cancer incidence in murcia, spain, in 1982: first results from a population-based cancer registry

7
Int. J. Cancer: 38, 1-7 (1986) 0 1986 Alan R. Liss, Inc. CANCER INCIDENCE IN MURCIA, SPAIN, IN 1982: FIRST RESULTS FROM A POPULATION-BASED CANCER REGISTRY Carmen NAVARRO”~, Doming0 PEREZ-FLORES~ and Michel P. COLE MAN^ ‘Consejeria de Sanidad, Consumo y Servicios Sociales, Seccibn de Epidemiologia, Ronda de Levante, 30008 Murcia; ’Depat?amento de Bioestadistica, Facultad de Medicina, 300 Murcia, Spain; 31mperial Cancer Research Fund, Cancer Epidemiology and Clinical Trials Unit, Gibson Building, Radclcfe Infirmary, Oxford Ox2 6HE, UK. This report presents the first findings of the Murcia Cancer Registry, a population-based cancer registry set up in May 1981 in the Murcia region of southeast Spain (955,487 inhabitants). Descriptive epidemiologi- cal methods have been applied to study cancer inci- dence in 1982. The validity of cancer registration has been assessed for both completeness and accuracy. Altogether 1,987 cases were registered in 1982, the crude (all-ages) annual incidence rates per 100,000 being 238 in males and 179 in females. Excluding non- melanoma skin cancer, which seems to be incom- pletely registered, age-standardized rates for Murcia are very similar to those of other registries in Spain in males, but lower in females. The risk was higher in males than in females for all sites and for the great majority of specific sites, especially larynx, oesopha- gus and bladder. The age-standardized rate for cancer of the larynx was 26 times higher in men than in women. Lung cancer was the most common cancer in males: the risk was 8 times that in females. Lung cancer rates were higher, in both males and females, than in other Spanish registries. Cancer of the larynx was the second most common site in males when either truncated rates (35-64 years) or cumulative in- cidence rates up to 64 years of age were used. The age-standardized rate (I8 per 100,000) supports pre- vious studies suggestingthat the risk for this cancer in Spain and other Mediterranean countries is very high. The lung/laqhx rate ratio in men was two. Cancer of the breast is the most common cancer among women, as in other registries in Spain and in most other coun- tries. The age-standardized rate (29.4 per 100,000) is lower than breast cancer rates elsewhere in Spain. This difference may be partly explained by incom- pleteness of ascertainment in Murcia. The incidence rate for cancer of the cervix uteri was 4.9 per 100,000, excluding carcinoma in situ. Despite the limitations of the data, cervical cancer incidence in Murcia is likely to be similar to that in other regions of Spain. Cancer accounted for 19.6% of all deaths in Spain in 1979. Cancer mortality rates have increased steadily since the beginning of the century, particularly in men. Between 1951 and 1979 this increase was mostly due to lung cancer in males and to breast cancer in females. Only mortality from stomach cancer in men and from stomach and uterus cancer in women decreased to any extent (Martinez, 1984). Crude cancer mortality rates per 100,OOO in Murcia in 1978 were 170.3 for men and 128.3 for women. The standardized mortality ratios (SMRs) for all cancers were 93.2 and 104.4, respectively (standard rates Spain 1978). Cancer mortality trends in Murcia region show a similar pattern to those in the rest of Spain (Navarro, 1981; Navarro and Ramos, 1982). The Murcia Cancer Registry was set up in May 1981 as a population-based cancer registry for Murcia re- gion, southeast Spain, in order to monitor the occur- rence of new cancers. At that time, mortality statistics showed increasing trends for a majority of cancers, and there were no incidence data for Murcia; else- where in Spain, incidence data were available from only two registries, Zaragoza and Navarra. The region constituted one of the recently created autonomous communities in Spain (Fig. l), with a stable population of nearly one million and good socio-economic data facilities (Caja de Ahorros de Alicante y Murcia, 1982, 1983). Pathologists and oncologists at the main public hospitals and private pathologists all supported the project and offered their collaboration. The Regional Department of Health funded the proposal. The aim of this report is to present the first results from the Murcia Registry, corresponding to 1982. The specific objectives are to outline the descriptive epide- miology of cancer in Murcia region, based on the incidence rates in 1982, and to compare Murcia with other regions, particularly in the Mediterranean area. MATERIAL AND METHODS Population at risk The population at risk comprises all residents of Murcia region. At the last census in March 1981 there were 955,487 inhabitants (469,339 males, 486,148 fe- males) (Instituto Nacional de Estadistica, 1982, 1984). A quarter of the population lives in rural areas. The mean population density is 85.71 inhabitants/km2. In 1982, 26.2% of the employed population worked in agriculture, a slight increase since 1976. This is in contrast to Spain as a whole, where the proportion working in agriculture fell from 21.5% to 18.3% over the same period. The percentage working in industry (25%) was the same in Murcia as in the rest of Spain. Unemployment was around 17% of the economically active population, similar to that in Spain as a whole (Caja de Ahorros de Alicante y Murcia, 1982, 1983). Murcia Cancer Registry The main objective of the Registry is to describe the nature and extent of cancer in the population of Mur- cia, and in particular to measure incidence rates and long-term cancer incidence trends. A secondary objec- 4T0 whom reprint requests should be sent. ’The following people comprise the Scientific Advisory Com- mittee of the Murcia Cancer Registry: A. Coll, A. Fernandez Nafria, J. Fedndez Perez, J. Garcia Esta~i, F. Hernrindez Cdvo, A. Morales, A. Navarrete, G. Ortufio, J.R. Rodriguez, M. Rodriguez Bermejo, J. Ruiz Jimbnez, M.J. Tormo. Received: November 19, 1985 and in revised form February 8, 1986.

Upload: carmen-navarro

Post on 11-Jun-2016

215 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

Int. J. Cancer: 38, 1-7 (1986) 0 1986 Alan R. Liss, Inc.

CANCER INCIDENCE IN MURCIA, SPAIN, IN 1982: FIRST RESULTS FROM A POPULATION-BASED CANCER REGISTRY Carmen NAVARRO”~, Doming0 PEREZ-FLORES~ and Michel P. COLE MAN^ ‘Consejeria de Sanidad, Consumo y Servicios Sociales, Seccibn de Epidemiologia, Ronda de Levante, 30008 Murcia; ’Depat?amento de Bioestadistica, Facultad de Medicina, 300 Murcia, Spain; 31mperial Cancer Research Fund, Cancer Epidemiology and Clinical Trials Unit, Gibson Building, Radclcfe Infirmary, Oxford Ox2 6HE, UK.

This report presents the f irst findings of the Murcia Cancer Registry, a population-based cancer registry set up in May 1981 in the Murcia region of southeast Spain (955,487 inhabitants). Descriptive epidemiologi- cal methods have been applied to study cancer inci- dence in 1982. The validity of cancer registration has been assessed for both completeness and accuracy. Altogether 1,987 cases were registered in 1982, the crude (all-ages) annual incidence rates per 100,000 being 238 in males and 179 in females. Excluding non- melanoma skin cancer, which seems to be incom- pletely registered, age-standardized rates for Murcia are very similar t o those of other registries in Spain in males, but lower in females. The risk was higher in males than in females for all sites and for the great majority of specific sites, especially larynx, oesopha- gus and bladder. The age-standardized rate for cancer of the larynx was 26 times higher in men than in women. Lung cancer was the most common cancer in males: the r isk was 8 times that in females. Lung cancer rates were higher, in both males and females, than in other Spanish registries. Cancer of the larynx was the second most common site in males when either truncated rates (35-64 years) or cumulative in- cidence rates up to 64 years of age were used. The age-standardized rate (I8 per 100,000) supports pre- vious studies suggesting that the risk for this cancer in Spain and other Mediterranean countries is very high. The lung/laqhx rate ratio in men was two. Cancer of the breast i s the most common cancer among women, as in other registries in Spain and in most other coun- tries. The age-standardized rate (29.4 per 100,000) is lower than breast cancer rates elsewhere in Spain. This difference may be partly explained by incom- pleteness of ascertainment in Murcia. The incidence rate for cancer of the cervix uteri was 4.9 per 100,000, excluding carcinoma in situ. Despite the limitations of the data, cervical cancer incidence in Murcia is likely to be similar to that in other regions of Spain.

Cancer accounted for 19.6% of all deaths in Spain in 1979. Cancer mortality rates have increased steadily since the beginning of the century, particularly in men. Between 1951 and 1979 this increase was mostly due to lung cancer in males and to breast cancer in females. Only mortality from stomach cancer in men and from stomach and uterus cancer in women decreased to any extent (Martinez, 1984).

Crude cancer mortality rates per 100,OOO in Murcia in 1978 were 170.3 for men and 128.3 for women. The standardized mortality ratios (SMRs) for all cancers were 93.2 and 104.4, respectively (standard rates Spain 1978). Cancer mortality trends in Murcia region show a similar pattern to those in the rest of Spain (Navarro, 1981; Navarro and Ramos, 1982).

The Murcia Cancer Registry was set up in May 1981 as a population-based cancer registry for Murcia re- gion, southeast Spain, in order to monitor the occur- rence of new cancers. At that time, mortality statistics

showed increasing trends for a majority of cancers, and there were no incidence data for Murcia; else- where in Spain, incidence data were available from only two registries, Zaragoza and Navarra. The region constituted one of the recently created autonomous communities in Spain (Fig. l) , with a stable population of nearly one million and good socio-economic data facilities (Caja de Ahorros de Alicante y Murcia, 1982, 1983). Pathologists and oncologists at the main public hospitals and private pathologists all supported the project and offered their collaboration. The Regional Department of Health funded the proposal.

The aim of this report is to present the first results from the Murcia Registry, corresponding to 1982. The specific objectives are to outline the descriptive epide- miology of cancer in Murcia region, based on the incidence rates in 1982, and to compare Murcia with other regions, particularly in the Mediterranean area.

MATERIAL AND METHODS

Population at risk The population at risk comprises all residents of

Murcia region. At the last census in March 1981 there were 955,487 inhabitants (469,339 males, 486,148 fe- males) (Instituto Nacional de Estadistica, 1982, 1984). A quarter of the population lives in rural areas. The mean population density is 85.71 inhabitants/km2. In 1982, 26.2% of the employed population worked in agriculture, a slight increase since 1976. This is in contrast to Spain as a whole, where the proportion working in agriculture fell from 21.5% to 18.3% over the same period. The percentage working in industry (25%) was the same in Murcia as in the rest of Spain. Unemployment was around 17% of the economically active population, similar to that in Spain as a whole (Caja de Ahorros de Alicante y Murcia, 1982, 1983). Murcia Cancer Registry

The main objective of the Registry is to describe the nature and extent of cancer in the population of Mur- cia, and in particular to measure incidence rates and long-term cancer incidence trends. A secondary objec-

4T0 whom reprint requests should be sent.

’The following people comprise the Scientific Advisory Com- mittee of the Murcia Cancer Registry: A. Coll, A. Fernandez Nafria, J. Fedndez Perez, J. Garcia Esta~i, F. Hernrindez Cdvo, A. Morales, A. Navarrete, G. Ortufio, J.R. Rodriguez, M. Rodriguez Bermejo, J . Ruiz Jimbnez, M.J. Tormo.

Received: November 19, 1985 and in revised form February 8, 1986.

Page 2: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

2 NAVARRO ET AL.

Asturias(l978) Guipuzwa(l982)

TarrapanaY1980)

- Autonomous community boundary Province boundary

FIGURE 1 - Population-based cancer registries in Spain, 1985 (first year of operation). Registries marked with an asterisk have published incidence rates.

tive is to assist epidemiological research. The registry will also provide cancer statistics to contribute to the understanding of cancer incidence in Spain and in the Mediterranean area.

A case for the registry is defined as a malignant neoplasm not previously registered in a person resident in Murcia at the time of diagnosis, or of death if a death certificate is the only source of data, whenever the diagnosis was made or the death occurred. Transi- tional-cell papillomas of the bladder are included, fol- lowing WHO recommendations (Muir, 1976).

Cases are sought actively by the Registry staff in the departments of pathology, haematology and radiother- apy of the public hospitals in the region. Hospital medical record files are used for completing the tu- mour record. One isolated radiotherapy clinic is vis- ited once a year. Private pathologists collaborate regularly with the registry; 2 out of 3 send the data to the Registry themselves; tumour records are com- pleted by the Registry staff at the private hospitals. Death certificates are reviewed monthly at the regional office of the Instituto Nacional de Estadistica (INE); every death certificate of a Murcia resident with a mention of cancer is abstracted and checked against the Registry’s files. Histological verification is sought in all cases.

The items of data collected and the registration pro- cedures are based on the recommendations of Mac- Lennan et al. (1978). The International Classification of Diseases for Oncology (ICD-0) (Organizacion Pan- americana de la Salud, 1977) is used for coding both site and morphology. Ancillary data sources

The National Children’s Cancer Registry (NCCR) has been used to assess completeness of case ascertain- ment for children, since its data collection is indepen- dent of the Murcia Registry. It is based on the voluntary participation of a group of collaborating centres that systematically notify all treated cases of cancer in

children aged 0-15 years to a central registry in Val- encia. The great majority of childhood cancer patients are treated in the highly specialized hospitals collabo- rating in the scheme (Registro Nacional de Tumores Infantiles, 1984). The NCCR provided a copy of each cancer registration for a child resident in Murcia.

The National Health Institute (INSALUD) maintains a file of patients who apply for a refund of travelling expenses for medical treatment and/or diagnosis out- side their region of residence. These records have been reviewed and the records of cancer patients first re- questing a refund in 1982 were abstracted.

Statistical methods Age- and sex-specific incidence rates were derived

for each cancer site and each 5-year age group from 5-84. Separate rates were obtained for infants (less than 1 year) and young children (1-4 years) and for persons aged 85 years and over.

No single summary measure of cancer incidence is completely adequate for all comparisons, and we have used 3 different standardized incidence rates. The age- standardized rate is obtained by direct standardization to the world standard population (Waterhouse et al . , 1982), and is used for comparisons with other regis- tries unless otherwise stated. Incidence rates are less reliable for older ages (Doll and Peto, 1981), however, and the truncated world standard population (35-64 years) is used for some comparisons. The cumulative rate is used to summarize incidence from birth to 64 years of age: this rate can be interpreted as the cumu- lative life-time risk of cancer up to that age, or as a rate directly standardized to a population with the same number of people in each age group (Day, 1976, 1982).

Completeness was assessed by several methods, based on death certificates (Muir, 1982), independent case identification (Goldberg et al . , 1980; Nwene and Smith, 1982) and historic data methods (Goldberg et al., 1980). Accuracy of registration was checked by

Page 3: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

3 CANCER INCIDENCE IN MURCIA, SPAIN

the proportion of cases with histological verification (Muir, 1982): this material has been published else- where (Navarro, 1985).

The data sets have been checked for internal consis- tency (ranges of codes, incompatible site and sex, etc .), both manually and automatically (Haenszel and Lourie, 1966; MacLennan et al . , 1978).

RESULTS

A total of 1,987 incident tumours were registered in 1982; 1,118 (56%) in males and 869 in females. Over half (1,062, 53 %) were registered in life, and all but 32 (3%) of these were histologically verified. The remainder (961, 47%) were ascertained from a death certificate in persons for whom no previous record of cancer diagnosis could be found. A further 337 cancer deaths, certified in 1982 in persons diagnosed in 1981 or earlier (and thus prevalent cases) are not included.

Of the 1,987 incident tumours, the great majority (1,644, 83%) were considered primary malignant tu- mours. There were 271 (14%) neoplasms of uncertain behaviour (benign or malignant), and 6 carcinomas in situ; 66 tumours ( 3 % ) were either metastatic or of uncertain primary site.

Histological verification was obtained for half of all tumours registered (51 % in males, 49% in females); the proportions were higher among patients under 65 (57 % , 62 %) and fell to less than a third in those aged 75 and over (33%, 31%). For many individual sites, however, more than 75% of tumours were verified histologically (Table I).

Incidence by age and sex, all sites The all-sites cancer incidence rate per 100,OOO was

238.2 in men and 178.8 in women; age-standardized rates were 206.0 and 133.5, respectively (Table 11). Male and female rates are broadly similar below age 50, but male rates exceed female rates at all ages above 50 years. Between the ages of 35 and 64 years, the all-

sites incidence rate increases with approximately the fourth power of age in both sexes (Fig. 2). Cancer incidence by site

The most common cancer in males was lung cancer (38.6 per l00,OOO), which comprised almost one in every 5 cancers (19%), and was twice as common as any other cancer in men (Table 11). The next 3 sites were of equal rank-large intestine including rectum, and stomach and bladder-each about 9% of the total. Laryngeal cancer ranked fifth, with 8.5%. Among women, breast cancer was the most common cancer (29.4 per 100,000), also providing a fifth (20%) of all female cancers. Cancer of the large intestine ranked second, with 12.8% of female cancers. Cancer risk in males exceeded that in females at most sites, particu- larly larynx, oesophagus and bladder. Trachea, bronchus and lung (ICD-0 162)

Lung cancer was the commonest tumour in men at all ages over 50. The crude (all-ages) incidence rate was 45.2 per lOO,OOO, more than 6 times the rate in females. The male-female ratio is almost 8:l when standardized rates are compared (Table 11). The cu- mulative incidence to age 64 was 3.25% in males and 0.35% in females. L a r y n ~ (ICD-0 161)

Although cancer of the larynx ranks fifth overall in males, it was as common as lung cancer in the 35-49 years age range. This is reflected in the truncated (35- 64 years) and cumulative (0-64 years) incidence rates, in which it ranks second to lung cancer (Table 11). The lung/larynx ratio is about 2: 1 for males at all ages, and only 1.8:l at ages below 64 years. Among females the risk is much lower than in males at all ages. The male- female ratio for laryngeal cancer is 25.9 (ratio of age- standardized rates)-higher than for any other cancer. Large intestine (ICD-0 153 and 154)

Cancers of the colon and rectum rank second in frequency in both sexes, although in males the age- standardized rate (18.7 per 100,000) is slightly lower

TABLE I - MURCIA CANCER REGISTRY, 1982. PERCENTAGE OF HISTOLOGICALLY VERIFIED CASES’ BY SITE AND SEX

Site Males Females ICD-0 code

140-9 Lip, buccal cavity and pharynx 80.4 77.8 150 Oesophagus 60.7 50.0 151 Stomach 43.3 44.8 153 Colon 58.9 51.6 154 Rectum 73.5 83.7 is5 Liver 156 Gall-bladder

25.0 19.4 33.3 75.0

157 Pancreas 35.0 23.3 161 Larynx 75.8 40.0 162 Trachea, lung and bronchus 22.9 13.9 169 Haematopoietic and reticulo- 2.9 13.6

173 Skin 96.8 90.5 Female breast - 46.3

180, 182 Cervix and corpus uteri - 78.0 174

183 185 Prostate 38.7 188 Urinary bladder 74.0 66.8 191-2 Nervous system 48.6 48.0 140-199 All sites 51.2 48.6

endothelial

Ovary - 60.9 -

‘“Death certificate only” cases (47%) are included in these figures. Overall, 97% of turnours registered in life (53% of all turnours) were histologically verified.

Page 4: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

4 NAVARRO ET AL.

TABLE I1 - MURCIA CANCER REGISTRY, 1982. THE 10 MOST FREQUENT CANCERS, MALES AND FEMALES

Site Truncated

rate3 (35-64 yrs)

Crude Cumulative Number of % of incidence ASR' incidence

1oO.ooO (0-64 vrs) cases total rate per (world) rate=

Males 1 Lung 2 Colon-rectum 3 Stomach 4 Bladder 5 Larynx 6 Prostate 7 Skin 8 Unknown primary 9 Nervous system

10 Haematopoietic Other sites All sites

Females 1 Breast 2 Colon-rectum 3 Stomach 4 Unknown primary 5 Skin 6 Liver I Lung 8 Uterus NOS 9 Pancreas

10 Cervix Other sites All sites

212 105 104 104 95 80 63 48 37 34

236 1118

175 111 66 56 42 36 36 34 30 26

257 869

19.0 9.4 9.3 9.3 8.5 7.2 5.6 4.3 3.3 3.0

21.0 100.0

20.1 12.8 7.6 6.4 4.8 4.1 4.1 3.9 3.5 3.0

29.6 100.0

45.2 38.6 22.2 18.7 22.2 19.1 22.2 19.0 20.2 18.1 17.0 13.2 13.4 11.3 10.2 8.0 1.9 7.4 7.2 5.9

238.2 206.0 - -

36.0 22.8 13.6 11.5 8.6 7.4 1.4 7.0 6.2 5.3

178.8 -

29.4 15.9 9.6 8.5 6.4 4.8 5.0 5.3 3.9 4.9

133.5

3.25 1.31 1.39 1.34 1.80 0.57 0.85 0.52 0.69 0.27

15.61 -

2.47 1.24 0.62 0.70 0.46 0.37 0.35 0.50 0.25 0.48

10.79 -

59.3 19.0 24.0 28.2 33.9 3.8

15.9 9.5

16.3 3.6

282.4

63.8 11.6 12.8 15.8 8.7 8.2 8.3 9.3 3.9

11.8

228.1

-

-

'ASR (world) = Age-standardized rate (world population).~*Cumulative incidence rate (per cent). -31ncidence rates per 100,000, standardized to the truncated population.

3000 2000

1000

Annual incidence rate per loo, 000

100

1c - 20 40 60 8 0

A g e ( yea rs )

FIGURE 2 - All-sites cancer incidence by age (20-80), Murcia Cancer Registry, 1982. Logarithmic scales.

than for bladder or stomach cancer. The rates for colon cancer slightly exceed those for rectal cancer in both sexes and at most ages: the ratio of age-standardized rates for colon and rectum is 1 . 1 in males and 1.3 in females. For the large intestine as a whole, male rates exceed female rates by a ratio of 1.2: 1.

Stomach (ICD-0 151) Stomach cancer ranked third in frequency in both

sexes. The crude incidence rate per 100,000 was 22.2 in males and 13.6 in females; the male excess was 2- fold when age-standardized rates were compared.

Breast (ICD-0 174 females, 175 males) Breast cancer represents the main risk of cancer in

women, with a crude incidence rate of 36 per 100,OOO. Age-specific incidence rates increase steadily until age 55-59, then remain at about 110 per 100,OOO until age 79. Breast cancer is the most common cancer in women at all ages except 70-84 years. The risk of breast cancer is about 6 times the risk of cervical cancer. Breast cancer was rare in men (3 cases), less than 2% of the frequency in women.

DISCUSSION

Cancer incidence statistics from Spain are already available for three population-based cancer registries in Navarra (Vinls, 1982), Zaragoza (Zubiri, 1982) and Tarragona (Borras et al . , 1984). Incidence rates age- standardized to the world standard population are about 230 and 160 per 100,OOO per year for males and fe- males, respectively. These figures are similar to those

Page 5: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

CANCER INCIDENCE IN MURCIA, SPAIN 5

for Poland, Norway, Hungary and Yugoslavia, but lower than those for the United Kingdom and Varese (Italy) (Waterhouse et al., 1982). Among individual sites of cancer in Spain, the most remarkable is cancer of the larynx in males, for which the incidence rates (15 per 100,OOO in Navarra and 11.5 in Tarragona) are among the highest in the world (Muir and Nectoux, 1982). Incidence rates of cancer of the cervix are much lower in Spain (Navarra 3.9 per 100,OOO; Tarragona 4.7; Zaragoza 5.6) than in many other countries (Par- kin et al., 1984), although the rate in Santa Cruz de Tenerife is 12.3 per 100,OOO (Sierra Lopez et al., 1984; Del Moral et al., 1984). These differences have not yet been explained. A high risk from cancer of the salivary glands in Spain has also been noted (Mould, 1983).

The all-sites cancer incidence rates in Murcia (206.0 per 100,000 in males, 135.5 in females) are somewhat lower than those recorded by other registries in Spain and those for the Mediterranean area (Waterhouse, 1982) (Table 111). Excluding skin cancer, which seems to be incompletely registered (Navarro, 1985), Murcia rates are similar to those of other registries in Spain in males, but lower in females.

The all-sites cancer risk is higher in males than in females, as in nearly all registries throughout the world. This is to be expected, since the risk is higher in males for the majority of individual sites. Truncated rates are also higher in males, again the most common situation. Registries with a male/female ratio of about unity for all-sites truncated rates are usually those with a high risk from breast cancer in women, given the predominance of breast cancer in premenopausal women (Chamberlain, 1982; Pike et al., 1983) relative to other sites. Thus the observed male/female ratio in Murcia for all sites is high ( lS ) , as female breast cancer rates are lower than in other European regis- tries (Table 111). On the other hand, the pattern of the age-specific rates by sex (Fig. 2 ) is similar to that in

other countries such as the United Kingdom (Office of Population Censuses and Surveys, 1981), and a major bias in the male/female ratio for all sites is unlikely.

Lung cancer Lung cancer rates in Murcia, in both males and

females, are higher than those recorded by other Span- ish registries, but lower than in many European coun- tries, such as the United Kingdom, Finland, Italy and Yugoslavia (Table 111).

It is not very likely that the higher risk in Murcia compared with other Spanish registries can be ex- plained only by the registration of prevalent cases, given the poor prognosis of cancer at this site. Mis- classification may partly account for the observed dif- ference, since a high proportion of cases are "death certificate only" (73.1% in males and 77.8% in fe- males) (Navarro, 1985). A recent study of the validity of death certificates in Murcia showed that there is some misclassification of deaths from other causes as due to lung cancer, but a similar number of deaths from lung cancer were certified as due to other causes (Navarro et al., 1984).

The risk of lung cancer is increasing in Spain, and there may be a small bias in comparing Murcia 1982 rates with Navarra or Zaragoza (1973-77 rates). How- ever, Tarragona rates correspond to 1981 and this registry is also new: the Murcia/Tarragona ratio of standardized rates for lung cancer is 1.3 for both males and females. Cancer of the larynx

Male incidence rates for cancer of the larynx in Murcia accord with previous studies which have sug- gested that the risk for this cancer in Spain, France and Italy is very high (Vifies, 1982; Zubiri, 1982; Borras et at., 1984; Muir and Nectoux, 1982; Izarzu- gaza, 1985). The lung/larynx ratio is 2: l in Murcia, while in the United States, for example, it is 1O:l

TABLE 111 - M I I R C I A CANCER KEGISTRY. 1982. CANCER INCIDENCE I N MURCIA ASD OTHER EUROPEAN CAN(W R E G I S T R I E S ~ . KATF PER ~w.ooo (AGE-STANDARDIZED TO WORLD WPULATIONJ. MALI.S A N D FEMALES. S E L t r r i i n SITES

Murria Navarra -.._. .I Slo\,cni.i Birmingham Vdrr.\e (Yugoslavia) (U.K.) (Italy) "erlrllarn

1973- I976 1973-1976 1976-1977 1973-1976 Site (Spain) (Spain) 1982 1973-1977

Males Stomach 19.1 34.8 17.1 42.0 22.1 Colon-rectum 18.7 20.4 36.0 18.9 33.0 Larynx 18.1 15.0 4.2 7.3 4.0 Lung 38.6 23.5 51.8 55.6 79.9 Prostate 13.2 17.6 23.6 15.8 18.6 Bladder 19.0 15.4 22.0 7.1 17.1 All sites 206.0 230.2 273.0 242.0 276.1 All sites 194.2 201.9 245.3 221.9 246.9

but skin Females

Stomach Colon-rectum Lung Breast Cervix Corpus uteri Bladder All sites All sites

9.6 18.2 8.9 18.0 10.1 15.9 14.6 29.2 14.5 24.9 5.0 2.6 12.1 6.2 13.7

29.4 38.2 58.8 34.2 56.4 4.9 3.9 23.0 17.1 12.0 4.4 12.4 13.0 10.2 9.4 1.3 1.6 5.6 1.4 4.5

133.5 157.6 242.3 173.1 213.0 126.6 145.5 225.1 154.5 193.6

38.5 35.6 16.0 70.4 22.8 24.6

344.3 318.4

19.1 26.0

5.8 57.6 11.7 14.2 3.1

221.3 208.0

but skin 'Source: Waterhouse ef al. (1982).

Page 6: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

6 NAVARRO ET AL.

(Austin, 1982). The age-standardized rate (18.1 per 100,000) is higher than any published in the most recent volume of CIFC, but as 24.2% of cases were “death certificate only” (Navarro, 1985), there were probably some prevalent cases diagnosed before 1982; these would inflate the observed incidence rate.

Taking truncated rates (35-64 years), which are more reliable, the rate (33.9 per 100,000) is still among the highest, being exceeded by only two registries (Doubs, France: 39.9 per 100,000, and Varese, Italy: 34.2). Even if all “death certificate only” cases were assumed to be prevalent, the truncated rate would be 28.6 per 100,000-still among the highest. The truncated (35- 64) incidence rate would be overestimated if tumours of the trachea and hypopharynx were misclassified as larynx cancer, but this is unlikely: 84.3% of laryngeal cancers in this age group were histologically verified. Laryngeal cancer incidence is therefore likely to be very high, despite the limitations of the data.

Cancers of mouth, hypopharynx and oesophagus, which are known to be related to alcohol consumption (Doll and Peto, 1981), do not present high rates in Murcia or other Spanish registries, compared with those from other registries (Waterhouse et al., 1982). However, the mouth and hypopharynx are rare sites with problems of small numbers. The role of wine as a risk factor in the Mediterranean area is under inves- tigation by the IARC (Tuyns, 1982, Tuyns et at., 1984).

Working in agriculture has been suggested as a risk factor for certain cancers (Burmeister, 1981 ; Burmeis- ter et al . , 1983; Buesching, 1984; Cantor and Blair, 1984), in particular laryngeal cancer (Flanders et al., 1984). This association could be investigated in Mur- cia, where a high proportion of the active population works in agriculture and the risk for cancer of the larynx is also high. Breast cancer

Cancer of the breast is the most common cancer among women in Murcia, as in other registries in Spain and most countries throughout the world (Table 111). However, the age-standardized rate (29.4 per 100,OOO) is lower than elsewhere in Spain (38.2 in Navarra, 36.5 in Zaragoza) and far below the rates in North America and north and west Europe, which are around 80 per 100,OOO (Muir, 1976).

It is difficult to decide whether the difference in risk between Murcia and the other Spanish registries could be explained by incompleteness of ascertainment in Murcia. Given the relatively good prognosis of this cancer [49-64% 5-year survival (Hirayama et al., 1980)], it is likely that some prevalent cases will have been registered; this would tend to inflate the Murcia rates, and the observed difference may well reflect a genuinely lower breast cancer risk in Murcia.

Other sites The rate for cancer of the cervix is low compared to

those from other registries in the world, but similar to

the rates in Zaragoza, Navarra and Tarragona. The risk of bladder cancer is high but not as high as in Tarragona where the male rate (23.8 per 100,OOO, excluding transitional-cell papilloma) is among the highest in the world (Borras et al., 1984).

In the evaluation of cancer incidence data from a new registry, both incompleteness of ascertainment and registration of prevalent cases must be carefully considered. As described elsewhere (Navarro, 1985), “death certificate only” (DCO) cases include 4 differ- ent groups, one of which comprises prevalent cases. Over a long period, inflation of incidence due to reg- istration of prevalent cases is probably balanced by under-registration or late registration of incident cases, unless the proportion of DCO registrations is changing rapidly (Muir, 1982). But in a new registry they may be a more serious problem. In this report, inflation of incidence due to DCO cases is unlikely to be important for tumours with very poor survival, such as stomach, lung, oesophagus, pancreas and liver cancers and leu- kaemia, especially since the registry was working for 8 months before 1982. Problems of interpretation of incidence rates do arise, however, for sites known to have relatively good survival (Hirayama et al., 1980), such as lip, non-melanoma skin cancer, salivary gland, breast, larynx or corpus uteri.

Other indirect indicators of validity (“deaths in pe- riod” ; registration ratios standardized to Navarra rates) have also suggested a degree of incompleteness in cancer ascertainment in Murcia (Navarro, 1985), which emphasizes the limitations of the data sources currently available to us, mostly histology departments and ra- diotherapy units, for complete case-finding at every cancer site. Cases diagnosed and treated outside the region (1.4%), which may be lost to registration, are not very important for all-sites incidence rates.

Several changes will now be made in registry prac- tice, after assessment of data collection methods and of the validity of the data-base built up during the first full year of operation. Completeness of registration will be improved by recruiting sources of information not so far included, and by the establishment of routine data exchange with the National Children’s Cancer Registry, in order to register childhood tumours treated outside Murcia. Validity will be improved by more thorough inquiry before registration of cases originally identified from a death certificate.

ACKNOWLEDGEMENTS

We are most grateful to all the hospitals, institutions, and people who collaborate with the Murcia Cancer Registry, as well as to the registry staff, Ms. J . Tortosa and Ms. G. Sanchez. We thank the Section of Sanitary Information of the Spanish Ministry of Health and the National Children’s Cancer Registry for providing data for this report. We also thank the University Comput- ing Center, the regional office of the Instituto Nacional de Estadistica and the regional office of the Instituto Nacional de la Salud.

REFERENCES

AUSTIN, D.F., Larynx. In: D. Schottenfeld and J.F. Fraumeni, Jr. (eds.), Cancer epidemiology and prevention, pp. 554-563, Saunders, Philadelphia (1982).

BORRAS, J . , CREUS, J . , CALBET, J . , ANGLADE, L.L., BOSCH, F.X., VILADIU, P., CAMPILLO, M. , and MARTIN, M. , El cancer en Tarrugona 1980-1981. Asociacidn Espaiiola contra el Cancer,

Page 7: Cancer incidence in Murcia, Spain, in 1982: First results from a population-based cancer registry

7 CANCER INCIDENCE IN MURCIA, SPAIN

Tarragona (1984).

BUESCHING, D.P., Cancer mortality among farmers. J. nat. Can- cer Inst., 72, 503 (1984). BURMEISTER, L.F., Cancer mortality in Iowa fanners, 1971-78. J. nat. Cancer Inst., 66,461-464 (1981). BURMEISTER, L.F., EVERETT, G.D., VAN LIER, S.F., and ISAC- SON, P., Selected cancer mortality and farm practices in Iowa.

Amer. J. Epidemiol., 118, 72-77 (1983). CMA DE AHORROS DE ALICANTE Y MURCIA. Murcia- 1981, Datos y Series Estadisticas. Caja de Ahorros de Alicante y Murcia, Alicante (1982). CAlA DE AHORROS DE ALICANTE Y MURCIA. Murcia- 1982, Datos y Series Estadisticas. Caja de Ahorros de Alicante y Murcia, Alicante (1983). CANTOR, K.P., and BLAIR, A., Farming and mortality from multiple myeloma: a case-control study with the use of death certificates. J. nat. Cancer Inst., 72, 251-255 (1984). CHAMBERLAIN, J., Carcinoma of the female breast. In: D.L. Miller and R.D.T. Farmer. Epidemiology of diseases, pp. 289- 301, Blackwell, Oxford (1982). DAY, N.E., A new measure of age standardised incidence, the cumulative rate. In: J. Waterhouse, C.S. Muir, P. Correa and J. Powell (eds.), Cancer incidence in five continents. Vol. 111, pp. 443-446, IARC Scientific Publication No. 15, IARC, Lyon (1976). DAY, N.E., Cumulative rate and cumulative risk. In: J. Water- house, C.S. Muir, P. Correa and J. Powell (eds.), Cancer inci- dence infive continents. Vol. N , pp. 668-670, IARC Scientific Publication No. 42, IARC, Lyon (1982). DEL MORAL, A,, ASCUNCE, E., and PEREZ DE RADA, M.E., Epidemiologia del cancer de cuello uterino en Navarra. In: Com- munications presentkes a la IF rkunion du Groupe pour I’Epi- dkmiologie et I’Enregistrement du Cancer duns les Pays de Langue Latine, pp. 101-104, Registre Genevois des Tumeurs, Geneve (1984). DOLL, R., and PETo, R., The causes of cancer. J. nat Cancer Inst., 66, 1191-1308 (1981). FLANDERS. W.D.. CANN. C.I.. ROTHMAN. K.J.. and FRIED. P., Work-related risk factors for laryngeal cancer. Amer. J. Epide- miol., 119, 23-32 (1984). GOLBERF, J., GELFAND, H.M., and LEVY, P.S., Registry evalu- ation methods: a review and case study. Epidemiol. Rev., 2 , 210-220 (1980). HAENSZEL, W., and LOURIE, W.I., Quality control of data in large-scale cancer register program. Meth. Inform. Med. , 5 , 67- 74 (1966). HIRAYAMA, T., WATERHOUSE, J.A.H., and FRAUMENI, J.F., Can- cer risk by site. UICC Technical Report Series, Vol. 41, UICC, Geneva (1980). INSTITUTO NACIONAL DE ESTADISTICA. Poblaciones de derecho y hecho de los municipos espatioles. Censo de Poblacion de 1981. Ministerio de Economia y Comercio, Madrid (1982).

INSTITUTO NACIONAL DE ESTADISTICA. Censo de Poblacion, 1981, Tomo 111. Resultados provinciales. Murcia la Parte. Ministerio de Economia y Comercio, Madrid (1984). IZARZUGAZA, I., Estudio piloto para un registro de cancer en Euskadi. Gaceta Med. Bilbao, 82, 1-36 (1985). MACLENNAN, R., MUIR, C.S., STEINITZ, R., and WINKLER, A., Cancer registration and its techniques. IARC Scientific Publica- tion No. 21, IARC, Lyon (1978). MARTINEZ, C., Mortalidad por cancer en Espaiia. Tendencia evolutiva. Bol. Epidemiol. Semanal. 16, 121-123 (1984). MOULD, R.F., Cancer Statistics. A. Hilger, Bristol (1983). MUIR, C.S., Classification. In: J. Waterhouse, C.S. Muir, P. Correa and J. Powell (eds.), Cancer incidence inJive continents,

Vol 111, pp. 15-17, IARC Scientific Publication No. 15, IARC, Lyon (1976). MUIR, C.S., Comparability of data and reliability of registration. In: J. Waterhouse, C.S. Muir, P. Coma and J. Powell (eds.), Cancer incidence in five continents, Vol. N , pp. 55-64, IARC Scientific Publication No. 42, IARC, Lyon (1982). MUIR, C.S., and NECTOUX, J., International patterns of cancer. In: D. Schottenfeld and J.F. Fraumeni, Jr. (eds.), Cancer epide- miology and prevention, pp. 119-137, Saunders, Philadelphia (1982). NAVARRO, C., Estudio de la mortalidad en Murcia en 1977. Bol. Salud Murcia, 1, 18-20 (1981). NAVARRO, C. (ed.), La incidencia de cancer en Murcia en 1982. Serie Monografias, No. 1, Consejeria de Sanidad, Consumo y Servicios Sociales. Murcia (1985). NAVARRO, C., and RAMOS, E., Mortalidad por cancer en la region Murciana, 1901-1976. Rev. San. High. Pub., 56, 977- 998 (1982). NAVARRO, C., SANCHEZ, J.A., and MOLINA, J.A., Validez del Boletin Estadistico de Defuncidn corno fuente de datos en las estadisticas sobre el cancer. Un estudio preliminar. Bol. Salud Murcia, 4, 177-180 (1984). NWENE, U., and SMITH, A., Assessing completeness of cancer registration in the North-Western region of England by a method of independent comparison. Brit. J. Cancer, 46,635-639 (1982). OFFICE OF POPULATION CENSUSES AND SURVEYS, Cancer statis- tics: Registrations, 1975. Series MBI No. 5 , HM Stationery Office, London (1981). ORCANIZACION PANAMERICANA DE LA SALUD, Clasijcucion Inter- national de Enfermedades para Oncologia. OPS Publicacion Cientifica No. 345. OPS, Panama (1977). PARKJN, D.M., STIERNSWARD, J . , and MUIR, C.S., Estimates of the worldwide frequency of twelve major cancers. WHO Bull.,

PIKE, M.C., KRAILO, M.D., HENDERSON, B.E., CASAGRANDE, J.T., and HOEL, D.G., “Hormonal” risk factors, “breast tissue age” and the age incidence of breast cancer. Nature (Lond.),

62, 163-182 (1984).

303,767-770 (1983). RECISTRO NACIONAL DE TUMORES INFANTILES. Anuario de Estad- isticas Basicas del RNTI 1 (1980-1982). Anu. ESP. Pediar., 20, 187-342 (1984). SIERRA LOPEZ, A., DORESTE ALONSO, J., and DEL MORAL, A,, La epidemiologia del cancer de cuello uterino en Tenerife y en Navarra. In: Communications presenties a la I Y rkunion du Groupe pour 1’Epidkmiologie et 1 ‘Enregistrement du Cancer dans les Pays de Langue Latine, pp. 95-100, Madrid (1984). TUYNS, A.J., Alcohol. In: D. Schottenfeld and J.F. Fraumeni, Jr. (eds.), Cancer epidemiology and Prevention, pp. 293-303, Saunders, Philadelphia (1982). TUYNS, A.J., ESTEVE, J., RIBOLI, E., ARSAN, A,, ZUBIRI, A,,

SANCHO-GARNIER. and BENHAMOU, Laryngeal and pharyngeal cancer in south-western Europe. In: International Agency for Research on Cancer. Annual Report 1984. pp. 66-68, IARC, Lyon (1984). VINES, J., Cancer incidence in Spain, Navarra. In: J. Water- house, C.S. Muir, P. Coma and J. Powell (eds.), Cancer inci- dence infive continents, Vol. IV, pp. 526-529, IARC Scientific Publication No. 42, IARC, Lyon (1982). WATERHOUSE, J., MUIR, C.S., CORREA, P., and POWELL, J. (eds.), Cancer incidence infive continents, Vol. N, IARC Sci- entific Publication No. 42, IARC, Lyon (1982).

ZUBIRI, A., Cancer incidence in Spain, Zaragoza. In: J. Water- house, C.S. Muir, P. Correa and J. Powell (eds.), Cancer inci- dence infive continents, Vol. W , pp. 530-533, lARC Scientific Publication No. 42, IARC, Lyon (1982).

DEL MORAL, A., TERRACINI, B., BERRINO, F., RAIMOND, L.,