Journal of Perinatology (2014), 1–5 © 2014 Nature America, Inc. All rights reserved 0743-8346/14 www.nature.com/jp
Serum caffeine concentrations and short-term outcomes in premature infants of ⩽ 29 weeks of gestation P Alur1, V Bollampalli1, T Bell2, N Hussain3 and J Liss1 OBJECTIVE: Caffeine is effective in the treatment of apnea of prematurity but it is not well known if the therapeutic concentration of the drug has an impact on other neonatal outcomes such as chronic lung disease (CLD). The aim of this study was to determine if there is an association between caffeine concentrations and the incidence of CLD in premature infants of ⩽ 29 weeks of gestation. STUDY DESIGN: A retrospective chart review of all the infants born ⩽ 29 weeks of gestation from 2007 to 2011, who survived until discharge or 36 weeks postmenstrual age, was conducted. Caffeine concentrations were obtained weekly on infants getting the drug. Average caffeine concentrations (ACCs) were determined for the duration of caffeine therapy and correlated with CLD, length of stay (LOS), oxygen at discharge (OD), duration of ventilation (DV) and total charges for hospitalization for each patient. RESULTS: Of the 222 eligible infants, 198 met the inclusion criteria. ACC for infants without CLD was 17.0 ± 3.8 µg ml − 1 compared with infants with CLD 14.3 ± 6.1 µg ml − 1 (P o0.001). Infants receiving high ACC (414.5 µg ml − 1) had lower incidence of patent ductus arteriosus, lesser number of days on ventilator and oxygen, lesser need for diuretics, lower incidence of CLD, were more likely to go home without supplemental OD and had lower LOS and lower total hospital charges (all differences were signiﬁcant Po 0.05) Multiple logistic regression modeling after adjusting for confounding variables indicated that higher caffeine concentrations were signiﬁcantly associated with decrease in CLD. Receiver operating curve analysis conﬁrmed a signiﬁcant predictive ability of caffeine concentration for CLD with a cutoff concentration of 14.5 µg ml − 1 (sensitivity of 42.6 and speciﬁcity of 86.8). The AUC (area under the curve) for the prediction of CLD was 0.632 (95% conﬁdence interval 0.56–0.69, P = 0.009). CONCLUSIONS: Caffeine concentrations 414.5 µg ml − 1 were strongly correlated with reduced CLD in infants born at ⩽ 29 weeks of gestation. Higher caffeine concentrations were associated with decreased total hospital charges, DV, OD and LOS. Additional randomized trials are needed to conﬁrm these ﬁndings, to identify ideal serum concentrations and determine possible long-term neurologic beneﬁts. Journal of Perinatology advance online publication, 18 December 2014; doi:10.1038/jp.2014.226
INTRODUCTION Caffeine, a nonspeciﬁc adenosine receptor antagonist, has been shown to be effective in treating apnea of prematurity.1,2 Recently, the ‘Caffeine for Apnea of Prematurity’ (CAP) trial showed reduced incidence of chronic lung disease (CLD) and improved neurological outcomes at 18–22 months with early use of caffeine.3,4 The CAP trial results prompted many to use caffeine for the prevention of CLD in addition to apnea management though it is not formally recommended for that indication. The serum caffeine concentrations in this group were not studied as caffeine is known to have a low toxicity potential.5 We are not aware of any studies that evaluated and correlated caffeine concentrations with neonatal outcomes. Recently, however, a small study indicated that higher caffeine concentrations (420 mg ml − 1) may be associated with higher concentrations of proinﬂammatory cytokines but no correlation with lung disease was demonstrated.6 Overall, scant literature exists on the effects of caffeine concentrations on health outcomes.7,8 An in vitro study showed that caffeine has a concentration-speciﬁc effect on cell cycle regulation, reactive oxygen species generation and cell survival in hyperoxic conditions with higher caffeine concentrations being deleterious.9 A large case series of fatalities due to caffeine intoxication in 1
adults found an average postmortem caffeine concentrations of 140.4 mg l − 1.10 It has been documented that caffeine concentrations vary based on gestational and post-conceptional age, but a correlation of this observation to neonatal outcomes has not been studied.11,12 The aim of this study was to determine if there is an association between serum caffeine concentrations and the incidence of CLD in premature infants born at ⩽ 29 weeks of gestation. METHODS A retrospective chart review was conducted at WellSpan Health neonatal intensive care unit of all the infants born at ⩽ 29 weeks of gestation from years 2007 to 2011. This study was reviewed and approved by the hospital institutional review board. All infants born at ⩽ 29 completed weeks of gestation who survived until discharge or 36 weeks postmenstrual age were included. Infants were excluded if they died or were transferred to another facility before reaching 36 weeks postmenstrual age. Postmenstrual age was calculated by adding the gestational age at birth (completed weeks) and chronological age (completed weeks). Demographic information including birth weight, gestational age, race, gender and multiplicity of gestation were collected. Prenatal information on chorioamnionitis, use of antenatal steroids and indication for premature delivery was obtained. Postnatal and hospital stay-related variables including weekly serum
Department of Pediatrics, York Hospital, WellSpan Health, York, PA, USA; 2Department of Pediatrics, Emig Research Center, York Hospital, Wellspan Health, York, PA, USA and Division of Neonatology, Department of Pediatrics, Connecticut Children’s Medical Center, University of Connecticut School of Medicine, Farmington, CT, USA. Correspondence: Dr P Alur, Department of Neonatology, York Hospital, WellSpan Health, 1001 St George Street, York, PA 17405, USA. E-mail: [email protected]
Received 16 May 2014; revised 27 October 2014; accepted 18 November 2014 3
Effect of serum caffeine in premature infants P Alur et al
2 caffeine concentrations, day of life when caffeine was started, duration of mechanical ventilation (DV) and information on short-term outcomes such as patent ductus arteriosus (PDA), bacterial sepsis, CLD, intraventricular hemorrhage (IVH), periventricular leukomalacia, necrotizing enterocolitis (NEC) and retinopathy of prematurity (ROP) were obtained. For the purpose of this study, CLD was deﬁned as persistent oxygen requirement at 36 weeks postmenstrual age.13 Nitric oxide is not prophylactically employed in our unit. PDA was deﬁned as any clinically suspected PDA conﬁrmed by echocardiogram. In our nursery, hemodynamically signiﬁcant PDA is treated with cyclooxygenase inhibitors and if unsuccessful, a persistent hemodynamically signiﬁcant PDA was surgically ligated. Our approach to management of PDA did not change during the study period. IVH was deﬁned and classiﬁed based on cranial ultrasound ﬁndings;14 periventricular leukomalacia was deﬁned as periventricular white matter densities identiﬁed by cranial ultrasound;15 NEC was deﬁned as Bell’s stage 2 or higher;16 and ROP was deﬁned as ⩾ Stage 1 retinopathy based on the ICROP classiﬁcation.17 ROP4Stage 2 was considered as severe ROP. IVH grades and ROP stages were determined by the highest level identiﬁed during hospital stay. Caffeine was prescribed at the discretion of the attending neonatologist shortly after birth while the infant was intubated and on mechanical ventilation or for management of apnea while on low ventilatory support or needing other noninvasive respiratory support such as CPAP. Even though caffeine is considered a relatively safe drug and routine monitoring of concentrations is not generally recommended11, our neonatal intensive care unit has continued to routinely monitor weekly caffeine concentrations to ensure that the concentrations do not exceed 30 µg ml − 1. Below that threshold, in ventilated infants, a caffeine bolus (5–15 mg kg − 1 caffeine citrate) was administered if there were persistent bradycardia episodes unrelated to mechanical problems and in nonventilated infants with symptomatic apnea and bradycardia, either caffeine boluses (5–15 mg kg − 1 caffeine citrate) were administered or maintenance doses (5–8 mg kg − 1 per day of caffeine citrate) were escalated to control the symptoms. Caffeine therapy and monitoring of concentrations were abruptly discontinued when the apneas were minimal and discharge was expected within 2 weeks. Caffeine concentrations that were obtained after discontinuation of the caffeine therapy were not included in the analysis. Caffeine concentrations were routinely obtained weekly for every infant on caffeine and more frequently in symptomatic patients. Average caffeine concentrations (ACC) were calculated by averaging all the caffeine concentrations that were obtained after caffeine was initiated up to the time the drug was discontinued.
Statistical analyses The relationship of CLD with ACC, length of stay, need for oxygen at discharge, duration of ventilation and total charges in dollars for hospitalization (obtained from hospital’s ﬁnancial data center) were evaluated by univariate analyses using t-tests, Χ2-tests and correlation coefﬁcients as appropriate. Variables of signiﬁcance in univariate analyses were subsequently analyzed using multivariable logistic regression analyses. The effectiveness of caffeine concentrations in predicting CLD was evaluated with a receiver operating curve (ROC) analysis of the area under the curve and was used to determine a cutoff point that could distinguish the groups with and without CLD. Subsequently, infant outcomes were compared based on the thresholds established by the cutoff points in the ROC analysis. Infants with ACC above and below the threshold concentration were compared using independent samples t-tests, Mann–Whitney U-test, Pearson Χ2- and Fisher’s Exact tests as appropriate.
RESULTS Of the 222 eligible infants in the database, 198 met inclusion criteria and 24 infants were eliminated because of death or transfer to another facility before they reached 36 weeks postmenstrual age. There were 18 deaths, among which nine expired within 48 h and did not receive caffeine, ﬁve deaths were due to NEC, two died due to severe sepsis, one died due to severe anomalies and one secondary to severe IVH. Demographic characteristics of infants The characteristics of the 198 infants were as follows: mean gestational age—27 ± 1.6 weeks (s.d.) and median gestational age Journal of Perinatology (2014), 1 – 5
Demographic characteristics of infants
Characteristic Gestational age (weeks) Birth weight (g) Gender Female Male
N = 198 27 ± 1.6 992.5 ± 298.5 100 (50.5%) 98 (48.5%)
Race African American Asian Hispanic Other White
38 (19.2%) 5 (2.5%) 22 (11.1%) 10 (5.1%) 123 (62.1%)
Ventilator use Infection CLD Oxygen at discharge Postnatal steroid use IVH PDA
128 (64.6%) 31 (15.7%) 47 (23.7%) 16 (8.1%) 28 (14.1%) 38 (19.2%) 92 (46.5%)
ROP Severe NEC Chorioamnionitis Antenatal steroid use Days on ventilator (days) Length of hospital stay (days) Total Charges (US $)
62 (31.3%) 7 (3.5%) 19 (9.6%) 172 (86.9%) 19 ± 28 77.8 ± 37.6 $294 537 ± 199 233
Abbreviations: CLD, chronic lung disease; IVH, intraventricular hemorrhage; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity.
27 weeks with an interquartile range of 2 weeks; mean birth weight was 992.5 ± 298.5 g (s.d.) and median birth weight of 980 g with an interquartile range 414 g. The male:female ratio was 98:100 and the ratio of whites to nonwhites was 123:75. See Table 1. Chronic lung disease Univariate analysis (see Table 2) of babies by CLD status showed that infants with CLD had lower gestational age and birth weight. It was notable that there was a signiﬁcantly higher ACC in infants without CLD (17.0 ± 3.8 µg ml − 1) compared with infants with CLD (14.3 ± 6.1 µg ml − 1; P o 0.001). Analysis of these variables showed that babies with CLD had signiﬁcantly more ventilator use, infections, postnatal steroid use, PDA, severe ROP, caffeine, days on ventilator, length of stay and total charges. Infants with CLD also had higher frequency of IVH, NEC, chorioamnionitis and antenatal steroid use. Multivariable analysis of outcomes Using variables signiﬁcant on univariate analyses, (gestational age at birth, birth weight, ventilator use, infection, postnatal steroid use, occurrences of IVH, treated PDA and high/low ACC concentration) multivariable logistic regression analysis was done for risk of CLD (Table 3). After controlling for these variables, low ACC was associated with a 4.5 times higher risk of CLD. Determination of optimal caffeine concentrations To determine the optimal caffeine concentrations that accounted for the most effect on CLD, a ROC curve was obtained. Analysis of the ROC showed a predictive ability of caffeine concentration for © 2014 Nature America, Inc.
Effect of serum caffeine in premature infants P Alur et al
3 Table 2.
Characteristics of infants by CLD
100 Gestational age (weeks) Birth weight (g) Gender Female Male Race African American Asian Hispanic Other White Ventilator use Infection Oxygen at discharge Postnatal steroid use IVH, severe PDA treated ROP, severe NEC Chorioamnionitis Antenatal steroid use
No CLD, N = 151
CLD, N = 47
27.4 ± 1.5
25.7 ± 1.5
1058.3 ± 279.7
780.9 ± 258.1
79 (51.7%) 73 (48.3%)
22 (46.8%) 25 (46.8%)
31 (20.5%) 3 (3.3%) 19 (12.6%) 7 (4.6%) 89 (58.9%) 84 (55.6%) 14 (9.3%) — 12 (7.9%) 7 (4.6%) 39 (25.8%) 8 (5.3%) 5 (3.3%) 14 (9.3%) 128 (84.8%)
7 (14.9%) — 3 (6.4%) 3 (6.4%) 34 (72.3%) 44 (93.6%) 17 (36.2%) 16 (34.0%) 16 (34.0%) 2 (4.3%) 32 (68.1%) 7 (14.9%) 2 (4.3%) 5 (10.6%) 44 (93.6%)
Caffeine concentration µg ml − 1 1–14.5 20 (13.2%) 20 (42.6%) ⩾ 14.6 131 (86.8%) 27 (57.4%) −1 ACC (µg ml ) 17 ± 3.8 14.3 ± 6.1 Days on ventilator 8.7 ± 13.2 52.3 ± 37.8 (days) Length of hospital stay 65.9 ± 20.4 116.0 ± 52.4 (days) Total Charges (US $) $225 493 ± 114 264 $516 357 ± 247 982
o0.001a o0.001 o0.001a o0.001 1.000a o0.001 0.052a 0.671a 0.780a 0.142a o0.001 o0.001 o0.001b
Multiple logistic regression analysis prediction CLD
95% CI for EXP(B)
Postnatal steroid use No steroid use Reference Steroid use 3.265
PDA Not treated Treated
Caffeine concentration (µg ml − 1) 14.6+ Reference 1–14.5 4.456
Birth weight Infection No infection Infection
Abbreviations: CI, conﬁdence interval; CLD, chronic lung disease; PDA, patent ductus arteriosus.
CLD at a cutoff concentration of 14.5 µg ml − 1 with sensitivity of 42.6 and speciﬁcity of 86.8 (Figure 1). The area under the curve for the prediction of CLD was 0.632 (95% conﬁdence interval 0.56– 0.69, P = 0.009). The study sample was then dichotomized using © 2014 Nature America, Inc.
60 40 100-Specificity
Figure 1. Receiver operating curve based on average caffeine concentration. At a cutoff point ⩽ 14.6, sensitivity = 42.6% and speciﬁcity = 86.8 (area under the curve = 0.632 and P = 0.009).
Abbreviations: ACC, average caffeine concentration; CLD, chronic lung disease; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity. aFishers exact test was used, all others used Pearson Χ2-test. bMann–Whitney U-test used, all others used Independent samples t-test.
this cutoff value into two groups, high ACC with concentrations 414.5 mg ml − 1 and low ACC with concentrations ⩽ 14.5 mg ml − 1. Other neonatal morbidities and caffeine concentrations Using the threshold value of 14.5 mg ml − 1 to dichotomize the study infants into high ACC and low ACC groups, univariate analysis showed that the two groups were similar in gestational age, birth weight, antenatal steroid use, race and gender characteristics and occurrence of RDS (Table 4). However, infants with high ACC had lower incidence of PDA, lesser number of days on ventilator and oxygen, lesser need for diuretics, lower incidence of CLD, were more likely to go home without supplemental oxygen at discharge and had lower total hospital charges (all differences were signiﬁcant P o 0.05) Nonrespiratory neonatal morbidities such as IVH, NEC and ROP were not signiﬁcantly different between the high and low ACC groups. DISCUSSION Caffeine is a safe drug in premature neonates with wide ranging beneﬁcial effects on the lung as well as the brain.18 Our study shows that there is a signiﬁcant effect from ACCs maintained during the infants’ initial hospital stay and predischarge outcomes in preterm infants born at ⩽ 29-week gestational age. Higher caffeine concentrations were associated with signiﬁcantly decreased days on ventilator, lower incidence of CLD, lesser days on oxygen, decreased length of hospital stay and lower total hospital charges in our study. Previous studies such as the CAP trial have shown that not only the use but the timing of use of caffeine can favorably affect short- and long-term outcomes in preterm infants.3,19,20 However, we believe that this is one of the ﬁrst reports of the relationship between caffeine concentrations and clinical outcomes in preterm infants. A small study of 26 preterm infants reported that proinﬂammatory cytokines were increased in infants with higher caffeine concentrations in the ﬁrst week of caffeine therapy.6 Moreover, proinﬂammatory cytokines were associated with prolonged Journal of Perinatology (2014), 1 – 5
Effect of serum caffeine in premature infants P Alur et al
4 Table 4.
Comparison of patient characteristics between low and high ACC Caffeine concentration µg ml − 1
Gestational age (weeks) Birth weight (g) Gender—female Race—White Ventilator use Infection Oxygen at 36 weeks Oxygen at discharge Postnatal steroid use IVH PDA ROP—any stage NEC Chorioamnionitis Antenatal steroid use Days on ventilator (days) Length of hospital stay (days) Total Charges (US $)
1–14.5, N = 40
14.6+, N = 158
26.63 ± 1.98 949.55 ± 371.04 20 (50%) 24 (60%) 31 (77.5%) 9 (22.5%) 20 (50%) 7 (17.5%) 10 (25%) 11 (27.5%) 26 (65%) 25 (62.5%) 3 (7.5%) 3 (7.5%) 36 (90%) 38.28 ± 42.05 89.38 ± 40.86 388 987 ± 264421
27.08 ± 1.54 1003.35 ± 277.55 80 (50.6%) 99 (62.7%) 97 (61.4%) 22 (13.9%) 27 (17.1%) 9 (5.7%) 18 (11.4%) 27 (17.1%) 66 (41.8%) 99 (62.7%) 4 (2.5%) 16 (10.1%) 136 (86.1%) 14.21 ± 21.46 74.83 ± 36.34 270 625 ± 172018
0.117 0.31 0.943 0.25 0.057 0.222a o0.001 0.023a 0.027 0.135 0.009 0.985 0.148a 0.614 0.770a o0.001b 0.029 0.001
Abbreviations: ACC, average caffeine concentration; CLD, chronic lung disease; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity. aFishers exact test was used, all others used Pearson Χ2-test. bMann–Whitney U -test used, all others used Independent samples t-test.
ventilator use and higher incidence of infection. But as the monitoring of caffeine concentrations and cytokine concentrations was limited to the ﬁrst week of caffeine therapy, the inﬂuence of caffeine concentrations on other short-term outcomes could not be evaluated. Data from a murine oleic acidinduced lung injury model suggest that there may be a dose (or concentration) at which caffeine is most beneﬁcial in decreasing lung injury and a higher or lower than optimal concentration may exacerbate injury through its effect on the adenosine A2a receptor.21 This study raises concern that not only high but even low concentrations of caffeine may have deleterious effects through lung inﬂammatory mechanisms and bolsters the argument for monitoring optimal caffeine concentrations. There is paucity of data even in adults on the concentration of caffeine affecting disease states with one study demonstrating that higher concentrations of caffeine have a beneﬁcial effect on dementia in adults.8 Human data on caffeine concentrations and outcomes in neonates are not available but inferences have been drawn from in vitro studies and study of laboratory animals. There have been some studies using caffeine dose as proxy for therapeutic effect22 but variability in caffeine metabolism based on age and maturity of the infant and pharmacogenetic differences make this assumption problematic.23–25. Caffeine pharmacokinetics in premature infants are based on a CYP1A2-mediated N7-demethylation process with changes in metabolism related to birth weight, gestational age, postmenstrual age and gender.11,24 Caffeine elimination and plasma half-life remains fairly constant in ﬁrst 2 weeks of life and then caffeine clearance starts to increase.5,26Adult rates of plasma clearance of caffeine are achieved by 3–4.5 months of age.12 Therefore, in infants who are on caffeine beyond 2 weeks of age, serum concentrations may vary depending not only on the infants’ pharmacogenetic proﬁle but also the chronological age, even if the dose per kg body weight is kept constant. It is well known that the timing of starting caffeine can inﬂuence the development of CLD with early (⩽3 days) initiation associated with less lung disease.19,27 Our ﬁndings would suggest that not only the timing but also the therapeutic concentration of caffeine may be important in the prevention of CLD. Journal of Perinatology (2014), 1 – 5
The mechanism of caffeine’s effect in decreasing CLD is unknown but there are animal studies that suggest putative pathways through the effect of caffeine on adenosine receptor blockade on inﬂammatory cells. Caffeine decreases the release of tumor necrosis factor-α and hypoxia-associated vascular leak through blockade of the A2B-adenosine receptor.28,29 Caffeine administration via inhibition of the A2a adenosine receptor may result in decreased hyperoxia induced CXC cytokines, decreased neutrophil migration and prevention of macrophage release of proinﬂammatory cytokines responsible for capillary leak.6,28–30 Caffeine not only affects lung cellular function but also has a positive effect on respiratory muscles.31,32 It has been shown in premature infants on ventilatory support that caffeine improves the force of contraction of diaphragm and thus impacts infant generated inspiratory pressures.32 This effect may facilitate early weaning on the ventilator and with concomitant decrease in lung injury. These and other mechanisms may be key to understanding the reasons for decreased incidence of CLD with caffeine use.33 Our study has certain limitations being a retrospective study. We have not followed our study infants to evaluate long-term neurological outcomes. We do not know the pharmacogenomic proﬁle of the study population to understand if the infants with higher caffeine concentrations were slow metabolizers or if they received more caffeine. As there were no differences in gestational age at birth, birth weight or other demographic variables at the time of initiation of caffeine therapy, we believe that some of the study biases were mitigated. It is also reassuring that similar to the ﬁndings in larger trials such as the CAP trial, we also found that infants on caffeine were extubated early and in addition we found that the infants, who were extubated early also had higher caffeine concentrations. Also similar to the CAP trial, where infants in the caffeine group received treatment for PDA less frequently than those in the placebo group, we observed a lower incidence of PDA in our group with higher caffeine concentration. Although the administration of caffeine to preterm sheep has no direct effect on ductal closure,31,34 it is tempting to speculate that there may be a dose-dependent response and higher caffeine concentrations may facilitate ductal closure through an indirect effect of decreased inﬂammatory mediators.35 It is also possible © 2014 Nature America, Inc.
Effect of serum caffeine in premature infants P Alur et al
5 that a higher incidence of PDA in the lower caffeine concentration group may have primarily contributed to increased rate of CLD in that group. Though our study has shown a strong association between caffeine concentrations and short-term outcomes, a large prospective randomized controlled study with long-term neurodevelopmental follow-up is needed to further clarify the relationship between caffeine concentrations and outcomes. There may also be a need for evaluating mathematical models for predicting caffeine concentrations that could potentially be used for longterm monitoring in individual infants.36 CONCLUSIONS Serum caffeine concentrations 414.5 µg ml − 1 are strongly correlated with reduced CLD in very low birth weight infants ⩽ 29 weeks of gestation. Higher caffeine concentrations were associated with decreased length of stay, total hospitalization charges, duration of ventilation and oxygen at discharge. Additional randomized trials are needed to conﬁrm our ﬁndings, to identify ideal serum concentrations and to determine possible long-term beneﬁts. CONFLICT OF INTEREST The authors declare no conﬂict of interest.
ACKNOWLEDGEMENTS We thank Ms Gaye Ludwig and Diane Myers for their proof reading and formatting help.
AUTHOR CONTRIBUTIONS PA was responsible for the conception, design, acquisition, analysis and interpretation of the data, and also for the drafting and ﬁnal revision of the manuscript. VB was responsible for the acquisition of data and interpretation. TB and NH performed the statistical analysis and interpreted the data. JL was responsible for the design, analysis and interpretation of data. VB and TB were responsible for the ﬁnal revision of the manuscript. NH was responsible for review and ﬁnal revision of the manuscript. JL was responsible for the drafting and ﬁnal revision of the manuscript. All authors have provided the ﬁnal approval of the manuscript.
REFERENCES 1 Jacobson KA. Introduction to adenosine receptors as therapeutic targets. Handb Exp Pharmacol 2009; (193): 1–24; PMCID: PMC3415694. 2 Rivkees SA, Zhao Z, Porter G, Turner C. Inﬂuences of adenosine on the fetusand newborn. Mol Genet Metab 2001; 74(1-2): 160–171. 3 Schmidt B, Roberts RS, Davis P, Doyle LW, Barrington KJ, Ohlsson A et al. Caffeine therapy for apnea of prematurity. N Engl J Med 2006; 354(20): 2112–2121. 4 Schmidt B, Roberts RS, Davis P, Doyle LW, Barrington KJ, Ohlsson A et al. Longterm effects of caffeine therapy for apnea of prematurity. N Engl J Med 2007; 357(19): 1893–1902. 5 Natarajan G, Botica ML, Thomas R, Aranda JV. Therapeutic drug monitoring for caffeine in preterm neonates: an unnecessary exercise? Pediatrics 2007; 119(5): 936–940. 6 Chavez Valdez R, Ahlawat R, Wills-Karp M, Nathan A, Ezell T, Gauda EB. Correlation between serum caffeine levels and changes in cytokine proﬁle in a cohort of preterm infants. J Pediatr 2011; 158(1): 57–64, 64 e51. 7 Du Y, Melchert HU, Knopf H, Braemer-Hauth M, Pabel E. Association of serum caffeine concentrations with serum glucose levels in caffeine-drug users and nonusers - results of German National Health Surveys. Diabetes Obes Metab 2007; 9(5): 756–758. 8 Cao C, Loewenstein DA, Lin X, Zhang C, Wang L, Duara R et al. High Blood caffeine levels in MCI linked to lack of progression to dementia. J Alzheimers Dis 2012; 30(3): 559–572.
© 2014 Nature America, Inc.
9 Tiwari KK, Chu C, Couroucli X, Moorthy B, Lingappan K. Differential concentrationspeciﬁc effects of caffeine on cell viability, oxidative stress, and cell cycle in pulmonary oxygen toxicity in vitro. Biochem Biophys Res Commun 2014; 450(4): 1345–1350. 10 Banerjee P, Ali Z, Levine B, Fowler DR. Fatal caffeine intoxication: a series of eight cases from 1999 to 2009. J Forensic Sci 2014; 59(3): 865–868. 11 al-Alaiyan S, al-Rawithi S, Raines D, Yusuf A, Legayada E, Shoukri MM et al. Caffeine metabolism in premature infants. J Clin Pharmacol 2001; 41(6): 620–627. 12 Aranda JV, Collinge JM, Zinman R, Watters G. Maturation of caffeine elimination in infancy. Arch Dis Child 1979; 54(12): 946–949. 13 Bancalari E, Abdenour GE, Feller R, Gannon J. Bronchopulmonary dysplasia: clinical presentation. J Pediatr 1979; 95(5 Pt 2): 819–823. 14 Papile LA, Burstein J, Burstein R, Kofﬂer H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr 1978; 92(4): 529–534. 15 DeReuck J, Chattha AS, Richardson EP Jr. Pathogenesis and evolution of periventricular leukomalacia in infancy. Arch Neurol 1972; 27(3): 229–236. 16 Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg 1978; 187(1): 1–7. 17 Patz A. New international classiﬁcation of retinopathy of prematurity. Pediatrics 1984; 74(8): 160–161. 18 Aranda JV, Beharry K, Valencia GB, Natarajan G, Davis J. Caffeine impact on neonatal morbidities. J Matern Fetal Neonatal Med 2010; 23(Suppl 3): 20–23. 19 Dobson NR, Patel RM, Smith PB, Kuehn DR, Clark J, Vyas-Read S et al. Trends in caffeine use and association between clinical outcomes and timing of therapy in very low birth weight infants. J Pediatr 2014; 164(5): 992–998. 20 Schmidt B, Davis PG, Roberts RS. Timing of caffeine therapy in very low birth weight infants. J Pediatr 2014; 164(5): 957–958. 21 Li J, Li G, Hu JL, Fu XH, Zeng YJ, Zhou YG et al. Chronic or high dose acute caffeine treatment protects mice against oleic acid-induced acute lung injury via an adenosine A2A receptor-independent mechanism. Eur J Pharmacol 2011; 654(3): 295–303. 22 Miller JL, Nandyal RR, Anderson MP, Escobedo MB. Relationship of caffeine dosing with serum alkaline phosphatase levels in extremely low-birth-weight infants. J Pediatr Pharmacol Ther 2012; 17(1): 58–66. 23 Ginsberg G, Hattis D, Sonawane B, Russ A, Banati P, Kozlak M et al. PBPK modelling of caffeine and theophylline in neonates and adults: implications for assessing children's risks from environmental agents. J Toxicol Environ Health Part A 2004; 67(4): 297–329. 24 Cazeneuve C, Pons G, Rey E, Treluyer JM, Cresteil T, Thiroux G et al. Biotransformation of caffeine in human liver microsomes from foetuses, neonates, infants and adults. Br J Clin Pharmacol 1994; 37(5): 405–412. 25 Thorn CF, Aklillu E, McDonagh EM, Klein TE, Altman RB. PharmGKB summary: caffeine pathway. Pharmacogenet Genomics 2012; 5(5): 389–395. 26 Leon AE, Michienzi K, Ma CX, Hutchison AA. Serum caffeine concentrations in preterm neonates. Am J Perinatol 2007; 24(1): 39–47. 27 Patel RM, Leong T, Carlton DP, Vyas-Read S. Early caffeine therapy and clinical outcomes in extremely preterm infants. J Perinatol 2013; 33(2): 134–140. 28 Chavez-Valdez R, Wills-Karp M, Ahlawat R, Cristofalo EA, Nathan A, Gauda EB. Caffeine modulates TNF-alpha production by cord blood monocytes: the role of adenosine receptors. Pediatr Res 2009; 65(2): 203–208. 29 Eckle T, Faigle M, Grenz A, Laucher S, Thompson LF, Eltzschig HK. A2B adenosine receptor dampens hypoxia-induced vascular leak. Blood 2008; 111(4): 2024–2035. 30 Sitkovsky M, Lukashev D. Regulation of immune cells by local-tissue oxygen tension: HIF1 alpha and adenosine receptors. Nat Rev Immunol 2005; 5(9): 712–721. 31 Crossley KJ, Allison BJ, Polglase G, Morley CJ, Harding R, Davis PG et al. Effects of caffeine on renal and pulmonary function in preterm newborn lambs. Pediatr Res 2012; 72: 19–25. 32 Kassim Z, Greenough A, Rafferty GF. Effect of caffeine on respiratory muscle strength and lung function in prematurely born, ventilated infants. Eur J Pediatr 2009; 168(12): 1491–1495. 33 Rivkees SA, Wendler CC. Adverse and protective inﬂuences of adenosine on the newborn and embryo: implications for preterm white matter injury and embryo protection. Pediatr Res 2011; 69(4): 271–278. 34 Clyman RI, Roman C. The effects of caffeine on the preterm sheep ductus arteriosus. Pediatr Res 2007; 62(2): 167–169. 35 Geraets L, Moonen HJ, Wouters EF, Bast A, Hageman GJ. Caffeine metabolites are inhibitors of the nuclear enzyme poly(ADP-ribose)polymerase-1 at physiological concentrations. Biochem Pharmacol 2006; 72(7): 902–910. 36 Micallef S, Amzal B, Bach V, Chardon K, Tourneux P, Bois FY. Sequential updating of a new dynamic pharmacokinetic model for caffeine in premature neonates. Clin Pharmacokinet 2007; 46(1): 59–74.
Journal of Perinatology (2014), 1 – 5