Asian and African babies
are at greater risk of rapidly metabolizing codeine
A class action suit over the death
of an apparently healthy Toronto newborn, who died last
year from opiate toxicity from breast milk, has renewed
the debate over prescribing Tylenol 3 to breastfeeding
mothers. After the baby's death, doctors at Toronto's
Hospital for Sick Children issued a warning that codeine
given for postnatal pain can produce deadly concentrations
of morphine in breast milk.
Tariq Jamieson was delivered vaginally
at full term and healthy weight everything appeared
normal. His mother Rani suffered some lingering pain
from an episiotomy so she was prescribed two tablets
of Tylenol 3 twice daily a common pain treatment
for mothers who have just given birth. Doctors halved
that dose after two days due to constipation and somnolence.
Tariq developed increasing lethargy
from the seven-day mark, and at 11 days was brought
to a pediatrician due to concerns about his skin colour
and poor feeding. He had, however, regained his birth
weight. But two days later the family called an ambulance.
Responders found the infant cyanotic and lacking vital
signs. Attempts at resuscitation failed.
On post mortem, the child was found
to have a blood concentration of acetaminophen at 5.9
µg/mL and morphine at 70 ng/mL. That morphine
concentration is about six times higher than would normally
be considered safe in a neonate.
Tylenol 3 contains 500mg of acetaminophen
and 30mg of codeine. Codeine is metabolized to morphine
in the body, but not all patients metabolize it at the
same rate. Ms Jamieson was genotyped and found to carry
three CYP 2D6 genes, which create the enzyme catalyzing
the O-demethylation of codeine to morphine. This essentially
made her an ultra-rapid metabolizer of codeine to morphine,
leading to an unexpectedly fast build-up of the opiate
in her breast milk.
This is the first reported case ever of a child dying
from opioid poisoning due to a breastfeeding mother's
use of codeine, and it was a fairly exceptional case.
Not only did Ms Jamieson have three CYP 2D6 genes, but
her husband and baby both had two, making all of them
But it is not so exceptional as
to be safely ignored. The ultrarapid metabolizer genotype
occurs in about 1% of Caucasians, but runs as high as
30% in some African and Asian populations. For every
baby whose life is threatened, many others may suffer
morbidity, depressed breathing, lethargy or poor feeding.
Even two CYP 2D6 genes can lead to unexpectedly high
concentrations of morphine in breastmilk.
The child's mother, not surprisingly, has said that
codeine "should not be used by nursing mothers under
any circumstance." But the experts at the Hospital for
Sick Children's Motherisk program, who still have to
deal with maternal pain somehow, are not quite ready
to go that far. Instead they suggest sensible approaches
to minimize the risk.
There are five strategies available,
they suggest. One is simply to avoid using codeine in
breastfeeding mothers. But this may leave the mother
with uncontrolled pain. Another option is to give the
codeine but avoid breastfeeding. No neonatologist, however,
is going to recommend stopping breastfeeding at this
crucial early stage if it can possibly be avoided.
A middle road is to give codeine,
but limit concentrations by not giving a high dosage
(240 mg/day codeine) for more than a few days. But the
Motherisk team worries that this may not control pain
adequately, and could still lead to toxic levels of
morphine in the milk of ultrarapid metabolizers.
The ideal solution would be to
genotype all mothers then limit codeine only in the
cases of fast metabolizers those with two or
three 2D6 genes. Unfortunately this would be very expensive,
and few centres currently have the facilities to do
That leaves old-fashioned clinical
judgement. The mother should be informed of the potential
for opioid toxicity, then she and the infant should
be monitored closely for danger signs. If symptoms appear,
administering naxolone, morphine's antidote, will generally
solve the problem and, in doing so, confirm it.
In the longer run, the question of codeine's safety
in breastfeeding mothers will have to be revisited.
The American Academy of Pediatrics guidelines on the
transfer of drugs into human milk, published in Pediatrics
in September 2001, list codeine as a "maternal medication
usually compatible with breastfeeding" and report no
evidence of symptoms in infants or effect on lactation.
It now appears as though that conclusion will have to
be revisited, as the Motherisk researchers openly attack
the guidelines for overlooking the risk of codeine in
breastfeeding "despite lack of sufficient published
data to support this recommendation." Motherisk is now
recruiting patients for its own study on the subject.