from haemoglobin A on acid electrophoresis. Another pigment migrating on alkaline paper electrophoresis behind J and in front of A is haemoglobin K. Haemo-.
MAY 21, 1960
HAEMOGLOBIN
"
SINGAPORE-BRISTOL"
test and the ultra-violet spectrum type which had been so decisive in the case of haemoglobin Bart's. The variant described here differs from all hitherto described human haemoglobins excepting that observed by Vella (1959) and then referred to as the F & P type. There is even one difference from that described by the Greek authors: the present pigment does not separate from haemoglobin A on acid electrophoresis. Another pigment migrating on alkaline paper electrophoresis behind J and in front of A is haemoglobin K. Haemoglobin K, however, moves slightly slower than the pigment described here and does not separate from haemoglobin A on chromatography. We cannot be sure that the present haemoglobin is truly a variant of haemoglobin F. It could conceivably be a recessive character in the adult series. After consultation with Dr. Vella (Siulgapore), Mwe have decided on the provisional designation haemoglobin "SingaporeBristol." Summary Of two " fast " haemoglobins seen in infants, but not in their parents, one moves more slowly than haemoglobin J on paper electrophoresis at alkaline pH, and the other faster. The second has been described as haemoglobin " Bart's " and is now thought to be the unnamed abnormal foetal haemoglobin seen by Fessas and Papaspyrou in 1957 (Fessas, 1959b). The other moves more slowly than haemoglobin J, and two new examples in infants of mixed marriages are described. This pigment has previously been referred to as the haemoglobin of Fessas and Papaspyrou (" F & P "); it is now redesignated haemoglobin "Singapore-Bristol," and its properties are described.
ADDENDUM.-Two relevant reports have appeared Virginia Minnich and her colleagues (1959) examined the cord blood of 200 infants (110 negroes and 90 Caucasian), and in nine of the negro babies and one of the Caucasians a fast-moving haemoglobin was found which moved more slowly than haemoglobin J on electrophoresis at alkaline pH. These authors state that they found this fraction to be relatively resistant to alkali denaturation, whereas we did not find this to be the case for haemoglobin Singapore-Bristol. Hendrickse and his colleagues (1960) found a " fast " haemoglobin fraction in 11 % out of 100 cord bloods examined in Nigeria. This fraction persisted for only the first two months of life. Three were examined further, and one of them was found to resemble haemoglobin Bart's and two the haemoglobin described in the present paper. since the above studies were completed.
We thank Dr. F. Vella for discussion and permission to reproduce results obtained with haemoglobin specimens sent to us by him; Professor C. B. Perry and Dr. J. Apley for permission to study their cases ; and especially Professor D. V. Hubble for his help in tracing the father in Case 2.
BMcM
RESPONSE TO GLUCAGON BY SUBJECTS Wi HYPERINSULINISM FROM ISLET-CELL TUMOURS BY
VINCENT MARKS, B.M., M.R.C.P.Ed. Department of Chemical Pathology, Institute of Neurology, National Hospital, Queen Square, London
The isolation and purification (Staub, Sinn, and Behrens, 1955) of glucagon-the hyperglycaemic-glycogenolytic factor-which contaminates many commercial preparations of insulin, enables its effects to be studied in man. Its exact physiological role in intact animals is still in doubt (Lancet, 1955; Brit. med. J., 1956), but it is generally agreed that working in conjunction with insulin it facilitates glucose transport from the liver to the peripheral tissues and its utilization there (Anderson, Monaco, Perfetto, and Termine, 1957). This paper records the results of administering glucagon to four subjects with hyperinsulinism.
Material and Methods Glucagon, 1 mg. in 1 ml., was given by intramuscular injection in subjects in the postprandial state, about six to eight hours after the last meal. Glucose concentrations were measured in capillary blood before and at 5, 10, 15, 30, 45, 60, 120, and 150 minutes after the injection by a glucose-oxidase method (Marks, 1959). Four subjects with organic hypoglycaemia from hyperinsulinism were studied in addition to normal subjects and patients suspected of hypoglycaemia but in whom the diagnosis was not substantiated. In three of the hypoglycaemic patients an islet-cell tumour was removed at operation. The fourth patient refused operation though the diagnosis of insulinism was not
in doubt clinically. Results The results in normal subjects resemble those observed by others (Hubble, 1955; Alivisatos and
120 100 0
00
~~~~~~~~~~Typical
bo
~~~~~NORMAL
E
O U
60
-J
40
REFERENCES
Ager, J. A. M., and Lehmann, H. (1958). Brit. med. J., 1, 929. Fessas, Ph. (1959a). In Abnormal Haemoglobins, edited by J. H. P. Jonxis and J. F. Delafresnaye, p. 134. Blackwell,
Oxford. (1959b). Brit. med. J., 2, 886. -and Papaspyrou, A. (1957). Science, 126, 1119. Hendrickse, R. G., Boyo, A. E., Fitzgerald, P. A., and Kuti, S. R. (1960). Brit. med. J., 1, 611. Minnich, V., Cordonnier, J. K., Jones, B., and Kingberg, W. G. (1959). A.M.A. J. Dis. Child., 98, 488. Singer, K., Chernoff, A. I., and Singer, L. (1951). Blood, 6, 429. Vella, F. (1959). Nature (Lond.), 184, 272.
1539
2
203 4 0
30
r"I m9I.M.
60
90
120
150 minutes
GLUCAGON
Response to 1 mg. of glucagon given intramuscularly in four subjects with byPernsulinsm from islet-cell tumour. The results from a typical normal individual are shown for companisoD.
1540 MAY 21, 1960
McCullagh, 1955); there is a rapid rise in blood sugar, about 50-100 mg./100 ml., which reaches a peak within half- to three-quarters of an hour, gradually returning to normal by two hours. The results in four patients with hyperinsulinism are shown in the Chart. Two had " chemical " hypoglycaemia-that is, blood sugar below 40 mg. /100 ml.prior to the glucagon injection, but only one (Case 4) displayed symptoms. In every case the injection was followed by a rapid rise in blood-sugar, and in the patient with symptomatic hypoglycaemia it. was accompanied by clinical improvement. Blood glucose reached a peak by the thirtieth minute and had fallen to, or below, the lower limit of normal by the ninetieth. At two hours severe hypoglycaemic symptoms necessitated termination of the test with glucose in two subjects. In the remaining two subjects hypoglycaemia symptoms appeared between two and two and a half hours; in neither case were they severe, but their hypoglycaemic nature was confirmed by a well-marked clinical improvement with administration of glucose. Discussion The response to glucagon administration in patients with hyperinsulinism differs from that observed in healthy subjects by the development of hypoglycaemia after an initial normal hyperglycaemic phase. This characteristic response was observed in all four patients studied and in the one other subject with hyperinsulinism in whom the effects of administering a small dose of glucagon have been published (Alivisatos and McCullagh, 1955). It is not observed in normal subjects under the conditions described: Bondy and Cardillo (1956) gave glucagon by continuous intravenous infusion over a two-hour period and described " hypoglycaemic symptoms" in four subjects; the blood-sugar levels, however, did not fall below 80 mg./ 100 ml. A hypoglycaemic phase was not observed to follow glucagon in the two patients with " functional" (reactive) hypoglycaemia studied by Alivisatos and McCullagh. Subjects with liver disease do not show the usual hyperglycaemic response to glucagon (Kibler, Taylor, and Myers, 1952), and it might be expected that this would be equally true in cases of hepatogenous hypoglycaemia; but more information is required on this point. Thus glucagon may be useful in distinguishing subjects with hyperinsulinism from normal subjects and in the differential diagnosis of the hypoglycaemic syndrome; further investigation in this direction seems
justified. Summary The response to the intramuscular injection of 1 mg. of glucagon in four patients with hyperinsulinism is recorded. A hypoglycaemic phase following a normal rise in blood sugar distinguishes it from the response observed in healthy subjects., The use of glucagon in the differential diagnosis of the hyperglycaemic syndrome is discussed. It is a pleasure to acknowledge the gift by Messrs. Eli Lilly and Company Ltd., Basingstoke, England, of the glucagon used in this work. REFERENCES
Alivisatos J. G., and McCullagh, E. P. (1955). J. Amer. med. Ass., 159, 1098. Anderson, G. E., Monaco, R. N., Perfetto, A. J., and Termine, C. M. (1957). Diabetes, 6, 239. Bondy, P. K., and Cardillo, L. R. (1956). J. clin. Invest., 35, 494.
IBRDIJJU
HYPERINSULINISM
Brit. med. J., 1956, 2, 288. Hubble, D. (1955). Diabetes, 4, 197. Kibler, R. F., Taylor, W. J., and Myers, J. D. (1952). Amer. J. Med., 13, 647. Lancet, 1955, 2, 912. Marks, V. (1959). Clin. chim. Acta, 4, 395. Staub, A., Sinn, L., and Behrens, 0. K. (1955). J. biol. Chem., 214, 619.
SERUM TRANSAMINASE I1 ENDOMYOCARDIAL FIBROSIS BY
JEAN CAMPBELL, B.Sc.
Biochemist, Uganda Medical Department AND
K. SOMERS, M.B., M.R.C.P. Lecturer in Medicine, Makerere College Medical School, Kampala, Uganda
The pathology of endomyocardial fibrosis essentially consists of fibrosis of the ventricular cavity endocardium and of the underlying myocardium, usually around the region of the apex. Extension of the lesion may sometimes involve the atrio-ventricular valves. Histologically, acute necrosis of myocardial fibres has not been seen (Davies and Ball, 1955). The usual presentation is the clinical picture of congestive cardiac failure, often with evidence of pure mitral and/or tricuspid incompetence, but it often presents as heart failure without other distinguishing features (Ball et al., 1954). Despite the recognition of a suggestive electrocardiographic pattern (Williams and Somers, 1960), and characteristic phonocardiographic signs (Somers and Williams, 1960), an exact diagnosis of the syndromes of endomyocardial fibrosis is difficult and can be confirmed only at necropsy. LaDue et al. (1954) first reported the elevation of the serum glutamic oxalacetic transaminase (G.O.T.) after necrosis of cardiac muscle in man. Agress et al. (1955) found from experimentally produced myocardial infarction in dogs that the increase of serum G.O.T. was proportional to the amount of infarction estimated at necropsy. Serial estimations of serum G.O.T. and other enzymes assist in the diagnosis of myocardial infarction where electrocardiographic and other diagnostic features are atypical or doubtful (Kattus et al., 1957; Baron et al., 1958; Dewar et al., 1958; Sampson, 1958; Goble and O'Brien, 1958; Keele et al., 1958). The object of the present study was to assess the value, if any, of estimations of serum G.O.T., which we shall refer to simply as transaminase, in endomyocardial fibrosis. Material and Methods Serum transaminase activity was estimated in a series of 24 patients admitted to Mulago Hospital during the first six months of 1959. There were 12 males and 12 females. Their ages ranged from 12 to 55 years. All the patients were in congestive heart failure when the first estimation was made. Subsequent estimations were carried out in various stages of recovery. In seven patients serum transaminase was estimated for periods exceeding two weeks. The investigation was done in patients with suspected endomyocardial fibrosis in whom the problem of diagnosis from hypertensive heart disease, heart failure due to severe anaemia, aortic valve disease, and other conditions was excluded on clinical findings. The diagnosis of endomyocardial fibrosis was based on clinical evidence of pure atrio-ventricular incompetence -mitral, tricuspid, or both-or heart failure and triple
