Minimizing Thermal Damage During Thulium Laser-Assisted Partial Arytenoidectomy: Pulsed Versus Continuous Cutting in an Ex-Vivo Calf Model

Introduction

Opening the posterior glottic airway in patients with airway obstruction due to bilateral vocal fold immobility can be achieved with partial arytenoidectomy using lasers.^1-4 Since the initial clinical use of the carbon dioxide (CO₂) laser by Strong and Jako^5,6 and Vaughan, ⁷ the variety of available lasers has expanded to include fiber-based lasers with hemostatic cutting and ablating properties.^8-11 All lasers commonly used in endolaryngeal surgery generate heat, and thermal effects of lasers on laryngeal tissue are well-documented in ex-vivo animal models^7,12-15 and human clinical series.^5,6,10,16,17 Excessive heat can lead to epithelial sloughing, direct damage to the tissue adjacent to the laser cut, and excessive scarring and fibrosis.^18-20 The anterior cut during most partial arytenoidectomy surgeries occurs near the vocal process, at or near the vocal fold membranous attachment, thereby releasing anteroposterior tension of the thyroarytenoid muscle and weakening the voice. Further vocal fold damage, such as would occur with collateral thermal damage to the superficial lamina propria (SLP) or other adjacent laryngeal tissue, should ideally be minimized during partial arytenoidectomy.

The 2 µm thulium laser has emerged as a useful fiber-based cutting tool during partial arytenoidectomy, but prior studies have shown that there can be significant collateral thermal damage when using this laser.^8,10,21 However, these studies utilized the thulium laser in a continuous cutting mode, and no prior study has measured the thermal effects with pulsed-mode thulium laser cutting during partial arytenoidectomy. This study measured temperature changes, time to complete standard trans-cartilage cuts, and thermal trauma during simulated partial arytenoidectomy using the thulium laser, with and without pulsing, in an ex-vivo calf model.

Materials and Methods

Twelve excised calf larynges were used within 12 hours of harvest from a local slaughterhouse, and maintained on ice until being warmed to room temperature at the time of laser cutting. The calf larynges were cut transversely into hemi larynges (N = 24). A 1.0 mm diameter thermistor (product AB6E5-GC14KA143L-37C, Thermometrics, Edison, New Jersey) was used to measure tissue temperatures. The thermistor was placed near the tip of the vocal process within the superficial lamina propria (SLP) between the phonatory mucosa and the thin vocal ligament separating the thyroarytenoid muscle from the lamina propria ²² (Figure 1A). The thermistor signal in volts was calibrated to provide temperature in centigrade, and the change in temperature was measured from immediately prior to laser application until the completion of the trans-cartilaginous cut from the digitized signal using DigiData 1440 hardware and Axoscope software (Molecular Devices, San Jose, CA). This thermistor position was chosen in order to measure the thermal effect within the SLP.

OPEN IN VIEWER

Figure 1.

Left calf hemi larynx showing placement of thermistor tip within the SLP near the tip of the vocal process (VP). (A) The dotted line 3 mm from the thermistor tip marks the site of proposed arytenoid cartilage cut perpendicular to the vocal fold. (B) Arytenoid cartilage cut has been completed through to the muscle. The thermal damage zone was measured from the side closest to the thermistor.

A 2 µm thulium laser (Revolix Jr, LISA Laser, Katlenburg-Lindau, Germany; 300 µm fiber, 5 W) was used to make controlled cuts through the vocal process of the arytenoid cartilage, simulating the cut made closest to the SLP during standard partial arytenoidectomy. Cuts were made by the senior author under direct visualization through an operating microscope (Leica F40 M525) with continuous wave (CW) power (N = 12), and at various pulsed wave (PW) power settings: 250 ms (N = 4), 500 ms (N = 4), 750 ms (N = 4) at a rate of 1 Hz. For each PW cut, the contralateral vocal fold received a CW cut to enable pairwise comparisons of PW versus CW cutting modes within the same larynx to control for anatomical variation and time post-mortem. The arytenoid cut was made 3 mm from the thermistor tip. Cuts were made until the arytenoid cartilage was ablated through to the muscle, creating an incision that was several millimeters deep and similar for all power settings (Figure 1B). The time to complete the cut was measured for each power setting.

Thermal damage was analyzed histologically by measuring the depth of lactate dehydrogenase inactivation perpendicular to the laser-cut edge at the vocal process (the edge of the cut closest to the thermistor). The incisions were marked with distinguishing tissue dyes, and the tissue blocks were then excised, embedded in water-soluble compound (Tissue-Tek OCT, 4583, Sakura Finetek USA Inc., Torrance CA), frozen in isopentane that was cooled in liquid nitrogen, cryostat sectioned at 10 µ, and mounted on glass slides. The tissue was stained for lactate dehydrogenase using nitro blue tetrazolium for 24 hours,^8,21 counter-stained with eosin and cover-slipped.

Digital photomicrographs (4608 × 3072 pixels) were taken of each specimen using an upright transmission-light microscope (SMZ1270, Nikon USA, Melville, NY) with digital camera (Lumix GH4, Panasonic Co., Osaka, Japan). Thermal damage was analyzed by measuring the elimination of LDH staining in a zone surrounding the incision site that represents the denaturing of LDH enzyme due to heating. To evaluate the average thickness for each zone of damage, measurements were made by 3 researchers blinded to cutting mode using a computer-based analysis software (ImageJ Ver 1.53K; http://imagej.nih.gov/ij). For each tissue section, the area of LDH depletion was outlined in the digital image and divided by the width of the laser cut from the medial epithelial surface to the lateral border of the laser-cut specimen, providing the average thickness of LDH depletion from the cut at the vocal process toward the anterior commissure (Figure 2).

OPEN IN VIEWER

Figure 2.

Nitroblue tetrazolium chloride-stained glottic tissue shows the LDH depletion (thermal damage) for pulsed and continuous wave cuts. The dotted line indicates the boundary of LDH depletion. The laser-cut edge is at the top of each image, with the medial edge of the SLP toward the left. The lamina propria separated from the underlying thyroarytenoid muscle in most specimens, so muscle is not visible in these representative sections and was not included in the LDH depletion measurements.

Measurements of thermal tissue damage, temperature rise during cutting, and time to procedure completion were compared using paired t-tests for the CW and PW cutting mode used on the left and right sides of each larynx, respectively. Regression analysis was used to calculate the correlation between the thermal damage depth and the peak temperature.

Results

The average laser energy (178 ± 23.8 mJ) delivered remained constant between the CW (179 ± 22.5 mJ) and PW (177 ± 25.9 mJ) cuts, confirming that the incision was similar for all power settings. The average total increase in SLP temperature was 6.5°C ± 3 for CW cuts which was lower than each of the PW cuts (250 ms—17.9°C ± 3.4; 500 ms—16.3°C ± 8.1; 750 ms—19.1°C ± 4) (P < .05) (Figure 3). Furthermore, the rate of temperature change (slope) was overall significantly lower during CW cuts (0.179°C/s ± 0.008) compared to the PW cuts (0.284°C/s ± 0.028; P < .05). When examined by PW group, the rate of temperature change was significantly lower for CW and 250 ms PW cuts than during the 750 ms PW cuts (P < .05). The average time to complete standard trans-cartilaginous cuts was significantly less for CW cuts compared to PW cuts at 250 and 500 ms pulse durations (P < .05) (Figure 4).

OPEN IN VIEWER

Figure 3.

Average total increase in temperature for continuous wave (CW) cuts paired with the corresponding pulsed wave cut. Temperature increase was significantly higher for the 3 pulsed wave (PW) cutting modes versus CW mode (P < .05).

OPEN IN VIEWER

Figure 4.

Average time to complete standard trans-cartilaginous cuts comparing CW and PW modes. Procedure time for the pulsed conditions was statistically longer for the 250 and 500 ms pulsed cutting modes compared to the continuous wave (CW) mode (P < .05), and borderline yet insignificantly lower for the 750 ms mode versus CW (P = .058).

The average depth of LDH depletion (thermal damage) was similar for all cutting modes (Figure 5). The transition between LDH presence versus absence based on nitroblue tetrazolium chloride staining was somewhat variable, but interobserver reliability was good among the 3 individuals measuring depletion, with the average pair-wise correlation being R ²  = .82. There was no significant correlation between rise in tissue temperature and the depth of tissue damage (R ²  = .0088; P > .05). No cartilage was present in the histologic specimens because the cartilage was either vaporized in the cutting process or the remnants fell off during processing, so thermal damage was measured only in the SLP.

OPEN IN VIEWER

Figure 5.

The average depth of LDH depletion (thermal damage) compared for each pulse duration versus the contralateral vocal fold cut in continuous mode. Two-tailed, pair-wise comparisons showed no difference in thermal tissue damage between PW and CW cutting modes for 250 and 750 ms, with 500 ms being borderline yet not statistically significant (P = .054).

Discussion

This study showed that pulsed wave (PW) thulium laser cuts raised glottic tissue temperature higher than continuous wave (CW) cuts while creating similar collateral thermal damage during simulated partial arytenoidectomy in an ex-vivo calf model. This was surprising, since prior studies had shown that delivering laser energy in a pulsed mode (thereby allowing time for the surrounding tissue to cool between pulses) could reduce thermal trauma.^9,16,17,23 However, direct comparisons of thermal trauma between laser cutting modes in prior studies most often controlled for the cut duration, thereby delivering less total energy during PW cuts versus CW cuts.^8,21 This study delivered similar total energy for all cuts (average 178 ± 23.8 mJ) because the endpoint of cutting was a trans-cartilaginous cut through the arytenoid no matter how long the cut took to complete. Therefore, this simulation more accurately reflects the thermal effects created during partial arytenoidectomy because the time to complete the cut is relevant. Even so, the pauses between pulses in this study apparently did not mitigate the accumulation of heat generated within each pulse, so the use of much shorter pulses delivered at higher repetition rates might have yielded less thermal accumulation than the pulses used in this study.^24,25

The rate of temperature change (slope) was significantly lower during CW cuts compared to the PW cuts. This is consistent with clinical practice, wherein the procedure time is so quick during CW cuts that the temperature has less time to peak. The 250 ms pulsed cuts also have a lower temperature slope due to the greater amount of cooling time between pulses; however, the time to complete the cut at 250 ms is so much greater than CW cuts that any advantage of tissue cooling from pulsing is negated. Similar thermal damage zones for all PW and CW cuts confirms this observation.

Utilizing the ex-vivo calf vocal fold is an effective and efficient way to obtain initial data regarding the thermal effects of lasers and the response to cooling. ²¹ While there are obvious limitations to this model, such as variations in soft tissue hydration, temperature, and turgor when compared to perfusing tissue, the data are useful in calibrating optimal parameters for lasering and simulating partial arytenoidectomy. The thermal damage recorded within the SLP support the concept that collateral thermal damage can occur unrelated to the primary treatment of disease. Data from this study may be underestimated due to sloughing of carbonized tissue along the cut edge. This debris (seen in Figure 1B, and consistently noted during all cuts) may not be present during histologic analysis. In fact, since no cartilage was present in the histologic specimens (either from vaporization in the cutting process or sloughing during processing), there might be greater thermal damage than what was measured herein due to loss of the true starting edge for LDH measurements. Since the trans-cartilaginous cut through the vocal process during partial arytenoidectomy is close to the SLP, every effort should be made to minimize collateral damage. This study showed that using CW thulium laser cuts were completed significantly faster than PW cuts and the thermal damage was similar. Results from this investigation can serve as a foundation for in-vivo studies that will ultimately translate into improved laser use in the clinical practice of endolaryngeal surgery.

Further study in a perfusing animal model, such as the rat larynx model proposed by Mallur et al, ²⁶ or by direct measurements of tissue temperatures during human surgery would overcome the limitations of an ex-vivo model and perhaps yield further information that would directly improve surgical outcomes. This study evaluated the thulium laser because it is the preferred surgical tool of the senior author during partial arytenoidectomy; but future studies should seek to compare thermal damage created with various lasers (such as the more commonly used CO₂ laser) at various settings during partial arytenoidectomy.

Conclusions