PEDIATRICS Vol. 119 No. 5 May 2007, pp. 936-940 (doi:10.1542/peds.2006-2986)
ARTICLE |
Therapeutic Drug Monitoring for Caffeine in Preterm Neonates: An Unnecessary Exercise?
a Divisions of Neonatology
c Biostatistics
d Clinical Pharmacology and Toxicology and the Pediatric Pharmacology Research Unit, Children's Hospital of Michigan, Detroit, Michigan
b Department of Pharmacy, Hutzel Women's Hospital, Detroit, Michigan
 |
ABSTRACT |
|---|
 TOP ABSTRACT PATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
OBJECTIVE. Our goal was to determine the value of measuring plasma caffeine levels in preterm neonates treated with caffeine for apnea. We evaluated plasma concentrations of caffeine attained in preterm neonates at standard doses, at varying postconceptual ages, with renal or hepatic dysfunction and when there was clinical lack of efficacy. We hypothesized that measurement of plasma caffeine concentrations during apnea therapy is not clinically helpful.
PATIENTS/METHODS. An observational study was conducted at Hutzel Women's Hospital between January 2000 and September 2005. Preterm neonates who were being treated with caffeine and who had a plasma caffeine level measured on at least 1 occasion were included.
RESULTS. A total of 231 caffeine blood levels were obtained from 101 preterm neonates with a median gestation of 28 weeks (range: 23–32 weeks) and birth weight of 1030 g (range: 540–2150 g). The caffeine citrate dose used ranged form 2.5 to 10.9 mg/kg (median: 5 mg/kg), and the levels ranged from 3.0 to 23.8 mg/L. Levels were between 5.1 and 20 mg/L in 94.8%, <5 mg/L in 2.1%, and >20 mg/L in 3.1%. Levels in the 5.1 to 20 mg/L range were attained on 91.3% of occasions when there was concomitant renal dysfunction (n = 23) and in all cases of hepatic dysfunction (n = 13). The median (25th, 75th quartiles) levels drawn for lack of efficacy (14.1 [10.2, 8.3] mg/L; n = 94) were comparable to those obtained for routine monitoring (13.7 [11, 9] mg/L; n = 107).
CONCLUSIONS. A majority of preterm neonates attain plasma caffeine levels between 5 and 20 mg/L, independent of gestation. This observation held even for the small number of subjects with elevated blood urea nitrogen, serum creatinine, or liver enzyme levels. Therapeutic drug monitoring is not necessary when caffeine is used for the treatment of apnea of prematurity in neonates.
Key Words: caffeine • methylxanthine • apnea • therapeutic drug monitoring
Abbreviations: PCA—postconceptional age • BUN—blood urea nitrogen
Caffeine is a methylxanthine used extensively as first-line pharmacotherapy in apnea of prematurity. Its efficacy in reducing the frequency of apneic episodes and the need for mechanical ventilation is well established.1,2 More recently, a large randomized, controlled trial among very low birth weight infants demonstrated a significant decrease in the rate of bronchopulmonary dysplasia in the caffeine-treated group, with a low incidence of short-term toxicity.3
The recommended standard dosing for caffeine citrate is 20 to 40 mg/kg loading followed by 5 to 8 mg/kg per day as maintenance.4 Larger doses up to 20 mg/kg day in the periextubation period have shown higher rates of successful extubation, without evidence of harm in the first year of life.5 Pharmacokinetic studies in premature neonates have established that the half-life of caffeine is prolonged to 102.9 ± 17.9 hours and remains prolonged for as long as 38weeks' gestation, which reflects the maturational deficit of its hepatic biotransformation in the newborn.6 The transition to adult levels of elimination occurs at 
3 to 4 months.7 Other factors such as cholestatic jaundice and breastfeeding seem to further prolong the half-life of caffeine.8 Clinically effective plasma concentrations vary over a wide range of 5 to 50 mg/L and overlap with subtherapeutic concentrations.6,8,9
Despite extensive pharmacokinetic data, the value of measuring caffeine plasma concentrations in neonates is unclear. Monitoring of serum concentrations of caffeine if there is lack of clinical response or if there is suspected toxicity has been recommended.10 Many centers also perform routine drug monitoring every week or 2 weeks, especially in extremely premature infants.8,11,12 We undertook an observational study to determine the value of measuring plasma caffeine levels in preterm neonates being treated with caffeine for apnea. Our specific aim was to evaluate the range of plasma caffeine levels attained in preterm neonates at varying postnatal and postconceptional ages (PCAs) or with renal or hepatic dysfunction, using standard maintenance doses. We also evaluated the plasma drug concentrations obtained in the subgroup of subjects in whom the indication for determining levels was lack of clinical efficacy. Because of the wide therapeutic index of caffeine and the dose-plasma concentration relationship in newborn infants with apnea, we hypothesized that measurement of plasma concentrations of caffeine during apnea therapy is not clinically helpful.
|
PATIENTS AND METHODS |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This was an observational study in which preterm infants treated with caffeine for apnea of prematurity, who had
1 blood levels obtained between January 2000 and September 2005 were identified. Clinical data, as well as plasma caffeine concentrations, were then included in the pharmacy database at Hutzel Women's Hospital, and the data were collated and analyzed. The institutional review board at Wayne State University approved the protocol. Data on the gestational age, birth weight, dose, and age at initiation of caffeine therapy; age at the time of obtaining the blood level; and renal and liver function profile within 3 days of obtaining the level were abstracted from the charts. The clinical indications for the drug level were noted from the medical chart when documented and classified as 1 of the following: lack of efficacy, suspected toxicity, or routine monitoring (trough level obtained on day 5 of therapy). The duration of caffeine therapy was noted. The decisions on when and whether to start caffeine, the doses, and need for and timing of plasma levels were taken by the attending neonatologist. Blood samples for assay of caffeine levels were usually drawn 2 hours before the next dose. Plasma caffeine levels were measured by the enzyme immunoassay technique (Dade Behring Inc, Deerfield, IL), and the results are reported in milligrams per liter.
SPSS 14 (SPSS, Chicago, IL) was used to obtain the median and quartile ranges of caffeine blood levels at different doses and PCAs.
|
RESULTS |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
A total of 137 infants were treated with caffeine during the study period. Thirty six of these were excluded because therapeutic drug monitoring was not performed, either because the apnea was well controlled at standard doses, the infant expired or was transferred out of the institution, or because of the clinical neonatologist's preference. Data from a total of 101 neonates were abstracted, on whom 231 caffeine levels were obtained.
Clinical Profile
The median gestational age at birth was 28 weeks (range: 23–32 weeks), and the median birth weight was 1030 g (range: 540–2150 g). Caffeine was initiated at a median age of 6 days (range: 1–70 days), with the third day of life being the most frequent day of initiation of therapy. The median dose of caffeine citrate was 5.0 mg/kg (range: 2.5–10.9 mg/kg). Caffeine levels were performed at a median PCA of 31 weeks (range: 24–41 weeks). The initial level was drawn at a median postnatal age of 14 days (range: 4–79 days) and a median PCA of 30 weeks (range: 24–39 weeks). Plasma drug concentrations of caffeine were obtained at a PCA of <28 weeks in 32 sample (13.8%), between 28 and 32 weeks in 132 (56.9%), and beyond 32 weeks in 67 (29.3%) of samples. The median number of levels per patient was 2 (range: 1–8). Of the 101 infants, 45 had a single level, 17 had 2 levels, 16 had 3 levels, 15 had 4 levels, 5 had 5 levels, and 1 infant each had 6, 7, and 8 levels obtained. The median duration of caffeine therapy was 29 days (range: 6–75 days).
Caffeine Plasma Drug Concentrations
The median caffeine concentration was 10.7 mg/L (SD: 4.0; range: 3–23.8 mg/L). The 25th quartile was 8.7 mg/L, and the 75th quartile was 14 mg/L. Figure 1 shows the distribution of levels in the study population and at different PCAs.
|
Effect of Renal and Hepatic Function on Caffeine Drug Concentrations
On the 23 occasions when serum creatinine was
1 mg/dL and/or blood urea nitrogen (BUN) was >30 mg/dL, the mean caffeine level was 13.9 (4.1) compared with 11.5 (4.1) mg/L on the 147 occasions when they were not. Plasma caffeine concentrations were within the 5 to 20 mg/L range on all 13 occasions when aspartate aminotransferase or alanine aminotransferase was higher than 60 U/L or direct bilirubin was >1 mg/dL (Table 1). The laboratory values were available within 3 days of only some samples. The elevation in BUN, serum creatinine, and liver enzyme levels was mild in most cases and related to hyperalimentation or drug effects. The highest BUN and creatinine in our data set was 78 mg/dL and 1.8 mg/dL, respectively. The peak aspartate aminotransferase and alanine aminotransferase in a single patient were 147 U/L and 268 U/L, respectively, and the direct bilirubin was 13.5 mg/dL.
|
Indications for Caffeine Therapeutic Drug Monitoring
The indications for obtaining caffeine levels were documented on 203 occasions. They were performed for lack of efficacy in 94 (46.3%) cases, routine monitoring in 107 (52.7%) cases, and for tachycardia (adverse effect) on 2 (1%) occasions. The median (25th, 75th quartiles) caffeine levels when performed for lack of efficacy and routine monitoring were 10.2 (8.3, 14.1) and 11 mg/L (9, 13.7), respectively. The levels on the 2 occasions with tachycardia were 13.9 and 22.7 mg/L.
Effect of Dose, Age, and Maturity on Caffeine Plasma Concentrations
The mean (SD) caffeine doses in the subgroups with plasma concentrations <5 mg/L, between 5.1 and 20 mg/L, and >20 mg/L were comparable at 5 (1.6), 5.2 (0.9), and 5.2 (1.1) mg/kg. The mean level obtained at a maintenance dose of 2.5 to 5 mg/kg was 11.18 mg/L (SD: 4.02 mg/L; n = 123), between 5.1 and 7.5 mg/kg was 12 mg/L (3.97 mg/L; n = 97), and >7.5 mg/kg was 9.86 mg/L (SD: 5.14 mg/L; n = 7). Table 2 shows the median and ranges of PCAs, caffeine dose, and multiple plasma levels. The number of infants who had 5 or levels were very few, which restricts the application of a repeated measures analysis of variance model to the smallest sample size in a cell level. For example, if we were to include level 5 data across the other previous levels, the sample size comparison would be restricted to a similar sample, that being an n of 8. We did, however, conduct the analysis by choosing a general linear model and found no statistically significant differences. Figure 2 depicts multiple levels at different PCAs, normalized to a dose of 1 mg/kg in the infants who had 3 levels.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our data suggest that the overwhelming majority (95%) of preterm neonates achieve plasma concentrations of caffeine between 5 and 20 mg/L at standard doses, independent of low gestation. This observation held even for the small number of infants who had elevated liver enzyme, serum creatinine, and/or BUN levels. In the subgroup of infants in whom caffeine plasma concentrations were obtained for lack of clinical efficacy, three quarters of the levels were <15 mg/L, which suggests that higher doses and plasma concentrations may be required for optimal efficacy.
Caffeine plasma concentrations achieved by the preterm neonates in our study were in accord with published pharmacokinetic data.6 Plasma concentrations between 7.4 and 19.4 mg/L (mean level: 13.7 mg/L) were reported at a maintenance dose (caffeine base) of 2.5 mg/kg in an early dose-finding study.6 A large population pharmacokinetic study reported mean serum caffeine concentrations of 7.4, 35.8, and 69 mg/L in 119 preterm neonates who were given 3, 15, and 30 mg/kg, respectively, once daily for a week.13 Plasma concentrations among specific ethnic groups (range: 3.6–28.4 mg/L with means between 10 and 20 mg/L) also fall within the same range.14 A previous population pharmacokinetic study based on routine therapeutic drug monitoring in 60 neonates reported that plasma concentrations within the traditional target range (5–20 mg/L) should be achieved in >70% of neonates.15 Another study on 16 preterm neonates found plasma levels within the 13 to 20 mg/L range in 69% of cases.16 Our study results were derived from a larger patient population in whom levels were obtained by the clinician for lack of efficacy, concerns about delayed clearance because of hepatic or renal dysfunction or extreme prematurity and rarely, adverse effects. It is reassuring that even in these specific situations, plasma concentrations of caffeine were as expected, and indeed within the recommended therapeutic range in >90% of infants.
This study also attempts to address the implication of a specific plasma level of caffeine. There is uncertainty on the precise desired plasma concentration and its correlation with efficacy. Although a decrease in apnea and increase in respiratory drive is known at plasma concentrations as low as 2.9 and 4 mg/L, optimal effect is at 10 mg/L.17 Higher doses and caffeine levels have been targeted in a few previous studies with some benefit and no adverse effects. A high dose regimen of 25 mg/kg loading followed by 6 mg/kg daily maintenance caffeine base generated mean (SD) plasma concentrations of 30.4 (4.0) mg/L and a more rapid reduction in apneas within 8 hours.16 Lee and associates,13 in a population pharmacokinetic study used daily caffeine citrate doses as high as 15 and 30 mg/kg over 7 days and attained mean caffeine levels of 35.8 and 69 mg/L, with no untoward effects. They suggested that preterm infants may require serum concentrations of 35 mg/L for effective prophylaxis against apnea after extubation. Our data set consisted of a large number of levels (n = 94), which had been obtained for a clinical concern of lack of efficacy. The median level in this subgroup was 10.2 mg/L, with a 75th quartile of 14.1 mg/L. These data taken collectively suggest that there are some preterm neonates who require daily doses of caffeine greater than previously recommended and that higher plasma levels are tolerated well in the short-term, although there are no long-term studies.
There was no correlation between caffeine levels and dose in our study within the relatively narrow dose range (5–8 mg/kg) used, probably because of the confounding effects of age and maturity. This relationship has been well established by other investigators, however.13,17 Plasma levels remained stable at later postnatal ages in infants in whom multiple levels were obtained. Because the PCA in most of our study group was <38 weeks, this was as expected.8
The limitations of our study are that the numbers of subjects who had renal or hepatic dysfunction at the time of a caffeine level were small. We were not able to evaluate efficacy in a systematic manner because the number of apneic episodes were not precisely recorded in the chart.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Routine monitoring of plasma concentrations of caffeine is unnecessary, even in extremely premature infants with renal or hepatic dysfunction or after prolonged use, because the overwhelming majority achieve concentrations in the range of 5 to 20 mg/L. In the subgroup of infants who do not show a clinical response to standard doses of caffeine, higher plasma levels may be targeted. Monitoring of plasma levels may be prudent in these cases. The risk–benefit of this approach needs to be studied in additional prospective clinical trials. Elimination of the practice of obtaining plasma levels would result in less blood draws and considerable cost-savings at a cost of $50 per assay.
|
FOOTNOTES |
|---|
Accepted Dec 29, 2006.
Address correspondence Girija Natarajan, MD, Division of Neonatology, Children's Hospital of Michigan, 3901 Beaubien Blvd, Detroit, MI 48201. E-mail: gnatara{at}med.wayne.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTPATIENTS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
- Erenberg A, Leff RD, Haack DG, Mosdell KW, Hicks GM, Wynne BA. Caffeine citrate for the treatment of apnea of prematurity: a double-blind placebo controlled study. Pharmacotherapy. 2000;20 :644 –652[CrossRef][Web of Science][Medline]
- Henderson-Smart DJ, Steer P. Methylxanthine treatment for apnea in preterm infants. Cochrane Database Syst Rev. 2001;(3):CD000140
- Schmidt B, Roberts RS, Davis P, et al. Caffeine for apnea of prematurity trial group: Caffeine therapy for apnea of prematurity.
N Engl J Med. 2006;354
:2112
–2221
[Abstract/Free Full Text] - Young TE, Mangum B. Caffeine citrate. In: Young TE, Mangum B, eds. Neofax. Raleigh, NC: Acorn Publishers, Inc; 2006:202–203
- Steer P, Flenady V, Shearman A, et al. High-dose caffeine citrate for extubation of preterm infants: a randomized controlled trial.
Arch Dis Child Fetal Neonatal Ed. 2004;89
:F499
–F503
[Abstract/Free Full Text] - Aranda JV, Cook CE, Gorman W, et al. Pharmacokinetic profile of caffeine in the premature newborn infant with apnea. J Pediatr. 1979;94 :663 –668[CrossRef][Web of Science][Medline]
- Aranda JV, Collinge JM, Zinman R, Watters G. Maturation of caffeine elimination in infancy.
Arch Dis Child. 1979;54
:946
–949
[Abstract/Free Full Text] - Le Guennec JC, Billon B, Pare C. Maturational changes of caffeine concentrations and disposition in infancy during maintenance therapy for apnea of prematurity: influence of gestational age, hepatic disease and breast feeding.
Pediatrics. 1985;76
:834
–840
[Abstract/Free Full Text] - Gorodischer R, Karplus M. Pharmacokinetic aspects of caffeine in premature infants with apnea. Eur J Clin Pharmacol. 1982;22 :47 –52[CrossRef][Web of Science][Medline]
- Pesce AJ, Rakhkin M, Kotagal U. Standards of laboratory practice: theophylline and caffeine monitoring.
Clin Chem. 1998;44
:1124
–1128
[Abstract/Free Full Text] - Falcao AC, Fernandez de Gatta MM, Delgado Iribarnegaray MF, et al. Population pharmacokinetics of caffeine in premature neonates. Eur J Clin Pharmacol. 1997;52 :211 –217[CrossRef][Web of Science][Medline]
- de Wildt SN, Kerkvliet KT, Wezenberg MG, et al. Use of saliva in therapeutic drug monitoring of caffeine in preterm infants. Ther Drug Monit. 2001;23 :250 –254[CrossRef][Web of Science][Medline]
- Lee TC, Charles B, Steer P, Flenady VJ, Debuse N. Population pharmacokinetics of intravenous caffeine in neonates with apnea of prematurity. Clin Pharmacol Ther. 1997;61 :628 –640[CrossRef][Web of Science][Medline]
- Lee HS, Khoo YM, Chirino-Barcelo Y, Tan KL, Ong D. Caffeine in apnoeic Asian neonates: a sparse data analysis. Br J Clin Pharmacol. 2002;54 :31 –37[CrossRef][Web of Science][Medline]
- Thomson AH, Kerr S, Wright S. Population pharmacokinetics of caffeine in neonates and young infants. Ther Drug Monit. 1996;18 :245 –253[CrossRef][Web of Science][Medline]
- Scanlon JEM, Chin KC, Morgan MEI, Durbin GM, Hale KA, Brown SS. Caffeine or theophylline for neonatal apnoea.
Arch Dis Child. 1992;67
:425
–458
[Abstract/Free Full Text] - Turmen T, Davis J, Aranda JV. Relationship of dose and plasma concentrations of caffeine and ventilation in neonatal apnea. Semin Perinatol. 1981;5326 –331:
PEDIATRICS (ISSN 1098-4275). ©2007 by the American Academy of Pediatrics
CiteULike 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  Caffeine for Apnea -- Why Check the Level? Journal Watch Pediatrics and Adolescent Medicine, June 6, 2007; 2007(606): 2 - 2. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
