Am J Ololaryngol 10:76-a4, 1989

Plasma-Mediated Excimer Laser Ablation of Bone: A Potential Microsurgical Tool

RAJABRATA SARKAR, BS, RICHARD L. FABIAN, MD, ROGER C. Nuss, MD, AND

CARMEN A. PULIAFITO, MD

A pulsed ultraviolet excimer laser was used to ablate bone in vitro at 193,248,308, and 351 nm and in vivo at 193 nm. Ablation was dependent on sufficient fluence (energy delivered per unit area per pulse) for plasma formation at the target site at all wavelengths. Adjacent tissue damage at various fluences for each wavelength was examined using a light mi- croscope. Damage was minimal at 193 nm (1 to 3 km) and most extensive at 351 nm (60 to 75 pm). This is in sharp contrast to the 1 to 3 mm of adjacent thermal damage produced when carbon dioxide lasers are used to ablate bone. Differences in the degree and type of damage to adjacent tissues among the wavelengths studied indicated that other ablation mechanisms and tissue interactions are involved in addition to simple plasma vaporiza- tion of bone. Bleeding during and after ablation demonstrated that the use of this laser does not cause thermal damage, which would cauterize adjacent vessels. Pre- and post- mortem lesions made at identical power and pulse settings were of equal depth, indicating that bleeding does not affect the ablation rate. The excimer laser has potential as a microsurgical instrument for the precise removal of bone with minimal damage to adja- cent structures. AM J OTOLARYNGOL 10:76-84. 0 1989 by W.B. Saunders Company. Key words: ultraviolet laser, bone ablation, plasma.

Laser technology is currently used in otolaryn- gology mainly for soft tissue surgery such as the removal of polyps, granulomas, and malignan- cies. The widely used continuous-wave carbon dioxide (CO,) laser uses water, generating heat which vaporizes soft tissues and provides con- comitant hemostasis by cauterizing adjacent ves- sels.

Many surgical procedures in otolaryngology require the removal of bone in areas adjacent to sensitive structures where heat generation would cause damage or impair healing. Exam- ples include the incision of the bony canal re- quired for optic or facial nerve decompression, skull base and middle ear surgery, and resection

Received July 25, 1988, from the Department of Otolaryn- gology-Head and Neck Surgery and the Laser Research Lab- oratory (Department of Ophthalmology), Massachusetts Eye and Ear Infirmary, Boston. Accepted for publication October 19, 1988.

Presented at the 1988 Midwest Medical Student Research Conference, Omaha, NE.

Address correspondence and reprint requests to Richard L. Fabian. MD. Deuartment of Otolarvngolo~-Head and Neck Surger$, Massachusetts Eye and l?ar-Infirmary, 243 Charles St, Boston, MA 02114. 0 1989 by W.B. Saunders Company. 0196-0709/89/1002-0004$5.00/O

of the ethmoid or sphenoid sinus disease in proximity to cranial nerves, major blood vessels, or the pituitary gland.

When continuous-wave 10.6~pm CO2 laser en-

ergy is applied to compact bone,* low water content causes desiccation followed by car-

bonization. Studies have shown a delay of ap-

proximately 2 weeks in the healing of osteoto-

mies made with a CO, laser, compared with me- chanical osteotomies.l Thermal necrosis and carbonization generated by the CO, laser adja- cent to the incision are responsible for this delay as well as for foreign body reactions to charred material. The zones of necrosis and charring ad- jacent to continuous wave CO, laser ablation sites made in bone measure 1 to 2 mm.’ Another drawback of CO, laser bone excision is the need to continuously cool the bone surface with ni- trogen gas to disperse tissue vapor and prevent tissue firesq3p4

Excimer lasers are high power sources of pulsed ultraviolet light. Investigations of poten- tial surgical applications have demonstrated their ability to make precise incisions in the cor- nea, lens, and skin.5-7 At 193 nm, there is a strik- ing absence of thermal damage to areas adjacent to the ablation site when an excimer laser is used

76

SARKAR ET AL

to ablate soft tissue. The 193-nm excimer laser is believed to ablate via “photodecomposition,” which involves the direct breakage of molecular bonds by high energy ultraviolet photons’ rather than by a thermal mechanism.

Our studies were designed to determine the minimum fluence for the effective ablation of bone, the efficiency and rate of ablation, the type and extent of damage adjacent to the ablation site, and the effect of bleeding from living tissue. Studies were performed at the four primary excimer laser wavelengths: 193, 248, 308, and 351 nm.

MATERIALS AND METHODS

In Vitro Studies

Fresh calvarium was removed from sacrificed Hartley strain guinea pigs. The dura mater and periosteum with its associated muscular attach- ments were removed. Specimens were kept in a saline-moistened gauze before and between ab- lations. The guinea pig calvarium consists of rel- atively thin (1 mm) compact bone with easily discernible medullary areas. Areas of the calvar- ium uniform in thickness and flatness were cho- sen for ablation. A total of four guinea pig crani- ums were required for in vitro study.

An excimer laser (Questek series 2200, Biller- ica, MA) was used to irradiate the bone speci- mens at wavelengths of 193 nm (ArF gas fill, 15 ns pulse duration), 248 nm (KrF gas fill, 20 ns pulse duration), 308 nm (XeCI gas fill, 20 ns pulse duration), and 351 nm (XeF gas fill, 20 ns pulse duration).

A g

1.6

3 p 1.2 z .z

s 0.8 !! E 'a 0.4 -m

9 0

_;--“d 2 3 4 5

3 1.6

Y

g 1.2 E 2

1.2

0.6

0.4

0 2 3 4 5 2 3 4 5

Fluence (log mJ/cm 2 ) Fluence (log mJ/cm 2 )

1.6

1.2

0.6

0.4

0

At each of the four wavelengths, lesions were made varying in fluence from the maximum available, which varied among wavelengths, to below the minimum necessary to ablate the bone. Each point on the ablation rate graphs (Fig 1) represents a lesion. When necessary, attenua- tors were used in conjunction with the laser mi- crocomputer to reduce the pulse energy to de- sired levels. All ablations were performed at a pulse rate of 10 Hz.

The experimental apparatus is illustrated in Fig 2. A fused silica, ultraviolet grade, 30.5 cm focal length, spherical lens was used to focus the beam onto a 1 mm by 200 pm mask placed di- rectly over the bone surface. The mask was ap- proximately 22 cm behind the lens. The imprint of the beam on a piece of unexposed photo- graphic paper was examined with a measuring microscope to assure uniform coverage of the slit area by the focused beam. Energy measurements were made with a power meter (Scientech #362/ 360001, Boulder, CO) placed behind the mask. Pulse to pulse energy variation was < 10%.

A silicon photodiode (EG+ G FNDlOOQ) was placed 50 cm behind the mask and sample. The output from the photodiode was displayed on an oscilloscope whose triggering signal was pro- vided by an external synchronization signal from the laser. The time necessary to perforate the specimen was obtained by timing the change in the oscilloscope signal. The thickness of the bone at the ablation site was determined with a needle point micrometer (Starrett #210-A, Athol, MA).

After ablation, the bone specimens were im- mediately fixed in a 2% glutaraldehyde, 4%

1.6

Figure 1. Ablation rate as a function of logarithm of flu- ence for the four excimer wavelengths: (A) 193 nm, (B) 248 nm, (C) 308 nm, and (II) 351 nm.

Volume 10

Number 2

March 1989

77

PLASMA-MEDIATED LASER BONE ABLATION

trigger signal ,

Excimer Laser

30

. . . . . . . . . . . . . . . .

Attenuators

Oscilloscope

Figure 2. Diagram of apparatus used to determine ablation rates. Attenuators ad- just laser output and lens focuses beam onto bone behind mask. Photodiode and oscilloscope respond to beam at moment of perforation.

Hask with sample

paraformaldehyde, 0.1 M sodium cacodylate buffer solution. Specimens were decalcified for four to seven days in hydrochloric acid decalci- fying solution, embedded in JB-4 glycol methac- rylate, cut at 3 km, and stained with Stevenol’s blue for light microscopy.

In Viva Studies

A single female Hartley strain guinea pig was placed under deep anesthesia with a 3:1 (vol/vol] mixture of 100 mg/mL ketamine per 20 mg/mL xylazine administered intramuscularly at a dos- age of 0.4 mL/kg. Calvarium was exposed by el- evating the overlying skin and connective tissue. Hemostasis was achieved with portable electro- cautery. The periosteum was elevated. All ex- posed tissues were kept moist with saline.

Lesions were made in the calvarium at a wave- length of 193 nm using the same optical appa- ratus as in the studies discussed above. Lesions were made at 10 Hz, the number of pulses was 1,200 for all lesions, and two fluences were used, 10,000 mJ/cm’ and 1,000 mJ/cm’. Additional an- esthesia was administered intermittently throughout the procedures. The animal was killed immediately with a l.O-mL intracardiac injection of 50 mg/mL sodium pentobarbital. The calvarium was removed and, in a different area, another series of lesions identical in fluence and number of pulses to those made before sacrifice were made for comparison. The specimen was fixed and processed for light microscopy as pre- viously described.

RESULTS

At all wavelengths, ablation of bone did not occur without plasma formation. Thus, the

American threshold fluence for plasma formation coin-

Journal tides with the threshold for ablation of the sam-

of ple. Plasma was characterized by a momentary,

Otolaryngolcgy often large (2 cm), orange-pink glow which ex-

78

panded from the sample surface toward the laser beam accompanied by a distinct photo-acoustic phenomenon.

The ablation rate v the logarithm of the fluence is plotted for each wavelength in Fig 1. The co- efficients of correlation (R’) were .884, .747,

.971, and .968 for 193 nm, 248 nm, 308 nm, and 351 nm, respectively.

Light microscopy of the lesions demonstrated that the degree of histologic damage adjacent to the lesion varied with both wavelength and flu- ence. A typical lesion is shown in Fig 3. Adja- cent damage was histologically determined to include fractures and fragmentation, loss of def- inition of the organic matrix, and changes in staining characteristics.

At 193 nm, an irregular darker zone measuring 2 to 3 pm in thickness was found at a fluence of 21,000 mJ/cm’ [Fig 4A). At a lower fluence, 950 mJ/cm’, the lesion edge was more uniform, with 1 to 2 km of adjacent damage. At 248 nm, there was 3 pm of damage at a fluence of 1,350 mJ/cm*. At a fluence of 16,500 mJ/cm*, the zone of dam- age measured 30 p,rn (Fig 4B). At 351 nm, ap- proximately 60 to 75 km of damage was seen at fluences of 6,000 mJ/cm* and 11,500 mJ/cm* (Fig 4C). At 308 nm, the relationship between fluence and the extent of adjacent damage was reversed. Lesions produced with a fluence of 17,500 mJ/ cm* showed 2 pm of damage, while a fluence of 3,850 mJ/cm* produced 15 p,m of damage as well as lines resembling fractures radiating into the damaged area (Fig 4D).

In summary, 351 nm produced the greatest amount of adjacent damage, approximately 60 to 75 pm regardless of fluence. 248 nm produced considerable (30 pm) damage at high fluences but at low fluences was comparable to the very slight damage (1 to 3 pm) seen for all fluences at 193 nm. Interestingly, at 308 nm, more damage was seen at lower fluences, but in all cases re- mained low (2 to 15 pm) compared with 248 and 351 nm. At all wavelengths, higher fluences al- tered the shape of the entire lesion from a narrow trench to a wider, more rounded incision.

SARKAR ET AL

Figure 3. Typical excimer lesion of bone; 308 nm, 3,800 mJ/cm’. (Steven01 blue stain; magnification X11.5.)

In the live animal, lesions made in visible medullary areas bled during and after ablation while those made in purely compact bone areas did not. Histologic damage was similar to in vitro specimens as discussed above. Light mi- croscopy showed that in vivo lesions (Fig 5, top) were more rounded in overall shape than their postmortem counterparts (Fig 5, bottom) and clots were seen adhering to the walls of the le- sion. The depth of lesions made before and after sacrifice was identical.

DISCUSSION

The highest ablation rate in our study (1.4 Fm/pulse) was achieved with 248 nm. The abla- tion rates at 248, 308, and especially 351 nm increase sharply with increasing fluence, which suggests that a more powerful laser could achieve a higher ablation rate at these wave- lengths.

The threshold fluence at which ablation be- gins varied among wavelengths, showing an in- crease with increasing wavelength. Three hun- dred fifty-one nanometers required 16 times the fluence of 193 nm (5,500 v 335 mJ/cm’), with 248

nm and 308 nm exhibiting threshold fluences of 1,350 and 2,250 mJ/cm’, respectively. The bone threshold fluence at 193 nm (335 mJlcm2) is sim- ilar to the threshold fluences at that wavelength for the ablation of aorta,g lens,” and skin.7 The necessity of plasma formation for the effective ablation of calcified tissue such as bone is sup-

ported by a correlation between our threshold fluences and those found for the plasma- mediated excimer ablation of calcified athero- sclerotic plaque at 308 and 351 nrn.l’.‘l

Plasma formation occurs when sufficient en- ergy is present to strip electrons from atoms, cre- ating free positive ions and electrons. The rela- tion of plasma formation to energy is non-linear; there exists a threshold below which there is in- sufficient energy for any plasma formation. The physical mechanisms by which plasma gener- ated at the laser target site removes tissue are multifold. Plasma surface temperature reaches 15,000”C focally and vaporizes solids. Pressure waves also contribute by causing phase change, eg, vaporization or melting, and thermal expan- sion of tissue. Plasma growth after generation of free electrons is caused by the absorption of in- coming photons by free electrons in a cascade process known as inverse Bremsstrahlung.‘2r’3 This absorption prevents transmission of pho- tons to the target surface and is known as plasma shielding.

Plasma shielding explains the relatively small increases in ablation rate with respect to in- creases in fluence seen in our studies. At 193 nm, increasing the fluence loo-fold resulted in only a fourfold increase in ablation rate. Thus, most of the energy delivered at higher fluences is expressed in increased plasma size. A stronger correlation was seen between ablation rate and fluence at 308 and 351 nm; however, plasma shielding effects may only be apparent at flu-

Volume 10

Number 2

March 1989

79

PLASMA-MEDIATRDLASERBONB ABLATION

American

Journal

of Otolaryngology

80

SARKAR ET AL

Figure 4. Damage to tissue at the lesion edge: (A) 193 nm. 21,000 mJ/cm’; (B) 248 nm, 16,500 mJ/cm’; (Cl 351 nm. mJ/cmZ; (0) 308 nm, 3,850 mJ/cm’. The circle indicates the zone of thermal damage adjacent to the ablation crater. (5 blue stain; magnification x 132.)

D , 11,50r

Volur

NW-r

March

ne

Ibe

19

10

r2

89

81

PLASMA-MEDIATED LASER BONE ABLATION

Figure 5. In vivo (top] premortem and [bottom) postmortem lesions. Both were 1,200 pulses at 1,000 mJ/cm* at 193 nm. Note the presence of clot and change in shape of the premortem lesion, but the equal depth of both lesions. (Steven01 blue stain; magnification X 33.)

ences significantly higher than those attainable in our study.

The maximum adjacent damage in our studies (60 to 75 pm) was observed at 351 nm, and is in

American sharp contrast to the 1 to 3 mm of adjacent dam-

Journal age noted in studies using CO, lasers to ablate

of bone.‘,’ This difference may be due to different

Otolaryngolcgy

82

ablation mechanisms. The continuous-wave CO, laser continuously heats the target tissue to achieve vaporization and thermal denaturation of adjacent areas, whereas the excimer produces a plasma that is tightly confined to the point of incision, and has a nanosecond duration that limits thermal dispersion into adjacent tissue.

SARKARETAL

Damage from consecutive pulses is not cumula- tive because tissue damaged by one pulse is re- moved by subsequent pulses.

Studies using excimer lasers at subplasma flu- ences to ablate skin, cornea, and lens have shown correlations between absorption by the target tissue at a particular wavelength and ab- lation rate, and between threshold fluence and damage to adjacent tissue.5-7 Greater absorption leads to deposition of the laser energy in a smaller volume of tissue, resulting in both va- porization at a lower energy level by thermal or “photodecomposition” mechanisms and less dispersion of energy into adjacent tissue. At flu- ences sufficient for plasma generation, adjacent damage should be constant regardless of wave- length or fluence. Varying degrees of damage ob- served in our studies indicate that mechanisms besides plasma formation play a role in the bone ablation process. The minimum damage seen at 193 nm is similar to findings in studies of exci- mer ablation of the lens6 and corneae5

The results of our in vivo studies support the histologic evidence that indicates plasma abla- tion minimally affects adjacent tissues. A lack of hemostasis indicates that, unlike the CO, laser, excimer-generated plasma does not cauterize ad- jacent blood vessels. The identical depth of pre- and postmortem lesions indicates that bleeding does not have an effect on the ablation rate.

Studies using the excimer laser to ablate skin in vivo at subplasma fluences showed that abla- tion was terminated by bleeding. This is due to the significant attenuation of 19%nm light by sa- line solutions7 In our study, it appears that the generation of plasma fluence at the target surface eliminates light attenuation. Thus, the ablation process continues unimpaired in the presence of blood at the ablation site.

The choice of excimer laser wavelength has clinically relevant effects. The least adjacent damage (1 to 3 pm) is produced by 193 nm, al- though even the maximum damage observed (60 to 75 pm at 351 nm) is acceptable for many po- tential microsurgical procedures. The ablation rate also varied with wavelength. All wave- lengths produced a rate of approximately 1 Fm/pulse at the fluences used in our studies. Real-time ablation rates can be increased by in- creasing the pulse rate. Excimer lasers that oper- ate at 100 pulse/s are commercially available and the potential for even higher pulse rates exists.

Fiber optic capability is also dependent on wavelength selection. Current fiber optic tech- nology is not capable of transmitting ultraviolet

wavelengths below 200 nm, and transmission of the higher peak-power pulses used in bone ab- lation is a technical challenge at all excimer wavelengths.”

Wavelength selection is important in UV- induced mutagenesis and carcinogenesis. At a wavelength of 248 nm, ablation of the cornea has been associated with a significant increase in py- rimidine dimer formation, suggesting increased DNA damage. Ablation at 193 nm, however, did not show a similar increase.14 UV-induced mu- tagenesis and carcinogenesis during plasma bone ablation should be less significant for two reasons: the opacity and relatively acellular na- ture of bone limits the dispersion of applied UV light to potential target cells, and the minimal and noncumulative adjacent damage associated with the plasma ablation mechanism ensures that the amount of potentially affected tissue is small. However, this hypothesis has not been confirmed in the laboratory.

Given the fine degree of control of ablation rate demonstrated by fluence adjustment in our studies (Fig l), the excimer laser would appear to be best suited to applications in otolaryngology that involve the removal of fine amounts of bone adjacent to heat-sensitive structures. Contain- ment of the excimer laser beam within an artic- ulated arm system coupled to an operating mi- croscope, similar to current CO2 laser systems, would be the most effective method of delivery. The development of a fiber delivery system would enhance the clinical feasibility of the ex- timer laser. Ablation of the bony canal for optic and facial nerve decompressions, procedures in the internal acoustic canal, and paranasal sinus surgery are areas in which excimer laser ablation could be useful.

The excimer laser used in our studies was de- signed for research use and operates on standard 220 VAC current. It requires a tap water line for cooling and an external air vent for exhaust fan discharge. Future surgical applications will re- quire additional experimental work relating to tissue healing, the mutagenic effects of different excimer wavelengths, and photoacoustic phe- nomena, as well as improved power capabilities and delivery systems. Necessary precautions for operating room personnel include plastic UV- absorbing safety glasses, which are inexpensive, widely available, and do not interfere with nor- mal vision.15

Our studies demonstrate that the excimer laser can ablate bone in a controlled and quantifiable manner without necrosis and significant adja-

Volume 10

Number 2

March 1989

83

PLASMA-MEDIATED LASER BONE ABLATION

cent damage associated with CO2 laser bone ab-

lation.

Acknowledgment. David Stern provided technical assistance with lasers and electronics. Jay Walsh pro- vided animals and assistance. Ernest Dobi, PhD, and Winifred Reidy provided assistance with histology and photography. Vivian Taylor-Ayoub assisted in the preparation of the manuscript.

References

1. Gertzbein SD, deDemeter D, Cruickshank B, et al: The effect of laser osteotomy on bone healing. Lasers Surg Med 1981; 1:361-373

2. Pao-Chang M, Xiou-qi X, Hui Z, et al: Preliminary report on the application of the CO, laser scalpel for opera- tions on the maxillofacial bone. Lasers Surg Med 1981; 1:375-384

3. Clauser C: Comparison of depth and profile of osteoto- mies performed by rapid superpulsed and continuous- wave CO, laser beams at high power output. J Oral Maxillofac Surg 1986; 44:425-430

4. Small IA, Osborn TP, Fuller T, et al: Observations of carbon dioxide laser and bone bur in the osteotomy of the rabbit tibia. J Oral Surg 1979; 37:159-166

5. Puliafito CA, Steinert RF, Deutsch TF, et al: Excimer la- ser ablation of the cornea and lens. Ophthalmology 1985; 92:741-748

6. Nanevicz TM, Prince MR, Gawande AA, et al: Excimer laser ablation of the lens. Arch Ophthalmol 1986; 104:1825-1829

7. Lane RJ, Linsker R, Wynne JJ, et al: Ultraviolet laser ab- lation of skin. Arch Dermatol 1985; 121:609-817

6. Srinivasan R, Leigh WJ: Ablative photodecomposition: Action of far-ultraviolet (193 nm) laser radiation on poly (ethylene terephthalate) films. J Am Chem Sot 1982; 104:6784-8785

9. Linsker R, Srinivasan R, Wynne JJ, et al: Far-ultraviolet laser ablation of atherosclerotic lesions. Lasers Surg Med 1985; 4:201-206

10. Murphy-Chutorian D, Seizer PM, Kosek J, et al: The in- teraction between excimer laser energy and vascular tissue. Am Heart J 1988; 112:739-745

11. Grundfest WS, Litvack F, Forrester JS, et al: Laser abla- tion of human atherosclerotic plaque without adjacent tissue injury. J Am Co11 Cardiol 1985; 5:929-933

12. Steinert RF, Puliafito CA: The Neodynium-YAG Laser in Ophthalmology: Principles and Clinical Applications of Photodisruption, ed 1. Philadelphia, Saunders, 1985, pp 22-30

13. Rulnick JM: Pulsed Laser Ablation of Biological Tissues at 355 nanometers. Thesis, Massachusetts Institute of Technology, 1986

14. Nuss RC, Puliafito CA, Dehm EJ: Unscheduled DNA syn- thesis following excimer laser ablation of the cornea in vivo. Invest Ophthalmol Vis Sci 1987; 28:287-294

15. Sliney DH, Wolbarsht M: Safety with Lasers and Other Optical Sources: A Comprehensive Handbook, New York, Plenum, 1980

American

Journal

of

Otolaryngology

84