Is Brain Cancer Mortality Increasing in Industrial Countries?

Is Brain Cancer Mortality Increasing in Industrial Countries?

DEVRA LEE DAVIS"

National Research Council Washington D. C. 20418

DAVID HOEL

National Institute of Environmental Sciences Research Triangle Park, North Carolina 27709

CONSTANCE PERCY

Cancer Surveillance Branch National Cancer Institute

Rockville, Maryland 20892

ANDERS AHLBOM

Karolinska Institute 18 Stockholm, Sweden

JOEL SCHWARTZ

United States Environmental Protection Agency Washington, DC 20460

INTRODUCTION

Brain cancer is not a common cancer, accounting for less than one percent of all cancers. A number of investigators have reported that braincancer and other nervous system cancers have increased rapidly over the past few decades in persons over age

"Address correspondence to: Devra Lee Davis, Ph.D., M.P.H., Scholar in Residence, National Academy of Sciences, 2101 Constitution Avenue NW, Washington, DC 20418.

l91

192 ANNALS NEW YORK ACADEMY OF SCIENCES

65.1-3 Some have suggested that this recordedincreaserepresents an artefact resulting from increased access to care and improved diagnostic technology, based on an analysis of well-ascertained causes of death (from the early 1950s through the 1970s) in Rochester, Minnesota, a relatively small but not heavily industrialized city!

Recently, this suggestion that increases in recorded brain cancer represent artefact was also echoed by those who conducted a study of brain cancer utilizing the Cancer Registry of the Swiss Canton of Vaud. With an average population of about 530,000 between 1974 and 1987, the overall age-adjusted rate for all malignant brain neoplasms did not change significantly, based on the world standard population adjusted to the 401 cases recorded during those 14 years, but the older age groups were not studied separately? Also, brain tumor incidence in Sweden has recently been reported not to have increased for specific brain tumor types, although there is a slight increase in males over age 70 for all brain tumors taken together.6

This paper explores recent trends in brain and other central nervous system cancer by examining age-specific mortality trends for a number of industrial countries, along with selected data on factors that may influence recording of mortality, such as diagnostic technology.

METHODS AND MATERIALS

Mortality Analyses

For the period 1968 through 198611987, annual data on deaths due to brain and other central nervoussystemcancer (coded ICD 191 and 1!?2), along withcorresponding population data separated into five-year age groups, were provided by the Statistics Unit of the WorldHealth Organizationin Genevafor Italy, France, the United States, the Federal Republic of Germany, and Japan. Data for the United Kingdom was also provided by their Office of Censuses and Population Surveys along with corresponding population data separated into five-year age groups.

These major industrialized countries have in common the existence of various forms of broad coverage for health services, and longstanding, nationwide systems for recordingdeathcertificates. Causeof deathis recorded by aphysician; thisinformation is then encoded by trained personnel who use detailed rules to convert thisinformation into a single code number. The code numbers and rules have evolved over nearly a century, with revisions at approximately 10-year intervals to accommodate changesin medical practice and concepts.’

Quality of Ascertainment Analyses

We considered whether changes in diagnostic technology may be involved in changes in recorded mortality by looking at detailed data from the United States and United Kingdom. Regarding the influence of changingdiagnosticpractices, national and international data on utilization of nuclear magnetic resonance (NMR), com-

DAVIS er al.: IS BRAIN CANCER MORTALITY INCREASING? 193

puterized axial tomography (CAT), and radiographic procedures were culled, where available. As to the influence of improved diagnostic accuracy, data from the U.S. National Cancer Institute, Surveillance, Epidemiology and End Results (SEER) program on changes in diagnostic confirmation of incident cases diagnosed while living were obtained and evaluated for specific age groups for three time periods: 1973-1977,1978-1982,1983-1987. The SEERdataprovide continuous coverage for about 10% of the U.S. population since 1973.

RESULTS

Mortality Analyses

FIGURE 1 shows the diverging trends in different age groups of U.S. males ages45- 84 by means of a smoothed three-dimensional curve developed in the SYSTAT program. Analysesof mortalitydatafortheUnitedStatesshowthatratesfor 1986for white males and females a g e s W fell more than 20% from those of 1969, while they increased more than 80% for persons ages65-84. FIGURE 2reveals that in 1986, brain and other central nervous system cancer mortality in US. white females increased

FIGURE 1. Mortality data for brain and other nervous system cancer (ICD 191,192) for males in the United States. Smoothed three-dimensional curve developed by the SYSTAT program.

10

0

110

00

40

_ .....

....

....

....

....

....

....

....

....

-....

......

......

......

......

......

...

- .....

......

......

......

......

......

..

....

...

......

.

......

. ....

....

...

....

....

...

......

......

......

......

......

......

......

......

......

......

......

......

. a’

a’

....

....

..

....

....

....

....

....

....

..

.

....

....

20

1 ...

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

......

/

L

0

-PO

- 40

iom

1970

197

1 19

711

i97

a 1

974

1976

197

8 19

77 i

o7

~

1979

1~

80

ig

ai I

QB

P ism

a 19

14 1

986

1988

198

7

FIG

UR

E 2.

Per

cent

cha

nge

in b

rain

and

oth

er n

ervo

us sy

stem

canc

er (I

CD

191

, 192

) mor

talit

y fo

r whi

te f

emal

es in

the U

nite

d St

ates

. 1%

9 m

orta

lity

rate

is

base

line.

(-

0-) ag

es 0

-24,

(- -k

-) ag

es 2

5-44

, (-*

-)

ages

45-

64, (

--n-)

ages

65-

84.

DAWS rt al.: IS BRAIN CANCER MORTALITY INCREASING? 195

30

88% over the 1969 rate, with increases continuing throughout the time period, while rates fell for females aged 64 and younger.

Comparison of male to female rates at different age groups reveals a consistent maleexcessinallcountries. TheU.S. rateformalesages65-84in 1986of20.4/100,000 still exceeded that of females, which was 14.2/100,000, although the female rate increased more rapidly during the time period. Even in the youngest age groups, an excess of male deaths is evident.

West Germany and the United Kingdom experienced similar age-specific trends to those in the United States (FIG. 3), with greater increases in the oldest age groups, greater rates of increase in the past decade, and a consistent excess of male deaths. In West GermansandBritonsaged65-84,ratesof braincancer andother nervoussystem mortality more than doubled; decreases occurred only in those ages 0-24. The rates of increase in older persons were similar for males and females. Males ages 65-84 in West Germany showed the higher age-specific rate in 1987, of 16/1OO,OOO, compared to 11.4/100,000 for females. In the United Kingdom, male rates increased more than threefold in this age group, from 5.0 to 16.5/100,000, with increases continuing throughout the last decade. Only the youngest group of West Germans ages &24 showed a decrease in deaths due to brain and other nervous system cancer throughout the period of study, with the rate for males higher than for females.

In contrast, over this same time period, France and Italy experienced increasing rates of brain and other nervous system cancers mortality in all age groups, with increases being smallest in the youngest age groups and greatest in the older groups

-

zo'

8

a

FIGURE 3. Mortality data for males in (a) Japan, (b) France, (c) West Germany, (d) Italy, (e) England. See FIG. 1 ffr details.

1%

30 i b

ANNALS NEW YORK ACADEMY OF SCIENCES

FIGURE 3. (Continued)

DAVIS el a&: IS BRAIN CANCER MORTALITY INCREASING? 197

d

e

FIGURE 3. (Continued)


(FIG. 4). Mortalityinmalesandfemalesages45-84in 1986increased at least 40% over the 1968 rate for this same age group. For those ages 65-84, increases in 1986 ranged from double to triple the 1968 rate. Rates of increase were greatest in females, although themaleratesremainhigher. The youngest agegroup,O-24, experienced the slowest rate of increase, with the rate for males higher than for females.

Quality ofAscertainment

To investigate the possible effects of diagnostic biases, analyses of 60,000 cancer deaths in 1985 and 1986 were compared to their microscopically confirmed hospital diagnosesin the SEER system. Data from theThirdNationalCancer Survey (TCNS) and SEER show that if the International Classification of Disease codes 191 and 192 are combined, the percent agreement between underlying cause of death from death certificate and SEER registry diagnosis is 93.1% overall. The percent agreement between original diagnosis of brain cancer and death certificates for persons under and over age 65 (in 1985-1986) was81.7% for the younger group and 83.3% for those over age 65.

Analyses of time trends in rates and types of diagnostic confirmation in the SEER program are illuminating. In each of three time periods (1973-1977,1978-1982, and 1983-1987) the overall rate of diagnostic confirmation for brain and other nervous

a

s L?

P

20

15

10

5

FIGURE 4. Mortality data for brain and other nervous system cancer (ICD 191,192) for femalcs in (a) Unites States, (b) Japan, (c) Francc, (d) West Germany, (e) Italy, and ( f ) England.

DAVIS ef al.: IS BRAIN CANCER MORTALITY INCREASING?

b

C

FIGURE 4 (Continued)


2o I d

i 5

2o I e


DAVIS et 01.: IS BRAIN CANCER MORTALITY INCREASING? 201

f


system cancer by other microscopic evaluations, such as biopsy or radiographic tests, such as CAT scan, remained above 96% for the oldest age groups, as well as for younger persons, diagnosed before death (TABLE 1). While 35.7% of persons ages 80- 84 had their brain and other central nervous system cancer diagnosed by radiography without microscopicconfirmationin 1973-1977,41% weresodiagnosedin 1983-1987.

DISCUSSION

Age-specific trends in brain and other nervous system mortality show markedly different patterns, with older age persons evidencing substantial increases in their 1986 rates compared with those for 1969 in the United States, France, Italy, West Germany, and the United Kingdom. These increases steadily continued during the past decade, with males experiencing higher rates than females.

While one cannot generalize readily between countries, an examination of such potentially confounding factors as changes in rate and types of diagnostic confirmation in the U.S. SEER program did not reveal major shifts in diagnostic technology that could, themselves, account for the recorded increases in mortality. Ninety-seven percent of these cancers have been microscopically or radiographically confirmed in incident cases throughout the time period in the United States. Moreover, these increases in elderly brain and other nervous system cancer mortality occurred before CAT scans and MRI were widely available in several major industrial countries in both sexes; also, they are continuing in recent years, when changesin access to care are unlikely to have occurred.

m2 ANNALS NEW YORK ACADEMY OF SCIENCES

Regarding potential diagnostic influences on these recorded increases in mortality, a recent United Nations publication indicates that by the mid-l980s, CAT and MRI scans became more widely available in all the countries examined in this paper.' These sophisticated imaging techniques are not used with equal enthusiasm in all countries and, in fact, are closely controlled in the United Kingdom and West

Examination of the slopes of brain cancer increases show that the greatest increases occurred well before the possible widespread use of CATscans, and that the increase steadily continues after their general introduction. In fact, during the period of study, all age groups had increased potential for receiving CAT scans; yet increases occurred only in the older population.

Since 1975, the number of radionuclide brain scans in the United States has continued to decline. The rate of increase in the use of CAT scans peaked in 1981.

TABLE I. Diagnostic Confirmation of Brain and Other Nervous System Cancers in the US. SEER Program (1973-1987)

Age Microscopic Confirmation Radiography Without Total Count (Year) (Percent) (Percent)

1913-1971 6 5 i 65-69 70-14

80-84 75-19

n5+

1918-1 982 65+ 65-69 70-74 15-79 80-84 85 +

1983-1987 65+ 65-69 70-74 15-79 80-84 85+

80.0 81.4 18.7 15.5 59.5 54.6

14.6

16.3 69.0 64.7 44.6

82.1

77.7 88.2 83.5 73.0 55.0 39.8

15.6 10.3 15.1 18.9 35.1 36.4

23.3 15.6 21.3 29.6 33.7 53.9

19.6 9.8

15.0 23.3 41.0 51.1

1,275 554 37 1 233

84 33

1,783 649 531 348 190 65

2,264 806 632 459 249 118

Cranial scans constituted 75% of such procedures in 1981, but 63% in 1983 in the United States." Increased rates in older persons could reflect the fact that such technology has allowed the easier detection of brain cancers, as compared with radionuclide-labeled brain scans that previously involved the invasive injection of radio-opaque materials. However, the change in use of such non-microscopic

DAVIS cr 01.: IS BRAIN CANCER MORTALITY INCREASING? 203

techniques in older persons is estimated to have been modest (TABLE 1). Other sources of error have not been evaluated in this paper. Thus, improved diagnoses of deaths previously misattributed to stroke and cerebrovascular disease may also contribute to these recorded trends in brain cancer.

CONCLUSIONS

The reported rates of brain and other central nervous system cancer mortality has increased significantly from 1%8 to 1986-1987 in older males and females in several major industrial countries. It seems likely that at least some of these increases reflect improved ascertainment, diagnostic technology, or other artefacts. Their occurrence simultaneously in older persons in the world’s leading trading nations should be carefully assessed, to identify causes, and to assist health planners in anticipating future demands for care.

SUMMARY

This paper analyzes recent age-specific trends in brain and other central nervous system cancer mortality from 1968 to 1986-1987 in the United States, United King- dom, Italy, France, and West Germany. It also examines changes in the use of diagnostic confirmation technology in the U.S. SEER program from 1973 to 1987 to estimate the influence of such factors on recent mortality trends. Other sources of error have not been evaluated in this paper. In the United States and Sweden, deaths due to brain and other central nervous system cancer, adjusted to the overall population, are unchanging. However, age-specific analyses of brain and other nervous system cancer in six major industrial countries show markedly different trends at different age groups, with drastic increases in brain tumor rates in the old: rates doubled in persons ages 75 to 84. In the United States, microscopic or radiographicconfirmationoccurred throughout this time periodin 96% of allincident casesof brain and othercentralnervoussystemcancersdiagnosed before death in the SEER program, with older persons receiving consistently more radiographic tests than younger persons. The use of diagnostic technology may change over time and across populations, but it is not known to what extent it accounts for these increasing trends, which require careful additional study.

REFERENCES

1. DAVIS, D. L. & J. SCHWARTZ. 1988. Trends in cancer mortality: US white males and females, 1968-83. Lancet i : 633-636.

2. National Cancer Institute. 1987.1986 Annual Cancer Statistics Review. National Institutes of Health. Bethesda, MD.

3. BAHEMUKA M., E. W. MASSEY & B. S. SCHOENBERG. 1988. International mortality from primary nervous system neoplasms: Distribution and trends. Int. J. Epidemiol. 171: 33-38.

4. ANNEGERS,J.F.,B.S.SCHOWBERG,H.OKAZAKI &L.T. K~~~~~~.1981.Epidemiologi~study of primary intracranial neoplasms. Arch. Neurol. 3& 217-219.


5. LEVI, F. & C. LA VECCHIA. 1989. Trends in brain cancer [letter]. Lancet ii: 917. 6. AHLBOM, A. & Y. RODVALL. 1989. Brain tumor trends [letter]. Lancet k12-72. 7. WHO Manual of the International Statistical Classification of Diseases, Injuries and Causes

of Death. 1979. World Health Organization. Geneva. 8. United Nations Scientific Committee on the Effects of Atomic Radiation. 1988. 1988

Report to the General Assembly, with Annexes Sources, Effects and Risks of Ionizing Radiation.

9. STEINBERG, E. P., J. E. SISK & K. E. bcm. 1985. X-ray CT and magnetic resonance imagers: Diffusion patterns and policy issues. New Engl. J. Med. 3Ue 859-864.

10. TERHORST, L. L. 1984. National survey of computed tomography unit capacity: An Update. Special report. Radiology 153t 207-210.

11. United Nations Scientific Committee on the Effects of Atomic Radiation. 1988. Sources, Effects and Risks of Ionizing Radiation. Report to the General Assembly, Geneva.

Is Brain Cancer Mortality Increasing in Industrial Countries?

Documents