22-22X/85/ 853-274$2./ THE JOURNAL OF INVESTIGATIVE DERMATOLOGY, 85:274-278, 1985 Copyright 1985 by The Willia ms & Wilkins Co. Printed in U.S.A. Lo-Fluene C 2 Laser Irradiation: Seletive Epidermal Damage to Human Skin* BRINDA R. KAMAT, M.D., STEPHEN V. TANG, M.D., KENNETH A. ARNDT, M.D., ROBERTS. STERN, M.D., JOEL M. NOE, M.D., AND SEYMOUR ROSEN, M.D. Departments of Pathology, Dermatology, Surgery, and the Laser Unit, Beth Israel Hospital and Harvard Medial Shool and the Charles A. Dana Researh Institute, Boston, Massahusetts, U.S.A. The interation of normal human skin ith lo-fluene C 2 laser irradiation as studied using a threephase approah. In phase one, freshly exised skin as observed immediately after impat. In phase to, skin irradiated 2 h prior to exision as studied. In phase three, human volunteers ere irradiated and biopsied at time zero, 24 h and 48 h. Seventy-five sites ere exposed and 6 biopsies ere performed. The earliest histologi hanges ere observed in the 6-1 J/m 2 fluene (radiant exposure) range and these hanges inluded spindle and vauolar hanges in the basal layer of the epidermis. Papillary dermal oagulation as present to a maximum of.3 mm. At fluenees of 1-25 J /m 2, superfiial dermal nerosis (.6-.8 mm) as observed. At fluenes above 25 J/m 2, transepidermal nerosis as present ith inreasing papillary dermal nerosis that as in proportion to the energy density delivered. At 2 h, basal vauolar hanges ere aompanied by diffuse keratinoyti ell death here ontat as maintained beteen the epidermis and dermis, hile here separation ourred limited keratinoyti death as observed. The earliest hanges ourred at loer threshold fluenes (4-6 Jjm 2 ). After 24 h, these doses resulted in extensive epidermal nerosis ith foal aute inflammatory infiltrates. At 48 h, the degree of epidermal "slough" as proportional to the energy density delivered and as maximal ith a fluene of 5. 7 J/m 2 delivered hereas ith a fluene of 3.8 J /m 2 thin slough (.2 mm) as observed. These findings suggest that lo-dose C 2 laser irradiation may provide a ne approah to seletively damage the epidermis ith minimal dermal damage. The linial use of C 2 laser on human skin involves exposure of target sites until a gross linial end point suh as utting, oagulation, harring, or vaporization is ahieved. Lo threshold effets suh as immediate or delayed erythema have reeived only limited investigation by U.S. government laser groups [1,2]. No studies have been arried out to examine the histologi dose response of human skin at timed intervals after exposure to energy fluenes (radiant exposures) ell belo those typially used in linial treatment. This investigation as designed to haraterize the dose/response of normal in vitro/in vivo human skin to lo-fluene C 2 laser, treatment Manusript reeived July 1, 1984; aepted for publiation February 28, 1985. * Presented in part at the Annual Meeting of The Soiety for Investigative Dermatology, In., Washington, D.C., May 7-9, 1984, and has been published in abstrat form (J Invest Dermatol 82:429, 1984). Reprint requests to: Seymour Rosen, M.D., Department of Pathology, Beth Israel Hospital, 33 Brookline Avenue, Boston, Massahusetts 2215. Abbreviations: D-E: dermal-epidermal dose levels at hih little or no immediate gross alterations are observed. The C2 laser avelength of 1.6 ~m is equally absorbed by all epidermal ells exept stratum orneum ithin the first 1-2!Lm of tissue, and subsequent damage to underlying tissue is by heat ondution. Thus, our hypothesis as that at lo fluene levels, seletive damage, limited primarily to epidermis, might be ahieved. Our studies indiate that (1) it is feasible to seletively damage the epidermis ith lo-dose C 2 at dosages here gross lesions may or may not be observed; (2) the dermal-epidermal (D-E) juntion seems to be a ritial interfae ith regard to the earliest disernable histologi lesions. MATERIALS AND METHODS The study design is shon in Table I. Equipment A oherent C 2 laser, model45xl, as used in defoussed gaussian mode ith irular spot sizes of 2 mm and 4.5 mm. Exposure times by shutter ranged beteen.1-1. s. Irradiane as varied from 2.5-2 W ith exposure to doses (fluenes) ranging from 1.6-23 Jfm 2 Material Freshly exised skin and human volunteers ere used in a three phase approah. The first phase utilized freshly exised normal breast tissue (1-15 min post exision kept on saline-soaked gauze at 3"C) to study immediate histologi hanges in 2 speimen sites exposed to fluenes 3.1-23 Jfm 2 The seond phase used in vivo breast skin at 35"C, exposed in 12 sites to fluenes 1.6-9.4 Jfm 2 2 h prior to mammoplasty to define appropriate dose ranges for more extensive in vivo studies. The third phase used 3 human volunteers exposed to fluenes 1.9-5.7 J/m 2 in 3 buttok skin sites at 33-34"C ith subsequent 3- to 4-mm punh biopsies taken at h (13 sites), 24 h (5 sites), and 48 h (12 sites). Dosing Parameters Phase I: Spot sizes 2 mm and 4.5 mm;.1 s for fluenes of 6.4, 15.8, 23.9, and 31 J/m 2 ;.2 s for fluenes of 3.1, 12.7, 31, 47.8, and 63 J/ m 2 ; 1. s for fluenes of 15.8, 31, 63, 159, and 23 J/m 2 Phase II: Spot size of 4.5 mm;.1 s for fluenes of 1.6 3.1, 4.7, 6.3, 7.9, and 9.4 J/m 2 ;.2 s for fluenes of 3.1, 6.3, 7.9, and 9'.4 J/m 2 Phpse III: Spot size of 4.5 mm;.1 s for fluenes of 1.9, 2.8, 3.8, 4.7, and 5.7 Jfm 2 Histology All tissue as fixed immediately in 1% formalin and proessed for histologi setions ut at 4-6!LID. Setions ere stained ith hematoxylin and eosin. RESULTS Gross hanges in Phase I involved initial hitening (oagulation) of skin at doses slightly above 1 J/m 2 to frank vaporization and harring of epidermis at higher fluenes. In Phase II, delayed (5-1 min) erythema ourred for fluenes 4-6 J / m 2. More obvious oagulation and blister formation ourred at higher fluenes. In Phase III, only delayed erythema ourred immediately hih ontinued into the 24-h sample time and resulted in mild blister formation at 48 h for the fluenes 5-6 J/m 2 The results are shon in Table I. 274
Sept. 1985 SELECTIVE EPIDERMAL DAMAGE BY LASER 275 TABLE I. Study Design and Su.mma r:v of Results PHASE I IN VITRO RESULTS Initial determination of fluene ra nge for da mage limited to epide rmis Use freshly exised breast s kin ( l) "Threshold" histologi injury observed in 6-1,J jm 2 fluene ra nge Time T emp 3' C (2) Diret rela tion beteen fluene de li v 3. 1-23 J /m 2 e red a nd dept h of derma l da ma"e a ll phases "' ' (l,ll,lll) (see Fig:)) PHASE II IN VIVO Trial of this fluene ra nge o n in vivo breast Time 2 h tissue irradiated 2 h prior to eletive a ug- T emp 35 mentatio n ma mmoplasty I.6-9.4 J jm 2 (1) "Threshold" histologi injury observed in 4-6 J / m 2 fluene ra nge PHASE Ill In vivo timed biopsies of butto k skin ith punh biopsies IN VIVO Time 24 h 48 h T emp. 33- :34 1.6-5. 7 J /m 2 (1) "Threshold" histologi injury observed in 4-6 J /m 2 flu ene ra nge a t h (2) Thikness of epiderma l slough af 48 h diretly related to flu ene delivered (see Fig 6) FIG 1. In vi t ro histologi ha nges. A, F'luene 6.3 J / m 2 B, F'luene 31.4 J / m 2 At lo ene rgy levels (A ) epiderma l struture an be defin ed but bulla fo rmation is noted on the ri ght; the papilla ry dermis is auterized in a band! ike area of approximately.2 mm t hikness. With highe r e ne rgy levels (B) the epidermis and superfiia l papill a ry dermis has been vaporized. (Hematoxylin and eosin, A x 1, B X 4.) Phase One (In Vitro) T he initial histologi hanges ere observed in t he 6-1 J / m 2 fluene range; belo this range both epidermis and dermis appeared histologially unaffeted. The earliest effet onsisted of a spindle and vauolar hange in t he basal layer. Papillary dermal hanges ere either absent or limited to a oagulative hange in the most superifial aspet (.1-.3 mm) of t he papillary dermis. The epidermis above the basal layer appeared to be unaffeted (Figs 1A,2A ). These seletive basal and papillary dermal hanges ere noted along prolonged lengths of skin (Fig 2A,B) and in some ases ere assoiated ith a spli t at the D-E juntion ith vesile formation (Fig 2C). The stripped epidermis forming the vesile ontained a morphologially unaffeted upper epidermis (Fig 2D ). At higher flu ene ranges (1-25 J /m 2 ), basal hanges desribed above ere aompanied by inreasing degrees of papillary dermal nerosis (.6-.8 mm). At fluene ranges above 25 J /m 2, the retiular dermis (up to.75 mm at 37 J /m 2 ) also shoed oagulative nerosis (Fig 1B). Thus the depth of dermal nerosis as related to the flu ene delivered (Fig 3). Phase To (in Vivo) (2 Hours) In this experiment, the in vivo thermal effets ere doumented at 2 h post laser. Here the threshold f1uene for basal
276 KAMAT ET AL FIG 2. Histologi observations at 2 h. A and B, Fluene 9.4 J/m 2. C and D, Fluene 7.9 J/ m 2 E, Fluene 7.9 J /m 2. At lo poer (A) diffuse hanges are noted at the D-E juntion. B, At higher magnifiation (area in A marked ith arro) basal vauolization and spindle hanges are noted ith an intat superfiial epidermis. In C, epidermal separation is more extensive than in (A) and bu llae formation is observed. D, Higher magnifiation (area in C marked ith arro) reveals hanges similar to (B) but the basal aspet or the epidermis is ell defined. Where ontinuity beteen epidermis and dermis is maintained, vauolar hanges are noted in the basal layer hile the upper epidermis has beome aidop hili ith loss of nulea r defi nition (E). (Hematoxylin and eosin, A X 1, B X 4, C X 4, D X 4, E X 4.) ~ ~..J!;(.5.4.3 5 1 2 4 8 1 15 2 3 C 2 LASER FLUENCE (J/CM 2 ).Rm. Sale FIG 3. At time equal h, depth of dermal damage (vertial axis in mm from the D-E juntion) vs COt laser flu ene (horizontal axis on natural log sale in J/m"). The plolted straight line has a orrelation oeffi ient or.9!)4 and a oeffi ient of determination of.91. Note the region of minimal papillary dermal damage for fluenes 1 J/m". This orresponds (see text) to the region of seletive epidermal damage. layer hanges as loered to 4-6 J/m 2. In addition, evidene of diffuse ell nerosis in the upper layers of the epidermis as noted in areas, here ontat ith dermis as maintained. The ytoplasm of these keratinoytes in this situation appeared to beome eosinophili ith foa l loss of nulear definition. (See Figs 2E, 4B.) In areas, separation of the epidermis from t he dermis as noted and t he superfiial epidermis appeared relatively intat (Fig 4C). Phase Three (In Vivo) (, 24, 48 Hours) At h, the basal hanges desribed above (i.e., vauolar and spindle hanges in the basal layer) ere again seen in t he 4-6 J/m 2 range and appeared similar to those observed at these doses hen biopsy as done 2 h after exposure (Phase II) (Fig 4A). At 24 h, hoever, t he superfiial epidermis appeared to undergo nerosis in large sheets (Fig 4D,E). In addition, a sparse inflammatory infiltrate (neutrophils and lymphoytes) not seen at or 2 h as noted, loated predominantly in the papillary dermis and overlying neroti epidermis. At 48 h, regenerative epithelium as observed groing in from the edges along t he surfae of t he denuded lasered area and lifting off the neroti epidermis ("epidermal slough") (Fig 4F). The thikness of the slough as diretly related to the fluene (Fig 5) at these lo levels. Thus ith a fluene of 3.8 J/m 2 the slough as thinnest and ith a tluene of 5.7 J /m 2 it as thikest (see Figs 5A- C,6). In addition, areas of omplete denudation ere present at higher tluenes (5.7 J/m 2 ) but not at loer f1uenes (3.8 J/m 2 ). DISCUSSION These results indiate that the C2 laser radiation administered at lo fluenes that give little or no immediate hanges in t he skin does ause substantial and seletive ha nges in the skin. Our studies sho t hat the greatest initial histologi effets ith lo-t1uene C 2 laser our in the basal layer and have a planar distribution. This finding is somehat unexpeted sine the C 2 laser energy is ell absorbed by any ater-ontaining ell hih inludes all epidermal ells exept the stratum orneum. Therefore, the superfiial epidermis is theoretially lia-
Sept. 1985 SELECTIVE EPIDERMAL DAMAGE BY LASER 277 FIG 4. Histologi hanges for flu ene 4.7-6.3 J /m 2 from in vivo buttok skin. Evolution over t ime: A = h, 4.7 J /m 2 ; B = 2 h, 4.7 J /m 2 ; C = 2 h, 6.3 J /m 2 ; D and E = 24 h, 4.7 J /m 2 ; F = 48 h, 5.7 J /m 2. At. t ime ze ro, i t h fluene 4.7 J / m 2, basal vauolar ha nges are noted (A). After 2 h, basal vauola r ha nges a re aompanied by keratinoyti nerosis (B). Where early bullae formation has ourred, t he most pronouned ha nges ou r in the basal region (C). At 24 h (dermis and epidermis), extensive epidermal ne rosis is noted aompanied by a foal aute inflammatory infiltrate. At fluene 5.7 J / m 2, extensive epidermal nerosis is noted at 48 h (F). (Hematoxylin and eosin, X 4.) FIG 5. Epidermal "slough" at 48 h from in vivo buttok skin: A= 3.8 J /m 2 ; B = 4.8 Jf m 2 ; C = 5.7 J /m 2. At flu ene of 3.8 J / m 2 (A) a t hin slough is noted (.2 mm). At. fluene of 4.8 J /m 2 (B) the slough is more extensive (.45 mm). At tluene of 5.7 J / m 2 (C) the entire epidermis is ne roti (.6 mm). (Hematoxylin and eosin, x 4.) ble to the highest exposure dose due to the proximity to irradiation and t he thermal diffusion limitation provided by the air-skin in terfae. The strutural a nd biologi inhomogeneity present at the D-E juntion may be fators for this prediletion for damage to the D-E juntion. Similar hanges have been observed at the D-E interfae using lo-dose irradiation ith the argon laser [3]. In addi tion, an inherent eakness in the region of the lamina Iuida may play a role in vesile formation [4]. At h, threshold injury as observed at loer flu ene levels in the vivo studies (3.8-5.7 J /m 2 ) as ompared ith in vitro studies (6-1 J /m 2 ). These hanges in threshold fluenes probably relate to t he effets of the higher temperatures main tain e d in vivo. At 2 h, extensive keratinoyte nerosis as observed. Hoever, it is interesting to note that histologi expression of keratinoyte ne rosis depended on the maintenane of ontat beteen dermis and epidermis; in ases here the epidermis as detahed, the upper epidermis failed to reveal nerosis at 2 h. lt thus appears that epidermal separation from the dermis delayed t he expression of histologi ell death. This effet may have to do ith prevention of edema (no exhange ith the dermal ompartment) or the lak of availability of ertain fators (e.g., alium) neessary for t he expression of injury 15]. At 24 h, t he extent of injury learly extended above t he basal layer. The nerosis of superfi ial epidermis at t his point must relate to nutrient fl o barriers via the damaged basal layer in
278 KAMAT ET AL l: (!) ::::>...J 1/)...J ii: ::::> l: o..: 1. 2. 3. 4. 5. 6. C 2 LASER FLUENCE (J/ CM 2 ) FIG 6. At time equal 48 h, fo r in vivo skin, t he thikness of epid!' rmal slough (vertial axis in mm) vs C 2 laser f1uene (horizontal axis in J / m 2 ). T he plolled straight line has a orrelation oefli ient of.924 a nd a oeffiient of determination of.854. addition to therm al damage. Regenerative ativity as also present at t his time. At 48 h, t he epidermis appeared reonst ituted but a superfi ial ne roti epidermal slough as present. The thikness of t hi s slough at lo flu ene levels as diretly related to absolu te!luene. At higher flu enes (5.8 J/m 2 ) in this range, regeneration as ino mplete and areas of transepidermal ellular loss ere present. We hypothesize that at loer fluene levels (3.8 J/m 2 ) epidermal "kill " is inomplete and that regeneration may our from surviving epidermal elements as ell as skin appendages and lateral unaffeted epidermis. While a lear relation beteen flu ene and depth of papillary dermal oagulation as demonstrated for f1uenes 3.1-23 J / m 2 (Fig 3), it must be noted t hat throughout t hese experiments, at t he loer t hres hold fluenes, papillary dermal involvement as minimal and o nsisted of mild oagulative hanges in t he superfiial areas to a maximum depth of.3 mm, hih t hen disappeared, so t hat at 48 h no morphologi ev idene of injury remained. These hanges may relate to loss of ater initially and subsequent rehyd ration. This study as onduted in 3 phases and had a number of variables. In part, the latter as due to finding the appropriate!luene ranges for the fin al experiments on human vo lunteers. T he differene in sites irradiated as of no onsequene sine epidermal thikness as equivalent in both a reas (.7-.8 mm). Spot size variation (2. mm and 4.5 mm) ere not important sine e ere dealing ith a planar heat effet of 2 1-Lm thikness; in Phase III experiments, a onstant spot size (4.5 mm) as utilized. Limitations in total poer available for our large spot sizes required that e use pulse durat ions of both.1 s and.2 s to ahieve a ide range of radiant exposures in P hase I. Despite these variations, e ere able to find exellent orrelation beteen t1uenes delivered and depth of dermal damage as ell as thikness of epidermal slough at 48 h. Phase II in vivo data together ith unpublished data shoed that histologi damage did not follo a reiproity relationship- that is to say, for a given radiant exposure, lengthening the exposure t ime ith reiproal loering of poer results in redued epidermal damage. This is onsistent it h a t hermal me hanism of damage and not ith a photohemial mehanism. This hypothesis is supported by the inrease in fluene required fo r "threshold" injury during our in vitro experiments and by subsequent manipulations of surfae temperature, hydration, and thikness of stratum orneum [6] hih e have initiated to eluidate a damage model based on putative t hermal mehanisms. While important to ompare the effets of knon t hermal injury to that of equivalent C 2 laser injury, delivery of a small amount of energy equivalent to.1 s using onventional heat soures may pose an implementation problem. Thus the C 2 laser may be used to indue seletive epidermal damage in normal skin ith lo fluenes that may or may not reate a linial (gross) hange. T he histologi evidene of thermal damage begins in the basal layer but is then ev ident in the entire epidermis, hih then rege nerates itself ithin 48 h. T he minimal involvement of the papillary dermis may help minimize t he possibility of sarring. This finding suggests that the lo-dose C 2 laser may be useful in t he t reatment of epidermal lesions. REFERENCES 1. Rokell RJ, Goldman L: Researh on human skin laser damage thresholds. U.S. Airfore Aerospae Medial Division, T exas, NTIS #AD-A1 2 73:83-95, 1974 2. Williams JPG, Carter SD, Pavkov KJ, Boatman JB: Investigation of laser skin hazards. Battelle Memori al Institute, Ohio, NTIS #AD 735-794, 1971 3. Greenald J, Rosen S, Anderson AR, Harrist T, MaFarland F, Noe J, Parrish JA: Compa rative histologial studies of the tunable dye (at 577 mm) laser and argon laser. The speifi vasular effets of t he dye laser. J Invest Dermatol 77:35-31, 1981 4. Klein GF, Hintner H, Shuler G, Fritsh P: Juntional blisters i.n aquired bullous disorders of the dermal-epidermal juntion zone: role of the lamina Iuida as t he mehanial lous minoris resistent iae. Br J Dermatol 19:499-58, 1983 5. Fa rber JL: Biology of disease: membra ne injury and alium homeostasis in the pathogenesis of oagulative nerosis. Lab Invest 47:114-23, 1982 6. T a ng SV, Kamat B, Arndt KA, Rosen S, Noe JM, Stern RS: Lo energy densi ty C 2 laser irradiation auses in traepidermal damage. Lasers Surg Med 3:329, 1984