Download www.clinchem.org - Clinical Chemistry

Clinical Chemistry 55:5
923–929 (2009)
Proteomics and Protein Markers
Variability of N-Terminal Probrain Natriuretic Peptide in
Stable Chronic Heart Failure and Its Relation to Changes
in Clinical Variables
Lutz Frankenstein,1* Andrew Remppis,1 Joerdis Frankenstein,2 Georg Hess,3 Dietmar Zdunek,4
Karen Slottje,1 Hugo A. Katus,1 and Christian Zugck1
BACKGROUND:
We investigated the variability of
N-terminal probrain natriuretic peptide (NT-proBNP)
and its relation to known confounding variables in patients with stable chronic heart failure who were on a stable optimized medication regimen.
METHODS:
At 4 sampling intervals (14-day, 1-month,
2-month, and 3-month) the results for NT-proBNP
measurements and several clinical variables were measured in samples from 41 patients with chronic systolic
dysfunction who met 21 prespecified criteria for
stability.
RESULTS:
Mean within-person NT-proBNP variabilities expressed as percentage CV were 17.6%, 18.9%,
15.5%, and 16.2% at 14-day, 1-month, 2-month, and
3-month follow-up, respectively, and the corresponding reference change values were 34.6%, 52.5%, 43.1%,
and 45.0%, respectively. Within-person variability of
NT-proBNP was not found to be associated with renal
function, weight, or waist circumference. Likewise, age,
sex, baseline NT-proBNP, New York Heart Association
functional class, and ejection fraction did not influence
variability of NT-proBNP. The index of individuality
ranged from 0.07– 0.15 depending on the time interval
between test results.
CONCLUSIONS: Although other reported studies have revealed variations in the range of 80%, in this prespecified stable heart-failure population variation of NTproBNP at 14-day, 1-month, 2-month, and 3-month
follow-up was lower and was not related to renal function or weight. In view of the low index of individuality
we observed, within-person variation is quite low compared to between-person variation. Consideration of
these facts is important for the interpretation of clinical
1
Department of Cardiology, Angiology, and Pulmonology; University of Heidelberg; Germany; 2 Lehrstuhl für Pharmazeutische Biologie; Friedrich-AlexanderUniversität Erlangen-Nürnberg; Germany; 3 Roche Diagnostics, Mannheim, Germany; 4 Roche Diagnostics, Rotkreuz, Switzerland.
* Address correspondence to this author at: Department of Cardiology, Angiology, Pulmonology, University of Heidelberg, Im Neuenheimer Feld 410,
D-69120 Heidelberg, Germany. Fax 0049-6221-56-6547; e-mail Lutz.
trials and the use of NT-proBNP in monitoring patients with heart failure.
© 2009 American Association for Clinical Chemistry
Serial measurement of natriuretic peptides in heart
failure patients has been suggested to have utility both
for the guidance of management strategies (1–5 ) and
for improved prognostication (6, 7 ). This application,
however, relies on knowledge about the corresponding
biological variation at the respective follow-up interval
chosen. To date, such intermediate-term variations are
largely unknown because studies assessing biological
variability in heart failure patients have either examined short-term (week-to-week) variations (8–11 ) or
long-term (year-to-year) variations (12, 13 ). Because
patients in the outpatient setting are rarely seen on a
weekly or a yearly basis, this gap in knowledge of
intermediate-term variability must be addressed.
Results of prior studies suggest that short-term
variability of natriuretic peptides is high (8–11 ) and
reference change values (RCVs)5—the percentage of
change of hormonal level considered to be of possible
diagnostic or prognostic importance— have been
found to exceed 100% of the initial value (11 ). These
high values are controversial (12 ) because they appear
to be related to the skewed distribution of measured
hormonal values and improve after normalizing transformation of the data (13 ). Furthermore, the biological
variation of brain natriuretic peptide (BNP) appears to
be greater than that of N-terminal pro-BNP (NTproBNP) (14 ).
Of those studies in which very large within-person
variations were observed, only 2 (9, 11 ) have included
patients with heart failure. Investigation of natriuretic
peptide variations in heart failure patients is important
[email protected].
Received June 5, 2008; accepted February 11, 2009.
Previously published online at DOI: 10.1373/clinchem.2008.112052
5
Nonstandard abbreviations: RCV, reference change value; BNP, brain natriuretic
peptide; NT-proBNP, N-terminal pro-BNP; NYHA, New York Heart Association;
GFR, glomerular filtration rate; MDRD, Modification of Diet in Renal Disease;
IOI, index of individuality.
923
because patients with heart failure have much higher
baseline hormonal values and when these patients are
clinically stable, their within-person variability expressed in terms of CV would be expected to be substantially lower than that of healthy individuals. In patients undergoing treatment for heart failure, stability
of medication and adherence to guidelines are important aspects to consider because, for example, administration of ␤-blockers could significantly influence
BNP levels (15 ). Furthermore, stability of clinical
state—the prerequisite for determination of biovariability—might be difficult to ascertain in heart failure
patients, especially with longer intervals between tests.
Because adherence to guidelines has been incomplete in previous studies and stability of chronic heart
failure has not always been confirmed by direct assessment, we prospectively investigated in an outpatient
setting the intermediate-term variability of NTproBNP in patients with stable chronic heart failure
who were undergoing treatment with individually optimized stable medication regimens.
Materials and Methods
This study constitutes the NT-proBNP arm of a multimarker project on neurohumoral pathways in chronic
heart failure. The local ethics committee approved the
project, which conforms to the principles of the declaration of Helsinki. Patients were recruited prospectively and continuously from the heart failure outpatient clinic of the University Hospital Heidelberg,
Germany. Between September 2006 and June 2007, all
patients being seen in follow-up visits to the clinic (n ⫽
766 patients) were screened for eligibility. A total of 90
patients fulfilled the criteria for inclusion in the study
described below. Of these, 43 patients consented to the
study protocol.
Two patients withdrew their consent after the first
visit. Therefore, the final study population consisted of
41 patients. At 3-month-follow-up, 3 patients showed
significant unexplained weight loss compared to
1-month follow-up. In accordance with the study protocol, they remained in the data set for 1-month
follow-up because they were stable in all other parameters up to that time, but were excluded from further
analysis at 2-month and 3-month follow-up, although they were stable according to all other criteria, including self-perceived clinical status. Therefore,
at 3-month-follow-up the final study populations consisted of 38 patients.
The diagnosis of heart failure was established according to published guidelines (16 ), and the etiology
was confirmed by results of cardiac catheterization,
which was performed in all patients during first contact
with our clinic. Patients were eligible if they met all of
924 Clinical Chemistry 55:5 (2009)
the following criteria: (a) chronic heart failure known
for at least 1 year followed at our clinic; (b) stable clinical condition with no hospitalization attributable to
worsening heart failure within the previous 6 months;
(c) New York Heart Association (NYHA) functional
class I–III; (d) subjectively stable clinical condition
since the last visit to the outpatient clinic, as judged
both by the patient and the physician; (e) ischemic or
idiopathic dilated cardiomyopathy as the etiology of
heart failure; (f) in case of ischemic origin, no revascularization performed within the previous 6 months and
no revascularization planned during the next 6
months; (g) age ⬎18 years; (h) complete adherence to
guidelines of medical treatment regarding class of
drugs; (i) individually optimized doses of guidelinerecommended drugs for at least 4 weeks before study
inclusion; (j) no recent involuntary change of weight
exceeding 2 kg within 4 weeks before inclusion; and (k)
serum sodium levels within reference intervals, indicating euvolemic state.
Patients were excluded if they met 1 or more of the
following criteria: (a) pregnancy, (b) significant renal
dysfunction as identified by a creatinine concentration
⬎176.8 ␮mol/L (2 mg/dL), (c) unstable renal function
as identified by a change in creatinine ⬎20% since the
last visit, (d) thyroid dysfunction, (e) history of pulmonary disease, (f) valvular heart disease, (g) inability to
cooperate or to comprehend the study protocol, (h)
active listing for cardiac transplantation, (i) primary
pulmonary hypertension, (j) in cases of ischemic origin: angina, angina-equivalent, or ischemia-induced
changes in the electrocardiogram during bicycle exercise testing of at least 7 MET (metabolic equivalent
test).
Blood samples for NT-proBNP analysis were
taken from an indwelling venous catheter after patients
had rested for a period of 30 min. The time of sample
collection was between 2 and 4 PM, because NTproBNP has been demonstrated to be stable during the
afternoon (9 ). After the initial visit (T0), patients were
seen again after 2 weeks and 4 weeks (T1 and T2) and
again after 12 weeks (T3). For T1 and T2, a delay of 2
days was allowed, and for T3, a delay of 4 days. EDTA
Vacutainer tubes were used, and samples were centrifuged at 4 °C immediately after collection to separate
the plasma, which was frozen at ⫺20 °C.
NT-proBNP was measured by use of a fully automated Elecsys® Roche Diagnostics 2010 analyzer. Assay precision, analytical sensitivity, interferences, and
stability for this method have been previously described (17, 18 ). To minimize the analytical variance,
all samples were analyzed together in 1 run. We determined within-run imprecision by using samples from
our study patients and synthetic controls. The CVs for
analysis of 21 aliquots of each sample—again, all sam-
Biovariability of NT-proBNP in Stable Heart Failure
ples within 1 assay run—were 1.6% (sample mean, 137
ng/L), 1.8% (sample mean, 150 ng/L), 1.0% (sample
mean, 581 ng/L), 1.2% (sample mean, 1222 ng/L), and
1.5% (sample mean, 5239 ng/L), giving a total variation
of 1.4%. NT-proBNP results are presented in nanograms per liter. To convert to picomoles per liter, multiply results by 0.118.
NT-proBNP was considered as a continuous variable for all calculations. Renal function was assessed
both by serum creatinine concentrations and glomerular filtration rate (GFR) as estimated by the simplified
Modification of Diet in Renal Disease (MDRD) formula (19, 20 ).
The total CV (CVt) and the analytical CV (CVa)
provided the basis for calculating the individual biological CV (CVi) where CVi ⫽ (CVt2 – CVa2)[1/2]. The CV
was calculated as CV ⫽ 100 ⫻ (SD/mean). RCVs were
calculated from median CVt values, according to the
formula: RCV ⫽ Z ⫻ 2[1/2] (CVa2/na ⫹ CVi2/ns)[1/2],
where Z ⫽ 1.96 (i.e., the Z-score for 95% confidence
with a 2-tailed P ⬍ 0.05); na is the number of replicate
assays; and ns is the number of patient samples to estimate each of the 2 homeostatic set points. The index of
individuality (IOI) was calculated as: IOI ⫽ (CVa2 ⫹
CVi2)[1/2]/CVG (⫽ intraindividual variance⫹analytical
variance/interindividual variance) (21, 22 ). Here, a
low IOI (⬍0.48) is considered to reflect strong individuality, which in turn indicates that an individual patient should be assessed with respect to his or her individual hormonal level. In contrast, a high IOI (⬎1.4)
indicates that this patient should be assessed with
respect to population-derived reference intervals
(21 ). Continuous data were tested by use of the
2-sample Wilcoxon test, Kruskal–Wallis-test, and
1-way ANOVA, as appropriate. Univariable and multivariable linear regressions were performed by using
uncorrected or log-transformed values for the dependent variable, as appropriate, and the Spearman
method was used to assess correlations between variables. The relation between relative change and hormonal level was assessed according to the method proposed by Bland and Altman (23, 24 ). An arbitrary level
of 5% statistical significance (P ⬍ 0.05; 2-tailed) was
assumed throughout. Calculations were performed
with SAS version 9.1.
Results
Our patients represent the average population at a tertiary reference center. Dilated cardiomyopathy was the
predominant clinical presentation in this population,
which included 80% men with a mean age of 61 years.
The majority of patients fell into NYHA functional
classes I and II. For complete characteristics see Table
1. At all 4 sampling intervals, patients were stable with
Table 1. Patient characteristics at study inclusion.
Parameter
Value
Patients, n
41
Men, n (%)
33 (80)
Age, y
61 ⫾ 10
Ischemic heart disease, n (%)
17 (41)
NYHA I, n (%)
9 (22)
NYHA II, n (%)
29 (71)
NYHA III, n (%)
3 (7)
LVEF,a %
33 ⫾ 10
Body mass index, kg/m2
28.2 ⫾ 4.6
102.5 ⫾ 11.1
Waist, cm
Systolic blood pressure, mmHg
111 ⫾ 17
Diastolic blood pressure, mmHg
72 ⫾ 10
Heart rate, beats/min
NT-proBNP, ng/L
68 ⫾ 11
597 (257–1306)
139 ⫾ 3
Sodium, mmol/L
Creatinine, ␮mol/L
108.0 ⫾ 25.7
MDRD-estimated GFR, mL/min
59 ⫾ 17
Medication
a
ACEI or ARB, n (%)
41 (100)
␤-Blocker, n (%)
40 (98)
Aldosterone antagonist, n (%)
28 (68)
Diuretics, n (%)
25 (61)
LVEF, left ventricular ejection fraction; ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker.
regard to weight (P ⫽ 0.99), waist circumference (P ⫽
0.94), renal function (P ⫽ 0.94 for creatinine; P ⫽ 0.90
for MDRD-estimated GFR), serum sodium concentration (P ⫽ 0.71), blood pressure (P ⫽ 0.82 for systolic;
P ⫽ 0.17 for diastolic), heart rate (P ⫽ 0.47), and NTproBNP (P ⫽ 0.99). No changes in medication or
NYHA functional class occurred. For a complete data
profile according to sample interval, see Table 2.
In this population of stable heart failure patients,
NT-proBNP showed biological variability of 11.3% to
18.9% depending on the time interval between tests.
Corresponding RCVs were between 31.4% and 52.5%.
The indices of individuality ranged from 0.07– 0.12.
More detailed results are available in Table 3. CVs for
the 4 test intervals showed a variable pattern of intercorrelation. Whereas 2-month coefficients did not correlate with other coefficients (r ⫽ 0.17, P ⫽ 0.28 vs
14-day; r ⫽ 0.21, P ⫽ 0.20 vs 1-month; r ⫽ 0.22; P ⫽
0.17 vs 3-month), all other coefficients were found to
be intercorrelated: 14-day coefficients correlated with
1-month (r ⫽ 0.66, P ⬍ 0.001) and 3-month (r ⫽ 0.45,
P ⬍ 0.01) coefficients and 1-month coefficients correClinical Chemistry 55:5 (2009) 925
Table 2. Patient characteristics during serial sampling.a
Parameter
Initial visit
14-days
28-days
90-days
NYHA I, n (%)
9 (22)
9 (22)
9 (22)
9 (22)
NYHA II, n (%)
29 (71)
29 (71)
29 (71)
29 (71)
NYHA III, n (%)
Body mass index, kg/m2
3 (7)
3 (7)
3 (7)
3 (7)
28.2 ⫾ 4.6
28.2 ⫾ 4.7
28.2 ⫾ 4.7
28.3 ⫾ 4.9
102.5 ⫾ 11.1
102.4 ⫾ 11.3
101.1 ⫾ 11.3
100.7 ⫾ 11.3
Systolic blood pressure, mmHg
111 ⫾ 17
112 ⫾ 17
110 ⫾ 17
111 ⫾ 18
Diastolic blood pressure, mmHg
72 ⫾ 10
69 ⫾ 8
67 ⫾ 10
69 ⫾ 11
Heart rate, beats/min
68 ⫾ 11
67 ⫾ 9
67 ⫾ 11
68 ⫾ 9
NT-proBNP, ng/L
597 (257–1306)
582 (272–1538)
590 (286–1183)
520 (215–1494)
Sodium, mmol/L
139 ⫾ 3
139 ⫾ 3
140 ⫾ 3
140 ⫾ 3
108.0 ⫾ 25.7
109.0 ⫾ 24.8
106.0 ⫾ 23.3
109.0 ⫾ 28.0
59 ⫾ 17
58 ⫾ 15
60 ⫾ 15
59 ⫾ 26
14.1 ⫾ 1.2
13.9 ⫾ 1.3
14.0 ⫾ 1.4
14.2 ⫾ 1.5
ACEIb or ARB, n (%)
41 (100)
41 (100)
41 (100)
41 (100)
␤-Blocker, n (%)
40 (98)
40 (98)
40 (98)
40 (98)
Aldosterone antagonist, n (%)
28 (68)
28 (68)
28 (68)
28 (68)
Diuretics, n (%)
25 (61)
25 (61)
25 (61)
25 (61)
Waist, cm
Creatinine, ␮mol/L
MDRD-estimated GFR, mL/min
Hemoglobin, mg/dL
Medication
a
b
No statistically significant differences were noted for any of the parameters across any of the test intervals.
ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker.
lated with 3-month coefficients (r ⫽ 0.33, P ⬍ 0.05).
We did not find any correlation between relative interval change in NT-proBNP and the corresponding interval changes in weight or waist circumference, renal
function (as indicated by either serum creatinine or
MDRD-estimated GFR), or heart rate at any of the 4
sampling intervals (data not shown). For hemoglobin,
only the relative interval change at 3-month follow-up
correlated with relative interval change in NT-proBNP
(r ⫽ 0.49, P ⬍ 0.01), with no correlation found at the
other sampling intervals.
Regression analyses did not reveal any significant
influence of baseline NT-proBNP concentrations on
the corresponding change values (P ⫽ 0.38, P ⫽ 0.69,
P ⫽ 0.18, and P ⫽ 0.48 for 14-day, 1-month, 2-month,
and 3-month values, respectively). Consequently, no
further subdivision according to NT-proBNP grouping was performed. The corresponding graphs can be
Table 3. Analytical and biological variation, reference change values, and index of individuality of NTpro-BNP
according to sampling interval.a
Measurement interval
14 Days (A)
14 Days (B)
14 Days (comb)
28 Days
62 Days
90 Days
b
14.2 ⫾ 0.5
13.9 ⫾ 0.9
14.1 ⫾ 0.6
28.1 ⫾ 1.2
63.4 ⫾ 2.9
91.5 ⫾ 2.9
CVt, %
18.5 (9.6–29.2)
11.4 (5.3–27.2)
17.7 (10.9–25.0)
19.0 (9.2–28.7)
15.6 (6.5–26.0)
16.3 (7.2–36.9)
CVi, %
18.4 (9.5–29.2)
11.3 (5.1–27.2)
17.6 (10.8–25.0)
18.9 (9.1–28.7)
15.5 (6.3–26.0)
16.2 (7.1–36.9)
RCV, %
51.1
31.4
34.6
52.5
43.1
45.0
tact, days
IOI
0.11
a
0.07
0.11
0.12
0.09
0.10
No statistically significant differences were noted for CV values and IOI across any of the test intervals CV values are given as median (interquartile range); (A),
first 14-day interval between measurements; (B), second 14-day interval between measurements; (comb), both 14-day intervals combined.
b
tact, Actually achieved time interval between measurements; CVt, total CV; CVi, individual biological CV.
926 Clinical Chemistry 55:5 (2009)
Biovariability of NT-proBNP in Stable Heart Failure
Fig. 1. Bland–Altman plot for NT-proBNP values of the 4 sampling intervals.
The dotted lines indicate 1.96 ⫻ SD of the mean of the differences (95% confidence with P ⬍ 0.05).
seen in Fig. 1. Likewise, NYHA functional class, sex,
age, weight or waist circumference, heart rate, hemoglobin levels, and ejection fraction were not found to
significantly influence the different change values (data
not shown). Renal function, as indicated by either serum creatinine or MDRD-estimated GFR, was found
to have a small but significant relationship with 14-day
change values (␤-coefficient, ⫺0.18; SE, 0.08; intercept, 0.42; SE, 0.1; r 2, 0.12; residual SD, 0.13; P ⫽ 0.03
for creatinine; ␤-coefficient, 0.003; SE, 0.001; intercept,
0.02; SE, 0.09; r 2, 0.12; residual SD, 0.13; P ⫽ 0.04 for
MDRD-estimated GFR) with no significant relationship with the other change values (P ⫽ 0.29, P ⫽ 0.60,
and P ⫽ 0.14 for 1-month, 2-month, and 3-month values of MDRD-estimated GFR, respectively).
Discussion
We determined the intermediate-term biological variation of NT-proBNP in stable chronic heart failure and
its relation to change in known confounding variables.
Importantly, we did so in an outpatient setting in patients who were on individually optimized stable medication regimens. We found that patients with stable
chronic heart failure and on stable medication had NTproBNP biological variation of 11% to 20% with sampling periods ranging from 14 days to 3 months. We
also found that the corresponding indices of individu-
ality for these sampling periods ranged from 0.07– 0.15
and indicated high individuality of values. Lastly, the
individual biological variability of NT-proBNP does
not appear to be influenced by known confounders for
NT-proBNP itself, such as NYHA functional class, sex,
age, weight or waist circumference, hemoglobin levels,
ejection fraction, or renal function.
We chose to investigate NT-proBNP because it is
more stable than BNP and its variation appears to be
lower (14 ). Also, to the best of our knowledge, this is
the first study to address biovariability of natriuretic
peptides during intermediate follow-up intervals and
only the second study to address indices of individuality in chronic heart failure. In patients the confounding
effects of clinical stability and titration of drugs must be
controlled to assess true biological variability. In this
context it is interesting to note that only approximately
one-ninth of patients attending our clinic were eligible
– or inversely, about 90% of patients in an outpatient
setting presented some degree of clinical instability.
Previous reported studies on short- or long-term
variability found RCVs around 100% (8, 9 ) (11 ), others found CVs between 20% and 30% (10, 12, 13,
25, 26 ). The key to understanding these results is the
rigidity of the definition of stability of heart failure. In
studies that expended large efforts to ascertain stability
(12, 25, 26 ) the CVs found were lower than in studies
with less strict criteria (9, 11 ). Consequently, we found
Clinical Chemistry 55:5 (2009) 927
CVs for NT-proBNP to be around 20%. Our study is a
valuable extension to current knowledge because
intermediate-term variation must be quantified to aid
interpretation of trials targeting guidance of medical
therapy or routine application as they relate to normal
follow-up duration in outpatient settings.
Important points must be considered with regard
to estimation of CVs for NT-proBNP. The skewed distribution of NT-proBNP values can be converted by
log-transformation to a normal distribution (12, 25 ), a
process that results in lower CVs and eliminates the
problem of a paradoxical decrease ⬎100%. It must be
emphasized, however, that CVs derived by this approach would apply only to log-transformed values,
thus limiting clinical applicability. After retransformation as suggested by Fokkema et al. (27 ), the CV
borders are comparable to those derived by use of
nontransformed values. In addition, the CV of logtransformed values is not independent of the scale of
the values, necessitating strict adherence to scale. A
more important consideration is that the concept of
biological variation around a homeostatic set-point
might not hold for NT-proBNP (14 ). This could be the
case because CVs for natriuretic peptides do not represent random variation, but rather the close balance between hemodynamics and neuroendocrine regulation
vs counterregulation (28 ); pulsatile secretion of BNP
(and consequently NT-proBNP) has already been
demonstrated in healthy individuals (29 ).
The interpretation of IOI has been reported previously (21 ) and is detailed above. So far, only 1 reported
study (25 ) has included data on IOI in heart failure
patients. As in the present study for intermediate
follow-up, the earlier study demonstrated that IOIs of
week-to-week variations were relatively low, paralleling our findings indicating that NT-proBNP shows
high interindividual variation with relative stability of
values for the individual patient. This relative stability
of NT-proBNP in the individual patient compared to
between-patient variability further strengthens the notion raised above that NT-proBNP might not vary randomly and could be an explanation for the difference
between high CVs observed in stable patients vs small
differences in serial NT-proBNP concentrations when
these indices are successfully used in therapy-guiding
trials (1, 2 ). Recently it has been suggested that the additional prognostic value of serial evaluation might be
more related to a categorical change with respect to a
cutoff concentration (6, 7 ) than to the use of relative or
absolute changes in concentration.
Despite the widespread use of natriuretic peptides,
there is a paucity of data on possible confounders for
serial measurements in patients with stable heart failure. We confirmed the only 2 available studies regarding the absence of influence of age, sex, and renal func928 Clinical Chemistry 55:5 (2009)
tion on short-term (25 ) and long-term (13 ) variations
of NT-proBNP and have extended these findings to
important confounders such as NYHA functional categories and obesity as represented by weight and waist
circumference. More importantly, we did not find any
relationship between change of NT-proBNP and
changes in renal function (either for serum creatinine
or MDRD-estimated GFR) or body weight (as indicated by weight or waist circumference) at any of the 4
sampling intervals. Only medium-term, but not shortterm, changes in hemoglobin were found to be related
to changes in NT-proBNP. Because hemoglobin influences NT-proBNP concentrations, it is possible that
the short-term changes in hemoglobin measured here
reflect biovariability or smaller changes in fluid intake
that would not exert sizeable influence on NT-proBNP
levels, whereas medium-term changes in hemoglobin
might be a reflection of more slowly developing patterns (e.g., evolving anemia), which could influence
NT-proBNP. These findings are important in clinical
practice, in which change of NT-proBNP will be interpreted in light of other variables. Furthermore, such
findings again provide indirect support for the notion
that NT-proBNP does not vary randomly.
Our study has several limitations. The relatively
small number of patients included may hinder adequate statistical analysis, but the size of our study population is comparable to previous, related studies. In
addition, the numerous and strict selection criteria
should give this study even greater value, although
these criteria might also adversely impact on clinical
variables and patient selection, effectively reducing the
number of individuals included in the study and possibly the clinical transferability of the findings. We
think that RCVs should be established only in clinically
stable patients to allow later application to more unstable settings. Loss of individuals from the study due to
patients dropping out could have introduced possible
bias to the study. We cannot completely rule out the
presence of hidden hemodynamic changes; it is possible that undetected intermittent changes of rhythm
contributed to the large variability seen in some patients and might explain some of the discrepancy between obvious clinical stability and large changes in
measured marker values. Because we did not perform
invasive testing or long-term Holter follow-up, we can
only speculate about this possibility. Consequently, despite the rigorous definition of stability used in our
study, it is still possible that we did not measure pure
biological variability but that to a certain extent clinically undetected variations in phenotype may have
been included in our determinations.
In conclusion, despite previous reports of values
around 80%, variation of NT-proBNP at 14-day,
1-month, 2-month, and 3-month follow-up is lower in
Biovariability of NT-proBNP in Stable Heart Failure
a prespecified population of patients with stable heart
failure and is not related to changes in renal function or
weight. In view of the low index of individuality we
observed, within-person variation is quite low compared to between-person variation. Consideration of
these facts is important for the interpretation of trials
and the use of NT-proBNP in monitoring patients with
heart failure.
Author Contributions: All authors confirmed they have contributed to
the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design,
acquisition of data, or analysis and interpretation of data; (b) drafting
or revising the article for intellectual content; and (c) final approval of
the published article.
Authors’ Disclosures of Potential Conflicts of Interest: Upon
manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: G. Hess, Roche Diagnostics; D.
Zdunek, Roche Diagnostics.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: C. Zugck, Novartis, Astrazeneca, PHTS, and Roche.
Research Funding: C. Zugck, PHTS, Roche, and AOK.
Expert Testimony: None declared.
Role of Sponsor: The funding organizations played no role in the
design of study, choice of enrolled patients, review and interpretation
of data, or preparation or approval of manuscript.
References
1. Troughton RW, Frampton CM, Yandle TG, Espiner
EA, Nicholls MG, Richards AM. Treatment of
heart failure guided by plasma aminoterminal
terminal brain natriuretic peptide (N-BNP) concentrations. Lancet 2000;355:1126 –9.
2. Jourdain P, Jondeau G, Funck F, Gueffet P, Le
Helloco A, Donal E, et al. Plasma brain natriuretic
peptide-guided therapy to improve outcome in
heart failure: the STARS-BNP Multicenter Study.
J Am Coll Cardiol 2007;49:1733–9.
3. Lainchbury JG, Troughton RW, Frampton CM,
Yandle TG, Hamid A, Nicholls MG, Richards AM.
NTproBNP-guided drug treatment for chronic
heart failure: design and methods in the
“BATTLESCARRED” trial. Eur J Heart Fail 2006;8:
532– 8.
4. Bettencourt P, Friões F, Azevedo A, Dias P, Pimenta J, Rocha-Gonçalves F, Ferreira A. Prognostic information provided by serial measurements
of brain natriuretic peptide in heart failure. Int
J Cardiol 2004;93:45– 8.
5. Miller WL, Hartman KA, Burritt MF, Grill DE,
Rodeheffer RJ, Burnett JC Jr, Jaffe AS. Serial
biomarker measurements in ambulatory patients
with chronic heart failure: the importance of
change over time. Circulation 2007;116:249 –57.
6. Latini R, Masson S, Wong M, Barlera S, Carretta
E, Staszewsky L, et al. Val-HeFT Investigators.
Incremental prognostic value of changes in
B-type natriuretic peptide in heart failure. Am J
Med 2006;119:70.e23–30.
7. Masson S, Latini R, Anand IS, Barlera S, Angelici
L, Vago T, et al. Val-HeFT Investigators. Prognostic value of changes in N-terminal pro-brain natriuretic peptide in Val-HeFT (Valsartan Heart Failure Trial). J Am Coll Cardiol 2008;52:997–1003.
8. Wu AH, Smith A. Biological variation of the natriuretic peptides and their role in monitoring
patients with heart failure. Eur J Heart Fail 2004;
6:355– 8.
9. Bruins S, Fokkema MR, Römer JW, Dejongste MJ,
van der Dijs FP, van den Ouweland JM, Muskiet
FA. High intraindividual variation of B-type natriuretic peptide (BNP) and amino-terminal proBNP
in patients with stable chronic heart failure. Clin
Chem 2004;50:2052– 8.
10. O’Hanlon R, O’Shea P, Ledwidge M, O’Loughlin
11.
12.
13.
14.
15.
16.
17.
18.
C, Lange S, Conlon C, et al. The biologic variability of B-type natriuretic peptide and N-terminal
pro-B-type natriuretic peptide in stable heart failure patients. J Card Fail 2007;13:50 –5.
Wu AH, Smith A, Wieczorek S, Mather JF, Duncan
B, White CM, et al. Biological variation for
N-terminal pro- and B-type natriuretic peptides
and implications for therapeutic monitoring of
patients with congestive heart failure. Am J Cardiol 2003;92:628 –31.
Cortes R, Rivera M, Salvador A, Bertomeu V, de
Burgos FG, Rosello-Lleti E, et al. Variability of
NT-proBNP plasma and urine levels in patients
with stable heart failure: a 2-year follow-up
study. Heart 2007;93:957– 62.
Schou M, Gustafsson F, Kjaer A, Hildebrandt PR.
Long-term clinical variation of NT-proBNP in stable chronic heart failure patients. Eur Heart J
2007;28:177– 82.
Clerico A, Carlo Zucchelli G, Pilo A, Passino C,
Emdin M. Clinical relevance of biological
variation: the lesson of brain natriuretic peptide
(BNP) and NT-proBNP assay. Clin Chem Lab Med
2006;44:366 –78.
Takeda Y, Fukutomi T, Suzuki S, Yamamoto K,
Ogata M, Kondo H, et al. Effects of carvedilol on
plasma B-type natriuretic peptide concentration
and symptoms in patients with heart failure and
preserved ejection fraction. Am J Cardiol 2004;
94:448 –53.
Swedberg K, Cleland J, Dargie H, Drexler H, Follath F, Komajda M, et al. Guidelines for the
diagnosis and treatment of chronic heart failure:
executive summary (update 2005): the Task Force
for the Diagnosis and Treatment of Chronic Heart
Failure of the European Society of Cardiology. Eur
Heart J 2005;26:1115– 40.
Yeo KT, Wu AH, Apple FS, Kroll MH, Christenson
RH, Lewandrowski KB, et al. Multicenter evaluation of the Roche NT-proBNP assay and comparison to the Biosite Triage BNP assay. Clin Chim
Acta 2003;338:107–15.
Clerico A, Prontera C, Emdin M, Passino C, Storti
S, Poletti R, et al. Analytical performance and
diagnostic accuracy of immunometric assays for
the measurement of plasma B-type natriuretic
peptide (BNP) and N-terminal proBNP. Clin Chem
2005;51:445–7.
19. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N,
Roth D. A more accurate method to estimate
glomerular filtration rate from serum creatinine: a
new prediction equation. Modification of Diet in
Renal Disease Study Group. Ann Intern Med
1999;130:461–70.
20. Levey AS, Greene T, Kusek JW, Beck GJ. A simplified equation to predict glomerular filtration
rate from serum creatinine [abstract]. J Am Soc
Nephrol 2000;11:A0828.
21. Petersen PH, Sandberg S, Fraser CG, Goldschmidt
H. Influence of index of individuality on false
positives in repeated sampling from healthy individuals. Clin Chem Lab Med 2001;39:160 –5.
22. Iglesias N, Petersen PH, Ricós C. Power function
of the reference change value in relation to cutoff points, reference intervals and index of individuality. Clin Chem Lab Med 2005;43:441– 8.
23. Bland JM, Altman DG. Statistical methods for
assessing agreement between two methods of
clinical measurement. Lancet 1986;1:307–10.
24. Bland JM, Altman DG. Measuring agreement in
method comparison studies. Stat Methods Med
Res 1999;8:135– 60.
25. Schou M, Gustafsson F, Nielsen PH, Madsen LH,
Kjaer A, Hildebrandt PR. Unexplained week-toweek variation in BNP and NT-proBNP is low in
chronic heart failure patients during steady state.
Eur J Heart Fail 2007;9:68 –74.
26. Troughton RW, Richards AM. BNP for clinical
monitoring of heart failure. Heart Fail Clin 2006;
2:333– 43.
27. Fokkema MR, Herrmann Z, Muskiet FA, Moecks J.
Reference change values for brain natriuretic
peptides revisited. Clin Chem 2006;52:1602–3.
28. Clerico A, Recchia FA, Passino C, Emdin M. Cardiac endocrine function is an essential component of the homeostatic regulation network:
physiological and clinical implications. Am J
Physiol Heart Circ Physiol 2006;290:H17–29.
29. Pedersen EB, Pedersen HB, Jensen KT. Pulsatile
secretion of atrial natriuretic peptide and brain
natriuretic peptide in healthy humans. Clin Sci
(Lond) 1999;97:201– 6.
Clinical Chemistry 55:5 (2009) 929