N E W S & A N A LY S I S

The long march of antisense

Antisense therapy has generated waves of enthusiasm and disappointment. With a new drug about to be submitted for
approval, is it on the cusp of becoming established as a platform technology?
Dan Jones
In 1978, Paul Zamecnik and Mary
Stephenson reported the first experiments
on antisense mechanisms of gene
silencing, using short synthetic antisense
oligonucleotides to inhibit replication
of the Rous sarcoma virus by binding
and blocking the action of 35s RNA. In
the following years, attention turned to
the possible therapeutic applications of
antisense technology yet, more than
threedecades later, just one antisense
therapy has reached the market: Isis
Pharmaceuticals fomivirsen, which was
approved in 1998 for retinitis induced by
cytomegalovirus, and discontinued in 2004
as the drugs marketshrank.
Despite substantial efforts since this
approval, clinical trials in this field have not
delivered. Two first-generation Isis products
aprinocarsen, which lowers expression of
protein kinase C (PKCa) for the treatment
of various cancers, and alicaforsen, which
lowers intercellular adhesion molecule1
(ICAM1) expression to treat Crohns disease
reached PhaseII/III trial before being
dropped because of lack of efficacy. Gentas
oblimersen, which targets the mRNA
encoding B cell lymphoma 2 (BCL2), was
filed for approval in the United States
for melanoma and chronic lymphocytic
leukaemia (in 2003 and 2006, respectively),
but was rejected for both indications.
As the early antisense buzz failed
to generate additional products, large
pharmaceutical companies lost interest in
the approach and smaller biotechnology
companies were left to carry the torch.
Yet, there are some signs the tide could be
ready to turn for classic antisense drugs,
which act by binding to target mRNA,
blocking protein translation and initiating
ribonuclease H-dependent degradation.
Last year OncoGeneX and Teva pushed
OGX011, a second-generation antisense
agent that was licensed from Isis and
targets clusterin, into PhaseIII trials for
prostate cancer. New PhaseIII data on
oblimersen in melanoma in a targeted
patient population are expected soon.
And Isis is set to submit its mipomersen,
which targets expression of apolipoprotein
B100 (APOB100) and thereby reduces

levels of low-density lipoprotein (LDL,

or bad cholesterol), for approval in both
the European Union and the United States
within months.
From product to platform
A key factor in the drawn-out history of
antisense therapies, says Stanley Crooke,
founder and CEO of Isis, has been the need
to develop the antisense approach into a
true platform technology. That is, despite
the early success with fomivirsen, all the
necessary elements were not in place at that
time to make multiple products. Indeed,
in recent decades, this has arguably only
been achieved for one novel therapeutic
modality monoclonal antibodies (mAbs).
Crooke therefore didnt expect a flood
of new antisense drugs in the immediate
years after fomivirsens introduction, and
nor did others in the industry. Fomivirsen
was widely seen as a one-off approval,
says Raymond Warrell, Chairman and
CEO of Genta. mAbs were widely seen as
the magic bullets of the 1980s, yet failed
to deliver for many years so time-wise,
I dont think the emergence of antisense
therapies is peculiar, he adds.
To create a viable antisense platform,
pioneers in the field have had to integrate
new technical knowledge about antisense
drugs with an increasing understanding
of the biology of their targets, within
the context of an evolving regulatory
environment. And such a process naturally
involves some early missteps, which, for
drug development, are timely and costly.
You cant really learn about a technology
until you get some drugs into development,
because its only then that you really find
out what the issues are, says Crooke.

mAbs were widely seen as the

magic bullets of the 1980s, yet
failed to deliver for many years
so time-wise, I dont think
the emergence of antisense
therapies is peculiar.

NATURE REVIEWS | DRUG DISCOVERY

This means making some bets on early

candidates, which may not pan out, and
then trying again.
In developing antisense into a platform
technology, much attention naturally
focused on the design of antisense
oligonucleotides. Sudhir Agrawal, CEO
and President of Idera Pharmaceuticals,
says that from day one there have been
three fundamental issues: stability,
delivery and off-target effects. With
regard to stability, for instance, natural
oligonucleotides have a phosphodiester
backbone that is susceptible to degradation
by nucleases, prompting Agrawal and
Zamecnik to develop the first degradationresistant synthetic oligonucleotides
phosphorothioate oligonucleotides (PSOs).
Subsequently, the team developed the
second-generation chemistry, which relies
on 2O-substitutions to provide greater
stability and minimize off-target effects.
These have been licensed to Isis, which is
now developing generation 2.5 molecules.
As oligonucleotide chemistry has
advanced, new classes of antisense drugs
which modulate gene expression by
harnessing routes other than the combination
of ribonuclease H degradation and steric
blocking have also emerged (BOX1).
More than just chemistry
Despite the advances in oligonucleotide
design, the importance of these new
chemistries for turning antisense into a
successful platform technology remains
debatable. The older chemistries may be
adequate if you fool around with them
enough, says Sidney Altman, a molecular
biologist at Yale University, Connecticut,
USA. Warrell, whose company uses the
early PSO technology, similarly argues that
oligonucleotide chemistry has not been
the fundamental obstacle within the field.
I dont view the generations of antisense
chemistries as very important, says
Warrell. We think that the first-generation
PSO chemistry is fine and has acceptable
targeting and binding characteristics.
Agrawal points out, however, that
the off-target effects of some antisense
compounds can be problematic. PSO
compounds interact with a number of
VOLUME 10 | JUNE 2011 | 401

2011 Macmillan Publishers Limited. All rights reserved

N E W S & A N A LY S I S
Box 1 | The many faces of antisense therapy
Antisense is often used broadly to describe all approaches that use RNA or DNA to block gene
expression as a result of complementary base pairing with a target RNA molecule in the cell. Yet, this
broad definition actually covers a range of mechanisms and activities. Classic antisense molecules
such as Isis Pharmaceuticals mipomersen, Gentas oblimersen and OncoGenix/Tevas OGX011
bind to target mRNAs, induce ribonuclease H (RNaseH)-dependent degradation and block protein
translation, explains Agrawal.
Other antisense agents, however, can work through various other mechanisms. AVI BioPharma,
for instance, works with phosphorodiamidate morpholino oligomers (PMOs). Although PMOs can
prevent the expression of target mRNAs (but without induction of RNaseH degradation), they
can also be used to alter the splicing of pre-mRNAs. AVI BioPharma is using such splice-switching
oligomers to achieve exon skipping to create functional proteins in Duchenne muscular dystrophy,
and their lead candidate, eteplirsen (AVI4658), has completed PhaseI/II trials in this indication.
Santaris Pharma, meanwhile, is working with locked nucleic acids (LNAs), which are synthetic
chemical analogues of RNA that can bind to different types of RNA targets. LNAs can inhibit
expression of target mRNAs via induction of RNaseH activity, but Santariss most advanced
LNA compound instead targets microRNA; miravirsen, which is due to enter PhaseII trials for
hepatitis C virus (HCV) in September 2011, inhibits miR122, a liver-specific microRNA that is
essential for HCV replication.
Idera Pharmaceuticals has also recently reported the creation of a novel class of compounds
they call gene-silencing oligonucleotides (GSOs; J.Med. Chem. 54, 30273036; 2011)
single-stranded DNA or RNA molecules that can target mRNAs and microRNAs, and seem to
modulate gene expression via RNA interference (RNAi) activity. Whereas most RNAi-inducing
agents are double-stranded, the fact that GSOs are single-stranded may be advantageous in
terms of their delivery.
Altman, meanwhile, is working on morpholino oligonucleotides that bind to target mRNAs and
activate degradation via RNaseP rather than RNaseH.

cellular factors, including Toll-like receptors

(TLRs), which detect free-floating DNA
or RNA and induce an immune response
that manifests as flu-like symptoms,
says Agrawal. If youre working with
an antisense compound, you need to be
sure that its activity is only related to its
antisense mechanism of action, and not
because of interactions with TLRs.
Others in the industry point to more
general rather than antisense-specific
issues as the major obstacles to success. One
fundamental challenge, says Warrell, has been
getting a better handle on the underlying
biology of diseases. If youre going to have
an effect with a single antisense agent, which
attacks only a single gene product, you have
to be sure that your target has an absolutely
central role in the complex biology of the
disease youre trying to treat, says Warrell.
Target criticality rather than general
problems of off-target effects and/or toxicity
appears to have played a crucial part in
the failure of some of Isiss early antisense
compounds as well. Crooke says that
aprinocarsen, Isiss anti-PKC drug, could

I dont view the generations

of antisense chemistries
as very important.

be given at very high doses as its side effects

were modest compared with other anticancer
agents. But it failed because PKC was a bad
choice of target, says Crooke. Its just not
very significant in the malignancies that we
looked at.
The usual problems with clinical
trial design and regulatory requirements
have also held up the field. The failure of
alicaforsen, Isiss ICAM1-targeting drug
for Crohns disease, probably relates to the
clinical trials Isis ran, says Crooke. We
learned afterwards that we probably ended
up with patients who had irritable bowel
syndrome, rather than Crohns.
The history of Gentas oblimersen
also underscores the importance of the
regulatory environment in explaining
some late-stage setbacks. Oblimersen, a
first-generation PSO, blocks the production
of BCL2, a central regulator of apoptosis.
BCL2 is overexpressed in numerous
cancers and protects cancer cells against
the cytotoxic effects of various therapies, so
giving cancer patients oblimersen ahead of
these therapies should, in theory, maximize
their cell-killing power.
Despite preclinical validation for this
approach, oblimersen failed to clear clinical
and regulatory hurdles. Oblimersen missed
its primary end point of overall survival in
melanoma, but was nevertheless filed in
this setting in 2003. Today, in 2011, no one

402 | JUNE 2011 | VOLUME 10

would submit having missed their primary

end point, but in 2003 it was not clear that
this would be a fatal issue, says Warrell.
Genta has nonetheless learnt some
important lessons about their patient
population and is running an additional
PhaseIII trial that used a biomarker to
select patients who are most likely to benefit.
Results are expected in the near future.
Light at the end of the tunnel?
Another product leading the field is
Isiss mipomersen, a second-generation
oligonucleotide that targets the expression
of APOB100 an essential protein for
the formation of LDL, which is a major
contributor to coronary heart disease.
We picked APOB100 as it is a genetically
validated target and it is also made in the
liver, an organ where we knew that about
12% of any dose we give accumulates,
says Crooke.
Mipomersen is being developed in
collaboration with Genzyme (now Sanofi)
for patients with severely high LDL despite
receiving maximum lipid-lowering therapy,
including patients with familial hypercholesterolaemia (FH). So far, late-stage clinical
data from the drug have looked promising.
In a PhaseIII trial in 51 patients with
homozygous FH (HoFH), for instance, the
drug reduced LDL levels by 25%, whereas
placebo only reduced levels by 3% (Lancet
375, 9981006; 2010). Although increased
levels of liver enzyme in this and other trials
have raised concern among some analysts,
Crooke points out that most of these cases
resolved with ongoing treatment and all
resolved after treatment discontinuation.
After several delays to its timeline driven
in part by regulatory requests for more
data to support a submission the drug is
due to be filed for approval with European
regulators this summer for HoFH and
severe heterozygous FH, and with the US
Food and Drug Administration later in the
year for HoFH.
If mipomersen is approved, it may not
only validate antisense as a therapeutic
modality, but also may be the first step to
demonstrating the commercial viability of
such products. Although the initial focus
for mipomersen is on patients with very
high-risk HoFH, the company hopes to
eventually expand the drugs market to
include the much larger population of
patients with high LDL levels. Similarly, if
the recent trials of oblimersen and OGX011
produce positive results, antisense may finally
have come of age as a genuine platform
technology.
www.nature.com/reviews/drugdisc

2011 Macmillan Publishers Limited. All rights reserved