|
|
| |||||||||||||||||||||||||||||||||||||||
Background Topical tretinoin (retinoic acid) modifies fine wrinkles and certain other features of human skin damaged by exposure to the sun (photodamage), but histologic changes do not account for this improvement. In mice with photodamage induced by ultraviolet light, effacement of fine wrinkles by tretinoin is correlated with dermal collagen synthesis but not with histologic changes. We investigated whether collagen synthesis was reduced in photodamaged human skin and, if so, whether it could be restored by treatment with topical tretinoin.
Methods Biopsies of photodamaged skin from the extensor aspect of the forearm and skin from the buttocks, which had been protected from the sun, were performed on 26 healthy subjects. In addition, 29 patients with photodamaged skin were treated for 10 to 12 months with a daily application of 0.1 percent tretinoin cream (15 patients) or vehicle cream (14 patients). Skin-biopsy specimens obtained at base line and after treatment were assessed immunohistologically for evidence of dermal collagen I formation (collagen synthesis).
Results Collagen I formation was 56 percent less in the papillary dermis of photodamaged skin than in skin protected from the sun (P<0.001) and was correlated with the clinical severity of photodamage (r = -0.58, P = 0.002). Treatment of photodamaged skin with tretinoin produced an 80 percent increase in collagen I formation, as compared with a 14 percent decrease in collagen formation with the use of vehicle alone (P = 0.006).
Conclusions The formation of collagen I is significantly decreased in photodamaged human skin, and this process is partly restored by treatment with tretinoin.
Mice exposed to ultraviolet irradiation simulating the sun acquire fine wrinkles similar to those seen in humans with photodamage10,11. When such mice are treated with topical tretinoin, effacement of the wrinkles occurs in association with the appearance of a subepidermal repair zone detectable by routine light microscopy. This repair zone was thought to represent the deposition of new collagen,12,13 and investigators postulated that it was responsible for the clinical improvement of wrinkles in mice. It was also assumed that a similar repair zone would prove to be the mechanism of wrinkle effacement in photodamaged human skin treated with tretinoin.
Unfortunately, despite early indications to the contrary,1 microscopical examination of the immediate subepidermal area in several large series of patients who were successfully treated with tretinoin for photodamage revealed scant evidence of such a repair zone2,5,14. Epidermal hyperproliferation was prominent in these studies but could not have been the cause of wrinkle effacement, since the epidermal changes occurred early, before clinical improvement, and since clinical improvement continued after the epidermis had returned to an almost normal thickness with long-term tretinoin treatment15. These observations suggest that some factor or factors other than the presence of a repair zone or epidermal hyperproliferation instigate clinical improvement.
Recently, Chen et al.16 demonstrated that wrinkle effacement in tretinoin-treated mice with photodamage was correlated with increased collagen synthesis but not with the development of a histologically detectable repair zone or epidermal hyperproliferation. The synthesis of collagen III is known to be increased in the subepidermal region of tretinoin-treated mice with photodamage17.
Collagen I predominates in human dermis, accounting for 85 percent of the total, whereas collagen III accounts for only 10 percent18. Thus, studies of collagen in human skin are more easily accomplished if directed at collagen I. The biosynthesis of collagen I in skin begins with the formation of procollagen I within dermal fibroblasts19,20 (Figure 1). After its secretion from fibroblasts, procollagen I is enzymatically cleaved of its aminopropeptide and carboxypropeptide in a one-to-one ratio19,21; the presence of propeptide provides an index of collagen I synthesis (Figure 1) 21. Therefore, an antibody that can recognize the aminopropeptide portion of procollagen I or its distal derivatives22 will both identify precursors of collagen I and provide an indirect measure of collagen I formation (Figure 1). Using such an antibody we determined whether the formation of collagen I is reduced in photodamaged human skin and is increased by topical tretinoin, as in mice. The finding of increased collagen I formation in photodamaged human skin treated with tretinoin suggests that tretinoin promotes clinical improvement by repairing dermal collagen.
|
Comparison of Photodamaged Skin and Skin Protected from the Sun
The overall severity of photodamage on the extensor aspect of the forearm was assessed in 26 healthy white subjects (9 men and 17 women; age range, 36 to 82 years; mean age, 56) with an established 10-point scale in which a grade of 0 indicates clinically detectable photodamage and a grade of 9 the most severe photodamage4,5. This scale is a composite score of the grades for wrinkles, surface roughness, actinic lentigines, and mottled hyperpigmentation4,5. In all 26 patients 2- or 4-mm punch-biopsy specimens were procured under local anesthesia from skin from the buttocks, which had been protected from the sun, and photodamaged skin from the extensor aspect of the forearm. In a subgroup of 11 patients, an additional biopsy specimen was obtained from underarm skin, which had been protected from the sun. This specimen was used to confirm that perceived differences in the extent of collagen I immunostaining between the forearm and buttock samples were due to differences in the degree of exposure to the sun rather than in the anatomical site.
Tretinoin and Vehicle Treatment of Photodamaged Skin
Twenty-nine white patients with photodamage (6 men and 23 women; mean age, 63 years) applied either 0.1 percent tretinoin cream (Retin-A, Ortho Pharmaceutical, Raritan, N.J.) or color-matched vehicle cream once daily to their forearms in a double-blind fashion. Patients were randomly assigned to treatment groups according to a computer-generated code. Eleven patients from a study of actinic dyspigmentation23 (7 assigned to tretinoin and 4 assigned to vehicle) were treated for 10 months, and 18 additional patients (8 assigned to tretinoin and 10 assigned to vehicle) were treated for 12 months. None of these patients were from the group of 26 subjects described above. Sunscreen with a sun-protection factor of 15 was provided and worn during the day, and the patients were advised to avoid strong sunlight. At base line and after treatment was completed, 2- or 4-mm punch-biopsy specimens were taken from all 29 patients while they were under local anesthesia. Post-treatment biopsy specimens were obtained close to the base-line biopsy sites but far enough away to avoid potential scar tissue in these areas.
Biopsy specimens from both studies were immediately oriented in Optimal Cutting Temperature embedding medium (Miles Laboratories, Naperville, Ill.), frozen in liquid nitrogen, and stored at -70 °C until use. All subjects gave written informed consent, and the studies were approved by the University of Michigan Medical Center Institutional Review Board.
Immunohistologic Analysis
For the immunohistologic analysis, 5-microm cryostat sections were placed on triple-welled microslides (Carlson Scientific, Peotone, Ill.) coated with poly-l-lysine transparent microslide adhesive (Sigma Chemical, St. Louis). The sections were fixed in chilled acetone and then incubated with a well-characterized mouse monoclonal IgG1 antibody (SP1.D8) that recognizes a 19-amino-acid portion of the aminopropeptide-cleavage site of human procollagen I22. This antibody recognizes the aminopropeptide whether it is cleaved from or still a part of procollagen or pN collagen I molecules -- i.e., collagen I precursors -- but does not recognize the triple helix of collagen I or the carboxypropeptide fragment (Figure 1). In this study, we use the term "collagen I immunostaining" to refer to tissue staining observed with this antibody. The hybridoma was obtained from the Developmental Studies Hybridoma Bank maintained by the Department of Pharmacology and Molecular Sciences, Johns Hopkins University School of Medicine, Baltimore, and the Department of Biology, University of Iowa, Iowa City. Appropriately diluted mouse IgG1 was used as a control. The samples were processed for staining with an immunoperoxidase technique (Vectastain ABC Kit, Vector Laboratories, Burlingame, Calif.), with 3-amino-9-ethyl carbazole (Sigma) as the chromogen, counterstained with hematoxylin (Lerner Laboratories, Pittsburgh), and wet-mounted with Aquamount (Lerner Laboratories).
Quantification of Immunostaining
The intensity and extent of staining were assessed with a six-point scale in which a score of 0 indicated no staining, a score of 1 minimal staining, a score of 2 low-to-moderate levels of staining, a score of 3 moderate staining, a score of 4 high levels of staining, and a score of 5 maximal staining. The patterns of collagen I immunostaining were extracellular, in a papillary dermal band (a narrow zone immediately beneath the epidermis), and intracellular, within fibroblasts. Four high-power fields from each section were graded under light microscopy. All slides were coded and randomized so that the persons grading them were unaware of the anatomical site from which the samples had been taken and whether the samples had been obtained before or after treatment with tretinoin or vehicle.
Statistical Analysis
The paired t-test was used to compare staining patterns in skin protected from the sun (buttocks) and photodamaged skin (forearm). The differences in the post-treatment change in extracellular and fibroblast collagen I immunostaining between tretinoin-treated and vehicle-treated skin were assessed with the two-sample t-test. The relation between the clinical severity of photodamage and the extent of extracellular collagen I immunostaining was assessed with the Pearson product-moment correlation analysis.
All P values are two-sided. Percentages were determined before rounding of the means. Summary data are expressed as means ±SE. Data were analyzed with the Michigan Interactive Data Analysis System, a statistical software package developed by the Center for Statistical Consultation and Research at the University of Michigan.
Results
Comparison of Photodamaged Skin and Skin Protected from the Sun
The average clinical severity of photodamage on the extensor aspect of the forearm was 5.4 ±0.3 (on the scale used, a grade of 9 indicated the most severe photodamage). Collagen I immunostaining in buttock skin protected from the sun could be categorized as a combination of extracellular, in a broad band in the upper papillary (superficial) dermis, and intracellular, within fibroblasts in the papillary and reticular (deep) dermis. Comparison of photodamaged skin from the extensor aspect of the forearm with skin from the buttocks, which had been protected from the sun, demonstrated a 56 percent reduction in extracellular collagen I immunostaining within photodamaged papillary dermis, with minimal staining (1.2 ±0.2) in photodamaged skin and moderate staining (2.7 ±0.3) in protected skin (P<0.001) (Figure 2). The extent of fibroblast immunostaining for collagen I within papillary or reticular dermis was not significantly different between protected skin and photodamaged skin (0.8 ±0.2 and 0.6 ±0.1, respectively; P = 0.37). There was a significant correlation between the clinical severity of photodamage and the amount of extracellular collagen I immunostaining in the papillary dermis (r = -0.58, P = 0.002), implying that the more severe the photodamage the greater the reduction in collagen I formation (Figure 3). Multiple linear regression modeling showed that the partial correlation of the clinical severity of photodamage and extracellular collagen I immunostaining, after adjustment for patient age, was significant (partial r = -0.55, P = 0.005). Conversely, however, the partial correlation of age was not significant after adjustment for the clinical severity of photodamage (partial r = 0.22, P = 0.3), suggesting that the degree of photodamage is a better predictor of collagen I formation than chronologic age.
|
|
Comparison of Tretinoin and Vehicle Treatment of Photodamaged Skin
Treatment of photodamaged skin with tretinoin produced a 119 percent increase in collagen I immunostaining in fibroblasts (from 1.3 ±0.3 at base line to 2.9 ±0.3 after treatment), as compared with an 18 percent decrease (from 1.5 ±0.2 to 1.3 ±0.3) with vehicle treatment (P<0.001) (Figure 4). Treatment with tretinoin also increased extracellular collagen I immunostaining within the papillary dermis by 80 percent (from 1.1 ±0.2 at base line to 2.0 ±0.3 after treatment), as compared with a 14 percent decrease (from 1.6 ±0.3 to 1.3 ±0.3) with vehicle treatment (P = 0.006) (Figure 5).
|
|
This study demonstrates that long-term exposure of skin to the sun is correlated with a reduction in the formation of collagen I within the papillary dermis. The relative reduction in collagen I formation in the extensor aspect of the forearm as compared with that in underarm skin protected from the sun is evidence that the observed decrease is probably a result of photodamage and is not due to the anatomical site. In addition to decreased formation of collagen I in photodamaged skin, there may also be increased extracellular breakdown of collagen I by collagenase released by inflammatory cells, mast cells, fibroblasts, or keratinocytes24,25,26. The significant correlation of reduced levels of extracellular collagen I immunostaining with the severity of photodamage and the known property of collagen to provide cutaneous tensile strength and resiliency raise the possibility that the reduction in collagen I formation may contribute to the production of wrinkles.
We detected increased collagen I immunostaining after treatment with tretinoin, both within fibroblasts (representing procollagen I) and extracellularly (representing collagen I precursors) in the papillary dermis. These observations point to a tretinoin-induced increase in collagen I formation in photodamaged human dermis and suggest that the clinical improvement reported in previous studies may result from this increase. We did not attempt to correlate clinical improvement directly with the formation of collagen I, since the biopsy specimens were obtained from the arm, a site in which it is difficult to assess improvements in wrinkles reliably. However, extensive evidence from multicenter clinical trials indicates that 10 to 12 months of treatment with tretinoin will improve facial wrinkles,1,2,3,4,5 and thus tretinoin will probably efface forearm wrinkles in the same period.
Although tretinoin can directly induce collagen synthesis,27,28 it can also decrease the breakdown of collagen by inducing tissue inhibitors of collagenase29,30. These two mechanisms (i.e., increased synthesis and reduced breakdown) working in concert would result in the increased accumulation of collagen during tretinoin treatment. However, since collagenase degrades mainly mature collagen and not its precursors31 and since the degree of immunostaining in this study reflected the extent of collagen formation, we were unable to determine whether tretinoin inhibited collagenase. The only other in vivo evidence that tretinoin affects collagen in humans is an ultrastructurally demonstrated increase in anchoring fibrils (collagen VII) at the epidermal-dermal junction after four months of treatment32. Anchoring fibrils play an important part in ensuring the adherence of epidermis to dermis,33 and any reduction in these fibrils, as may occur with photodamage, may also contribute to skin wrinkling.
The results of the current study are an indication that in humans, as in mice, reduced dermal collagen formation in photodamaged skin is partly restored by treatment with tretinoin. The design of this study did not permit a direct correlation of collagen I formation with tretinoin-induced clinical improvement. However, it is reasonable to presume that the two may be correlated, as has been reported in mice16. Thus, the ability of topical tretinoin to reduce the effects of photodamage clinically is not simply cosmetic but rather is probably based on the active repair of dermal collagen.
Supported by a grant from the R.W. Johnson Pharmaceutical Research Institute, Raritan, N.J., by the Babcock Dermatologic Endowment, Ann Arbor, Mich., and by a contract (N01-HD-6-2915) from the National Institute of Child Health and Human Development.
During part of the study Dr. Voorhees served as a consultant to the Johnson and Johnson Corporation (of which the Ortho Pharmaceutical Corporation, manufacturer of the study drug, is a subsidiary).
We are indebted to Elyse S. Rafal, M.D., for supplying some of the tretinoin-treated and vehicle-treated tissue and to Dale Yessian for assistance in the preparation of the manuscript.
Source Information
From the Dermatopharmacology Unit, Department of Dermatology, University of Michigan Medical Center, Ann Arbor. Presented in part at the annual meeting of the Society for Investigative Dermatology, Washington, D.C., April 28-30, 1993.
Address reprint requests to Dr. Voorhees at the Department of Dermatology, University of Michigan Medical Center, 1910 Taubman Ctr., 1500 E. Medical Ctr. Dr., Ann Arbor, MI 48109-0314.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Effect of Tretinoin on Collagen Formation in Photodamaged Skin
Whitmore S. E., Krupsky M., Goldstein R. H., Griffiths C. E.M., Hamilton T. A., Voorhees J. J.
Extract |
Full Text
N Engl J Med 1993;
329:2038-2039, Dec 30, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
