ORIGINAL ARTICLES
`
`Comparative Physical Properties of Hyaluronic Acid Dermal
`Fillers
`yz
`JEFFREY KABLIK, GARY D. MONHEIT, MD,
`
`AND JULIA GERSHKOVICH
`
`LIPING YU, PHD, GRACE CHANG,
`
`BACKGROUND Hyaluronic acid (HA) fillers are becoming the material of choice for use in cosmetic soft
`tissue and dermal correction. HA fillers appear to be similar, but their physical characteristics can be
`quite different. These differences have the potential to affect the ability of the physician to provide the
`patient with a natural and enduring result.
`
`OBJECTIVE The objective of this article is to discuss the key physical properties and methods used in
`characterizing dermal fillers. These methods were then used to analyze several well-known commer-
`cially available fillers.
`
`METHODS AND MATERIALS Analytical methods were employed to generate data on the properties
`of various fillers. The measured physical properties were concentration, gel-to-fluid ratio, HA gel
`concentration, degree of HA modification, percentage of cross-linking, swelling, modulus, and particle size.
`
`RESULTS The results demonstrated that commercial fillers exhibit a wide variety of properties.
`
`CONCLUSION Combining the objective factors that influence filler performance with clinical experience will
`provide the patient with the optimal product for achieving the best cosmetic result. A careful review of these gel
`characteristics is essential in determining filler selection, performance, and patient expectations.
`
`Jeffrey Kablik, LiPing Yu, Grace Chang, and Julia Gershkovich are employees of Genzyme Corporation. The
`materials used in this study were provided by Genzyme.
`
`I n recent years, hyaluronic acid (HA)-based fillers
`
`have become the material of choice for use in soft
`tissue and dermal correction, for the most part
`replacing collagen fillers such as Zyderm, Zyplast,
`Cosmoderm, and Cosmoplast (Allergan, Irvine, CA).1–3
`Although the HA fillers appear to be similar, their
`physical characteristics and methods of manufacture
`are not the same.2 These differences have clinical
`ramifications for the physician in that they can affect
`injection technique, usage, and the quality of the
`outcome. Often fillers are pragmatically evaluated,
`with consideration given to the results of the appli-
`cation. Questions such as whether the material is
`easy to deliver; whether the duration of correction is
`appropriate; whether the material bruises, swells,
`and creates inflammation; and whether the results
`
`look natural are frequently the only means of charac-
`terizing a filler.
`
`There is no universal filler that is appropriate for
`every application or for every patient. Understanding
`physical properties of HA fillers and how they in-
`teract provides significant information about the
`expected clinical outcome and the corresponding
`best cosmetic result for a patient. Therefore, it is
`important to take an objective approach in assessing
`factors that may influence HA filler performance,
`such as total HA concentration, modulus, swelling,
`particle size, cross-linking, and extrusion force.
`
`Scientists and engineers use a variety of methods to
`design materials that have the desired final proper-
`
` Genzyme Corporation, Cambridge, Massachusetts; yTotal Skin and Beauty Dermatology Center, P.C., Birmingham,
`Alabama; zDepartment of Dermatology, University of Alabama at Birmingham, Birmingham, Alabama
`& 2009 by the American Society for Dermatologic Surgery, Inc.  Published by Wiley Periodicals, Inc. 
`ISSN: 1076-0512  Dermatol Surg 2009;35:302–312  DOI: 10.1111/j.1524-4725.2008.01046.x
`
`3 0 2
`
`Page 1
`
`

`

`K A B L I K E T A L
`
`ties. Consequently, the filler designers make use of
`characteristics such as raw material properties,
`cross-linking schemes, HA concentration, and
`rheological properties to achieve the end results.
`Although the results may vary, manufacturers take
`similar approaches to the design of their fillers. Un-
`derstanding the means employed by manufacturers
`to design and characterize their fillers should provide
`useful insight as to the ability to clinically provide
`the patient an enduring, natural-looking result.
`
`Recent review articles describe important physical
`characteristics of HA-based fillers.2,4 In this review,
`we discuss the key physical properties and methods
`used to design and characterize dermal fillers. We
`then employ these methods to analyze several well-
`known commercially available fillers.
`
`HA Dermal Filler Properties
`
`Hyaluronic Acid
`
`HA is a glycosaminoglycan disaccharide composed
`of alternately repeating units of D-glucuronic acid
`and N-acetyl-D-glucosamine (Figure 1). At physio-
`logic pH, HA exists mostly as a sodium salt; this is
`the most common form of commercially available
`HA. HA is naturally occurring in the extracellular
`matrix found in many human tissues, including skin,
`synovial fluid of joints, vitreous fluid of the eye, and
`scaffolding within cartilage.5,6 The average 70-kg
`man has roughly 15 g of hyaluronan in his body,
`one-third of which is turned over (degraded and
`
`OH
`
`O
`
`O
`
`H
`
`OH
`
`O
`
`O
`HO
`
`ONa
`
`O
`
`CH3
`
`NH
`
`O
`
`CH2OH
`
`OH
`
`N
`D-Glucuronic Acid N-Acetyl-D-Glucosamine
`
`is a glycosaminoglycan
`Figure 1. Hyaluronic acid (HA)
`disaccharide composed of repeating units of D-glucuronic
`acid and N-acetyl-D-glucosamine. The molecular weight of
`HA is proportional
`to the number of
`these repeating
`disaccharides.
`
`synthesized) every day.5,7 The largest amount of HA
`resides in skin tissue (7–8 g per average human
`adult); thus approximately 50% of the total HA in
`the body is found in the skin.5,7 HA is a polyanionic
`polymer at physiologic pH and is therefore highly
`charged. The highly charged nature of HA renders it
`soluble and allows it to bind water extensively.
`
`Molecular Weight
`
`The molecular weight of HA is proportional to the
`number of repeating disaccharides in the HA mol-
`ecule (Figure 1). When discussing the molecular
`weight (MW) of HA, it is most often the average
`MW of a sample that is reported. As a result, the
`polydispersity or range of molecular weights found
`in a sample is also a consideration. The HA used in
`manufacturing dermal fillers can range from 500 to
`6,000 kDa. Commercial preparations of hyaluronan
`are usually supplied as the sodium salt and have a
`disaccharide MW of approximately 401 Da. There-
`fore, a 1,000,000-MW polymer of HA will have ap-
`proximately 2,500 repeating disaccharide units, all of
`which are negatively charged at physiologic pH.
`
`Sometimes the term ‘‘MW’’ is applied generally to
`properties of dermal fillers. This is technically
`incorrect, because a typical filler comprises HA
`molecules cross-linked to form a gel. As a result, the
`MW of a HA gel is enormous and is essentially
`immeasurable. Because the MW of the final HA
`gel is so large, small differences in MW of the start-
`ing HA have little effect on the final properties of the
`gel. Although we cannot effectively speak of the MW
`of a gel, the number of cross-links and the percentage
`of modification are important considerations when
`characterizing HA gels.
`
`Modification and Crosslinking
`
`In its natural state, HA exhibits poor biomechanical
`properties as a dermal filler. HA has excellent bio-
`compatibility and affinity for water molecules, but it
`is a soluble polymer that is cleared rapidly when
`injected into normal skin (Figure 2A).5,7 Therefore,
`to provide the ability to lift and fill wrinkles in the
`
`3 5 : S 1 : F E B R U A RY 2 0 0 9
`
`3 0 3
`
`Page 2
`
`

`

`P R O P E R T I E S O F H A D E R M A L F I L L E R S
`
`A
`
`B
`
`C
`
`Uncrosslinked HA = Liquid
`
`Crosslinked HA = Gel
`
`Modified HA with Pendant Crosslinker
`
`Figure 2. When dissolved in water, hyaluronic acid (HA) be-
`haves as a fluid, with excellent biocompatibility but poor
`mechanical properties (A). Modification of HA molecules by
`cross-linking improves mechanical properties by creating
`gels that have a firmer structure and are able to resist de-
`gradation (B). Modification does not necessarily cross-link
`HA to other HA molecules, resulting in a pendant cross-
`linker (C). Such structures often result in softer gels.
`
`skin, chemical modification is required to improve
`its mechanical properties (Figure 2B) and residence
`time at the implant site. The two most common
`functional groups that can be modified in HA are the
`carboxylic acid and the hydroxyl (alcohol). Cross-
`linking strategies attempt to improve biomechanical
`properties while maintaining biocompatibility and
`biological activity. The literature reports many
`methods for cross-linking HA.2,8 Biomaterials have
`been produced through modification to the carboxyl
`acid group by esterification and through the use of
`cross-linkers such as dialdehydes and disulfides.8
`The most commonly employed cross-linkers for
`dermal fillers are divinyl sulfone (Hylaform, Cap-
`tique, and Prevelle Genzyme Co., Cambridge, MA)
`and diglycidyl ethers (Restylane, Q-Med, Uppsala,
`Sweden; Juvederm, Allergan, Irvine, CA; and Belo-
`tero, Anteis SA, Geneva, Switzerland) or bis-epox-
`ides (Puragen, Mentor, Santa Barbara, CA).1,2,4,9
`
`An assessment of the degree of modification must go
`beyond determining the amount of cross-links in a
`
`material. Bifunctional cross-linkers do not necessar-
`ily react at both ends to connect two different
`strands of HA. Often the cross-linker will bond only
`at one end, leaving the other end pendant (Figure
`2C). Thus the total degree of modification can be
`defined as;
`
`Total % Degree of Modification
`¼ % Crosslink þ % Pendant
`
`Whether chemical modification results in formation
`of a cross-link (a bond between two strands of HA)
`or a pendant group is a function of the reaction
`conditions used by different manufacturers of HA
`fillers.
`
`The degree of modification can have a significant
`effect on the properties of a filler material. As the
`cross-link density of a gel increases, the distance
`between the cross-linked segments becomes shorter.
`When a load is applied, these shorter segments re-
`quire a greater force to deflect. Thus, increasing
`cross-link density strengthens the overall network,
`thereby increasing the hardness or stiffness of the gel.
`However, when the gel comprises all or mostly
`pendant HA modification, a low cross-link-density
`network is formed, resulting in softer gels.
`
`In general in vivo degradation of HA occurs through
`enzymatic degradation and reaction with reactive
`oxygen species (e.g., superoxide, peroxynitrite). In
`each case, HA molecular strands are cleaved to
`smaller oligosaccharides that are more amenable to
`metabolism and clearance from the body. Thus, a
`network of cross-linked HA retains its structure until
`sufficient degradation has occurred at the gel surface
`to form soluble oligosaccharides that can be metab-
`olized and cleared from the body.5,7 This simplistic
`approach provides a general overview of the degra-
`dation of HA, although specific cross-linking re-
`agents and conditions used in the cross-linking
`process can affect the degradation rate of cross-
`linked HA hydrogels. Also, other physical properties
`such as gel concentration and degree of swelling can
`affect the rate of degradation.
`
`3 0 4
`
`D E R M AT O L O G I C S U R G E RY
`
`Page 3
`
`

`

`A
`
`B
`
`Figure 3. Concentration is a measure of the amount of
`hyaluronic acid (HA) in a gel. Given the same degree of
`cross-linking, low concentrations will result in softer gels
`(A), whereas higher concentration gels result in stiffer gels
`(B). It also stands to reason that, because there is more
`cross-linked HA in higher-concentration gels that it should
`last longer.
`
`Concentration
`
`When manufacturers convey the concentration of a
`filler, they are articulating the total amount of HA
`found in the filler, typically expressed in mg/mL
`(Figure 3). The total HA concentration consists of
`insoluble HA gel and soluble-free HA. Manufactur-
`ers may provide free HA as a soluble fluid compo-
`nent to the gel to facilitate the extrusion of the filler
`through fine-bore needles. Although not all manu-
`facturers add HA fluid to their fillers, a fluid com-
`ponent is often present. This fluid component
`contains unmodified and modified soluble HA that is
`
`K A B L I K E T A L
`
`generated during the manufacturing process when
`HA fragments are formed as a side-product of the
`chemical modification. These soluble fluids are easily
`metabolized and do not contribute to the extended
`duration and effectiveness of the product. Only the
`cross-linked HA resists enzymatic and radical de-
`gradation and therefore extends the filler’s presence
`in the dermis, contributing to its effectiveness. Con-
`sequently, it is important to understand how much of
`the filler’s HA concentration is gel or cross-linked
`HA and how much is soluble fluid or free HA.
`
`Modulus
`
`Most HA-based dermal fillers are viscoelastic, con-
`taining elastic (solid) and viscous (liquid) compo-
`nents that can be evaluated using dynamic testing.
`The rheological characteristic that describes this
`property is the complex modulus (G ), which defines
` can
`the material’s total resistance to deformation. G
`also be defined as sum of the elastic modulus (G0)
`and the viscous modulus (G00). Elastic modulus is
`also called storage modulus because it describes the
`storage of energy from the motion in the structure.
`The magnitude of the G0 is dependent upon the
`elastic interaction and the strength of the interaction
`in the sample. Viscous modulus is also labelled loss
`modulus, and it describes the energy that is lost as
`viscous dissipation. Thus the value of G00 is a mea-
`sure of the flow properties for a structured sample.
`
`The elastic modulus G0 is most often used to char-
`acterize the firmness of a gel. Because the elastic
`modulus or G0 of a material describes the interaction
`between elasticity and strength, it provides a quan-
`titative method for characterizing the hardness or
`softness of a gel. G0 represents the amount of stress
`required to produce a given amount of deformation.
`
`G0 ¼ stress
`strain
`
`Another way of thinking of this is that elastic mod-
`ulus is a measure of a material0s ability to resist de-
`formation. As an example, a stiffer material will
`have a higher modulus; it will take a greater force to
`
`3 5 : S 1 : F E B R U A RY 2 0 0 9
`
`3 0 5
`
`Page 4
`
`

`

`P R O P E R T I E S O F H A D E R M A L F I L L E R S
`
`deform the material a given distance. For most ma-
`terials, G0 is dependent upon the speed (frequency) at
`which the force is applied. Intuitively, this makes
`sense; for instance, a material will resist deformation
`if the load is applied at a rapid rate, resulting in a
`higher modulus than if the load were applied at a
`slower rate. Thus it is important to ensure that there
`is parity in methods of measurement when compar-
`ing modulus values for different materials.
`
`The degree of cross-linking and gel concentration play
`important roles in defining the modulus of the gel, and
`many manufacturers use these parameters to influence
`the hardness or softness of their fillers. Higher gel
`concentration produces more molecular entanglements
`and in so doing increases the modulus of the gel. A gel
`with a lower degree of cross-linking but higher gel
`concentration could have a similar modulus as a lower
`concentration gel with a much higher degree of cross-
`linking. A gel with a lower number of cross-links (co-
`valent bonds) has a greater length of the HA molecule
`between links, thus requiring less of a force to deform
`the gel (Figure 4A). As the network is tightened by
`increasing the number of cross-links, the gel will be-
`come stiffer (Figure 4B). HA gels with pendant-type
`modification have a small effect on modulus because
`they do not form a cross-linked network (Figure 2).
`
`Gels with higher G0 (higher stiffness) have a better
`ability to resist dynamic forces occurring during
`
`A
`
`B
`
`Lightly cross-linked Gel = soft
`
`Highly cross-linked Gel = firm
`
`facial muscle movement and thus may provide better
`support and lift and longer duration of correction in
`areas such as nasolabial folds and marionette lines.
`Gels with low G0 are probably better suited to areas
`with static and superficial wrinkles, where resistance
`to deformation is not critical, or areas where
`anatomy does not require stiffness but volume and
`softness are important, such as in lips. Although all
`HA gels vary in elastic modulus, even the ones with
`the highest G0 are much softer than the elastic
`modulus of human dermis, which has G0 in the
`3-MPa range.10
`
`Swelling
`
`HA at physiological pH is hydrated extensively by
`water. The three-dimensional structure of HA has a
`significant influence on the water-binding capabili-
`ties of HA. In solution, the coil-like structure of a
`HA molecule occupies a large domain in comparison
`with its molecular weight. When in a physiologically
`neutral solution, water forms hydrogen bonds with
`the N-acetyl and carboxyl groups. The dipole at-
`traction of the hydrogen bond with carboxyl group
`results in HA’s affinity for retaining water. With re-
`peating disaccharide units, the longer the HA mol-
`ecule, the more water molecules are bound per unit
`of polymer.
`
`A filler’s predisposition for swelling is a function of
`whether the HA filler has reached its equilibrium for
`bound water. A HA gel’s capacity for swelling will
`vary from product to product and is dependent upon
`concentration, cross-link density, and the process
`used to hydrate the gel. Fully hydrated or equilib-
`rium gels have already reached their hydration ca-
`pacity; thus they will not swell when injected into the
`dermis. Nonequilibrium gels tend to swell postin-
`jection, and consideration must be given to under-
`filling when performing a correction with these gels.
`
`Particle Size and Extrusion Force
`
`Figure 4. Gels with fewer cross-links have a greater length
`between links, requiring less of a force to deform the gel (A).
`Increasing the number of cross-links shortens the distance
`between cross-links, resulting in a stiffer gel (B).
`
`The cross-linked gels that constitute dermal fillers
`must be of sufficient particle size that they can be
`injected easily through an appropriately sized needle.
`
`3 0 6
`
`D E R M AT O L O G I C S U R G E RY
`
`Page 5
`
`

`

`K A B L I K E T A L
`
`In efforts to reduce undesired side effects such as
`pain, bruising, bleeding, and edema, small-bore
`needles (27-g and 30-g) are employed. Thus, the gel
`particles must be appropriately sized to be able to
`pass through these fine-bore needles with an ac-
`ceptable extrusion force.
`
`The HA filler manufacturers employ various meth-
`ods of particulating the gels based on their modulus
`to obtain an appropriate extrusion force. This results
`in gels that have various particles sizes and broad or
`narrow ranges of distribution. The ultimate goal is to
`size the HA gel particles and define their modulus so
`that the final gel can be easily administered to the
`site of application.
`
`When characterizing the particle sizes of a HA gel,
`consideration must be given to the average particle
`size, as well as the particle size distribution. Because
`larger gel particles are more difficult to push through
`a small-bore needle, a filler with a high average
`particle size will be more difficult to extrude. The
`average extrusion force of the filler can be decreased
`by reducing the average particle size, but if the dis-
`tribution of particles still includes a number of larger
`particles, there is the potential that they may cause
`interrupted or sporadic flow of the product through
`the needle.
`
`Gel hardness or G0 plays an important role in how
`the gels must be sized for easy delivery through fine-
`bore needles. Firm gels, with a high ability to resist
`deformation, must be sized to small particles and
`should have a narrow distribution range to be easily
`injected through a thin-bore needle. On the other
`hand, soft gels with low G0 can have a broader dis-
`tribution of particle sizes because the softer particles
`can be easily deformed to pass through the needle.
`Regardless of whether a gel is firm or soft, particle
`size uniformity is preferred to avoid ‘‘stop and go’’
`action during injections and for better control of gel
`placement.
`
`force of a filler. Rheological properties such as
`modulus of the filler have an effect. The degree of
`modification, the amount of cross-linked and un-
`cross-linked HA, concentration, and the degree of
`hydration affect these rheological properties. Thus,
`extrusion force is the result of a combination of
`properties that are integral to the design of the filler.
`
`Methods
`
`Percentage Modification Measurement
`
`HA filler samples were degraded using Strep-
`tomyces-derived hyaluronidase (VWR Scientific,
`Bridgeport, NJ) for 72 hours at pH 5.0 (acetate
`buffer) and 371C. This species of hyaluronidase
`depolymerizes HA using a unique mechanism that
`introduces a double bond into the resulting
`oligosaccharide.11 Exhaustive digestion of unmodi-
`fied hyaluronate results in a mixture of tetra- and
`hexasaccharides.11 These resulting oligosaccharides
`were analyzed using high-performance liquid chro-
`matography (HPLC).12,13 When chemically modified
`or cross-linked HA is subjected to this enzymatic
`digestion and analysis, one observes higher-MW
`oligosaccharides that reflect the chemical modifica-
`tion of the gel.
`
`Conditions of HPLC analysis:
`
`Column: Anionic exchange (4  250 mm, CarboPac PA
`100, Dionex Corporation, Sunnyvale, CA)
`
`Mobile Phase-
`
`A: water
`B: 0.4 M sodium phosphate,
`pH 5.8
`
`Flow rate: 0.8 mL/min
`Gradient: step linear
`
`Gradient Table:
`
`Time (min)
`
`0
`5
`55
`57
`
`% A
`
`90
`90
`20
`90
`
`% B
`
`10
`10
`80
`10
`
`UV detectionF232 nm
`
`As can be surmised from the previous discussions, it
`is not particle size alone that affects the extrusion
`
`Injection Volume: 50 mL/each injection
`
`3 5 : S 1 : F E B R U A RY 2 0 0 9
`
`3 0 7
`
`Page 6
`
`

`

`P R O P E R T I E S O F H A D E R M A L F I L L E R S
`
`The method separates the digest fragments based on
`the overall anionic charge of the oligosaccharides. A
`HPLC analysis of an incomplete hyaluronidase
`digest was used to generate an elution profile based
`on oligosaccharide size. Because the hyaluronidase
`could not digest the cross-linker, after complete di-
`gestion of the modified HA, the detection of those
`peaks that elute at retention times greater than or
`equal to those of the octasaccharides are the result of
`covalent cross-linking of these particular
`oligosaccharides. Therefore, percentage of cross-
`linking is determined to be the sum total for all of
`these late-eluting peaks. Because the integrated peak
`area is proportional to the concentration of each
`fragment, the relative percentage of cross-linked or
`pendant modification was determined.
`
`Rheology Measurements
`
`Rheological characterization was performed using
`an automated Controlled Stress Rheometer (Malvern
`Instruments LTD, Worcestershire, UK), using a par-
`allel-plate, cone-and-plate, or cylinder-and-cup
`measuring system at 251C. The elastic (G0) and vis-
`cous (G00) moduli and phase angle (1) were deter-
`mined using a frequency sweep test. The experiments
`were performed within the range of the linear visco-
`elastic region. The phase difference between the
`stress and strain in an oscillatory deformation is
`measured as a phase angle that is equal to tan-1 (G00/
`G0). The G0 measured at frequency 5 Hz for these
`gels were compared in this study.
`
`Swelling–Dilution Durability Assay
`
`In the dilution study, test samples were diluted with
`various volume ratios of sample to phosphate
`buffered saline (PBS) ratio ranging from 1:0.33
`(33% dilution) to 1:4 (400% dilution). For each
`sample lot, three to four different dilutions were
`made. The diluted gels or solutions were mixed and
`then tested for rheological properties. The phase
`angle of the sample at different dilutions was deter-
`mined on a Bohlin CVO-50 rheometer (Malvern
`Instruments LTD) using an oscillation test at a fre-
`quency of 1 Hz. The percentage change in phase
`
`angle for each sample was calculated and plotted
`against the percentage dilution. The percentage di-
`lution at which the phase angle increased to 50% of
`its original value was defined as the dilution dura-
`bility. The dilution durability can be interpreted as
`the maximum swelling of the gel before phase sep-
`aration.
`
`Concentration Measurement According to
`Hexuronic Acid Assay
`
`All samples were diluted with 2N sulfuric acid and
`then heated in an oven for 1 hour at approximately
`951C. After being cooled to ambient temperature,
`the samples were diluted with deionized water to a
`final concentration of approximately 10 to 75 mg/mL
`hexuronic acid.
`
`A Bran Luebbe Flow Injection Autoanalyzer 3 Sys-
`tem (SEAL Analytical Inc., Mequon, WI) was used to
`measure the total concentration of hexuronic acid as
`glucuronic acid. The sulfuric acid–treated HA sam-
`ples and various concentrations (10, 25, 50, and
`75 mg/ml) of glucuronic acid standards were injected
`in sequence through the autoanalyzer. In the auto-
`analyzer, each sample is first mixed with sulfuric
`acid–borate and heated at 951C and then mixed with
`0.1% carbazole–ethanol and heated at 951C again.
`At the end, a pink color forms that is quantified by
`measuring the absorbance at 530 nm. The HA con-
`centration of each injected sample was calculated by
`comparison with authentic standards of glucuronic
`acid.
`
`Gel-to-Fluid Ratio Using Size-Exclusion
`Chromatograph with the Multi-Angle Laser
`Light Scattering Measurement
`
`Size Exclusion Chromatograph (SEC) with the
`Multi-Angle Laser Light Scattering (MALLS, Wyatt
`Technology Corporation, Santa Barbara, CA) and
`refractive index detection can provide direct MW
`and concentration measurement of soluble polymer
`in the sample. Dermal filler products were diluted
`with PBS, thoroughly agitated, and then centrifuged
`to separate the gel phase from the supernatant. The
`
`3 0 8
`
`D E R M AT O L O G I C S U R G E RY
`
`Page 7
`
`

`

`K A B L I K E T A L
`
`supernatant, which corresponds to the fluid portion
`of each sample, was filtered through a 0.45-mm filter
`and then injected into the SEC/MALLS system to
`determine MW and HA concentration. The gel-to-
`fluid ratio could be calculated using the following
`equation:
`Gel=Fluid Ratio ¼
`f½Total HA ConcŠ ½Soluble HA ConcŠg=
`½Soluble HA ConcŠ
`
`sult, companies provided line extensions that pur-
`ported to extend product duration by increasing the
`HA concentration or particle size. Examples of
`products with greater HA concentration include
`Juvederm 18 and Juvederm 24 by Allergan (formerly
`Corneal) and Belotero Soft/Belotero Basic by Merz,
`whereas Hylaform/Hylaform Plus by Genzyme and
`Restylane/Perlane by Q-Med are examples of differ-
`entiation by particle size.
`
`Particle Size
`
`Particle size and distribution measurements were
`performed on a Malvern Master Sizer Longbed-S
`particle analyzer (Malvern Instruments LTD). Test
`samples were placed in a saline suspension in the
`particle analyzer and scanned for mean particle size
`and distribution.
`
`Results
`
`A summary of the properties of various fillers is
`available in Table 1
`
`Discussion
`
`HA Filler Performance
`
`For many years, scientists and physicians have de-
`bated which parameter has the most influence on
`HA filler duration. In the past, HA concentration
`and gel particle size were thought to be the most
`important differentiating parameters.1,3,4,14 As a re-
`
`TABLE 1. Properties of Fillers in the Study
`
`The clinical evaluation of Hylaform and Hylaform
`Plus as well as Restylane and Perlane, the HA gels
`with the same chemical formulation but different
`particles sizes, demonstrated that larger particle size
`does not extend duration of those formulations.15–20
`One explanation for these results is that the particle
`sizes are not sufficiently different (B700 m for
`Hylaform Plus and Perlane, vs 500 and 300 m for
`Hylaform and Restylane, respectively) to translate
`into a discernible clinical effect. Therefore, large-
`particle-size fillers could be beneficial for filling
`deeper wrinkles, although one should not expect a
`longer duration than with a small particle filler of
`the same composition.
`
`HA concentration is a principal parameter in influ-
`encing product duration, although as previously
`discussed, it is not the total HA concentration that
`affects duration, but rather the amount of cross-
`linked HA gel that plays an important role in filler
`performance. Unmodified HA is completely metab-
`olized a few days after injection.5 Table 1 lists values
`for the free HA concentration and cross-linked HA
`
`Hylaform Hylaform Plus Prevelle Restylane Perlane Juvederm 30 HV
`
`Total HA concentration (mg/mL)
`5.5
`Gel-to-fluid ratio
`98:2
`HA gel concentration (mg/mL)
`5.4
`Degree of HA modification (%)
`23
`Percentage cross-linked HA
`12
`Dilution durability/percentage swelling o25
`G0 modulus (Pa)
`140–220
`Average particle size (mm)
`500
`
`5.5
`98:2
`5.4
`23
`12
`o25
`140–220
`700
`
`20
`5.5
`75:25
`98:2
`15.0
`5.4
`3
`23
`1.2
`12
`o25
`50
`230–260 660
`350
`300
`
`20
`75:25
`15.0
`3
`1.4
`50
`588
`650
`
`24
`60:40
`14.4
`10
`2
`300
`105
`300
`
`HA = hyaluronic acid.
`
`3 5 : S 1 : F E B R U A RY 2 0 0 9
`
`3 0 9
`
`Page 8
`
`

`

`P R O P E R T I E S O F H A D E R M A L F I L L E R S
`
`gel concentration for commercially available dermal
`fillers. Hylaform/Prevelle has 98% or 5.4 mg/mL of
`cross-linked HA gel, whereas Restylane/Perlane has
`75% or 15.0 mg/mL, and Juvederm 30 HV has only
`60% or 14.4 mg/mL of cross-linked HA gel compo-
`nent contributing to their duration.
`
`Another important characteristic that affects clinical
`performance is the degree of cross-linking that was
`introduced earlier. Quite frequently, the degree of
`cross-linking is used interchangeably with the degree
`of total modification when describing HA dermal
`fillers. We need to remember that total modification
`includes the percentage of cross-link plus the per-
`centage of pendant. The cross-link ratio can be de-
`fined as the ratio of percentage of cross-linking to the
`percentage of total modification and can be used as a
`way of characterizing a particular gel. For example,
`the ratio of cross-linked HA to modified HA is ap-
`proximately 50% for Hylaform/Prevelle, 40% for
`Restylane, and as low as 20% for Juvederm 30 HV,
`as described in Table 1. This ratio is dependent on
`reaction conditions used to produce these products.
`The HA modified with predominantly pendant
`groups forms gels that are held together by physical
`entanglement due to interchain hydrogen bonding.
`These gels are not as strong as the ones produced by
`creating a covalently cross-linked network. There-
`fore, when comparing HA gels with the same con-
`centration and total degree of modification, gels with
`a high cross-link-to-pendant ratio should provide
`better resistance to degradation and deformation and
`thus should maintain longer duration of effect than
`those with predominantly pendant groups.
`
`To further understand performance of the HA fillers
`in the clinical setting, it may be useful to combine the
`HA gel concentration and degree of cross-linking
`together. Table 1 shows that Restylane/Perlane and
`Juvederm 30 HV have similar HA gel concentration
`(15.0 and 14.4 mg/mL, respectively) and percentage
`cross-linking (1.3% and 2%, respectively). Evalua-
`tion of these three products in the controlled clinical
`studies showed duration of effect of 6 months in the
`majority of patients.17,18,21 Although these products
`
`were not tested side by side in the clinic, the clinical
`trial designs were similar, allowing us to postulate
`that comparable results are due to similar concen-
`tration and percentage cross-linking exhibited in the
`two products.
`
`Pendant modification is a result of the reaction
`conditions and is not specific to a bifunctional cross-
`linker. Pendant modification can change the confor-
`mation of the HA molecule, rendering it less soluble
`than unmodified HA, although this type of modifi-
`cation does not produce strong covalent bonds to
`retard the degradation and deformation of HA net-
`work and therefore is more likely to contribute to gel
`swelling than to its longevity. This could partly ex-
`plain high swelling of Juvederm 30 HV, as shown by
`our in vitro testing, because this product mostly
`contains the pendant type of HA modification (8%)
`and less cross-linked type (2%) (Table 1) and is
`supported by the clinical experience.9,21
`
`Another reason for swelling of Juvederm 30 HV is its
`nonequilibrium state that forces the formulation to
`hydrate by attracting fluids after injection.9 The
`same is true for Restylane/Perlane, although in this
`case, the nonequilibrium hydration state is most
`likely the reason for continued HA gel swelling when
`implanted.
`
`Hylaform/Hylaform Plus and Prevelle also have a
`substantial percentage of pendant-type modification
`(11%), which constitutes 50% of the total modifi-
`cation of the HA in these formulations. However,
`these products do not swell as much (Table 1), in
`part because of the high degree of cross-linked HA
`network holding the structure together and because
`of the fully hydrated state of the product.
`
`The analysis of HA