HPE Drug stability: What do we need to know? | Page 9

taken from a 10ml vial, leaving 40mg unused. For
a public price of 14,300€ for a 50mg/vial (286€/mg)
in France, this would amount to a loss of 11,400€.
In the centralised unit, the manipulations are
made under strict and controlled aseptic conditions
(using laminar flow hoods in class 5 environment
or isolators) by well-trained manipulators, so the
infectious risk is insignificant, and retaining the
residual content for another preparation is possible.
Furthermore, new modes of treatment such as long-
term continuous infusion by ambulatory pumps
require the administration of concentrated solutions
over a long period of time. Finally, the development
of standardised doses (dose-banding) necessitates
advance preparation of drug batches.
Other unpredictable situations, such as breaking
the cold-chain (for example, failure of a refrigerator),
long-term storage of a bag outside the fridge in
the ward, or long-distance transportation in poor
conditions, can arise. 1–3
It is therefore critical for the pharmacist who
has final responsibility for the quality of the drug
to have well-documented data about the stability
of the opened drug, of the formulation following
reconstitution of a lyophilised product, or after
dilution in various vehicles. Unfortunately, these
data are not always available, or the published
results are old, contradictory, or limited to a
particular vehicle or a specific bag. Moreover, the
manufacturer Summary of Product Characteristics
(SPC) frequently quotes post-dilution stability
data such as ‘stable for 24 or 48 hours’, which
is unhelpful because it only considers possible
bacterial contamination or the fact that stability
tests have been conducted over very short periods
during the development of the drug. Indeed, the
international requirements for stability testing
of manufactured drugs only cover test batches
of the final product and possibly of the freeze-
dried products after their reconstitution. 4,5 This is
obviously insufficient for most practical situations
as previously explained.
Moreover, clinical protocols are frequently
contradictory to the stability data provided by the
drug manufacturer. As an example, a Phase II study
of prolonged ambulatory infusion carboplatin and
oral etoposide for patients progressing through
hormonal therapy for prostate cancer has been
published. 6 The protocol required continuous
administration of carboplatin for 21 days, the pump
reservoir being refilled every week. However, when
the paper was published, no data were available on
1-week stability of carboplatin at body temperature.
Interestingly, the SPC indicates that carboplatin
stability in 5% dextrose has been demonstrated
for 12 hours but should be used immediately ‘for
microbiological reasons’ and that its non-immediate
use is ‘the responsibility of the user’, and that ‘the
storage should not exceed 24 hours’. Obviously, if
the stability data provided by the manufacturer were
followed, this protocol is unfeasible.
Thus, the data provided by a manufacturer do
not necessarily reflect the real drug stability. The
main reason appears to be exercising a ‘cautious
principle’, based on the idea that the preparation
is made under poor aseptic conditions, but without
consideration of the real degradation rate of the
product. Because it is difficult to predict all the
possible conditions under which the product will be
opened, diluted, reconstituted and stored, etc, the
user is responsible for maintaining the quality of the
product that is administered to the patient. From
the manufacturer’s side, only chemical and physical
in-use stability can be claimed for the respective
period. From a microbiological point of view, the
product should be used immediately and, if not used
immediately, in-use storage time and conditions
prior to use are the responsibility of the user.
The best demonstration is given by the two
formulations of the originator trastuzumab
(Herceptin ® , Roche). For a long time, before the
arrival of biosimilars, only a vial containing 150mg
lyophilised trastuzumab was available in Europe.
The product was reconstituted by 7.1ml of sterile
water prior to final dilution. The SPC indicated
that the stability of this reconstituted solution
had been demonstrated for 48 hours but, from
a bacteriological point of view, ‘it should be used
immediately’. However, for the 420mg formulation,
containing exactly the same formulation but also
an ampoule of bacteriostatic water and available
only outside of Europe, the same manufacturer
indicated that the reconstituted solution is stable
for 28 days if using the bacteriostatic water, but
only 48 hours is non- bacteriostatic water was
used. Clearly, this stability does not reflect the real
physicochemical behaviour of the molecule. There
are now several biosimilars of trastuzumab on
the market and, interestingly, the manufacturers
propose a 28-day stability based on a company-based
study, and the originator company has recently
modified this in the SPC. Surprisingly, the product
which was previously supposedly stable only for a
few days now has a stability of 15 days. 7 Moreover,
the stability data proposed in the new SPC are very
confusing. The statement that a vial of Herceptin ®
reconstituted aseptically with sterile water for
injection is physicochemically stable for 48 hours
at 2°C to 8°C after reconstitution is contradictory
to the following statement that, after aseptic
dilution into polyvinylchloride, polyethylene or
polypropylene bags containing sodium chloride
9mg/ml, stability had been demonstrated for 30 days
at 2–8°C. 7 However, this extended stability has been
previously demonstrated
by independent studies
for the originator 8 and
biosimilars. 9
Unreal stability data are therefore
The same can be
seen with the originator
unacceptable because large financial
rituximab Mabthera ® ,
resources, mainly paid through social
commercialised at the
insurance systems, can be wasted in
beginning of the 2000s
with a stability limited
cases of very costly products
to 24 hours. However,
at the beginning of
2018, when biosimilars
of rituximab were introduced in the European
market claiming one-month stability after dilution,
Mabthera ® stability was increased to seven days in its
SPC. 10 However, and similar to trastuzumab, several
studies had already demonstrated that rituximab
(originators and biosimilars) was stable for months
after dilution. 11–13
The same problem of discrepancy between SPC
indications and real stability can be observed for
the classical anticancer drugs. As an example, since
the first commercialisation of the originator of
paclitaxel (Taxol ® ), stability data remain unchanged
with 27-hour stability: three hours for the
preparation and transportation and 24 hours for
continuous infusion after dilution.
Clearly, pharmaceutical companies perform
stability studies (final product and in-use) only to
obtain regulatory approval for their products and
for their own interest. Frequently the results of
these studies are largely insufficient, do not reflect
the real stability of a drug, and do not meet the
hospitalpharmacyeurope.com | 2019 | 9