Download Report

NDT Advance Access published February 5, 2009
Nephrol Dial Transplant (2009) 1 of 4
doi: 10.1093/ndt/gfp021
Original Article
False low parathyroid hormone values secondary to sample
contamination with the tissue plasminogen activator
Brigitte Schiller1,2 , Amy Wong1 , Marty Blair1 and John Moran2
1
Satellite Laboratory Services, Redwood City and 2 Satellite Healthcare, Mountain View, CA, USA
Keywords: haemodialysis; haemodialysis access;
parathyroid hormone (PTH); tissue plasminogen activator
(tPA)
Correspondence and offprint requests to: Brigitte Schiller, Clinical Director, Satellite Laboratory Services, 1400 Industrial Way, Redwood City,
CA 94063, USA. Tel: +1-650-404-3640; Fax: +1-650-625-6240; E-mail:
[email protected]
Introduction
An 82-year-old Caucasian female with end-stage renal
disease secondary to diabetes mellitus and hypertension
had undergone maintenance haemodialysis for 6 years.
Quarterly parathyroid hormone (PTH) results had ranged
between 230 and 480 pg/mL. When PTH increased to
834 pg/mL, 30 mg oral cinacalcet was added to her thrice
weekly IV vitamin D and oral phosphate binder regimen.
Three months later, PTH had decreased to 142 pg/mL, a surprisingly dramatic decrease of >80%, higher than the usual
decrease of 30–40% following the administration of a calcimimetic [1]. Over the next 7 months, PTH values ranged
between 48 and 471 pg/mL. Inconsistency of calcimimetic
intake including non-adherence to the recommended 12-h
dose interval prior to blood draw was suspected [2]. However, a PTH value of 17 pg/mL made this assumption questionable. Laboratory quality control on precision and accuracy were validated.
Fortuitously, at this time a stability study of PTH measurements was being performed in our laboratory in order
to assess any effect of the delay in specimen receipt with
centralized laboratory services common in dialysis practice. Left-over plasma samples from the routine monthly
blood draw in a nearby dialysis unit were measured immediately after draw and 24, 48 and 72 h later. The data
revealed very satisfactory consistency of results over time,
with only one outlier out of 85 samples. This sample measured at baseline 180 pg/mL, but declined to 78 pg/mL after
24 h, and PTH was no longer detectable after 48 h. In the
absence of any explanation the sample was further investigated and identified as coming from the patient described
above.
On investigating the medical history of this patient, it
became apparent that she dialyzed with a tunnelled central venous catheter and intermittently received alteplase
R
R
R
(Cathflo
Activase
, Genentech, CA, USA; Actilyse
,
Boehringer Ingelheim Pharma, Germany) as a catheter lock
to maintain patency. The blood samples had been drawn
from the catheter. The patient’s medication history confirmed that the catheter had been locked with alteplase at
the end of the previous dialysis treatment prior to the day
of the blood draw in question.
C The Author [2009]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: [email protected]
Downloaded from http://ndt.oxfordjournals.org/ by guest on October 6, 2014
Abstract
Background. Fluctuating parathyroid hormone values
(PTH) are common in patients undergoing haemodialysis.
Widely varying PTH results in an 82-year-old haemodialysis (HD) patient could not be explained. When PTH in
the same blood sample was no longer detectable 24 h after
blood draw, it was hypothesized that contamination with the
catheter lock solution containing tissue plasminogen activator (tPA, alteplase) caused degradation of PTH in vitro.
Methods. Leftover samples from 21 patients on maintenance HD as well as control samples from healthy volunteers (n = 3) were incubated at 4◦ C with small amounts of
tPA (25 and 50 µL). In addition, pooled samples from HD
patients with various PTH levels were incubated with 6.5,
12.5 and 25 µL of tPA and analysed with two different PTH
assays with incubation times up to 48 h.
Results. A rapid decline of PTH values to 2.5–33.5% of
the original baseline was observed after 24 h with a further
decrease to <1–15% after 48 h. The two different assays
gave very similar results when the samples were incubated
with tPA.
Conclusion. Minimal contamination of a blood sample with
tPA results in degradation of PTH in a time-dependent manner. The tPA is therefore unique as a contaminant since its
enzymatic activity means that even tiny amounts of contamination will lead to major errors in PTH results by digestion
of the protein. This phenomenon was independent of the
assay used. Strict attention to the technique when drawing
a blood sample from a catheter is mandatory to prevent
contamination and avoid spurious test results.
2
B. Schiller et al.
We therefore hypothesized that contamination of the
blood sample with alteplase was the cause of the decreasing
PTH values noted in the specimen in the stability study.
Methods
Results
Twenty-one samples of left-over plasma from patients with
varying PTH baseline values ranging from <100 pg/mL
to levels >2000 pg/mL were incubated at 4◦ C with 25 or
50 µL alteplase. A dramatic decline to 2.5–33.5% of the
baseline PTH levels was noted after 24 h (Figure 1) with
a further decline to <1–15% after 48 h (data not shown).
Fig. 1. The dramatic decline of PTH values after incubation with 25 µL
alteplase is shown in this graph (y-axis, PTH in pg/mL, x-axis: baseline
and 24-h result). Independent of the baseline level, PTH results decreased
by 66.5–98.5% after 24-h incubation.
The degree of decline was not significantly enhanced with
the higher dose of 50 µL alteplase at the times measured
(data not shown). A similar effect was noted in the three
healthy control subjects, where initial PTH baseline values
of 17, 38 and 39 pg/mL all decreased to <2.5 pg/mL after
24 h incubation, equivalent to a 93–98% reduction from the
baseline value.
The second set of experiments comparing the PTH results with the two different PTH assays and incubation with
various amounts of alteplase revealed that indeed the PTH
results declined with both assays, resulting in a dramatic
decline of the PTH results over time independent of the assay used. When comparing the control samples without any
alteplase, it was evident that both assays performed very
well, with consistent values over the 48 h evaluation time,
the variation being well below the 12.5% allowable error
for accuracy limits.
To assess the influence of alteplase incubation we defined any change >12.5% being significant to differentiate
from the inherent test variation. With this measure in mind,
decreased PTH values occurred as early as 1 h of incubation in some samples spiked with the higher amounts of
alteplase (12.5 and 25 µL). After 3 h, all samples incubated
with 12.5 or 25 µL showed a significant decline of their
PTH measures by 17–53% of their baseline value in both
assays. After 6 h, all the samples showed PTH values decreased by 14–72%, independent of the alteplase amount.
After 12 h, the PTH values had decreased by 35–88% compared to their baseline measure; after 24 h the change was
between 70 and 96%. PTH was barely detectable after 48 h
with both assays measuring only 1–9% of the baseline PTH
value prior to alteplase incubation. The exact percentages
of decrease in PTH values are shown in Table 1 for both
the Bayer and Roche assays. While the trend of decline was
Downloaded from http://ndt.oxfordjournals.org/ by guest on October 6, 2014
The leftover plasma samples from routine monthly blood
draws of 21 haemodialysis patients with known PTH levels,
representing PTH values commonly found in patients on
maintenance HD, ranging from 68 to 2340 pg/mL were
selected. Each sample (750 µL) was incubated with 25 or
50 µL of alteplase at 4◦ C for a total of 48 h to mimic
transportation time in a cooled container while in route
to the central laboratory. In addition, three samples from
healthy volunteers were treated similarly.
After 24 and 48 h, each sample was re-tested with the
R
Bayer intact PTH assay using the ADVIA Centaur
system (Bayer, Norwood, MA, USA). The ADVIA Centaur Intact PTH assay is a two-site sandwich immunoassay using
direct chemiluminometric technology. A polyclonal goat
anti-human PTH labelled with acridinium ester binds in the
region of the N-terminal 1–34 AA. The second antibody is
a biotinylated polyclonal goat anti-human PTH binding in
the 39–84 region. Streptavidin in the solid phase is covalently coupled to paramagnetic latex particles as the capture
mechanism.
To confirm that such an effect was not specific to the
Bayer assay but rather a more general phenomenon, the
study was repeated measuring PTH values in parallel with
both the Bayer assay as outlined above and the Roche PTH
assay (Roche Diagnostics, Indianapolis, IN, USA).
The Roche PTH assay is an electrochemiluminescence
immunoassay also based on a sandwich test principle utilizing, however, different antibodies and detection systems. A
biotinylated monoclonal PTH-specific antibody reacts with
the N-terminal fragment (1–37) and another monoclonal
antibody labelled with a ruthenium complex reacts with
the C-terminal fragment (38–84). The respective epitopes
recognized by the antibodies correspond to the amino acid
regions 26–32 and 55–64. Streptavidin-coated microparticles bind the complex to the solid phase via interaction of
biotin and streptavidin.
Pooled samples from haemodialysis patients with PTH
levels in the ranges of 100–300 pg/dL, 300–600 pg/dL,
600–1200 pg/dL and >1200 pg/dL were incubated with
various amounts of tPA including 0 (control) 6.5, 12.5 and
25 µL and measured with both PTH assays at different time
points (1, 3, 6, 12 and 24 h) after addition of alteplase to
further study the kinetics of this process and to test for dose
dependence.
tPA falsely decreases PTH levels
3
Table 1. Percentage change compared with initial baseline PTH value after various incubation times with 0 (control), 6.5, 12.5 and 25 µL alteplase in
all four pooled samples
PTH sample
pg/mL
100–300
300–600
600–1000
>1200
Amount of
alteplase (µL)
% Diff
(BL-3 h)
% Diff
(BL-6 h)
% Diff
(BL-12 h)
% Diff
(BL-24 h)
% Diff
(BL-24 h)
Roche
Bayer
Roche
Bayer
Roche
Bayer
Roche
Bayer
Roche
Bayer
Roche
Bayer
1.0
−1.5
−5.5
−12.9
−0.7
−4.2
−7.4
−15.4
4.3
−3.0
−3.2
−10.6
−1.0
−2.9
−2.8
−11.5
−5.8
−0.6
−16.5
−18.7
−2.9
−11.9
−13.3
−23.8
−4.4
5.5
−9.6
−24.2
1.4
−5.0
−6.9
−25.9
−3.5
−8.8
−20.9
−35.4
−2.1
−9.5
−24.4
−43.5
4.1
−5.5
−17.1
−32.0
−1.2
−5.9
−17.8
−31.3
−8.5
−10.1
−28.1
−36.9
−0.2
−19.3
−40.0
−53.3
−7.6
−1.9
−37.8
−47.8
−2.6
−10.5
−31.8
−45.9
−2.5
−13.8
−35.3
−46.5
−3.1
−22.8
−49.3
−57.3
2.0
−21.2
−44.1
−55.6
−1.4
−20.9
−43.7
−53.8
−6.3
−29.4
−49.3
−58.5
2.6
−47.8
−64.9
−72.4
−2.7
−32.0
−62.9
−68.6
−4.4
−36.6
−58.0
−67.8
−0.2
−35.2
−55.1
−63.6
−2.7
−51.6
−69.3
−74.8
−3.7
−50.3
−67.5
−73.3
−2.8
−47.7
−65.1
−69.6
−3.9
−57.6
−71.2
−73.3
−0.4
−74.6
−83.8
−87.8
−5.5
−67.6
−81.6
−84.9
2.1
−69.6
−78.9
−83.0
−4.3
−69.4
−78.5
−81.1
−6.7
−81.9
−87.5
−89.3
−2.1
−81.5
−87.0
−88.8
−9.9
−78.4
−84.0
−86.2
−3.9
−82.0
−88.2
−90.0
−1.0
−93.7
−95.0
−95.7
−5.3
−90.3
−94.0
−94.4
−3.4
−89.2
−91.9
−93.5
−3.2
−90.9
−92.4
−92.9
−4.2
−95.3
−96.2
−96.2
−0.5
−95.7
−96.3
−96.6
−6.2
−93.3
−94.3
−94.7
−0.9
−94.5
−97.0
−95.9
−2.0
−98.0
−98.6
−98.9
−2.6
−98.3
−98.6
−98.5
2.9
−97.7
−97.8
−98.0
Fig. 2. Representative graphs for PTH levels measured with both the Roche (top row) and Bayer (lower row) assays are shown for the pooled samples
with PTH levels of 300–600 (left column) and 600–1200 pg/mL (right column) demonstrating similar curves of PTH decline over time.
very similar in both assays, the decrease of PTH analysed
with the Bayer assay appeared to be slightly accelerated,
resulting in larger declines with time. The similar course
of the gradual PTH decline over time in all pooled samples
demonstrates that the effect is independent of the amount
of substrate present but rather, as one would expect with
an enzymatic reaction, depends on the amount of the alteplase present in both assays in a time-dependent manner
(Figure 2).
Discussion
Thirty-two to seventy-five percent of errors in laboratory
medicine are accounted for by pre-analytical specimen errors. The majority of such errors occur due to improper
handling of the specimen, including failure to centrifuge a
specimen, haemolyzed or clotted samples, or quantity not
sufficient [3], very similar to nationally reported data in
annual CAP surveys [4]. In patients suffering from ESRD
Downloaded from http://ndt.oxfordjournals.org/ by guest on October 6, 2014
0
6.5
12.5
25
0
6.5
12.5
25
0
6.5
12.5
25
0
6.5
12.5
25
% Diff
(BL-1 h)
4
in the area of AA 19–20. It is possible that the labelled
polyclonal antibody in the Bayer assay which binds in the
area of AA 1–34—a position which includes the valine–
arginine bond—will no longer be captured, and therefore,
the recovery of the PTH fragment may be hindered resulting in tremendously decreased results. In the case of the
Roche assay, neither of the monoclonal antibodies binds in
the critical region of the AA19-20 of the PTH molecule.
However, the assay is nonetheless affected implying an as
yet undetermined pathway of alteplase interference with
this assay. As PTH results often vary secondary to clinicalphysiological causes, one needs to be aware of externally
induced variability due to a contaminated sample. Only a
strict protocol when sampling blood from a catheter to avoid
any contamination and interference with assays will prevent
such errors.
Acknowledgement. Roche Diagnostics provided generously the test kits
for the PTH measurements to allow comparison with an alternative PTH
assay.
Conflict of interest statement. None declared.
References
1. Block GA, Martin KJ, de Francisco AL et al. Cinacalcet for secondary
hyperparathyroidism in patients receiving hemodialysis. N Engl J Med
2004; 350: 1516–1525
2. Goodman WG, Hladik GA, Turner SA et al. The calcimimetic agent
AMG 073 lowers plasma parathyroid hormone levels in hemodialysis
patients with secondary hyperparathyroidism. J Am Soc Nephrol 2002;
13: 1017–1024
3. Bonini P, Plabani M, Ceriotti F et al. Errors in laboratory medicine.
Clin Chem 2002; 48: 691–698
4. College of the American Pathologists. Q-track Survey Quarter 1,
2008—SLS comparison by CAP to National US Laboratories
5. Reddan D, Klassen P, Frankenfield DL et al. National profile of
practice patterns for hemodialysis vascular access in the United States.
J Am Soc Nephrol 2002; 13: 2117–2124
6. Hopson S, Frankenfield D, Rocco M et al. Variability in reasons for
hemodialysis catheter use by race, sex, and geography: findings from
the ESRD Clinical Performance Measures Project. Am J Kidney Dis
2008; 52: 753–760
7. Jaffer Y, Selby NM, Taal MW et al. A meta-analysis of hemodialysis catheter locking solutions in the prevention of catheter-related
infection. Am J Kidney Dis 2008; 51: 233–241
8. Mandolfo S, Borlandelli S, Elli A. Catheter lock solutions: it’s time
for a change. J Vasc Access 2006; 7: 99–102
9. Ball CL, Tobler K, Ross BC et al. Spurious hyperphosphatemia due
to sample contamination with heparinized saline from an indwelling
catheter. Clin Chem Lab Med 2004; 42: 107–108
10. Schiller B, Virk B, Blair M et al. Spurious hyperphosphatemia in
patients on hemodialysis. Am J Kidney Dis 2008; 52: 617–620
11. Lamb EJ, Abbas NA. Spurious hypernatraemia and Citra-Lock. Ann
Clin Biochem 2007; 44: 579
12. Souberbielle JC, Boutten A, Carlier MC et al. Inter-method variability
in PTH measurement: implication for the care of CKD patients. Kidney
Int 2006; 70: 345–350
13. Madison EL, Coombs GS, Corey DR. Substrate specificity of tissue
type plasminogen activator. Characterization of the fibrin independent
specificity of t-PA for plasminogen. J Biol Chem 1995; 270: 7558–
7562
Received for publication: 17.8.08
Accepted in revised form: 9.1.09
Downloaded from http://ndt.oxfordjournals.org/ by guest on October 6, 2014
undergoing maintenance haemodialysis or peritoneal
dialysis, the reasons for rejecting specimens are very comparable (internal data on file). In addition to these causes,
blood samples drawn from a central catheter can have additional issues.
A significant number of patients undergoing HD in the
USA are treated at some time with a tunnelled catheter as
access [5,6]. In order to prevent the two most significant
complications in these patients, infections and clotting, a
variety of catheter lock solutions are used. These include
solutions containing anticoagulants such as heparin and
citrate or antibiotics such as gentamicin or vancomycin
[7,8].
There are various mechanisms by which catheter lock
solutions can affect laboratory results if not completely
removed from the catheter. Dilution of the sample is an
obvious cause and can often be verified by assessing the
sample and the simultaneous decrease in other analytes. In
other instances, the result may be falsified by the solution
binding the analyte, thereby preventing its recovery, such as
citrate binding calcium. Spurious elevation of phosphate results through contamination with heparin and alteplase has
been reported through the addition of phosphate contained
in the solution or by interference with the assay [9,10]. Spurious hypernatraemia has also been described secondary to
trisodium citrate contamination by a catheter lock solution
[11].
Our study describes a previously unreported mechanism,
whereby PTH results can be falsely decreased due to the
contamination of the pre-analytical specimen with small
amounts of alteplase. The incubation of plasma samples occurring during transit of a sample to the laboratory even at
4◦ C results in a time-dependent digestion of PTH into fragments that are no longer detectable by the PTH assay. This
effect was noted both in patients undergoing haemodialysis
and in healthy control subjects indicating that it is independent of the uraemic state.
This effect was reconfirmed with two different PTH assays available for automated platforms. Variations in the
PTH result dependent on the assay utilized are well documented, and the difference between the two assays seen in
our studies correlates well with published data [12]. The results, however, show that both assays are similarly affected
by the addition of alteplase when analysing both the kinetics of the effect and the dose dependence. Both assays studied here are based on a sandwich immunoassay, one using
chemiluminescence and the second one electrochemiluminescence. The effect is seen with both assays even though
they use completely different antibodies with affinity to different epitopes and different detection systems. Although
the only substrate of alteplase under physiological conditions is plasminogen [13], these experiments demonstrate
that contamination with alteplase can lead to digestion of
PTH with prolonged ‘incubation’ in vitro and consequent
extreme variability in results. Alteplase activates the plasminogen pro-enzyme by splitting a single valine–arginine
peptide bond. The PTH protein contains one such bond
B. Schiller et al.