from new molecules to
leads for innovation
studies on the post-innovation
learning cycle for pharmaceuticals
Lay-out inside work:
The work presented in this thesis was performed at the Division of Pharmaco-
epidemiology & Pharmacotherapy, Utrecht Institute for Pharmaceutical Sciences,
Faculty of Science, Utrecht University, The Netherlands.
CIP-gegevens Koninklijke Bibliotheek, Den Haag
From new molecules to leads for innovation – Studies on the Post-Innovation
Learning Cycle for Pharmaceuticals
Thesis Utrecht University – with ref. – with summary in Dutch
© 2008 Pieter Stolk
Francis te Nijenhuis, zonnezijn creaties, 's-Hertogenbosch
Francis te Nijenhuis, zonnezijn creaties, 's-Hertogenbosch
Optima Grafische Communicatie, Rotterdam
From new molecules to leads for innovation –
Studies on the Post-Innovation Learning Cycle
Van nieuwe moleculen tot aanwijzingen voor innovatie –
Studies aan de post-innovatie leercyclus voor geneesmiddelen
(met een samenvatting in het Nederlands)
ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van
de rector magnificus, prof.dr. J.C. Stoof, ingevolge het besluit van het college
voor promoties in het openbaar te verdedigen op maandag 15 september 2008
des middags te 4.15 uur
geboren op 21 augustus 1978 te Rotterdam
Prof.dr. H.G.M. Leufkens
Dr. E.R. Heerdink
Dit proefschrift werd (mede) mogelijk gemaakt met financiële steun van het
Ministerie van Volksgezondheid, Welzijn en Sport.
Embedding new drugs in the health system
2.1 Changes in the defined daily dose: CYP2D6/CYP3A
metabolism as an indicator for dose setting problems
2.2 Variability in use of newly approved drugs with or
without an orphan designation
2.3 Rare essentials: drugs for rare diseases as essential
Effects of policy (interventions) on the use of drugs in clinical
3.1 Impact analysis of the discontinuation of reimbursement:
the case of oral contraceptives
3.2 Variable access to clopidogrel in a harmonised EU market
3.3 Between country variation in the utilisation of
antihypertensive agents: guidelines and clinical practice
3.4 Seige-cycles as a learning device in pharmacovigilance
Pharmacoepidemiology as a learning device in pharmaceutical
4.1 The association between exposure to COX-2 inhibitors
and schizoprenia deterioration: a nested case-control
4.2 Taking low dose aspirin is associated with more stable
drug treatment for lithium users
List of co-authors 211
List of publications 217
About the author 221
The pharmaceutical arena is evolving constantly: new drugs enter the market
while older ones are discontinued, clinical practice changes, health care budgets
rise and fall, and public health needs are redefined. Against this background, the
achievement of three main policy goals related to pharmaceuticals has to be
1. To ensure patient access to safe and effective medicines that are used in
the context of a high quality delivery system;
2. To allocate scarce resources in a health system in such a way that
pharmaceutical spending remains sustainable, while optimal health
outcomes for the individual patient are achieved;
3. To create an environment where innovation is rewarded, and the aims of
(national) industrial policies are achieved.
The ways in which societies reconcile these three goals are manifold, leading to a
variety of health systems, allocation policies, and usage patterns.
several important issues have converged that pose important challenges to the
fulfilment of the main policy goals. For example, there are growing concerns
about the ability of the current system to satisfactorily respond to issues
surrounding drug safety. This may lead to a decrease in the trust that patients and
health professionals have in the health system as a vehicle to deliver high quality
expenditures due to demographic developments and the increasing per unit
treatment costs of new drugs.
containment. Also, the number of innovative drugs coming to the market is
decreasing and profit margins for companies are under pressure, which will
change the structure of the industry in the coming years.
extent or source of the dearth of drug innovation is unknown, changing
regulatory requirements may be a key issue.
discussion about whether or not innovation is matched to real public health
needs. This was analysed in detail in the 2004 report of the World Health
Organisation (WHO) ‘Priority medicines for Europe and the world’.
mismatch not only relates to diseases that are highly prevalent in low- and
middle-income countries, but also to diseases that affect patients around the
world (e.g. Alzheimer’s disease, stroke and the orphan diseases).
6,7 Furthermore, governments are faced with growing health care
8 This leads to strong incentives for cost
9,10 Although the precise
11 Moreover, there is a continuing
Between these challenges that confront regulators worldwide there are many
cross-links: interventions to address one issue often have a broader impact as
well. An example for this can be found in a recent retrospective cohort study of
patients with acute coronary syndrome, discharged from hospital and receiving
posthospital treatment with clopidogrel (Plavix
study, an increase in mortality in the first 90 days after the discontinuation of
clopidogrel was found.
alternatives such as acetyl salicylic acid, this drug has been a frequent target for
cost containment interventions by regulators and third party payers. Therefore,
reimbursement policies may have played a role in determining the duration of
use for individual patients. One of the authors of the study, Dr. John Rumsfeld,
remarked: “Cost must be an issue here. If clopidogrel cost the same amount as
aspirin, perhaps we would be recommending indefinite use of this drug as
This example shows that careful assessment of the possible impact of regulatory
intervention on the usage environment is warranted. Ideally, proposals for policy
changes should be backed up by information about current use or the impact of
prior interventions. In his evaluation of regulatory requirements during the
development phase, Rawlins has proposed two criteria to evaluate all
regulation? And, does each regulatory requirement offer value for money? To
answer these questions in a satisfactory way, information from the usage
environment is needed.
Furthermore, data form the usage environment can also provide important
information and incentives for drug development and innovation. The study on
the clinical impact of the discontinuation of clopidogrel described above is such
an example. New insights from drug use in clinical practice is one of the three
main routes for ‘post-innovation innovation’: the discovery of new indications
after market entry.
example through pharmacoepidemiological studies, can contribute to the learn-
confirm cycle of drug development.
To assist in thinking about the usage environment of drugs in a comprehensive
manner that includes regulation, clinical outcomes, and incentives for
innovation, we propose a conceptual learning cycle for pharmaceuticals that
incorporates all these elements. Results from studies on this learning cycle can be
®), an antiplatelet drug. In the
13 Because of the high costs of clopidogrel compared to
15 is there a clear evidence-base to support the continuing of the
17 In this way, insights from use in clinical practice, for
helpful in designing future policies, as well as identifying opportunities for
optimising drug use and the innovation process.
THE LEARNING-CYCLE FOR PHARMACEUTICALS – A CONCEPTUAL
The cycle that we want to use as the conceptual framework is described in
Figure 1. The framework contains a ‘pre-innovation’ phase and a ‘post-
innovation’ phase. The distinction between innovation and post-innovation is
based on the work of David Banta, who places the moment of ‘innovation’ at
the start of clinical use.
part of the cycle that begins with the embedding of a new drug in the existing
health care system, and ends with the leads for innovation that arise from use in
19 In this thesis we want to focus on the ‘post-innovation’
FIGURE 1 - A proposed learning cycle for pharmaceuticals
The first step in the cycle is when a drug receives a market authorisation by a
regulator such as the European Medicines Agency (EMEA) or the US Food and
Drug Administration (FDA) (Section 1). At this moment, the new drug or drug
class must be embedded in the existing regulatory and health system. For
example, payers have to make a decision about whether or not the drug should
be reimbursed, and professional organisations have to make a decision about the
role of a drug in clinical practice.
In the next section of the cycle, the drug is taken up and used in clinical practice
by patients and health professionals (Section 2). During this period, more
information comes available about the benefits and risks of the new drug. Use in
clinical practice is heavily influenced by reimbursement policies and guidelines.
Based on the position that the drug attains in clinical practice and the outcomes
of drug treatment, the therapeutic needs of the population may change or leads
for new indications or future drugs are discovered; both of these provide
incentives for pharmaceutical research and development (Section 3) .
Next, we come to the drug development (pre-innovation) phase of the cycle
(Section 4). This section contains all drug discovery and clinical development
within academia and industry, including all pre-marketing activities of regulators
such as FDA, the EMEA and national authorities. Ideally, the product of this
process is marketed as a new drug, returning us to Section 1 of the learning cycle.
In this thesis we will use a variety of analysis tools to study the links between the
first three sections of the learning cycle for pharmaceuticals, the phase for post-
innovation learning. We will evaluate how these tools can best be used and
adapted for this purpose. The tools used in this thesis include
pharmacoepidemiological methods, multicountry comparisons and case studies.
We are especially interested in the international context, making use of the
natural variability of policy environments and health care systems.
POST-INNOVATION LEARNING FOR PHARMACEUTICALS (PILLS) -
OBJECTIVES OF THIS THESIS
A primary objective of this thesis is to develop a set of analytical tools to study
the post-innovation learning cycle of pharmaceuticals. With these tools we aim
to provide an evidence base for the formulation of policies that want to achieve a
sustainable balance between providing good quality health care, stimulating the
optimal allocation of scarce resources, and fostering an environment where
innovation is adjusted to real public health needs. Finally, we want to identify
directions for future research.
OUTLINE OF THIS THESIS
This thesis contains nine studies divided in three chapters. Each chapter is located
on a section of the post-innovation learning cycle presented in Figure 1.
In Chapter 2 we will focus on several challenges for policymakers that arise
from specific characteristics of new molecules, Section 1 in the cycle. An example
that we will study is the Defined Daily Dose (DDD), which plays an important
role in the price-setting systems of many countries. Also, we will explore the
implications of new therapeutic groups for policies of global organisations such as
the WHO. Finally we look at how the use of orphan drugs varies in different EU
Chapter 3 focuses on the relationship between regulation and use in clinical
practice, represented by Section 2 in the cycle. Firstly, we look at a policy
interventions at the micro level: the discontinuation of the reimbursement of oral
contraceptives. In the second study we use a macro perspective by looking at the
relationship between guideline preferences and the use of antihypertensives in
clinical practice in a multicountry setting. Another macro level study looks at
challenges to the payers in the health system. Here we look at the differences
between the EMEA market authorisation, the national reimbursement conditions
and actual use in nine EU member states. Finally, we use a life-cycle perspective
to look at how signals from the usage environment are translated into a
regulatory response for two drug safety cases, the market withdrawal of
Serotonin Inhibitors (SSRIs) in children.
In Chapter 4 we present two studies that show how information from actual use
in clinical practice can provide leads for drug development, the third section in
the cycle. We concentrate on the field of psychiatry and two classes of drugs that
have been linked to possible beneficial effects in patients with severe psychiatric
®), an expensive antithrombotic drug that poses special
®) and the safety issues surrounding suicide and Selective
illness: cyclooxygenase-2 (COX-2) inhibitors in schizophrenia, and non-steroidal
anti-inflammatory drugs (NSAIDs) in bipolar disorder.
In Chapter 5 the results from the earlier chapters are discussed in the context of
the post-innovation learning cycle and, based on this, we will provide a synthesis
and directions for future research.
1. Mossialos E, Mrazek M, Walley T (editors). Regulating pharmaceuticals in Europe:
striving for efficiency, equity and quality. Maidenhead (UK): Open University
2. Ess SM, Schneeweiss S, Szucs TD. European healthcare policies for controlling
drug expenditure. Pharmacoeconomics 2003;21:89-103.
3. Garattini L, Cornago D, De Compadri P. Pricing and reimbursement of in-patent
drugs in seven European countries: a comparative analysis. Health Policy
4. Goossens H, Ferech M, Vander Stichele R, Elseviers M, ESAC Project Group.
Outpatient antibiotic use in Europe and association with resistance: a cross-national
database study. Lancet 2005;365:579-587.
5. Bégaud B, Bergman U, Eichler HG, Leufkens HG, Meier PJ. Drug reimbursement:
indicators of inappropriate resource allocation. Br J Clin Pharmacol 2002;54:528-
6. Fontanarosa PB, Rennie D, DeAngelis CD. Postmarketing surveillance – lack of
vigilance, lack of trust. JAMA 2004;292:2647-2650.
7. Couzin J. Gaps in the safety net. Science 2005;307:196-198.
8. PriceWaterhouseCoopers. Pharma 2020 - The vision: Which path will you take?
PriceWaterhouseCoopers; 2007. Available from: http://www.pwc.com/extweb/
pwcpublications.nsf/docid/91BF330647FFA402852572F2005ECC22 (Accessed 17
9. Hughes B. 2007 FDA drug approvals: a year of flux. Nat Rev Drug Disc 2008;107-
10. Beyond the pill. Economist, 2007 October 27, p. 76.
11. Rawlins MD. Cutting the cost of drug development? Nat Rev Drug Disc
12. World Health Organisation. Priority medicines for Europe and the world. Geneva:
World Health Organisation; 2004.
13. Ho PM, Peterson ED, Wang L, Magid DJ, Fihn SD, Larsen GC, Jesse RA,
Rumsfeld JS. Incidence of death and acutre myocardial infarction associated with
stopping clopidogrel after Acute Coronary Syndrome. JAMA 2008;299:532-539.
14. Hughes S. Rebound Effect on Stopping Clopidogrel. Medscape Medical News.
Available from: http://www.medscape.com/viewarticle/569917 (Accessed 17 June
15. Rawlins MD. Cutting the cost of drug development. Nat Rev Drug Disc
16. Garrison LP, Neumann PJ, Erickson P, Marshall D, Mullins CD. Using real-world
data for coverage and payment decisions: The ISPOR real-world data task force
report. Value in Health 2007;10:326-335.
17. Gelijns AC, Rosenberg N, Moskowitz AJ. Capturing the unexpected benefits of
medical research. NEJM 1998;339:693-698.
18. Sheiner LB. Learning versus confirming in clinical drug development. Clin
Pharmacol Ther 1997;61:275-291.
19. Banta HD. Social science research on medical technology: utility and limitations.
Soc Sci Med 1983;17:1363-1369.
Changes in the defined daily dose:
CYP2D6/CYP3A metabolism as
an indicator for dose
Eur J Clin Pharmacol 2005;61:243-246
Interindividual variability is common at all stages of drug absorption, distribution,
pharmacodynamics, metabolism and elimination. In this study, we focused on
two enzymes involved in phase-I drug metabolism as markers of pharmacological
variability: the CYP3A and CYP2D6 subsystems of cytochrome P450. The main
aim of our study was to determine whether substrate drugs for CYP2D6 and/or
CYP3A enzymes, showing high interindividual metabolic variability, are more
prone to post-marketing adjustments of defined daily dose (DDD).
A case-control design was used. We identified all DDD changes between 1982
and May 2004 through the website of the World Health Organisation
Collaborating Centre for Drug Statistics Methodology. Cases were drugs with a
DDD change and controls were other drugs with unchanged DDDs. Information
about metabolism pathway, introduction year, literature exposure and
administration route was retrieved.
We included 88 cases and 176 controls. Of the 88 cases, 51 were dosage
decreases (58.0%). Overall, DDD changes were not associated with
CYP2D6/CYP3A metabolism (odds ratio [OR] 1.92; 95% confidence interval
[95%CI] 0.78–4.72). However, DDD decreases were associated with
CYP2D6/CYP3A metabolism (OR 3.21; 95%CI 1.25–8.26). Adjusting for
introduction year weakened this effect (OR 2.78; 95%CI 0.98–7.90).
Our study indicates that CYP2D6 and CYP3A substrates are more likely to
require a DDD decrease after granting of market authorisation. However, this
effect was diminished by adjusting for period of introduction. The implication of
this finding is that variability indicators, as is demonstrated in this study for
CYP2D6/CYP3A metabolism, can exert their influence on a wide variety of
drug measures, such as the DDD.
CYP2D6/CYP3A metabolism as an indicator for dose-setting problems
Various authors have stressed the provisional nature of assessments about
effectiveness and safety of medicines at the time when a new active substance is
introduced into clinical practice.
the basis of pre-authorisation studies.
Dosing of pharmaceuticals is a dynamic process, in which recommended dosages
may undergo changes over time. When Cross et al. studied labelling changes of
New Molecular Entities in the United States approved by the Food and Drug
Administration between 1980 and 1999, they found that one in five compounds
underwent a dosage change after marketing.
on 115 changes in the defined daily dose (DDD) – a dose measure developed and
maintained by the World Health Organisation (WHO) – and came basically to
the same conclusion.
antibiotics, while cardiovascular drugs underwent more dose decreases. An
important finding from both studies was that newer drugs were more susceptible
to post-marketing dose changes than older drugs.
Thus, optimising dosage strategies remains an important challenge for drug
life-cycle. Both pharmaceutical, clinical and economical determinants of
variability in dosing have been reported.
marketing dose reduction of captopril, where the initially recommended dose
was much higher than necessary for the vast majority of patients being prescribed
the drug in routine clinical practice.
In pharmacology, interindividual variability is common at all stages of drug
absorption, distribution, pharmacodynamics, metabolism and elimination. For
metabolism, interindividual variation in the phase-I metabolising enzymes of the
cytochrome P450 (CYP) system is a widely recognized source of between-
patient differences regarding drug therapy response,
in drug safety. Here, we want to focus on two enzymes involved in phase-I drug
metabolism as markers of pharmacological variability, namely the CYP3A and
CYP2D6 subsystems of CYP. The CYP2D6 and CYP3A enzymes together are
responsible for about 60–75% of phase-I reactions undergone by all drugs
metabolised through the CYP system and show extensive interindividual
1,2 The same holds for finding the right dose on
3 Our group has previously reported
4 DDD increases were most frequently associated with
5 Causes for dosage changes may be found anywhere in the drug
6,7 A classic example is the post-
9 and plays an important role
9 Also, these enzymes play an important role in many drug interactions.
The pathway from being a substrate of a metabolising enzyme system to
complexities in clinical practice wit finding the ‘right’ dose, and as a possible
consequence a change in a dose measure such as the DDD, is long and may be
full of erratic features (e.g. publication of new clinical trial results, marketing by
the industry and changes in good practice guidelines). However, although the
DDD is not an average recommended dose by definition, a DDD change can be
seen as a reflection of ‘noise’ surrounding the dosing of drugs in daily practice,
indicating situations where the actual prescribed dose has departed significantly
from the labelled use at the moment of setting the DDD. Accordingly, we use
the DDD as a measure for problematic dose setting in this study.
We hypothesise that drugs metabolised through these variable enzyme systems
are more susceptible to changes in dosage after marketing authorisation and,
consequently, the need for a DDD change. Therefore, the main aim of our study
was to determine whether substrate drugs for CYP2D6 and/or CYP3A enzymes
are more prone to post-marketing DDD adjustments.
From a drug safety perspective, special interest goes out to post-marketing DDD
decreases, since these represent cases where the average prescribed dose was
lowered over time, possibly instigated by safety concerns or the ‘overdosing’ of
the drug after introduction.
We used a design comparable to that used in the previously mentioned study on
Cases were drugs with a DDD change between 1982 and May 2004. Only the
first DDD change of a drug was included. Controls were randomly selected from
all other drugs for which a DDD was available. Excluded from the analysis were
drugs with a topical action, laxatives, drugs acting on the respiratory tract and
stomatological preparations, since these drugs probably have a limited systemic
absorption. Products with multiple active ingredients and drugs that show a high
between-patient dose variability due to a strong relationship between dose and
disease state (e.g. insulin, anti-anaemic preparations) were also excluded.
Furthermore, we excluded drugs with a low volume of use in Europe; only
4 Again, we made use of a case-control design for the analysis.
CYP2D6/CYP3A metabolism as an indicator for dose-setting problems
drugs that were marketed in the Netherlands were included in the analysis. For
each case, two controls were randomly selected.
Information about the DDD changes, drugs and route of administration was
retrieved from the website of WHOCC-DSM.
drug was ascertained, since this was a determinant in the previous study.
Information on whether drugs were substrates of the CYP2D6/CYP3A enzymes
was retrieved from the so-called ‘Flockhart’ CYP drug-interaction table.
To adjust for possible bias introduced by the prominence of a drug in the
scientific literature, we calculated a measure of ‘attention exposure’ in medical
journals. For all cases and controls, the sum of citations in MEDLINE as a
fraction of the total number of MEDLINE citations in 2 years before an index
year was ascertained.
All results were calculated using a logistic regression model and presented as odds
ratios (ORs) with 95% confidence intervals (95%CIs). In the model, metabolism
pathway, route of administration and decade of registration were entered as
categorical variables, ‘attention exposure’ in medical journals as a continuous
10 Year of global introduction of a
We included 88 cases and 176 controls. Of the 88 cases, 51 were dosage
decreases (58.0%). An overview of the cases and controls is displayed in Table 1.
The distribution of the fraction of MEDLINE publications (not shown in table)
indicated that drugs with reported DDD changes were more widely covered
than controls in the medical literature (p=0.036).
Being a substrate for either the CYP2D6 or CYP3A4 metabolism pathway was
not significantly associated with any DDD change (OR 1.92; 95%CI 0.78–4.72).
However, when we looked at only dosage decreases, the unadjusted OR for
metabolism through CYP2D6 or CYP3A (9 of 51 cases versus 11 of 176
controls) was 3.21 with a 95%CI from 1.25 to 8.26. Table 2 shows the
unadjusted and adjusted ORs for DDD decreases.
When CYP2D6/CYP3A metabolism results were entered into the regression
model together with the influence of the administration route, the effect of the
metabolic pathway on DDD decreases remained basically the same (OR 3.97;
95%CI 1.35–11.67). Adjustment for the number of publications on a specific
compound had no large effect (OR 3.03; 95%CI 1.17–7.87). CYP2D6/CYP3A
metabolism and decade of registration were also included in the logistic model,
since early registration may be associated with less information about the
metabolic pathway, fewer changes in dosing due to more experience with the
drug, and not having had a post-DDD-setting review by the WHO during the
study period. In this variant, the effect of CYP2D6/CYP3A on dosage decreases
lost significance (OR 2.78; 95%CI 0.98–7.90).
TABLE 1 — Characteristics of cases of Defined Daily Doses changes (all) and
n= =88 (100%)
n= =176 (100%)
Period of introduction
1971 — 1980
1981 — 1990
4.01 (1.94 — 8.31)
4.38 (2.15 — 8.90)
3.38 (1.56 — 7.32)
1.34 (0.79 — 2.30)
CYP2D6 or/and CYP3A
CYP3A4 and CYP2D6
6 ( 6.8%)
5 ( 5.7%)
1 ( 1.1%)
11 ( 6.4%)
7 ( 4.0%)
4 ( 2.3%)
0 ( 0.0%)
1.92 (0.78 — 4.72)
1.77 (0.58 — 5.42)
2.59 (0.68 — 9.90)
OR = Odds Ratio; NA = not applicable
TABLE 2 — Odds Ratios (OR) for Defined Daily Doses decreases in drugs
metabolised by CYP2D6 or CYP3A4
Unadjusted 3.21 (1.25 — 8.26)
Adjusted for: − route of administration
− exposure in medical journals
− period of introduction
3.97 (1.35 — 11.67)
3.03 (1.17 — 7.87)
2.78 (0.98 — 7.90)
CYP2D6/CYP3A metabolism as an indicator for dose-setting problems
Optimisation of drug dosing is key to successful drug development.
that being a substrate for either CYP2D6 or CYP3A (i.e. combining the
numbers for both groups) makes a drug about three times more prone to require
its DDD to be decreased after market authorisation has been granted. Although
this effect is diminished when adjusting for decade of introduction, a direction of
effect remains. This implicates that patients with early prescriptions of these drugs
immediately after introduction into clinical practice may be exposed to
inappropriate dose regimens.
Phase-I metabolism is a process susceptible to a great amount of variance.
Differences in metabolising activity and henceforth larger variability in plasma
levels and half-life can result in a greater variation in patient responses. The
nature of the variability is different for CYP2D6 and CYP3A metabolism: for
CYP2D6 it is caused by different alleles, resulting in a multi-modal distribution
of enzyme activity; the origin of variability in CYP3A metabolism is still the
subject of discussion, but a multi-gene or gene-environment interaction is
suggested by the unimodal nature of enzyme activity.
tried to link these intrinsic drug properties, while disparate in character, to
outcomes that are relevant from both clinical and regulatory perspectives.
Of course, there are many other sources of variability besides CYP3A and
CYP2D6 that influence dosing of drugs and that need to be explored further in
the future. For example, variability in other metabolising pathways, in
absorption, in distribution or in drug targets.
Recently, Kircheiner et al.
response for antidepressants and antipsychotics and pointed to the complex
nature and consequences that multigenic and gene-environment interactions at
different stages may have on treatment recommendations; response to
antihypertensive agents is also known to be dependent on phenotype.
For certain, there are other non-drug factors that are also of influence. In this
study, the decade of registration was a strong predictor for undergoing a DDD
change; this was also found in our previous study.
that, because of the extensive experience with drugs marketed before 1970, good
dosing strategies had already been developed in clinical practice long before the
DDD methodology was introduced in the mid-1970s.
5,8 We found
12 In this study, we have
13 reviewed the influence of phenotype on drug
4 The reason for this could be
Although we consider the methodology employed here useful to gain insight
into the studied process, there are some limitations that have to be addressed.
First of all, the DDD changes included in this study possibly represent an
underestimation of changes from the initial prescribed dose in daily practice. To
minimize effects on drug utilisation studies, the number of DDD changes is kept
as low as possible. Often only changes in the average maintenance dose of 50%
or more warrant a DDD change. For recently established DDDs and major
drugs, an exception is made; here smaller DDD adjustments are possible. Also,
some changes might be made for pragmatic reasons.
Bias may also have been introduced by the fact that well-known drugs are more
likely to undergo a change in the DDD and also have their metabolism pathway
elucidated. We tried to adjust for this by introducing the fraction of MEDLINE
publications as a parameter, including the influence of the decade of registration
and excluding drugs not often used in Europe. Furthermore, the table used for
determining enzyme substrates of drugs does not indicate the primary metabolic
pathway and may not be complete. However, the table is widely used as a
reference guide and provides the evidence base for the reported metabolic
pathways by referring to relevant publications.
One may argue that every need for adjustment of the dose after marketing is a
failure of drug development and/or the regulatory system.
predictability of pre-marketing research with respect to patient outcomes after
drug approval has improved significantly during the last decades. This paper
warrants an ongoing strive to invest in the linking of in vitro data on, for
example, metabolising properties of drugs with evaluations in real clinical
4 For sure,
In conclusion, our study indicates that CYP2D6 or CYP3A substrates are more
likely to require a DDD decrease after granting of market authorisation.
However, this effect was diminished by adjusting for period of introduction. The
implication of this finding is that variability indicators, as is demonstrated in this
study for CYP2D6/CYP3A metabolism, can exert their influence on a wide
variety of drug measures, such as the DDD.
CYP2D6/CYP3A metabolism as an indicator for dose-setting problems
For the future, the interactions between variability in dosing and variability
indicators, whether pharmacodynamic or pharmacokinetic, warrant further
Acknowledgements. The authors would like to thank Professor John Urquhart and
Kees de Joncheere (WHO-Europe) for their helpful comments in the design and
interpretation phase of this study.
1. Jefferys DB, Leakey D, Lewis JA, Payne S, Rawlins MD. New active substances
authorized in the United Kingdom between 1972 and 1994. Br J Clin Pharmacol
2. Stricker BH, Psaty BM. Detection, verification, and quantification of adverse drug
reactions. BMJ 2004;329:44–47.
3. Cross J, Lee H, Westelinck A, Nelson J, Grudzinskas C, Peck C. Postmarketing
drug dosage changes of 499 FDA-approved new molecular entities, 1980–1999.
Pharmacoepidemiol Drug Saf 2002;11:439–446.
4. Heerdink ER, Urquhart J, Leufkens HG. Changes in prescribed drug doses after
market introduction. Pharmacoepidemiol Drug Saf 2002;11:447–453.
5. Tucker GT, Houston JB, Huang SM. Optimizing drug development: strategies to
assess drug metabolism/transporter interaction potential – towards a consensus. Br J
Clin Pharmacol 2001;52:107–117.
6. Richter A, Anton SE, Koch P, Dennett SL. The impact of reducing dose frequency
on health outcomes. Clin Ther 2003;25:2307–2335.
7. Weinshilboum R. Inheritance and drug response. N Engl J Med 2003;348:529–
8. Temple R. Dose-response and registration of new drugs. In: Lasagna L, Erill S,
Naranjo CA (editors). Dose-response relationships in clinical pharmacology.
Amsterdam: Elsevier; 1989. p. 145–167.
9. Ingelman-Sundberg M. Pharmacogenetics of cytochrome P450 and its applications
in drug therapy: the past, present and future. Trends Pharmacol Sci 2004;25:193–
10. WHO Collaborating Centre for Drug Statistics Methodology. About the
ATC/DDD system. Available from: http://www.whocc.no/atcddd/atcsystem.html
(Accessed 17 June 2008).
11. Indiana University Department of Medicine, Division of Clinical Pharmacology.
Cytochrome P450 drug interaction table. Available from:
http://medicine.iupui.edu/flockhart/table.htm (Accessed 17 June 2008).
12. Lamba JK, Lin YS, Schuetz EG, Thummel KE. Genetic contribution to variable
human CYP3A-mediated metabolism. Adv Drug Deliv Rev 2002;54:1271–1294.
13. Kirchheiner J, Nickchen K, Bauer M, Wong ML, Licinio J, Roots I, et al.
Pharmacogenetics of antidepressants and antipsychotics: the contribution of allelic
variations to the phenotype of drug response. Mol Psychiatry 2004;9:442–473.
14. Schelleman H, Stricker BH, de Boer A, Kroon AA, Verschuren MW, van Duijn
CM, et al. Drug-gene interactions between genetic polymorphisms and
antihypertensive therapy. Drugs 2004;64:1801–1816.
15. Ito K, Brown HS, Houston JB. Database analyses for the prediction of in vivo
drug-drug interactions from in vitro data. Br J Clin Pharmacol 2004;57:473–486.
Variability in use of
newly approved drugs with
or without an orphan designation
Regulators and third party payers have to strike a balance between the needs of
the individual patient and the optimal allocation of scarce resources. Orphan
drugs, are a special group of in this context because of high per unit costs and for
usually not being able to fulfil the standard cost-effectiveness criteria. Our
objective is to determine how utilisation of centrally authorised drugs varies
across a selection of EU (European Union) member states, and in particular to
determine whether drugs that have received an orphan drug designation show a
different level of variability in use than drugs without an orphan drug
We selected five orphan drugs and nine other centralised drugs that were
centrally authorised in the EU between 1 January 2000 and 30 November 2006
and could also be used in the ambulatory setting. We compared utilisation of
these drugs in seven EU member states: Austria, Denmark, United Kingdom
(represented by England), Finland, Portugal, The Netherlands, and Sweden.
Utilisation data was expressed as Defined Daily Doses (DDDs) per 1000 persons
per year. For each drug relative standard deviations (RSD) across countries were
computed as a measure of variability in use. Per treatment costs and
innovativeness for each drugs were determined.
Drugs with an orphan drug designation are, in general, more expensive and have
a higher innovation score than drugs without an orphan drug designation. We
found no association between orphan drug designation status and variability in
use across countries.
Orphan drugs show no larger variability in utilisation than drugs without an
orphan drug designation. Heterogeneity in use may be a feature of the drug
market in the EU in general, and not restricted to one class of drugs.
Usage variability for orphan and non-orphan drugs
In every health system regulators and third party payers have to strike a balance
between the needs of the individual patient and the optimal allocation of scarce
resources. For new pharmaceuticals, national regulations and traditions are
important determinants for how individual drugs are embedded in the system.
Therefore, studying the variation in uptake of drugs across health systems can
provide information on how access to new therapies differs from one country to
another. Drugs indicated for the treatment or rare diseases, so called orphan
drugs, are a group of special interest in this context because of their high per unit
costs and for usually not being able to fulfil the standard cost-effectiveness
criteria, that are often used in reimbursement decisions.
such as the European Organisation for Rare Diseases (Eurordis), have called
attention to the heterogeneous and incomplete availability of Orphan Drugs in
the European Union (EU).
At the EU level, the European Medicines Agency (EMEA) provides a centralised
marketing authorisation procedure for new medicinal products, with a
harmonised Summary of Product Characteristics (SmPC), for the whole EU
since 1995. Therefore, the EMEA centralised procedure allows the comparison
of the response of different EU health systems for drugs of which the quality,
safety and efficacy was assessed by one and the same institution.
Currently, the centralised procedure is mandatory for biotechnology drugs and
for all medicines intended for the treatment of HIV/AIDS, cancer, diabetes,
neurodegenerative diseases, autoimmune and other immune dysfunctions, and
viral diseases. The procedure is mandatory for Orphan Drugs (ODs) as well. A
centralised procedure is optional for medicinal products constituting a
therapeutic, scientific or technical innovation (e.g. new chemical entities).
Our objective in this study is to determine how utilisation of centrally authorised
drugs varies across a selection of EU member states, and in particular to
determine whether drugs that have received an orphan drug designation, and
therefore may be more vulnerable to heterogeneity in access and use, show a
different level of variability in use than centrally authorised medicines without an
orphan drug designation.
1 As a result, organisations
We selected fifteen drugs that were centrally authorised in the EU between 1
January 2000 and 30 November 2006 and could also be used in the ambulatory
setting. We randomly selected five ODs: imatinib mesilate (Glivec
drugs: levetiracetam (Keppra
adefovir dipivoxil (Hepsera
In our initial selection we also included apomorfine (Uprima
market authorisation was not renewed for this drug in 2006, and because the
drug was only marketed in a few of the countries in this study, we excluded it
from our final analysis.
We retrieved information about the utilisation of these drugs in seven European
Union member states countries: Austria, Denmark, United Kingdom
(represented by England), Finland, Portugal, The Netherlands, and Sweden.
These countries represent a selection of EU member states from different regions
and with different health systems.
®) and sodium
®), zinc acetate dihydrate (Wilzin
®). In addition, we randomly selected ten other/non-orphan
®), desloratidine (Aerius
®), emtricitabine (Emtriva
®), oxybutinin (Kentera
®), abacavir/lamuvidine (Kivexa
®), nitisinon (Orfadin
®), apomorfine (Uprima
®), pregabalin (Lyrica
®), but since the
Utilisation rates of drugs included in the study
Utilisation data was requested for all countries included in the study. We
calculated drug utilisation rates used as a measure of uptake in the health system.
We determined utilisation rates for the year 2006, as this was the latest full
calendar year in the study period.
Utilisation rates were expressed as the number of Defined Daily Doses (DDD)
per 1000 inhabitants per year. The DDD is a standard dosage measure defined by
the World Health Organisation.
defined the DDD ourselves based on information about the average daily dose
contained in the official drug label.
3 If DDDs were not available for a drug we
Between country variability was determined by calculating relative standard
deviation (RSD) for the utilisation rates of individual drugs across countries. This
measure for variability was calculated as follows:
Usage variability for orphan and non-orphan drugs
Standard deviation of class utilisation as % of total antihypertensive use
100 × International average of class utilisation as % of total antihypertensive use
This method for calculating the variability in utilisation was used elsewhere as
4 Utilisation rates equal to zero were excluded from further analysis.
Innovativeness of individual drugs was rated according to a system based on an
algorithm designed by Motola et al.
in five classes according to two dimensions of therapeutic innovation; availability
of other treatments and the therapeutic effect. Scores for availability of other
treatments ranged from 5: drugs for diseases without recognised standard
treatment to 1: mere technological innovation. Scores for therapeutic effect
ranged from 3: major benefit on clinical endpoints or validated surrogate
endpoints to 1: minor or temporary benefit on some aspects of the disease. All
drugs included in the study were ranked according to the two dimensions of
innovation in this system. Where available, we used a list compiled Motola et al.
which was available from their website.
available, the innovativeness was rated by two of the authors (HH and PS). We
calculated the product of both scores as numeric indicator of therapeutic
innovativeness for all products in the study.
5 This algorithm divides newly marketed drug
6 For drugs for which no score was
As an indicator for cost differences between drugs we sampled the prices for each
drug in three of the countries in this study (Denmark, The Netherlands and
Sweden). Within each country, we ranked all drugs according to their price per
DDD. Based on the average within-country price ranking of in each of the
countries we determined an overall price rank for each drug. A score of 1 was
assigned to the cheapest drug; a score of 14 was assigned to the most expensive
The basic characteristics of the seven countries included in this study are shown
in Table 1. This Table also provides information about the data sources from
which the utilisation data in this study was retrieved.
TABLE 1 — Characteristics of countries (Source: OECD Health Data 2007 and EuroStat; all data for 2006)
Population aged 65
and over (%)
GDP per capita (US
32 896 a
as a share of GDP
spending per capita
(US Dollar PPP)
spending (% of GDP)
for reimbursement 9
economic evaluation may
one of the
analysis is used
to justify a
(Table 1 continued)
Claims data of
primary care prescribing
of drug costs
data from the
data from the
and IMS for
GDP = Gross Domestic Product; PPP = Purchasing Power Parity
a) Data for whole United Kingdom.
b) Indicates that data was not available from OECD.
TABLE 2 — Overview of the drugs included in the study
WITH > >0
11 Nov 2001
Chronic myeloid leukaemia, acute
diseases, advanced hypereosinophilic
syndrome (HES), chronic eosinophilic
leukaemia (CEL), metastatic
malignant gastrointestinal stromal
400 mg a
15 May 2002
Pulmonary arterial hypertension
zinc acetate dehydrate 13 Oct 2004
21 Feb 2005
Hereditary tyrosinemia type 1
13 Oct 2005
Cataplexy in adult patients with
29 Sep 2000
19 Apr 2002
65 mg c
(Table 2 continued)
WITH > >0
24 Oct 2003
HIV-1 infected adults and children in
combination with other antiretroviral
6 Mar 2003
Chronic Hepatitis B
26 Feb 2004
Treatment of the symptoms of urge
incontinence and/or increased
6 Jul 2004
Treatment of peripheral and central
20 Sep 2004
Treatment of moderate to severe
chronic plaque psoriasis
17 Dec 2004
Combination therapy for the
treatment of HIV infection
0.9 g c
21 Sep 2000
The relief of symptoms associated
with: allergic rhinitis and chronic
DDD = defined daily doses; RSD = relative standard deviations; HCT = hydrochlorothiazide; HIV = Human Immunodeficiency Virus
a) DDD determined by authors based on literature, not WHO-CC (World Health Organisation Collaborating Centre for Drug Statistics Methodology).
b) Innovativeness score determined by authors based on algorithm by Motola et al.5
c) Combination preparation, DDD determined according to WHO guidelines.
Table 2 provides an overview of the fourteen drugs included in the final analysis.
For each drug, the date of EU market authorisation, the indication, the DDD
used in the analysis, the innovation score, the price ranking, and the RSD for the
utilisation rate across the seven countries are reported. As the Table shows, drugs
with an orphan drug designation are, in general, more expensive and have a
higher innovation score than drugs without an orphan drug designation.
FIGURE 1 - Variability in use of drugs in 2006 in an innovativeness vs. price matrix
Horizontal axis describes therapeutic innovativeness, while vertical axis describes average price
rank. The size of the bubbles describes the variation in utilisation across the countries in the
study. No statistically significant difference in variation in utilisation can be observed (p=0.80).
In Figure 1 we have displayed the relationship between, cost, innovativeness and
variability in utilisation for each of the drugs. The innovativeness score for each
drug is depicted on the x-axis, the y-axis shows the cost ranking, while the
bubble sizes denotes variability in utilisation. The dark gray bubbles are orphan
Usage variability for orphan and non-orphan drugs
drugs, the light gray bubbles are drugs without an orphan drug designation. The
Figure can be divided in roughly four quadrants. The lower left quadrant
contains drugs with a low score for innovativeness and low treatment costs. The
upper right quadrant contains drugs with a high innovativeness score and high
treatment costs. An independent samples t-test showed that there is no
association between variability in utilisation and orphan drug designation status
The results from this study show that the variability in use for Orphan Drugs
appears to be comparable to the other newly authorised drugs that were included
in the analysis. This means that, although strong heterogeneity in access may
exist, orphan drugs may not be a ‘special’ group in this respect; heterogeneity
may be an intrinsic aspect of the drug market in the European Union as a whole.
Orphan Drugs rated higher on an innovativeness rating system than the drugs
without an Orphan Drug designation. This is not surprising, since the
requirements for an orphan designation are strongly congruent with the
requirements for an ‘Important’ innovation in the model of Motola et al.
the requirements for a drug to be eligible for an orphan drug designation is that
“there exists no satisfactory method of diagnosis, prevention or treatment of the
condition in question that has been authorised in the Community or, if such
method exists, that the medicinal product will be of significant benefit to those
affected by that condition.”
Given the small sample sizes, it is very difficult to disentangle the specific
contributions of innovativeness and cost to the variability in utilisation.
However, this study gives an indication that neither cost, nor innovativeness or
an orphan status influences the variation in utilisation between the countries in
the study. Therefore, this study indicates that access to orphan drugs should be
viewed in a broader context of access to medicines in general. Furthermore,
access should not be measured as a binary variable. Whenever studying access to
certain drugs, especially when comparing multiple health systems that are
characterised by diverging rules and regulations, access should be considered in
the light of actual use in clinical practice.
5 One of
Certain limitations apply to this study. When including per unit treatment costs
in our analysis we used a method of ranking drugs according to their cost per
DDD which only included costs in three countries, disregarding costs in the four
other countries in the study. However, given the small variation in relative costs
between these three European countries, it’s unlikely that relative costs in the
other four countries in the study would differ significantly. We have no reason to
believe that the relative prices for the drugs in this study would show large
variations when these other countries would have been included as well.
Furthermore, we only looked at the national level in this study and did not take
regional variation into consideration. Access to, and use of, drugs may show large
regional variability, for example, depending on policy and budget considerations
of individual hospitals or insurance companies.
Another limitation is that we did not include the prevalence of the diseases into
our analysis. For some of the primary indications of drugs included in this study
the prevalence may vary across countries, thus leading to an overestimation of
the relative standard deviation measure for the utilisation of these drugs. Also, for
zinc acetate, compounded alternatives may be available, which would make our
measurement of the utilisation of this drug an underestimate.
The data sources used in this study consisted of reimbursement and dispensing
data. Therefore, our results may have been influenced by differences between
reimbursement and utilisation within countries or by differences by the method
of procurement (public pharmacy, hospital pharmacy or both). As this limitation
is an inherent characteristic of the data, we have indicated the source of all our
data in Table 2. It also indicates that there is a need for a harmonised method of
data collection on drug utilisation within Europe.
Finally, the countries included constitute a selection of EU member states.
Results from this study should be extrapolated to other EU member states with
caution, in particular for those countries that joined the EU on or after 1 May
2004. These countries may well have specific challenges with access and use of
newly authorised drugs.
In conclusion, we found that orphan drugs show no larger variability in
utilisation than drugs without an orphan drug designation. Therefore,
heterogeneity in use may be a feature of the drug market in the EU in general,
and not restricted to one class of drugs. We would like to argue for studies on
access issues to take this variability in use into account. Future studies looking at
Usage variability for orphan and non-orphan drugs
access issues should also take into account the actual utilisation for a
comprehensive assessment of this topic.
1. Drummond MF, Wilson DA, Kanavos P, Ubel P, Rovira J. Assessing the
economic challenges posed by orphan drugs. Int J Technol Assess Health Care
2. Eurordis. Eurordis survey on orphan drugs 2007. Available from: http://
www.eurordis.org/article.php3?id_article=1013 (Accessed 14 June 2008).
3. WHO Collaborating Centre for Drug Statistics Methodology. About the
ATC/DDD system. Available from: http://www.whocc.no/atcddd/ (Accesssed 14
4. Stolk P, Van Wijk BL, Leufkens HG, Heerdink ER. Between-country variation in
the utilization of antihypertensive agents: Guidelines and clinical practice. J Hum
5. Motola D, De Ponti F, Rossi P, Martini N, Montanaro N. Therapeutic innovation
in the European Union: analysis of the drugs approved by the EMEA between
1995 and 2003. Br J Clin Pharmacol 2004;59:475-478.
6. Centro Regionale di Valutazione e Informazione sui Farmacia. Tharapeutic inno-
vation. Available from: http://www.crevif.it/innovation/index.htm (Accessed 14
7. Regulation (EC) No 141/2000 of the European Parliament and of the Council of
16 December 1999 on Orphan Medicinal Products.
8. Zwart-van Rijkom JE, Leufkens HG, Simoons ML, Broekmans AW. Variability in
abciximab (ReoPro) prescribing: evidence
Pharmacoepidemiol Drug Saf 2002;11:135-141.
9. The PPRI Network. Pharmaceutical Pricing and Reimbursement Information
(PPRI) report 2008. Available from: http://ppri.oebig.at/ (Accessed 14 June 2008).
based or budget driven?
drugs for rare diseases
as essential medicines
This chapter is based on an invited discussion paper
for the 14th meeting of the WHO Expert Committee
on the Selection and Use of Essential Medicines
(7-11 March 2005, Geneva, Switzerland)
Bull World Health Organ 2006;84:745-751
Since 1977, the World Health Organisation (WHO) Model List of Essential
Medicines (EML), has provided advice for Member States that struggle to decide
which pharmaceutical technologies should be provided to patients within their
public health systems. Originating from outside WHO, an incentive system has
been put in place by various governments for the development of medicines for
rare diseases (‘orphan drugs’). With progress in pharmaceutical research (e.g.,
drugs targeted for narrower indications), these medicines will feature more often
on future public health agendas. However, when current definitions for selecting
essential medicines are applied strictly, orphan drugs cannot be part of the WHO
Essential Medicines Programme, creating the risk that WHO may lose touch
with this field. In our opinion WHO should explicitly include orphan drugs in
its policy sphere by composing a complementary Orphan Medicines Model List
as an addition to the EML. This complementary list of ‘rare essentials’ could aid
policy-makers and patients in, for example, emerging countries to improve access
to these drugs and stimulate relevant policies. Furthermore, inconsistencies in the
current EML with regard to medicines for rare diseases can be resolved. In this
paper we propose selection criteria for an Orphan Medicines Model List that
could form a departure point for future work towards an extensive WHO
Orphan Medicines Programme.
In all health-care systems, there is a struggle to decide which technologies should
be provided to patients within the system. Criteria such as efficacy, need,
prevalence and cost-effectiveness are used in this selection process. These
struggles are particularly acute when considering pharmaceuticals. Since 1977,
World Health Organisation (WHO) has provided advice for countries by
defining a WHO Model List of Essential Medicines (EML).
EML as normative guidance and technical support has helped over 150 countries
to establish the principle that essential medicines save lives and improve health,
but only when they are available, affordable, of good quality, and properly used.
The fourteenth edition of the EML was published recently.
outside WHO, an ‘orphan drugs’ movement has developed primarily in affluent
countries since the early 1980s to create incentives for the development of
medicines for rare diseases.
are not attractive for pharmaceutical companies to develop and market.
While both are systems of prioritising resources and allocating incentives for
pharmacotherapy, the orphan drug movement and the WHO Essential
Medicines Policy have many differences in background, goals and conceptual
frame. However, it is becoming increasingly clear that they share common
ground, i.e. there are essential medicines for rare diseases. Although orphan drugs
have not been on the priority agenda of WHO because there are urgent
population health needs with a high disease burden to be met, this may change as
more orphan drugs come onto the market. For example, orphan drugs currently
constitute about 15% of new centralised authorizations in the European Union
(EU), there is increasing attention for ‘rare diseases’ in emerging countries (e.g.
Egypt, India) and more spin-offs of orphan drug innovations with implications
for drug treatment in general (e.g. imatinib mesylate, used for the treatment of
chronic myeloid leukaemia).
fields of orphan drugs and essential medicines, and propose how WHO may
develop an approach to provide useful advice to Member States that want to
improve access to treatments using orphan drugs. For this purpose, we would
like to recommend the creation of a complementary WHO Model List for
Orphan Medicines as an addition to the current EML. Furthermore, we aim to
1 The concept of the
3 Originating from
4 Because of their small market potential, such drugs
5 In this paper, we review recent advances in the
provide a framework for analysing future questions surrounding the selection of
‘essential orphan medicines’ or ‘rare essentials’.
MEDICINES FOR RARE DISEASES: SMALL NUMBERS WITH IMPACT
“Which diseases are classified as rare?” is not an easy question to answer, as we
have to deal with a complex mosaic of hard-to-categorise conditions. Many rare
diseases have a genetic basis. Often this is a monogenic modification, as in the
case of X chromosome-linked haemophilia or the defect in transmembrane
chloride ion transportation that causes cystic fibrosis.
Currently, several criteria to identify and classify rare diseases are found in orphan
drug legislation, which provides incentives for the development and marketing of
medicinal products for diseases that may otherwise suffer from non-viability of
the market. These market failures are mainly caused by scientific deficiencies
(e.g. small numbers of subjects for clinical trials, lack of knowledge about the
cause of the disease, absence of valid biomarkers), greater regulatory demands on
new drugs in terms of safety and effectiveness, possible obstacles in patenting, and
a lack of public awareness of the issue.
legislation was introduced in the United States of America (USA) in 1983. Other
countries (e.g. Australia, Japan, Singapore) followed in the 1990s, and in 2000
the EU established its own orphan drug legislation. Table 1 provides an overview
of the main features of orphan drug systems in the EU and USA. Methods used
in regulations to stimulate research and development of orphan drugs include
extended regulatory guidance and advice, waivers of regulatory fees and market
exclusivity. It is important to note that there are differences between the USA
and EU definitions of a rare disease. In the USA Orphan Drug Act, the
definition relates to an absolute number (<200 000 patients in the USA), while
the European regulation uses a relative measure (<5 cases per 10 000 inhabitants)
and requires disorders to be life-threatening and/or chronically debilitating.
When these definitions are used, it is estimated that between 5000 and 7000
conditions qualify as rare diseases, bringing the total number of patients suffering
from these diseases in Europe and the USA alone to 55 million.
other countries data are scarce, but the prevalence of rare diseases is likely to be
6 In response to this, the first orphan drug
4,7 For many
TABLE 1 — Features of the USA and EU Orphan Drug incentive system 11
Program established 1983 — Orphan Drug Act modified
the Federal Food, Drug and
2000 — Orphan Medicinal
for rare disease
Less than 200 000 patients in USA
Life-threatening or chronically
debilitating disorder that affects
less than 5:10 000 in EU
Rare disease, or R&D costs cannot
be recovered in 7 years
Rare disease, or product unlikely
to be developed without
incentives, or new product will
be of significant benefit
Products eligible for
Drugs and biologicals (including
vaccines and in vivo diagnostics)
Drugs and biologicals (including
vaccines and in vivo diagnostics)
Market exclusivity 7 years; prevents same product
being approved for the same
indication unless clinical
superiority is shown
10 years; can be reduced to 6 if
orphan criteria are no longer
Regulatory fee waivers, 50% tax
credit on clinical research after
designation; grants for clinical
research (pharma and academia
eligible); protocol assistance;
faster review if indication
warrants; research grants for
medical devices and medical food
Regulatory fees can be reduced
or waived; access to centralised
procedure; protocol assistance.
Individual Member States have
to implement measures to
stimulate the development of
orphan medicinal products.
USA = United States of America; EU = European Union
To prioritise limited public health resources it is important to possess reliable data
on disease burden, course of disease and long-term prognosis. This has been a
difficult task for rare diseases. A primary reason why sound epidemiological data
is often lacking is the absence of proper classification and coding for the disease
and the absence of registration of the patients suffering from rare conditions.
Although International Classification of Disease (ICD) codes are available for
some of the better-known rare diseases, such as thalassaemia, cystic fibrosis and
haemophilia, many orphan drugs are not included in medical registries and
databases. Often these rare disorders are grouped under higher classification levels
such as ‘endocrine metabolic disorders’. A second reason for the lack of reliable
epidemiological data is the frequent absence of appropriate biochemical and
genetic diagnostic data. Generally speaking, indicators to quantify disease burden,
such as the disability-adjusted life year (DALY), are not very useful in the case of
rare diseases, as the low prevalence brings DALY estimates for these diseases to
the bottom of any list created on the basis of burden of disease.
The impact of the Orphan Drug Act on drug development and public health in
the USA was evaluated in 2003, the 20th anniversary of its establishment.
the introduction of this legislation, about 1100 drugs have received an orphan
drug designation. Of these, 231 were marketed, providing an estimated 11
million patients in the USA with a new treatment for their disease. In the EU,
the first 5 years of orphan drug legislation were recently evaluated by the
European Medicines Agency (EMEA). Overall, the experience was positive; by
April 2005, more than 260 products had acquired an orphan drug designation,
and 22 of these received a marketing authorisation, creating new treatment
options for more than one million patients in the EU.
ACCESS TO AND AFFORDABILITY OF MEDICINES FOR RARE
Despite this progress, no effective and safe treatment is available for many rare
diseases. Furthermore, when treatments are available, obstacles are encountered
that hinder access and use of these drugs.
? Challenges in assessing clinical relevance and cost-effectiveness. The methodology for
evaluating orphan drug treatments is often still in an experimental phase,
hampering positioning in clinical practice.
? Lack of knowledge and training. For many rare diseases, available information is
inadequate. Health professionals are often deficient in appropriate training and
awareness to be able to diagnose and adequately treat these diseases. The aim
of initiatives like Orphanet
? Deficient diagnostic systems. For many diseases no diagnostic methods exist, or
diagnostic facilities are unavailable. In these cases, diagnosis may be
problematic. Consequently, validity, coding and reproducibility are problems.
? High prices. Prices of orphan drugs per treatment episode can be very high.
For example, the cost of treatment with enzyme replacement therapies may
reach more than US$ 150 000 per treatment-year. The affordability of orphan
drugs has become a major issue for payers and is a strong driver of tensions
7 is to address this issue.
between the different stakeholders.
by developing programmes to facilitate access to orphan drugs.
9 Some companies have responded to this
These obstacles to treating rare diseases with orphan drugs exemplify and mirror
the global debate of deficiencies in bringing new drugs to patients who need
them. The recent WHO report Priority medicines for Europe and the World
gives a thoughtful account of this and has provided a priority listing of gaps in
antibiotics. This crisis was linked to the orphan drug issue in a more general
context in Science magazine: “Will all drugs become orphans in the future, not
because of the rareness of the disease, but because other factors hinder investment
in drug discovery and development?”
Furthermore, advances in pharmacogenomics may lead to treatments benefiting a
small subgroup of patients.
an increasing number of drugs specifically indicated, and effective, for rare
diseases, these medicines will feature more often on future public health agendas.
ESSENTIAL MEDICINES: BIG NUMBERS WITH IMPACT
11 One of these gaps is the crisis in the development of new
13 Whatever the outcome, it seems inevitable that with
In 1977, the first Essential Drug List was published, containing medicines that
were indispensable for the health needs of the majority of the population.
2002, the definitions of the EML had changed. From then on essential medicines
were selected with ‘priority conditions’ in mind: they had to be evidence-based,
safe and cost-effective. Priority conditions were selected considering current and
future public health relevance.
published together: a ‘core’ list representing the minimum medicine needs for a
basic health-care system, and a ‘complementary’ list for medicines that address
priority health-care needs, but require specialised facilities/services, or are costly.
Within the context of the EML, medicines for ‘neglected diseases’ may be
included in the list on the basis of the criteria described above since they meet
the priority needs of a specific population (e.g. local high-prevalence conditions
such as trypanosomiasis), in contrast to ‘rare diseases’ (diseases with a low
Three major functions for the EML (and other WHO medicines policies) have
been identified: operational, educational and symbolic purposes.
14 The EML consists of two sections, which are
15 As an
Chapter 2.3 Download full-text
operational tool, the EML is an important guide for policy-makers and
programme managers to identify medicines that require priority attention in
terms of production, and access. Furthermore, the list is an educational tool for
health professionals and policy-makers, not only through improvement of
formulary building and utilisation, but also through the procedures used to select
WHO committee members and candidate medicines for the EML. Finally, the
list has a significant symbolic value. Classification as an essential medicine confers
worldwide recognition, preferred position in pharmaceutical management and
may stimulate related policies (e.g. production, infrastructure investments or the
establishment of quality systems).
While selection occurs at a global level, the EML concept should be
implemented nationally. Countries are invited and encouraged to formulate
national policies with the EML as a model to be adapted. This results in separate
national lists, which vary from the WHO list due to local circumstances such as
demographics, epidemiology, public health relevance, financial resources or
capacity of the health system. Whether a medicine is included in a national list
can be considered as an indicator for the level of adoption and dissemination of
the EML. A comprehensive overview of the differences between the EML and
national lists can be found in an analysis published in the Lancet.
is an ongoing debate about the impact of these lists on national drug use, the
balance sheet for the EML, particularly in less affluent countries, looks very
ORPHAN DRUGS AND ESSENTIAL MEDICINES
1 Although there
Although the fields of essential medicines and orphan drugs share principles of
social justice and equity, Table 2 lists some important ways in which the two
groups of medicines differ.
Two recent examples illustrate the tensions in the discussion about orphan drugs
within the WHO Expert Committee on the Selection and Use of Essential
Medicines: the cases of fludrocortisone and factor VIII/IX concentrates.
Fludrocortisone, indicated for adrenal insufficiency, was deleted from the EML
in 2003, because its rare indication did not meet the criterion of “satisfying the
priority health-care needs of the population”, it was on few national lists, and