9L

TABLE 23.—Relative risk of cancer of the larynx for men, comparing cigar, pipe, and cigarette
smokers with nonsmokers. A summary of retrospective studies
Relative Risk Ratio and Percentage of Cases and Controls by Type of Smoking

 

 

 

          
  
  

 

 

Author (Reference) Number
Non- Cigar Pipe Total pipe Cigarette Mixed
smoker only only and cigar only
Schrek et al. (244): Relative risk .....0.0.c.cccccccesecscsceseueese 10 0 Mo eceen 23
Oy 73 Percent cases...........0...c.cccceececeese secs 14 0 To llteeeaeee 80
Controls 2.0... ccc ccecacncveeeeeeeaees 522 —- Percent controls 24 10 Wo ee, 59
Sadowsky et al. (230): Relative risk 10 2.2 2.3 3.7 4.1
CaBOS oe cee cees eee eseeceecenes 273 4 2 5 60 29
Controls 000.0... ccc cccccecceeenecens 615 Percent controls 13 3 7 53 23
Relative risk 1.0 24.6
209 “Percent cases 5 86
209 Percent controls 11 74
Relative risk 10 6.3
60 Percent cases 5 47
271 ~~“ Percent controls 24 36
Relative risk 1.0 22.0 16.0
142 Percent cases ........ wee 1 62 16
220 ~—- Percent controls 16 45 16
Relative risk ...............ccccccceees vee 10 8.7 3.2
546 Percent cases ..........0..0.0.00cccceee vee 7 78 4
713° Percent controls ‘ 39 50 7
Relative risk ........000000000.. wee 1.0 $0.2 00
207 Percent cases ..........00....00ccee tee 5 BB,
912 Percent controls 17 Gb ee.
Svoboda (261): Relative risk 10 WO
COBB occ cce ceca tcencensceceaes 205 Percent cases 3 95
Controds 0.0... cccececceeaeeeseees 320 ~— Percent controls 22 71
Stell (254): Relative risk 1.0 24
Cases oo. cece eee cee sees eneees 190 = Percent cases li v9

Controla 2.0... eee cece cena ee 190 Percent controla .................... ee 17
CASES = 239
CONTROLS = 4725

CIGARETTES
PER DAY

 

OUNCES OF ALCOHOL PER DAY

FIGURE 25.—Relative risks of larynx cancer by daily
consumption of alcohol and cigarettes for

males
* Not significant.
SOURCE: McCoy et al. (779).

respiratory tract. Several recent experiments have been performed
(23, 24, 72, 73, 125, 126, 133).

Cigarette smoke inhalation has not been found to induce laryngeal
tumors in other rodents. Such tumors have been induced, however,
by direct application of carcinogens known to be present in cigarette
smoke. This is accomplished by the intratracheal instillation of
benzo[a]pyrene in combination with particulates into hamster lungs.
In this animal model, laryngeal tumors, as well as tumors in other
parts of the respiratory tract, are induced (184, 231, 232). One study
has recently reported a synergy of alcohol and benzo[a]pyrene
injection (257).

Conclusion

1. Cigarette smoking is the major cause of laryngeal cancer in the
United States. Cigar and pipe smokers experience a risk for
laryngeal cancer similar to that of a cigarette smoker.

2.The risk of developing laryngeal cancer increases with in-
creased exposure as measured by the number of cigarettes
smoked daily as well as other dose measurements. Heavy

V7
smokers have laryngeal cancer mortality risks 20 to 30 times
greater than nonsmokers.

3. Cessation of smoking reduces the risk of laryngeal cancer
mortality compared to that of the continuing smoker. The
longer a former gijioker is off cigarettes the lower the risk.

4.Smokers who use filtered lower tar cigarettes have lower
laryngeal cancer risks than those who use unfiltered higher tar
cigarettes.

5.The use of alcohol in combination with cigarette smoking
appears to act synergistically to greatly increase the risk for
cancer of the larynx.

Oral Cancer
Introduction

Cancers of the oral cavity include malignant tumors of the lip,
tongue, salivary gland, floor of the mouth, mesopharynx, and
hypopharynx. It is estimated that in 1982 there will be 26,800 new
cases and 9,150 deaths due to these tumors (2). Males are affected
more commonly than females (by about threefold). Several authors
(29, 175) have reported geographic differences in mortality. In the
southeast, females living in urban and rural areas have mortality
rates that exceed those of northern females by 30 and 90 percent
respectively.

Cancer of the Buccal Cavity and Pharynx, Excluding Lip?

From 1950 to 1967, the age-adjusted rate remained stable at 2.8
per 100,000. The increase in the age-adjusted death rate from 2.8 to
2.9 per 100,000 between 1967 and 1968 resulted in part from changes
in coding procedures in the International Classification of Diseases.
From 1968 to 1977, the age-adjusted rate rose from 2.9 to 3.1. Total
deaths from cancer of these sites increased from 1,461 in 1950 to
8,291 in 1977.

While the age-adjusted death rate of white males fell slightly over
the study period (Figure 26), rates of white females and of males and
females of races other than white increased. The largest increases
occurred among other than white males, whose mortality rates rose
from 4.1 to 7.7 per 100,000 between 1950 and 1977. The white male to
female.mortality ratio fell gradually over the study period, from 4.09
to 2.93. In contrast, the mortality sex ratio (male/female) in the
other than white population increased from 2.56 to 3.85. The
mortality ratio of other than white males to white males increased
from 0.91 to 1.75, while the mortality ratio of other than white
females to white females decreased slightly, from 1.45 to 1.33.

® Cancer of the lip is causally associated with smoking, particularly pipe smoking. However, because this cancer
site represents so few deaths in the United States, only 163 in 1977, it is excluded from this review.

78
1965 1970 1975

CALENDAR YEARS

1960

    

NONWHITE FEMALES
1955

NONWHITE MALES

WHITE MALES
WHITE FEMALES

+
*
®
o

 

 

1950

° wo wo ~
RATES/100.000
FIGURE 26.—Age-adjusted* mortality rates for cancer of
the buccal cavity plus oral pharynx, by race
and sex, United States, 1950-1977

* This graph is age-adjusted to the U.S. population as enumerated in 1970; all rates cited within the text of the
Report, however, are adjusted to the population as enumerated in 1940.
SOURCE: National Cancer Institute (798).

79
The death rates of white males 35 to 54 years of age and of those at
least 75 years old were lower in 1977 than in 1960 (Figure 27), but
rates were higher among white males between 55 and 74 years of
age, as well as among white females in the Same age range. In
contrast, among other than white males in every 10-year age group
from 35 through 84, as well as among females between 35 and 64,
death rates were higher in 1977 than in 1960; the average increase in
mortality in these age groups was 60 percent (Figure 28).

When age-specific death rates are plotted by calendar year and age
(Figures 29 and 30), a three-dimensional graph is produced, which
can be examined from 1950 to 197 7, or from the reverse Perspective,

Squamous cell cancer is the most common histological type of ora]
cancer and comprises about 90 percent of these tumors. The 5-year
survival for cancer of the floor of the mouth, tongue, and pharynx
ranges from 25 to 45 percent.

Numerous epidemiological and experimental studies have estab.
lished a close association between smoking and oral cancer. Alcohol
has an incompletely understood but important synergistic role with
tobacco in increasing disease incidence and mortality.

Causal Significance of the Association

Consistency of the Association

More than 25 retrospective studies have examined the relation-
ship between smoking and the development of cancer of the oral
cavity (269, 276).

These studies have been done in many countries, in different
areas, and have involved diverse study methods. Almost uniformly,
they show an association between cigarettes and other forms of
tobacco use and cancer of the oral cavity and pharynx. The TNCS
study (299) and the Hawaiian Study of Five Ethnic Groups (173)
reported similar findings.

Six of the major prospective studies examined the relationship
between smoking and oral cancer. These data, presented in Table 24,
show a close association between smoking and oral cancer.

Strength of the Association

The relative risks for oral cancer among smokers were substantial-
ly greater compared with nonsmokers in the retrospective studies.
Similarly, in the prospective studies, the mortality ratios for cancer
of the oral cavity among smokers ranged from 1.22 among Japanese
females to over 13 in the U.S. Veterans and British Physicians
studies (Table 24),

A dose-response relationship was noted in many of the retrospec-
tive and prospective studies (Table 25) (64, 98, 120, 131, 276). The
American Cancer Society 25-State Study (155) reported a reduction

80
FEMALES

AGE IN YEARS (BY 5-YEAR AGE GROUPS)

1950-1956
1957-1963
1964-1970
1971-1977

+
x
o
a

 

 

78
60
4
3

AGE IN YEARS (BY 5-YEAR AGE GROUPS)

 

 

60

w
~ mn -

75

RATES/100.000

FIGURE 27.—Age-specific mortality rates for whites in the
United States for cancer of the buccal cavity

plus oral pharynx
SOURCE: National Cancer Institute (798).

81
80

 

 

 

oo
tu
a on
< ro
= a
ut >
w °
a
oOo
o
ua
3
a
Ot
wa
ul
>
'
w
o
>
Oo
wo
oa
Og
ul
omonr >
worn
nanan z
ee me on
torts Nn
om wn ud
wm or Qo
Qa a a x
ae
noun e
+* ©O8
°o
uo o wo o wo a
~ o vr n -
RATES/ 100,000
wo =~
i o
= s
= °
a
oS
uu
a
x
a
c
ld
-
1
ire)
=
o
wn
or
x
wi
>»
z
w
o
«
w o w o
é © + “

 

RATES/100,¢00

FIGURE 28.—Age-specific mortality rates for nonwhites in
the United States for cancer of the buccal

cavity plus oral pharynx
SOURCE: National Cancer Institute (798).

82
- 90

  
     
   

   
   

/G0

 

   
    

Le) f
HN
{/
LEST
5

\S 130
LEE AA

MORTALITY RATE / 100,000

   
  

 

 

  

~~ ee LLY)

a
|
LELLLILD Looe /

 

i
Eye -90

HPP LLL AL TLITLEILELL ELT P60

MORTALITY RATE/ 100,000

SALILLIL TEL LL LLL P30

1975

1955 gg bo HO <9

   

1965

FIGURE 29.—Age-specific mortality rates by 5-year age
groups for cancer of the buccal cavity and
pharynx for white males, United States, 1950-
1977

SOURCE: National Cancer Institute (198).

in risk for cancer of the buccal cavity and pharynx among smokers of
lower tar and nicotine cigarettes, but the reduction was not
statistically significant. Wynder and Hoffmann (376) reported simi-
lar findings in a retrospective study of smokers of filter cigarettes
versus smokers of nonfilter cigarettes.

83
PEEUEEded deeded

PED Peed

MORTALITY RATE, 100,000

 

-30

¥
a

 
  
  
    

PHT peo

MORTALITY RATE/100,000

AULT bo

  

0
1975

  

1955 go 60 Yo 20

FIGURE 30.—Age-specific mortality rates by 5-year age
groups for cancer of the buccal cavity and
pharynx for white females, United States,

1950-1977

SOURCE: National Cancer Institute (198).

Specificity of the Association

The prospective studies have reported mortality data for a large
number of diseases. Specificity, which is related to the magnitude of
the association between smoking and oral cancer, is evidenced by the
differences in the mortality ratios (smokers versus nonsmokers) of
oral cancer and other cancers (Appendix Tables A and B). These

84
TABLE 24.—Mortality ratios for cancer of the oral cavity—
prospective studies

 

 

Number
of Cigarette
Study Population size deaths Nonsmokers smokers Comments
ACS 9-State 188,000 males 55 1.00 5.06 Only 3 deaths
Study among
nonsmokers
British 34,000 males 38 1.00 13.00 Includes lip,
Physicians tongue, mouth,
pharynx, larynx,
and trachea
US. Veterans 290,000 61 1.00 4.22 Buccal cavity
1.00 14.05 Pharynx
ACS 25-State 358,000 males 167 1.00 6.52 Buccal cavity
Study 483,000 females 65 1.00 3.25 and pharynx
California males 68,000 males 19 1.00 2.76
in 9 occupations
Japanese 122,200 males 43 1.00 2.88 males Data for mouth
Study 142,800 females ll 1.00 1.22 females only
Swedish 55,000 males 15 Mortality ratios not 5 deaths in
Study and females published nonsmoking
males;
10 deaths in

smoking males

 

differences are even greater when comparisons are made with the
mortality ratios of heavy smokers.

Temporal Relationship of the Association

Evidence for a temporal relationship of this association is provided
by the prospective studies in which populations of apparently
disease-free smokers and nonsmokers were followed over time for
oral cancer mortality. In addition, the finding of premalignant oral
mucosal changes in greater proportions of smokers than nonsmokers
provides evidence for the temporality of the association (see below).

Coherence of the Association

Dose-Response Relationship

The finding of a dose-response relationship between smoking and
oral cancer mortality in both retrospective and prospective studies
lends support to the causal nature of the association.

85
TABLE 25.—Oral cancer mortality ratios by amount
smoked—prospective studies

 

 

ay,
Amount Smoked
Study Population per Day Comments
Males Females
British 40,000 NS 1.00 NS 1.00 = Male data
Physicians 1-14 5.00 1-14 _ by grams
15-24 7.00 15-24 4.00 of tobacco
25 + 33.00 25 + 6.50 per day
US. Veterans 290,000 NS 1.00 *Based
1-9 2.92* on fewer
10-20 2.87 than 20
21-39 6.15 deaths.
40 + 12.40°
Japanese in 29 265,000 NS 1.00 Hypopharynx
Health Districts 1-19 1.20 only
20-29 5.50
30 + 9.10
ACS 9-State 188,000 NS 1.00 Includes
Study males 1-9 7.00 larynx
10-20 6.00 and
20 + 787 esophagus
California males 68,000 NS 1.00
in 9 occupations males <“_ pack 3.69
1 pack 117
1, pack 5.52

 

NOTE: NS: Nonsmoker.

Correlation of Sex Differences in Oral Cancer With Different
Smoking Habits

Oral cancer is predominantly a disease of males, but the difference
between male and female rates of disease is narrowing. This finding
is consistent with the differences in the smoking trends of males and
females noted above. As with laryngeal and esophageal cancer, there
is a strong association between oral cancer and alcohol consumption.
This must be considered as contributing to the excess ratio of male to
female oral cancer mortality (see below).

Correlation of Oral Cancer Mortality Rates Among Populations With
Different Tobacco Consumption

In populations with low proportions of smokers (e.g., Mormons and
Seventh Day Adventists), the incidence and mortality rates of cancer
of the gum, mouth, tongue, and pharynx are substantially reduced
(79, 165, 166, 211, 294).

86
 

 

N= 497

 

RELATIVE RISK

 

 

 

 

 

 

 

 

PRESENT 1-3 4-6 7-10 11-15 16+ NON-
SMOKERS SMOKERS

FIGURE 31.—Relative risk of male ex-smokers for cancer
of the oral cavity by years since quitting

smoking
SOURCE: Wynder and Stellman (326).

Oral Cancer Mortality and Cessation of Smoking

In the U.S. Veterans Study (224), ex-smokers had approximately
40 percent of the risk for oral cancers of current smokers. Data from
the American Health Foundation study found that the risk of cancer
of the oral cavity among former smokers declined with the number
of years off cigarettes when compared to the risk of continuing
smokers. After 16 or more years of cessation, the risk of oral cancer
approaches that of nonsmokers (Figure 31). This is consistent with
the causal nature of the association.

Smoking and Histological Changes in the Oral Mucosa

Leukoplakia is an abnormal thickening and keratinization of oral
mucosa and is recognized as a precursor of malignancy of the oral
cavity (124). A few studies have established a relationship between
smoking in various forms and leukoplakia (269).

Oral Cancer and Non-Cigarette Tobacco Use

The oral cavity and pharynx are the sites most consistently
exposed to tobacco smoke. A summary of the data from the
prospective epidemiological studies is presented in Table 26. They
demonstrate that cigar and pipe smokers experience a significant
risk of developing cancer of the oral cavity compared with nonsmok-
ers. This risk is approximately equal for all smokers whether an
individual uses a pipe, cigar, or cigarette.

Several authors have reported a relationship between chewing
tobacco and/or snuff dipping (the placement and retention of fine

87
TABLE 26.—Mortality ratios for oral cancer in cigar and
pipe smokers. A summary of prospective
epidemiological studies

 

 

Smoking Type
Study
Non- Cigar Pipe Total Pipe Cigarette Mixed
Smoker Only Only and Cigar Only

ACS 9-State Study ' 1.00 5.00 3.50 _— 5.06 —
British Physicians * 1.00 _ _ 29.00 13.00 11.00

ACS 25-State Study 1.00 _ — 4.94 M 6.52 ~

F 3.75 _—

US. Veterans Study

Oral* o0....... 1.00 4.11 3.12 4.20 4,22 3.79
Pharynx ........., 1.00 - 1.98 7.76 14.05 7.75

 

‘Combines data for oral, larynx, and esophagus.

? Figures for all non-lung respiratory cancers.

* Mortality ratios for ages 45 to 64 only as presented.
* Excludes pharynx.

ground or powdered tobacco in the oral vestibule between the gums
and cheek) and oral cancer (36, 186, 207, 234, 299, 301, 310). A recent
report found a fourfold increase in risk for oral cancer among female
snuff dippers compared to nontobacco users (301). The excess risk for
cancers of the cheek and gum was nearly fiftyfold among long-term
users. The authors estimated 87 percent of these tumors were related
to snuff use. In the Third National Cancer Survey, Williams and
Horm (299) noted an excess relative risk for cancers of the gum and
mouth in male and female users of chewing tobacco or snuff.
However, this risk was only statistically significant for males.

A few epidemiological investigations have demonstrated an associ-
ation between the combined use of alcohol and pipe or cigar smoking
and the development of oral cancer (135, 172, 173, 310). Heavy pipe
and/or cigar smoking and heavy drinking are associated with higher
rates of oral cancer than are seen with either habit alone.

Synergistic Role of Alcohol and Cigarettes for Oral Cancer

Oral cancer occurs more commonly in heavier users of alcohol (37,
88, 136, 227, 283, 301, 310). A recent study (179) noted an interaction
(Figure 32) for oral cavity cancer in white males who use both
alcohol and cigarettes. Nonsmokers who consumed 7 ounces or more
of alcohol per day had a relative risk of 2.5. Those cigarette smokers
who consumed 7 ounces or more of alcohol per day had a relative risk
of 5.1 if they smoked one-half a pack or less daily, 20.5 if they smoked
11 to 20 cigarettes per day, and 24.0 if they smoked more than one
pack of cigarettes per day. A distinct synergy (a multiplicative effect)
of alcohol and cigarette smoking has been described elsewhere (271).
The mechanism by which these two factors interact is unclear.

88
CASES = 384
CONTROLS = 4725

CIGARETTES
PER DAY

 

0 1-6 7+
OUNCES OF ALCOHOL PER DAY

FIGURE 32.—Relative risks of oral cavity cancer by daily
consumption of alcohol and cigarettes for
males

* Not significant.
SOURCE: McCoy et al. (179).

Experimental Studies

A useful animal model for the experimental study of oral
carcinogenesis has not been found. Cigarette smoke and cigarette
smoke condensates generally fail to produce malignancies when
applied to the oral cavity of mice, rabbits, or hamsters. Mechanical
factors, such as secretion of saliva, interfere with the retention of
carcinogenic agents. However, positive results have been obtained
with benzo[a]pyrene, 20-methyl-cholanthrene, 9,10-dimethy]-1,2 ben-
zanthracene, and other tobacco smoke carcinogens when applied to
the cheek pouch of hamsters. The cheek pouch, however, lacks
salivary glands, and its structure and function differ from those of
the oral mucosa. These studies have been reviewed in previous
reports of the U.S. Public Health Service (272, 276).

Conclusion

1. Cigarette smoking is a major cause of cancers of the oral cavity
in the United States. Individuals who smoke pipes or cigars
experience a risk for oral cancer similar to that of the cigarette
smoker.

89
2. Mortality ratios for oral cancer increase with the number of
cigarettes smoked daily and diminish with cessation of smok-
ing.

3. Cigarette smoking and alcohol use act synergistically to
increase the risk of oral cavity cancers.

4. Long term use of snuff appears to be a factor factor in the
development of cancers of the oral cavity, particularly cancers
of the cheek and gum.

Carcinoma of the Esophagus
Introduction

Carcinoma of the esophagus is a rapidly fatal neoplasm; there is a
median survival of less than 6 months following diagnosis and a 5-
year survival rate of 3 percent.

The number of deaths caused by esophageal cancer rose from 3,866
in 1950 to 7,283 in 1977. The age-adjusted death rate increased from
2.3 to 2.6 over this period (Figure 33).

In the United States in 1977, 3,924 white males and 1,520 white
females died from esophageal cancer; in the other than white
population, 1,404 males and 435 females died from this disease.
While these figures represent only a slight increase in age-adjusted
mortality in the white population, they do reflect nearly a twofold
increase in the other than white population from 1950 to 1977.

The ratio of the age-adjusted death rate of the other than white
population to that of the white population increased over the study
period. In 1977, the death rate from this cause among other than
white males between the ages of 35 and 44 years was eight times that

“among white males of the same age. The death rate of other than

white females in this age group was 13 times the corresponding rate
of white females. Mortality ratios by race (white/ other-than-white)
decreased with age in both males and females.

Among whites, the mortality sex ratio (male/ female) declined
slightly between 1968 and 1977. In the other than white group, there
was also a greater relative increase in the age-adjusted death rate of
females than in those of males.

Among white males and females, age-specific death rates from
cancer of the esophagus (Figure 34) increased in each succeeding 10-
year age group to the end of the lifespan. In other than white males,
mortality peaked between ages 65 and 74 (Figure 35). The pattern
was irregular in other than white females, varying with age group
and time span over the 1950-1977 period.

A three-dimensional graph of age-specific death rates for white
males and females for cancer of the esophagus over the period 1950-
1977 is shown in Figures 36 and 37.

90
       

 

 

 

wo 7
ud
wm
ow « 4
wt £ io
wo a1 & w ~
wager ke 40
a = bal
ecu w WW
meee 4
ote
ww x= =
-e Fe X =
“= ZZ 4
xz ros
xX z2Z2z
uo oo out 4
+* OB
o
&
190
~_
4
Ww
42m
Oe
“uu
J} >
we
a
4 6s
=z
uw
J 4
«
oO
o
Je
n
—_
q
4
B
12m
_
4
4
#4
°
1 1 { We
wo N a wo on
-_ “7
RATES/100,000

FIGURE 33.—Age-adjusted* mortality rates for cancer of
the esophagus, by race and sex, United
States, 1950-1977

* This graph is age-adjusted to the U.S. population as enumerated ir; ‘‘'70; all rates cited within the text of the
Report, however, are adjusted to the population as enumerated in 1940.
SOURCE: National Cancer Institute (798).

1
377-310 0 - 82 - 8 9
36

(861) ANINSU] Ja0UeD [BUOTEN :-ADUNOS

snZeydosa ay) Jo 190U¥d I0J s97879G pazIUgQ

THA JO} Soyer APT eVIOU ayreds-33y— Fe qunola

T

ay} w

000° 001 /S31ba

50

40

30

20

 

10

MALES

20 30 40 50 60 70
AGE IN YEARS (BY 5-YEAR AGE GROUPS)

60

000°00)/saluy

50

40

30

20

 

FEMALES

+ =1950-1956
=1957~-1963
© =1964-1970
51971-1977

to 20 30 40 50 60 70
RAGE IN YEARS (BY 5-YEAR AGE GRouPS)

60
FEMALES

 

AGE IN YEARS (BY 5-YEAR AGE GROUPS}

1950-1956
1957-1963
1971-1977

1964-1970

+
x
0
QO

 

oO o
~ N

RATES/100.000

 

50
40

 

 

AGE IN YEARS (BY 5-YEAR AGE GROUPS}

10

 

— a
o o
” N

RATES/100,0C0

50
40
* 10
a

FIGURE 35.—Age-specific mortality rates for nonwhites in
the United States for cancer of the esophagus

SOURCE: National Cancer Institute (198).

93
 

-

LTP Elder

 
       

PITELTTTEP ETE E PTET 60

  
       
 
 
 

  

LILI SD 2
KAIF ALOIS B
PEPPE ETE tel SELLS SIGE

- 30
LLL IM Yh

MORTALITY RATE /100,000

   

 

PN

 
  

           
 

 

 

             

 

  

 

 

 

 

 

 

 

 

 

LLL AAT KL LLL Le
SSSo LO? LLL 2
. lee IIT.
59 9 60 80
___-—-+ 90
TT
HHITITIT PTET Go
HITE R30
. °
\ =
1955 gg Go 4O 20

FIGURE 36.—Age-specific mortality rates by 5-year age
groups for cancer of the esophagus for white
males, United States, 1950-1977

SOURCE: National Cancer Institute (198).

It is estimated that in 1982 in the United States there will be 8,900
new cases and 8,300 deaths from this disease (2).

A number of epidemiological and experimental studies have
established an association between smoking and esophageal cancer.

94
 

 
 
  
   
  

    
   
  

   
       

     

  

 

 

      
  
  

    

  

——7 30
8
HteP EP Edi beened i
PAN MI :
vin QPL pA, E
LIN SLU | §
Zilog EE ;
, . SEI ee Go Pll, ZL1 A
° 1955 tds EL ALTE |
lp 1965 1975 9 20 4o SO 80
<> 1 30
8
PEE EELELE EE EOE CETTE LEPEELT ET 20 8
AULT bao 2
fe g
0
1975 G

 

gg Go YO 20

 

1955

FIGURE 37.—Age-specific mortality rates by 5-year age
groups for cancer of the esophagus for white
females, United States, 1950-1977

SOURCE: National Cancer Institute (198).

Causal Significance of the Association

Consistency of the Association

At least 10 retrospective studies have examined the relationship
between smoking and esophageal cancer (276). Regardless of method-
ology, risk ratios were consistently increased. Data from the major
prospective studies (Table 27) also demonstrate consistently in-
creased mortality ratios for male smokers as compared with non-

95
TABLE 27.—Mortality ratios for cancer of the esophagus—
prospective studies

 

 

Number of Cigarette
Study Population size deaths Nonsmokers — smokers Comments
ACS 9-State 188,000 1 nonsmoker 1.00 5.06 Esophagus and
Study 33 smokers other respiratory
sites
British 34,000 males 65 1.00 4.70 Esophagus and
Physicians other respiratory
sites
US. Veterans 290,000 119 1.00 6.43
ACS 25-State 398,000 males 116 1.00 3.96
Study 483,000 females 48 1.00 4.89
California males 68,000 males 32 1.00 1.82
in 9 occupations
Japanese 122,200 males 215 1.00 2.35
Study
Swedish 55,000 males 1 nonsmoker
Study and females 12 smokers 1.00 _

 

smokers. The ACS 25-State Study showed similar results for female
smokers and cancer of the esophagus.

Strength of the Association

Mortality ratios in the retrospective studies ranged from 1.3 to
11.1 among heavy smokers; mortality ratios in the prospective
studies ranged from 1.8 to 6.4. In four of the large prospective
studies, a dose-response relationship was demonstrated (Table 28), A
reduced risk for esophageal cancer among female but not male
smokers of lower tar and nicotine cigarettes has also been reported
(158).

Specificity of the Association

Specificity of the association between smoking and esophageal
cancer is evidenced by substantial differences in the mortality ratios
(smokers versus nonsmokers) for esophageal cancer compared to
other smoking-related cancers (Appendix Tables A and B).

Temporal Relationship of the Association

The temporal relationship of this association is supported by the
prospective studies in which populations of initially disease-free
subjects were followed for the development of esophageal carcinoma.
In addition, there are histological data suggesting that smoking

96
TABLE 28.—Mortality ratios for cancer of the esophagus
; by amount smoked—prospective studies

 

 

Study Population Size Cigarettes/Day Ratio Comments
British 34,000 males Nonsmoker 1.00 Grams of
Physicians 1-14 4.00 tobacco

15-24 4.33 per
25 + 10.00 day
US. Veterans 290,000 Nonsmoker 1.00 *Based on
1-9 3.06* fewer than
10-20 4.34 20 deaths
21-39 12.42
40 + 9.20*
Japanese in 29 122,200 males Nonsmoker 1.00
Health Districts 1-19 2.20
20-29 2.80
30 + 3.20
California males in 68,000 Nonsmoker 1.00
9 occupations about /. pk 1.27
about 1 pk. 1.69
about 1'4 pk 1.82

 

antedates premalignant and malignant transformation of esopha-
geal epithelium (73, 76).

Coherence of the Association
Dose-Response Relationship

There is a dose-response relationship between smoking and
esophageal cancer mortality in retrospective and prospective studies
(276).

Esophageal Cancer Mortality and Cessation of Smoking

Several of the prospective studies noted reduced risks for cancer of
the esophagus after quitting smoking. The U.S. Veterans Study found
that the mortality ratio for ex-smokers decreased to 2.41 compared to
6.43 for continuing smokers. For the British Physicians Study, the
corresponding ratios were 1.66 and 5.33, respectively. Thus, ex-
smokers had only about one-third the risk for esophageal cancer of
current smokers.

Figure 38 presents data from the American Health Foundation
study for esophageal cancer mortality risk by the number of years off
cigarettes. After quitting smoking for 4 years or more, former
smoker rates were not substantially above those of nonsmokers.

97
 

RELATIVE RISK

 

PRESENT 1-3 4-6 7-10 11-15 16+ NON-
SMOKERS SMOKERS

FIGURE 38.—Relative risk of male ex-smokers for cancer
of the esophagus by years since quitting

smoking
SOURCE: Wynder and Stellman (326).

Correlation of Sex Differences in Esophageal Cancer With Different
Smoking Habits

Esophageal cancer is predominantly a disease of males. The sex
differences observed for esophageal cancer mortality are compatible
with the sex differences in smoking patterns. As with oral and
laryngeal cancer, esophageal cancer has also been related to
excessive alcohol consumption. This must be considered as contribut-
ing to the excess ratios of male to female esophageal cancer
mortality (see page 101).

Correlation of Esophageal Cancer Mortality Among Populations
With Different Tobacco Consumption

In populations with low proportions of smokers (e.g., Mormons and
Seventh Day Adventists), the mortality rates from esophageal cancer
are substantially reduced (79, 165, 1 66, 211, 294).

98
TABLE 29.—Mortality ratios for cancer of the esophagus in
cigar and pipe smokers—a summary of
prospective epidemiological studies

 

 

 

Smoking type
Non- Cigar Pipe Total pipe Cigarette
Study smoker only only and cigar only Mixed

ACS 9State Study! 1.00 5.00 3.50 = 5.06 _
British Physicians 1.00 — _ 3.70 4.70 9.0
ACS 25-State Study 1.00 _- - 3.97 males 3.96? _

females 4.89? —
U.S. Veterans 1.00 5.33 1.99 4.05 6.43 =

 

‘ Combines data for oral, larynx, and esophagus.
? Mortality ratio for ages 45 to 64.

Smoking and Histologic Changes in the Esophagus

Examination of 12,598 histologic sections of esophageal autopsy
tissue from 1,268 men showed histologic findings which were similar
to the abnormalities generally accepted as being premalignant in
respiratory tract epithelium (76). Only 2.5 percent of the slides from
current smokers exhibited no atypical cells, compared with 93.5
percent of slides from nonsmokers. The finding of 60 percent or more
atypical cells was rare in the tissue of nonsmokers (0.3 percent), but
much more common in tissue of smokers (17.7 percent).

Esophageal Cancer and Non-Cigarette Tobacco Use

The esophagus is not directly exposed to inhaled tobacco smoke,
but tobacco smoke constituents condense on the mucous membranes
of the mouth and pharynx and are swallowed, thus contacting
esophageal cells. The esophagus also receives mucous cleared from
the lungs by the ciliary mechanism or by coughing which is also
swallowed. Variations in the inhalation of the smoke of different
tobacco products may not appreciably alter the degree of exposure of
the esophagus. This possibility is suggested by the prospective and
retrospective epidemiological studies which demonstrate similar
mortality rates for cancer of the esophagus in smokers of cigars,
pipes, and cigarettes. These data are presented in Table 29.

Several retrospective investigations have examined the association
between smoking in various forms and cancer of the esophagus
(Table 30). These studies suggest that cigar, pipe, and cigarette
smokers develop cancer of the esophagus at rates substantially
higher than do nonsmokers and that little difference exists between
these rates observed in smokers of pipes, cigars, or cigarettes.
Histologic changes in the esophagus have been related to smoking of

99
cigarettes and other forms of tobacco (76). Several retrospective
studies conducted in the United States and other countries have
examined the synergistic role of tobacco use and heavy alcohol
intake and the risk of mortality from cancer of the esophagus. At
least four of these investigations contain data on pipe and cigar
smoking (33, 172, 173, 307). It appears that smoking in any form in
combination with heavy drinking results in especially high rates of
cancer of the esophagus.

TABLE 30.—Relative risk of cancer of the esophagus for
men, comparing cigar, pipe, and cigarette
smokers with nonsmokers. A summary of
retrospective studies

 

Relative risk ratio and percentage of cases
and controls by type of smoking

Non- Cigar Pipe Total pipe Cigarette

Author, reference Number

 

 

smoker only only * and cigar only Mixed
Sadowsky (296): Relativ: risk 1.0 48 3.8 5.1 38. 33
104 Percent. cases 4 5 3 6 6 1g
615 Percent controls 18 3 7 4 53 19
Relative risk 100031 21 ween 26 A
39 = Percent cases 18 15 18 see 51 3
115 Percent controls 24 9 16 tee 36 1B
Relative risk 10 .... 30 tee 27 59
202 ~~ Percent cases Wos.... OF cee 59 18
713 Percent controls 39 .... 5 50 7
Schwartz (247): Relative risk 10.1... (26 tee 11.7 86
Cases... 249 Percent cases 2 .... 2 tee & 7
Controls............... 249° «Percent controls 18 .... «7 67 7
Wynder and Bross
(907): Relative risk 1000 (36 9.0 6.0 28 37
Cages... le, 150 Percent cases 5 19 9 4 51 11
Controls............... 150 Percent controls 15 16 3 2 55 9
Bradshaw and
Schonland (83): Relative risk 100.2... «448 ee 23
Cases... 117“ Percent cases wo... Ad tee 68
Controls............... 366 Percent controls 32 .... 18 an 58
Martinez (172): Relative risk 10 20 2... OL, 15 22
Cases... 120 Percent cases 8 9 Lee. 31 43
Controls. .............. 360 Percent controls 14 8 tee ee 34 M
Martinez! (179): Relative risk 10020 28 ae 17 25
Cases... 346 = Percent cases 21 10 15 34 xu
Centrols............... 346 Percent controls 22 9 1 tae Bs] 3

 

‘This study combines data for oral cancer and cancer of the esophagus.

100