| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Medicine (T.C.R.P., C.F., T.G.Y., A.M.R., E.A.E.) and Paediatrics (B.A.D.), Christchurch School of Medicine and Health Sciences, and Neonatal Service (B.D., B.A.D.), Christchurch Women’s Hospital, Christchurch 8140, New Zealand
Address all correspondence and requests for reprints to: Timothy Prickett, Department of Medicine, Christchurch School of Medicine and Health Sciences, P.O. Box 4345, Christchurch Mail Centre, Christchurch 8140, New Zealand. E-mail: tim.prickett{at}chmeds.ac.nz.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objectives: Our objective was to determine the effects of prematurity, gender, and antenatal steroids on plasma NTproCNP at birth, and serial changes in hormone concentrations, linear growth, and markers of bone turnover in the first month of postnatal life.
Design and Setting: This is a prospective study of newborn infants admitted to an intensive care unit.
Subjects: A total of 48 infants (four gestation groups) were enrolled. Umbilical cord samples were also obtained from 39 healthy term infants.
Main Outcome Measures: Plasma NTproCNP and CNP were measured in cord plasma. In enrolled neonates, serial measurements of hormone concentrations and markers of bone turnover were related to tibial growth velocity as measured by knemometry.
Results: Cord plasma NTproCNP was inversely related to gestational age (r = –0.35; P = 0.003) and was higher in males (P < 0.001). Plasma NTproCNP (P = 0.016) and CNP (P < 0.001) increased within the first week of life, the increase relating inversely to gestational age (r = –0.64; P < 0.001). Plasma NTproCNP at 1 wk was strongly correlated with linear growth velocity (r = 0.49; P < 0.001), and also at 2–4 wk, the relation being stronger than observed between bone turnover markers and growth velocity.
Conclusions: In neonates with diverse disorders affecting growth and nutrition, plasma NTproCNP was strongly correlated with linear growth during the first 4 wk of postnatal life and may prove to be a novel marker of growth plate activity in neonates.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
GA was determined by the best estimate from a combination of the first day of mother’s last menstrual period and an early ultrasound scan. Antenatal glucocorticoids, if required, were administered to the mother as betamethasone 12.5 mg im daily for 2 d before parturition. Cord blood (arterial) was obtained in 27 of the babies. Postnatal venous blood samples (an additional 1.5 ml) coincided with blood collections required for clinical care. Accordingly, timing of blood sampling varied across the cohort studied, was dependent on length of stay, and was more frequent in wk 1–4. To minimize blood sampling, no more than three samples were drawn from neonates less than 28-wk gestation. Weight and lower leg length (knemometry) were determined every 3–4 d.
Neonatal nutrition
All infants in the study were treated according to the CNU protocols. Our policy is to encourage enteral feeds with mother’s own breast milk. Generally, for infants not tolerating enteral feeds, parenteral nutrition was commenced on d 1–2, comprising 1 g/kg·d amino acid solution in 10–15% dextrose and increasing as tolerated to 3.5 g/kg·d; and 20% Intralipid (Kabi Pharmacia AB, Stockholm, Sweden) 1 g/kg·d and increasing as tolerated to 3 g/kg·d. Fat (Vitlipid N; Kabi Pharmacia AB; 4 ml/kg·d) and water-soluble (Soluvit N; Kabi Pharmacia AB; 1 ml/kg·d) vitamins were added to the Intralipid (7), and a standard trace element solution was added to the amino acid dextrose solution. Particularly in preterm infants, calorie intake was monitored on a daily basis, and intake increased as appropriate to achieve optimal growth.
Linear growth
Postnatal growth velocity was measured using change in lower leg length as monitored by knemometry (8). The knemometer is an electronic caliper that measures the distance between two plates. Infants were positioned either supine or lying on the left side with the right leg free and exposed and bent to 90° at hip and knee. The right knee was placed against the fixed plate and the leg held gently in place as the sliding arm was advanced until the predetermined pressure was achieved, and the recording device triggered. Measurements were repeated five times, and the mean and variance of readings calculated. Readings were accepted if the variance of measurements was less than 0.8 mm. If greater, the process was repeated. Of all knemometry measurements, 95% were made by the same investigator. In this study the error (determined by calculating the SD of the variance at each reading) was 0.485 mm, equivalent to 1.18-d lower leg growth. In each infant the slope of the regression line, calculated from all leg measurements (median number of measurements per infant was nine, range 3–30) over a period of 3–6 wk (median 37 d), was used as the index of linear growth velocity. The range of r values was 0.37–0.99 (median 0.97; n = 47). To provide growth measurements during the first 3 months of life, crown-heel length (as measured by the community nurse) was obtained from the infant’s postnatal record.
For comparative purposes, cord plasma samples were also obtained from 39 healthy term infants (23 females) delivered (by cesarean section in 20) at Christchurch Women’s hospital over a similar time period.
Plasma assays
Blood samples were collected into standard blood collection tubes containing EDTA (7.5 mg/ml, Vacutainer; BD, Franklin Lakes, NJ), centrifuged at 4 C, and the plasma was stored at –75 C before analysis for CNP, NTproCNP, type 1 collagen C-telopeptide [β-CrossLaps (CTx)], and procollagen type 1 amino-terminal propeptide (PINP). The collagen markers were measured using an Elecsys 2010 analyzer (Roche Diagnostics, Mannheim, Germany).
RIA for amino-terminal proCNP
NTproCNP was assayed in duplicate as previously described (3, 6) using the primary rabbit antiserum (J39) raised against NTproCNP(1–15) (100 µl, 1:6000 diluted antiserum per assay tube). Peptide standards were made from synthetic human proCNP(1–19) considering the purity data supplied (Chiron Technologies Pty. Ltd., Parkville, Australia). Within- and between-assay coefficients of variation were 4.9 and 6.4%, respectively, at 22 pmol/liter.
RIA for CNP
CNP was assayed as previously described (9) using a commercial antiserum (catalog no. RAB-014-03; Phoenix Pharmaceuticals, Inc., Belmont, CA). The rabbit antiserum raised against proCNP(82–103) shows 100% cross-reactivity with CNP-22 and hCNP-53 (Phoenix Pharmaceuticals data sheet). Within- and between-assay coefficients of variation were 3.6 and 8.3%, respectively, at 7.5 pmol/liter. Because of limited volume of sample, all CNP measurements were singletons.
Statistical methods
Data are presented as mean ± SEM where appropriate. The paired Student’s t test was used to analyze differences in analyte levels. ANOVA with repeated measures was used to assess changes in biochemical measurements with time. Pearson product-moment correlation coefficient was used to determine correlation between variables. Statistical significance was assumed when P < 0.05.
All procedures in this study were approved by the Canterbury Ethics Committee and were undertaken only after informed parental consent.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Clinical details are listed in Table 1
. Of the 49 neonates enrolled and studied in the CNU (45% female), one infant died on d 6. Data from this infant were not included in the analysis. A second infant (included in the study results) died after the study had been completed. A total of 37 infants were exposed to at least one antenatal injection of betamethasone (12.5 mg), 12 of which received one or more injections within 72 h of birth. Postnatal infections were diagnosed in 20 infants. Eight infants received total parenteral nutrition for 20 or more days, and 19 required surgery on at least one occasion. Of the 48 infants, 23 (48%) weighed less than 1500 g at birth; six (five males) were classified as small for GA (<10th centile for birth weight). Plasma creatinine concentrations were within the normal reference range in all neonates throughout the study period.
|
Combining results from the CNU and normal term infants (n = 66), the cord concentration of NTproCNP was significantly (and inversely) related to GA (r = –0.36; P = 0.003), birth weight (r = – 0.36; P = 0.003), and birth length (r = –0.41; P = 0.001) (Fig. 1). When birth weight or length was corrected for GA (i.e. expressed as SD scores), the relationships with NTproCNP were lost. Plasma CNP concentration in cord plasma was also inversely correlated with GA (r = –0.21; P = 0.09). However, the correlation between the two CNP forms was not significant (r = –0.04; n = 66).
|
). The difference could not be attributed to maturity because birth weight and GA in the two groups did not differ significantly.
|
In the CNU study group, no association of umbilical cord plasma concentration of calcium, phosphate, or alkaline phosphatase was found with NTproCNP concentration.
Postnatal changes (CNU study)
Sequential changes in plasma NTproCNP after birth are shown in Fig. 3A. Concentrations during the first week of life (241 ± 25 pmol/liter) were significantly higher than mean umbilical plasma NTproCNP concentration (180 ± 11 pmol/liter; P < 0.016)) in neonates in whom both time points were sampled (n = 24). Similarly, plasma CNP increased during the same time interval (10.3 ± 1.3 vs. 4.7 ± 0.3 pmol/liter at birth; P < 0.001). The step-up in plasma NTproCNP concentration during the first week of postnatal life was strongly (and inversely) related to gestation (r = –0.66; P < 0.001; n = 24; Fig. 3B). Birth weight (r = –0.62; P < 0.001) and GA (r = –0.59; P < 0.001; Fig. 4), but not z score, were also correlated with plasma NTproCNP concentration during the first week of life. Whereas NTproCNP concentrations in the first week were lower in infants recently exposed to steroids than in the untreated matched group [238 ± 25 (n = 12) vs. 307 ± 51 pmol/liter (n = 8)], the difference was not significant (P = 0.2).
|
|
Skeletal growth and NTproCNP
The mean increase in leg length was 0.41 mm ± 0.15 SD/d during the study period (median 37 d). As shown in Fig. 5, strong positive relationships between lower leg length and postmenstrual age (r = 0.91; P = 0.001) and body weight (r = 0.96; P < 0.001) were identified. Growth velocity was inversely correlated with birth weight (r = –0.68; P < 0.001; Fig. 6A), lighter babies growing faster over the period of study, but unrelated to z score (r = –0.1).
|
|
. As shown in Fig. 6B
, plasma NTproCNP concentration in the first week of life was strongly correlated with lower leg growth velocity in the subsequent 3- to 4-wk period of study (r = 0.49; P < 0.001; n = 44). Plasma NTproCNP was also significantly correlated with growth velocity in wk 2–4 (Table 2), whereas correlations between CNP and growth velocity were lower and inconsistent. Relating to growth velocity, when "head to head" comparison of plasma NTproCNP was made with the C/P ratio in infants having both measurements, similar correlations were found, although the relationship was stronger for NTproCNP (Table 2). Neither plasma NTproCNP nor the C/P ratio correlated with growth velocity over 3 months as calculated from birth length and length at 3 months of postnatal age.
|
In keeping with lower birth weight, four sets of bichorial twins exhibited significantly higher concentrations of plasma NTproCNP in the first postnatal week (Fig. 4). Thus, lower birth weight (1218 ± 120 g; n = 8) compared with other infants in the CNU study (1983 ± 163 g; n = 42; P = 0.05) but similar GA [211 ± 7 (n = 8) vs. 226 ± 5 pmol/liter (n = 42); P = 0.2] and cord concentrations [211 ± 26 (n = 6) vs. 177 ± 13 pmol/liter (n = 21); P = 0.23] was associated with increased plasma NTproCNP at 1 wk (334 ± 26, n = 8, vs. 223 ± 16, n = 36; P = 0.004). The C/P ratio at 1 wk in the twins (mean 0.14 ± 0.02; n = 6) was also significantly different (P = 0.03) from singletons (mean 0.33 ± 0.04; n = 20). However, growth velocity, although greater in twins (0.45 ± 0.03 vs. 0.42 ± 0.02 mm/d), did not differ between the two groups (P = 0.46).
ANOVA failed to show any significant effects of surgery, parenteral nutrition, infections, or insulin treatment on plasma CNP forms.
Plasma CNP and NTproCNP interaction
As noted previously in cord plasma, there was no correlation between the two CNP forms in postnatal samples (r = 0.08; n = 47). In cord plasma, the mean ratio of NTproCNP to CNP was 46.8 ± 2.7 (n = 66). The ratio was unaffected by gestation and was similar (mean 43.1 ± 4.5) in infants after 4 wk of life (n = 17).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
In keeping with the decreasing fetal growth rate in the second half of pregnancy, we found an inverse relationship between plasma NTproCNP in cord plasma and GA. Similar inverse relationships with NTproCNP were noted with birth weight and birth length, but these disappeared when normalized for GA. The source of these high circulating concentrations in the fetus is unknown. NTproCNP, a stable product of CNP synthesis in tissues (4), is likely to be cosecreted with (bioactive) CNP-53, which presumably mediates expansion of the growth plate (1, 12). Transcripts for all known components of the CNP signaling pathway have been identified in fetal long bones of rodents (13), and the fetal mouse tibia is highly responsive to exogenous CNP ex vivo (14). Together, these findings suggest that the fetal growth plate itself is one important source of circulating NTproCNP. However, genetic manipulations or spontaneous mutations in most (1, 15, 16, 17, 18) but not all cases (10, 19, 20), although markedly impacting on postnatal growth, do not appear to affect size at birth. Even if not always the driver of fetal endochondral growth, our data are consistent with the involvement of CNP at some point in the growth process. The inhibiting effect of glucocorticoids on NTproCNP (4, 12), the higher concentrations in males (4), and the similar response of the fetal (21) and postnatal lamb (5) to undernutrition provide further support for this view. In 4-wk-old lambs, glucocorticoids promptly and reversibly inhibit CNP synthesis and metacarpal growth velocity, the effects of which abate within several days once the steroid is withdrawn (4). Although we did not conduct a randomized prospective study, the lower concentrations of NTproCNP in infants exposed to betamethasone within 72 h of birth are consistent with our previous findings in young lambs.
The strong inverse relation of NTproCNP with GA noted at birth continued and, in fact, was strengthened by observations made in the first postnatal week. Whereas significant increase in plasma NTproCNP within 1 wk occurred in the group as a whole, the increase above cord concentrations was most remarkable in premature infants, and was highly dependent on GA. Together, these findings constitute direct evidence that de novo production of CNP occurs after birth in humans, and indicate that the placenta is unlikely to be a major source of CNP in the fetal circulation. Overall, the findings suggest that CNP synthesis in the fetus is suppressed during parturition, concentrations returning to those appropriate for GA within a 1- to 2-wk period after birth. Notably, the dramatic increase occurs during a period of relative undernutrition (mean weight gain at 1 wk was –36 ± 16 g), which in the healthy fetal and postnatal lamb impairs CNP synthesis. Although the mechanisms underpinning these changes, and factors regulating CNP synthesis in general, are not known, it is possible that perinatal changes in thyroid hormone concentrations or activity (22, 23) play at least a permissive role. Thyroid hormones are crucial to growth and maturation of the growth plate (24), increase within a few days of birth, and could affect the cellular activity of growth plate-related tissues from which CNP is sourced (1).
Clarifying the link, if any, between postnatal plasma NTproCNP concentration and linear growth was an important objective in this study for at least two reasons. First, coupling NTproCNP concentrations with the rapid changes in growth that normally characterize the new born infant would be an important contribution to our understanding of CNP’s role in human biology. Second, linear growth is an accepted index of well-being yet cannot be assessed prospectively. Therefore, a biochemical marker could be useful to clinicians in settings in which growth was in doubt. Using the growth rate of the tibia, we found that the plasma NTproCNP concentration within the first week accounted for approximately half of the variation in growth velocity in individual infants over the first 3- to 4-wk postnatal life. Even stronger correlations were found using the plasma NTproCNP concentration in subsequent weeks. These highly significant associations (r values ranging from 0.49–0.62) are similar to those linking NTproCNP with metacarpal growth in lambs (0.55) or height increments in older children (0.57) (4). A variety of biochemical markers, including markers of bone turnover and growth factors (e.g. IGF-I and associated binding proteins), have been used to assess growth velocity in previous studies of neonates. Kajantie et al. (25) found that the ratio of carboxyl-terminal telopeptide of type 1 collagen (a marker of bone resorption) to PINP (a marker of bone formation) was the most accurate predictor of impaired growth in newborn infants with very low birth weight (VLBW), infants with higher ratios exhibiting lower growth rates. Using similar indices, in agreement with Kajantie et al., we found the ratio C/P superior to either alone. However in our population the ratio was less consistent than plasma NTproCNP, particularly within the first 2 wk. Possibly relevant here is the different pattern of growth velocity in the two studies. In our series the lighter infants at birth grew faster (Fig. 6A), whereas the opposite was true in infants with VLBWs. Furthermore, we measured tibial length using the slope of the regression line (lower leg length mm/d) over a median of 37 d, in contrast to Kajantie et al. (25), who measured changes at intervals of 1 wk. Although growth velocity was highly linear in our study, these different methods as well as different growth patterns could affect outcomes.
In contrast to our previous studies, in many of which CNP concentrations were closely correlated with simultaneously drawn NTproCNP values (4, 5), no significant correlation between the two CNP forms was found in cord or postnatal samples in the current study. CNP is more susceptible than NTproCNP to degradation at ambient temperature, which could explain some of the discrepancy in cord samples (in which some delay in sampling was inevitable), but not in postnatal samples. Several other factors are known to affect CNP, including uptake and degradation by the clearance receptor (natriuretic peptide receptor C) and neprilysin (4), respectively. The changing activity of these pathways and fluctuating concentrations of competitive substrates such as BNP (26, 27) across parturition (28) could contribute to our findings. Until further work clarifies the discrepancy, our findings strongly favor the measurement of NTproCNP rather than CNP in neonatal studies.
The present studies undertaken in neonates with diverse disorders affecting growth and nutrition show that in the newborn infant, plasma NTproCNP is inversely related to GA and is higher in males at birth. Plasma NTproCNP increases abruptly within the first week of postnatal life, increases that are inversely related to GA. As found in older children, the concentration of NTproCNP in neonates is strongly correlated with linear growth, suggesting that it may be a novel marker of growth velocity, not just in neonates but during childhood (4) and adolescence (29). It is instructive to compare these results using NTproCNP with those studies using plasma IGF-I concentration, a well-accepted circulating (GH dependent) growth factor of hepatic origin that is also reported to reflect short-term growth velocity in some settings (30). In strong contrast to NTproCNP, plasma IGF-I concentrations are low at birth and increase slowly during childhood to peak during adolescence (31). Similarly, IGF-I concentrations increase as the fetus matures, and are lower in infants born prematurely (32, 33). It follows that, with the possible exception of VLBW infants (30) or children receiving GH treatment (34), measurements of IGF-I cannot be expected to reflect concurrent growth velocity in the individual (preadolescent) child. Recent reports confirm this (35) and emphasize the inadequacy of serum bone markers in specifically reflecting linear growth (36). Whether this desirable goal can be met using a marker of CNP synthesis, a highly potent cartilage growth factor and presumably sourced from the actively growing skeleton, now requires prospective studies, both in normal infants and children as well as those with disorders of growth.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Disclosure Statement: The authors have no conflicts to disclose.
First Published Online October 30, 2007
Abbreviations: ANP, Atrial natriuretic peptide; BNP, brain natriuretic peptide; CNP, C-type natriuretic peptide; C/P, β-CrossLaps to procollagen type 1 amino-terminal propeptide; CTx, β-CrossLaps; GA, gestational age; NTproCNP, amino-terminal pro C-type natriuretic peptide; PINP, procollagen type 1 amino-terminal propeptide; VLBW, very low birth weight.
Received August 14, 2007.
Accepted October 24, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. C R Prickett, G. K Barrell, M. Wellby, T. G Yandle, A M. Richards, and E. A Espiner Effect of sex steroids on plasma C-type natriuretic peptide forms: stimulation by oestradiol in lambs and adult sheep J. Endocrinol., December 1, 2008; 199(3): 481 - 487. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
