- Email: [email protected]
Morphine metabolism in acutely ill preterm newborn infants
Morphine metabolism in acutely ill preterm newborn infants R a m a Bhat, MD, M o h a m a d A b u - H a r b , MD, G o p a l Chari, PhD, a n d Anil Gulati, MD, PhD From the Division of Neonatology, Department of Pediatrics, and the Department of Pharmacodynamics, University of Illinois at Chicago, Chicago, Illinois To e x a m i n e the manner in which morphine is m e t a b o l i z e d in acutely ill premature infants, we measured the levels of morphine, morphine-3- and -6-glucuronides, and c o d e i n e in timed urine specimens and paired plasma specimens at 4 hours and 24 hours after a single dose of morphine in 16 preterm infants (<32 weeks of gestational age). A large amount of unmetabolized morphine was found in the urine in 13 (8t.2%) of the 16 infants at 4 hours; in 12 of them, morphine was e x c r e t e d even at 24 hours. Urinary morphine levels varied greatly; the coefficient of variation was 130% at 4 hours and 1t8% at 24 hours. C o d e i n e was not found in any of the infants. In t0 (62.5%) of the t6 infants, at least one m e t a b o l i t e was found in either plasma or urine. Plasma and urinary levels of morphine conjugates also varied greatly a m o n g these 10 infants (coefficient of variation: 109% to 3t7%). All six infants (37.5%) who had no metabolites e x c r e t e d large amounts of u n m e t a b o l i z e d morphine in the urine for up to 24 hours. Birth weight, gestational age, postnatal age, systemic b l o o d pressure, and other clinical or physi o l o g i c variables did not differ significantly b e t w e e n the t0 infants who had morphine conjugates and the six who did not. We c o n c l u d e that (t) nearly two thirds of a c u t e l y ill preterm infants born at <32 weeks of gestational a g e conj u g a t e morphine; (2) irrespective of their ability to p r o d u c e morphine conjugates, preterm infants e x c r e t e large amounts of morphine unmetabolized, as late as 24 hours after a single dose; (3) morphine handling patterns are highly variable a m o n g premature infants, and no obvious factors a c c o u n t for the variability; and (4) such variability in morphine handling in general, and the production of the highly potent morphine-6-glucuronide in particular, could explain the variance in morphine p h a r m a c o k i n e t i c measures and in the clinical responses to morphine during the newborn period. (J PEDIATRt992;t20:795-9)
Morphine is one of the most frequently used drugs for neonatal analgesia,1 and yet its metabolism in newborn infants has not been well studied. Adults conjugate morphine predominantly as morphine-3-glucuronide and, to a small extent, as a highly potent analgesic, morphine-6-glucuronidefl Preliminary work in this area suggests that premature infants can conjugate morphine, 3 but the extent of conjugation and the magnitude of various subfractions of morphine Supported in part by a grant from the St. Joseph Hospital and Health Care Center, Chicago, Ill. Submitted for publication Sept. 24, 1991; accepted Dec. 20, 1991. Reprint address: Rama Bhat, MD, Department of Pediatrics (m/c 856), 840 South Wood St. Chicago, IL 60612. 9/25/35837
metabolites have not been studied in acutely ill premature infants. We noted that, in comparison with term infants, infants born at <30 weeks of gestational age had prolonged morphine half-life and cleared the drug more slowly. The intersubject variability for these pharmacokinetic meal
M3G M6G CV
Morphine-3-glucuronide Morphine-6-glucuronide Coefficientof variation
]
sures was very large4; we speculated that these findings reflect variations in morphine handling among preterm in, fants. To explore further the metabolic fate of morphine in acutely ill preterm infants after administration of a single 795
796
B h a t et al.
The Journal o f Pediatrics May 1992
Table I. Plasma and urine metabolites in 16 preterm
dose, in this study we measured its major metabolites in the blood and urine, using a modified, extremely sensitive, high-pressure liquid chromatography assay. We sought to determine (1) the predominant morphine metaholite in preterm infants, (2) whether preterm infants produce M6G, (3) to what extent morphine is excreted unmetabolized in the urine, and (4) the reasons for the intersubject variability in morphine handling.
20 to 24 hours after morphine dosing. The bladder was emptied by the Cr6d6 technique before the dose, and a pediatric urine collection bag (U-Bag; Hollister Inc., Libertyville, Ill.) was placed for urine collection. One milliliter of blood was obtained via the indwelling catheter or from an arterial puncture, 4 hours and 24 hours after the dose. Blood samples were spun, and the plasma was separated and stored at - 7 0 C ~ along with the paired urine samples, for later analysis of morphine, M3G, M6G, and codeine. Morphine, M3G, M6G, and codeine were measured by high-performance liquid chromatographic method developed in our laboratory. For morphine and its metabolites, the lower limit of assay sensitivity was 1 ng/ml. The details of the assay have been recently reported. 5 The data for the groups were compared by means of the t test for normally distributed continuous variables and Mann-Whitney U test for highly skewed continuous variables. A paired t test was used to compare the changes in respective measures between 4 and 24 hours in the total groups and within subgroups. A p value of <0.05 was considered to be statistically significant. Data are presented as mean _+ SD, and where required the coefficient of variation (CV) (100 X SD/ mean) values are provided.
METHODS
RESULTS
Subjects. Preterm infants admitted to the intensive care nursery and receiving mechanical ventilation were eligible for the study if they required a single intravenously administered dose of morphine for a minor surgical procedure, such as percutaneous venous catheter insertion or removal of a chest tube. The decision to administer morphine was made by the attending neonatologist. The study protocol was approved by the institutional review board, and written informed consent was obtained from the parents. Two additional infants who received multiple doses of morphine for sedation along with pancuronium bromide also were studied. Because these infants were mature and received multiple doses, their data are presented separately. Morphine administration. During administration of morphine, the heart rate and blood pressure were monitored continuously by means of an indwelling arterial catheter and electronic monitor in all except three infants, in whom noninvasive measurements were obtained. The intravenously administered morphine dose was 50 #g/kg for infants weighing <750 gin, and 100 g g / k g for infants weighing >750 gin. Morphine sulfate was diluted 1:5 to 1: 10 in sterile water; thus the final concentration was 0.2 to 0.4 mg/ml. For each infant the dosing vials were prepared separately; after morphine administration via the peripheral vein, the infusion line was flushed to ensure complete drug delivery. Two sets of 4-hour timed urine samples were collected from each subject, the first up to 4 hours and the other from
Clinical data. We studied 16 preterm infants (8 male infants) whose mean birth weight was 843 _+ 332 gm (range 690 to 1220 gm) and gestational age 28.2 _+ 2.2 weeks (range 25 to 32 weeks). The postnatal age at study was 9.7 _+ 17 days; 13 of the infants were <5 days of age, and the other three were 19, 31, and 66 days of age, respectively. The extent of assisted ventilatory support required by the 16 infants varied. The mean dose of morphine administered was 91.8 _+ 18.7/~g/kg. Plasma and urinary morphine concentrations (Table I). In 13 of the infants (81.2%), plasma morphine concentrations were detected at 4 hours, and 12 of these had detectable levels at 24 hours. The 24-hour mean plasma morphine concentration was significantlylower than the 4-hour mean concentration. In two of the three infants who had no plasma morphine at 4 hours, the 24-hour plasma specimens revealed small amounts of morphine (6.1 ng/ml and 9.7 ng/ml). Large quantities of unmetabolized morphine (3% to 15% of the administered dose) were found in the urine of 13 (81.2%) of the 16 study infants at 4 hours, and in 12 (75%) at 24 hours. The mean urinary morphine levels at 24 hours were significantly lower than at 4 hours. The intersubject variability was very large, with CV of 130% at 4 hours, a n d 118% at 24 hours. Metabolites. Large quantities of M3G were detected in the plasma and urine (Table I), even 24 hours after the initial dosing. The urinary M3G values were significantly
infants Variables
4-Hr values
Morphine (ng/ml) Plasma 34.1 + 20.2 Urine 371.2 _+ 484 M3G (ng/ml) Plasma 11.3 _+ 22.1 Urine 17.5 _+ 27.5 M6G (ng/ml) Plasma 0 Urine 18.2 + 52.7
24-Hr values
5.0 _+ 5.5* 18.8 _+ 2.3t 3.7 _+ 9.9 7.3 _+ 24.2t 0.8 + 3.1 1.4 + 3.5
*p <0.001,as comparedwiththe corresponding4-hourvalues(paired t test). tP <0.05,as comparedwiththe corresponding4-hourvalues(paired t test).
Volume 120 Number 5
lower at 24 than at 4 hours (p <0.05). The CV values for plasma and urine M3G at 4 and 24 hours ranged between 157% and 222%. At 4 hours M6G was not detectable in the plasma, but a small amount was detected at 24 hours. However, amounts of this potent analgesic were significant in the urine at 4 hours and dropped to a lower level by 24 hours in all patients; however, the variability in the urine M6G level was very large--CV was 290% at 4 hours, and 400% at 24 hours. Subgroup analyses. Of the 16 study infants, 10 (62.5%) had one or both of the two conjugates. Six of these ten had only M3G, three had only M6G, and the other had both conjugates. Thus M3G was found in 7 (43.75%) of the 16 infants, and M6G in 4 (25%). There was no difference between the infants who were <5 days of age and those >2 weeks of age with regard to the frequency of morphine metabolites (Table II). None of the infants had codeine in plasma or urine. All of the six infants (37.5%) who had no morphine metabolites either in the plasma or in the urine at any time during the study were <5 days of age (Table II). To evaluate the reasons for the differences in morphine handling, we analyzed study variables in two subgroups--one having morphine metabolites and the other having no metabolites (Table III). The dose of morphine did not differ significantly between these subgroups. Both groups of infants excreted large amounts of morphine at 4 hours. Except for the absence of metabolites in the first subgroup, mean morphine concentrations in the plasma and the morphine excretion patterns in the urine at 4 and 24 hours did not differ significantly between the two subgroups (Table III). In both subgroups the 24-hour values for plasma and urine morphine were significantly lower than the corresponding 4-hour values. The CV values also were large in both subgroups. The birth weight, gestational age, male/female ratio, and postnatal age (despite the inclusion of three older infants in one group) did not differ significantlybetween the subgroups. Systemic blood pressure and arterial blood gas values were normal in both groups, and the frequencies of maternal administration of steroids, phenobarbital, or tocolytic agents and of maternal substance abuse (heroin and cocaine) also were similar. Thus we could find no factor that accounted for the presence or absence of morphine metabolites. Among the 10 infants who had metabolites, there were no differences in the mean morphine concentrations (in urine and plasma) and in the concentrations of morphine metabolites between the seven infants who were <5 days and the three who were >2 weeks of age (data not shown). Multiple doses. The two infants who received multiple morphine doses were both 1 day of age, were born at 36 and 39 weeks of gestational age, and at birth weighed 2519 and 3550 gm. As in the 16 index cases, they received an initial morphine dose, 100 t~g/kg, before the 4-hour sample
Morphine in acutely ill preterm infants
797
Table II. Frequency of morphine metabolites Variables
A g e I-5 d a y s (n = 13)
A g e > 2 wk (n = 3)
All c a s e s (n = 16)
Codeine Only M3G Only M6G M3G and M6G No metabolites
0 4 2 1 6
0 2 1 ---
0 6 (37.5%) 3 (18.8%) 1 (6.2%) 6 (37.5%)
Valuesare frequenciesand (%) of totalcases. The frequenciesincludepresence of metabolitein plasmaor urineor both at 4 and/or 24 hoursafter a singledose of morphine. collection. Subsequently, because of a need for sedation, they were given two or three additional doses of morphine before the 24-hour measurements. Four hours after the initial dose, the unmetabolized urine morphine value was 1100 ng/ml in the term infant and 52.2 ng/ml in the 36-week preterm infant; these levels dropped to 12.7 ng/ml in the term infant and 6.3 ng/ml in the preterm infant. The term infant excreted 89.2 ng of M3G and 28.6 ng of M6G per milliliter at 4 hours, and the concentrations dropped to undetectable and 2.1 ng/ml, respectively, at 24 hours. In the preterm infant, no detectable metabolites were found in plasma or urine, either in the 4-hour samples or in the 24hour samples. DISCUSSION We have shown that about one third of acutely ill infants born at <32 weeks of gestational age did not conjugate morphine but eliminated it unmetabolized in the urine. The infants who did conjugate morphine also eliminated large quantities of unconjugated drug. The predominant metabolite was M3G, but the potent M6G was found in 25% of infants. Marked variability in most of the measured morphine metabolites was the rule rather than the exception. No obvious clinical or physiologic characteristics were associated with the great variability in morphine handling. No infant metabolized morphine to codeine. Choonara et al. 6 examined morphine metabolites in premature infants receiving a continuous infusion of morphine and found M3G to be the predominant metabolite. Although M6G was found in some urine samples, the infants' gestational ages were not provided. Further, the assay for M6G used by these investigators was sensitive to 40 ng/ml, in contrast to a sensitivity of 1 ng/ml in our method. Another study showed that fetal lambs rapidly metabolize morphine to large quantities of M3G, but not to M6G. 7 In adults, morphine is conjugated to M3G (54% to 74%), M6G (<1%), and as ethereal sulfate 0.5% to 5%. Additionally, about 5% is oxidized to normorphine, 1.6% is methylated to codeine, and 7.5% to 12% is excreted in the urine unmetabolized. 8 The elimination and biologic effects of metabolites could
798
Bhat et al.
The Journal of Pediatrics May 1992
Table llh Plasma and urine metabolites in the subgroups Metabolites present (n=40) Variables
Morphine (ng/ml) Plasma Urine M3G (ng/ml) Plasma Urine M6G (ng/ml) Plasma Urine Morphine dose (#g/kg) Birth weight (gm) Gestational age (wk) Study age (days)
No metabolites (n=6)
4 Hours
24 Hours
4 Hours
32.1 + 25.4 454.3 + 537.1
4.9 + 6.5* 14.5 _+ 22.2t
37.5 _+ 5.9 233.0 _+ 384.3
18.0 + 26.1 28.0 _+ 30.5
5.9 + 9.9 11.6 + 16.6
0 0
0 0
0 29.0 + 65.4
0.8 _+ 3.1 2.3 _+ 4.3
0 0
0 0
96.0 842 28.2 13.4
+ 9.9 _+ 295 _+ t.9 + 20.9
24 Hours
5.14 _+ 4.1t 25.9 _+ 22.6"t"
85.0 951 28.2 2.8
_+ 27.9 -+ 211 + 2.8 + 1.8
*p <0.01. Paired t test between corresponding values at 4 hours and 24 hours within each subgroup. Plasma and urinary morphine and postnatal age were not significantlydifferent between subgroups (Mann-Whitney O test), and other clinical variables also were similar (Student t test). tP <0.05. Information as for p <0.01, above. influence the magnitude and duration of the drug effect. For instance, M 3 G is not an analgesic but has an excitatory effect and can cause seizures. 7 On the other hand, M 6 G is a highly potent analgesic and sedative--more potent than morphine itself. 2, 9 High concentrations of M 6 G have been detected in the cerebrospinal fluid of adults receiving morphine. I~ Because morphine glucuronides have prolonged terminal elimination half-lives, even small amounts of M 6 G in the plasma could significantly enhance the analgesic effect of the drug, particularly because of its blood-brain barrier permeability. A number of factors might modulate morphine metabolism. Although neonates have low or poor conjugating capacity, which improves during the first few days, we could not demonstrate an effect of postnatal age on morphine metabolism; a larger population sample with a greater range of postnatal ages will be needed to confirm this observation. Similarly, between 25 weeks and 32 weeks, we found no evidence of an effect of gestational age on either the frequency or extent of morphine disposition. In addition, of the two mature infants who received multiple doses, only one had evidence of morphine conjugation, but both excreted unmetabolized morphine. Although all infants were physiologically stable at the time of our study, they were acutely ill and required ventilatory support; other factors not examined by us could have influenced the rate and extent of morphine handling. Therefore our findings cannot be generalized to all preterm infants or to all neonates. From a practical point of view, however, only ill infants usually require morphine, so the findings of this study, particularly the great variation in morphine handling, are of clinical importance. Variability in morphine metabolism could also explain the variability in clinical response to morphine in acutely ill
newborn infants. Perhaps in those producing large amounts of M 6 G and in those who clear it at lower rates, the analgesic and sedative effects are greater and more prolonged because the highly potent M 6 G can penetrate the bloodbrain barrier easilyJ ~ This might be one of the reasons for greater morphine sensitivity in newborn infants than in adults. However, because we did not measure the analgesic effect of morphine in relation to M 6 G levels and its clearance, these explanations remain speculative. Previously we found 20% of morphine to be protein bound in premature infants 4 as compared with 32% 4n adults. 11 This could be another explanation for the higher morphine sensitivity in neonates. These findings underscore the need to monitor the concentrations of both morphine and M 6 G to determine clinical efficacy. The presence of plasma morphine in 2 of the 16 study infants at 24 hours but not at 4 hours might be explained by enterohepatic recirculation of morphine. However, the levels were low, confirming our previous observation that the frequency and magnitude of enterohepatic recirculation of morphine in the neonatal period are probably small. 4 We measured morphine and its metaboIites in only two timed urine samples and in a pair of plasma samples. Continuous urinary collection for a longer period would have provided greater insights into the rates of excretion of morphine and its metabolites. It is also possible that some of the differences in morphine excretion were related to differences in renal function and creatinine clearance, which were not measured in this study. Because of constraints of blood volume, we did not measure other morphine metabolites, such as normorphine and morphine sulfates. Despite these limitations, we conclude that morphine handling is greatly variable among acutely ill premature infants. A large amount of unmetabolized morphine also is
Volume 120 Number 5
found in the urine, irrespective of morphine conjugation. W e could not find any differences between infants who metabolized morphine and those who did not, nor could we explain the large differences in the concentrations of urinary morphine and its conjugates. Our findings may provide some explanations for the great variability in morphine pharmacokinetics and for the degree of morphine sensitivity in premature infants. Further studies are needed to explore the relationships between concentrations of serum and cerebrospinal fluid morphine and M 6 G and their respective clinical effects.
Morphine in acutely ill preterm infants
4.
5.
6. 97.
We thank Drs. Tonse N. K. Raju for assistance in statistical analyses and manuscript preparation, Kristine McCutloch for manuscript editing, and Dharmapuri Vidyasagar for guidance during these studies. REFERENCES
1. Purcell-Jones G, Dormon MB, Sumner BM. The use of opioids in neonates: a retrospective study of 933 cases. Anesthesia 1987;42:1316-20. 2. Osborne R J, Joel SP, Trew D, Slevin ML. Morphine and metabolite behavior after different routes of morphine administration: demonstration of importance of active metabolite morphine-6-glucoronide. Clin Pharmacnl Ther 1990;47:12-9. 3. Pacifici GM, Sawe J, Kager L, Rane A. Morphine glucuroni-
8. 9.
10.
11.
799
dation in human fetal liver and adult liver. Eur J Clin Pharmacol 1982;22:553-8. Bhat R, Chari G, Gulati A, Aldana O, Velamati R, Bhargava H. Pharmacokinetics of a single dose of morphine in preterm infants during the first week of life. J PEDIATR 1990;117:47781. Chari G, Gulati A, Bhat R, Tebbet JR. High-performance liquid chromatographic determination of morphine, morphine3-glucuronide, morphine-6-glucuronoside, and codeine in biological samples using a multiwave-length forward optical detection. J Chromatogr 1991;571:263-70. Choonara IA, McKay P, Hain R, Rane A. Morphine metabolism in children. Br J Clin Pharmacol 1989;28:599-604. Olsen GD, Sommer KM, Wheeler PL, Boyea SR, Michetson SP, Sheek DB. Accumulation and clearance of morphine 3-flD-glucuronide in fetal lambs. J Pharmacol Exp Ther 1988;247:576-84. Boerner U, Abbott S, Roe RL. The metabolism of morphine and heroin in man. Drug Metab Rev 1975;4:39-73. Shimomura K, Kamata O, Ueki S, Oguri IS, Yoshimura H, Tsukamoto H. Analgesic effect of morphine glucuronides. Tohoku J Exp Med 1971;105:45-52. Hand CW, Blunnie WP, Claffey LP, McShane AJ, McQuay H J, Moore RA. Potential analgesic contribution from morphine-6-glucuronoside in CSF [Letter]. Lancet 1987;2:1207-8. Pradhan SN, Datta SN. Agonists and antagonists. In: Pradhan SN, ed. Pharmacology in medicine: principles and practice. Bethesda, Maryland: SP Press International, 1986:20723.
Clinical and laboratory observations Cardiotoxic effects of astemizole overdose in children J a m e s F. W i l e y II, MD, M a r c i a L. G e l b e r , RPh, CSPI, JD, Fred M. H e n r e t i g , MD, C a t h e r i n e C. Wiley, MD, S a t i n d e r S a n d h u , MD, a n d John Loiselle, MD From the Sections of General Pediatrics and Cardiology, St. Christopher's Hospital for Children, the Division of General Pediatrics, Children's Hospital of Philadelphia, and the Delaware Valley Regional Poison Control Center, Philadelphia, Pennsylvania Astemizole, a nonsedating antihistamine, caused a prolonged corrected QT interval, ventricular dysrhythmias, and atrioventricular heart block after overdose in five children. Cardiotoxic effects lasted an a v e r a g e of 21~ days. Children poisoned with astemizole need emergent medical evaluation, a 12-1ead electroc a r d i o g r a m with calculation of the corrected QT interval, and continuous card i a c monitoring for 24 hours. (J PEDIATR1992;120:799-802)
Presented in part at the combined American Association of Poison Control Centers, American Board of Medical Toxicology, American Association of Clinical Toxicology and Canadian Association of Poison Control Centers Annual Meeting, Toronto, Ontario, Canada, Oct. 1, 1991.
Submitted for publication Nov. 1, 1991; accepted Jan. 21, 1992. Reprint requests: James F. Wiley II, MD, Division of Emergency Medicine, St. Christopher's Hospital for Children, Front Street at Erie Avenue, Philadelphia, PA 19134. 9/26/36517 799
Morphine is one of the most frequently used drugs for neonatal analgesia,1 and yet its metabolism in newborn infants has not been well studied. Adults conjugate morphine predominantly as morphine-3-glucuronide and, to a small extent, as a highly potent analgesic, morphine-6-glucuronidefl Preliminary work in this area suggests that premature infants can conjugate morphine, 3 but the extent of conjugation and the magnitude of various subfractions of morphine Supported in part by a grant from the St. Joseph Hospital and Health Care Center, Chicago, Ill. Submitted for publication Sept. 24, 1991; accepted Dec. 20, 1991. Reprint address: Rama Bhat, MD, Department of Pediatrics (m/c 856), 840 South Wood St. Chicago, IL 60612. 9/25/35837
metabolites have not been studied in acutely ill premature infants. We noted that, in comparison with term infants, infants born at <30 weeks of gestational age had prolonged morphine half-life and cleared the drug more slowly. The intersubject variability for these pharmacokinetic meal
M3G M6G CV
Morphine-3-glucuronide Morphine-6-glucuronide Coefficientof variation
]
sures was very large4; we speculated that these findings reflect variations in morphine handling among preterm in, fants. To explore further the metabolic fate of morphine in acutely ill preterm infants after administration of a single 795
796
B h a t et al.
The Journal o f Pediatrics May 1992
Table I. Plasma and urine metabolites in 16 preterm
dose, in this study we measured its major metabolites in the blood and urine, using a modified, extremely sensitive, high-pressure liquid chromatography assay. We sought to determine (1) the predominant morphine metaholite in preterm infants, (2) whether preterm infants produce M6G, (3) to what extent morphine is excreted unmetabolized in the urine, and (4) the reasons for the intersubject variability in morphine handling.
20 to 24 hours after morphine dosing. The bladder was emptied by the Cr6d6 technique before the dose, and a pediatric urine collection bag (U-Bag; Hollister Inc., Libertyville, Ill.) was placed for urine collection. One milliliter of blood was obtained via the indwelling catheter or from an arterial puncture, 4 hours and 24 hours after the dose. Blood samples were spun, and the plasma was separated and stored at - 7 0 C ~ along with the paired urine samples, for later analysis of morphine, M3G, M6G, and codeine. Morphine, M3G, M6G, and codeine were measured by high-performance liquid chromatographic method developed in our laboratory. For morphine and its metabolites, the lower limit of assay sensitivity was 1 ng/ml. The details of the assay have been recently reported. 5 The data for the groups were compared by means of the t test for normally distributed continuous variables and Mann-Whitney U test for highly skewed continuous variables. A paired t test was used to compare the changes in respective measures between 4 and 24 hours in the total groups and within subgroups. A p value of <0.05 was considered to be statistically significant. Data are presented as mean _+ SD, and where required the coefficient of variation (CV) (100 X SD/ mean) values are provided.
METHODS
RESULTS
Subjects. Preterm infants admitted to the intensive care nursery and receiving mechanical ventilation were eligible for the study if they required a single intravenously administered dose of morphine for a minor surgical procedure, such as percutaneous venous catheter insertion or removal of a chest tube. The decision to administer morphine was made by the attending neonatologist. The study protocol was approved by the institutional review board, and written informed consent was obtained from the parents. Two additional infants who received multiple doses of morphine for sedation along with pancuronium bromide also were studied. Because these infants were mature and received multiple doses, their data are presented separately. Morphine administration. During administration of morphine, the heart rate and blood pressure were monitored continuously by means of an indwelling arterial catheter and electronic monitor in all except three infants, in whom noninvasive measurements were obtained. The intravenously administered morphine dose was 50 #g/kg for infants weighing <750 gin, and 100 g g / k g for infants weighing >750 gin. Morphine sulfate was diluted 1:5 to 1: 10 in sterile water; thus the final concentration was 0.2 to 0.4 mg/ml. For each infant the dosing vials were prepared separately; after morphine administration via the peripheral vein, the infusion line was flushed to ensure complete drug delivery. Two sets of 4-hour timed urine samples were collected from each subject, the first up to 4 hours and the other from
Clinical data. We studied 16 preterm infants (8 male infants) whose mean birth weight was 843 _+ 332 gm (range 690 to 1220 gm) and gestational age 28.2 _+ 2.2 weeks (range 25 to 32 weeks). The postnatal age at study was 9.7 _+ 17 days; 13 of the infants were <5 days of age, and the other three were 19, 31, and 66 days of age, respectively. The extent of assisted ventilatory support required by the 16 infants varied. The mean dose of morphine administered was 91.8 _+ 18.7/~g/kg. Plasma and urinary morphine concentrations (Table I). In 13 of the infants (81.2%), plasma morphine concentrations were detected at 4 hours, and 12 of these had detectable levels at 24 hours. The 24-hour mean plasma morphine concentration was significantlylower than the 4-hour mean concentration. In two of the three infants who had no plasma morphine at 4 hours, the 24-hour plasma specimens revealed small amounts of morphine (6.1 ng/ml and 9.7 ng/ml). Large quantities of unmetabolized morphine (3% to 15% of the administered dose) were found in the urine of 13 (81.2%) of the 16 study infants at 4 hours, and in 12 (75%) at 24 hours. The mean urinary morphine levels at 24 hours were significantly lower than at 4 hours. The intersubject variability was very large, with CV of 130% at 4 hours, a n d 118% at 24 hours. Metabolites. Large quantities of M3G were detected in the plasma and urine (Table I), even 24 hours after the initial dosing. The urinary M3G values were significantly
infants Variables
4-Hr values
Morphine (ng/ml) Plasma 34.1 + 20.2 Urine 371.2 _+ 484 M3G (ng/ml) Plasma 11.3 _+ 22.1 Urine 17.5 _+ 27.5 M6G (ng/ml) Plasma 0 Urine 18.2 + 52.7
24-Hr values
5.0 _+ 5.5* 18.8 _+ 2.3t 3.7 _+ 9.9 7.3 _+ 24.2t 0.8 + 3.1 1.4 + 3.5
*p <0.001,as comparedwiththe corresponding4-hourvalues(paired t test). tP <0.05,as comparedwiththe corresponding4-hourvalues(paired t test).
Volume 120 Number 5
lower at 24 than at 4 hours (p <0.05). The CV values for plasma and urine M3G at 4 and 24 hours ranged between 157% and 222%. At 4 hours M6G was not detectable in the plasma, but a small amount was detected at 24 hours. However, amounts of this potent analgesic were significant in the urine at 4 hours and dropped to a lower level by 24 hours in all patients; however, the variability in the urine M6G level was very large--CV was 290% at 4 hours, and 400% at 24 hours. Subgroup analyses. Of the 16 study infants, 10 (62.5%) had one or both of the two conjugates. Six of these ten had only M3G, three had only M6G, and the other had both conjugates. Thus M3G was found in 7 (43.75%) of the 16 infants, and M6G in 4 (25%). There was no difference between the infants who were <5 days of age and those >2 weeks of age with regard to the frequency of morphine metabolites (Table II). None of the infants had codeine in plasma or urine. All of the six infants (37.5%) who had no morphine metabolites either in the plasma or in the urine at any time during the study were <5 days of age (Table II). To evaluate the reasons for the differences in morphine handling, we analyzed study variables in two subgroups--one having morphine metabolites and the other having no metabolites (Table III). The dose of morphine did not differ significantly between these subgroups. Both groups of infants excreted large amounts of morphine at 4 hours. Except for the absence of metabolites in the first subgroup, mean morphine concentrations in the plasma and the morphine excretion patterns in the urine at 4 and 24 hours did not differ significantly between the two subgroups (Table III). In both subgroups the 24-hour values for plasma and urine morphine were significantly lower than the corresponding 4-hour values. The CV values also were large in both subgroups. The birth weight, gestational age, male/female ratio, and postnatal age (despite the inclusion of three older infants in one group) did not differ significantlybetween the subgroups. Systemic blood pressure and arterial blood gas values were normal in both groups, and the frequencies of maternal administration of steroids, phenobarbital, or tocolytic agents and of maternal substance abuse (heroin and cocaine) also were similar. Thus we could find no factor that accounted for the presence or absence of morphine metabolites. Among the 10 infants who had metabolites, there were no differences in the mean morphine concentrations (in urine and plasma) and in the concentrations of morphine metabolites between the seven infants who were <5 days and the three who were >2 weeks of age (data not shown). Multiple doses. The two infants who received multiple morphine doses were both 1 day of age, were born at 36 and 39 weeks of gestational age, and at birth weighed 2519 and 3550 gm. As in the 16 index cases, they received an initial morphine dose, 100 t~g/kg, before the 4-hour sample
Morphine in acutely ill preterm infants
797
Table II. Frequency of morphine metabolites Variables
A g e I-5 d a y s (n = 13)
A g e > 2 wk (n = 3)
All c a s e s (n = 16)
Codeine Only M3G Only M6G M3G and M6G No metabolites
0 4 2 1 6
0 2 1 ---
0 6 (37.5%) 3 (18.8%) 1 (6.2%) 6 (37.5%)
Valuesare frequenciesand (%) of totalcases. The frequenciesincludepresence of metabolitein plasmaor urineor both at 4 and/or 24 hoursafter a singledose of morphine. collection. Subsequently, because of a need for sedation, they were given two or three additional doses of morphine before the 24-hour measurements. Four hours after the initial dose, the unmetabolized urine morphine value was 1100 ng/ml in the term infant and 52.2 ng/ml in the 36-week preterm infant; these levels dropped to 12.7 ng/ml in the term infant and 6.3 ng/ml in the preterm infant. The term infant excreted 89.2 ng of M3G and 28.6 ng of M6G per milliliter at 4 hours, and the concentrations dropped to undetectable and 2.1 ng/ml, respectively, at 24 hours. In the preterm infant, no detectable metabolites were found in plasma or urine, either in the 4-hour samples or in the 24hour samples. DISCUSSION We have shown that about one third of acutely ill infants born at <32 weeks of gestational age did not conjugate morphine but eliminated it unmetabolized in the urine. The infants who did conjugate morphine also eliminated large quantities of unconjugated drug. The predominant metabolite was M3G, but the potent M6G was found in 25% of infants. Marked variability in most of the measured morphine metabolites was the rule rather than the exception. No obvious clinical or physiologic characteristics were associated with the great variability in morphine handling. No infant metabolized morphine to codeine. Choonara et al. 6 examined morphine metabolites in premature infants receiving a continuous infusion of morphine and found M3G to be the predominant metabolite. Although M6G was found in some urine samples, the infants' gestational ages were not provided. Further, the assay for M6G used by these investigators was sensitive to 40 ng/ml, in contrast to a sensitivity of 1 ng/ml in our method. Another study showed that fetal lambs rapidly metabolize morphine to large quantities of M3G, but not to M6G. 7 In adults, morphine is conjugated to M3G (54% to 74%), M6G (<1%), and as ethereal sulfate 0.5% to 5%. Additionally, about 5% is oxidized to normorphine, 1.6% is methylated to codeine, and 7.5% to 12% is excreted in the urine unmetabolized. 8 The elimination and biologic effects of metabolites could
798
Bhat et al.
The Journal of Pediatrics May 1992
Table llh Plasma and urine metabolites in the subgroups Metabolites present (n=40) Variables
Morphine (ng/ml) Plasma Urine M3G (ng/ml) Plasma Urine M6G (ng/ml) Plasma Urine Morphine dose (#g/kg) Birth weight (gm) Gestational age (wk) Study age (days)
No metabolites (n=6)
4 Hours
24 Hours
4 Hours
32.1 + 25.4 454.3 + 537.1
4.9 + 6.5* 14.5 _+ 22.2t
37.5 _+ 5.9 233.0 _+ 384.3
18.0 + 26.1 28.0 _+ 30.5
5.9 + 9.9 11.6 + 16.6
0 0
0 0
0 29.0 + 65.4
0.8 _+ 3.1 2.3 _+ 4.3
0 0
0 0
96.0 842 28.2 13.4
+ 9.9 _+ 295 _+ t.9 + 20.9
24 Hours
5.14 _+ 4.1t 25.9 _+ 22.6"t"
85.0 951 28.2 2.8
_+ 27.9 -+ 211 + 2.8 + 1.8
*p <0.01. Paired t test between corresponding values at 4 hours and 24 hours within each subgroup. Plasma and urinary morphine and postnatal age were not significantlydifferent between subgroups (Mann-Whitney O test), and other clinical variables also were similar (Student t test). tP <0.05. Information as for p <0.01, above. influence the magnitude and duration of the drug effect. For instance, M 3 G is not an analgesic but has an excitatory effect and can cause seizures. 7 On the other hand, M 6 G is a highly potent analgesic and sedative--more potent than morphine itself. 2, 9 High concentrations of M 6 G have been detected in the cerebrospinal fluid of adults receiving morphine. I~ Because morphine glucuronides have prolonged terminal elimination half-lives, even small amounts of M 6 G in the plasma could significantly enhance the analgesic effect of the drug, particularly because of its blood-brain barrier permeability. A number of factors might modulate morphine metabolism. Although neonates have low or poor conjugating capacity, which improves during the first few days, we could not demonstrate an effect of postnatal age on morphine metabolism; a larger population sample with a greater range of postnatal ages will be needed to confirm this observation. Similarly, between 25 weeks and 32 weeks, we found no evidence of an effect of gestational age on either the frequency or extent of morphine disposition. In addition, of the two mature infants who received multiple doses, only one had evidence of morphine conjugation, but both excreted unmetabolized morphine. Although all infants were physiologically stable at the time of our study, they were acutely ill and required ventilatory support; other factors not examined by us could have influenced the rate and extent of morphine handling. Therefore our findings cannot be generalized to all preterm infants or to all neonates. From a practical point of view, however, only ill infants usually require morphine, so the findings of this study, particularly the great variation in morphine handling, are of clinical importance. Variability in morphine metabolism could also explain the variability in clinical response to morphine in acutely ill
newborn infants. Perhaps in those producing large amounts of M 6 G and in those who clear it at lower rates, the analgesic and sedative effects are greater and more prolonged because the highly potent M 6 G can penetrate the bloodbrain barrier easilyJ ~ This might be one of the reasons for greater morphine sensitivity in newborn infants than in adults. However, because we did not measure the analgesic effect of morphine in relation to M 6 G levels and its clearance, these explanations remain speculative. Previously we found 20% of morphine to be protein bound in premature infants 4 as compared with 32% 4n adults. 11 This could be another explanation for the higher morphine sensitivity in neonates. These findings underscore the need to monitor the concentrations of both morphine and M 6 G to determine clinical efficacy. The presence of plasma morphine in 2 of the 16 study infants at 24 hours but not at 4 hours might be explained by enterohepatic recirculation of morphine. However, the levels were low, confirming our previous observation that the frequency and magnitude of enterohepatic recirculation of morphine in the neonatal period are probably small. 4 We measured morphine and its metaboIites in only two timed urine samples and in a pair of plasma samples. Continuous urinary collection for a longer period would have provided greater insights into the rates of excretion of morphine and its metabolites. It is also possible that some of the differences in morphine excretion were related to differences in renal function and creatinine clearance, which were not measured in this study. Because of constraints of blood volume, we did not measure other morphine metabolites, such as normorphine and morphine sulfates. Despite these limitations, we conclude that morphine handling is greatly variable among acutely ill premature infants. A large amount of unmetabolized morphine also is
Volume 120 Number 5
found in the urine, irrespective of morphine conjugation. W e could not find any differences between infants who metabolized morphine and those who did not, nor could we explain the large differences in the concentrations of urinary morphine and its conjugates. Our findings may provide some explanations for the great variability in morphine pharmacokinetics and for the degree of morphine sensitivity in premature infants. Further studies are needed to explore the relationships between concentrations of serum and cerebrospinal fluid morphine and M 6 G and their respective clinical effects.
Morphine in acutely ill preterm infants
4.
5.
6. 97.
We thank Drs. Tonse N. K. Raju for assistance in statistical analyses and manuscript preparation, Kristine McCutloch for manuscript editing, and Dharmapuri Vidyasagar for guidance during these studies. REFERENCES
1. Purcell-Jones G, Dormon MB, Sumner BM. The use of opioids in neonates: a retrospective study of 933 cases. Anesthesia 1987;42:1316-20. 2. Osborne R J, Joel SP, Trew D, Slevin ML. Morphine and metabolite behavior after different routes of morphine administration: demonstration of importance of active metabolite morphine-6-glucoronide. Clin Pharmacnl Ther 1990;47:12-9. 3. Pacifici GM, Sawe J, Kager L, Rane A. Morphine glucuroni-
8. 9.
10.
11.
799
dation in human fetal liver and adult liver. Eur J Clin Pharmacol 1982;22:553-8. Bhat R, Chari G, Gulati A, Aldana O, Velamati R, Bhargava H. Pharmacokinetics of a single dose of morphine in preterm infants during the first week of life. J PEDIATR 1990;117:47781. Chari G, Gulati A, Bhat R, Tebbet JR. High-performance liquid chromatographic determination of morphine, morphine3-glucuronide, morphine-6-glucuronoside, and codeine in biological samples using a multiwave-length forward optical detection. J Chromatogr 1991;571:263-70. Choonara IA, McKay P, Hain R, Rane A. Morphine metabolism in children. Br J Clin Pharmacol 1989;28:599-604. Olsen GD, Sommer KM, Wheeler PL, Boyea SR, Michetson SP, Sheek DB. Accumulation and clearance of morphine 3-flD-glucuronide in fetal lambs. J Pharmacol Exp Ther 1988;247:576-84. Boerner U, Abbott S, Roe RL. The metabolism of morphine and heroin in man. Drug Metab Rev 1975;4:39-73. Shimomura K, Kamata O, Ueki S, Oguri IS, Yoshimura H, Tsukamoto H. Analgesic effect of morphine glucuronides. Tohoku J Exp Med 1971;105:45-52. Hand CW, Blunnie WP, Claffey LP, McShane AJ, McQuay H J, Moore RA. Potential analgesic contribution from morphine-6-glucuronoside in CSF [Letter]. Lancet 1987;2:1207-8. Pradhan SN, Datta SN. Agonists and antagonists. In: Pradhan SN, ed. Pharmacology in medicine: principles and practice. Bethesda, Maryland: SP Press International, 1986:20723.
Clinical and laboratory observations Cardiotoxic effects of astemizole overdose in children J a m e s F. W i l e y II, MD, M a r c i a L. G e l b e r , RPh, CSPI, JD, Fred M. H e n r e t i g , MD, C a t h e r i n e C. Wiley, MD, S a t i n d e r S a n d h u , MD, a n d John Loiselle, MD From the Sections of General Pediatrics and Cardiology, St. Christopher's Hospital for Children, the Division of General Pediatrics, Children's Hospital of Philadelphia, and the Delaware Valley Regional Poison Control Center, Philadelphia, Pennsylvania Astemizole, a nonsedating antihistamine, caused a prolonged corrected QT interval, ventricular dysrhythmias, and atrioventricular heart block after overdose in five children. Cardiotoxic effects lasted an a v e r a g e of 21~ days. Children poisoned with astemizole need emergent medical evaluation, a 12-1ead electroc a r d i o g r a m with calculation of the corrected QT interval, and continuous card i a c monitoring for 24 hours. (J PEDIATR1992;120:799-802)
Presented in part at the combined American Association of Poison Control Centers, American Board of Medical Toxicology, American Association of Clinical Toxicology and Canadian Association of Poison Control Centers Annual Meeting, Toronto, Ontario, Canada, Oct. 1, 1991.
Submitted for publication Nov. 1, 1991; accepted Jan. 21, 1992. Reprint requests: James F. Wiley II, MD, Division of Emergency Medicine, St. Christopher's Hospital for Children, Front Street at Erie Avenue, Philadelphia, PA 19134. 9/26/36517 799