Original article
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
1
Department of Pediatrics,
All India Institute of Medical
Sciences (AIIMS), Bhopal,
Madhya Pradesh, India
2
Department of Pediatrics,
Division of Nephrology,
Montreal Children’s Hospital
and McGill University, Montreal,
Quebec, Canada
3
Department of Pediatrics,
All India Institute of Medical
Sciences (AIIMS), Bhubaneswar,
Odisha, India
Correspondence to
Dr Girish Chandra Bhatt,
Department of Pediatrics,
All India Institute of Medical
Sciences (AIIMS), Bhopal,
Madhya Pradesh, India;
​drgcbhatt@​gmail.c​ om
Received 6 July 2018
Revised 26 November 2018
Accepted 22 January 2019
© Author(s) (or their
employer(s)) 2019. No
commercial re-use. See rights
and permissions. Published
by BMJ.
To cite: Bhatt GC, Gogia P,
Bitzan M, et al. Arch Dis Child
Epub ahead of print: [please
include Day Month Year].
doi:10.1136/
archdischild-2018-315805
Theophylline and aminophylline for prevention of
acute kidney injury in neonates and children: a
systematic review
Girish Chandra Bhatt,1 Priya Gogia,1 Martin Bitzan,2 Rashmi Ranjan Das3
Abstract
Objective  To compare the efficacy and safety of
theophylline or aminophylline for prevention of acute
kidney injury (AKI) in neonates and children.
Design  Systematic review and meta-analysis with
application of Grading of Recommendations, Assessment,
Development and Evaluation system.
Data sources  PubMed/MEDLINE, Embase, Google
Scholar and Cochrane renal group were searched from
1970 to May 2018.
Eligibility criteria  Randomised clinical trials and
quasi-randomised trials comparing the efficacy and
safety of prophylactic theophylline or aminophylline
for prevention of AKI in neonates and children were
included. The primary outcomes were: incidence of AKI,
serum creatinine levels and all-cause mortality.
Results  A total of nine trials were included in the
qualitative synthesis. Six trials including 436 term
neonates with birth asphyxia who received a single dose
of theophylline were finally included in the meta-analysis.
The pooled estimate showed 60% reduction in the
incidence of AKI in neonates with severe birth asphyxia
(RR: 0.40; 95% CI 0.3 to 0.54; heterogeneity: I2=0%)
(moderate quality evidence), decrease in serum creatinine
over days 2–5 (very low to low quality evidence) without
significant difference in all-cause mortality (RR: 0.88;
95% CI 0.52 to 1.50; heterogeneity: I2=0%) (very low-
quality evidence). A significant difference in the negative
fluid balance, increase in GFR and decrease in urinary β2
microglobulin was seen in favour of theophylline.
Conclusion and relevance  A single dose of
prophylactic theophylline helps in prevention of AKI/
severe renal dysfunction in term neonates with severe
birth asphyxia (moderate quality evidence) without
increasing the risk of complications and without affecting
all-cause mortality (very low-quality evidence).
Trial registration number  CRD 42017073600.
Introduction
Acute kidney injury (AKI) is defined as a rapid
loss in kidney function (hours to days), resulting
in derangements in fluid balance, electrolytes and
waste products.1 AKI may result from impaired
renal perfusion, exposure to nephrotoxic drugs,
sepsis or ischaemia during surgery, such as cardio-
pulmonary bypass.2 In a recent prospective national
cohort study using electronic alert system, the inci-
dence of hospital and community acquired AKI in
children was found to be 40.1% and 29.4%, respec-
tively.3 The reported incidence of neonatal AKI
ranges from 8.4% to 34.5%3–5 and is associated
Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805
What is already known on this topic?
►► Acute kidney injury (AKI) is associated with
increased mortality, longer duration of hospital
stay and an increased cost.
►► Care for neonates and children with AKI
remains supportive.
What this study adds?
►► A single dose of adenosine antagonists
reduces the incidence of AKI in term neonates
with severe birth asphyxia by 60% without
increasing the risk of complications (moderate
quality evidence).
►► Prophylactic theophylline given to neonates
with severe birth asphyxia also decreases serum
creatinine, maintains negative fluid balance and
increases glomerular filtration rate.
with increased mortality, longer duration of hospital
stay and an increased cost.6 7 A systematic review
of large cohort studies conducted between 2004 to
2012 showed a pooled incidence of AKI in 33.7%
children and a pooled mortality rate of 13.8%.8
Despite increased mortality, there are few modali-
ties for prevention and treatment of AKI.
Some of the previous studies and randomised
controlled trials (RCTs) have shown renoprotec-
tive role of theophylline and aminophylline in
term9–13 and preterm neonates with severe birth
asphyxia,14 children undergoing cardiac surgery15 16
and in preterm neonates with respiratory distress
syndrome.17 A previous systematic review had
shown improved outcome with use of a single
dose of theophylline.18 Since then, two RCTs have
been published.9 19 The previous review focused
mainly on neonates with severe birth asphyxia;
the description of outcome data was limited, and
treatment emergent complications were excluded
from the final meta-analysis. Thus, we conducted
this updated systematic review including RCTs
and quasirandomised trials to explore the present
evidence for the use of adenosine antagonists for
prevention of AKI. We used the Grading of Recom-
mendations, Assessment, Development and Evalua-
tion (GRADE)20 approach to rate quality of evidence
for primary outcomes and present our results with
GRADE summary tables, which was also lacking in
the previous systematic review.
   1
 Original article
Objective
To describe the efficacy and safety of theophylline or aminophyl-
line compared with standard therapy or placebo for prevention
of AKI in neonates and children.
Method
We included RCTs and quasi-RCTs (trials where allocation is not
truly random, eg, by alternation, use of alternate medical record
numbers, date of birth or other methods with a potential for
selection bias).
Types of participants
Neonates and children (0–18 years) who received theophylline
or aminophylline for prevention of AKI and compared with stan-
dard therapy or placebo were eligible. Exclusion criterion were:
children with pre-existing renal disease, history of tachyarrhyth-
mias, seizures, cardiac transplant recipients, small for gestational
age, neonates requiring mechanical ventilation or history of
drug intake by mother causing fetal and neonatal depression,
children/infants requiring renal replacement therapy, deranged
liver enzymes (>3 times normal) and coagulopathy.
The definition of AKI was taken as used in the individual study
(table 1).
Search methods for identification of studies
Cochrane Central Register of Controlled Trials, PubMed/
MEDLINE, Embase, Google Scholar and Cochrane renal group
were searched from 1970 to May 2018. The following search
strategy was applied: ((((((adenosine receptor antagonist) OR
adenosine antagonist) OR theophylline) OR aminophylline)
OR caffeine)) AND ((((((((acute kidney injury) OR acute renal
failure) OR acute Kidney Failure) OR renal failure) OR acute
renal insufficiency) OR renal insufficiency) OR renal dysfunc-
tion) OR acute renal injury)) AND (((((((Neonates) OR Newborn
Infant) OR infant) OR preschool) OR toddler) OR paediatric)
OR children). The paediatric age group includes patients up to
18 years.
Types of outcome measures
The primary and secondary outcomes were defined before the
collection of data. The outcome measures were collected day
wise up to 5 or 7 days as specified in individual trials.
Primary outcome measures
1. Incidence of AKI/severe renal dysfunction in initial 7-day
period.
2. Serum creatinine (mg/dL) levels daily up to 5 days.
3. All-cause mortality.
Secondary outcome measures
1. Fluid balance daily up to 5 days.
2. Estimated glomerular filtration rate (eGFR; mL/min/1.73 m2)
daily up to 5 days.
3. Urinary β2 globulin levels (mg/L) during initial 5 days.
4. Disease complications and treatment emergent adverse
events.
Treatment emergent adverse events has been defined as
adverse events that have first occurred or worsened in severity
after initiation of treatment.21
Data extraction (selection and coding)
Data were collected using a pilot tested data extraction form.
Two authors (GCB and PG) independently extracted data
2 Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
including author, type of population, exposure/intervention
(theophylline or aminophylline vs standard therapy or placebo),
results (outcome measures, effect and significance) and sources
of funding/support. Any disagreement in the extracted data was
resolved by discussion with the third author (MB).
Risk of bias (quality) assessment
Two review authors (GCB and RRD) independently assessed the
quality of the included trials by using Cochrane methodological
quality assessment forms.22
Strategy for data synthesis
The data were pooled and expressed as mean difference
(MD) with 95% CI for continuous data and risk ratio
(RR) with 95% CI for categorical data. A p value <0.05
was considered significant. I 2 statistics was used for assess-
ment of heterogeneity. In case of high-level heterogeneity
(>50%), we tried to explore the cause. A fixed effect model
was initially conducted and if significant statistical hetero-
geneity existed between trials (I 2 >50%), potential sources
of heterogeneity were considered and where appropriate a
random effects model was used. RevMan (Review Manager)
V.5.2 was used for all the analyses. 23
Publication bias
For publication bias, we used inverted funnel plot as suggested
by Egger et al.24
Grade of evidence
For assessment of the quality of evidence we used GRADE
Profiler software (V.3.2). 25 GRADE assessment of evidence
quality reflects confidence in the estimates of benefits or
harms. GRADE is implemented with four levels of evidence
quality, namely, high, moderate, low and very low. Rating
is made for each outcome based on study design, risk of
bias, imprecision, indirectness and magnitude of effect. 26
‘Summary of findings (SOF)’ tables were constructed by
using GRADE profiler, and only primary outcomes were
included in SOF tables.
Sensitivity analysis
We evaluated the impact of methodological quality by removing
the trials at high or unclear risk of bias for random sequence
generation and allocation concealment.
Results
A total of 88 citations were retrieved by using the search
strategy; after duplicate removal, 84 articles were available
for screening of which 73 were excluded (title and/or abstract
not relevant). The full texts of the remaining 11 articles were
assessed for eligibility (figure 1). After further exclusion of two
articles (one has evaluated the role of aminophylline in patients
with chronic kidney disease (CKD), one has experimental drug
in both the arms), nine trials were finally included in the quali-
tative synthesis (table 1). Six trials,9–13 19 with 436 participants,
evaluated the effect of prophylactic theophylline for the preven-
tion of severe renal dysfunction in term neonates with severe
birth asphyxia. One trial17 evaluated the role of theophylline in
preterm neonates <32 weeks’ gestation with respiratory distress
syndrome, two trials evaluated the role of aminophylline—one
in preterm neonates with birth asphyxia14 and another in infants
and children undergoing cardiac surgery with cardiopulmo-
nary bypass.15 Only two trials have low risk of bias in all the
 Table 1  Characteristics of included studies
Study author Setting, country Participants Intervention Outcomes measured Comments
13
Jenik et al Three centres, Numbers (n=51): theophylline group (n=24); placebo group (n=27). Theophylline group received single 1. Severe renal dysfunction. Blinding of outcome assessor not
Argentina Age: neonates. dose of 8 mg/kg over 5 min by infusion 2. Serum creatinine. clear. Theophylline levels were
Inclusion criterion: neonates (term and post-term) with severe birth asphyxia within first hour of birth. 3. Fluid balance. measured.
defined as: history of fetal distress, 5 min Apgar score of 6 or lower and Placebo group received 5% dextrose 4. GFR
requirement of immediate neonatal ventilation. in water. 5. Urinary sodium excretion.
Exclusion criterion: condition not related to asphyxia, small for gestational 6. Urinary β2 microglobulin.
age, congenital abnormalities of kidney/urinary tract, cardiovascular pathology Severe renal dysfunction was defined
not related to prenatal asphyxia, exposure to medications modifying by following criteria: serum creatinine
haemodynamics and renal function, polycythaemia, clinical evidence of >1.5 mg/dL for at least two consecutive
potential antenatal injury and pharmacological depression. days or rising levels of serum creatinine
(0.3 mg/day) per day.
Bakr12 Single-centre Numbers (n=40): theophylline group (n=20), placebo group (n=20). Theophylline group received single 1. Severe renal dysfunction. Random sequence generation,
NICU, Age: term and post-term newborns. dose of 5 mg/kg theophylline 2. Serum creatinine. allocation concealment and binding
Egypt Inclusion criterion: newborns with severe birth asphyxia defined by: 5 min intravenously over 5 min. 3. Fluid balance. of outcome assessor not clear.
Apgar score of 6 or lower, base deficit >15 mEq/L in cord or admission blood Placebo group received 2 mL of 10% 4. Creatinine clearance. Theophylline levels not done.
gas, requirement of vigorous resuscitation. dextrose. 5. GFR.
Exclusion criterion: preterm and small for gestational age and neonates with 6. Urinary β2 microglobulin and
congenital abnormality and dysmorphism. haematuria.
Severe renal dysfunction was defined as:
serum creatinine of >1.5 mg/dL for two
consecutive days.
Bhat et al10 Single centre Numbers (n=70): theophylline group (n=40); placebo group (n=30). Theophylline group received 8 mg/kg 1. Severe renal dysfunction. Random sequence generation,

Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805

(NICU), Age: term and post-term neonates. intravenous theophylline over 5 min. 2. Serum creatinine. allocation concealment and binding
India. Inclusion criterion: neonates with severe perinatal asphyxia defined by: history Placebo group received equal amount 3. Urinary sodium excretion. of outcome assessor not clear.
of fetal distress, need for immediate ventilation and a 5 min Apgar score of ≤6, of 5% dextrose in water. 4. Fluid balance. Theophylline levels not done.
base deficit ≥15 mEq/L in cord blood or admission or cord blood PH <7. 5. Weight change.
Exclusion criterion: drugs used by mother that can affect renal haemodynamics 6. Urine output.
and renal function, any condition unrelated to asphyxia, cardiovascular disease Severe renal dysfunction was defined
unrelated to asphyxia, congenital malformation of the kidneys or urogenital as: serum creatinine >1.5 mg/dL for
tract, polycythaemia, microcephaly and chromosomal disorders or severe two consecutive days and rising serum
intrauterine growth disorders. creatinine level (0.3 mg/kg/day).
Cattarelli et al17 Single centre, Numbers (n=50): theophylline group (n=25); placebo group (n=25). Experimental group was given 1. Urine output. The trial has low risk of bias in all
Italy Age: preterm neonates. intravenous theophylline 1 mg/kg for 2. Serum creatinine. domains. Theophylline levels were
Inclusion criterion: preterm neonates <32 weeks gestation who developed 3 days. 3. GFR. measured.
respiratory distress syndrome within 6 hours and needed mechanical ventilation Control group was given equal volume 4. Serum electrolytes.
or nasal continuous positive pressure ventilation. of placebo, that is, 5% dextrose in 5. Urinary β2 microglobulin.
Exclusion criterion: kidney or urinary tract congenital anomalies, congenital water for 3 days. 6. Complications.
heart defects, prenatal exposure to inhibitors of ACE or non-steroidal anti-
inflammatory drugs and chromosomal disorders or multiple malformations.
Eslami et al11 Single centre, Numbers (n=36): theophylline group (n=17); placebo group (n=19). Experimental group received 1. Acute kidney failure. Blinding of participants, caregiver
Iran Age: neonates. single intravenous dose of 5 mg/kg 2. Serumand urinary creatinine. and outcome assessor along with
Inclusion criterion: term and post-term neonates weighing 2500 g or more with theophylline slowly. 3. GFR. allocation concealment unclear.
severe birth asphyxia. Placebo group received 2 mL of 4. Urinary sodium excretion. Theophylline levels not done.
Severe birth asphyxia was defined by: Apgar score of ≤3 in first minute and dextrose solution. 5. Serum electrolytes.
≤6 in the fifth minute, base deficit >15 mEq/L in cord or arterial sample or need Acute kidney failure was defined as:
for severe resuscitation. increase in serum creatinine ≥0.3 mg/dL or
Exclusion criterion: preterm delivery, small for gestational age, congenital serum creatinine level >1.5 mg/dL for at
anomalies, need for drugs affecting kidney function, hypotension, requirement least two consecutive days.
of ventilator, seizures, cerebral attacks, severe kidney dysfunction and oliguria.

Continued
Original article

3
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
4
Table 1  Continued 
Study author Setting, country Participants Intervention Outcomes measured Comments
Original article

Merrikhi et al14 Single centre, Numbers (n=22): aminophylline group (n=11); placebo group (n=11). Experimental group: received single 1. Urine output. Random sequence generation,
Iran Age: preterm neonates. dose of aminophylline 5 mg/kg slowly. 2. GFR. blinding of participants, caregiver
Inclusion criterion: preterm infants diagnosed with perinatal asphyxia. Placebogroup received 5% dextrose 3. β2 microglobulin levels. and outcome assessor along with
Neonatal asphyxia was defined by: Apgar scores less than or equal to 6 at water 4. NAG levels. allocation concealment unclear.
5 min plus umbilical artery pH values less than 7.0 according to AAP guideline. Theophylline levels not measured.
Exclusion criterion: preterm neonates with major congenital malformations,
chromosomal or renal abnormalities, breech presentation and those with
maternal gestational diabetes.
Axelrod et al15 Single centre, Numbers (n=144): aminophylline group (n=72); placebo group (n=72). Experimental group received 1. Development of AKI in the first five Study drug was discontinued in
USA Age: patients less than 18 years old with congenital heart defects undergoing aminophylline 5 mg/kg intravenous postoperative days. 24% in experimental arm and 15%
cardiac surgery with cardiopulmonary bypass. over 30 min followed by 1.8 mg/kg 2. Median time between postoperative in placebo group (but intention-to-
Exclusion criterion: history of tachyarrythmia, seizures, aspartate intravenous every 6 hours for 72 hours CVICU admission and first successful treat analysis was done). The trial
aminotransferase or alanine aminotransferase greater than 3 times normal, (13 doses). endotracheal extubation. has low risk of bias in all domains.
coagulopathy, sepsis, fever or hypothyroidism, cardiac transplant recipients, Placebo group received intravenous 3. Mean percentage fluid overload. Theophylline trough levels were
neonates less than 36 weeks’ corrected gestational age and patients requiring infusion of normal saline for the same 4. Total fluid balance. measured.
renal replacement therapy or extracorporeal membrane oxygenation. duration. 5. Urine output.
6. Inotropic support.
Raina et al9 Single centre, Numbers (n=159): theophylline group (n=78); placebo group (n=81). Experimental group received single 1. Severe renal dysfunction. Blinding of participants, caregiver
India Age: term neonates. dose of intravenous theophylline 5 mg/ 2. Serum creatinine. and outcome assessor unclear.
Inclusion criterion: term neonates with weight more than 2500 g with severe kg over 5 min period. 3. GFR. Follow-up not done. Theophylline
birth asphyxia. Control group received equal volume 4. Urinary sodium excretion. levels were not measured.
Severe birth asphyxia was defined as having any two of the following criteria: of placebo (0.25 mL/kg normal saline). 5. Fluid balance.
Apgar score of up to 3 at 1 min or up to 5 at 5 min, positive pressure ventilation Severe renal dysfunction was defined as per
for more than 10 min, seizure, severe hypotonia and a Thomson score of more the neonatal KDIGO guidelines, that is, rise
than 15. in creatine >1.5 times the lowest previous
Exclusion criterion: small for gestation age, presence of congenital anomalies, value.
mother taking drugs causing fetal or neonatal depression baby requiring
mechanical ventilation and having culture proven sepsis.
Ghazala et al19 Single centre, Numbers (n=80): theophylline group(n=40); placebo group (n=40). Experimental group received single 1. Severe renal dysfunction. Allocation concealment, blinding of
Pakistan Age: term neonates. dose of intravenous theophylline 5 mg/ 2. Serum creatinine. participants, caregiver and outcome
Inclusion criterion: term neonates with weight more than 2500 g with severe kg slowly. 3. Serum electrolyte. assessor as well as selective
birth asphyxia. Control group received 2 mL/kg of 10% 4. Fluid balance. outcome reporting were unclear. The
Severe birth asphyxia was defined as having any four of the following criteria: dextrose solution. 5. Urine output. outcomes were not defined. Follow-
history of fetal distress, requirement of neonatal resuscitation by bag and mask Severe renal dysfunction was defined as up not done. Theophylline levels
or intubation >2 min after birth, 5 min Apgar score of <6, PH <7.0 as measured per the neonatal KDIGO guidelines, that were not measured.
in NICU and neurological manifestations in immediate neonatal period. is, rise in creatinine >1.5 times the lowest
Exclusion criterion: preterm delivery <37 weeks, small for gestational age, previous value.
congenital anomalies (polycystic kidney disease, ectopic kidneys and horseshoe
kidneys), those on medication (amikacin, acyclovir and vancomycin) and
hypotension (mean arterial pressure ≤40 mm Hg).
AAP, American Academy of Pediatrics; AKI, acute kidney injury; CVICU, cardiovascular intensive care unit; GFR, glomerular filtration rate; KDIGO, kidney disease improving global outcomes; NAG, N-acetyl-beta-D glucosaminidase; NICU, neonatal
intensive care unit.

Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805

Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.

Figure 1  PRISMA flow diagram. PRISMA, Preferred Reporting Items
for Systematic Reviews and Meta-Analyses. CKD, chronic kidney disease.
domains,15 17 and the rest seven trials have unclear risk of bias
in one more domains (random sequence generation, allocation
concealment, blinding and selective reporting) without any
domain for high risk of bias (online supplementary figure 1).
The risk of bias for random sequence generation was unclear
in three trials.10 12 14 The risk of bias for allocation concealment
was unclear in five trials.10–12 14 19 The risk of performance bias
was unclear in four trials.9 11 14 19 The risk of detection bias was
unclear, attrition and other bias was low, in all the nine trials.
Reporting bias was low in all except one trial.19
Theophylline versus placebo for prevention of AKI in term
neonates with severe birth asphyxia
Theophylline was given in a single doses of 5mg/kg in four
trials9 11 12 19 and 8mg/kg in two trials.10 13
Primary outcome measures
1. Incidence of AKI: six trials,9–13 19 with 436 participants re-
ported this outcome. The pooled estimate showed 60% re-
duction in the incidence of AKI in the neonates with severe
birth asphyxia receiving single dose of prophylactic theoph-
ylline compared with placebo (RR: 0.35; 95% CI 0.25 to
0.49; I²=0%) (figure 2).
Figure 2  Forest plot showing incidence of AKI in neonates with severe birth asphyxia (theophylline vs placebo). AKI, acute kidney injury.
Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805
Original article
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
2. Creatinine levels during initial 5 days: this outcome was
reported by six trials.9–13 19 A high degree of unexplained
heterogeneity was noted for this outcome. There was no
significant difference between the creatinine levels on day
1 (mean difference (MD): −0.24; 95% CI −0.72 to 0.24;
I2=98%). Over the next 2–5 days, there was a significant
decrease in serum creatinine levels in neonates who received
theophylline (day 5 level: [MD: −0.38; 95% CI −0.56 to
–0.21; I2=85%]) (figure 3).
3. All-cause mortality: five trials9–13 with 356 participants
reported 47 deaths. The pooled estimated showed 12%
reduction of mortality in neonates who received prophylac-
tic theophylline. However, the difference between the two
groups was not significant (RR: 0.88; 95% CI 0.52 to 1.50;
I2=0%) (figure 4).
Secondary outcome measures
1. Fluid balance during the initial 5 days: four trials reported
this outcome,9 10 12 13 and the pooled estimate showed a sig-
nificant negative fluid balance in the neonates who received
a single dose of prophylactic theophylline as compared with
placebo. There was no significant difference between the
two groups on day 1 (MD: −0.22, 95% CI −0.51 to 0.08;
I2=95%),day 2 (MD: 0.79; 95% CI 0.04 to 1.53; I2=96%)
and day 4 (MD: 0.20; 95% CI −0.22 to 0.62; I2=89%).
However, the difference was significant on day 3 (MD: 0.5;
95% CI 0.11 to 0.89;I2=86%) and day 5 (MD: 0.32; 95% CI
0.20 to 0.44; I2=34%). Another trial reported a negative flu-
id balance in favour of the theophylline group over days 2–5
compared with placebo9 (figure 5 and online supplementary
figure 2).
2. eGFR during the initial 5 days: four trials reported this out-
come.9 11–13 Although there was no significant difference
in the eGFR on days 1 (MD: 1.17; 95% CI −2.8 to 5.16;
I2=47%) and 4 (MD: 3.9; 95% CI −13.9 to 21.8; I2=96%),
the difference was significant on days 2 (MD: 9.4; 95% CI
6.1 to 12.68;I2=0%), 3 (MD: 14.3; 95% CI 11.73 to 16.87;
I2=0%) and 5 (MD: 10.1; 95% CI 5.81 to 14.29; I2=47%)
(online supplementary figure 3).
3. Urinary β2 globulin levels during the initial 5 days: three tri-
als reported this outcome.10 12 13 The pooled results showed
significantly lower mean urine β2 microglobulin (B2M)
concentrations, used as a marker of tubular injury27 in the
treatment group compared with the control (MD: −7.07;
95% CI −8.93 to –5.22; I2=32%) (online supplementary
figure 4).
4. Complications: five trials reported this outcome9–13 (online
supplementary figure 5).
5
 Original article

Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
Figure 3  Forest plot showing serum creatinine in neonates with severe birth asphyxia (theophylline vs placebo).

a. Requirement of ventilatory support: two trials with 195
participants reported this outcome. There was no signif-
icant difference between treatment and control group
(RR: 0.75; 95% CI 0.43 to 1.31; I2=44%).
b. Requirement of inotropic support: four trials9 10 12 13 with
320 participants reported this outcome without signifi-
cant difference between groups (RR: 0.92; 95% CI 0.68
to 1.25; I2=0%).
c. Seizures: four trials9 11–13 with 286 participants report-
ed this outcome without significant difference between
groups (RR: 0.79; 95% CI 0.56 to 1.12; I2=0%).
Theophylline for preterm infants with respiratory distress
syndrome
We identified a single, double-blind randomised, placebo
controlled trial examining the effect of intravenous theophylline
in preterm neonates with respiratory distress syndrome at a dose
of 1 mg/kg for 3 days.17 The incidence of oliguria was decreased
in preterm neonates receiving theophylline on day 1 (RR: 0.14;
95% CI 0.02 to 1.05), but the difference did not extend to the
subsequent 10 days. Serum creatinine levels on day 24 were
insignificantly lower in neonates who received theophylline
compared with placebo (MD: −22.50; 95% CI −39.25 to 5.75).
Similarly, creatinine levels on days 3 to10 did not differ signifi-
cantly between groups. All-cause mortality and other secondary
outcomes were not reported.
Aminophylline for preterm neonates with perinatal asphyxia
One trial examined the effect of a single dose of aminophylline
on renal function and markers of AKI (N-acetyl-glucosaminidase
and B2M) on days 1, 4 and 7 of life.14 Aminophylline (5 mg/kg)
was given to preterm neonates with perinatal asphyxia(n=11)

Figure 4  Forest plot showing all-cause mortality in neonates with severe birth asphyxia (theophylline vs placebo).
6 Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805

Figure 5  Forest plot showing fluid balance in neonates with severe birth asphyxia (theophylline vs placebo).
within the first hour after birth and compared with placebo
(n=11). Aminophylline-treated infants demonstrated signifi-
cantly higher eGFR on day 4 compared with placebo with no
difference in NAG and B2M levels and increased urine output
from day 1. Incidence of AKI, all-cause mortality and other
secondary outcomes were not reported.
Aminophylline for children less than 18 years with congenital
heart defects undergoing cardiac surgery
One trial reported use of aminophylline in postoperative chil-
dren after congenital heart surgery with cardiopulmonary
bypass.15 This was a double-blind randomised placebo controlled
trial where the experimental group (n=72) received intravenous
aminophylline (5 mg/kg loading dose followed by 1.8 mg/kg every
6 hours for 72 hours) (n=72) and the control group isotonic
saline (placebo) matched by volume and appearance. Study drug
was discontinued in 24% of the patients in the experimental arm
and 15% of subjects in the placebo group. There was no differ-
ence in the incidence and stage of serum creatinine-based AKI
or urine output between the two groups. A subgroup analysis
(<3 months old) showed no difference between the outcomes,
but this analysis was insufficiently powered. Adverse event rates
were similar in both groups.
Treatment emergent adverse events
These were reported in seven9–13 15 17 of nine trials. They were
mainly related to the disease process and occurred with similar
frequency in either group. Table 2 shows treatment emergent
Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805
Original article
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
adverse events and whether the treatment emergent adverse
events were defined as one of the outcomes.
Sensitivity analysis
As all the six included trials of theophylline versus placebo for
prevention of AKI in neonates with severe birth asphyxia were
having one or more domains of unclear risk of bias, we could not
perform sensitive analysis.
Publication bias and small effect
Due to the paucity of published studies (<10), a funnel plot was
not constructed.
Grade of evidence
The grade evidence generated for the comparison of theoph-
ylline versus placebo for prevention of AKI in neonates with
birth asphyxia was as follows: ‘moderate’ quality for inci-
dence of AKI/severe renal dysfunction, and ‘very low’ to ‘low’
quality for serum creatinine levels on days 1–5 and all-cause
mortality (table 3). This was because of serious study limita-
tions (unclear risk of bias in one or more domains of random
sequence generation, allocation concealment, blinding and
selective reporting), a very high and statistically significant
heterogeneity and the 95% CI includes null effect and/or
appreciable benefit or harm.
7
 Original article

Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
Table 2  Summary of adverse events in the included trials
Method of collection of adverse Priori inclusion of adverse effects in
Study effect data primary/secondary outcomes Adverse effects or complications reported*
13
Jenik et al Routine monitoring. Not mentioned. 13/27 infants in control group and 9/24 in theophylline group developed
seizures. Authors stated no difference in frequency and severity of CNS,
pulmonary, heart and gastrointestinal involvement.
Bakr12 Routine monitoring. Yes. Seizures: 15/20 in placebo and 13/20 in theophylline group.
Authors recorded ventilatory requirement, Pulmonary involvement: 6/20 in placebo and 4/20 in theophylline group.
seizures and other brain involvement.
Bhat et al10 Routine monitoring. Not mentioned. Author stated, ‘No side effects occurs in infants receiving theophylline’.
Figure showed no significant involvement of CNS, cardiac or GIT
between the two groups.
Cattarelli et al17 Routine monitoring. Yes. PDA: 6/24 in placebo and 3/23 in theophylline group.
Complications such as PDA, intraventricular PVL: 2/24 in placebo and 2/23 in theophylline group.
haemorrhage, PVL, ROP, BPD and NEC were ROP: 2/24 in placebo and 2/23 in theophylline group.
considered. BPD: 4/24 in placebo and 1/23 in theophylline group.
NEC: 1/24 in placebo and 0/23 in theophylline group.
Eslami et al11 Spontaneous reporting. Not mentioned. Ventilatory requirement: 7/19 in placebo and 2/17 in theophylline group.
Seizures: 6/19 in placebo and 2/17 in theophylline group.
Cerebral haemorrhage: 1/19 in placebo and 0/17 in theophylline group.
Haematuria: 13/19 in placebo and 7/17 in theophylline group.
Blood transfusion: 3/19 in placebo and 2/17 in theophylline group.
Merrikhi et al14 Questionnaire (whether the Not mentioned. No.
questionnaire contained items for
adverse events not mentioned).
Axelrod et al15 Routine monitoring. Patients with history of seizures, Eleven adverse events occurred in 10 patients in the treatment group
tachyarrhythmias and liver dysfunction were and 13 adverse events occurred in 13 patients in the placebo group.
excluded. Trial was monitored by data safety
monitoring board with access to unblinded
data. No prior inclusion of adverse event in
method section.
Raina et al9 Routine monitoring. Not mentioned. Seizures: 10/81 in placebo group and 9/78 in theophylline group.
Authors stated, ‘No major differences in CNS, pulmonology, GIT
or cardiac were observed between two groups. No adverse effects
attributable to theophylline occurred in the intervention group’.
Ghazala et al19 Not clear. Not mentioned. No.
*Includes both authors’ attempt to detect adverse events as well as complications related to disease process.
BPD, bronchopulmonary dysplasia; CNS, central nervous system; GIT, gastrointestinal tract; NEC, necrotising enterocolitis; PDA, patent ductus arteriosus; PVL, periventricular
leukomalacia; ROP, retinopathy of prematurity.

Discussion
Summary of evidence
An extensive literature search in the present systematic review
revealed nine trials to be eligible for inclusion. Of these, the data
from six trials enrolling 436 neonates who were given prophy-
lactic theophylline as compared with placebo were meta-anal-
ysed. The findings of the present systematic review indicate that
single dose of prophylactic theophylline given to term neonates
with severe birth asphyxia decreases the incidence of AKI/severe
renal dysfunction by 60% (moderate quality evidence) compared
with placebo and decreases serum creatinine levels from days
2 to 5 (very low to low-quality evidence) without significant
difference in mortality between the two groups (very low quality
evidence). For the secondary outcomes, there was a significant
difference in the negative fluid balance in favour of theophylline,
increase in glomerular filtration rate (GFR), decrease in urinary
B2M and without significant difference in complication rate
between the two groups. Thus, the modest transient increase in
diuresis and negative fluid balance associated with theophylline
may be beneficial as they can possibly act as surrogate markers of
hard clinical outcomes such as mortality, ICU and hospital length
of stay, and long-term chronic kidney disease, which needs to be
determined in future studies.28
AKI has been reported to range from 30% to 70% in neonates
with birth asphyxia.29 30 AKI in the perinatal asphyxiated
neonates increases the mortality to as high as 60%, especially
8 Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805
in neonates with low Apgar score at 5 min, oliguria and fluid
overload of more than 20%.31 32
A recent trial9 that recruited 159 severely asphyxiated term
neonates showed that the neonates receiving single prophylactic
dose of theophylline had a lower creatinine levels, had higher
creatinine clearance and significantly decreased risk of AKI/severe
dysfunction. The authors also found a lower mortality rate in the
neonates who received prophylactic theophylline, though the
difference was not statistically significant. A secondary analysis from
the Assessment of Worldwide Acute Kidney Injury Epidemiology
in Neonates (AWAKEN) study found that caffeine administration
in preterm neonates was associated with reduced incidence and
severity of AKI.33
Experimental studies in animals have shown that renal adenosine
act as a vasoconstrictive metabolite in the kidney after hypoxia
causing fall in GFR and filtration factor.34 Thus, non-specific
adenosine receptor antagonists such as theophylline inhibits vaso-
constriction produced by adenosine. Another experimental study
using 8-cyclopentyl-1,3-dipropylxanthine, a specific antagonist of
the A1 adenosine receptors found it to be less effective than theoph-
ylline in preventing hypoxia-induced renal failure, thus implicating
that the later agent may act on the other targets than adenosine A1
receptors.35
Theophylline has wide variation in metabolism and a narrow ther-
apeutic window, and hence, it is essential to measure theophylline
levels to avoid toxicity.36 Only three13 15 17 of nine included studies
 Original article

Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
Table 3  Theophylline versus placebo for prevention of acute kidney injury (AKI) in neonates with birth asphyxia
Patient or population: neonates with severe birth asphyxia.
Settings: neonatal intensive care unit (NICU).
Intervention: theophylline.
Control: placebo.
Illustrative comparative risks* (95% CI)
Assumed risk Relative effect No of participants Quality of the evidence
Outcomes Control Corresponding risk Theophylline (95% CI) (studies) (GRADE)
Incidence of AKI/severe renal 516 per 1000 RR 0.4 436 ⊕⊕⊕⊝ Moderate.†
dysfunction 206 per 1000 (155 to 279). (0.3 to 0.54). (six studies).
Serum creatinine level on day 1 The mean creatinine levels on day 1 ranged from 436 ⊕⊝⊝⊝ Very low.†‡
0.86 to 1.99 in the placebo group. (six studies).
Serum creatinine level on day 2 The mean creatinine levels on day 2 ranged from 161 ⊕⊝⊝⊝ Very low.†§
1.41 to 1.56 in the placebo group. (three studies).
Serum creatinine level on day 3 The mean creatinine levels day 3 ranged from 356 ⊕⊕⊝⊝ Low.†
1.06 to 1.94 in the placebo group. (five studies).
Serum creatinine level on day 4 The mean creatinine levels on day 4 ranged from 161 ⊕⊝⊝⊝ Very low.†§
1.59 to 1.62 in the placebo group. (three studies).
Serum creatinine level on day 5 The mean creatinine levels on 5 ranged from 0.73 436 ⊕⊝⊝⊝ Very low.† §
to 1.57 in the placebo group. (six studies).
All-cause mortality 141 per 1000. 124 per 1000 RR 0.88 356 ⊕⊝⊝⊝ Very low.†¶
(73 to 212). (0.52 to 1.5). (five studies).
GRADE Working Group grades of evidence.
High quality: further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: we are very uncertain about the estimate.
*The basis for the assumed risk (eg, the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk
in the comparison group and the relative effect of the intervention (and its 95% CI).
*†The trials have unclear risk of bias in one or more domains (random sequence generation, allocation concealment, blinding and selective reporting).
‡The 95% CI include null effect.
§A very high and statistically significant heterogeneity is present.
¶The 95% CI includes null effect and appreciable benefit or harm.

in the qualitative synthesis measured the theophylline levels. The
pooled estimate showed no significant difference in the complica-
tion rates between the two groups in the present systematic review.
Strengths and limitations
Strengths: (1) it is an updated systematic review with broader
search question and inclusion of two recent trials; and (2)
rigorous use of GRADE approach to rate quality of evidence.
Limitations: (1) single trial studying the effect of adenosine
antagonists for prevention of AKI in infants undergoing cardiac
surgery, preterm neonates with birth asphyxia and respiratory
distress syndrome; (2) no long-term follow-up beyond neonatal
period in most of the studies; and (3) role of theophylline or
aminophylline not evaluated in neonates receiving therapeutic
hypothermia, which has now emerged as standard of care for
perinatal asphyxia.
Conclusions
This updated meta-analysis showed a benefit (decreased inci-
dence of AKI/severe renal dysfunction) of theophylline in term
neonates with severe birth asphyxia (moderate quality evidence)
without significant adverse effects or effect on the mortality (very
low-quality evidence). However, the long-term effect of these
outcomes as well as potential toxicity if any needs to be inves-
tigated in further trials and in the setting of therapeutic hypo-
thermia for birth asphyxia. More good quality trials reporting all
the outcomes are needed for the beneficial effect of theophylline
in preterm neonates with birth asphyxia and in infants under-
going cardiopulmonary bypass surgery.
Acknowledgements  The authors would like to thank Dr Nishant P Jaiswal,
Scientist C, Indian Council of Medical Research (ICMR) Advanced Centre for Evidence
Based Child Health, PGIMER, Chandigarh, for his help in the database search. GCB
was supported by the ICMR International Fellowship award.
Contributors  GCB designed the research; GCB and PG wrote the paper; GCB, RRD
and MB performed the research; GCB and RRD analysed the data and MB supervised
the paper; all authors read and approved the final manuscript.
Funding  The authors have not declared a specific grant for this research from any
funding agency in the public, commercial or not-for-profit sectors.
Competing interests  None declared.
Patient consent for publication  Not required.
Provenance and peer review  Not commissioned; internally peer reviewed.
References
1 Selewski DT, Charlton JR, Jetton JG, et al. Neonatal acute kidney injury. Pediatrics
2015;136:e463–e473.
2 Levey AS, James MT. Acute Kidney Injury. Ann Intern Med 2017;167:ITC66-
ITC80:ITC66.
3 Holmes J, Roberts G, May K, et al. The incidence of pediatric acute kidney injury is
increased when identified by a change in a creatinine-based electronic alert. Kidney
Int 2017;92:432–9.
4 Bolat F, Comert S, Bolat G, et al. Acute kidney injury in a single neonatal intensive care
unit in Turkey. World J Pediatr 2013;9:323–9.
5 Carmody JB, Swanson JR, Rhone ET, et al. Recognition and reporting of AKI in very
low birth weight infants. Clin J Am Soc Nephrol 2014;9:2036–43.
6 Jetton JG, Boohaker LJ, Sethi SK, et al. Incidence and outcomes of neonatal acute
kidney injury (AWAKEN): a multicentre, multinational, observational cohort study.
Lancet Child Adolesc Health 2017;1:184–94.
7 Bhatt GC, Das RR. Early versus late initiation of renal replacement therapy in
patients with acute kidney injury-a systematic review & meta-analysis of randomized
controlled trials. BMC Nephrol 2017;18:78.
8 Susantitaphong P, Cruz DN, Cerda J, et al. World incidence of AKI: a meta-analysis.
Clin J Am Soc Nephrol 2013;8:1482–93.

Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805 9

 Original article
9 Raina A, Pandita A, Harish R, et al. Treating perinatal asphyxia with theophylline at
birth helps to reduce the severity of renal dysfunction in term neonates. Acta Paediatr
2016;105:e448–51.
10 Bhat MA, Shah ZA, Makhdoomi MS, et al. Theophylline for renal function in term
neonates with perinatal asphyxia: a randomized, placebo-controlled trial. J Pediatr
2006;149:180–4.
11 Eslami Z, Shajari A, Kheirandish M, et al. Theophylline for prevention of kidney
dysfunction in neonates with severe asphyxia. Iran J Kidney Dis 2009;3:222–6.
12 Bakr AF. Prophylactic theophylline to prevent renal dysfunction in newborns
exposed to perinatal asphyxia-a study in a developing country. Pediatr Nephrol
2005;20:1249–52.
13 Jenik AG, Ceriani Cernadas JM, Gorenstein A, et al. A randomized, double-blind,
placebo-controlled trial of the effects of prophylactic theophylline on renal function in
term neonates with perinatal asphyxia. Pediatrics 2000;105:E45:e45.
14 Merrikhi AR, Ghaemi S, Gheissari A, et al. Effects of aminophyllinein preventing
renal failure in premature neonates with asphyxia in Isfahan-Iran. J Pak Med Assoc
2012;62:S48–51.
15 Axelrod DM, Sutherland SM, Anglemyer A, et al. A double-blinded, randomized,
placebo-controlled clinical trial of aminophylline to prevent acute kidney injury in
children following congenital heart surgery with cardiopulmonary bypass. Pediatr Crit
Care Med 2016;17:135–43.
16 Axelrod DM, Anglemyer AT, Sherman-Levine SF, et al. Initial experience using
aminophylline to improve renal dysfunction in the pediatric cardiovascular ICU. Pediatr
Crit Care Med 2014;15:21–7.
17 Cattarelli D, Spandrio M, Gasparoni A, et al. A randomised, double blind, placebo
controlled trial of the effect of theophylline in prevention of vasomotor nephropathy
in very preterm neonates with respiratory distress syndrome. Arch Dis Child Fetal
Neonatal Ed 2006;91:F80–4.
18 Al-Wassia H, Alshaikh B, Sauve R. Prophylactic theophylline for the prevention of severe
renal dysfunction in term and post-term neonates with perinatal asphyxia: a systematic
review and meta-analysis of randomized controlled trials. J Perinatol 2013;33:271–7.
19 Ghazala K, Sidra K, Hamid M, et al. Efficacy of theophylline for prevention of kidney
dysfunction in neonates with severe birth asphyxia. PJMHS;18:1346–8.
20 Schünemann HJ, Schünemann AH, Oxman AD, et al. Grading quality of evidence
and strength of recommendations for diagnostic tests and strategies. BMJ
2008;336:1106–10.
21 Nilsson ME, Koke SC. Defining treatment-emergent adverse events with the Medical
Dictionary for Regulatory Activities (MedDRA). Drug Inf J 2001;35:1289–99.
10
Arch Dis Child: first published as 10.1136/archdischild-2018-315805 on 23 February 2019. Downloaded from http://adc.bmj.com/ on 24 February 2019 by guest. Protected by copyright.
22 ​Handbook.​cochrane.​org. Assessing risk of bias in included studies. http://​handbook.​
cochrane.o​ rg/​chapter_8​ /​8_​assessing_​risk_o​ f_​bias_​in_​included_​studies.​htm
(Accessed 30th Aug 2016).
23 RevMan 5. RevMan 5 | Cochrane Community. /tools/review-production-tools/
revman-5 (Accessed 13th Nov 2017).
24 Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a
simple, graphical test. BMJ 1997;315:629–34.
25 GRADEpro GDT. GRADEpro GDT | Cochrane community. /tools/review-production-
tools/gradepro-gdt (Accessed 26th Aug 2017).
26 Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduction-
GRADE evidence profiles and summary of findings tables. J Clin Epidemiol
2011;64:383–94.
27 Argyropoulos CP, Chen SS, Ng YH, Y-h N, et al. Rediscovering Beta-2 microglobulin as
a biomarker across the spectrum of kidney diseases. Front Med 2017;4.
28 Askenazi D. Should neonates with perinatal asphyxia receive a single dose
of IV theophyline to prevent acute kidney injury? Acta Paediatr Oslo Nor
2016;1992:1125–6.
29 Askenazi DJ, Ambalavanan N, Goldstein SL. Acute kidney injury in critically ill
newborns: what do we know? What do we need to learn? Pediatr Nephrol
2009;24:265–74.
30 Shah P, Riphagen S, Beyene J, et al. Multiorgan dysfunction in infants with post-
asphyxial hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed
2004;89:152F–5.
31 Jayashree G, Dutta AK, Sarna MS, et al. Acute renal failure in asphyxiated newborns.
Indian Pediatr 1991;28:19–23.
32 Durkan AM, Alexander RT. Acute kidney injury post neonatal asphyxia. J Pediatr
2011;158:e29–33.
33 Harer MW, Askenazi DJ, Boohaker LJ, et al. Association between early caffeine citrate
administration and risk of acute kidney injury in preterm neonates: results from the
AWAKEN Study. JAMA Pediatr 2018;172:e180322.
34 Gouyon J-B, Guignard J-P. Theophylline prevents the hypoxemia-induced renal
hemodynamic changes in rabbits. Kidney Int 1988;33:1078–83.
35 Prévot A, Mosig D, Rijtema M, et al. Renal effects of adenosine A1-receptor blockade
with 8-cyclopentyl-1,3-dipropylxanthine in hypoxemic newborn rabbits. Pediatr Res
2003;54:400–5.
36 Day YJ, Huang L, McDuffie MJ, et al. Renal protection from ischemia mediated
by A2A adenosine receptors on bone marrow-derived cells. J Clin Invest
2003;112:883–91.
Bhatt GC, et al. Arch Dis Child 2019;0:1–10. doi:10.1136/archdischild-2018-315805