Pilot Ultrastructural Evaluation of Human Preauricular Skin Before and After High-Energy Pulsed Carbon Dioxide Laser Treatment

Background  Carbon dioxide laser resurfacing has recently come into favor for the treatment of photodamaged skin. While the clinical and histologic effects of high-energy short-pulse carbon dioxide lasers on human skin have been investigated, the ultrastructural effects of these lasers have not been documented. Our objective was to study the ultrastructural effects of a high-energy pulsed carbon dioxide laser on photodamaged human skin.

Observations  Before laser surgery, the ultrastructural changes characteristic of photodamaged skin were evident. Immediately after treatment, there was extensive coagulation necrosis of the epidermis and papillary dermis. Thirty days after treatment, there was no evidence of intercellular or intracellular edema, and ordered differentiation of the epidermal keratinocytes, with a loss of keratinocyte dysplasia, was seen. Increased numbers of desmosomes and tonofibrils were noted. New deposition of collagen was present in the papillary dermis. The ultrastructural findings seen at 90 days after treatment were similar to those seen at 30 days, apart from increased organization of collagen fibers in the papillary dermis.

Conclusions  Treatment with the high-energy pulsed carbon dioxide laser appears to reverse the epidermal and dermal changes of photoaging on an ultrastructural level. These changes appear morphologically to be consistent with previously described clinical and histologic changes following laser resurfacing.

LONG-TERM exposure to UV radiation produces characteristic clinical and histologic changes in the skin.^1-8 Actinically damaged skin tends to develop fine and coarse wrinkling, mottled pigmentation, telangiectases, loss of elasticity, and premalignant and malignant lesions. Histologically, epidermal atrophy, keratinocyte dysplasia, solar elastosis, and increased melanocyte activity are generally evident. The ultrastructural changes seen in photodamaged skin, including epidermal spongiosis, degeneration of basal and suprabasal keratinocytes, vacuolization of the dermoepidermal junction, disorganization of collagen bundles and collagen and elastic fiber degradation in the papillary dermis, and inactive fibroblasts with scant cytoplasm and few organelles, have also been described.^8-14

Treatment with α-hydroxy acids and topical tretinoin has been shown to improve the clinical and histologic changes associated with photoaging.^15-24 Reversal of the ultrastructural changes seen in photodamaged skin with the use of these topical agents has also been demonstrated.^{8,15,20,23,24} α-Hydroxy acid–treated keratinocytes appear to be connected by fewer desmosomes and exhibit less clumping of tonofibrils. Both α-hydroxy acids and topical tretinoin seem to increase the numbers of anchoring fibrils at the dermoepidermal junction of treated photoaged skin.^15,20,23 Normalization of the structure and organization of papillary dermal collagen, reduction in the amount of degenerated microfibrillar material, and increased activity and numbers of fibroblasts have also been described after treatment with topical tretinoin.^20,24,25

Dramatic clinical and histologic improvement in photoaged skin may also be produced with medium-depth and deep chemical peeling agents and dermabrasion.^26-37 When properly performed, both procedures result in architectural and cytologic normalization of the epidermis, as well as an expanded papillary and reticular dermis composed of dense, parallel arrays of collagen bundles, otherwise known as the dermal repair zone. Nelson et al³⁴ were the first to document the ultrastructural changes produced by medium-depth chemical peels on photodamaged facial skin. Three months after a single 35% trichloroacetic acid peel, markedly decreased intracytoplasmic vacuoles within and between keratinocytes were seen, as were activated fibroblasts with increased cytoplasm and organelles and abundant deposition of new collagen with clearly defined cross striations arranged in an orderly parallel fashion. Decreased solar elastosis was also noted within the papillary dermis.

Carbon dioxide laser resurfacing has recently come into favor as a means of treating photodamaged human skin safely and effectively.^38-53 Clinical improvement in facial rhytides and photodamage with the new generation of high-energy pulsed carbon dioxide lasers has been well documented.^39-48,51,52 The precise control over the extent of tissue vaporization results in minimization of thermal damage to the skin, thereby reducing the potential risks of scarring and hyperpigmentation, while maximizing therapeutic efficacy.^39,49-52

The histologic changes seen after pulsed laser resurfacing have recently been detailed.^50,51 Extensive epidermal necrosis and coagulative changes in the superficial papillary dermis are seen 24 hours after laser administration. Partial or complete reepithelialization is usually seen by day 3, although evidence of papillary dermal collagen damage is still seen. By 90 days after laser treatment, the epidermis is completely intact and dysplasia is absent. A papillary dermal repair zone composed of dense compact collagen bundles aligned in a parallel fashion with the epidermal surface is seen. Decreased numbers of thin elastic fibers oriented perpendicular to the dermoepidermal junction are noted in the superficial papillary dermis, with thicker, more haphazardly arranged fibers in the upper reticular dermis. To our knowledge, the ultrastructural changes seen after treatment with the pulsed carbon dioxide lasers have not yet been described. We sought to determine what changes could be seen not only in the epidermis and dermis, but also at the dermoepidermal junction, using a recent study by Cotton et al⁵⁰ as a model for our work.

Preauricular skin samples were obtained from 4 patients (3 men and 1 woman), aged 65 to 80 years, who were scheduled to undergo elective laser resurfacing for multiple actinic keratoses. Informed consent was obtained after the nature of the study had been fully explained. There was no preoperative protocol for the patients. The patients did not apply α-hydroxy acids or tretinoin cream before laser treatment.

A single 2.0 × 2.0-cm preauricular (sun-exposed) area of skin on each of the patients was treated with a high-energy pulsed carbon dioxide laser (Ultrapulse 5000, Coherent Laser Corporation, Palo Alto, Calif) for a total of 2 passes as follows: A collimated handpiece was used to deliver a 3-mm-spot-size beam. The laser was set at 500 mJ and 4 to 6 W. Laser pulses were placed adjacent to one another with less than 10% overlap to minimize char formation. After the first pass with the laser, the residual coagulated skin was wiped away with moistened gauze. The second pass was performed in the same fashion as the first. Immediately after treatment, a topical antibiotic (bacitracin ointment) was applied, followed by a dry sterile dressing. The patient was instructed to remove the dressing after 24 hours and to clean the wound twice daily thereafter with hydrogen peroxide, followed by reapplication of bacitracin ointment and a dry sterile adhesive strip (Band-Aid).

One biopsy specimen was obtained from the treated preauricular area of each patient at each time point, ie, before laser treatment, immediately after laser treatment, and at days 30 and 90 after laser treatment, for a total of 4 biopsy specimens per patient. The biopsy site was anesthetized with 1% lidocaine with 1:100000 epinephrine, and a 1.25-mm punch biopsy specimen was obtained using a metal hair transplant punch. After the punch had been inserted to the level of the subcutaneous fat, the punch was withdrawn, leaving a cylindrical column of tissue attached by a pedicle. The pedicle was cut with curved iris scissors, and hemostasis was achieved with pressure or 35% aluminum chloride solution. Each specimen was divided into 2 or 3 small pieces before fixation. Tissue specimens were fixed in 1.6% glutaraldehyde in 0.1-mol/L Sorensen buffer at a pH of 7.3 (98 mL of 27.22-g potassium phosphate per liter of distilled water plus 102 mL of 28.39-g sodium phosphate per liter of distilled water) for 20 minutes. The fixative was then replaced by 0.1-mol/L Sorensen buffer at a pH of 7.3 for 3 rinses of 20 minutes each. The samples were placed into microcentrifuge tubes with positive sealing lids (snap caps) (Eppendorf tubes, Brinkmann, Westbury, NY) filled with Sorensen buffer at a pH of 7.3 and mailed (via Federal Express) to France on ice blocks. The tissue was then postfixed in 1% osmium tetroxide, dehydrated through a graded ethanol series, and embedded in epoxy resin (Epon). Thin sections were cut on an ultramicrotome (Nova microtome, LKB-Produktter AB, Bromma, Sweden), double stained with uranyl acetate–lead citrate, and observed in a transmission electron microscope (Elmiskop I, Siemens Corp, Iselin, NJ).

Three patients (all men) completed the study. One patient (a woman) did not complete the study, as she did not undergo a 90-day biopsy; however, her pretreatment, posttreatment, and 30-day biopsy results were included in our study.

Prior to laser surgery, ultrastructural findings characteristic of photodamaged skin were noted. There was marked epidermal disorganization, as well as prominent intercellular edema with loss of desmosomal connections and intracellular vacuolization of epidermal keratinocytes (Figure 1). On higher magnification, epidermal atypia was clearly evident (Figure 2). Prominent nucleoli were seen, surrounded by a nucleoplasmic area containing highly dispersed chromatin, reflective of heightened nuclear synthetic activity (Figure 2). Condensed chromatin clumps were rare. Sparse tonofibrils were found within the keratinocytes of the stratum spinosum, and the cytoplasm had a cytolytic appearance, reflective of organellar degradation (Figure 2). The dermoepidermal junction exhibited a flattened contour, as relatively few "footlike" processes of the basal cells were seen (Figure 3). Melanocytes were rarely identified. In the superficial dermis, amorphous material consistent with degraded collagen and elastic tissue was seen (Figure 3). Disorganization of the papillary dermal collagen was present (Figure 3). Inactive fibroblasts with relatively few organelles were occasionally seen.

Immediately after laser treatment, there was extensive epidermal coagulation necrosis as well as coagulative change in the superficial papillary dermis (Figure 4). The reticular dermis did not exhibit coagulation necrosis.

Thirty days after treatment, marked changes were seen in both the dermis and the epidermis. The keratinocytes were arrayed in a more organized fashion, and widened intercellular spaces were no longer evident (Figure 5). On higher magnification, innumerable desmosomes filled the intercellular spaces between keratinocytes (Figure 5, inset). Numerous, tightly organized bundles of tonofibrils were uniformly distributed throughout the cytoplasm of all keratinocytes (Figure 6). Atypical keratinocytes were no longer evident, and keratinocyte nuclei exhibited abundant condensed chromatin and less prominent nucleoli. Intracytoplasmic vacuoles were no longer seen. The basal cells were precisely aligned along the dermoepidermal junction, with convolution of the junction due to the increased numbers of footlike processes of the basal cells (Figure 7). Scattered normal-appearing melanocytes were identified in the suprabasal and basal layers. The basement membrane zone exhibited newly prominent anchoring fibrils (Figure 7, inset). New deposition of collagen fibers organized in parallel arrays was present in the papillary dermis (Figure 7). A decreased amount of amorphous degraded material was seen in the papillary dermis. Fibroblasts exhibited increased numbers of organelles, most notably rough endoplasmic reticulum and mitochondria.

By 90 days after treatment, the epidermal keratinocytes were still arrayed in an organized fashion (Figure 8). Their intercellular spaces were not as tightly packed with desmosomes, although bundles of tonofibrils were uniformly distributed within their cytoplasm. Intracytoplasmic vacuolization of the keratinocytes was not seen. The basal cells were precisely aligned at the dermoepidermal junction, and prominent footlike processes were again noted. Occasional normal-appearing melanocytes were seen. Prominent anchoring fibrils were identified in the basement membrane zone, comparable in quantity to those seen at 30 days. An even more highly organized network of collagen and elastic fibers was present in the papillary dermis, and amorphous degraded material was absent (Figure 8). Fibroblasts with increased numbers of organelles were noted in the dermis.

In recent years, high-energy short-pulse carbon dioxide lasers have grown in popularity for the treatment of photoaged facial skin, both for the treatment of actinic damage and rhytides. The clinical and histologic effects of these lasers have been studied, and it has been shown that the histologic effects of laser resurfacing are microscopically similar to those of phenol peeling, such that, at 90 days after laser treatment, epidermal atypia and dysplasia are corrected and epidermal polarity is restored, the epidermis being then "indistinguishable from that of younger, normal skin."⁵¹ The presence of a subepidermal repair zone consisting of new subepidermal collagen at 3 months after laser treatment, comparable to that seen after medium-depth chemical peels or dermabrasion, has also been described.^50,51 The ultrastructural changes seen in our small group of patients at 30 and 90 days after laser resurfacing appear to correlate with these histologic findings.

It is known that ablation of photodamaged epidermis and upper dermis, whether by chemical or physical means, allows reepithelialization from deeper, less photodamaged cells, resulting in the restored structural and functional integrity of epidermal keratinocytes.^26-38,50 The increased number and organization of the tonofibrils in epidermal keratinocytes that we have seen at 30 and 90 days after laser treatment appears therefore to be significant, as these findings correlate with the normalization of keratinocyte differentiation from the basal layer to the stratum corneum seen histologically after laser resurfacing. Additionally, the loss of intracellular and intercellular epidermal vacuolization and the presence of folded and tightly apposed intercellular spaces of adjacent keratinocytes studded at intervals with innumerable desmosomes 30 and 90 days after treatment appear to be significant, as these findings are characteristically seen in normal squamous epithelium.⁵⁴

Clinical studies have revealed a measurable and reproducible decrease in fine wrinkling and improvement in skin texture after laser resurfacing, but the mechanism by which these changes occur is not yet clear.^39-48 Fitzpatrick et al³⁹ have postulated that, when the carbon dioxide laser interacts with tissue, 3 zones of tissue damage are produced: a vaporized zone, a zone of irreversible thermal necrosis, and a zone of reversible thermal damage, in which collagen shrinkage is thought to take place. It has been documented that thermal damage to collagen itself results in shrinkage, but it is not clear how great a role this shrinkage plays in generating clinical improvement in wrinkles.^55-57 It is also thought that repair of this layer during healing may account for tightening of sagging skin and improvement of creases.^39,46,47 The clinically evident tightening of sagging skin after laser surfacing may be in part due to the formation of new collagen, the decrease in the amount of amorphous debris in the papillary dermis, and the increased activity and number of fibroblasts that have been documented histologically and ultrastructurally after laser and other resurfacing procedures.^26-36,38,50

Nelson et al³⁴ have in fact found that collagen's striation periodicity is reduced after medium-depth chemical peels, resulting in a more compact architecture, and that the diameter of individual fibrils is more variable, consistent with recent production of collagen by activated fibroblasts. Precise quantitative studies will be required to substantiate whether compression of the collagen bundles occurs immediately after laser resurfacing and whether the striation periodicity of the collagen is immediately changed after treatment. If, in fact, the striation periodicity of the collagen in the papillary dermis is reduced at 30 and 90 days after treatment, this may partially explain the clinical perception of tighter, smoother skin in patients treated with resurfacing lasers. The new deposition of dermal collagen that we have seen ultrastructurally after laser resurfacing seems to correspond to the papillary dermal repair zone described by others, but will need to be further characterized, both ultrastructurally and biochemically. While it does appear that the papillary dermal collagen is more organized after 90 days than after 30 days after laser resurfacing, and that individual collagen fibrils may exhibit more variability in diameter at 30 days than at 90 days after laser resurfacing, it is important to realize that these are subjective impressions and that these findings must be documented both qualitatively and quantitatively when future studies of larger numbers of patients are performed.

It has been postulated that increased numbers of anchoring fibrils may help to produce increased adherence of the epidermis to the dermis, resulting in a pulling, or "tenting," that may decrease wrinkling, as seen in patients treated with tretinoin and α-hydroxy acids.^15,23,25 While increased convolution of the dermoepidermal junction and increased numbers of anchoring fibrils may play a role in the increased smoothness and tautness of the skin seen clinically in patients treated with laser resurfacing, further studies will be needed to verify and quantitate any true increase in the number of anchoring fibrils. It is possible that such a phenomenon may somehow also contribute to the clinical improvement in superficial rhytides after laser resurfacing.

It is impossible to draw any conclusions about the ultrastructural effects of laser resurfacing on melanocytes at this time. Very few melanocytes were seen in our pretreatment specimens, but it is probable that this paucity of melanocytes was a result of sampling error. While normal-appearing melanocytes were seen in the basal layer of the epidermis at 30 and 90 days after treatment, it would be premature to make any generalizations regarding melanocyte activity or number at this time. This will be an important area of future study, especially given the recent reports of delayed hypopigmentation occurring after laser resurfacing.⁵⁸

Despite the small size of our study, it appears that treatment with the high-energy pulsed carbon dioxide laser appears to reverse epidermal and dermal photoaging changes on an ultrastructural level. These changes appear morphologically to be consistent with previously described clinical and histologic changes following laser resurfacing. Further studies will be necessary to correlate more precisely the clinical, histologic, and ultrastructural changes that result from laser resurfacing. Such studies will undoubtedly provide us with a wealth of useful information.

Accepted for publication January 5, 1998.

We would like to acknowledge the financial support of the Center for Surgical Dermatology, Lutherville, Md, during the early stages of this study.

Presented in poster form at the 24th Annual Clinical and Scientific Meeting of the American Society for Dermatologic Surgery, Boston, Mass, May 8-11, 1997.

We thank Craig Thomas, MS, for his assistance in the preparation of the manuscript for this article.

Reprints: Désirée Ratner, MD, Department of Dermatology, Columbia-Presbyterian Medical Center, 161 Fort Washington Ave, New York, NY 10032.