Introduction

Therapeutic hypothermia is the standard treatment for hypoxic-ischemic encephalopathy (HIE), but despite its widespread use, the rates of mortality and neurodevelopmental impairment for moderate to severe HIE remain around 30%1. To address this, neonatologists are exploring adjunctive therapies to further improve outcomes. Preclinical studies suggest methylxanthines, such as caffeine and aminophylline, have potential neuroprotective effects in the setting of hypoxic-ischemic injury2,3. Mechanisms by which methylxanthines may offer neuroprotection are a reduction in neuronal cell death and anti-inflammatory and anti-oxidative properties which protect against the cascade of injuries following reperfusion4,5. Furthermore, methylxanthines may protect against acute kidney injury (AKI) associated with HIE, and the Kidney Disease Improving Global Outcomes (KDIGO) guidelines recommend a single dose of theophylline for HIE-related AKI6,7. Data on the safety and efficacy of combining therapeutic hypothermia with methylxanthines in infants with HIE is limited to a few small studies8,9. We evaluated in-hospital outcomes for infants with HIE exposed to concomitant use of methylxanthines and hypothermia.

Methods

We obtained data from the Pediatrix Medical Group Clinical Data Warehouse, encompassing clinical information from > 300 neonatal intensive care units10. We included inborn infants who received therapeutic hypothermia for HIE on postnatal day 0 or 1, received methylxanthines (caffeine or aminophylline) on postnatal day 0–3, and were discharged between 2007 and 2020. We excluded infants with major congenital anomalies. Other data included prenatal characteristics, demographics, medication exposures, in-hospital clinical outcomes, and length of hospital stay. Neuroimaging data were not available for inclusion in this analysis. We performed a literature review of therapeutic hypothermia clinical trials to compare outcomes to this population, if available.

Definitions

We examined the day of initiation and duration of methylxanthine exposure for each infant. We examined creatinine rise > 0.3 mg/dL over 2 days during the first postnatal week per KDIGO criteria11, prothrombin time (PT) > 16 s, aspartate aminotransferase (AST) > 200 U/L and alanine aminotransferase (ALT) > 100 U/L12,13, and mechanical ventilation for ≥ 5 days. We identified exposures to anti-epileptic drugs (AEDs) and examined gastrostomy tube placement prior to discharge.

Statistical analysis

We used standard summary statistics to describe study variables and performed comparisons among infants exposed to only one methylxanthine using Fisher’s exact or Wilcoxon rank sum tests. The use of de-identified data was approved with waiver of consent by the Duke University Health System Institutional Review Board (IRB). The study was performed in accordance with the relevant guidelines and regulations and in accordance with the Declaration of Helsinki.

Results

Among 5909 infants with HIE receiving therapeutic hypothermia between 2007 and 2020, 52 (0.9%) received methylxanthine therapy. Twenty-eight infants received caffeine only, 22 received aminophylline only, and two who received both drugs were included in summary statistics for both cohorts. Four infants (14%) in the caffeine group and 4 infants (20%) in the aminophylline group died, while 10% of infants not exposed to methylxanthines died (Table 1). Infants receiving aminophylline were born at slightly later gestational age (median 38 weeks [25th–75th percentile: 38–39] vs. 36 weeks [35–39]; p = 0.06) and had larger birth weight (3.25 kg [2.54–3.60] vs. 2.77 kg [2.12–3.38]; p = 0.03) than caffeine recipients. Caffeine treatment ranged from 1 to 18 days, while aminophylline treatment ranged from 1 to 4 days. There were no statistically significant differences in incidence of renal failure (41% vs. 19%; P = 0.18) or median peak creatinine levels (1.3 mg/dL [0.9–1.7] vs. 1.0 mg/dL [0.7–1.3]; P = 0.06) in aminophylline-treated infants compared to caffeine-treated infants. There were no significant differences in median baseline creatinine in the caffeine (0.85 mg/dL), aminophylline (0.91 mg/dL), and no methylxanthine (0.90 mg/dL) groups (P = 0.38). The aminophylline and caffeine groups had a similar frequency of AED exposure in the first postnatal week (42% vs. 37%; P = 0.77).

Table 1 In-hospital mortality and morbidities in infants with HIE who received methylxanthines.
Full size table

Discussion

Previous studies investigating methylxanthines in the setting of HIE have primarily focused on kidney outcomes. A meta-analysis of six trials (n = 436) conducted in term infants with severe birth asphyxia in settings which did not have therapeutic hypothermia available found that a prophylactic dose of theophylline reduced AKI incidence compared to placebo without increasing mortality or seizures14. A retrospective study of 16 infants with HIE receiving therapeutic hypothermia found that early aminophylline use was associated with increased urine output and lower serum creatinine8. Finally, a prospective, open-label trial of caffeine in 17 infants with HIE receiving therapeutic hypothermia at a single center found an 18% incidence of AKI with frequencies of mortality and seizures similar to those reported in the intervention arm of the Whole-Body Hypothermia for HIE trial9,12.

The current study provides further insights into methylxanthines use in infants with HIE undergoing therapeutic hypothermia. The frequency of mortality and morbidities were similar to those of infants undergoing therapeutic hypothermia without adjunctive therapies in clinical trials12,13,15,16,17. Of note, the most recent clinical trials demonstrated lower mortality in the standard hypothermia arm compared to earlier trials, likely due to the widespread implementation of hypothermia as the standard of care for treating HIE15,17. While inconsistent definitions of renal failure across studies make direct comparisons challenging, a large cohort study utilizing a similar definition as our study found a frequency of AKI in 38% of infants receiving hypothermia for HIE18. Because the precise timing of methylxanthine administration in our population is not known, we were not able to report whether the medication was administered soon after delivery as recommended by the KDIGO guidelines6. While we are unable to determine the indication for methylxanthine use in this dataset, the shorter duration of aminophylline use and the higher peak creatinine in the aminophylline group compared to the caffeine group may reflect a decision by the clinician to use aminophylline for AKI prevention or treatment, while caffeine may have been used for respiratory support to prevent or treat apnea. The increase in mortality and length of stay in the methylxanthine group compared to infants who did not receive methylxanthines likely reflects confounding by indication whereby infants with more severe disease may have been more likely to receive the intervention.

Considering caffeine's association with a reduced seizure threshold and the FDA label's mention of seizures as a potential side effect of caffeine overdose, we explored AED receipt as a proxy for clinical seizures and found that AED exposures were similar to those reported in trials19,20. While our study leveraged observational data from an extensive clinical database, shedding light on methylxanthine administration during therapeutic hypothermia for HIE, some limitations must be acknowledged. Most importantly, we were not able to capture MRI brain results from the database in infants exposed to methylxanthines. Studies investigating the therapeutic utility of methylxanthines in the setting of HIE will need to include neuroimaging and neurodevelopmental follow up data. Missing data and infant transfers limited our ability to ensure complete outcome uniformity. In addition, the uncertainty in dosing, timing, and indication of caffeine and aminophylline administration in our study limits our ability to draw conclusions on the utility of methylxanthines in the setting of HIE.

Conclusion

Methylxanthines have potential neuroprotective effects in the setting of HIE. In this retrospective study of infants exposed to methylxanthines and therapeutic hypothermia, we found that in-hospital mortality and morbidities were similar to those described in trials of infants receiving therapeutic hypothermia alone. Clinical trials are needed to ascertain the safety and efficacy of methylxanthines in infants undergoing therapeutic hypothermia for HIE.