2.1.

Laser, Cryogen, and CSC Delivery Device

The flashlamp-pumped pulsed dye laser Candela Sclero-Plus $™$ (Candela Inc., Wayland, MA) was used to irradiate the ex vivo normal dark human skin samples and in vivo rabbit external ears. This laser provided user-specified discrete radiant exposures between 4 to $15\mathrm{J}∕{\mathrm{cm}}^2$ . The wavelength was tuned to 585 or $595\mathrm{nm}$ . The pulse duration was fixed at $1.5\mathrm{ms}$ , spot size $7\mathrm{mm}$ .

The cryogen types used in the study were R134-a and R-404a (Dupont Corp., Wilmington, DE). While there are various refrigerants (Table 1 ) and we have previously reported on the use of R-407c and R-22, ^{17, 18, 19, 20} we chose to compare R-134a (currently used in clinical setting) with R-404a based on the following criteria: (1) environmental compatibility, (2) a sufficiently low boiling point but not inducing cryoinjury in the absence of laser irradiation, and (3) practical delivery (e.g., our previous attempts at delivering frozen ${\mathrm{CO}}_2$ produced an intense sound, similar to gun shot, when injected through a valve). Based on these criteria, R-22 and dry ice did not serve as candidate materials. R-407c would have been a reasonable choice; however, it has a slightly warmer boiling point than R-404a.

Table 1

Examples of refrigerants and their boiling points at atmospheric pressure.

Cryogen R-134a (1,1,1,2-tetrafluoroethane) is environmentally compatible, nontoxic with its boiling point of $-26.2°\mathrm{C}$ at atmospheric pressure, and has been approved by the Food and Drug Administration for use in cutaneous laser treatment. The composition of the R-404a as a percentage of weight is 52% trifluoroethane (R-143a), 44% pentafluoroethane (R-125), and 4% 1,1,1,2-tetrafluoroethane (R-134a). The boiling point of R-404a is $-46.5°\mathrm{C}$ at atmospheric pressure. Cryogen R-404a is also environmentally compatible and classified as a high-pressure refrigerant used in commercial applications of low temperature cooling systems.²¹

A commercial fuel injector (Standard Motor Products, 800-1257N, Long Island City, NY) was used to deliver R-134a or R-404a. A nozzle with a 1-mm diameter orifice was attached to the injector to produce a uniform spray cone. The injector was aimed toward the sprayed surface at an angle of $30\mathrm{deg}$ with respect to the normal. The injector-to-surface distance was $85\mathrm{mm}$ for all experiments, which was the optimized spray distance for the maximum temperature reduction within a skin phantom for this particular cryogen delivery device.²² A previous study demonstrated that various cryogen delivery devices produce different amounts of $Q$ [see Eq. 1].²² The “optimum” injector-to-surface distance is, therefore, device dependent. In all experiments, the cryogen spurt duration as well as the time delay between the termination of cryogen spurt and the onset of laser pulse (in ex vivo and in vivo studies) were controlled by using a programmable digital delay generator (DG 535, Stanford Research System, Sunnyvale, CA).

2.2.

Skin Phantom

A skin phantom made of epoxy resin (EP30, Master Bond Inc., Hackensack, NJ) with an embedded $30\text{-}\mathrm{\mu }\mathrm{m}$ diameter type $K$ thermocouple $\left({\mathrm{Chromega}}^{\circledR }\text{-}{\mathrm{Alomega}}^{\circledR }\right)$ (Omega Engineering, Inc., Stamford, CT) positioned at a $100\text{-}\mathrm{\mu }\mathrm{m}$ $(\pm 5\mathrm{\mu }\mathrm{m})$ depth below the phantom surface was used to acquire the temperature profiles in response to CSC. The thermal diffusivity of the phantom was $0.843\times {10}^{-7}{\mathrm{m}}^2{\mathrm{s}}^{-1}$ , which is within the range of human skin ( $6.9\times {10}^{-8}{\mathrm{m}}^2{\mathrm{s}}^{-1}$ to $1.1\times {10}^{-7}{\mathrm{m}}^2{\mathrm{s}}^{-1}$ ).²³ Such a phantom has also been used by other investigators ^{24, 25, 26, 27, 28, 29, 30, 31, 32} to investigate the effects of cryogen droplet size, spray density, droplet velocity, diameter, and film height on the CSC-induced heat removal from the phantom. Good agreements between computed temperature profiles and those measured in an epoxy resin phantom have been obtained,^{24, 26} indicating the effectiveness of this in vitro model. In addition, another type $K$ thermocouple with a $60\text{-}\mathrm{\mu }\mathrm{m}$ diameter sensing bead was used to measure the cryogen film temperature on the surface of the skin phantom. Both thermocouples were connected to an InstruNet digital data acquisition system (Omega Engineering, Inc., Stamford, CT). The data acquisition rate during the temperature measurements was $1000\mathrm{Hz}$ .

The skin phantom, initially at room temperature $(\approx 21°\mathrm{C})$ , was sprayed with R-134a or R-404a spurts at the durations of ${\tau }_{\mathrm{CSC}}=100$ , 200, and $300\mathrm{ms}$ . All temperature measurements were performed in triplicate for each setting. Measured temperature profiles were subsequently used to estimate CSC-induced heat removal from skin phantom using an algorithm that solved an inverse heat conduction problem.²⁶ Additional details about the performance and accuracy of the algorithm can be found in Ref. 22.

2.3.

Ex Vivo Dark Human Skin Sample

Normal ex vivo human skin samples of Fitzpatrick types V-VI were obtained from the abdomens of seven consenting adult females undergoing the transverse rectus abdominis myocutaneoues (TRAM) flap procedures in the department of plastic surgery at the University of Texas M.D. Anderson Cancer Center (MDACC). The TRAM procedure extracts a skin and muscle section from the lower part of the abdomen for breast reconstruction. The redundant skin sections from TRAM procedure were then used for the study. The protocol to obtain the skin samples was approved by the Institutional Review Board at MDACC and the Rice University. The skin samples were harvested with epidermis, dermis, and subcutaneous fat as a whole. Once the skin samples were harvested, they were stored in biohazard plastic bags to prevent the loss of water content and then delivered to the laboratory. Temperatures of the skin samples were approximately $21°\mathrm{C}$ (room temperature) before experimentation. Each ex vivo human skin sample was irradiated by 595-nm wavelength at the radiant exposures of $D_{\mathrm{0,595}}=4$ , 6, 8, 10, 13, and $15\mathrm{J}∕{\mathrm{cm}}^2$ in conjunction with CSC (R-134a or R-404a) at the spurt durations of ${\tau }_{\mathrm{CSC}}=100$ , 200, and $300\mathrm{ms}$ , respectively. Based on the parameters used in clinic settings,¹³ the time delay between the termination of cryogen spurt and the onset of laser pulse was $20\mathrm{ms}$ throughout the ex vivo experiments. This time delay is believed to be sufficient to ensure a relatively calm cryogen film to form on the sprayed surface, but sufficiently short to allow for the majority of heat removal to occur during the spurt duration.³³ Biopsy specimens were taken from the irradiated sites by 6-mm punches, fixed in 10% buffered formalin, processed for histological sectioning, and stained with hematoxylin and eosin (H&E). Thermal injury to the epidermis was evaluated by H&E histological observations.

2.4.

In Vivo Rabbit External Ear

Two New Zealand Albino white rabbits were used for the in vivo study as a model for cutaneous anomalies because the skin on a rabbit’s external ear contains plentiful dermal vasculature. It was observed from the histological sections of nonirradiated control sites that rabbit ears contained vasculature ranging from capillaries with diameters $<10\mathrm{\mu }\mathrm{m}$ to blood vessels with a diameter of $250\mathrm{\mu }\mathrm{m}$ . The thicknesses of the rabbit ears ranged from 920 to $1190\mathrm{\mu }\mathrm{m}$ . The protocol to carry out the animal experiments was approved by the Institutional Animal Care and Use Committee at the Rice University.

The animals were fed with standard laboratory diet and water ad libitum. At the time of the experiments, the rabbits weighed between 2.5 and $3.0\mathrm{kg}$ . Prior to laser irradiation, animals were anesthetized with 3% isofluorane in a lucite chamber until the animals reached the unconscious state. Once unconscious, the animals were maintained under anesthesia with 2% isofluorane delivered through a face mask. The animals were irradiated by 585-nm wavelength at two radiant exposures of $D_{\mathrm{0,585}}=10$ and $13.5\mathrm{J}∕{\mathrm{cm}}^2$ in conjunction with CSC (R-134a or R-404a) at the spurt durations of ${\tau }_{\mathrm{CSC}}=100$ , 200, and $300\mathrm{ms}$ , respectively. The 585-nm wavelength was used here because the blood absorption at $585\mathrm{nm}$ $(\approx 191{\mathrm{cm}}^{-1})$ is higher than that at $595\mathrm{nm}$ $(\approx 43{\mathrm{cm}}^{-1})$ .³⁴ The time delay between the termination of cryogen spurt and the onset of laser pulse was also $20\mathrm{ms}$ throughout the in vivo experiments. One animal was sacrified $2\mathrm{h}$ after laser irradiation and the other 4 days after irradiation. Biopsy specimens were taken from the irradiated sites by 6-mm punches, fixed in 10% buffered formalin, processed for histological sectioning, and stained with H&E. Thermal injury to the dermal vasculature was evaluated by H&E histological observations.

3.1.

Temperature Profiles and Heat Removal in Response to CSC With R-134a or R-404a

Figure 1 compares the averaged temperature measurements $(n=3)$ of a cryogen film on the surface of a skin phantom in response to R-134a and R-404a spurts at the durations of 100, 200, and $300\mathrm{ms}$ . During cryogen spraying time, regardless of spurt duration, cryogen film temperature reached and remained almost constant at approximately $-55°\mathrm{C}$ for R-134a and $-68°\mathrm{C}$ for R-404a, well below the corresponding boiling points of the cryogens due to the evaporation of cryogen droplets during the flight from the injector to the skin phantom. Following cryogen spurt termination, liquid cryogen film remained on the phantom surface for a period of time (on the order of seconds), and the cryogen film temperatures during this period were near the boiling points of the cryogens: $-26.2°\mathrm{C}$ for R-134a and $-46.5°\mathrm{C}$ for R-404a. The lifetime of an after-spurt cryogen film was defined as the time between the cryogen spurt termination and the complete evaporation of the cryogen film. We quantified it from our temperature measurements using the $60\text{-}\mathrm{\mu }\mathrm{m}\text{-}\mathrm{bead}$ type $K$ thermocouple placed on the surface of a skin phantom. The complete evaporation time was identified as the time when the recorded temperature began to increase above the respective boiling points for R-134a and R-404a. The lifetime of after-spurt cryogen film ${\tau }_{\mathrm{a}}$ was directly related to the cryogen spurt duration ${\tau }_{\mathrm{CSC}}$ . For the same spurt duration ${\tau }_{\mathrm{CSC}}$ , the ${\tau }_{\mathrm{a}}$ for R-404a was much shorter than that for R-134a. When ${\tau }_{\mathrm{CSC}}$ was $100\mathrm{ms}$ , ${\tau }_{\mathrm{a}}$ was approximately $530\mathrm{ms}$ for R-404a and $1833\mathrm{ms}$ for R-134a [Fig. 1a]. When ${\tau }_{\mathrm{CSC}}$ was increased to $200\mathrm{ms}$ , ${\tau }_{\mathrm{a}}$ were, respectively, 1147 and $2645\mathrm{ms}$ for R-404a and R-134a [Fig. 1b]. When ${\tau }_{\mathrm{CSC}}$ was further increased to $300\mathrm{ms}$ , ${\tau }_{\mathrm{a}}$ were 1447 and $3421\mathrm{ms}$ for R-404a and R-134a, respectively [Fig. 1c]. These differences may be due to the much lower boiling point of R-404a and subsequently the larger temperature difference between cryogen film and the surface temperature of the skin phantom, leading to a higher evaporation rate for R-404a film.

Fig. 1

Comparison of cryogen film temperature on the surface of the skin phantom in response to R-404a and R-134a spurts: (a) ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , (b) ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , (c) ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ .

Figure 2 depicts the measured internal temperature profiles of skin phantom at the depth of approximately $100\mathrm{\mu }\mathrm{m}$ below the phantom surface in response to R-134a or R-404a spurts at the spurt durations of 100, 200, and $300\mathrm{ms}$ . For ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , the temperature at the depth of $100\mathrm{\mu }\mathrm{m}$ was $0.8°\mathrm{C}$ (temperature reduction $\mathrm{\Delta }T=21-0.8=20.2°\mathrm{C}$ ) for R-404a and $5.2°\mathrm{C}$ $(\mathrm{\Delta }T=15.8°\mathrm{C})$ for R-134a immediately after the spurt termination [Fig. 2a]. In response to ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , the corresponding temperatures were $-10.6°\mathrm{C}$ $(\mathrm{\Delta }T=31.6°\mathrm{C})$ for R-404a and $-6.8°\mathrm{C}$ $(\mathrm{\Delta }T=27.8°\mathrm{C})$ for R-134a immediately after the spurt termination [Fig. 2b]. With ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ , the corresponding temperatures were $-18.6°\mathrm{C}$ $(\mathrm{\Delta }T=39.6°\mathrm{C})$ for R-404a and $-13.7°\mathrm{C}$ $(\mathrm{\Delta }T=34.7°\mathrm{C})$ for R-134a immediately after the spurt termination [Fig. 2c].

Fig. 2

Comparison of internal temperature at the depth of $100\mathrm{\mu }\mathrm{m}$ below the surface of the skin phantom in response to R-404a and R-134a spurts: (a) ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , (b) ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , (c) ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ .

Using an algorithm that solved a one-dimensional inverse heat conduction problem,²⁶ we estimated the CSC-induced heat removal from the skin phantom in response to R-404a or R-134a spurts at various spurt durations from the measured internal temperatures (Fig. 3 ). The heat removal profiles of R-404a and R-134a in response to 300-ms spurts suggest that a 100, 200, or 300-ms spurt of R-404a extracts, respectively, 12.1, 11.1, or 11.7% (6.4 versus $5.6\mathrm{kJ}∕{\mathrm{m}}^2$ , 11.5 versus $10.2\mathrm{kJ}∕{\mathrm{m}}^2$ , or 15.4 versus $13.6\mathrm{kJ}∕{\mathrm{m}}^2$ ) more heat than a R-134a spurt of equal duration. Therefore, the comparative increase in heat removal at the end of the cryogen spurt due to CSC with R-404 is approximately constant regardless of the spurt durations from 100 to $300\mathrm{ms}$ .

Fig. 3

Comparison of CSC-induced heat removal from the skin phantom in response to R-404a and R-134a spurts: (a) ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , (b) ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , (c) ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ .

3.2.

Histological Evaluation of Thermal Injury to the Epidermis

Figure 4 shows the H&E histological sections of a control site (without CSC and laser irradiation) and a site sprayed with a 300-ms R-404a spurt alone (without laser irradiation). The two sites were from the same skin sample. From the comparison of the two histological sections, it is observed that a 300-ms R-404a spurt did not induce obvious CSC-induced morphological changes within skin. Analysis of the experimental results of all ex vivo human skin samples was consistent with this finding.

Fig. 4

H&E histological sections of (a) a control site (without CSC and laser irradiation) and (b) a site sprayed by a $300\mathrm{ms}$ R-404a spurt without laser irradiation from the same ex vivo human skin sample. Bars: $100\mathrm{\mu }\mathrm{m}$ .

The degrees of thermal injury to the epidermis were scored in the range of 0 to 5, corresponding to the range from the no visible thermal injury to maximum degree of thermal injury to the epidermis observed from H&E histological sections: 0—no visible thermal injury [Fig. 5a ], 1—basal cell elongation and hyperchromia [Fig. 5b], 2—cytoplasmic vacuolization [Fig. 5c], 3—multiple basal lacunae formation [Fig. 5d], 4—partial basal layer separation [Fig. 5e], and 5—complete epidermal ablation [Fig. 5f].

Fig. 5

Representative H&E histological sections showing various degrees of laser-induced thermal injury to the epidermis: (a) no damage, (b) basal cell elongation and hyperchromia (curves, $D_0=6\mathrm{J}∕{\mathrm{cm}}^2$ , cryogen type R-404a, ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ ), (c) cytoplasmic vacuolization (arrows, $D_0=6\mathrm{J}∕{\mathrm{cm}}^2$ , cryogen type R-404a, ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ ), (d) multiple basal lacunae formation (arrows, $D_0=10\mathrm{J}∕{\mathrm{cm}}^2$ , cryogen type R-404a, ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ ), (e) partial basal layer separation (arrows, $D_0=13\mathrm{J}∕{\mathrm{cm}}^2$ , cryogen type R-404a, ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ ), (f) complete basal layer separation ( $D_0=15\mathrm{J}∕{\mathrm{cm}}^2$ , cryogen type R-134a, ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ ). Bars: $100\mathrm{\mu }\mathrm{m}$ .

Figure 6 shows the histological sections of two skin sites irradiated at the same laser irradiation parameters $(D_0=8\mathrm{J}∕{\mathrm{cm}}^2,{\tau }_{\mathrm{laser}}=1.5\mathrm{ms},\lambda =595\mathrm{nm})$ but one sprayed with R-134a [Fig. 6a] and the other with R-404a [Fig. 6b]. Cryogen spray durations in both cases were $300\mathrm{ms}$ . Partial basal layer separation was observed on the site sprayed with R-134a [Fig. 6a], but no obvious thermal injury occurred on the site sprayed with R-404a [Fig. 6b].

Fig. 6

H&E histological sections of two skin sites irradiated at the same laser irradiation parameters ( $D_0=8\mathrm{J}∕{\mathrm{cm}}^2$ , ${\tau }_{\mathrm{laser}}=1.5\mathrm{ms}$ , $\lambda =595\mathrm{nm}$ ) but precooled with R-134a and R-404a at ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ , respectively: (a) R-134a, (b) R-404a. Oval: Epidermal damage of partial basal layer separation. Bars: $100\mathrm{\mu }\mathrm{m}$ .

Figure 7 presents the average threshold radiant exposures for irreversible thermal injury to the epidermis $D_{0,\mathrm{thr}}$ with $D_{0,\mathrm{thr}}$ defined as the minimum radiant exposure that induced a thermal injury to the epidermis with a score $\ge 2$ . It was considered that an epidermal damage score of 2 could possibly lead to pigmentation change, and an epidermal damage score higher than 2 would induce cell necrosis.¹⁵ With ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , values of $D_{0,\mathrm{thr}}$ were $4.8\mathrm{J}∕{\mathrm{cm}}^2$ for R-134a, and $5.2\mathrm{J}∕{\mathrm{cm}}^2$ for R-404a [Fig. 7a], but the difference was not statistically significant (one-tail $t$ test $p=0.3$ ). With ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , values of $D_{0,\mathrm{thr}}$ were 4.8 and $6.8\mathrm{J}∕{\mathrm{cm}}^2$ for R-134a and R-404a, respectively [Fig. 7b], and the difference was statistically significant (one-tail $t$ test $p=0.04$ ). With ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ , values of $D_{0,\mathrm{thr}}$ were 6.4 and $9.4\mathrm{J}∕{\mathrm{cm}}^2$ , respectively, for R-134a and R-404a [Fig. 7c], and the difference was statistically significant (one-tail $t$ -test $p=0.04$ ).

Fig. 7

Comparison of the threshold radiant exposures at 595-nm wavelength for irreversible thermal injury to the epidermis $D_{0,\mathrm{thr}}$ in response to CSC with R-134a and R-404a at various spurt durations: (a) ${\tau }_{\mathrm{CSC}}=100\mathrm{ms}$ , (b) ${\tau }_{\mathrm{CSC}}=200\mathrm{ms}$ , (c) ${\tau }_{\mathrm{CSC}}=300\mathrm{ms}$ .

3.3.

Histological Evaluation of Thermal Injury to Dermal Blood Vessels

We assessed the effects of CSC on laser photothermolysis of dermal blood vessels using the in vivo skin on rabbit external ears. Figures 8a, 8b, 8c are the H&E histological sections of the sites irradiated at the same irradiation parameters $D_0=13.5\mathrm{J}∕{\mathrm{cm}}^2$ , $\lambda =585\mathrm{nm}$ , but the site in Fig. 8a was without CSC, while the sites in Figs. 8b and 8c were cooled with a 100-ms cryogen spurt of R-134a and R-404a, respectively. All sites were excised $2\mathrm{h}$ after the irradiation. Only the 2-h sites were selected here as the residual thermal injury in the 4-day sites was minimal and the differences between the 4-day sites were not distinguishable due to significant wound healing effects in rabbit ear vasculature.

Fig. 8

Influence of CSC on the thermal injury to dermal blood vessels in rabbit ears: (a) no CSC, (b) $100\mathrm{ms}$ R-134a spurt, (c) $100\mathrm{ms}$ R-404a spurt. Ovals: the most superficial blood vessels that were thermally damaged. $D_0=13.5\mathrm{J}∕{\mathrm{cm}}^2$ . $\lambda =585\mathrm{nm}$ . Bars: $200\mathrm{\mu }\mathrm{m}$ . The sites were excised $2\mathrm{h}$ after the laser irradiation.

It was found that CSC increased the depth at which the most superficial thermal injury to the blood vessel occurred (regardless of the degree of thermal injury) (ovals), implying that CSC applied at these parameters resulted in some cooling of superficial blood vessels. The depths of the most superficial thermal damage to dermal blood vessels for the sites shown in Figs. 8a, 8b, 8c were approximately 70, 165, and $270\mathrm{\mu }\mathrm{m}$ , respectively. Table 2 also indicates a general tendency that the sites with cooling (sites 1 to 12) show deeper locations of the most superficial thermal injury to the blood vessel than those without cooling (sites 13 and 14). Fluctuations in the depths of the most superficial vascular damage in Table 2 with increased cryogen spurt durations were believed to be caused by the biological variability, e.g., the distributions of dermal blood vessels in different sites. There were no significant differences in the depths of the most superficial thermal injury to the blood vessel between the sites cooled by R-134a and those by R-404a (paired $t$ test for comparing means,³⁵ $p=0.96$ ), implying that using R-404a instead of R-134a may not interfere with the spatial selectivity of CSC.

Table 2

Effect of CSC on the depth of thermally damaged blood vessels.