Lasers in Surgery and Medicine 9556-571 (1989) Characteristics of 308 nm Excimer Laser Activated Arterial Tissue Photoemission Under Ablative and Non-Ablative Conditions Gunther Laufer, MD, Gregor Wollenek, MD, Benno Rueckle, PhD, Martin Buchelt, MD, Christian Kuckla, MD, Helmut Ruatti, MS, Peter Buxbaum, MD, Roland Fasol, MD, and Peter Zilla, MD Department of Surgery 11 (G. L., G. W., M. B., C. K., R. F., PZ.) and Bone-Biomaterials Res. Lab., Histological-Embryological lnsfitute (H. R.), University of Vienna, A- 1090 Vienna, Austria, Technolas Res. Lab., 0-8032, Munich-Graefelfing, W-Germany (6. R.) The present study was designed to assess the characteristics of tissue photoemission obtained from normal and atherosclerotic segments of human postmortem femoral arteries by 308 nm exci- mer laser irradiation of 60 ns pulsewidth. Three ablative (20, 30, and 40 mJ/pulse) and three non-ablative (2.5,5, and 10 mJ/pulse) energy fluences were employed. Both the activating laser pulses and the induced photoemission were guided simultaneously over one and the same 1,000 pm core optical fiber that was positioned in direct tissue contact perpendicular to the vascular surface. The spectral lineshape of normal arterial and noncalcified ath- erosclerotic structures was characterized by a broad-continuum, double-peak emission of relevant intensity between wavelengths of 360 and 500 nm, with the most prominent emission in the range of 400-415 (407 nm peak) and 430-445 nm (437 nm peak). Fibrous and lipid atherosclerotic lesions, however, exhibited a signifi- cantly reduced intensity at 437 nm compared to normal artery layers (P < 0.001), expressed as a 4071437 nm ratio of 1.321 k 0.075 for fibrous and 1.392 2 0.104 for lipid lesions. Normal artery com- ponents presented with approximately equal intensity at both emission peaks (407/437 nm ratio: intima, 1.054 f 0.033; media, 1.024 f 0.019; adventitia, 0.976 +. 0.021). Comparison of spectral lineshape obtained under various energy fluences within a group of noncalcified tissues disclosed no substantial difference using the 407/437 nm ratio (P > 0.05). In contrast, calcified lesions re- vealed high-intensity multiple-line (397, 442, 461, and 528 nm) emission spectra under ablative energy fluences, whereas a low- intensity broad-continuum, single-peak spectrum resulted from irradiation beyond the ablation threshold. Thus, these findings suggest fluorescence phenomena for broad-continuum spectra, and plasma emission for multiple-line spectra as an underlying photodynamic process. Regardless of the activating energy flu- ence, spectral analysis of 308 nm activated photoemission pro- vides accurate information about the laser target under stan- dardized in vitro conditions. It is demonstrated that direct contact ablation and simultaneous spectral imaging of the target tissue via the same optical fiber is feasible. Key words: excimer laser, spectroscopy, tissue photoemission, laser angioplasty Address reprint requests to Gunther Laufer M.D., Depart- ment of Surgery 2, The University of Vienna, Spitalgasse 23, A-1090 Vienna, Austria. Accepted for publication June 16, 1989. Funds were received from “Gemeinde Wien,3. 0 1989 Alan R. Liss, Inc.
Transcript
Lasers in Surgery and Medicine 9556-571 (1989)
Characteristics of 308 nm Excimer Laser Activated Arterial Tissue
Photoemission Under Ablative and Non-Ablative Conditions
Gunther Laufer, MD, Gregor Wollenek, MD, Benno Rueckle, PhD, Martin Buchelt, MD, Christian Kuckla, MD, Helmut Ruatti, MS, Peter Buxbaum, MD, Roland Fasol, MD, and Peter Zilla, MD
Department of Surgery 11 (G. L., G. W., M. B., C. K., R. F., PZ.) and Bone-Biomaterials Res. Lab., Histological-Embryological lnsfitute (H. R.), University of Vienna, A- 1090 Vienna,
Austria, Technolas Res. Lab., 0-8032, Munich-Graefelfing, W-Germany (6. R.)
The present study was designed to assess the characteristics of tissue photoemission obtained from normal and atherosclerotic segments of human postmortem femoral arteries by 308 nm exci- mer laser irradiation of 60 ns pulsewidth. Three ablative (20, 30, and 40 mJ/pulse) and three non-ablative (2.5,5, and 10 mJ/pulse) energy fluences were employed. Both the activating laser pulses and the induced photoemission were guided simultaneously over one and the same 1,000 pm core optical fiber that was positioned in direct tissue contact perpendicular to the vascular surface. The spectral lineshape of normal arterial and noncalcified ath- erosclerotic structures was characterized by a broad-continuum, double-peak emission of relevant intensity between wavelengths of 360 and 500 nm, with the most prominent emission in the range of 400-415 (407 nm peak) and 430-445 nm (437 nm peak). Fibrous and lipid atherosclerotic lesions, however, exhibited a signifi- cantly reduced intensity at 437 nm compared to normal artery layers (P < 0.001), expressed as a 4071437 nm ratio of 1.321 k 0.075 for fibrous and 1.392 2 0.104 for lipid lesions. Normal artery com- ponents presented with approximately equal intensity at both emission peaks (407/437 nm ratio: intima, 1.054 f 0.033; media, 1.024 f 0.019; adventitia, 0.976 +. 0.021). Comparison of spectral lineshape obtained under various energy fluences within a group of noncalcified tissues disclosed no substantial difference using the 407/437 nm ratio (P > 0.05). In contrast, calcified lesions re- vealed high-intensity multiple-line (397, 442, 461, and 528 nm) emission spectra under ablative energy fluences, whereas a low- intensity broad-continuum, single-peak spectrum resulted from irradiation beyond the ablation threshold. Thus, these findings suggest fluorescence phenomena for broad-continuum spectra, and plasma emission for multiple-line spectra as an underlying photodynamic process. Regardless of the activating energy flu- ence, spectral analysis of 308 nm activated photoemission pro- vides accurate information about the laser target under stan- dardized in vitro conditions. It is demonstrated that direct contact ablation and simultaneous spectral imaging of the target tissue via the same optical fiber is feasible.
Address reprint requests to Gunther Laufer M.D., Depart- ment of Surgery 2, The University of Vienna, Spitalgasse 23, A-1090 Vienna, Austria.
Accepted for publication June 16, 1989. Funds were received from “Gemeinde Wien,3.
0 1989 Alan R. Liss, Inc.
308 nm Excimer Laser Activated Arterial Tissue 557 I NTR ODU CTI 0 N
Laser irradiation of cardiovascular tissue is associated with an electromagnetic emission from the laser target [l]. According to different op- tochemical processes at the site of laser-tissue in- teraction, varying types of tissue emission may be expected. The most important phenomena include laser-induced fluorescence, laser-induced plasma emission, and Raman-light scattering [ 11. Spec- troscopic analysis of these tissue emission phe- nomena have been demonstrated to provide diag- nostic information about the type and chemical composition of the tissue at the target site [2,31. The potential employment of spectroscopic tissue characterization in laser angioplasty procedures implements control of the ablation process by re- striction to atherosclerotic artery layers only.
In vitro studies have indicated that excimer laser energy is capable of precise removal of dif- ferent sorts of atherosclerotic tissue [41; simulta- neously, the electromagnetic emission that is re- leased from the tissue target carries sensitive information about the composition of the artery wall 151. The quantitative and qualitative charac- teristics of 308 nm excimer laser-induced tissue emission under different ablative and non-abla- tive energy fluences, applied by fiberoptic deliv- ery in a direct contact mode, have not yet been determined. The present investigation deals with spectral intensities and patterns of 308 nm exci- mer laser irradiation excited arterial tissue emis- sion observed under six different energy fluences, employing a single optical fiber for guidance of the laser pulse and the tissue photoemission.
MATERIALS AND METHODS Laser Source
A MAX-10 excimer laser (Technolas, Mu- nich-Graefelfing, FRG) provided 308 nm wave- length laser irradiation using a xenon-chlorine gas mixture as active medium. A pulse length of 60 ns duration was applied in single pulse and multiple pulse exposures of 10 Hz repetition rate. The laser power output was modified by adjust- ment of the high voltage according to the desired energy fluence at the fiber tip (see next para- graph). This laser source was additionally equipped with an helium-neon pilot guiding beam of 632.8 nm wavelength.
Fiberoptic Delivery and Mode of Exposure The native laser pulse as emitted from the
laser source was conducted over a beam-splitting,
semireflective mirror and guided through an op- tical lens with a focal length of 10 cm. The fo- cussed beam was coupled to a flexible optical quartz fiber (Diaguide ST-U1000H-SY, Mitsu- bishi, Tokyo, Japan) with a core diameter of 1,000 pm and a length of 1.5 m. The overall fiber diam- eter including the cladding and coating layer re- vealed 2,250 pm. The distal tip was used in a bare fiber tip configuration. Prior to each experiment energy fluence was measured by an energy meter at the distal fiber tip and adapted to the desired values. Six different energy fluences, three non- ablative (2.5,5, and 10 mJ/pulse, corresponding to 3.2, 6.4, and 12.8 mJ/mm2/pulse) and three abla- tive (20, 30, and 40 mJ/pulse, corresponding to 25.6, 38.5, and 51.3 mJ/mm2/pulse) were em- ployed. All experiments were performed in direct contact between fiber tip and irradiated vascular tissue (direct contact mode) with the distal por- tion of the fiber in a perpendicular position ac- cording to the surface of the irradiated vascular segment. A target area of circular shape and 1.0 mm in diameter (0.78 mm2 square area) was thus obtained. Gentle mechanical pressure was ap- plied to the fiber in order to accomplish direct contact throughout single and multiple pulse ex- posures. Experiments with non-ablative energy f luence were carried out as single-pulse exposure exclusively. Air was used as the environmental medium.
Tissue Emission Analysis Emission released from the tissue target was
sampled over the same quartz fiber as used for delivery of the laser beam, guided through the semireflective mirror, focussed by a lens, and cou- pled to a spectrograph and 0-SMA optical simul- taneous multichannel spectral analyzer system (Spectroscopy Instruments, Munich, FRG) via an optical fiber. An ultraviolet-cutoff filter that blocked all wavelengths below 300 nm totally and between 300 and 350 nm partially, was put in front of the spectrograph to protect the diode ar- ray from the reflected laser emission. The analyz- ing unit consisted of the following subunits: tissue emission was spectrally dispersed by a 27 cm spectrograph (holographic concave diffraction grating 150 limm, JA-27, Spectroscopy Instru- ments, Munich, W-Germany) and covered a re- gion of 200-1,000 nm. The wavelength resolution was approximately 1 nm. The spectral pattern was imaged onto an optical multichannel ana- lyzer detector head (IRY-1024, Princeton Instru- ments Inc., Princeton, NJ), consisting of a linear
558 Laufer et a1
A W-blWkhtg filter
I spectrograph I 1- detectorhead I m detector controller H
0P;ical lens
308 nm excimerlaser
Fig. 1. Schematic arrangement for irradiation and activation of tissue photoemission by pulsed 308 nm excimer energy. A single optical fiber in direct tissue contact is used for guidance of the activating beam and the induced photoemission. Laser pulse and photoemis- sion are separated by a semireflective (beam-splitting) mirror.
diode array of 1,024 elements. Diode signal out- puts were handled by a detector controller (ST- 120, Spectroscopy Instruments, Munich, FRG) and further processed, collected, and stored by a computer (IBM-AT) with appropriate software. All analyses were carried out in real time mode. Photoemission spectra were displayed on a com- puter screen and traced by a plotter. Background photoemission activity was subtracted from each spectrum for the purpose of eliminating noise sig- nals from different sources. The total photoemis- sion released from the tissue target was recorded by triggering the spectral analyzer to the onset of the laser pulse and by collecting and utilizing the entire spectral emission of the following 5 ps (zero delay, 5 ps gate).
A diagrammatic representation of the exper- imental setup is displayed in Figure 1.
Calibration Wavelength calibration was accomplished by
utilizing the photoemission of a mercury and neon vapor light source as well as the filter-attenuated 308 nm excimer and the 632.8 nm He-Ne laser wavelengths. Measurement of background activ- ity during blank laserpulse delivery was carried out for each individual experiment. As these background spectra were subtracted from the spectra obtained under tissue exposure, the sam- pled spectra were regarded as tissue-related pho-
toemission phenomena. The intensity of spectral emissions was quantified in arbitrary units.
Tissue Samples
Fresh (6-12 hs) specimens of human post- mortem superficial femoral @FA) and common femoral arteries (CFA) were used as laser targets. Arterial segments were sampled from 9 individu- als aged 19-76 years. Tissue from six individuals presented with severe atherosclerotic disease; three individuals had normal gross and micro- scopic arterial morphology. Samples were split longitudinally and repeatedly flushed with and stored in isotonic saline solution until the exper- iments were performed. Tissue layers were irra- diated in situ, i.e., in context with the natural artery wall architecture. According to macro- scopic and microscopic examination, the site of laser irradiation was classified in categories of normal intima, media, or adventitia, and plaques consisting predominantly of lipid ( = yellow plaque), fibrous ( = white plaque), or calcified ma- terial (=hard plaque). In order to characterize these categories of tissue, spectra were excluded from analysis if the tissue layer in the laser target was found to represent two or more different cat- egories of tissue on microscopic examination. Fur- thermore, atherosclerotic lesions associated with thrombotic formations on the luminal surface and intramural hemorrhage were not utilized for
308 nm Excimer Laser Activated Arterial Tissue 559 spectral investigation because of the potential in- terference with hemoglobin. Spectra were ob- tained and further analyzed from normal media, adventitia, and intima, lipid, fibrous, and calci- fied atherosclerotic lesions. After macroscopic ex- amination specimens were fixed in formalin and embedded in paraffin or methyl-methacrylate, if calcified tissue was expected. Afterwards samples were sectioned and stained with hematoxylin- eosin, Goldner’s trichrom, Elastica-acid Orcein, and Kossa’s stain for undecalcified specimens [6].
Spectra Evaluation and Statistics Histological appearance of the irradiated tis-
sue surface was compared and correlated with the corresponding spectrum of single-pulse exposures or with the spectrum obtained from the final laser pulse among multiple-pulse exposures. Spectra were evaluated for gross lineshape and wave- length range of emission maxima. Intensity of peak emission was measured and data grouped in terms of underlying histology and activating en- ergy fluence. The ratio was calculated for the most prominent emission maxima in order to characterize spectral lineshape by a single nu- meric value. Data were summarized by means of descriptive statistics and were expressed by the mean _+ SD. Testing for differences between groups was accomplished by the multiple compar- ison corrected Student-Newman-Keuls test [7]. Significance was assumed at a P-value < .05.
RESULTS Performance of the Experimental Setup
Although the entire experimental configura- tion worked quite satisfactorily throughout the experiments, destruction of the fiber tip occurred once among four attempts at direct contact abla- tion and spectroscopy of a calcified lesion at 40 mJ/pulse. Microscopic examination confirmed ab- lation among all categories of tissue at 20,30, and 40 mJ/pulse and lacking evidence of ablation at 2.5, 5, and 10 mJ/pulse.
Spectral Lineshape and Its Relation to Energy Fluence Within Groups of Tissue-Noncalcified Tissue
Normal intima, media, and adventitia, fi- brous and lipid lesions (noncalcified tissue) dem- onstrated a broad-continuum photoemission be- tween the wavelengths of 360 and 700 nm and the most prominent intensity in the range of 360-500 nm. Regardless of the ablative or non-ablative na-
ture (energy fluence) of the exciting laser pulse, the gross spectral lineshape within one histologi- cal category of tissue was indistinguishable. A de- tailed analysis is given in the following section.
Media. The typical spectra released from a media layer under ablative and nonablative con- ditions are depicted in Figure 2a. A total of 44 spectra from normal media wall layers were ana- lyzed and consisted of a broad-continuum photo- emission from 360 to 700 nm with 2 major inten- sity ranges from 400 to 415 nm and from 430 to 445 nm that culminated in peak wavelengths of 407 and 437 nm. Both peaks were of approxi- mately equal size with a slightly decreased mean value for the longer wavelength (Fig. 2b). Appear- ing thus as an “equal intensity double peak” pat- tern, spectral lineshapes were identical regard- less of the ablative or non-ablative character of the activating laser pulse. Calculating the 407/437 nm intensity ratio revealed no significant difference (P > 0.05) between groups of various ablative and non-ablative energy fluences (Fig. 2b). The ratio for all spectra derived from media tissue yielded 1.024 2 0.019.
Adventitia. Two typical emission spectra from adventitia are illustrated in Figure 3a. Forty-six spectra from this tissue layer were char- acterized by a broad-continuum photoemission that strongly resembled the lineshape of a media spectrum with the exception of a slightly higher emission intensity at the 430-445 nm than the 400-415 nm range. Spectral lineshape was not influenced by the strength of the activating laser pulse, which is also reflected in the comparison to 4071437 nm ratios between groups of different en- ergy fluence (P > 0.05, Fig. 3b). Summarizing the ratio for 46 adventitia spectra yielded a mean value of 0.976 k 0.021.
Intima. Spectra from normal intima were obtained exclusively under non-ablative energy fluence. Spectra under ablative conditions were excluded, as a single pulse of energy fluence of 20 mJ or higher ablated the entire intimal target structures and a tiny layer of the adjacent media. Analysis of 25 spectra revealed again the spectral lineshape of broad-continuum emission with the two major intensity ranges culminating at wave- lengths of 407 and 437 nm. Two representative spectra are depicted in Figure 4a. The 407/437 nm ratio of all intima derived spectra was 1.054 * 0.033, thus indicating a decreased intensity in the long wavelength range (Fig. 4b). Comparison be- tween the ratios from groups of 2.5, 5, and 10 mJ/pulse disclosed no significant difference.
560
.s 1.1 - E
1,o - 0 9 - 0.8
Laufer et a1
‘ 1 1 I I I I
- 8 . 1 - 1 . 1 . 1 . 1 . 1 . I
MEDIA
2700
2400
2100 - In .u 5 1800 h
1500 )1 - e z 1200
g 900 % = u C CI -
600
300
0 200 300 450 600 750 900 nm
wavelength
L W
Media
y 1600 h
.- C 3
-1 ,b I , I . , . I ; I I 400 0 5 10 15 20 25 30 35 40
energy fluence (mJ/pulse)
Fig. 2. a: Photoemission spectra obtained from normal media layers of a SFA (19-year-old male). Spectrum 1 resulted from 30 mJipulse (ablative) irradiation, spectrum 2 from 2.5 mJ/pulse (nonablative) irradiation. b: Characteristics for me-
Fibrous Plaque. Forty-six spectra were re- corded and analyzed from tissue targets with f i - brous atherosclerosis and exhibited the same
broad-continuum emission with a peak in the 400- 415 nm range. Intensity in the 430-445 nm range, however, was attenuated and always
308 nm Excimer Laser Activated Arterial Tissue
2700
ADVENTITIA
.
2400
2100 .
561
wavelength
1800
-1 ,b I , I , I . ,; , %:i‘., I 400
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
energy fluence (xnJ/pulse) energy fluence (mTlpulse)
Fig. 3. a: Adventitia spectra (25-year-old male, SFA) induced by 30 mJ/pulse (spectrum 1) and 10 mJ/pulse (spectrum 2), respectively. b: Absolute photoemission intensity at 407 and 437 nm and 407/437 nm ratio (mean % SD) for adventitia spectra according to incremental energy fluence.
markedly smaller than in the short wavelength range. This changed the spectral lineshape, re- sembling a “single peak” feature with an attenu- ated or minimal second peak (Fig. 5a). This was
confirmed by computation of the 407/437 nm in- tensity ratio that revealed 1.321 & 0.075 for all spectra from fibrous lesions. Again significant dif- ferences in this ratio and thus in spectral line-
Fig. 4. a: Intima spectra released from SFA of a 19-year-old male by activation with 10 mJ/pulse (spectrum 1) and 5 mJipulse (spectrum 2), respectively. b: Absolute photoemission intensity at 407 and 437 nm and 4071437 nm ratio (mean f SD) for intima spectra according to incremental energy fluence.
shape in relation to the exciting energy fluence were not demonstrable (Fig. 5b).
Lipid Plaque. Figure 6a displays represen- tative spectra of a lipid atherosclerotic lesion.
Forty-five spectra were analyzed and their line- shape closely resembled those of fibrous plaques. Exciting energy fluence did not influence spec- tral lineshape, yielding a P-value > 0.05 for com-
Fig. 5. a: Spectrum 1 was obtained from a fibrous lesion in the CFA of a 76-year-old woman with strong ablative (40 mJ/pulse) energy fluence. Spectrum 2 is from the same plaque with 10 mJ/pulse activating energy. b: Absolute photoemission intensity a t 407 and 437 nm and 4071437 nm ratio (mean 2 SD) for spectra from fibrous lesions according to incremental energy fluence.
564 Laufer et a1
wavelength
2000 Lipid plaque
!
f 4 407nm
o 437nm
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 b energy fluence (&/pulse) c energy fluence (&/pulse)
Fig. 6. a: The 30 mJ/pulse energy fluence revealed spectrum 1 (lipid lesion, SFA, 55-year- old male). Spectrum 2 achieved under 2.5 mJ/pulse, lipid lesion, CFA, 55-year-old male. b: Absolute photoemission intensity a t 407 and 437 nm and 407/437 nm ratio (mean * SD) for spectra from lipid lesions according to incremental energy fluence.
308 nm Excimer Laser Activated Arterial Tissue 565 parison of the 407/437 nm intensity ratio between groups of different energy fluence (Fig. 6b). The 4071437 nm ratio for all lipid plaque spectra was 1.392 * 0.104.
Spectral Lineshape and Its Relation to Energy Fluence Within Groups of Tissue-Calcified Tissue
Twenty-nine spectra from calcified athero- sclerotic lesions exhibited a spectral lineshape quite different from that of noncalcified arterial tissue (Fig. 7a). Activating energy fluences of 20, 30, and 40 mJ/pulse resulted in a strong multiple- line emission at wavelengths of 397,442,461, and 528, superimposing a broad-continuum emission between 360 and 500 nm. 2.5, 5 , and 10 mJ/pulse failed to produce a strong line-emission and spec- tral analysis revealed a “single peak” broad-con- tinuum emission from 360 to 700 nm with the peak intensity range from 430 to 460 nm. Figure 7c demonstrates a histologic example of a calci- fied lesion corresponding to spectrum 1 (Fig. 7a).
Comparison of the Spectral Lineshape Between Groups of Tissue
Differentiation between calcified lesions and noncalcified arterial wall layers was feasible us- ing the characteristic spectrum obtained under ei- ther ablative or non-ablative energy fluence. All other spectra from noncalcified tissue layers dem- onstrated a broad-continuum emission reflecting the double-peak lineshape. This characteristic lineshape feature, however, was attenuated by a less prominent peak at 437 nm in cases of lipid lesions. Distinction was attempted employing the relationship between the short and the long wave- length peak at 407 and 437 nm, respectively. Lipid plaques (highest mean value for the 407/437 nm ratio) and fibrous plaques revealed a highly significant P < 0.001 compared with all 3 normal arterial wall layers. Borderline significance (P = 0.05) was achieved for comparison of fibrous and lipid lesions. Among structures of normal arterial architecture, intima and media layers differed significantly from the adventitia, yielding P < 0.01. However, we were not able to detect signif- icant spectral differences (P > 0.05) between in- tima and media by comparing intensity ratios.
Photoemission Intensity and Its Relation to Energy Fluence and Groups of Tissue
Noncalcified Samples. The absolute in- tensity of tissue photoemission between the wave- lengths of 360 and 500 nm was influenced by the strength of activating laser pulses, although spec- tral lineshape of noncalcified tissue was indepen- dent of energy fluence. Intima, media, adventitia, fibrous, and lipid plaques exhibited the lowest mean values of peak intensity at 2.5 mJ/pulse continuously increasing to the highest mean val- ues at 40 mJ/pulse (Figs. 2-6b). The range of in- dividual peak intensity values, however, showed a considerable amount of variability that pre- cluded a precise determination of the underlying activating energy fluence from this value for a definite type of tissue (Table 1). Tissue groups of intima and fibrous lesions had the highest vari- ability of intensity values whereas lipid lesions had the lowest.
Calcified Samples. Generally, a higher la- ser energy fluence was associated with higher mean values of peak intensities for both types of emission spectra (Fig. 7b). Signals obtained from calcified specimens under ablative conditions dis- closed an approximately 5 -6-fold increased inten- sity at wavelengths of 397, 442, and 461 nm op- posed to the 407 nm intensity peak of noncalcified arterial layers. This contrasted with the spectral intensity during non-ablative energy f luence, which was of a similar order of magnitude to that in noncalcified tissue.
Comparison of Spectral Lineshape and Photoemission Intensity Between Groups of Repetition Rate
Stratification of spectra according to mode of exposure yielded subgroups of single pulse and multiple pulse (10 Hz) exposure within groups of tissue and groups of ablative energy fluence. Comparison of spectral lineshape or photoemis- sion intensity between such corresponding sub- groups did not exhibit any substantial difference (P > 0.05) based on stratification according to rep- etition rates of the exciting beam. Spectra ob- tained from non-ablative excitation were de- signed exclusively as single pulse exposure.
Employing an arbitrary 4071437 nm ratio of 1.2 as Background Photoemission a cutoff value non-calcified atherosclerotic lesions were differentiated from normal wall layers at This consisted of environmental electromag- subablation energy fluences with a sensitivity of netic activity and laser pulse-related background 100% and specificity of 97.8% and at ablative en- activity that involved fluorescence and backscat- ergy fluences with 100% and 97.9%, respectively. tering of the 308 nm excimer- and He-Ne laser
566
CALCIFIED PLAQUE
Laufer et a1
6300
5600
4900 - - 5 4200 C J
2 3500 2 n .L - 2800
v
Y - 2 2100
1400
V C *. -
a 70C
0 200 300 450 600 750 900 nm
wavelength
1800
0.0 2.5 5.0 7.5 10.0
energy fluence (mJ/pulse)
Fig. 7. a: Ablative (10 Hz, 40 mJ/pulse-spectrum 1) and non- ablative (single pulse, 2.5 mJ/pulse-spectrum 2) energy flu- ence revealed a striking difference in spectral lineshape for calcified plaques. Spectrum 1: Calcified lesion, SFA, 73-year- old male, corresponding tissue layer in panel c (facing page)(asterisk). Spectrum 2: Calcified lesion, SFA, 73-year- old male. b: Absolute photoemission intensity at wavelength
T Calcified plaque 12000
397nm o 4 4 2 n m
o 528nm
15 20 25 30 35 40 energy fluence (mJ/pulse)
range of 430-460 nm (nonablative excitation) and 397, 442, 461, and 528 nm (ablative excitation), expressed as mean t SD for spectra from calcified plaques according to incremen- tal energy fluence. c: Calcified lesion, SFA, 73-year-old male. The calcified tissue is blackened and embedded in fibrous material (F). Arrow, laser pulse direction. Kossa staining, x 120.
308 nm Excimer Laser Activated Arterial Tissue 567 nal. The apparent zero photoemission intensity below wavelengths of 300 nm was caused by total filter suppression.
Figure 7c.
from the fiber and guiding optics. Background photoemission was subtracted and therefore did not appear on the recorded tissue spectra.
Laser Backscattering at the Tissue Target Both excimer and He-Ne laser light were
scattered at the tissue target, as with any laser treatment of a tissue site. Scattered laser light that appeared in spectra of this investigation had the characteristics of Rayleigh scattering with the scattered light presenting the same wave- length as the native laser beam [ll. According to the extent of this process, line emission of varying intensity at 308 and 632.8 nm occurred in nearly all individual spectral recordings (Figs. 2-7a).
Ultraviolet Filter
The selected cutoff filter suppressed wave- lengths below 300 nm completely and between 300 and 350 nm partially. Thus the reflected la- ser pulse appeared attenuated in the spectral sig-
DISCUSSION
Lack of appropriate control of the laser abla- tion process is the predominant factor for arterial perforation during laser angioplasty [8] . An ap- proximately 5 -20% rate of perforation has been reported for peripheral and coronary trials in ex- perimental and clinical studies [9-111. Thus, ar- terial perforation represents one of the major problems of laser angioplasty. Although ul- trashort-pulsed laser energy has been demon- strated in vitro and in vivo to provide vascular tissue ablation with minimal amount of thermal injury to adjacent layers [ 12-15], high-quality ab- lation itself does not prevent arterial perforation as the ablating pulse is not able to distinguish between normal and pathological arterial struc- tures. The often eccentric nature and heteroge- neous constitution of human atherosclerotic dis- ease determines a path of least resistance that leads to fiber penetration into the adventitia [161.
Identification of the laser target by means of laser-induced spectroscopy has been proposed by various groups [3,17-191 to overcome the critical problem of perforation by target-specific ablation. We have demonstrated in vitro that lens-focussed 248 nm excimer-laser pulses are capable of nor- mal and atherosclerotic arterial tissue ablation with subsequent identification of the laser target by analysis of the photoemission, which is excited by the ablating laser pulse [ 5 ] . In contrast to this previous work the present investigation evalu- ates spectral characteristics basically under dif- ferent ablative and nonablative 308 nm energy fluence (as opposed to a single 248 nm ablative fluence of 50 mJ/mm2/pulse) and the use of a sin- gle optical fiber in a direct contact mode for elec- tromagnetic transmission (as opposed to a lens- focussed non-contact setup).
Due to the difficulty of coupling and trans- mission of 248 nm excimer laser pulses through optical fibers, the longer excimer wavelengths of 308 and 351 nm (XeCL and XeF, respectively) in various pulse durations are more frequently em- ployed [20]. In the setting presented, 308 nm ir- radiation of 60 ns (moderately stretched) pulse duration was applied over a 1,000 pm fiber with up to 40 mJlpulse energy fluence in a direct con- tact mode. We even achieved ablation of severely calcified lesions in direct contact mode, but ob-
568 Laufer et a1 TABLE 1. MinimumlMaximum Values of Peak Intensity at Wavelengths of 407 and 437 nm
Midmax Lipid plaque Intima Media Adventitia Fibrous plaque
served one destruction of the fiber tip at 40 mJ/pulse. This may reflect either inappropriate strong mechanical pressure on the fiber tip or, quite conceivably, fiber damage by strong plasma formation in the laser target, which might limit the use of higher energy fluence via optical fibers in the direct contact mode. Obviously, it indicates that 40 mJ/pulse energy fluence is close to fiber tip destruction threshold for fibers of this type. Fiberoptic delivery of the ablating excimer laser pulse and simultaneous guidance of the excited photoemission over the same fiber under condi- tions of direct contact as well as real time spectral analysis provide the necessary configuration for successful application in an in vivo system. Al- though this experimental setup was used for in vitro studies, it may be considered as a configu- ration which is suitable for clinical use. This sug- gests the basic feasibility of spectroscopic target tissue characterization by an intraoperative ap- proach or by catheterization techniques and con- struction of a real-time, spectroscopically guided device for laser atherolysis.
Spectral analysis in this investigation re- vealed a spectral lineshape quite similar to that obtained by short pulse 248 nm excitation [51, al- though the entire spectrum shifted into the longer wavelength range. Spectra from noncalcified ar- terial layers disclosed a broad-continuum emis- sion with distinct peak intensities in the range of 400-415 and 430-445 nm, used as detection bands. The relation between both intensity peaks allowed precise spectroscopic differentiation of fi- brous and lipid layers from normal components of the artery wall as well as differentiation of ad- ventitia from media and intima. The gross “dou- ble-peak” lineshape of spectral signals from arte- rial target tissue with approximately equal intensity for normal layers and always lower in- tensity of the long wavelength peak for noncalci- fied atherosclerotic lesions has also been de- scribed for other activating laser wavelengths.
Andersson et al. [171 used a pulsed nitrogen laser (lambda = 337 nm> to induce vascular tissue emis- sion and found intensity maxima at different wavelengths but the same relation between the short and the long wavelength maximum. Kit- trell et al. [18] reported intensity peaks at 550 and 600 nm under 480 nm laser excitation. Dis- eased samples again showed the same two peaks but always reduced intensity at 600 nm. Employ- ing 476 nm argon-ion laser energy for fluores- cence excitation, similar results were obtained by Hoyt et al. [31: this author described peak emis- sion at 520 and 600 nm, the peak intensity of normal tissue being approximately three times higher than for early atherosclerotic plaques. The “double-peak” lineshape was also attenuated for plaques by marked reduction of peak intensity at 600 nm. Additional studies (unpublished results) on human coronary arteries removed from the re- cipient heart during cardiac transplantation, thus reflecting in vivo conditions, indicated a similar lineshape. Therefore, it is very unlikely that the typical “double-peak” lineshape is an artifact re- lated to postmortem changes such as hemoglobin diffusion in the artery wall. Using 337 nm pulsed nitrogen laser excitation, Deckelbaum et al. [19] reported contrasting findings with a spectral line- shape that lacks the characteristic “double-peak” pattern of other investigations. This emphasizes an important influence of the activating wave- length, although a close resemblance of gross spectral lineshape is evident in most studies.
All these previous investigators [3, 17-19] used an activating laser beam that operated far below the ablation threshold. Therefore spectral analysis was confined to the luminal surface of the vessel, as the laser light had no access to fur- ther peripheral wall layers. This implicates the necessity for a second, different laser for tissue removal. During laser angioplasty deep layers of the atherosclerotic lesion or surrounding normal arterial structures are exposed by the ablating
308 nm Excimer Laser Activated Arterial Tissue 569 laser beam and ablation itself may induce changes of the spectral properties of the exposed target tissue. This may be very likely if ablation is associated with thermal injury to the adjacent tissue. Under these circumstances analysis of photoemission stimulated by a low-power laser beam may not be conclusive and the results influ- enced by the quality of the ablation process. Al- though the experimental results of these investi- gators suggest identification and localization of plaques by spectroscopic surface scanning, the value of the concept of low-power activating laser beams for continuous spectroscopic imaging of the target tissue during ablation remains undefined. Additionally, the employment of a second, abla- tive laser pulse combined with the exciting low- power laser pulse requires a more complex system than the use of a single beam for ablation and simultaneous excitation.
Spectral properties of the target tissue may be theoretically altered not only by an ablating laser beam, which causes significant thermal in- jury, but also by excimer laser pulses under cer- tain conditions. Classic histologic signs of ther- mal injury occur if excimer irradiation is applied repeatedly in a defocussed manner that results in a subablation energy fluence [211. In this case the activating, but not ablating excimer pulse might be able to induce spectral alterations in the target tissue on repeated exposure. For this reason ex- periments with non-ablative fluence were carried out exclusively as single-pulse exposure without previous protracted, non-ablative tissue irradia- tion. In contrast 308 nm excimer pulses at abla- tive energy fluences of 20, 30, and 40 mJ/pulse were associated with the high quality of tissue ablation as described in the literature [22] that left the spectral tissue attributes unchanged.
Spectra obtained within one category of or- ganic tissue (without calcium deposits) exhibited an identical spectral lineshape, expressed as 4071437 nm intensity ratio without significant differences between groups of various energy flu- ences. Therefore, with organic arterial tissue, the same type of tissue photoemission is produced re- gardless of whether the irradiated laser target is simultaneously ablated or not (Figs. 2a-6a). Spec- tral lineshape and information are independent of ablation and this suggests fluorescence phenom- ena as the basic photodynamic process. As fluo- rescence phenomena occur as a photoemission of laser-excited, but nondegraded molecules, these photodynamic processes must be located at tissue sites remaining at the basis of the laser crater in
front of the fiber tip in the ablative setting. It is emphasized that analysis of fluorescence during ablation supplies information about the tissue layers exposed by the laser pulse rather than about the tissue which is simultaneously ablated.
These findings with noncalcified arterial structures contrasted with the results from calci- fied lesions. Striking spectral changes were asso- ciated with ablation of these lesions (Fig. 7a) and provide supplementary evidence that plasma for- mation occurs during excimer ablation of calcified atheromas and that the resulting, high-intensity, multiple-line emission is a plasma emission [231. Plasma formation is attributable to the presence of inorganic calcium deposits, and a relevant amount is not generated during ablation of bio- logical polymers. The observed emission lines were identical with that obtained during 248 nm excitation [51. Low energy fluence (2.5, 5, and 10 mJ/pulse) did not initiate ablation or concomitant plasma formation in calcified lesions; thus, the recorded signals obtained under non- ablative en- ergy fluence may represent the fluorescence of calcified lesions, which is superimposed by plasma emission under ablative conditions. Re- gardless of the activating energy fluence, the spectral pattern of multiple-line or single-peak broad-continuum emission allowed differentia- tion from noncalcified arterial layers.
The intensity of photoemission within a tis- sue category increases according to the strength of the activating laser pulse. As the intensity of tissue fluorescence or plasma emission is related to the extent of molecule or particle excitation by the laser pulse, energy fluence impacts on quan- titative spectral parameters. However, there is a considerable sample-to-sample variability in the peak photoemission intensity of spectra obtained from a definite category of tissue under a constant energy fluence. Firstly, this variability may be caused by changes in the fiberoptic transmission, which may occur during angulation of the fiber and/or even minor variations in fiber-tip posi- tions. Secondly, although of identical histology, the tissue target itself may consist of a variable amount of fluorescing molecules. This may ex- plain the high standard deviation for intensity values of fibrous plaques. For this reason we do not consider determination of absolute photoemis- sion intensities as a helpful tool in providing ad- ditional diagnostic information about the biolog- ical structure of the target that is irradiated by pulses of 308 nm laser energy. This statement is definitely not valid for calcified lesions irradiated
570 Laufer et a1 with ablative energy fluence because of ex- tremely high (4-5-fold increased) peak values.
Despite the feasibility of simultaneous abla- tion and spectroscopic tissue identification using a single optical fiber in direct contact mode, a crit- ical determination of the potential value for laser angioplasty and endarterectomy is indicated. The major critical questions remaining have been al- ready addressed [51 and include simultaneous presence of normal and atherosclerotic tissue in the target area with resulting mixed spectra, in- terference with blood and attenuation of the sur- rounding normal arterial structures in severely diseased artery segments. It will be the task of further investigation to elucidate the potential value of this method under in vivo conditions of human atherosclerotic disease.
In conclusion, the results presented provide supplementary information for the understand- ing of the basic photodynamic processes of exi- merlaser activated target tissue emission. Arte- rial tissue without calcified deposits exhibit a fluorescence emission that allows target classifi- cation by analyzing the 407/437 nm ratio regard- less of the activating energy fluence. Calcified plaque tissue reveals energy f luence dependent emission lines that strongly suggest plasma emis- sion. It has been demonstrated that tissue abla- tion and simultaneous spectral imaging of this process are feasible with the same optical fiber guiding the laser beam as well as the tissue flu- orescence under direct contact conditions. The re- maining critical questions encourage further evaluation of this system in a blood-filled envi- ronment and in a coaxial beam direction.
REFERENCES
1.
2.
3.
4.
5.
Clarke RH, Isner JM, Gauthier T, Nakagawa K, Cerio F, Hanlon E, Gaffney E, Rouse E, DeJesus S. Spectroscopic characterization of cardiovascular tissue. Lasers Surg Med 1988; 8(1):45-59. Sartori M, Sauerbrey R, Kubodera S, Tittel FK, Roberts R, Henry PD. Autofluorescence maps of atherosclerotic human arteries-a new technique in medical imaging. IEEE J Quant Elect 1987; QE-23(10):1794-1797. Hoyt CC, Richards-Kortum RR, Costello B, Sacks BA, Kittrell C, Ratliff NB, Kramer JR, Feld MS. Remote bio- medical spectroscopic imaging of human artery wall. La- sers Surg Med 1988; 8(1):1-9. Isner JM, Donaldson RF, Deckelbaum LI, Clarke RH, Laliberte SM, Ucci AA, Salem DN, Konstam MA. The excimer laser: Gross, light microscopic and ultrastruc- tural analysis of potential advantages for use in laser therapy of cardiovascular disease. J Am Coll Cardiol 1985; 6(5):1102-1109. Laufer G, Wollenek G, Hohla K, Horvat R, Henke KH,
Buchelt M, Wuzl G, Wolner E. Excimer laser induced simultaneous ablation and spectral identification of nor- mal and atherosclerotic arterial tissue layers. Circula- tion 1988; 78:1031-1039.
6. Plenk H, Jr . The microscopical evaluation of hard tissue implants. In Williams DF, (ed): “Techniques in Biocom- patibility Testing,” Vol. 1. Boca Raton, FL: CRC Press, 1986, pp 37-92.
7. Glantz SA. “Primer of Biostatistics.” New York: McGraw- Hill, 1987, pp 91-96.
8. Forrester JS, Litvack F, Grundfest WS. Laser angio- plasty and cardiovascular disease. Am J Cardiol 1986;
9. Grundfest WS, Litvack IF, Goldenberg T, Sherman T, Morgenstern L, Carroll R, Fishbein M, Forrester J , Mar- gitan J , McDermid s, Pacala TJ, Rider DM, Lauden- slager JB. Pulsed ultraviolet lasers and their potential for safe laser angioplasty. Am J Surg 1985; 150:220-226.
10. Abela GS, Normann SJ, Cohen DM, Franzini D, Feldman RL, Crea F, Fenech A, Pepine CJ, Conti CR. Laser reca- nalization of occluded arteries in vivo and in vitro. Cir- culation 1985; 71(2):403-411.
11. Ginsburg R, Wexler L, Mitchell RS, Profitt D. Percuta- neous transluminal laser angioplasty for treatment of pe- ripheral vascular disease. Radiology 1985; 156:619-624.
12. Deckelbaum LI, Isner JM, Donaldson RF, Clarke RH, Laliberte S, Aharon AS, Bernstein JS. Reduction of laser- induced pathologic tissue injury using pulsed energy de- livery. Am J Cardiol 1985; 56:662-667.
13. Murphy-Chutorian D, Selzer PM, Kosek J, Quay SC, Profitt D, Ginsburg R. The interaction between excimer laser energy and vascular tissue. Am Heart J 1986; 112: 739-745.
14. Farrell EM, Higginson LAJ, Nip WS, Walley VM, Keon WJ. Pulsed excimer laser angioplasty of human cadav- eric arteries. J Vasc Surg 1986; 3:284-287.
15. Prevosti LG, Leon MB, Smith PD, Dodd JT, Bonner RF, Robinowitz M, Clark RE, Virmani R. Early and late heal- ing responses of normal canine artery to excimer laser irradiation. J Thorac Cardiovasc Surg 1988; 96:150- 156.
16. Isner JM, Donaldson RF, Funai JT, Deckelbaum LI, Pan- dian NG, Clarke RH, Konstam MA, Salem DN, Bernstein JS. Factors contributing to perforations resulting from laser coronary angioplasty: Observations in an intact hu- man postmortem preparation of intraoperative laser cor- onary angioplasty. Circulation 1985; 72(11):191-199.
17. Andersson PS, Montan S, Svanberg S. Multispectral sys- tem for medical fluorescence imaging. IEEE J Quant Elect 1987; QE-23: 1798-1805.
18. Kittrell C, Willett RL, de 10s Santos-Pacheo C, Ratliff NB, Kramer JR, Malk EG, Feld MS. Diagnosis of fibrous arterial atherosclerosis using fluorescence. Appl Opt 1985; 24:2280, 2281.
19. Deckelbaum LI, Lam JK, Cabin HS, Clubb KS, Long MB. Discrimination of normal and atherosclerotic aorta by la- ser-induced fluorescence. Lasers Surg Med 1987; 7:330- 335.
