Awaken Study

Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Research

JAMA Pediatrics | Original Investigation

Association Between Early Caffeine Citrate Administration


and Risk of Acute Kidney Injury in Preterm Neonates
Results From the AWAKEN Study
Matthew W. Harer, MD; David J. Askenazi, MD, MSPH; Louis J. Boohaker, MPH; J. Bryan Carmody, MD, MPH;
Russell L. Griffin, PhD; Ronnie Guillet, MD, PhD; David T. Selewski, MD; Jonathan R. Swanson, MD, MSc;
Jennifer R. Charlton, MD, MSc; for the Neonatal Kidney Collaborative (NKC)

Supplemental content
IMPORTANCE Acute kidney injury (AKI) occurs commonly in preterm neonates and is
associated with increased morbidity and mortality.

OBJECTIVES To examine the association between caffeine citrate administration and AKI in
preterm neonates in the first 7 days after birth and to test the hypothesis that caffeine
administration would be associated with reduced incidence and severity of AKI.

DESIGN, SETTING, AND PARTICIPANTS This study was a secondary analysis of the Assessment
of Worldwide Acute Kidney Injury Epidemiology in Neonates (AWAKEN) study, a
retrospective observational cohort that enrolled neonates born from January 1 to March 31,
2014. The dates of analysis were October 2016 to December 2017. The setting was an
international, multicenter cohort study of neonates admitted to 24 participating level III or IV
neonatal intensive care units. Participants met the original inclusion and exclusion criteria of
the AWAKEN study. Additional exclusion criteria for this study included participants greater
than or equal to 33 weeks’ gestation at birth, admission after age 7 days, use of theophylline
in the neonatal intensive care unit, or lack of data to define AKI. There were 675 preterm
neonates available for analysis.

EXPOSURE Administration of caffeine in the first 7 days after birth.

MAIN OUTCOMES AND MEASURES The primary outcome was the incidence of AKI (based on
the modified neonatal Kidney Disease: Improving Global Outcomes [KDIGO] definition) in the
first 7 days after birth. The hypothesis that caffeine administration would be associated with
reduced AKI incidence was formulated before data analysis.

RESULTS The study cohort (n = 675) was 55.4% (n = 374) male, with a mean (SD) gestational
age of 28.9 (2.8) weeks and a mean (SD) birth weight of 1285 (477) g. Acute kidney injury
occurred in 122 neonates (18.1%) in the first 7 days after birth. Acute kidney injury occurred
less frequently among neonates who received caffeine than among those who did not (50 of
447 [11.2%] vs 72 of 228 [31.6%], P < .01). After multivariable adjustment, administration of
caffeine remained associated with reduced odds of developing AKI (adjusted odds ratio,
0.20; 95% CI, 0.11-0.34), indicating that for every 4.3 neonates exposed to caffeine one case
of AKI was prevented. Among neonates with early AKI, those receiving caffeine were less
likely to develop stage 2 or 3 AKI (adjusted odds ratio, 0.20; 95% CI, 0.12-0.34).

Author Affiliations: Author


CONCLUSIONS AND RELEVANCE Caffeine administration in preterm neonates is associated affiliations are listed at the end of this
with reduced incidence and severity of AKI. Further studies should focus on the timing and article.
dosage of caffeine to optimize the prevention of AKI. Group Information: The members of
the Neonatal Kidney Collaborative
(NKC) are listed at the end of this
article.
Corresponding Author: Jennifer R.
Charlton, MD, MSc, Division of
Nephrology, Department of
Pediatrics, University of Virginia, PO
JAMA Pediatr. 2018;172(6):e180322. doi:10.1001/jamapediatrics.2018.0322 Box 800386, Charlottesville, VA
Published online April 2, 2018. Corrected on May 7, 2018. 22908 ([email protected]).

(Reprinted) 1/9
© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Research Original Investigation Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates

I
n recent decades, advances in the care of preterm neo-
nates have dramatically lowered mortality rates.1-3 As sur- Key Points
vival has improved, clinicians have been increasingly fo-
Question Is early caffeine citrate administration associated with
cused on decreasing the short-term and long-term sequelae reduced incidence or severity of acute kidney injury in preterm
of prematurity.4 Acute kidney injury (AKI) occurs frequently neonates?
in preterm neonates and is associated with increased morbid-
Findings In this multicenter cohort study of 675 preterm
ity and mortality.5-7 In 2014, the Neonatal Kidney Collabora-
neonates, those who received caffeine in the first 7 days after birth
tive (NKC) initiated an international, multicenter (24 centers developed acute kidney injury less frequently than neonates who
in 4 countries) study designed to evaluate the incidence, risk did not (11.2% vs 31.6%).
factors, and outcomes associated with neonatal AKI called the
Meaning Caffeine administration in the first 7 days after birth may
Assessment of Worldwide Acute Kidney Injury Epidemiology
be associated with less frequent acute kidney injury in preterm
in Neonates (AWAKEN) study.8 This study demonstrated that neonates; further studies on dosage, the timing of administration,
30% of neonates admitted to a neonatal intensive care unit and long-term outcomes are needed.
(NICU) developed AKI and that those with AKI had 4.8 times
higher adjusted odds of mortality compared with neonates
without AKI.9 However, there are few specific strategies to pre- Setting and Participants
vent or ameliorate AKI beyond supportive measures, such as The AWAKEN study was a retrospective observational cohort
avoidance of nephrotoxins and optimization of blood pres- investigation that enrolled neonates born from January 1 to
sure and fluid balance.10-12 Therefore, identifying therapies to March 31, 2014. The dates of analysis were October 2016 to De-
prevent or reduce the severity of AKI are of paramount impor- cember 2017. The setting was an international, multicenter co-
tance. hort study of neonates admitted to 24 participating level III or
Methylxanthines are adenosine antagonists that act via IV NICUs. Data were collected from the time of NICU admis-
A1 and A2A receptors present in the brain, heart, blood ves- sion until discharge, transfer, death, or 120 days after birth,
sels, respirator y system, gastrointestinal tract, and whichever came first. Inclusion criteria for the AWAKEN study
kidneys.13 The results of clinical trials using theophylline or were admission to the NICU and administration of intrave-
aminophylline have suggested that methylxanthines pre- nous fluids for at least the first 48 hours after admission. The
vent AKI or improve renal function in special populations of original exclusion criteria for the AWAKEN study were indi-
high-risk neonates and infants, including those with perina- viduals with admission 14 days or longer after birth, those with
tal hypoxia/ischemia or prematurity and undergoing cardiac congenital heart disease requiring repair less than 7 days af-
surgery. 14-20 However, these medications are no longer ter birth, infants with lethal chromosomal anomaly, and neo-
widely used in the general neonatal population: a 2014 nates with death within 48 hours after birth. Specific to the
study21 found that caffeine citrate (another methylxanthine) present study, the following additional exclusion criteria ap-
accounted for 96% of methylxanthine use in 2010. Whether plied: gestational age at least 33 weeks, admission more than
exposure to caffeine might be associated with decreased 7 days after birth, admission to a NICU that used theophylline
incidence of AKI was explored by a single-center retrospec- rather than caffeine, less than 1 day of measured urine output
tive study22 of very low-birth-weight (VLBW) neonates (birth (UOP) on days 2 to 7 after birth, and fewer than 2 serum cre-
weight, <1500 g). Although the results of this study sug- atinine (sCr) measurements (Figure). Analysis was limited to
gested that early caffeine administration might prevent AKI, participants younger than 33 weeks’ gestation because few neo-
to date there has been no comprehensive multicenter evalu- nates with gestational age at least 33 weeks received caffeine
ation of caffeine administration and its association with AKI (owing to the low incidence of apnea of prematurity in this
in preterm neonates. group).
To address this knowledge gap, we performed a second-
ary analysis of preterm neonates born at less than 33 weeks’ Variables of Outcomes, Exposures, Confounders,
gestation enrolled in the AWAKEN study. The aim of this analy- and Effect Modifiers
sis was to examine the association of caffeine administration The primary outcome of interest was early AKI occurring in the
with AKI in the first 7 days after birth. We hypothesized that first 7 days after birth. Neonatal AKI was defined by the modi-
neonates given caffeine would have reduced incidence and se- fied neonatal Kidney Disease: Improving Global Outcomes
verity of AKI. (KDIGO) definition (Table 1).23 This is the consensus defini-
tion for neonatal AKI based on the recommendation of a re-
cent National Institutes of Health/National Institute of Diabe-
tes and Digestive and Kidney Diseases workshop.24 Secondary
Methods outcomes of interest were the severity of AKI, defined by modi-
A comprehensive description of the NKC and the methods used fied neonatal KDIGO staging, and the incidence of AKI within
to collect the data for the AWAKEN study8 has been pub- the entire data collection period.
lished. The University of Alabama at Birmingham Institu- The exposure of interest was administration of caffeine be-
tional Review Board approved this collaborative study, and fore AKI. If administration of caffeine occurred in the first 7
each participating center received approval from their respec- days after birth but after AKI occurred, these neonates were
tive institutional review boards. considered not to have received caffeine.

2/9 JAMA Pediatrics June 2018 Volume 172, Number 6 (Reprinted) jamapediatrics.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates Original Investigation Research

Figure. Study Flow Diagram Table 1. Definition of Acute Kidney Injurya


UOP During the
4273 Assessed for eligibility Stage Serum Creatinine (sCr) Past 24 h
0 No change in sCr or >1 mL/kg/h
rise <0.3 mg/dL
2111 Excludeda
1 sCr rise ≥0.3 mg/dL within 48 h or Between >0.5
1793 IV fluids <48 h
sCr rise between ≥1.5 and 1.9 times reference sCr and ≤1 mL/kg/h
278 Age >14 d
within 7 d
157 Other
68 Congenital heart disease 2 sCr rise between ≥2 and 2.9 times reference sCr Between >0.3
and surgery within 7 d and ≤0.5
mL/kg/h
3 sCr rise ≥3 times reference sCr or ≤0.3 mL/kg/h
2162 Enrolled in AWAKEN sCr ≥2.5 mg/dLb or
receipt of dialysis

1487 Excludeda Abbreviation: UOP, urine output.


1449 GA ≥33 wk SI conversion factor: To convert serum creatinine value to micromoles per liter,
140 <2 SCr measurements multiply by 88.4.
or no UOP data a
42 Age >7 d Reference sCr is the lowest prior sCr value.
14 No caffeine use at b
This value is lower than the original Kidney Disease: Improving Global
study center Outcomes (KDIGO) definition because an sCr value of 2.5 mg/dL in neonates
suggests an estimated glomerular filtration rate less than 10 mL/min/1.73 m2.
675 Included in analysis
204 GA 22-27 6/7 wk
471 GA 28-32 6/7 wk nonnormally distributed variables, the medians (interquar-
tile ranges) were reported, and groups were compared using
the Mann-Whitney test.
447 Received caffeine 228 Did not receive caffeine To evaluate the association between caffeine administra-
tion and AKI—both early (ie, in the first 7 days after birth) and
within 120 days—a generalized linear mixed model using a logit
199 No caffeine in 29 Received caffeine link and binary distribution and a random intercept by study
first 7 d but after AKI site (to account for possible clustering by study site) was used
to calculate unadjusted odds ratios (ORs) and associated 95%
Acute kidney injury (AKI) was defined as occurring in the first 7 days after birth, CIs. Maternal, neonatal, and clinical characteristics with P < .20
and caffeine citrate administration was defined as occurring before an AKI in bivariate analysis were considered candidate variables for
event. AWAKEN indicates Assessment of Worldwide Acute Kidney Injury
Epidemiology in Neonates; GA, gestational age; IV, intravenous; SCr, serum the final adjusted model. Using a backward selection pro-
creatinine; and UOP, urine output. cess, multivariable models were generated that included only
a
Because exclusion criteria are not mutually exclusive, some potential variables with P < .05 to create the most parsimonious model
participants could have been excluded for multiple reasons and are counted in to estimate adjusted ORs. To assess possible effect modifica-
each exclusion category.
tion by gestational age, an interaction between gestational age
and caffeine administration was entered into the final ad-
justed model to examine whether the observed associations
To characterize the cohort and to assess potential con- between caffeine administration and AKI differed statisti-
founders, both maternal and neonatal data were collected, in- cally by gestational age. In secondary analyses, the incidence
cluding the following: maternal age, mode of delivery, out- of AKI was compared between caffeine groups using a gener-
born delivery, pregnancy medications, gestational age, birth alized linear mixed model with a logit link and binomial dis-
weight, small for gestational age status, sex, resuscitation ef- tribution, and AKI staging was compared between caffeine
forts, Apgar scores, admission diagnoses, ventilation, patent groups using a generalized linear mixed model with a cumu-
ductus arteriosus therapies, initial sCr values, type of center, lative logit link and multinomial distribution for both early AKI
and neonatal medications in the NICU. Similar to the classifi- and AKI within 120 days. All regression models were ad-
cation of caffeine administration, exposure to other neonatal justed for gestational age, sepsis evaluation, type of center, and
medications, including antimicrobials, diuretics, and vaso- antibiotic use.
pressors, was defined as administration of the medication be- The number needed to be exposed (NNE), corresponding
fore AKI. The updated Clinical Risk Index for Babies (CRIB II) to the number of patients who would need to be treated with
score was used to assess the severity of illness on admission.25 caffeine to prevent one case of AKI after adjustment for con-
founding variables, was calculated from the adjusted OR
Statistical Analysis obtained in logistic regression using statistical methods
Comparisons were made using the χ2 test for categorical vari- described previously.26 For all analyses, 2-sided P < .05 was
ables and the Fisher exact test for continuous variables. All con- considered statistically significant. All statistical analyses
tinuous variables were tested for normality using the Shapiro- were performed centrally at the University of Alabama at Bir-
Wilk test. For normally distributed continuous variables, the mingham using a software program (SAS, version 9.4; SAS
means (SDs) were reported and analyzed using the t test. For Institute Inc).

jamapediatrics.com (Reprinted) JAMA Pediatrics June 2018 Volume 172, Number 6 3/9

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Research Original Investigation Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates

caffeine with early AKI as defined by decreased UOP without


Results a rise in creatinine (adjusted OR, 0.40; 95% CI, 0.15-1.06)
(Table 4). Among neonates who developed early AKI, those
Participants receiving caffeine had an 80% decrease in the odds of stage
The AWAKEN study enrolled 2162 participants. After the ap- 2 or 3 AKI (adjusted OR, 0.20; 95% CI, 0.12-0.34). Stage 3
plication of the additional exclusion criteria for this study, 675 AKI occurred approximately 8 times more frequently among
preterm neonates were eligible for our secondary analysis neonates who did not receive caffeine (1.3% [6 of 447] vs
(Figure). Of the 675 participants, 447 (66.2%) received at least 10.5% [24 of 228], P < .001).
one dose of caffeine in the first 7 days after birth and before Similar associations were observed for AKI that occurred
AKI. Of the 228 neonates considered to have received no caf- after the initial 7 days (before NICU discharge, transfer, death,
feine, 199 received no caffeine at all in the first 7 days after birth, or age 120 days). The incidence of AKI through the entire data
while 29 received caffeine after AKI. For neonates included in collection period was 27.6% (186 of 675). Among patients re-
the study, 36.9% (249 of 675) received their first dose of caf- ceiving caffeine, the odds of developing AKI were reduced by
feine on day 1 after birth, and 53.8% (363 of 675) received their 44% during the data collection period (crude OR, 0.56; 95%
first dose within the first 2 days after birth. Bivariate analyses CI, 0.38-0.84), an association that persisted after multivari-
were conducted to compare the severity of illness between able adjustment (adjusted OR, 0.27; 95% CI, 0.16-0.47). The
those who received caffeine on day 1 or day 1 and day 2 with NNE to prevent one case of AKI during the participants’ hos-
those who did not receive caffeine (eTable 1 and eTable 2 in pitalization (up to 120 days) was 4.4.
the Supplement). Neonates who did not receive caffeine had
older gestational age, higher birth weight, greater Apgar scores Differences in Outcomes by Groups Stratified
at 5 minutes, and lower CRIB II scores, suggesting that these by Gestational Age
neonates were not sicker than those who received caffeine. As summarized in eTable 4 and eTable 5 in the Supplement,
exposure to caffeine remained associated with decreased fre-
Comparison of Infants by Caffeine Administration quency of AKI in both gestational age strata. In particular, an
Compared with neonates who did not receive caffeine, neo- 87% decreased odds of early AKI was observed among ex-
nates who received caffeine had younger gestational age and tremely preterm neonates (adjusted OR, 0.13; 95% CI, 0.06-
lower birth weight (Table 2). In addition, these neonates were 0.31; NNE, 2.2), and a 73% decreased odds of early AKI was ob-
more likely to be intubated in the delivery room and had lower served for very preterm neonates (adjusted OR, 0.27; 95% CI,
median Apgar scores at 1 minute and 5 minutes. On admis- 0.13-0.56; NNE, 8.1). There was no statistical interaction be-
sion, those receiving caffeine were more likely to have a diag- tween gestational age and caffeine administration, suggest-
nosis of respiratory distress, undergo a sepsis evaluation, and ing that there was no difference in the association of caffeine
receive either invasive or noninvasive respiratory support. Neo- administration with early AKI by gestational age category. Simi-
nates receiving caffeine were more likely to be given a neph- lar, although slightly attenuated, associations were observed
rotoxic antimicrobial agent and a nonsteroidal anti- for all AKI, with an observed 76% decreased odds of any AKI
inflammatory drug in the NICU. among extremely preterm neonates (adjusted OR, 0.24; 95%
CI, 0.10-0.58; NNE, 3.1) and a 68% decreased odds among very
Association of Caffeine Administration preterm neonates (adjusted OR, 0.32; 95% CI, 0.16-0.62; NNE,
With Early AKI (Primary Outcome) 8.0). As with early AKI, no statistical interaction was ob-
Acute kidney injury occurred in the first 7 days after birth in served between gestational age and caffeine.
18.1% (122 of 675) of neonates. Demographics of participants
with early AKI compared with those without AKI are listed in
eTable 3 in the Supplement. Neonates who received caffeine
were less likely to develop early AKI compared with those who
Discussion
did not (11.2% [50 of 447] vs 31.6% [72 of 228], P < .01; unad- In this secondary analysis of the international, multicenter
justed OR, 0.28; 95% CI, 0.18-0.44) (Table 3). The eFigure in AWAKEN study cohort, preterm neonates who were exposed
the Supplement shows a Kaplan-Meier AKI-free survival curve to caffeine were less likely to develop AKI. This association oc-
for neonates in the first 7 days after birth. After multivariable curred despite the presence of traditional risk factors for AKI
analysis (adjusting for gestational age, sepsis evaluation, type among neonates receiving caffeine (including younger gesta-
of center, and nephrotoxic antimicrobial use), neonates who tional age, lower birth weight, and higher illness severity) and
received caffeine remained less likely to have AKI than those persisted even after multivariable adjustment. Neonates ex-
who did not receive caffeine (adjusted OR, 0.20; 95% CI, 0.11- posed to caffeine had an almost 3-fold reduction in the inci-
0.34; P < .001). The NNE to prevent a single case of AKI was 4.3. dence of AKI and an 8-fold reduction in stage 3 AKI in the first
week after birth, supporting the hypothesis that early expo-
Association of Caffeine Administration sure to caffeine is associated with a reduction in both AKI fre-
With Secondary AKI Outcomes quency and severity in preterm infants.
Neonates who received caffeine were less likely to develop The field of neonatal AKI has undergone a renaissance, with
early AKI based on sCr values alone (adjusted OR, 0.20; 95% numerous studies5,7,9,27-31 demonstrating that AKI in criti-
CI, 0.11-0.37); however, there was no association of use of cally ill neonates is common and consistently associated with

4/9 JAMA Pediatrics June 2018 Volume 172, Number 6 (Reprinted) jamapediatrics.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates Original Investigation Research

Table 2. Characteristics of Patients Stratified by Caffeine Citrate Administration


Whole Cohort Caffeine No Caffeine
Variable (N = 675) (n = 447) (n = 228) P Valuea
Perinatal Factors
HTN gestational disease, No. (%)b 177 (26.2) 113 (25.3) 64 (28.1) .43
Maternal diabetes, No. (%) 79 (11.7) 53 (11.9) 26 (11.4) .86
Maternal age, mean (SD), y 28.9 (6.3) 28.8 (6.5) 29.0 (6.0) .77
Assisted conception, No. (%) 92 (13.6) 58 (13.0) 34 (14.9) .49
Cesarean delivery, No. (%) 420 (62.2) 277 (62.0) 143 (62.7) .14
Multiple birth, No. (%) 216 (32.0) 153 (34.2) 63 (27.6) .08
Outborn delivery, No. (%) 181 (26.8) 121 (27.1) 60 (26.3) .83
Pregnancy medications
Betamethasone, No. (%) 494 (73.2) 337 (75.4) 157 (68.9) .07
NSAID, No. (%) 45 (6.7) 34 (7.6) 11 (4.8) .17
Antihypertensive, No. (%)c 122 (18.1) 89 (19.9) 33 (14.5) .08
Neonatal Factors
Gestational age, mean (SD), wk 28.9 (2.8) 28.5 (2.6) 29.6 (2.9) <.001
Birth weight, mean (SD), g 1285 (477.0) 1207 (413.4) 1436 (552.0) <.001
SGA status, No. (%) 97 (14.4) 61 (13.6) 36 (15.8) .47
Male, No. (%) 374 (55.4) 249 (55.7) 125 (54.8) .75
Ethnicity, No. (%)
Hispanic 87 (12.9) 61 (13.6) 26 (11.4)
Non-Hispanic 502 (74.4) 331 (74.0) 171 (75.0) .67
Unknown 86 (12.7) 55 (12.3) 31 (13.6)
Race, No. (%) Abbreviations: CRIB, Clinical Risk
White 367 (54.4) 238 (53.2) 129 (56.6) Index for Babies; HTN, hypertension;
Black 149 (22.1) 102 (22.8) 47 (20.6) IQR, interquartile range; NICU,
.70
neonatal intensive care unit; NSAID,
Other 159 (23.6) 107 (23.9) 52 (22.8) nonsteroidal anti-inflammatory drug;
Delivery room PDA, patent ductus arteriosus; SGA,
Intubation, No. (%) 308 (45.6) 222 (49.7) 86 (37.7) .003 small for gestational age; sCr, serum
creatinine.
Chest compressions, No. (%) 37 (5.5) 19 (4.3) 18 (7.9) .049
SI conversion factor: To convert
Apgar score at 1 min, median (IQR) 6 (3, 7) 5 (3, 7) 7 (3, 8) .001 serum creatinine value to micromoles
Apgar score at 5 min, median (IQR) 8 (6, 9) 8 (6, 8) 8 (7, 9) .005 per liter, multiply by 88.4.
a
Admission diagnoses, No. (%) Based on χ2 test for categorical
variables or t test for continuous
Respiratory distress 594 (88.0) 415 (92.8) 179 (78.5) <.001
variables.
Sepsis evaluation 407 (60.3) 284 (63.5) 123 (53.9) .02 b
Preeclampsia, eclampsia, or chronic
CRIB II score, mean (SD) 5.5 (4.6) 5.9 (4.3) 4.7 (5.1) .001 hypertension.
Respiratory support, No. (%) c
Propranolol, atenolol, carvedilol,
Invasive 387 (57.3) 278 (62.2) 109 (47.8) <.001 metoprolol, esmolol hydrochloride,
labetalol hydrochloride, amlodipine,
Noninvasive 490 (72.6) 355 (79.4) 135 (59.2) <.001
felodipine, isradipine, nicardipine
Treated PDA, No. (%) 94 (13.9) 67 (15.0) 27 (11.8) .26 hydrochloride, hydralazine
Initial sCr value, mean (SD), mg/dL 0.8 (0.3) 0.8 (0.2) 0.8 (0.5) .13 hydrochloride, minoxidil, sodium
nitroprusside, or clonidine
No. of sCr values measured, mean (SD) 11.4 (15.3) 11.6 (15.3) 11.0 (15.3) .63
hydrochloride.
Type of center, No. (%) d
Acyclovir, amphotericin B,
Children’s hospital 180 (26.7) 111 (24.8) 69 (30.3) aminoglycosides, piperacillin-
Perinatal center 115 (17.0) 62 (13.9) 53 (23.2) .005 tazobactam, or vancomycin
hydrochloride.
Surgical perinatal center 380 (56.3) 274 (61.3) 106 (46.5) e
Bumetanide, chlorothiazide,
Neonatal medications in the NICU, No. (%) furosemide, or spironolactone.
Nephrotoxic antimicrobial agentd 539 (79.7) 387 (86.6) 152 (66.7) <.001 f
Dobutamine hydrochloride,
Diuretice 24 (3.6) 15 (3.4) 9 (3.9) .69 dopamine hydrochloride,
epinephrine, milrinone lactate, or
Vasopressorf 71 (10.5) 45 (10.1) 26 (11.4) .59
norepinephrine bitartrate.
NSAIDg 83 (12.3) 71 (15.9) 12 (5.3) <.001 g
Indomethacin or ibuprofen.

adverse outcomes. It is now acknowledged that preterm in- ney disease (CKD),32-40 and the findings of other studies33,41-43
fants have a higher long-term risk of developing chronic kid- have suggested that preterm infants who experience AKI are

jamapediatrics.com (Reprinted) JAMA Pediatrics June 2018 Volume 172, Number 6 5/9

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Research Original Investigation Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates

Table 3. Primary Acute Kidney Injury (AKI) Outcomes Stratified by Caffeine Citrate Administrationa

No./Total No. (%) OR (95% CI) Abbreviations: NNE, number needed


Variable Caffeine No Caffeine Unadjusted Adjustedb NNE to be exposed; OR, odds ratio.
Early AKI, ≤7 dc a
Based on generalized linear mixed
Overall 50/447 (11.2) 72/228 (31.6) 0.28 (0.18-0.44) 0.20 (0.11-0.34) 4.3 model with logit link and binary
distribution.
Extremely preterm, <27 wk 30/149 (20.1) 38/55 (69.1) 0.07 (0.03-0.16) 0.13 (0.06-0.31) 2.2 b
Adjusted for gestational age, sepsis
Very preterm, 28-32 wk 20/298 (10.1) 34/173 (19.7) 0.31 (0.16-0.61) 0.27 (0.13-0.56) 8.1 evaluation, type of center, and
Any AKI, ≤120 dd antibiotic use.
c
Overall 103/447 (23.0) 83/228 (36.4) 0.56 (0.38-0.84) 0.27 (0.16-0.47) 4.4 P for interaction of gestational age
and caffeine administration = .23.
Extremely preterm, <27 wk 69/149 (29.5) 44/55 (80.0) 0.12 (0.05-0.30) 0.24 (0.10-0.58) 3.1
d
P for interaction of gestational age
Very preterm, 28-32 wk 34/293 (11.6) 39/170 (22.9) 0.52 (0.29-0.94) 0.32 (0.16-0.62) 8.0
and caffeine administration = .62.

Table 4. Secondary Acute Kidney Injury (AKI) Outcomes Stratified by Caffeine Citrate Administration

No. (%)
Caffeine No Caffeine
Variable (n = 447) (n = 228) Adjusted OR (95% CI)
Early AKI, ≤7 d Abbreviations: NNE, number needed
to be exposed; OR, odds ratio;
AKI, sCr plus UOP 50 (11.2) 72 (31.6) 0.20 (0.11-0.34)a
sCr, serum creatinine; UOP, urine
AKI, sCr 47 (10.5) 60 (26.3) 0.20 (0.11-0.37)a output.
AKI, UOP 8 (1.8) 17 (7.5) 0.40 (0.15-1.06)a a
Adjusted for assisted conception,
Stage 1 27 (6.0) 32 (14.0) gestational age, birth weight, race,
Apgar score at 1 minute, invasive
Stage 2 17 (3.8) 16 (7.0) 0.20 (0.12-0.34)b respiratory support, type of center,
Stage 3 6 (1.3) 24 (10.5) antibiotic use, and nonsteroidal
Any AKI, ≤120 d anti-inflammatory drug use.
b
Adjusted for assisted conception,
AKI, sCr 100 (22.4) 71 (31.1) 0.28 (0.16-0.49)a
birth weight, race, Apgar score at 1
Stage 1 54 (12.1) 40 (17.5) minute, invasive respiratory
Stage 2 34 (7.6) 17 (7.5) 0.30 (0.19-0.48)b support, type of center, antibiotic
use, and nonsteroidal
Stage 3 15 (3.4) 26 (11.4)
anti-inflammatory drug use.

at particular risk of subsequent diagnosis of CKD. In light of plasia, and treatment of patent ductus arteriosus.21 Given
the association of AKI with both short-term and long-term ad- these data, it is possible that decreased frequency of AKI
verse outcomes, it is imperative to develop strategies to pre- among infants receiving caffeine is not mediated by any
vent or ameliorate AKI. direct association of caffeine with kidney function but is
The results of this multicenter study confirm the associa- instead conferred through benefits in neonatal respiratory
tion of caffeine administration with AKI in preterm neonates status or hemodynamic stability.
described in smaller studies, with an effect magnitude simi- There are several potential mechanisms of action
lar to that found previously in a single-center study 22 of through which caffeine and possibly other methylxanthines
caffeine administration and AKI in VLBW neonates (NNE, 2.9 could directly reduce AKI. In investigations involving new-
for both cohorts). However, while Carmody et al22 found that born rabbits exposed to caffeine or theophylline, Gouyon and
most AKI reduction after caffeine administration occurred in Guignard46 reported increased renal blood flow, enhanced
neonates with stage 1 AKI (with no association with more se- sodium excretion, and a higher glomerular filtration rate. A
vere AKI stages), the findings of the present study suggest that subsequent study47 by the same authors demonstrated that
the benefit of caffeine extends to more severe AKI. theophylline counteracted hypoxemia-induced renal
In VLBW infants, caffeine use almost doubled between hemodynamic changes by maintaining renal vascular resis-
1997 to 2010, from 40% to 70%.21 This dramatic increase tance. Another potential mechanism of caffeine-mediated
coincides with clinical trials, such as the Caffeine for Apnea renal protection may involve attenuation of oxidative stress
of Prematurity Trial,44,45 in which infants randomized to and injury on endoplasmic reticulum. 48 Using a rodent
early treatment with caffeine were less likely to develop hyperoxia model, Teng et al48 demonstrated that caffeine, at
bronchopulmonary dysplasia. Over the same period, age at clinically appropriate doses, protected against hyperoxia-
initiation of caffeine administration fell from a mean of 10 induced impairment of alveolar formation, improved radial
to 12 days after birth (median, 4-5 days) to a mean of 4 days alveolar count and secondary septation, increased vascular-
(median, 1 day).21 In addition, the early use of caffeine has ity, normalized several signaling pathways, maintained nor-
been associated with significant decreases in early-onset mal endoplasmic reticulum and mitochondrial structure, and
and late-onset sepsis, pulmonary interstitial emphysema, decreased apoptosis. Whether any or all of these mechanisms
intraventricular hemorrhage, retinopathy of prematurity, could be relevant in the reduction of AKI seen in the present
duration of mechanical ventilation, bronchopulmonary dys- study requires further evaluation.

6/9 JAMA Pediatrics June 2018 Volume 172, Number 6 (Reprinted) jamapediatrics.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates Original Investigation Research

Limitations
Although this study included a large, diverse sample en- Conclusions
abling control for multiple potential confounders, there are sev-
eral limitations. First, because this was an observational study, This large multicenter study found that caffeine administra-
participants were not randomized to receive caffeine, and the tion in preterm infants is associated with reduced risk and se-
indication for caffeine administration could not be ascer- verity of AKI. Given the established benefits, widespread use,
tained. Second, although we created multivariable models to and safety of early caffeine treatment in neonates younger than
adjust for potential confounders, patients who received caf- 28 weeks’ gestational age, it is no longer possible to ethically
feine may nonetheless differ systematically from those who conduct a randomized clinical trial of caffeine vs placebo for
did not in unmeasured ways that could account for differ- protection against neonatal AKI. While these extremely pre-
ences in AKI occurrence. Third, we were unable to evaluate a term infants are at the highest risk of neonatal AKI, more ma-
dose-dependent effect because medication dosages and sys- ture neonates remain at risk: AKI occurred in almost 20% of
temic levels of caffeine were not collected in the AWAKEN the AWAKEN study participants born at 29 to 35 weeks’ ges-
study. Fourth, in this study, data collection was limited to a tation in the first week after birth. Because of the benefits and
maximum of 120 days. Therefore, we could only assess short- favorable adverse effect profile of caffeine, it may be reason-
term associations of caffeine administration with kidney func- able to consider routine use of prophylactic caffeine in neo-
tion. Whether exposure to caffeine might confer long-term re- nates of 28 to 32 weeks’ gestational age to prevent or reduce
nal benefits was beyond the scope of this study. Fifth, there is AKI, even when apnea of prematurity or the need for positive
also the possibility of misclassification bias because sCr and pressure respiratory support is absent. For extremely pre-
UOP values were not recorded every day on every infant. In term neonates, evaluation of the optimal dose, duration, and
addition, the definition of neonatal AKI used in this study does timing of initiation of caffeine therapy to prevent and reduce
not take into account maternal sCr values or the possibility of the severity of AKI should be explored with a standard proto-
misquantification of UOP. However, we believe that such mis- col for evaluating renal function and injury. Long-term out-
classification, if present, would occur equally between both comes of renal function and rates of CKD in those treated with
groups. caffeine are also needed to inform clinical practice.

ARTICLE INFORMATION Administrative, technical, or material support: Birmingham (UAB), as well as support for 2 of the
Accepted for Publication: January 31, 2018. Askenazi, Charlton. AWAKEN study investigators at UAB (Drs Askenazi
Study supervision: Harer, Askenazi, Charlton. and Griffin and Mr Boohaker). PICAN is part of the
Published Online: April 2, 2018. Department of Pediatrics at UAB and is funded by
doi:10.1001/jamapediatrics.2018.0322 Conflict of Interest Disclosures: All authors
reported no real or perceived conflicts of interest Children’s of Alabama hospital, UAB Department of
Correction: This article was corrected on May 7, Pediatrics, UAB School of Medicine, and UAB
that could affect the study design; the collection,
2018, to fix omissions in the Group Information Center for Clinical and Translational Sciences
analysis, and interpretation of data; the writing of
section of the end matter. (National Institutes of Health grant UL1TR001417).
the report; or the decision to submit the manuscript
Author Affiliations: Division of Neonatology, for publication. For full disclosure, we provide the Finally, the AWAKEN study at The University of New
Department of Pediatrics, School of Medicine and following additional list of authors’ other Mexico was supported by the Clinical and
Public Health, University of Wisconsin–Madison commitments and funding sources that are not Translational Science Center at The University of
(Harer); Division of Nephrology, Department of directly related to this study. Dr Askenazi reported New Mexico (National Institutes of Health grant
Pediatrics, The University of Alabama at serving on the speaker board for Baxter and for the UL1TR001449) and by The University of Iowa
Birmingham (Askenazi, Boohaker, Griffin); Division Acute Kidney Injury Foundation; he also reported Institute for Clinical and Translational Science (grant
of Nephrology, Department of Pediatrics, Eastern receiving grant funding for studies not related to U54TR001356). The AWAKEN study investigators
Virginia Medical School, Norfolk (Carmody); this work: grant R01 DK103608 from the National at the Canberra Hospital at the Australian National
Division of Neonatology, Department of Pediatrics, Institutes of Health/National Institute of Diabetes University Medical School were supported by the
Golisano Children’s Hospital, University of and Digestive and Kidney Diseases and grant R01 Canberra Hospital Private Practice Fund, and
Rochester, Rochester, New York (Guillet); Division FD005092 from the National Institutes of Health/ investigators at University of Virginia Children’s
of Nephrology, Department of Pediatrics, C.S. Mott US Food and Drug Administration. Dr Guillet Hospital were supported by a 100 Women Who
Children’s Hospital, University of Michigan, Ann reported receiving grant funding unrelated to this Care Grant from the 100 Women Charitable
Arbor (Selewski); Division of Neonatology, study: grant CTSI UL1 TR002001 from the Patient- Foundation.
Department of Pediatrics, University of Virginia, Centered Outcomes Research Institute. Dr Charlton Role of the Funder/Sponsor: The funding sources
Charlottesville (Swanson); Division of Nephrology, reported being the co-owner of Sindri for this study had no role in the design and conduct
Department of Pediatrics, University of Virginia, Technologies, LLC, which is not related to this study of the study; collection, management, analysis, and
Charlottesville (Charlton). and reported receiving grants 3P50DK096373, interpretation of the data; preparation, review, or
Author Contributions: Dr Charlton had full access R01DK110622, and R01DK111861 from the National approval of the manuscript; and decision to submit
to all of the data in the study and takes Institutes of Health. No other disclosures were the manuscript for publication.
responsibility for the integrity of the data and the reported. Group Information: The members of the NKC are
accuracy of the data analysis. Funding/Support: Cincinnati Children’s Hospital listed below. The following individuals served as
Study concept and design: Harer, Askenazi, Guillet, Center for Acute Care Nephrology provided funding collaborators and site investigators for the
Selewski, Swanson, Charlton. to create and maintain the Assessment of AWAKEN study and deserve a PubMed citation.
Acquisition, analysis, or interpretation of data: All Worldwide Acute Kidney Injury Epidemiology in They collaborated in protocol development and
authors. Neonates (AWAKEN) study Medidata Rave review, local institutional review board submission,
Drafting of the manuscript: All authors. electronic database. The Pediatric and Infant Center and data collection and participated in drafting or
Critical revision of the manuscript for important for Acute Nephrology (PICAN) provided support for review of the manuscript; Namasivayam
intellectual content: Harer, Askenazi, Carmody, web meetings and for the Neonatal Kidney Ambalavanan, MD (Department of Pediatrics, The
Griffin, Guillet, Selewski, Swanson, Charlton. Collaborative (NKC) steering committee annual University of Alabama at Birmingham); Subrata
Statistical analysis: Askenazi, Boohaker, Griffin. meeting at The University of Alabama at Sarkar, MD (C. S. Mott Children’s Hospital,

jamapediatrics.com (Reprinted) JAMA Pediatrics June 2018 Volume 172, Number 6 7/9

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Research Original Investigation Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates

University of Michigan, Ann Arbor); Alison Kent, Morales, MD, and Denisse Cristina Pareja Valarezo, 7. Askenazi D, Patil NR, Ambalavanan N, et al.
MD, and Jeffery Fletcher, PhD (Centenary Hospital MD (Holtz Children’s Hospital, University of Miami, Acute kidney injury is associated with
for Women and Children, Canberra Hospital, Miami, Florida); Sarah Cashman, BS, and Madeleine bronchopulmonary dysplasia/mortality in
Australian National University Medical School, Stead, BS (University of Iowa Children’s Hospital, premature infants. Pediatr Nephrol. 2015;30(9):
Canberra); Carolyn L. Abitbol, MD, Marissa Iowa City); Jonathan Davis, MD, and Julie Nicoletta, 1511-1518.
DeFreitas, MD, and Shahnaz Duara, MD (Holtz MD (Floating Hospital for Children at Tufts Medical 8. Jetton JG, Guillet R, Askenazi DJ, et al; Neonatal
Children’s Hospital, University of Miami, Miami, Center, Tufts University School of Medicine, Boston, Kidney Collaborative. Assessment of Worldwide
Florida); Patrick Brophy, MD, Carl D’Angio, MD, Massachusetts); Alanna DeMello (British Columbia Acute Kidney Injury Epidemiology in Neonates:
Ayesa Mian, MD, and Erin Rademacher, MD Children’s Hospital, Vancouver, Canada); Lynn Dill, design of a retrospective cohort study. Front Pediatr.
(Golisano Children’s Hospital, University of RN, and Emma Perez-Costas, PhD (The University 2016;4:68.
Rochester, Rochester, New York); Maroun J. of Alabama at Birmingham); Ellen Guthrie, RN
Mhanna, MD, Rupesh Raina, MD, and Deepak (MetroHealth Medical Center, Case Western 9. Jetton JG, Boohaker LJ, Sethi SK, et al; Neonatal
Kumar, MD (MetroHealth Medical Center, Case Reserve University, Cleveland, Ohio); Nicholas L. Kidney Collaborative (NKC). Incidence and
Western Reserve University, Cleveland, Ohio); Ayse Harris, BS, and Susan M. Hieber, MSQM (C.S. Mott outcomes of neonatal acute kidney injury
Akcan Arikan, MD, Christopher J. Rhee, MD (Texas Children’s Hospital, University of Michigan, (AWAKEN): a multicentre, multinational,
Children’s Hospital, Baylor College of Medicine, Ann Arbor); Katherine Huang and Rosa Waters observational cohort study. Lancet Child Adolesc
Houston); Stuart L. Goldstein, MD, and Amy T. (University of Virginia Children’s Hospital, Health. 2017;1(3):184-194.
Nathan, MD (Cincinnati Children’s Hospital Medical Charlottesville); Judd Jacobs, Ryan Knox, BS, Hilary 10. Jetton JG, Sorenson M. Pharmacological
Center, Cincinnati, Ohio); Juan C. Kupferman, MD, Pitner, MS, and Tara Terrell (Cincinnati Children’s management of acute kidney injury and chronic
Alok Bhutada, MD, and Shantanu Rastogi, MD Hospital Medical Center, Cincinnati, Ohio); Nilima kidney disease in neonates. Semin Fetal Neonatal
(Maimonides Medical Center, Brooklyn, New York); Jawale, MD (Maimonides Medical Center, Brooklyn, Med. 2017;22(2):109-115.
Elizabeth Bonachea, MD, John Mahan, MD, Arwa New York); Emily Kane (Australian National 11. Coulthard MG. The management of neonatal
Nada, MBBCH, and Susan Ingraham, MD University, Canberra); Vijay Kher, DM, and Puneet acute and chronic renal failure: a review. Early Hum
(Nationwide Children’s Hospital, Columbus, Ohio); Sodhi, MBBS (Medanta Kidney Institute, Medanta Dev. 2016;102:25-29.
Jennifer Jetton, MD, Tarah T. Colaizy, MD, and the Medicity, Gurgaon, Haryana, India); Grace Mele
Jonathan M. Klein, MD (University of Iowa (New York College of Osteopathic Medicine, 12. Jo SK, Rosner MH, Okusa MD. Pharmacologic
Children’s Hospital, Iowa City); F. Sessions Cole, MD, Westbury); Patricia Mele, DNP (Stony Brook treatment of acute kidney injury: why drugs haven’t
and T. Keefe Davis, MD (Washington University in St Children’s Hospital, Stony Brook, New York); worked and what is on the horizon. Clin J Am Soc
Louis, Missouri); Joshua Dower, BS, Lawrence Charity Njoku, Tennille Paulsen, and Sadia Zubair Nephrol. 2007;2(2):356-365.
Milner, MD, and Alexandra Smith, MD (Tufts (Texas Children’s Hospital, Baylor College of 13. Dobson NR, Hunt CE. Pharmacology review:
University School of Medicine, Boston, Medicine, Houston); Emily Pao (University of caffeine use in neonates: indications,
Massachusetts); Kimberly Reidy, MD, and Frederick Washington, Seattle Children’s Hospital, Seattle); pharmacokinetics, clinical effects, outcomes.
J. Kaskel, MD (The Children’s Hospital at Becky Selman, RN, and Michele Spear, CCRC Neoreviews. 2013;14(11):e540-e550. doi:
Montefiore, Bronx, New York); Jason Gien, MD, (University of New Mexico Health Sciences Center, 10.1542/neo.14-11-e540
Katja M. Gist, DO (University of Colorado, Children’s Albuquerque); Melissa Vega, PA-C (The Children’s 14. Bakr AF. Prophylactic theophylline to prevent
Hospital Colorado, Aurora); Mina H. Hanna, MD Hospital at Montefiore, Bronx, New York); and renal dysfunction in newborns exposed to perinatal
(University of Kentucky, Lexington); Craig S. Wong, Leslie Walther, RN (Washington University in asphyxia: a study in a developing country. Pediatr
MD, Catherine Joseph, MD, Tara DuPont, MD, Amy St Louis, Missouri). No compensation was received. Nephrol. 2005;20(9):1249-1252.
Staples, MD, and Robin Ohls, MD (University of
New Mexico Health Sciences Center, Albuquerque); REFERENCES 15. Bhat MA, Shah ZA, Makhdoomi MS, Mufti MH.
Surender Khokhar, MD (Apollo Cradle, Gurgaon, Theophylline for renal function in term neonates
1. Stoll BJ, Hansen NI, Bell EF, et al; Eunice Kennedy with perinatal asphyxia: a randomized,
Haryana, India); Sofia Perazzo, MD, Patricio E. Ray, Shriver National Institute of Child Health and
MD, and Mary Revenis, MD (Children’s National placebo-controlled trial. J Pediatr. 2006;149(2):
Human Development Neonatal Research Network. 180-184.
Medical Center, George Washington University Trends in care practices, morbidity, and mortality of
School of Medicine and the Health Sciences, extremely preterm neonates, 1993-2012. JAMA. 16. Jenik AG, Ceriani Cernadas JM, Gorenstein A,
Washington, DC); Cherry Mammen, MD, and Anne 2015;314(10):1039-1051. et al. A randomized, double-blind,
Synnes, MDCM (British Columbia Children’s placebo-controlled trial of the effects of
Hospital, Vancouver, Canada); Pia Wintermark, MD, 2. Manuck TA, Rice MM, Bailit JL, et al; Eunice prophylactic theophylline on renal function in term
and Michael Zappitelli, MD (Montreal Children’s Kennedy Shriver National Institute of Child Health neonates with perinatal asphyxia. Pediatrics. 2000;
Hospital, McGill University Health Centre, Montreal, and Human Development Maternal–Fetal Medicine 105(4):E45.
Quebec, Canada); Sidharth K. Sethi, MD (Kidney Units Network. Preterm neonatal morbidity and
mortality by gestational age: a contemporary 17. Raina A, Pandita A, Harish R, Yachha M, Jamwal
and Urology Institute, Medanta the Medicity, A. Treating perinatal asphyxia with theophylline at
Gurgaon, Haryana, India); Sanjay Wazir, MD cohort. Am J Obstet Gynecol. 2016;215(1):103.e1-
103.e14. birth helps to reduce the severity of renal
(Neonatology, Cloudnine Hospital, Gurgaon, dysfunction in term neonates. Acta Paediatr. 2016;
Haryana, India); Smriti Rohatgi, MD (Medanta the 3. Pierrat V, Marchand-Martin L, Arnaud C, et al; 105(10):e448-e451.
Medicity, Gurgaon, Haryana, India); Danielle E. EPIPAGE-2 Writing Group. Neurodevelopmental
Soranno, MD (University of Colorado, Children’s outcome at 2 years for preterm children born at 22 18. Cho HJ, Jeong IS. The renal effect of
Hospital Colorado, Aurora); Aftab S. Chishti, MD to 34 weeks’ gestation in France in 2011: EPIPAGE-2 prophylactic aminophylline therapy after cardiac
(University of Kentucky, Lexington); and Robert cohort study. BMJ. 2017;358:j3448. surgery in infants. J Crit Care. 2015;30(4):852.
Woroniecki, MD, and Shanty Sridhar, MD (Stony 4. Patel RM. Short- and long-term outcomes for 19. Al-Wassia H, Alshaikh B, Sauve R. Prophylactic
Brook School of Medicine, Stony Brook, New York). extremely preterm infants. Am J Perinatol. 2016;33 theophylline for the prevention of severe renal
Additional Contributions: We acknowledge the (3):318-328. dysfunction in term and post-term neonates with
outstanding work of the following clinical research perinatal asphyxia: a systematic review and
5. Srinivasan N, Schwartz A, John E, Price R, Amin meta-analysis of randomized controlled trials.
personnel and colleagues for their involvement in S. Acute kidney injury impairs postnatal renal
the AWAKEN study: Ariana Aimani, Samantha J Perinatol. 2013;33(4):271-277.
adaptation and increases morbidity and mortality in
Kronish, Ana Palijan, MD, and Michael Pizzi very low-birth-weight infants. Am J Perinatol. 2018; 20. Cattarelli D, Spandrio M, Gasparoni A, Bottino
(Montreal Children’s Hospital, McGill University 35(1):39-47. R, Offer C, Chirico G. A randomised, double blind,
Health Centre, Montreal, Quebec, Canada); Laila placebo controlled trial of the effect of theophylline
Ajour, BS, and Julia Wrona, BS (University of 6. Koralkar R, Ambalavanan N, Levitan EB, McGwin in prevention of vasomotor nephropathy in very
Colorado, Children’s Hospital Colorado, Aurora); G, Goldstein S, Askenazi D. Acute kidney injury preterm neonates with respiratory distress
Melissa Bowman, RN (University of Rochester, reduces survival in very low birth weight infants. syndrome. Arch Dis Child Fetal Neonatal Ed. 2006;
Rochester, New York); Teresa Cano, RN, Marta G. Pediatr Res. 2011;69(4):354-358. 91(2):F80-F84.
Galarza, MD, Wendy Glaberson, MD, Aura Arenas

8/9 JAMA Pediatrics June 2018 Volume 172, Number 6 (Reprinted) jamapediatrics.com

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019


Association Between Caffeine Administration and Acute Kidney Injury in Preterm Neonates Original Investigation Research

21. Dobson NR, Patel RM, Smith PB, et al. Trends in 30. Selewski DT, Charlton JR, Jetton JG, et al. 40. Iacobelli S, Loprieno S, Bonsante F, Latorre G,
caffeine use and association between clinical Neonatal acute kidney injury. Pediatrics. 2015;136 Esposito L, Gouyon JB. Renal function in early
outcomes and timing of therapy in very low birth (2):e463-e473. childhood in very low birthweight infants.
weight infants [published correction appears in J 31. Mammen C, Al Abbas A, Skippen P, et al. Am J Perinatol. 2007;24(10):587-592.
Pediatr. 2014;164(5):1244]. J Pediatr. 2014;164(5): Long-term risk of CKD in children surviving 41. Chaturvedi S, Ng KH, Mammen C. The path to
992-998.e3. episodes of acute kidney injury in the intensive care chronic kidney disease following acute kidney
22. Carmody JB, Harer MW, Denotti AR, Swanson unit: a prospective cohort study. Am J Kidney Dis. injury: a neonatal perspective. Pediatr Nephrol.
JR, Charlton JR. Caffeine exposure and risk of acute 2012;59(4):523-530. 2017;32(2):227-241.
kidney injury in a retrospective cohort of very low 32. Low Birth Weight and Nephron Number 42. Abitbol CL, Bauer CR, Montané B, Chandar J,
birth weight neonates. J Pediatr. 2016;172:63-68.e1. Working Group. The impact of kidney development Duara S, Zilleruelo G. Long-term follow-up of
23. Jetton JG, Askenazi DJ. Acute kidney injury in on the life course: a consensus document for extremely low birth weight infants with neonatal
the neonate. Clin Perinatol. 2014;41(3):487-502. action. Nephron. 2017;136(1):3-49. renal failure. Pediatr Nephrol. 2003;18(9):887-893.
24. Zappitelli M, Ambalavanan N, Askenazi DJ, 33. Maqsood S, Fung N, Chowdhary V, Raina R, 43. Harer MW, Pope CF, Conaway MR, Charlton JR.
et al. Developing a neonatal acute kidney injury Mhanna MJ. Outcome of extremely low birth Follow-up of Acute Kidney Injury in Neonates
research definition: a report from the NIDDK weight infants with a history of neonatal acute During Childhood Years (FANCY): a prospective
neonatal AKI workshop. Pediatr Res. 2017;82(4): kidney injury. Pediatr Nephrol. 2017;32(6):1035-1043. cohort study. Pediatr Nephrol. 2017;32(6):1067-1076.
569-573. 34. Luyckx VA. Preterm birth and its impact on 44. Schmidt B, Roberts RS, Davis P, et al; Caffeine
25. Parry G, Tucker J, Tarnow-Mordi W; UK renal health. Semin Nephrol. 2017;37(4):311-319. for Apnea of Prematurity Trial Group. Caffeine
Neonatal Staffing Study Collaborative Group. CRIB 35. Stritzke A, Thomas S, Amin H, Fusch C, Lodha therapy for apnea of prematurity. N Engl J Med.
II: an update of the Clinical Risk Index for Babies A. Renal consequences of preterm birth. Mol Cell 2006;354(20):2112-2121.
score. Lancet. 2003;361(9371):1789-1791. Pediatr. 2017;4(1):2. 45. Schmidt B, Roberts RS, Davis P, et al; Caffeine
26. Bender R, Blettner M. Calculating the “number 36. Starzec K, Klimek M, Grudzień A, Jagła M, for Apnea of Prematurity Trial Group. Long-term
needed to be exposed” with adjustment for Kwinta P. Longitudinal assessment of renal size and effects of caffeine therapy for apnea of prematurity.
confounding variables in epidemiological studies. function in extremely low birth weight children at 7 N Engl J Med. 2007;357(19):1893-1902.
J Clin Epidemiol. 2002;55(5):525-530. and 11 years of age. Pediatr Nephrol. 2016;31(11): 46. Gouyon JB, Guignard JP. Renal effects of
27. Stoops C, Sims B, Griffin R, Askenazi DJ. 2119-2126. theophylline and caffeine in newborn rabbits.
Neonatal acute kidney injury and the risk of 37. Nishizaki N, Hirano D, Nishizaki Y, et al. Pediatr Res. 1987;21(6):615-618.
intraventricular hemorrhage in the very low birth Increased urinary angiotensinogen is an effective 47. Gouyon JB, Guignard JP. Theophylline prevents
weight infant. Neonatology. 2016;110(4):307-312. marker of chronic renal impairment in very low the hypoxemia-induced renal hemodynamic
28. Carmody JB, Charlton JR. Short-term birth weight children. Clin Exp Nephrol. 2014;18(4): changes in rabbits. Kidney Int. 1988;33(6):1078-1083.
gestation, long-term risk: prematurity and chronic 642-648. 48. Teng RJ, Jing X, Michalkiewicz T, Afolayan AJ,
kidney disease. Pediatrics. 2013;131(6):1168-1179. 38. Abitbol CL, Rodriguez MM. The long-term renal Wu TJ, Konduri GG. Attenuation of endoplasmic
29. Madsen NL, Goldstein SL, Frøslev T, and cardiovascular consequences of prematurity. reticulum stress by caffeine ameliorates
Christiansen CF, Olsen M. Cardiac surgery in Nat Rev Nephrol. 2012;8(5):265-274. hyperoxia-induced lung injury. Am J Physiol Lung
patients with congenital heart disease is associated 39. Bonamy AK, Källén K, Norman M. High blood Cell Mol Physiol. 2017;312(5):L586-L598.
with acute kidney injury and the risk of chronic pressure in 2.5-year-old children born extremely
kidney disease. Kidney Int. 2017;92(3):751-756. preterm. Pediatrics. 2012;129(5):e1199-e1204.

jamapediatrics.com (Reprinted) JAMA Pediatrics June 2018 Volume 172, Number 6 9/9

© 2018 American Medical Association. All rights reserved.

Downloaded From: by a Tulane University User on 01/05/2019

You might also like