Journal of Generic Medicines: The Business

Journal for the Generic Medicines Sector

http://jgm.sagepub.com/

Deficiencies in generic product dossiers as submitted to the WHO Prequalification of Medicines

Programme
Wondiyfraw Z Worku, John Gordon, Matthias MS Stahl and Lembit Rägo
Journal of Generic Medicines: The Business Journal for the Generic Medicines Sector 2012 9: 63
DOI: 10.1177/1741134312448062

The online version of this article can be found at:

http://jgm.sagepub.com/content/9/2/63

Published by:

http://www.sagepublications.com

Additional services and information for Journal of Generic Medicines: The Business Journal for the Generic Medicines Sector can be
found at:

Email Alerts: http://jgm.sagepub.com/cgi/alerts

Subscriptions: http://jgm.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://jgm.sagepub.com/content/9/2/63.refs.html

>> Version of Record - May 24, 2012

What is This?

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Original Paper
Journal of Generic Medicines
9(2) 63–74

Deficiencies in generic product dossiers as ! The Author(s) 2012

Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
submitted to the WHO Prequalification of DOI: 10.1177/1741134312448062
jgm.sagepub.com
Medicines Programme

Wondiyfraw Z Worku1, John Gordon2, Matthias MS Stahl1 and

Lembit Rägo1

Abstract
This study was undertaken to determine the type and extent of deficiencies in generic product dossiers in the
therapeutic areas of HIV/AIDS, tuberculosis, malaria and reproductive health, as submitted to the WHO
Prequalification of Medicines Programme. There were considerably more quality-related deficiencies in
tuberculosis, malaria and reproductive health dossiers compared to HIV dossiers, especially in the category
specification of active pharmaceutical ingredients, development pharmaceutics, manufacturing method and
finished pharmaceutical product specifications. The deficiencies related to the efficacy/safety portion of the
dossiers displayed a trend similar to that observed in the quality portion in that the most critical deficiencies
such as an incorrect study design, the use of an unacceptable comparator or the failure to include a study
occurred considerably more frequently in the tuberculosis, malaria and reproductive health dossiers than in
the HIV dossiers. The frequency of dossier-related deficiencies as determined on screening and assessment
of the dossiers seemed to be inversely related to the number of product dossiers that had been prequalified by
the end of 2010. The results of this study stress the need for continued capacity building of local generic
manufacturers, further development of pharmacopoeial monographs by WHO (PhInt) and other pharmaco-
poeial commissions, not least to promote development of generic products, as well as development of new
guidelines (WHO guidelines for development of generic and paediatric products and a technology transfer
guidance document are currently being finalized). To our knowledge, this is the first comprehensive review of
the quality and efficacy/safety portions of generic product dossiers, originating from pharmaceutical
companies in emerging markets, and comparison of dossier deficiencies across four critically important
therapeutic areas.

Keywords
Prequalification, dossier deficiencies, generic medicines, quality, bioequivalence, HIV/AIDS, tuberculosis,
malaria, reproductive health

Introduction
The World Health Organization (WHO)
Prequalification of Medicines Programme started in 1
WHO Prequalification of Medicines Programme, Quality
2001 in response to the absence of harmonized drug Assurance and Safety of Medicines, Essential Medicines and
regulatory requirements across developing countries.1 Pharmaceutical Policies, World Health Organization, Geneva,
Switzerland
The task of the programme is to assure that medicinal 2
Division of Biopharmaceutics Evaluation 2, Bureau of
products – mostly generic (multisource) products – Pharmaceutical Sciences, Therapeutic products Directorate,
supplied for procurement by United Nations Health Canada, Ottawa, Canada
Organization, governments, etc. meet WHO norms
and standards with respect to quality, efficacy and Corresponding author:
safety (refer PQP website). Today, the scope of the pro- Matthias Stahl, WHO Prequalification of Medicines Programme,
Quality Assurance and Safety of Medicines, Essential Medicines
gramme has expanded from HIV/AIDS (human and Pharmaceutical Policies, World Health Organization, Avenue
immunodeficiency virus/acquired immune deficiency Appia 20, Geneva 27, CH-1211, Switzerland
syndrome) products to also cover medicinal products Email: [email protected]

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

64 Journal of Generic Medicines 9(2)

to treat tuberculosis (TB), malaria (MA), influenza and of deficiencies in the dossiers submitted in the differ-
some neglected tropical diseases as well as reproductive ent therapeutic areas was made.
health (RH) products and zinc products for diarrhoea To our knowledge, this is the first comprehensive
(refer PQP website). Generic medicines play an review of the quality and efficacy/safety portions of
important role in public health as they are well known generic product dossiers, originating from pharma-
to the medical community and are usually more afford- ceutical companies in emerging markets, and compari-
able due to competition. The quality and efficacy/safety son of dossier deficiencies across four critically
of a generic medicine must conform to the same stand- important therapeutic areas. Similar studies by US
ards as required of the originator’s (comparator) prod- Food and Drug Administration (USFDA) and
ucts. Specifically, the generic product should be European Medicines Agency (EMA) investigated
therapeutically and pharmaceutically equivalent to quality and/or efficacy/safety deficiencies in
and thus, interchangeable with, the comparator prod- Abbreviated New Drug Applications (ANDAs) or
uct.2 Therapeutic equivalence is usually demonstrated new drug applications, but did not indicate the thera-
by testing the bioequivalence (BE) between a product peutic areas studied, nor the origin of the dossiers.3–9
and a suitable comparator product in a well-
standardized pharmacokinetic study with a limited
Materials and methods
number of usually healthy subjects, thus obviating the
need for clinical trials involving many patients to prove This retrospective study encompassed all generic
safety and efficacy.2 product dossiers in the therapeutic areas of HIV, TB,
Inspection of manufacturing plants and laboratory MA and RH received by the WHO Prequalification of
quality control analysis only do not guarantee product Medicines Programme between April 2007 and
quality and safety.1 All processes involved in the manu- December 2010. April 2007 was chosen as start of
facture of the active pharmaceutical ingredients (APIs) the study since a formalized screening procedure of
and the finished dosage forms need to be controlled.1 dossiers submitted to the programme was initiated at
Therefore, to ensure that the manufacturer has that time.
built quality into the product from the beginning,
assessment of the product dossier prior to its accept-
Screening
ance is paramount.1
The assessment of product dossiers also provides Reports of screening of the submitted dossiers were
manufacturers possessing limited experience interact- reviewed for deficiencies identified. Submitted dossiers
ing with stringent national regulatory authorities with were screened with respect to completeness of the
an opportunity to understand the requirements and information on the quality and efficacy/safety of the
processes involved in such a process, since the product, including completed WHO quality and effi-
expectations and requirements for prequalification cacy/safety summary forms (the latter forming the
are similar to the requirements of stringent national basis for the subsequent assessment reports generated
regulatory authorities. As the assessment and accept- by the assessors). Minimum expectations at the time of
ance of a product dossier are necessary for the pre- screening were: availability of process and analytical
qualification of a product, the time required to validation data, at least 6 months accelerated and 6
prequalify a product following application for prequa- months long-term stability data for three batches of
lification will depend heavily on the content and at least pilot scale and BE data or data justifying a
organization of the product dossier. The higher the biowaiver.
quality of a product dossier, the more quickly and
efficiently the product can be assessed and prequali-
Assessment
fied. Dossiers possessing a large number of deficien-
cies will necessitate more interaction between the Reports generated following the first round of assess-
programme and the manufacturer during the assess- ment of the quality and efficacy/safety parts (typically
ment process, thus increasing the time to BE study data in the latter case) of dossiers that had
prequalification. passed screening were reviewed for data deficiencies
This study was undertaken to determine the type identified during the assessment. Specifically, issues
and extent of deficiencies in generic product dossiers that were forwarded to the applicant following an
in the therapeutic areas of HIV/AIDS, TB, MA and assessment, and hence affecting the final acceptance
RH, as submitted to the WHO Prequalification of of the finished pharmaceutical product (FPP) quality
Medicines Programme. Deficiencies in both of the and efficacy/safety parts of the dossier, were recorded.
two main portions of a product dossier, quality and Deficiencies related to the FPP quality part were
efficacy/safety, were examined. Further, a comparison grouped into 10 categories (API specification,

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Worku et al. 65

development pharmaceutics, manufacturing process stability data and comments on preferred conditions
and controls, process validation, excipients, finished for long-term stability studies (in most cases 30 C/
product specification, analytical methods and valid- 75% relative humidity) were not included in the ana-
ation, container closure system, stability data and lysis, as such comments were sent out for nearly all the
product labelling). These categories were chosen reviewed dossiers.
since they represent the FPP quality sections of a The status, as at the end of 2010, of dossiers
standard dossier. The efficacy-related deficiencies accepted for assessment are tabulated by therapeutic
were similarly categorized into four main categories area, in terms of number of dossiers prequalified, still
(study design/administration, clinical portion of under assessment or cancelled/withdrawn.
the study, bioanalytical portion of the study and the
statistical/pharmacokinetic analysis portion of the
study). Deficiencies in the information available Results
regarding the auditing and monitoring of the overall
study were also tracked.
Screening
The results for the efficacy/safety portion are sum- Between April 2007 and December 2010, a total of
marized and presented as the percentage of the total 245 generic product dossiers were received and
number of dossiers surveyed that contained each iden- screened by the WHO Prequalification of Medicines
tified deficiency. The results for the quality part are Programme. Approximately 80% of HIV, 85% of TB,
presented as average number of deficiencies per dos- 35% of MA and 30% of RH dossiers received were
sier since all the reviewed dossiers had deficiencies in submitted by manufacturers in India. Manufacturers
all the 10 quality categories. in China and Indonesia accounted for 25% of MA and
API information can be provided to PQP as a cer- 33% of RH dossiers, respectively. The remaining dos-
tificate of suitability issued by The European siers were submitted by generic product manufacturers
Directorate for the Quality of Medicines, as an active located in Argentina, Belgium, Brazil, Chile, Cyprus,
pharmaceutical ingredient master file (APIMF), Egypt, Eritrea, Ghana, Iran, Jordan, Kenya, Nigeria,
within the framework of the FPP dossier itself or Pakistan, the Philippines, Russia, Senegal, South
recently through the use of a prequalified API.10 Africa, Switzerland, Thailand, Vietnam and
Further, in some cases, the API-related data had Zimbabwe.
already been assessed by PQP and considered accept- The outcome of the screening of the dossiers is
able in relation to a previously prequalified dossier. summarized in Table 1 and deficiencies observed
Depending on the option in each case, the extent of during screening are shown in Figure 1.
the API assessment by PQP will vary. For this reason,
this review covered only deficiencies related to phy-
Assessment
sico-chemical characterization and specification of
APIs as used by the FPP applicant, since these two Of the 245 generic dossiers received and screened, a
areas were common to all dossiers. total of 178 dossiers were accepted for assessment (85
Further, common questions such as requests for HIV, 48 TB, 25 MA and 20 RH dossiers). A total of
justifications of proposed specifications, updated seven dossiers were cancelled/withdrawn before

Table 1. Number of generic product dossiers received, screened and accepted for assessment between April 2007 and
December 2010

2007 2008 2009 2010

HIV TB MA RH HIV TB MA RH HIV TB MA RH HIV TB MA RH

Dossiers received 15 15 7 16 46 17 16 5 31 18 12 3 20 15 6 3
Dossiers accepted during initial screening 3 0 1 6 19 8 1 1 9 10 5 0 9 5 5 2
Dossiers accepted after at least one round 8 12 2 4 24 5 8 3 8 5 3 3 5 3 0 1
of WHO queries
Pending dossiers 0 0 0 0 0 0 0 0 10 0 1 0 6 4 1 0
Rejected dossiers 4 3 4 6 3 4 7 1 4 3 3 0 0 3 0 0
Dossiers accepted for assessment 11 12 3 10 43 13 9 4 17 15 8 3 14 8 5 3
HIV: human immunodeficiency virus; TB: tuberculosis; MA: malaria; and RH: reproductive health.
Number of rejected/cancelled dossiers during the same period has also been included.

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

66 Journal of Generic Medicines 9(2)

% of received dossiers
0 10 20 30 40 50 60 70

Stability data; insufficient number of

months % HIV
% TB
Stability data; insufficient number of % Malaria
batches % RH

No stability data submitted at all

Process validation data on primary

batches missing

Method validation data not submitted

Submission without BE or biowaiver

data

Dossier not properly compiled (missing

pages etc)

WHO summary form not properly filled

or not submitted.

Not in EOI

Figure 1. Deficiencies in generic product dossiers as observed during screening presented as percentage dossiers with a
certain deficiency per therapeutic area.
HIV: human immunodeficiency virus; TB: tuberculosis; RH: reproductive health; BE: bioequivalence; EOI: expression of interest; and
WHO: World Health Organization.

assessment could commence and nine were pending considered acceptable during the first round of assess-
assessment as at the end of December 2010. ment, as were two HIV reports (biowaiver applications
Therefore, in total, 162 dossiers with a quality and/ for oral solutions).
or efficacy/safety assessment report were available for The quality and efficacy/safety parts of the same
analysis (79 HIV, 44 TB, 23 MA and 16 RH dossiers). dossier are generally assessed at different time points,
For the quality part, 147 reports were available for which may explain the absence of one of the two types
analysis (74 HIV, 42 TB, 18 MA and 13 RH). Most of of assessment reports for any given dossier.
the 147 quality assessment reports concerned tablet The deficiencies observed during assessment are
or capsule dosage forms 64 (86%) of HIV, 37 shown in Figures 2 (quality) and 3 (efficacy/safety).
(88%) of TB, 14 (78%) of MA and 7 (54%) of RH In Figure 3, deficiencies encompass study design/
dossiers. The remaining dossiers were for injectable administration (the first category on the x-axis of the
preparations 3 (4%) of HIV, 5 (12%) of TB, 4 graph), the clinical portion of the study (categories
(22%) of MA and 6 (46%) of RH dossiers. Oral solu- 2–5), bioanalysis (categories 6 and 7) and the statis-
tions/suspensions accounted for 7 (9%) of the HIV tical/pharmacokinetic analysis (categories 8 and 9).
dossiers. The final category on the x-axis (auditing/monitoring)
For the efficacy/safety part, 151 were reports avail- falls across categories as all the data should have been
able for analysis (76 HIV, 42 TB, 20 MA, 13 RH dos- audited.
siers). Out of the 151 reports, 3 HIV, 5 TB and 3 RH The status, as of the end of December 2010, of
reports concerned biowaiver applications for injectable the 178 dossiers accepted for assessment is presented
products administered as solution. These were all in Table 2.

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Worku et al. 67

Number of deficiencies per dossier (mean)

6.0

HIV n=74
5.0
TB, n=42
Malaria, n=18
4.0 RH, n=13

3.0

2.0

1.0

0.0

n
n
n

ts

rs
s

ls
ds
s

ls

ilit
tio
io

er
io

tic

be
en

e
ro

ho
at
at

th
in

ab
eu

ca
nt

La
pi
id
iz

ta
et

O
St
ifi
ci
ac

co
er

al

on
m
ec
Ex
sv
rm
t
ac

C
al
sp
an

es
ha

ic
ar

yt
P
oc
ch

s
tp

FP

al
es

Pr
en

An
d

oc
an

pr
op
n

g
tio

el

rin
ca

ev

tu
fi

ac
ci
pe

uf
an
Is

M
AP

Figure 2. Deficiencies observed in generic product dossiers on the assessment of the quality (chemistry–pharmaceutical)
part of the dossier, presented as the mean number of quality deficiencies per dossier and therapeutic area, by each of the 10
main categories.
Deficiencies are related to incomplete or incorrect information provided for the identified category.
HIV: human immunodeficiency virus; TB: tuberculosis; RH: reproductive health; API: active pharmaceutical ingredient; and FPP: finished
pharmaceutical product.

Deficiencies such as improperly compiled dossiers,

Discussion missing dossier pages or incomplete WHO summary
To our knowledge, this is the first comprehensive forms were frequent across all therapeutic areas. Such
review of the quality and efficacy/safety portions of deficiencies, although not a reason for non-acceptance
generic product dossiers, originating from pharma- or rejection of a dossier, are significant in delaying the
ceutical companies in emerging markets, and compari- start of the assessment of the dossier.
son of dossier deficiencies across four critically
important therapeutic areas. Similar studies by
Assessment
USFDA and EMA investigated quality and/or effi-
cacy/safety deficiencies in ANDAs or new drug appli- The review of the assessed generic product dossiers
cations, but did not indicate the therapeutic areas submitted to the WHO Prequalification of Medicines
studied, nor the origin of the dossiers.3–9 Programme during April 2007 to December 2010
revealed considerably more critical quality-related
deficiencies in TB, MA and RH dossiers compared
to HIV ones (Figure 2). The same was true for the
Screening
efficacy/safety part of the dossiers, in that the most
Of the submitted 245 dossiers, 45 (18%) were rejected critical deficiencies such as an incorrect study design,
either at first screening, i.e. if the product was not the use of an unacceptable comparator, or the failure
invited to the programme or later due to the applicant’s to include a study, occurred considerably more fre-
failure to respond to the PQP queries in a timely fashion quently in the TB, MA, and RH dossiers than in the
(maximum 1 year) (Table 1). The proportion of HIV ones (Figure 3).
rejected dossiers was found to be higher for MA
(34%), RH (26%) and TB (20%) than for HIV
Quality
(10%). Major deficiencies, such as inadequate stability
data, incomplete or missing process and analytical val- As shown in Figure 2, deficiencies related to API
idation data or missing BE data, resulting either in and FPP specifications and manufacture of the
delayed acceptance or ultimate rejection of the dossiers, FPP were the most frequent ones. This is similar to
were more frequent in MA and TB dossiers (Figure 1). EMA findings.3 The type of deficiencies noted below

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

68 Journal of Generic Medicines 9(2)

100.0

90.0
HIV, n=76
80.0 TB, n=42

Percentage of Dossiers (%)

Malaria, n=20
70.0
RH, n=13
60.0

50.0

40.0

30.0

20.0

10.0

0.0

rm

ns

n
ria
t
n

n
io

n
g

uc

io
io

io

io
io
n

fo
at

ite

at
si

od
at

at

at
at
m

rm
is

cr
rm

rm

l
rm
ar
pr

cu
or
m

fo
E
m
fo

fo
nf

fo

l
ca
y

tB

in
m
in

to

in
ti

in
ud

su

g
ra

uc

ee
ts
y

is

n
st

rin
ud

ta
pa

io
ys
od

tm
O
or

ito
at

/s
st

al
m

H
Pr

PK
lid

no
gn

on
W

an
co
al

va
ic
si

t/m
do
e
e
lin
de

pl
bl

di
C

ts
ho
am
ta
y

Au
ul
ep
ud

et
ts

es
m
cc
st

ec

R
na

al
ct

bj

ic
U
rre

Su

yt
al
co

An
In

Figure 3. Deficiencies observed on the assessment of the efficacy/safety (i.e. BE/biowaiver) part of the dossier, presented
as percentage of the total number of dossiers surveyed that contained each identified deficiency, per therapeutic area and
by each of the four main categories.
Unless otherwise specified, deficiencies are related to incomplete or incorrect information provided for the identified category.
HIV: human immunodeficiency virus; TB: tuberculosis; RH: reproductive health; BE: bioequivalence; PK/stat: pharmacokinetic/statistical;
and WHO: World Health Organization.

Table 2. Status of dossiers as of the end of December 2010 (only dossiers accepted for assessment; numbers and
percentage of total within each therapeutic area)

Cancelled % Under assessment % Prequalified % Total

HIV 15 18 42 49 28 33 85
TB 4 8 42 88 2 4 48
MA 7 28 16 64 2 8 25
RH 11 55 9 45 0 0 20
HIV: human immunodeficiency virus; TB: tuberculosis; MA: malaria; and RH; reproductive health.

is likewise similar to that reported by USFDA and 67%) or RH dossiers (range 24–69%) (data not
EMA.3–7 shown). This may be due to a combination of avail-
ability of public information such as public assessment
API specification and characterization. There were reports for HIV comparator products, the greater
fewer deficiencies in HIV dossiers as compared to stringent regulatory authority (SRA) experience
other therapeutic areas (Figure 2). Inadequate data amongst HIV manufacturers and the fact that many
or information on physico-chemical characteristics of of the HIV manufacturers produce their own APIs.
the API, missing or inadequate control on key MA APIs are often water insoluble, which may
parameters, such as particle size distribution (PSD) contribute to the frequent deficiencies in MA dossiers.
or inadequate control of impurities, and batch analysis In the case of RH dossiers, the frequent deficiencies
data for the API batches used in the submission/exhi- may be related to content uniformity issues, requiring
bit batches were less frequent in HIV dossiers (range additional controls.
11–24% of HIV dossiers with each of the above defi- For water-insoluble APIs, the rate of dissolution/
ciencies) than TB (range 26–42%), MA (range 33– solubility and the subsequent absorption of the active

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Worku et al. 69

ingredient depends on the physico-chemical properties plan and absence of data to support hold times for
of the API. Consequently, if different batches of API intermediate products (ranges HIV 14–22%, TB
exhibit different polymorph forms or different PSDs, it 19–36%, MA 17–39% and in up to 69% of RH dos-
may result in variable absorptions, possibly influencing siers). It is of note that 50% of the reviewed MA and
the efficacy and/or safety of the product. Needless to RH dossiers were either compiled as a translated dos-
say, if the active ingredient contains more impurities, sier or concerned injectable dosage forms for which
these may be carried over to the final product and the manufacturing process and its related records are
result in exposure of the patient to potentially toxic required to be more detailed than for solid or oral
levels of impurities. API-related impurities should liquid dosage forms. This may contribute to the overall
therefore be controlled or their levels should be limited higher number of deficiencies in MA and RH dossiers
to an acceptable or safe level. (Figure 2). Moreover, the RH manufacturers studied
seemed to have limited SRA experience, in particular
Development pharmaceutics. As noted in Figure 2, of SRA good manufacturing practice (GMP) inspec-
the number of deficiencies in TB and RH dossiers tions. Satisfactory documentation of manufacturing
was considerably higher compared to HIV and MA process records and execution of process validations
dossiers. Common deficiencies such as missing or is directly related to the level of GMP compliance.
inadequate data on API–API/API–excipient compati- Again, deficiencies with respect to the manufactur-
bility, insufficient information on selection and opti- ing process, controls and its validation may lead to
mization of formulation and process parameters, variability in batch-to-batch quality, which could influ-
missing or incomplete comparative dissolution profile ence the product’s in vivo performance. It is essential
data, inadequate data on qualification of proposed that the manufacturer identifies and controls the
container closure system and absence of justification details of the process parameters, equipments, end
and supporting data for overages and score lines were points and in-process tests. Such controls ensure
observed in higher numbers in TB (range 26–50%) production of batches with uniform and consistent
and RH dossiers (range 0–77%) compared to HIV quality.
(range 4–26%) and MA dossiers (range 6–22%).
Presence of such deficiencies indicates that the manu- Excipients. Commonly observed deficiencies such as
facturer is not in control of the quality attributes of the lack of identification tests for colourants/flavours, fail-
product. In such cases, batch-to-batch quality variabil- ure to submit supporting certificate of analysis and
ity is likely, with possibly variable treatment outcome. failure to submit the necessary TSE/BSE (transmis-
Since most of the TB and RH products are not new sible spongiform encephalopathy/bovine spongiform
to the manufacturers, generation of development data encephalopathy) free certification for excipients of
was not done in consideration of the current guide- animal origin showed no obvious difference between
lines. Such incomplete study reports raise questions the therapeutic areas (Figure 2 and data not shown).
which may be addressed by the submission of a As mentioned above, most of the reviewed dossiers
detailed annual product review for several production were for immediate-release tablets, capsules or solu-
size batches. It is noted that all RH dossiers except one tions. Excipients used in such dosage forms are well
were submitted by manufacturers with limited or no established and there are readily available pharmaco-
SRA experience. poeial specifications across therapeutic areas.

Manufacturing process, controls and process FPP specification and analytical methods. This
validation. There were a higher number of deficiencies category represented the majority of deficiencies in
for MA and RH dossiers compared to HIV dossiers, as terms of numbers. Overall, the deficiencies in TB
illustrated in Figure 2. Common deficiencies were and HIV dossiers were fewer compared to MA and
observed across the therapeutic areas, including HIV RH product dossiers (Figure 2).
(range 4–34%), TB (range 12–50%), MA (range Deficiencies such as absence of additional identifi-
11–66%) and RH (range 8–38%) dossiers. These defi- cation test for active ingredients, unjustified/unaccept-
ciencies were related to missing executed and blank able dissolution limits, unjustified limits for moisture
records, inadequate description of equipments, pro- content and insufficient supporting data for microbial
cess parameters and end-point determination, inad- skip testing proposals tended to be more frequent in
equate description of sterile processes, unsatisfactory MA dossiers (39–72%, mean 60%) compared to other
in-process tests and their frequency or acceptability of therapeutic areas but were common also in HIV (range
intermediate product specification. Also observed 16–50%, mean 33%), TB (range 22–71%, mean 39%)
were missing protocols for process validation on and RH (range 8–38%, mean 23%) dossiers.
production batches, inadequate sampling and testing With regard to control of degradation products, there

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

70 Journal of Generic Medicines 9(2)

were less frequent deficiencies in HIV (34%) and TB (54% of the dossiers) but was also seen in HIV
(29%) dossiers compared to MA (72%) and RH (24%), TB (35%) and MA (33%) dossiers.
(46%) dossiers. Containers with immediate contact with the product
Absence of certain tests or unacceptably wide should be confirmed to be non-toxic as there could be
acceptance limits suggest potential batch-to-batch migration of substances from the container to the
variability. For example, wide dissolution limits may product. Thus, absence of control of the identity
allow marketing of batches with significantly different (nature and quality) of the primary container compo-
dissolution results compared to the batch used in nents may allow use of unapproved potentially toxic
clinical or bio-studies. Such differences may translate containers.
to differences in in vivo bioavailability. Similarly, if the
acceptance limits for impurities are not sufficiently Stability data. Deficiencies were of similar frequency
tight, batches with higher impurity contents may be across the HIV, TB and MA dossiers (Figure 2).
marketed, leading to potential toxicity. Failure to provide adequate discussion on observed
Concerning validation of analytical methods, most trends, variability, lack of mass balance and out-
deficiencies related to inadequate demonstration of of-specification results was observed in 50% of the
specificity, quantitation limits, repeatability and accur- MA dossiers (21% of HIV, 22% of TB and 37% of
acy for assay, dissolution and related substance RH dossiers). An additional common deficiency that
methods were more frequent in MA dossiers (range was observed in 75% of RH dossiers was failure to
55–66%) compared to HIV (range 5–18%), TB include critical stability indicating parameters such as
(range 26–35%) and RH (range 8–46%) dossiers. related substances and dissolution. The RH product
Deficiencies in terms of validation will put in manufacturers studied tended to design their stability
question the validity or reliability of the analytical studies based on minimum test parameters as specified
method used to test the quality of the product. in pharmacopoeial monographs which do not neces-
Unreliable test results means that the efficacy/safety sarily include tests for related substances or dissolution
profile of the product cannot be assured. and therefore did not comply with stringent regulatory
Presence of public assessment reports on compara- requirements.
tor products, well-defined API specifications along A pharmaceutical product should have a certain
with well-executed development studies and experi- shelf life (validity period) within which the approved
ence with SRA submission enable the development quality characteristics and ultimately the approved
of FPP specifications, which may explain the rela- safety and efficacy profiles remain acceptable. If a
tively low number of deficiencies for HIV dossiers. batch of the FPP fails to maintain its approved quality
In respect of TB, there is an increasing availability characteristics after release for marketing or distribu-
of monographs for first-line TB products in USP, tion, then it is no longer possible to assure the efficacy/
BP and PhInt. For MA products, there were only a safety profile of that batch. Stability data collected
few pharmacopoeial monographs for artemisinin- from studies on a certain number of batches are used
based FPPs until recently and manufacturers were to extrapolate a shelf life for every future batch. It is
expected to develop and validate their own specifica- essential that such stability results are rigorously ana-
tions and analytical methods. In the area of RH prod- lysed and discussed to make sure that results are con-
ucts, despite the existence of monographs for some of fidently extrapolated to future batches.
the products, tests such as related substances and
dissolution were not included. In such cases, manu- Product labelling. Quality-related information
facturers will be expected to develop in-house test included in the summary of product characteristics/
methods. Most of the RH manufacturers studied, patient information leaflet and outer and immediate
due to limited SRA experience, tend to limit their container labels are assessed as part of the quality
specifications to the minimum pharmacopoeial assessment of the dossier. Common deficiencies were
requirements. The absence of analytical validation failure to include storage statements that reflect the
data for related substances and dissolution may stability results, failure to include appropriate instruc-
explain the fewer questions on analytical method val- tion on dosage administration, failure to indicate the
idation for RH dossiers. full excipient list and failure to include complete infor-
mation on the nature and contents of the product
Container closure systems. The number of deficien- container.
cies in this category was similar across the therapeutic
areas (Figure 2). Exclusion of identity testing from Other/miscellaneous quality deficiencies. Other/mis-
primary container specifications was the major defi- cellaneous deficiencies concerned discrepancies in
ciency observed most frequently in RH dossiers data provided in the WHO summary form for quality

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Worku et al. 71

(i.e. pharmaceutical quality information form) as information submitted does not adequately establish
compared to the submitted dossier, missing prod- the safety and efficacy of that product. For example,
uct samples or submission of samples which did a small percentage of dossiers in each treatment area
not represent the product features proposed in the included studies whose data did not meet the BE
dossier. acceptance criteria. Such results do not prove that
the proposed product and comparator are inequiva-
lent, but they fail to prove the objective, BE (and
Efficacy/safety ultimately therapeutic equivalence).
Study design. As noted in Figure 3, TB and MA In addition to the submission of incorrectly
dossiers were more often deficient in this area than designed studies or the failure to submit a study at
HIV or RH dossiers. Earlier, MA dossiers tended to all, other deficiencies were noted which precluded
include information related to some limited clinical the useful evaluation of the proposed FPP’s perform-
trials or clinical experience with the proposed product. ance. A significant number of TB and RH dossiers
These trials were not sufficient to establish the efficacy/ included studies that employed an unacceptable com-
safety of the proposed product without comparison to parator product. In all these cases, the efficacy/safety
a comparator product and, if the trials were compara- portion of the dossier was unacceptable as an assess-
tive in nature, they were not sufficiently robust and/or ment of the proposed product’s performance could
were not properly designed to detect product perform- not be completed. BE is meaningful only when the
ance differences. Regarding TB dossiers, many of the comparator product was approved based on clinically
TB products are fairly old products and the data established safety and efficacy information. The ten-
provided in support of some of these products were dency for this deficiency to occur most frequently
found to be deficient in relation to the current guide- with TB and RH products may be related to the age
lines. These issues were not observed with the HIV or of these products, i.e. they tend to be older products
RH product dossiers. than those falling into the HIV and MA areas,
Deficiencies with respect to the study design will and hence, studies conducted with local comparator
question the reliability of the final pharmacokinetic products conducted originally for purposes other
results and BE and ultimately the therapeutic equiva- than submission to this programme tend to be more
lence of the proposed formulation to the acceptable common.
comparator product.
Bioanalytical information. This category was divided
Clinical study information. In keeping with the into two subcategories: information related to the ana-
findings in the quality part of the dossier, overall, the lysis of the clinical subject samples and analytical
efficacy/safety portion of HIV dossiers contained fewer method validation information.
deficiencies than dossiers submitted in the TB and As illustrated in Figure 3, dossiers in all therapeutic
MA categories. This is highlighted in Figure 3 where areas contained deficiencies in both these subcate-
it can be seen that 60% of submitted TB and MA gories. With regard to subject sample analyses, defi-
dossiers, contained deficiencies in the information pro- ciencies related to calibration data, quality control
vided in support of the clinical portion of the studies data and the selection of samples for repeat analyses
submitted, while the percentage of HIV (or RH) dos- were all observed in 5–10% of the dossiers for all
siers containing significant deficiencies in the informa- therapeutic areas, with the exception that issues
tion related to the clinical portion of studies was lower related to the selection of samples for repeat samples
(37%). A notable proportion of the deficiencies found were noted in approximately 15% of the RH dossiers
in the clinical study information were related to cor- (data not shown). Incomplete documentation with
rectable issues such as the failure to provide complete respect to subject sample analyses, an issue that
information on the drug products that were employed would delay completion of the dossier assessment,
in the study or failure to provide adequately prepared was observed in between 7% (for HIV dossiers) and
WHO summary forms for BE, i.e. the BE trial infor- 15% (for MA dossiers) of the submitted dossiers (data
mation form. Such deficiencies were typically correct- not shown).
able and hence delayed the acceptance of the efficacy/ With regard to information related to the analytical
safety portion of dossiers but did not prevent it. method validation, seriously deficient or no method
There were other deficiencies related to the design validation was provided for between 7% (for HIV
and conduct of efficacy/safety studies that resulted in dossiers) and 19% (for TB dossiers) of the submitted
dossiers being found unacceptable. An unacceptable dossiers (data not shown). It is interesting to note
dossier does not necessarily indicate that the proposed that none of the RH dossiers were considered to
FPP itself is unacceptable; it may just be that the possess seriously deficient or no method validation

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

72 Journal of Generic Medicines 9(2)

information, although the method validation informa- during the study was minimal, in most cases, sufficient
tion in 15% of the RH dossiers was considered to be auditing/monitoring did occur.
unclear or inconclusive in some respects. The most
significant deficiencies in the submitted method valid-
ation information for all the treatment categories were
Status of dossiers as of the end of 2010
related to stability data. Between 15% (for MA and Table 2 presents the status, as of the end of December
RH dossiers, respectively) and 29% (for HIV dossiers) 2010, of the 178 dossiers accepted for assessment.
of the submitted dossiers were deficient in stability A total of 32 (18%) had been prequalified, 109
data (data not shown). The frequent deficiency was (61%) were under assessment and 37 (21%) had
related to inadequate or missing long-term storage been cancelled/withdrawn. At the end of Dec 2011
stability data under the correct study conditions for (data not shown), the number of prequalified products
plasma samples containing the moiety of interest. It had increased to 60 (HIV: 49, TB: 6, Malaria: 3 and
is likely that this deficiency is more commonly present RH: 2). The number of cancelled/withdrawn dossiers
in the dossiers related to the HIV and MA products, had increased from 37 to 54 (HIV: 22, TB: 9, Malaria:
because, again, they tend to be newer products relative 11 and RH: 12).
to the TB and RH products and the data were still in
the process of being generated at the time of dossier
Limitations of this study
submission.
The bioanalytical deficiencies highlighted above are The number of MA and RH dossiers in this material
similar to those identified as the most common defi- was relatively small. Due to the absence of a forma-
ciencies in ANDA applications to USFDA.8,9 lized screening procedure for dossiers submitted prior
Needless to say, the above deficiencies, unless cor- to April 2007, data from older dossiers could not be
rected or clarified, question the acceptability of the easily merged with data from more recent dossiers
final pharmacokinetic results and ultimately the thera- (post-April 2007). Dossiers submitted to the pro-
peutic equivalence of the proposed product. gramme prior to April 2007 were therefore excluded.
Further, since the quality of dossiers submitted to the
Pharmacokinetic/statistical analyses. Dossiers from programme has improved significantly since 2001 and
all therapeutic areas contained some deficiencies assessment practices and guidelines (WHO and
related to the pharmacokinetic/statistical analysis of others) likewise have developed, a study of a more
the study data, although a higher percentage of these recent population of dossiers would better reflect cur-
deficiencies were noted in TB dossiers than in the rent dossier related issues within the WHO
other therapeutic areas. The majority of the deficien- Prequalification of Medicines Programme. RH is a
cies in this area in the TB dossiers were related to new therapeutic area in PQP and the number of
insufficient information being provided in order to RH dossiers submitted has been low since their first
understand the procedures employed by the investiga- invitation to PQP (late 2006). For all therapeutic
tors and a failure to discuss the observed results in the areas, the dossier-related deficiencies were those iden-
context of other data available from studies related to tified following the first assessment of the dossier. The
the pharmacokinetics/bioavailability of the moiety of applicant’s response to the issues raised was therefore
interest. The RH dossiers were again somewhat not included in this review. Also, with respect to the
unique in that 15% of the dossiers included incorrect API, only the API specifications used by the FPP
pharmacokinetic/statistical calculations that required applicant were included. Therefore, any additional
revision and re-calculation. This deficiency was issues identified as part of the assessment of the cor-
observed in 10% or less of the dossiers in other treat- responding APIMF/DMF (drug master file) which
ment categories (data not shown). may have affected the final acceptability of the API
specification, such as control of additional impurities,
Audit/monitoring information. Approximately 30% of have not been covered. For some of the efficacy/safety
the dossiers submitted in the HIV and TB treatment categories (Figure 3), the differences between the
categories were found to be deficient in the informa- therapeutic areas may reflect the differences between
tion provided in relation to the auditing and monitor- contract research organizations (CROs) rather than
ing activities that took place in relation to the studies manufacturers. However, certain deficiencies are less
undertaken (Figure 3). This factor was also found to dependent on the CRO chosen and may thus better
be the case in 15% of the RH and 5% of the MA reflect differences between therapeutic areas (for
dossiers (Figure 3). Although in some cases it was example, availability of acceptable comparator product
found that the level of monitoring that occurred or incorrect study design).

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

Worku et al. 73

Conflict of interest
Conclusion
The authors have no conflicts of interest that are directly
In this review of generic product dossiers submitted to relevant to the content of this article. The views expressed
the WHO Prequalification of Medicines Programme in this article are the personal views of the authors and may
during April 2007 to December 2010, there were not be used or quoted as being made on behalf of, or reflect-
considerably more quality-related deficiencies in TB, ing the position of, WHO or Health Canada.
MA and RH dossiers compared to HIV dossiers,
especially in the categories specification of APIs, devel-
opment pharmaceutics, manufacturing method and References
FPP specifications. This may be related to the applicant’s 1. Rägo L and Santoso B. Drug regulation: history, present and
experience of interacting with SRAs, availability of infor- future. In: van Boxtel CJ, Santoso B and Edwards IR (eds) Drug
benefits and risks: International textbook of clinical pharmacology,
mation on comparator products, availability of pharma- revised 2nd edn. Amsterdam: IOS Press; Uppsala: UMC, 2008,
copoeial monographs and the age of the product. pp.65–77.
The deficiencies related to the efficacy/safety portion 2. Gordon J, Potthast H, Stahl M, et al. Bioequivalence require-
of the dossiers displayed a trend similar to that observed ments. The World Health Organization. In: Kanfer I and
in the quality portion in that the most critical deficien- Shargel L (eds) Generic drug product development. International
regulatory requirements for bioequivalence. Drugs and the
cies, such as an incorrect study design, the use of an
Pharmaceutical Sciences. Vol. 201, New York: Informa
unacceptable comparator or the failure to include a Healthcare, 2010, pp.282–300.
study, occurred considerably more frequently in the 3. John JB, Jean-Louis R, George W, et al. Where is industry get-
TB, MA, and RH dossiers than in the HIV dossiers. ting it wrong? A review of quality concerns raised at day 120 by
Similarly, less critical but still important deficiencies the Committee for Medicinal Products for Human Use during
European Centralised Marketing Authorisation Submissions for
in the clinical study information also occurred less fre-
chemical entity medicinal products. J Pharm Pharm Sci 2009;
quently in the HIV dossiers than in the dossiers for the 12(2): 181–198.
other therapeutic areas. Again, these findings may be 4. Paul S. Drug Master file review issues at The Office of Generic
attributable to the fact that the HIV products tend to be Drugs. J Generic Med 2006; 3(4): 280–286.
newer products than some of the other products, the 5. Aloka S and Robert I. Common deficiencies in Abbreviated
New Drug Applications: Part 1: Drug substance. Pharm
TB and RH products, in particular, and hence, there
Technol 2010; 34(1): 50–59.
are fewer old studies in existence for these products that 6. Aloka S, Robert I and Devinder SG. Common deficiencies
may not have been designed to satisfy the requirements in Abbreviated New Drug Applications: Part 2: Description,
of the more SRAs and/or employed local comparators composition, and excipients. Pharm Technol 2010; 34(8):
that do not meet programme requirements. Dossiers 45–51.
7. Aloka S, Robert I and Devinder SG. Common deficiencies
for all therapeutic areas showed a similar frequency
in Abbreviated New Drug Applications: Part 3: Control of
for deficiencies related to the bioanalytical portion of the drug product and stability. Pharm Technol 2011; 35(2):
the submitted studies. The most common deficiency in 58–67.
this area was the failure to provide complete study-rele- 8. Suman D, Barbara MD, Svetlana AC, et al. Common deficien-
vant stability data. cies with bioequivalence (BE) submissions in Abbreviated New
Drug Applications (ANDA). In: AAPS annual meeting and
The frequency of dossier-related deficiencies as
exposition, Los Angeles, CA, 8–12 November 2009.
determined on screening and assessment of the dos- 9. Williamson LN, Conner DP, Stier EM, et al. Common bioana-
siers seemed to be inversely related to the number of lytical deficiencies with bioequivalence (BE) submissions in
the product dossiers that had been prequalified by the Abbreviated New Drug Applications (ANDAs), http://
end of 2010 (Table 2). www.aapsj.org/abstracts/AM_2010/W4424.pdf.
10. World Health Organization. Guideline on submission of docu-
The results of this study stress the need for contin-
mentation for multisource (generic) finished pharmaceutical
ued capacity building of local generic manufacturers, product: quality part (draft guideline for comment), http://
further development of pharmacopoeial monographs apps.who.int/prequal/info_general/documents/generic_guide/
by WHO (PhInt) and other pharmacopoeial commis- GenericGuideline_Quality.pdf.
sions, not least to promote the development of generic
products, as well as development of new guidelines
(WHO guidelines for development of generic and Author’s Biographies
paediatric products and a technology transfer guidance Wondiyfraw Z Worku is a pharmaceutical assessor in
document are currently being finalized). the WHO Prequalification of Medicines Programme.

Funding John Gordon is a PhD and bioequivalence assessor at

No sources of funding were used to assist in the preparation Health Canada and a consultant to the WHO
of this article. Prequalification of Medicines Programme.

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014

74 Journal of Generic Medicines 9(2)

Matthias MS Stahl is an MD and clinical pharmacol- Lembit Rägo is an MD and PhD with clinical
ogist with a background in the pharmaceutical indus- pharmacology background. He is coordinator of
try in early drug development and drug safety. He is the WHO quality and safety of medicines team,
Head of assessments in the WHO Prequalification of which includes the Prequalification of Medicines
Medicines Programme. Programme.

Downloaded from jgm.sagepub.com at UNIV MASSACHUSETTS BOSTON on November 10, 2014