The present study revealed that a correlation between number of apneic episodes and caffeine serum concentrations was not established under current dosing regimens. A significant association between heart rate and concentrations was found among the other PD parameters, consistent with the fact that the probability of tachycardia increased as caffeine serum concentration increased. A high dose regimen 40/5 mg/kg q12h led to similar re-intubation rate but significantly higher percentage of patients having tachycardia than standard regimen 20/5 mg/kg q24h, agreeing well with the PK/PD relationships found above. The total re-intubation rate of the standard regimen in our patients (24.5 %) was similar to that reported in the literature (24.0 %) [16].
The lack of correlation between efficacy and caffeine serum concentrations under current dosing regimens was in agreement with the trial that granted caffeine citrate label approval by FDA, which reported no association between success of???50 % reduction or elimination in apnea events and mean daily caffeine concentrations.3 Skouroliakou et al.’s study also revealed that methylxanthine concentrations were not significantly associated with number of apneic events per day in neonates with GA??33 weeks [17]. A previous study comparing 10 mg/kg and 5 mg/kg maintenance doses in neonates with GA??32 weeks showed similar efficacy in reducing apnea spells despite significantly higher frequency of tachycardia in the 10 mg/kg group [18]. It could be partially due to the underlying multifactorial etiology of apnea, with prematurity being a prerequisite for the indication [19]. Apnea of prematurity is known to have an incidence inversely related to GA and could regress with the maturation of the newborn [19]. This is echoed in our results that the number of apnea episodes reduced significantly as PMA increased as well as in patients with higher GA (Fig. 2).
The variable PK/PD response in neonates to caffeine citrate therapy is also likely to be attributed to variability in caffeine metabolism in individuals. Caffeine metabolism by hepatic enzymes is usually limited in neonates. Maturation of metabolic enzymes could lead to the improvement in metabolic function, which is significantly associated with the increase in PNA and varies extensively among individuals (range 1 %–41 %) [20]. N7-demethylation, which produces theophylline, acts as the predominant metabolic pathway in premature neonates (range 1 %–37 %) [20]. Theophylline is the active metabolite that is partially responsible for side effects such as tachycardia [18], the variable caffeine-theophylline conversion rate could lead to PD response variability. Several other factors, including genetic variation in hepatic metabolic enzymes and genetic variations in caffeine receptors may also contribute to the variability in PD responses [21–23]. Due to the variability in the caffeine metabolism and the resulting wide range of half-lives among individuals [10], the measurement of caffeine serum concentrations at a postnatal age of 15 (10–24) days in our study population may not necessarily reflect concentrations at steady state. Thus, caffeine concentrations may vary considerably and make it difficult to find other significant associations between PD responses and concentrations.
The 2-fold higher-than-standard dosing regimen led to similar re-intubation rate but significantly higher percentage of patients having tachycardia than standard regimen. This is similar to Steer et al.’s findings on the use of 3–6 fold higher maintenance doses for a course of 7 days in neonates of similar GA range (32 weeks) to our study’s (33 weeks) [24]. A significant reduction in re-ventilation was shown when 4-fold higher maintenance doses were used for the duration of averagely 1 month in neonates with GA??30 weeks [16]. This effect was more evident in the stratified subgroup of neonates with GA??28 weeks that a significant reduction in the days on mechanical ventilator (average 8 days) was observed [16]. A retrospective study in neonates??28 weeks GA revealed that patients receiving caffeine citrate??7.9 mg/kg/day were associated with a decreased need for clinical interventions in terms of dose adjustments compared to those receiving???7.9 mg/kg/day doses [25]. It suggests that neonates with lower GA, especially extremely low-gestational-age neonates (GA??28 weeks), could benefit more from high doses of caffeine citrate. Dosing regimens stratified by GA is warranted for more systematic trial evaluation.
None of the neonates died or had a severe reaction under current dosing regimens, however, TDM may be helpful in suspected toxicity to diagnose caffeine-related adverse events, based on our findings on a significant association between tachycardia and caffeine serum concentrations. The use of high dose caffeine citrate inevitably increases the risk of tachycardia in neonates, this need to be taken into consideration combined with other elements of therapy, such as efficacy and requirement for respiratory support, to determine treatment priority in lieu of medical cost and facility resources available. Caffeine TDM is valuable in the setting of clinically-significant tachycardia to assist differential diagnosis with respect to other potential etiologies.
This study has several limitations associated with its nature of retrospective chart evaluation. First, the TDM was done at the discretion of the medical team and may have patient selection bias towards sicker patients or patients who had adverse events. Second, vital signs were measured during the physical exam in the day of TDM, thus, they were the approximate rather than the exact values at TDM sampling time. Third, the records were physician notes that were not verified by audits of electronic monitoring, however, they are the critical information used by clinicians for patient management and decision making.