|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From Interventional Pulmonary Services, Pulmonary and Critical Care Medicine Division, University of California, San Diego Medical Center, San Diego, CA.
Correspondence to: Henri G. Colt, MD, FCCP, Director, Interventional Pulmonary Services, UCSD-La Jolla, 9310 Campus Point Dr, San Diego, CA 92037; e-mail: hcolt{at}ucsd.edu
| Abstract |
|---|
|
|
|---|
Design: Experimental in vitro study simulating a patient-care environment.
Methods: Uncovered and covered metal Wallstent (Schneider; Zurich, Switzerland) and Dumon (Bryan Corporation; Woburn, MA) silicone stents were deployed in the tracheobronchial tree of a ventilated and oxygenated (fraction of inspired oxygen, 40%) heart-lung block of a dead canine. Rigid bronchoscopic Nd-YAG (1,064 nm) laser procedures were performed in order to deliver laser energy using fiber-to-target distances of 10 mm and 20 mm, and noncontact, continuous-mode, 1-s pulses at power settings of 10 W, 30 W, and 40 W. The major outcome measure was laser-induced stent damage, defined as discoloration, ignition, or breakage. This was assessed using six power densities: 75 W/cm2, 172 W/cm2, 225 W/cm2, 300 W/cm2, 518 W/cm2, and 690 W/cm2.
Results: The uncovered Wallstent and the silicone stent remained intact at power densities of 75 W/cm2 (10 W, 20 mm) and 172 W/cm2 (10 W, 10 mm), but were damaged at power densities > 225 W/cm2 (30 W, 20 mm). The covered Wallstent was damaged at all power densities tested.
Conclusion: Uncovered Wallstent and
silicone stents are not damaged when Nd-YAG laser energy is delivered
using power densities
172 W/cm2 (10 W, 10 mm). Covered
Wallstents, however, had a high likelihood of ignition at all power
densities studied.
Key Words: bronchoscopy complications laser physics laser/tissue interactions metal stents Nd-YAG laser resection power density silicone stents
| Introduction |
|---|
|
|
|---|
The paucity of data pertaining to laser/stent interactions is increasingly important because recurrent obstruction from either granulation tissue or tumor occurs in 9 to 67% of patients with metal stents1 2 3 4 5 6 7 8 and in 6 to 15% of patients with silicone stents.9 10 11 In these patients, laser-induced stent damage may cause substantial morbidity from airway burn injury in case of stent ignition, or airway wall and vascular perforation in case of metal stent rupture.
In patients without indwelling stents, it is generally recommended that
laser resection be performed using laser pulse durations of 1 to 2
s, a power output of
40 W, and a fraction of inspired oxygen
(FIO2)
0.40.12
Using
these recommendations as a starting point, the purpose of this study
was to identify margins of safety defined by the classic measures of
power density and pulse frequency within which Nd-YAG laser energy
could be delivered without damaging indwelling silicone and metal
stents.
| Materials and Methods |
|---|
|
|
|---|
![]() |
is the thermal conductivity, tP the
duration of irradiation,
the density of the material,
CV the specific heat per unit volume, R the reflectivity,
and I0 the laser intensity. The expression
tP1/2- is the thermal diffusivity or penetration
depth of the laser beam, while (1-R) is the absorption
coefficient. For example, at low power densities, the poor absorption of the Nd-YAG laser and its pronounced Mie and Rayleigh scattering result in slow homogeneous heating of a large volume of tissue without serious mechanical damage to the tissue surface.16 At high power densities, however, the temperature 2 to 3 mm below the tissue surface rises rapidly, prompting vaporization of water content and a pocket of steam with a pressure high enough to rupture overlying tissues.17 Based on these principles, we hypothesized that certain quantities of laser energy, defined by power density, would cause local heating without the extreme temperature elevation required to disrupt stent integrity or prompt stent ignition.
In Vitro Experimental Model
In order to enhance applicability of our results to the clinical
setting, the following experimental environment simulating a rigid
bronchoscopic intervention in patients was designed. An intact
heart-lung-tracheobronchial tree was harvested from a dead adult canine
and cleaned thoroughly. The entire structure was suspended at a slope
of 60° (Fig 1 ). Throughout the experiment, this lung-tracheobronchial model was
intubated with a rigid bronchoscope (EFER-Dumon Rigid Bronchoscope;
Bryan Corporation; Woburn, MA) and connected to a Harvard respirator
pump for animals (Harvard Apparatus; Holliston, MA) delivering a
tidal volume of 400 mL per breath. Respiratory frequency was set at 12
cycles per minute, and oxygen delivery through the system was set at an
FIO2 of 40%. The rigid telescope,
laser fiber, and suction catheters were introduced through sealed
rubber obturator caps, maintaining a closed system so that two-lung
ventilation was maintained throughout the intervention. This protocol
was reviewed and approved by the University of California-San Diego
(UCSD) Institutional Animal Care and Use Committee (IACUC) [IACUC
tissue request No. T 99154]. No animals were killed for the sole
purpose of this study; tissue was transferred from a previously killed
animal in a separate IACUC-approved protocol.
|
For all experiments, a 100-Wcapacity Nd-YAG laser (model 813; Laserscope; San Jose, CA) and saline solution-cooled, 0.6-mmdiameter coaxial fiber (Endostat; Laserscope) were used. The fiber was placed through the rigid bronchoscope and directed tangentially toward the inner side of the proximal extremity of each indwelling stent. The helium-neon focusing laser beam was used to ensure that the laser was fired parallel to the airway wall in the direction of the target surface.
Using noncontact, 1-s duration, and continuous mode, laser energy was delivered at power settings of 10 W, 30 W, and 40 W. At least 1 s separated each pulse of laser energy delivered. The distance between the distal aspect of the laser fiber and the target was precisely measured and maintained by fixing the fiber to a ruler attached to the obturator cap of the rigid bronchoscope. For each of the three power settings, laser fiber-to-target surface distances of 10 mm and 20 mm were used.
The average power density delivered for each setting was computed based
on the formula18
: PAW = 0.8635 x power
(watts)/
r2 where
r2
is the spot size, and PAW is average power density. Because
the laser beam diverges 11° around the fiber core, the spot size at a
distance was calculated using the following equation: spot size =
[(fiber radius in centimeters) + (distance in centimeters) x tan
5.5°]2. Using fiber-to-target distances
of 10 mm and 20 mm, therefore, spot sizes were 0.05
cm2 and 0.155 cm2,
respectively. Computed power densities for each experimental condition
are shown in Table 1
.
|
The total number of pulses and the total energy required to damage a stent during each experimental condition were recorded and expressed as mean ± SD. Comparisons of mean number of pulses and mean total energy delivered were made using paired Students t test at a level of significance of 0.05.
| Results |
|---|
|
|
|---|
225 W/cm2 (30 W,
20 mm; Table 2
).
|
225 W/cm2, black discoloration
was immediately followed by perforation of the covering of the
Wallstent (Fig 2
). This occurred after a mean of only 2.6 pulses were delivered. At the
highest power density tested, 690 W/cm2 (less
than two pulses at 40 W, 10 mm), the stent abruptly burst into flames.
|
225 W/cm2, a slight discoloration was noted
at the impact site after a mean of
3.6 pulses (Table 2) . A blister
formed, accompanied by focal black discoloration at the impact site at
a power density of 690 W/cm2 after less than two
pulses were delivered, although the stent did not ignite (Fig 3
).
|
|
| Discussion |
|---|
|
|
|---|
40%. Results clearly
identify a margin of safety for laser resection when power density is
172 W/cm2, or when a power setting of 10 W is
applied with a laser fiber-to-target distance of at least 10 mm. Our observation that 60 pulses at 10 W applied 10 mm from the target did not damage either the uncovered Wallstent or the silicone stent reinforces that of Witt et al,19 who found that uncovered Wallstent, Strecker, and Palmaz metal stents were destroyed when noncontact Nd-YAG laser was used at power settings > 10 W. In fact, these investigators19 found that metal stents were destroyed after just one 25-W pulse delivered 4 mm from the target, prompting them to conclude that Nd-YAG laser should never be used in patients with indwelling metal stents.
A few researchers might contend that lower power densities are not frequently used in bronchoscopic practice in which some operators are accustomed to higher power settings and shorter laser fiber-to-target distances. We suggest that the effects of laser resection using these settings should be further investigated. For example, in an Nd-YAG dosimetry study performed in porcine bladder, Cos and Di Santagnese20 demonstrated that 10 W applied for 10 s produced far more effective penetration of laser energy than 100 W for 1 s. Although these investigators20 used a fiber-to-target distance of only 2 mm, it is possible that the effect would be similar if applied at a greater distance because of the dynamic change of absorption with repetitive irradiation.16 21 22
Although spectroscopic studies23 24 have shown similar alteration in optical properties of organic polymers with repetitive irradiation, the paucity of adverse effects from prolonged irradiation at 10 W may be related to the cooling times of the stent materials. Before metals and organic polymers are uniformly heated, absorbed laser energy is equitably partitioned among electrons. Expressed as "cooling time," this equipartition may be as short as 10-13 s in metals, and between 10-12 to 10-6 s in nonmetals.25 26
In our study, it is noteworthy that the covered Wallstent was damaged at an even lower power density (75 W/cm2; 10 W, 20 mm). The covering of the original model of the self-expanding Wallstent is a thermoplastic elastomer made with a urethane backbone. This material does not have the high-temperature resistance of conventional vulcanized rubber and cannot tolerate temperatures of 88°C to 100°C.27 This is probably why the covered Wallstent ignited immediately at a power setting of 40 W.
We also found that uncovered Wallstents, covered Wallstents, and silicone stents were all damaged at power densities > 225 W/cm2 and at power settings > 30 W. The thermal conductivity of most metals ranges from 0.66 to 4.180 W/cm2/°C/cm,28 while that of silicone rubber is 6 to 12 W/cm2/°C/cm.27 Comparatively, the thermal conductivity of water, a major constituent of biological tissues, is only 0.005 W/cm2/°C/cm.28 A radiant power density of 1460 W/cm2 is required to vaporize biological tissues.22 Despite their greater thermal conductivities, all stents were damaged at power densities commonly used to coagulate rather than to vaporize tissues. This might indicate that stent materials have greater absorptivity of Nd-YAG laser energy than biological tissues. Sainte-Catherine and coworkers29 observed, for example, that metal surfaces had greater absorptivity for Nd-YAG laser energy than CO2 laser energy. This finding is contrary to the known behavior of Nd-YAG and CO2 laser systems in human tissues.
We believe the potential for increased absorptivity of laser energy by stents is real because, as in the case of the covered Wallstent, temperatures rise rapidly, prompting stent ignition without warning. Obviously, the presence of dark pigment such as blood in or around the stent would further increase the absorption of Nd-YAG laser radiation, precipitating stent ignition and airway burn injury.
The silicone stent, although becoming discolored, never ignited. Perhaps this is because the polydimethylsiloxane chain, the basic skeleton of the silicone stent, can withstand temperatures ranging from - 70°C to + 250°C.27 We ceased laser firing, however, the moment we saw discoloration rather than continuing until stent ignition. The reason for this is that in actual clinical practice, the severely grave consequences of airway burn injury must be avoided, and it is well recognized that Nd-YAG laser irradiation of darkly pigmented or charred material increases surface absorption.30
Although our study suggests parameters that might allow granulation or tumor tissue surrounding a silicone stent to be resected without removing the stent, careful attention must be paid to these parameters and to the rules governing safe bronchoscopic laser resection. The importance of laser/tissue interactions was well demonstrated by Scherer,31 who, in a recently published in vitro study, was able to ignite blood or soot-covered silicone stents using multiple laser power settings, whereas clean silicone stents could not be ignited regardless of power density or oxygen concentration.
Our study might raise a question pertaining to the effect of local temperature on laser/tissue interaction. The threshold temperature at which materials are damaged by laser energy is more easily reached when the baseline temperature of a target material is higher and nearer the threshold temperature. The laser energy delivered in our study was applied to tissues residing within an experimental model at less than true body temperature, but greater than room temperature. Predictably, the stents placed in our model were also at higher baseline temperature, and thus more easily damaged by laser energy than if operated on at room temperature.
Another potential concern about our study is that the laser beam was focused on each stent, rather than on surrounding tissues. During tissue removal of granulation tissue, the laser beam is usually focused on the tissue itself, which may or may not be directly adjacent to the indwelling stent. Because laser energy is easily scattered and reflected, however, collateral absorption is frequent. Laser energy delivered onto surrounding tissues, therefore, will also be frequently absorbed by adjacent foreign bodies. In our study, stents were frequently damaged after only one or two impulses at higher power densities, suggesting that substantial variation of target points during bronchoscopic laser resection would be necessary to prevent inadvertent stent damage.
Because uncovered Wallstent and silicone stents were damaged after just
a few pulses at 30 W and 40 W, but were not affected by prolonged
irradiation using 10 W, questions are raised pertaining to the concept
of energy density per pulse (defined as power density
[watts/centimeters squared] multiplied by pulse duration in seconds)
as a defining factor of laser safety. We cannot address this issue,
however, because we did not vary impact duration times in our study. In
the study by Scherer,31
impact duration significantly
contributed to the likelihood of silicone stent ignition, prompting
that authors recommendation to maintain impact duration to
1 s
when using power settings of
40 W.
Finally, both in our study and in that of Witt et al,19 there is an absence of a significant relationship between power density and total energy (in joules) delivered before stent damage occurs. Probably, this is related to the lack of influence of distance on total energy delivered. It is tempting to presume that the total energy needed to damage a stent decreases as power density increases. As shown in Figure 4 , however, at various power densities damaging all three stents, the total energies delivered were similar. For instance, at power densities of 300 W/cm2 (40 W, 20 mm) and 690 W/cm2 (40 W, 10 mm), differences in total energy delivered to damage an uncovered Wallstent were not statistically significant (54.3 ± 21.8 J vs 56.3 ± 18.5 J, p < 0.85). Similar observations were made for silicone stents damaged at power densities > 225 W/cm2, and covered Wallstents, which were damaged at all power densities.
One reason why total energy may be an inconsequential measure of outcome for any given power density is that power density involves not only the time rate at which energy is expended but also the spatial extent of that energy.18 In biological tissues, for example, a wide range of power densities can achieve similar macroscopic effects (ie, tissue blanching). Although total energies delivered to achieve these effects might be similar, each power density will be different in terms of tissue margins and depth of coagulation. Energy delivered from a site near the target tissue, for instance, achieves narrower lateral margins and a deeper depth of coagulation than energy delivered from a site far from the target tissue (dependent on wavelength absorption characteristics).
In summary, our results suggest that a wide range of power densities can be damaging to an airway stent, and that total energy alone should not be used to delineate margins of safe laser resection. Applying our findings to the clinical setting, bronchoscopists should be most aware of power density, rather than concerned with power settings or total energy only, when contemplating laser/stent interactions.
At an FIO2
40%, a margin of
safety exists for Nd-YAG laser applications onto or around an uncovered
Wallstent or silicone stent using power densities
172
W/cm2, or a power setting of 10 W with a
fiber-to-target distance of 10 mm. The common practice of removing
these stents, specifically the silicone stent, prior to laser resection
may in this case be unnecessary. A margin of safety is not as readily
apparent for covered Wallstents, which at higher power densities ignite
easily and without warning.
| Acknowledgements |
|---|
| Footnotes |
|---|
Funded in part by University of California-San Diego Academic Senate Grant No. 07427A.
Received for publication September 11, 2000. Accepted for publication March 26, 2001.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Ernst, D. Feller-Kopman, H. D. Becker, and A. C. Mehta Central Airway Obstruction Am. J. Respir. Crit. Care Med., June 15, 2004; 169(12): 1278 - 1297. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |