[Ip-health] Costs of drug development

[James Love]

Richard Tren's use of the Tufts' $802 million figure for drug development
costs is just the most recent instance when this number enters policy
debates on medicines.  Estimating the costs of drug development can be a
tricky thing, because there are a lot of different questions to ask and ways
of thinking about this.  Clearly a lot can be spent on the pre-clinical
stage to find a treatment for a particular illness, and also one can get
lucky and spend very little.  Clinical trials can cost a few thousand
dollars to several million dollars.  Drug development costs for different
classes of drugs vary a lot, or even drugs within the same class.   One
useful study that probably has not been given enough attention is the
excellent report by the TB Alliance.  It is useful because it provides more
detail than any previous published study to justify its findings, it
provides estimates of developed and developing country costs of clinical
trials, and it also deals with preclinical expenditures.       This study
does not address the allocation of those costs between public and private
funders.   Here are a few sections from the report, but the best way to
appreciate the study is to read the whole thing.... it is packed with data.

http://www.tballiance.org/3_per.cfm?rm=economics&sub=per

The TB Alliance report on drug development costs is a pretty good document
to work with.

Drug Development Costs

      The costs--including the costs of failure--to develop a new chemical
entity (NCE) to treat TB are estimated to be approximately $76 million to
$115 million, including preclinical development, pharmaceutical development,
and clinical trials. Including rough estimates of discovery brings the
estimated total to between $115 million and $240 million (including the
costs of failure). (All values are in U.S. dollars.)

      Estimating Discovery Costs
      Discovery costs for an anti-TB drug are roughly estimated to range
from $40 million to $125 million (including failure costs).

      As suggested by the breadth of this range, discovery costs are
difficult to estimate. However, it generally is accepted that discovery and
development of a new drug to treat TB will require an international,
collaborative effort. In this way, discovery costs can be shared by multiple
organizations, ultimately lowering the investment burden borne by a single
agency or company.

      Estimating Preclinical Development Costs
      The total costs of proposed preclinical studies required to support
registration based on a clinical dosing period of 3 to 6 months range from
$4.9 million to $5.3 million.

      The preclinical studies used to calculate these costs are those likely
to be needed for an NCE that has not been previously evaluated in
preclinical or clinical studies. Proposed studies include adequate
toxicology studies to allow at least 6 months of clinical administration as
well as to satisfy all of the requirements for regulatory approval. Also
proposed are pharmacokinetic and absorption, distribution, metabolism, and
elimination (ADME) studies. Cost estimates were obtained from a survey of
contract research organizations specializing in microbiology, toxicology,
and drug metabolism. Due to these organizations' lower recharge costs, these
costs might appear lower than average costs at larger, research-based
pharmaceutical companies.

      Estimating Pharmaceutical Development Costs
      In general, the overall costs for the chemistry, manufacturing, and
controls (CMC) development program are estimated to be at least $5.3
million, although Paraxel's 1999 Pharmaceutical R&D Statistical Sourcebook
suggests that these costs could be as high as $8 million.

      The estimated costs should be considered an approximation based on
previous development costs for marketed drugs. Factors such as development
timelines, the complexity of the synthetic route, cost of chemical
intermediates, amounts of drug product needed for clinical testing, or other
factors could significantly alter the development costs.

      Estimating Clinical Trials Costs
      In an established economy, clinical trials for an NCE to treat TB are
estimated to cost $26.6 million and to take 7 to 10 years to complete.
Comparable studies in a developing economy are estimated to cost $9.9
million. The costs related to the probability of failure are not included in
these cost estimates.

      From the initial Phase I trial of a new TB agent in humans, it is
estimated that it will take approximately 10 years to gain regulatory
approval. Time savings of up to 3 years might be possible if some Phase I
trials are conducted concurrently, or if provisional regulatory approval is
sought following completion of the 6-month drug therapy or after completion
of 6 months of the total 24-month follow-up period.



--------------------
James Love, mailto:james.love@cptech.org, http://www.cptech.org
voice +1.202.387.8030, mobile +1.202.361.3040, fax +1.202.234.5176