And here we have very
similar opinion: the author is trying in a very correct and polite manner
prove that the targeted approach is a bunch of BS:
Try as they might, the scientific
sleuths working for big drugmakers and aspiring biotechs have, generally, had a
tough time in recent years developing a large cache of new medications.
Certainly, there have been successes, but the industry continues to lament the
challenges in filling its pipeline. Meanwhile, there are complaints and
concerns about the ensuing costs involved. So what might be the best approach –
targeting diseases or mechanisms of action? Drug discovery is not an
all-or-nothing pursuit, but which might provide a better outcome? Frank
Sams-Dodd, chief scientific advisor to the Aslepios Bioresearch venture capital
firm, co-founder of the Willingsford device maker and a former head of
psychopharmacology at Boehringer Ingelheim has this to say…
For the past 20 years the pharmaceutical industry has invested billions
of dollars into drug discovery and development. However, in spite of these
investments, the industry has not been able to develop sufficient numbers of
new drugs to replace existing drugs coming off patent and has therefore not
been able to maintain its profitability. The result has over the past 5-7 years
been increasing lay-offs, site closures and program terminations.
So, the key question is: what is the reason for this productivity
failure? There have been many explanations, but a consistent observation is
that the majority of new drugs fail during development due to lack of efficacy.
This in turn indicates that it is how the industry conducts the drug discovery
process that is the fundamental cause of the problem.
For most diseases, we do not understand the cause of the disease, we do
not understand how the disease process affects the body and we do not
understand how the majority of the effective drugs on the market interfere with
the disease process or the symptoms to exert their therapeutic effects. A good
example of the difficulty in developing new drugs is the current crisis with
antibiotic resistant bacteria – we understand the cause of an infection (we can
look at it in a microscope), but in spite of the fact that we have full disease
insight we cannot develop an effective treatment.
The consequence of this is that the development of new treatments
essentially is a question of how many new mechanisms, i.e. components in the
body that a drug affects, or combinations of mechanisms that we can test for
therapeutic efficacy in a particular disease – the more mechanisms the higher
the probability of finding a treatment. It is not known how many possible ways
of affecting the body there exist, but we do know that new biological
mechanisms and complex interactions between different systems in the body are
constantly discovered, so it is safe to assume that there is a very large
number and that it includes many that we have not yet discovered.
Drug discovery can essentially be conducted using two different
approaches: 1) a disease-based approach, where drugs are screened directly in a
disease model for their ability to reduce the disease process or symptoms; and
2) a target-based approach, where the researcher first selects the mechanism
that the drug should affect, next develops a drug with these properties and
finally tests the drug in a disease model (often the same model as used for the
disease-based approach) to demonstrate that the target-selective drug has
therapeutic effects.
The advantage of the disease-based approach is that we do not need to
understand how the drug works – the biology will show us whether the drug has
the desired effect and this is enough. In contrast, the target-based approach
requires that we select the mechanism whereby to treat the disease, which means
that this approach is limited by how well we understand the disease and the
biology of the body. The disadvantage of the disease-based approach is that it
is time consuming to screen large numbers of drugs in complex disease models,
whereas for the target-based approach thousands of drugs can be screened every
week for selectivity against the target before final validation of therapeutic
efficacy in a disease model.
It is the ability to automate and screen large numbers of drugs that has
made the target-based approach the dominant approach in industry, but what if
we compare the two approaches based on the number of mechanisms that are
evaluated per drug and the approximate costs per evaluated mechanism – because,
as said above, what really matters is the number of mechanisms being evaluated.
For the disease-based approach each drug will typically affect anywhere from 1
to 3 or more different mechanisms and the costs of screening may be anywhere
from $200 to $50,000 per drug depending upon the complexity of the model and
can take anywhere from 1 day to 2 months.
For the target-based approach, it is normally necessary first to develop
a selective drug, which affects the chosen target, before the drug can be
tested in a disease-model. It is during this drug-optimisation process that
thousands of compounds are screened, but at the end of the day, the entire
exercise will still only have resulted in 1 single mechanism having been tested
for possible therapeutic effects. For the target-based approach the costs per
tested mechanism will therefore range from $1 mil to $50 mil, depending upon
how many chemists that have been allocated to the project, and it may take
anywhere from 6 months to 2-3 years.
So, if we compare the two methods based on costs and time per mechanism
tested there is no doubt that the disease-based approach is much cheaper, but
is it also more effective at finding treatments due to the fact that a larger
number of mechanisms are tested? Yes, a study analysed the research strategy
that had been used to develop first-in-class drugs during the period of
1999-2008. During this period 75 new first-in-class drugs were approved by the
FDA and of these only 40% had been developed using the target-based approach.
Considering that the target-based approach has almost completely dominated drug
discovery during this period, this shows that the target-based approach is less
likely to lead to an approved drug.
Two additional studies indirectly support this conclusion. One study
found that the attrition rate for phase I, II and III increased substantial
during the period from 1990 to 2004, when the target-based approach has
increasingly been dominating drug discovery. The second found that only 3% of
drug discovery projects focusing on a novel target will lead to a drug being
taken into pre-clinical development (toxicology), meaning that 97% of the
projects fail during discovery.
So, why is the target-based approach the dominant drug discovery
paradigm in industry? Historically, it developed as a response to the
limitations of the disease-based approach, where it is not possible to screen
large numbers of drugs and where the drug discovery process essentially is by
trial-and-error. By selecting the mechanism through which the drug should act,
it becomes possible to stream-line and to industrialize the discovery process,
and thereby to screen thousands of drugs per week. At a theoretical level the
approach is very attractive, i.e. the disease is treated by targeting the cause
of the disease, thereby avoiding side-effects, and it seems more scientific by
appearing to be based on an understanding of the disease and how the drug
works.
However, what was not taken into account was the extremely low
likelihood of selecting the correct target due to our limited insight into
diseases and the complexity of the body. Also, by reducing disease treatment to
a question of targets, it was no longer necessary to incorporate the true complexity
of the disease and the clinic into the drug discovery programs, whereby the
research lost touch with the clinical reality.
More recently, the screening throughput of the two approaches have also
become more comparable. For screening using the target-based approach, it has
become more common to work with smaller drug-libraries consisting of carefully
selected drugs instead of huge libraries that include everything and the
technology for the automated analysis of cell-based disease models and animal
behavior has evolved considerably. Altogether, this means that the original
arguments for using the target-based approach are beginning to disappear.
The intention here is not to say that the target-based approach is
wrong. There are definitely cases, where this is the optimal approach for
developing a new treatment, because it does have a number of advantages, but
only for those cases, where a target can be selected, which has a high
probability of being therapeutically effective. This is unfortunately not the
case for the majority of diseases and in these cases other approaches are
likely to be more successful because more drug mechanisms can be tested over a
shorter duration of time.
The target-based approach also has a competitive disadvantage, which
rarely is considered: if a company develops a drug using the target-based
approach, their competitors can easily copy this drug because it is known to
everyone how it works. The market exclusivity period is therefore only a few
months for target-based drugs whereas it is years for drugs developed by the
disease-based approach, because the mechanism of action is unknown and
competitors can therefore not easily copy the drug.
So, let us return to our initial question. If the Pharma industry
chooses a drug discovery approach, where the evaluation of each single
mechanism costs between $1 million and 50 million dollars and may take between
6 months and 2 years and where the likelihood of success for each mechanism is
close to zero, because we do not understand the disease biology, then this in
itself will place a severe limitation on the number of drugs that can advance
from research into clinical development each year and can in part explain why
the increased level of investments in R&D have not resulted in more new
drugs.
No comments:
Post a Comment