HomeMy WebLinkAboutMINUTES - 05051987 - S.4 TO: BOARD OF SUPERVISORS
FROM: Supervisor Nancy C. Fanden Contra
Costa
April 28, 1987 for Consideration �Oun+�
DATE: May 5, 1987 l
SUBJECT: DR. DONALD F. AUSTIN'S COMMENTS
ON THE SHELL CANCER STATEMENT .
SPECIFIC REQUEST(S) OR RECOMMENDATION(S) & BACKGROUND AND JUSTIFICATION
REQUEST:
That Dr. Donald F. Austin' s comments on the Shell Cancer
Statement on "Lung Cancer in Contra Costa County" be
referred to Dr. Wendel Brunner for review and response.
That the Board might consider placing this report and Dr.
Brunner' s response on the Cancer Summit Meeting agenda.
BACKGROUND:
On March 11, 1987 , a Lung Cancer Statement given to me by
Mr. Ron Banducci, Manager of Shell, was referred to Donald
Austin, M.D. , Chief, Cancer Epidemeology, State of
California, and Dr. -Wendel Brunner, Director of
Environmental Health, for their evaluation.
Attached is a copy of Dr. Donald F. Austin' s comments on
this matter.
CONTINUED ON ATTACHMENT: YES SIGNATURE:
RECOMMENDATION OF COUNTY ADMINISTRATOR RECOMMENDATION OF BOARD COMMITTEE
APPROVE OTHER
SIGNATURE(S)
ACTION OF BOARD ON May 5, 1987 APPROVED AS RECOMMENDED x _ OTHER
VOTE OF SUPERVISORS
X UNANIMOUS (ABSENT _ ) I HEREBY CERTIFY THAT THIS.IS A TRUE
AYES: NOES: AND CORRECT COPY OF AN ACTION TAKEN
ABSENT: ABSTAIN: AND ENTERED ON THE MINUTES OF THE BOARD
OF SUPERVISORS ON THE DATE SHOWN.
CC: Health Services Director ATTESTED ,... _
Environmental Health Phil Batchelor, Clerk of the Board of
County Administrator t Supervisors and County Administrator _
M382/7-83 BY DEPUTY
L, 1
STATE OF CALIFORNIA—HEALTH AND WELFARE AGENCY GEORGE DEUKMEJIAN, Governor
DEPARTMENT OF HEALTH SERVICES
5850 SHELLMOUND STREET, SUITE 200 ? ee
EMERYVILLE, CA 94608 `
(415) 540-3500
April 15 , 1987
Nancy Cardinalli Fahen
Contra Costa County Board of Supervisors
805 Las Juntas
Martinez , CA 94553
Dear Supervisor Fanden:
Enclosed are my comments on the Shell Cancer Statement you asked me to evaluate
in your March 11 memo. I hope you find the comments useful.
Sincerely,
Donald F. Austin, M.D. , M.P.H. , Chief
Cancer Prevention Section and
California Tumor Registry
Enclosure(1 )
cc: Wendell Brunner, M.D. , M.P .H.
Comments on "Short Statements on Lung Cancer in Contra Costa County"
1 . Cancer mortality in white males in Contra Costa County has remained
consistently slightly below U.S. rates since 1960.
Comment: This statement is correct and is supported by data published by
the Environmental Protection Agency and National Cancer
Institute. However, for cancer mortality in white females, the
mortality rates are higher (although not significantly so)
during the 1970's than the U.S. rates.
2. Lung cancer mortality rates for white males in Contra Costa County were
13% higher than U.S. rates for the period 1950-1959; 7% higher than U.S.
rates for the period 1960-1969; but have been lower than U.S. rates for
the period 1970-1979 right up to the present.
Comment: These data are correct and are .supported by data published by
the Environmental Protection Agency and the National Cancer
Institute. For white females, the rates are higher for Contra
Costa County than for the U.S. from 1960-79. For non-white
males the Contra Costa rates are lower in both the 1960's and
19701s.
3. From above, if lung cancer mortality in white males in Contra Costa County
is no higher than the general U.S. population, yet there is a _35% gradient
between the so-called "industrial area" residents of the county and the
"non-industrial area" residents of the county, then this is likely to be
due to a deficit of lung cancer in the "non-industrial" residents rather
than a major excess in "industrial" residents.
Comment: Lung cancer mortality rates for white males in Contra Costa
County in the 1970' s have been just below the rates for the
entire country. This is supported by published data from the
Environmental Protection Agency and the National Cancer
Institute. Given this finding it is most likely that the
industrial area is. somewhat above the U.S. rate and the non-
industrial area is somewhat below the U.S. rate for lung cancer
in white males in order for the entire county to be just below
the U .S. rate. The 35% "gradient" in the industrial area over
the non-industrial area is in the incidence of lung cancer.
While one would expect that mortality rates are similarly
different, no one has addressed that issue.
4. For the Bay area as a whole, Contra Costa County lung cancer rates fall
below those of Alameda County and San Francisco County. They are very
similiar to Marin County lung cancer rates.
Comment: This is correct and is shown in unpublished data from the
California Tumor Registry for the time period 1980-84. However,
for males Contra Costa has higher rates than Alameda, San
Francisco or Marin and for females it has lower rates than for
those three counties.
5 . A similar cancer gradient was identified a few years ago in Harris County,
Texas. A major case-control study of lung cancer in Harris County and .
four surrounding counties investigated air pollution data, smoking habits,
and occupation. This University of Texas study by Dr. Patricia Buffler
showed conclusively that the gradient seen was overwhelmingly due to
socioeconomic and lifestyle differences between neighborhoods and was
unrelated to either particulate or sulfate pollution measurements. The
conclusion of this study was that less than 2% of the lung cancer cases
could even possibly be explained by pollution.
Comment: I have been unable to find a report of the results and con-
clusions of this study. I have found a publication of a
presentation during September, 1981 of this study, but at that
time no results were available.
6. From above, the cancer gradient seen in Contra Costa County is not unique,
and socioeconomic and lifestyle differences are the most probable explana-
tions for it.
Comment: We know of a similar cancer gradient in Los Angeles that was
reported by investigators from USC. They later presented in an
obscure symposium the only description of their further
investigation which showed that lifestyle differences
(particularly smoking) could explain the differences found. A
similar study of lung cancer mortality rates in Pittsburg, PA. ,
found that differential smoking habits could explain the
apparent relationship with air pollution. In fact, the Depart-
ment's investigation in Contra Costa County a few years ago
reported that the only differences between individuals with lung
cancer in the county and individuals in the county with similar
characteristics but without lung cancer, were lifestyle
factors. These investigations make it likely that the conten-
tion in this short statement is correct.
7. Shell Martinez employees do not experience a high mortality due to lung
cancer. In fact, for the period 1973-1982 for the Martinez complex as a
whole, retired and active Shell employees died from lung cancer at a rate
300 lower than the U.S. as a whole. Deaths from all cancers in these same
Shell, employees for the same period were 27% lower than U .S. rates.
Comment: While I do not have access to these data, I do not doubt the
statement. Individuals -too ill to work or pass pre-employment
physical examinations are known to be at increased risk of death
from all causes compared to individuals who are employed, a
phenomenon so well .established it has been called the "Healthy
2
Worker Effect." While the Healthy Worker Effect is somewhat
weaker for cancer than for other causes, such as respiratory
diseases, it is still present, particularly in the early years
of employment, and would slightly decrease death rates from all
cancers and from lung cancer.
There is also another factor which may be even more important
here than the Healthy Worker Effect. For safety reasons smoking
is not permitted in oil refineries and, one assumes, has not
been allowed in Shell Martinez for many years, if ever.
Smokers, particularly heavy smokers, are likely to find such
working conditions unacceptable and, for that reason, have been
found to be underrepresented in the work force of other oil
refineries (and presumably in Shell Martinez) . Since smokers,
and particularly heavy smokers, are a very high risk group for
lung cancer and a number of other different cancers, any group
such as employees and former employees of Shell Martinez which
contained relatively few of these high risk individuals would
have lower death rates from lung cancer and, because smoking
related cancers are relatively common causes of death in com-
parison to other cancers, would also have lower death rates from
all cancers combined.
8. Shell employees are hot unique in experiencing a low mortality rate due to
lung cancer in Contra Costa County. The Chevron Richmond Refinery
employees experienced an all cancer mortality rate 22o lower than the
U.S. as a whole for the period 1950-1980 and a lung cancer rate 38% lower
than U.S. rates.
Comment: While we do not have access to these data, they also appear
likely to be correct. In addition to the theoretical considera-
tions given under statement #7 above, a recently published study
of employees in 16 refineries, in which present and former
employees of that refinery made up over 20% of the study popula-
tion, found that the death rates for lung cancer were 15% lower
and for all cancers were 13% lower.
9. At most, it has been estimated that occupational factors are responsible
for only 4% of the cancers in this country. Presumably, this also is true
for Contra Costa County. One would expect to see evidence of an occupa-
tional contribution to lung cancer in Contra Costa County in occupational
cohort studies. We have not found this in Shell employees, nor has
Chevron found this in their employees.
Comment: This statement is probably correct. There have been much higher
estimates, but the ones with the most credibility that are most
generally accepted by the scientific community are of the order
of magnitude given in the short statement.
In most situations it can be assumed that, if a material is
3
carcinogenic, individuals occupationally exposed have higher
exposures than anyone else and therefore would be expected to
show the highest risk of developing or dying from cancer as a
result of this exposure. In such cases, the failure to find an
increased risk of cancer among those occupationally exposed is
strong evidence that those with lower exposures are also not at
detectably increased risk. This is the sole justification for
this statement.
It is theoretically possible, however, that the general public
immediately downwind of industrial emissions may have higher
exposures because ventilation practices tend to protect the
workers by blowing off these materials in higher concentrations
directly into the outside exhuast. It is also theoretically
possible that changes made chemically, in an effort to
deactivate potentially dangerous materials, may create, for
thefirst time, carcinogens which would appear in the exhaust pro-
ducts only. In such unlikely situations, the exposures would be
higher in those individuals living downwind from the exhaust,
and, particularly in the latter mechanism, there would be no
occupationally increased risk but a possibly increased risk for
those living downwind. On the other hand, employees tend to
live near 'their places of employment and, hence, would be likely
to have high levels of these theoretical and environmental
exposures. If these theoretical situations were the case,
results of studies similar to those reported in statements #7
and #8 would be unlikely. Further, we found no evidence for
this theoretical situation existing in our earlier lung cancer
study in Contra Costa County in which we measured certain known
carcinogens and tested all particulates for mutogenicity using
the Ames test, a common method of screening for potential
carcinogenicity.
10. If the primary concern in Contra Costa County is with respect to lung
cancer and the 35% gradient between two areas within the county, then a
large scale coordinated look at all new lung cancers in Contra Costa
County with an assessment of duration of residence, lifestyle factors
(including smoking and dietary factors) , occupational exposures, and other
factors of concern compared to a control population within the county
would appear to offer a speedy (within 5 years) , relatively inexpensive
(unlikely to exceed one million dollars) , and definitive answer to the
questions now being asked.
Comment: The study as described does seem likely to offer an answer
(although I do not concur with the definitions of "speedy" and
"relatively inexpensive" given in statement #10) . Such an
answer would probably be the most definitive one that could be
provided. Of course, such a "speedy, relatively inexpensive"
study was carried out by the Department several years ago, and
there is reason to believe that the findings of this new study
would be similar. However, if the questions now being asked are
considered important enough that the funds are provided by
Shell, Contra Costa County or any other source, the answers
could be obtained by the method suggested.
4