Otolaryngology–Head and Neck Surgery (2010) 143, 375-378

ORIGINAL RESEARCH–HEAD AND NECK SURGERY

Esophageal pathology in patients after treatment

for head and neck cancer
D. Gregory Farwell, MD, Catherine J. Rees, MD, Debbie A. Mouadeb, MD,
Jacqueline Allen, MD, Allen M. Chen, MD, Danny J. Enepekides, MD, and
Peter C. Belafsky, MD, PhD, Sacramento, CA; Winston-Salem, NC; and Toronto,
Ontario, Canada
Sponsorships or competing interests that may be relevant to con-
tent are disclosed at the end of this article.
ABSTRACT
OBJECTIVE: To determine the prevalence of esophageal pa-
thology following treatment for primary head and neck cancer
(HNCA).
STUDY DESIGN: Case series with planned data collection.
SETTING: Academic medical practice.
SUBJECTS AND METHODS: Subjects comprised HNCA
survivors. Esophagoscopy was prospectively performed on 100
patients at least three months after treatment for HNCA. Patient
demographics including cancer stage, cancer treatment, use of
reflux medications, symptoms surveys, and esophageal findings
were prospectively determined.
RESULTS: The mean age of the cohort was 64 (⫾ 10) years; 75
percent were male. The mean time between the end of treatment
and endoscopy was 40 (⫾ 51) months. Eighty-one percent of
HNCA was advanced stage (3 or 4). The distribution of site of the
primary HNCA was as follows: oropharynx (38%), larynx (33%),
oral cavity (17%), unknown primary (10%), hypopharynx (1%),
and nasopharynx (1%). Treatment modalities included surgery
alone (15%), surgery with radiation (34%), radiation alone (6%),
chemoradiation alone (24%), and chemoradiation with surgery
(20%). The findings on esophagoscopy included peptic esophagitis
(63%), stricture (23%), candidiasis (9%), Barrett metaplasia (8%),
gastritis (4%), and carcinoma (4%). Only 13 percent had a normal
esophagoscopy.
CONCLUSION: Esophageal pathology is extremely common in
patients treated for HNCA. These findings support routine esoph-
ageal screening after HNCA treatment.
© 2010 American Academy of Otolaryngology–Head and Neck
Surgery Foundation. All rights reserved.
S wallowing is an incredibly complex process involving
multiple levels of the upper aerodigestive tract. Dyspha-
gia is common after treatment for primary head and neck
cancer (HNCA). Recently, there has been a renewed focus
on patients’ swallowing function. There are several poten-
tial causes for swallowing dysfunction in HNCA patients.
Direct injury to tissues within the surgical or radiation field,
including the cervical esophagus, is implicated as a likely
cause. Similarly, delayed sequelae, such as fibrosis, stric-
ture, and neural dyscoordination, may play a role in the
swallowing dysfunction these patients experience.
Oropharyngeal dysfunction and aspiration have been
demonstrated in up to 59 percent of patients treated for
HNCA with chemoradiotherapy.1 Stricture of the hypophar-
ynx and cervical esophagus are relatively common,2 with
some studies demonstrating a rate as high as 11 percent after
radiation for HNCA3 and 13 to 21 percent after chemora-
diation for HNCA.4,5 However, there is relatively little
information available about the prevalence of esophageal
pathology in HNCA survivors. The purpose of this study
was to evaluate the esophagus endoscopically in a large
series of patients having undergone treatment for HNCA at
a single institution.
Methods
This prospective study was approved by the University of
California Davis School of Medicine Institutional Review
Board. One hundred consecutive subjects were recruited
from routine tumor surveillance visits if they had 1) a
history of HNCA and 2) finished all treatment at least three
months prior to recruitment. Subjects were asked if they
perceived any problem with their swallowing and were
classified as symptomatic or asymptomatic. All participants
were given the 10-item Eating Assessment Tool (EAT-10).
The EAT-10 is a validated, self-administered, disease-spe-
cific outcome instrument for dysphagia. Subjects were of-
fered evaluation of their esophagus via transnasal esopha-
goscopy regardless of the presence of dysphagia symptoms
present on direct questioning.6 Demographic data and tumor
details were collected from the medical chart, including:
gender, age, primary cancer site, cancer stage, time since
treatment, and treatment modality. The American Joint

Received March 3, 2010; revised April 20, 2010; accepted May 6, 2010.

0194-5998/$36.00 © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
doi:10.1016/j.otohns.2010.05.006
376 Otolaryngology–Head and Neck Surgery, Vol 143, No 3, September 2010
Committee on Cancer tumor node metastasis staging system
was used for tumor staging.
Transnasal esophagoscopy (TNE) with topical anesthesia
was performed in the office setting for all subjects. Subjects
were placed in the seated position, and the nose was anes-
thetized and decongested with topical four percent lido-
caine/oxymetazoline. The Pentax EE-1580K (KayPentax,
Lincoln Park, NJ) 5.1-mm transnasal esophagoscope was
passed through the more patent nare and advanced into the
hypopharynx. After visualization of the larynx and pharynx,
the endoscope was advanced through the esophagus and
into the stomach, to include a retroflexed view of the gastric
cardia and fundus. The squamocolumnar junction was ex-
amined for signs of esophagitis, hiatal hernia, Barrett esoph-
agus, and neoplasm. The rest of the esophagus was exam-
ined as the endoscope was withdrawn. If indicated, 1.8-mm
cup forceps were used through the working channel of the
endoscope to obtain mucosal biopsies. The Los Angeles
Classification system was used to characterize findings of
erosive esophagitis.7 “Grade 0” esophagitis was used to
describe nonerosive edema and inflammation of the distal
esophagus, also known as grade M esophagitis in the Jap-
anese Classification system.8,9
Pathology was obtained on all histologic specimens bi-
opsied at the time of endoscopy. Patient clinical informa-
tion, and pathologic and esophagoscopy findings were en-
tered into SPSS 17.0 for the Macintosh (SPSS, Inc.,
Chicago, IL). Categorical data were compared with the ␹2
test. The independent samples t test was utilized to evaluate
statistical significance between continuous data. A Bonfer-
roni correction was used to adjust for the evaluation
of multiple comparisons. A probability of Type I error
(␣) ⫽ 0.05 was used to ascertain statistical significance.
Results
One hundred subjects were prospectively enrolled in this
study (75% male). The mean age of the cohort was 64 years
(range 38-85 yrs). Cancer staging was available for 95
subjects (95%). Eleven percent of patients had American
Joint Committee on Cancer (AJCC) stage 1 disease, while
eight percent had stage 2, and the vast majority of the
subjects (77 of 95, 81%) had advanced, AJCC stage III or
IV disease. The primary site was definitively known in 90
subjects. The most common primary tumor site was the
oropharynx (38 of 100, 38%), followed by the larynx (33 of
100, 33%) and oral cavity (17 of 100, 17%). Less common
sites included carcinomas from unknown primary sites (10
of 100, 10%), the nasopharynx (1 of 100, 1%), and the
hypopharynx (1 of 100, 1%). Time since treatment was
available for 99 subjects, averaging 40.3 months (3-240
months). Treatment modality was available for all subjects:
35 subjects were treated with surgery and radiation therapy;
24 subjects were treated with combined chemoradiation
(without surgery); and 20 subjects had surgery in combina-
tion with chemoradiation. Fifteen subjects had surgery
alone and six subjects had radiation alone. Eighty-five sub-
jects (85%) had radiation therapy at some point in their
treatment. Fifty-nine subjects (59%) had chemotherapy as
part of their treatment regimen.
TNE was successfully performed for all 100 subjects
with topical anesthesia in the office. Biopsy was performed
in 44 of 100 cases. Overall, only 13 of 100 (13%) exami-
nations were described as normal. The most common find-
ing was peptic esophagitis (63 of 100, 63%). Esophageal
stricture was noted in 23 subjects (23%). Other findings
included esophageal candidiasis (9%), biopsy-proven Bar-
rett metaplasia (8%), gastritis (4%), and carcinoma (4%). Of
the carcinomas identified, one was in the hypopharynx, two
were in the esophagus, and one was in the gastric cardia.
Approximately one half of subjects (50 of 99) were on a
proton pump inhibitor (PPI) at the time of TNE. In the
subjects with stricture, 12 of 23 subjects were on a PPI at the
time of TNE (52%). Thirty subjects in the esophagitis group
were already being treated with a PPI (48%). There was no
association between PPI use and the presence of esophagitis
on endoscopy (P ⬎ 0.05). In the subjects with esophagitis,
26 subjects were classified as grade 0 (41%), nine as grade
A (14%), 25 as grade B (40%), and three as grade C (5%).
Of those subjects with signs of nonerosive reflux disease
(classified as grade 0, n ⫽ 26), two had candida, one had
Barrett metaplasia, and one had stricture. If the subjects
with the finding of grade 0 esophagitis only are included in
the “normal” group, 35 of 100 TNE examinations were
normal (65% abnormal).
Twenty-two subjects (22%) reported that they had no
problem with swallowing (asymptomatic). Of these 22 sub-
jects who denied symptoms, only four had normal esopha-
goscopy. Sixteen of these subjects (73%) had esophagitis
(eight grade 0, two grade A, five grade B, one grade C), two
had candidiasis, and one had stricture. The mean (SD)
EAT-10 for patients with a normal endoscopy was 17.5
(15.9) compared to a mean (SD) EAT-10 of 18.6 (12.5) in
persons with an abnormal examination. The severity of
patients’ symptoms as documented by the EAT-10 was not
predictive of a normal endoscopy (P ⬎ 0.05). The EAT-10,
however, was predictive of the presence of an esophageal
stricture, as the mean (SD) of patients with a stenosis was
29.0 (8.8) in comparison to an EAT-10 of 15.5 (12.4) in
persons without an esophageal stricture (P ⬍ 0.001).
Fifteen subjects had no exposure to radiation therapy
during their treatment course. Only two of these subjects
(13%) had normal esophageal examinations. Three of the 15
had grade B esophagitis, and six had grade 0 esophagitis.
Other findings included stricture (2 subjects), Barrett esoph-
agus (1 subject), candida esophagitis (1 subject), and hypo-
pharyngeal carcinoma (1 subject).
The stage of disease was not associated with the presence
of a normal esophagoscopy. Of patients with early-stage
disease (stages 1 and 2), 15.8 percent had a normal endos-
copy, in comparison to 14.9 percent of persons with late-
stage (stages 3 and 4) disease (P ⬎ 0.05). Stage had no
association with the prevalence of esophagitis or Barrett
Farwell et al Esophageal pathology in patients after treatment . . .
metaplasia (P ⬎ 0.05), but advanced stage was associated
with a diagnosis of candida esophagitis, as all persons with
candida (nine) had late-stage disease (P ⬍ 0.05). Although
insufficient numbers precluded a statistical comparison,
late-stage disease also appeared to be associated with the
diagnosis of a second primary, as all patients with a cancer
identified on TNE had Stage 3 or 4 disease.
Discussion
Many patients complain of swallowing dysfunction after
treatment for HNCA, and in current practice most do not
have an endoscopic evaluation of the esophagus. The results
of this study indicate that esophageal pathology is common
in this population, with 87 percent having some abnormality
on esophagoscopy.
Peptic esophagitis was the most common finding in this
study. The ability of the endoscopist to identify nonerosive
mucosal changes (grade 0 esophagitis) has been questioned
in the literature.8,9 The endoscopies in this study were per-
formed by two physicians experienced in TNE (C.J.R. and
P.C.B.), limiting possible variability in this diagnosis. Nev-
ertheless, it is somewhat controversial how significant this
observation may be. When cases of isolated grade 0 esoph-
agitis are excluded, the rate of abnormal esophagoscopy in
this population is still 65 percent. The clinical importance of
low-grade peptic esophagitis (grades A and B) is also un-
known. However, this high prevalence of low-grade esoph-
agitis must be considered in the context of the high rate of
swallowing dysfunction in this patient population, and,
therefore, this finding may be more important in this group.
Esophageal changes after treatment for HNCA are likely
multifactorial and related to changes in bacterial flora, mu-
cosal injury from chemoradiation therapy, fibrosis, and/or
xerostomia and its resultant change in pH. Salivary bicar-
bonate is important for neutralizing refluxed acid in the
esophagus. In the setting of post-radiation xerostomia, this
protective mechanism is lost, possibly predisposing these
patients to esophageal injury. As it is our practice to be
increasingly aggressive in the prophylactic use of PPIs in
our patients, it was interesting to note that approximately
half of the subjects who had peptic esophagitis or stricture
were already on PPI therapy for reflux. We would have
expected to see less esophagitis in the group treated with
PPI, but the use of a PPI was not associated with the
endoscopic diagnosis of esophagitis in this cohort (P ⬎
0.05). It is unclear if the severity of the esophagitis would
have been worse had they not been on the PPI.
Surprisingly, radiation exposure was not predictive of
esophageal pathology in this group (P ⬎ 0.05). Of the small
subset of 15 subjects who had never been treated with
radiation, only two had normal examinations (eight normal
examinations, if grade 0 esophagitis is excluded). These
data suggest that altered swallowing anatomy secondary to
surgery or preexisting esophageal disease may also contrib-
ute to risk for pathology after treatment. Many of these
patients have had chronic tobacco and/or ethanol exposure.
377
It is possible that the damage from these carcinogens pre-
disposes these patients to chronic esophageal pathology.
Four carcinomas were identified in this series, and only
one (hypopharynx) would have been found on a routine
laryngoscopy. The other three carcinomas were located in
the esophagus (two) and gastric cardia (one). In a pooled
analysis of 13 cancer registries, the standardized incidence
rate of second primary carcinoma for all patients with
HNCA was 1.86, with a 20-year cumulative risk of 36
percent.10 Most of the second primaries in that study were
lung cancers. In our series, 75 percent of the four carcino-
mas identified were in the esophagus or stomach, suggesting
that esophagoscopy may be a useful adjunct to routine
oncologic surveillance in HNCA survivors. All patients
who had a second cancer identified on TNE had advanced-
stage disease. This topic deserves further research.
The severity of patients’ self-reports of dysphagia was
associated with the diagnosis of an esophageal stricture but
was not associated with the prevalence of a normal esopha-
goscopy. These data suggest that the decision for routine
screening in this population cannot be based on the patient
symptoms of dysphagia alone. Severe dysphagia (EAT-
10 ⬎ 25), however, is suggestive of the presence of an
esophageal stricture (P ⬍ 0.001).
Limitations of this study include lack of pretreatment
flexible esophagoscopy. Rigid esophagoscopy is routinely
employed as part of the staging work-up of a new HNCA.
Because rigid esophagoscopy frequently does not provide a
thorough distal esophageal evaluation, erosive esophagitis,
Barrett esophagus, and distal esophageal strictures may not
be identified at the pretreatment rigid endoscopy. In a study
of pretreatment flexible esophagogastroduodenoscopy in
570 patients with oral cancer, Kesting et al identified patho-
logic findings in 30 percent.11 It is possible that many of the
findings noted in the current study were present pretreat-
ment. The high prevalence of esophageal pathology re-
ported in this investigation must be confirmed by other
institutions before these data can be generalized to other
HNCA populations. Future studies will also evaluate esoph-
ageal mucosal pathology with flexible endoscopy both be-
fore and after HNCA treatment.
A second potential limitation is the wide range of time
points from the end of treatment to the esophagoscopy.
All patients were at least three months out from the end
of their treatment, allowing us to control for and exclude
the expected acute injury from their treatment. The
extremely high prevalence of esophageal pathology
throughout the cohort suggests that patients are at risk for
esophageal pathology long after the completion of ther-
apy. The study population was extremely heterogeneous.
The wide range of time since treatment (3-240 months)
must be taken into account when generalizing these data
to different populations.
The final limitation was the relatively small (22%) num-
ber of patients that considered themselves asymptomatic.
This was very informative to our group and reinforces how
378 Otolaryngology–Head and Neck Surgery, Vol 143, No 3, September 2010
common swallowing dysfunction is in this patient popula-
tion. This prevalence makes it incumbent upon the practi-
tioner to take the initiative to ask the patient how he or she
swallows and to consider preemptive esophagoscopy in this
high-risk group.
Conclusion
Esophageal pathology is extremely common in patients
treated for HNCA. The severity of patient dysphagia symp-
toms is not predictive of a normal endoscopy, and the
decision to perform screening esophagoscopy in this popu-
lation must not be based on patient symptoms alone.
Author Information
From the Department of Otolaryngology–Head and Neck Surgery, Univer-
sity of California Davis School of Medicine (Drs. Farwell, Mouadeb,
Allen, and Belafsky), Sacramento, CA; Department of Otolaryngology,
Wake Forest University School of Medicine (Dr. Rees), Winston-Salem,
NC; Department of Radiation Oncology, University of California Davis
School of Medicine (Dr. Chen), Sacramento, CA; and Department of
Otolaryngology–Head and Neck Surgery, Sunnybrook Medical Center (Dr.
Enepekides), Toronto, Ontario, Canada.
Corresponding author: D. Gregory Farwell, MD, Department of Otolaryn-
gology–Head and Neck Surgery, University of California Davis, 2521
Stockton Blvd, Suite 7200, Sacramento, CA 95817.
E-mail address: gregory.farwell@ucdmc.ucdavis.edu.
Author Contributions
D. Gregory Farwell, hypothesis, data acquisition, manuscript preparation;
Catherine J. Rees, data acquisition, manuscript preparation; Debbie A.
Mouadeb, data acquisition, manuscript preparation; Jacqueline Allen,
data acquisition, manuscript preparation; Allen M. Chen, data acquisition,
manuscript preparation; Danny J. Enepekides, data acquisition, manu-
script preparation; Peter C. Belafsky, hypothesis, data acquisition, manu-
script preparation.
Disclosures
Competing interests: Catherine J. Rees, consultant: Boston Scientific
Corporation; Peter C. Belafsky, consultant: Olympus, Gyrus ACMI, and
Nestle; research support and consultant: Medtronic.
Sponsorships: None.
References
1. Nguyen NP, Frank C, Moltz CC, et al. Aspiration rate following
chemoradiation for head and neck cancer: an underreported occur-
rence. Radiother Oncol 2006;80:302– 6.
2. Laurell G, Kraepelien T, Mavroidis P, et al. Stricture of the proximal
esophagus in head and neck carcinoma patients after radiotherapy.
Cancer 2003;97:1693–700.
3. Nguyen NP, Smith HJ, Moltz CC, et al. Prevalence of pharyngeal and
esophageal stenosis following radiation for head and neck cancer. J
Otolaryngol Head Neck Surg 2008;37:219 –24.
4. Lawson JD, Otto K, Grist W, Johnstone PA. Frequency of esophageal
stenosis after simultaneous modulated accelerated radiation therapy
and chemotherapy for head and neck cancer. Am J Otolaryngol 2008;
29:13–9.
5. Lee WT, Akst LM, Adelstein DJ, et al. Risk factors for hypopharyn-
geal/upper esophageal stricture formation after concurrent chemora-
diation. Head Neck 2006;28:808 –12.
6. Belafsky PC, Mouadeb DA, Rees CJ, et al. Validity and reliability of
the Eating Assessment Tool (EAT-10). Ann Otol Rhinol Laryngol
2008;117:919 –24.
7. Armstrong D, Bennett JR, Blum AL, et al. The endoscopic assessment
of esophagitis: a progress report on observer agreement. Gastroenter-
ology 1996;111:85–92.
8. Dent J. Endoscopic grading of reflux oesophagitis: the past, present
and future. Best Pract Res Clin Gastroenterol 2008;22:585–99.
9. Miwa H, Yokoyama T, Hori K, et al. Interobserver agreement in
endoscopic evaluation of reflux esophagitis using a modified Los
Angeles classification incorporating grades N and M: a validation
study in a cohort of Japanese endoscopists. Dis Esophagus 2008;21:
355– 63.
10. Chuang SC, Scelo G, Tonita JM, et al. Risk of second primary cancer
among patients with head and neck cancers: a pooled analysis of 13
cancer registries. Int J Cancer 2008;123:2390 – 6.
11. Kesting MR, Schurr C, Robitzky L, et al. Results of esophagogas-
troduodenoscopy in patients with oral squamous cell carcinoma—
value of endoscopic screening: 10-year experience. J Oral Maxillofac
Surg 2009;67:1649 –55.