in controlled human trials. A Phase II
study with 80 subjects in 2005 showed
that a single 6 day treatment of
new-onset patients with otelixizumab
helped reduce insulin needs for up to
18 months. According to one source,
some of these benefits continue to
the present day—nearly four years af-
ter the drug was given. ‘‘The most ex-
citing thing is that we are targeting
what causes the disorder,’’ says Lou
Vaickus, chief medical officer of
Tolerx, contrasting that approach
with insulin replacement therapy,
where ‘‘the bottom line is, you are
destroying an organ that is irreplace-
able, and then you play catch up after-
wards.’’ Macrogenics’s teplizumab
had a similar success story in 2005,
with the treatment group in trial of 42
new-onset patients showing a sig-
nificant improvement in the insulin
response two years after a 12–14 day
course of treatment.
Studies suggest that anti-CD3 ther-
apy has a dual effect: it not only
destroys autoreactive T cells immedi-
ately, but also creates regulatory
T cells that continue to keep the rogue
cells in check for a long time. ‘‘What
makes anti-CD3 attractive is that it is
not continuous immune suppression,’’
says Kevan Herold, who led the tepli-
zumab clinical trial. ‘‘We give an immu-
nomudulator for a brief period of time,
with lasting effect—it is the first step in
the direction of tolerance.’’
One of the drawbacks with systemic
immune modulators such as anti-CD3
is the way they are administered: pa-
tients have to undergo an intravenous
infusion of the drug for up to 2 weeks.
Another potential drawback is the
transient reactivation of existing infec-
tions, although neither trial reported
any serious problems of this kind.
Finally, the benefits of the treatment,
though long lasting, are likely to dimin-
ish with time. ‘‘To be fair, it is not
a cure,’’ acknowledges Herold. ‘‘But
the overall clinical experience is much
better for the patients.’’
As an alternative to systemic inter-
ventions, a number of antigen-specific
methods are being tried. Their goal is
to retrain the autoreactive immune
system with a vaccine based on anti-
gens associated with b cells. In theory,
such a targeted approach could fix the
underlying problem in type 1 diabetes
while having minimal impact on normal
immune function. And as with any
vaccine, the effect would be sus-
tained. One vaccine that has shown
clinical efficacy in type 1 diabetes is
based on glutamic acid decarboxylase
(GAD), a major autoantigen implicated
in the disease. Tested in a placebo-
controlled trial of 70 patients in 2006,
the drug showed that it could help
preserve insulin levels for up to 2 years
without major side effects. In contrast
to anti-CD3 therapy, the GAD vaccine,
developed by a Swedish company,
Diamyd Medical, is given in two or
three simple injections. ‘‘And it seems
to be just as effective, at least after
21 months of data,’’ says Anders
Essen-Moller, the company’s CEO.
The antigen-specific approach has
its own drawbacks. In theory, it could
precipitate the very immune reaction
that it is supposed to prevent, as has
been observed with a similar treatment
for multiple sclerosis. (However, no
such adverse event has been reported
with for GAD vaccine.) Further, since
the pathophysiology of type 1 diabetes
is still poorly understood, it is not en-
tirely clear which antigen(s) would be
the best to target. Other than Diamyd’s
GAD vaccine, most antigen-specific
approaches tried so far have been clin-
ically disappointing. Some experts fear
that the antigens involved may vary be-
tween patients and with time, and pres-
ent a moving target for immunotherapy.
Could systemic and antigen-specific
agents be combined? In 2006, Matthias
von Herrath and other researchers from
the La Jolla Institute for Allergy and Im-
munology in San Diego showed that
aninsulin-relatedpeptidevaccinecould
work in synergy with an anti-CD3 com-
pound to cure type 1 diabetes in mouse
models. For their experiments, the re-
searchers used molecules developed
specifically for mice. ‘‘We need to de-
velop this kind of therapy for humans,’’
says Herrath. ‘‘In order to do this, we
need flexibility on the part of companies
tosharetheirPhase3productsforcom-
bination trials.’’ However, this is unlikely
to happen very soon, says Vaickus. He
points out that it took decades for drugs
in oncology to be used in combination.
‘‘You can’t just throw in two or three
agents together when they are in devel-
opment,’’ he says. ‘‘You need to first
find out what each does singly.’’
Both types of therapies share one
major limitation: neither succeeds in
inducing normoglycemia, or normal
blood sugar levels, without insulin. In
contrast, by some estimates more
than 200 therapies have succeeded
in producing a complete remission of
type 1 diabetes in the NOD and other
rodent models. Some experts attribute
this to the differences between the
rodent immune system, which is well
characterized, and the human immune
system, which is still something of a
mystery. ‘‘Our knowledge of immunol-
ogy is a drop in the bucket compared
to the complexity of the human im-
mune system,’’ says Vaickus. Others
point out that it is much harder to study
type 1 diabetes in humans, since the
b cells that are key to the disease lie
hidden in the pancreas. ‘‘We know a
lot about the mouse, but we lack a
good understanding of human type 1
diabetes,’’ says Insel.
Despite these challenges, a massive
worldwide effort is underway to find
more effective therapies for the dis-
ease. Besides nearly a dozen com-
pany-sponsored trials, several studies
are being coordinated by TrialNet, an
international clinical trial network for
type 1 diabetes. Some of these studies
are looking at nutritional interventions.
Others are testing drugs that have
already been approved for other auto-
immuneconditionssuchasrheumatoid
arthritis. Cellular approaches based on
dendritic cells could also emerge in the
near future, according to TrialNet study
chairman Jay Skyler. ‘‘We may see
a whole variety of treatments blossom-
ing in the next few years,’’ he says.