A Prospective, Split-Face, Randomized, Comparative

Study of Safety and 12-Month Longevity of Three

Formulations of Hyaluronic Acid Dermal Filler for Treatment
of Nasolabial Folds
WELF PRAGER, MD,* ESTHER WISSMUELLER, MD,* ISABEL HAVERMANN, MD,* EVA K. BEE, MD,
DAVID J. HOWELL, PHD, RRT, INA ZSCHOCKE, MD, AND JEANNETTE SIMON

BACKGROUND Data regarding several hyaluronic acids (HAs) used identically for facial tissue augmentation
have heretofore been unavailable.
OBJECTIVES This prospective, split-face, randomized, two-armed study sought to determine the long-term
safety and effectiveness of three HAs (HA-1 (Belotero Basic/Balance), HA-2 (Restylane), and HA-3 (Juvederm
Ultra 3/Juvederm Ultra Plus XC) in the treatment of nasolabial folds (NLFs).
METHODS Twenty participants in Arm A received HA-1 in one NLF and HA-2 in the other. In Arm B, 20
participants received HA-1 in one NLF and HA-3 in the other. Injection was at visit 2, with follow-up visits at 1,
6, 9, and 12 months. Mean volume of HA was slightly <1.5 mL/NLF.
RESULTS Adverse events were unremarkable across all HAs, with injection site erythema being the most
frequent adverse event. Mean pretreatment NLF severity rating for both arms was 2.3; at 12 months, mean
posttreatment severity rating was 1.5 for HA-1/HA-2 and 1.6 for HA-1/HA-3. Although not statistically
significant, participants tended to show a preference for HA-1.
CONCLUSION All three HAs provided essentially equivalent results, except for 4-week evenness results,
which favored HA-1. Injection volumes of the three HAs were also similar.
Merz Pharmaceuticals provided the necessary products and supplies for execution of this study and has
compensated Dr. Prager for presentations to the medical community. David Howell is a medical writer paid by
Merz Pharmaceuticals to assist in the preparation of this manuscript.

iological aging of the skin, involving the loss of

HA, subcutaneous fatty tissue, and collagen
and elastic fibers differs among individuals. The
aging process and facial wrinkle formation depend
on genetic predisposition, chronic exposure to
ultraviolet (UV) radiation, deleterious environmental factors, nicotine, and alcohol. Unwanted wrinkles can be corrected using a variety of methods
depending on their nature and causation. An effective method of treating facial folds is augmentation,
in which degradable or permanent dermal fillers of

various types are implanted or injected. The most

commonly used dermal filler is hyaluronic acid
(HA).
HA, or hyaluronate, is a polysaccharide (glycosaminoglycan) found in vertebrate extracellular tissue
matrix. The chemical structure of pure HA is always
the same regardless of species and type of tissue
non-cross-linked, a polymer of linear structure, a
chain moleculebut with different molecular
weight distributions. The free functional groups of

*Dermatologikum, Hamburg, Germany; Private Practice, Muenster, Germany; Medical Writer, San Francisco,
California; SCIderm GmbH, Hamburg, Germany; Merz Pharamaceuticals, GmbH, Frankfurt am Main, Germany
2012 by the American Society for Dermatologic Surgery, Inc.  Published by Wiley Periodicals, Inc. 
ISSN: 1076-0512  Dermatol Surg 2012;38:11431150  DOI: 10.1111/j.1524-4725.2012.02468.x
1143

COMPARATIVE SAFETY AND LONGEVITY OF THREE HAs

this polymer can react with other substances and can

bind water, which gives the skin suppleness and
volume. When injected into the connective tissue,
native, non-cross-linked HA is absorbed with a halflife of 1 to 2 days.1 The short half-life of non-crosslinked HA makes it undesirable and unfeasible for
volume augmentation and wrinkle treatment.
To prevent the rapid degradation of HA and prolong
its persistence in the skin, cross-linked HAs have
been synthesized. The various HA preparations
differ in origin of HA used, molecule length, and
production techniques. They are extracted from
animal tissue (e.g., rooster combs) or produced using
gene technology or fermentative processes from
equine streptococcal strains not pathologic to
humans.1 These chemically modified, cross-linked
polymers are biologically inert, water insoluble, and
more or less viscoelastic. They are much more
viscous than the non-cross-linked HAs and are thus
suitable as biologically compatible filler material for
the treatment of lines and wrinkles.
This report presents longevity findings on three
cross-linked HAs: Belotero Basic2 (Belotero Balance
in the United States; HA-1, Merz Aesthetics,
Frankfurt am Main, Germanymanufactured
by Anteis), Restylane3 (HA-2; Q-MED, Uppsala,
Swedendistributed in the United States by Medicis
Aesthetics, Scottsdale, AZ), and Juvederm4 Ultra 3
(HA-3; Juvederm Ultra Plus XC in the United States;
Allergan, Inc., Irvine, CA). HA-1 and HA-3 are
nonsieved, cohesive gels but are produced in
different ways, and HA-2 is a particulated, sieved
gel.5 Sieved gels, such as HA-2, contain HA particles
of relatively uniform size. Depending on indication,
preparations containing different particle sizes may
be used. No particle sizing is designed for gels
manufactured using a nonsieved process (HA-1 and
HA-3). HA-1 has a HA sodium concentration of
22.5 mg/mL, with a unique production process that
delivers different density zones, leading to a cohesive
polydensified matrix without particles; HA-2 has a
HA sodium concentration of 20 mg/mL; and HA-3
has a HA sodium concentration of 24 mg/mL. All

1144

DERMATOLOGIC SURGERY

dermal fillers are manufactured in a cross-linking

process with 1,4-butanediol-diglycidyl-ether.
HA-2 and HA-3 were compared with HA-1 in two
separate but nearly identically designed, randomized, split-face, prospective arms of a single study;
these arms are the focus of this article. Four-week
results of Arm A (HA-1/HA-2) have been reported
elsewhere6 but are discussed at length in this report,
along with long-term results as well. This study used
an innovative standardized tool called Phase-shift
Rapid In-Vivo Measurement of Skin (PRIMOS, GF
Messtechnik GmbH, Teltow/Berlin, Germany) to
measure intra-individual differences in depth of
nasolabial folds (NLFs) treated with HA-1 and
HA-2. The 20-participant study supported greater
evenness of the HA-1 NLF than the HA-2 NLF at
4 weeks, probably because of the reported greater
degree of tissue and dermal integration of HA-1 than
of HA-2 after injection and resulting in more even,
or smoother, outcomes.7
An earlier study compared Juvederm (HA-3) with a
standard HA (Restylane, HA-2).8 Cosmetic results
were comparable, but 95% of participants preferred
HA-3 for overall injection comfort. (The HA-3
preparation contained lidocaine.) Finally, a multicenter, prospective, noncontrolled study of HA-1 in
109 volunteers injected bilaterally with the product
has also been published.9 Favorable results of HA-1
out to 36 weeks were reported for cosmetic and
participant satisfaction outcomes.
No cross-HA trials comparing the cosmetic outcomes of HA-2 or HA-3 with HA-1 have been
reported. This article explains the study design and
then reports safety and longevity findings at 4 weeks
and 6, 9, and 12 months for the two arms of the
study, one with HA-1 and HA-2 and the other with
HA-1 and HA-3, in treatment of NLFs.

Study Approval and Objective

The three HA fillers investigated in this study carry
a Conformite Europeenne mark, indicating that

PRAGER ET AL

they have undergone a conformity assessment

procedure. Study details were submitted to the
responsible independent ethics committee, and
approval was obtained before the study began in
October 2008.

Methods and Materials

Participant Population
All individuals who gave written consent to participate in the study and who fulfilled all other
inclusion and exclusion criteria were enrolled in the
study. At initial diagnosis, participants in both
studies had to have bilateral, symmetrical NLFs and
a desire for cosmetic correction. Men and women
aged 30 to 65 who agreed not to have any other
treatment to correct NLFs during the study period
and had documented Merz Severity Scale score of 3
or 4 for NLFs on a five-point scale met the criteria
for enrollment.10

Individuals with scars in their NLFs; known hypersensitivity to HA, lidocaine, or other amide-type
local anesthetics; previous or concomitant therapy;
any nonpermanent filler implantation or correction
<6 months before start of study; any permanent
filler or implant in the NLF region; any treatment
with fat grafting in the NLF; and any facial hair in
the NLF were excluded.
In Arm A, 20 participants (19 female, 20 Caucasian)
aged 30 to 60 (mean age 45.8) were treated with
HA-1 and HA-2 in each NLF to full cosmetic
correction. All pretreatment NLFs were rated as
moderate (15 participants) or severe (5 participants)
using the Merz Scale (1, absence of visible folds; 2,
mild folds; 3, moderate folds; 4, severe folds; 5,
extreme folds; Figure 1).10
In Arm B, 20 participants (19 female, 20 Caucasian)
aged 35 to 64 (mean age 45.9) were treated with
HA-1 and HA-3. All pretreatment NLFs were rated

Figure 1. Merz Nasolabial fold severity scale.7

38:7 PART II:JULY 2012

1145

COMPARATIVE SAFETY AND LONGEVITY OF THREE HAs

as moderate (13 participants) or severe (seven

participants) using the five-point Merz Scale.

massage the dermal filler at home but were allowed

to resume normal activities. Pain reliever was
allowed after injection (e.g., acetaminophen).

Study Design
This two-arm prospective, randomized, singlecenter, Phase IV clinical comparison trial was
conducted in one country (Germany) pursuant to
Good Clinical Practices. Each arm was originally
designed for 4 weeks but was extended to
12 months to assess cosmetic effectiveness and
participant satisfaction of the HAs under investigation. The initial phase of both arms consisted of
three visits, one screening visit (V1); the baseline
visit (V2), during which injection of the fillers took
place; and the 1-month visit (week 4 5 days, V3).
Subsequent follow-up visits to assess cosmetic effects
were at 6, 9, and 12 months.
Treatment Protocol
Before treatment, the NLFs were cleaned with
antiseptic. Each participant was then injected with
HA-1 in one NLF and HA-2 (Arm A) or HA-3 (Arm
B) in the contralateral NLF, according to a split-face
design, with allocation of fillers to the side of the
face randomized. An allergy test was not necessary
before the injection because the fillers are pure
polysaccharides. In both studies, the filler was to be
injected according the instructions for use in the mid
to deep dermis. Treatment was to full correction as
assessed by the treating physician; overcorrection
was not allowed.
Dosages of the three HA fillers differed between
individuals, depending on the depth of the NLF,
with the goal of optimal correction of both NLFs
during a single injection session of each HA filler.
Injection was performed using a 27G 0.5-inch needle
using a threading or multiple puncture technique.
No local anesthetics, systemic painkillers, or local
ice cooling were used to reduce pain during injection, although one of the dermal fillers, HA-3,
contains lidocaine as part of the formulation. After
the injection, participants were instructed not to

1146

DERMATOLOGIC SURGERY

Evaluation
Safety
Frequency and severity of adverse events (AEs; e.g.,
bruising, itching, firmness, redness, swelling, pain,
discoloration, numbness, lumps and bumps) during
and after the implantation were evaluated, as criteria
for the tolerability of all three HA fillers. The
injection sites were examined at every visit to check
for erythema, swelling, papules, acneiform pustules,
telangiectasis, and nodules.
NLF Severity
Visual assessment of NLF severity on a five-point
scale (with specimen photographs illustrating scores
of 15) is a recognized method of standardizing the
raters judgment. For Arm A and Arm B, the
investigator used the Merz 5-point scale of fold
severity to measure baseline values and changes at
4 weeks and 6, 9, and 12 months.10 Arm A also
relied on the PRIMOS system to measure evenness
of depth of injected HAs. (Investigators did not
have access to the PRIMOS during the Arm B
portion of the study.) In addition to depth, effectiveness of NLF improvement for Arms A and B
was defined as improvement of at least 1 point
from baseline to 1-month follow-up visit on the
Merz Scale, as assessed by a subinvestigator who
was blinded to which HA was used in which NLF
for all participants. The blinded rating for Arms A
and B was performed using standard photographs
taken by the injecting dermatologist at V1, V4, V5,
V6, and V7 (FotoFinder, Teachscreen Software,
Bad Birnbach, Germany). The blinded rater was
not aware of the scores of the initial baseline
rating. Participants were not told which product
was being injected into which NLF, although the
product injected into each NLF was later identified
4 weeks after treatment following assessment of
patient satisfaction.

PRAGER ET AL

Volume Dosages and Participant Satisfaction

NLF Severity

Investigators kept records of the amount of HA

injected into each NLF in both arms of the study.
Participants were asked at 4 weeks to indicate
preference for one HA over the other HA in both arms
of the study. The investigators tabulated the results.

The NLFs were treated on both sides with different

products: HA-1 and HA-2 in Arm A and HA-1 and
HA-3 in Arm B. The NLFs were evaluated before
treatment and 1, 6, 9, and 12 months after treatment.
Arm A Mean NLF scores from individual participant visits were calculated for each study. The
results of the first group treated with HA-1 and HA2 (Arm A; Figure 2) were similar to those of the
second group (Arm B). Before treatment, the mean
value of severity of the NLFs was 2.3; 1 month after
treatment, the mean value had fallen to 0.8. It
increased slightly with time, to 1.0 at 6 months, 1.4
at 9 months, and 1.5 at 12 months.

Results
Safety
In general, the safety profile of all three HA fillers was
favorable. No deaths or serious AEs occurred during
the study (4-week or extended phases), and no AEs
led to the premature withdrawal of any participant.
Most of the treatment-emergent AEs were injection
site related (hematoma, erythema, and swelling).
These NLF site-specific AEs were mild to moderate in
severity, although swelling was noted in 25% of the
participants treated with HA-3 and none with HA-1.
Lidocaine was premixed in HA-3. The vasodilatory
properties of lidocaine may account in part for the
greater swelling, or the swelling may have been due to
other variables not yet identified. AEs for the two
arms are shown in Tables 1 and 2.

In addition to assessing the mean value of the NLFs,

this study arm also used the PRIMOS system to
assess mean depth and evenness of the NLFs treated
with HA-1 and those treated with HA-2. Specifically, the intra-individual depth differences between
HA-1 and HA-2 were a mean change of 109.5 lm
in HA-1 and a mean change of 71.8 lm in
HA-2 (p < .0001).6 Figure 3 shows the PRIMOS
measurements, originally presented elsewhere.6

TABLE 1. Adverse Events in Order of Frequency in Arm A (Hyaluronic Acid (HA)-1, HA-2)
HA-1 (n = 20)
Adverse Event

Mild

General disorders and administration site conditions

Application site erythema
Implant site hematoma
Implant site swelling
Implant site infection

9
7
1
1

Moderate

HA-2 (n = 20)
Severe

Mild

Moderate

Severe

9
6
3
1

TABLE 2. Adverse Events in Order of Frequency in Arm B (HA-1, HA-3)

HA-1 (n = 20)
Adverse Event

Mild

Moderate

General disorders and administration site conditions

Application site erythema
Implant site hematoma
Implant site swelling
Pain
Induration
Numbness

9
7
2

HA-3 (n = 20)
Severe

Mild

Moderate

16
8
3
5

Severe

1
1

1
1

38:7 PART II:JULY 2012

1147

COMPARATIVE SAFETY AND LONGEVITY OF THREE HAs

Figure 2. Arm A (hyaluronic acid (HA)-1 and HA-2) longevity

results, with lower scores at 4 weeks representing most
improvement and identical matching scores at 6, 9, and
12 months.

Figure 4. Arm B (hyaluronic acid (HA)-1 and HA-3) longevity

results, with lower scores at 4 weeks representing most
improvement and identical matching scores at 6, 9, and
12 months.

from 0.88 to 1.86 mL. Mean volumes were 1.41 mL

for HA-1 and 1.37 mL for HA-2.6
Arm B (HA-1/HA-3) The volumes of HA-1 ranged
from 1.0 to 2.0 mL; volumes of HA-3 ranged from
0.9 to 2.1 mL. Mean volumes were 1.42 mL for
HA-1 and 1.43 mL for HA-3. In Group 1, the right
NLF was injected with HA-1, followed by injection
of HA-3 in the left NLF. In Group 2, injection in the
right NLF was with HA-3, followed by injection of
HA-1 in the left NLF.
Figure 3. Sample of depth changes in hyaluronic acid (HA)1 and HA-2 after 4 weeks (Arm A) in one participant. Fifty
profile lines with fixed relative position, oriented across the
wrinkles, were taken from each three-dimensional (3-D)
profile, and the system calculated the total waviness
parameter Wt for each line. The mean total waviness of all
50 lines from a 3-D profile (mean Wt, [lm]) was the variable
for statistical analysis. The nasolabial folds (NLFs) treated
with HA-1 showed greater evenness and less depth after
treatment, whereas the NLFs treated with HA-2 roughly
paralleled the original depth of the skin. This is especially
visible from depths of 0.40 mm to 1.0 mm for the HA-1.

Arm B The results of the group treated with HA-1

and HA-3 are shown in Figure 4. Before injection,
mean NLF score was 2.3. After treatment, that
decreased to 0.9 and then to 1.2 at 6 months, 1.5 at
9 months, and 1.6 at 12 months. Because of the
high conformity between the two groups, statistical
analysis was not necessary.

1148

Participant Satisfaction
In Arm A, 11 of 20 participants indicated that HA-1
and HA-2 were both acceptable, five preferred HA-1,
two preferred HA-2, and two did not state a preference. In Arm B, five of 20 participants recommended
HA-1 and HA-3, eight recommended HA-1, and
seven recommended HA-3. Statistical significance
was not achieved, perhaps because of the small
sample size.
Representative pre- and post-treatment photographs
of participants from Arm A and Arm B are shown in
Figures 5 and 6, respectively.

Volume Dosages

Conclusion

Arm A (HA-1/HA-2) The volumes of HA-1 ranged

from 0.94 to 1.88 mL; volumes of HA-2 ranged

This prospective, randomized, single-center Phase IV

clinical study, with two arms involving 40

DERMATOLOGIC SURGERY

PRAGER ET AL

Figure 5. Results of nasolabial fold (NLF) correction in a 50-year-old woman. Participant received 2.0 mL of hyaluronic acid
(HA)-1 in the left NLF and 2.0 mL of HA-2 in the right NLF. Photographs shown are at baseline and 1, 6, and 12 months. Merz
severity scores were 4 at pretreatment, 2 at 1 month, 2 at 6 months, and 2 at 12 months.

participants with bilateral, symmetrical NLFs and a

wish for correction, investigated efficacy and safety
parameters. AEs were essentially benign and comparable between all three HAs. As expected in HA
applications, aesthetic results at 4 weeks yielded
better scores than results at later time points.
Although 16 of the 40 participants expressed no
preference for one HA over another, when
preference was expressed, participants in both arms
preferred HA-1.
All three products were injected the same way, into
the mid to deep dermis. PRIMOS measurements
showed considerably more evenness in the surface
area in the HA-1-treated NLF than in the HA-2treated NLF. It is likely that the fact that HA-1 is
nonparticulated and monophasic, with a cohesive
polydensified matrix, with different density zones
present in the formulation, and that HA-2 is particulated, with free HA that is needed for lubrication can
explain the evenness in part. The free HA resorbs after
several days, leaving particulated filler. The different
density zones in HA-1 allow the filler to adapt to the
dermal space. High-density zones will be located in
the deeper wrinkles, whereas the low-density zones
will integrate into the smaller fine wrinkles. It is likely

that further studies with three-dimensional and histologic assessments would further elucidate the role
that the nonparticulate property plays in product
distribution in the mid to deep dermis.
Viscosity and elasticity are also considerations in the
performance of the HAs. Viscosity can affect how
the filler flows from the needle, and elasticity can
affect the products response to deformation after
injection. The lower viscosity and lower elasticity of
HA-1 than of HA-2 and HA-3 may be factors in
integration of HA-1 into the dermis and therefore a
smoother result. Recent studies published by Flynn
and colleagues demonstrate a unique and more
homogeneous distribution of the HA-1, extended
out to 114 days, than of HA-2 and HA-3 in human
dermal punch biopsy samples,7 which may explain
the more even results of NLF effacement demonstrated in the PRIMOS study data.
From an aesthetic perspective, using the Merz NLF
Scale, all three HAs resulted in significant improvement, but the PRIMOS measurements at 4 weeks
showed statistically significantly greater evenness in
the HA-1-treated NLF than in the HA-2-treated
NLF. Essentially identical volumes were used for all

38:7 PART II:JULY 2012

1149

COMPARATIVE SAFETY AND LONGEVITY OF THREE HAs

Figure 6. Results of nasolabial fold (NLF) correction in a 51-year-old woman. Participant received 1.7 mL of hyaluronic acid
(HA)-1 in the right NLF and 1.8 mL of HA-3 in the left NLF. Photographs shown are at baseline and 1, 6, and 12 months. Merz
severity scores were 3 at pretreatment, 1 at 1 month, 2 at 6 months, and 2 at 12 months.

three HAs to reach full correction, even though the

investigators had the option of using as much
material as necessary to bring participants NLFs to
full correction. This translates roughly into equivalence in cosmetic correction for HA-1, HA-2, and to
HA-3 and to equivalence in longevity for all HAs at
all time points throughout the study.

tolerability of Restylane versus Zyplast for the correction of

nasolabial folds. Dermatol Surg 2003;29:58895.
5. Kablik J, Monheit GD, Yu L, Chang G. Comparative physical
properties of hyaluronic acid dermal fillers. Dermatol Surg
2009;35(Suppl 1):30212.
6. Prager W, Steinkraus V. A prospective, rater-blind, randomized
comparison of the effectiveness and tolerability of Belotero Basic
versus Restylane for correction of nasolabial folds. Eur
J Dermatol 2010;20(6):15.
7. Flynn TC, Sarazin D, Bezzola A, Terrani C, et al. Comparative
histology of intradermal implantation of mono and biphasic
hyaluronic acid fillers. Dermatol Surg 2011;37:63743.

References
1. Lemperle G. Hyaluronsaure zur Faltenunterspritzung. Hylaform
(USA), Restylane (S), Rofilan (NL). Arzneimittel Ther Kritik
2000;3:6357.
2. Narins RS, Coleman W, Donofrio L, Jones D, et al. Nonanimal
sourced hyaluronic acid-based dermal filler using a cohesive
polydensified matrix technology is superior to bovine collagen in
the correction of moderate to severe nasolabial folds: results from
a 6-month, randomized, blinded, controlled multicenter study.
Dermatol Surg 2010;36:73040.
3. Pinksy MA, Thomas JA, Murphy DK, Walker PS. Juve derm
injectable gel: a multicenter, randomized, double-blind
study of safety and effectiveness. Aesthet Surg J 2008;28:
1723.
4. Narins RS, Brandt F, Leyden J, Lorene P, et al. A randomized,
double-blind multicenter comparison of the efficacy and

1150

DERMATOLOGIC SURGERY

8. Levy PM, De Boulle K, Raspaldo H. A split-face comparison of a

new hyaluronic acid facial filler containing pre-incorporated
lidocaine versus a standard hyaluronic acid facial filler in the
treatment of nasolabial folds. J Cosmet and Laser Ther 2009;11
(3):16973.
9. Dirting K, Lampe H, Wolters M, Prager W, et al. Hyaluronic
acid filler for correction of nasolabial groovesresults of a
clinical study. J Dtsch Dermatol Ges 2008;6(Suppl 2):1014.
10. Merz Aesthetics. Frankfurt am Main, Germany. 2009.

Address correspondence and reprint requests to: Welf

Prager, MD, Dermatologikum Hamburg, Stephansplatz 5,
D-20354 Hamburg, Germany, or e-mail:
prager@dermatologikum.de