Comparative Clinical Trial of 2 Carbon Dioxide Resurfacing Lasers With Varying Pulse Durations: 100 Microseconds vs 1 Millisecond

Objectives  To compare the clinical and histological effects of 2 carbon dioxide lasers with different pulse durations and to evaluate the effect of carbon dioxide laser pulse duration on postprocedure erythema, wound healing, and efficacy of wrinkle treatment.

Design  Prospective, randomized, comparative clinical trial.

Setting  A university-affiliated hospital-based laser center.

Patients  Thirty-five patients with facial wrinkles were enrolled in the study. Treatment sites included 15 perioral, 14 periorbital areas, and 6 full face.

Intervention  A 2-sided comparison was performed. One side of the study site was treated with the TruPulse laser (Tissue Technologies, Palomar Medical Products Inc, Lexington, Mass). The other side of the study site was treated with the UltraPulse 5000 laser (Coherent Medical Inc, Palo Alto, Calif). The 2 sides were treated to equivalent tissue effects rather than maintaining the number of passes.

Main Outcome Measures  Photographs of the treatment areas at baseline, week 1, week 2, month 2, and month 6 were evaluated by a 5-member panel for degree of erythema, amount of edema, and percentage of wrinkle improvement. Silicon skin casts for profilometry measurements before and after the treatment were compared. To evaluate skin shrinkage, surface area before and after treatment of square tattoos on both cheeks of the full-face patients were computed using a digital imaging system. Histological sections before and after the procedure were analyzed.

Results  At week 1, 75% of the patients had more erythema on the UltraPulse than TruPulse sides. The difference in erythema (TruPulse less than UltraPulse) between the 2 treatment sides was clinically mild yet statistically significant for weeks 1 (P=.05) and 2 (P=.05). Although observed results favored the UltraPulse over the TruPulse, the difference in efficacy between the 2 lasers did not reach statistical significance.

Conclusions  Compared with the longer pulse–duration carbon dioxide laser, the shorter pulse–duration carbon dioxide laser, used with higher energy and more passes, caused slightly less erythema while maintaining efficacy. The longer pulse–duration laser required lower energy and fewer number of passes to achieve an equivalent depth of ablation, level of residual thermal damage, and degree of efficacy. The shorter TruPulse allows for more superficial tissue damage per pass and therefore is best suited for situations requiring superficial or more controlled ablation. The longer UltraPulse achieves a desirable depth of tissue damage with fewer passes. The data did not support the long-term presence of tissue collagen shrinkage in the treated areas.

PULSED OR SCANNED carbon dioxide lasers with high-peak powers and short laser-tissue interaction times can safely and effectively treat cutaneous photodamage, including mild to moderate wrinkles.^1,2 Carbon dioxide laser energy, at 10600 nm, is strongly absorbed by water and in this way heats and destroys tissue. Continuous wave carbon dioxide lasers not only ablate tissue but also cause charring and a surrounding zone of thermal damage 0.2- to 1-mm thick.³ The newer pulsed or scanned lasers have high-peak power that can ablate tissue without leaving behind a charred surface and short pulse durations that limit the amount of residual thermal damage. Therefore, only a thin layer of tissue (30-80 µm) is removed with each pass of the laser. Besides ablation and smoothing of uneven skin texture, new collagen formation and shrinkage of collagen have been proposed as mechanisms for decreasing the depth of wrinkles.

Following a carbon dioxide laser resurfacing procedure, some patients experience an extended period, 6 to 12 weeks, of postprocedure erythema. Therefore, the amount of residual tissue thermal damage may impact the degree of erythema. A laser with a shorter pulse duration that further limits the heat absorbed by the tissue may decrease postprocedure erythema. The objective of this study was to compare the clinical and histological effects of 2 carbon dioxide lasers with different pulse durations and to determine if a laser with a shorter pulse duration results in less postprocedure erythema while maintaining efficacy of the treatment.

Thirty-five patients were enrolled in the study to receive laser treatment for facial sun-damaged skin. Inclusion criteria were age 18 to 90 years and the desire to have facial wrinkles improved. Exclusion criteria were having an active infection, an immunocompromised condition, or an anticoagulation disorder. Patients were treated in 1 of 3 anatomical locations: perioral, periorbital, or full face. Six patients were chosen for a full-face treatment so that serial digital images could be taken of their cheeks, as described later. Our study was approved by the Subcommittee of Human Studies at Massachusetts General Hospital, Boston. Informed consent was obtained from all study participants after the objectives, design, and risks of the study had been explained.

Patient pretreatment evaluations included a medical evaluation, review of enrollment criteria, rating of wrinkle severity, and determination of skin type (Fitzpatrick scale I-III). Rhytides were evaluated by the following scale: 1, superficial fine lines with minimal textural changes; 2, clearly visible, sharply defined lines with moderate textural change; and 3, clearly visible, sharply defined lines with more severe textural change and redundant skin with creases and folds.

After enrollment in the study, patients began a pretreatment regimen with 0.025% tretinoin cream (Retin-A, Ortho Pharmaceutical Corporation, Raritan, NJ) and 3% hydroquinone solution (Melanex, Neutrogena Dermatologics, Los Angeles, Calif) for at least 2 weeks and 250 mg of dicloxacillin 4 times a day and 125 mg of famciclovir twice a day for 7 days starting 1 day before the procedure. Patients were advised to avoid exposure to sunlight before and after the treatment.

The lasers compared in this study were the TruPulse (Tissue Technologies, Palomar Medical Products Inc, Lexington, Mass) and the UltraPulse 5000 (Coherent Medical Inc, Palo Alto, Calif). The TruPulse has a pulse duration of 60 to 100 microseconds, a square spot size of 3 mm, and a maximum pulse energy of 500 mJ. The UltraPulse has a pulse duration of 600 microseconds to 1 millisecond, a collimated beam with a circular 3-mm spot size, and a maximum pulse energy of 500 mJ.

For each patient, one side of the study area was treated with the TruPulse and the other side was treated with the UltraPulse. Laser assignments alternated from left to right side of the treatment site for each consecutive patient and were independent of severity of wrinkles. Both laser treatments were completed at the same session. Two physicians were involved in the patient treatments.

Treatment settings are described in Table 1. The number of passes and laser fluence varied depending on the laser used, severity of wrinkles, and anatomical location. Therefore, the same final tissue effect was obtained on both sides of the treatment site, yet fluence and number of passes were not held constant. Tissue response to the laser, reduction of wrinkles, and color of the tissue were all variables used to determine the final tissue effect. For all treatments, the TruPulse laser was used at an equivalent or higher fluence and with more passes than the UltraPulse.

Anesthesia was obtained with nerve blocks and local dermal infiltration using 2% lidocaine with 1:100000 epinephrine (Abbott Laboratories, North Chicago, Ill). Patients undergoing a full-face procedure received oral diazepam (5-10 mg), intramuscular meperidine (Demerol, Elkins-Sinn Inc, Cherry Hill, NJ) (25-50 mg), and promethazine hydrochloride (Phenergan, Elkins-Sinn Inc) (12.5-25 mg).

Postprocedure wound care was the same for both treatment sides and included Second Skin (Spenco Medical Ltd, West Sussex, England) dressings, dilute vinegar soaks followed by application of vaseline or healing ointment (Aquaphor, Beiersdorf Inc, Norwalk, Conn), and continuation of oral antibiotics and antivirals for a total of 1 week.

Photographs of the treatment areas were taken before treatment, immediately after treatment, and at 1 week, 2 weeks, 2 months, and 6 months. The same 35-mm camera (Dine Macro-light system Model II, LA Dine Inc, Palm Beach Gardens, Fla) and ASA 100 film were used for all photographs. Standard photographic views (en face, 45°, and 90°) were taken at each patient visit. All film was processed by the same laboratory.

A 5-member panel, blinded to study objectives and laser assignments, evaluated the clinical photographs. All panel members were trained for the outcomes evaluation by reviewing nonstudy photographs. Erythema and edema were scored, using a continuous numeric scale of 1 to 10, for the right and left sides of the treatment areas at week 1, week 2, and month 2 follow-up intervals. Before treatment and 2 and 6 months after treatment photographs were shown to evaluate efficacy. Percentage of wrinkle improvement, on a scale of 0% to 100% improvement from baseline, was scored for the right and left sides of the treatment areas.

Profilometry measurements using silicone (Silfo, Developments Ltd, England) skin casts were used to quantitate wrinkle improvement. Skin casts of a wrinkle on each treatment side were obtained before the treatment and at 2 and 6 months. Placement of follow-up casts was based on photographic localization of the baseline cast over a specific wrinkle. All wrinkle casts were analyzed by the same laboratory using depth, area, and shadowing measurements.⁴

On the 6 full-face patients, 4 dot tattoos in the shape of a 1-cm square were applied to each middle to lateral aspect of the cheek before the treatment. The dot tattoos were applied with dermal injections of india ink. Digital images of the square tattoos were taken before the treatment and at 1 week, 2 weeks, 2 months, and 6 months. Square areas of the tattoo images were computed by a digital imaging system (IPLab Spectrum 10, 3.0, Signal Analytics Corporation, Vienna, Va), and preprocedure and postprocedure square areas were compared to determine if a decrease in surface area, representative of "collagen shrinkage," could be detected. The dot tattoos were removed at the end of the study period using a 1064-nm Nd:YAG laser.

Six patients had 2-mm skin biopsies on each side of the treatment area: 2 patients before and immediately after the treatment, 2 patients before and 1 week after the treatment, and 2 patients before and 6 months after the treatment. The before and after treatment biopsy specimens were taken from adjacent skin. The tissue samples, stained with hematoxylin-eosin and elastic tissue stains, were evaluated for degree of pretreatment photodamage, depth of thermal damage, and new collagen formation (measured increase in width of collagen in the papillary dermis).

All comparative analyses were based on analysis of variance evaluations with repeated measures or paired t tests. Significant differences were based on the analysis of variance calculations. For the profilometry scores, mean percentage of improvement was calculated using the north-south shadowing measurements. All comparative analyses used a Student paired t test calculation. Tattoo measurements were based on surface area calculations of the square.

Data analysis was based on the following number of completed patient follow-ups: 32 patients for weeks 1 and 2, 35 patients for month 2 healing, 34 patients for month 2 efficacy, and 33 patients for month 6 efficacy. Thirty-four patients were women; 1 was a man. The average patient age was 50 years. Skin types included 42% type I, 50% type II, and 8% type III. No individuals with skin types IV to VI were enrolled in this study. Pretreatment wrinkle severity did not differ significantly (P=.57) from right to left side of the study areas.

Healing, efficacy, and profilometry data are presented in Table 2 and tattoo surface areas in Table 3. The degree of erythema for each laser and the difference in erythema between the 2 lasers diminished from week 1 to week 2 and from week 2 to month 2. At week 1, 75% of the patients had more erythema on the UltraPulse than TruPulse sides. The difference in erythema and edema between the 2 treatment sides was clinically mild (Figure 1, Figure 2, Figure 3, and Figure 4) yet statistically significant for only week 1 (P=.02). At week 1, the erythema induced by the TruPulse was mildly less than that caused by the UltraPulse. All panel members considered the UltraPulse sides to have improved slightly more than the TruPulse sides at 2 months. The difference in efficacy between the 2 lasers reached statistical significance for month 2 (P=.02), yet no difference was noted at 6 months (Figure 5 and Figure 6). Adverse effects were the same for both laser treatment areas and included 0% scarring, 0% hypopigmentation, 20% hyperpigmentation, and 8% infection (2 patients with a positive culture for Staphylococcus aureus and 1 patient with a herpetic infection). All infections occurred after the full course of the preprocedure and postprocedure medication had been completed.

Histological sections demonstrated a larger amount of residual thermal damage (defined as the depth of altered collagen measured from the top of the ablated surface) with increasing numbers of passes with the TruPulse: 1 pass, 0-5 µm, and 5 passes, 85 µm. Additional histological data from nonstudy patients treated with the TruPulse using the same protocol found additional levels of thermal damage: 2 passes, 17 µm; 3 passes, 25 µm; and 4 passes, 40 µm (Figure 7). The UltraPulse laser (Figure 8) caused slightly more residual thermal damage: 2 passes, 40 µm; 4 passes, 60 µm with focal areas 80 to 100 µm. Damage to endothelial cells was visualized at depths of 250 µm. Six-month postprocedure biopsy specimens from 1 patient showed a band of new collagen in the papillary dermis measuring 30 µm on the TruPulse side (pretreatment band, 8 µm; 207% increase) and 40 µm on the UltraPulse side (pretreatment band, 13 µm; 208% increase). A second patient with pretreatment and posttreatment biopsy specimens in the periorbital area showed no signifcant increase in collagen for either laser. The epidermis in all postprocedure biopsy specimens showed less maturation disarray, fewer dyskeratotic cells, and a less flattened rete ridge pattern.

Evaluation of the square tattoos before and after laser resurfacing did not demonstrate any significant decrease in surface area (Table 3). The measurements showed no evidence of long-term collagen shrinkage.

The carbon dioxide lasers with a short pulse duration can effectively treat photodamaged skin and wrinkles.^1-3,5,6 The possibility of pronounced postprocedure erythema, which on average lasts as long as 8 to 12 weeks but may last as long as 6 months,² deters some patients from electing to undergo this procedure. One week after resurfacing biopsy specimens from this study demonstrate that the erythema correlates histologically with a superficial perivascular lymphocytic infiltrate, ectatic blood vessels, and neovascularization. These findings are normal wound healing responses, and new collagen formation, a potential component of wrinkle improvement, is dependent on this inflammatory phase of wound healing.

The question arises as to whether an equally efficacious outcome can be achieved if the residual thermal damage is limited, and in this way the erythematous, or inflammatory, phase of wound healing is diminished. The results from this study found that a shorter pulse duration laser caused slightly less erythema and no significant decrease in efficacy at weeks 1 and 2. The difference in erythema was clinically mild and diminished over time. Panel evaluations of efficacy showed a trend toward the UltraPulse causing more wrinkle reduction, but the difference was not statistically significant. In comparison, the profilometry measurements at month 2 demonstrated that the UltraPulse caused more wrinkle reduction, yet by month 6 there was no significant difference in outcome.

One should note that the shorter pulse–duration laser was generally used at higher fluences and with more passes than the longer pulse–duration laser. A comparative trial using the 2 lasers at equivalent fluences and number of passes would probably have caused a more distinct difference in both the amount of erythema and the efficacy. Furthermore, profilometry data can be difficult to interpret because a slight alteration from baseline in patient expression, orientation, or wrinkle identification can bias the results. In this study, it is unlikely that the UltraPulse sides had less wrinkle reduction at month 6 than month 2.

The histopathological presence of a wrinkle is often subtle. Some wrinkles are due to repetitive stressing of the skin and may be associated with no visible histological findings.⁷ More pronounced frown lines caused by muscular contraction have deeper thickened hypodermal septae.⁸ Rhytides arising from photodamage are associated with decreased collagen in the papillary dermis and an accumulation of elastotic material in the midreticular dermis.⁹ The finer collagen fibers of the papillary dermis diminish, causing the epidermis to rest on a more condensed papillary dermis. Wrinkles form because the superficial dermis has diminished in size and the epidermis, which has maintained or increased its length, adjusts by crinkling.¹⁰ Chronological aging causes progressive disappearance of the superficial network of perpendicular oxytalan fibers that also may cause the skin to wrinkle because of decreased contractility.¹¹ This study found several pretreatment wrinkles to contain focal areas of mid-dermal elastolysis and loss of the normal architecture of the superficial papillary dermal elastic structure containing oxytalan and eulanin fibers.

With these concepts of wrinkle development in mind, proposed mechanisms of action of resurfacing for wrinkle improvement include leveling of the skin surface by tissue ablation, new collagen formation resulting from thermal injury, and collagen shrinkage that tightens the skin.² For each laser, increasing the number of passes correlated with a greater depth of residual thermal damage. The shorter pulse–duration laser caused less thermal damage with each pass, yet to achieve the same tissue effect, more passes had to be performed. The additional passes caused increased amounts of thermal damage to the same depth similar to that seen with fewer passes with the longer pulse–duration laser. This study showed that using the lasers to achieve the same treatment tissue effect led to similar long-term clinical outcomes. As slightly less thermal damage per pass was induced, the shorter pulse–duration laser may be better suited for more superficial resurfacing where the level of ablation and depth of thermal damage need to be more precisely controlled. Superficial resurfacing is appropriate in the periorbital area, for mild photodamage and wrinkles, in darker skin-type individuals, and potentially, with further study, in the neck region with its delicate skin. Increasing the number of passes, or using a longer pulse–duration laser, can achieve deeper thermal damage, which is more desirable when treating more extensive photodamage.

Analysis of the skin following dermabrasion and deep chemical peeling shows a new layer of dense, compact collagen in the papillary dermis arranged in parallel alignment to the epidermis (repair zone), similar to a superficial scar. Behin et al¹² found that the band of new collagen that formed following both phenol peeling and dermabrasion increased from week 2 to week 16. Kligman et al¹³ noted an area of dermal collagen 2 to 3 mm wide on skin that had received a phenol peel 15 to 20 years earlier. Thin, randomly located elastic fibers were present within the collagen band. The epidermis is thicker and has restoration of the normal rete ridge pattern, less maturation disarray, and fewer dyskeratotic cells. The long-term effect of laser resurfacing on collagen remodeling is less well documented. Cotton et al¹⁴ examined resurfaced skin 3 months following resurfacing and found a dermal repair zone in an unspecified percentage of 4 patients' biopsy specimens. No quantitative comparison with pretreatment skin was noted. There was more fibrosis with higher laser energies, yet this increase was not statistically significant. It is important to note that there is a thin dermal repair zone in photodamaged skin itself, secondary to years of exposure to UV light.¹⁵ Seckel¹⁶ found a new layer of collagen present at 1 and 2 months but not at 3 months after resurfacing, suggesting that the zone of repair is not a long-term finding. Other clinical studies¹⁷ have found a maintained growth in the collagen repair zone. Wrinkle improvement in patients who have undergone resurfacing appears to increase between months 1 and 6, suggesting that fibroblast activity and collagen remodeling occur during this period to form a sustained band of fibrosis. Documenting increased collagen formation in resurfaced skin by in situ hybridization and Western blot analysis, similar to 12 weeks postdermabrasion results in which Nelson et al¹⁸ showed a 3-fold increase in type I procollagen, would provide more evidence to support this proposed mechanism.

Ideally, one would like to establish a dose-response association between depth of thermal damage and amount of new collagen formation after the procedure, yet not all the 6-month biopsy specimens in this study showed a significant band of new collagen. The limited new collagen seen on the postprocedure biopsy specimens may be because patients, especially the one treated in the periorbital area, had relatively superficial wrinkles and were treated with fewer laser passes. The mild thermal damage may have induced a less visible amount of new collagen, suggesting that superficial resurfacing causes minimal dermal changes. Limited statistical analysis can be completed because only 4 six-month biopsy specimens were obtained in this study.

Shrinkage of the skin is visible during resurfacing, and heat from the laser probably induces denaturation of some collagen fibers, an event that occurs at 55°C to 62°C.¹⁹ The uncoiled and shorter collagen fibers may form a condensed "shrunken" scaffolding on which new collagen forms.² The role of collagen shrinkage in the long-term effects of laser skin resurfacing is debatable since denatured, shrunken collagen is mostly degraded over several weeks and therefore not involved in the collagen remodeling process. The lack of evidence in this study of collagen shrinkage at 2 and 6 months after the procedure, as demonstrated by minimal change in the surface area of resurfaced skin measured to the pixel unit by a digital imaging system, supports a limited role of collagen shrinkage in wrinkle improvement. In contrast, measurements of collagen bands on electromagnetic images of resurfaced skin have shown a decrease in the length of collagen fibers by 27% up to 2 and 3 months after the procedure.¹⁵ Further studies evaluating more specifically the length of collagen fibers at long-term resurfacing follow-ups may be more definitive regarding the issue of collagen shrinkage.

Compared with the longer pulse–duration carbon dioxide laser, the shorter pulse–duration carbon dioxide laser, used with higher energy and more passes, caused slightly less erythema in the first 2 weeks while maintaining efficacy. The difference in erythema in the first 2 weeks was statistically significant yet clinically less relevant. The longer pulse–duration laser required lower energy and fewer number of passes to achieve an equivalent depth of ablation, level of residual thermal damage, and degree of efficacy. With either laser, increasing the number of passes increased the thermal damage and correlated with increased postprocedure erythema. The shorter TruPulse allows for more superficial tissue damage per pass and therefore appears better suited for situations requiring superficial ablation akin to erbium:YAG laser skin resufacing.²⁰ The longer UltraPulse achieves a desirable depth of tissue damage with fewer passes. The data on a limited number of patients did not support the long-term presence of collagen shrinkage and found a thin band of new papillary dermal collagen in some posttreatment sites. Further studies involving larger numbers of long-term follow-up skin specimens of laser-treated patients would provide more information regarding the mechanisms of action of laser skin resurfacing.

Accepted for publication July 23, 1998.

This study was supported by a financial grant from Palomar Medical Technologies, Lexington, Mass.

Presented as an abstract at the American Society for Laser Medicine and Surgery Annual Meeting, Phoenix, Ariz, April 4, 1997.

We thank Erin Kammann for help with the statistical analysis and Rowena Bonoan for her assistance in editing the manuscript.

Reprints: Joop M. Grevelink, MD, PhD, Massachusetts General Hospital, Dermatology Laser Center, POB 503, 275 Cambridge St, Boston, MA 02114 (e-mail: grevelink.joop@mgh.harvard.edu).