- Email: [email protected]
The biomechanics of the medialization laryngoplasty(thyroplasty type 1) in an ex vivo canine model
Joumal of Voice Vol. 12. No. 3, pp. 372-382 © 1998 Singular Publishing Group. Inc.
The Biomechanics of the Medialization Laryngoplasty (Thyroplasty Type 1) in an ex vivo Canine Model J. Pieter Noordzij, David A. O p p e r m a n , D o n a l d E Perrault, Jr., and P e a k W o o Department of Otolao,ngology--Head and Neck Surgeo" Tufts Universit3, School of Medicine, New England Medical Centel; Boston, Massachusetts, U.S.A.
Summary: The biomechanics of medialization laryngoplasty are not well understood. An excised canine larynx model was used to test the effects of various sized silicon implants. The vocal fold length, position, and tension were measured. Medialization laryngoplasty did not affect vocal fold length. At the midmembranous vocal fold, larger shims resulted in greater medialization and tension. Medialization laryngoplasty neither medialized nor stiffened the vocal process to resist iateralizing forces. We conclude that medialization laryngoplasty provides bulk and support for defects of the membranous region of the vocal fold, but does not appear to close a posterior glottal gap. The selection of a surgical procedure to treat glottal incompetence should take into account the unique biomechanical properties of the anterior (membranous vocal folds) and posterior (cartilaginous portion) glottis. Key Words: Medialization laryngoplasty~Thyroplasty type l--Biomechanics--Unilateral vocal fold paralysis-Glottis--Glottal incompetence--Medialization---Phonosurgery--Vocal fold-Vocal cord--Tension.
INTRODUCTION
symptoms of reduced vocal intensity, restricted frequency and intensity range, increased vocal perturbation, breathiness, shortness of breath, impaired audibility, pitch elevation, and fatigue (1). Glottal incompetence may be due to unilateral vocal fold paralysis (UVFP), vocal fold atrophy or bowing as seen in sulcus vocalis and presbylaryngis, or deficient vocal fold tissue caused by trauma or surgery (2,3). Of these pathologic conditions, UVFP is by far the most common cause of glottal incompetence. Much research has been done to find solutions to the problem of glottal incompetence caused by the above-mentioned diseases, and multiple effective surgical procedures have been developed. The goals of surgical therapy for glottal incompetencies are many. The procedure may include medial displacement of the vocal fold in order to close the glottal
Otolaryngologists are often presented with patients that suffer from dysphonia secondary to glottal incompetence. These glottal incompetencies may be due to abnormal position, shape, level, and/or tension of the vocal folds. These patients may present with
Accepted for publication, November 19, 1997. Presented at The Voice Foundation's 25th Annual Symposium: Care of the Professional Voice and The International Association of Phonosurgeons' IVth International Symposium on Phonosurgery, Philadelphia, Pennsylvania, June 8, 1996. Address correspondence and reprint requests to Peak Woo, MD, Department of Otolaryngoiogy--Head and Neck Surgery, Mt. Sinai Medical Center, One Gustave Levy Place, P.O. Box 1189, New York, NY.
372
THE BIOMECHANICS OF 7¥IE MEDIALIZATION LARYNGOPLASTY
gap, straightening of the vocal fold edge along the length, and restoring tension to the vocal fold. Augmentation of the bulk of the vocal fold may be appropriate when there is atrophy or paralysis (4). Medial displacement of the vocal fold is considered most important in order to achieve good phonation in cases of UVFP (3). Various laryngoplastic phonosurgical techniques for glottic incompetence have been developed and refined over time (2). At present, the techniques to correct glottic incompetence can be classified into three major types: (a) injection laryngoplasty, (b) medialization laryngoplasty (ML), and (c) arytenoid adduction (AA) (5). One of the most common of these surgical therapies is ML (6). ML has been shown to be an effective treatment for glottal incompetence through both clinical and histological studies (7-10). A study by Koufman found that I0 out of 11 patients who underwent ML for UVFP showed improvements in voice quality (8). Other long-term studies involving patients who have undergone ML show significantly higher fundamental frequencies, moderation of habitual vocal intensity, and longer maximum phonation times (11). However, Woodson cites failure of the ML procedure to correct deficient vocal fold length in cases of UVFP (12). Netterville, Stone, Luken, Civantos, and Ossoff also reported that ML does not always close a posterior glottal gap (13). Despite these observations, the biomechanical properties of ML have not been investigated. Some investigators have examined the biomechanics of the larynx but not in the setting of the ML procedure. Haji, Moil, Omorin, and Isshiki explored the mechanical properties of the vocal fold in freshly excised human larynges (14). Tran, Berke, Gerratti, and Kreiman devised a method of intraoperatively measuring stiffness (Young's Modulus) for the human vocal fold (15). However, the biomechanical properties of the larynx in the setting of ML have not been studied. A better understanding of the biomechanics of ML may be useful in clinical management issues. Moreover, an increased knowledge of the biomechanical effects of ML may help in deciding whether ML is the appropriate choice of therapy in various cases of glottal incompetence. A number of questions regarding the effects of ML on the larynx exist. What is the effect of shim size on glottal configuration? Do implanted shims effect vo-
373
cal fold length? Does an implanted shim medialize the entire vocal fold? Why is ML effective in treating bowed or atrophied vocal folds? Does the vocal process of the arytenoid cartilage change position with varying shim sizes? Can longer shims obliterate a posterior glottal gap? Does ML increase vocal fold stiffness? Does ML increase the ability of the arytenoid cartilage to resist lateralizing forces? The present study was designed to. answer these questions. A model was devised that used excised canine larynges to examine the biomechanics of ML. The purpose of this model was to examine the effect of various sized shims on glottal configuration (vocal fold length and position) and vocal fold tension at the midmembranous vocal fold and at the vocal process. Our goal was to determine if these biomechanical properties could be correlated with clinical observations. M A T E R I A L AND M E T H O D S
Specimen preparation Seven canine larynges were immediately harvested after sacrifice from a variety of breeds. All dogs were adults and weighed between 30-40 lbs. Supraglottic tissues were removed above the vocal folds and tracheal rings were removed inferiorly. The larynx was then mounted to a cylindrical plexiglass base using eight 0-0 silk structure ligatures, securing the cricoid cartilage to the plexiglass base. The plexiglass base was fastened to facilitate force measurements without cricoid cartilage movement. According to the procedure developed by Nobuhiko Isshiki for medialization laryngoplasty, a window was created in the thyroid cartilage (2). After identifying the thyroid notch and the lower border of the thyroid cartilage, the midpoint between these structures was determined. Next, using a caliper, the superior anterior comer of the window was marked 5 mm lateral to the noted midpoint. A window 5 m m in height by 14 mm in length was created using a #11 scalpel blade. A great deal of care was used during this procedure to ensure minimal damage to the underlying perichondrium. Four shim implants were carved from a Silastic ® block (Dow Coming, Midland, MI). Each shim had a wedge shape as depicted in Fig. 1. Two of the dimensions of the wedge were varied among the shims, resulting in four different sized shims. Shim A was the smallest, whereas Shim D was the largest. Jounml of Voice, Vol. 12, No. 3, 1998
374
J. PIETER NOORDZIJ ET AL.
6 X Y
Shim X
3
Y
16
A
mm m m
Shim
B
3 mm 20
m m
Shim
C
6 mm 16
m m
Shim 6
20
D m m
mm
F I G . 1. D i a g r a m o f S i l a s t i c ® s h i m .
Force measurements Vocal fold tension was defined as the force necessary to lateralize the vocal fold (a movement that also occurs during physiological vocal fold vibration). Vocal fold tension was measured at the midmembranous vocal fold and vocal process. The vocal fold was lateralized by a steel rod (2.5 mm diameter, 20 mm in length) mounted to a load cell force transducer. Collectively, the steel rod and load cell were called force sensors. Force sensor #1, a high sensitivity load cell (LCL 227, Omega, Stamford, CT) calibrated to measure a maximum force of 40 g with a ___5% error, was used to measure tension at the midmembranous vocal fold. Force sensor #2 (LCL 816 g, Omega, Stamford, CT), calibrated to measure a maximum force of 200 g with a ---5% error, was used to measure the larger tensions at the vocal process. Each force sensor was mounted to a micrometer translation stage (Newport, Irvine, CA). The micrometer stage could be adjusted in discrete intervals, allowing for movement of the force sensors into the vocal folds. The force sensors were connected to a data acquisition system which consisted of signal conditioning amplifiers (Burr Brown, Tucson, AZ) interfaced to a personal computer with an A/D Converter (NB-MIO-16, National Instruments Austin, TX) controlled by data acquisition software (Labview, National Instruments, Austin, TX) (Fig. 2).
Experimental method Repeated trials using the four different shims were performed for each of the excised larynges. First, voJournal of Voice, VoL 12, No. 3. 1998
cal fold tension at the midmembranous vocal fold and vocal process was measured prior to the creation of the thyroid window (control state). Next, a window was created in the thyroid cartilage as described above. Vocal fold tension measurements were recorded prior to shim insertion (second control state). Next, shims were inserted between the inner perichondrium and the thyroid cartilage rim through the created window. Vocal fold tension measurements were then recorded for each successive shim (four in all). This experimental procedure was repeated for seven canine larynges. The measurements for the vocal tension were performed in the following manner: The steel rod of the force sensor was oriented perpendicular to the vocal fold and positioned so as to be just touching the affected vocal fold. The midpoint of the steel rod was aligned with the vocal fold edge. The steel rod was then laterally moved against the vocal fold via the micrometer translation stage which moved in 1-mm increments over a 5-mm interval. The resultant vocal fold tension was continuously measured (via the load cell) as a function of time over the entire 5-mm excursion at a rate of 10 samples per sec. The measurement procedure was carried out at the midmembranous vocal fold and vocal process with force sensors #1 and #2, respectively. Because the force sensors were subject to very slight flexion as the micrometer moved the force sensors against the vocal fold, the vocal folds were not lateralized in exactly 1-mm increments. The spring constants for both force sensors #1 and #2 were found to be constant.
Glottal configuration analysis A Panasonic solid-state video camera was mounted above the larynx in each experiment to document the effects of the implanted shim on the vocal fold configuration. A personal computer with a frame grabber was used to digitize these top view images from each canine experiment for each shim implant. A software program (Optimas Bioscan 5.1, Seattle, WA) was then used to measure the length and degree of medialization of the affected and opposite (control) vocal folds. Length of the membranous vocal fold was measured by first marking the anterior commissure, the affected vocal process, and control vocal process.
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
375
I Camera I
/
\ Vocal Cord
Vocal Fold Tension Sensor and Micrometer I Video Digitizer
Cartilage I
I
-
I
II
I1
These markers gave each point "x,y" coordinates m the axial plane of the glottis. Length could then be calculated from these "x,y" coordinates using the Pythagorean theorem. From the digitized images, degree of vocal fold medialization was determined using an angle technique (Fig. 3). Degree of vocal fold medialization was measured for the midmembranous vocal fold and the vocal process. Zero percent medialization was defined as no movement, whereas 100% medialization was defined movement of the vocal fold exactly to the midline. Again, specific points on the image were marked and thus assigned "x,y" coordinates: the anterior and posterior commissure (AC and PC), the affected and control midmembranous vocal fold edge (AMVF and CMVF), and the affected and control vocal process (AVP and CVP). For each image, four angles were calculated: 1. Angle 2. Angle 3. Angle 4. Angle
Aff.MVF: Con.MVF" Aff.ve: Con.vP:
Data Acquisition Computer
FIG. 2. Schematic o f medialization laryngoplasty b i o m e • chanics experiment. (Adapted from Ingo R. Titze. Principles o f Voice Production. Copyright 1994 by Allyn and Bacon. Reprinted by permission.)
AMVF-AC-PC CMVF-AC-PC AVP-AC-PC CVP-AC-PC
Degree of medialization could then be calculated using these equations:
Degree of med. for the AMVF = (Angle con.MV~ -- Angle Aft.MVF) Angle con.MVF× 100% Degree of med. for the AVP = (Angle co.. vP - Angle An.VP) Angle co.. vP × 100%
Data analysis The raw data was filtered with a low pass Bessel filter software program to reduce the noise level on the signal. The filtered data, vocal fold tension as a function of time, consisted of a series of increasing peaks which exponentially decayed to a steady state value. These steady state values corresponded with the vocal fold tension at a given force sensor deflection distance. These values were measured and then plotted as a function of implant size. Statistical methods Means and standard errors were calculated from the data collected from the seven canine larynx experiments. The Kruskal-Wallis test (nonparametric one-way ANOVA [analysis of variance]) was used to compare vocal fold length and degree of medialization for the various shim implants. The Friedman test (nonparametric two-way ANOVA) was used to compare vocal fold tension for the various shim implants. Journal of Voice, Vol. 12, No. 3, 1998
376
J. PIETER NOORDZIJ ET AL.
Mid-Mem. Vocal Fold AC
AMVF
i
\ ngle Con. •
Angle Aft.
Vocal Process
Angle Con.
When either the Kruskal-Wallis test or Friedman test was significant, the Wilcoxon signed rank test was used to make selective clinically relevant pairwise comparisons (Tables 1 and 2). A resulting p (probability) level of 0.05 or less was defined to indicate a statistically significant difference between the compared measurements. RESULTS Visual inspection of the larynges after insertion of each shim implant showed the following (Fig. 4): All shim implants tested increased the general thickness of the affected membranous vocal fold without a noticeable change in vocal fold length. The resting position of the affected vocal fold was more medialized with the larger shim implants compared to the smaller shim implants. The firmness to palpation of the affected membranous vocal fold appeared to increase with the implantation of the larger shim implants. In contrast to the visual changes seen at the mid vocal fold, the implantation of shim implants resulted in no appreciable anatomical changes at the vocal process. The addition of larger shim implants failed to result in any noticeable medialization of the vocal process (compare Fig. 4-a with Fig. 4-b). The posterior glottal gap was not closed despite increasing implant size. Mean normalized membranous vocal fold length as a function of implant type is shown in Fig. 5. The affected vocal fold length was normalized by dividJournal of Voice, Vol. 12, No. 3, 1998
FIG. 3. Degree of medialization was calculated using an angle technique. First, the top view image of the each larynx was marked with six necessary points. Next, the angles of medialization for midmembranous vocal fold and vocal process were calculated. Using these angles, the degree of medialization could be determined.
Angle Aft.
ing it by the affected vocal fold length in the control state (no window created). Likewise, the control (opposite) vocal fold was normalized by dividing it by the control (opposite) vocal fold length in the control state (no window created). This figure shows that the length of both the affected and control (opposite) membranous vocal folds did not significantly change with the addition of any shim type as evidenced by Kruskal-Wallis test (p equalled 0.75 and 0.60, respectively). Vocal fold medialization as a function of implant type is shown in Fig. 6. Medialization of the affected vocal fold was measured at the midmembranous vocal fold and at the vocal process. Zero percent medialization is defined as no movement toward the midline, whereas 100% medialization is defined as movement of the vocal fold exactly to the midline. For the midmembranous vocal fold, there was a significant trend of greater medialization with larger implants (Kruskal-Wallis test, p = 0.0003). Each shim implant resulted in a significant amount of midmembranous vocal fold medialization compared to the control (see Table 1). The greatest amount of medialization of the midmembranous vocal fold occurred with the second largest shim, Shim C (43% medialization). At the vocal process, there was a trend toward greater medialization with larger implants. However the amount of medialization was relatively small and not statistically significant (Kruskal-Wallis test, p = 0.23). The greatest amount
THE BIOMECHANICS OF 7*I-IEMEDIALIZATION LARYNGOPLASTY T A B L E 1. Degreeof midmembranous vocal fold medialization comparisons: control vs. various shim implants using the Wilcoxon signed rank test y Implant Type
T A B L E 2. Midmembranous vocal fold tension comparisons: control vs. various shim implants (A 3 mm force sensor deflection distance) using the Wilcoxon signed rank test y
p Value
Control vs. Window
377
Implant Type
p Value
0.08
Control vs. Shim A
0.03
Control vs. Window
0.24
Control vs. Shim B
0.02
Control vs. Shim A
0.87
Control vs. Shim C
0.02
Control vs. Shim B
0.61
Control vs. Shim D
0.03
Control vs. Shim C
0.03
Control vs. Shim D
0.03
-Significant p values were defined as less than 0.05.
tSignificant p values were defined as less than 0.05.
t. ,,', i<-! ,,.
P
4
-e.~
$
1. No Implant (control)
2. Shim B (small)
3. Shim D (large)
FIG. 4. Frame #1 shows the canine larynx with no shim implant (control), frame #2 shows the canine larynx with Shim B (small implant) in place, and frame #3 shows the canine larynx with Shim D (large implant) in place.
of medialization of the vocal process occurred with shim C (15% medialized). Fig. 7 shows mean midmembranous vocal fold tension as a function of shim type. This figure shows the following: tension at the midmembranous vocal fold increased with the addition of larger shim implants. These changes were significant (Friedman test, p = 0.002). The greatest amount of tension was seen with the implantation of the largest shim. Shim D resulted in 41% greater resistance to lateralization as compared to the control state at the 3 mm deflection dis-
tance. Only shims C and D resulted in significantly greater midmembranous vocal fold tension than in controls (see Table 2). Fig. 8 shows the mean vocal fold tension at the vocal process as a function of shim type. This figure shows that vocal fold tension at the vocal process was not significantly affected by shim implant size (Friedman test, p = 0.31). In summary, vocal fold length was not affected by implant size. Larger implants medialized the mid= membranous vocal fold, more than smaller ones, but Journal of Voice, Vol. 12, No. 3, 1998
378
J. P I E T E R N O O R D Z I J E T AL.
Mean Normalized Vocal Fold Length v. Implant Type 1.01 .=
1.00
d)=l
¢ 0.99 ,.J •o 0.98 o
It. - - 0.97
- - Affected Vocal Fold
o
o
0.96
.N 0.95 m
E 0.94 O
Z
Control Vocal Fold
0.93
=} 0.92 0.91 0.90 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 5. Mean normalized vocal fold length plotted vs. implant type with standard error bars included.
Vocal Fold Medialization v. Implant Type 100 90 80 70 A
60 Midmembranous Vocal Fold
O 50
mm
.N 40
±
30
m
~- 20 I,I. > 10
T
m
m m
m
l_ ,,
Window Only
Shim A
Shim B
m I ml
Shim C
-10
Implant Type FIG. 6. Mean vocal fold mediali~,ation plotted vs. implant type with standard error bars included.
Journal ofVoice, Vol. 12, No. 3, 1998
Shim D
Vocal Process
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
379
Mean Mid-Membranous Vocal Fold Tension v. Implant Type 25.0
Deflection Distance
g
20.0
|o "o
131 mm
15.0
0 ~L ¢U 0 0
>
mm
1~2 mm [ ] 3 mm
10.0
.Q
134 mm
E
[ ] 5 mm
5.0
0.0 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 7. Mean midmembranous vocal fold tension plotted vs. implant type with standard error bars included.
Mean Vocal Fold Tension at Vocal Process v. Implant Type 60.0 A v
=
50.0
~"
Deflection Distance
2 a.
40.0
I0
m ° 0
1~2 mm
30.0 c
i1~!3 mm B 4 mm
o i.~ O It.
mm
[31 mm
20.0
---
10.0
[ ] 5 mm
-
o
O >
0.0 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 8. Mean vocal fold tension at the vocal process plotted vs. implant type with standard error bars included. Journal of Voice, Vol. 12, No. 3, 1998
380
J. PIETER NOORDZIJ ET AL.
did not medialize the vocal process. Thus the posterior glottal gap was not significantly closed by the ML procedure. Vocal fold tension increased with larger implants at the midmembranous vocal fold, but not at the vocal process. DISCUSSION The evaluation and treatment of glottal incompetence due to UVFP, vocal fold atrophy, and trauma have seen much change in the last 30 years. Many advances have taken place in the treatment of this disease. The goals of treatment in these patients are to reduce aspiration and restore phonatory function to nearly normal level. A variety of treatment modalities have been developed that restore phonatory function to near normal. The current treatment options are voice therapy, injection laryngoplasty, laryngoplastic phonosurgery, or laryngeal reinnervation. Voice analysis and assessment by a speech-language pathologist specializing in voice are useful in nearly all patients with dysphonia (16). Very often though, speech therapy alone will not restore phonatory function to a satisfactory level. Once surgery becomes a desirable option, there are a variety of procedures available. The training and experience of the surgeon, as well as patient factors (age and other significant medical problems) play a major role. The appearance of the larynx on videostrobolaryngoscopy is also an important factor in choosing a specific surgical technique. A better understanding of the biomechanics of the larynx before and after surgery would aid the otolaryngologist in choosing a surgery or combination of surgeries for the patient with glottal incompetence. Injection laryngoplasty can be used as the primary medialization therapy to correct glottic incompetence or as an adjunct to enhance medialization results. A variety of substances have been used, including Teflon, gelfoam, collagen, and fat. This technique is the method of choice in patients with incurable malignancies or whose medical status makes them high risk for surgery (5). Although simple to perform, injection techniques have major disadvantages in being irreversible and difficult to tailor. Teflon granuloma and Teflon overinjection can result in deterioration of voice which is difficult to correct (17). Journal of Voice, Vol. 12, No. 3, 1998
Laryngoplastic phonosurgery, including ML, allows for a more precise treatment of glottic incompetence. Popularized by Isshiki, the ML (originally termed thyroplasty type I ) is the most commonly performed laryngoplasty procedure in the setting of UVFP. This surgical technique has been refined by others (8). Technically, the procedure is precise and straightforward in concept. However, the ML is not without its difficulties. Complications of this procedure include implant extrusion, airway distress, and poor voice result. Objective voice results from ML have not been reported in sufficient detail to compare to other procedures. However, many authors, including Isshiki (2), have noted the inability of the medialization procedure to close a posterior glottal gap. Ideal patients to be treated for ML are those with preoperative findings of a mid-cord gap or a bowed vocal fold. When a large posterior glottal gap is present, other surgical procedures should be considered. Although subjective voice improvements in patients with UVFP who undergo phonosurgery are important, a more basic understanding of the physical phenomenon that occurs due to phonosurgery is essential. Optimizing the biomechanics (configuration and tension) of the paralyzed vocal fold should result in improved phonatory ability (3). Green, Berke, and Ward (18) studied the differences in vocal efficiency (acoustic power/subglottic power) between AA surgery and ML in an in vivo canine model and found that AA was superior to ML in vocal efficiency, traveling wave motion, and acoustics. Yet the actual biomechanical effects of ML on the larynx remain to be examined. The present study uncovered several biomechanical properties of ML in the excised canine model. ML does not affect vocal fold length; it medializes and significantly stiffens the midmembranous vocal fold. It would appear that this enhanced midmembranous stiffness is due to medial compression of the vocal fold by the shim implant. Although the vocal processes became slightly medialized with ML, this did not result in significant alterations in glottal dimensions. Additionally, ML does not stiffen the arytenoid cartilage so as to resist lateral movement. ML appears to be efficient in medializing the anterior membranous vocal fold, but is a poor substitute for closure of the posterior glottis and an inadequate substitute for the arytenoid adduction procedure.
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
This study demonstrates that the phonatory larynx can be divided into two biomechanical subunits when considering surgical rehabilitation for glottal incompetence. These anatomic biomechanical subunits are conveniently divided into anterior and posterior subunits. The anterior subunit is comprised of the membranous vocal fold. The posterior subunit is comprised of the arytenoid cartilage and the intrinsic muscles of the larynx that insert on the arytenoid cartilage. There is also clinical relevance to this subdivision of the larynx. In patients with UVFP, the glottis may have a variety of anterior and/or posterior configurational changes. In patients with UVFP or presbylaryngis, the membranous vocal fold may have variable amounts of bowing and atrophy. Similarly, sulcus vocalis may present with atrophy of the membranous vocal fold. UVFP may result in shortening of the vocal folds to varying extents. Vocal fold shortening leads to compression of the contralateral vocal fold (in a compensatory effort to approximate the vocal processes during phonation), which in turns results in increased glottal incompetence (12). In patients with UVFP, the size of the posterior glottal gap and the position of the arytenoid cartilage can be variable as well (12). In the case of deficient vocal fold tissue secondary to intubation trauma, only the posterior subunit is involved. In some patients with high vagal nerve injury, a mid-cord gap, anterior arytenoid rotation, and a posterior gap are present, thus involving both subunits of the glottis. We feel surgical procedures designed to correct glottal incompetence should be individualized to correct the defect within each subunit or be combined to address all the affected subunits. Within the anterior subunit of a larynx affected by glottal incompetence, medializing, lengthening, and stiffening of the membranous vocal fold may be beneficial. These changes will help restore oscillation of the vocal fold with a mucosal wave (body cover theory). In patients with a mid-cord gap secondary to vocal fold atrophy, ML should effectively medialize and stiffen the membranous vocal fold. ML is an excellent choice of surgery to correct a gap and lack of stiffness within the anterior laryngeal subunit. However, ML does not appear to increase vocal fold length, which would be beneficial to some patients with UVFP. In cases of sulcus vocalis, where stiffness around the sulcus is the dominant feature, ML
381
may exaggerate already poor vibratory characteristics. Similarly, in cases of trauma or surgery where the dominant feature is excess scar formation and increased vocal fold stiffness, ML may not be beneficial in improving voice. Within the posterior subunit of a larynx affected by glottal incompetence, the position and tension of vocal process of the arytenoid cartilage are very important in restoring function. A midline vocal process which resists lateralizing movement from the opposite vocal fold should achieve better glottic closure and valving during phonation. A better valve will resist airflow leakage and allow the larynx to serve as a more efficient airflow oscillator. A more efficient valve should allow for a better build-up of subglottic pressure resulting in increased acoustic power. This should translate into more efficient vocalization with better loudness. In this important component of arytenoid function, ML appears to be ineffective. Other surgical therapies, such as reinnervation or AA should be considered if arytenoid position rehabilitation is desired. Our experimental findings are coorborated by clinical observations. Other authors have reported AA as an alternative to ML (19). ML and AA have been used in conjunction to rehabilitate UVFP (13). The clinical impression of the posterior glottal gap being difficult to rehabilitate by ML is supported by findings from this study (13). Furthermore, the foreshortened VF with arytenoid rotation also appears to be difficult to rehabilitate with ML procedure alone (12). CONCLUSIONS When properly done, the ML procedure appears to medialize and the stiffen the midmembranous fold. Thus, the ML appears to be a good choice of surgery to correct bowing or augment atrophic vocal folds due to UVFP, presbylaryngis, or suicus vocalis. However, ML does not appear to significantly lengthen the vocal fold and does not significantly medialize or stiffen the arytenoid cartilage. ML cannot be used to effectively close a posterior glottal gap. In this case, AA should be considered, as it has been shown to medialize and stiffen the arytenoid cartilage. We conclude that it is important to individually consider the anterior and posterior biomechanical subunits of the larynx when selecting a surgical therapy for the Jounzal of Voice, Vol. 12, No. 3, 1998
382
J. PIETER NOORDZIJ ET AL.
rehabilitation of patients with UVFP or other forms of glottal incompetence. REFERENCES 1. Rubin JS, Sataloff RT, Korovin GS, Gould WJ. Diagnosis and treatment of voice disorders. New York: Igaku-Shoin Ltd, 1995:384-5. 2. Isshiki N. Phono-surger3,. Japan: Springer-Verlag, 1989: 82-95. 3. Hirano M. Surgical alteration of voice quality. In: Cummings CW, ed. Otolar3'ngology----Head and Neck SurgepT, Update I. St. Louis: CV Mosby Co, 1989:249-62. 4. Arnold GE. Vocal rehabilitation of paralytic dysphonia. Arch OtolaiTngol 1962;76:358-68. 5. Benninger MS, Crumley RL, Ford CN, Gould WJ, Hanson DG, Ossoff RH, Sataloff RT. Evaluation and treatment of the unilateral paralyzed vocal fold. OtolaD,ngol--Head Neck Surg 1994; I I 1:497-508. 6. Isshiki N, Taira T, Kojima H, Shoji K. Recent modifications in thyroplasty type I. Ann Otol Rhfllol Laiyngol 1989;98: 777-9. 7. Sasaki CT, Leder SB, Petcu L. Friedman CD. Longitudinal voice quality changes following Isshiki thyroplasty type I: The yale experience. Laryngoscope 1990; 100:849-52. 8. Koufman JA. Laryngoplasty for vocal cord medialization: an alternative to teflon. Laryngoscope 1989;96:726-31. 9. Maves MD, McCabe BE Gray S. Phonosurgery: indications and pitfalls. Ann Otol Rhinol Lal3,ngol 1989;98:577-80.
Journal of Voice, VoL 12, No. 3, 1998
10. Isaacson G, Kim JH, Kirchner JC, Kirchner JA. Histology of Isshiki thyroplasty type 1. Ann Oto! Rhh~ol l_xwyngol 1990; 99:42-5. 11. Leder SB, Sasaki CT. Long-term changes in vocal quality following Isshiki thyroplasty type I. La~3,ngoscope 1994: 104:275-7. 12. Woodson GE. Configuration of the glottis in laryngeal paralysis I: clinical study. Laryngoscope 1993;103:1227-34. 13. Netterville JL, Stone RE, Luken, ES, Civantos FJ, Ossoff RH. Silastic medialization and arytenoid adduction: the Vanderbilt experience. Ann Otol Rhinol Lan3,ngol 1993;102: 413-24. 14. Haji T, Mori K. Omori K, lsshiki N. Mechanical properties of the vocal tbld. Stress-strain studies. Acta Otolatyngol (Stockh) 1992; I 12:559-65. 15. Tran QT, Berke GS, Gerratt BR, Kreiman J. Measurement of Young's modulus in the in vivo human vocal folds. Ann Otol Rhinol Lal3,ngol 1993;102:584-91. 16. Sataloff RT. Professional voice: The science and art of clinical care. New York: Raven Press, 199 I. 17. Crumley RL. Teflon versus thyroplasty versus nerve transfer: a comparison. Ann Otol Rhhlol Laryngol 1990;99:759-63. 18. Green DC, Berke GS, Ward PH. Vocal fold medialization by surgical augmentation versus arytenoid adduction in the in vivo canine model. Ann Otol Rhinol l.xu3,ngol 1991; 100: 280-7. 19. Isshiki N, Tanabe M, Sawada M. Arytenoid adduction for unilateral vocal cord paralysis. Arch OtolaJyngol 1978;104: 555-8.
The Biomechanics of the Medialization Laryngoplasty (Thyroplasty Type 1) in an ex vivo Canine Model J. Pieter Noordzij, David A. O p p e r m a n , D o n a l d E Perrault, Jr., and P e a k W o o Department of Otolao,ngology--Head and Neck Surgeo" Tufts Universit3, School of Medicine, New England Medical Centel; Boston, Massachusetts, U.S.A.
Summary: The biomechanics of medialization laryngoplasty are not well understood. An excised canine larynx model was used to test the effects of various sized silicon implants. The vocal fold length, position, and tension were measured. Medialization laryngoplasty did not affect vocal fold length. At the midmembranous vocal fold, larger shims resulted in greater medialization and tension. Medialization laryngoplasty neither medialized nor stiffened the vocal process to resist iateralizing forces. We conclude that medialization laryngoplasty provides bulk and support for defects of the membranous region of the vocal fold, but does not appear to close a posterior glottal gap. The selection of a surgical procedure to treat glottal incompetence should take into account the unique biomechanical properties of the anterior (membranous vocal folds) and posterior (cartilaginous portion) glottis. Key Words: Medialization laryngoplasty~Thyroplasty type l--Biomechanics--Unilateral vocal fold paralysis-Glottis--Glottal incompetence--Medialization---Phonosurgery--Vocal fold-Vocal cord--Tension.
INTRODUCTION
symptoms of reduced vocal intensity, restricted frequency and intensity range, increased vocal perturbation, breathiness, shortness of breath, impaired audibility, pitch elevation, and fatigue (1). Glottal incompetence may be due to unilateral vocal fold paralysis (UVFP), vocal fold atrophy or bowing as seen in sulcus vocalis and presbylaryngis, or deficient vocal fold tissue caused by trauma or surgery (2,3). Of these pathologic conditions, UVFP is by far the most common cause of glottal incompetence. Much research has been done to find solutions to the problem of glottal incompetence caused by the above-mentioned diseases, and multiple effective surgical procedures have been developed. The goals of surgical therapy for glottal incompetencies are many. The procedure may include medial displacement of the vocal fold in order to close the glottal
Otolaryngologists are often presented with patients that suffer from dysphonia secondary to glottal incompetence. These glottal incompetencies may be due to abnormal position, shape, level, and/or tension of the vocal folds. These patients may present with
Accepted for publication, November 19, 1997. Presented at The Voice Foundation's 25th Annual Symposium: Care of the Professional Voice and The International Association of Phonosurgeons' IVth International Symposium on Phonosurgery, Philadelphia, Pennsylvania, June 8, 1996. Address correspondence and reprint requests to Peak Woo, MD, Department of Otolaryngoiogy--Head and Neck Surgery, Mt. Sinai Medical Center, One Gustave Levy Place, P.O. Box 1189, New York, NY.
372
THE BIOMECHANICS OF 7¥IE MEDIALIZATION LARYNGOPLASTY
gap, straightening of the vocal fold edge along the length, and restoring tension to the vocal fold. Augmentation of the bulk of the vocal fold may be appropriate when there is atrophy or paralysis (4). Medial displacement of the vocal fold is considered most important in order to achieve good phonation in cases of UVFP (3). Various laryngoplastic phonosurgical techniques for glottic incompetence have been developed and refined over time (2). At present, the techniques to correct glottic incompetence can be classified into three major types: (a) injection laryngoplasty, (b) medialization laryngoplasty (ML), and (c) arytenoid adduction (AA) (5). One of the most common of these surgical therapies is ML (6). ML has been shown to be an effective treatment for glottal incompetence through both clinical and histological studies (7-10). A study by Koufman found that I0 out of 11 patients who underwent ML for UVFP showed improvements in voice quality (8). Other long-term studies involving patients who have undergone ML show significantly higher fundamental frequencies, moderation of habitual vocal intensity, and longer maximum phonation times (11). However, Woodson cites failure of the ML procedure to correct deficient vocal fold length in cases of UVFP (12). Netterville, Stone, Luken, Civantos, and Ossoff also reported that ML does not always close a posterior glottal gap (13). Despite these observations, the biomechanical properties of ML have not been investigated. Some investigators have examined the biomechanics of the larynx but not in the setting of the ML procedure. Haji, Moil, Omorin, and Isshiki explored the mechanical properties of the vocal fold in freshly excised human larynges (14). Tran, Berke, Gerratti, and Kreiman devised a method of intraoperatively measuring stiffness (Young's Modulus) for the human vocal fold (15). However, the biomechanical properties of the larynx in the setting of ML have not been studied. A better understanding of the biomechanics of ML may be useful in clinical management issues. Moreover, an increased knowledge of the biomechanical effects of ML may help in deciding whether ML is the appropriate choice of therapy in various cases of glottal incompetence. A number of questions regarding the effects of ML on the larynx exist. What is the effect of shim size on glottal configuration? Do implanted shims effect vo-
373
cal fold length? Does an implanted shim medialize the entire vocal fold? Why is ML effective in treating bowed or atrophied vocal folds? Does the vocal process of the arytenoid cartilage change position with varying shim sizes? Can longer shims obliterate a posterior glottal gap? Does ML increase vocal fold stiffness? Does ML increase the ability of the arytenoid cartilage to resist lateralizing forces? The present study was designed to. answer these questions. A model was devised that used excised canine larynges to examine the biomechanics of ML. The purpose of this model was to examine the effect of various sized shims on glottal configuration (vocal fold length and position) and vocal fold tension at the midmembranous vocal fold and at the vocal process. Our goal was to determine if these biomechanical properties could be correlated with clinical observations. M A T E R I A L AND M E T H O D S
Specimen preparation Seven canine larynges were immediately harvested after sacrifice from a variety of breeds. All dogs were adults and weighed between 30-40 lbs. Supraglottic tissues were removed above the vocal folds and tracheal rings were removed inferiorly. The larynx was then mounted to a cylindrical plexiglass base using eight 0-0 silk structure ligatures, securing the cricoid cartilage to the plexiglass base. The plexiglass base was fastened to facilitate force measurements without cricoid cartilage movement. According to the procedure developed by Nobuhiko Isshiki for medialization laryngoplasty, a window was created in the thyroid cartilage (2). After identifying the thyroid notch and the lower border of the thyroid cartilage, the midpoint between these structures was determined. Next, using a caliper, the superior anterior comer of the window was marked 5 mm lateral to the noted midpoint. A window 5 m m in height by 14 mm in length was created using a #11 scalpel blade. A great deal of care was used during this procedure to ensure minimal damage to the underlying perichondrium. Four shim implants were carved from a Silastic ® block (Dow Coming, Midland, MI). Each shim had a wedge shape as depicted in Fig. 1. Two of the dimensions of the wedge were varied among the shims, resulting in four different sized shims. Shim A was the smallest, whereas Shim D was the largest. Jounml of Voice, Vol. 12, No. 3, 1998
374
J. PIETER NOORDZIJ ET AL.
6 X Y
Shim X
3
Y
16
A
mm m m
Shim
B
3 mm 20
m m
Shim
C
6 mm 16
m m
Shim 6
20
D m m
mm
F I G . 1. D i a g r a m o f S i l a s t i c ® s h i m .
Force measurements Vocal fold tension was defined as the force necessary to lateralize the vocal fold (a movement that also occurs during physiological vocal fold vibration). Vocal fold tension was measured at the midmembranous vocal fold and vocal process. The vocal fold was lateralized by a steel rod (2.5 mm diameter, 20 mm in length) mounted to a load cell force transducer. Collectively, the steel rod and load cell were called force sensors. Force sensor #1, a high sensitivity load cell (LCL 227, Omega, Stamford, CT) calibrated to measure a maximum force of 40 g with a ___5% error, was used to measure tension at the midmembranous vocal fold. Force sensor #2 (LCL 816 g, Omega, Stamford, CT), calibrated to measure a maximum force of 200 g with a ---5% error, was used to measure the larger tensions at the vocal process. Each force sensor was mounted to a micrometer translation stage (Newport, Irvine, CA). The micrometer stage could be adjusted in discrete intervals, allowing for movement of the force sensors into the vocal folds. The force sensors were connected to a data acquisition system which consisted of signal conditioning amplifiers (Burr Brown, Tucson, AZ) interfaced to a personal computer with an A/D Converter (NB-MIO-16, National Instruments Austin, TX) controlled by data acquisition software (Labview, National Instruments, Austin, TX) (Fig. 2).
Experimental method Repeated trials using the four different shims were performed for each of the excised larynges. First, voJournal of Voice, VoL 12, No. 3. 1998
cal fold tension at the midmembranous vocal fold and vocal process was measured prior to the creation of the thyroid window (control state). Next, a window was created in the thyroid cartilage as described above. Vocal fold tension measurements were recorded prior to shim insertion (second control state). Next, shims were inserted between the inner perichondrium and the thyroid cartilage rim through the created window. Vocal fold tension measurements were then recorded for each successive shim (four in all). This experimental procedure was repeated for seven canine larynges. The measurements for the vocal tension were performed in the following manner: The steel rod of the force sensor was oriented perpendicular to the vocal fold and positioned so as to be just touching the affected vocal fold. The midpoint of the steel rod was aligned with the vocal fold edge. The steel rod was then laterally moved against the vocal fold via the micrometer translation stage which moved in 1-mm increments over a 5-mm interval. The resultant vocal fold tension was continuously measured (via the load cell) as a function of time over the entire 5-mm excursion at a rate of 10 samples per sec. The measurement procedure was carried out at the midmembranous vocal fold and vocal process with force sensors #1 and #2, respectively. Because the force sensors were subject to very slight flexion as the micrometer moved the force sensors against the vocal fold, the vocal folds were not lateralized in exactly 1-mm increments. The spring constants for both force sensors #1 and #2 were found to be constant.
Glottal configuration analysis A Panasonic solid-state video camera was mounted above the larynx in each experiment to document the effects of the implanted shim on the vocal fold configuration. A personal computer with a frame grabber was used to digitize these top view images from each canine experiment for each shim implant. A software program (Optimas Bioscan 5.1, Seattle, WA) was then used to measure the length and degree of medialization of the affected and opposite (control) vocal folds. Length of the membranous vocal fold was measured by first marking the anterior commissure, the affected vocal process, and control vocal process.
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
375
I Camera I
/
\ Vocal Cord
Vocal Fold Tension Sensor and Micrometer I Video Digitizer
Cartilage I
I
-
I
II
I1
These markers gave each point "x,y" coordinates m the axial plane of the glottis. Length could then be calculated from these "x,y" coordinates using the Pythagorean theorem. From the digitized images, degree of vocal fold medialization was determined using an angle technique (Fig. 3). Degree of vocal fold medialization was measured for the midmembranous vocal fold and the vocal process. Zero percent medialization was defined as no movement, whereas 100% medialization was defined movement of the vocal fold exactly to the midline. Again, specific points on the image were marked and thus assigned "x,y" coordinates: the anterior and posterior commissure (AC and PC), the affected and control midmembranous vocal fold edge (AMVF and CMVF), and the affected and control vocal process (AVP and CVP). For each image, four angles were calculated: 1. Angle 2. Angle 3. Angle 4. Angle
Aff.MVF: Con.MVF" Aff.ve: Con.vP:
Data Acquisition Computer
FIG. 2. Schematic o f medialization laryngoplasty b i o m e • chanics experiment. (Adapted from Ingo R. Titze. Principles o f Voice Production. Copyright 1994 by Allyn and Bacon. Reprinted by permission.)
AMVF-AC-PC CMVF-AC-PC AVP-AC-PC CVP-AC-PC
Degree of medialization could then be calculated using these equations:
Degree of med. for the AMVF = (Angle con.MV~ -- Angle Aft.MVF) Angle con.MVF× 100% Degree of med. for the AVP = (Angle co.. vP - Angle An.VP) Angle co.. vP × 100%
Data analysis The raw data was filtered with a low pass Bessel filter software program to reduce the noise level on the signal. The filtered data, vocal fold tension as a function of time, consisted of a series of increasing peaks which exponentially decayed to a steady state value. These steady state values corresponded with the vocal fold tension at a given force sensor deflection distance. These values were measured and then plotted as a function of implant size. Statistical methods Means and standard errors were calculated from the data collected from the seven canine larynx experiments. The Kruskal-Wallis test (nonparametric one-way ANOVA [analysis of variance]) was used to compare vocal fold length and degree of medialization for the various shim implants. The Friedman test (nonparametric two-way ANOVA) was used to compare vocal fold tension for the various shim implants. Journal of Voice, Vol. 12, No. 3, 1998
376
J. PIETER NOORDZIJ ET AL.
Mid-Mem. Vocal Fold AC
AMVF
i
\ ngle Con. •
Angle Aft.
Vocal Process
Angle Con.
When either the Kruskal-Wallis test or Friedman test was significant, the Wilcoxon signed rank test was used to make selective clinically relevant pairwise comparisons (Tables 1 and 2). A resulting p (probability) level of 0.05 or less was defined to indicate a statistically significant difference between the compared measurements. RESULTS Visual inspection of the larynges after insertion of each shim implant showed the following (Fig. 4): All shim implants tested increased the general thickness of the affected membranous vocal fold without a noticeable change in vocal fold length. The resting position of the affected vocal fold was more medialized with the larger shim implants compared to the smaller shim implants. The firmness to palpation of the affected membranous vocal fold appeared to increase with the implantation of the larger shim implants. In contrast to the visual changes seen at the mid vocal fold, the implantation of shim implants resulted in no appreciable anatomical changes at the vocal process. The addition of larger shim implants failed to result in any noticeable medialization of the vocal process (compare Fig. 4-a with Fig. 4-b). The posterior glottal gap was not closed despite increasing implant size. Mean normalized membranous vocal fold length as a function of implant type is shown in Fig. 5. The affected vocal fold length was normalized by dividJournal of Voice, Vol. 12, No. 3, 1998
FIG. 3. Degree of medialization was calculated using an angle technique. First, the top view image of the each larynx was marked with six necessary points. Next, the angles of medialization for midmembranous vocal fold and vocal process were calculated. Using these angles, the degree of medialization could be determined.
Angle Aft.
ing it by the affected vocal fold length in the control state (no window created). Likewise, the control (opposite) vocal fold was normalized by dividing it by the control (opposite) vocal fold length in the control state (no window created). This figure shows that the length of both the affected and control (opposite) membranous vocal folds did not significantly change with the addition of any shim type as evidenced by Kruskal-Wallis test (p equalled 0.75 and 0.60, respectively). Vocal fold medialization as a function of implant type is shown in Fig. 6. Medialization of the affected vocal fold was measured at the midmembranous vocal fold and at the vocal process. Zero percent medialization is defined as no movement toward the midline, whereas 100% medialization is defined as movement of the vocal fold exactly to the midline. For the midmembranous vocal fold, there was a significant trend of greater medialization with larger implants (Kruskal-Wallis test, p = 0.0003). Each shim implant resulted in a significant amount of midmembranous vocal fold medialization compared to the control (see Table 1). The greatest amount of medialization of the midmembranous vocal fold occurred with the second largest shim, Shim C (43% medialization). At the vocal process, there was a trend toward greater medialization with larger implants. However the amount of medialization was relatively small and not statistically significant (Kruskal-Wallis test, p = 0.23). The greatest amount
THE BIOMECHANICS OF 7*I-IEMEDIALIZATION LARYNGOPLASTY T A B L E 1. Degreeof midmembranous vocal fold medialization comparisons: control vs. various shim implants using the Wilcoxon signed rank test y Implant Type
T A B L E 2. Midmembranous vocal fold tension comparisons: control vs. various shim implants (A 3 mm force sensor deflection distance) using the Wilcoxon signed rank test y
p Value
Control vs. Window
377
Implant Type
p Value
0.08
Control vs. Shim A
0.03
Control vs. Window
0.24
Control vs. Shim B
0.02
Control vs. Shim A
0.87
Control vs. Shim C
0.02
Control vs. Shim B
0.61
Control vs. Shim D
0.03
Control vs. Shim C
0.03
Control vs. Shim D
0.03
-Significant p values were defined as less than 0.05.
tSignificant p values were defined as less than 0.05.
t. ,,', i<-! ,,.
P
4
-e.~
$
1. No Implant (control)
2. Shim B (small)
3. Shim D (large)
FIG. 4. Frame #1 shows the canine larynx with no shim implant (control), frame #2 shows the canine larynx with Shim B (small implant) in place, and frame #3 shows the canine larynx with Shim D (large implant) in place.
of medialization of the vocal process occurred with shim C (15% medialized). Fig. 7 shows mean midmembranous vocal fold tension as a function of shim type. This figure shows the following: tension at the midmembranous vocal fold increased with the addition of larger shim implants. These changes were significant (Friedman test, p = 0.002). The greatest amount of tension was seen with the implantation of the largest shim. Shim D resulted in 41% greater resistance to lateralization as compared to the control state at the 3 mm deflection dis-
tance. Only shims C and D resulted in significantly greater midmembranous vocal fold tension than in controls (see Table 2). Fig. 8 shows the mean vocal fold tension at the vocal process as a function of shim type. This figure shows that vocal fold tension at the vocal process was not significantly affected by shim implant size (Friedman test, p = 0.31). In summary, vocal fold length was not affected by implant size. Larger implants medialized the mid= membranous vocal fold, more than smaller ones, but Journal of Voice, Vol. 12, No. 3, 1998
378
J. P I E T E R N O O R D Z I J E T AL.
Mean Normalized Vocal Fold Length v. Implant Type 1.01 .=
1.00
d)=l
¢ 0.99 ,.J •o 0.98 o
It. - - 0.97
- - Affected Vocal Fold
o
o
0.96
.N 0.95 m
E 0.94 O
Z
Control Vocal Fold
0.93
=} 0.92 0.91 0.90 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 5. Mean normalized vocal fold length plotted vs. implant type with standard error bars included.
Vocal Fold Medialization v. Implant Type 100 90 80 70 A
60 Midmembranous Vocal Fold
O 50
mm
.N 40
±
30
m
~- 20 I,I. > 10
T
m
m m
m
l_ ,,
Window Only
Shim A
Shim B
m I ml
Shim C
-10
Implant Type FIG. 6. Mean vocal fold mediali~,ation plotted vs. implant type with standard error bars included.
Journal ofVoice, Vol. 12, No. 3, 1998
Shim D
Vocal Process
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
379
Mean Mid-Membranous Vocal Fold Tension v. Implant Type 25.0
Deflection Distance
g
20.0
|o "o
131 mm
15.0
0 ~L ¢U 0 0
>
mm
1~2 mm [ ] 3 mm
10.0
.Q
134 mm
E
[ ] 5 mm
5.0
0.0 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 7. Mean midmembranous vocal fold tension plotted vs. implant type with standard error bars included.
Mean Vocal Fold Tension at Vocal Process v. Implant Type 60.0 A v
=
50.0
~"
Deflection Distance
2 a.
40.0
I0
m ° 0
1~2 mm
30.0 c
i1~!3 mm B 4 mm
o i.~ O It.
mm
[31 mm
20.0
---
10.0
[ ] 5 mm
-
o
O >
0.0 Control
Window Only
Shim A
Shim B
Shim C
Shim D
Implant Type FIG. 8. Mean vocal fold tension at the vocal process plotted vs. implant type with standard error bars included. Journal of Voice, Vol. 12, No. 3, 1998
380
J. PIETER NOORDZIJ ET AL.
did not medialize the vocal process. Thus the posterior glottal gap was not significantly closed by the ML procedure. Vocal fold tension increased with larger implants at the midmembranous vocal fold, but not at the vocal process. DISCUSSION The evaluation and treatment of glottal incompetence due to UVFP, vocal fold atrophy, and trauma have seen much change in the last 30 years. Many advances have taken place in the treatment of this disease. The goals of treatment in these patients are to reduce aspiration and restore phonatory function to nearly normal level. A variety of treatment modalities have been developed that restore phonatory function to near normal. The current treatment options are voice therapy, injection laryngoplasty, laryngoplastic phonosurgery, or laryngeal reinnervation. Voice analysis and assessment by a speech-language pathologist specializing in voice are useful in nearly all patients with dysphonia (16). Very often though, speech therapy alone will not restore phonatory function to a satisfactory level. Once surgery becomes a desirable option, there are a variety of procedures available. The training and experience of the surgeon, as well as patient factors (age and other significant medical problems) play a major role. The appearance of the larynx on videostrobolaryngoscopy is also an important factor in choosing a specific surgical technique. A better understanding of the biomechanics of the larynx before and after surgery would aid the otolaryngologist in choosing a surgery or combination of surgeries for the patient with glottal incompetence. Injection laryngoplasty can be used as the primary medialization therapy to correct glottic incompetence or as an adjunct to enhance medialization results. A variety of substances have been used, including Teflon, gelfoam, collagen, and fat. This technique is the method of choice in patients with incurable malignancies or whose medical status makes them high risk for surgery (5). Although simple to perform, injection techniques have major disadvantages in being irreversible and difficult to tailor. Teflon granuloma and Teflon overinjection can result in deterioration of voice which is difficult to correct (17). Journal of Voice, Vol. 12, No. 3, 1998
Laryngoplastic phonosurgery, including ML, allows for a more precise treatment of glottic incompetence. Popularized by Isshiki, the ML (originally termed thyroplasty type I ) is the most commonly performed laryngoplasty procedure in the setting of UVFP. This surgical technique has been refined by others (8). Technically, the procedure is precise and straightforward in concept. However, the ML is not without its difficulties. Complications of this procedure include implant extrusion, airway distress, and poor voice result. Objective voice results from ML have not been reported in sufficient detail to compare to other procedures. However, many authors, including Isshiki (2), have noted the inability of the medialization procedure to close a posterior glottal gap. Ideal patients to be treated for ML are those with preoperative findings of a mid-cord gap or a bowed vocal fold. When a large posterior glottal gap is present, other surgical procedures should be considered. Although subjective voice improvements in patients with UVFP who undergo phonosurgery are important, a more basic understanding of the physical phenomenon that occurs due to phonosurgery is essential. Optimizing the biomechanics (configuration and tension) of the paralyzed vocal fold should result in improved phonatory ability (3). Green, Berke, and Ward (18) studied the differences in vocal efficiency (acoustic power/subglottic power) between AA surgery and ML in an in vivo canine model and found that AA was superior to ML in vocal efficiency, traveling wave motion, and acoustics. Yet the actual biomechanical effects of ML on the larynx remain to be examined. The present study uncovered several biomechanical properties of ML in the excised canine model. ML does not affect vocal fold length; it medializes and significantly stiffens the midmembranous vocal fold. It would appear that this enhanced midmembranous stiffness is due to medial compression of the vocal fold by the shim implant. Although the vocal processes became slightly medialized with ML, this did not result in significant alterations in glottal dimensions. Additionally, ML does not stiffen the arytenoid cartilage so as to resist lateral movement. ML appears to be efficient in medializing the anterior membranous vocal fold, but is a poor substitute for closure of the posterior glottis and an inadequate substitute for the arytenoid adduction procedure.
THE BIOMECHANICS OF THE MEDIALIZATION LARYNGOPLASTY
This study demonstrates that the phonatory larynx can be divided into two biomechanical subunits when considering surgical rehabilitation for glottal incompetence. These anatomic biomechanical subunits are conveniently divided into anterior and posterior subunits. The anterior subunit is comprised of the membranous vocal fold. The posterior subunit is comprised of the arytenoid cartilage and the intrinsic muscles of the larynx that insert on the arytenoid cartilage. There is also clinical relevance to this subdivision of the larynx. In patients with UVFP, the glottis may have a variety of anterior and/or posterior configurational changes. In patients with UVFP or presbylaryngis, the membranous vocal fold may have variable amounts of bowing and atrophy. Similarly, sulcus vocalis may present with atrophy of the membranous vocal fold. UVFP may result in shortening of the vocal folds to varying extents. Vocal fold shortening leads to compression of the contralateral vocal fold (in a compensatory effort to approximate the vocal processes during phonation), which in turns results in increased glottal incompetence (12). In patients with UVFP, the size of the posterior glottal gap and the position of the arytenoid cartilage can be variable as well (12). In the case of deficient vocal fold tissue secondary to intubation trauma, only the posterior subunit is involved. In some patients with high vagal nerve injury, a mid-cord gap, anterior arytenoid rotation, and a posterior gap are present, thus involving both subunits of the glottis. We feel surgical procedures designed to correct glottal incompetence should be individualized to correct the defect within each subunit or be combined to address all the affected subunits. Within the anterior subunit of a larynx affected by glottal incompetence, medializing, lengthening, and stiffening of the membranous vocal fold may be beneficial. These changes will help restore oscillation of the vocal fold with a mucosal wave (body cover theory). In patients with a mid-cord gap secondary to vocal fold atrophy, ML should effectively medialize and stiffen the membranous vocal fold. ML is an excellent choice of surgery to correct a gap and lack of stiffness within the anterior laryngeal subunit. However, ML does not appear to increase vocal fold length, which would be beneficial to some patients with UVFP. In cases of sulcus vocalis, where stiffness around the sulcus is the dominant feature, ML
381
may exaggerate already poor vibratory characteristics. Similarly, in cases of trauma or surgery where the dominant feature is excess scar formation and increased vocal fold stiffness, ML may not be beneficial in improving voice. Within the posterior subunit of a larynx affected by glottal incompetence, the position and tension of vocal process of the arytenoid cartilage are very important in restoring function. A midline vocal process which resists lateralizing movement from the opposite vocal fold should achieve better glottic closure and valving during phonation. A better valve will resist airflow leakage and allow the larynx to serve as a more efficient airflow oscillator. A more efficient valve should allow for a better build-up of subglottic pressure resulting in increased acoustic power. This should translate into more efficient vocalization with better loudness. In this important component of arytenoid function, ML appears to be ineffective. Other surgical therapies, such as reinnervation or AA should be considered if arytenoid position rehabilitation is desired. Our experimental findings are coorborated by clinical observations. Other authors have reported AA as an alternative to ML (19). ML and AA have been used in conjunction to rehabilitate UVFP (13). The clinical impression of the posterior glottal gap being difficult to rehabilitate by ML is supported by findings from this study (13). Furthermore, the foreshortened VF with arytenoid rotation also appears to be difficult to rehabilitate with ML procedure alone (12). CONCLUSIONS When properly done, the ML procedure appears to medialize and the stiffen the midmembranous fold. Thus, the ML appears to be a good choice of surgery to correct bowing or augment atrophic vocal folds due to UVFP, presbylaryngis, or suicus vocalis. However, ML does not appear to significantly lengthen the vocal fold and does not significantly medialize or stiffen the arytenoid cartilage. ML cannot be used to effectively close a posterior glottal gap. In this case, AA should be considered, as it has been shown to medialize and stiffen the arytenoid cartilage. We conclude that it is important to individually consider the anterior and posterior biomechanical subunits of the larynx when selecting a surgical therapy for the Jounzal of Voice, Vol. 12, No. 3, 1998
382
J. PIETER NOORDZIJ ET AL.
rehabilitation of patients with UVFP or other forms of glottal incompetence. REFERENCES 1. Rubin JS, Sataloff RT, Korovin GS, Gould WJ. Diagnosis and treatment of voice disorders. New York: Igaku-Shoin Ltd, 1995:384-5. 2. Isshiki N. Phono-surger3,. Japan: Springer-Verlag, 1989: 82-95. 3. Hirano M. Surgical alteration of voice quality. In: Cummings CW, ed. Otolar3'ngology----Head and Neck SurgepT, Update I. St. Louis: CV Mosby Co, 1989:249-62. 4. Arnold GE. Vocal rehabilitation of paralytic dysphonia. Arch OtolaiTngol 1962;76:358-68. 5. Benninger MS, Crumley RL, Ford CN, Gould WJ, Hanson DG, Ossoff RH, Sataloff RT. Evaluation and treatment of the unilateral paralyzed vocal fold. OtolaD,ngol--Head Neck Surg 1994; I I 1:497-508. 6. Isshiki N, Taira T, Kojima H, Shoji K. Recent modifications in thyroplasty type I. Ann Otol Rhfllol Laiyngol 1989;98: 777-9. 7. Sasaki CT, Leder SB, Petcu L. Friedman CD. Longitudinal voice quality changes following Isshiki thyroplasty type I: The yale experience. Laryngoscope 1990; 100:849-52. 8. Koufman JA. Laryngoplasty for vocal cord medialization: an alternative to teflon. Laryngoscope 1989;96:726-31. 9. Maves MD, McCabe BE Gray S. Phonosurgery: indications and pitfalls. Ann Otol Rhinol Lal3,ngol 1989;98:577-80.
Journal of Voice, VoL 12, No. 3, 1998
10. Isaacson G, Kim JH, Kirchner JC, Kirchner JA. Histology of Isshiki thyroplasty type 1. Ann Oto! Rhh~ol l_xwyngol 1990; 99:42-5. 11. Leder SB, Sasaki CT. Long-term changes in vocal quality following Isshiki thyroplasty type I. La~3,ngoscope 1994: 104:275-7. 12. Woodson GE. Configuration of the glottis in laryngeal paralysis I: clinical study. Laryngoscope 1993;103:1227-34. 13. Netterville JL, Stone RE, Luken, ES, Civantos FJ, Ossoff RH. Silastic medialization and arytenoid adduction: the Vanderbilt experience. Ann Otol Rhinol Lan3,ngol 1993;102: 413-24. 14. Haji T, Mori K. Omori K, lsshiki N. Mechanical properties of the vocal tbld. Stress-strain studies. Acta Otolatyngol (Stockh) 1992; I 12:559-65. 15. Tran QT, Berke GS, Gerratt BR, Kreiman J. Measurement of Young's modulus in the in vivo human vocal folds. Ann Otol Rhinol Lal3,ngol 1993;102:584-91. 16. Sataloff RT. Professional voice: The science and art of clinical care. New York: Raven Press, 199 I. 17. Crumley RL. Teflon versus thyroplasty versus nerve transfer: a comparison. Ann Otol Rhhlol Laryngol 1990;99:759-63. 18. Green DC, Berke GS, Ward PH. Vocal fold medialization by surgical augmentation versus arytenoid adduction in the in vivo canine model. Ann Otol Rhinol l.xu3,ngol 1991; 100: 280-7. 19. Isshiki N, Tanabe M, Sawada M. Arytenoid adduction for unilateral vocal cord paralysis. Arch OtolaJyngol 1978;104: 555-8.