Effect of Medialization Thyroplasty on Glottic Airway Anatomy: Cadaver Model

Effect of Medialization Thyroplasty on Glottic Airway Anatomy: Cadaver Model *,†Tulika Shinghal, *,†Jennifer Anderson, ‡Janet Chung, §Aaron Hong, and kAditya Bharatha, *yxkToronto and

zMississauga, Ontario, Canada

Summary: Objectives. The purpose of this study was to investigate the change in airway dimensions after medialization thyroplasty (MT) using a cadaveric model. Helical computerized tomography (CT) was performed before and after placement of a silastic block in human larynges to investigate the effect on airway anatomy at the level of the glottis. Tissue density (TD) of the medialized vocal fold (VF) was documented to understand the effect on tissue displacement. Study Design. This is a cadaveric study. Methods. Thirteen human cadaveric larynges underwent fine-cut CT scan before and after MT was performed using carved blocks in two sizes (small block and large block [LB]). Clientstream software was used to measure laryngeal dimensions: intraglottic volume (IGV), cross-sectional area (CSA), posterior-glottic diameter (PGD), VF density (in Hounsfield units [HUs]), and anterior-posterior diameter (APD). Eight sequential axial sections 0.625 mm cuts) at the level of the true VFs were analyzed. Results. There was a significant decrease between the three conditions for IGV (P < 0.0001) and CSA (P < 0.0001). TD of the VF was increased after MT as indicated by HU increase (P ¼ 0.0003). APD was not significantly changed. PGD was significantly different between the no block to LB placement (P ¼ 0.0012). Conclusions. MT significantly changes the IGV and CSA at the level of the glottis. Density in the true VF was significantly increased. These findings have important implications for understanding volumetric effects of MT. Key Words: Medialization thyroplasty–Postthyroplasty intubation–Vocal fold paralysis–Complication–Airway compromise–Glottic volume.

INTRODUCTION Unilateral vocal fold paralysis (UVFP) is an important clinical problem in laryngology that not only affects the voice and swallowing but also airway function.1 One of the mainstays of surgical management of UVFP for the past 25 years has been medialization thyroplasty (MT), which was initially described by Payr in 1915.2 This technique was modified by Isshiki et al, who also proposed using alloplastic materials for medialization.3,4 MT enables better voice production by repositioning the immobile vocal fold (VF) into an adducted posture. In addition, MT has been shown to improve the associated swallowing dysfunction with a reduction in pneumonia and shorter lengths of stay in hospital for UVFP patients.5 Various materials have been used for medialization including silastic block, Montgomery implant (Boston Medical Products, Westborough, MA), GORE-TEX (W.L. Gore and Associates, Elkton, MD), and other types of implants. The complication

Accepted for publication August 12, 2015. This work was done in the Otolaryngology–Head and Neck Surgery and Medical Imaging Departments of St. Michael’s Hospital, Toronto, ON, Canada. This article was presented as a poster presentation at American Laryngological Association (ALA), COSM, Boston, USA, on the 22nd of April 2015. From the *Department of Otolaryngology–Head and Neck Surgery, University of Toronto, Toronto, Ontario, Canada; yDepartment of Otolaryngology–Head and Neck Surgery, St. Michael’s Hospital, Toronto, Ontario, Canada; zDepartment of Otolaryngology–Head and Neck Surgery, Trillium Health Partners, Mississauga, Ontario, Canada; xDepartment of Anesthesia, St. Michael’s Hospital, Toronto, Ontario, Canada; and the kDepartment of Medical Imaging, St. Michael’s Hospital, Toronto, Ontario, Canada. Address correspondence and reprint requests to Tulika Shinghal, Department of Otolaryngology–Head and Neck Surgery, St. Michael’s Hospital, 30 Bond Street, 8C-121, Toronto, ON M5B 1W8, Canada. E-mail: [email protected] Journal of Voice, Vol. -, No. -, pp. 1-6 0892-1997/$36.00 Ó 2015 The Voice Foundation http://dx.doi.org/10.1016/j.jvoice.2015.08.009

rate has been estimated to be between 8% and 15%, which are largely minor complications such as VF edema and wound infections, although extrusion, hematoma, and rarely tracheotomy5–7 have also been reported. According to a survey in 1998, the most common major complications were implant migration and failure to improve voice quality.8 Airway compromise requiring intervention in laryngeal framework is observed to be around 2.2%7 and is more likely with after arytenoid adduction than after medialization laryngoplasty.1 By the nature of the procedure, MT patients will have alteration in the shape of their airway at the level of the true VFs after implant placement. Certain patients will go on to have further surgical procedures, and one consideration after MT is related to future endotracheal intubation. There is a paucity of data regarding the potential risk of perioperative airway complications in post-MT patients undergoing general anesthesia and intubation. The largest report reviewed 74 post-MT patients who underwent general anesthetic compared to 79 procedurematched controls. Perioperative complications were found in 6.8% (five patients). Three patients were managed conservatively (steroids, racemic epinephrine) with glottic edema postoperatively. One patient required conversion from a laryngeal mask to an endotracheal intubation (no further issues). One patient required a tracheotomy for airway management. No airway complications were documented in the control group. The authors concluded that the risk of airway problems in post-MT patients undergoing general anesthesia with endotracheal intubation was nonnegligible.9 To date, no published report has compared laryngeal volumetric anatomy with silastic block placement before and after MT. Hence, the objective of our study was to investigate the

2 change in airway dimensions at the level of the glottis before and after silastic block insertion and to understand the effects on tissue displacement in a human cadaveric model. MATERIALS AND METHODS This study was approved by the St. Michael’s hospital Research Ethics Board (13-241). Thirteen human cadaver larynges (seven women and six men) were used for this study. The human cadavers were prepared by the Anatomy Department at the University of Toronto. A soft floppy embalming solution technique was used (1500 cc of phenol 90%, 2 gallons of water, 0.5 gallons of propylene glycol, and 2 gallons of alcohol; 15 to 20 L was used per cadaver). The cadaver is embalmed in the solution for 2–3 days. The larynx was then excised. Each specimen included the complete larynx from the hyoid to the fifth or sixth tracheal ring. The excised larynges were prepared for MT by the senior author (J.A.) who was assisted by two of the authors (T.S. and J.C.). All specimens had the MT performed on the left side. The left thyroid ala was exposed and the thyrotomy window was created (4 by 8 mm for women; 5 by 10 mm for men). The window was placed approximately 10 mm posterior to the anterior midline of the larynx with the superior border at the midpoint of height of the cartilage. The inferior border of the window was parallel to the inferior aspect of the thyroid ala with a 3- to 4-mm strut of cartilage below the window. Once the outer perichondrium and cartilage was removed (assisted with a drill if necessary), the inner perichondrium was elevated. Two sizes of silastic blocks were carved for each specimen (thyroplasty block with a length of 2 cm [women] and 2.5 cm [men]; Figure 1A). Arbitrarily, the silastic blocks were carved in two sizes with a medialization dimension of 8 and 10 mm for the female larynges. The male specimens had silastic blocks carved with a 10 and 12 mm of medialization. The size of large silastic blocks used in the study is close to the maximum that would be used in a patient undergoing MT.

Journal of Voice, Vol. -, No. -, 2015

Care was taken to not damage the underlying soft tissue when the blocks were positioned into the window just before imaging the specimens. Helical computerized tomography (CT) scanning was performed of all 13 larynges under three test conditions: no block (NB), small block (SB), and large block (LB) implant in place. A General Electric (GE) Light speed Volume Computed Tomography (GE Healthcare, Little Chalfont, United Kingdom) scanner was used. CT scans were performed using the following technical parameters: 120 kVp, 200 mA, 0.625 mm slice thickness with helical acquisition, 0.516:1 pitch, 0.7 second rotation time, 40 mm detector coverage, small body field of view (FOV), and display FOV 140 mm2. The larynges were scanned in a caudocranial direction and carefully positioned in a foam holder such that glottis was in axial plane with the thyroid notch in the midline anteriorly. The foam holder was scanned before the study to determine its density was negligible and would not interfere with tissue scanning. The axial images were analyzed using Carestream Client software (Version 11.4.1, Carestream Client; Carestream Health Inc., Rochester). Parameters collected included crosssectional area (CSA), posterior-glottic diameter (PGD), and anterior-posterior diameter (APD) (Figure 1B). The freehand region of interest (ROI) tool was used to outline the intraglottic airway at the first cut below the inferior laryngeal ventricle to obtain the CSA. A total of eight consecutive axial sections (0.625 mm cuts) from this initial cut were analyzed for CSA. These eight CSAs were summated to calculate the intraglottic volume (IGV). The PGD of the airway was measured at the level of minimum CSA axial cut between the left and right the vocal process tips. The APD of the airway was also measured at this level (Figure 1C). The tissue density (TD) of the true VF immediately anterior to the vocal process as at the level of minimal CSA cut was recorded using the oval ROI tool. TD was expressed in Hounsfield units (HUs). Minimum CSA from each larynx was compared to the CSA of the

FIGURE 1. (A) Thyroplasty block; (B) (i) CSA with silastic block, (ii) APD and PGD from medial aspect of vocal processes with silastic block, (iii) CSAwithout silastic block, and (iv) APD and PGD without silastic block; (C) Actual CT scan demonstrating measures done in a larynx without a block and a larynx with a silastic block in place.

Tulika Shinghal, et al

Effect of Type 1 Thyroplasty: Glottic Anatomy

3

FIGURE 2. There was a statistically significant difference in IGV between the three conditions [F ¼ 23.5 (2,38) (P < 0.0001)]. outer diameter (OD) of standard adult endotracheal tubes (ETTs) (Mallinckrodt, Covidien, Mansfield). Statistical methods Descriptive statistics are reported as means, standard deviations, and standard errors. A one-way ANOVA with repeated measures was performed to compare NB, SB, and LB with IGV, CSA, APD, PGD, and HU. A means comparison test was also conducted to compare subsets of the means, corrected by the Tukey’s test to evaluate all possible pairs of means. A two-way ANOVA was used to evaluate the effect of gender, comparing NB, SB, and LB in the male versus female larynges. Statistical significance was defined as P < 0.05. All analyses were performed using GraphPad Prism (Version 6.0, GraphPad Software; La Jolla, California). RESULTS Thirteen cadaveric larynges were used in this study, including seven female (age: 79.7 ± 15.4 years) and six male (age: 81.5 ± 10.59 years) larynges.

FIGURE 4. There was no statistically significant difference in APD between the three conditions [F ¼ 1.419 (2,38) (P ¼ 0.2624)]. There was a statistically significant difference in IGV between the three conditions [F ¼ 23.5 (2,38) (P < 0.0001)] (Figure 2). Similarly, CSA had a statistically significant difference between the three conditions [F ¼ 34.65 (2,38) (P < 0.0001)] (Figure 3). There was no significant difference in APD between the three conditions (P ¼ 0.6856; Figure 4). There was a statistically significant difference in PGD between the three conditions [F ¼ 12.38 (2,38) (P ¼ 0.0012)]. However, there was no statistically significant difference between the SB (7.00 [±1.77]) and LB (6.57 [±1.58]) condition for PGD (P ¼ 0.0776; Figure 5). It was noted that the reduction in glottis airway CSA was dramatically less than the CSA of the block implanted. This suggests that tissues are being displaced other than into the airway at that level and/or tissues are compressed. As proxy for an increase in tissue compression resulting from MT, the

FIGURE 5. There was a statistically significant difference in FIGURE 3. There was a statistically significant difference in minimal CSA between the three conditions [F ¼ 34.65 (2,38) (P < 0.0001)].

PGD between the three conditions [F ¼ 12.38 (2,38) (P ¼ 0.0012)]. However, there was no statistically significant difference between the SB and LB condition (P ¼ 0.0776).

4

Journal of Voice, Vol. -, No. -, 2015

FIGURE 6. There was a statistically significant difference in density (HU) between the three conditions [F ¼ 17.29 (2,38) (P ¼ 0.0003)]. TD of the true VF immediately anterior to the vocal process was analyzed from the axial scans with the minimum CSA and recorded in HUs. The density in the three different conditions in HU was 33.42 HU (±11.2) for the VF with NB (without the implant), 44.1 HU (±11.92) for the SB, and 56.16 HU (±13.83) for the LB. There was a statistically significant difference in density between the three conditions [F ¼ 17.29 (2,38) (P ¼ 0.0003); Figure 6; Table 1). The parameters for male and female larynges were evaluated separately using Tukey’s multiple comparison test to examine the possible effect of gender (Table 2). There was a significant interaction between block size and gender for IGV [F(2, 22) ¼ 5.718; P ¼ 0.0100] and CSA [F(2,22) ¼ 9.394; P ¼ 0.0011]. In the male larynges, IGV and CSA were significantly reduced between NB versus SB and NB versus LB conditions. In the female larynges, IGV was statistically reduced when NB versus LB placement was compared. CSA comparison in the female larynges showed a significant reduction between NB versus SB and NB versus LB condition. Gender did not have a significant effect on change in APD [F(2,22) ¼ 0.3588; P ¼ 0.7025], PDG [F(2,22) ¼ 0.1424; P ¼ 0.8681], or density (HU) [F(2,22) ¼ 0.04831; P ¼ 0.9529] in the three test conditions. The CSAs of a standard ETT sizes 6, 7, and 7.5 are 52.81 mm2, 72.38 mm2, and 81.71 mm2, respectively. All larynges had a minimum CSA larger than a size 6 ETT, and all male larynges had a minimum CSA greater than a size 7.5 ETT.

DISCUSSION MT is an effective treatment option for VF paralysis and glottal incompetence. Since its first description, MT has become a relatively common procedure. Complication rates are reported in range from 8 to 15%.5–7,10 Laryngeal framework surgery with the intent of altering the shape of the glottis and airway anatomy has the potential to lead to airway compromise. Acute intervention for airway distress after MT has been reported in 2.2% of patients.7 The long-term potential effects of MT on future intubations risks have not been studied extensively. This study evaluated laryngeal airway anatomy in a cadaveric model at the level of the true VFs. Helical CT scans were performed and analyzed to assess IGV, minimal CSA, PGD, and APD with and without MT using different size silastic blocks in male and female larynges. Not surprisingly, the study has shown that all these parameters excluding APD are significantly reduced after MT with a silastic block implant. In men, the minimal CSA was reduced from 174.3 mm2 (±37.06) to 124.4 mm2 (±30.92) when the largest size silastic implant was in place. In women, the minimal CSA was reduced from 92.38 mm2 (±15.36) to 70.17 mm2 (±16.82). The CSAs of a standard number 6, 7.5, and 8 ETT are 52.8 mm2, 81.7 mm2, and 95.0 mm2, respectively. Although many factors affect intubation, CT scan assessment of airway contour and size is one method of evaluating the glottic lumen. The female larynges all had a CSA larger than a size 6 tube and size 7.5 for male larynges. On the basis of our study, we suggest a size 6 ETT in women and a size 7.5 in men after MT, if endotracheal intubation is necessary. Avoiding the use of an ETT larger than 7.5 in women and 8.5 in men after MT is also recommended. This study has limitations including the relatively small number of larynges, which reduces the power of our study. In addition, the fixation solution may have affected the soft TD of the larynges. Most importantly, applying results derived from a cadaveric model is inherently challenged given it is a static anatomic model. In reference to intubation, multiple factors for safe intubation must, as per usual standard of practice, be taken into consideration. Mallampati score, thyrohyoid distance, upper airway anatomy (tonsils, tongue base) are all factors, which influence safe intubation. A safe intubation and extubation after thyroplasty is also influenced by the movement of the contralateral presumably normal VF, and both cricothyoid joints can

TABLE 1. Volume, Minimum CSA, APD, PGD, and Density of Different Block Condition All Larynges

Volume (mm3)

Minimum CSA (mm2)

APD (mm)

PGD (mm)

Density (HU)

No block Small block Large block P value

724 (±232.8) 625 (±186.4) 578.2 (±173.3) <0.0001

130.2 (±49.97) 108.5 (±41.42) 95.21 (±36.5) <0.0001

20.26 (±3.79) 20.65 (±3.4) 20.67 (±3.49) 0.2624

7.68 (±1.89) 7.00 (±1.77) 6.57 (±1.58) 0.0012

32.33 43.22 55.18 0.0003

Tulika Shinghal, et al

5

Effect of Type 1 Thyroplasty: Glottic Anatomy

TABLE 2. Volume, Minimum CSA, APD, and PGD in Men and Women Volume (mm3) All Larynges

Men

Women

Minimum CSA (mm2) Men

Women

APD (mm) Men

Women

PGD (mm) Men

Women

No block 926.4 (±170.3) 550.6 (±90.39) 174.3 (±37.06) 92.38 (±15.36) 23.62 (±2.50) 17.39 (±1.63) 9.04 (±1.88) 6.52 (±0.87) Small block 774 (±165) 497.3 (±74.91) 142.4 (±35.55) 79.4 (±15.58) 23.82 (±1.88) 17.94 (±1.27) 8.49 (±1.27) 5.73 (±0.89) Large block 718.4 (±137.5) 458 (±88.3) 124.4 (±30.92) 70.17 (±16.82) 23.78 (±2.18) 18.01 (±1.56) 8.00 (±1.04) 5.34 (±0.51)

passively rotate to accommodate ETT. The cadaveric model cannot reproduce the dynamic changes in a patient undergoing MT under local anesthesia wherein the intraoperative findings (voice quality and endoscopy) would guide the size and placement of the thyroplasty material. However, typically the width of medialization would rarely be larger than the dimensions of the LB used in this study. Other clinical considerations would include the addition of arytenoid adduction or arytenoidopexy, which could potentially further decrease the glottic CSA, was not addressed in this study and is an area of future investigation. There has been some suggestion of higher airway complication rates after arytenoid adduction.5,11 Evaluation of patients before and after MT using CT and clinical parameters such as pulmonary function testing will further our understanding of how MT influences airway function. There are no published reports, which use radiological imaging for predicting a difficult endotracheal intubation based on axial views after thyroplasty. Most radiological studies for predicting difficult endotracheal intubation evaluate anatomic measures such as distance from base of tongue to posterior pharyngeal wall (PPW), epiglottis to PPW, uvula to PPW, VFs to PPW, length of epiglottis, and size of tongue.12,13 These factors would not change after MT and hence are not beneficial when evaluating differences pre-MT to post-MT. A recent study designed to develop a predictive model for implant dimensions on the basis of a CT imaging in 11 patients and in three cadaver specimens was reported by Benninger et al.14 This study evaluated glottic dimensions from preoperative CT scans of patients and included consideration of soft tissue compression in MT to custom design silastic implant shape. As mentioned previously, complications arising immediately after the placement of the implant in MT have been well studied, but long-term studies of airway problems related to intubation after MT are rare, and only two reports were identified in the English literature. Friedlander et al reported on 17 patients who required additional surgical procedures after MT. The ETT sizes ranged from size 6–9. None of these patients required perioperative steroids. All were successfully extubated without any postintubation airway complication.15 However, as discussed earlier, the largest series reported on 74 after MT anesthesia encounters

found perioperative anesthesia complications in 6.8% of the patients.9 This report indicates a significant reduction in CSA and IGV after MT in a cadaver model. TD of the true VF was found to significantly increase after implant placement, which likely represents tissue compression. CONCLUSIONS In this study, standardized silastic implants sizes were used for male and female laryngeal specimens to evaluate effects on airway measures. The cadaver MT model shows significant changes the volume, CSA, and PGD at the level of the glottis. Tissue displacement and some degree of tissue compression must also take place after MT to explain the discrepancy between block size and change in CSA at the level of the glottis. On the basis of this pilot study, a future study is planned (cadaver/live model), using CT imaging of the larynges to design individualized implant dimensions, which enable achievement, the ideal VF position. Acknowledgments There are no financial disclaimers. There is no conflict of interest to disclose. REFERENCES 1. Misono S, Merati AL. Evidence-based practice: evaluation and management of unilateral vocal fold paralysis. Otolaryngol Clin North Am. 2012;45:1083–1108. 2. Payr E. Plastik am Schildknorpel zur Behebung der Folgen einseitiger Stimmbandl€ahmung [in German]. Dtsch Med Wochenschr. 1915;43: 1265–1270. 3. Isshiki N, Okamura H, Ishikawa T. Thyroplasty type I (lateral compression) for dysphonia due to vocal cord paralysis or atrophy. Acta Otolaryngol. 1975;80:465–473. 4. Isshiki N, Taira T, Kojima H, Shoji K. Recent modifications in thyroplasty type I. Ann Otol Rhinol Laryngol. 1989;98:777–779. 5. Abraham MT, Gonen M, Kraus DH. Complications of type I thyroplasty and arytenoid adduction. Laryngoscope. 2001;111:1322–1329. 6. Cotter CS, Avidano MA, Crary MA, Cassisi NJ, Gorham MM. Laryngeal complications after type 1 thyroplasty. Otolaryngol Head Neck Surg. 1995;113:671–673. 7. Young VN, Zullo TG, Rosen CA. Analysis of laryngeal framework surgery: 10-year follow-up to a national survey. Laryngoscope. 2010;120:1602–1608. 8. Rosen CA. Complications of phonosurgery: results of a national survey. Laryngoscope. 1998;108:1697–1703. 9. Lin HW, Bhattacharyya N. Incidence of perioperative airway complications in patients with previous medialization thyroplasty. Laryngoscope. 2009; 119:675–678.

6 10. Netterville JL, Stone RE, Luken ES, Civantos FJ, Ossoff RH. Silastic medialization and arytenoid adduction: the Vanderbilt experience. A review of 116 phonosurgical procedures. Ann Otol Rhinol Laryngol. 1993;102: 413–424. 11. Weinman EC, Maragos NE. Airway compromise in thyroplasty surgery. Laryngoscope. 2000;110:1082–1085. 12. Samra SK, Schork MA, Guinto FC Jr. A study of radiologic imaging techniques and airway grading to predict a difficult endotracheal intubation. J Clin Anesth. 1995;7:373–379.

Journal of Voice, Vol. -, No. -, 2015 13. Randell T, Hakala P, Kytta J, Kinnunen J. The relevance of clinical and radiological measurements in predicting difficulties in fibreoptic orotracheal intubation in adults. Anaesthesia. 1998;53:1144–1147. 14. Benninger MS, Chota RL, Bryson PC, Drake RL. Custom implants for medialization laryngoplasty: a model that considers tissue compression. J Voice. 2015;29:363–369. 15. Friedlander P, Aygene E, Kraus DH. Prevention of airway complications in thyroplasty patients requiring endotracheal intubation. Ann Otol Rhinol Laryngol. 1999;108:735–737.