E4

.7 v ‘ ‘4 my
.

'I-fv

5f

‘ a, ,«g
—'.‘§-~.~IJ us

.x‘ :3: an»;
‘ “.3“:

‘L

«9:»:

n ‘ I‘
,.:....m‘:.. :cémx
a. N1. mum” Wm .
1—”;
'x‘".‘.':u‘.
.

mam
{£33393} 1

53." if»?!

.‘I

< Lem:

f. gémrzm
M
1%:

a ‘f
l "C

“3::

 

m
"umprcx'n
‘ 1‘ <~ ~..

y 1”"
L.J.a
J“ Li’

)4 ‘ 4'5"
'.
a“ Rh

.7.»

. ’._
V .
. \‘fi ‘

‘ . at:
a _ ' ‘- ‘3,
F . .

”.1

”a“. 3.
1 "fl“:

.- . . “.r -fifi'avfi- gxé’
M r "’mm vs: 33 7‘?" w” =5»
7“ ‘Eggzgéziif‘l '~ ' Tm‘Y-fi‘ ;A£‘~:£‘I$§(g£‘
4. >1 '- .W“: - ~ W ‘ ” '
. figxflgfi a? was?
at: ,2" ‘ mug '4' A
n'A
I.
It:

5?
13:73.
11.51:: .

‘wk
23::

 

(ӣ85

IHIWIW\||\W“dHliIHWth"fllHfllHiHlWll

3 1293 010251

This is to certify that the

dissertation entitled

Objective Pressure Measurements in Two

Groups of Laryngectomees

presented by

Jan Susan Lewin

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in Audiology and

 

 

Speech Sciences

 

 

 

   

Ejor professor

Datégfig: W

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE II REI’URN BOX to romovo this checkout from your record.

TO AVOID FINES return on or baton onto duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU lo An Affirmative Action/Equal Opportunity Inflation
mun-on

m1

 

OBJECTIVE PRESSURE MEASUREMENTS IN TWO GROUPS OF
LARYNGECTOMEES

BY

Jan Susan Lewin

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the
requirements for the degree of

DOCTOR OF PHILOSOPHY

Department of Audiology and Speech Sciences

1994

ABSTRACT

OBJECTIVE PRESSURE MEASUREMENTS IN TWO GROUPS OF
LARYNGECTOMEES

BY

Jan Susan Lewin

In order to establish the intraluminal pressures associated
with tracheojejunal (TJ) speech production, objective air
insufflation was performed transtracheally in 4 subjects
with jejunal interposition and TJ puncture following
laryngopharyngectomy during sustained vowel production and
counting. Peak pressure measurements were compared to the
intraesophageal peak pressures of 9 laryngectomees with

tracheoesophageal (TE) puncture who served as controls.

TJ subjects demonstrated significantly greater pressure than
TE subjects during sustained vowel production. No
significant pressure difference was found between groups
during counting; however, experimental subjects did

demonstrate higher pressure than controls.

Decreased pressure was associated with longer lengths of
time from surgery. No significant differences were found
based on gender. Stronger correlations for females
demonstrated decreased pressure associated with longer time
since surgery, whereas males showed less pressure

variability during counting.

A significant difference in pressure variability between
groups was found during sustained vowel production. Both
groups demonstrated greater pressure variability during
counting than during sustained vowel production. Higher
variances and small sample size may have prevented findings

of statistical significance.

A power analysis of the data projected a minimum requirement
of 12 subjects per group in order to detect significance
with an 80% chance of accuracy. Given the high mortality
and morbidity rates and the rare incidence of carcinoma of
the laryngopharynx requiring jejunal interposition,
longitudinal study with retrospective analysis appears more

practical for ongoing study in this population.

None of the experimental subjects used the voice prosthesis
for communication. All of the controls used the voice
prosthesis to speak. While the results of this study do not
provide conclusive evidence to indicate increased pressure
as the sole reason for failure of TJ speakers to use the
voice prosthesis for communication, increased pressure was a
limiting factor. A discussion of possible influencing
factors is provided. Further questions need to be answered
in order to determine the efficacy of TJ puncture and voice
restoration as a viable alaryngeal alternative for speech
production in patients with jejunal interposition following

laryngopharyngectomy.

Copyright by
Jan Susan Lewin
1994

To my mom and dad, Rosalyn and Melvin Lewin,
my best friends and wisest teachers

ACKNOWLEDGEMENTS

There is no doubt that many long hours have been spent in
the preparation and completion of this manuscript. However,
to suggest that this dissertation belongs only to me would
not be true nor would it be fair. There are several people
who shared in the work and the stress to whom I am truly

grateful.

I am indebted to the Chair of my committee, Dr. Peter
LaPine, who quietly and patiently guided me through the
challenges and obstacles one encounters during the pursuit
of a doctoral degree. His commitment to my education will

never be forgotten.

I would like to further acknowledge the members of my
committee -- Drs. Paul Cooke, David Daly, and Leo Deal --
who willingly agreed to accept me as their student and gave
freely of their time and knowledge. They allowed me the
freedom to pursue my interests through independent projects
while demanding excellence in the quality of the results. I

am truly grateful to them for their support of my endeavors.

I would like to thank my colleague and friend, Mr. Marc

Haxer who covered the clinic, worked in the extra patients,

vi

ran to the library to pick up an article, and postponed his
vacations when the demands of classes and writing became

overwhelming. Without his help, I am not sure I would have
been able to meet the deadlines and fulfill the requirements

in a timely manner. He remains a true friend.

This manuscript would never have been possible without the
continued help of Ms. Yvonne Beerens whose mastery of
computer word processing transformed the thoughts and
handwritten notes into a professionally produced manuscript.
Her expedient assistance and calm management of never-ending

dilemmas will always be appreciated.

I would also like to acknowledge Mr. James Bruce for
transforming my "stick-figure" scrawlings into artistically
drawn figures in a moment's notice. He remains responsible
for the graphic precision of this document. He is a fine

artist and a true friend.

To my family, my mother and father, my sister, Carol, and my
friends who never failed me when I needed them, I will be
eternally grateful. There is no way to measure or to repay
the hours of listening, the perpetual support, and the
undying faith in me they continued to show. I wish that my
grandparents could be here to see the completion of this

work because I know that they too would be proud.

Finally, I would like to thank the faculty of the Department

of 0tolaryngology at the University of Michigan Medical
vii

Center for allowing me the time to complete my doctorate and
supporting my research endeavors. Most importantly, I
respectfully acknowledge all of the laryngectomees who
agreed to participate in this investigation, for without

them this research could not have been done.

viii

TABLE OF CONTENTS

LIST OF TABLES . . . . . . . . . . . . . . . . . . .

LIST OF FIGURES . . . . . . . . . . . . . . . . . . .

CHAPTER I. INTRODUCTION . . . . . . . . . . . . . .

CHAPTER II. METHODS . . . . . . . . . . . . . . . .

subjects 0 O O O O O O O I O O O O O O O O O O 0
Surgical Technique . . . . . . . . . . . . . . .
Assessment . . . . . . . . . . . . . . . . . . .

CHAPTER III. RESULTS . . . . . . . . . . . . . . . .

Subject Characteristics . . . . . . . . . . . .
Correlation Analysis by Subject Characteristics
Power Analysis . . . . . . . . . . . . . . . . .
Correlation Analysis by Gender . . . . . . . . .
Peak Pressure Measurement Analysis . . . . . . .
Power Analysis . . . . . . . . . . . . . . . . .

CHAPTER IV. DISCUSSION . . . . . . . . . . . . . . .

CHAPTER V. CONCLUSIONS AND FUTURE DIRECTIONS FOR

RESEARCH O O O O O O O O O O O O O O I O O
summary 0 O O O O I O O O O O O O O O O O O O 0
Future Research . . . . . . . . . . . . . . . .

APPENDICES

A. Objective Air Insufflation Form . . . . . . .

B. Michigan State University Informed Consent
for Experimental Procedure . . . . . . . . .

C. The University of Michigan Medical Center
Informed Consent for Experimental Procedure.

LIST OF REFERENCES . . . . . . . . . . . . . . . . .

ix

.30
31
31
33

37
.37
.43

45

46

48

54

56

67

.67

.69

72

.73

.75

.77

1a.
1b.

2a.

2b.

4a.

4b.

5a.

5b.

LIST OF TABLES

- Subject Characteristics - Control Group .

Subject Characteristics - Experimental Group .

Central Tendency and Dispersion Data - Control

Group ; n = 9 O O O O O O O O O O O O O 0

Central Tendency and Dispersion Data -
Experimental Group; n = 4 . . . . . . . .

Correlation Matrix Demonstrating the

Relationship Between Subject Characteristics

and Task Performance for All Subjects . .

Correlation Matrix Demonstrating Relationship

Between Subject Characteristics and Task
Performance for Female Subjects . . . . .

Correlation Matrix Demonstrating Relationship

Between Subject Characteristics and Task
Performance for Male Subjects . . . . . .

Individual Peak Pressure Recordings in mmHg for

/a/ and Counting - Control Group . . . . .

Peak Pressure Recordings in mmHg for /a/ and

Counting - Experimental Group . . . . . .

Mann-Whitney Test for Tasks and Variability in

Task Performance Between Groups . . . . .

Two-sample t-test Analysis of Intraluminal

Pressures Associated with Task Performance .

Power Analysis for Counting - Independent
Groups and One-Tail Test for Unequal ns .

.38

40

.41

42

.44

.47

.48

49

.50

.51

53

.55

LIST OF FIGURES

Diagram of Testing Apparatus Demonstrating
Catheter Placement, Attachment to Pressure
Recorder, and Compressed Airflow Source . . . . . . .34

Diagram of Testing Apparatus Demonstrating

Catheter Placement, Attachment to Pressure

Recorder, and Insufflation Source for

Preoperative Prediction of Postoperative TE

Speech . . . . . . . . . . . . . . . . . . . . . . . 63

xi

CHAPTER I

Introduction

Surgical/prosthetic restoration of speech dates back to the
time of the first laryngectomy performed by Billroth in
1874. The early technique of laryngectomy was to create a
fistula in the gullet through the neck. Gussenbauer, a
resident under Billroth, developed a tubular device with a
vibrating reed, much like the present day fenestrated
tracheostomy tube, which was inserted into the hypopharynx
and trachea, respectively. An obturator was inserted to
block the pharyngeal extension during deglutition while a
plug was inserted for pharyngo-tracheal respiration. The
patient inserted a "flutter valve" (Montgomery & Toohill,
1968, p. 500) for speech. The device not only prevented
aspiration but also restored the patient's voice. As the
surgical technique for laryngectomy changed so, too, did the
design of the prosthetic devices used for vocal
rehabilitation. The vocal prostheses in general became more‘
complicated and harder to manage with variable patient and
physician satisfaction (Panje, 1981; Montgomery & Toohill,

1968).

A review by Holinger (1975) suggests that surgical
procedures for the removal of the larynx initially involved

the creation of both a laryngostome into the patient's

trachea as well as a pharyngostome into the pharynx.

Primary closure of the pharynx following laryngectomy as
pioneered by Gluck and Sorensen in 1894 eliminated the need
for the external pharyngostome, but it also eliminated the
use of the prosthetic devices which depended on both
openings to shunt air from the trachea into the pharynx for
alaryngeal speech production. Following the introduction of
primary closure and the elimination of the prosthetic shunts
for speech production, patients learned that vibration of
the esophageal walls could be used to produce voice. Soon
esophageal speech production became the chief method for
alaryngeal voice restoration for the laryngectomized
patient. It has continued to remain a viable alternative

for alaryngeal speech production to the present day.

Speech rehabilitation of the laryngectomee remained
basically unchanged until 1952 when Briana proposed the
creation of a controlled pharyngocutaneous fistula into the
pharyngoesophagus with the use of a rubber tube to join the
trachea and pharynx so that air could again be shunted into
the esophagus for voice production. The procedure was
complicated by aspiration of food and secretions. In an
attempt to reduce the contamination from the esophagus to
the trachea, Conley (1958) and colleagues (Conley, DeAmesti,
& Pierce, 1959) offered two methods for shunting air from
the trachea into the esophagus. The first method involved

the creation of a mucosal tunnel taken from the lining of

the esophagus that was looped over the omohyoid muscle.

Upon swallowing, the tunnel was compressed by muscular
contraction and depression on the tunnel. The procedure was
complicated by stenosis of the tract and aspiration. It
also had limited use in irradiated patients. In 1959,
Conley switched to the use of an autogenous vein graft to
replace the tubed muscular tunnel; but again the usefulness
of the procedure was limited by the same complications:

stenosis and aspiration.

Asai, in 1965, modified Conley's approach and offered a
three-stage laryngoplasty in which a dermal tube connected a
pharyngostome and tracheostome for internal shunting of
pulmonary air for voice production. The procedure was
introduced to the United States in 1967 by Miller.
Unfortunately, both procedures continued to be complicated
by aspiration and stenosis of the dermal tube. Various
modifications of both procedures were offered by others.
Again, all were limited by the same problems, aspiration and
stenosis (Karlan, 1968; Montgomery & Toohill, 1968; McGrail
& Oldfield, 1971; Calcaterra & Jafek, 1971; Komorn, Weyeer,

Sisson & Malone, 1973; Amatsu, 1980).

Continued interest in laryngeal surgery with primary voice
restoration sparked other investigators to attempt new
procedures. Staffieri (1969) and Arslan and Serafini (1972)
pioneered the neoglottic procedure which created a primitive

valved glottis with a flap of mucosal remnant. The

4

procedure was actually a tracheo-hypopharyngeal shunt in
which the trachea was transected at the subcricoid level
with the hyoid bone left in place. A flap of posterior
cricoid mucosa was created. The flap was placed over the
top of the trachea and the trachea was slanted at the top to
avoid aspiration. Arslan and Serafini (1972) modified the
technique by preserving a remnant of the epiglottis and
suturing it to the anterior edge of the top of the trachea.
The procedure was limited by the extent of cancer.
Complications also included aspiration, limited usefulness
in irradiated patients, need for revision surgery, and
failure to develop voice. Sisson, Bytell, Becker, McConnel,
and Singer (1978) introduced Staffieri's technique to the
United States. They reported their experience with 12
cases. Although strict surgical criteria were advocated for
the procedure in order to ensure complete tumor removal,
three patients developed recurrent tumor and only 6 achieved
permanent vocalization. Leipzig, Griffiths, and Shea (1980)
reported 4 of 30 patients (13%) recurred in the neck and 12
of the 30 patients (40%) aspirated, half of which
necessitated surgical closure of the neoglottis. Attempts
at neoglottic reconstruction continued, although the

usefulness of these procedures was limited.

The high rate of failure of the shunt and neoglottic
procedures prompted a renewed interest in the use of

prosthetic devices for voice production. In 1972, Taub and

5

Spiro introduced their removable air bypass prosthesis, the
"VoiceBak," a valved bypass cannula that allowed air from
the tracheostome into the esophagus for voice production
through an esophagotomy. While the device did restore
speech to some, it was cumbersome, expensive, and had
limited use in radiated patients or patients with neck
dissection because of the potential injury to the carotid
artery (Henley & Souliere, 1986). Sisson, McConnel,
Logemann, and Yeh (1975) reported 26 cases in which two
deaths occurred from carotid rupture. They reported that
the main problem with the device was the development of a

salivary fistula.

Tracheoesophageal (TE) puncture and voice restoration
following total laryngectomy, as described by Singer and
Blom in 1980, has probably become the most common method of
prosthetic alaryngeal speech restoration. The method
involves the surgical creation of a controlled fistula which
forms a communication between the trachea and esophagus. A
small, one-way valved prosthesis stents the puncture tract
and allows pulmonary airflow into the esophagus for voice
production but prevents aspiration of food and saliva. The
TE puncture as originally described was performed as a
secondary procedure following laryngectomy. It was done
under general anesthesia and required a short
hospitalization (Singer & Blom, 1980; Maniglia, Lundy,

Casiano, & Swim, 1989). However, the procedure can also be

6

performed as a primary procedure at the time of laryngectomy
immediately after laryngeal resection (Lau, Wei, Ho, & Lam,
1988; Maniglia, 1982; Maves & Lingeman, 1982; Hamaker,

Singer, Blom, & Daniels, 1985).

The vibration of the pharyngoesophageal mucosa functions as
a substitute sound generator for the larynx. TE puncture is
reportedly simple with successfully restored speech
production between 50 percent and 97 percent (Donegan,
Gluckman, & Singh, 1981; Juarbe, Shemen, Eberle, Klatsky, &
Fox, 1986; Juarbe et a1., 1989; Panje, 1981). In those
cases in which TE speech production has not been achieved,
pharyngeal constrictor myotomy, pharyngeal constrictor
myectomy, or pharyngeal plexus neurectomy has been performed
to facilitate conversational speech production (Baugh,

Lewin, & Baker, 1987).

Historically, interest has continued in identifying those
patients who would be good candidates for surgical and
prosthetic voice restoration prior to surgical intervention.
Various investigators have introduced air into the esophagus
prior to surgical and prosthetic intervention in an attempt
to determine the vibratory response of the esophagus to air
insufflation. Air-blowing, or insufflation, has been used
to determine the location of esophageal vibration for the
production of esophageal sound for speech production. It
has also been used to assess possible obstructions to the

flow of air that might impede or prevent esophageal

7

vibration within the cervical esophagus. These obstructions

are often not visualized during a routine barium swallow.

Reports have suggested that speech production resulting from
preoperative air insufflation of the esophagus is a good
indicator of postoperative TE speech production ability
(Blom, Singer, & Hamaker, 1985; Singer & Blom, 1980; Singer
& Blom, 1981; Taub, 1975, 1981). VandenBerg and Moolenaar-
Bijl (1959) were the first, followed by Seeman (1967), to
use objective measurements of intraesophageal pressures to
predict esophageal voice production. Taub (1975, 1981)
reintroduced esophageal insufflation to determine the
location of maximum sound production in the esophagus for
placement of his VoiceBak prosthesis. He later used it to
determine the patient's potential for successful speech
results using his device. He suggested that the
insufflation test could be used to determine the esophageal
speech potential for all laryngectomees. However, the
variations in esophageal insufflation including insufflation

force and sound duration were not addressed.

Singer and Blom (1980) included esophageal air insufflation
as part of their preoperative assessment for TB voice
restoration following total laryngectomy and again as part
of their self-insufflation test in 1985. Both methods
lacked clear operational definitions as to the level of
airflow and amount of air presentation as well as

definitions for postoperative TE speech success versus

8

failure. Its advocates, Singer and Blom (1980, 1981),
experienced both false positive and false negative results.
They also reported data with inconclusive results. Others
found esophageal air insufflation to be of limited
prognostic value (Panje, 1981; Donegan et a1., 1981; Wood,
Tucker, Rusner, & Levine, 1981; Johns & Cantrell, 1981).
The lack of objectivity and specificity of esophageal air
insufflation testing prompted suggestion of its abandonment

(Panje, 1981).

In 1987, Lewin, Baugh, and Baker introduced an objective
method of air insufflation using intraesophageal pressure
measurements to predict those patients who would acquire
fluent TE speech following TE puncture and those patients
who would require myotomy of the pharyngeal constrictor
musculature to achieve conversational fluency. Patients
were insufflated preoperatively, and their peak intra-
esophageal pressures were recorded. Following TE puncture,
three groups of TE speakers were identified and were found
to be statistically different (p < .001) based upon their
pressure measurements using the Kruskal-Wallis One-Way
Analysis of Variance by Ranks. Patients with low
intraesophageal peak pressure measurements (< 20 mm Hg)
obtained fluent TE speech without the need for myotomy.
Fluent speakers demonstrated abilities comparable to normal
laryngeal speakers (i.e., greater than 10 seconds of

uninterrupted vowel duration and a minimum production of

9

10-15 syllables per breath). Non-fluent speakers
demonstrated sound and word production abilities below the
fluent criterion. Non-speakers demonstrated sound durations
below one second and no word production capability. Both
the non-fluent speakers and non-speakers had preoperative

intraesophageal peak pressure measurements above 20 mm Hg.

An inability to sustain sound for ten seconds and/or produce
ten to 15 syllables was interpreted as being associated with
elevated pharyngeal constrictor muscle resistance. A
pharyngeal nerve plexus blockade of 1-3 cc of 1% lidocaine
injection into the paraesophageal and parapharyngeal tissues
was performed. Testing was repeated. A reduction in
intraesophageal pressure measurements with improved ability
to perform speech tasks clinically confirmed pharyngeal
constrictor hypertonicity and the need for myotomy to

achieve fluent speech post-TE puncture.

The authors therefore offered a method of assessment to
quantify and predict TE speech fluency as well as to
identify those patients in need of myotomy to achieve
conversational speech ease.‘ Since testing was performed
prior to TE puncture, a negative test result supported
myotomy at the time of the TE puncture rather than at a
later date after the patient experienced conversational
speech failure. Therefore, an additional surgery could be
avoided along with patient frustration associated with

further speech delay.

10

The majority of current documentation concerning surgical
and prosthetic methods of alaryngeal voice production
focuses on tracheoesophageal voice restoration following
simple laryngectomy. However, new reconstructive procedures
for radical surgery of the laryngopharynx and cervical
esophagus have resulted in a limited experience with vocal
rehabilitation of patients undergoing such extended
surgeries. Realistically the goal for rehabilitation has
been "to provide an adequate conduit to allow peroral
alimentation at a single operation synchronous to the
resection" (Coleman, 1993, p. 85). Success of the procedure
has been determined by the ability of the patient to swallow

following surgery (Shumrick & Savoury, 1988).

Malignant tumors arising in the hypopharynx and cervical
esophagus, or laryngopharynx, are associated with the worst
prognosis amongst neoplasia arising in the upper
aerodigestive tract. The overall incidence of cancer of the
hypopharynx in the United States is approximately 8 per 1
million population. The highest incidence occurs twice as
often in males usually in the seventh and eighth decades of

life (Shah, 1993).

Discussion of the surgical management of hypopharyngeal and
cervical esophageal carcinomas has centered around the ideal
method of reconstruction following tumor ablation (Wenig,
Mulooly, Levy, & Abramsom, 1989). Long-term survival

following radical surgery for advanced stage III and IV

11

carcinoma of the laryngopharynx and cervical esophagus is
extremely poor with reported two and five year survival
rates of zero to 50% (Gluckman, Weissler, & McCafferty,
1987). Consequently, the optimal reconstructive procedure
must provide the lowest mortality and morbidity, the
shortest hospitalization, and the highest rate and most
rapid functional restoration of oral alimentation (Surkin,
1984). Since speech restoration is not critical to patient
survival, it has not been a primary consideration in the

choice of surgical reconstruction.

Various methods have been used for functional reconstruction
of the pharynx, larynx, and cervical esophagus after
extirpative tumor resections. These techniques include
prostheses, skin grafts, cervical flaps, tubed cutaneous and
myocutaneous chest flaps, visceral reconstruction with
stomach, colon, and free jejunal autografts. The problems
associated with skin grafts and flaps are well documented.
They include use for limited pharyngoesophageal resections,
multi-staged procedures, and prolonged hospitalization.

They have high postoperative morbidity and mortality rates
and are frequently associated with flap necrosis and distal
anastomotic stricture. They result in a poor swallowing
mechanism (Biel & Maisel, 1987; Schechter, Baker, &
Gilbert, 1987; de Vries et a1., 1989; Gluckman et a1., 1985;
Gluckman et a1., 1987; Urken, 1989; Fisher, Cole, Meyers, &

Seigler, 1985).

12

Immediate jejunal autograft interposition was first proposed
by Seidenberg, Rosenak, Hurwitt, and Son in 1959 for
reconstruction of defects of the hypopharynx and cervical
esophagus. During the mid 19703, experience with and
refinement of microvascular techniques and instrumentation
facilitated the use of free jejunal autografts to
successfully reconstruct complete pharyngoesophageal defects
resulting from total laryngopharyngectomy and cervical
esophagectomy (Fisher et a1., 1985). Biel and Maisel (1987)
reported that in addition to allowing for extensive local
tumor resections, the method has the advantage of being a
single-staged procedure with low mortality and morbidity
rates and with relatively short hospitalization and
restoration of near-normal swallowing function. Bradford,
Esclamado, and Carroll (1992) have corroborated these

findings.

Physiologically, the jejunal autograft is well suited for
the reconstruction of circumferential pharyngoesophageal
defects. Its luminal diameter of approximately 3.0 cm is
comparable to that of the hypopharynx and cervical
esophagus. It is mucosally lined and therefore has the
advantage of mucous secretion which enhances food passage
through the segment. In contrast, skin flaps, for example,
do not secrete mucus and result in an adynamic, dry passage
which does not assist food transit. Transplanted jejunum

also tolerates irradiation with good preservation of

13

function. The intrinsic peristalsis associated with the
jejunum may also facilitate swallowing and food transit
(Jacob, Francone, & Lossow, 1982, Ch. 14; Gullane, Havas,

Patterson, Todd, & Boyd, 1987; Fisher et al., 1985).

Alternatively, the viability of the jejunal segment to serve
as a vibratory source for alaryngeal voice production has
received much less attention because of the lack of emphasis
on speech restoration and is, therefore, not well known.

The twofold necessity of adequate tumor extirpation and
physiologic restoration of swallowing function has
overshadowed other considerations. To date, little
attention has been given to vocal rehabilitation of patients
undergoing removal of the larynx and significant portions of

the hypopharynx and cervical esophagus (Wenig et al., 1989).

The few studies that have examined vocal restoration have
focused on surgical/prosthetic methods of voice restoration
using either tracheojejunal (TJ) shunts or tracheojejunal
puncture. Recently, Denk, Grasl, Frank, Deutsch, and
Ehrenberger (1992) and Grasl (1993) proposed a surgical
technique for creating a tracheojejunal shunt which includes
immediate cervical anastomoses with previously chosen
arteries and veins. The procedure, similar to the
tracheoesophageal shunt first proposed by Conley et al. in
1958 for voice restoration following total laryngectomy,
requires an end-to-end anastomosis of the graft to the

remaining tracheal stump located above the tracheostome. It

14

is then pulled up and a loop is formed with its turn close
to the floor of the mouth where it is fixed to the digastric
muscles. The distal end of the transplant is then drawn
inferiorly and inserted end-to-side into the reconstructed
wall of the hypopharynx which contributes to the anterior

circumference of the distal part of the neopharynx.

Kinishi, Amatsu, Tahara, and Makino (1991) suggested
alternative surgical modifications including tubed mucosal
flaps to create the tracheojejunal shunt. However, the
method of "shunting" air into the reconstructed hypopharynx

for voice production remains the same.

The exact location of the vibratory section in patients with
jejunal graft interposition remains under investigation
(Kinishi et al., 1991; de Vries et al., 1989; Grasl, 1993).
Denk et al. (1992) suggested that "the sound source for
voice production seems to be a pseudoglottis in the area of
the anastomosis between the hypopharynx and the jejunal
graft" (p. 249). Following tracheo-hypopharyngeal shunt
with jejunal transplantation, the authors described
postoperative voice rehabilitation in 32 of the 40 patients
who underwent the procedure as successful with adequate

communication for everyday purposes.

Kinishi et al. (1991) reported acquisition of voice
capability in three of three patients with the

tracheojejunal shunt operation and consistent use of

15

tracheojejunal shunt speech by all patients. Zeismann,
Boyd, Manktelow, and Rosen (1989) reported good speech
results in two of three patients who underwent neoglottic
and neopharyngeal reconstruction using jejunum.
Unfortunately, an operational definition of the term "good"
is lacking. Two other investigations reported similar
results in their studies of patients who underwent a
tracheohypopharyngeal shunt for voice restoration with free
jejunal autografts. Handl-Zeller et al. (1992) and Grasl
(1993) reported successful voice rehabilitation in 19 of 20
patients (95%) and 31 of 39 patients (79%) respectively

using the "speech-siphon" (p. 98).

The studies that have examined vocal restoration following
tracheojejunal puncture are also limited in number and
methodological description. Wenig et al. (1989) reported
satisfactory communication in 5 patients, 4 of whom had
jejunal grafts and 1 a radial forearm graft to replace a
complete hypopharyngeal/esophageal defect. Although the
investigators reported "coarse" vocal quality and lack of
projection, all patients did achieve fluent voice as judged
by 2 "voice" professionals and one independent non-
professional listener. The investigators felt that the
bowel segment did not provide as good a vibratory source as
the residual pharyngeal wall. Conversely, de Vries et al.
(1989) reported that only 2 of 17 patients who underwent

free jejunal interposition were able to speak with

16

Blom-Singer voice prostheses through a TJ fistula. One
patient used neoglottic speech, whereas four relied on an
electronic speech aid and one on paper and pencil to
communicate. Eight of the 17 patients died of their disease
or other causes, whereas one patient remained alive at the
time of publication with metastatic lesions in the lung.

The authors concluded a better potential for speech

rehabilitation in patients after free jejunal interposition.

In a study by Salamoun et al. in 1987, three of 32 patients
who had undergone free jejunal transfer were using TJ speech
through a voice prosthesis. Unfortunately, no other
information was provided as to vocal quality or prosthetic

use .

Medina, Nance, Burns, and Overton (1987) compared the vocal
quality of 10 patients who used a TE voice prosthesis and
had previously undergone total laryngopharyngectomy or
laryngopharyngoesophagectomy with the vocal quality of 10
patients who had a total laryngectomy and TE puncture who
served as the controls. Five of the 10 experimental
subjects underwent laryngopharyngectomy. Only one subject
had reconstruction of the circumferential defect using a
free jejunal graft. The remaining four were reconstructed
with either a myocutaneous flap or colon interposition. The
authors reported adequate voice and an ability to carry on a
conversation for all patients; however, all subjects

exhibited lower overall pitch and loudness than the

17

laryngectomized control group as well as a "wet" vocal
quality. Although the authors concluded that the TE voice
prosthesis could be used for speech rehabilitation following
total laryngopharyngectomy, the long term use by such

patients for daily communication was not addressed.

Similar vocal results were reported by Mendelsohn, Morris,
and Gallagher in 1993. The vocal qualities of 22
laryngectomees and seven laryngopharyngectomees with jejunal
graft reconstruction were compared with that of a group of
10 normal control subjects. All experimental subjects used
a voice prosthesis for speech production. The
laryngopharyngectomized patients consistently scored lower
in fundamental frequency, intelligibility and social
acceptability than both the laryngectomized group and normal
subjects (p 5 0.05). The vocal quality of the
laryngopharyngectomized patients was perceived as "wet."

The investigators concluded that the operative decision of
total laryngopharyngectomy may profoundly affect the

patient's ability to communicate.

Juarbe et al. (1989) reported an overall 50% success rate
for speech production in 10 patients who underwent total
laryngopharyngectomy. Nine patients who had flap closure
initially achieved acceptable voice with TE puncture. The
single patient reconstructed with jejunal graft had
difficulty voicing and ultimately removed his voice

prosthesis permanently. He was unable to produce esophageal

18

speech as well and ultimately used an electronic speech aid

for communication.

These findings contrast the findings of Schechter et al.
(1987) who reported on a 12-year experience involving the
functional evaluation of 115 patients who had
pharyngoesophageal resections for cancer treatment. Each
patient received reconstruction by one of four major
techniques: deltopectoral flaps (n = 43), pectoralis
myocutaneous flaps (n = 36), gastric pull-ups (n = 19), and
free jejunal autografts (n = 17). Their findings suggested
that the major functional problem with the jejunal autograft
was the failure to develop adequate neoesophageal speech.
They attributed the failure to the large lumen of the
jejunal segment which limits effective vibration for speech
by standard techniques of oral air implosion. Fisher et al.
(1985) also concluded that a disadvantage of the free
jejunal graft interposition is that "it causes limited

esophageal speech" (p. 752).

The two previous studies assessed alaryngeal speech
production using standard methods of oral air implosion as
opposed to surgical/prosthetic alternatives. The critical
point is the inability to achieve voice production following
jejunal graft interposition. The only reports of successful
voice restoration in patients with jejunal graft
interposition have involved surgical and prosthetic methods.

This finding may imply the need for a greater insufflation

19

force beyond that which can be produced using the standard
methods of oral implosion to initiate and maintain tissue
vibration for speech production following jejunal graft

interposition.

Although documentation indicates some successful voice
production using surgical/prosthetic methods, the sole
reliance on TJ voice for conversational speech purposes
remains unclear. Acquisition does not equal functional use
for daily conversational needs. As Baugh, Lewin, and Baker
(1990) suggested, "The correlation between long-term
tracheoesophageal prosthetic use and final tracheoesophageal
speech fluency suggests that the restoration of
communication skills to near pre-laryngectomy speech status
may be an important factor for long-term tracheoesophageal
prosthetic use" (p. 72). Just as tracheoesophageal speech
has been compared to and striven for pre-laryngectomy status
as the standard of speech fluency, so must tracheojejunal
voice attempt to meet the same criteria to remain a viable

alaryngeal speech alternative.

A major factor associated with conversational speech fluency
involves the ease in which speech production can be
produced. If speech cannot be readily and easily spoken, it
is unlikely that the method will prove a viable alternative
for functional restoration of communication. Various
studies have been undertaken to determine those factors

which might be associated with postoperative TE voice

20

production following laryngectomy. Schuller, Jarrow, Kelly,
and Miglets (1983) proposed a formula that reportedly
predicted successful development of TE speech based upon two
parameters: alcohol abuse and stoma size. While both
alcohol abuse and inadequate stoma size may negatively
influence successful TE prosthetic use, neither is a
determinant factor in TE sound production, the fundamental
requirement for the development of successful TE speech.

The authors also studied eight other parameters: current
mode of communication, concomitant medical problems, eye-
hand coordination, educational level, age, hearing acuity,
work status, and living environment. Again, these factors
should be considered extrinsic complications to functional
TE speech restoration. Donegan et al. (1981) also suggested
that such personal characteristics as motivation,
persistence and a high energy level are important extrinsic
determinants of successful long-term TE speech use.

However, personal characteristics like all extrinsic factors
are relevant only if TE methods of voice production are
successful in restoring communication skills to the fluency

and communicative ease of pre-laryngectomy status.

The shunting of pulmonary air into the esophagus induces
oscillations of the lining of the pharynx and upper
esophagus thereby causing a vibrating column of air to be
produced. The sound thus produced is articulated by the

structures of the oral cavity. The amount of oscillation of

21

the lining of the pharynx and cervical esophagus then is the
major determinant of successful TE voice production and
conversation fluency. Hence, various methods of
intraesophageal air insufflation have been proposed to
measure the intraesophageal pressures associated with
tracheoesophageal voice production during the superior
egress of airflow for speech purposes (Singer & Blom, 1980;
Baugh et al., 1987; Lewin et al., 1987; Blom, Singer, &

Hamaker, 1985).

Three etiologies of tracheoesophageal speech failure have
been identified: the hypotonic pharyngoesophageal segment,
hypopharyngeal stricture, and pharyngeal hypertonicity of
the constrictor musculature which Singer and Blom (1980)

first referred to as "pharyngoesophageal spasm."

The breathy voice is the clinical correlate of the flaccid
or hypotonic pharyngoesophageal segment (Perry, Cheesman,
McIvor, and Charlton, 1987). As the severity of the
hypotonia increases, the egress of air from the pharynx and
esophagus may not generate sufficient tissue vibration to
result in TE sound. Digital pressure to the neck segment or
the use of a pressure band around the neck will usually
facilitate tissue vibration and improve sound production

(Cheesman, Knight, McIvor, & Perry, 1986).

Hypertonicity of the pharyngeal constrictor musculature, or

"pharyngoesophageal spasm," has remained the most common

22

reason for TE speech failure (Blom, Singer, & Hamaker, 1986;
Perry et al., 1987; Baugh, Baker, & Lewin, 1988). Surgical
myotomy or myectomy of the pharyngeal constrictors or
neurectomy of the pharyngeal plexus compromises the
pharyngeal contraction which occurs in response to
esophageal distension, thereby allowing fluent TE speech

production (Singer & Blom, 1981; Baugh et al., 1990).

A constriction or stricture of the hypopharynx can be an
important cause of failure to achieve TE speech. A
stricture is best classified as a structural or anatomical
abnormality which hinders TE speech ability by impeding
airflow from the pharynx and esophagus thereby limiting
sound production. Simpson, Smith, and Gordon (1972)
referred to one type of permanent, organic stricture above
the pharyngoesophageal junction as a constant area of
narrowing that does not distend even during swallowing.

This type of stricture is most likely related to loss of
excised mucosa or fibrosis and scarring secondary to delayed
healing. Ogura and Thawley (1979) reported various types of
organic strictures. However, management usually entails
repeated dilation or surgical reconstruction of the stenotic
area. Organic strictures do not improve with relaxation or

voice therapy.

In the cases of hypertonicity of the constrictor musculature
and hypopharyngeal stricture, increased resistance to

airflow has been associated with decreased TE speech fluency

23

and increased intraesophageal pressures. The result is in
an inferior egress of airflow with gastric filling and
limited to no volitional speech production. Speech therapy
has not been successful in relieving the complaints of
increased vocal effort, discomfort, and lack of
conversational speech fluency associated with pharyngeal
constrictor hypertonicity or stricture. Surgical relief
improves tracheoesophageal speech fluency (Baugh et al.,
1990; Singer & Blom, 1981; Singer, Blom, & Hamaker, 1986;

Henley & Souliere, 1986).

Mendelsohn et al. (1993) reported that six of seven patients
(86%) who had undergone jejunal graft reconstruction were
unable to sustain phonation for longer than eight seconds,
compared with 16 of the 22 patients (71%) who had undergone
laryngectomy. An explanation for this finding may be
related to the swallowing difficulties some patients
experience following jejunal graft reconstruction. It has
been estimated that 40% of patients with jejunal graft
interposition have some degree of delay in bolus transit
time through the grafted segment (Kerlin, McCafferty,
Robinson, & Thiele, 1986). The motility problems in these
grafts seem to result from continued, unmodulated
peristalsis and segmental contractions mediated by the
myenteric plexus. Although the grafts are extrinsically

denervated, these autonomous plexi continue to function.

24

Findings such as these have led some investigators to
propose that the intrinsic contractions of the jejunal graft
could potentially lead to dysphagia and could inhibit
tracheojejunal phonation. It has been suggested that
jejunal myotomy may assist TJ voice production similar to
the performance of pharyngeal constrictor myotomy to improve
TE speech production by relieving graft contractions. Wenig
et al. (1989) advocated myotomy above and below the
pharyngoesophageal defect in combination with a tightly
stretched jejunal segment. The authors advised against
redundant or loose tissue in order to create a resonating
area, or "neoglottis." The investigators felt that failure
to establish these conditions would hinder phonatory and

swallowing ability.

Haughey and Forsen (1992) developed a canine model to assess
the functional effects of myotomy on transplanted jejunum.
Five of nine animals underwent complete longitudinal myotomy
and four underwent a sham procedure. Results demonstrated
no significant difference in motility between myotomized
grafts and sham-operated grafts. Furthermore, pressure
parameters between the myotomy group and the sham group were
not significantly different as indicated by manometric
testing. The authors concluded that their results suggested
that myotomy was ineffective in eliminating the intrinsic

contractions seen in free jejunal graft transplantations.

25

Juarbe et al. (1989) reported failure to develop speech in a
patient with jejunal graft reconstruction despite the
performance of a posterior myotomy. The patient ultimately

removed his prosthesis because of his inability to use it.

History has shown that most tracheoesophageal puncture
patients acquire the fluency necessary for effective daily
communication without specific intervention. However, early
studies reported removal of the voice prosthesis in voice
failures without myotomy with spontaneous closure of the TE
puncture. The majority of these failures were associated
with limited speech production as a result of the increased
pharyngeal resistance to the superior egress of voluntary
airflow associated with either hypertonicity of the
pharyngeal constrictor musculature or hypopharyngeal
stricture. However, other reasons for removal of the voice
prosthesis included an inability to care for and manage the
prosthesis, an unrealistic expectation of voice production,
difficulty removing and reinserting the voice prosthesis and
motivational factors (Singer & Blom, 1980; Wood et al.,
1981; Donegan et al., 1981; Atkinson, 1984; Wetmore,

Krueger, Wesson, & Blessing, 1985).

Several authors have suggested that the resistance of the

voice prosthesis to airflow through it should influence the
efficiency with which TE voice is produced. In 1981, Panje
described a surgical—prosthetic method of voice restoration

for patients who received laryngectomy. Similar to other

26

surgical methods, a tracheoesophageal fistula is created and
a Voice Button is inserted seven to 14 days later. The
Panje method of voice restoration also requires occlusion of
the tracheostome to divert pulmonary air into the esophagus
to produce sound. In 1982, Weinberg compared airflow
resistance values for the Panje Voice Button with the Blom-
Singer voice prosthesis, the pharyngoesophageal segment
during voicing, and the human larynx during vowel
production. Average laryngeal airway resistance as reported
by Smitheran and Hixon (1981) is about 35 cm HZO/L/s. The
airway resistance offered by the Voice Button was about ten
times greater than that normally produced by the larynx
during vowel production. The Voice Button also always
exceeded that offered by Blom-Singer voice prostheses.
Airway resistance of Blom-Singer prostheses ranged from 45-
120 cm HZO/L/s. The Panje Voice Button was 2.5 to ten times
greater than that offered by Blom-Singer prostheses.
Weinberg also demonstrated that airway resistance offered by
the Voice Button also exceeded that offered by the voicing
source, the pharyngoesophageal segment, used by esophageal
speakers during speech production. Weinberg, Horii, Blom,
and Singer (1982) reported a source resistance between 155-
270 cm HZO/L/s for five esophageal speakers who had

laryngectomies.

Weinberg’s results revealed a substantial airway resistance

offered by the Panje Voice Button ranging between 285 and

27

440 cm HZO/L/s. He commented, "These results are
discouraging. Speakers using this method must work to
overcome larger opposition to airflow...This circumstance
would be expected to greatly reduce the efficiency with
which tracheoesophageal voice is produced" (p. 500).
Personal experience suggests that current use of the Panje
Voice Button for TE voice restoration has markedly if not
completely declined. Too much effort on the part of the
patient was required to sustain easy conversational speech

production using the device.

In patients who have undergone free jejunal graft
interposition following laryngopharyngectomy and cervical
esophagectomy, little is known about the vibratory
properties of the jejunal mucosa, specifically airflow
resistance factors. Ehrenberger et al. (1985) suggested a
low resistance to airflow with an average "total phonation
pressure" between 50 and 60 cm H20. The authors stated,
"Acting together with the air chamber effect of the jejunal
graft, the low phonation pressure produces a corresponding
phonation" (p. 222). Unfortunately this statement is
difficult to interpret with relationship to restoration of
functional conversational speech since methodology and

terminology are not well defined.

Grasl (1993) also attempted to measure airflow resistance
for voice production in patients with "speech-siphons." In

vivo airflow resistance was measured with a U-shaped, open,

28

water-filled measurement device coupled to the tracheostome.
Values were measured at the start of phonation and expressed
in kilopascals (kPa). A mean maximum phonation time of 8.3
seconds with a range of 5-14 seconds and a mean of 5.2 kPa
with a range of 3.5-8.5 kPa was reported. The study also
reported a mean of 17 syllables per exhalation with a range
of 12-23 syllables. Again, the relationship of these
results to everyday reliance of TJ speech for functional

communication is unclear.

The purpose of this investigation was to determine the
intrajejunal pressure measurements associated with
tracheojejunal speech production by comparing the pressures
associated with speech production in two groups of
laryngectomees. Previous studies have documented the

ability of the jejunum to vibrate.

This study was also undertaken to examine tracheojejunal
vibration as a viable alternative for conversational speech
purposes. It was hypothesized that the patient’s use of TJ
voice as the method for communication should be a good
indicator of functional restoration of speech production.
It was further hypothesized that increased intrajejunal
pressures during speech tasks should correlate with
increased effort to produce and maintain speech production,
thus reducing patient use of TJ voice production for
conversational speech purposes. Given the low survival

rates for patients with this type of advanced carcinoma, it

29

is critical that the selected method of alaryngeal voice
restoration provide efficient speech production which is
comparable to the patient’s premorbid speech abilities.
Time is critical. It must not be wasted on less than
optimal methods which lack the potential for functional,

timely restoration of speech.

CHAPTER II
Methods

Objective air insufflation was performed transtracheally
through the TJ puncture in 4 subjects with jejunal autograft
interposition following laryngopharyngectomy in order to
obtain intrajejunal pressure measurements associated with TJ
voice production. Intrajejunal pressure measurements were
compared to the intraesophageal pressure measurements
obtained from 9 laryngectomized subjects with fluent TE

speech who served as the controls.

Fluent speech was defined as an ability to produce speech
comparable to normal laryngeal speakers. Fluent speech is
greater than 10 seconds of uninterrupted sound duration and
a minimum production of 10 to 15 syllables per breath (Blom

et al., 1985; Baugh et al., 1987; Lewin et al., 1987).

Results were analyzed based upon the criteria published by
Lewin et al. (1987). The criteria were 1) intraesophageal
peak pressures g 20 mm Hg associated with fluent TE speech
production; 2) greater than 10 seconds of uninterrupted
sound duration; and 3) a minimum production of 10 syllables
per breath. Data from both groups were compared as to their
central tendency and dispersion measures and then

statistical tests of significance were applied.

30

31

Subjects

Four patients, between 51 and 62 years of age, with free
jejunal autograft interposition following
laryngopharyngectomy and TJ puncture as described in the
Surgical Technique Section served as subjects. Nine
laryngectomized patients whose primary communication mode
was TE speech served as controls. All subjects were
referred following medical clearance by their
otolaryngologist. No evidence of stricture was reported by
the physician at the time of testing. All subjects were
able to eat soft to regular diets, thus indicating no overt
evidence of stricture. All subjects were previously fit
with and wore an appropriately sized Blom-Singer, low-
pressure voice prosthesis (International Health
Technologies, Inc., formerly Baxter-Mueller, Santa Barbara,
California) to ensure uniformity in experimental design and

prosthesis type.

Surgical Technique

The surgical technique involves the harvest by the general
surgery team as well as placement of a gastrostomy tube.

The length of jejunum harvested corresponds to that supplied
by the mesenteric pedicle, including the artery and vein.
The graft is transferred to the neck and revascularized
using standard microvascular techniques. The appropriate
length of jejunum required for reconstruction of the

circumferential pharyngeal defect is assessed. Redundant

32

length of jejunum is to be avoided as it can lead to
postoperative dysphagia. The jejunum is divided into
reconstructive and monitoring segments, but both portions
are left pedicled on the mesenteric blood supply. The
reconstructive portion of the jejunum is sewn into place to
restore pharyngeal continuity in the isoperistaltic
direction. The monitoring segment of jejunum is brought out
through the neck incision for postoperative assessment of
viability by checking Doppler pulses. The exteriorized
jejunal segment is removed at the bedside on postoperative

day seven.

Secondary tracheoesophageal puncture is performed in the
operating room. The cervical esophagoscope is introduced to
the level of the planned puncture, usually 5-8 mm inferior
to the superior edge of the stoma in the midline. A 14
gauge angiocath is introduced at the planned puncture site.
The needle is withdrawn and the hub is removed. The
puncture site is dilated by passing a filiform then a
follower in sequence through the puncture site. The 14 Fr
red rubber Robinson catheter is sewed to the follower and
all are brought out the mouth. The last step involves
directing the red rubber Robinson catheter inferiorly in the
esophagus. This is accomplished by partial withdrawal of
the catheter under direct visualization with the

esophagoscope followed by directing it distally with a cup

33

forceps.

Assessment

A 14 French catheter (Davol, Inc., Ganston, Rhode Island)
was passed transtracheally through the puncture tract into
the jejunum in subjects with jejunal interposition or into
the esophagus in the laryngectomized controls. The length
of the subject's voice prosthesis was used to determine the
distance of catheter insertion in order to be sure that the
distal tip end of the catheter was sufficiently stenting the
TJ wall. The catheter was connected to a standard
compressed airflow source and a pressure manometer (W.A.
Baum Company, Inc., New York City, NY) via a silicone "Y"
connector (Seamless Hospital Products Company, Wallingford,
Connecticut) and plastic connecting tubing (Sherwood
Medical, St. Louis, Missouri). The testing apparatus is

illustrated in Figure 1.

The method of assessment as previously described by Lewin et
al. (1987) was used to assess subjects. Insufflation was
performed under two conditions: sustained vowel production
and counting for a minimum of 10 seconds during each trial.
Each condition was performed three times. Following each
trial the catheter was removed and rinsed to assure that it
was clean and clear of debris. Presentation of compressed
air from a Thorpe flow meter (Puritan-Bennett Corporation,

Kansas City, Missouri) at 3L per minute was used during the

34

two conditions. This flow rate was chosen based on
published findings suggesting that 3L of flow sustained
esophageal insufflation throughout testing at a rate
sufficient to produce vibration and to register intraluminal

pressure changes on a pressure manometer (Lewin et al.,

1987).
INSUFFLATION
SOURCE
d5.
03
g: PRESSURE
: RECORDER

 

30—1

 

ho

 

      

0

I

 

  

C4? ‘
((qlttwfi

Figure 1. Diagram of testing apparatus demonstrating
catheter placement, attachment to pressure
recorder, and compressed airflow source.

35

Prior to insufflation, the patient was instructed to keep
the mouth open as if producing the sound /a/ throughout
testing. Airflow was initiated slowly. During each trial
of insufflation, pressure levels fluctuated throughout the
10 seconds of sustained phonation of the phoneme /a/. The
peak pressure was recorded. The maximum level of pressure,
peak pressure was recorded since it was felt that the peak
level of pressure should most closely represent the maximal
jejunal response to ongoing insufflation during speech
production. Phonatory duration was monitored using a stop
watch and was dependent upon the length of time from the
onset to the end of voicing. Air presentation was then
stopped to allow the subject to rest. The catheter was
disconnected from the "Y" connector and pressure manometer
to allow the pressures to return to a baseline of zero. The
catheter was then removed and rinsed prior to the next
trial. The catheter was reconnected and the next testing

was completed.

During the second condition, insufflation was performed as
in the previous condition, except that the patient was
instructed to count to 10 as sound was produced. The
maximum number counted as well as the peak level of pressure
were recorded. Again, air presentation was stopped, the
patient was allowed to rest and the catheter was
disconnected to allow baselines to return to zero. The

catheter was removed, rinsed, and reconnected. The next

36

testing was completed. Again, the task was performed three
times. Insufflation was terminated upon completion of each
experimental condition or sooner, if the patient reported

discomfort.

CHAPTER III

Results

Following continuous air insufflation during two conditions,
sustained vowel production and counting, the peak pressure
measurements of the control and experimental groups were
analyzed to determine whether the groups were significantly
different. All individuals were able to sustain sound for a
minimum duration of ten seconds and were able to count to
ten during ongoing air insufflation. Data were
statistically analyzed using the Mann-Whitney test,
Student’s t-test and Pearson product-moment correlation
comparisons because these tests were felt to be the most

sensitive tests of samples of such small size (n = 13).

Subject Characteristics

Tables 1a and 1b demonstrate subject demographic and speech
characteristics for the control and experimental groups.
Three females and 6 males served as controls. Age ranged
between 41 and 78 years with a mean of 63 years of age. The
length of time since total laryngectomy (TL) and TE puncture
to the date of testing for each of the control subjects
ranged between 13 and 115 months since TL with a mean of
67.78 months and 13 and 105 months with a mean of 63.22
months since TE puncture. Four of the 9 subjects received
their tracheoesophageal puncture primarily at the time of

total laryngectomy as reflected by the same number of months

37

38

in both the TL and TE columns of Table 1a. It is
interesting to note that the youngest subject was not the
most recent in terms of length of time from either surgery;
however, the oldest subject was also the longest in terms of
time since both surgeries. Three of the 9 control subjects
had not received radiation therapy as part of their overall

course of treatment.

Table 1a

Subject Characteristics - Control Group

 

 

 

 

 

 

 

 

 

 

 

 

Subject Gender Age TL TE RT UPC
1 F 59 42 22 Yes Yes
2 F 41 68 68 Yes Yes
3 M 60 77 75 No Yes
4 M 61 13 13 Yes Yes
5 M 67 88 88 Yes Yes
6 F 62 63 62 No Yes
7 M 71 51 51 Yes Yes
8 M 68 93 85 Yes Yes
9 M 78 115 105 NO Yes

Mean - 63 67.78 63.22 - -

 

 

 

 

 

 

 

 

 

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

TE = Number of months from tracheoesophageal puncture to
date of testing

RT = Receipt of radiation therapy

UPC = Use of voice prosthesis for communication

39

As Table 1a demonstrates, no information regarding jejunal
interposition (JI) is supplied. For statistical comparison,
the data for time from TL to the date of testing were used
to supply data for comparisons with the experimental group
since the control group’s surgeries did not include jejunal

interposition.

Three males and 1 female served as experimental subjects.
The youngest subject was 51 years old. Two subjects who
were 62 years of age were the oldest. Mean age was 59
years. As Table 1b demonstrates, Subjects 3 and 4 had the
shortest times from all surgeries to the time of testing.
Subjects 3 and 4 also had all three surgical procedures
including total laryngectomy, jejunal interposition and
tracheojejunal puncture completed at the same time as
indicated by the same number in each of the three columns.
Subjects 1 and 2 underwent multi-staged procedures on
separate dates. The length of time since total laryngectomy
to the date of testing ranged between 1 and 61 months with a
mean of 22.75 months. The length of time since jejunal
interposition ranged between 1 and 56 months with a mean of
20 months. The range of time since tracheojejunal puncture
was between 1 and 54 months with a mean of 19 months. Half
of the subjects had received radiation therapy as part of
their overall treatment course, whereas the other two had

not.

40

Table 1b
Subjegt Charactegistics - Experimenual Grgup

 

 

 

 

 

 

Subject Gender Age TL J I TJ RT UPC
1 M 62 61 56 54 Yes NO
2 F 51 25 19 17 NO NO
3 M 62 1 1 1 NO NO
4 M 61 4 4 4 Yes No

Mean - 59 22.75 20 19 - -

 

 

 

 

 

 

 

 

 

 

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

JI = Number of months from jejunal interposition to date of
testing

TJ = Number of months from tracheojejunal puncture to date
of testing

RT = Receipt of radiation therapy

UPC = Use of voice prosthesis for communication

Subjects in both groups wore the voice prosthesis at the
time of testing, but only the control group used the voice
prosthesis for daily communication. None of the
~experimental subjects reported routine use of the voice
prosthesis for speech purposes. Although Subjects 3 and 4
in the experimental group had limited practice using the
device given the short amount of time since TE puncture to
the time of testing, one and four months respectively, no
attempt to speak with the device was made by either subject.
Functional alaryngeal speech restoration with TJ puncture

appears equivocal for this group.

41

Table 2a shows central tendency and dispersion data for age,
time from total laryngectomy and TE puncture for subjects in
the control group. Mean age was 63 years with a standard
deviation of 10.27 years. The average time since total
laryngectomy to the time of testing was 67.78 months with a
standard deviation of 30.27 months. A mean of 63.22 months
since tracheoesophageal puncture to the time of testing with

a standard deviation of 30.35 months occurred.

Table 2a

Central Tendency and Dispersion Data - qutrol Grguu; n = 9

 

 

 

 

Minimum Maximum Mean Standard Deviation
Age 41.00 78.00 63.00 10.27
TL 13.00 115.00 67.78 30.27
TE 13.00 105.00 63.22 30.35

 

 

 

 

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

TE = Number of months from tracheoesophageal puncture to
date of testing

The mean age for the experimental group was 59 years with a
standard deviation for the 4 subjects of 5.35 years. The
time since total laryngectomy averaged 22.75 months to the
time of testing with a standard deviation of 27.65 months.
The average time since jejunal interposition (JI) to the
time of testing was 20 months with a standard deviation of

25.26 months. The mean time from tracheojejunal puncture to

42

the time of testing was 19 months with a standard deviation
of 24.35 months. Central tendency and dispersion data for
age, time from total laryngectomy, jejunal interposition and

TJ puncture for the experimental group are summarized in

 

 

 

 

 

Table 2b.
Table 2b
Centua; Tendency gnd Dispersiou Duua - Euuegiuentgl Quouu; n
=4
Minimum Maximum Mean Standard Deviation
Age 51.00 62.00 59.00 5.35
TL 1.00 61.00 22.75 27.65
JI 1.00 56.00 20.00 25.26
TJ 1.00 54.00 19.00 24.35

 

 

 

 

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

JI = Number of months from jejunal interposition to date of
testing

TJ = Number of months from tracheojejunal puncture to date
of testing

A comparison between the control and experimental groups
reveals slightly less variability overall in the
experimental group. However, given an n of only 4, this
finding is not surprising. The recent use of jejunal
interposition as a routine surgical alternative for
reconstruction following resections for carcinoma of the

laryngopharynx further decreases the potential variability

43

of time from surgery in the experimental group. In general,
both groups appear to be similar in age; however, the
control group demonstrated greater time since all surgical
procedures. This might imply longer and complete surgical
recovery and greater opportunity for practice and use with
the voice prosthesis for speech purposes for the control
subjects. Given the decreased time since surgery to the
time of testing for 2 of the experimental subjects, factors
related to incomplete recovery may have influenced results.
For example, the influence of unresolved edema on TJ voice
production and its use for speech purposes should be

considered when making comparisons to the control group.

Cgrrelation Analysis by Subject Characteristics

A correlation of the relationship between subject
characteristics and experimental variables is illustrated in
Table 3 for both groups of subjects. Inverse relationships
between the length of time since each of the three surgical
procedures and the intraluminal pressure associated with
vowel prolongation and counting and the length of time and
intraluminal pressure variability are noted. As length of
time from surgery increased, the intraluminal pressures
associated with each task and the variability of the
pressure decreased. Although none of the correlations were
found to be statistically significant (p g 0.05, tabled
value < 0.5529), stronger correlations are demonstrated

between length of time since surgery and the intraluminal

44

pressures associated with task performance during sustained
/a/ production. The correlation between TL and /a/ was r =
-0.3926, r = -0.4150 for TJ and /a/, and r = -0.4594 for TE.
The correlation between TL and counting was r = -0.1137, r =
-0.1306 for TJ and counting, and r = -0.1974 for TE.
Comparatively, weaker correlations are found between length
of time since surgery and the variability of the
intraluminal pressure measurements associated with task
performance. The correlation between TL and the variability
of /a/ was r = -0.1443, r = -0.1901 for TJ and the
variability for /a/, and r = -0.1503 for TE. The
correlation between TL and the variability of counting was r
= -0.0962, r = -0.1218 for TJ and the variability of
counting and r = -0.0942 for TE. Statistical significance
remains difficult to detect because of the small sample

size.

45
Table 3
Qprzeiauion Matrix Demonstrating the Relationship Seuweep

Supjegp Quuragteyistics and Task Pepfoguapce f9; Ail
Subjects

 

 

 

 

 

 

 

 

 

 

 

 

 

Task
/a/ Count /a/SD Count SD
Age 0.118 0.2628 -0.4139 0.0327
TL -0.3926 -0.1137 -0.1443 -0.0962
TJ -0.4150 -0.1306 -0.1901 -0.1218
TE -0.4594 -0.1974 -0.1503 -0.0942
N = 13
DF = 11

R 9 0.0500 = 0.5529

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

TJ = Number of months from tracheojejunal puncture to date
of testing

TE = Number of months from tracheoesophageal puncture to
date of testing

Power Analysis

A power analysis of the data suggests that in order to have
an 80% chance of detecting a significant correlation at the
0.05 level for a two-tailed test, the correlation
coefficient would need to assume a value of 0.70 or higher
in a similar study with 13 cases. Alternatively, in order
to have an 80% chance of detecting a significant correlation
at the 0.05 level for a two-tailed test using a correlation
coefficient of approximately 0.40, the number of subjects

would need to be increased beyond the current 13 to a total

46

of 45. Given the particular subject characteristics of this
population, statistical significance may be difficult to

achieve in either circumstance.

Correlation Anaiysis by Gender

In order to determine the strength and direction of the
relationship between intraluminal pressure and task
performance based upon gender, a correlation analysis was
applied separately for males and females. Tables 4a and 4b
illustrate the data. Inspection of the data presented for
female subjects in Table 4a suggests stronger correlations
for female subjects particularly between the time since all
surgeries to the date of testing and both of the tasks
including sustained /a/ and counting. The correlation
between TL and /a/ was r = -0.8939, r = -0.8794 for TJ and
/a/, and r = -0.8814 for TE. The correlation between TL and
counting was r = -0.9151, r = -0.8823 for TJ and counting,
and r = -0.9839 for TE. The inverse relationship suggests
that longer times from surgical procedures are correlated
with lower pressures associated with voicing and speech
production. Only the males demonstrated an inverse
relationship between the time since any of the surgeries and
the variability in pressure on the counting task. This
might suggest that there is less variability in pressure
associated with counting the further one is from surgery for
male subjects. Although the time from TE puncture and the

peak pressures associated with counting was found to be

47

statistically significant, none of the other correlations
were found to be significant for the female subjects (p 5

0.05, tabled value < 0.9500). None of the correlations were

found to be statistically significant for the male subjects

(p g 0.05, tabled value 5 0.6664) on any variable.

Table 4a

Correlation Matrix Demonstrating Relationship Between
Subject Characteristics and Task Perfogpance for Female
m

 

 

 

 

 

 

 

 

 

 

 

 

 

/a/ Count /a/SD Count SD
Age 0.5857 0.4299 -0.1799 0.7008
TL -0.8939 -0.9151 -0.3234 0.3189
TJ -0.8794 -0.8823 -0.3910 0.3139
TE -0.8814 -0.9839 -0.0513 0.3665
n = 4
BF = 2

R 9 0.0500 = 0.9500

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

TJ = Number of months from tracheojejunal puncture to date
of testing

TE = Number of months from tracheoesophageal puncture to
date of testing

48
Table 4b

gobrelabion Matrix Demonsbrating Relationship Between
Subject Characteristics and Task Perfobmanee for Male
Subjecbs

 

 

 

 

 

 

 

 

 

 

 

 

/a/ Count /a/SD Count SD
Age -0.0284 -0.0653 -0.1781 -0.2599
TL -0.3304 0.0070 -0.0827 -0.1553
TJ -0.3482 -0.0052 -0.1186 -0.1914
TE -0.3821 -0.0388 -0.1165 -0.1867
n = 9
DP = 7

R 9 0.0500 = 0.6664

Age = Number of years old at time of testing

TL = Number of months from total laryngectomy to date of
testing

TJ = Number of months from tracheojejunal puncture to date
of testing

TE = Number of months from tracheoesophageal puncture to
date of testing

Peek Pressure Measurement Analysis

Individual peak pressure measurements are presented for both
the control and experimental groups in Tables 5a and 5b. Of
interest are the peak pressure measurements above 20 mmHg
for the control group during both tasks. Previously
published data by Lewin et al. (1987) suggested a level of
20 mmHg as the maximal elevation in pressure for fluent TE
speakers during similar tasks of sustained vowel duration
and counting. It should be noted that while peak pressures

did demonstrate mild elevations above 20 mmHg for some

control subjects, no simultaneous phonatory deterioration in

49

task performance occurred during either task for any of the
subjects. All subjects in the control group were able to
sustain /a/ for a minimum of 10 seconds during insufflation

and were able to count to 10 without interruption.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 5a

Individual Peak Pressure Recordings in uuflg go; [at and

Counting - Control Group

/a/ Counting
Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3

1 22 20 22 1 24 20 22
2 20 8 8 2 10 8 8
3 8 8 8 3 18 14 14
4 16 16 16 4 15 14 16
5 14 12 10 5 14 15 14
6 22 8 26 6 8 7 22
7 20 21 14 7 20 12 14
8 20 26 22 8 28 26 28
9 18 18 20 9 18 20 18

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mean 17.778 15.222 16.222 mean 17.222 15.111 17.333

Peak pressure recordings in the experimental group as shown
in Table 5b demonstrated higher peak pressure measurements
during sustained /a/ compared to the control group. None of
the experimental subjects demonstrated the same peak
pressure elevations during any of the 3 trials under either

condition, sustained /a/ or counting, whereas Subjects 3 and

50

4 in the control group produced the same peak pressure
elevations during all three trials of sustained /a/ as shown

in Table 5a.

Table 5b

geek Eressure Recorgings in mmHg for [at apd gouptipg -
Euperiuental Group

 

 

 

 

 

 

 

 

 

/a/ Counting
Trial 1 Trial 2 Trial 3 Trial 1 Trial 2 Trial 3
1 30 38 12 1 40 14 18
2 20 40 12 2 20 24 20
3 22 18 20 3 10 22 12
4 22 32 32 4 30 28 28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mean 23.500 32.000 19.000 mean 25.000 22.000 19.500

The means for each of the 3 trials for each task were
computed and compared. To determine whether the 2 groups
were significantly different from one another based on their
pressure measurements during either task, a Mann-Whitney
test was performed. Results are illustrated in Table 6.
The statistically significant difference that was found
between the groups during sustained /a/ production, (U=2, p
5 0.01), was extrapolated from Siegel (1956) in Table K (p.
275). The experimental group demonstrated significantly
greater pressure measurements than the control group during
sustained vowel production. No significant difference in

peak pressure measurements was found during counting nor in

51

pressure variability during either task using the Mann-
Whitney test (U=9, p z 0.05) as extrapolated from Siegel

(1956) in Table K (p. 277).

Table 6

nn- 'tne est or Tasks and Variabi it i T s

Eegformance between Groups

 

 

 

 

 

Variable U-Statistic P-value (one-tailed)
la/ 2.00 p g 0.01
count 9.00 p Z 0.05
/a/SD 6.50 p _>__ 0.05
count SD 8.50 p z 0.05

 

 

 

 

 

Further analysis of the data using a two-sample t-test
confirmed findings of the Mann-Whitney test. Data are
presented in Table 7. Statistical significance was found
between groups during the sustained vowel production (0.0106
= p g 0.05). T-test analysis also revealed a statistically
significant difference in the pressure variability between
groups during sustained /a/ production (0.0402 = p g 0.05).
Although no significant difference was found between groups
in peak pressure measurements during counting, an
examination of the means for both groups (control 2 =
16.556; experimental x = 22.167) reveals higher pressure
associated with counting by the TJ speakers than the TE

speakers.

52

Table 7 also shows higher variances associated with counting
(control variance = 30.083; experimental variance = 34.185)
than with sustained vowel production (control variance =
23.383; experimental variance = 14.037) for both groups.
These results suggest higher pressure variation for both
groups during the counting task. The higher variances along
with the small sample size reduce the ability to
statistically demonstrate a significant difference between

groups on the counting task.

53

 

 

 

 

 

 

 

maoum Housuafluumxu u u
acoum Houucoo u 0
e m z
mmmo.o w.m mhom.m n owh.mn ooma.m mocowho>
mmmd.o Ha mmmm.HI mmwm.m cmn¢.~ :00: am HCSOU
e m z
mmso.o m.m mam¢.m u mas.mm va.oa mocmfium>
movo.o AH memn.~u Hmsm.m mmmo.n cum: om\m\
c m z
Hamm.o m.m mmma.a u mma.¢m mmo.om OOCMHHM>
wNNH.o Ha mahm.al th.N~ www.ma cow: uafioo
w m z
nhom.o m.w mmww.H n bmo.vH mmm.mm mocowuc>
moao.o HH mabo.m fl mmm.¢m 50¢.0H £602 \m\
GOEMOflmwcmfim mo ofiumfiuoum Home 0 0 msouu manowuo>

 

 

 

 

 

 

 

h OHQMB

 

54

A statistically significant difference between groups in
pressure variability was detected during sustained vowel
production as represented in Table 7 by the variable /a/SD,
throughout the three trials using the t-test analysis
(0.0402 = p 5 0.05). An examination of the means for /a/SD
for both groups and their variances (control 2 = 3.0589,
variance = 10.432; experimental R = 8.8781, variance =
35.795) demonstrates less variability in pressure for the
controls during sustained vowel production throughout the
three trials. The greater variability for the experimental
group may suggest greater difficulty in regulating and
controlling sound production by individuals with jejunal
interposition as compared to tracheoesophageal speakers.
This variability may also be the result of the small sample

size.

Eewer Analysis

A power analysis for independent groups and one-tail test
for unequal ns was performed in order to project the number
of cases needed to extract a valid number for significance
between groups during counting. The results of the analysis
are presented in Table 8. The table shows that the present
results allow for a 50% chance measurement of significance
based on the available raw data. Assuming these data are
representative of the true means of the total population,
the number of subjects would need to be increased to 12 in

each group in order to have an 81% chance of detecting a

55

significant difference between groups. A subject pool of 20
in each group would allow for a 95% chance of detecting
significance, whereas 30 in each group would allow for a 99%
chance of finding a significant difference between groups.
The extremely poor rates for long-term survival in this
population, as well as the relatively rare occurrence of
such advanced stage III and IV carcinomas of the
laryngopharynx which necessitate jejunal interposition,
suggest the practicality of locating even a subject pool of
12 per group may be difficult. The results of the present
power analysis suggest that longitudinal study with
retrospective analysis of this population may be the more

practical investigative design.

Table 8
Poyer Apelysis for Sguntiug - Indepengepb Srgups end One-

tail Test for Unegual ns

 

t-test Parameters

 

 

 

 

 

 

 

Significance level 0.05

Mean in Group 1 16

Mean in Group 2 22

Standard deviation 5.656854

Number of cases in Group 1 9.0 12 20 30
Number of cases in Group 2 4.0 12 20 30
Power 0.50 0.81 0.95 0.99

 

 

 

 

 

 

 

CHAPTER IV

Discussion

These findings provide support for the hypothesis that
following jejunal interposition and tracheojejunal puncture,
TJ voice and speech production are associated with increased
pressure measurements when compared to pressure measurements
associated with tracheoesophageal speech production in
patients following total laryngectomy and TE puncture.

While all of the control subjects used the voice prosthesis
for communication, none of the four TJ speakers in this
study used the voice prosthesis for speech purposes.
Although functional speech restoration as indicated by the
patient's use of the voice prosthesis for communication
appears reduced for patients with TJ puncture, the findings
do not provide absolute confirmation of increased pressure
as the sole reason for lack of reliance on TJ voice for
communication. The recency of surgery for half of the
experimental subjects and the limited time using the
prosthesis along with the possibility of incomplete recovery
from all surgeries may have limited vocal competency and,
therefore, the use of the prosthesis for communication.
Possible delays in postoperative healing and associated
complications such as continued edema and reduced
familiarity with the prosthesis must be considered when
using the results of this study to examine the efficacy of

this method for alaryngeal speech restoration in patients
56

57

with jejunal interposition and TJ puncture. These factors
and the number of experimental subjects in this study make

it difficult to statistically substantiate findings.

Despite the limitations of the subject pool, several
significant findings and trends are clear. As the length of
time from surgery increased, the intraluminal pressures
associated with each task and the variability of the
pressure decreased. This inverse relationship between
length of time from surgery and pressure suggests that the
greater the length of recovery from surgery the easier it
becomes to speak. One possible explanation for this finding
may involve the stability and tone of the musculature
following surgery. Immediately following total
laryngectomy, a temporary but complete compromise of the
musculature occurs as a result of the resection and removal
of the larynx. However, over time vascularization and
partial reinnervation occurs concomitantly with healing,
improving the tone and stability of the resected
musculature. These changes may, in fact, assist phonatory

vibration by providing greater consistency in muscle tone.

While statistical analysis did not reveal a strong
relationship between length of time from surgery and the
variability of pressure measurements, there was an overall
decrease in the pressure variability. Two of the control
subjects demonstrated the same peak pressure measurements

over three trials of sustained vowel production suggesting

58

greater stability, tone and control of the vibratory segment

during phonation with longer postoperative recovery.

One other explanation for the decreased pressure variability
in TE speakers compared to TJ speakers may be that the
pharyngeal wall has a better ability to vibrate than the
jejunal wall making it a superior vibratory replacement for
alaryngeal sound production. Pharyngeal physiology may in
itself provide more regularity and stability than the

jejunum during vibration.

Greater pressure variation associated with speech production
may occur in patients following jejunal interposition as a
result of the continued, unmodulated peristalsis and
segmental contractions which occur in the transposed graft.
Although the grafts are extrinsically denervated, the
intrinsic contractions of the graft may continue for several
months post-surgery (Kerlin et al., 1986) and potentially
inhibit tracheojejunal phonation. Interestingly,
experimental Subject 1 who had the longest duration since
all surgeries, approximately 5 years, was continuing to
demonstrate intrinsic peristaltic activity as indicated by
his physician following recent direct laryngoscopy. The
same subject also demonstrated consistently higher pressure

elevations during the two phonatory tasks.

The results of this study do not support any significant

differences between males and females based on pressure

59

measurements during performance of either task. Although
correlations appear to be stronger for female subjects
between the length of time from surgery and lower pressure
levels for phonatory tasks, male subjects demonstrated
stronger correlations between length of time from surgery
and decreased pressure variability during counting. Female
subjects also demonstraed a statistically significant
correlation between the time from TE puncture and peak
pressure measurements during counting. None of the other
correlations for any task by either gender were found
significant. These findings may actually be a result of the
small size of the sample rather than a true finding in the

population.

Intraesophageal peak pressure measurements varied between 8
and 26 mmHg during sustained vowel production for the TE
speaker group, whereas the TJ speakers produced a sustained
vowel production associated with a pressure range of 12-40
mmHg. Both groups were able to sustain /a/ for 10 seconds
without vocal deterioration. Statistical analysis of the
difference in pressure ranges for both groups was found to
be statistically significant. Again, this finding supports
the hypothesis that TJ sound production is associated with

higher pressures than TE sound production.

Individual peak pressure measurements between 7 and 28 mmHg
occurred for the control group during counting. Peak

pressure measurements occurred between 10 and 40 mmHg for

60

the 4 experimental subjects during the counting task.
Although no significant difference was found, higher
pressure measurements were associated with counting by the
TJ speakers as compared to the control group of TE speakers.
Despite a lack of a statistically significant difference,
the pressure variation during counting for both the control
and experimental subjects was consistently higher than the
pressure variation during sustained vowel production by both
groups. This finding may suggest an influence of

articulation on connected speech sounds.

Before examining the pressure differences between the two
groups during task performance, it is interesting to note
the particular range of pressure associated with fluent TE
speech for the control group during performance of both
tasks. Previous documentation by Lewin et al. (1987)
suggests pre-TE puncture pressure measurements less than or
equal to 20 mmHg are a good predictor of fluent
postoperative TE speech. Patients who obtained fluent TE
speech without myotomy post-TE puncture had low
intraesophageal peak pressure measurements (< 20 mmHg), a
minimum of 10 seconds of uninterrupted sound production and
an ability to produce at least 10 syllables per breath
preoperatively. Patients whose peak pressure measurements
were above 20 mmHg also demonstrated phonatory and speech
deterioration below the established criterion prior to TE

puncture. They achieved fluent speech production following

61

myotomy of the pharyngeal constrictor musculature. Pressure
measurements post-myotomy were similar to preoperative
pressure measurements for the fluent speakers

(< 20 mmHg) pre-TE puncture.

All subjects who served as controls in the current study
were at least 1.5 years post-TE puncture. The subject with
the longest time since TE puncture was 8.75 years. All
subjects were fluent speakers. No vocal deterioration
occurred during air insufflation in subjects whose peak
pressure measurements were above 20 mmHg as had occurred in
non-fluent subjects prior to TE puncture in the Lewin et al.

(1987) study. Several explanations may be considered.

In the original study, subjects were tested prior to
surgery. Those patients who received myotomy were then
retested shortly after surgery. Again, it is possible that
changes in the tone of the musculature continues to improve
with the passing of time post-surgery. Therefore, the
current elevations in peak pressure measurements during
insufflation may represent the pressure needed to overcome
the resting tone of the pharyngoesophageal (PE) segment
post-healing in patients who served as control subjects in
the present study. Unlike subjects in the original study,
elevated peak pressure measurements in the present subjects
were not associated with deterioration in vocal tasks nor
did pressure measurements continue to rise as occurred in

patients with hypertonicity of the PE segment in the

62

original study. Clearly, the elevations in peak pressure
measurements in control subjects in the current study are
different from the elevations associated with deterioration

in speech tasks by patients in the original study.

Another possible explanation for the extended pressure range
in the present control subjects may involve possible over-
insufflation of air. Given the presumed improvement in
muscular tone of the vibratory segment over time post-
surgery, 3L per minute of airflow may have been excessive
resulting in elevated pressure measurements and an inferior
egress of the unused portion to the stomach. In the study
by Lewin et al. (1987), gastric filling with air was
reported by subjects and was indicated by marked elevations
in pressure. In the present study, air eructation was noted

in some subjects following air insufflation.

The method of air insufflation may also be a factor
affecting intraluminal pressure measurements. Transtracheal
air insufflation was performed in the present study, whereas
transnasal air insufflation, as illustrated in Figure 2, was
used in the original study by Lewin et al. (1987). While
both methods were found to adequately insufflate the
esophagus for sound production, the placement of the
catheter into the esophagus may inadvertently affect

resulting pressure measurements.

63

PRESSURE
RECORDER

 

J...

 

 

 

 

INSUFFLATION
SOURCE

 

 

Figure 2. Diagram of testing apparatus demonstrating
catheter placement, attachment to pressure
recorder, and insufflation source for preoperative
prediction of postoperative TE speech.

64

Many TE speakers are also able to speak using traditional
esophageal methods of oral air implosion. It has been
suggested that self-injection of air through oral implosion
as used by traditional esophageal speakers may interfere
with external methods of air insufflation by falsely
elevating pressure measurements. Again, inadvertent non-
volitional control over the vibrating segment may have
resulted in an increase in pressure measurements in control

subjects.

Returning to the current results, a statistically
significant difference in peak pressure measurements between
the control and experimental subjects during sustained vowel
production was found. A trend towards increased pressure
associated with counting by the experimental subjects
compared to lower pressure generated on the same task by the
control subjects was also found. Both findings support the
original hypothesis of elevated pressure associated with TJ

speech production.

The need for increased air pressure by TJ speakers for
jejunal vibration may be a reason for the increased effort
often reported by TJ speakers to maintain TJ speech
production. While increased intraluminal pressure may not
be the sole reason for failure of the TJ speakers in this
study to use the voice prosthesis for routine communication,
it can certainly be considered a barrier to the ease of

speech production. When combined with the need to overcome

65

the resistance of the various voice prostheses to airflow
(Weinberg, 1982), increased patient effort to speak may be a
limiting factor to the use of voice prostheses for TJ
communication. Patient effort is an important consideration
when evaluating the method as a viable and efficacious
alternative for communication in patients with jejunal

interposition.

Greater pressure variability was also demonstrated for both
groups during counting than during sustained vowel
production. Greater variation in pressure measurements
during speech tasks may reflect the influences of phrasing
and articulatory valving on continuous airflow compared to a
sustained phonatory production. The significance of
possible influences on TJ speech production warrants further

investigation.

A significant difference between groups in the pressure
variability during sustained /a/ throughout the three trials
was demonstrated. Less variability in pressure occurred in
control subjects than experimental subjects. Although the
jejunum is well suited in its physical similarities to the
pharynx and esophagus as an appropriate substitute for
reconstruction of the laryngopharynx and cervical esophagus
after extirpative tumor resections, the inherent differences
in its physiology, lumen diameter and wall thickness most
likely provide greater resistance to phonatory vibration.

Again, the greater variability for the experimental group

66

may suggest greater difficulty in the regulation and control
of sound production for individuals with jejunal

interposition as compared to TE speakers.

Although no significant difference was found in pressure
measurements between the control and experimental groups
during counting, the limited number of subjects in the
sample may have reduced the actual detection of a real
difference. Based on a power analysis of the data, the
possible recruitment of a minimum of 12 subjects per group
may allow for a significant detection of a real difference
between the groups. However, the high mortality and
morbidity rates in patients with advanced Stage III and IV
carcinoma of the laryngopharynx and cervical esophagus, as
well as the relatively rare occurrence of such tumors, make
the practicality of locating even a subject pool of 12 per
group difficult for assessment in a single short term study.
Furthermore, the variation in surgical technique from
surgeon to surgeon -- including the placement of the TJ
puncture -- offer further challenges to obtaining similar
samples of adequate size. It appears that the answers to
the questions posed in this investigation based on the
results of the present power analysis may be better found
through longitudinal study and retrospective analysis of

this population.

CHAPTER V

Conclusions and Future Directions for Research

Summary

This investigation has attempted to examine tracheojejunal
voice production by assessing the intraluminal pressure
associated with its production and comparing the results to
the intraesophageal pressure produced by tracheoesophageal
speakers during the production of two tasks in two groups of
laryngectomees. It has been the experience of this examiner
that patients who have undergone tracheojejunal puncture
following jejunal interposition for extirpative removal of
malignant tumors involving the laryngopharynx are able to
produce tracheojejunal phonation. However, the use of TJ
sound production for routine communication by TJ speakers is
markedly less than the use of TE sound production for speech
purposes by patients following simple laryngectomy and TE
puncture. A review of the literature reveals a myriad of
subjective descriptions of tracheojejunal sound production;
however, little objective evaluation of TJ voice production
is available to either support or refute the method as a
viable, efficacious alternative for alaryngeal speech
restoration in patients with extended tumor resections
involving the hypopharynx and cervical esophagus. This
study, therefore, proposed that increased pressure

measurements were associated with TJ speech production and

67

68

may be one factor which might negatively affect the use of

TJ voice production for routine communication.

Although the results of this study generally supported the
hypothesis of increased pressure associated with
tracheojejunal voice production, only the task associated
with sustained vowel production was found to be
significantly different based on pressure measurements
between the two groups of laryngectomees. Although
statistical significance was not found between the groups
during the counting task, the experimental group did
demonstrate greater pressures associated with task
production supporting the proposed hypothesis. Greater
variability in pressure responses during counting and the
small size of the sample made the detection of significance

during this task difficult.

The possibility of differences in pressure associated with
task performance based on gender could not be definitively
determined. Stronger relationships which demonstrated
decreased pressure variability over time associated with
counting was found for male subjects. Female subjects
demonstrated less pressure associated with the production of
either task the longer they were recovered from the time of
surgery. Whether this distinction between genders is an
accurate reflection of the population is questionable given
the limitations of the sample. Personal experience has not

found this distinction to be true.

69

Questions regarding the tone and stability of the vibratory
segment in relationship to the length of time from surgery
and other factors -- including adequate healing of tissue
and adequate practice using the voice prosthesis -- suggest
that a longitudinal analysis of subjects may be needed to
better assess these factors. A power analysis of the data
has suggested that a minimum of 12 subjects in each group
would be needed in order to have approximately an 80% chance
of finding a significant difference between groups based on

pressure measurements .

Future Research

Given the poor rates for long-term survival in patients with
carcinoma of the hypopharynx and cervical esophagus as well
as the extremely rare incidence of these types of tumors
which require jejunal reconstruction, a multi-institutional
investigation involving longitudinal study with
retrospective analysis is probably the more practical
investigative method to answer questions regarding
statistical significance. Despite a lack of statistically
significant results, the clinical relevance of findings
should not be overlooked. A trend in results may be as
important to decisions regarding quality of life as finding

statistical significance in this population.

As part of the assessment method of this investigation,
three successive trials per task were completed by each

patient. Only 2 control subjects were able to replicate the

70

same pressures on each trial per task. All other patients
produced different peak pressure measurements during each
trial of either task. Future research is needed to
investigate the factors which influence the intra-subject
variability in pressure associated with TE and TJ voice
production. An understanding of these factors may offer the
possibility of greater volitional control over the vibratory
segments following laryngectomy or laryngopharyngectomy,
thus enhancing the efficiency and success of speech

restoration in both populations of laryngectomees.

The results of the present study demonstrated a difference
in the level of air pressure required to produce a steady
state vowel versus speech production. Further investigation
is needed to determine the factors affecting the changes in
intraluminal air pressure associated with each type of
production so that voice prostheses which offer the least
resistance to airflow can be chosen to meet individual

patient requirements for speech production.

A better physiological understanding of the jejunal
autograft including its continued ability to produce
peristaltic contractions is also suggested as an avenue for
further research with respect to its viability as a
phonatory source for alaryngeal voice production. Mucous
secretion, which is one of the advantages of the jejunal

autograft for enhancing the act of swallowing, may in fact

 

71

be responsible for the "wet" or liquid quality of the

resulting voice which many patients find unattractive.

Future studies are also needed to answer questions regarding
the location of TJ vibration in patients with TJ puncture
following jejunal interposition. The results may provide
valuable information regarding surgical decisions as to the
actual placement of the puncture site during surgical
resections. Specific information as to the location of
tracheojejunal vibration may also help speech pathologists
provide more appropriate treatment to facilitate
tracheojejunal speech production. Such information may also
be beneficial in optimizing current prosthetic designs while

prompting new types of voice prostheses.

In this study none of the experimental subjects used their
voice prostheses for routine speech production. While the
results of this study do not provide conclusive evidence to
contradict the method of tracheojejunal puncture and voice
restoration for patients with jejunal interposition, they do
provoke further questions which need to be answered in order
to ultimately provide the optimal method of communication
while avoiding patient frustration and communicative failure

in this population.

APPENDICES

APPENDIX A

OBJECTIVE AIR INSUFFLATION FORM

APPENDIX A

SPEECH-LANGUAGE PATHOLOGY
Department of Otolaryngology — Head & Neck Surgery
University of Michigan Hospitals

 

 

 

 

 

 

(313) 763-4003
Patient:
Reg #:
Date of Birth:
Date of Evaluation:
Referral Source:
0 TIVE AIR IN ATI

Catheter Placement: Transtracheal catheter distance - _cm

Sustained la/

Tri 1 11m Idalfi
_mmHg _sec. _mmHg _sec. _mmHg _sec.
_mmHg _sec. _mmHg _sec. _mmHg _sec.
_mmHg _sec. _mmHg _sec. _mmHg _sec.

Counting

Tri 1 :LriaLz Trial}

#3 mmHg #3 mmHg #3 mmHg
#5 mmHg #5 mmHg #5 mmHg
#3 mmHg #3 mmHg #5 mmHg

 

 

 

72

APPENDIX B

MICHIGAN STATE UNIVERSITY INFORMED CONSENT FOR EXPERIMENTAL
PROCEDURE

 

APPENDIX C

The University of Michigan Medical Center
INFORMED CONSENT FOR EXPERIMENTAL PROCEDURE
(Version January 1992)

1. Title of research project.

Objective Pressure Measurements in Two Groups of laryngectomees
2. Names of the researchers.

Jan S. Lewin, M.S.

3. Purpose of the research.

This investigation is being done to compare the pressure measurements of individuals
who have undergone laryngectomy with the pressure measurements of individuals who
have undergone laryngopharyngectomy with free jejunal graft reconstruction. This
information will provide a better understanding of tracheojej unal sound production as
an alternative for alaryngeal speech production.

4. Number of subjects included in the study.

Ten patients who have had laryngopharyngectomy with jej unal reconstruction and ten
patients who have had laryngectomy.

5. Description of the experiments.

A small catheter will be gently inserted into the puncture site and air will gently
insufflate the esophagus ancl/or Jejunum. The catheter will be connected to a pressure
recorder. Pressure measurements will be recorded during two tasks. You will be
asked to keep your mouth open while air is gently insufflated simulating the production
of an ”ah." During the second task you W111 be asked to count during gentle
insufflation. Again, pressure measurements will be recorded.

6. Risks, discomforts and inconveniences of the research.

There are no known risks associated with air insufflation. Some patients may
experience gastric filling with air and experience mild discomfort associated with air in
the stomach.

7. Measures to be taken to minimize risks, discomforts and inconveniences.
At any time should you complain of discomfort the procedure will be terminated. You

will be tested in the University of Michigan Hospital, 0tolaryngology Clinic. A
physician will be available at any time within the clinic, should assistance be needed.

8. Expected benefits to subjects or to others.

It is important that the selected method of voice restoration attempt to obtain rapid and
effective vocal rehabilitation. This investigation will provide information as to the
value of tracheojejunal voice for alaryngeal speech purposes. The results of this study
may also have ramifications to surgical decisions.

9. Appropriate alternative procedures.

This assessment procedure is done routinely as part of the preoperative voice
examination. If you decide that you do not wish to participate in this study, you will
still receive this examination as part of your preoperative assessment.

75

 

76

10. Costs to subject or insurance carrier resulting from participation in the
study.

I understand that my participation in this research project will not involve any
additional costs to me or my health care insurer.

ll. Payments to subject for participating in the study.

Subject participation will be strictly voluntary. Subjects will not be paid for research
participation.
12. Confidentiality of information collected.

I understand that I will not be identified in any reports on this study. The records will
be kept confidential to the extent provided by Federal, State, and local low.

13. Management of physical injury.

You will be tested in the 0tolaryngology Clinic. In the event of a physical injury,
which may result from research procedures, the University of Michigan will provide
first-aid medical treatment. Additional medical treatment will be provided in
accordance with the determine by the University of its responsibility to provide such
treatment. However, the University does not provide compensation to a person who is
injured while participating as a subject in research.

14. Availability of further information.

If significant new knowledge is obtained during the course of this research which may
related to your willingness to continue participation, you will be informed of this
knowledge. To find out more about any aspect of this study, including your rights,
olu may contact the persons whose names, addresses, and telephone numbers appear
ow:

Jan S. Lewin, M.S.
Department of Otolaryngology
(313) 936-8013

If you have any questions or concerns regarding your rights as a research subject, you
may also contact the Office of Patient/Staff Relations, 1500 E. Medical Center Dnve,
C246 Med Inn Building, Ann Arbor, MI 48109-0822, telephone #(313) 763-5456.

15. Voluntary nature of participation.

I understand that my participation in this project is voluntary and that I may refuse to
participate in or withdraw from the study at any time without penalty or loss of benefits
to which I may otherwise be entitled.

16. Documentation of the consent.

One copy of this document will be kept together with our research records on this
study. A second copy will be placed in your Hospital record. A third copy will be
given to you to keep.

17. Consent of the subject.

I have read the information given above. I understand the meaning of this information.
Jan S. Lewin has satisfactorily answered my questions concerning the study. I hearby
consent to participate in the study.

18. Names and signatures of consenting persons and witnesses.

 

 

Signature of Participant Date Jan S. Lewin, M.S., CCC-SLP Date
Witness

 

APPENDIX C

THE UNIVERSITY OF MICHIGAN MEDICAL CENTER INFORMED CONSENT
FOR EXPERIMENTAL PROCEDURE

LIST OF REFERENCES

LIST OF REFERENCES

Amatsu, M. (1980). A one stage surgical technique for
postlaryngectomy voice rehabilitation. Lenynggeeepe SQ,
1378-1386.

Arslan, M. & Serafini, I. (1972). Restoration of laryngeal
functions after total laryngectomy. Report on the first 25
cases. Lar n osco e, SS, 1349-1360, .

Asai, R. (1965). Asai's New Voice grogucbion Methog: A
Substitution for Human Speech. Paper presented at the 8th
International Congress of 0torhinolaryngology, Tokyo, Japan.

Atkinson, S.I. (1984). Tracheo-esophageal puncture -
literature review and case reports. Journal of

Otolanyngoiogy, 13(5). 315-317.

Bavry, J.L. (1984, 1987, 1991). Stat-power Statistical

Design Analysis System. In J.L. Barry, Usez's Guide, i991,
Seeeng_Egibien. Chicago: Scientific Software, Inc.

Baugh, R.F., Baker, S.R., & Lewin, J.S. (1988) Surgical
treatment of pharyngoesophageal spasm. Lezyngoseope,
2S(10), 1124-1146.

Baugh, R.F., Lewin, J.S., & Baker, S.R. (1987).
Preoperative assessment of tracheoesophageal speech.
La n s o , 21(4), 461-466.

Baugh, R.F., Lewin, J.S., & Baker, S.R. (1990). Vocal
rehabilitation of tracheoesophageal speech failures. Head

and_Negk. 12(1). 69-73.

Biel, M.A. & Maisel, R.H. (1987). Free jejunal autograft
reconstruction of the pharyngoesophagus. Review of a 10-

year experience. Qtoianyngeiggy Heeg end Neck Suzgeny,
21(4), 369-375.

Billroth, T. In Gussenbauer, C. (1874). Ueber die erste
durch Th. Billroth an Menschen ausgeffihrte Kehlkopf—
Exstirpation, und die Anwendung eines kfinstlichen
Kehlkopfes.

77

78

Blom, E.D., Singer, M.I. & Hamaker, R.C. (1986). A
prospective study of tracheoesophageal speech. Anchives e;

0tolaryngology Head ang Neck Suzgezy, 11;, 440-447.

Blom, E.D., Singer, M.I., & Hamaker, R.C. (1985). An
improved esophageal insufflation test. Anenives gf

Qtolaryngology, iii, 211-212.

Bradford, C.R., Esclamado, R.H., & Carroll, W.R. (1992).
Monitoring of revascularized jejunal autografts. AIQDIYES

e; Otelazyngoiegy Sean and Neek Sungeny, iiS(10), 1042-1044.

Calcaterra, T.C. & Jafek, B.W. (1971). Tracheo-esophageal
shunt for speech rehabilitation after total laryngectomy.
A chives of 0tolar n olo , 21(2), 124-128.

Cheesman, A.D., Knight, J., McIvor, J., & Perry, A. (1986).
Tracheo-esophagheal "puncture:" An assessment technique for
failed oesophageal speakers. Journ of r n o an

Obology, iog(2), 191-199.

Coleman, J.J. (1993). Reconstruction of the pharynx and
cervical esophagus. In J.T. Johnson and M.S. Didolkar

(Eds.), Head end Neck Cancen: Voiune 111. (pp. 85-88).

Amsterdam: Excerpta Medica.

Conley, J.J. (1959). Vocal rehabilitation by autogenous
vein graft. nn 3 of Otolo ' o ,
SS, 990-995.

Conley, J.J., De Amesti, F., Pierce, M.K. (1958). A new
surgical technique for vocal rehabilitation after

laryngectomy. Archives of Otolaryngoiegy, SS, 499-505.

de Vries, E., Stein, D.W., Johnson, J.T., Wagner, R.L.,
Schusterman, M.A., Myers, E.N., Shestak, R., Jones, N.F.,
and Williams, S. (1989). Hypopharyngeal reconstruction: A
comparison of two alternatives. Lanyngoseope, 22, 614-617.

Denk, D.M., Grasl, M.C., Frank, F., Deutsch, W., &
Ehrenberger, K. (1992). Surgical voice rehabilitation after
laryngopharyngectomy. Functional results of tracheo-
hypopharyngeal shunts by jejunal transplantation. Anubiyee

of Otonnineiaryngoiogy, 242(5), 248-252.

Donegan, J.O., Gluckman, J.L., & Singh, J. (1981).
Limitations of the Blom-Singer technique for voice

restoration. Annals of Otology Rhinolegy and Laryngology,
SQ, 492-494.

Ehrenberger, R., Wicke, W., Piza, H., Roka, R., Grasl, M., &
Swoboda, H. (1985). Jejunal grafts for reconstructing a
phonatory neoglottis in laryngectomized patients. Anchiyes

of Otgnbinoiezyngelogy, 252, 217-223.

79

Fisher, S.R., Cole, B.C., Meyers, W.C., 8 Seigler, H.F.
(1985). Pharyngoesophageal reconstruction using free

jejunal interposition grafts. Archixes_2f_9tglarxnselggx.
ili, 747-752.

Gluckman, J.L., McDonough, J., Donegan, J.O., Crissman, J.,
Fullen, W., 8 Shumrick, D. (1985). The free jejunal graft
in head and neck reconstruction. Leryngoseope, 2;, 1887-
1895.

Gluckman, J.L., Weissler, M.C., 8 McCafferty, G. (1987).
Partial vs. total esophagectomy for advanced carcinoma of
the hypopharynx. rch ves o Oto n olo Head and Neck

Subgery, 113, 69-72.

Grasl, M.C. (1993). Free jejunal autografts for immediate
voice restoration following laryngectomy. La sco e,
103, 96-100.

Gullane, P., Havas, T., Patterson, A., Todd, T., 8 Boyd, B.
(1987). Pharyngeal reconstruction: Current controversies.

Soubnei of 0tolaryngology, iS(3), 169-173.

Hamaker, R.C., Singer, M.I., Blom. E.D., 8 Daniels, H.A.
(1985). Primary voice restoration at laryngectomy.

Anchives gf 0toleryngoiogy, iii, 182-186.

Handl-Zeller, L., Hohenberg, G., Frasl, M., Stanek, C.,
Piza, H., Roka, R., 8 Ehrenberger, K. (1992). Radiation of
jejunal interposition in T3-T4 upper aerodigestive tumors.

Eurepean Jeunnai of geneen, ;S(1), 50-53.

Haughey, B.H. 8 Forsen, J.W. (1992). Free jejunal graft

effects on longitudinal myotomy. Anneie_pfi_gbgiegy
Rhinoiegy end Lenyngology, 10;, 333-338.

Henley, J. 8 Souliere, C. (1986). Tracheoesophageal speech
failure in the laryngectomee: The role of constrictor
myotomy. Laryngoseope, SS, 1016-1020.

Holinger, P.H. (1975). A century of progress of
laryngectomies in the northern hemisphere. banyngoscope,
ee, 322-332.

Jacob, S.W., Francone, C.A., 8 Lossow, W.J. (1982).
Stnucture and Function in Man. Philadelphia, W.B. Saunders

Company.

Johns, M.E. 8 Cantrell, R.W. (1981). Voice restoration of
the total laryngectomy patient, the Singer-Blom technique.

0tolaryngoiogy Heed end Neck Surgery, S2(1), 82-86.

 

80

Juarbe, C., Shemen, L., Wang, R., Anand, V., Eberle, R.,
Sirovatka, A., Malanaphy, K., 8 Klatsky, I. (1989).
Tracheoesophageal puncture for voice restoration after

extended laryngopharyngectomy. Arehives of Otolaryngoiogy
Head and Neck Surgery 115, 356-359.

Juarbe, C., Shemen, L., Eberle, R., Klatsky, I., 8 Fox, M.
(1986). Primary tracheoesophageal puncture for voice

restoration. American Jouznai of Sungery, isz, 464-466.

Karlan, M.S. (1968). Two stage Asai laryngectomy utilizing

a modified Tucker valve. American Journal of Sungery,
i1§(4), 597-599.

Kerlin, P., McCafferty, G., Robinson, D.W., 8 Theile, D.
(1986). Function of a free jejunal "conduit" graft in the

cervical esophagus. Gastnoenbezeiogy, 2Q, 1956-1963.

Kinishi, M. Amatsu, M., Tahara, S., 8 Makino, K. (1991).
Primary tracheojejunal shunt operation for voice restoration
following pharyngolaryngoesophagectomy. Annals of Otoiggy
Rhinology end Lanyngelogy, iuu, 435-438.

Komorn, R.M., Weyeer, J.S, Sisson, R.B., 8 Malone, P.E.
(1973). Vocal rehabilitation with a tracheoesophageal

shunt. Anubiyes pf Otolaryngoiegy, 21, 303-305.

Lau, W.F., Wei, W.I., Ho, C.M., 8 Lam, K.H. (1988).
Immediate tracheoesophageal puncture for voice restoration
in laryngopharyngeal resection. Am 'can nal o
Surgery, ibb, 269-272.

Leipzig, B., Griffiths, C.M., 8 Shea, J.P. (1980).
Neoglottic reconstruction following total laryngectomy: The
Galveston experience. Annals_2f_9tolosx_8hinologx_and
Laryngology, 19S, 77-78.

Lewin, J. S. Baugh, R. F., 8 Baker, 8. R. (1987). An objective
method for prediction of tracheoesophageal speech

gigdgegion. lour__l_2f_§neech_and_flear_ng_nisorders 52(3).

Maniglia, A.J. (1982). Vocal rehabilitation after total
laryngectomy: A flexible fiberoptic endoscopic technique
for tracheoesophageal fistula. Lazyngoscope, S3, 1437-1439.

Maniglia, A.J., Lundy, D.S., Casiano, R.C., 8 Swim, S.C.
(1989). Speech restoration and complications of primary
versus secondary tracheoesophageal puncture following total

laryngectomy. Larxnsoscone. 22. 489-491-

 

81

Maves, M.D. 8 Lingeman, R.E. (1982). Primary vocal
rehabilitation using the Blom-Singer and Pange voice
prosthesis. o ' o o ,
22, 458-460.

McGrail, J.S. 8 Oldfield, D.L. (1971). One-stage operation
for vocal rehabilitation at laryngectomy. Transamerican

neegeny pf Optnainoiogy and Otelabyngoiegy, 1S(3), 510-512.

Medina, J.E., Nance, A., Burns, L., 8 Overton, R. (1987).
Voice restoration after total laryngopharyngectomy and
cervical esophagectomy using the duckbill prosthesis.

American Journal of Surgery, 154, 407-410.

Mendelsohn, M., Morris, M., 8 Gallagher, R. (1993). A
comparative study of speech after total laryngectomy and

total laryngopharyngectomy. nncnives pf Obelagyngoiogy Heeg
and Neck Surgery, i19(5), 508-510.

Miller, A.H. (1967). First experiences with the Asai
Technique for vocal rehabilitation after total laryngectomy.
Annals of Otology, 1S, 829.

Montgomery, W.W. 8 Toohill, R.J. (1968). Voice
rehabilitation after laryngectomy. Anchives of

Otoiaryngeiogy, SS, 499-506.

Ogura, J.H. 8 Thawley, S.E. (1979). Complications of
laryngeal surgery. In J.J. Conley (Ed.), Conpiieabiens pf
Head and Neck Sungery (pp. 246-268), Philadelphia: W.B.
Saunders.

Panje, W.R. (1981). Prosthetic voice rehabilitation

following laryngectomy. Annais o: Otology Rhinplogy and
LQIYDQQIQQY: 22(2), 116‘120-

Perry, A., Cheesman, A.D., McIvor, J., 8 Charlton, R.
(1987). British experience of surgical voice restoration
techniques as a secondary procedure following total

laryngectomy. W. 191. 155-
163.

Salamoun, W., Swartz, W.M., Johnston, J.T., Johnes, N.F.,
Myers, E.N., Schramm, V.L., Jr., 8 Wagner, R.L. (1987).
Free jejunal transfer for reconstruction of the

laryngopharynx. Obolenynggiogy Heag ang Neck Surgery, SS,
149-150.

Schechter, G.L., Baker, J.W., 8 Gilbert, D.A. (1987).
Functional evaluation of pharyngoesophageal reconstructive
techniques. es 0 t o o ead and e

Sungery, 113, 40-44.

 

82

Schuller, D.E., Jarrow, J.E., Kelly, D.R., 8 Miglets, A.W.
(1983). Prognostic factors affecting the success of

duckbill vocal restoration. Obeiezyngpiegy flead eng Neek
WI 21: 396-398 0

Seeman, M. (1967). Rehabilitation of laryngectomized
subjects. nebe Qbo-ienyngolegica, S5, 235-241.

Seidenberg, B., Rosenak, S.S., Hurwitt, E.S., 8 Som, M.L.
(1959). Immediate reconstruction of the cervical esophagus
by a revascularized isolated jejunal segment. Annals of

§Q£9§£Yr 1&2: 162-171-

Shah, J.P. (1993). Cancer of the hypopharynx:
introduction. In J.T. Johnson and M.S. Didolkar (Eds.),

n ° V II . (pp. 83-84). Amsterdam:
Excerpta Medica.

Shumrick, D.A. 8 Savoury, L.W. (1988). Recent advances in
laryngo, pharyngo, esophageal reconstruction. ncta Oto-

ienyngologica, Supplement SSS, 190-193.

Siegel, S. (1956). Nonparametric Sbebistics for the
Sehavioral Scieneee. New York: McGraw-Hill Book Company.

Simpson, I.C., Smith, J.C.S., 8 Gordon, M.T. (1972).
Laryngectomy: The influence of muscle reconstruction on the
mechanism of oesophageal voice production. Sounnal o:

LQIYDQQlQQY_§n§_Q§QlQQYI §§7 951-

Singer, M.I. 8 Blom, E.D. (1980). An endoscopic technique
for restoration of voice after laryngectomy. Anneie_gg
Otolo Rhinolo and La n ol , SS, 529-533.

Singer, M.I. 8 Blom, E.D. (1981). Selective myotomy for
voice restoration after total laryngectomy. Abcnives gt

Qtolaryngology, i0], 670-673.

Singer, M.I., Blom, E.D., 8 Hamaker, R.C. (1986).
Pharyngeal plexus neurectomy for alaryngeal speech
rehabilitation. L s o e, SS, 50-54.

Sisson, G.A., McConnel, M.S, Logemann, J.A., 8 Yeh, S., Jr.
(1975). Voice rehabilitation after laryngectomy. nneniyee

of Otoianyngeiogy, 191, 178-181.

Sisson, G.A., Bytell, D.E., Becker, S.P., McConnel, F.M., 8
Singer, M.I. (1978). Total laryngectomy and reconstruction
of a pseudoglottis: Problems and complications.

Laryngoscope, SS, 639.

83

Smitheran, J. R. 8 Hixon, T. J. (1981). A clinical method
for estimating laryngeal airway resistance during vowel

production. WW
138-146.

Staffieri, M. (1969). Laringectomia totalie con
reconstruzione di glottide fonatoria. ourn ti
S c e M C r a 8 na.

Surkin, M.I., Lawson, W., 8 Biller, H.F. (1984). Analysis
of the methods of pharyngoesophageal reconstruction. Heed

ang neck Surgery, May/June.

Taub, S. (1975). Bypass voice prosthesis for vocal
rehabilitation of laryngectomies. Annals o; Oboiogy

Rhinolpgy and Leryngoiggy, S5, 45-48.

Taub, S. (1981). Air-bypass voice prosthesis for vocal

rehabilitation of laryngectomees. E§I_NQ§§_QEQ;IDI2§L
Journal, SS, 42-54.

Taub, S. 8 Spiro, R.H. (1972). Vocal rehabilitation of
laryngectomees: Preliminary report of a new technique.

American Journal of Surgery, 124, 87-90.

Urken, M.L., Weinberg, H., Vickery, C., Buchbinder, D., 8
Biller, H.F. (1989). Free flap design in head and neck
reconstruction to achieve an external segment of monitoring.

Ancnives of Otoieryngology Sead and Neck Suzgeny, 11S, 1447-

1453.

VandenBerg, J. 8 Moolenaar-Bijl, A.J. (1959).
Cricopharyngeal sphincter, pitch, intensity, and fluency in

oesophageal speech. Ezeebiee Obo-rhino-ieryngelegie, Si,
298-315.

Weinberg, B. (1982). Airway resistance of the voice button.
Ancbives of Oboienynggiogy, i08, 498-500.

Weinberg, B., Horii, Y., Blom, E., 8 Singer M. (1982).
Airway resistance during esophageal phonation. ourn of

Speech end Heaning Disorders, 51(2), 194-199.

Wenig, B.L., Mullooly, V., Levy, J., 8 Abramson, A.L.
(1989). Voice restoration following laryngectomy: The role
of primary versus secondary tracheoesophageal puncture.

Annals of Otology Sbinoiogy and Laryngoiogx 28. 70-73.

Wetmore, S.J., Krueger, K., Wesson, K., 8 Blessing, M.L.
(1985). Long-term results of the Blom-Singer Speech

Rehabilitation Procedure. Anchives pf Qtoienyngoiogy, 1 1,
106-109.

 

84

Wood, B.G., Tucker, H.M., Rusner, M.G., 8 Levine, H.L.
(1981). Tracheoesophageal puncture for alaryngeal voice

restoration. Annais of Otelogy Sninoiogy and Lanyngology,

29. 492-494.

Zeismann, M., Boyd, B., Manktelow, R.T., 8 Rosen, 1.8.
(1989). Speaking jejunum after laryngopharyngectomy with
neoglottic and neopharyngeal reconstruction. Anenican

Seurnei pf Surgery, 158, 321-323.

‘lllllllllllllllllll“