SciELO - Scientific Electronic Library Online

 
vol.84 issue3MastocytosisAn estimate of the cost of treating mlanoma disease in the state of Sao Paulo - Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Article

Indicators

Related links

Share


Anais Brasileiros de Dermatologia

On-line version ISSN 1806-4841

An. Bras. Dermatol. vol.84 no.3 Rio de Janeiro July 2009

http://dx.doi.org/10.1590/S0365-05962009000300003 

INVESTIGATION

 

Sugar (sucrose) and fatty acid compounds with triglycerides on the treatment of wounds: experimental study in rats

 

 

William César CavazanaI; Maria de Lourdes Pessole Biondo SimõesII; Sergio Ossamu YoshiiIII; Ciomar Aparecida Bersani AmadoIV; Roberto Kenji Nakamura CumanV

IAssistant Professor, Discipline of Plastic Surgery, Universidade Estadual de Maringá – Maringá (PR), Brazil
II
Full Member of Colégio Brasileiro de Cirurgiões (CBC) and member of Sociedade Brasileira de Cirurgia Plástica (SBCP). Faculty Professor, Discipline of Scientific Methodology, Pontifícia Universidade Católica do Paraná, Associate Professor
IIIAssociate Professor, Discipline of Clinical Pathology, Universidade Federal do Paraná – Maringá (PR), Brazil. Associate Professor, Discipline of Clinical Pathology, Pontifícia Universidade Católica do Paraná – Maringá (PR), Brazil IV Associate Professor, Discipline of Pharmacology, Universidade Estadual de Maringá (UEM) – Maringá (PR), Brazil
IVDepartment of Surgery, Discipline of Surgical Technique and Experimental Surgery, Universidade Federal do Paraná – Maringá (PR), Brazil
VPh.D. in Science – Pharmacology, Universidade de São Paulo (USP). Associate Professor, Discipline of Pharmacology, Universidade Estadual de Maringá (UEM) – Maringá (PR), Brazil

Mailing Address

 

 


ABSTRACT

BACKGROUND: In the last 10 years, the use of essential fatty acids (EFA) compounds for the treatment of wounds has increased in Brazil, while there has been reducing indication for the use of sugar.
OBJECTIVE: The aim of this study was to evaluate the response to saline, sugar and EFA on induced wounds in rats.
METHODS: A wound of 400 mm2 was produced on the back of each Wistar rat. The rats were divided into three groups, each being treated with either saline, sugar or EFA. All the animals received a closed dressing on the wounds, changed daily. Measures were taken in four moments, and the values of wound area reduction by healing, cellular inflammatory response, collagen ordering and types I and III collagen density were assessed.
RESULTS: The wound healing was equal in all the three groups, but sugar promoted an inflammatory response modulation between the 7th and 14th days. On the 20th post-operative day, there were no differences between the three treated groups concerning types I and III collagen.
CONCLUSIONS: The wounds healed in the three groups. The sugar group promoted effective cellular inflammatory response modulation. There were no differences between all the treated groups regarding types I and III collagen at the end of this study.

Keywords: Wound Healing; Skin


 

 

INTRODUCTION

Many aspects of second intention healing of pressure ulcers and skin wounds have been studied. These studies aim at standardizing the techniques that contribute to healing process with fast evolution and good quality scar.

The first clinical indications of sugar were described in the old age using honey or syrup 1. The first scientific description was published by Gozenbach and Hoffman in 1936.2

Sugar regulates the inflammatory response 3 and it is also bacteriostactic and bactericide owing to the high osmolarity of syrup that is formed after few hours from its application over the wound 4.

According to Biondo-Simões et al.,5 topical use of sugar on skin wounds of Wistar rats accelerates the healing process. When comparing its effect to that of acexamic acid it was observed that it promoted acceleration of the healing process in the first seven days 6.

In Brazil, treatment of pressure ulcers and wounds using medium-chain essential fatty acids and triglycerides (EFA-TG) was popularized after the publication of Declair, in 1994, who observed the clinical effects of EFA-TG in prevention of bed sores 7. The same agents have also proved to be more effective in bed sore management than topical polyvinylpyrrolidone-iodine (PVPI) 8.

In recent decades, there has been an increase in clinical indication of EFA-TG by Brazilian healthcare professionals dedicated to managing wounds, whereas the indications to use of sugar have decreased based on the assumption that sugar could become a culture medium to bacteria in the treated tissues 9. However, there is no evidence in the scientific literature that validates this assumption, even though it is at large transmitted in many medical and nursing services.

Pieper et al.10 reported that in Brazil products such as sugar, papaya/ papain and essential fatty acids are widely used in wounds and in scientific studies there are varied protocols but low evidence level. These studies have assumed that healthcare professionals may respond for the situations in which these not standardized products have been used.

The purpose of the present study was to assess the effect of using sodium chloride saline solution, sugar and EFA-TG to treat experimentally-induced skin wounds in rats.

 

MATERIAL AND METHODS

The experimental protocol of the present study was approved by the Ethics Committee for Experimenting in Animals, Universidade Estadual de Maringá, under number 043-2006.

We used 96 rats (Rattus norvegicus albinus, Rodentia mammalia), Wistar line, male and aged approximately 120 days, mean weight of 233. 8 grams and standard deviation of 49.94.

Animals were maintained in the sector animal laboratory, in polypropylene cases standardized for their species and housing four animals per case. They had free access to water and standard feed for the species and were maintained under regular environmental humidity, with light/dark cycles of 12 hours and controlled temperature.

Animals were randomly distributed into three groups of 32, forming the samples. Group A was named the control group; group B, sugar; group C, EFA-TG (Trigliceril CM®). Surgical procedure was started by anesthesia applied intramuscularly with hydrochloride -(2,6-xylidine)-5,6-dihydro-4H-1,3-thiazine (Ronpun®) and ketamine hydrochloride (Ketalar®) in 1:1 ratio, employing a association of 1mk/Kg-1 of body weight, followed by manual epilation of the animal back. Next, we used a 400 mm2 silicone mold reinforced with polypropylene to define the area to be resected on the dorsal medial region. We made a quadrangular incision over the marked area with resection of dorsal-medial region. The bloody areas were treated with two milliliters of sodium chloride saline solution in animals of the control group, two grams of crystallized sugar in the sugar group, and two milliliters of EFA-TG solution in the animals in the EFA-TG group. In each animal we made an occlusive dressing over the wound. Dressings were changed daily once a day at the same time and under the same anesthetic protocol.

Euthanasia of eight randomly selected animals per group was made on days 3, 7, 14 and 20 with lethal dose of the anesthetic. After the animals were dead, the subgroups were formed. Animals in the subgroups control 1-8, sugar 1-8 and EFA-TG 1-8 corresponded to day 3. Subgroups control 9-16, sugar 9-16 and EFA-TG 9-16 to day 7; subgroup control 17-24, sugar 7-24 and EFA-TG 17-24 to day 14; and animals of subgroups control 25-32, sugar 25-32 and EFA-TG 25-32 corresponded to day 20 of the checking. When the animals were anesthetized, we took pictures of the wounds using a camera Canon EOS Rebel G and placed a transparent plastic sheet over the lesions, to obtain planigraphies. Next, each lesion was surgically removed, including the subcutaneous area, with regular margins of skin measuring 5 mm, followed by euthanasia. The material was fixed with formalin and a central segment of the larger lesions or covering the lesion extension was included in paraffin. From 4 micrometer-thick histological sections we prepared three slides to each animal, stained with hematoxylin-eosin, Mallory trichrome and Sirus-red.11

The photographs enabled registration of the healing progression of wounds in the studied groups.

Planigraphy images were digitalized and their area was quantified by software Auto CAD R14® (Auto Desk®), in square millimeters.

Histological analyses of inflammatory reaction and collagen order were performed by the laboratory of Pathology, Pontifícia Universidade Católica do Paraná – Curitiba, PR.

In sections stained with hematoxylin-eosin we analyzed cell inflammatory response. We read the highest cellularity field and, in it, 100 points in the 400x magnification. Predominance of polymorphonuclear (PMN) characterized the acute cell inflammatory response and predominance of monomorphonuclear (MMN) characterized chronic cellular response.

In sections stained with trichrome, collagen fibers got blue and collagen order was assessed based on semi-quantitative analysis after the section had been visualized in magnification of 40x and 100x, classifying the collagen fiber order as disorganized and organized, according to the parallel arrangement among them.

In sections stained with Sirius-red, the percentage of types I and III collagen fibers was obtained by reading under polarized light microscopy, corresponding to the lesion and 400 times magnification, with images captured by digital camera, transferred to a computer and analyzed by software Optimas® 6.0 (Media Cybernetics, Inc., USA). Type I collagen fibers are thick and have an orange to red color, and collagen fibers type III are fine and obtain a green color. Histological analyses were performed in the laboratory of experimental pathology at Pontifícia Universidade Católica do Paraná – Curitiba, PR.

Descriptive results of the study were expressed by mean, standard deviation and standard error. To assess the effect of the group and the checking day over the studied variables Anova methodology was used (analysis of variance ) with two factors (group and checking day), followed by Tukey-Kramer test using software Statistica 7.0 (StatSoft, 2004). To assess collagen order, we used the comparison of two ratios for independent samples using software Primer of Biostatistics (Glantz S.A.). The statistically significance level adopted was 0.05.

 

RESULTS

There were four deaths of animals treated with sugar and four deaths of animals treated with EFA-TG during the experiment, resulting from the anesthetic act.

The visual analysis of photos, used to assess the conditions of wounds, showed that they had clinically presented as expected for this type of lesion, with solution of continuity in the first days after the surgery that progressed to scar at the final stages of the repair.

The percentage of reduction in the wound area did not suffer influence of any treatment (p=0.101) and there was no interaction between treatment and time (p = 0.5084), which led to equivalent results in the three groups. The only effect that proved to be strongly significant was time (p < 0.001) (Table 1).

In the assessment of inflammatory response, the influence of time was significant (p=0.0048), but the effect of treatment occurred only under interaction (p = 0.0042). Therefore, we observed that there was no significant difference in relation to treatment between the EFA-TG group and the control group, but the treatment with sugar provided an increase in acute stage inflammatory cells as of day 3 up to day 7, which caused significant reduction on day 14, which coincided with increase in cells of chronic inflammatory response (p < 0.001) (Table 2).

Concerning collagen order, there were no statistically significant differences between the treatment groups and the checking days (Figure 1).

When we analyzed the density of collagen fibers type I and type III, the effect of time (p = 0.7884) and treatment (p = 0.8604) was not significant, differently from the effect of interaction (p < 0.001). On day 20 there was higher density of collagen type I and lower density of collagen type III (p=0.0027) in EFA-TG group when compared to the control group. However, there were no statistically significant differences when the groups sugar and EFA-TG (p=0.098) or sugar and control (p=0.1104) were compared that day. However, the amount of type I collagen in the sugar group on day 20 was significant in relation to the control group, demonstrating that treatment with sugar had a trend of better healing conditions than treatment with sodium chloride saline solution (Table 3).

 

DISCUSSION

Literature data concerning the topical use of sugar 12 and EFA-TG in wound healing are scarce, as confirmed by comprehensive literature review.

Regular commercial sugar is disaccharide (glucose/ fructose, 1-2 binding) under the form of granules measuring 0.16 to 0.80 µm.13 According to Barnes,13 topical application of sugar promoted healing in almost 80% of the 334 bed sores studied in five years, which made them conclude that sugar is bactericide.

Honey had similar properties to sugar and Blomfield14 emphasized its benefits and low cost.

The antibacterial activity of sugar was studied by Rahal et al.,4 and showed in vitro bactericide activity to Staphylococcus aureus, Escherichia coli, Pseudomonas aeruginosa and Klebsiella enterobacter and bacteriostactic activity to Bacillus subtilis.

Herszage et al.15 reported 120 patients, among which 6 were diabetic, treated with application of sucrose over the suppurative lesions in which healing quickly progressed with disappearance of agents such as Streptococcus, Klebsiella, Escherichia and Clostridium. According to them, wound management with sugar would eventually be consolidated as an option owing to the antimicrobial properties of mono and disaccharide.

Shi et al.16 demonstrated that a paste containing 70% sugar and 3% povidone-iodine sped up reepithelialization of ulcers induced in diabetic rats infected with methylcyclin-resistance Staphylococcus aureus and reduced the amount of colony forming units in these ulcers compared to non-treated wounds.

Water activity (Aw) refers to the amount of free water in the medium. Aw of 0.6 means the medium is free from microorganisms. Bacterial growth, according to Chirife et al.17, depends on water flow into the bacteria and efficacy of sugar in managing infected wounds is due to high local osmolarity (low Aw). In a clinical trial, these authors analyzed Aw in an abdominal wound treated with sugar. After 10 hours, Aw was 0.951, which to them continued to be low Aw, effective to inhibit the growth of most human pathogens. These data suggest that dressing changes with sugar may be safely made in longer than 10-hour intervals, which puts an end to the need for multiple changes within shorter intervals, which was broadly diffused. In our study, dressings were changed every 24 hours, without impairing healing.

Concerning hyperosmolarity of the wound, Forrest 18 questioned the possibility of damaging the tissue cells, and Duffet et al.19 published that these cells are protected from high sugar concentrations thanks to intercommunication with other cells of the same organism, which enables water exchange among them.

To assess the effects of sucrose in three different concentrations (0.01; 0.1 and 1 M) in the formation of granulation tissue in a standard wound, Kössi et al.20 analyzed the amount of granulation tissue and the quantity and distribution of collage types I and III. They concluded that the different concentrations of solutions containing sugar had not qualitatively modify the granulation tissue originated in the treated wounds. In our study, on day 20, both sugar and EFA-TG group presented equivalent type I collagen levels, suggesting a trend of both treatments to lead to scars with good resistance.

The modulating role of inflammatory response of sugar was studied by Nakao et al.,3 exposing keratinocytes and fibroblasts in a mixture containing 70% sugar and 3% povidone-iodine, with positive results both for antibacterial action of sugar and its modulating action of inflammatory reaction. This study has also indicated that sugar acted as a modulator of inflammatory reaction and not simply as a harmful agent to bacteria by a mechanism of hyperosmolarity.

Concerning essential fatty acids, Nutegeren et al. 21 demonstrated that essential fatty acids that were radiolabeled over the normal skin reached the deeper layers of the epidermis, incorporated to membrane phospholipids and acylglucosylceramides. Thus, Marcelo et al.22 observed that acylglucosylceramides suffer enzymatic lipo-oxygenation to form the acylceramides that form the water barrier of the skin, with production of peroxide. Peroxides may activate guanylate cyclase, promoting increase in level of cyclic GMP in a site appropriate to the epidermis, influencing the balance between proliferation and epidermal cell differentiation. It has also been demonstrated by White et al.23 that inhibin, empirical name of an antibacterial agent present in honey, is hydrogen peroxide. Therefore, the peroxide originated both from sugar and essential fatty acids contribute to improving the healing process of proinflammatory agents. Moreover, essential fatty acids preserve the morphological characteristics of the skin cover layer, because to Horrobin,24, changes caused on the skin by deficiency of EFA are similar in all species, delaying normal healing by the formation of defective collagen and marked transepidermal water loss. According to Hansen and Jensen,25 linoleic acid is essential to maintain the skin water barrier, regardless of its role as precursor of arachdonic acid and eicosanoids.

In surgically-induced wounds in rats, Cardoso et al.26 demonstrated that linolenic fatty acids (n-3), linoleic (n-6) and oleic (n-9) fatty acids modulate the repair. Linoleic acid, as proinflammatory agent, has promoted significant improvement in wounds, with peaks of nitric oxide production 48 hours after surgery. Conversely, oleic acid has strongly inhibited the production of nitric oxide in the wound and linolenic acid has also statistically significantly delayed the healing of wounds. In the present study, sugar was superior to EFA-TG as stimulator of inflammatory response.

Greca et al.,27 irrigating the intraluminal mucosa of rats submitted to colonic anastomosis with isosmolar solution of short chain fatty acids and isotonic saline solution for 7 days, concluded that treatment with short-chain FA promoted greater concentration of type I collagen and total concentration on the suture line of the rectal stump, when compared to the control group. Our study has also indicated better effect of EFA-TG in relation to isotonic saline solution maintained through soaked gauzes over the wounds.

Araújo et al.28 noticed that on days 3 and 14 of postoperative care mild mineral oil (Agarol®) and Trigliceril® influenced the healing process of skin wounds. On postoperative day 7, there was increase in quantity of granulation tissue in groups treated with Agarol® , with reduced neovascularization in group treated with Trigliceril®. The results presented in this experiment did not manage to demonstrate the effects of EFA-TG that would justify the different beneficial properties that were attributed to them.

Comparing the topical application of medium chain triglycerides, sunflower oil and isotonic saline solution in rats, Rocha et al.29 concluded that medium chain triglycerides as well as sunflower seed oil modify the process of formation of a beneficial scar when compared to the control group. To these authors, therefore, there was no difference in the type of oil that contained essential fatty acids, which in a way expands the access of needy populations to this product, as opposed to regular commercial prices.

Semi-quantitative histological analysis of collagen fiber order showed that healing quality was equivalent in the three studied groups, with gradual replacement of disarrangement between collagen fibers owing to its ordered matching in the end of healing.

In summary, the analysis of results led to the conclusion that in all groups there was equivalent wound healing during the experiment; sugar promoted modulation of cell inflammatory response as of day 3 and 7 and day 14 (there was replacement of PMN by MMN on day 20); and even though group EFA-TG presented higher amount of type I collagen and less amount of type III collagen in relation to the control group, when sugar and EFA-TG groups were compared there were no significant differences.

 

CONCLUSION

In this experimental study we demonstrated the therapeutic advantages of sugar in relation to EFA-TG and to saline solution and sodium chloride in treating acute skin wounds in guinea pigs (rats). Clinical trials in humans should be performed to confirm the possible benefit of using sugar in treatment of wounds in human beings.

 

REFERENCES

1. Early history of wound treatment. J R Soc Med. 1982;75:198-205         [ Links ]

2. Prata MB, Haddad CM, Goldenberg S, Simões MJ, Moura LA, Trabulsi LR. Uso tópico do açúcar em ferida cutânea: estudo experimental em rato. Acta Cir Bras. 1988;3:43-8         [ Links ]

3. Nakao H, Yamazaki M, Tsuboi R, Ogawa H. Mixture of sugar and povidone--iodine stimulates wound healing by activating keratinocytes and fibroblast functions. Arch Dermatol Res. 2006;298:175-82         [ Links ]

4. Rahal F, Mimica I, Pereira V, Athié E. O açúcar no tratamento local das infecções de feridas cirúrgicas. Rev Col Bras Cir. 1983;10:135-6         [ Links ]

5. Biondo-Simões MLP, BaretaJr VC, Ferreira LF, Collaço LM. Efeito do açúcar na cicatrização por segunda intenção: estudo experimental em ratos. Acta Cir Bras. 1991;61:65         [ Links ]

6. Biondo-Simões MLP, Lima EJB, Rosário MAK, Marques LO, Adur RC, Cavazana WC, et al. Açúcar e ácido acexâmico na cicatrização de feridas cutâneas em ratos. Acta Cir Bras. 1993;8:83-6         [ Links ]

7. Declair V. Aplicação do triglicerídeos de cadeia média (TCM) na prevenção de úlceras de decúbito. Rev Bras de Enferm. 1994;47:27-30         [ Links ]

8. Goldmeier S. Comparação dos triglicerídeos de cadeia média com ácidos graxos essenciais, com polivinilpirrolidona- iodo no tratamento de úlceras de decúbito em pacientes cardiopatas. Rev Paul Enferm. 1997;16:30-4         [ Links ]

9. Santos KA, Neves RCS, de Jesus MCP. Uso terapêutico do açúcar mascavo em ulcerações. HU Rev. 1996;22:52-67         [ Links ]

10. Pieper B, Caliri MH. Nontraditional wound care: a review of the evidence for use of sugar, papaya/papain, and fatty acids. J Wound Ostomy Continence Nurs. 2003;30:175-83         [ Links ]

11. Junqueira LC, Bignolas G, Brentani RR. Picrosirius staining plus polarization microscopy, a specific method for collagen detection in tissue sections. Histochem J. 1979;11:447-55         [ Links ]

12. Borges EL, Latine F, Carvalho DV, Machado AM. A utilização do açúcar no tratamento de feridas cutâneas: revisão bibliográfica dos últimos 20 anos. Rev Med Minas Gerais. 2002;12:163-73         [ Links ]

13. Barnes WB. Sugar sweetens the lot of patients with bedsores. JAMA. 1973;223:122         [ Links ]

14. Blomfield R. Honey for decubitus ulcers. JAMA. 1973;224:905         [ Links ]

15. Herszage L, Montenegro J, Joseph A. Traitement des plaies suppurées par application de saccharose. Nouv         [ Links ]

16. 1982;12:940. 16. Shi CM, Nakao H, Yamazaki M Tsuboi R, Ogawa H. Mixture of sugar and povidine-iodine stimulates healing of MRSA-infected skin ulcers on db/db mice. Arch Dermatol Res. 2007;299:449-56         [ Links ]

17. Chirife J, Herszage L, Joseph A, Kohn ES. In vitro study of bacterial inhibition in concentrated sugar solutions: microbiological basis for the use of sugar in treating infected wounds. Antimicrob Agents Chemother. 1983;23:766-73         [ Links ]

18. Forrest RD. Sugar in the wound. Lancet.1982;1:861         [ Links ]

19. Duffet Am,ViauF. Traitement des plaies et escarres par le sucre cristalisé. Rev Infirm.1986;36:37-8         [ Links ]

20. Kössi JA, Ekfors TO, Aaltonen V, Laato M. Sucrose has no beneficial effects on wound healing in rats. Eur J Surg. 2000;166:818-2         [ Links ]

21. Nugteren DH, Christ-Hazelhof E, van der Beek A, Houstmuller UM. Metabolism of linoleic acid and other essential fatty acids in the epidermis of the rat. Biochim Biophys Acta. 1985;834:429-36         [ Links ]

22. Marcelo CL, Duell EA, Stawiski MA, Anderson TF, Voorhees JJ. Cyclic nucleotide levels in psoriatic and normal keratomed epidermis. J Invest Dermatol. 1979;72:20-4         [ Links ]

23. White JW Jr, Subers MH, Schepartz AI. The identification of inhibine, the antibacterialfactor of honey, as hydrogen peroxide and its origin in a honey glucose-oxidase system. Biochim Biophys Acta. 1963;73:57-70         [ Links ]

24. Horrobin DF. Essential fatty acids in clinical dermatology. J Am Acad Dermatol. 1989;20:1045-53         [ Links ]

25. Hansen HS, Jensen B. Essential function of linoleic acid esterified in acylglucosylceramide e acylceramide in maintaining the epidermal water permeability barrier. Evidence from feeding studies with oleate, linoleate, arachidonate, columbinate e alpha-linoleate. Biochim Biophys Acta. 1985;834:357-63         [ Links ]

26. Cardoso CR, Souza MA, Ferro EA, Favoreto S Jr, Pena JD. Influence of topical administration of n-3 and n-6 essential and n-9 nonessential fatty acids on the healing of cutaneous wounds. Wound Repair Regen. 2004;12:235-43         [ Links ]

27. Greca FH, Biondo-Simões MLP, Collaço LM, Martins VDM, Tolazzi ARD, Gasparetto EL, et al. A ação dos ácidos graxos de cadeia curta na cicatrizaçäo de anastomoses colônicas: estudo experimental em ratos. Acta Cir Bras. 2000;15(Supl3):12-6         [ Links ]

28. Araújo CFR, Souza FZA, Greca FH, Guerreiro MHCPM, Leite AL, Mansur AEC, et al. Efeitos do Agarol® e do Trigliceril® sobre a cicatrização de pele: estudo expe imental em ratos. Acta Cir Bras. 1998;13:232-7         [ Links ]

29. Rocha RP, Rocha ELP, Hames RL, Sposeto TB. Estudo comparativo do processo de cicatrização com o uso de óleo de semente de girassol e triglicérides de cadeiamédia: modelo experimental em ratos. Sci Med. 2004;14:203-8        [ Links ]

 

 

Mailing Address:
William César Cavazana
Rua Dr. Nagib Daher 885
86800 040 Apucarana - Paraná
Tel./Fax: 43 3423 1339 3422-8577
E-mail: wcavazana@uol.com.br

 

 

Conflict of interest: None
Financial funding: Fundação de apoio ao desenvolvimento científico da Universidade Estadual de Maringá – Maringá (PR), Brasil.
How to cite this article: Cavazana WC, Biondo-Simões MLP, Yoshii SO, Bersani-Amado CA, Cuman RKN. Açúcar (sacarose) e triglicerídeos de cadeia média com ácidos graxos essenciais no tratamento de feridas cutâneas: estudo experimental em ratos. An Bras Dermatol. 2009;84(3):229-36.