International Journal of Pharmaceutics 475 (2014) 110122

Contents lists available at ScienceDirect

International Journal of Pharmaceutics

journal homepage: www.elsevier.com/locate/ijpharm

Development of performance matrix for generic product equivalence of

acyclovir topical creams
Yellela S.R. Krishnaiah a , Xiaoming Xu a , Ziyaur Rahman a , Yang Yang a ,
Usha Katragadda a , Robert Lionberger b , John R. Peters b , Kathleen Uhl b ,
Mansoor A. Khan a, *
a
Division of Product Quality Research, Ofce of Testing and Research, Ofce of Pharmaceutical Sciences, Center for Drug Evaluation and Research, Food and
Drug Administration, Silver Spring, MD, USA
b
Ofce of Generic Drugs, Center for Drug Evaluation and Research, Food and Drug Administration, Silver Spring, MD, USA

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 14 May 2014
Received in revised form 10 July 2014
Accepted 24 July 2014
Available online 30 July 2014

The effect of process variability on physicochemical characteristics and in vitro performance of

qualitatively (Q1) and quantitatively (Q2) equivalent generic acyclovir topical dermatological creams was
investigated to develop a matrix of standards for determining their in vitro bioequivalence with reference
listed drug (RLD) product (Zovirax1). A fractional factorial design of experiment (DOE) with triplicate
center point was used to create 11 acyclovir cream formulations with manufacturing variables such as pH
of aqueous phase, emulsication time, homogenization speed, and emulsication temperature. Three
more formulations (F-12F-14) with drug particle size representing RLD were also prepared where the
pH of the nal product was adjusted. The formulations were subjected to physicochemical
characterization (drug particle size, spreadability, viscosity, pH, and drug concentration in aqueous
phase) and in vitro drug release studies against RLD. The results demonstrated that DOE formulations
were structurally and functionally (e.g., drug release) similar (Q3) to RLD. Moreover, in vitro drug
permeation studies showed that extent of drug bioavailability/retention in human epidermis from F-12
F-14 were similar to RLD, although differed in rate of permeation. The results suggested generic acyclovir
creams can be manufactured to obtain identical performance as that of RLD with Q1/Q2/Q3.
Published by Elsevier B.V.

Keywords:
Acyclovir
Topical cream
Process variables
Quality metrics
Sameness
Q1/Q2/Q3

1. Introduction
Zovirax1 cream was approved by US/FDA in 2002 for the
treatment of recurrent herpes labialis (cold sores) in adults and
adolescents. It is a topical dermatological product containing 5%
w/w of acyclovir in aqueous cream base formulated with
cetostearyl alcohol, mineral oil, poloxamer 407, propylene glycol,
sodium lauryl sulfate, water, and white petrolatum as inactive
ingredients (Zovirax, 2002). Acyclovir is a synthetic purine
nucleoside analog with in vitro and in vivo inhibitory activity
against herpes simplex virus types 1 (HSV-1), 2 (HSV-2), and
varicella-zoster virus (VZV) (Acosta and Flexner, 2011). There are
no generic acyclovir topical dermatological cream products
available at this time in the market. The possible generic products

* Corresponding author at: FDA/CDER/OPS/OTR/DPQR, White Oak, LS Building 64,

Room 1070, 10903, New Hampshire Ave, Silver Spring, MD 20993-002, USA.
Tel.: +1 301 796 0016.
E-mail address: mansoor.khan@fda.hhs.gov (M.A. Khan).
http://dx.doi.org/10.1016/j.ijpharm.2014.07.034
0378-5173/ Published by Elsevier B.V.

of acyclovir topical cream have to conform to the same standards of

quality as that of Zovirax1 cream (reference listed drug product,
RLD) and demonstrate clear bioequivalence (BE) by in vivo or in
vitro methodologies. The availability of product quality metrics is
critical to demonstrate that generic pharmaceutical drug products
are therapeutically equivalent and interchangeable with their
associated innovator's product.
A list of in vivo and in vitro methods have been provided to
establish the BE under regulation 21CFR320.24(b) 44 (21CFR320.1,
2014; 21CFR320.24, 2014). In vivo studies in humans comparing
drug/metabolite concentrations in an accessible biological uid, in
vivo testing in humans of an acute pharmacological effect, and
controlled clinical BE trials in humans to establish ability to
achieve an equivalent clinical endpoint with no evidence of
differing safety prole are to be chosen as the rst, second, and
third approaches. The in vitro methods are to be chosen as the next
available choices. The sponsors may choose any other rational
approach and provide data to convince FDA on the use of such
approach in demonstrating bioequivalence. One or more of these
approaches might be used to demonstrate BE. For example, the

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

bioequivalence of solid oral dosage forms intended for systemic

delivery is established by in vivo pharmacokinetic (PK) studies
with a support of comparative in vitro drug release data. This
approach has been successfully applied to a large number of drug
products (Kryscio et al., 2008). However, the conventional in vivo
BE study with PK endpoints such as Cmax and AUC is neither
appropriate nor feasible for establishing BE of topically applied
dermatological products. Determination of topical bioequivalence
for locally acting drugs in skin is more complicated as local drug
concentrations cannot be measured directly. The guidance on
bioavailability and bioequivalence drafted by Committee of
Proprietary Medical Products (CPMP) of the European regulatory
authorities stated for medicinal products not intended to be
delivered into the general circulation, the common systemic
bioavailability approach cannot be applied (EMA, 2000). The US
FDA provided certain recommendations with respect to the
establishment of BE for such specic products (FDA, 2010). Draft
guidance documents on locally acting topical drug products such
as cyclosporine ophthalmic emulsion and acyclovir ointment have
been developed by FDA to provide recommendations to sponsors
to meet statutory and regulatory requirements (FDA, 2012, 2013).
Generally, FDA addresses the issue on a case by case basis as
outlined by the drug-specic guidance. Therefore, it is necessary to
identify the key scientic principles for consistent and efcient
identication of bioequivalence methods for locally acting topical
dermatological products.
The current regulation requires conducting clinical endpoint
trials for demonstrating BE between topical generic and RLD
products when alternative methods, such as pharmacodynamic
endpoint measures are not feasible (21CFR320.1, 2014;
21CFR320.24, 2014). Topical glucocorticoids (Chang et al., 2013b)
are an example of products where a clear pharmacodynamic
endpoint (skin blanching) is possible. Clinical endpoint bioequivalence studies with topical drug products are lengthy and expensive
(Shah et al., 1998). These studies are subjected to greater variability
than other in vivo methods for determining bioequivalence. Thus,
the large inter-subject variability and dichotomous nature of these
clinical endpoint bioequivalence studies demand the enrollment of
several hundred subjects to achieve sufcient statistical power
(Bhandari et al., 2002; Donner and Eliasziw, 1994). In order to
determine BE of acyclovir topical cream products for treating
herpes simplex labialis, the primary endpoint is the time to
complete healing of lesions. This is particularly challenging for
three reasons: (1) the severity of lesions is confounding; (2) lesions
last a short period of time and heal rapidly regardless of treatment;
and (3) the effectiveness of therapy is related to the rapidity with
which treatment is initiated. In two clinical studies conducted for
Zovirax1 cream, no signicant difference was observed between
subjects receiving Zovirax1 cream or vehicle (Zovirax, 2002). The
mean duration of the recurrent herpes labialis episode was
approximately half a day shorter in the subjects (n = 682) treated
with Zovirax1 cream (4.5 days) compared with subjects (n = 702)
treated with placebo (5 days). The considerable variability in
clinical endpoints is common and renders the BE clinical design
difcult to detect the small difference in therapeutic response
between generic and RLD (Chang et al., 2013b; Yacobi et al., 2014).
A variety of surrogate methods such as skin stripping/
dermatopharmacokinetics (DPK), dermal microdialysis (DMD),
in vitro permeation studies and near infrared (NIR) spectroscopy
have been explored to demonstrate the BE of topical dermatological products (Lionberger, 2008; Narkar, 2010; Yacobi et al., 2014).
Yet these surrogate methods are even more prone to failures in
detecting low drug concentration in skin due to their limited
sensitivity, technical difculty, and high variability. The scope and
limitations associated with these techniques have been reviewed
(Herkenne et al., 2008; Mateus et al., 2013; Narkar, 2010; Yacobi

111

et al., 2014). For example, skin stripping has been used for testing
BE of topical dermatological products acting in stratum corneum
(N'Dri-Stempfer et al., 2008, 2009; Navidi et al., 2008; Parry et al.,
1992). But this is unsuitable for studying the BE of topical
dermatological products whose site of action is a compromised
skin (e.g., cold sores due to herpes labialis).
The in vitro drug permeation across human skin and in vitro
drug release testing may be suitable to test the sameness (Q3) of
Q1/Q2 equivalent topical dermatological products with respect to
their performance. Such in vitro tests have been recommended to
test the product sameness under certain scale-up and postapproval changes (SUPAC) as it is believed to collectively reect any
differences due to several physicochemical properties such as
solubility, particle size of drug, and rheological properties of
vehicle (FDA, 1997). The present study was carried out to
understand and identify the appropriate in vitro quality metrics
that can discriminate the effect of process and formulation
variables on critical quality attributes (CQA) of possible generic
acyclovir topical cream formulations having the same Q1/Q2 to
that of Zovirax1.
Quality by design (QbD) approach was used to study the effect
of process and formulation variables on CQA of acyclovir topical
cream formulations. The preparation of acyclovir cream typically
involves homogenization of oil-soluble and water-soluble components along with the drug to form oil-in-water cream at 70
C.
Based on the preliminary process understanding, three process
parameters (emulsication time, homogenization speed, and
temperature of oil/water phases) were identied as critical
process parameters (CPP). Moreover, the HSV-1 infection and
replication occurs in the basal cell layer of the epidermis (Parry
et al., 1992). Therefore effectiveness of acyclovir topical dermatological creams depends on drug permeation across skin and
drug retention in epidermis (DRE), which in general is a function
of the aqueous phase drug concentration (thermodynamic
activity). The equilibrium water solubility of acyclovir was
reported to be inuenced by pH with the highest solubility
being at pH <3 and at pH >9 (Shojaei et al., 1998). Thus pH of
acyclovir cream products was chosen as a formulation variable in
addition to the above three CPPs for studying their inuence on
CQA. A fractional factorial design (241) with triplicate center
point was chosen to study the effects of process and formulation
variability on product CQA. This is a Resolution IV design where
estimation of the main effects is not confounded by two-factor
interactions (Chang et al., 2013a). Accordingly 11 formulations
were prepared and subjected to physicochemical characterization
and in vitro performance testing to test the sameness (Q3) of
Q1/Q2 equivalent acyclovir creams.
2. Materials and methods
2.1. Materials
Zovirax1 cream was obtained from Bradley Drugs, Bethesda,
MD, USA. Acyclovir (>99%) was purchased from RIA International
LLC, East Hanover, NJ, USA. Propylene glycol USP, white petrolatum
USP, mineral oil USP, glacial acetic acid USP and sodium lauryl
sulfate (SLS) NF were purchased from Fischer Scientic, Norcross,
GA, USA. Poloxamer 407 NF and cetostearyl alcohol NF were
purchased from Spectrum Chemical Manufacturing Co., New
Brunswick, NJ, USA.
2.2. Preparation of acyclovir cream formulations
Four process/formulation variables (pH of aqueous phase,
emulsication time, homogenization speed and emulsication
temperature) were studied using a fractional factorial design with

112

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

Table 1
Fractional factorial design (241) to assess the process variables of acyclovir cream formulations.
ID

Emulsication time
(min)

Homogenization speed (rpm)

Temperature
(
C)

pH of aqueous phase

DOE-1
DOE-2
DOE-3
DOE-4
DOE-5
DOE-6
DOE-7
DOE-8
DOE-9
DOE-10
DOE-11

15
15
30
30
22.5
22.5
22.5
30
15
15
30

2500
5000
2500
5000
3750
3750
3750
2500
5000
2500
5000

90
90
90
90
80
80
80
70
70
70
70

8.5
5
5
8.5
6.75
6.75
6.75
8.5
8.5
5
5

triplicate center points. Based on this Design of Experiment (DOE),

11 acyclovir formulations (DOE 1 to 11) were prepared (Table 1).
Briey, aqueous phase and oil phase were mixed and homogenized. Aqueous phase was prepared by dissolving poloxamer 407
in water by mechanical stirrer (RW-20 digital, IKA, Wilmington
USA) at 500 rpm for 30 min, adding propylene glycol and adjusting
its pH (5.5, 6.75, or 8.5). Acyclovir was dispersed in the aqueous
phase by mechanical stirring at 500 rpm at 70, 80, or 90
C for
10 min. Similarly, oil phase was prepared by melting white
petrolatum, cetostearyl alcohol, and mineral oil at 70, 80, or
90
C. Sodium lauryl sulphate was dispersed in the oil phase by
Silverson homogenizer (L5M-A, Silverson, Baltimore, MD) at
2000 rpm for 2 min. Aqueous phase was added to oil phase and
homogenized by Silverson Homogenizer for 15, 22.5, or 30 min at
2000, 3750, or 5000 rpm. The cream was allowed to cool at room
temperature while being homogenized at 1000 rpm for 2 h.
Placebo cream was also prepared in the same way but without
drug and pH adjustment. Drug was physically added to the placebo
cream to make its composition equal to DOE formulations. Three
more formulations (F-12F-14) were prepared with a slight
modication in the preparation of aqueous and oil phases using
acyclovir drug particles representing Zovirax1 cream. The aqueous
phase pH was adjusted after dissolving SLS and dispersing
acyclovir, and the oil phase was without SLS. The aqueous and
oil phases were homogenized at 5000 rpm and 70
C for 15 min
(process conditions representing as those of DOE-9). The pH of
acyclovir cream formulations was measured using a pH meter. All
the acyclovir cream formulations were packed (5 g in quantity) in
multiple-dose aluminum tubes, and stored in a chamber at
25
C/60% RH until used for further studies.
2.3. Drug content uniformity
Topically applied semisolid drug products such as acyclovir
cream may show physical separation during manufacturing
process and during their shelf life. To ensure their integrity, it is
essential to evaluate the uniformity of the nished product with
respect to visual uniformity and uniformity of active ingredients.
This was carried out as per the procedure described in USP (USP36NF31, 2013a). The bottom tube seal was cut off and a vertical cut
was made from the bottom to the top of the tube. The tube around
the upper rim was cut to open the two aps and aps laid open to
expose the product. The product was inspected for the presence of
phase separation, and change in physical appearance and texture
(e.g., color change). An appropriate amount of accurately weighed
product (100 mg) was removed from the top, middle, and bottom
portions of the tube, and transferred to a ask containing 400 ml
solvent (pH 9.2 borate buffer). The contents were homogenized at
7600 rpm for 15 min (25
C), ltered, diluted suitably, and injected
into HPLC column for determining acyclovir concentration. This

was used to calculate the amount of acyclovir in the samples

obtained from the tube.
2.4. Drug concentration in aqueous phase
The acyclovir cream formulations including Zovirax1
were lled into Eppendorf tubes (capacity 5 ml) without air gaps.
These tubes were centrifuged at 14,000 rpm for 5 h (sample
temperature set to 25
C). The top layer of oil phase separated from
the cream was scooped out carefully, and remaining aqueous phase
centrifuged at 14,000 rpm for 1.5 h to clearly separate the traces of
oil phase. The resultant aqueous phase was suitably diluted,
ltered through 0.45-mm syringe lter and injected into HPLC for
determining acyclovir concentration.
2.5. X-ray powder diffraction (XRPD)
Acyclovir physical forms in the cream were conrmed by XRPD.
Diffractogram were collected using a Bruker D8 Advance with
DaVinci design (Bruker AXS, Madison, Wisconsin) equipped with
the LYNXEYE scintillation detector and Cu Ka radiation
(l = 1.5405 ) at a voltage 40 kV and current 40 mA. About
500 mg sample were placed in the sample holder and diffractogram was collected over 2u range of 440
with an increment of
0.0114 at 1 s per step (3000 total steps). Sample holder was rotated
during run to get the average diffractogram of the sample. The
XRPD operation, data collection, and data analysis were achieved
through Diffrac.SuiteTM (V2.2).
2.6. Particle size analysis
Cream samples were applied onto a glass slide and spread evenly
using a cover slip. Images were acquired using an Olympus BX51
polarized light microscopy (Olympus America Inc. Melville, New
York). On average about 10 microscopy images (about
200500 particles) were acquired for each sample (200 magnication). In each image, particles were manually counted and
measured using Olympus cellSens software (Olympus America Inc
Melville, New York) to obtain particle size and elongation information which were then imported into Excel to generate statistical
information (i.e., D10, D50 and D90) with a percentile function. Particle
size of acyclovir API powder was determined using a HALEOS laser
diffraction instrument (Sympatec, Clausthal-Zellerfeld, Germany),
equipped with a R5 lens (0.5875 mm). For each measurement,
approximately 100 mg of dry sample was dispersed at a feed rate of
50% by a controlled feeder (VIBRI/L), with dry dispersal (RODOS)
attachments set at a main pressure of 1.0 bar. The triggering
condition was set at 0.2% optical concentration. Data were analyzed
using Fraunhofer theory with WINDOX 5 software (Sympatec,
Clausthal-Zellerfeld, Germany).

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

113

2.7. Rheological characterization

2.9. Preparation of heat separated human epidermis

Rheological behaviors of creams were evaluated using a

stress-controlled hybrid rheometer (DHR-3, TA Instruments, New
Castle, Delaware, USA) equipped with a step-peltier stage (25
C)
and a 25 mm sandblasted parallel plate. For each test, approximately 0.3 ml of cream sample was placed on the lower plate,
before slowly lowering the upper plate to the preset trimming gap
of 550 mm. After trimming excess material, the geometry gap was
set at 500 mm. The following procedures were performed in
sequence on each sample to characterize its rheological behavior:
(1) 10 min equilibration (time sweep mode) to allow the material
to fully recover from the shear applied during sample preparation
(monitored at 0.1% strain and 1 Hz oscillation); (2) assessment of
the yield stress with strain sweep method (0.0550% strain at
1 Hz), determined by plotting storage modulus vs. oscillation
stress and nding the onset point; (3) 10 min equilibration to
allow the material to recover from the shear applied during
previous step (monitored at 0.1% strain and 1 Hz oscillation); and
(4) steady-state-ow method to characterize the ow property
(104100 s1) and to obtain viscosity values at low (0.01 s1),
medium (1 s1), and high (56 s1) shear rate. For the temperature
sweep study, sample was rst equilibrated at 16
C for 20 min
(monitored at 0.1% strain at 1 rad/s), and followed by step increase
of the temperature from 16 to 46
C at 1
C interval. At each
temperature, there was a 30 s soak time (equilibration time)
before oscillation test was performed (0.1% strain at an angular
velocity of 1 rad/s). All rheological studies were performed in
triplicate and results were expressed as mean  SD.

Full thickness human cadaver skin was obtained from The

National Disease Research Interchange (NDRI, Philadelphia, PA).
The protocol for the use of human cadaver skin samples was
approved by Research Involving Human Subjects Committee
([RIHSC], RIHSC #11077-D) of US Food and Drug Administration,
and is in accordance with tenets of the Declaration of Helsinki
promulgated in 1964. The samples were either used immediately
or stored at 20
C and used within 2 weeks of storage. Prior to
use, skin samples were thawed for 2 h at room temperature. The
epidermis was prepared using a heat separation technique
(Krishnaiah et al., 2006; Zhao and Singh, 1999). The skin samples
were soaked in water at 60
C for 60 s, followed by careful removal
of epidermis. The heat separated epidermis was washed with
water and examined for physical damage by using a magnifying
lens. The barrier integrity of the excised human epidermis was
also tested by measuring transepidermal water loss (TEWL) using
a vapometer (Deln Technologies, Kuopio, Finland). The epidermis free from physical damage and having a TEWL of about
6 g/cm2  h was used for in vitro drug permeation studies
(Andersen et al., 2006).

2.8. In vitro drug release studies using immersion cell apparatus

Immersion cell model A (exposed area of 2 cm2; Agilent
Technologies, USA) was used with USP apparatus 2 to determine
the in vitro release of acyclovir from DOE formulations as per the
procedure described in USP (USP36-NF31, 2013b). A small
dissolution vessel (200 ml) was used. Appropriate modications
including holders for small vessels and replacement of standard
paddles with appropriate paddles were made to the USP apparatus
2. The bath temperature was set at 32  0.5
C. The components of
enhancer cell model A were prepared and assembled as per the
manufacturer's directions. The polysulfone membrane (Tuffryn1,
PALL Life Sciences HT-450, 0.45 mm average pore size) was soaked
in dissolution medium (alkaline borate buffer, pH 9.2) 30 min
before loading of enhancer cells into the dissolution system. The
reservoir of the cell was lled completely and uniformly with
cream formulation using a spatula. The weight of cream lled into
these enhancer cells was determined by subtracting the weight of
empty cell assembly from the weight of cells completely lled with
cream. The mean weight of cream used in the drug release testing
was 2 g. The soaked polysulfone membrane was placed over the
surface of sample compartment without the formation of wrinkles
in the membrane. The immersion cell components were assembled
as specied by manufacturer. The assembled enhancer cell was
placed at the bottom of the dissolution vessel with the membrane
facing up. Before loading the enhancer cell, the paddle height was
set at 1.0  0.2 cm above the surface of the membrane. The
pre-heated (32  0.5
C) dissolution medium was added to the
dissolution vessel containing the loaded immersion cell to start
the test with automatic sampling at pre-determined time points
up to 6 h. The amount of acyclovir released in the samples was
determined by validated HPLC method (injection volume of 5 ml)
as described below. The cumulative amount of drug dissolved per
cm2 of exposed membrane area was plotted against square root of
time (Higuchi plot) to determine the in vitro release rate (slope of
Higuchi plot) of acyclovir from the cream (USP36-NF31, 2013b).

2.10. Drug retained in epidermis (DRE) after in vitro drug permeation

study
An in vitro drug permeation study was carried out with selected
acyclovir cream formulations (F-12F-14 and Zovirax1) using an
automated vertical Franz-type glass diffusion cell system (Microette-Plus, Hanson Research Corporation, CA). The receptor
chamber of the diffusion cell (12 ml; diffusion area of 1.767 cm2)
was lled with a receptor medium (alkaline borate buffer, pH 9.2)
and stirred at 400 rpm. The heat separated epidermis, with the
stratum corneum facing up, was sandwiched between the dosage
wafer and receptor chamber. The dosage wafer of the diffusion cell
was lled with accurately weighed quantity of acyclovir cream
(0.4 g), applied uniformly with a spatula and covered with a glass
top. The temperature of the diffusion cells was maintained at
32.0  0.5
C by means of circulating water bath. During the
experiments, the receptor medium was continuously stirred with a
magnetic stirrer at 400 rpm. Permeate samples (1 ml) were
collected at pre-determined time intervals up to 24 h and injected
(50 ml) into HPLC column for determining acyclovir concentration.
The acyclovir concentration in the skin permeates was corrected
for sampling effects according to the equation described by Hayton
and
Chen
and
Chen,
1982):
(Hayton 
C 1n C n V T =V T  V S  C 1n =C n1 , where C 1n is the corrected
concentration of nth sample, Cn is the measured concentration
of acyclovir in nth sample, Cn  10 is the measured concentration of
acyclovir in (n  1)th sample, VT is total volume of receiver uid,
and Vs is the volume of sample drawn. The rate of drug
permeation (ux) was calculated from the plot of cumulative
amount of drug permeated per cm2 against time.
At the end of in vitro drug permeation study, the cream was
removed from the epidermis using cotton-tipped applicators, and
epidermis washed thrice briey for 15 s with methanol to remove
the traces of adhering cream. The resultant epidermis was added to
a polypropylene conical tube (Becton Dickinson Labware, Franklin
Lanes, NJ, USA) containing 10 ml of extraction solvent (0.5% v/v
acetic acid), screw-capped and sonicated in a bath sonicator
(Branson 8210, Branson Ultrasonics, Danbury, CT, USA) for 99 min
at 32
C. At the end of sonication cycle, samples were withdrawn,
ltered through nylon syringe lter (0.45 mm) and injected (50 ml)
into HPLC for determining acyclovir concentration. The values of
acyclovir concentration were used to calculate the amount of drug
(acyclovir) retained in epidermis (DRE).

114

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

2.11. HPLC analysis of acyclovir

An Agilent 1260 Series high-performance liquid chromatography (HPLC) system equipped with binary solvent pump, autosampler, photodiode array detector, thermostated column
compartment and Chemstation chromatographic software was
used for estimating concentration of acyclovir. The method
described elsewhere in the literature was used with a few
modications (Parry et al., 1992). Waters SunFireTM C18 column
(5 mm, 4.6  150 mm) maintained at 25
C was used to elute
acyclovir. The mobile phase used was an isocratic mixture of glacial
acetic acid and water (0.5% v/v). The ow rate was 1.2 ml/min.
Standard solutions (5 or 50 ml) containing 0.05 to 10 mg/ml of
acyclovir were injected into the HPLC column, and the eluting
acyclovir solutions were detected at 254 nm. The peak areas were
obtained and subjected to regression analysis. A good linear
relationship was observed between the peak area of acyclovir
standard solutions and their concentration with a high correlation
coefcient (r2 = 1.0). The HPLC analytical method was validated
according to USP Validation of Compendial Methods (USP36-NF31,
2013c). The method was precise (intra- and inter-day variation was
<1.0%) and accurate (mean recovery 99.8%). The standard curve,
constructed as described above, was used for determining the
acyclovir quantity in the samples of content uniformity, drug
release, in vitro permeation and DRE studies.
2.12. Statistical analysis
Data collected for responses in each run was analyzed using JMP
software (version 10, SAS, NC, USA) and tted into linear regression
model. F-ratio was calculated to measure the error in the model.
Larger values of F-ratio indicate the smaller error in the model while
smaller values of F-ratio indicate larger error in the model. Model
correctness was assessed by lack of t test, with replicate
experiments (pass if statistically insignicant). Analysis of variance

(ANOVA) was used to compare the multiple means of the responses

by analyzing the variance due to pure random error and differences
between means. In all the cases, HolmSidak test was used for pairwise multiple comparisons (SigmaPlot for Windows Version 12.5,
Systat Software, Inc. USA). A value of p < 0.05 was considered
statistically signicant.
3. Results and discussion
It is essential to identify the scientically sound and clinically
relevant CQA in the approval of Q1/Q2 equivalent generic acyclovir
topical dermatological cream products. Patient tolerance (acceptance) of cream drug products for topical application differs from
those intended for oral or parenteral administrations. In particular,
apart from the therapeutic effect, the acceptance level of creams is
governed by a series of properties unique to the formulation,
collectively known as product's texture prole, which may include
appearance, odor, extrudability (if applicable), initial sensation upon
contact with the skin, spreading properties, tackiness, and residual
greasiness after application, etc., (Barry and Grace, 1972). Of special
interest,spreadabilityofacreamproducthasbeenshownto berelated
to its rheological characteristics through yield stress (s 0), the
minimum shear stress required to initiate irreversible plastic ow.
Yield stress affects the spreading behavior of the cream, and also
plays an important role in the product's physical stability
(e.g., against sedimentation). For this reason, the yield stress was
identied as one of the CQA of acyclovir cream formulations. As
explained under introduction, the functional ability of acyclovir
topicaldermatologicalcreamproductsislikelytobeaffectedbypHof
the formulations (Parry et al., 1992; Shojaei et al., 1998). Thus pH,
acyclovir concentration in aqueous phase, drug retention in
epidermis after 24 h of in vitro drug permeation study across human
epidermis and drug release rate of acyclovir cream formulations
were measured as CQA. The compendial requirement of content
uniformity was also measured in addition to the above CQA.

Fig. 1. X-ray powder diffraction of acyclovir alone, placebo cream and placebo cream + acyclovir.

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

3.1. Acyclovir physical form

X-ray powder diffraction of acyclovir drug powder, placebo
cream, physically mixed drug in the placebo cream, and DOE
formulation are shown in Fig. 1. The drug was crystalline as it
showed characteristics diffraction peaks at 6.28, 6.85, 10.46, 13, 16,
21, 23.90, 26.15 and 29.2
. On the other hand, placebo cream
showed broad and low intensity peaks at 6.35, 21.24, and 21.58

suggesting its partly crystalline nature. Physically mixed drug in
the placebo cream showed peaks of drug and placebo cream.
However, intensity of drug peaks in the placebo cream was much
lower when compared with raw API peaks. This was probably due
to dilution of the drug substance with the placebo cream. Similarly,
DOE formulations showed identical diffraction pattern as that of
physically mixed drug in the placebo cream. This suggested that
processing of drug during cream manufacturing has not changed
its polymorphic/physical forms.
3.2. PH, drug content uniformity, and drug concentration of aqueous
phase
In the nal dosage form of acyclovir cream drug product,
acyclovir may exist in both oil and aqueous phases as soluble form
(aqueous and oil) and suspended form (aqueous and oil). Because
the amount of drug present in the formulation (5% w/w) greatly
exceeds the equilibrium solubility of the drug (24 mg/ml between
pH 29), majority of the drug is expected to be in the aqueous
phase as suspended form (Shojaei et al., 1998). However, with
respect to the product performance and its therapeutic outcome,
solubilized drug in aqueous phase is the most relevant parameter.

115

As shown in Fig. 2, the amount of drug dissolved in aqueous phase

(total aqueous drug) is determined by two equilibriums: solubilization and partitioning (log P = 1.56) (Kristl et al., 1993), which
should remain relatively constant considering that both solid drug
amount and oil phase concentration remain unchanged. Depending on the actual pH of aqueous phase, the total dissolved drug in
aqueous drug phase may present as three different species:
cationic, zwitterionic, and anionic, each of which may have slight
different skin permeation potential (Shojaei et al., 1998). For this
reason, in the current study, pH of aqueous phase was identied as
one risk factor that may impact the product performance (i.e., drug
retention in epidermis) and was incorporated into the experimental design.
The measured pH of acyclovir cream DOE formulations
remained relatively constant ranging from 7.92 to 8.46 despite
the intentional change in pH of the aqueous phase (to 5.00, 6.75, or
8.50). The statistical analysis showed no effect of investigated
processing variables on the nal pH of DOE formulations. The pH of
Zovirax1 was also almost at similar level (7.92). The lack of change
in the pH of DOE formulations is most likely due to the pH altering
effect of formulation ingredients, particularly the SLS (pH > 9) and
the drug addition method. To investigate such a possibility, three
additional acyclovir cream formulations (F-12F-14) were prepared wherein both the drug and SLS were added to aqueous
phase, and then pH adjusted to a varying degree (5.00, 6.75, or
8.50). As expected, when pH was adjusted in the presence of SLS,
the pH matched the design values (pH of F-12F-14 were 5.11, 6.75,
and 8.43, respectively).
The top, middle, and bottom portions of acyclovir cream in the
tubes showed no evidence of phase separation or change in physical

Fig. 2. Role of drug partitioning/solublization and pH on skin permeation and retention in epidermis from acyclovir cream formulations.

116

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

Fig. 3. Polarized light microscopy images of various acyclovir cream formulations (200 magnication, the bar represents 50 mm). At least 10 images were taken for each
sample with total of 200500 particles in order to calculate the size distribution.

appearance and structure. When tested for drug content uniformity,

all the formulations including Zovirax1 complied with USP
specications (USP36-NF31, 2013c). The drug concentration in
aqueous phase of cream formulations was almost the same ranging
from 3.4 to 4.0 mg/ml, which was slightly higher than the aqueous
solubility of acyclovir (2.5 mg/ml at 37
C) (Zovirax, 2002). This was
attributed to the solubilizing effect of formulation ingredients such
as Poloxamer 407, SLS (surfactants), and propylene glycol (cosolvent). Absence of statistically signicant difference (p > 0.05) in
aqueous phase drug concentration of the investigated acyclovir
cream formulations conrms that processing variables have no
signicant impact on the drug concentration in the aqueous phase.
3.3. Effect of various factors on particle size determined by polarized
light microscopy

44.7  8.1 mm using polarized light microscope (Table 2) or

62.7  3.1 mm using HALEOS laser diffraction, both of which were
very similar to those observed in 11 DOE formulations. This
suggests that manufacturing processes have little or no impact on
the particle size of the drug in the nal product. To further conrm
this hypothesis, the raw API was milled using a model 00 Jet-OMizer (Glen Mills Inc. Clifton, NJ) to comparable sizes to the
Zovirax1 (D90 = 19.6  0.8 mm determined using HALEOS or
23.7  2.4 mm using polarized light microscope). Three additional
formulations (F12F14) were prepared with the milled API. The
resulted formulations exhibited similar particle size as compared
to the milled API as well as Zovirax1, conrming that the nal
product particle size is solely dependent on the API particle size
and is less likely to be affected by the manufacturing conditions.
3.4. Rheological characterization

Under polarized light microscope, acyclovir crystal exhibited a

rectangular/square shape as can be seen in Fig. 3. All eleven DOE
formulations exhibited very similar particle size values (Table 2),
and no processing factors showed a signicant effect on the drug
particle size. ANOVA conrmed that the model was not signicant
and independent factors had no relationship with the response
(p > 0.05). However, it is clear that DOE formulations showed
signicantly (p < 0.05) larger particle size as compared to
Zovirax1, most likely due to the size differences of the raw API.
Particle size of the API used for Zovirax1 is unknown, but particle
size (D90) of the raw API used in this study was determined to be

For semi-solid pharmaceutical dosage forms such as creams,

ointments, gels, and lotions, it is highly important to fully
understand the rheology of the product as it may affect both its
application (the feel) and delivery of the active agent to and across
the skin (the therapeutic effectiveness). The rheology of these
semi-solids is highly dependent on the product's characteristic
microstructure, which may be affected by the manufacturing
conditions even for those formulations that are Q1/Q2 equivalent.
For this reason, attempts were made to using rheology tests to
assess the effect of processing conditions on several key

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

117

Table 2
Particle size of acyclovir in various samples, determined using polarized light microscopy.
Sample

Particle count

Length Min.
(mm)

Length Max.
(mm)

D10
(mm)

D50
(mm)

D90
(mm)

Zovirax1
DOE-1
DOE-2
DOE-3
DOE-4
DOE-5
DOE-6
DOE-7
DOE-8
DOE-9
DOE-10
DOE-11

200
279
283
261
433
372
450
408
433
338
417
417

2.5
4.0
4.4
3.1
5.1
3.0
3.6
3.5
2.3
7.0
2.8
3.1

43.4
210.8
211.4
147.2
255.1
197.5
165.4
234.4
197.2
168.5
181.9
176.5

4.3  0.4
9.4  0.3
8.7  1.6
8.7  2.1
8.6  0.7
8.0  0.4
13.3  2.1
13.6  1.4
9.5  1.2
14.3  0.4
10.6  3.3
10.5  0.5

9.3  1.7
18.4  1.0
17.2  2.3
17.5  2.2
18.9  1.4
26.7  3.5
33.8  1.2
37.3  3.0
23.9  2.3
27.9  1.6
25.7  3.6
25.4  2.0

21.8  2.0
60.5  11.9
68.7  17.9
61.6  11.5
55.6  8.4
59.4  7.6
65.9  1.6
78.6  14.1
58.3  2.8
50.9  1.7
54.6  3.7
53.0  3.7

API-raw
API-milled

206
217

4.1
4.5

199.4
52.8

11.9  3.9
9.2  0.6

23.3  2.6
14.3  0.4

44.7  8.1
23.7  2.4

F-12
F-13
F-14

403
401
410

3.2
4.0
2.9

54.4
60.0
61.3

7.1  1.0
7.6  0.3
6.6  0.9

14.1  0.4
14.1  0.1
13.0  1.8

27.5  0.4
27.8  1.7
26.0  5.4

rheological properties: viscoelastic behavior, yield stress, effect of

shear rate on cream viscosity (viscosity at low, medium, and high
shear rate), etc.
Zovirax1 exhibited a typical viscoelastic behavior when
subjecting to increasing strain. As shown in Fig. 4A, the storage
modulus of Zovirax1 cream remained constant up to approximately 1% strain, after which it started to decline. This region is
generally identied as the linear viscoelastic region (LVR) within
which any disturbance to the microstructure is instantaneously
recovered (reversible process). Note that for Zovirax1 cream, the
storage modulus (G0 ) is signicantly higher than the loss modulus
(G00 ), suggesting that the microstructure of the cream is highly
organized and dominated by cohesive forces. Overall material
behaviors like a solid. As the shear strain increases, both G0 and G00
decrease and the material becomes progressively more uid-like
and eventually G00 exceeds G0 . As discussed earlier, when plotting G0
as a function of shear stress exerted on the material during the
oscillation strain sweep, the onset of G0 curve indicates an
irreversible plastic ow of the material and corresponds the
material's dynamic yield stress (s 0). With respect to the
viscoelastic behavior, all DOE formulations exhibited similar LVR
as the Zovirax1 (data not shown).
Experimentally, yield stress can be measured via several
different ways. One commonly used technique is to determine

the stress at the viscosity maximum during a stress ramp

experiment (Kryscio et al., 2008). In this type of test, the viscosity
maximum occurs as a result of two competing effect: (1) time
dependent viscosity build-up of the viscoelastic material; and (2)
viscosity decrease due to structure break-down with increasing
stress (Barnes, 1999; Franck, 2014). Though fairly reproducible, the
measurement result of this technique is sensitive to the selection of
ramp rate. For this reason, in this study another method was
chosen, by determining the onset of storage modulus (G0 ) versus
shear stress curve during an oscillation stress/strain sweep
experiment (Fig. 4B) (Pal 1999).
3.4.1. Effect of various factors on product's yield stress
All DOE formulations exhibited similar yield stress values
(ranging from 45.9 to 91.6 Pa) to both blank (57.6  1.9 Pa) and
Zovirax1 (73.0  13.7 Pa) as shown in Fig. 5. Statistical analysis
conrmed that the investigated processing parameters showed no
signicant (p > 0.05) impact on the product yield stress. Though the
yield stress might be affected by formulation variables, it is not
affected by the process variables investigated in the present study.
The formulation variables were kept constant for all the DOE
formulations (Q1/Q2 equivalent). Three additional formulations
(F12-F14) exhibited similar yield stress values compared to
Zovirax1, blank, and DOE formulations, suggesting that all samples

Fig. 4. Strain sweep test (0.0550% strain at 1 Hz) for a Zovirax1 cream sample: (A) assessment of linear viscoelastic region; (B) determination of yield stress using onset of
storage modulus.

118

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

Fig. 5. Yield stress of various formulations (n = 5 for Zovirax1 and n = 3 for all the other samples).

should have similar spreading behavior over the skin. A yield stress
value between 10 and 100 Pa requires very minimal force to initiate
the ow (110 mN force over 1 cm2 area), and hence the cream is
expected to be very easily applied onto and spread over the skin.
3.4.2. Effect of shear rate on sample viscosity
All tested formulations exhibited similar plastic ow behavior
as compared to Zovirax1 (Fig. 6 and 7). Under the low shear
condition, the cream displays very high viscosity (10,000 Pa  s),
giving the rmness feel to the product. As the shear increases,
product viscosity quickly reduces (to <10 Pa  s) to allow easy
spreading over the skin to cover large area. This type of prole is an

important characteristic of the product and can only be obtained

when tested at various shear rate conditions. With regard to the
investigation of the effect of processing conditions, the tested
parameters were determined to have no impact on the viscosity
prole (p > 0.05). However, one formulation (DOE-3) showed
slightly higher viscosity at low (0.001 s1) and medium (1 s1)
shear rate conditions but slightly lower viscosity at high (>20 s1)
shear rate, suggesting that this particular formulation would have
higher rmness feel to it under the low shear but easier to ow
during spreading than Zovirax1. Three additional formulations
(F12-F14) exhibited similar viscosity proles compared to all the
DOE formulations.

Fig. 6. Viscosity prole of acyclovir cream formulations as a function of shear rate (n = 5 for Zovirax1 and n = 3 for other samples).

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

119

Fig. 7. Viscosity of acyclovir cream formulations at three different shear rates (n = 5 for Zovirax1 and n = 3 for other samples).

3.5. Cream stability: rheological considerations

It was noticed that during storage all of the DOE formulations
(DOE-1 to DOE-11) were accidentally exposed to elevated
temperature (3238
C) for an unknown period of time due to
the air-conditioner malfunctioning at the facility. After such an
exposure, the yield stress and viscosity of DOE samples were
signicantly lower (data not shown). This suggests a destabilization of the cream structure. To understand the effect of
temperature on cream stability, temperature ramping tests were
performed on a series of new samples (Zovirax1 and three new
formulations F-12F-14). As shown in Fig. 8, three new formulations behaved similar to the Zovirax1: the storage modules
remained relatively constant until temperature reached approx.
32
C, above which sample behaved more like a liquid. Storage

Fig. 8. Temperature effect on storage modulus of acyclovir cream formulations.

under this elevated temperature conditions may accelerate phase

separation under static environment, causing irreversible change
in the rheological prole. On another interesting note, the cream's
storage modulus was signicantly lower at temperature above
32
C (skin temperature). This makes it even easier to spread the
cream over the skin (in addition to the shear rate effect).
3.6. In vitro drug release study
The in vitro drug release studies were carried out to detect
the effect of minor changes in process variables involved in the
manufacturing of acyclovir cream formulations (FDA, 1997). The
percent of drug transported from acyclovir cream formulations
into the dissolution medium ranged from 1.74  0.23 to 2.10  0.16
based on the total drug amount (dissolved and undissolved), or
31.26  3.37 to 41.69  3.18 based on aqueous drug amount
(assuming acyclovir concentration remains constant in aqueous
phase). At the end of the test, there was still a large portion of drug
available for release. In order to enhance the passive transport of
drug across the skin, the thermodynamic activity of drug(s) is
maintained high with large quantity of drug in topical and
transdermal products (Allen Jr et al., 2005; Davis and Hadgraft,
1991; Iervolino et al., 2000; Pellett et al., 1994). When the amount
of drug diffused per cm2 of membrane was plotted against square
root of time (Higuchi diffusion kinetics), there was a linear
relationship with high correlation coefcient (0.9687 to 0.9987)
(USP36-NF31, 2013b). The drug release rate for the acyclovir cream
formulations was ranging from 0.32  0.03 to 0.38  0.02 mg/
cm2  h0.5. The RLD also showed almost the same drug release rate
of 0.35  0.04 mg/cm2  h0.5. Statistical analysis of the DOE model

120

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

Table 3
Equivalence testing based on 90% condence interval for T/R ratio (expressed as %) of drug release rate of acyclovir from acyclovir cream formulations against Zovirax1
(Reference, R).
Test formulation
(T)
DOE-1
DOE-2
DOE-3
DOE-4
DOE-5
DOE-6
DOE-7
DOE-8
DOE-9
DOE-10
DOE-11
F-12
F-13
F-14
a

90% Condence Interval for T/R ratio (expressed as %) of drug release rate
First stage testing (n = 6)

Compliance in rst stagea (Yes/No)

Second stage testing (n = 18)

Compliance in second stagea (Yes/No)

61.3106.0
74.998.9
87.9126.4
93.0122.2
96.8123.0
90.6119.8
88.9117.5
93.4123.4
88.1113.2
100.1129.5
78.7100.3
73.697.9
81.0107.1
90.5118.9

No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes

79.6101.5
82.699.5

84.294.5

Yes
Yes

Yes

Complies if values are between 75% and 133.3%.

showed that none of the processing variables or pH of aqueous

phase affected drug release rate of acyclovir from the topical cream
formulations. Even direct comparison of drug release rates
between individual DOE cream formulations and RLD showed
no signicant difference (ANOVA with HolmSidak test).
Because outliers are expected to occur on occasion with in vitro
drug release testing (for example, due to an air bubble between the
product sample and the membrane), a nonparametric statistical
technique was used as described in USP and FDA guidance document
(FDA, 1997; USP36-NF31, 2013b). Since the investigated acyclovir
cream formulations (DOE-1DOE-11, F-12F-14) are Q1/Q2 equivalent, the sameness of these formulations with respect to their
performance (in vitro drug release) may be considered as level 1
change (due to possible minor variation in processing parameters).
The in vitro drug release rate of DOE formulations, F-12F-14 (test
formulations) were compared against Zovirax1 (Reference formulation) as a two-stage test. If the 90% condence interval for the ratio
of test to reference release rates does not fall within 0.75 and 0.13333
(or 75133.33%) in rst stage (6 cells each for test and reference), four
additional tests (12 cells each for test and reference) were carried out.
All the acyclovir cream formulations except DOE-1, DOE-2, and F-12
complied with the sameness of drug release rate in rst stage.
However, the DOE-1, DOE-2, and F-12 formulations also complied
with the sameness of drug release rate at the second stage of testing
(Table 3). These results showed that that none of the processing
variables affected the drug release performance of acyclovir
formulations indicating that Q1/Q2 equivalent generic acyclovir
cream formulations can be prepared to match the performance of
Zovirax1 (RLD).
3.7. Drug deposition (or retained) in human epidermis
The in vitro drug permeation studies across human skin can
detect the difference in topical delivery of generic acyclovir topical

creams which vary in formulation composition, and therefore used

in this study for detecting the effect of process variables on Q1/Q2
equivalent acyclovir cream formulations (Trottet et al., 2005). The
functional ability of acyclovir topical dermatological products
depends on the ability of the formulation to retain the drug at the
site of action in basal layers of epidermis. Drug retained in
epidermal layers after 24 h of in vitro skin permeation study was
able to detect the functionality difference between 1% penciclovir
cream and 5% acyclovir cream (Hasler-Nguyen et al., 2009). Hence,
the amount of acyclovir retained at the end of 24 of in vitro drug
permeation study (DRE) was determined in the present study.
Since none of the investigated process variables or pH (a
formulation variable) affected the structural characteristics
(viscosity, yield stress, particle size, drug concentration of aqueous
phase) and performance (in vitro drug release rate) of acyclovir
cream formulations, DRE studies were carried out only with F-12
F-14 cream formulations against Zovirax1 cream (RLD). All the
three formulations (F-12F-14) showed similar structural characteristics (viscosity, yield stress, particle size, drug concentration of
aqueous phase) and drug release rate prole as that of RLD, but
differ only with respect to pH of aqueous phase. The observed pH of
F-12F-14 acyclovir cream formulations were 5.11, 6.95, and 8.43,
respectively whereas the pH of Zovirax1 was 7.92. As discussed
above, the pH of acyclovir cream formulations play a critical role in
their functional ability to retain the drug in the targeted layers of
human epidermis (Shojaei et al., 1998).
The mean (S.E.M.) in vitro permeation parameters of
Zovirax1, F-12F-14 acyclovir cream formulations across human
epidermis (n = 9) are shown in Table 4. There was a steady state
drug ux from acyclovir topical cream formulations (F12, F-13, F14 and Zovirax1) across human epidermis. The quantity of drug
permeated in 24 h from F-12 and F-13 were signicantly (p < 0.05)
higher when compared to Zovirax1 (about 50% higher). Accordingly the steady-state ux of acyclovir from F-12 and F-13

Table 4
Mean (S.E.M.) in vitro permeation parameters of Zovirax1, F-12F-14 acyclovir cream formulations across human epidermis (n = 9).
Formulation
Zovirax
F-12
F-13
F-14
*
**

Quantity of drug permeated at 24 h (mg/cm2)

Flux (mg/cm2  h)

Drug retained in epidermis, DRE (mg/cm2) after 24 h of in vitro permeation

12.04  1.10
19.41  1.53**
17.89  1.99*
16.25  2.24

0.78  0.06
1.22  0.12**
1.14  0.13*
1.07  0.15

2.20  0.34
2.91  0.43
3.42  0.49
2.16  0.41

p < 0.05 when compared to Zovirax1.

p < 0.01 when compared to Zovirax1.

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

Fig. 9. Mean (S.D.) drug release rate of acyclovir from acyclovir topical cream
formulations (n = 12 to18).

formulations was signicantly higher than Zovirax1 (Fig. 9).

However, neither the ux nor the amount of drug permeated at
24 h from F-14 was signicantly different to Zovirax1.
The signicantly higher skin permeation from F-12 and F-13
compared to Zovirax1 as shown in Fig. 10 can be explained on the
basis of the observed differences in their pH values. The pKa values
of acyclovir are 2.27 and 9.25, giving its isoelectric point at pH 5.76
(Shojaei et al., 1998). The pH of F-12 (5.11) is closest to the drug's
isoelectric point, resulting in acyclovir less charged than in F-13
(pH 6.75), F-14 (pH 8.43), and Zovirax1 (pH 7.92). The polarity of
acyclovir in these four cream formulations increases with an
increase in the pH of the product. As the non-polar small molecules
are more skin permeable than the polar counterparts (Subedi et al.,
2010), the less charged acyclovir in F-12 achieved the highest skin
permeation, followed by F-13 and F-14, as compared to Zovirax1.
The pH of F-14 is the closest to Zovirax1, demonstrating similar
skin permeation proles (Table 4).
Therapeutic performance of topically applied dermatological
creams depends on their ability to act locally in epidermal/dermal
layers of skin. In case of acyclovir topical dermatological creams,
the site of therapeutic action is basal epidermal layers (Parry et al.,
1992). Hence, it is necessary to determine the amount of drug
retained/deposited in epidermis (DRE) at the end of 24 h of in vitro

121

permeation study. This represents the best quality metrics to

assess the in vitro performance of acyclovir topical dermatological
creams. The mean (S.E.M.) amount of drug retained in epidermis
(DRE) with Zovirax1 was 2.20  0.34 mg/cm2. The DRE values with
F-12F-14 were 2.91  0.43, 3.42  0.49, and 2.16  0.41 mg/cm2,
respectively. However, there was no statistical signicant difference in DRE values of F-12F-14 formulations when compared to
that observed with Zovirax1 (Table 4). Although not statistically
signicant, F-12 and F-13 exhibited greater DRE and skin
permeation ux than Zovirax1, also indicating the pH effect on
their in vitro performance. The ability of epidermis to retain the
drug in its layers is a saturation process. Since acyclovir cream
formulations are in contact with the epidermis for 24 h, it is
possible that their functional ability reached to saturation levels,
and thus the DRE values are not signicantly different. The DRE
values were also expressed as the amount (mg) of acyclovir
retained per cm3 (ml) of epidermis. The volume of exposed
epidermis was calculated by multiplying the average thickness of
human epidermis (0.007127 cm) with exposed area (1.717 cm2).
The mean (S.E.M.) DRE with Zovirax1 was 308.4  47.9 mg/ml
whereas the DRE values with F-12F-14 were 408.4  60.8,
497.7  68.3 and 303.7  57.2 mg/ml, respectively. These drug
concentrations available in epidermal layers are in far more than
the desired IC50 values of acyclovir (0.0213.5 mg/ml) to produce
therapeutic effect (Zovirax, 2002). The results of in vitro drug
permeation studies across human epidermis suggest that Q1/Q2
equivalent generic acyclovir topical cream products show sameness in terms of functional ability to provide the desired
therapeutic drug concentrations at the targeted site of action in
epidermal layers.
It is well known that change in formulation composition of
acyclovir topical creams affects their therapeutic efcacy. For
example, it was reported that generic acyclovir topical creams
containing varying levels of propylene glycol exhibited differences
in therapeutic efcacy (Trottet et al., 2005). However, in the
present study formulation excipients and their composition were
kept the same (Q1/Q2). The results suggested generic acyclovir
creams can be manufactured to obtain identical performance as
that of RLD with Q1/Q2/Q3.
4. Conclusions
Quality by Design approach was used to identify the effect of
process variability on structural and functional sameness (Q3) of
qualitatively (Q1) and quantitatively (Q2) equivalent generic
acyclovir topical dermatological cream products in comparison
with RLD (Zovirax1). The investigated critical process variables
(homogenization speed, homogenization time, emulsication
temperature) did not affect the structural and functional characteristics with respect to content uniformity, particle size, spreadability on skin (yield stress), viscosity (at low, medium, and high shear
rates), drug concentration in aqueous phase, and in vitro drug
release. Although not statistically signicant, acyclovir creams
with a pH of 5.11 (F-12) and 6.75 (F-13) exhibited greater DRE and
skin permeation ux than Zovirax1 indicating the pH effect on
their in vitro performance. The thermo-rheological characterization suggested that storage conditions above 30
C should be
avoided as it may bring irreversible changes in storage modulus
affecting product storage stability. The study concluded that
generic Q1/Q2 equivalent acyclovir topical creams can be
manufactured with similar Q3 matching to RLD.
Conict of interest

Fig. 10. Mean (S.E.M.) amount of drug permeated from Zovirax1, F-12F-14
acyclovir cream formulations across human epidermis (n = 9).

This scientic contribution is intended to support regulatory

policy development. The views presented in this article have not

122

Y.S.R. Krishnaiah et al. / International Journal of Pharmaceutics 475 (2014) 110122

been adopted as regulatory policies by the Food and Drug

Administration at this time.
Acknowledgements
Ofce of Generic Drugs is gratefully acknowledged for providing
funding to carry out this research. The authors would like to thank
Robert Hunt for his help with the in vitro drug release studies. The
National Disease Research Interchange (NDRI, Philadelphia, PA,
USA) is also acknowledged for providing the human cadaver skin.
References
21CFR320.1, 2014. Title 21: Food and Drugs: Part 320Bioavailability and
Bioequivalence Requirements, Subpart AGeneral Provisions. Available at
http://ecfr.gpoaccess.gov/cgi/t/text/text-idx?c=ecfr&sid=a2c8242fba4332eaa1b90947931245a3&rgn=div5&view=text&node=21:5.0.1.1.7&idno=21#21:5.0.1.1.7.1.1.1H (accessed 30.3.14).
21CFR320.24, 2014. Title 21: Food and Drugs: PART 320Bioavailability and
Bioequivalence Requirements, Subpart BProcedures for Determining the
Bioavailability or Bioequivalence of Drug Products. Available at http://www.
ecfr.gov/cgi-bin/text-idx?SID=301bbb87cfab41036bc82f497f1aa9aa&node=pt21.5.320&rgn=div5#se21.5.320_124 (accessed 30.3.2014).
Acosta, E.P., Flexner, C., 2011. Chapter 58: Antiviral agents (Nonretroviral), In:
Brunton, L.L., Chabner, B.A., Knollmann, B.C. (Eds.), Goodman & Gilman's The
Pharmacological Basis of Therapeutics. twelfth ed. McGraw Hill, New York
Available
at
http://accessmedicine.mhmedical.com/content.aspx?bookid=374&Sectionid=41266269 (accessed 22.4.14).
Andersen, F., Hedegaard, K., Petersen, T.K., Bindslev-Jensen, C., Fullerton, A.,
Andersen, K.E., 2006. The hairless guinea-pig as a model for treatment of
cumulative irritation in humans. Skin. Res. Technol. 12, 6067.
Barnes, H.A., 1999. The yield stress a review or pi alpha nu tau alpha rho epsilon
iota' everything ows? J Non-Newton. Fluid. Mech. 81, 133178.
Barry, B.W., Grace, A.J., 1972. Sensory testing of spreadability: investigation of
rheological conditions operative during application of topical preparations. J.
Pharm. Sci. 61, 335341.
Bhandari, M., Lochner, H., Tornetta, P., 2002. Effect of continuous versus
dichotomous outcome variables on study power when sample sizes of
orthopaedic randomized trials are small. Arch. Orthop. Traum. Su. 122, 9698.
Chang, R.K., Raw, A., Lionberger, R., Yu, L., 2013a. Generic development of topical
dermatologic products, part II: quality by design for topical semisolid products.
AAPS J. 15, 674683.
Chang, R.K., Raw, A., Lionberger, R., Yu, L., 2013b. Generic development of topical
dermatologic products: formulation development, process development, and
testing of topical dermatologic products. AAPS J. 15, 4152.
Davis, A.F., Hadgraft, J., 1991. Effect of supersaturation on membrane-transport. 1.
Hydrocortisone acetate. PInt. J. Pharm. 76, 18.
Donner, A., Eliasziw, M., 1994. Statistical implications of the choice between a
dichotomous or continuous trait in studies of interobserver agreement.
Biometrics 50, 550555.
EMA, 2000. Committee for Proprietary Medicinal Products: note for guidance on the
investigation of bioavailability and bioequivalence. Available at http://www.
ema.europa.eu/docs/en_GB/document_library/Scientic_guideline/2009/09/
WC500003519.pdf (accessed 30.03.14).
FDA, 1997. Guidance for Industry. Nonsterile semisolid dosage forms, scale-up and
postapproval changes: chemistry, manufacturing, and controls; in vitro release
testing and in vivo bioequivalence documentation. Available at http://www.fda.
gov/downloads/drugs/guidancecomplianceregulatoryinformation/guidances/
ucm070930.pdf (accessed 30.03.14).
FDA, 2010. Guidance for Industry Bioequivalence Recommendations for Specic
Products. Available at http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM072872.pdf (accessed 30.03.14).
FDA, 2012. Draft guidance on acyclvir. Available at http://www.fda.gov/downloads/
Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM296733.
pdf (accessed 30.03.14).
FDA, 2013. Draft guidance on cyclosporine. Available at http://www.fda.gov/
downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/
UCM358114.pdf (accessed 30.03.14).
Franck, A., Understanding Rheology of Structured Fluids (Technical Note AAN016;
TA Instruments). Available at http://www.tainstruments.com/pdf/literature/
AAN016_V1_U_StructFluids.pdf (accessed 30.03.14).
Hasler-Nguyen, N., Shelton, D., Ponard, G., Bader, M., Schaffrik, M., Mallefet, P., 2009.
Evaluation of the in vitro skin permeation of antiviral drugs from penciclovir 1%

cream and acyclovir 5% cream used to treat herpes simplex virus infection. BMC
Dermatol. 9, 3.
Hayton, W.L., Chen, T., 1982. Correction of perfusate concentration for sample
removal. J. Pharm. Sci. 71, 820821.
Herkenne, C., Alberti, I., Naik, A., Kalia, Y.N., Mathy, F.X., Preat, V., Guy, R.H., 2008. In
vivo methods for the assessment of topical drug bioavailability. Pharmaceut.
Res. 25, 87103.
Iervolino, M., Raghavan, S.L., Hadgraft, J., 2000. Membrane penetration enhancement of ibuprofen using supersaturation. Int. J. Pharm. 198, 229238.
Krishnaiah, Y.S.R., Al-Saidan, S.M., Chandrasekhar, D.V., Rama, B., 2006. Effect of
nerodilol and carvone on in vitro permeation of nicorandil across rat epidermal
membrane. Drug Dev. Ind. Pharm. 32, 423435.
Kristl, A., Vesnaver, G., Mrhar, A., Kozjek, F., 1993. Evaluation of partitioning and
solubility data for some guanine derivatives in terms of mutual miscibility of
octanol and water phase. Pharmazie 48, 608610.
Kryscio, D.R., Sathe, P.M., Lionberger, R., Yu, L., Bell, M.A., Jay, M., Hilt, J.Z., 2008.
Spreadability measurements to assess structural equivalence (Q(3)) of topical
formulations a technical note. AAPS Pharmscitech 9, 8486.
Lionberger, R.A., 2008. FDA critical path initiatives: opportunities for generic drug
development. AAPS J. 10, 103109.
Mateus, R., Abdalghafor, H., Oliveira, G., Hadgraft, J., Lane, M.E., 2013. A new
paradigm in dermatopharmacokinetics confocal Raman spectroscopy. Int. J.
Pharm. 444, 106108.
N'Dri-Stempfer, B., Navidi, W.C., Guy, R.H., Bunge, A.L., 2008. Optimizing metrics for
the assessment of bioequivalence between topical drug products. Pharmaceut.
Res. 25, 16211630.
N'Dri-Stempfer, B., Navidi, W.C., Guy, R.H., Bunge, A.L., 2009. Improved bioequivalence assessment of topical dermatological drug products using dermatopharmacokinetics. Pharmaceut. Res. 26, 316328.
Narkar, Y., 2010. Bioequivalence for topical products an update. Pharmaceut. Res.
27, 25902601.
Navidi, W., Hutchinson, A., N'Dri-Stempfer, B., Bunge, A., 2008. Determining
bioequivalence of topical dermatological drug products by tape-stripping. J.
Pharmacokinet. Phar. 35, 337348.
Pal, R., 1999. Yield stress and viscoelastic properties of high internal phase ratio
emulsions. Colloid. Polym. Sci. 277, 583588.
Parry, G.E., Dunn, P., Shah, V.P., Pershing, L.K., 1992. Acyclovir bioavailability in
human skin. J. Invest. Dermatol. 98, 856863.
Pellett, M.A., Davis, A.F., Hadgraft, J., 1994. Effect of supersaturation on membranetransport. 2. Piroxicam. Int. J. Pharm. 111, 16.
Shah, V.P., Flynn, G.L., Yacobi, A., Maibach, H.I., Bon, C., Fleischer, N.M., Franz, T.J.,
Kaplan, S.A., Kawamoto, J., Lesko, L.J., Marty, J.P., Pershing, L.K., Schaefer, H.,
Sequeira, J.A., Shrivastava, S.P., Wilkin, J., Williams, R.L., 1998. Bioequivalence of
topical dermatological dosage forms methods of evaluation of bioequivalence.
Pharmaceut. Res. 15, 167171.
Shojaei, A.H., Berner, B., Li, X.L., 1998. Transbuccal delivery of acyclovir: I In vitro
determination of routes of buccal transport. Pharmaceut. Res. 15, 11821188.
Subedi, R.K., Oh, S.Y., Chun, M.K., Choi, H.K., 2010. Recent advances in transdermal
drug delivery. Arch. Pharm. Res. 33, 339351.
Trottet, L., Owen, H., Holme, P., Heylings, J., Collin, I.P., Breen, A.P., Siyad, M.N.,
Nandra, R.S., Davis, A.F., 2005. Are all aciclovir cream formulations bioequivalent? Int. J. Pharm. 304, 6371.
USP36-NF31, 2013a. General Chapter: <3>Topical and Transdermal Drug ProductsProduct Quality Tests. Available at http://www.uspnf.com/uspnf/pdf/download?usp=36&nf=31&s=2&q=usp36nf31s2_c3.pdf&ofcialOn=December
(accessed 30.3.14) US Pharmacopeia, Rockville, MD, USA.
USP36-NF31, 2013b. General chapter <1724> Semisolid drug products-Performance
tests.
Available
at
http://www.uspnf.com/uspnf/pdf/download?
usp=36&nf=31&s=2&q=usp36nf31s2_c1724.pdf&ofcialOn=December
(accessed 30.3.14). US Pharmacopeia, Rockville, MD, USA, pp. 5778-5788.
USP36-NF31, 2013c. General Chapter <621> Errata to second supplement,
Chromatography. Available at http://www.uspnf.com/uspnf/pub/index?
usp=36&nf=31&s=1&ofcialOn=August%201,%202013 (accessed 30.3.14). US
Pharmacopeia, Rockville, MD, USA, pp. 1-7.
Yacobi, A., Shah, V.P., Bashaw, E.D., Benfeldt, E., Davit, B., Ganes, D., Ghosh, T., Kanfer,
I., Kasting, G.B., Katz, L., Lionberger, R., Lu, G.W., Maibach, H.I., Pershing, L.K.,
Rackley, R.J., Raw, A., Shukla, C.G., Thakker, K., Wagner, N., Zovko, E., Lane, M.E.,
2014. Current challenges in bioequivalence, quality, and novel assessment
technologies for topical products. Pharmaceut. Res. 31, 837846.
Zhao, K.D., Singh, J., 1999. In vitro percutaneous absorption enhancement of
propranolol hydrochloride through porcine epidermis by terpenes/ethanol. J.
Control. Release 62, 359366.
Zovirax1, 2002. Prescribing information: ZOVIRAX1 (acyclovir) Cream 5%. Available
at
http://www.accessdata.fda.gov/drugsatfda_docs/label/2002/21478_zovirax_lbl.pdf(accessed 30.03.14).