INSIGHTS | PERSPECTIVES
1064 6 MARCH 2015 • VOL 347 ISSUE 6226 sciencemag.org SCIENCE
discovery phase, where rapid diagnostics
would enable the development and clinical
adoption of pathogen-targeted antibiotics.
The wide genetic diversity of pathogenic
bacteria (16) has been an obstacle to the
development of broad-spectrum drugs.
Molecules that target individual pathogens
of particular importance should be easier
to identify and optimize for selectivity
and toleration, and more rapid to develop
in targeted patient populations. One such
pathogen could be Acinetobacter baumannii, which is often resistant to current treatments and is associated with high mortality
rates (17). Such narrowly targeted drugs
should be premium priced, as they would
be used to treat a small number of patients
with serious, otherwise untreatable infections (18).
Rapid molecular diagnostics paired with
pathogen-targeted antibiotics would usher
in an era of “personalized health care” for
patients suffering from bacterial infections. Most patients would continue to be
treated with older, inexpensive, and still
effective antibiotics. Rapid diagnostics
could help identify those patients requiring new-generation drugs that target highly
pathogenic or resistant strains, limiting unnecessary use and slowing the emergence of
resistance: altogether, a more sustainable
practice, both in terms of clinical care and
antibiotic stewardship. ■
REFERENCES AND NOTES
1. H. W.Boucher et al.,Infectious Diseases Societyof
America, Clin. Infect. Dis. 56, 1685 (2013).
2. D. M. Livermore, Int.J.Antimicrob.Agents 39, 283 (2012).
3. I. Karaiskos, H. Giamarellou, Expert Opin. Pharmacother.
15, 1351 (2014).
4. P. M. Hawkey, A. M. Jones, J. Antimicrob. Chemother. 64
(suppl. 1), i3 (2009).
5. S.Baker, Science 347,1064(2015).
6. See www.imi.europa.eu/content/translocation.
7. See https://longitudeprize.org/.
8. See www.nih.gov/about/director/09182014_statement_
9. D. J. Payne et al ., Nat. Rev. Drug Discov. 6, 29 (2007).
10. J. M. Pagès, C. E. James, M. Winterhalter, Nat. Rev.
Microbiol. 6, 893 (2008).
11. S. Biswas et al ., Nat. Struct. Mol. Biol. 14, 1108 (2007).
12. L. L. Ling etal .,Nature517, 455 (2015).
13. N. P. Braykov, M. R. Eber, E. Y. Klein, D. J. Morgan, R.
Laxminarayan, Infect. Control Hosp. Epidemiol. 34, 259
14. F. C. Tenover et al ., J. Clin. Microbiol. 51, 3780 (2013).
15. E. Mylonakis et al ., Clin. Infect. Dis. 10.1093/cid/ciu959
16. N. R. Pace,Science 276, 734 (1997).
17. A. Y. Peleg, H. Seifert, D. L. Paterson, Clin.Microbiol.Rev.
21, 538 (2008).
18. B. Spellberg, J. H. Rex, Nat. Rev. Drug Discov. 12, 963
19. Data reproduced with permission from IMS Health
International MIDAS Data 1999-2010.
I thank P. Bradford for critical review. The views expressed in this
article are those of the author and do not necessarily represent
the views of, nor should be attributed to, AstraZeneca.
After many years out of the limelight, antimicrobial resistance (AMR) in bacteria is firmly back on the interna- tional political and scientific agenda (1, 2). The potential impact of AMR on hospital-acquired bacterial infections such as Staphylococcus aureus and
Acinetobacter baumannii in higher-income
countries has created both fear and a surge
of motivation aimed at providing new solutions for the problem (3, 4). The political
will and momentum to tackle AMR lies in
higher-income countries, but the
medical, social, and economic
effects of AMR are likely to be
felt more in lower-income countries, particularly those in South
and Southeast Asia and in sub-Saharan Africa. The identification and development of new
drugs is a potential solution but
is challenging and costly; any
novel therapies introduced into low-income
settings without a suitable infrastructure to
understand and prevent the rapid development of resistance will likely be expensive
In many countries at the lower end of the
global economic ladder, infections caused
by multidrug resistant (MDR) and extended
drug resistant (XDR) bacteria are a common
reality. Variants of bacterial pathogens carry-
ing novel AMR mechanisms disproportion-
ately originate in lower-income countries,
with downstream consequences both within
and outside the region in which they ap-
pear. This phenomenon was highlighted in
2008 by the emergence of the carbapenem
resistance–inducing New Delhi metallo-β-
lactamase–1 (NDM-1) (5). This gene induces
broad resistance against carbapenems and
other β-lactams and was first identified in a
Klebsiella pneumoniae strain isolated from
a Swedish national upon returning from
India. The plasmids carrying this gene have
since become common and are having dra-
matic impact on the efficacy of carbapenems
and other β-lactams in hospitals. A recent
report described hospital outbreaks of
Klebsiella pneumoniae in children on high-
dependency wards in South Asia (6). These
outbreaks were caused by particularly viru-
lent variants, which induced a rapid-onset
bacteremia resulting in a 75% mortality rate
in the infected children. The presence of
NDM-1 within an already broadly antimicro-
bial-resistant and highly virulent strain se-
verely restricted the treatment options, with
a direct impact on patient mortality. This
and many other studies have shown that
AMR genes thrive in low-income settings
and can combine effortlessly with other re-
sistance mechanisms. Further, these wide-
ranging combinations of drug resistance
mechanisms can be maintained
and then transferred within and
between numerous bacterial
The reasons behind the apparent amplification of the
current risk in AMR infections
in lower-income countries are
intricate and occasionally geographically driven, but there are
common themes that highlight the key issues. First, the bacterial pathogens found
in lower-income settings (such as typhoid
fever and tuberculosis meningitis) typically
cause more severe infections than those in
higher-income countries. Second, antimicrobials are widely available for purchase
in the community without medical consultation and without government policies restricting their use; community overuse and
underdosing are common. Third, the medical treatment, range of available antimicrobials, and health care facilities are generally
better in higher-income countries; the risk
associated with having a poor outcome after infection with a resistant pathogen is
therefore greater in lower-income countries. Fourth, very few patients receive any
form of conclusive diagnostic testing before, or indeed after, they are treated with
an empiric antimicrobial regime. For example, febrile infections across Asia are commonly treated with a fluoroquinolone or a
A return to the
By Stephen Baker1,2,3
The effects of antimicrobial resistance will be felt most
acutely in lower-income countries
1 Hospital for Tropical Diseases, Wellcome Trust Major
Overseas Programme, Oxford University Clinical Research
Unit, Ho Chi Minh City, Vietnam. 2Centre for Tropical Medicine,
Oxford University, Oxford, UK. 3 London School of Hygiene and
Tropical Medicine, London, UK. E-mail: email@example.com