876
anterior infarction may have been related to the much more extensive Q wave change detected in placebo-group patients with anterior infarction than in those with inferior infarction. However, a 12-lead system only was used, and there may have been some variability in the positioning of the electrodes making these measurements less valid than those of 35-lead systems used in another study, which did show a reduction of infarct size with GL enzyme.6 Although in our study the reduction in mortality did not reach statistical significance, the results seem sufficiently encouraging, when taken in association with the other two studies reported today,6,7 to justify a larger scale trial of this apparently safe drug to determine whether mortality is reduced.
to
We thank the medical, nursing, technical, and secretarial staff of the care unit and academic cardiology unit for their collaboration in the study; Prof. D. Newell and Dr Judy Simpson of the Department of Medical Statistics, University of Newcastle upon Tyne, for their advice; and Biorex Laboratories for supplies ofGL enzyme and placebo.
coronary
Correspondence should be addressed to D. G. J. REFERENCES 1. Hillis LD, Maroko PR, Braunwald E, Fishbein MC. Influence of the time interval between coronary artery occlusion and the administration of hyaluronidase on myocardial salvage. Circulation 1976; 54 (supplement II): 161. 2. Maclean D, Maroko PR, Fishbein MC, Carpenter CB, Braunwald E. Reduction of infarct size up to 21 days after coronary occlusion in the rat. Circulation 1976; 54
(supplement II): 161. RA, Braunwald E, Maroko PR. Long-term preservation of ischemic myocardium in the dog by hyaluronidase. Circulation 1978; 58: 220-26. 4. Maroko PR, Hillis LD, Muller JE, et al. Favorable effects of hyaluronidase on electrocardiographic evidence of necrosis in patients with acute myocardial infarction. N Engl J Med 1977; 296: 898-903. 5. Kind LS, Roffler S. Allergic reactions to hyaluronidase. Proc Soc Exp Biol Med 1961; 3 Kloner
106: 548-50. S, Robinson PS, Coltart DJ, Webb-Peploe MM, Croft DN. Effect of early administration of a highly purified hyaluronidase preparation (GL enzyme) on myocardial infarct size. Lancet 1982; i: 867-71. 7. Flint EJ, De Giovanni J, Cadigan PJ, Lamb P, Pentecost BL. GL enzyme (a high purified hyaluronidase preparation) in the management of myocardial infarction. Lancet 1982; ii: 871-74.
epithelium and finally to cancer. Thermal injury resulting from the drinking of very hot beverages, physical injury caused by ingesting very coarse food, and deficiences of riboflavine, vitamin A, and zinc are proposed as risk factors for these precursor lesions. Introduction
endoscopic survey in a high-risk population for oesophageal cancer in north-east Iran’ showed an association between a high prevalence of asymptomatic chronic oesophagitis and a high incidence of oesophageal cancer. To assess the precancerous nature of chronic oesophagitis, it was considered of crucial importance that its prevalence be verified in another high-risk population as well as in a lowrisk group, and that subjects with such lesions be followed up. These studies have been conducted in China, where the incidence of oesophageal cancer in some areas, notably Linxian, Henan Province, is as high as or even higher than AN
that in north-east Iran.2 Why the incidence of oesophageal cancer in Iran and China is so high is still unknown. Some of the recognised risk factors for this cancer, such as alcohol and tobacco, play only a very small role.2,3 In Iran, suspicion falls on specific nutritional deficiences which are prevalent, such as those of riboflavine and vitamins A and C, and on opium tar. 3-5 ’In China, the same nutritional deficiencies have been suspected, and so have nitrosamines formed in mouldy foodstuffs.2 The survey was conducted in Linxian in April, 1980, and selected patients were re-examined in May, 1981.
6. Saltissi
PRECURSOR LESIONS OF OESOPHAGEAL CANCER IN HIGH-RISK POPULATIONS IN IRAN AND CHINA
N. MUÑOZ A. GRASSI SHEN QIONG
M. CRESPI WANG GUO QING LI ZHANG CAI
International Agency for Research on Cancer, Lyon, France; Regina Elena Institute, Rome, Italy; Cancer Institute of Beijing Beijing; Henan Tumour Research Institute, Henan Province; and Linxian County Hospital, Henan Province, People’s Republic of China
To determine the prevalence of precursor lesions of oesophageal cancer in Linxian, People’s Republic of China, which has an extremely high incidence of oesophageal cancer, 527 subjects were interviewed by a questionnaire about exposure to suspected risk factors. They also underwent oesophagoscopy, cytology, and biopsy. 84% had a chronic oesophagitis which had similar endoscopic and histological characteristics to that reported in Iran. The oesophagitis was accompanied by atrophy of the epithelium in 10% of these subjects and by dysplasia in 8%. Re-examination of 20 selected subjects 1 year later showed progression of the oesophagitis to atrophy and dysplasia in 4 individuals and to cancer in 4 others. The surveys in Iran and China suggest that the natural history of oesophageal cancer starts with an oesophagitis which, in a few individuals, progresses to atrophy and dysplasia of the
Summary
Subjects and Methods selected for examination in the Chen Guam of Linxian. 299 of these had been examined in 1975 by cytological screening by the Chinese balloon technique and 201 had been diagnosed as having cytological evidence of dysplasia. The remaining 326 subjects were selected from members of the same commune who had not been screened in 1975; 299 were matched by sex and age with those who had been screened, and the others were additional. 98 of those selected refused endoscopic examination, leaving 527 subjects made up as follows: 169 subjects with cytological diagnosis of dysplasia in 1975,81 with a normal cytology in 1975, and 277 not screened in 1975. The examination of each 625
subjects
were
commune
subject included: (1) A questionnaire to obtain demographic data and information on personal habits (smoking, drinking, dietary habits before 1950 and in 1979), family history, and personal medical history, focused on symptoms of upper gastrointestinal disease. (2) A physical examination, including evaluation of general health and signs of specific vitamin deficiencies. Blood was obtained from a sub-sample of 110 individuals for measurement of riboflavine, &bgr;-carotene, retinol, and zinc levels, and hair was collected from 254 subjects for zinc determinations. (3) Endoscopic examination, with a fibreoptic forwardviewing oesophagogastroscope, and guided biopsies. The same protocol and classification criteria used in the Iran surveyl were adopted, but in addition cytological specimens were taken with a small brush through the biopsy channel of the endoscope from anywhere in the middle and lower thirds of the oesophagus or from macroscopic lesions whenever present. Additional biopsy specimens were taken for evaluation of cell kinetics in 200 individuals from representative lesions and surrounding mucosa.
The biopsy specimens were processed as previously described! and the slides were interpreted according to the same criteria
877 described in the previous report, without knowledge of clinical data. Biopsy specimens for cell kinetics were immediately incubated with tritiated thymidine (1-5 pCi/ml) in basal medium for 1-3 h and fixed in methanol. 20 subjects were re-examined in May, 1980 survey, to
procedure used in the oesophageal lesions.
1981, according to the same assess
the evolution of the
Results The results reported here relate to the 527 subjects who examined endoscopically. 61% of them had a close relative with oesophageal cancer; 83% reported that they did not drink alcohol, and those who did drink did so only on special occasions; and 55% were non-smokers, and most of those who smoked consumed less than twenty cigarettes per day. Of the 105 subjects in whom serum vitamin levels were measured, 96% had abnormally low values of riboflavine, 43% of retinol, 25% of zinc, and 5% of &bgr;-carotene.
were
The
general physical
condition
was
deemed excellent in
2%, good in 94%, poor in 1%, and for 3% of the subjects it recorded. Oral leukoplakia was noted in 20% and angular stomatitis in 6%. The dental condition was considered excellent in 18%, good in 54%, and poor in 26%, and in 2% it was not recorded. No cases of tylosis were observed. 24% of subjects reported intermittent dysphagia with solids; only 4 individuals reported continuous dysphagia with both solids and liquids, and 42% complained of regurgitation, mainly acid. 42% reported a long history of intermittent epigastric pain and 4% retrosternal pain. 6 subjects (1%) reported having night blindness. Melaena was reported by 2% and loss of weight by 5%; 33% were was not
symptom-free. Endoscopic Findings The findings (table I) are very similar to those previously reported in Iran. The inflammation affected the middle and lower third of the oesophagus in 85% of the subjects. As in Iran, involvement of the lower third was not accompanied by disease in the precardial region or signs of gastric reflux. In 34% of the cases the oesophagitis was accompanied by diffuse or scattered whitish patches. Oesophagitis was present in 77% of the 320 subjects with a family history of oesophageal cancer and in 81 -6% of the 207 without. 82% of the 263 subjects who reported dysphagia or regurgitation had varying degrees of oesophagitis, compared with 79% of the 264 who did
not
report such symptoms. In
11 % the Z line (oesophageal gastric
Oesophageal varices, present in 7 -2% ofthe men and 9 -4% of the women, appeared as slightly raised bluish spots 2-5 mm in, diameter. The corresponding figures for Iran were 13’2% and 18’ 4%. 81 % of the varices were in the middle and lower thirds. As in Iran, their occurrence bore no relation to the degree of oesophagitis. The prevalence of incompetent cardia was 9 -2% in men and 9 -8% in women, compared with 7’3% and 8’5%, respectively, in Iran; that of hiatus hernia was 0 - 7% in men, compared with 0 - 9% in Iran. Oesophageal cancer was diagnosed endoscopically in 4 subjects (confirmed histologically in 3 and cytologically in 1) and suspected in 1 case (not confirmed by cytology or histology). Cancer of the cardia was diagnosed endoscopically in 5 subjects (all confirmed histologically), and stomach cancer in 4 (also confirmed histologically). Histological Findings (Table II)
whitish mucosa seen at endoscopy. Chronic oesophagitis, characterised by papillomatosis, infiltration of lymphocytes and plasma cells, and proliferation and dilatation of blood vessels of the submucosa and epithelium, was observed in 65% of the men and 63 -5% of the women. The corresponding observations in Iran were 83 - 1% and 76, 2%. In 11 -607o of the men and 9 -8% of the women (12’707o and or
8,3%, respectively, in Iran) the chronic oesophagitis
parentheses refer to
nos.
of subjects examined.
was
accompanied by atrophy of the epithelium. Dysplasia was diagnosed in 7 - 9% of the men and 8.1% of the women in Linxian, compared with 4 -7% and 2,9%, respectively, in Iran. Squamous-cell carcinoma of the oesophagus was diagnosed in 5 subjects, and adenocarcinoma ofthe cardia in 5 and ofthe stomach in 4.1 of the cancers ofthe oesophagus was. diagnosed only after histological examination, and 1 only at cytology. Cell kinetics studies Relation Between
are
in progress.
Endoscopic and Histological Findings
(Table III) There is good correlation between positive endoscopic and 1980
————————————————————
m
.
The commonest lesion was clear-cell acanthosis, present in 80’8% of the men and 72’4% of the women, compared with 66 . 20% and 64.9% respectively, in Iran. The lesion was characterised by squamous epithelium thickened by swollen clear cells which, in most instances, were periodic-acid/Schiff negative, and it was the lesion producing the whitish patches
TABLE I-ENDOSCOPIC FINDINGS IN LINXIAN,
Nos.
junction) was higher than
normal.
878 TABLE II-HISTOLOGICAL FINDINGS IN LINXIAN
Nos.
in
parentheses refer to
nos.
1980
of subjects examined. TABLE III-CORRELATION BETWEEN ENDOSCOPIC AND HISTOLOGICAL FINDINGS
histological findings, but a large number of endoscopically normal subjects have microscopic oesophagitis. There is little correlation between the results of the cytological examination in 1975 and the cytological and histological findings in this survey (table IV). All but 1 of the cancer cases occurred in those not previously screened, which may indicate that those at highest risk in 1975 were not available for the study because they had been treated for cancer, or had died. More detailed cytological evaluation is in progress.
Follow-up Study The precancerous nature of the oesophageal lesions described can only be demonstrated by a follow-up study. In May, 1981, 20 subjects from the 1980 survey, selected from among those who had a presumptive cytological diagnosis of
dysplasia or severe vitamin deficiences were re-examined: in 12 of them the lesions had not changed (6 mild oesophagitis confirmed histologically, 3 clear-cell acanthosis, and 3 oesophagitis with atrophy and mild dysplasia). In 3 subjects mild oesophagitis had progressed to oesophagitis with atrophy and dysplasia. The remaining four subjects now had cancer. In 2 of them we had diagnosed oesophagitis and dysplasia in 1980; the next year an in-situ squamous-cell carcinoma of the oesophagus had developed in one and an early adenocarcinoma of the cardia in the other. In the third subject an oesophagitis with atrophy was diagnosed in 1980 and an in-situ squamous-cell carcinoma of the oesophagus in 1981. The fourth subject had a mild oesophagitis in 1980 and an invasive squamous-cell carcinoma in 1981. The possibility of false-negative diagnoses of cancer in the first examination is unlikely but it cannot be ruled out.
TABLE IV-CORRELATION BETWEEN HISTOLOGICAL FINDINGS AND CYTOLOGICAL DIAGNOSIS IN
Numbers in italics
are
percentages.
1975
879
Discussion
similarity in the very high prevalence of chronic oesophagitis in two populations at high risk for oesophageal cancer living in widely separated geographical regions, north-east Iran and central China, is striking. These lesions are rare among patients attending gastrointestinal clinics in Europe, where oesophageal cancer is uncommon. 6,7 The high prevalence of oesophagitis, even in the younger age-groups, and the finding that oesophagitis and cancer afflict the same sections of the oesophagus, suggest that these two lesions are The
associated. The oesophagitis in these two high-risk populations differs from that reported in low-risk populations. In Iran and China the oesophagitis, which generally does not produce symptoms, is characterised by an irregular, friable mucosa with varying degrees of oedema, hyperaemia, and leukoplakia, but without ulceration, and it usually involves the middle and lower thirds of the oesophagus, leaving the precardial region free. In contrast, the oesophagitis in lowrisk populations in Europe is characterised by erosions and ulcerations, which usually involve the precardial region since the condition is in general due to reflux. Histologically, oesophagitis in Iran and China is characterised by lymphocyte and plasma-cell infiltration of the submucosa and mucosa, papillomatosis, and vascular proliferation and dilatation. The clear-cell acanthosis which very often accompanies the oesophagitis is usually the lesion corresponding to the whitish patches or whitish mucosa described endoscopically, and may be the result of thermal injury, since the habit of drinking very hot beverages is widespread in both China and Iran. In both places atrophy is the second most common epithelial lesion accompanying the oesophagitis, followed by dysplasia. Little information is available on predysplastic lesions of the human oesophagus; only dysplasia has been proposed as a possible precancerous lesion.8-13 Our observations, however, suggest that the natural history of oesophageal cancer is as follows:
injury due to very coarse food (in both Iran and China wheat contains silica fibres and millet bran contains silica plates 16 [O’Neill CH, Clarke GD, personal communication]); and deficiences of vitamins B2 and A. There is some experimental evidence in baboons that atrophic lesions in the oesophagus can be induced by riboflavine deficiency, 17 and riboflavine deficiency is widespread in both the Iranian and Chinese
high-risk populations.8 An understanding of the mechanisms of carcinogenesis will determine whether prevention should be aimed primarily at the suspect carcinogen, or at modification of the co-factors. There would therefore be great interest in an intervention study to clarify the role of selected vitamin deficiencies in the development of precancerous and cancerous lesions. This should be a randomised blind trial in which subjects with different stages of precancerous lesions are treated with either riboflavine, vitamin A, or placebo, and are assessed endoscopically, cytologically, and histologically before and after treatment. Plans for such a study in China are under way. We thank Dr Li Bing and Dr Qu Chuan Yan, Chinese Academy of Medical Sciences in Beijing, and Dr Yang Wen Xian, Henan Tumour Research Institute, for organising the survey; Dr Li Jung-Yao, Dr Liu Fu-Shen, Dr Zheng Su Fang, Dr Cai Hai-Ying, Dr Zhang Jing Sheng, and Dr Fan Ying, Cancer Institute of Beijing, Dr Yang Guan Rui, Dr Huang-He, Dr Qiu Zong Liang, Dr Qiao Si Jie, Dr Lu Jian Bang, and Dr Wu Qing Cheng, Henan Medical College, and Dr Zhao Heng Zhong, Linxian County Hospital, for their assistance during the field studies ; Dr David Thurnham, London School of Hygiene, for vitamin determinations; Dr M Hambidge, University of Colorado, for zinc analysis; Dr Allen Linsell, I.A.R.C., for his helpful comments on reviewing the manuscript; Annie Arslan and Brigitte Kajo, I.A.R.C., for their technical assistance; and ACM (Germany) for the loan of an
endoscope. Correspondence should be addressed to N. M., Analytical Epidemiology Programme, Division of Epidemiology & Biostatistics, International Agency for Research on Cancer, 150 Cours Albert-Thomas 69372, Lyon Cedex 2, France. REFERENCES 1. 2.
Crespi M, Muñoz N, Grassi A, et al. Oesophageal lesions in northern Iran; a premalignant condition? Lancet 1979; ii: 217-21. Yang CS. Research on esophageal cancer in China: a review. Cancer Res 1980; 40: 2633-44.
oesophagitis-atrophy- dysplasia- cancer suggestion is supported by laboratory observations
chronic
This both in rats treated with N-methyl-N-nitrosoaniline14 and in non-human with treated primates 1-methyl-lnitrosourea15-oesophageal abnormalities similar to those seen in Iran and China preceded the development of oesophageal tumours. Further insight into the natural history of the human disease could come from similar studies in lowrisk populations and more extensive follow-up studies of subjects with different stages of precursor lesions.
The following are two possible theories about the aetiology of the precursor lesions and of the cancer of the oesophagus in Iran and China.
(a) As in the experimental models, both the precursor lesions and the cancer may be due to a carcinogen(s). In Iran, the carcinogen may be opium tar, but in China the carcinogen remains to be identified. The effect of the carcinogen could be facilitated or potentiated by co-factors (see below). (b) The precursor lesions could be caused by co-factors, and cancer may result when a carcinogen acts on the precursor lesions. Co-factors that might be considered are: thermal injury caused by very hot beverages, although it has not been established that lesions can be induced in this way; physical
3.
Joint Iran/IARC Study Group. Oesophageal cancer studies in the Caspian littoral of Iran: results of population studies—a prodrome J Nat Cancer Inst 1977; 59:
1127-38. 4. Cook-Mozaffari
PJ, Azordegan F, Day NE, et al. Oesophageal cancer studies in the littoral of Iran: results of a case-control study Br J Cancer 1979; 39: 293-309. 5. Kmet J, McLaren DS, Siassi F. Epidemiology of esophageal cancer with special reference to nutritional studies among the Turkoman of Iran. Adv Modern Hum Nutr 1980; 343-65. 6. Brunnen PL, Karmody A, Needham CD. Severe peptic oesophagitis. Gut 1969; 10: 831-37. 7. Skinner DB. Symptomatic esophageal reflux. Am JDig Dis 1966; 11: 771-79. 8. Day N, Muñoz N. Esophagus. In: Shottenfeld D, Fraumeni JF, Eds. Cancer epidemiology and prevention Philadelphia: W.B. Saunders, 1982: 596-623. 9. Kolicheva VJ. Data on the epidemiology and morphology of precancerous changes and of cancer of the oesophagus in Kazakhstan, USSR, Doctoral Thesis, Alma Ata, 1974. 10. Ushigome S, Spjut HJ, Noon GP. Extensive dysplasia and carcinoma in situ of esophageal epithelium. Cancer 1967; 20: 1023-34. 11. Postlethwait RW, Wendell Musser A. Changes in the esophagus in 1000 autopsy specimens. J Thorac Cardiovasc Surg 1974; 68: 953-56. 12. Mukada T, Sato E, Sasano N. Comparative studies on dysplasia of esophageal epithelium in four prefectures in Japan (Miyagi, Nara, Wakayama and Aomori) with reference to risk of carcinoma. Tohoku Exp Med 1976; 119: 51-63. J 13. Mandard AM, Chasle J, Marnay J. Cancer of the oesophagus and dysplasias (preliminary results). Eur J Cancer 1978. 1: 15-26. 14. Napalkov NP, Pozharisski K Morphogenesis of experimental tumors of the oesophagus J Nat Cancer Inst 1969; 42: 922-40 15. Adamson RH, Krolikowski FJ, Correa P, et al. Carcinogenicity of 1-methyl-1nitrosourea in non-human primates. J Nat Cancer Inst 1977; 59: 415-22. 16. O’Neill CH, Hodges GM, Riddle PN, et al. A fine fibrous silica contaminant of flour in the high oesophageal cancer area of north-east Iran. Int J Cancer 1980, 26: 617-28. 17 Foy H, Gillman T, Kondi A. Histological changes m the skin of baboons deprived of riboflavin. Medical primatology 1972. Proceedings of the Conference on Experimental Medicine and Surgery in Primates, Lyon 1972, part II. Basel: Karger, 1972: 159-68.
Caspian