EYE CANCER IN ADULTS IN UZBEKISTAN, 1978-1998 MOURATOVA T.* ABSTRACT vée a été remarquée dans le groupe ethnique ’’non- local’’. Les tumeurs de la paupière sont les plus fré- Incidence rate, age-adjusted and standardized inci- quentes. Après viennent les tumeurs conjonctiva- dence and mortality of malignant eye tumours (METs) les, intra-oculaires et orbitaires. L’incidence parmi in Uzbekistan in 1978-1998 and professional, eth- les femmes urbaines et rurales est supérieure à cel- nic, environmental, and geographical factors which le des hommes. Au cours de 10 ans l’incidence stan- could influence it are analyzed. Before 1978, there dardisée a augmenté 2.3 fois et la mortalité 6.0 fois. was no information about patients with METs in L’incidence la plus élevée a été remarquée dans les Uzbekistan. We gathered 4872 primary new cases, régions de Karakalpakstan et de Khorezm. of whom 79.7% (3882 patients) were treated in our department. All samples were statistically tested. KEY WORDS Age-adjusted MET incidence has a peak in the 60- 69 age subgroup. The highest incidence was in ’’non- Incidence. Localization. Malignant eye local’’ ethnic group. Eyelid tumours occurred most tumours. Mortality. Pollution. frequently, followed by conjunctival, intraocular and orbital tumours. Urban and rural females had high- MOTS-CLÉS er incidence than urban and rural males. For 10 years, standardized incidence has increased 2,3 and Incidence. Localisation.Tumeurs malignes mortality 6,0 times. The highest incidence and mor- d’oeil. Mortalité. Pollution. tality were in Karakalpakstan and Khorezm Wiloyat. In cherished memory of Ikhtibor Yuldasheva RÉSUMÉ Une analyse portant sur les taux d’incidence et de mortalité, par âge et standardisés de la tumeur ma- ligne de l’oeil (TMO) ainsi que sur les facteurs pro- fessionnels, ethniques, environnementaux et géo- graphiques pouvant exercer une influence sur cette affection a été effectuée en Ouzbékistan pour la pé- riode de 1978-1998. Avant 1978, l’on ne dispo- sait d’aucune information sur les patients atteints de TMO en Ouzbékistan. Nous avons recueilli 4872 nouveaux cas primaires, dont 79.7% (3882 pa- tients) ont été traités dans notre département. Tous les cas ont été testés statistiquement. L’incidence par âge de la TMO a son pic dans le sous-groupe de 60-69 ans. L’incidence la plus éle- zzzzzz * Department of Ophthalmology, Institute of Oncology and Radiology, Academy of Sciences, Uzbekistan. received: 25.03.04 accepted: 31.08.04 Bull. Soc. belge Ophtalmol., 294, 25-34, 2004. 25 INTRODUCTION The purpose of this paper is to analyze inci- dence rate, age-adjusted and standardized in- Epidemiological study of METs is very impor- cidence and mortality of METs in Uzbekistan tant in identifying the relationship between var- in 1978-1998 and to study professional, ethnic, ious external factors and incidence in order to environmental, and geographical factors which determine methods of prevention and effective could influence MET incidence and mortality. treatment. In doing so, the authors writing on the subject use standardized indices (10, 21), age-standardized incidence (2), mean age-ad- PATIENTS AND METHOD justed incidence (26), average annual incidence (8), average annual age-adjusted incidence (11), Before 1978, there was no information about average annual age standardized incidence (9), the number of patients with METs in Uzbeki- incidence (15), age-specific rates (22), and stan. Since 1978, the Institute of Oncology and specific incidence (5). Most authors calculate Radiology (IOR) of Uzbekistan’s Academy of incidence rates per 100,000 inhabitants but Sciences has been collecting the information some calculate incidence per 1,000,000 in- about primary new cases of all malignant tu- habitants (9, 11, 26). In addition, types and mours after histopathologic confirmation, in- localizations of studied METs differ greatly, e.g., cluding METs of adults and children. In joint “eye” (8, 10, 11), “eye, orbit, related adnexa” collaboration with the IOR’s Resource Depart- (2), “intraocular” (5, 10, 15), “intraocular, con- ment, we have gathered MET related informa- junctival, orbital” (9), “ocular malignant mel- tion from 1978 to 1998. The information in- anoma” (21), “uveal melanoma” (26), and “ma- cludes patient’s name, age, ethnic background, lignant melanomas of the choroid and ciliary address, profession, diagnosis, method of treat- body” (1). Nonetheless, the total incidence of ment, results of treatment and, where applica- intraocular METs is not more than 1.0 per ble, date and cause of death. During that pe- 100,000 inhabitants (5, 10, 15). Moravcova riod, 4872 patients have been registered - 2377 et al (15) found that incidence of intraocular males (48.8%) and 2495 females (51.2%). tumours in the Slovak Republic was 8.2 per 79.7% (3883 patients) were treated in the IOR’s 100,000 inhabitants in 1968-1985. In the Department of Ophthalmology. The rest 20.2% same period and country (5), a specific inci- (983 patients) were treated in regional cancer dence of intraocular tumours for males was 1.0 dispensaries, less than 0.1% (5 patients) in and for females 0.8 per 100,000 inhabitants, Russia (Moscow and St. Petersburg), Ukraine i.e. more than ten times less. MET incidence (Odessa) and one patient was treated in Israel. is not more than 1.0 per 100,000 inhabitants Histopathologic confirmation has been done in (5, 8, 10) in 1966-1986. Incidence of malig- the IOR’s Department of Histopathology or in nant melanoma of the choroid and the ciliary regional cancer dispensaries. body is also less than 1.0 (0.72 per 100,000 For statistical purposes, the patients were di- inhabitants) (1). Concerning ethnic groups of vided into 15-19, 20-24, 25-29, 30-34, 35- patients with METs, most authors argue that 39, 40-44, 45-49, 50-54, 55-59, 60-64, 65- basal cell carcinoma is the most common ma- 69, 70-74, 75-79,80-84, 85-89, and 90 and lignant eyelid tumour in whites (4), that eye more age subgroups. We have calculated inci- cancer occurs more frequently in whites than dence rate, age-adjusted and standardized in- in blacks (8), that primary eyelid tumours of cidence. Incidence rate is defined as ’’a num- any type are rare in blacks (13), that age-ad- ber of new cancers of a specific site/type oc- justed rates were higher for whites than for curring in a specified population during a year, blacks and declined from 1973-1977 to 1983- usually expressed as a number of cancers per 1987 in whites (22), that most cases (97.8%) 100,000 population’’ (18). Age-adjusted inci- of uveal melanomas occurred in the white pop- dence is ’’a weighted average of the age-spe- ulation (26), that the incidence rate was ap- cific rates, where the weights are the propor- proximately fivefold higher among white males tion of persons in the corresponding age groups (27), and that in white men the incidence rate of a standard population’’ (18). The standard- increased by 295% within 27 years (35). ized incidence is calculated first by estimating 26 the age-specific rates and then applying these Climatic and geographic characteristics of MET rates to the reference population, which is the incidence have been studied by plotting inci- standard world population (34). To calculate in- dence and climatic zone data presented in sin- cidence rate, age-adjusted and standardized in- gle scale on Uzbekistan’s maps. Climatic maps cidence, similar number of population divided and maps of water and soil pollution were ob- by the same age subgroups was used. Inci- tained from the Central Asian Regional Re- dence in different ethnic groups was compared search Hydrometeorological Institute. Air pol- to the number of the same ethnic groups in lution data were provided by the State Com- Uzbekistan. Population data were obtained from mittee of Nature Protection of Uzbekistan. Com- Uzbekistan’s Statistics State Committee. Con- puter application for statistical data manipula- fidence intervals were calculated using the log tion was developed by the Central Asian Re- rank test to determine differences among the gional Computing Centre and the Institute of groups of interest. All rates were statistically Nuclear Physics. tested and considered statistically significant if intersecting the 95% confidence interval (CI RESULTS 95%, P < 0,05). To analyze MET incidence, the International Classification IX, 1984, was In the subgroups between 15-19 and 30-34 used: eyelid skin (172.1, 173.1), conjunctiva years (Figure 1), the age-adjusted incidence did (190.3), retina (190.5), uveal tract (190.6), not change significantly but beginning from the orbit except for bones (190.1), and orbital bones 35-39 to 60-64 age subgroups the age-adjust- (170.0). Other malignant tumours of eyelid as ed incidence increased and then declined in well as malignant tumours of lacrimal gland both sexes in the 65-69 to 90 and more age and malignant Meibomian tumours were in- subgroups. cluded in the sections 190.1 and 173.1 re- Of all 125 ethnic groups in Uzbekistan in 1978- spectively. For statistical purposes, all METs 1987, the highest MET incidence rate was in were divided into four subgroups: eyelid, con- Jews, Germans and Russians, followed by Ukrai- junctiva, intraocular and orbit. For mortality nians, Tartars, Armenians, Turkmens, Azeris, rates, only those cases were considered where Koreans, Uzbeks, Karakalpaks, Kazakhs, Tajiks, death has been caused by a MET. and Kyrgyz. If dividing patients into ’’local’’ Fig 1. Sex and age-adjusted MET incidence per 100,000 in 1978-1998. 27 Fig 2. Percentage of METs by patients’ professional activities. Fig 3. Percentage of malignant eye tumours by localization (Uzbeks, Karakalpaks, Kazakhs, Tajiks, Kyrgyz, tinued to be 1,8 times than the one in ’’local’’ and Turkmens) and ’’non-local’’ (Russians, Ukrai- subgroup.