Different hyaluronan concentration determinate distinctive physicochemical and biological properties of the intra-articular gel but does not affect satisfactory clinical outcome
Article History
-
Received: ,
Accepted: ,
Abstract
Abstract
Intra-articular injection of hyaluronic acid (HA) has been established as a suitable treatment for osteoarthritic (OA) joints, able to effectively relieve the disease symptoms. However, HA-based medical devices available on the market differ in their key properties that finally might radically influence their clinical effectiveness. This study aimed to evaluate and compare the physicochemical and biological properties of two different hyaluronic acid-based gels intended for intra-articular injection into large synovial joints and subsequently determine if observed differences have reflection in clinical outcome related to their use.
Samples of Biolevox HA and Biolevox HA ONE (Biovico, Poland) provided from standard production batches were included in the analysis. The HA-based products were analyzed in order to assess their physicochemical properties, cytotoxicity, and the ability to rescue a negative impact of the pro-inflammatory-inducing agent on chondrocyte gene expression levels. Additionally, the clinical recognition of the HA products included in this study was evaluated.
HA gels tested in this study differ from each other in terms of their physicochemical characteristics, with a higher product viscosity as a result of higher molecular weight, and higher concentration of hyaluronan in Biolevox HA ONE. Tested HA gels were not able to negatively affect the cell viability-regardless gel concentration used, what excluded any cytotoxic properties. Analysis of the gene expression showed that tested HA gels were able to effectively restore the down-regulated expression rate of bCol2a1 and bAcan in chondrocytes stimulated with rBoIL-1. Finally, the use of both Biolevox HA and Biolevox HA ONE gels was evaluated by the majority of physicians as very well-tolerated and painless by the patients.
We show here that precisely tailored physicochemical properties of the HA products can provide desired biological properties of the gel and guarantee beneficial outcome related to the product.
Keywords
This article was originally published by Reed Elsevier India Pvt. Ltd. and was migrated to Scientific Scholar after the change of Publisher.
1 Introduction
Osteoarthritis (OA) is one of the most common diseases of the musculoskeletal system, as it is estimated that the number of patients with OA has already reached 300 million worldwide and it's still rising.1,2
Regrettably, today, the only fully effective therapy for an end-stage OA is a total joint replacement surgery. Alternative, non-surgical therapeutic approaches predominantly address OA symptoms, like pain and function, along with limited ability for recovery and the structural improvement of affected joint tissues. Nevertheless, none of the existing non-surgical therapy is able to fully arrest the progression of the disease.3 Therapeutic approaches, including intra-articular therapy with hyaluronic acid (HA), platelet-rich plasma (PRP), or mesenchymal stem cells (MSC) may slow down the progression of the disease.4
Procedure of intra-articular (IA) injection of the exogenous HA into the OA affected joint called viscosupplementation (VS) brings immediate mechanical effect as it contributes to the improvement of the viscosity and elasticity of the synovial fluid, thus reducing pain and increases the functionality of the joint.4 Moreover, HA upon intra-articular delivery in known to provide a potent biological and pharmacological effect as it reduces the production of proinflammatory mediators and matrix degradations enzymes, and reduces nerve impulses and nerve sensitivity associated with OA pain.5 Importantly, exogenously delivered HA protects chondrocytes against apoptosis, and it stimulates the synthesis of endogenous HA and proteoglycans, what brings a long-lasting clinical effect.6
Consequently, according to numerous of meta-analysis and professional guidelines, VS is a very well-established treatment option for OA as the effectiveness and safety of the therapy has been proven by wide range of clinical studies.7–10 Literature show a beneficial effect of the therapy on pain reduction, function improvement, and patient global assessment recovery at different time points after the injection.11
Nevertheless, the efficacy of HA products may vary regarding the different production method, molecular weight (MW), viscoelasticity, rheological properties, half-life in the joint cavity, and the number of injections in each treatment cycle. All variables may have impact on clinical effectiveness.12 As a consequence it remains unclear however, if different HA-based products should be considered as a one, uniform group of therapeutics or whether their characteristics would distinguish them clinically. Noteworthy, the leaders of HA-based products for intra-articular injection in Poland, namely Biolevox HA and Biolevox HA ONE, contain different amount of HA in the injection thus they are dedicated for multiple (Biolevox HA) or one injection therapy (Biolevox HA ONE).
In view of this, the aim of the study was to evaluate and compare the physicochemical and biological properties of two different HA-based product (Biolevox HA and Biolevox HA ONE) dedicated for intra-articular injection into the large synovial joints and determine if possible differences result in particular clinical functioning of each HA gel.
2 Material and methods
2.1 Hyaluronic acid samples
Biolevox HA and Biolevox HA ONE (Biovico, Gdynia, Poland) products included in analysis were chosen from standard production batches. Detailed characteristics of mentioned products is attached in Supplementary materials (Table S1).
2.2 Physicochemical properties
Viscosity analysis was performed using a Kinexus Ultra + rheometer (Malvern). In the method of determining the viscosity of the gels, a PU20 spindle (diameter 20 mm) was used, the measurement gap was established (2 mm) as well as temperature measurement was 37 °C (imitating the conditions in the human body). The measurement of hyaluronan concentration of the Biolevox HA and Biolevox HA ONE was determined using a HPLC (Agilent 1100, Agilent, USA) with DAD (G1315B) and Phenomenex PolySep-GFC-P Linear, LC Column 300 × 7.8 mm column (Phenomenex, USA). The molecular weight of both products was determined using a HPLC with an SEC column and MALS detector (HPLC Agilent 1100, Shodex OHpack LB-806 M 8.0 × 300 mm column, MALS DAWN 8+ from Wyatt, respectively).
2.3 Cytotoxicity assay
Cytotoxicity analysis of Biolevox HA and Biolevox HA ONE products was performed on WI-38 human primary fibroblast cell line (ATCC, USA). Samples were analyzed in triplicates and the obtained results were averaged. Additional detailed description is included in the Supplementary materials.
2.4 Effect of Biolevox HA and Biolevox HA ONE on chondrocyte mRNA expression in the presence of an inflammation-inducing agent
Primary chondrocytes were isolated from bovine joint cartilage and seeded at a density of 0.5 mln/mL. Cells were cultured in DMEM/F12 medium (Capricorn) supplemented with 10 % FBS (ATCC) and 1 % Pen/Strep (Capricorn) on 6-well plates (Corning). After reaching a full adherence, the FBS was withdrawn from the culture. Afterward the 24 hours of culture, a 3 mL of Biolevox HA or Biolevox HA ONE gel with 4 ng of rBoIL-1 (Sigma-Aldrich) was applied to the plate inserts (Greiner). The negative control was cells treated with culture medium only, and the positive control was cells cultured in medium containing 4 ng of IL-1β. The detailed description of gene expression analysis and primers sequences (Table S2) are attached in the Supplementary materials. The results of mRNA expression analysis for studied genes were presented as -ΔCt values.
2.5 Clinical outcome
In years between 2018 and 2020, a clinical research program entitled "Practical aspects of administration and experience with use of Biolevox HA for the knee joint and preferred parameters of intra-articular products with hyaluronic acid" was carried out in Poland.
In the research program a 60,515 injections of Biolevox HA and 7235 injections of Biolevox HA ONE were performed. During the program, 20,172 and 7235 patients responded to the questionnaire after conducted treatment with Biolevox HA and Biolevox HA ONE, respectively. Characterization of the included population is attached in the Supplementary materials (Table S3). After the injections, physicians filled the questionnaire that aimed to assess their experience with hyaluronic acid used for this IA therapy. Program allowed to broaden the knowledge of the hyaluronic acid products used for IA injection, as it served as confirmation of safety, effectiveness and tolerance of therapy. The following research is an evaluation of the functionality of injectable hyaluronic acids based on the laboratory results and experience of users, thus it is worth noting that the study presented here is not a clinical trial, and therefore does not indicate a direct clinical effect at the time points.
2.6 Statistical analysis
Detailed description of statistics used in the study are included in the Supplementary materials.
3 Results
3.1 Biolevox HA and Biolevox HA ONE are different products in terms of their physicochemical properties
The measured content of sodium hyaluronate in ready to use product showed values of 21.72 ± 0.32 mg/mL for Biolevox HA (Fig. 1A) and 24.45 ± 0.38 mg/mL for Biolevox HA ONE (Fig. 1A). The viscosity analysis of intra-articular gels showed that for Biolevox HA viscosity it was 257.2 ± 7.83 Pa∗s (Fig. 1B) and subsequently for Biolevox HA ONE the measured viscosity was 534.8 ± 26.43 Pa∗s (Fig. 1B). The determined molecular weight reached the following mean values: 1388 ± 28.86 kDa for Biolevox HA and 1604 kDa ± 26.80 kDa for Biolevox HA ONE (Fig. 1C).

3.2 Biolevox HA and Biolevox HA ONE possess equivalent biological properties
The influence of different concentration of HA gels on WI-38 cells morphology and cell viability was evaluated.
Results showed that the cells morphology when exposed to increasing concentrations of Biolevox HA and Biolevox HA ONE remained unchanged. Cells were viable in all concentrations of the product as they remained their fibroblast-like shape and morphology (Figs. S1A and S2A). Cells were characterized with similar growth rate (>90 %) during incubation in 12.5 % and 25 % concentrated HA gels. The analyzed products showed a lack of cytotoxic effect towards WI-38 cells. The only minor adverse effect on cell growth was achieved, when the researched products were at 100 % concentration and the viability of the WI-38 cells reached 83.96 ± 2.91 % and 76.80 ± 5.90 % for Biolevox HA and Biolevox HA ONE, respectively (Figs. S1B and S2B). This stayed in contrast when compared to the control where the evident cell death was related to the increasing concentration of DMSO, and it was clearly visible already at low concentration of the used buffer. Cells exposed to increasing concentrations of DMSO lost their fibroblast-like cell shape, became spherical, and cell death was certainly noticeable (Figs. S1 and S2).
Analysis of bCol2a1 and bAcan expression in chondrocytes showed that 4 ng of IL-1β was able to significantly reduce expression levels of both genes in chondrocytes. For bCol2a1 and bAcan the reduction was 1.27 Ct and 1.3 Ct, respectively (Fig. 2A and B). However, when the cells were treated with either of Biolevox gels in the presence of IL-1β, the expression levels of bCol2a1 and bAcan increased when compared to cytokine-treated samples. For bCol2a1 this increase in the expression level was 0.84 Ct in chondrocytes treated with Biolevox HA and 1.22 Ct in cells treated with Biolevox HA ONE (Fig. 2A). For bAcan this increase in the expression level was 0.98 Ct in cells treated with Biolevox HA and 1.24 Ct in cells treated with Biolevox HA ONE (Fig. 2B).

3.3 Biolevox HA and Biolevox HA ONE provide equal beneficial outcome
After intra-articular injection with Biolevox HA or Biolevox HA ONE a clinical outcome of the therapy was evaluated with appropriate questionnaire that contained series of questions that guided this evaluation (Fig. 3).

Therapy with both of the products is considered as a very well tolerated. This is based on the fact that 99.6 % of specialist determined the intra-articular injection with Biolevox HA as well tolerated and 99.9 % of the specialist described the intra-articular injection with Biolevox HA ONE as well tolerated therapy (Fig. 3A).
After injections of Biolevox HA hyaluronic acid, the vast majority of specialists (98.2 %) did not recommend use of NSAIDs. Only 1.8 % recommended diclofenac or ibuprofen for their patients to be taken after the therapy based on HA (Fig. 3B). Furthermore, NSAIDs was not recommended by 98.8 % of the specialists who administered Biolevox HA ONE (Fig. 3B). In this case, only 1.2 % of physicians recommended the use of NSAIDs, such as aceclofenac or celecoxib, after the therapy with this hyaluronic acid (Fig. 3B).
For the majority of patients (86.1 %), the injection procedure with Biolevox HA was not related with any pain and only for 13.9 % of them the procedure was painful (Fig. 3C). For this reason, in most cases, so in 88.8 % physicians decided to administer Biolevox HA hyaluronic acid without anesthesia (Fig. 3D). Similar situation was observed in the case of Biolevox HA ONE, where the vast majority of patients assessed the viscosupplementation procedure as painless (78.5 %) and only 21.5 % of patients experienced pain during the injection (Fig. 3C). Therefore, 84 % of the physicians did not use anesthesia during the administration of Biolevox HA ONE (Fig. 3D).
4 Discussion
The vast majority of clinical reports identify IA injections with HA as first line, non-surgical treatment for OA patients, as they are able to alleviate pain and increase functionality of the joint affected by the OA-related deleterious changes.13 Moreover, recent epidemiological data show that intra-articular injections of HA do not only relieve symptoms of knee OA but also have the potential to delay the necessity for total knee arthroplasty (TKA).4,14
Nevertheless, the market of HA dedicated for IA injection is already considerable large, providing numerous various products that can differ in volume, molecular weight (MW), concentration, amount of hyaluronic acid in the injection, viscoelastic properties, cross-linking and presence of any additional ingredients.15 Importantly, the clinical effectiveness of HA-based products can be highly affected by those differences, making it sometimes clinically ineffective. Regrettably, current literature provides inconsistent results and conclusions regarding this problem and thus regarding a IA-HA injection as OA treatment method itself.16 Consequently, several popular guidelines for clinical practice (GCPs) have provided inconsistent and contradictory recommendations for therapy based on IA-HA.7,10,17,18 Unfortunately, there is general lack of understanding regarding the differences in the intrinsic properties of IA-HA products and relation of those to their clinical efficacy.
Motivated by this general problem, in the current article we decided to determine if the occurrence of divergence in physicochemical properties of two studied HA products will influence their biological characteristics and thus affect their clinical recognition, already as therapeutical agents.
We confirmed here that Biolevox HA ONE is characterized with higher MW than Biolevox HA. Consequently, greater MW value of Biolevox HA ONE together with higher concentration of hyaluronan in the gel, determines higher product viscosity when compared to the viscosity of Biolevox HA. This stays in line with previous investigations, which shows that product MW parameter is positively correlated with product viscosity.19
MW of HA is strictly connected with the linear structure of the hyaluronan molecule,20 and importantly it is crucial for HA molecule biological functionality. Scientific data show that HA MW parameters play crucial role in HA involvement on biological, thus clinical functioning at the molecular level.21 In this scope, the relation of HA and cellular activity is directly related to signaling via specific cell membrane receptor – CD44 (cluster of differentiation receptor 44). The connection of HA with CD44 receptor and later proper multiple receptor clustering plays a critical role in the induction of anti-catabolic processes in OA affected joint.22 It is demonstrated that the beneficial clinical outcome is closely bounded to the certain range of HA molecules MW that should stay between 1.3 and 1.7 MDa.23 Thus, regarding the correlation of the best physicochemical properties and subsequent clinical outcome related to the therapy, a selected range of MW of HA molecules seems to play a key role. Based on this, we identify that both of the HA gels investigated in this study meet the criteria to be characterized as products with optimal MW properties. This is strictly connected with proper clustering with CD44 receptors in order to achieve the biological functionality and thus long-lasting clinical recognition.
Within this framework, we have been able to ascertain the beneficial impact of examined HA gels with defined MW on bCol2a1 and bAcan genes expression, whose down-regulation was induced by pro-inflammatory cytokine IL-1β, in cartilage cells in-vitro culture. Above mentioned genes are responsible for biosynthesis of type II collagen and aggrecan molecules, respectively, which means that expression of both genes is critical for production of articular cartilage extra-cellular matrix (ECM). During OA a strong pro-inflammatory cytokines production and release occurs from the chondrocytes and synoviocytes, which leads to inhibition of both mentioned genes, amongst many others, what subsequently leads to the progression of tissue degradation and OA development.24 Here we show that both, Biolevox HA and Biolevox HA ONE with defined MW able to group CD-44 receptors, used in the study led to rescue of bCol2a1 and bAcan expression in the presence of IL-1β. Considering those results, both investigated products characterized with appropriate MW value express a direct beneficial biological effect.
The most common type of the treatment procedure with HA is a course of three consecutive injections, with 7 days intervals (like for Biolevox HA). An alternative might be a single injection of more concentrated HA gel25 (like Biolevox HA ONE). This different allocation of the products are strictly correlated to the differences in their concentration as well as in their viscosity. However, here we show that, despite the diversity of physicochemical properties between Biolevox HA and Biolevox HA ONE, the therapeutic functionality of both products is highly similar in all studied clinical outcome aspects. This is confirmed by the literature, since many clinical studies demonstrate that there is no difference in clinical effectiveness between single or multiple course of IA HA therapy.26,27 Moreover, comprehensive meta-analysis show that both therapies based on single or multiple injection of HA (2–4 injections) courses showed clinical effectiveness, which lasts for at least 6 months.28
All intra-articular therapies used in OA affected joint might be subjected to significant questions about their safety.29 Therapy based on HA is no exception. The literature shows that HA of avian origin shows much higher incidence of effusion upon the IA injection than HA derived from bacterial fermentation.30 Thus, the HA obtained by bacterial fermentation is safer than HA of animal origin.12 In current research, two included HA have its origin from the bacterial fermentation, which proves of their high safety. This was also confirmed by in vitro study where a direct impact of both studied HA gels on cell cultures was shown and no significant influence of HA products on cells viability has been observed. This observation is in line with current literature that prove no influence of increase in HA concentration on cell viability.31 Furthermore, the IA application of the analyzed HA products, were connected with satisfactory clinical outcome measured with high therapy tolerance, which confirms of their clinical recognition.
Nevertheless, as authors of this research were aware of some limitations accompanying this paper. Worth noting limitations will be a lack of designed follow-up periods after the IA therapy with investigated products. In order to determine the clinical efficiency of the injectable products, it is important to investigate the effect of the therapy after chosen time intervals. Nevertheless, despite that, the presented study presented here is not a clinical trial, but a clinical recognition of the products, which can also provide significant information about their behavior when used in clinical situation. On that account, authors are finalizing a full clinical study with suitable follow up.
5 Conclusion
In summary, the current paper focuses on the differences in physicochemical parameters of two analyzed products (Biolevox HA and Biolevox HA ONE) that may have an impact on the biological functioning of HA and finally on their clinical recognition in OA patients. Products included in this research are different in the number of doses in course, what strictly determines their initial properties like concentration, molecular weight and finally viscosity. However, despite those differences, both products have the same anti-catabolic biological potential and analogous clinical recognition.
Funding
This study was supported by the Biovico Sp. z o.o. company (Gdynia, Poland) based on own funding. The funders had no role in study design, decision to publish, or preparation of the manuscript.
Ethical approval and patient consent
The submitted research is not a clinical trial, but a clinical recognition assessment of used HA-products, thus the included results are in form of questionnaire answers of physicians about the performed viscosupplementation. Thus, the authors indicated that, there is no need for an ethical statement.
Declaration of patient consent form
The submitted research is not a clinical trial, but a clinical recognition assessment of used HA-products, thus the included results are in form of questionnaire answers of physicians about the performed viscosupplementation. Thus, the authors of this study has not include patient's consent.
Patient's concent statement
The submitted research is not a clinical trial, but a clinical recognition assessment of used HA-products, thus the included results are in form of questionnaire answers of physicians about the performed viscosupplementation. Thus, the authors of this study has not include patient's consent.
CRediT authorship contribution statement
Julián B. Troncoso: Conceptualization, Project administration, Writing – original draft. Javier R. Fernández: Conceptualization, Methodology, Writing – original draft. Jakub Hubeny: Writing – original draft, Visualization, Formal analysis. Marta Grzywacz: Writing – original draft. Anna Silwanowicz: Investigation, Validation. Magdalena Bunikowska: Investigation, Data curation. Patryk Kozłowski: Investigation, Data curation. Marcin Martyniak: Methodology, Funding acquisition, Project administration, Supervision, Writing – review & editing.
References
- The global burden of hip and knee osteoarthritis: estimates from the Global Burden of Disease 2010 study. Ann Rheum Dis. 2014;73(7):1323-1330.
- [Google Scholar]
- Osteoarthritis: pathology, diagnosis, and treatment options. Med Clin North Am. 2020;104(2):293-311.
- [Google Scholar]
- Prospects for disease modification in osteoarthritis. Nat Rev Rheumatol. 2006;2(6):304-312.
- [Google Scholar]
- Effectiveness of intra-articular therapies in osteoarthritis: a literature review. Therap Advan Musculoskelet. 2017;9(8):183-196.
- [Google Scholar]
- Intra-articular hyaluronan (hyaluronic acid) and hylans for the treatment of osteoarthritis: mechanisms of action. Arthritis Res Ther. 2003;5(2):54-67.
- [Google Scholar]
- Longterm effects of intraarticular hyaluronan on synovial fluid in osteoarthritis of the knee. J Rheumatol. 2006;33(5):946-950.
- [Google Scholar]
- OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis Cartilage. 2019;27(11):1578-1589.
- [Google Scholar]
- Non-surgical management of knee osteoarthritis: comparison of ESCEO and OARSI 2019 guidelines. Nat Rev Rheumatol. 2021;17(1):59-66.
- [Google Scholar]
- An algorithm recommendation for the management of knee osteoarthritis in Europe and internationally: a report from a task force of the European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis (ESCEO) Semin Arthritis Rheum. 2014;44(3):253-263.
- [Google Scholar]
- A systematic review of current clinical practice guidelines on intra-articular hyaluronic acid, corticosteroid, and platelet-rich plasma injection for knee osteoarthritis: an international perspective. Orthopaedic J Sport Med. 2021;9(8)
- [Google Scholar]
- Product differences in intra-articular hyaluronic acids for osteoarthritis of the knee. Am J Sports Med. 2016;44(8):2158-2165.
- [Google Scholar]
- Development and prospect of intra-articular injection in the treatment of osteoarthritis: a review. J Pain Res. 2020;13:1941-1955.
- [Google Scholar]
- Hyaluronan and synovial joint: function, distribution and healing. Interdiscipl Toxicol. 2013;6(3):111-125.
- [Google Scholar]
- The efficacy of multiple versus single hyaluronic acid injections: a systematic review and meta-analysis. BMC Muscoskel Disord. 2017;18(1):542.
- [Google Scholar]
- The effect of intra-articular hyaluronic acid treatment on gait velocity in older knee osteoarthritis patients: a randomized, controlled study. Arch Gerontol Geriatr. 2012;55(2):310-315.
- [Google Scholar]
- AAOS clinical practice guideline: treatment of osteoarthritis of the knee: evidence-based guideline, 2nd edition. J Am Acad Orthop Surg. 2013;21(9):577-579.
- [Google Scholar]
- 2022
- [Google Scholar]
- Effects of molecular weight of hyaluronic acid on its viscosity and enzymatic activities of lysozyme and peroxidase. Arch Oral Biol. 2018;89:55-64.
- [Google Scholar]
- Differential activation of ERK and rac mediates the proliferative and anti-proliferative effects of hyaluronan and CD44. J Biol Chem. 2008;283(46):31823-31829.
- [Google Scholar]
- Expression of CD44 in articular cartilage is associated with disease severity in knee osteoarthritis. Mod Rheumatol. 2013;23(6):1186-1191.
- [Google Scholar]
- High-molecular-weight hyaluronic acid attenuated matrix metalloproteinase-1 and -3 expression via CD44 in tendinopathy. Sci Rep. 2017;7(1)
- [Google Scholar]
- Hyaluronic acid counteracts interleukin-1-induced inhibition of collagen biosynthesis in cultured human chondrocytes. Pharmacol Res. 2006;54(4):275-281.
- [Google Scholar]
- Exploring reasons for the observed inconsistent trial reports on intra-articular injections with hyaluronic acid in the treatment of osteoarthritis: meta-regression analyses of randomized trials. Semin Arthritis Rheum. 2016;46(1):34-48.
- [Google Scholar]
- Intra-articular hyaluronic acid in the symptomatic treatment of knee osteoarthritis: a meta-analysis of single-injection products. Curr Ther Res. 2019;90:39-51.
- [Google Scholar]
- Efficacy of single high-molecular-weight versus triple low-molecular-weight hyaluronic acid intra-articular injection among knee osteoarthritis patients. BMC Muscoskel Disord. 2020;21(1):550.
- [Google Scholar]
- The efficacy of multiple versus single hyaluronic acid injections: a systematic review and meta-analysis. BMC Muscoskel Disord. 2017;18(1):542.
- [Google Scholar]
- Viscosupplementation for knee osteoarthritis: systematic review and meta-analysis. BMJ July 6, 2022
- [Google Scholar]
- Comparison of avian and nonavian hyaluronic acid in osteoarthritis of the knee. ORR. 2010;2:5-9.
- [Google Scholar]
- In vitro evaluation of cytotoxicity of hyaluronic acid as an extracellular matrix on OFCOL II cells by the MTT assay. Oral Surg Oral Med Oral Pathol Oral Radiol. 2014;117(6):e423-e428.
- [Google Scholar]

