Systemic Isotretinoin: Eflects on Dermal Wound Healing in a Rabbit Ear Model In Vivo RONALD L. MOY, M.D. LARRY S. MOY, M.D. RICHARD G. BENNETT, M.D. JOHN A. ZITELLI, M.D. JOUNI UITTO, M.D., PH.D.
Abstract. Clinical observations have suggested that wound healing may be altered in patients treated with systemic isotretinoin. In this study, we examined the effects of systemic isotretinoin on dermal wound healing and connective tissue metabolism in a rabbit ear model. Forty 6mm punch-biopsy wounds were created in the ears of two control rabbits as well as two experimental animals fed isotretinoin, 4 mg/kg per day. Clinical inspection and histologic examination revealed no difference between the control and isotretinoin-treated rabbits in terms of the time required for complete wound healing or the appearance of the final scar. The tissue removed from the wound site at days 0, 7, 14, and 21 after wounding was subjected to anal-
Ronald L. Moy, M.D., Chief of Mohs Micrographic Surgery and Cutaneous Oncology Unit, Assistant Professor; and Larry S. Moy, M.D., Chief Resident, are from the UCLA Division of Dermatology, Los Angeles, California. Richard G. Bennett, M.D., is in private practice and is a consultant in Medicine (Dermatology), University of Southern California, School of Medicine, Los Angeles, California. John A. Zitelli is staff physician from the Department of Medicine and Surgery, Montefiore Hospital, University Health Center of Pittsburgh, Pittsburgh, Pennsylvania. Jouni Uitto, M.D., Ph.D., is Chairman, Department of Dermatology, Thomas Jefferson University, Philadelphia, Pennsylvania. Address reprint requests to: Ronald L. Moy, M.D., UCLA Mohs Micrographic Surgery Unit, Division of Dermatology, Los Angeles, CA 90024. Supported in part by the United States Public Health Service, NIH Grants AR-28450, GM-28833, and AR-35297, by a grant from the American Society of Dermatologic Surgery awarded by the Dermatology Foundation in 1986, and by the Dermatology Research Foundation of California, Inc. Isotretinoin was kindly provided by Hoffmann-LaRoche. The Authors acknowledge the expert secretarial assistance of Kathy Casey.
ysis of a collagen production and collagen gene expression. Collagen production, determined by the synthesis of [3H]hydroxyproline after incubation of tissue slices with [3H]proline in vitro or by the measurement of the steadystate levels of types I and 111 procollagen mRNAs, was not significantly different between the two groups. The results indicate that systemic administration of isotretinoin does not affect collagen synthesis in the rabbit ear model of wound healing. J Dermatol Surg Oncal 1990;16:1142-1146.
Much retinoid research has focused on epidermal effects, however, retinoids are known to produce dermal alterations.',* In vitro studies have demonstrated that isotretinoin can selectively inhibit procollagen production by skin fibroblasts in culture.3~~ The excessive granulation tissue formation sometimes seen in isotretinoin patients could be explained by the fact that isotretinoin inhibits collagen degradative enzymes in fibroblast culture^.^-^ Previous experimental studies have shown that vitamin A stimulates incisional wounds in rats to heal more rapidly and prevents the inhibitory effects of cortisone on the healing of open wounds.9 Retinoids, synthetic derivatives of vitamin A, may have differential effects on the growth and differentiation of cells, and on the wound healing process. Few in vivo studies have investigated the effects of isotretinoin on the growth and differentiation of cells, or on the dermal wound healing process.2 Such studies are important because reports of keloids developing on patients taking isotretinoin who had full face dermabrasions may suggest a possible effect on collagen synthesis and on dermal wound healing.l0 In the present study, we examined the effect of systemic isotretinoin on dermal wound healing and
0 1990 by Elsevier Science Publishing Co., Inc. 0148-0812/90/$3.50
MOY ET AL
connective tissue metabolism in a rabbit ear model." We explored collagen biosynthesis by measuring the synthesis of protein bound radioactive hydroxyproline after incubating the tissue with radioactive proline and by measuring type I and type I1 procollagen mRNA levels in tissue after wounding.
MATERIALS AND METHODS WOUND HEALING MODEL
We used a rabbit ear model for wound healing in this study. Two New Zealand rabbits (each weighing approximately 2 kg) were administered 4 mg/kg per day of isotretinoin for 3 weeks. This dose of isotretinoin was chosen based on four prior studies (McFarland, unpublished observations).12McFarland had previously determined that the same oral dosage of isotretinoin in rabbits, given in the same manner as our study at the same institution, gave an average serum level of 1420 ng/mL in blood specimens. Our preliminary study demonstrated that this dose caused decreased food consumption and decreased body weight gain in two rabbits fed retinoid for 3 weeks. Previous rabbit studies showed that this dosage range was probably teratogenic and fetotoxic.l* Administration of isotretinoin at 5 mg/kg per day to four rabbits demonstrated a reduction in comedones compared with control rabbits13; this dose was at least two times higher than that used clinically. Two control rabbits were not fed any retinoid. All animals were housed in separate cages. After three weeks of daily isotretinoin therapy, the four rabbits were prepared for wounding. The rabbits were anesthetized with ketamine (5 mg/kg). The ears were prepped with alcohol and betadine, and a 6-mm punchbiopsy instrument was used to create equal diameter wounds in each rabbit ear. The wounds were of uniform thickness down to the perichondrium. Four wounds were created on each ear. Biopsies of the wounds were performed on days 0, (immediately after wounding and adjacent to the wound), 7, 14, and 21. These biopsies were obtained for collagen assay and procollagen mRNA levels after wounding. The rate of epithelialization and complete granulation of tissue was determined visually and by photographs daily. The isotretinoin was continued throughout the wounding period. Histologic examination was performed on day 28. COLLAGEN ASSAYS
Biopsy tissues were placed in Dulbecco's modified Eagle's medium containing Hepes buffer, pH 7.6, 200 U/mL of penicillin, 200 pg/mL of streptomycin, and 25 pg/mL of ascorbic acid. After 1 hour preincuJ Dermatol Surg Oncol 16:12 December 1990
bation, the samples were labeled with [3H]proline ([2,3- 3H]proline; Amersham Corp., Arlington Heights, IL), 430 pCi/lOO mg for 8 hours at 27°C. At the end of the labeling period, the media were removed, cooled to 4"C, and protease inhibitors added to give the following final concentrations: 20 mmol Na ethylenediaminetetro-acetic acid (EDTA), 10 mmol Nethylmaleimide, and 1 mmol phenylmethylsulfonyl fluoride. The tissue samples were then homogenized using a Polytron homogenizer. To quantitate the synthesis of [3H]hydroxyproline and incorporation of total H-radioactivity into proteins, samples labeled with [3H]proline were dialyzed against running tap water to remove the free proline. The dialysates were then hydrolyzed in 6 mol HC1 in sealed tubes at 120 C for 18 hours, and assayed for [3H]hydroxyproline and total [3H]radioactivity using a specific radiochemical method.14 ASSAY OF PROCOLLAGEN MESSENGER RNA LEVELS
For isolation of total RNA, the other half of the biopsy tissue specimens were homogenized in buffer containing 4 mols guanidinium thiocyanate, 5 mmol sodium citrate, 0.5% sarkosyl, 0.1 2-mercaptoethanol, and 0.1% Antifoam A (Sigma Chemical, St. Louis, MO). The homogenates were layered on top of a 2.5 mL cushion of 5.7 mols CsCl, and total RNA was recovered by centrifugation at 35,000 rpm for 16 hours at 15°C using a Beckman SW 40.1 rotor.15.16 The pellet containing RNA was dissolved in sterilized distilled water, and re-precipitated with 70% ethyl alcohol containing 0.4 mols NaCl. The final pellet was dissolved in distilled water, and the concentration of total RNA was determined by the absorbance at 260/280 nmol. For specific procollagen mRNA level determinations, varying amounts of total RNA were dotted onto nitrocellulose filters using a commercial vacuum manifold (Schleicher & Schuell). We immobilized RNA onto the filters by heating under vacuum at 78°C for 90 minutes. The filters were then prehybridized and hybridized with type I and type I11 procollagen specific cDNA probes which were radioactively labeled by nick translation.17-19The prehybridization and hybridization conditions have been previously shown to exclude cross-hybridization between type I and type I11 procollagen cDNA probes and their corresponding mRNA species.12 The specific mRNA [32P]cDNA-,RNA hybrids were visualized by autoradiography using x-ray cassettes equipped with intensifying screens, and the levels of the specific mRNAs quantitated by scanning densitometry at 700 nmol.20 The procollagen mRNA levels were then expressed as absorbance units per mg total RNA. For type I/III procollagen mRNA ratio determinations, the absorb-
ISOTRETINOIN/DERMAL WOUND HEALING
FIGURE 1. A 6-mm punch biopsy instrument was used to
create four equal diameter wounds in each rabbit ear. The wounds were a uniform thickness, down to the perichondrium.
ance values were corrected for the length and specific activity of the cDNA probes, as well as for type I and 111 procollagen chain composition.19 RESULTS Full-thickness wounds in the rabbit ear were used to study the effects of isotretinoin treatment on the rate of dermal wound healing. After the wounding procedure, daily observations suggested that the rate of granulation and epithelialization proceeded at equal rates, both in rabbits treated with isotretinoin and in the controls. Almost complete healing of the wounds was observed by day 28 (Figs. 1 and 2). There was no significant difference observed histologically between the treated and control animals. In order to quantitate biochemically the wound healing processes, the rate of collagen synthesis was determined by incubation of punch biopsy specimens of healing wound sites with radioactive proline, and the synthesis of radioactive hydroxyproline was taken as an index of collagen production. There was a decrease in [3H]hydroxyproline synthesis from day 7 to 14, which has been confirmed in other studies. Prior studies by Jacksonz1 demonstrated that when using this radiochemical method on wound granulation tissue there was a rapid rise in specific activity followed by a slower exponential fall suggesting some oscillation and metabolic pools. Others have also demonstrated that collagen synthesis is maximal prior to day 7, then decreased over time in pig and rat studies.21-23 The results of this study indicated there was no
FIGURE 2. Almost complete healing of the wound was observed by day 28 in the isotretinoin treated rabbits and the control rabbits. Granulation and epithelialization proceeded at equal rates, both in rabbits treated with isotretinoin and in the controls.
significant difference between the synthesis of [3H]hydroxyproline in isotretinoin treated rabbits and the controls (Fig. 3). Furthermore, no significant differences in type I and type 111 procollagen mRNA levels were noted between the treated and untreated animals (Table 1).There was an increase in type I and type I11 procollagen mRNA levels in the treated rab-
DAYS AFTER WOUNDING FIGURE 3. Collagen production was determined by the syn-
thesis of [3H]hydroxyprolineafter incubation of tissue specimens with [3H]proline in vitro. The synthesis of radioactive hydroxyproline was taken as an index of collagen production. The values are means of 3 parallel determinations from 3 separate tissue specimens in each animal, at each point. The data are expressed as [3H]hydroxyproline (mean k SEM).
1Dermatol Surg Oncol 161.2 December
MOY ET AL
TABLE 1 Type I and Type I11 Procollagen Messenger RNA Levels in the Skin of Rabbits Treated with Systemic 13-cis-retinoic acid Type I mRNA*
Type 111 mRNA*
Days after wounding
0 (control) 7 14
213 157 157
415 271 175
215 175 150
400 190 180
The values are expressed as densitometric units per pg total RNA used for slot blot hybridization. The values are mean of four paralleled determinations from four separate tissue specimens at each point in time. Rabbits were treated with 13-cis-retinoic acid (4 mg/kg/day) for 4 weeks and wounds were created in the ears of treated animals and untreated animals left as controls, as described in Materials and Methods. At the time points indicated, tissue samples from wounds were used for isolation of total RNA and the specific type I and type 111 pro-collagen mRNA levels were determined.
bits prior to wounding compared with untreated rabbits. These observations suggest that the systemic treatment of rabbits with isotretinoin at a dose of 4 mg/kg/day does not alter the wound healing of dermal wounds up to 6 weeks of treatment, although it may increase the procollagen mRNA levels prior to wounding. DISCUSSION Previous in vitro studies suggested that isotretinoin is a potent inhibitor of procollagen production in fibroblast cultures.3 At the same time, isotretinoin inhibits the expression of collagenase, the enzyme that initiates collagen degradation in tissues.sThus, a balance between the inhibition of the synthetic pathway and the degredative processes would determine alterations in the tissue deposition of collagen in animals treated with retinoids. In addition, there are a few cases reported that suggest excessive scarring after full face dermabrasions that were attributed to treatment with systemic isotretinoin.1,24,25 However, our experience, as well as that of others treating patients with full face dermabrasion, failed to disclose any problems in wound healing, and there has been no evidence of keloidal scarring among 14 patients taking isotretinoin at the time of surgery and 36 patients who had taken isotretinoin within 6 months prior to surgery (Lask G, David L. No scarring from dermabrasion from Accutane. Schoch Letter, 1985, 35:21; Mandy s, Miami, Florida, Personnal Communication, 1986; Moy RL. Unpublished observation).20.23,24 Systemic isotretinoin did not effect epidermal wound healing in dog dermabrasion wound or in full face dermabrasion patients treated with isotretinoin (Lask and David, Mandy, and Moy).21 This suggests that delayed epidermal wound healing Dermatol Surg Oncol 16:12 December 1990
does not play a role in the development of the keloids. In the present study, no differences in the rate of collagen synthesis or in the expression of genetically distinct collagen genes, as determined by messenger RNA level measurements, could be noted in animals treated with 4 mg/kg per day of isotretinoin. Although other authors suggested that excision and closure should be deferred until the termination of retinoid use because of experimental evidence suggesting inhibition of collagen formation, we demonstrated in this wound healing model that there is no alteration in the wound heaIing process.26
REFERENCES 1. Williams ML, Elias PM. Nature of skin fragility in patients receiving retinoids. Arch Dermatol 117:611-615, 1981. 2. Peck GL. Retinoids in clinical dermatology. Prog Dis Skin 1:227-269, 1981. 3. Oikarinen H, Oikarinen AI, Tan EML, et al. Modulation of procollagen gene expression by retinoids. J Clin Invest 75:15451553, 1984. 4. Abergel RP, Meeker CA, Oikarinen H, Oikarinen Al, Uitto J. Retinoid modulation of connective tissue metabolism in keloid fibroblast cultures. Arch Dermatol 121:632-635, 1985. 5. Puig L, Moreno A, Llistosella E, Nuguera X, deMorgas JM. Granulation tissue proliferation during isotretinoin treatment. lnt J Dermatol 25:191- 193, 1986. 6. Exner JH, Dahod S, Pochi PE. Pyrogenic granuloma-like acne lesions during isotretinoin therapy. Arch Dermatol 119:808811, 1983. 7. Robertson DB, Kubiak E, Gomez EC. Excess granulation tissue responses associated with isotretinoin therapy. Br J Dermatol 111:689-694, 1984. 8. Bauer EA, Seltzer JL, Eisen AZ. Inhibition of collagen degradation enzymes by retinoic acid in vitro. J Amer Acad Dermatol 6~603-607,1982. 9 . Hunt TK, Ehrlich HI', Garcia JA, Duphy JE. Effects of vitamin A on reversing the inhibitory effects of cortisone on healing of open wounds in animals and man. Ann Surg 170:633-637, 1969. 10. Rubenstein R, Roenigk HH, Stegman SJ, Hanke CW. Atypical keloids after dermabrasion of patients taking isotretinoin. J Amer Acad Dermatol 15:280-285, 1986. 11. Stark RB. The rabbit ear as an experimental tool. Plast Reconst Surg 19:28,29, 1957. 12. Kamm J. Toxicology, carcinogenicity and teratogenicity of some orally administered retinoids. J Am Acad Dermatol 6:652-659, 1982. 13. Weinstein M, Mills 0, Berger R. Evaluating the comedolyticactivity of 13-cis-retinoic acid in the rabbit ear assay. Clin Res 36:380A, 1988. 14. Juva K, Prockop D]. Modified procedure for the assay of 3H or "C labelled hydroxyproline. Anal Biochem 15:77- 83, 1966. 15. Chwgwin JM, Przbyla AE, McDonald RJ, Rutter WJ. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18:5294-5299, 1983. 16. Chu ML, Myers JC, Bernard MP, Ding JF, Ramirez F. Cloning and characterization of five overlapping cDNAs specific for human pro l(1)collagen chain. Nucleic Acids Res 10:5925-5934, 1982.
ISOTRETINOIN/DERMAL WOUND HEALING
17. Chu ML, Weil D, Devet W, Bernard MI', Sippola M, Ramirez F. Isolation of cDNA and genomic clones encoding human pro la(II1) collagen. Partial characterization of the 3'-end region of the gene. J Biol Chem 260:4357-4363,1985. 18. Thomas PS. Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose.Proc Natl Acad Sci (USA) 775201 -5205, 1980. 19. Uitto J, Perejda AJ, Abergel RP, Chu ML, Ramirez F. Altered steady-state ratio of type I/III procollagen biosynthesis in cultured fibroblasts. Proc Natl Acad Sci (USA) 82:5935-5939, 1985. 20. Abergel RP, Pizzurro D, Meeker CA, et al. Biochemical composition of the connective tissue in keloids and analysis of collagen metabolism in keloid fibroblast cultures. J Invest Dermatol 84~384-390,1985. 21. Jackson DS. The healing wound as an experimental model for
24. 25. 26.
biochemical studies of connective tissues. In: Wound healing, New York, The MacMillian Company, 1964, pp 30-38. Hebda PA, Kao W, Mertz PM, Eaglstein WH. Collagen biosynthesis and collagen content as a measure of dermal healing in experimental wounds in domestic swine. In: Swine in Biomedical Research, New York, Plenum Press, 1986, pp 303-311. Clore JN, Cohen IK, Diegelmann RF. Quantitative of collagen types I and I11 during wound healing in rat skin. Proc Soc Exp Biol Med 161:337-340, 1979. Roenigk HH, Pmski JB, Robinson JK, Hanke CW. Acne, retinoids and dermabrasion. J Dermatol Surg Oncol 11:396-398, 1985. Zachariae H. Isotretinoin may cause keloid formation after dermabrasion. Br J Dermatol 118~703-706,1988. Dzubow LM, Miller WH. The effect of 13-cis-retinoic acid on wound healing in dogs. J Dermatol Surg Oncol 13265-268, 1986.
Dermatol Surg Oncol 1 6 1 2 December 1990