hours (5 days) d One aliquot at 30°C for 24 hours (1 day) e One aliquot at 30°C for 120 hours (5 days) *The freezer used for sample storage maintained a temperature of ⫺24°C, and was set to alarm at ⫺15°C. This is consequently referred to as ⫺20°C. At the completion of the storage phase, samples were held at ⫺20°C until assayed. Except where indicated, each analyte was measured in duplicate in samples from a minimum of 6 different volunteers for every collection method and storage condition. All samples from individual volunteers were measured in the same assay for each analyte. Samples collected for brain natriuretic peptide (BNP) had values at or near the detection limit of the assay. Therefore further samples were collected from heart failure subjects into EDTA or heparin, and stored as either whole blood or, after separation, at room temperature for 0, 6, 24, 48 and 72 hours. These samples were also analysed for N-terminal BNP (NT-BNP).
2.4. Statistical analysis 2.4.1. Anticoagulants and serum For each analyte, a two-way analysis of variance with subject and anticoagulant as the factors, was performed on log-transformed results from the ⫺20°C reference sample for each collection method. The results from serum, heparin and fluoride collections were each compared to those measured in EDTA plasma, and if there was a significant ( p ⬍ 0.05) difference, and if the mean difference was greater than 10%, this was considered a significant effect of anticoagulant. A difference of 10% was selected because changes of less than this are unlikely to be clinically significant. 2.4.2. Storage For each analyte (except BNP and NT-BNP), the results at 0, 24 and 120 hours for each temperature was expressed as a percent of the value in the reference aliquot (⫺20°C) for each anticoagulant and serum. The mean rate of change (k) was estimated by fitting to the data the equation: Y ⫽ e kt
(1)
where y is the fraction of initial immunoactivity remaining after time t. Fitting was performed using the non-linear least squares curve fitting feature of SigmaStat (Jandel Scientific, Chicago). Loss of immunoreactivity with time gave negative k values and increases positive values. Hormones were considered stable for the number of hours required for the median immunoreactivity, interpolated between observations using equation 1, to reach either 90% or 110% of the original value. For BNP and NT-BNP the statistical method was the same (although temperature and times were different) and analytes were considered to be stable up to the time at which the median loss reached 10%.
2.3. Hormones measured and methodology
3. Results
The following hormones were tested and a summary of the assay details are shown on Table 1: Adrenocorticotrophic hormone (ACTH), aldosterone, alpha-subunit (gonadotrophin), total (␣-subunit, total) and free (␣-subunit, free), arginine vasopressin (anti-diuretic hormone) (AVP), brain natriuretic peptide (BNP), corticotrophin-releasing hormone (CRH), C-peptide, follicle-stimulating hormone (FSH), glucagon, growth hormone (GH), insulin-like growth factor (IGF-1), IGF-binding protein 3 (IGFBP-3), insulin, leptin, luteinizing hormone (LH), N-terminal BNP (NT-BNP), estradiol, pancreatic polypeptide (PPP), parathyroid hormone (PTH), prolactin, urine C-peptide, vasoactive intestinal peptide (VIP).
3.1. Effect of anticoagulant Table 2 shows the effect of each anticoagulant for each of the analytes tested, as measured in the ⫺20°C samples and compared to the EDTA sample. Measured concentrations were significantly altered by the method of collection, with significant differences recorded for 6 of 22 analytes when serum was the matrix , 4 of 20 when fluoride was the anticoagulant, and 3 of 22 when heparin was the anticoagulant. For one analyte, aldosterone, all 3 matrices were significantly different from the result with EDTA. Results are statistically significant ( p ⬍ 0.05) with a difference of ⬍⫺10% or ⬎⫹10%.
Method
2-site radioimmunometric assay (Bioclone Australia PTY, Marrickville, NSW, Australia) ria after extraction of samples with acid/ethanol and a cryoprecipitation step. Antiserum from Dr B Breier, University of Auckland, New Zealand ria kit procedure (Nichols Institute, Ca, USA) Abbott IMX insulin kit (Abbott Laboratories, Diagnostic Division, Ill, USA) ria kit for human Leptin (LINCO Research Inc, Cat # HL81K, St Charles, Mo USA) As described for FSH ria (“in house” rabbit anti-NT-BNP antiserum), NT-BNP (1–21) as standard (Peninsula Laboratories Inc) ria (Estradiol-2, Clinical Assay™, Sorin Biomedica, Italy) ria (rabbit anti-human antiserum Eli Lilly and Company, Indianapolis, Ind USA) 2-site immunoradiometric assay (IRMA) (Nichols Institute, Ca, USA) As described for FSH ria after extraction using in acetone/acetic acid/water (80:19:1), “inhouse” antiserum
Separation
Units
I-AVP purified on HPLC I-BNP purified by HPLC
PEG
125
125 I-human VIP (Peninsula Labs Inc) PEG plus anti purified on HPLC ␥globulin
I-Bovine PPP (Eli Lilly) purified on HPLC
Solid phase anti ␥globulin
PEG
PEG
PEG
PEG
PEG Solid phase anti ␥globulin
NT-BNP(1-13)125I-Tyr14 purified on HPLC (9)
I-labelled hormone (IGF-1 GroPep, Adelaide, Australia), purified on HPLC.
125
125 I-Human glucagon (Sigma Chemical Co), purified on HPLC
125 I-C-Peptide (human [tyr0] CPeptide (Sigma Chemical Co), purified by HPLC
I-CRH purified by HPL
125
125
125
13 3–8 4 4.8
g/L g/ml pmol/L g/L
pmol/L mIU/L pmol/L
pmol/L ng/L
8.1
2.9
7 7.3
1.9 2.2
3
g/L
IU/L pmol/L
12.5
2.5
5.9
6
6 8.6
free 9 total 4
3.4 6.8
Intra assay CV %
ng/L
UI/L
pmol/L
pmol/L
pmol/L pmol/L
pmol/L Aldosterone-3, carboxymethyloxime- dextran-coated charcoal pmol/L [125I]histamine, HPLC purification 125 PEG plus anti g/L I-hCG ␣-subunit (Biogenesis UK), purified by HPLC ␥globulin
Tracer
0.7 3.8
DL mean
10–64 77–568
Range of values
4.1 4.8 14.5
5 14
5.5 5.8
6.8
10.4 4.1
10
5.4
18.6
5.5
5.3
8.5
0.1 ⱕ14.4 2.9
2 17.7
9–57 137–530 9–18.5
15–1140 91–396
1.5–25.9 12–628
ⱕ0.5 1.8
0.1
2.6–5.1 1.6– 13.51 3.8–49.5
119–296
0.4–21.0
69–167
3.5–47.9
471– 1427
1.0–1.9
0.10 6
1.3
0.19
13
0.2
2.0
0.2
free 13 free 0.2 free total 12 total 0.2 0.4–2.7 total 0.9–4.5 12 0.3 1.1–3.1 6.5 8.3 34–153
14.8 14
Inter assay CV %
b
ria, radioimmunoassay “In house” preparation of 125I-labelled hormone used the chloramine T method (8) and 125Iodine (Amersham, UK). c Purification of 125I-labelled hormone was by reverse phase high performance liquid chromatography (HPLC) using a Perkin-Elmer HPLC with a Brownlee RP300 column, unless otherwise stated. d PEG, polyethylene glycol e CV% calculated for concentrations within the range measured in the study (except for insulin) f Detection limit was defined as the smallest value which could be distinguished from zero at the 95% confidence limit. g Where the range of concentrations measured in this study are given for each analyte, these are taken from the EDTA, ⫺20°C sample results. 1 2 of 6 subjects reading lower than the calculated detection limit.
a
PTH Prolactin VIP
E2 PPP
LH NT-BNP
Leptin
IGF-BP3 Insulin
IGF-1
GH
Glucagon
FSH
C-Peptide
SELECT kitset with microparticle enzyme immunoassay (MEIA) technology (Abbott Laboratories, Ill, USA) ria using an “in house” antiserum
ria (4) after acetonitrile extraction of samples, “in house” antiserum ria (modified from (5) after extraction on C18 Sep-Pak cartridges (Waters Corp., Mass. USA), antiserum (RAS 9086 rabbit anti-BNP32 (human), Peninsula Lboratories Inc) ria (6,7) following extraction of samples and standards (human/rat CRH41, Peninsula Laboratories Inc) with methanol, rabbit anti-ovine CRH antiserum (gift from P Lowry, University of Reading, UK) ria (Go2o antiserum, Guildhay), synthetic human C-Peptide (Sigma Chemical Co, St Louis, Mo, USA) as standard
AVP BNP
CRH
ria: 2 “in-house” antisera, 1 binding free ␣-subunit, the other both the free and bound hormone
␣-subunit
ACTH two-site immunoradiometric assay (IRMA) (Nichols Institute, Ca, USA) Aldosterone ria (3)
Hormone
Table 1 Summary of assay details
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112 109
110
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
Table 2 Comparison of the effect of anticoagulant on hormone concentration as compared to EDTA in ⫺20°C samples (frozen immediately)
Table 4 Duration of stability of BNP and NT-BNP in hours when stored as whole blood or plasma at room temperature
Hormone
Heparin
Serum
Fluoride
Plasma
ACTH Aldosterone ␣-subunit (total) ␣-subunit (Free) AVP BNP* CRH C-peptide FSH Glucagon GH IGF-1 IGFBP-3 Insulin Leptin LH NT-BNP* Estradiol PPP PTH Prolactin VIP
⫺74% ⫹17% ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫺20%
⬃ ⫹15% ⬃ ⬃ ⫺42% ⬃ ⫹46% ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫹11% ⫹19% ⬃ ⬃ ⫺22%
⬃ ⫺12% ⫺13% ⬃ ⬃ NT ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫺28% ⬃ ⬃ ⬃ NT ⬃ ⫺11% ⬃ ⬃ ⬃
EDTA
Heparin
EDTA
Heparin
18 7
10 8
20 5
4
*n ⫽ 3 NT ⫽ not tested ⬃ ⫽ no significant difference (p ⬎ 0.05 and/or difference ⬍10%)
3.2. Effect of storage conditions and collection method Table 3 shows the stability of hormones collected into different anticoagulants and stored at 4°C and 30°C ex-
BNP NT-BNP
Whole Blood
pressed in hours, based on time taken to decrease or increase 10% from the reference value, as calculated from k values (see Methods). Measurements were not continued after 120 hours. Estradiol was not tested for storage conditions. Table 4 shows the stability of BNP and NT-BNP stored at room temperature as either whole blood or as plasma collected into EDTA or heparin. Urine C-peptide was stable at 4°C for ⬎120 hours, and at 30°C for 19 hours.
4. Discussion The aim of this study was to determine the effect of different anticoagulants (EDTA, heparin, sodium fluoride and serum) and time and temperature of storage on the results obtained for a range of hormones analysed in our laboratory (22 hormones in total). The anticoagulants selected are those commonly used in the blood collection process. BNP and NT-BNP were stored under conditions replicating collection in a physician’s surgery.
Table 3 Duration of stability of hormones in hours after collection into different anticoagulants and stored at 4°C or 30°C At 4°C
ACTH Aldosterone ␣subunit total ␣subunit free AVP CRH C-Peptide FSH Glucagon GH IGF-1 IGF-BP3 Insulin Leptin LH PPP PTH Prolactin VIP
At 30°C
EDTA
Heparin
Serum
Fluoride
EDTA
Heparin
Serum
Fluoride
18 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120
NT ⬎120 ⬎120 ⬎120 ⬎120 100 ⬎120 ⬎120 74 ⬎120 114 NT ⬎120 ⬎120 ⬎120 ⬎120 101 ⬎120 ⬎120
3 ⬎120 ⬎120 ⬎120 ⬎120 50 ⬎120 ⬎120 34 ⬎120 ⬎120 NT ⬎120 ⬎120 ⬎120 ⬎120 55 ⬎120 ⬎120
NT ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 NT ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120
8 14 ⬎120 60 96 18 ⬎120 ⬎120 69 62 ⬎120 ⬎120 31 ⬎120 ⬎120 ⬎120 36 109 25
NT 13 ⬎120 ⬎120 28 14 ⬎120 ⬎120 42 64 ⬎120 NT 21 69 59 53 33 ⬎120 13
1 16 ⬎120 86 28 11 ⬎120 ⬎120 30 66 ⬎120 NT 32 ⬎120 82 64 3 84 20
NT 15 ⬎120 96 39 15 ⬎120 ⬎120 114 61 88 NT 10 60 88 67 16 87 20
NT ⫽ Not tested. Duration of stability is the time elapsing until the median immunoreactivity, interpolated between observations using equation 1, reached 90 or 110% of the initial value.
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
When interpreting these results, it should be noted that different assay methods and antibodies may give different results [1, 10], there were small numbers of samples in each group, and that samples were handled and stored in a nonsterile manner after collection. The selection of a change of 10% as the limit of stability is arbitrary but has been used previously [1]. In some cases, results for hormones measured following an extraction step showed an apparent increase in immunoreactivity after storage, suggesting assay interference was a factor, or possibly release of bound hormone from a protein factor. When the different collection methods were compared to collection into EDTA, a significant effect was shown for several hormones. Serum affected the results of 6 hormones (aldosterone, AVP, CRH, oestradiol, PPP and VIP), fluoride affected 4 hormones (aldosterone, ␣-subunit (total), IGFBP-3, and PPP), and heparin affected 3 hormones (ACTH, aldosterone, and VIP). AVP measurements in heparin samples appeared to be lower (⫺36% of EDTA) but this was not significant. Some of the results were altered by as much as 40% and could be expected to affect the interpretation of the result if compared to a reference interval based on samples collected into EDTA. This suggests that samples sent to a laboratory should be collected into the appropriate anticoagulant for that laboratory’s methods. In our laboratory, we recommend the collection of samples into EDTA (unless otherwise recommended by a specific kit manufacturer) as our reference intervals are established using samples collected in this way. Using the ⫺20°C sample as the reference value, EDTA samples kept at 4°C for 24 to 120 hours showed no significant effect on the results for most hormones. The exception to this was ACTH, where incubations at 4°C for 24 hours had a significant negative effect. The requirement for ACTH to be stored and transported frozen has been shown previously [11, 12]. The use of dry ice is recommended for transport of ACTH samples, or alternatively, shipment in special containers of “pink ice”.1 In either case, transport time is limited to the time until the dry ice is fully sublimed or the “pink ice” completely thawed. Collection of samples into heparin similarly showed that most samples were stable at 4°C for 24 –120 hours (ACTH not tested), although CRH, glucagon, IGF-1 and PTH were stable for less than 120 hours. Serum samples were also stable for 24 hours at 4°C, except for ACTH, while incubation for 120 hours at this temperature affected ACTH, CRH, glucagon and PTH. When fluoride was used as the anticoagulant, all hormones (ACTH not tested) were stable for 120 hours at 4 °C. From these results it would appear that when compared to 1 Preparation of “pink ice”: 186 g ammonium chloride (industrial grade, Harcoss Chemicals, Christchurch) dissolved in 814 g water and sufficient rhodamine B added to colour the solution pink. When frozen this 18.6% solution melts at ⫺15.8 °C and since it is the eutectic mixture it maintains this temperature until it is nearly all melted.
111
the corresponding ⫺20°C sample, fluoride and EDTA were the most effective anticoagulants for storage and transport of samples at 4°C for up to 120 hours, followed by heparin then serum. However it must be noted that in some cases these different collection methods also affected the ⫺20°C results, and that the effect of storage relates to the ⫺20 °C sample of each anticoagulant, and not to the EDTA value. These results are similar to those of an earlier study from our laboratory using different analytical methods [1] that found that LH, FSH, GH, prolactin and insulin in EDTA plasma were stable at 4°C for at least 8 days. As might be expected, incubation of samples at 30 °C had a more marked effect on the measurements. For the EDTA samples, results for ACTH, aldosterone, and CRH were significantly affected within 24 hours, and by 120 hours, ␣-subunit (free), AVP, glucagon, GH, insulin, PTH, prolactin and VIP were also affected. With collection into heparin, aldosterone, CRH, insulin and VIP were significantly affected within 24 hours (no measurement made for ACTH), and by 120 hours, very few analytes were stable. Similarly for serum, incubation at 30°C for 24 hours affected ACTH, aldosterone, CRH, PPP, PTH and VIP, with most analytes unstable after 5 days. When fluoride was used as the anticoagulant and after 24 hours at 30 °C, aldosterone, CRH, insulin, PPP, PTH and VIP (ACTH not measured) were affected, and by 120 hours, very few hormones were stable. Livesey et al [1] similarly found that in EDTA plasma stored at 37°C, LH was stable for about 3 days, prolactin and GH for about 2 days, and insulin for less than 1 day. Kubasik et al [2] found that in serum, LH, FSH and prolactin were stable at either 4°C or 22°C for up to 14 days, but in our assays, only FSH was stable at 30°C for more than 1 day in serum. Similarly to the present study, Kubasik et al [2] showed that insulin in serum was unstable above 4°C. In a study investigating the effect of storing heparinized whole blood on hormone stability, Diver et al [13] showed no clinically significant changes in FSH, GH, oestradiol or prolactin stored up to 48 hours at room temperature, and up to a week at 4°C. They also found that ACTH was stable for less than 24 hours at room temperature. Our stability data for BNP at room temperature is in agreement with that published by others [14] and is greater than at 37°C [5]. The results contrast with those of Tsuji et al. [15] who found aprotinin was necessary to prevent rapid metabolism of BNP in plasma. The difference in apparent stability between various methods may be due to differences in stability of epitopes of BNP that are recognised by the different antisera. The stability of NT-BNP in this study agrees closely with a preliminary study previously published by us [9] showing measurable loss of NT-BNP immunoreactivity over 6 hours. Results in two published studies indicate greater stability of NT-BNP in plasma [16, 17] than those reported here. These differences are again likely to result from the use of different antisera and indeed agree with our unpublished results (Yandle) using antisera to different NT-BNP epitopes.
112
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
In summary, this study has shown that for many hormones, collection into EDTA or fluoride, and storage and transport at 0 – 4 °C is practicable. The exception to this guideline is ACTH which requires frozen transport and storage. Acknowledgments The authors thank Sarah L Norton, Endolab, Christchurch Hospital for co-ordinating the collection and storage of the samples and collating the data. We also thank the Endocrine Test Centre Nurses, Christchurch Hospital for collecting the samples, the staff members who volunteered to donate blood, and all the technical staff of Endolab, Christchurch Hospital, for performing the assays. References [1] Livesey JH, Hodgkinson SC, Roud HR, Donald RA. Effect of time, temperature and freezing on the stability of immunoreactive LH, FSH, TSH, growth hormone, prolactin and insulin in plasma. Clin Biochem 1980;13:151–155. [2] Kubasik NP, Ricotta M, Hunter T, Sine HE. Effect of duration and temperature of storage on serum analyte stability: Examination of 14 selected radioimmunoassay procedures. Clin Chem 1982:28:164 – 165. [3] Lun S, Espiner EA, Nicholls MG, Yandle TG. A direct radioimmunoassay for aldosterone in plasma. Clin Chem 1983;29:268 –271. [4] Sadler WA, Lynskey CP, Gilchrist NL, Espiner EA, Nicholls MG. A sensitive radioimmunoassay for measuring plasma antidiuretic hormone in man. NZ Medical J 1983;96:959 –963. [5] Yandle TG, Richards AM, Gilbert A, Fisher S, Holmes S, Espiner EA. Assay of brain natriuretic peptide (BNP) in human plasma— evidence for high molecular weight BNP as a major plasma component in heart failure. Clin Chem 1993;76:832– 838.
[6] Ellis MJ, Livesey JH, Donald RA. Circulating plasma corticotrophin releasing factor-like immunoreactivity. J Endocrinol 1988;117:299 – 307. [7] Ellis MJ, Schmidli RS, Donald RA, Livesey JH, Espiner EA. Plasma corticotrophin-releasing factor and vasopressin responses to hypoglycaemia in normal man. Clin Endocrinol 1990;32:93–100. [8] Landon J, Livanou T, Greenwood FC. The preparation and immunological properties of 131I labelled adrenocorticotropin. Biochem J 1967;105:1075–1083. [9] Hunt PJ, Richards AM, Nicholls MG, Yandle TG, Doughty RN, Espiner EA. Immunoreactive amino-terminal pro-brain natriuretic peptide (NT-proBNP)—a new marker of cardiac impairment. Clin Endocrinol 1997;47:287–296. [10] Donald RA. Radioimmunoassay of corticotropin (ACTH). In: Abrahams G, Ed. Handbook of Radioimmunoassay. Pp.319 –390. Marcel Dekker, New York, 1977. [11] Meakin JW, Tingey WH, Nelson DH. Catabolism of adrenocorticotropic hormone: The stability of adrenocorticotropic hormone in blood, plasma, serum and saline. Endocrinology 1960;66:59 –72. [12] Purves HD, Sirett NE. The stability of endogenous and exogenous corticotrophin in rat plasma as established by a bioassay in intact rats. J Endocrinol 1968;41:491– 497. [13] Diver MJ, Hughes JG, Hutton JL, West CR, Hipkin LJ. The long-term stability in whole blood of 14 commonly-requested hormone analytes. Ann Clin Biochem 1994;31:561–565. [14] Buckley MG, Marcus NJ, Yacoub MH, Singer DRJ. Prolonged stability of brain natriuretic peptide—importance for non-invasive assessment of cardiac function in clinical practice. Clinical Science 1998;95:235–239. [15] Tsuji T, Imagawa K, Masud H et al. Stabilization of human brain natriuretic peptide in blood samples. Clinical Chemistry 1994;40: 672– 673. [16] Karl J, Borgya A, Gallusser A et al. Development of a novel, NTerminal-proBNP (NT-proBNP) assay with a low detection limit. Scandinavian Journal of Clinical & Laboratory Investigation 1999;59 (Suppl 230);177–181. [17] Downie PF, Talwar S, Squire IB, Davies JE, Barnewtt DB, Ng LL. Assessment of the stability of N-terminal pro-brain natriuretic peptide in vitro: implications for assessment of left ventricular dysfunction. Clinical Science 1999;97:255–258.
Effect of anticoagulants and storage temperatures on stability of plasma and serum hormones Margaret J. Evans*, John H. Livesey, M. Jane Ellis, Timothy G. Yandle Endolab, Department of Endocrinology, Christchurch Hospital, PO Box 4710, Christchurch, New Zealand Received 25 October 2000; received in revised form 1 February 2001; accepted 6 February 2001
Abstract Objectives: To determine the effect of different anticoagulants and storage conditions on the stability of hormones in plasma and serum. Design and methods: Human blood samples were collected from volunteers into EDTA, lithium heparin, sodium fluoride/potassium oxalate, or tubes without anticoagulant, plasma and serum left at ⫺20°C, 4°C or 30°C for 24 and 120 hours then assayed for ACTH, aldosterone, ␣-subunit, AVP, CRH, C-peptide, estradiol, FSH, glucagon, GH, IGF-1, IGFBP-3, insulin, leptin, LH, PPP, PTH, prolactin and VIP, or at room temperature for 0 to 72 hours (BNP, NT-BNP)(n ⫽ 6 per condition). Results: The anticoagulant altered the measured concentrations for 9 hormones when compared to EDTA. All hormones except ACTH were stable for ⬎120 hours in EDTA or fluoride at 4°C, but only 13 hormones were stable in all anticoagulants. At 30°C, 8 hormones were stable for ⬎120 hours in EDTA, and 3 hormones in all anticoagulants. BNP and NT-BNP were stable for ⬍24 hours when stored in EDTA or heparin at room temperature. Discussion: Storage of samples in EDTA plasma at 4°C is suitable for most hormones (except ACTH) for up to 120 hours. © 2001 The Canadian Society of Clinical Chemists. All rights reserved. Keywords: Plasma; Storage; Hormones; Anticoagulant; Serum; Temperature; Stability
1. Introduction With the development of referral centres specialising in the measurement of hormones, there has been an increase in the number of samples that require shipping from the place of collection to the laboratory where the analysis takes place. In our own laboratory, the most frequent enquiries from collection centres relate to the suitability of particular anticoagulants, and whether or not the sample needs special transport conditions. The latter, in particular, impacts on the convenience and cost of using a specialist laboratory outside the area from which the sample is collected. Previous studies have looked at the effect of different storage temperatures on a smaller number of analytes measured in either EDTA [1], or serum [2]. In the present study, we extended these observations to 22 hormones, and undertook to examine the effect of three of the most commonly used anticoagulants and serum, as well as the
* Corresponding author. Tel.: ⫹64-33640848; fax: ⫹64-33640818. E-mail address: [email protected] (M.J. Evans).
effect with each of these collection methods of storing samples frozen, at 4°C or at 30°C for 24 and 120 hours (1 and 5 days). These temperatures and time points were selected to differentiate between analytes which require immediate freezing, analytes which could be sent on ice (4°C) to arrive either overnight or within 5 days, or analytes which were stable either overnight or for 5 days at ambient temperature (likely maximum 30°C). Samples (whole blood, serum and plasma) from heart failure subjects for natriuretic hormone measurements were stored at room temperature at intervals from 0 to 72 hours to reproduce collection procedures in a general practitioner’s surgery and transport to the laboratory before or after separation.
2. Methods 2.1. Collection of samples Blood samples (approximately 260 ml) were collected from non-fasting volunteers between 0800 and 1000 hours at the Outpatient Bleeding Service, Department of Endocri-
0009-9120/01/$ – see front matter © 2001 The Canadian Society of Clinical Chemists. All rights reserved. PII: S 0 0 0 9 - 9 1 2 0 ( 0 1 ) 0 0 1 9 6 - 5
108
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
nology. Volunteers gave informed consent, and the ethical committee for our institution approved the study. Sequential samples were collected into each of the following four blood collection tubes: (a) EDTA (ethylenediaminetetra-acetic acid); (Vacuette威, Greiner Labortechnik, Austria); (b) Serum (Vacutainer威, Becton Dickinson Vacutainer System USA, Rutherford, NJ, USA); (c) Lithium heparin (Vacutainer威), and (d) Sodium fluoride/potassium oxalate (Vacutainer威). 2.2. Storage of samples After collection, samples were centrifuged at 4°C immediately (plasma) or after 10 min (serum) and the plasma and serum were separated. Five aliquots of each of the 4 samples from each volunteer were prepared on ice/water and immediately placed in the following storage conditions: a One aliquot at ⫺20°C* (reference temperature) b One aliquot at 4°C for 24 hours (1 day) c One aliquot at 4°C for 120 hours (5 days) d One aliquot at 30°C for 24 hours (1 day) e One aliquot at 30°C for 120 hours (5 days) *The freezer used for sample storage maintained a temperature of ⫺24°C, and was set to alarm at ⫺15°C. This is consequently referred to as ⫺20°C. At the completion of the storage phase, samples were held at ⫺20°C until assayed. Except where indicated, each analyte was measured in duplicate in samples from a minimum of 6 different volunteers for every collection method and storage condition. All samples from individual volunteers were measured in the same assay for each analyte. Samples collected for brain natriuretic peptide (BNP) had values at or near the detection limit of the assay. Therefore further samples were collected from heart failure subjects into EDTA or heparin, and stored as either whole blood or, after separation, at room temperature for 0, 6, 24, 48 and 72 hours. These samples were also analysed for N-terminal BNP (NT-BNP).
2.4. Statistical analysis 2.4.1. Anticoagulants and serum For each analyte, a two-way analysis of variance with subject and anticoagulant as the factors, was performed on log-transformed results from the ⫺20°C reference sample for each collection method. The results from serum, heparin and fluoride collections were each compared to those measured in EDTA plasma, and if there was a significant ( p ⬍ 0.05) difference, and if the mean difference was greater than 10%, this was considered a significant effect of anticoagulant. A difference of 10% was selected because changes of less than this are unlikely to be clinically significant. 2.4.2. Storage For each analyte (except BNP and NT-BNP), the results at 0, 24 and 120 hours for each temperature was expressed as a percent of the value in the reference aliquot (⫺20°C) for each anticoagulant and serum. The mean rate of change (k) was estimated by fitting to the data the equation: Y ⫽ e kt
(1)
where y is the fraction of initial immunoactivity remaining after time t. Fitting was performed using the non-linear least squares curve fitting feature of SigmaStat (Jandel Scientific, Chicago). Loss of immunoreactivity with time gave negative k values and increases positive values. Hormones were considered stable for the number of hours required for the median immunoreactivity, interpolated between observations using equation 1, to reach either 90% or 110% of the original value. For BNP and NT-BNP the statistical method was the same (although temperature and times were different) and analytes were considered to be stable up to the time at which the median loss reached 10%.
2.3. Hormones measured and methodology
3. Results
The following hormones were tested and a summary of the assay details are shown on Table 1: Adrenocorticotrophic hormone (ACTH), aldosterone, alpha-subunit (gonadotrophin), total (␣-subunit, total) and free (␣-subunit, free), arginine vasopressin (anti-diuretic hormone) (AVP), brain natriuretic peptide (BNP), corticotrophin-releasing hormone (CRH), C-peptide, follicle-stimulating hormone (FSH), glucagon, growth hormone (GH), insulin-like growth factor (IGF-1), IGF-binding protein 3 (IGFBP-3), insulin, leptin, luteinizing hormone (LH), N-terminal BNP (NT-BNP), estradiol, pancreatic polypeptide (PPP), parathyroid hormone (PTH), prolactin, urine C-peptide, vasoactive intestinal peptide (VIP).
3.1. Effect of anticoagulant Table 2 shows the effect of each anticoagulant for each of the analytes tested, as measured in the ⫺20°C samples and compared to the EDTA sample. Measured concentrations were significantly altered by the method of collection, with significant differences recorded for 6 of 22 analytes when serum was the matrix , 4 of 20 when fluoride was the anticoagulant, and 3 of 22 when heparin was the anticoagulant. For one analyte, aldosterone, all 3 matrices were significantly different from the result with EDTA. Results are statistically significant ( p ⬍ 0.05) with a difference of ⬍⫺10% or ⬎⫹10%.
Method
2-site radioimmunometric assay (Bioclone Australia PTY, Marrickville, NSW, Australia) ria after extraction of samples with acid/ethanol and a cryoprecipitation step. Antiserum from Dr B Breier, University of Auckland, New Zealand ria kit procedure (Nichols Institute, Ca, USA) Abbott IMX insulin kit (Abbott Laboratories, Diagnostic Division, Ill, USA) ria kit for human Leptin (LINCO Research Inc, Cat # HL81K, St Charles, Mo USA) As described for FSH ria (“in house” rabbit anti-NT-BNP antiserum), NT-BNP (1–21) as standard (Peninsula Laboratories Inc) ria (Estradiol-2, Clinical Assay™, Sorin Biomedica, Italy) ria (rabbit anti-human antiserum Eli Lilly and Company, Indianapolis, Ind USA) 2-site immunoradiometric assay (IRMA) (Nichols Institute, Ca, USA) As described for FSH ria after extraction using in acetone/acetic acid/water (80:19:1), “inhouse” antiserum
Separation
Units
I-AVP purified on HPLC I-BNP purified by HPLC
PEG
125
125 I-human VIP (Peninsula Labs Inc) PEG plus anti purified on HPLC ␥globulin
I-Bovine PPP (Eli Lilly) purified on HPLC
Solid phase anti ␥globulin
PEG
PEG
PEG
PEG
PEG Solid phase anti ␥globulin
NT-BNP(1-13)125I-Tyr14 purified on HPLC (9)
I-labelled hormone (IGF-1 GroPep, Adelaide, Australia), purified on HPLC.
125
125 I-Human glucagon (Sigma Chemical Co), purified on HPLC
125 I-C-Peptide (human [tyr0] CPeptide (Sigma Chemical Co), purified by HPLC
I-CRH purified by HPL
125
125
125
13 3–8 4 4.8
g/L g/ml pmol/L g/L
pmol/L mIU/L pmol/L
pmol/L ng/L
8.1
2.9
7 7.3
1.9 2.2
3
g/L
IU/L pmol/L
12.5
2.5
5.9
6
6 8.6
free 9 total 4
3.4 6.8
Intra assay CV %
ng/L
UI/L
pmol/L
pmol/L
pmol/L pmol/L
pmol/L Aldosterone-3, carboxymethyloxime- dextran-coated charcoal pmol/L [125I]histamine, HPLC purification 125 PEG plus anti g/L I-hCG ␣-subunit (Biogenesis UK), purified by HPLC ␥globulin
Tracer
0.7 3.8
DL mean
10–64 77–568
Range of values
4.1 4.8 14.5
5 14
5.5 5.8
6.8
10.4 4.1
10
5.4
18.6
5.5
5.3
8.5
0.1 ⱕ14.4 2.9
2 17.7
9–57 137–530 9–18.5
15–1140 91–396
1.5–25.9 12–628
ⱕ0.5 1.8
0.1
2.6–5.1 1.6– 13.51 3.8–49.5
119–296
0.4–21.0
69–167
3.5–47.9
471– 1427
1.0–1.9
0.10 6
1.3
0.19
13
0.2
2.0
0.2
free 13 free 0.2 free total 12 total 0.2 0.4–2.7 total 0.9–4.5 12 0.3 1.1–3.1 6.5 8.3 34–153
14.8 14
Inter assay CV %
b
ria, radioimmunoassay “In house” preparation of 125I-labelled hormone used the chloramine T method (8) and 125Iodine (Amersham, UK). c Purification of 125I-labelled hormone was by reverse phase high performance liquid chromatography (HPLC) using a Perkin-Elmer HPLC with a Brownlee RP300 column, unless otherwise stated. d PEG, polyethylene glycol e CV% calculated for concentrations within the range measured in the study (except for insulin) f Detection limit was defined as the smallest value which could be distinguished from zero at the 95% confidence limit. g Where the range of concentrations measured in this study are given for each analyte, these are taken from the EDTA, ⫺20°C sample results. 1 2 of 6 subjects reading lower than the calculated detection limit.
a
PTH Prolactin VIP
E2 PPP
LH NT-BNP
Leptin
IGF-BP3 Insulin
IGF-1
GH
Glucagon
FSH
C-Peptide
SELECT kitset with microparticle enzyme immunoassay (MEIA) technology (Abbott Laboratories, Ill, USA) ria using an “in house” antiserum
ria (4) after acetonitrile extraction of samples, “in house” antiserum ria (modified from (5) after extraction on C18 Sep-Pak cartridges (Waters Corp., Mass. USA), antiserum (RAS 9086 rabbit anti-BNP32 (human), Peninsula Lboratories Inc) ria (6,7) following extraction of samples and standards (human/rat CRH41, Peninsula Laboratories Inc) with methanol, rabbit anti-ovine CRH antiserum (gift from P Lowry, University of Reading, UK) ria (Go2o antiserum, Guildhay), synthetic human C-Peptide (Sigma Chemical Co, St Louis, Mo, USA) as standard
AVP BNP
CRH
ria: 2 “in-house” antisera, 1 binding free ␣-subunit, the other both the free and bound hormone
␣-subunit
ACTH two-site immunoradiometric assay (IRMA) (Nichols Institute, Ca, USA) Aldosterone ria (3)
Hormone
Table 1 Summary of assay details
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112 109
110
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
Table 2 Comparison of the effect of anticoagulant on hormone concentration as compared to EDTA in ⫺20°C samples (frozen immediately)
Table 4 Duration of stability of BNP and NT-BNP in hours when stored as whole blood or plasma at room temperature
Hormone
Heparin
Serum
Fluoride
Plasma
ACTH Aldosterone ␣-subunit (total) ␣-subunit (Free) AVP BNP* CRH C-peptide FSH Glucagon GH IGF-1 IGFBP-3 Insulin Leptin LH NT-BNP* Estradiol PPP PTH Prolactin VIP
⫺74% ⫹17% ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫺20%
⬃ ⫹15% ⬃ ⬃ ⫺42% ⬃ ⫹46% ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫹11% ⫹19% ⬃ ⬃ ⫺22%
⬃ ⫺12% ⫺13% ⬃ ⬃ NT ⬃ ⬃ ⬃ ⬃ ⬃ ⬃ ⫺28% ⬃ ⬃ ⬃ NT ⬃ ⫺11% ⬃ ⬃ ⬃
EDTA
Heparin
EDTA
Heparin
18 7
10 8
20 5
4
*n ⫽ 3 NT ⫽ not tested ⬃ ⫽ no significant difference (p ⬎ 0.05 and/or difference ⬍10%)
3.2. Effect of storage conditions and collection method Table 3 shows the stability of hormones collected into different anticoagulants and stored at 4°C and 30°C ex-
BNP NT-BNP
Whole Blood
pressed in hours, based on time taken to decrease or increase 10% from the reference value, as calculated from k values (see Methods). Measurements were not continued after 120 hours. Estradiol was not tested for storage conditions. Table 4 shows the stability of BNP and NT-BNP stored at room temperature as either whole blood or as plasma collected into EDTA or heparin. Urine C-peptide was stable at 4°C for ⬎120 hours, and at 30°C for 19 hours.
4. Discussion The aim of this study was to determine the effect of different anticoagulants (EDTA, heparin, sodium fluoride and serum) and time and temperature of storage on the results obtained for a range of hormones analysed in our laboratory (22 hormones in total). The anticoagulants selected are those commonly used in the blood collection process. BNP and NT-BNP were stored under conditions replicating collection in a physician’s surgery.
Table 3 Duration of stability of hormones in hours after collection into different anticoagulants and stored at 4°C or 30°C At 4°C
ACTH Aldosterone ␣subunit total ␣subunit free AVP CRH C-Peptide FSH Glucagon GH IGF-1 IGF-BP3 Insulin Leptin LH PPP PTH Prolactin VIP
At 30°C
EDTA
Heparin
Serum
Fluoride
EDTA
Heparin
Serum
Fluoride
18 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120
NT ⬎120 ⬎120 ⬎120 ⬎120 100 ⬎120 ⬎120 74 ⬎120 114 NT ⬎120 ⬎120 ⬎120 ⬎120 101 ⬎120 ⬎120
3 ⬎120 ⬎120 ⬎120 ⬎120 50 ⬎120 ⬎120 34 ⬎120 ⬎120 NT ⬎120 ⬎120 ⬎120 ⬎120 55 ⬎120 ⬎120
NT ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 NT ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120 ⬎120
8 14 ⬎120 60 96 18 ⬎120 ⬎120 69 62 ⬎120 ⬎120 31 ⬎120 ⬎120 ⬎120 36 109 25
NT 13 ⬎120 ⬎120 28 14 ⬎120 ⬎120 42 64 ⬎120 NT 21 69 59 53 33 ⬎120 13
1 16 ⬎120 86 28 11 ⬎120 ⬎120 30 66 ⬎120 NT 32 ⬎120 82 64 3 84 20
NT 15 ⬎120 96 39 15 ⬎120 ⬎120 114 61 88 NT 10 60 88 67 16 87 20
NT ⫽ Not tested. Duration of stability is the time elapsing until the median immunoreactivity, interpolated between observations using equation 1, reached 90 or 110% of the initial value.
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
When interpreting these results, it should be noted that different assay methods and antibodies may give different results [1, 10], there were small numbers of samples in each group, and that samples were handled and stored in a nonsterile manner after collection. The selection of a change of 10% as the limit of stability is arbitrary but has been used previously [1]. In some cases, results for hormones measured following an extraction step showed an apparent increase in immunoreactivity after storage, suggesting assay interference was a factor, or possibly release of bound hormone from a protein factor. When the different collection methods were compared to collection into EDTA, a significant effect was shown for several hormones. Serum affected the results of 6 hormones (aldosterone, AVP, CRH, oestradiol, PPP and VIP), fluoride affected 4 hormones (aldosterone, ␣-subunit (total), IGFBP-3, and PPP), and heparin affected 3 hormones (ACTH, aldosterone, and VIP). AVP measurements in heparin samples appeared to be lower (⫺36% of EDTA) but this was not significant. Some of the results were altered by as much as 40% and could be expected to affect the interpretation of the result if compared to a reference interval based on samples collected into EDTA. This suggests that samples sent to a laboratory should be collected into the appropriate anticoagulant for that laboratory’s methods. In our laboratory, we recommend the collection of samples into EDTA (unless otherwise recommended by a specific kit manufacturer) as our reference intervals are established using samples collected in this way. Using the ⫺20°C sample as the reference value, EDTA samples kept at 4°C for 24 to 120 hours showed no significant effect on the results for most hormones. The exception to this was ACTH, where incubations at 4°C for 24 hours had a significant negative effect. The requirement for ACTH to be stored and transported frozen has been shown previously [11, 12]. The use of dry ice is recommended for transport of ACTH samples, or alternatively, shipment in special containers of “pink ice”.1 In either case, transport time is limited to the time until the dry ice is fully sublimed or the “pink ice” completely thawed. Collection of samples into heparin similarly showed that most samples were stable at 4°C for 24 –120 hours (ACTH not tested), although CRH, glucagon, IGF-1 and PTH were stable for less than 120 hours. Serum samples were also stable for 24 hours at 4°C, except for ACTH, while incubation for 120 hours at this temperature affected ACTH, CRH, glucagon and PTH. When fluoride was used as the anticoagulant, all hormones (ACTH not tested) were stable for 120 hours at 4 °C. From these results it would appear that when compared to 1 Preparation of “pink ice”: 186 g ammonium chloride (industrial grade, Harcoss Chemicals, Christchurch) dissolved in 814 g water and sufficient rhodamine B added to colour the solution pink. When frozen this 18.6% solution melts at ⫺15.8 °C and since it is the eutectic mixture it maintains this temperature until it is nearly all melted.
111
the corresponding ⫺20°C sample, fluoride and EDTA were the most effective anticoagulants for storage and transport of samples at 4°C for up to 120 hours, followed by heparin then serum. However it must be noted that in some cases these different collection methods also affected the ⫺20°C results, and that the effect of storage relates to the ⫺20 °C sample of each anticoagulant, and not to the EDTA value. These results are similar to those of an earlier study from our laboratory using different analytical methods [1] that found that LH, FSH, GH, prolactin and insulin in EDTA plasma were stable at 4°C for at least 8 days. As might be expected, incubation of samples at 30 °C had a more marked effect on the measurements. For the EDTA samples, results for ACTH, aldosterone, and CRH were significantly affected within 24 hours, and by 120 hours, ␣-subunit (free), AVP, glucagon, GH, insulin, PTH, prolactin and VIP were also affected. With collection into heparin, aldosterone, CRH, insulin and VIP were significantly affected within 24 hours (no measurement made for ACTH), and by 120 hours, very few analytes were stable. Similarly for serum, incubation at 30°C for 24 hours affected ACTH, aldosterone, CRH, PPP, PTH and VIP, with most analytes unstable after 5 days. When fluoride was used as the anticoagulant and after 24 hours at 30 °C, aldosterone, CRH, insulin, PPP, PTH and VIP (ACTH not measured) were affected, and by 120 hours, very few hormones were stable. Livesey et al [1] similarly found that in EDTA plasma stored at 37°C, LH was stable for about 3 days, prolactin and GH for about 2 days, and insulin for less than 1 day. Kubasik et al [2] found that in serum, LH, FSH and prolactin were stable at either 4°C or 22°C for up to 14 days, but in our assays, only FSH was stable at 30°C for more than 1 day in serum. Similarly to the present study, Kubasik et al [2] showed that insulin in serum was unstable above 4°C. In a study investigating the effect of storing heparinized whole blood on hormone stability, Diver et al [13] showed no clinically significant changes in FSH, GH, oestradiol or prolactin stored up to 48 hours at room temperature, and up to a week at 4°C. They also found that ACTH was stable for less than 24 hours at room temperature. Our stability data for BNP at room temperature is in agreement with that published by others [14] and is greater than at 37°C [5]. The results contrast with those of Tsuji et al. [15] who found aprotinin was necessary to prevent rapid metabolism of BNP in plasma. The difference in apparent stability between various methods may be due to differences in stability of epitopes of BNP that are recognised by the different antisera. The stability of NT-BNP in this study agrees closely with a preliminary study previously published by us [9] showing measurable loss of NT-BNP immunoreactivity over 6 hours. Results in two published studies indicate greater stability of NT-BNP in plasma [16, 17] than those reported here. These differences are again likely to result from the use of different antisera and indeed agree with our unpublished results (Yandle) using antisera to different NT-BNP epitopes.
112
M.J. Evans et al. / Clinical Biochemistry 34 (2001) 107–112
In summary, this study has shown that for many hormones, collection into EDTA or fluoride, and storage and transport at 0 – 4 °C is practicable. The exception to this guideline is ACTH which requires frozen transport and storage. Acknowledgments The authors thank Sarah L Norton, Endolab, Christchurch Hospital for co-ordinating the collection and storage of the samples and collating the data. We also thank the Endocrine Test Centre Nurses, Christchurch Hospital for collecting the samples, the staff members who volunteered to donate blood, and all the technical staff of Endolab, Christchurch Hospital, for performing the assays. References [1] Livesey JH, Hodgkinson SC, Roud HR, Donald RA. Effect of time, temperature and freezing on the stability of immunoreactive LH, FSH, TSH, growth hormone, prolactin and insulin in plasma. Clin Biochem 1980;13:151–155. [2] Kubasik NP, Ricotta M, Hunter T, Sine HE. Effect of duration and temperature of storage on serum analyte stability: Examination of 14 selected radioimmunoassay procedures. Clin Chem 1982:28:164 – 165. [3] Lun S, Espiner EA, Nicholls MG, Yandle TG. A direct radioimmunoassay for aldosterone in plasma. Clin Chem 1983;29:268 –271. [4] Sadler WA, Lynskey CP, Gilchrist NL, Espiner EA, Nicholls MG. A sensitive radioimmunoassay for measuring plasma antidiuretic hormone in man. NZ Medical J 1983;96:959 –963. [5] Yandle TG, Richards AM, Gilbert A, Fisher S, Holmes S, Espiner EA. Assay of brain natriuretic peptide (BNP) in human plasma— evidence for high molecular weight BNP as a major plasma component in heart failure. Clin Chem 1993;76:832– 838.
[6] Ellis MJ, Livesey JH, Donald RA. Circulating plasma corticotrophin releasing factor-like immunoreactivity. J Endocrinol 1988;117:299 – 307. [7] Ellis MJ, Schmidli RS, Donald RA, Livesey JH, Espiner EA. Plasma corticotrophin-releasing factor and vasopressin responses to hypoglycaemia in normal man. Clin Endocrinol 1990;32:93–100. [8] Landon J, Livanou T, Greenwood FC. The preparation and immunological properties of 131I labelled adrenocorticotropin. Biochem J 1967;105:1075–1083. [9] Hunt PJ, Richards AM, Nicholls MG, Yandle TG, Doughty RN, Espiner EA. Immunoreactive amino-terminal pro-brain natriuretic peptide (NT-proBNP)—a new marker of cardiac impairment. Clin Endocrinol 1997;47:287–296. [10] Donald RA. Radioimmunoassay of corticotropin (ACTH). In: Abrahams G, Ed. Handbook of Radioimmunoassay. Pp.319 –390. Marcel Dekker, New York, 1977. [11] Meakin JW, Tingey WH, Nelson DH. Catabolism of adrenocorticotropic hormone: The stability of adrenocorticotropic hormone in blood, plasma, serum and saline. Endocrinology 1960;66:59 –72. [12] Purves HD, Sirett NE. The stability of endogenous and exogenous corticotrophin in rat plasma as established by a bioassay in intact rats. J Endocrinol 1968;41:491– 497. [13] Diver MJ, Hughes JG, Hutton JL, West CR, Hipkin LJ. The long-term stability in whole blood of 14 commonly-requested hormone analytes. Ann Clin Biochem 1994;31:561–565. [14] Buckley MG, Marcus NJ, Yacoub MH, Singer DRJ. Prolonged stability of brain natriuretic peptide—importance for non-invasive assessment of cardiac function in clinical practice. Clinical Science 1998;95:235–239. [15] Tsuji T, Imagawa K, Masud H et al. Stabilization of human brain natriuretic peptide in blood samples. Clinical Chemistry 1994;40: 672– 673. [16] Karl J, Borgya A, Gallusser A et al. Development of a novel, NTerminal-proBNP (NT-proBNP) assay with a low detection limit. Scandinavian Journal of Clinical & Laboratory Investigation 1999;59 (Suppl 230);177–181. [17] Downie PF, Talwar S, Squire IB, Davies JE, Barnewtt DB, Ng LL. Assessment of the stability of N-terminal pro-brain natriuretic peptide in vitro: implications for assessment of left ventricular dysfunction. Clinical Science 1999;97:255–258.