Saturday 8 May
A-bomb survivors. IS Beebe et a1.I9 demonstrated this for the years 1962-66, but it was not apparent for 1954-62. A preliminary survey 20 had shown that carcinogenic effects of radiation were especially prominent in survivors less than ten years old at the time of the bombs (A.T.B.). We have examined more completely the evidence for radiation carcinogenesis among survivors, the oldest of whom is now only thirty-five.
CANCER IN JAPANESE EXPOSED AS CHILDREN TO ATOMIC BOMBS SEYMOUR
KIYOSHI TACHIKAWA ARTHUR STEER
JABLON
JOSEPH L. BELSKY
Departments of Statistics, Medicine, and Pathology, Atomic Bomb Casualty Commission, Hijiyama Koen, Hiroshima,
Japan The incidence of cancer has been studied in a cohort of 20,609 people who were less than ten years old at the time of the atomic bombs in Hiroshima and Nagasaki in 1945, including 15,584 exposed survivors and 5025 unexposed controls. Among survivors who received 100 rad or more or whose radiation doses are unknown no cancers were diagnosed before 1958. However, were alive the 1109 who survivors among high-dose in January, 1955, 8 had died of leukæmia by 1969 as compared with 0.42 such deaths expected and 8 more had died of cancer other than leukæmia as compared with 1.1 expected. Both differences are
Summary
highly significant statistically.
Additionally,
I97I
an
excessive incidence of solid tumours was found for carcinoma of the thyroid (5 cases compared with 2.2 expected) and other clinically diagnosed cancers (6
Methods
Population The population studied under
consists of those who
were
years of age A.T.B.
(Hiroshima, Aug. 6, and Nagasaki, Aug. 9, 1945) and who are members of the cohort called the J.N.I.H.-A.B.C.C. life-span study (L.s.s.) extended sample.19 This cohort includes nearly ten
109,000 persons, of whom about 82,000 are survivors of bombings, exposed at all distances, while nearly 27,000 are persons who entered the cities after the bomb-
the
ings
and who form one comparison group (" not in city ", group). Details of the definition and selection of the
N.I.C.
cohort
are
given elsewhere.18,19
I-THE J.N.I.H.-A.B.C.C. LIFE-SPAN STUDY (EXTENDED) TABLE COHORT (AGE < 10 YR. A.T.B.) BY DISTANCE FROM HYPOCENTRE AND BY SEX AND BY CITY
observed, only 1.3 expected). Introduction
ONE of the main purposes of the continuing studies of the Atomic Bomb Casualty Commission (A.B.C.C.) on the survivors of the Hiroshima and Nagasaki atomic bombings is to obtain information about radiation carcinogenesis in man. Leukxmia was an as identified radiation effect among important early but evidence for the radiation induction survivors,1-9 of other tumours has been slow to accumulate, and the effects observed have been less striking. Data from the Hiroshima and Nagasaki tumour registries for the period 1957-59 showed that survivors who had been located within 1500 m. of the hypocentres had a significantly elevated incidence of malignant neoplasms, excluding leukasmia and lymphoma. 10 Thyroid carcinoma was found to be increased in survivors who had received large doses.11-15 Lung cancer 16 and cancer of the female breast 17 have been at least suggestively linked with radiation exposure in the A.B.C.C. studies. However, no other specific tumour has yet been shown to be radiation related among the survivors. On the other hand, studies of mortality, using death certificates, have suggested excessive mortality from cancer (apart from leukaemia) among 7706
20,609 members of the entire cohort years of age A.T.B. (table i), and we have incidence of tumours in this subgroup.
under
ten
investigated
the
were
of Information on Tumours Transcripts of death certificates
Sources
are routinely obtained for all deaths in the cohort.211 Death-certificate notifications in these two cities are about 90% accurate as to cancer as a class,22 even if they are not always correct as to specific site. Moreover, they are virtually complete: the special arrangements made by the Ministry of Health and Welfare for the collaborative life-span study ensure that eventually we learn about more than 99% of the deaths in sample members, anywhere in Japan.211 However, about 5% of the transcripts, pertaining to persons who have moved to other places in Japan, are sometimes delayed as much as three years. We believe that our death notifications are essentially complete up to the end of 1968, but a small proportion of 1969 deaths may be still .
unreported.
,
’
We also searched the necropsy and
surgical-pathology
928 files of A.B.C.C., Hiroshima and Nagasaki University Medical Schools, and all hospitals in the two cities which do their own pathology. Finally, we have utilised the A.B.C.C. medical files and the Hiroshima and Nagasaki tumour registries. The medical files at A.B.C.C. pertain to the members of the sample for the A.B.C.C.-J.N.I.H. adult health study (A.H.S.),23 a subgroup of 20,000 persons out of the larger cohort of 109,000. The tumour registries, with which A.B.C.C. cooperates, are activities of the Medical Associations of the two cities. Sources of diagnoses such as tumour registrations and surgical-pathology files, unlike death notices, are incomplete : if the patient does not seek treatment for a slowgrowing tumour or if medical care is given outside the two cities, we do not ordinarily learn of the diagnosis unless it later figures as a cause of death. A further advantage of death-certificate data is that direct comparison with published official vital statistics is possible, whereas there are no comparable external data for rates based on other notification systems. We therefore treat separately diagnoses found on death certificates and those obtained from other sources.
TABLE III-SPECIFIC CAUSES OF MORTALITY FROM MALIGNANT NEOPLASMS OTHER THAN LEUKaeMIA 1955-69, L.S.S. COHORT YR. A.T.B. AGE < 10
Dosimetry The dosimetry system used by AB.C.C.24 rests upon theoretical and experimental work done at the Oak Ridge *
TABLE II-NUMBER OF DEATHS ATTRIBUTED TO LEUKIEMIA AND TO YR. OTHER MALIGNANT NEOPLASMS, L.S.S. COHORT AGE < 10 A.T.B. (ALL RADIATION CATEGORIES)
Located 1866
m.
t Located 1339
m.
the
from hypocentre; dose probably very small. from hypocentre; dose probably over 100 rad.
"
total dose "-i.e., the sum of the collision) and gamma doses in rad.
neutron
(iirst
Results
*
From Oct. 1, 1950.
National Laboratory 25 and the National Institute of Radiological Sciences of Japan. 26** Many thousands of field interviews were done by A.B.C.C. investigators to obtain the detailed information about each survivor required to enable estimates of the " air dose" at the place of exposure and the attenuation attributable to the shielding configuration in which he was located. The dosimetry system provides individual dose estimates that are believed to be accurate to a factor of ±30% or so. Separate estimates are made of the dose from gamma radiation and from neutrons. However, neutrons were of importance only in Hiroshima,24 and here we use simply
Death Attributed to Cancer Table 11 shows the number of deaths reported as being due to leukaemia and to other cancer for 1950-69 among those less than ten years of age A.T.B. Leukaemias far outnumbered other cancers from 1950 to about 1960, but from 1965 forward the situation was reversed. The first cancer death not ascribed to leukaemia was in 1954, when a girl died in Hiroshima from lymphosarcoma. However, this girl was not an A-bomb survivor, but a member of the N.I.C. component (control) of the sample. We shall henceforth confine our attention to the period beginning Jan. 1, 1955, since, apart from leukaemia, there was no evidence of radiation carcinogenesis before that time. The 22 deaths during 1955-69 from malignant neoplasms other than leukaemia are listed in table III in order of occurrence. The diagnoses are those given
TABLE IV-OBSERVED AND EXPECTED MORTALITY FROM LEUK1MIA AND OTHER MALIGNANT L.S.S. COHORT AGE < 10 YR. A.T.B.
NEOPLASMS, 1955-69,
* Based on Japanese national death-rates for 1962 using rates for 5-year age intervals applied to the cohort experience. t A small proportion of the survivors are credited with suspiciously high estimated doses. We have therefore arbitrarily used
estimated doses of 600 rad or more. t Differs from the total number 20,609 shown in tablei because of 92 deaths from Oct. 1, 1950, persons with
to Dec.
31, 1954.
600 rad for all 51
929 as
underlying
cause
of death
However, in 14 of the 22
on
the certificates.
cases ante-mortem
surgical
were available at the time of death. 5 of the cases came to necropsy, including 3 in whom no surgical pathology had been done. 6 have had surgery for the tumour, including 2 cases for whom neither surgical-pathology nor necropsy reports were available. In short, in all except 3 cases, records of confirmatory pathological or surgical evidence were found, the exceptions being cases A-6, A-13, and A-20. In two cases the radiation dose could not be estimated by the dosimetry system. 1 girl (case A-17) was 1866 m. from the Nagasaki hypocentre, where the dose in the open was about 30 rad; however, she was in an air-raid shelter, where shielding was probably quite effective. A boy (case A-20) was 1339 m. from the Nagasaki hypocentre, where the air dose was about 230 rad. He was in the open, but shielded by a large tree in the direction of the bomb burst; the dose cannot be calculated but it must have exceeded 100 rad.
pathology diagnoses
During
the
fifteen-year period 1955-69,
among
those with dose estimates of 10-99 rad, there was a small excess of mortality from leukaemia but not from other cancers (table iv); in fact, there were no nonleuksmic cancer deaths, although 4 would have been expected at Japanese national rates.2’ For survivors with doses exceeding 100 rad, both leukaemia and other cancer mortality was excessive, the excess being highly significant (P<0-0001) in each instance. The ratio of observed to expected mortality was more than twice as great for leukaemia as for other cancers in this high-dose group, but the differences between observed and expected deaths is about the same: for leuksemia, there were 8 deaths contrasted with 0.42 expected, and for other cancers there were also 8 deaths, with 1-11 expected. We have " arbitrarily included the dose unknown " group with those of known dose over 100 rad, although some of them certainly had smaller doses. Most of these persons were close to the hypocentres, where the air
dose was high. However, they were in concrete structures, crowded street cars, or other physical configurations that made estimation of the attenuation due to shielding impossible. With respect to type of cancer, there was no striking difference between the high dose (more than 100 rad or dose unknown) and low dose (less than 10 rad) cases. The 8 high or unknown dose cases included 2 each of stomach cancer and osteogenic sarcoma, 1 each of reticulosarcoma, rhabdomyosarcoma of the prostate, and carcinoma of the pancreas, and a case of carcinoma metastatic to the liver, primary site unknown. The 14 low-dose cases included 7 stomach cancers, 1 osteogenic sarcoma, a case of Hodgkin’s disease, and 5 other single cases of malignancies not duplicated in the high-dose group. The only suggestive features of the comparison are the relatively low frequency of stomach cancer in the high-dose group (2/8) as compared with the low-dose group (7/14), and the relatively high proportion of sarcomas (4/8 as compared with 2/14).
Thyroid Cancer Thyroid cancer is difficult to evaluate. Although some rare forms grow rapidly, metastasise widely, produce symptoms early, have a short clinical course, and are easily diagnosed, other more common forms grow slowly, tend to have restricted, local metastases, produce few symptoms until late, and are generally diagnosed early only when nodular thyroid enlargements are specifically searched for by an alerted physician or because the patient seeks help. Since 1958 thyroid cancer has been a subject of diligent search at A.B.C.C. in the adult-health-study (A.H.S.) examinations, which are done every two years on a subsample of about a fifth of the larger L.s.s. sample. The probability of detection of thyroid cancer in members of this subsample is, therefore, quite differfrom that in the remainder of the L.s.s. cohort. In fact, thyroid carcinomas have been diagnosed in 8 out of the 1819 members of the A.H.s. subsample who were less than ten years of age A.T.B., and in ent
TABLE V-THYROID CARCINOMAS IN L.S.S. COHORT AGE <
10
YR. A.T.B.,
1955-69
’ Death was certified as due to acute spinitis (probably acute myelitis). t Carcinoma of thyroid an incidental autopsy finding; cause of death, leukaemia. t Death certified to symptomatic heart-disease ". Principal necropsy diagnosis was adenocarcinoma of the thyroid. § This patient died in 1969. Death was certified as caused by " cardiac asthma ". The principal autopsy diagnosis, however, thyroid, with widespread metastases. A Autopsy. S Surgical pathology. Med. ref. = Medical referral. "
"
"
=
=
was
carcinoma of the
930
truly six times that in the baseline group, still only 0-8 cases would be expected. The fact that no cases have yet been noted in this group, therefore, does not necessarily imply that the risk actually is
only 7 out of 18,698 other members of the L.s.s. sample. The 15 cases are listed in table v. None was
were
detected before 1955. Of the 8 cases of thyroid cancer among members of the A.H.s., 5 (cases B-2 to B-6) were first detected at an A.B.C.C. clinical examination. In case B-1, carcinoma of the thyroid was found at necropsy, and 2 ’more patients (B-7 and B-8) sought medical care because of their thyroid problem. Among the 7 cases not in the A.H.s., 5 came to attention because they sought medical care, and in 2 instances the thyroid malignancies were found at necropsy. Since the case-finding procedures were so different in the A.H.s. portion and the remainder of the i.,.s.s. cohort, the occurrence of thyroid carcinoma is analysed separately in the two components of the sample (table vi). A relationship with radiation, significant at the 5% level, was found within each component. In the A.H.s. 10-99 rad dose group, the relative risk, compared with the N.I.C. and 0-9 rad group as a baseline, was nearly two, and in the high dose group 100 rad or unknown) was nearly six. In the (over " other " component of the L.s.s. cohort, the relative risk was more than six in the 10-99 rad group, and no cases occurred in the higher dose group. However, the numbers are very small: even if, in the " other " component; the relative risk in the 100+ rad group
low.
Other tumours Besides the malignant neoplasms certified as underlying cause of death there were, thyroid carcinomas apart, 24 other persons who were diagnosed as having a cancer
calculating expectation. 6 malignant tumours were diagnosed among the approximately 1100 children with doses exceeding 100 rad (or unknown). The cancer-rate was more than six times that in the children who were not in the cities or who had doses estimated to be less than
TABLE VI-OBSERVED AND EXPECTED NUMBERS OF CARCINOMAS OF THE THYROID
* Survivors as of Jan. 1, 1955. t Expectations obtained by applying the total rates (8/1819 and 7/18,698) t Risk relative to the N.I.C. or 0-9 rad group as standard.
TABLE VII-DIAGNOSES OF MALIGNANT
(table vn).
The earliest case, C-l, was a Nagasaki boy, with an estimated dose of 47 rad, who died in 1952 from a brain tumour. This death was classed as a " neo" plasm of unspecified nature according to the International Classification of Diseases (l.c.D.), and hence is not included in table n as a death attributed to a malignant neoplasm. The distinction is, perhaps, somewhat artificial, but we had to follow the I.C.D. coding rules so that national death-rates, which are based on these rules, might be used as a basis for
to
GLAND,
the individual
TUMOURS,*
1955-69,
L.S.S. COHORT AGE <
10
YR. A.T.B.
population sizes.
L.S.S. COHORT AGE <
10
YR. A.T.B.
*
Excludes leukxmia, thyroid cancer, and deaths certified as due to malignancy; see text. S Surgical pathology. R Radiological report. - Basis of diagnosis: A = Necropsy. t Dose unknown; located 1534 m. from hypocentre, where air dose was about 110 rad. =
=
C = Clinical
diagnosis,
none
of preceding.
931 TABLE VIII-CANCER NOTIFICATIONS* OTHER THAN DEATH CERTIFIYR. A.T.B. CATES BY RADIATION DOSE L.S.S. COHORT AGE <10
dose groups.
However, strong arguments brought against this point of view:
can
be
1. Members of the N.i.c. group are, very largely, a combination of immigrants from rural areas and persons who returned after the end of the war from Korea, Manchuria, China, and South-East Asia. The contrast between such persons and the city dwellers is probably much larger than the contrast between residents of two different parts of the same city, especially in Japan where urban geographic stratification of residence by income levels is not
prominent.
Excludes leukxmia, thyroid cancer, and deaths certified
malignancy; see text. Survivors to Jan. 1, 1960. t Risk relative to the N.i.c.
or
0-9 rad group
as
as
due
2. The contrast in rates, in either absolute or relative terms, between high and low dose groups of survivors, is much larger than that between those not in the cities and the low-dose group. to
standard.
10 rad (table vm). The rate increased rapidly after 1960 in all dose classes as the children came into adulthood: in 1960-64 there were 7 tumours, includding 2 in the high-dose group, and in 1965-69 there were 14 tumours, of which 3 were in the high-dose group. The kinds of tumours represented vary widely, and those diagnosed among the children who received large radiation doses do not seem to differ remarkably from those diagnosed in the other groups. Discussion
It is striking that, despite wide differences as to ascertainment and nature of tumour, the results are fairly consistent (tables iv, vi, and VIII): the relative risks for the high-dose group are 7-3 for death attributed to malignant neoplasms, 4-5 for clinically detected thyroid carcinomas, and 6.1 for all other malignant tumour notifications. The increased risk associated with high radiation doses is not likely to be due to chance, since the significance levels for tables iv and vm are both less than 0.0001. There are, however, some puzzling features. In tables iv and vm the rates are much higher in the N.l.c. group than in the 0-9 rad group. These discrepancies are too large to ascribe to chance, and we cannot explain them. The difference is not explicable in terms of dissimilarities by city, sex, or medical It is of interest, however, that services used. Mewissen 2S has reported the incidence of thymic lymphosarcoma to be decreased in mice experimentally irradiated with 9 rem of neutrons as compared with the incidence in controls. The cohorts were defined on the basis of early censuses and, in any case, were closed by 1962, before the occurrence of most of the illness reported here. The classification by radiation dose for these children is practically equivalent to a classification by place of residence in 1945, and there are undoubtedly differences in socioeconomic status among the various dose-groups, perhaps especially as regards the N.I.c. groups as contrasted with the survivors. If we choose to regard probable socioeconomic factors as a likely explanation of the high tumour-rates in the N.i.c. group, it would seem to follow that such factors might also explain the high rates in the high-radiation case
3. An extensive body of evidence exists to show that radiation is, in fact, a carcinogen in animals and in man, including uranium miners 29 and radiologists. 30
Therefore, although it is ences
not
unlikely that differ-
among the groups somewhat affect the
measures
of relative risk, it seems probable that those children who received radiation doses of 100 rad or more have indeed, as a consequence, been experiencing a significantly elevated rate of tumour induction. The number of tumours observed during the period 1955-69 is not large: in what we have called the high-dose group there were 8 cancer deaths and 8 from leukxmia; there were also 5 carcinomas of the thyroid and 6 other cancers, not reported on death certificates, that we know about. However, although there are only 19 malignant neoplasms in total other than leukaemia, these have occurred in a group of only 1109 persons, the oldest of whom is now only thirty-five. By contrast, even in the N.i.c. group, where rates seem to be abnormally high, there were only 18 cancers discovered among 5010 persons, and among the survivors with doses estimated in the range 0-99 rad, only 24 cancers in 14,398 persons. The rates, per 1000, were 17-1, 3-6, and 1.7 in the high-dose group, N.I.C. group, and 0-99 rad group,
respectively. In short, after a (i.e., since 1960)
latent period of about fifteen years children who received radiation doses of 100 rad or more began to develop an excessive number of malignant neoplasms and now, twenty-five years after exposure, the accumulated increase is striking with no evidence yet that a peak has been reached. During the next ten years these persons will be entering upon ages when, ordinarily, cancer-rates begin to increase strikingly. Clearly these people must be followed up closely and continuously. A.B.C.C. is a cooperative research agency of the U.S. National Academy of Sciences and the Japanese National Institute of Health, with funds provided by the U.S. Atomic Energy Commission and the Japanese Ministry of Health and Welfare. We thank the following persons and organisations who have kindly allowed us to make use of their material: Hiroshima University (Dr. S. Iijima, Dr. A. Yamada, Dr. T. Dodo, and Dr. T. Tsubokura); Hiroshima City Medical Association (Dr. H. Sawachika); Nagasaki University (Dr. 1. Hayashi, Dr. H. Tsuchiyama, Dr. S. Matsuoka, Dr. I. Nishimori, and Dr. T. Itoga); and Nagasaki City Medical Association (Dr. S. Miyagi). Requests for reprints should be addressed to the Editor, Atomic Bomb Casualty Commission, 5-2 Hijiyama Koen, Hiroshima 730, Japan.
References overleaf
932
LEVODOPA IN PARKINSONIAN PATIENTS WITH HEART-DISEASE A. HOLLMAN G. M. STERN
K. R. HUNTER D. R. LAURENCE
University College Hospital and Medical School, London
cardiac dysrhythmias has been described in parkinsonian patients not taking levodopa.A comprehensive assessment of cardiac abnormalities attributable to levodopa will require a long-term survey of many patients, but we felt that a preliminary study of the extent of possible hazards was indicated.
W.C.1 Patients and Methods to 20
Levodopa given parkinsonian patients with heart-disease and observations were continued for up to 57 weeks. 1 patient had a recurrence of angina; 2 developed dysrhythmias but only 1 required specific treatment with a &bgr;adrenoreceptor blocker. 1 patient died after 8 weeks. It is suggested that this degree of risk is acceptable in patients with disabling parkinsonism but they should be admitted to hospital for institution of therapy. was
Sum ary
Introduction
LEVODOPA is the most effective drug available for the majority of patients with parkinsonism and is currently prescribed unless a specific contraindication is present. During early trials cardiac dysrhythmias 1-4 and myocardial infarction 1,5 were noted in certain patients taking levodopa; these abnormalities have been attributed to therapy, although a high incidence of
All patients attending the neurological department of University College Hospital with parkinsonism (other than drug-induced syndromes) and abnormal electrocardiograms (E.C.G.S) or histories of heart-disease were assessed by a cardiologist and were admitted to this study if there was a significant cardiac abnormality. 20 patients fulfilled this criterion; details are shown in the accompanying table. Many were taking anticholinergic and other drugs. In contrast to our usual practice, all the above patients were admitted to hospital. Pulse-rate and supine and erect blood-pressure were recorded four times daily. In most of the patients with angina pectoris or a previous myocardial infarction, E.C.G.s were continuously displayed using chest leads attached to a Hewlett Packard model 708-7A monitor. A ten-second sample was permanently recorded every fifteen minutes, and an additional permanent -trace was automatically printed if the heart-rate deviated beyond preset limits. Recordings were continued until the appropriate daily dose of levodopa was reached. Patients were later followed in the clinic and serial E.C.G.s were recorded.
Results 1. 2. 3. 4. 5. 6. 7.
8. 9. 10. 11. 12.
13.
14. 15. 16. 17. 18. 19.
Folley, J. H., Borges, W., Yamawaki, T. Am. J. Med. 1952, 13, 311. Lange, R. D., Moloney, W. C., Yamawaki, T. Blood, 1954, 9, 574. Moloney, W. C. New Engl. J. Med. 1955, 253, 88. Moloney, W. C., Kastenbaum, M. A. Science, 1955, 121, 308. Wald, N. ibid. 1958, 127, 699. Heyssel, R. M., Brill, A. B., Woodbury, L. A., Nishimura, E. T., Ghose, T., Hoshino, T., Yamasaki, M. Blood, 1960, 15, 313. Brill, A. B., Tomonaga, M., Heyssel, R. M. Ann. intern. Med. 1962, 56, 590. Bizzozero, O. J., Jr., Johnson, K. G., Ciocco, A. New Engl. J. Med. 1966, 274, 1095. Ishimaru, T., Hoshino, T., Ichimaru, M., Okada, H., Tomiyasu, T., Tsuchimoto, T., Yamamoto, T. Radiat. Res. 1971, 45, 216. Harada, T., Ide, M., Ishida, M., Troup, G. M. Hiroshima Igaku, 1969, 22, 1084. Hollingsworth, D. R., Hamilton, H. B., Tamagaki, H., Beebe, G. W. Medicine, Baltimore, 1963, 42, 47. Socolow, E. L., Neriishi, S., Niitani, R., Hashizume, A. New Engl. J. Med. 1963, 268, 406. Wood, J. W., Tamagaki, H., Neriishi, S., Sato, T., Sheldon, W. F., Archer, P. G., Hamilton, H. B., Johnson, K. G. Am. J. Epidemiol. 1969, 89, 4. Angevine, D. M., Jablon, S. Ann. N.Y. Acad. Sci. 1964, 114, 823. Sampson, R. J., Key, C. R., Buncher, C. R., Iijima, S. J. Am. med. Ass. 1969, 209, 65. Wanebo, C. K., Johnson, K. G., Sato, K., Thorslund, T. W. Am. Rev. resp. Dis. 1968, 98, 778. Wanebo, C. K., Johnson, K. G., Sato, K., Thorslund, T. W. New Engl. J. Med. 1968, 279, 667. Jablon, S., Ishida, M., Yamasaki, M. Radiat. Res. 1965, 25, 25. Beebe, G. W., Kato, H., Land, C. E. A.B.C.C. techn. Rep. 1970, no. 11-70.
Jablon, S., Belsky, J. L. 10th int. Cancer Congr. (in the press). Beebe, G. W., Ishida, M., Jablon, S. Radiat. Res. 1962, 16, 253. Beebe, G. W., Yamamoto, T., Matsumoto, Y. S., Gould, S. E. Hiroshima Igaku, 1968, 21, 729. 23. Atomic Bomb Casualty Commission-Japanese National Institute of Health. A.B.C.C. techn. Rep. 1962, no. 11-62. 24. Milton, R. C., Shohoji, T. ibid. 1968, no. 1-68. 25. Auxier, J. A., Cheka, J. S., Haywood, F. F., Jones, T. D., Thorngate, J. H. Hlth Phys. 1966, 12, 425. 26. Hashizume, T., Maruyama, T., Shiragai, A., Tanaka, E., Izawa, M., Kawamura, S., Nagaoka, S. ibid. 1967, 13, 149. 27. Division of Health and Welfare Statistics. Vital Statistics of Japan, 1962, vol. i. Health and Welfare Minister’s Secretariat, Tokyo. 28. Mewissen, D. J. in Radiation Dose and Effects (edited by R. J. Cloutier, C. L. Edwards, and W. S. Snyder); p. 413. Oak Ridge, Tennessee, 1970. 29. Wagoner, J. K., Archer, V. E., Lundin, F. E., Jr., Holaday, D. A., Lloyd, J. W. New Engl. J. Med. 1965, 273, 181. 30. Seltser, R., Sartwell, D. E. Am. J. Epidemiol. 1965, 81, 2. 20. 21. 22.
In 16 of the 20 patients no change in cardiac state was found except that in case 4 ventricular extrasystoles stopped. In the remaining 4 patients changes were seen which may have been related to levodopa. Case 5: Recurrent Paroxysmal Tachycardia In 1937 this patient was found to have extrasystoles when he complained of palpitations: during the ensuing thirty years he had at least two further attacks. In the assess-
Case 5: E.C.G.
showing short
paroxysm of atrial
tachycardia.
of his parkinsonism of five years’ duration, an E.C.G. showed slight ST depression in leads V and V After thirty-six weeks of levodopa therapy he suddenly felt faint and giddy at home and was found to have a heart-rate of 140 per minute : he was given phenobarbitone and recovered within hours. A similar attack three days later was treated with oral propranolol. When he was referred to hospital five days later his E.C.G. showed sinus rhythm with bursts of atrial tachycardia (see accompanying figure); propranolol was discontinued; he was digitalised and the dose of levodopa was halved to 2-0 g. daily. Three days later the E.c.G. was unchanged except that ventricular ectopic beats were also present. Digoxin was stopped and oral practolol 100 mg. b.d. started; after three days practolol was increased to 100 mg. t.d.s. and levodopa to 3-0 g. daily because his parkinsonism had markedly deteriorated. After ten days on practolol his E.C.G. was found to have reverted to sinus rhythm; practolol was continued for eight weeks. ment
Case 9: Atrial Ectopic Beats After total thyroidectomy twenty-two years
previously