70 Science & technology The Economist April 16th 2022
Personalisedmedicine
An ounce of
prevention
P
eterley, a retiredcivil servant who
lives in London, was diagnosed with co
lon cancer in 2017. An operation to remove
the tumour was successful. But the chemo
therapy that followed caused a severe reac
tion that required a twoweek hospital stay
and a pause in his cancer treatment.
All that could have been avoided had a
simple test been done. The test examines a
gene that encodes a liver enzyme called di
hydropyrimidine dehydrogenase (or dpd
for short). The enzyme breaks down sever
al common cancer drugs. Without it, toxic
levels of the drugs build up in the body,
sometimes with fatal results. A complete
inability to make dpdis rare, but there are
four mutations in the dpdregulating gene
that are known to reduce its production. As
it turned out, Mr Ley had one of them.
Screening for such “pharmacogenes” is
an idea that is catching on among doctors.
Several big hospitals in America are testing
their patients for a dozen or more of them.
Separate pilot projects are under way in at
least seven of the European Union’s mem
ber states. Britain’s National Health Service
(nhs) is doing screening tests for some pa
tients being prescribed cancer and hiv
drugs. A report on March 29th by the Brit
ish Pharmacological Society (bps) and the
Royal College of Physicians (rcp) proposed
widening that testing to cover the 40 drugs
among the 100 mostprescribed that are
known to be affected by pharmacogenes.
The report’s authors reckon testing could
feasibly be rolled out across the nhsas
soon as 2023.
Genetic screening promises big bene
fits. Mutations can affect drugs in all sorts
of ways, determining a pill’s efficacy, toxic
ity, how well it is absorbed, and how well it
is broken down. Some genetic variants af
fect several drugs at once, because they al
ter common enzymes in widely used meta
bolic pathways. Britain’s 100,000 Genomes
project has shown that almost 99% of peo
ple carry at least one pharmacogene; 25%
have four. About 9% of Caucasian people
have, like Mr Ley, dpddeficiency; one in
200 lack the enzyme completely. Roughly
8% of Britain’s population get little pain re
lief from codeine, because they lack an en
zyme responsible for metabolising the
drug into morphine (they instead metabol
ise it into other substances that have little
influence on pain).
All told, scientists have identified about
120 such druggene pairs so far. Roughly
half of them are “actionable”, says Henk
Jan Guchelaar, a pharmacologist at the
University of Leiden in the Netherlands—
meaning that changing the dose or replac
ing the drug can lead to a better clinical
outcome. And most people will be pre
scribed at least one of those drugs at some
point in their lives. In Britain people over
the age of 70 have around 70% chance of
taking at least one drug whose safety or ef
ficacy is compromised by their genes, says
Munir Pirmohamed, a pharmacologist and
geneticist at the University of Liverpool.
Currently, clashes between a patient’s
genome and his drug regimen are dealt
with by trialanderror prescribing. But
that is timeconsuming, and may be harm
ful. If a drug is being prescribed for high
blood pressure or arteryclogging levels of
cholesterol, time spent trying different
drugs means time in which a stroke, heart
attack or organ damage may occur. And
cleverer prescribing would have benefits
for the healthcare system overall, as well
as for individual patients. Adverse drug re
actions account for 6.5% of hospital ad
missions in Britain.
The chief issue, as ever, is cost. In the
Netherlands a test for 50 pharmacogenes
costs about €200 ($217). In Britain a panel
test for 40 such genes costs £100150 ($130
195). Carrying out testing on an entire pop
ulation would, therefore, be extremely ex
pensive. Some light on whether it is worth
the cost will be shone later this year when
prepare, a study that began in 2017, pub
lishes its results. The project, which is led
by Dr Guchelaar, recruited 7,000 people
across seven European countries for a stu
dy of mutations affecting 42 different
drugs. Half the participants were screened,
and given cards listing the drugs flagged
up. That information, in turn, was made
available to doctors, pharmacists and the
like. Dr Guchelaar and his colleagues are
analysing how much this reduces adverse
drug reactions compared with the un
screened participants—and, crucially, the
healthcare costs averted as a result.
Such costbenefit analyses will be vital
in making the argument that governments
or insurance firms should pay for wide
spread genetic testing. In the meantime,
though, doctors are already pondering
ways to get the most bang for their buck.
The bps and rcpstudy suggests several
ways to expand pharmacogene screening.
One is to test for the genes the first time a
drug known to be susceptible is pre
scribed. Another option is to offer that test
to everyone over a certain age, perhaps
50—though the nhsis also pondering the
idea of comprehensive geneticscreening
for all newborn babies. Thatcouldpay off
handsomely later in their lives.n
Genetic screening can make drugs
work better
Two of these wavelengths are also in the
infrared and red colour range. That allows
some of the reemitted light to emerge
from the paste to where it can be analysed
by an external instrument called a spec
trometer. By precisely measuring the dif
ference between the two wavelengths, the
temperature of particles that are emitting
it—and thus of the paste as a whole—can
be worked out.
Thus far, the researchers have tested the
technique only on pieces of pig intestine
that they obtained from a slaughterhouse.
Soldering a cut is done in a matter of min
utes. Similar “ex vivo” tests of the strength
and permeability of the bond will also be
needed for human tissue, followed by clin
ical tests on actual pigs and, eventually,
humans. But the researchers are optimis
tic. At the conference, they were cagey
about exactly what the fluorescing parti
cles are made of. They are applying for a
patent, which could be quite valuableif the
tools at the ready in a doctor’sofficeone
day include a laser soldering gun.n