Figure 1. Vancouver Scar Scale questionnaire used in our study. The questionnaire includes assessment of vascularity, pigmentation, pliability, and height.
Figure 2. Patient and Observer Scar Assessment Scale10 used in our study. Hypo indicates hypopigmentation; hyper, hyperpigmentation.
Figure 3. Flowchart demonstrating participant accrual to the study.
Figure 4. Statistically significant improvement in pretreatment and posttreatment Vancouver Scar Scale observer scores (Wilcoxon signed rank test, P = .002).
Figure 5. Clinical photographs of a mature burn scar before (A) and after (B) 3 treatments with the fractional carbon dioxide laser.
Figure 6. Patient and Observer Scar Assessment Scale scores. There was statistically significant improvement in pretreatment and posttreatment Patient (Wilcoxon signed rank test, P = .002) (A) and Observer (Wilcoxon signed rank test, P = .004) (B) Scar Assessment Scale scores.
Figure 7. Herovici-stained specimen. Note the shift from type I collagen (red) predominance before fractional carbon dioxide laser treatment (A) to type III collagen (blue) after treatment (B) (original magnification ×20).
Figure 8. Hematoxylin-eosin–stained specimen. There is significant improvement in dermal collagen from pretreatment (A) to posttreatment (B) specimens, which have finer and more fibrillar collagen (original magnification ×20).
Customize your JAMA Network experience by selecting one or more topics from the list below.
Ozog DM, Liu A, Chaffins ML, et al. Evaluation of Clinical Results, Histological Architecture, and Collagen Expression Following Treatment of Mature Burn Scars With a Fractional Carbon Dioxide Laser. JAMA Dermatol. 2013;149(1):50–57. doi:10.1001/2013.jamadermatol.668
Objective To assess mature burn scars treated with a fractional carbon dioxide laser for changes in histological architecture, type I to III collagen ratios, density of elastic tissue, and subjective measures of clinical improvements.
Design Uncontrolled, prospective study of patients with mature burn scars, from a clinical and histological perspective. Biopsy specimens were obtained before and 2 months after 3 treatment sessions. The tissue was prepared with Verhoff von Giesen (VVG) stain to discern elastic tissue and Herovici stain to differentiate types I and III collagen.
Setting Subjects were recruited from the Grossman Burn Centers.
Participants Of 18 patients with mature burn scars, 10 completed the entire treatment protocol.
Intervention Participants received 3 treatments with a fractional carbon dioxide laser.
Main Outcome Measures Vancouver Scar Scale and Patient and Observer Scar Assessment Scale survey scores. In histological analysis, imaging software was used to measure changes in collagen subtype and elastic tissue. A rating scale was developed to assess normal vs scar architecture.
Results The first hypothesis that significant histological improvement would occur and the second hypothesis of a statistically significant increase in type III collagen expression or a decrease in type I collagen expression were confirmed. There were no significant changes in elastic tissue. Statistically significant improvements were seen in all survey data.
Conclusions Treatment with a fractional carbon dioxide laser improved the appearance of mature burn scars and resulted in a significant improvement in collagen architecture following treatment. Furthermore, in treated skin specimens, a collagen subtype (types I and III collagen) profile resembling that of nonwounded skin was found.
In the United States, approximately 500 000 patients are treated yearly for burn injuries, with a survival rate of 94.4%.1 After medical stabilization, patients who experienced burns can develop lifelong scars that have a devastating emotional and physical impact. Significant functional impairment can result from contractures. Various sources of thermal burns include fire, chemicals, scalding water, grease, and electricity.
Wound healing involves the release of cytokines and inflammatory mediators, leading to eventual re-epithelialization and deposition of type III collagen, followed by type I collagen, within the dermis. Mature burn scars have an abundance of type I collagen, typically arranged in thick bundles similar to hypertrophic scars. Fetal skin primarily consists of type III collagen, while adult skin has more abundant type I collagen.
The management of burn scars has focused on addressing the aberrant collagen deposition seen in mature scars. Treatments used today include scar excision, dermabrasion, intralesional corticosteroids, chemical peels, silicone gel sheeting, compression therapy, skin flaps, split-thickness and full-thickness skin grafts, autografting, cadaver skin transplants, and lasers. Recently, epidermal and dermal skin substitutes from human donor skin, animal-derived collagen, and hyaluronic acid have been used in wound healing.2 Newer treatment options include controlled enzymatic debridement, stem cells, and intelligent films.3
Ablative lasers can remove and debulk scar tissue. Fractional ablative lasers have been shown to be well tolerated for scar treatment, even in darker-skinned individuals.4,5 To our knowledge, possible changes in ratios of collagen types after fractional ablative laser treatment have not been previously described.
The staining technique described by Herovici in 1963 has been used to distinguish type I collagen (red) and type III collagen (blue) in the skin.6 This technique has been used to subjectively analyze keloid scars and Duputyren contractures.7,8 The Herovici stain has been used to quantify the amount of types I and III collagen within mature burn scars, providing an objective analysis.9
Despite anecdotal reports that ablative lasers improve scars, to our knowledge, no formal investigation has quantified improvement or possible underlying mechanisms. The present prospective study analyzed patients with mature burn scars treated with a fractional carbon dioxide laser. Potential histological architectural changes in treated mature burn scars were evaluated. It was hypothesized that treated areas would trend toward normal collagen architecture. The second objective was to evaluate the collagen subtype profile (types I and III collagen) and density of elastic tissue in burn scars before and after treatment. We hypothesized that type I to III collagen ratios would trend toward normal. The clinical objective was to examine whether patients noted a significant improvement after treatment.
This was an uncontrolled, prospective study, from both a clinical and histological perspective, of patients with mature burn scars treated with a fractional carbon dioxide laser. The study was conducted in coordination with the Moy-Fincher Medical Group (E.F.F., L.K.C., and R.L.M.), Grossman Burn Centers (P.H.G.), and 3 dermatopathologists (M.L.C., A.H.O., and J.C.P.)from the Department of Dermatology, Henry Ford Hospital. This study was approved by the Western International Review Board. All subjects provided written informed consent.
Before treatment and 2 months after the final treatment, the patients and treating physicians completed the scar rating scales. Patients completed 2 clinical assessment scales: (1) The Vancouver Scar Scale, the most widely used assessment scale which measures vascularity, pigmentation, pliability, and height (Figure 1), and (2) the patient portion of the Patient and Observer Scar Assessment Scale, which asks more concrete questions regarding the scar such as pruritus, pain, and stiffness. The treating physician completed the observer portion of the Patient and Observer Scar Assessment Scale (Figure 2).
Pretreatment biopsy specimens were collected from December 2009 through May 2010. The area of scar that was biopsied was carefully marked and photographed. All patients were treated with fractional carbon dioxide resurfacing using a fractional carbon dioxide laser (Lumenis Ltd). Tumescent anesthesia was initially performed for all patients receiving treatment. However, owing to pain experienced by several patients during tumescent anesthesia, a combination of topical medications and local injections was instead used to painlessly achieve adequate anesthesia.
The laser settings were developed from clinical experience with prior scar and resurfacing treatments. No pilot study was performed. All areas selected for inclusion had predominantly hypertrophic scarring and were treated with an Active FX (Lumenis Ltd) (energy settings, 80-100 mJ [53- to 79-μm ablation], 200 Hz, 75 MTZ [microthermal treatment zone]/cm2 [density, 68%-100%], and 1.3-mm spot size). The thicker, banded areas within these scars were treated with both Active FX (same energy settings) and Deep FX (Lumenis Ltd) (average energy settings, 20 mJ [600-μm ablation], 300 Hz, 361 to 529 MTZ/cm2 [density, 10%-15%], and 0.12-mm spot size). Adjustments were made within the described parameters for patient comfort.
Treatments were performed at 3 sessions scheduled at 2- to 3-month intervals. Treatment sizes ranged from 76 to 560 cm2. At follow-up 2 months after the final treatment, 4-mm punch biopsies were performed on the treated areas and sent for tissue processing and staining. These posttreatment biopsy specimens were taken adjacent to the pretreatment biopsy sites, as identified photographically.
Tissue blocks were fixed in 10% buffered formalin, embedded in paraffin, and sectioned in standard fashion. The stains included hematoxylin-eosin (H&E), Verhoff von Giesen (VVG), and a technique described by Herovici that allows distinction between type I collagen (red) and type III collagen (blue).
The H&E-stained slides were independently evaluated by 3 blinded dermatopathologists (M.L.C., A.H.O., and J.C.P) and graded according the appearance and pattern of dermal collagen deposition. The following grading criteria were used: 0, normal; 1, collagen is fine and fibrillar; 2, combination of 1 and 3; 3, collagen is fibrotic, vessels perpendicular to epidermis; 4, combination of 3 and 5; and 5, collagen is extremely sclerotic and compacted in thick bundles.
Similarly, the VVG-stained slides were also analyzed independently by the 3 blinded dermatopathologists. Scores were assigned to each slide based on the appearance of the dermal elastic tissue. The following grading criteria were used: 0, normal; 1, short fragmented elastic fibers; 2, intermediate between 1 and 3; 3, fibrillar elastic fibers, parallel to epidermis; 4, intermediate between 3 and 5; and 5, absent or nearly absent.
All VVG- and Herovici-stained specimens were viewed using an Olympus BX50 light microscope (Olympus Corp), and digital images were taken of the lower papillary to reticular dermis at ×20 magnification with an Olympus DP20 microscope camera. The image files were saved as high-resolution tag image file format (TIFF) files. Image Pro Plus 7.0 (Media Cybernetics) was used to identify and measure the number of red (type I collagen) and blue (type III collagen) pixels. This was achieved by computer software measurement of the number of pixels of each respective color. All data were stored in Excel files (Microsoft Corp) for statistical analysis. Similarly, the density of black VVG-stained elastic tissue in pretreatment and posttreatment specimens were evaluated using Image Pro Plus 7.0.
Eighteen patients with mature burn scars from various causes were recruited from the Los Angeles, California, area. Ten patients with ages ranging from 20 to 53 years were included in the study (Table 1). The causes of the burns included hot water (n = 2), acid (n = 1), and fire (n = 7). The mean total body surface area of involvement was 30.2% and ranged from 4% to 65%. Previous treatments included intralesional corticosteroid injections, skin grafts, and other surgical interventions.
Eight patients completed only the first treatment and were excluded. Six of these patients reported that the process was too involved and reported pain with tumescent anesthesia, 1 patient developed a biopsy site infection, and 1 patient developed an ulceration on the leg after the first treatment (Figure 3). Pain during tumescent anesthesia was thought to be due to the restrictive nature of hypertrophic burn scars when injected with large volumes of tumescent. Subsequently, the method of anesthesia was changed to a combination of topical anesthetics and local injections, with a decrease in pain observed. Because the patients dropped out early before clinical effects were expected, these patients were not considered to be nonresponders.
Pretreatment evaluation of scars revealed persistent vascularity and erythema in many subjects despite the significant length of time since injury. Most scars had increased pigmentation compared with surrounding noninvolved skin. Hypertrophy was noted in all treatment areas. Immediately after treatment, the skin surface displayed a white-gray frost, which on close inspection revealed a pinpoint pattern corresponding to individual laser columns. No bleeding was seen. Two months after treatment, there was improvement in vascularity, pigmentation, and scar thickness, as reported by both clinical evaluator and patient participants.
The mean (SD) initial Vancouver Scar Scale score was 8.5 (1.5), while the mean (SD) posttreatment Vancouver Scar Scale score was 4.9 (1.6) (Wilcoxon signed rank test, P = .002) (Figure 4). These results correlated with the observed clinical improvements (Figure 5). The mean (SD) initial patient portion score of the Patient and Observer Scar Assessment Scale was 39.7 (7.5), while the mean (SD) posttreatment patient score was 22.8 (6.2) (Wilcoxon signed rank test, P = .002) (Figure 6A). The mean (SD) initial observer portion score of the Patient and Observer Scar Assessment Scale was 32.2 (8.1), while the mean (SD) posttreatment observer score was 19.2 (6.1) (Wilcoxon signed rank test, P = .004) (Figure 6B).
Statistical analysis of type I and III collagen and elastic tissue densities, as well as dermatopathologists' ratings of H&E- and VVG-stained slides, are presented in Table 2. The interrater agreement was moderate among the 3 dermatopathologists involved in the study. The corresponding intraclass correlation coefficients are 0.57 for the prereadings and 0.54 for the postreadings.
A statistically significant decrease in type I collagen (red) was seen in the posttreatment biopsy specimens. The mean density of type I collagen in pretreatment specimens was 632 360 red pixels per image, while the mean density following laser treatment was 347 178 red pixels per image (Wilcoxon signed rank test, P = .002) (Table 2). In contrast, a statistically significant increase in type III collagen (blue) was identified in tissue specimens following treatment with the fractional carbon dioxide laser (Figure 7). The mean pretreatment type III collagen density was 198 505 blue pixels per image, and the mean posttreatment density was 302 061 blue pixels per image (Wilcoxon signed rank test, P = .01) (Table 2).
Furthermore, a statistically significant H&E mean rating decrease (improvement) was detected from pretreatment to posttreatment samples (Figure 8). The mean initial rating for H&E-stained slides was 4.00, while the mean for posttreatment ratings was 3.00 (Wilcoxon signed rank test, P = .004). With regard to the VVG-stained slides, neither the elastic tissue patterns nor densities were significantly changed on evaluation by the dermatopathologists or computer software analysis.
A significant improvement was seen in subjective clinical parameters, and the hypotheses regarding histological architecture and collagen subtypes were confirmed. The quantitative changes in collagen subtypes had not been previously described after treatment of hypertrophic burn scars with fractional carbon dioxide laser treatment. The change toward normal collagen histological architecture with a more haphazard arrangement is most likely responsible for the comments by several patients of increased mobility, particularly on the neck. Although there was improvement in the appearance of the scars observed by both subjects and investigators, this was not obvious photographically. This is likely because of improvement in collagen architecture without restoration of adnexal structures including hair follicles, sebaceous glands, and eccrine ducts.
Ablative devices including the Er:YAG and carbon dioxide lasers have been found effective in improving the appearance of scars, including mature burn scars. The carbon dioxide laser has also been combined with other interventions to achieve optimal results. In 2007, a case report described a large burn scar treated using the pinhole method with a carbon dioxide laser, combined with collagen induction therapy using a microneedle therapy system.11 Following 5 monthly sessions, improvement was seen in color and texture, as well as a decrease in scar contracture. This combined method is believed to cause both thermal and physical damage of scar collagen, resulting in its regeneration and realignment. A study in 2010 described 16 patients with mature burn scars who were treated with a combination of carbon dioxide laser followed by placement of split-thickness skin grafts.12 Although color and texture mismatch was seen in many of the study patients, overall, the treatments made the scars much more cosmetically acceptable.
Combined treatment with both Er:YAG and carbon dioxide lasers has also been studied for 327 patients with different types of hypertrophic scars, including hypertrophic burn scars.13 After 1 to 5 sessions at 6-month to 1-year intervals, improvements were noted based on subjective self-assessment and the Vancouver Scar Scale. The treated scars appeared softer, flatter, and less erythematous, although no improvement in scar contracture was observed. Another small report of 3 patients with facial burn scars also found the Er:YAG laser by itself to be an effective treatment modality.14
Manstein et al15 introduced the concept of fractional photothermolysis for inducing nonablative dermal remodeling. In this process, microthermal treatment zones composed of columns of destruction are created. The surrounding islands of untreated tissue promote rapid healing and re-epithelization. This technology has proven to be effective for treating mature burn scars. One mature burn scar was treated 5 times on a monthly basis using a fractional nonablative 1550-nm laser.16 Improvements in scar appearance and contracture were seen. Similarly, a randomized controlled trial investigated 17 patients with mature burn scars treated with a fractional nonablative 1540-nm erbium:glass laser.17 Following 3 monthly treatments, considerable improvement in skin texture was noted.
Although the adverse effect profile of fractional carbon dioxide lasers is improved compared with fully ablative devices, there are still reports of scarring, particularly in areas of thinner skin and decreased adnexal structures such as the neck. These adverse effects may be due to bulk heating of tissue and penetration of laser energy beneath the dermis. For this reason the treatment parameters used in this study had been previously used on many aesthetic patients in the practice with no adverse sequelae.
Most patients began treatment with baseline hyperpigmentation in their scars and had significant lightening both by self-report and clinician observer scoring. In no cases did the lightening effect cause hypopigmentation relative to the patients' normal uninvolved skin. Delayed hypopigmentation was a concern with fully ablative devices but has not to our knowledge been reported with fractional devices. However, it is unclear whether multiple treatments for conditions such as severe scarring will eventually lead to such cases. The clinical improvement in both hyperpigmentation and erythema may be related to the histological improvement in the appearance of superficial blood vessels. The vessels within the remodeled collagen now lie in both perpendicular and parallel planes. This may result in less vascular trapping.
The fractional carbon dioxide laser has been shown to be effective for burn and acne scars, even in individuals with darker skin types.4,5 In these reports, no long-lasting effects such as dyschromia or scar exacerbation were observed. Other case reports have also documented successful outcomes in burn scars after treatment with the fractional ablative carbon dioxide laser.18,19 No severe or permanent adverse effects were described in these reports. The fractional ablative Er:YAG laser has also been studied and successfully used for managing burn scars.20 In this report by Bowen,20 2 patients with mature burn scars were treated with a fractional Er:YAG laser combined with a method called the “selective objective fractional technique (SOFT).” This method adjusted the depth of treatment according to the thickness of the scar.
Compared with normal skin, collagen within scars demonstrates an aberrant appearance and deposition pattern. A large objective histological study using image analysis software of mature scars found collagen to be arranged in a more parallel fashion compared with normal skin.21 In 1995, a study involving 5 patients with hypertrophic burn scars present for 5 to 18 months evaluated the tissue specimens using in situ hybridization and immunohistochemical analysis.22 Compared with normal skin, the scar tissue samples demonstrated significant elevation of types I and III procollagen messenger RNA (mRNA), type I procollagen protein, and transforming growth factor-β1 mRNA and protein levels. There is currently no ideal method for determining and measuring the collagen subtypes in tissue samples. This may account for the inconsistent data regarding collagen deposition within scars.
A study in 2009 examined types I and III collagen ratios within mature burn scars using immunohistochemical analysis and laser confocal microscopy.23 Among the 17 patients there was a significant elevation of type III collagen in hypertrophic scars relative to normal skin. No difference in type I collagen ratios between scars and normal skin was detected. In contrast, another study involving mature burn scars and burn scar contractures found that scar tissues possessed higher type I to type III collagen ratios compared with normal skin.9 Specifically, scars demonstrated both elevated type I collagen levels and lower levels of type III collagen. These findings may differ from the previously mentioned study owing to differences in their methods for determining collagen subtype. As in our study, collagen subtype assessment was performed using the Herovici stain, whereby type I collagen is stained red while type III collagen is stained blue. As fetal skin of up to 6 months gestational age is known to possess the ability to heal without notable, if any, scarring at all, animal models have been developed to further examine this unique feature. A study using an ovine (sheep) animal model of partial-thickness burns producing scarless fetal healing examined differences in collagen subtype expression.24 Using the picrosirius dye stain combined with polarized light microscopy to visualize collagen fibers, Cuttle et al24 found that type I collagen displays a yellow to red birefringence, while type III collagen shows green birefringence. Combined with imaging software analysis, the authors found that with increasing fetal age, there was an increase in type I collagen level, decrease in type III collagen level, and increased deposition of thicker collagen fibers. Following exposure to burns, the collagen subtype ratios did not change, but while fetal tissue retained normal collagen organization, more mature lamb tissue displayed irregular collagen architecture.
Limitations of this study include the small sample size, lack of objective clinical measures for scarring, and relatively short follow-up time. The number of patients who dropped out after the first treatment was higher than expected. Several patients reported pain during anesthesia administration as a reason for not wanting to continue with additional treatments. These patients had received tumescent anesthesia as opposed to topical and intralesional injections. It was found that the restrictive nature of these hypertrophic burn scars increased the pain with large volumes of tumescent. After these patients experienced difficulty, the method of anesthesia was changed to include the use of topical anesthetics and local injections. A decrease in pain was observed, while still maintaining adequate anesthesia during treatment. As the patients dropped out early before clinical effects were expected, these patients were not considered to be nonresponders. The lack of objective measures for scar improvement is a barrier in scar intervention studies. A longer follow-up period may have resulted in additional collagen remodeling and subjective clinical improvement.
A single biopsy site infection was seen on the lower extremity of 1 patient. For future studies with biopsies, we would consider antibacterial prophylaxis. Larger treatment areas may also warrant antibiotic prophylaxis. If the patient has a history of herpes simplex infection and facial lesions are treated, particularly periorally, antiviral prophylaxis is recommended.
In our study, burn scars treated with the fractionated carbon dioxide laser demonstrated a return toward a fetal collagen profile, with increased type III collagen and decreased type I collagen levels. This pattern of collagen deposition resembling fetal skin may account for the improved clinical appearance of treated scars. Histological evaluation of the H&E-stained slides demonstrated an improvement in the collagen architecture, more closely resembling that of normal skin. The authors in a previous study stressed caution in interpreting their data because small regrowing fibers may be mistaken for type III collagen.24 This may also be an issue in the interpretation of our data as well.
We believe that the fractional carbon dioxide laser is an effective modality for improving the appearance and minimizing the morbidity associated with mature burn scars. While there have been several case reports of burn scars successfully treated with this device, our prospective study is the first to correlate these findings with histological data as well. Given the excellent tolerability and lack of prolonged adverse effects, the fractional carbon dioxide laser should be further evaluated for the treatment of mature burn scars. During this study, unsolicited patient comments regarding increased range of motion were received, specifically from patients who received treatment on the neck. Therefore, future work should include range of motion studies, since functional improvement in patients with restrictive movement is often critical.
Correspondence: David M. Ozog, MD, Division of Mohs Micrographic Surgery, Department of Dermatology, Henry Ford Hospital, 3031 W Grand Blvd, Ste 800, Detroit, MI 48202 (firstname.lastname@example.org).
Accepted for Publication: August 7, 2012.
Published Online: October 15, 2012. doi:10.1001/2013.jamadermatol.668
Author Contributions: All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Ozog, Chaffins, Ormsby, Fincher, and Moy. Acquisition of data: Ozog, Liu, Chaffins, Ormsby, and Chipps. Analysis and interpretation of data: Ozog, Liu, Chaffins, Ormsby, Mi, Grossman, and Pui. Drafting of the manuscript: Ozog, Liu, Ormsby, and Mi. Critical revision of the manuscript for important intellectual content: Ozog, Liu, Chaffins, Fincher, Chipps, Grossman, Pui, and Moy. Statistical analysis: Ormsby. Obtained funding: Ozog. Administrative, technical, and material support: Ozog, Chipps, Mi, and Grossman. Study supervision: Ozog, Chaffins, Ormsby, Fincher, Chipps, and Moy.
Financial Disclosure: None reported.
Funding/Support: This study was supported in part by a grant from Lumenis.
Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.
Previous Presentation: Findings from this study were presented at the American Society for Laser Medicine and Surgery conference; April 1, 2011; Grapevine, Texas.
Additional Contributions: Melody Eide, MD, MPH, helped review the manuscript; Elizabeth Farhat, MD, helped organize slides and acquire imaging software; and Gordon Jacobsen, MS, assisted with statistical analysis.
Create a personal account or sign in to: