Cancer of mouth, pharynx and nasopharynx in Asian and Chinese immigrants resident in Thames regions

Cancer of mouth, pharynx and nasopharynx in Asian and Chinese immigrants resident in Thames regions

Oral Oncology 35 (1999) 471±475 www.elsevier.com/locate/oraloncology Cancer of mouth, pharynx and nasopharynx in Asian and Chinese immigrants reside...

123KB Sizes 12 Downloads 47 Views

Oral Oncology 35 (1999) 471±475

www.elsevier.com/locate/oraloncology

Cancer of mouth, pharynx and nasopharynx in Asian and Chinese immigrants resident in Thames regions K.A.A.S. Warnakulasuriya a,*, N.W. Johnson a, K.M. Linklater b, J. Bell b a

Department of Oral Medicine and Pathology/WHO Collaborating Centre for Oral Cancer and Precancer, Guy's, King's and St. Thomas' School of Dentistry, Denmark Hill Campus, Caldecot Road, London SE5 9RW, UK b Thames Cancer Registry, Guy's, King's and St. Thomas' Schools of Medicine and Dentistry, Capitol House, Weston Street, London SE1 3QD, UK Received 12 February 1999; accepted 19 February 1999

Abstract Studies on migrants can generate important clues on the etiology of cancer. The purpose of the present study was to determine the relationship between ethnic origin and the incidence of oral and pharyngeal cancers among residents of the Thames regions in southern England. Records from the Thames Cancer Registry during the period 1986±91 were examined and south Asians and Chinese ethnic immigrants ¯agged using their place of birth and names. Computation of relative incidence among head and neck cancers (n ˆ 7222) showed that oral cancer was signi®cantly higher among Asians (95/232=40.9%) and nasopharyngeal cancer among Chinese (45/67=67.2%). Some di€erences in the intra-oral site of cancer and ethnic origin were noted. The ethnic migrants were signi®cantly younger (Asians 51.6‹34.8 years, Chinese 47.6‹14.8 years) compared to the rest of the population (64.8‹15.6 years) at the time of cancer diagnosis ( p ˆ 0:0) but no signi®cant di€erences were found for the stage of presentation. The mean survival period for a cancer of the head and neck was 2.2 years and signi®cant di€erences in cumulative rates of survival were noted among the three groups studied ( p ˆ 0:003). A strong correlation was noted between the incidence of oral cancer and local authorities with a high percentage of Asian residents. The south Asian and Chinese ethnic minorities constitute important high risk groups for oral and nasopharyngeal cancer, for whom targeted prevention is indicated. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Cancer incidence; Immigrants; Asian; Chinese; Head and neck; Oral

1. Introduction There are very large di€erences in the relative frequency of site-speci®c cancers by geographic regions of the world [1]. The relative importance of the di€erent cancers, therefore, are somewhat di€erent in world regions. For example, 75% of oropharyngeal cancers are estimated to occur in developing countries and the largest contribution in the world total are from Southern Asia [2]. On the other hand, cancers at this site are relatively uncommon in the UK, amounting to less than 2% of all new registrations each year [3]. The enormous geographic variations referred to above pose important questions about the patterns of

* Corresponding author. Tel.: +44-171-346-3608; fax: +44-171346-3624. E-mail address: [email protected] (K.A.A.S. Warnakulasuriya)

disease in migrant populations. For example, to what extent are early environmental exposures and cultural practices important in the etiology of these cancers? In spite of this, there has been a surprising lack of attention to analytical methods for migrant data [4]. Among Indians living in the Malay peninsula, for example, the overall incidence of oral cancer is considerably higher than for Malays or Chinese [5]. Similar trends were noted among Indian migrants to Natal in South Africa [6]. Studies on cancer incidence among migrant Asian populations settled in Western Europe are meagre. Moreover, the few studies that have been conducted have looked at a range of neoplasms and, therefore, the numbers of oral and pharyngeal cancers examined in each study are small. Marmot et al. [7] reported on 15 oral cancer deaths in England and Wales between 1970 and 1972 for male Indian ethnic migrants and concluded that the proportional mortality ratio in this ethnic group was higher than expected. Donaldson and

1368-8375/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S1368-8375(99)00019-6

472

K.A.A.S. Warnakulasuriya et al. / Oral Oncology 35 (1999) 471±475

Clayton [8] reported on a small cohort of oral cancers incident during 1976±82 in Asian named individuals in Leicestershire. A further study examined 21 deaths from malignancies a€ecting the buccal cavity and pharynx in migrants from the Indian subcontinents in England and Wales during 1975±77 [9]. The largest study so far on Indian born ethnic migrants to the UK [10] examined only 30 oral cancers between 1979 and 1985 compared to 4564 natives from England and Wales who died of oral cancer during the same period. They con®rmed that the odds ratio (OR) for deaths among Indian ethnic migrants was signi®cantly higher. The present study examines the incidence of, and outcome from, common head and neck cancers in Asians and Chinese living in the Thames regions and registered during 1986±91 with particular reference to oral, oropharyngeal and nasopharyngeal cancer. 2. Population and methods The Thames Cancer Registry (TCR) is the largest population-based registry in Western Europe and covers 25% of the population of England and Wales, amounting to about 13.8 million residents [11]. Registerable cases are collected from a variety of sources. Within the Thames Health Regions, peripatetic registration ocers trained and employed by the TCR visit designated hospitals and other healthcare facilities in their catchment areas. Individual hospital medical information systems are used to capture cases and to record a minimum set of data on each case. Death certi®cates are used for veri®cation and auditing, and for updating cancer registrations when incident entries are missing. Country of birth is recorded when known but duration of residence is not known. Data on ethnicity were not collected until 1993. The incident cases normally resident in Thames regions, diagnosed between 1986 and 1991 and noti®ed to the Registry data base were tabulated by age, sex and cancer site (ICD 10). The latter was bridge-coded using the data from the original ICD 9 entries. This covered lip, all sites of mouth, pharynx, nasal cavities as well as larynx and bronchus. The stage of cancer (TCR 1±4) was based on a modi®ed TNM system adopted by the Registry since 1965 which successfully stages about 80% of all solid cancers. TCR stage is based on whether the disease is: (1) local (stage 1); (2) has extension beyond the organ of origin (stage 2); (3) has regional lymph node enlargement (stage 3); or (4) has distant metastasis (stage 4) [11]. When known to the Registry, date of death was used to assess percentage survival with or without disease. Asians for this study were taken as natives from India, Pakistan, Bangladesh, Nepal, and Sri Lanka. Records on place of birth were incomplete for some

incident cases. Three Asian researchers together manually coded Asian-sounding names (using surnames and place of birth supplemented by ®rst names and maiden names where appropriate) of all cancers ¯agged by these sites for the period of this study. Similarly, Chinesesounding names [12] to include Chinese, natives from Hong Kong and Vietnamese were also separately coded. The study also looked at any correlation between the incidence of oral cancers [average age standardised rates (world) ASR (W) for the period 1986±91] and the percentage of Asian residents in District Health Authorities (DHAs) in north and south Thames areas (according to the 1991 census). There were some changes in the boundaries of individual DHAs during the period covered by the study but the raw data allowed reassignment. Cross-tabulations were performed using the Statistical Package for Social Sciences SPSS package. Ethnicity was grouped under Asian, Chinese and other residents of southern England. Any associations between ethnicity and the mean age of diagnosis were compared by Student's t-test. For the comparison of categorical variables between ethnic groups the w2-squared test was used. Overall survival of the three groups from the beginning of treatment was examined by the Kaplan± Meier method. 3. Results During the period 1986±91, 7521 cancers at these sites were recorded in the TCR data base. Of these, 232 were Asians (3.1%) and 67 were Chinese (0.9%). 5072 of these were in males (67.4%) and 2449 (32.6%) were in females. Mean age at diagnosis for the total sample was 64.3‹16.8 years. This was lowest among the Chinese (47.6‹14.8 years). Among the Asians the mean age was 51.6‹34.8 years. The mean ages of both ethnic migrant groups were signi®cantly di€erent ( p ˆ 0:00) from other residents (64.8‹15.6 years). 20.6% of cases were in stages 3 and 4 at the time of diagnosis but no signi®cant di€erences in the stage at presentation were found in ethnic migrants compared to other natives ( p ˆ 0:57). 2173 cancers of the mouth and oropharynx and 287 cancers of the nasopharynx were recorded in the Registry for this period. Site-speci®c distribution of cancers of lip, mouth and oropharynx, nasopharynx, nasal cavity, paranasal sinuses and larynx in the Asians, Chinese and other natives is shown in Table 1. No lip cancers were recorded either in Asians or Chinese. 95/232 Asians (40.9%) had their cancers in mouth or pharynx while 45/67 Chinese (67.2%) had nasopharyngeal cancer. A signi®cantly higher proportion of cancers occurred at these two sites among the ethnic migrant groups compared to other natives (for oral among Asian vs other 2 ˆ 13:6, p ˆ 0:00; for nasopharynx among

K.A.A.S. Warnakulasuriya et al. / Oral Oncology 35 (1999) 471±475

473

Table 1 Site distribution of head and neck neoplasms reported by ethnicity Site

Chinese

Lip Oral/oropharynx Nasopharynx Nasal and paranasal Larynx Other

0 7 45 0 5 10

All

67

Asian

Others

Total (%)

0 95 14 15 50 58

197 2071 228 524 2402 1800

197 2173 287 539 2457 1868

232

7222

7521 (100)

(2.6) (28.9) (3.8) (7.2) (32.7) (24.8)

Chinese vs other 2 ˆ 192:7; p ˆ 0:00). Compared to other residents the intra-oral sites underrepresented among Asians were the ¯oor of mouth and palate. In all other intra-oral sites the relative frequency of recorded cancers was higher among the Asians. Reported cancers in the oropharynx or in the larynx were comparatively lower in the Chinese. The plotted results showed there was a strong correlation between the incidence of oral cancer and local authorities with a high percentage of Asian residents (Fig. 1 ). Percentage cumulative survival for the three population groups known to the Registry is shown in Figs. 2±4. Mean survival for the total group was 2.2 years [95% con®dence interval (CI) 2.1±2.2]. Survival from head and neck cancer (Fig 2) was signi®cantly di€erent for ethnic migrants compared to other residents ( p ˆ 0:005). Asians and Chinese demonstrated a better outcome from oropharyngeal (Fig. 3) and nasopharyngeal cancer (Fig. 4) ( p < 0:04).

Fig. 2. Five-year survival from head and neck cancer (all ethnic groups).

Fig. 3. Five-year survival following oral cancer (all ethnic groups).

4. Discussion Research into the incidence of cancer in speci®c ethnic groups in the UK is hampered by the fact that entry of ethnic group for an incident case only became part of the contract minimum data set in 1993. Furthermore,

Fig. 1. Correlation between oral cancer incidence (1986±91) and the percentage Asian population in the Local Authorities in North and South Thames.

Fig. 4. Five-year survival following nasopharyngeal cancer by ethnic groups.

entry of ethnicity in hospital notes is still far from complete. Ethnic groups can be identi®ed by place of birth and by their names. The reliability of such coding can be questionable. However, it has been demonstrated that Asian ethnic groups can be identi®ed by their names with a high sensitivity and speci®city [13]. This method of coding for ethnicity has been used in several migrant studies on cancer and is employed here. When conducted by experienced researchers who have an awareness of Asian and Chinese names, and when substantiated by cross-checking against place of birth, the method can provide useful data for studies of this nature. Anglo-Indians and Indians with English names (mostly Christians) may be misclassi®ed as being of

474

K.A.A.S. Warnakulasuriya et al. / Oral Oncology 35 (1999) 471±475

English descent but this constitutes only a very small group. We have previously questioned the completeness of oral cancer registration in the UK [14] but there is no reason to believe that certi®cation practice would di€er by ethnic groups as data capture for incident cases in TCR is through hospital sources. Past studies on cancer in Indian ethnic migrants have largely been on cancer mortality rather than on its incidence. For cancers of the mouth this can be unreliable and incomplete. We have previously shown that only 47% of death certi®cates mention oral cancer in patients known to have a mouth cancer during their lifetime and many die from another cause such as bronchopneumonia. Mortality data also su€er from undernoti®cation of non-fatal cases. If the aggressiveness of speci®c cancers were to di€er among ethnic groups, this also may be re¯ected in mortality data. In this study we therefore examined incident cases to overcome these complexities. A higher risk for death for oral or pharyngeal cancer (OR 2.2 95%, CI 1.5±3.1) among Indians in England and Wales was reported by Swerdlow et al. [10]. They argued that this may be related to the habit of betel quid chewing remaining prevalent after migration. Recent studies have reported that approximately half of adult south Asians interviewed in north and east London claimed to chew betel quid and a large majority of these subjects use tobacco in preparation of their quids [15,16]. The present study also con®rms that natives from South Asia are at a signi®cant risk of developing mouth cancer compared to others in the Thames regions, providing crude evidence of underlying lifestyle risk factors. These risk habits operating in south Asian populations are a matter of concern and appropriate educational and intervention programmes need to be developed for prevention of oral cancer among these ethnic migrants [17,18]. It is of interest to note that migrants from India and Sri Lanka to New South Wales, Australia, were not found to have an excess risk of oral cancer compared to natives in Australia [19]. These authors had explained this contradiction on the basis that these Asian migrants to Australia are generally of high socioeconomic status and may have never chewed tobacco or stopped chewing following migration in contrast to Asian migrants to England and Wales. Asian immigrants to the UK, on the other hand, particularly immigrants from Pakistan and Bangladesh have come, in general, from impoverished circumstances [20]. Migration in general may result in alteration in lifestyle and acclimatisation to Western culture, habits and dietary practices. Among the older Asians this appears not be the case and perhaps this explains why oral cancer rates are not changeable on migration. Further studies on incidence among younger Asians, particularly the children of migrants, are required to examine the

e€ect of the period of residence and Western education on the incidence of cancer at these speci®c sites. If large di€erences in incidence persisted in the secondgeneration Asians this may imply that an inherited susceptibility may underlie some of the variation. The interpretation of these trends would be compounded by the fact that betel quid habits among Asian adolescents living in the UK are still prevalent; these habits appear to continue into their second generation in another country long after migration [21]. A signi®cant relationship between site of involvement and ethnic origin was also noted. Tongue and gum cancer were more predominant among Asians, while sites such as lip and ¯oor of mouth were underrepresented. This may relate to risk habits and sites of mutagenic action and di€erent ways in which tobacco is consumed by these populations. Age at detection was signi®cantly lower for both Chinese and Asians. In high risk countries such as India the median age for carcinoma of mouth is reported to be at least 5 years younger compared to global ®gures [22]. The reasons for the di€erences observed were unclear. In the present study there was no signi®cant di€erence in tumour stage and this appears to suggest that no particular delays exist among Asians and Chinese in presenting for care. However, signi®cant di€erences were noted in the rates of survival. This may suggest that these neoplasms among ethnic migrants are generally less biologically aggressive but this is open to speculation. Age and site di€erences in the Asian cohort may have in¯uenced these observed di€erences in survival. Further prospective studies are needed to con®rm the variations reported and to determine their causes. Acknowledgements We thank Drs S. Osman and K. Shetty for assistance with coding the data set.

References [1] Parkin DM, Pisani P, Ferlay J. Estimates of the worldwide incidence of eighteen major cancers in 1985. International Journal of Cancer 1993;54:594±606. [2] Johnson NW. A global view of the epidemiology of oral cancer. In: Johnson NW, editor. Risk Markers for Oral Diseases. Oral Cancer. Detection of Patients and Lesions at Risk, Vol. 2. Cambridge: Cambridge University Press, 1991. p. 3±26. [3] Johnson NW, Warnakulasuriya KAAS. Epidemiology and etiology of oral cancer in the United Kingdom. Community Dental Health 1993;10(Suppl. 1):13±29. [4] Parkin DM, Khlat M. Studies of cancer in migrants: rationale and methodology. European Journal of Cancer 1996;32A:761±71. [5] Omar-Ahmed U, Ramanathan K. Oral carcinoma-racial variations in the Indians, Malays and Chinese. Medical Journal of Malaysia 1968;22:172±80.

K.A.A.S. Warnakulasuriya et al. / Oral Oncology 35 (1999) 471±475 [6] Van Wyk CW, Stander I, Padayachce A, et al. The areca nut chewing habit and oral squamous-cell carcinoma in South African Indians. South African Medical Journal 1992;22: 425±9. [7] Marmot MG, Adelstein AM, Bulusu L. Immigrant mortality in England and Wales 1970-78. Oce of Population Census and Surveys. Studies on Medical and Population Subjects No. 47. London: HMSO, 1984. [8] Donaldson LJ, Clayton DG. Occurrence of cancer in Asian and non-Asians. Journal of Epidemiology and Community Health 1983;38:203±7. [9] Balarajan R, Bulusu L, Adelstein AM, Shukla V. Patterns of mortality among migrants to England and Wales from the Indian subcontinent. British Medical Journal 1984;289:1185±7. [10] Swerdlow AJ, Marmot MG, Grulich AE, Head J. Cancer mortality in Indian and British ethnic immigrants from the Indian subcontinent to England and Wales. British Journal of Cancer 1995;72:1312±9. [11] Thames Cancer Registry. Cancer in South East England 1996. London: Thames Cancer Registry, 1997. [12] Choi BC, Hanley AJ, Holowaty EJ, Dale D. Use of surnames to identify individuals of Chinese ancestry. American Journal of Epidemiology 1993;138:723±4. [13] Nicoll A, Bassett K, Ulljaszek SJ. What's in a name. Accuracy of using surnames and forenames in ascribing Asian ethnic identity in English populations. Journal of Epidemiology and Community Health 1986;40:364±8. [14] Warnakulasuriya KAAS, Acworth P, Bell J, Johnson NW. Incompleteness of oral cancer registration in south-east England, 1971±87. British Journal of Cancer 1994;70:736±8.

475

[15] Warnakulasuriya KAAS. Ethnicity, race and oral cancer. In: Bedi R, Bahl V, Rayan RR, editors. Dentists, Patients and Ethnic Minorities. London: Faculty of General Dental Practicioners of the Royal College of Surgeons, 1996. p. 57±65. [16] Pearson N, Croucher R, Heath R. Pan-chewing habits in a sample of Bangladeshi adults in Tower Hamlets. Journal of Dental Research 1996;75,1151. [17] Warnakulasuriya KAAS, Johnson NW. Epidemiology and risk factors for oral cancer: rising trends in Europe and possible e€ects of migration. International Dental Journal 1996; 46(Suppl. 1): 245±50. [18] Bedi R. Betel-quid and tobacco chewing among United Kingdom's Bangladeshi community. British Journal of Cancer 1996; 29(Suppl.):573±7. [19] Grulich AE, McCredie M, Coates M. Cancer incidence in Asian migrants to New South Wales, Australia. British Journal of Cancer 1995;71:400±8. [20] Robinson, V. Boom and gloom: the success and failure of South Asians in Britain. In: Clarke C, Peach C, Vertovec S, editors. South Asians Overseas: Migration and Ethnicity. Cambridge: Cambridge University Press. [21] Osman S, Warnakulasuriya S, Cooper D, Gelbier S. Betel quid chewing and tobacco habits among Asian schoolchildren in Bedfordshire. Journal of Dental Research 1997; 76:1054. [22] Sankaranarayanan R. Oral cancer in India: an epidemiological and clinical review. Oral Surgery Oral Medicine Oral Pathology 1990;69:325±30.