244
BRITISH MEDICAL JOURNAL
colon.30 Whether the lactase inhibition found in hamsters is due to a direct effect of the antibiotic on the mucosa or is secondary to bacterial overgrowth in the small bowel remains conjectural. No histological changes were found in the mucosa on light or electronmicroscopy. Similar reversible disaccharidase intolerance arises after acute infective gastroenteritis, which has been shown to cause mucosal damage. Possibly, therefore, the effects of the antibiotics are not direct but result from superficial bacterial damage to the mucosa. Alternatively, bacteria may themselves depress the small bowel mucosal function in some way, possibly by their influence on bile salts,31 or by affecting micelle formation, which precedes fat absorption.32 Together with an unusually high concentration of free bile salts in the gut lumen these effects may lead to steatorrhoea and diarrhoea.33 It follows that an unaltered gut flora contributes to normal intestinal physiology and function, provided the mucosa can resist the continual challenge of the bacterial load in the lumen. Some protection comes from the immunoglobulins secreted by the mucosa, a local immune response that is largely independent of circulating antibody. By far the most important is the IgA synthesised by plasma cells in the lamina propria34 and concentrated in and between epithelial cells in the lower two-thirds of the crypts. The stimulus to the production of secretory IgA lies in bacterial antigens to which the mucosa has previously been exposed. Some IgM is also present, but nevertheless the degree of mucosal protection to challenge from unusual or virulent micro-organisms, especially when present in increased numbers, would almost certainly be small. So possibly those patients who are particularly susceptible to developing antibiotic diarrhoea are those with low levels of secretory IgA and IgM. In these patients an increased or altered bacterial load present in the small bowel and possibly extending far more proximally than usual might lead to alterations in bile salt metabolism and induce lactose intolerance as a result of intralumenal or mucosal toxic effects; and superimposed upon these might be a direct effect of the antibiotic on the mucosa itself. The end result of this combination would be looseness of the stools or frank diarrhoea. 1 Fekety, F R, of the Amnerican Medical Association, 1968, 203, 216. Stanley, R J, Melson, G L, and Tedesco, F J, Radiology, 1974, 111, 519. 3 Kay, A W, Richards, R L, and Watson, A J, British Journal of Surgery, 1958, 46, 45. 4 Goulston, S J M, and McGovern, V J, Guit, 1965, 6, 207. 5 Klotz, A P, Palmer, W L, and Kirsner, J B, Gastroenterology, 1953, 25, 54. 6 Reiner, L, Schlesinger, M J, and Miller, G M, Archives of Pathology, 1952, 54, 39. 7 Shapiro, R L, and Newman, A, Radiology, 1973, 108, 263. 8 Keating, J P, et al, Amlerican Journal of Diseases of Childretn, 1974, 128, 369. 9 Garrod, L P, Lambert, H P, and O'Grady, F, Antibiotic and Chetotherapy, 4th edn. Edinburgh, Churchill Livingstone, 1973. 10 Geddes, A M, and Williams, J D, eds Cuirrent Antibiotic Therapy. Edinburgh, Churchill Livingstone, 1973. 11 British Medical Journal, 1968, 4, 402. 12 Percy-Robb, I W, and Collee, J G, British Medical Journal, 1972, 3,813. 13 British Tuberculosis Association, British Medical Jm?urnal, 1968, 4, 511. 14 Tedesco, F J, American Journal of Digestive Diseases, 1975, 20, 295. 1 Bass, J W, et al,J_7ournal of Pediatrics, 1973, 83, 106. 16 Hinton, N A, Current Therapetutic Research, 1970, 12, 341. 17 Bass, J W, et al, Clinical Paediatrics, 1974, 13, 273. 18 Fitzpatrick, J J, and Topley, H E, Amiterican Journal of the Medical Sciences, 1966, 252, 310. 19 Benner, E J, and Tellman, W H, American Journal of Gastroenterology, 1970, 54, 55. 20 Viteri, A L, Howard, P H, and Dyck, W P, Gastroenterology, 1974, 66, 1137. 21 British Medical3Journal, 1974, 4, 65 22 Hubbard, W N Jr, Lancet, 1974, 1, 172. 23 Tedesco, F J, Barton, R W, and Alpers, D H, Annals of Internal Medicine, 1974, 81, 429. 24 Scott, A J, Nicholson, G I, and Kerr, A R, Lancet, 1973, 2, 1232. 25 Dyck, W P, Viteri, A L, and Howard, P H, Larncet, 1974, 1, 273. 26 Dobbins, W O, Herrero, B A, and Mansbach, C M, American Journal of the Medical Sciences, 1968, 255, 63. 2
J7ournal
1 NOVEMBER 1975
27 28
Paes, I C, et al, Gastroenterology, 1967, 53, 49. Faloon, W W, et al, Annals of the Nez York Academny of Sciences, 1966, 132, 879. 29 Bywater, R J, and Smith, C C, unpublished work in preparation. 30 Neale, G,Jrournal of Clinical Pathology, 1971, 24, suppl no 5, p 22. 31 Garbutt, J T, et al, Gastroenterology, 1970, 59, 553. 32 Dawson, A M,J7ozurnal of Clinical Pathology, 1971, 24, suppl no 5, p 77. 33 O'Grady, F, and Vince, A, J7ournal of Clinical Pathology, 1971, 24, suppl no 5, p 130. 34 Crabb6, P A, Carbonara, A 0, and Heremaus, J F, Laboratory Investigation, 1965, 14, 235.
Abortion again Last week Mrs Castle announced (p 293) that the Government had accepted the recommendations made by the House of Commons select committee which had been examining the Abortion (Amendment) Bill. She added that Parliament would be given a chance to re-establish the select committee in the coming session-opening up a prospect of a repetition of all the familiar arguments on the topic. The real priorities have been plain ever since the Lane Committee reported.1 Wider availability of reliable contraceptive advice and facilities could help to reduce the numbers of unwanted pregnancies (currently estimated at 200 000 a year in Britain). There is still inadequate provision for early termination of pregnancy within the NHS in many parts of the country. More use of outpatient procedures, supported by sympathetic and effective counselling services, could eliminate the circumstances in which women have to shop around for weeks in order to arrange a termination, so introducing unnecessary medical hazards and psychological stresses. Almost all the objectionable practices occur in the private sector, which exists mainly because of deficiencies in the NHS. Very few women pay for their abortions as a matter of choice. Changes in the law rarely correct abuses, and they would not be needed if all parts of the country had the same facilities as those freely available in units which have given the problem its proper priority. Report of the Commnittee on the Working of the Abortion Act (Chairman, Mrs Justice Lane), Volumes I, II, and III, Cmnd 5538. London, HMSO, 1974.
Blood transfusion and iron overload In children with chronic aplastic anaemia or homozygous thalassaemia the maintenance of haemoglobin levels by regular blood transfusion is the only effective way to maintain life. A regimen which maintains the haemoglobin concentration above 9 g 'dl results in a normal growth pattern and a quality of life that goes a long way towards justifying permanent dependence on blood transfusion. Its short-term benefits are beyond dispute, for fully treated children with thalassaemia are indistinguishable from normal up to about the age of puberty. After this time the clinical effects of iron deposition in the tissues appear and toxicity is usually progressive, causing death during adolescence or early adult life. Buja and Roberts' have described 19 patients with cardiac
BRITISH MEDICAL JOURNAL
1 NOVEMBER 1975
iron deposits and concluded that extensive siderosis occurred in those patients who received more than 100 units of blood. The deposits were always more extensive in the ventricular myocardium than in the atria and were noted in the contractile cells rather than the conducting system. Five of the patients had chronic congestive cardiac failure, and atrial iron deposits were associated with supraventricular arrhythmias. Sanyal et a12 have recently reported the ultrastructural findings in the heart of a young girl with congenital aplastic anaemia who had received 14 units of blood up to the time of her terminal illness at the age of 14 years. Regardless of the pattern of deposition within myocardial cells some iron was consistently present within the nuclei and mitochondria. The terminal myocardial insufficiency was associated with widespread swelling and disruption of mitochondria together with an apparent reduction of myofibrils. Iron overload also causes toxic lesions in other organs, particularly the liver, and, though mitochondrial or lysosomal dysfunction, denaturation of cellular proteins, and membrane damage have all been suggested to account for the pathological changes, the mechanism is unknown. In recent years an attempt has been made to reduce these unwanted side effects of hypertransfusion by continuous chelation therapy. Barry et a13 have studied children with thalassaemia major and shown that daily injections of desferrioxamine maintained over a period of years can reduce the amount of iron accumulating in the liver. An encouraging feature of this study was the apparently static nature of the hepatic fibrosis in the chelator-treated group compared with progressive liver damage in the untreated controls. However, even with chelation therapy the liver iron concentration remained greatly increased, and serum ferritin concentrations were 50-100 times higher than normal.4 Desferrioxamine has to be administered by daily intramuscular injection, and the occurrence of occasional adverse effects is a further disadvantage. These factors and the probability that the drug cannot entirely prevent iron accumulating to toxic levels have stimulated a new search for other iron-chelating agents. Graziano et a15 have used the hypertransfused rat as an experimental model and shown that 2,3 dihydroxybenzoic acid (2,3-DHB) is a potentially useful drug for iron chelation. It has the particularly valuable property of being well absorbed after oral administration, it is specific for iron, and appears to be non-toxic. It not only increases iron excretion in the urine and faeces but decreases iron absorption from the gut. The important question remaining is whether 2,3-DHB will remove enough iron from iron-loaded patients to have an effect on long-term toxicity. Further studies of benzoic acid derivatives, hydroxamic acids and other compounds, are in progress, and it is to be hoped that potent therapeutic agents will become available for the promotion of iron excretion within the next few years. For the presernt, treatment must be directed to preventing iron accumulation by all available means. The quantitative assessment of body iron stores has been simplified by use of the serum ferritin assay.6 An increase of stores to pathological levels is probably an indication for starting chelation therapy with desferrioxamine, though Modell and Beck7 have shown that treatment may be fully effective only when there is already a considerable degree of overload. The maintenance of a high haemoglobin concentration by transfusion reduces iron absorption from the gut. Theoretically the availability of dietary iron might be further reduced by the exclusion of meat from the diet and an increased intake of those foods which have a high content of phytate, phosphate, or tannin, all of which
245
reduce the availability of iron. The effect of splenectomy in reducing transfusion requirements has yet to be fully assessed, but there is some indication8 that it may be effective. The combination of all these may make it possible to retard iron accumulation while new potential methods of treatment are assessed. 'Buja, L M, and Roberts, W C, American Journal of Medicine, 1971, 51, 209. Sanyal, S K, et al, Pediatrics, 1975, 55, 336. 3Barry, M, et al, British Medical3Journal, 1974, 2, 16. 4 Letsky, E A, et al, Journal of Clinical Pathology, 1974, 27, 652. 5Graziano, J H, Grady, R W, and Cerami, A, Journal of Pharmacology and Experimental Therapeutics, 1974, 190, 570. 6 Jacobs, A, and Worwood, M, British3Journal of Haematology, 1975, 31, 1. 7Modell, C B, and Beck, J, Annals of the New York Academy of Sciences, 1974, 232, 201. 8 Blendis, L M, et al, British Journal of Haematology, 1974, 28, 77.
2
Statistical review of cancer in England and Wales Cancer registration is meant to produce accurate statistics of the incidence of various types of cancer as well as certain epidemiological data; but in England and Wales it is a voluntary scheme, inaccuracy or failure to report carries no penalty, and the efficiency of the process depends on the enthusiasm and skill of those in charge and the interest it can arouse in the mandarins of the regional hospital board. The latest survey, covering 1968-70, has just been published,' and clearly reporting varies widely from region to region. If the top end of the range represents the true picture then those at the bottom end of the league should be asked why they are failing to notify so many cases of cancer. The registration scheme could be used as a basis for the epidemiological study of cancer in our very complex environment and as a yardstick for monitoring progress, but the present level of inaccuracy might mask the subtle clues that emerge from time- and space-cluster analysis and by record linkage with other data held by the Office of Population Censuses and Surveys. In fairness, it should be said that our cancer registration statistics are considerably better than those available for many parts of Europe and North America, but that is no grounds for complacency. The size of the problem may be seen from the 155 096 cases registered in 1970, a rate of 333-2 and 300 7 per 100 000 in males and females, respectively. Lung cancer heads the list in men with a rate of 99 9 per 100 000, and in women the rate for breast cancer is 72-2; these two diseases account for 29% and 24% of all cases of cancer in the two sexes, respectively. The registration of gastric cancer is falling slowly but not as fast as the decline in mortality might suggest, while the rates of other tumours of the alimentary tract are increasing (oesophagus, pancreas, and large bowel); cancers of the lip and tongue are on the wane. The rise of Hodgkin's disease and myeloma may reffect the growing interest in these conditions and the impact of various national trials. Clearly, registrations of borderline conditions such as myelofibrosis and polycythemia are gross underestimates of their frequency. A similar and more serious problem is the considerable variation in the apparent incidence from one region to another of carcinoma in situ of the cervix: Newcastle reported 15-9 per 100 000 while Liverpool said the rate was 1 1. Such divergence rules out the use of this data in
BRITISH MEDICAL JOURNAL
colon.30 Whether the lactase inhibition found in hamsters is due to a direct effect of the antibiotic on the mucosa or is secondary to bacterial overgrowth in the small bowel remains conjectural. No histological changes were found in the mucosa on light or electronmicroscopy. Similar reversible disaccharidase intolerance arises after acute infective gastroenteritis, which has been shown to cause mucosal damage. Possibly, therefore, the effects of the antibiotics are not direct but result from superficial bacterial damage to the mucosa. Alternatively, bacteria may themselves depress the small bowel mucosal function in some way, possibly by their influence on bile salts,31 or by affecting micelle formation, which precedes fat absorption.32 Together with an unusually high concentration of free bile salts in the gut lumen these effects may lead to steatorrhoea and diarrhoea.33 It follows that an unaltered gut flora contributes to normal intestinal physiology and function, provided the mucosa can resist the continual challenge of the bacterial load in the lumen. Some protection comes from the immunoglobulins secreted by the mucosa, a local immune response that is largely independent of circulating antibody. By far the most important is the IgA synthesised by plasma cells in the lamina propria34 and concentrated in and between epithelial cells in the lower two-thirds of the crypts. The stimulus to the production of secretory IgA lies in bacterial antigens to which the mucosa has previously been exposed. Some IgM is also present, but nevertheless the degree of mucosal protection to challenge from unusual or virulent micro-organisms, especially when present in increased numbers, would almost certainly be small. So possibly those patients who are particularly susceptible to developing antibiotic diarrhoea are those with low levels of secretory IgA and IgM. In these patients an increased or altered bacterial load present in the small bowel and possibly extending far more proximally than usual might lead to alterations in bile salt metabolism and induce lactose intolerance as a result of intralumenal or mucosal toxic effects; and superimposed upon these might be a direct effect of the antibiotic on the mucosa itself. The end result of this combination would be looseness of the stools or frank diarrhoea. 1 Fekety, F R, of the Amnerican Medical Association, 1968, 203, 216. Stanley, R J, Melson, G L, and Tedesco, F J, Radiology, 1974, 111, 519. 3 Kay, A W, Richards, R L, and Watson, A J, British Journal of Surgery, 1958, 46, 45. 4 Goulston, S J M, and McGovern, V J, Guit, 1965, 6, 207. 5 Klotz, A P, Palmer, W L, and Kirsner, J B, Gastroenterology, 1953, 25, 54. 6 Reiner, L, Schlesinger, M J, and Miller, G M, Archives of Pathology, 1952, 54, 39. 7 Shapiro, R L, and Newman, A, Radiology, 1973, 108, 263. 8 Keating, J P, et al, Amlerican Journal of Diseases of Childretn, 1974, 128, 369. 9 Garrod, L P, Lambert, H P, and O'Grady, F, Antibiotic and Chetotherapy, 4th edn. Edinburgh, Churchill Livingstone, 1973. 10 Geddes, A M, and Williams, J D, eds Cuirrent Antibiotic Therapy. Edinburgh, Churchill Livingstone, 1973. 11 British Medical Journal, 1968, 4, 402. 12 Percy-Robb, I W, and Collee, J G, British Medical Journal, 1972, 3,813. 13 British Tuberculosis Association, British Medical Jm?urnal, 1968, 4, 511. 14 Tedesco, F J, American Journal of Digestive Diseases, 1975, 20, 295. 1 Bass, J W, et al,J_7ournal of Pediatrics, 1973, 83, 106. 16 Hinton, N A, Current Therapetutic Research, 1970, 12, 341. 17 Bass, J W, et al, Clinical Paediatrics, 1974, 13, 273. 18 Fitzpatrick, J J, and Topley, H E, Amiterican Journal of the Medical Sciences, 1966, 252, 310. 19 Benner, E J, and Tellman, W H, American Journal of Gastroenterology, 1970, 54, 55. 20 Viteri, A L, Howard, P H, and Dyck, W P, Gastroenterology, 1974, 66, 1137. 21 British Medical3Journal, 1974, 4, 65 22 Hubbard, W N Jr, Lancet, 1974, 1, 172. 23 Tedesco, F J, Barton, R W, and Alpers, D H, Annals of Internal Medicine, 1974, 81, 429. 24 Scott, A J, Nicholson, G I, and Kerr, A R, Lancet, 1973, 2, 1232. 25 Dyck, W P, Viteri, A L, and Howard, P H, Larncet, 1974, 1, 273. 26 Dobbins, W O, Herrero, B A, and Mansbach, C M, American Journal of the Medical Sciences, 1968, 255, 63. 2
J7ournal
1 NOVEMBER 1975
27 28
Paes, I C, et al, Gastroenterology, 1967, 53, 49. Faloon, W W, et al, Annals of the Nez York Academny of Sciences, 1966, 132, 879. 29 Bywater, R J, and Smith, C C, unpublished work in preparation. 30 Neale, G,Jrournal of Clinical Pathology, 1971, 24, suppl no 5, p 22. 31 Garbutt, J T, et al, Gastroenterology, 1970, 59, 553. 32 Dawson, A M,J7ozurnal of Clinical Pathology, 1971, 24, suppl no 5, p 77. 33 O'Grady, F, and Vince, A, J7ournal of Clinical Pathology, 1971, 24, suppl no 5, p 130. 34 Crabb6, P A, Carbonara, A 0, and Heremaus, J F, Laboratory Investigation, 1965, 14, 235.
Abortion again Last week Mrs Castle announced (p 293) that the Government had accepted the recommendations made by the House of Commons select committee which had been examining the Abortion (Amendment) Bill. She added that Parliament would be given a chance to re-establish the select committee in the coming session-opening up a prospect of a repetition of all the familiar arguments on the topic. The real priorities have been plain ever since the Lane Committee reported.1 Wider availability of reliable contraceptive advice and facilities could help to reduce the numbers of unwanted pregnancies (currently estimated at 200 000 a year in Britain). There is still inadequate provision for early termination of pregnancy within the NHS in many parts of the country. More use of outpatient procedures, supported by sympathetic and effective counselling services, could eliminate the circumstances in which women have to shop around for weeks in order to arrange a termination, so introducing unnecessary medical hazards and psychological stresses. Almost all the objectionable practices occur in the private sector, which exists mainly because of deficiencies in the NHS. Very few women pay for their abortions as a matter of choice. Changes in the law rarely correct abuses, and they would not be needed if all parts of the country had the same facilities as those freely available in units which have given the problem its proper priority. Report of the Commnittee on the Working of the Abortion Act (Chairman, Mrs Justice Lane), Volumes I, II, and III, Cmnd 5538. London, HMSO, 1974.
Blood transfusion and iron overload In children with chronic aplastic anaemia or homozygous thalassaemia the maintenance of haemoglobin levels by regular blood transfusion is the only effective way to maintain life. A regimen which maintains the haemoglobin concentration above 9 g 'dl results in a normal growth pattern and a quality of life that goes a long way towards justifying permanent dependence on blood transfusion. Its short-term benefits are beyond dispute, for fully treated children with thalassaemia are indistinguishable from normal up to about the age of puberty. After this time the clinical effects of iron deposition in the tissues appear and toxicity is usually progressive, causing death during adolescence or early adult life. Buja and Roberts' have described 19 patients with cardiac
BRITISH MEDICAL JOURNAL
1 NOVEMBER 1975
iron deposits and concluded that extensive siderosis occurred in those patients who received more than 100 units of blood. The deposits were always more extensive in the ventricular myocardium than in the atria and were noted in the contractile cells rather than the conducting system. Five of the patients had chronic congestive cardiac failure, and atrial iron deposits were associated with supraventricular arrhythmias. Sanyal et a12 have recently reported the ultrastructural findings in the heart of a young girl with congenital aplastic anaemia who had received 14 units of blood up to the time of her terminal illness at the age of 14 years. Regardless of the pattern of deposition within myocardial cells some iron was consistently present within the nuclei and mitochondria. The terminal myocardial insufficiency was associated with widespread swelling and disruption of mitochondria together with an apparent reduction of myofibrils. Iron overload also causes toxic lesions in other organs, particularly the liver, and, though mitochondrial or lysosomal dysfunction, denaturation of cellular proteins, and membrane damage have all been suggested to account for the pathological changes, the mechanism is unknown. In recent years an attempt has been made to reduce these unwanted side effects of hypertransfusion by continuous chelation therapy. Barry et a13 have studied children with thalassaemia major and shown that daily injections of desferrioxamine maintained over a period of years can reduce the amount of iron accumulating in the liver. An encouraging feature of this study was the apparently static nature of the hepatic fibrosis in the chelator-treated group compared with progressive liver damage in the untreated controls. However, even with chelation therapy the liver iron concentration remained greatly increased, and serum ferritin concentrations were 50-100 times higher than normal.4 Desferrioxamine has to be administered by daily intramuscular injection, and the occurrence of occasional adverse effects is a further disadvantage. These factors and the probability that the drug cannot entirely prevent iron accumulating to toxic levels have stimulated a new search for other iron-chelating agents. Graziano et a15 have used the hypertransfused rat as an experimental model and shown that 2,3 dihydroxybenzoic acid (2,3-DHB) is a potentially useful drug for iron chelation. It has the particularly valuable property of being well absorbed after oral administration, it is specific for iron, and appears to be non-toxic. It not only increases iron excretion in the urine and faeces but decreases iron absorption from the gut. The important question remaining is whether 2,3-DHB will remove enough iron from iron-loaded patients to have an effect on long-term toxicity. Further studies of benzoic acid derivatives, hydroxamic acids and other compounds, are in progress, and it is to be hoped that potent therapeutic agents will become available for the promotion of iron excretion within the next few years. For the presernt, treatment must be directed to preventing iron accumulation by all available means. The quantitative assessment of body iron stores has been simplified by use of the serum ferritin assay.6 An increase of stores to pathological levels is probably an indication for starting chelation therapy with desferrioxamine, though Modell and Beck7 have shown that treatment may be fully effective only when there is already a considerable degree of overload. The maintenance of a high haemoglobin concentration by transfusion reduces iron absorption from the gut. Theoretically the availability of dietary iron might be further reduced by the exclusion of meat from the diet and an increased intake of those foods which have a high content of phytate, phosphate, or tannin, all of which
245
reduce the availability of iron. The effect of splenectomy in reducing transfusion requirements has yet to be fully assessed, but there is some indication8 that it may be effective. The combination of all these may make it possible to retard iron accumulation while new potential methods of treatment are assessed. 'Buja, L M, and Roberts, W C, American Journal of Medicine, 1971, 51, 209. Sanyal, S K, et al, Pediatrics, 1975, 55, 336. 3Barry, M, et al, British Medical3Journal, 1974, 2, 16. 4 Letsky, E A, et al, Journal of Clinical Pathology, 1974, 27, 652. 5Graziano, J H, Grady, R W, and Cerami, A, Journal of Pharmacology and Experimental Therapeutics, 1974, 190, 570. 6 Jacobs, A, and Worwood, M, British3Journal of Haematology, 1975, 31, 1. 7Modell, C B, and Beck, J, Annals of the New York Academy of Sciences, 1974, 232, 201. 8 Blendis, L M, et al, British Journal of Haematology, 1974, 28, 77.
2
Statistical review of cancer in England and Wales Cancer registration is meant to produce accurate statistics of the incidence of various types of cancer as well as certain epidemiological data; but in England and Wales it is a voluntary scheme, inaccuracy or failure to report carries no penalty, and the efficiency of the process depends on the enthusiasm and skill of those in charge and the interest it can arouse in the mandarins of the regional hospital board. The latest survey, covering 1968-70, has just been published,' and clearly reporting varies widely from region to region. If the top end of the range represents the true picture then those at the bottom end of the league should be asked why they are failing to notify so many cases of cancer. The registration scheme could be used as a basis for the epidemiological study of cancer in our very complex environment and as a yardstick for monitoring progress, but the present level of inaccuracy might mask the subtle clues that emerge from time- and space-cluster analysis and by record linkage with other data held by the Office of Population Censuses and Surveys. In fairness, it should be said that our cancer registration statistics are considerably better than those available for many parts of Europe and North America, but that is no grounds for complacency. The size of the problem may be seen from the 155 096 cases registered in 1970, a rate of 333-2 and 300 7 per 100 000 in males and females, respectively. Lung cancer heads the list in men with a rate of 99 9 per 100 000, and in women the rate for breast cancer is 72-2; these two diseases account for 29% and 24% of all cases of cancer in the two sexes, respectively. The registration of gastric cancer is falling slowly but not as fast as the decline in mortality might suggest, while the rates of other tumours of the alimentary tract are increasing (oesophagus, pancreas, and large bowel); cancers of the lip and tongue are on the wane. The rise of Hodgkin's disease and myeloma may reffect the growing interest in these conditions and the impact of various national trials. Clearly, registrations of borderline conditions such as myelofibrosis and polycythemia are gross underestimates of their frequency. A similar and more serious problem is the considerable variation in the apparent incidence from one region to another of carcinoma in situ of the cervix: Newcastle reported 15-9 per 100 000 while Liverpool said the rate was 1 1. Such divergence rules out the use of this data in
