206 / 220
Likewise, other prior research has reported conflict-
ing results. A European case-control study of 1,774
HNSCC cases found no association between heartburn
and specific HNSCC tumor sites, except for an inverse
association with hypopharyngeal cancer (OR
5
0.64;
95% CI 0.44, 0.93).
13
Another epidemiologic study (1,303
cases) reported that ever-smoker/ever-drinker HNSCC
subjects were not more likely to have had a history of
GERD than never-smoker/never-drinker HNSCC
patients.
14
A small case-control study (120 cases) exam-
ined associations between
H. pylori
infection, a cause of
GERD, and odds of laryngopharyngeal cancer but found
no association (OR
5
1.53; 95% CI 0.69, 3.41).
15
Studies of the etiologic role of GERD in laryngeal
cancer have arrived at different conclusions. A meta-
analysis concluded that GERD was associated with an
increased odds of laryngeal cancer (OR
5
2.21; 95% CI
1.53, 3.19) but not pharyngeal cancer, although the
pharynx OR was extremely imprecise.
16
A literature
review that was not a meta-analysis could not conclude
that GERD caused laryngeal cancer, but noted that con-
founding by alcohol and tobacco consumption were inad-
equately controlled.
17
A large case-control study
conducted in the Veterans Health Administration system
(14,449 cases) found no association between GERD and
laryngeal cancer (OR
5
1.01; 95% CI 0.92, 1.12).
18
An important strength of our study was examina-
tion of two different measures of GERD exposure: 1)
self-reported history of symptoms and medical diagnosis
and 2) development of HNSCC. A medical diagnosis of
GERD is more likely to indicate substantial GERD mor-
bidity than a self-reported history of having had fre-
quent heartburn, resulting in less misclassification.
Second, our analysis was based on a large, population-
based case-control study, making it representative of a
well-defined source population and increasing the preci-
sion of the effect estimates. Third, CHANCE has
detailed information on alcohol and tobacco consump-
tion, important causes of HNSCC not well measured in
a number of previous studies that examined relation-
ships between GERD and HNSCC.
17
This enabled us to
appropriately control for the effects of tobacco and
alcohol.
In terms of limitations, our assessment of GERD by
self-report was not as accurate as an objective measure-
ment such as pH monitoring would be,
5
even when the
self-reported measure was medical diagnosis of GERD
rather than self-assessment of GERD symptoms. Medi-
cal diagnosis of GERD was ascertained by asking sub-
jects whether they were ever diagnosed with GERD by a
doctor rather than abstracting the information from
medical records. Furthermore, even among self-reported
measures of GERD, our simple one-question assessment
might not be as accurate or reliable as validated multi-
question instruments such as the Reflux Symptom
Index.
19
However, alternative measures of self-reported
GERD were not available in CHANCE. There is the pos-
sibility of misclassification of the history of GERD, espe-
cially if subjects are not aware of the criteria for
frequency and severity of symptoms used to diagnose
GERD.
20
For example, subjects who are not aware of
frequency criteria and assume that their symptoms do
not occur frequently enough to warrant being considered
a medical diagnosis could falsely report not having had
GERD exposure, thereby possibly attenuating estimates
of an association between GERD and HNSCC.
Future research on GERD and HNSCC must con-
sider the differing study designs and inconsistent find-
ings of results reported to date. A larger study may be
beneficial to further elucidate this association. Such a
study would need to provide adequate control for tobacco
and alcohol consumption as well as obesity, as was done
here.
5
It would be informative to compare the effect of
GERD when measured by self-report, medical diagnosis
as ascertained by medical records, and by pH monitoring
or another objective measurement. Multiple measures of
self-reported GERD could be used for purposes of com-
parison, including questionnaires such as the Reflux
Symptom Index.
19
Because a few of our site-specific
associations suggested greater risk, estimates of the
association of GERD with HNSCC should be conducted
for both overall HNSCC as well as individual tumor
sites, as was done here.
CONCLUSION
In summary, we find no general pattern of associa-
tion between GERD and the development of HNSCC.
Subgroup analysis of subjects who were neither heavy
drinkers nor heavy smokers does not show an associa-
tion between GERD and the development of laryngo-
pharyngeal cancer, a finding that conflicts with prior
research. However, whereas none of our associations is
statistically significant, the patterns of some point esti-
mates, such as the larynx result, are suggestive and
should be further investigated in future larger studies.
Such additional work would help to resolve the inconsis-
tencies observed in this literature.
BIBLIOGRAPHY
1. Spechler SJ. Clinical manifestations and esophageal complications of
GERD.
Am J Med Sci
2003;326:279–284.
2. Williams JL. Gastroesophageal reflux disease: clinical manifestations.
Gas-
troenterol Nurs
2003;26:195–200.
3. Kapoor H, Agrawal DK, Mittal SK. Barrett’s esophagus: recent insights
into pathogenesis and cellular ontogeny.
Transl Res
2015;166:28–40. doi:
10.1016/j.trsl.2015.01.009.
4. Wang RH. From reflux esophagitis to Barrett’s esophagus and esophageal
adenocarcinoma.
World J Gastroenterol
2015;21:5210–5219.
5. Herbella FA, Neto SP, Santoro IL, Figueiredo LC. Gastroesophageal reflux
disease and nonesophageal cancer.
World J Gastroenterol
2015;21:815–
819.
6. Langevin SM, Michaud DS, Marsit CJ, et al. Gastric reflux is an inde-
pendent risk factor for laryngopharyngeal carcinoma.
Cancer Epidemiol
Biomarkers Prev
2013;22:1061–1068.
7. Johnston N, Dettmar PW, Strugala V, Allen JE, Chan WW. Laryngophar-
yngeal reflux and GERD.
Ann N Y Acad Sci
2013;1300:71–79.
8. Kuo CL, Chen YT, Shiao AS, Lien CF, Wang SJ. Acid reflux and head and
neck cancer risk: A nationwide registry over 13 years.
Auris Nasus Lar-
ynx
2015;42:401–405. doi:
10.1016/j.anl.2015.03.008.9. Johnston N, Yan JC, Hoekzema CR, et al. Pepsin promotes proliferation of
laryngeal and pharyngeal epithelial cells.
Laryngoscope
2012;122:1317–
1325.
10. Kelly EA, Samuels TL, Johnston N. Chronic pepsin exposure pro-
motes anchorage-independent growth and migration of a hypophar-
yngeal squamous cell line.
Otolaryngol Head Neck Surg
2014;150:
618–624.
11. Divaris K, Olshan AF, Smith J, et al. Oral health and risk for head and
neck squamous cell carcinoma: the Carolina Head and Neck Cancer
Study.
Cancer Causes Control
2010;21:567–575.
12. Stingone JA, Funkhouser WK, Weissler MC, Bell ME, Olshan AF. Racial
differences in the relationship between tobacco, alcohol, and squamous
Laryngoscope 126: May 2016
Busch et al.: GERD and Head and Neck Cancer
186