ORIGINAL ARTICLE. Histologic Evaluation of Laser Lipolysis Comparing Continuous Wave vs Pulsed Lasers in an In Vivo Pig Model

Transcription

1 ORIGINAL ARTICLE Histologic Evaluation of Laser Lipolysis Comparing Continuous Wave vs Pulsed Lasers in an In Vivo Pig Model Jessica R. Levi, MD; Anna Veerappan, BA; Bo Chen, PhD; Mirko Mirkov, PhD; Ray Sierra, PhD; Jeffrey H. Spiegel, MD Objectives: To evaluate acute and delayed laser effects of subdermal lipolysis and collagen deposition using an in vivo pig model and to compare histologic findings in fatty tissue after continuous wave diode (CW) vs pulsed laser treatment. Methods: Three CW lasers (980, 1370, and 1470 nm) and 3 pulsed lasers (1064, 1320, and 1440 nm) were used to treat 4 Göttingen minipigs. Following administration of Klein tumescent solution, a laser cannula was inserted at the top of a 10 -cm rectangle and was passed subdermally to create separate laser tunnels. Temperatures at the surface and at intervals of 4-mm to 20-mm depths were recorded immediately after exposure and were correlated with skin injury. Full-thickness cutaneous biopsy specimens were obtained at 1 day, 1 week, and 1 month after exposure and were stained with hematoxylin-eosin and trichrome stain. Qualitative and semiquantitative histopathologic evaluations were performed with attention to vascular damage, lipolysis, and collagen deposition. Results: Skin surface damage occurred at temperatures exceeding 46 C. Histologic examination at 1 day after exposure showed hemorrhage, fibrous collagen fiber coagulation, and adipocyte damage. Adipocytes surrounded by histiocytes, a marker of lipolysis, were present at 1 week and 1 month after exposure. Collagen deposition in subdermal fatty tissue and in reticular dermis of some specimens was noted at 1 week and had increased at 1 month. Tissue treated with CW laser at 1470 nm demonstrated greater hemorrhage and more histiocytes at damage sites than tissue treated with pulsed laser at 1440 nm. There was a trend toward more collagen deposition with pulsed lasers than with CW lasers, but this was not statistically significant. Histopathologic comparison between results of CW laser at 980 nm vs pulsed laser at 1064 nm showed the same trend. Hemorrhage differences may result from pulse duration variations. A theoretical calculation estimating temperature rise in vessels supported this hypothesis. Conclusions: Pulsed lasers with higher peak powers provided better hemostatic effects than CW lasers. The degree of lipolysis depended on wavelength, laser power, and energy density. Subdermal laser irradiation can stimulate collagen deposition in subdermal tissue and reticular dermis. Arch Facial Plast Surg. 2011;13(1):41-50 Author Affiliations: Division of Facial Plastic and Reconstructive Surgery, Department of Otolaryngology Head and Neck Surgery, Boston University School of Medicine, Boston (Drs Levi and Spiegel and Ms Veerappan), and Cynosure, Inc, Westford (Drs Chen, Mirkov, and Sierra), Massachusetts. IN THE PAST DECADE, ALTERNAtives to and improvements in liposuction have been explored. Among many innovative techniques that have emerged, perhaps the most promising is laser lipolysis. Delivered through a thin cannula, laser energy is converted to heat energy within the subcutaneous layer. Adipocytes absorb the energy, resulting in apoptosis and necrosis. In addition to removal of fat using laser lipolysis, heating of subdermal and dermal tissues can theoretically be used to treat tissue laxity. 1-4 This technique requires a balance of local and bulk tissue heating. Noninvasive studies 5-7 have shown that heating the dermis through the epidermis, limiting surface temperature to approximately 40 C, correlates with a dermal temperature of 60 C to 70 C and results in collagen contraction with improvement in skin texture. Several studies 7,8 have shown that administering the laser to the epidermis and limiting surface temperature to 40 C will produce changes in the dermis. Most notably, transepidermal laser administration causes collagen contraction and improved skin texture that lasts for months after treatment. 4 Direct delivery of laser energy deep to the dermis potentially provides a new method for treating tissue laxity through stimulation of collagen deposition in subdermal and dermal tissues. To further the evolution of such techniques and equipment, this study evaluated the acute and delayed tissue effects of subdermal laser application at various 41

2 1440 nm, 1470 nm, 6 W 1470 nm, 1470 nm, 6 W 1440 nm, 1470 nm, 6 W 1470 nm, 12 W specimen T1 Incision point M1 M2 Figure 1. Seventeen 10 -cm rectangles were tattooed on each pig, with a -cm space between adjacent areas. B1 Laser channels wavelengths and pulse formations. The study objective was to determine the best wavelength and pulse formation for safe and efficient collagen deposition with subcutaneous lipolysis and minimal bleeding. METHODS STUDY DESIGN Institutional approval for animal use was obtained before our study. Four adult female Göttingen minipigs weighing 25 to 35 kg each were used. Each pig underwent 4 procedures. On day 0, pigs were anesthetized using inhaled anesthetic, prepared with 5% povidone-iodine solution (Betadine; Perdue Frederick Co, Norwalk, Connecticut), and draped via sterile technique. Seventeen 10 -cm rectangles were tattooed on each pig, with a -cm space between adjacent areas (Figure 1). Klein tumescent solution (saline with bicarbonate [10 mg/l] and lidocaine, 0.1%, with epinephrine [0.65 mg/l]) was injected subcutaneously until the skin region was firm to palpation, indicating proper tumescence of subcutaneous tissue. Approximately 15 ml of Klein tumescent solution was injected per treatment rectangle. After 5 minutes to allow for vasoconstriction and diffusion of Klein tumescent solution, a small stab incision was made at the top of the first rectangle. A 1-mm-diameter steel cannula with a fiberoptic laser cable extending 2 mm distal to the cannula tip was introduced. The laser was made operational, and the cannula was repeatedly passed in the subdermal layer with care to administer the laser at a constant speed. This was performed such that passes were in a fanlike distribution, with tissue proximal to the point of entry receiving a more concentrated distribution of laser passes than the more distal bottom of the rectangles. This procedure was repeated with the remaining treatment rectangles. Each area underwent laser administration with 1 wavelength and at 1 energy setting, with 3 rectangles each receiving the same dosing. One rectangle on each pig was used as a control, and a cannula was passed through this rectangle without laser administration. Total administered energy per treatment site was measured, as well as temperature at the surface using a thermal camera (ThermaCAM E45; FLIR Thermography, Niceville, Florida) and deeper temperature extending subdermally to 2 cm deep using a needletip thermistor (Cynosure Inc, Westford, Massachusetts). Temperature measurements were obtained immediately following completion of laser administration to each treatment area. After laser treatment was completed, incisions were sutured with 2-0 chromic, and bacitracin containing petrolatum Figure 2. positions and laser channels. B1 indicates bottom; M1, second; M2, third; and T1, top. ointment was applied. The animals then recovered from the anesthetic. On days 1, 7, and 28 following laser administration, biopsies were performed on each pig. Pigs were anesthetized with inhaled anesthetic in addition to 2-mL local infiltration of lidocaine, %, with epinephrine 1: per biopsy site. Four biopsy specimens, elliptically shaped along a transverse axis of the rectangle and evenly spaced, were obtained from each rectangle and were labeled T1, M1, M2, and B1, with T1 referring to the top section of the rectangle, M1 the second, M2 the third, and B1 the bottom (Figure 2). Each biopsy site was closed with 3-0 nylon suture. Daily examination and photography were used to monitor for cutaneous injury. s were obtained for histologic examination and were analyzed using hematoxylin-eosin and trichrome stain. At 4 weeks, pigs were killed following the last biopsy session. LASERS Three continuous wave diode (CW) lasers and 3 pulsed Nd:YAG lasers were used in this study. The CW laser wavelengths were 980, 1370, and 1470 nm. The pulsed laser wavelengths were 1064, 1320, and 1440 nm with maximum repetition rates of 40 Hz. Pulse width for the pulsed laser was approximately 350 microseconds. Maximum energy per pulse for pulsed lasers with 1440, 1320, and 1064 nm were 375, 500, and 750 mj, respectively. Powers of lasers were as follows: 1440 nm at 20 W, 1320 nm at 20 W, 1064 nm at 30 W, 1470 nm at 15 W, 1370 nm at, and 980 nm at 2 0W. Laser energy was conducted to adipose tissue through a 600-µm optical fiber delivery system via a stainless steel microcannula. RESULTS Treatment involved a predetermined number of laser passes per tattooed rectangle; each pass occurred at a predefined speed and power based on an estimation of heating adipose tissue to approximately 65 C (coagulation temperature) in front of the fiber tip. The 6 laser wavelengths were divided into 2 groups according to absorption in fatty tissue. 42

3 A B C Gross Observation of Skin Damage 1 0 Skin damage Surface Temperature, C Figure 3. Surface temperature threshold of skin injury using 20-W 1320-nm pulsed laser with a surface temperature of 46 C. A, Surface temperature vs gross observation of skin damage (1 indicates yes; 2, no). B, Lesion on pig skin at 1 day after laser irradiation. C, Hematoxylin-eosin stain of skin surface lesion. Arrowhead indicates epidermolysis; arrow, focal thermal denaturation of collagen fibers. The high-absorption group included the 1370-, 1440-, and 1470-nm lasers, which have absorption coefficients of 1.71, 6.36, and 5.28 cm 1 in fatty tissue, respectively. For this group, 15 passes (with an attempt to evenly space them within the treatment area) were performed in each tattooed rectangle. Because of absorption differences, moving speeds of the cannula were 1 cm/s for 1370 nm and 2 cm/s for 1440 and 1470 nm. Various laser powers from 0 W (moving the cannula at the same speed without firing the laser) to 16 W were tested to obtain a correlation between laser power and tissue response. The low-absorption group included the 980-, 1064-, and 1320-nm lasers, which have absorption coefficients of 9, 6, and 0.33 cm 1 in fatty tissue, respectively. For this group, maximum allowable power for each laser was used to obtain 65 C in tissue in a single pass with practical moving speed. The set powers were 20 W for 980 nm, 30 W for 1064 nm, and 20 W for 1320 nm. For this group, 5 passes were performed in each tattooed rectangle. Various moving speeds were tested from 0.25 to 0 cm/s to test effects of different energy levels. Total energy administered per treated rectangle was calculated via dose meters on the laser equipment. These values served as a point of comparison among passes. Data were obtained for several measured variables, which are discussed in the following subsections. SURFACE TEMPERATURE THRESHOLD OF SKIN INJURY Injury observed on the skin surface was correlated with surface temperature measured using a thermal camera. Temperature threshold of skin injuries was 46 C to 48 C, as shown in Figure 3A. The threshold temperature did not relate to wavelength. Because subdermal temperatures were kept close to 65 C to achieve vessel coagulation and fibrosis, skin injury did not correlate with subdermal temperature per se. Skin injury was correlated with total energy delivered to the rectangle. Cutaneous injury manifested as flat red spots (Figure 3B). Most lesions appeared during laser administration, indicating acute thermal injury. As power levels approached the estimated thresholds, redness developed on skin several seconds after laser irradiation and persisted through the first postoperative day. Control sites had the same number of passes, and the cannula was moved at the same speed without active administration of laser energy. No injury was found at control sites. On histologic examination, cutaneous injuries were characterized by epidermolysis at the epithelial-dermal junction and by dermal collagen denaturation; epidermal cells within the thermal lesion were detached, granular, and shrunken. Collagen fibers were often fragmented and in various stages of partial or complete denaturation (Figure 3C). TISSUE RESPONSE VS LASER POWER AND ENERGY DENSITY Tissues treated with the 1440-nm pulsed laser at various powers were studied to evaluate acute and delayed laser effects on lipolysis and collagen deposition, as well as hemorrhage. Figure 4A shows a typical tissue response to 1440-nm laser irradiation at 1 day after administration. Inflammation and mild hemorrhage are visible, and adipocytes are in various shapes. In general, local hemorrhage, thermal coagulation of collagen fibers, and adipocyte injury were observed in subcutaneous tissue of each treated rectangle to varying degrees but not in control rectangles. Open cavities, typically less than mm in diameter, were noted in biopsy specimens (Figure 4A). Most inflammation, mild hemorrhage, and thermal coagulation of collagen fibers were observed around cavities. At control sites treated by mechanically moving the cannula at the same speed without laser application, scattered areas of open cavities and local hemorrhage were also observed, but no collagen coagulation was found. Blind semiquantitative histopathologic grading was conducted based on a scale of 0 (none) to 3 (severe) to evaluate the extent of subcutaneous hemorrhage at 1 day after exposure, the number of histiocytes at 1 week, and the degree of collagen deposition at 1 month. These variable and time combinations were selected because they showed the greatest differences compared with 2 other biopsy times; therefore, hemorrhage was most pronounced at day 1 compared with 1 week and 1 month. 43

4 A 3 B µm 500 µm C Grading of Hemorrhage D Mean over whole quadrant T1 M2 M1 B1 0 1 Hemorrhage at 1 d 2 3 Grading of Fibrosis Fibrosis at 1 mo 2 3 Grading of Histiocytes Histiocytes at 1 wk Figure 4. Tissue response vs laser power and energy density using 10-W 1440-nm pulsed laser. A-C, Hematoxylin-eosin original magnification 100. A, Subcutaneous tissue at 1 day after laser irradiation at biopsy position M2. 1 Indicates open cavity; 2, thermal coagulation of collagen fiber; 3, hemorrhage; and 4, various shapes of adipocytes. B, Subcutaneous tissue at 1 month after laser irradiation at biopsy position M1. C, Grading of hemorrhage, fibrosis, and histiocytes. D, Power relative to hemorrhage at 1 day, histiocytes at 1 week, and fibrosis at 1 month. B1 indicates bottom; M1, second; M2, third; and T1, top. In contrast, histiocytes took a few days to migrate to the damaged area and then gradually decreased over time, and collagen required at least a few weeks to develop. A score of 0 indicates no tissue change, 1 indicates mild change that is less than 10% of the observed area, 2 indicates moderate change that covers 10% to 50% of the 44

5 area, and 3 indicates severe change that includes more than 50% of the area within the specimen (Figure 4C). The 4 biopsy positions were interpreted to represent 4 levels of energy densities. T1 was obtained close to the cannula entrance; therefore, tissue at T1 received the highest energy density at the center and the least at the sides. In comparison, tissue obtained from B1 was evenly treated over the whole area and had an even energy density across the biopsy specimen. The grading scores at each laser power and biopsy position are shown in Figure 4D. Collagen deposition in subdermal tissue and in reticular dermis of some specimens was identified at 1 week and was more apparent at 1 month after treatment. At 1 month, scattered areas of tissue damage were still healing. At control sites, no collagen regeneration was noted at 1 week or at 1 month. Overall, there was more collagen deposition with increased power levels. At day 1 after exposure, it was hard to distinguish thermal damage to adipocytes from the artifact of histologic preparation. Adipocytes surrounded by histiocytes, a marker of fat necrosis, were present at 1 week and 1 month after exposure (Figure 5). At control sites, few histiocytes were noted at 1 week and 1 month after exposure. Overall, there were more histiocytes with increased powers and energies. Histiocyte quantity did not significantly correlate with laser wavelength. Degree of subcutaneous hemorrhage correlated most with period after exposure, with hemorrhage being most severe at 1 day, much reduced at 1 week, and essentially gone at 1 month. Hemorrhage was also found at control sites. There was no clear correlation of hemorrhage with lipolysis and collagen deposition. There was no correlation of hemorrhage with energy or power, unlike adipocyte damage and collagen deposition, which were clearly correlated with power. Hemorrhage also correlated with laser wavelength, with shorter wavelengths demonstrating greater hemorrhage than longer wavelengths. TISSUE RESPONSE TO CW AND PULSED LASER IRRADIATION 500 µm Figure 5. Subcutaneous tissue at 1 week after laser irradiation using 10-W 1440-nm pulsed laser at biopsy position B1 (hematoxylin-eosin original magnification 400. Blue circle indicates estimated outline of damage zone. To study effects on tissue, histopathologic grading scores of tissue responses to 1470-nm CW laser and 1440-nm pulsed laser at various powers were compared (Figure 6A). Because the 2 wavelengths had a similar absorption coefficient in tissue, the difference in tissue effects can be attributed to pulse formation. The comparison was performed by pairing tissue biopsy specimens at the same laser power and biopsy position from 1470-nm CW laser vs 1440-nm pulsed laser. At 1 day after exposure, 9 of 16 paired biopsies showed more hemorrhage from 1470-nm CW laser than from 1440-nm pulsed laser. Two pairs showed more hemorrhage from 1440-nm pulsed laser; the remainder showed the same amount of hemorrhage from the 2 lasers. Wilcoxon signed rank test was used to calculate the significance. Histiocyte grading at 1 week and collagen deposition at 1 month were also tested by Wilcoxon signed tank test. The results and P values are given in Table 1 and Table 2. A similar comparison was conducted to test another paired group, 980-nm CW laser and 1064-nm pulsed laser. These 2 wavelengths also had similar absorption coefficients; therefore, the difference in tissue response is due to pulse formation. The comparisons are shown in Figure 6B, with the results and statistical evaluation summarized in Table 3 and Table 4. Overall, there was more hemorrhage with CW lasers than with pulsed lasers, reaching significant values. Although there appeared to be more histiocytes with CW lasers than with pulsed lasers, this reached significance for 1 pair only, 1470-nm CW laser vs 1440-nm pulsed laser. There was a trend toward more collagen deposition with pulsed lasers, but this was not statistically significant. In 2 instances, once with the 1320-nm pulsed laser and once with the 1440-nm pulsed laser, the irradiation perforated skin. This indicated that the laser was closer to skin than intended and farther from subdermal target areas, that the vector or force was incorrect, or that energy administration was too localized. In addition, sutures used at biopsy sites seem to have irritated the animals, who periodically rubbed against the sides of the cages to relieve their discomfort. This may have worsened the appearance of the sample areas. COMMENT Surface temperatures exceeding 46 C to 48 C produced visible skin damage. This result is consistent with findings by DiBernardo et al. 9 Skin injury occurred when the surface temperature exceeded this threshold temperature regardless of the laser wavelength delivered to the subdermal layer. The laser was administered through a straight fiber; therefore, most energy was deposited directly to subdermal tissue, with only a small portion scattered to skin. Heating of the skin surface was mainly due to thermal diffusion from heated subdermal tissue underneath; therefore, absorption characteristics of skin to various wavelengths did not significantly affect the threshold temperature of skin injury. Lasers with low coefficients of absorption (980, 1064, and 1320 nm) had greater thermal diffusion. Temperature on the surface was proportional to total energy delivered in each rectangle (as measured by the dose meter on the laser equipment) (Figure 7). 45

6 AHemorrhage at 1 d, grading score BHemorrhage at 1 d 1470-nm CW 1440-nm Pulsed 980-nm CW 1064-nm Pulse 12 W Laser Power and Position 6W 30 s 20 s Speed and Position 10 s Histocytes at 1 wk, grading score Histocytes at 1 wk 12 W Laser Power and Position 6 W 30 s 20 s Speed and Position 10 s Fibrosis at 1 mo, grading score Fibrosis at 1 mo 12 W Laser Power and Position 6 W 30 s 20 s Laser Power and Position 10 s Figure 6. Tissue response to continuous wave diode (CW) laser and pulsed laser irradiation. A, Laser power and biopsy position relative to hemorrhage at 1 day, histiocytes at 1 week, and fibrosis at 1 month. s at are the means of 2 tests. B, Speed and biopsy position relative to hemorrhage at 1 day, histiocytes at 1 week, and fibrosis at 1 month. B1 indicates bottom; M1, second; M2, third; and T1, top. Level of adipocyte damage characterized by histiocyte quantity present at 1 week after exposure depended on laser power and energy density. As shown in Figure 4D, higher laser power and energy density were associated with more observed adipocyte damage. Figure 5 shows tissue effects from 1 pass of a 10-W 1440-nm pulsed laser at 2 cm/s. Damage to adipocytes was confined to approximately a 2-mm-diameter range around the tip of the fiber. The 1440-nm pulsed laser was characterized by high absorption and created localized thermal damage in adipose tissue. This was also true for other highabsorption lasers (ie, 1370 and 1470 nm), while lowerabsorption lasers demonstrated more diffuse tissue responses. Degree of collagen deposition at 1 month was proportional to the amount of collagen initially seen in tissue at 1 week after exposure; for the 1440-nm pulsed laser, the degree of collagen at 1 month was also proportional to the number of histiocytes at 1 week. At 1 month, lasers administered at the setting of 16 W induced significant collagen deposition (grade 3). On gross examination, tissue demonstrated significant tightening, but it was nodular and bumpy in appearance with an irregular surface, which was not ideal for cosmetic purposes. With this in mind, a 1440-nm pulsed laser at settings between 6 and 12 W was analyzed. At a speed of 2 cm/s across all powers between 6 and 12 W, tissue demonstrated no nodularity; therefore, this laser at these set- 46

7 Table 1. Blind Semiquantitative Histopathologic Grading of Tissue Response to 1470-nm Continuous Wave Diode Laser and to 1440-nm Pulsed Laser a Extent of Subcutaneous Hemorrhage at 1 d After Exposure Hemorrhage Hemorrhage T T1 0 M1 1 M1 0 M2 1 M2 0 B1 0 B T T1 1 M1 2 M1 2 B1 1 B T T1 1 M1 2 M1 1 M2 2 M2 0 B1 1 B T T1 1 M1 0 M1 1 M2 1 M2 1 No. of Histiocytes at 1 wk After Exposure Histiocytes T T1 2 M1 3 M1 2 M2 3 M2 3 B1 1 B T T1 2 M1 2 M T T1 1 M1 2 M1 1 M2 2 M2 2 B1 2 B T T1 2 M1 3 M1 1 Histiocytes Degree of Collagen Deposition at 1 mo After Exposure Collagen Deposition T T1 2 M1 1 M1 2 M2 1 M2 1 B1 1 B T T1 2 M1 1 M1 2 B1 1 B T T1 1 M1 1 M1 1 M2 2 M2 2 B1 1 B T T1 2 M1 2 M1 2 M2 3 M2 1 B1 3 B1 1 Abbreviations: B1, bottom; M1, second; M2, third; T1, top. a s at are the means of 2 tests. Collagen Deposition 47

8 Table 2. Statistical Significance of Data in Table 1 by Wilcoxon Signed Rank Test No. of Paired Samples That Have Higher s Tissue Response 1440-nm Pulsed Laser 1470-nm CW Laser Statistical Significance P Value Hemorrhage at 1 d 2 9 Yes.03 Histiocytes at 1 wk 1 8 Yes.02 Collagen deposition at 1 mo 7 3 No.67 Abbreviation: CW, continuous wave diode. Table 3. Blind Semiquantitative Histopathologic Grading of Tissue Response to 980-nm Continuous Wave Diode Laser and to 1064-nm Pulsed Laser Time, s Extent of Subcutaneous Hemorrhage at 1 d After Exposure Hemorrhage Time, s Hemorrhage T T1 1 M1 2 M T T1 2 M1 2 M1 2 B1 2 B T T1 0 M1 1 M1 0 M2 1 M2 1 Time, s No. of Histiocytes at 1 wk After Exposure Histiocytes T T1 3 M1 1 M1 3 M2 2 M2 3 B1 3 B T T1 1 M1 2 M1 2 M2 2 M2 2 B1 3 B T T1 1 M1 2 M1 2 Histiocytes Time, s Degree of Collagen Deposition at 1 mo After Exposure Collagen Deposition T T1 2 M1 3 M1 2 M2 3 M T T1 3 M1 2 M1 3 M2 2 M2 3 B1 1 B T T1 0 M1 1 M1 1 M2 2 M2 2 Abbreviations: B1, bottom; M1, second; M2, third; T1, top. Collagen Deposition 48

9 Table 4. Statistical Significance of Data in Table 3 by Wilcoxon Signed Rank Test No. of Paired Samples That Have Higher s Tissue Response 1064-nm Pulsed Laser 980-nm CW Laser Statistical Significance P Value Hemorrhage at 1 d 1 8 Yes.02 Histiocytes at 1 wk 3 4 No.99 Collagen deposition at 1 mo 5 4 No.56 Abbreviation: CW, continuous wave diode. Surface Temperature, C nm Linear (1440 nm) R 2 = Energy, J Figure 7. Surface temperature vs energy. Blood Vessel Temperature, C Higher probability of vessel coagulation Time, ms Figure 8. Model prediction of blood vessel temperature rise vs time. Pulsed laser:, 40 Hz, 250 mj/pulse, 0.35 ms pulse duration. Continuous wave diode laser:, 250 mj delivered in an equivalent pulse duration of 25 ms. Vessel diameter: 50 µm. Analytical thermal model by Mirkov et al. 10 tings represents a safe and effective way to create moderate collagen deposition underneath skin in this animal model. A similar degree of collagen deposition using lasers from the low-absorption group can be achieved by administering with higher power, slower speed, or both. In this group, tissue absorbed energy less efficiently; therefore, to achieve the same temperature in front of the laser tip and to induce the same effect on fibrosis, the laser would have to be moved at slower speed (because higher powers were impossible). Tissue treated with 30-W 1064-nm pulsed laser or 20-W 980-nm CW laser achieves a collagen deposition grade exceeding 1 ( 10% of tissue exhibiting fibrotic changes) at 1 month. However, because of the use of higher power and lower speed for these wavelengths compared with the high-absorption group, total energy delivered under skin was much higher and resulted in a much higher temperature rise on the skin surface. For example, the surface temperature reached 60 C using a 30-W 1064-nm pulsed laser with moving speed of 0.33 cm/s. Skin ulceration was noted at 1 day after exposure. Considering skin safety and efficacy in creating collagen deposition, a high-absorption wavelength with a low power is a better choice. Overall, laser application created mechanical and thermal damage to blood vessels within subdermal fat layers, which could cause bleeding. This was clearly seen in tissue treated with the 1470-nm CW laser, which demonstrated grade 2 hemorrhage (Figure 6). In contrast, tissue treated with a 1440-nm pulsed laser showed minimal hemorrhage at 1 day after exposure (grade 1 or less) (Figure 4D and Table 1). The difference was statistically significant by Wilcoxon rank sign test (P=.03). Therefore, pulsed lasers with high peak power provided a better hemostatic effect than CW lasers. Hemostasis is likely due to quick coagulation of blood vessels from pulsed laser at high peak power, which prevents further bleeding of broken blood vessels. In other words, pulsed lasers heat up blood vessels quickly, causing coagulation without losing much heat to surrounding tissues via thermal diffusion, as is seen with CW lasers. Pulsed lasers are characterized by peak powers in kilowatts, which is 2 orders of magnitudes greater than power levels of CW lasers (Figure 8). To prove this hypothesis of vessel coagulation, a mathematical model was created that confirms vessel coagulation results from application of high peak power as seen with pulsed lasers. Figure 8 shows temperature rise in a 50-µg vessel from a 10-W pulsed laser and CW laser. During a single pulse from pulsed laser, 250 mj of energy are delivered, vessel temperature rises beyond coagulation temperature (65 C), and hemostasis is subsequently achieved. In contrast, it took 25 milliseconds (71 times longer than for pulsed laser) for CW laser to deliver the same amount of energy; therefore, temperature rise was too low to coagulate damaged vessels. Similar calculations were performed to examine the relationship of blood vessel diameter to coagulation. For vessels 10 µm in diameter, there is low probability that pulsed laser or CW laser will produce coagulation. However, for vessels between 50 and 200 µm in diameter, there is high probability that pulsed laser will produce coagulation but still low probability that CW laser will cause coagulation. For vessels 300 µm in diameter or larger, both laser types have high probability of producing coagulation (Figure 9). Study design limitations must be considered. There is inherent variability associated with laser administra- 49

10 Blood Vessel Temperature, C Pulsed CW Pulse Duration, ms Higher probability of vessel coagulation Lower probability of vessel coagulation 100 Vessel diameter 10 µm 50 µm 100 µm 300 µm Figure 9. Model prediction of vessel coagulation. Pulsed laser:, 40 Hz, 250 mj/pulse, 0.35 ms pulse duration. Continuous wave diode (CW) laser: was, 250 mj delivered in an equivalent pulse duration of 25 ms. Analytical thermal model by Mirkov et al. 10 tion by humans. Although timing of each laser pass was consistent, precise location of laser application within the rectangle and exact depth of laser penetration were more difficult to fully control. Consequently, some areas of a rectangle received more laser heat than others, sometimes leading to focal laser application within the rectangle. This uneven heat distribution produced unintended variation in histologic findings; for example, areas receiving more heat than intended showed more localized and extensive laser effects. To control for this, at least 3 rectangles had the same laser settings for any given set of variables. In this study, we provided an accurate profile of immediate, short-term, and longer-term laser effects on tissue. To truly characterize long-term effects, specimens should be examined 3 to 6 months following laser exposure. In conclusion, subdermal laser application can be effective in changing skin appearance and histologic findings. Variations in wavelength, laser power, and energy density alter the resulting amount of lipolysis and ultimately the appearance of skin. For safety, skin temperature should be kept below 46 C to 48 C regardless of wavelength. Furthermore, pulsed lasers provide a better hemostatic effect than CW lasers. This is likely due to their short duration of exposure with high peak power. Considering the factors of skin safety, hemostasis, and efficacy, low-power (6-12 W) pulsed laser at highabsorption wavelength (eg, 1440 nm) is a good option for treating skin laxity by creating controllable collagen deposition in subdermal tissue and reticular dermis. With lasers at low-absorption wavelength, similar collagen deposition could be achieved but at the detriment of overall skin appearance. With these lasers, higher power with slower speed applied significant energy to tissues and a surface temperature in excess of 46 C. Overall, we demonstrated that subdermal laser irradiation can create a range of collagen deposition effects in subdermal tissue and reticular dermis without damaging the epidermis and papillary dermis. Therefore, it is a potential tool for controllable skin tightening. Accepted for Publication: June 27, Correspondence: Jeffrey H. Spiegel, MD, Division of Facial Plastic and Reconstructive Surgery, Department of Otolaryngology Head and Neck Surgery, Boston University School of Medicine, Boston, MA Author Contributions: Study concept and design: Levi, Veerappan, Chen, Mirkov, Sierra, and Spiegel. Acquisition of data: Levi, Veerappan, Chen, Mirkov, Sierra, and Spiegel. Analysis and interpretation of data: Levi and Chen. Drafting of the manuscript: Levi, Veerappan, Chen, and Spiegel. Critical revision of the manuscript for important intellectual content: Levi, Veerappan, Chen, Mirkov, Sierra, and Spiegel. Statistical analysis: Levi and Chen. Obtained funding: Spiegel. Administrative, technical, and material support: Levi, Veerappan, Chen, Mirkov, and Spiegel. Study supervision: Levi, Sierra, and Spiegel. Financial Disclosure: None reported. Funding/Support: This study was funded by Cynosure Inc (Dr Spiegel). Role of the Sponsor: Cynosure Inc had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript. Previous Presentation: This study was presented at the American Academy of Facial Plastic and Reconstructive Surgery Fall Meeting, October 1, 2010; San Diego, California. REFERENCES 1. Goldman A. Submental Nd:YAG laser-assisted liposuction. Lasers Surg Med. 2006; 38(3): Goldman A, Gotkin RH. Laser-assisted liposuction. Clin Plast Surg. 2009;36(2): Goldman A, Schavelzon DE, Blugerman GS. Laser lipolysis: liposuction using Nd- YAG laser. Rev Soc Bras Cir Plast. 2002;17(1): Kim KH, Geronemus RG. Laser lipolysis using a novel 1,064 nm Nd:YAG laser. Dermatol Surg. 2006;32(2): Alam M, Hsu TS, Dover JS, Wrone DA, Arndt KA. Nonablative laser and light treatments: histology and tissue effects a review. Lasers Surg Med. 2003;33(1): Badin AZ, Moraes LM, Gondek L, Chiaratti MG, Canta L. Laser lipolysis: flaccidity under control. Aesthetic Plast Surg. 2002;26(5): Chan HH, Lam LK, Wong DS, Kono T, Trendell-Smith N. Use of 1,320 nm Nd: YAG laser for wrinkle reduction and the treatment of atrophic acne scarring in Asians. Lasers Surg Med. 2004;34(2): DeHoratius DM, Dover JS. Nonablative tissue remodeling and photorejuvenation. Clin Dermatol. 2007;25(5): DiBernardo BE, Reyes J, Chen B. Evaluation of tissue thermal effects from 1064/ 1320-nm laser-assisted lipolysis and its clinical implications. J Cosmet Laser Ther. 2009;11(2): Mirkov M, Sherr EA, Sierra RA, Lloyd JR, Tanghetti E. Analytical modeling of laser pulse heating of embedded biological targets: an application to cutaneous vascular lesions. J Appl Phys. 2006;99:e /japiau/v99/i11/p114701_s1?isauthorized=no. Accessed October 19,