INTRODUCTION. The main action of paracetamol (acetaminophen) is presumed to be in the central nervous system. The central nervous system penetration of paracetamol has been described in children with intracranial pathologies but not in children with an intact blood-brain barrier.
OBJECTIVE. We investigated the cerebrospinal fluid penetration of paracetamol in 32 healthy children, aged 3 months to 12 years, who were undergoing surgery in the lower body using spinal anesthesia.
MATERIALS AND METHODS. In this open-label prospective study, children were given a single intravenous injection of paracetamol (15 mg/kg). Cerebrospinal fluid and venous blood samples were obtained between 5 minutes and 5 hours after injection. Paracetamol concentrations were determined from the cerebrospinal fluid and plasma by using a fluorescence polarization immunoassay.
RESULTS. Paracetamol was detected in cerebrospinal fluid from the earliest sample at 5 minutes, although in this sample paracetamol concentration was below the limit of quantification of 1.0 mg/L. Subsequent paracetamol concentrations in cerebrospinal fluid ranged between 1.3 and 18 mg/L (median: 7.2 mg/L), plasma concentrations ranged between 2.4 and 33 mg/L, and cerebrospinal fluid/plasma ratios ranged between 0.06 and 2.0. The highest CSF paracetamol concentration was detected at 57 minutes.
CONCLUSIONS. Paracetamol permeates readily into the cerebrospinal fluid of children. This fast and extensive transfer enables the rapid central analgesic and antipyretic action of intravenous paracetamol.
Paracetamol (acetaminophen) is the most commonly used antipyretic analgesic in children for the symptomatic treatment of acute pain and fever. At sufficient enteral dosage, paracetamol is effective alone for the treatment of mild and moderate pain,1 but in the acute situation, the intravenous preparation is more convenient and may perform better.2,3 With intravenous administration, the onset of analgesic and antipyretic action is rapid, with the analgesic action occurring within 15 minutes4 and fever reduction occurring within 30 minutes.5
The mechanism of action of paracetamol has not been fully established.6 However, paracetamol has analgesic and antipyretic properties and weak antiinflammatory activity, thus it is likely that analgesic action may be attributable to inhibition of prostaglandin synthesis in the central nervous system (CNS) and in peripheral tissues.7,8 The analgesic action of paracetamol is likely to be linked also with the serotonergic system.9 Serotonin has an important role in pain modulation in the CNS. In adult volunteers the serotonin receptor 5-HT3 antagonists tropisetron and granisetron completely block the analgesic effect of paracetamol.10
To have effects within the CNS, paracetamol must penetrate the blood-brain barrier (BBB). The penetration of paracetamol into cerebrospinal fluid (CSF) was evaluated in adults,11,12 as well as in 2 studies in children.13,14 However, the children studied had significant pathology, such as head trauma, raised intracranial pressure, or tumors, which may affect the BBB permeability. We designed this study to investigate the CSF penetration of intravenous paracetamol (15 mg/kg) in healthy children undergoing surgery in the lower part of the body using spinal anesthesia.
MATERIALS AND METHODS
The study protocol was approved by the research ethics committee of the Hospital District of Northern Savo, Kuopio, Finland (No. 120/2004). The Finnish National Agency for Medicines was notified (No. 161/2004); the trial was recorded in the EudraCT database (No. 2004-001702-27), and it was conducted in accordance with the latest revision of the Declaration of Helsinki. The parents and children, if old enough to understand the trial protocol and the interventions, were informed, and the parents gave written consent and children assented.
Thirty-four children were asked to participate, but the parents of 2 children refused to consent (not wanting anything “extra” for their child), hence leaving a group of 32 children (19 boys and 13 girls) in the study population. The children were scheduled for genitourologic surgery (15 children), herniotomy (11 children),11 and orthopedic surgery (6 children),6 all to be performed under spinal anesthesia. Children were excluded if they had contraindications to the use of paracetamol or spinal anesthesia
The children were premedicated with buccal midazolam (0.375 mg/kg up to 7.5 mg) and ketamine (1.25 mg/kg up to 25 mg) 15 to 30 minutes before anesthesia. All children were sedated with midazolam, thiopental, and/or propofol before lumbar puncture for spinal anesthesia (n = 28) or combined spinal-epidural anesthesia (n = 4).
The children received an intravenous infusion of paracetamol (15 mg/kg) (Perfalgan 10 mg/mL, lot 5H00568, expiration date August 2007 [Bristol-Myers Squibb AB, Bromma, Sweden]) over 10 minutes into a dorsal hand vein 5 minutes to 5 hours before lumbar puncture. One mL of CSF was collected into a polypropylene tube during lumbar puncture before the injection of local anesthetic. An indwelling catheter was inserted in a dorsal foot vein, and a 3-mL blood sample was obtained for paracetamol assay into a heparinized tube. Plasma was obtained by centrifugation at 3000g at 20°C for 10 minutes. The plasma was divided into 2 polypropylene tubes to obtain 2 samples of at least 0.5 mL each. One sample was tested, and another was kept as a control. The plasma and CSF were protected from light and stored at −72°C.
After the surgery, children were transferred to the postanesthesia care unit (PACU) for monitoring of vital signs, pain, and adverse effects. The pain intensity at rest and with movement was assessed by a research nurse on an 11-point numeric rating scale (0 [no pain] to 10 [worst possible pain]) and was recorded at every hour after the end of surgery. After the surgery, the children received an intravenous injection of ketoprofen (1 mg/kg). If the child was in pain (pain score at rest ≥3 or with movement ≥5), fentanyl (1 μg/kg, intravenously) or oxycodone (0.05 mg/kg, intravenously) was given for rescue analgesia.
Paracetamol concentrations in plasma and CSF samples were determined in 1 run using fluorescence polarization immunoassay technology (TDxFLx; Abbott Laboratories, Abbott Park, IL). The sensitivity of the paracetamol assay, defined as the lowest measurable concentration that can be distinguished from 0 with 95% confidence, was 1.0 mg/L. Within-run variations for controls at paracetamol concentrations of 15, 35, and 150 mg/L were 4.9% (n = 4), 3.6% (n = 4), and 3.8% (n = 4), respectively.
No formal sample size calculation was performed, but a sample of 30 children was considered to provide sufficient information on CSF penetration of paracetamol in healthy children.
Data were entered and analyzed with the SPSS 13.0 (SPSS Inc, Chicago, IL). Correlations between paracetamol concentrations and patient characteristics were tested with the Pearson correlation test. A gender difference in mean CSF-paracetamol concentration was tested with independent samples t tests, including Levene's test for equality of variances, and both pooled- and separate-variances t tests for equality of means. A P value of .05 was considered as the limit of statistical significance.
The patient characteristics are presented in Table 1. There were few minor protocol deviations that were unlikely to affect the study results: 1 child (patient 1) had paracetamol (20 mg/kg), 1 child (patient 4) had paracetamol (12.5 mg/kg), and for 2 children (patients 26 and 15), the infusion time was 30 and 40 minutes, respectively. All other children received the study medication as defined in the protocol, and all CSF and plasma samples were collected as defined in the protocol. In 1 child (patient 5), the CSF sample was blood stained, but the paracetamol concentration was similar to others and, therefore, it was included in the analysis. All the other CSF samples were visually clear.
The individual sampling times and paracetamol concentrations are presented in Table 2 and Fig 1. Paracetamol was detected from the earliest CSF sample taken at 5 minutes after injection, but in this sample the concentration was below the sensitivity of the assay (1.0 mg/L). In the other CSF samples, paracetamol concentrations ranged between 1.3 and 18 mg/L (median: 7.2 mg/L), and plasma concentrations were between 2.4 and 33 mg/L (14 mg/L).
The CSF to plasma concentration ratio ranged between 0 and 2 (median: 0.8). There was a positive correlation between sampling time and CSF to plasma concentration ratio (r = 0.89; P < .001) (Fig 2).
There was no correlation between CSF paracetamol concentration and age, height, or weight. However, in a posthoc analysis there was a significant gender difference in the CSF paracetamol concentrations, with girls having higher concentrations than boys (P = .001) (Fig 3).
Six children developed adverse effects: 3 children were agitated in the recovery room, 1 vomited, 1 had nausea, and 1 developed shivering. None of the children complained of pain during paracetamol injection.
For pain treatment in the PACU, the children received ketoprofen (1 mg/kg intravenously; n = 17), naproxen (5 mg/kg by mouth; n = 2), ibuprofen (10 mg/kg by mouth; n = 1), or a second dose of paracetamol (15 mg/kg by mouth; n = 13). Four children had an epidural infusion. Ten children had significant pain in the PACU (numeric rating scale >3 at rest or >5 with movement), and they were given a single dose of opioid: fentanyl (1 μg/kg; n = 5) or oxycodone (0.05 mg/kg; n = 5) intravenously for rescue analgesia.
Our study indicates that paracetamol enters readily through an intact BBB in children. Paracetamol was detected from the earliest CSF samples, and only in the CSF sample taken 5 minutes after the intravenous injection was paracetamol below the lower limit of quantification 1.0 mg/L. The highest CSF-paracetamol concentration, 18 mg/L, was measured 57 minutes after the injection. Thereafter, CSF and plasma concentrations were similar, with CSF/plasma ratios between 0.78 and 2. This relatively rapid CNS penetration enables the fast analgesic and antipyretic onset of intravenous paracetamol observed in clinical studies.2–5
The CNS penetration of paracetamol in children was measured in 2 previous reports, which found that similar peak CSF concentrations are attained with enteral paracetamol doses 2 to 3 times higher than our study using intravenous paracetamol (15 mg/kg). Anderson et al13 administered paracetamol (40 mg/kg) by nasogastric tube to 9 ventilator-dependent children with head trauma or other CNS pathology. In these 9 children, the highest CSF and plasma concentrations (20 and 21 mg/L, respectively) were similar to those detected in the present study (18 and 33 mg/L, respectively). However, the peak CSF paracetamol concentration was attained at 210 minutes in that study in contrast to our study where the peak CSF concentrations were detected 57 minutes after intravenous injection. In another study by Anderson's group,14 paracetamol (30 mg/kg) was administered rectally to 41 children undergoing insertion or revision of ventriculoperitoneal shunt or insertion of ventricular drain. The highest CSF and plasma paracetamol concentrations observed 2 hours after the paracetamol administration were 21 and 33 mg/L, respectively, which are similar to those observed in our study.
The mechanism by which paracetamol exerts its antinociceptive effects is not established, but recent studies indicate that the analgesic action of paracetamol may be multimodal and that activity in the CNS are essential for its analgesic action. In the CNS, paracetamol may act through several different pathways. First, it was shown that paracetamol attenuates prostaglandin synthesis through a weak cyclooxygenase inhibition,15 and there is evidence to suggest both peripheral16 and central sites17 of action that may involve inhibition of cyclooxygenase. Second, animal studies indicate that paracetamol antinociceptive action may also involve spinal nitric oxide pathways, which are associated with spinal glutamate N-methyl-D-aspartate receptor activation.18 Finally, both animal and human experimental pain models consistently indicate that paracetamol acts at the CNS by serotonergic mechanisms.10
In previous trials in children, we evaluated the CSF penetration of 2 nonsteroidal antiinflammatory drugs (NSAIDs), indomethacin19 and ketoprofen.20,21 The onset of analgesic action of these 2 NSAIDs in children is fast, and both also penetrate readily into the CSF, because high CSF concentrations are detected early.19–21 After intravenous indomethacin, CSF concentrations are uniform from the first minutes after injection up to 4 hours, which could explain both the rapid analgesic action and the high incidence of CNS adverse effects.
However, there seem to be a major difference in BBB penetration between paracetamol and NSAIDs. Forty-fifty minutes after intravenous administration, paracetamol concentrations in CSF are similar or higher to those in plasma. On the contrary, because of a high degree of plasma protein binding with indomethacin and ketoprofen, CSF concentrations are <0.1% of total plasma concentrations19–21 but similar to protein-free plasma concentrations. Thus, although paracetamol activity on cyclooxygenase is weak,15 we suggest that the relatively high CSF concentrations are sufficient to inhibit prostaglandin E2 in the CNS to a degree comparable to NSAIDs.
Our study shows that in children paracetamol may penetrate the BBB more readily than in adults, which may explain the effective analgesic action of paracetamol in pediatric pain management. Bannwarth et al11 evaluated CSF penetration of intravenous propacetamol (2 g), a prodrug of paracetamol, in 43 healthy adults having lumbar puncture for myelography. At this dose, corresponding to 1 g of paracetamol (15 mg/kg), peak CSF paracetamol concentration was found at 45 minutes, which is comparable to 57 minutes in our study. However, the peak CSF concentration of 9 mg/L was only half of that observed in children.13,14 In our study, 12 of 32 children had CSF-paracetamol concentrations >9 mg/L, which was the highest concentration observed in adults.
In our study, an interesting finding was that girls developed significantly higher CSF paracetamol concentrations than boys. In adults, gender differences in drug response and adverse effects were noted, but it is unclear whether these differences are pharmacokinetic or pharmacodynamic.23 The present finding warrants additional studies on gender differences in pharmacokinetics of analgesics in pediatric populations.
In our study, intravenous paracetamol was well tolerated. Six children developed adverse effects. Most common was emergence agitation, which occurred in 3 children. Emergence agitation is common in children after inhalation anesthesia24 but rarely reported after spinal anesthesia.25 Paracetamol-related agitation was not reported earlier, and in our present study it was more likely with the associated drugs given for performing the spinal anesthesia (midazolam, thiopental, or propofol). One child experienced shivering in the PACU, and this was probably because of the spinal anesthesia used.26
It is considered unethical to recruit healthy volunteers for studies on pediatric pharmacology; furthermore, performing a lumbar puncture involves risks. We collected the CSF samples during an already-indicated lumbar puncture. The volume of the sample (1 mL of CSF) was always less than the volume of injected local anesthetic. Therefore, the study protocol was well incorporated into the standard care in our institution. Spinal anesthesia is routinely used in our department, and all children having surgery are provided nonopioid analgesic for the prevention and treatment of postoperative pain. We believe that this should have been the reason why most parents and children asked (32 of 34) consented to participation.
Our study indicates that paracetamol (acetaminophen) permeates readily into the CSF of healthy children. The relatively fast and extensive CNS permeation enables the rapid CNS analgesic and antipyretic effects of intravenous paracetamol.
- Accepted December 6, 2006.
- Address correspondence to Hannu Kokki, MD, PhD, Department of Anesthesiology and Intensive Care, PO Box 1777, FI-70211 Kuopio, Finland. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
Some of these results were presented at the XXV Annual ESRA Congress; September 6–9, 2006; Monte Carlo, Monaco (see Kokki H, Kumpulainen E, Heikkinen M, Laisalmi M, Halonen T. Reg Anesth Pain Med. 2006;31:56).
- ↵Moller PL, Sindet-Pedersen S, Petersen CT, et al. Onset of acetaminophen analgesia: comparison of oral and intravenous routes after third molar surgery. Br J Anaesth.2005;94 :642– 688
- ↵Boutaud O, Aronoff DM, Richardson JH, Marnett LJ, Oates JA. Determinants of the cellular specificity of acetaminophen as an inhibitor of prostaglandin H(2) synthases. Proc Natl Acad Sci USA. 2002;99 :7130– 7135
- ↵Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ, Vane JR. Selectivity of nonsteroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci USA.1993;90 :11693– 11697
- ↵Colletti AE, Vogl HW, Rahe T, Zambraski EJ. Effects of acetaminophen and ibuprofen on renal function in anesthetized normal and sodium-depleted dogs. J Appl Physiol.1999;86 :592– 597
- ↵Bjorkman R. Central antinociceptive effects of non-steroidal anti-inflammatory drugs and paracetamol: experimental studies in the rat. Acta Anaesthesiol Scand.1995;103(suppl) :1– 44
- ↵Mannila A, Kumpulainen E, Lehtonen M et al. Plasma and cerebrospinal fluid concentrations of indomethacin in children after intravenous administration. Clin Pharmacol.2007;44 :94– 100
- ↵Ciccone GK, Holdcroft A. Drugs and sex differences: a review of drugs relating to anaesthesia. Br J Anaesth.1999;82 :255– 265
- Copyright © 2007 by the American Academy of Pediatrics