Clinical Science (1990) 78,.469-474
469
Splenic responses to acute insulin-induced hypoglycaemia in humans B. M. FISHER', G. GILLEN*, D. A. HEPBUR", H. J. DARGIE3, E. BARNETT4 AND B. M. FRIER' Departments of I Diabetes, 'Medical Physics, Tardiology and 4Radiology,Western Infirmary/Gartnavel General Hospital, Glasgow, Scotland, U.K.
(Received 27 October 1989/12 January 1990; accepted 18 January 1990) SUMMARY 1. The effects of acute hypoglycaemia on the spleen were examined in normal humans using radioisotopic techniques, complemented by ultrasonic examination of the spleen. Hypoglycaemia had a modest effect on splenic area, measured by ultrasonography, which declined to 6 2 f 6% ( r n e a n f ~of~ the ~ ) basal value after the onset of the acute hypoglycaemic reaction. 2. Hypoglycaemia had a pronounced effect on the splenic radioactivity, which decreased significantly to a mean of 10 f 7% of basal radioactivity at 15 min after the onset of hypoglycaemia. The splenic image completely disappeared at some time after hypoglycaemia in all subjects. 3. The reduction of splenic radioactivity was abolished during non-selective a-adrenergic blockade with phentolamine, but was unaffected by p-adrenergic blockade with propranolol, or cholinergic blockade with atropine, which suggests that the response of vessels perfusing the spleen is mediated by a-adrenoceptors. Key words: hypoglycaemia, phentolamine, spleen, ultrasound. INTRODUCTION
Acute hypoglycaemia in humans provokes stimulation of the autonomic nervous system, including sympathoadrenal activation and the release of catecholamines [ 11. The effects of acute hypoglycaemia on heart rate and blood pressure are well defined: hypoglycaemia causes an increase in heart rate, a rise in systolic and a fall in diastolic blood pressure, with no change in mean arterial blood pressure [2]. Cardiac function in response to hypoglycaemia has been examined using non-invasive techniques: cardiac output [3] and myocardial contractility, Correspondence: Dr B. M. Fisher, Wards 4/5, Royal Infirmary, Glasgow G4 OSF, Scotland,U.K.
estimated by left ventricular ejection fraction [4], both increase significantly. These profound haemodynamic changes are thought to occur secondary to sympathoadrenal activation [3]. Knowledge of the physiological effects of hypoglycaemia on other vascular systems and perfusion of organs is, however, fragmentary. T h e aim of the present study was to examine the mechanisms underlying the responses of the spleen to acute hypoglycaemia. The role of autonomic mechanisms was assessed by pharmacological blockade: (a) with propranolol, pro.ducing non-selective blockade of p-adrenoreceptors, (b) with phentolamine, producing non-selective blockade of a-adrenoreceptors, and (c) with atropine, producing blockade of parasympathetic and sympathetic cholinergic neurotransmission. METHODS Subjects
The study was approved by the local medical ethical advisory committee and informed consent was obtained from all subjects. Each study group comprised six normal healthy male subjects, aged 21-30 years, all of whom had a normal body mass index, and none of whom was taking medications. A total of 30 experimental studies were performed in 24 subjects, with the six subjects in the control group being studied twice. As the changes in splenic radioactivity were greater than the changes in splenic area, the role of autonomic mechanisms was assessed by pharmacological blockade with radioisotopic imaging, and changes in splenic area were not measured during pharmacological blockade to avoid unnecessary exposure to hypoglycaemia. Protocol
T h e studies were divided into four groups as follows: (1) Control studies (n= 6): acute hypoglycaemia with no pharmacological blockade. The cardiac effects of
470
B. M. Fisher et al.
acute hypoglycaemia on the left ventricular ejection fraction in this control group have been published previously [4]. Each control subject was studied twice, on two separate occasions. On the first day, radioisotopic imaging of the spleen was performed. On the second day, ultrasonic imaging of the splenic area was performed with no pharmacological blockade. (2) a-Adrenergic blockade (n= 6): acute hypoglycaemia with radioisotopic studies during a-adrenergic blockade with 5 mg of phentolamine administered as an intravenous bolus injection, followed by 30 mg of phentolamine/h as an intravenous infusion [5]. (3) B-Adrenergic blockade (n= 6 ) : acute hypoglycaemia with radioisotopic studies during non-selective /3-adrenergic blockade with 10 mg of propranolol administered as an intravenous bolus injection, followed by 3 mg of propanolol/h as an intravenous infusion [6-81. (4) Cholinergic-blockade (n= 6): acute hypoglycaemia with radioisotopic studies during cholinergic blockade with 1.2 mg of atropine administered as an intravenous injection 30 min before insulin, and followed by 0.6 mg of atropine 1 h after the initial dose [9]. Subjects were fasted overnight and studied after remaining in a recumbent position for 1 h. The consumption of alcohol, tobacco, tea and coffee was avoided for 12 h before each study. At 08.30 hours, basal venous blood samples were withdrawn from an indwelling Teflon cannula which had been inserted into an antecubital vein. Pharmacological blockade was commenced after basal blood sampling, and intravenous soluble insulin (Human Actrapid; Novo, Basingstoke, Hants, U.K.) was administered as a single bolus at 09.00 hours in a standard dose of 0.15 unit/kg body weight. Serial estimation of haemodynamic variables and blood glucose were made every 5 min until the onset of the acute autonomic reaction (‘R), which was determined by the sudden increase in heart rate and the development of autonomic symptoms, including sweating, and coincided with the nadir of blood glucose. To eliminate the individual variability in the time taken to reach the nadir of blood glucose after the intravenous injection of insulin, subsequent blood sampling and cardiovascular measurements were timed from ‘R. Sampling was continued until ‘ R + 90 min when the study was terminated. The heart rate was measured continuously using precordial electrodes with the heart rate displayed on an oscilloscope (Life Trace 12; Albury Instruments Ltd, London). Blood pressure was measured using a mercury sphygmomanometer by a single observer with diastolic pressure being measured by Korotkoff’s 5th sound. Blood glucose was measured with the Cobas Bio Centrifugal analyser (Roche Diagnostica, Basel, Switzerland) using a hexokinase method. Splenic radionuclide imaging was performed using the multiple-gated technique with yymT (800 MBq). Erythrocytes were labelled iri vivo by the injection of isotope 30 min after the intravenous injection of stannous pyrophosphate. Subjects were studied supine using the Siemens LEM mobile gamma camera (Siemens Ltd, Des Plaines, IL, U.S.A.) in the left anterior oblique position.
Data were obtained at a rate of 20 frames/cycle with acquisition of data during 5 min. To determine the splenic radioactivity, a region of interest was drawn around the spleen. Splenic radioactivity was determined by a single observer who was unaware of the timing of the measurements. The basal splenic radioactivity was designated as 10Oo/~,and subsequent measurements were compared with the basal value. The intra-observer coefficient of variation for this technique is 6%. Ultrasonic imaging of the spleen was performed using a ATL real-time B Scanner using a 3.5 mHz medium focus probe (Squibb Medical Systems Ltd, Hemel Hempstead, Herts, U.K.). Longitudinal scans were made in the left axilla with the subject lying supine, using a fixed depth of focus. The splenic perimeter was outlined using electronic cursors and the maximum area of the longitudinal section was calculated. The intra-observer coefficient of variation for this technique is 8%. All ultrasonic measurements were made by a single observer (E.B.), who was unaware of the timing of the measurements with respect to hypoglycaemia. Statistics Statistical analyses were performed using the Statview 5 12 package (Brainpower Inc, Calabasas, CA, U.S.A.) on an Apple Macintosh Plus Computer (Apple Computer UK Ltd, Hemel Hempstead, H e m , U.K.). Initial comparisons were made using analysis of variance with post hoc comparisons of means using Student’s t-test for unpaired data. The results are presented as means k SEM. P values less than 0.05 were considered significant. RESULTS
During phentolamine infusion subjects developed nasal congestion and flushing of the extremities. No particular symptoms were reported after the administration of propranolol. After the injection of atropine all subjects developed a dry mouth and blurring of vision. The subjects all experienced an acute hypoglycaemic reaction (‘R) associated with the onset of autonomic and neuroglycopenic symptom, which coincided with the nadir of blood glucose. Sweating was observed at the time of the acute autonomic reaction, except in subjects exposed to pharmacological blockade with atropine. An increase in heart rate was noticed at ‘R,except during the propranolol studies. The typical blood pressure changes of acute hypoglycaemia were observed in the control studies. During a-adrenergic blockade the basal blood pressures were similar to the control group, but after hypoglycaemia a fall in diastolic blood pressure was observed, with lesser falls in systolic and mean arterial pressures, and widening of the pulse pressure. During P-adrenergic blockade, the blood pressure changes after hypoglycaemia in the present study were concurrent increases in systolic, mean and diastolic blood pressures, as described previously by others [6]. During blockade with atropine, the blood pressure changes after hypoglycaemia were similar to changes observed in the
Splenic responses to hypoglycaemia 140
47 1
1
-m
120-
2
100-
v)
c
8 .-21
-
v
.>, a
80-
60-
mV
.Y C
cn" 0'
200-
I
Basal
40-
15
'R'
'R'+ 15 'R'+30'R'+45 'R' + 60 'R' t 75 'R'+90
Basal
+5
+15
'R' 'R'+15'R'+30'R'+45'R'+60'R'+75'R+90
Time (min)
Time (min)
Fig. 1. Changes in blood glucose ( m e a n k s m ) in response to insulin-induced hypoglycemia in the control study (0)and during treatment with phentolamine ( A ), propranolol (0)or atropine (0).Insulin was given immediately after basal sampling, and 'R' denotes the acute autonomic reaction.
Fig. 3. Changes in splenic activity (mean ~ S E M ) in response to insulin-induced hypoglycaemia in the control study (0)and during treatment with phentolamine ( A ) , propranolol (0)or atropine (0). Insulin was given immediately after basal sampling, and 'R' denotes the acute autonomic reaction. No significant differences in blood glucose recovery were observed between the groups treated with either phentolamine o r atropine when compared with the control values.
Splenic activity
Fig. 2. Changes in the splenic radionuclide image in response to insulin-induced hypoglycaemia in one control study. Images were taken after basal sampling ( a ) , at 'R' ( b ) ,at 'R'+15 min ( c )and at 'R' + 60 min (d). control studies. No differences were observed between the study protocols in the mean time taken to reach 'R' after the administration of the bolus of insulin.
In the control studies splenic radioactivity declined suddenly at 'R', and the splenic image completely disappeared during hypoglycaemia in all six subjects (Fig. 2). The disappearance of the splenic image occurred at 'R' in four subjects and at ' R + 15 min in two subjects. The mean splenic activity fell to 16.8 k 10.5% of basal radioactivity at 'R, and to 10.8k6.7% of basal activity at 'R' + 15 min (Fig. 3). T h e acute changes of splenic radioactivity were abolished during pharmacological blockade with phentolamine, but were unaffected by blockade with either propranolol or atropine (Fig. 3). Ultrasonographyof spleen The change in splenic size on ultrasound in one of the control studies is shown in Fig. 4. The mean splenic area fell to a nadir of 6 2 k 6 % of basal value at 'R'+15 min (Fig. 5). DISCUSSION
Blood glucose (Fig. 1) In the control radioisotopic study, the mean blood glucose fell significantly from 4.6 kO.l to 2.1 *0.2 mmol/l at 0 + 15 min, and reached a nadir of 1.O k 0.2 mmol/l at 'R' (IW0.05).In the group treated with propranolol, the blood glucose nadir was similar at 1.0 0.1 mmol/l, but thereafter recovery was slightly slower than the other groups, and the blood glucose was lower at all subsequent times of measurement (one-way analysis of variance, P < 0.05), as described previously by others [7].
*
In humans, acute insulin-induced hypoglycaemia stimulates hypothalamic centres to promote activation of the autonomic nervous system. Hypoglycaemia is one of the most potent stimuli for the release of adrenaline in humans [ 101 and, in conjunction with the concurrent activation of parts of the sympathetic nervous system, is believed to be primarily responsible for the major cardiovascular changes which occur in response to the hypoglycaemia [3]. The effects of hypoglycaemia on the blood flow of many major organs and vascular systems has not
472
B. M. Fisher et al.
Fig. 4. Changes in the ,ultrasonic image of the spleen in response to insulin-induced hypoglycaemia under control conditions in one subject. Images were taken after basal sampling (a),at 'R' (b),at 'R' + 15 min (c)and at 'R' + 6 0 min (d).
received cietailed examination in humans. An increase in the total blood flow to the limbs has been described [I, 111, comprising increments in both skin [ 121 and muscle [13] blood flow. Modest increases in cerebral [14] and total hepatic [ 151 blood flow have been demonstrated in humans. A fall in the overall peripheral vascular resistance occurs during hypoglycaemia, with no change in splanchnic vascular resistance [3].In the present study a sudden and dramatic decline in splenic radioactivity was observed in response to hypoglycaemia. The radioactivity was associated with erythrocytes, which were labelled by the injection of isotope into the peripheral vascular system, but it was not possible in the present study to determine exactly how much of the splenic radioactivity was related to the arterial blood supply and how much was related to erythrocytes stored in the spleen. The liver is of central importance to glucose homoeostasis, being vital for the rapid production of glucose in response to hypoglycaemia, and a rapid reduction in splenic blood flow could promote preferential diversion of blood to the liver by the hepatic artery. A slight increase in hepatic blood flow after hypoglycaemia could accelerate the delivery of substrates to the liver for gluconeogenesis, and the outflow of glucose via the hepatic veins to the systemic circulation. This diversion of blood to the liver by diminishing splenic perfusion may represent a further minor homoeostatic mechanism to promote blood glucose recovery and protect the individual from hypoglycaemia if other glucose counter-regulatory mechanisms were impaired.
In the present study an apparent dissociation was demonstrated between the effects of hypoglycaemia on the splenic radioactivity and on splenic size using different but complementary techniques. Allowing for the severe limitations of ultrasonography, which measured the two-dimensional area of the spleen and not its actual volume, and could include parenchymal tissue in addition to erythrocytes, the profound reduction in splenic radioactivity was associated with a modest reduction in splenic size to about half of the resting volume. A direct contraction of the spleen in response to autonomic neural stimulation is unlikely in humans, as the capsule of the human spleen consists mainly of fibroelastic tissue, and contains only a small muscular component; contraction and distension of the spleen are thought to depend on constriction or relaxation of the splenic vasculature which alter the volume of blood within the organ [16]. This suggests that the primary physiological mechanism after hypoglycaemia was a reduction in the volume of blood entering the spleen with a consequential reduction in splenic size as a secondary effect. Although a comparison with a euglycaemic-hyperinsulinaemic clamp was not performed in the present study, it seems unlikely that insulin had a direct effect on the spleen as a bolus of insulin was given which would be cleared rapidly from the circulation. Any direct effect would be expected to occur in the first 5-10 min after injection, when the direct effects of insulin on cardiac contractility were noted [4], and this was not the finding in the present study. Similarly,
Splenic responses to hypoglycaemia
OJ
, Basal
I
'R'
I
'R'+15 'R'+30 'R'+45 'R'+60 'R'+75 'R'+90
Time (min)
Fig. 5. Changes in splenic activiiy ( 0 ) and& the ultrasonic area of the spleen )(. (means f SEM) in the control study in response to insulin-induced hypoglycaemia. Insulin was given immediately after basal sampling, and ' R denotes the acute autonomic reaction.
it would be unlikely that autonomic blockade would affect the response if this were caused directly by insulin, but a minor direct effect of insulin cannot be completely excluded. Various clinico-pathological conditions have been utilized as experimental models to attempt elucidation of the autonomic mechanisms which modulate changes in the blood flow in different vascular systems in response to hypoglycaemia and other stimuli. These have included studies of patients with a pre-ganglionic sympathectomy caused by traumatic transection of the cervical spinal cord [17], patients with a therapeutic splanchnic sympathectomy [I], patients with bilateral adrenalectomy [ 181 and patients with unilateral sympathectomy for localized vascular disease [ 191. By using pharmacological blocking agents we have previously demonstrated that the effects of hypoglycaemia on the heart are mediated by padrenergic mechanisms [20]. In the present study the changes in the splenic radioactivity after hypoglycaemia have been shown to be mediated by an a-adrenergic mechanism. It is not possible, however, to determine whether this represented a response to circulating adrenaline or to the local release of noradrenaline as a neurotransmitter. The present results in intact humans are inconsistent with the elegant studies of Ayers et al. [21], who examined the splenic vascular resistance and splenic volume in vitro in human spleens which had been removed surgically. In that study the splenic vascular responses to sympathetic neural stimulation and to the infusion of catecholamines, polypeptides and acetylcholine were examined and were compared with the responses of isolated canine spleens. In the human spleen a pronounced increase in splenic vascular resistance occurred in response to sympathetic nerve stimulation which caused vasoconstriction of splenic blood vessels, but the change in splenic volume was minimal. The increase in splenic vascular resistance was simulated by the infusion both of adrenaline and noradrenaline, and the responses to either of three
473
catecholamines and to sympathetic nerve stimulation were abolished completely by a-adrenergic blockade with phenoxybenzamine. By contrast, the increase in splenic vascular resistance in the canine spleen was associated with a marked reduction in splenic volume, indicating a definite interspecies variation in response [2 11. The responses of the intact human spleen in the present study, with a marked reduction in splenic radioactivity and minimal changes in splenic volume, were similar to the responses of the isolated human spleen described previously [21], and supports the role of an a-adrenergic mechanism in producing this effect. The absence of an effect by atropine on splenic blood flow precludes a role for any putative sympathetic cholinergic innervation of the spleen. Similarly, Ayers et al. [21] could not demonstrate any effect of infused acetylcholine on the isolated perfused spleen. Acute hypoglycaemia has previously been associated with changes in peripheral blood cell counts [22]. A n increase in the packed cell volume was observed, with an early rise in the lymphocyte count and a later rise in granulocyte count. Similar increases in the packed cell volume and leucocyte counts were observed in splenectomized subjects, and it was concluded that the peripheral blood cell changes did not involve splenic contraction, although no imaging of the spleen was performed during that study [22]. We recently demonstrated that during hypoglycaemia with a-adrenergic blockade the increases in packed cell volume and lymphocyte counts after hypoglycaemia were abolished [23]. The results of the present study indicate that the responses of these cells and of the spleen to hypoglcycaemia may share similar autonomic mechanisms. Alternatively, in normal subjects contraction of the spleen may indeed be contributing to the rise in blood cells which occurs after hypoglycaemia.
ACKNOWLEDGMENTS
Sincere thanks are given to Ms E. Henderson and Mr J. Wilson for expert technical assistance with the radionuclide studies, and to the staff of the Department of Pathological Biochemistry, Gartnavel General Hospital, Glasgow, for practical assistance with the studies.
REFERENCES 1. Cannon, W.B., Mclver, M.A. & Bliss, S.W. Studies on the conditions of activity in endocrine glands. XIII. A sympathetic and adrenal mechanism for mobilizing sugar in hypoglycemia. Am. J. Physiol. 1924; 69,46-66. 2. French, E.B. & Kilpatrick, R. The role of adrenaline in hypoglycaemia reactions in man. Clin. Sci. 1955; 14, 639-51. 3. Hilsted, J., Bonde-Petersen, F., Nmgaard, M.-B. et al. Haemodynamic changes in insulin-induced hypoglycaemia in normal man. Diabetologia 1984; 26,328-32. 4. Fisher, B.M., Gillen, G., Dargie, H.J., Inglis, G.C. & Frier, B.M. The effects of insulin-induced hypoglycaemia on cardiovascular function in normal man: studies using radionuclide ventriculography. Diabetologia 1988; 30, 841-45.
474
B. M. Fisher et al.
5. Clarke, W.L., Santiago, J.V., Thomas, L., Ben-Galim, E., Haymond, M.W. & Cryer, P.E. Adrenergic mechanisms in recovery from hypoglycemia in man: adrenergic blockade. Am. J. Physiol. 1979; 236, E147-52. 6. Davidson. N. McD., Corrall, R.J.M., Shaw,T.R.D. & French, E.B. Observations in man of hypoglycaemia during selective and non-selective beta-blockade. Scot. Med. J. 1977; 22, 69-72. 7. Corrall, R.J.M., Frier, B.M., Davidson, N.McD. & French, E.B. Hormonal and substrate responses during recovery from hypoglyceamia in man during betal-selective and nonselective beta-adrenergic blockade. Eur. J. Clin. Invest. I98 I ; 1I , 279-83. 8. Frier, B.M., Corrall, R.J.M., O'Brien, I.A.D., Lewin, I.G., Hay, I.D. & Roland, J. Hypoglycemia during adrenergic beta-blockade: evidence against mediation via a deficiency of lactate for gluconeogenesis. Metabolism 1985; 34, 1039-43. 9. Corrall, R.J.M., Frier, B.M., Davidson, N.McD., Hopkins, W.M. & French, E.B. Cholinergic manifestations of the acute autonomic reaction to hypoglycaemia in man. Clin. Sci. 1983; 64,49-53. 10. Cryer, P.E. Physiology and pathophysiology of the human symprtthoadrenal neuroendocrine system. N. Engl. J. Med. 1980; 303,436-44. 1 I . Abramson, D.L., Schkloven, N., Margolis, M.N. & Mirsky, LA. Influence of massive doses of insulin on peripheral blood flow in man. Am. J. Physiol. 1939; 128, 124-32. 12. Allwood, M.J., Birchall, 1. & Staffurth, J.S. Circulatory changes in the forearm during insulin hypoglycaemia studied by regional ?"a clearance and by plethysmography. J. Physiol. (London) 1958; 143,332-42. 13. Allwood, M.J., Hensel, H. & Papenberg, J. Muscle and skin blood flow in the human forearm during insulin hypoglycaemia. J. Physiol. (London) 1959; 147,269-73.
14. Neil, H.A.W., Gale, E.A.M., Hamilton, S.J.C., LopezEspinoza, I., Kaura, R. & McCarthy, S.T. Cerebral blood flow increases during insulin-induced hypoglycaemia in Type 1 (insulin-dependent) diabetic patients and control subjects. Diabetologia 1987; 30,305-9. 15. Bearn, A.G., Billing, B.H. & Sherlock, S. The response of the liver to insulin in normal subjects and diabetes mellitus: hepatic vein catheterization studies. Clin. Sci. 1952; 1 1 , 151-65. 16. Warwick, R. & Williams, P.L. In: Warwick, R. & Williams, P.L., eds., Gray's anatomy, 35th edn. Edinburgh: Longmans, 1973: 71 8-22. 17. Mathias, C.J., Frankel, H.L., Turner, R.C. & Christensen, N.J. Physiological responses to insulin hypoglycaemia in spinal man. Paraplegia 1979; 17,319-26. 18. Ginsburg, J. & Paton, A. Effects of insulin after adrenalectomy. Lancet 1956; ii, 491-4. 19. Middleton, W.G. & French, E.B. Studies of the peripheral vasodilator response to acute insulin-induced hypoglycaemia in man. Clin. Sci. Mol. Med. 1974; 47,46 1-70. 20. Fisher, B.M., Gillen, G., Hepburn, D.A., Dargie, H.J. & Frier, B.M. Cardiac responses to acute insulin-induced hypoglycemia in humans. Am. J. Physiol. 1990 (In press). 21. Ayers, A.B., Davies, B.N. & Withrington, P.G. Responses of the isolated, perfused human spleen to sympathetic nerve stimulation, catecholamines and polypeptides. Br. J. Pharmacol. 1972; 44, 17-30. 22. Frier, B.M., Corrall, R.J.M., Davidson, N.McD., Webber, R.G., Dewar, A. & French, E.B. Peripheral blood cell changes in response to acute hypoglycaemia in man. Eur. J. Clin. Invest. 1983; 13,33-9. 23. Fisher, B.M., Hepburn, D.A., Smith, J.G. & Frier, B.M. The effect of alpha-adrenergic blockade on responses of peripheral blood cells to acute insulin-induced hypoglycaemia in humans. Eur. J. Clin. Invest. 1990 (In press).
469
Splenic responses to acute insulin-induced hypoglycaemia in humans B. M. FISHER', G. GILLEN*, D. A. HEPBUR", H. J. DARGIE3, E. BARNETT4 AND B. M. FRIER' Departments of I Diabetes, 'Medical Physics, Tardiology and 4Radiology,Western Infirmary/Gartnavel General Hospital, Glasgow, Scotland, U.K.
(Received 27 October 1989/12 January 1990; accepted 18 January 1990) SUMMARY 1. The effects of acute hypoglycaemia on the spleen were examined in normal humans using radioisotopic techniques, complemented by ultrasonic examination of the spleen. Hypoglycaemia had a modest effect on splenic area, measured by ultrasonography, which declined to 6 2 f 6% ( r n e a n f ~of~ the ~ ) basal value after the onset of the acute hypoglycaemic reaction. 2. Hypoglycaemia had a pronounced effect on the splenic radioactivity, which decreased significantly to a mean of 10 f 7% of basal radioactivity at 15 min after the onset of hypoglycaemia. The splenic image completely disappeared at some time after hypoglycaemia in all subjects. 3. The reduction of splenic radioactivity was abolished during non-selective a-adrenergic blockade with phentolamine, but was unaffected by p-adrenergic blockade with propranolol, or cholinergic blockade with atropine, which suggests that the response of vessels perfusing the spleen is mediated by a-adrenoceptors. Key words: hypoglycaemia, phentolamine, spleen, ultrasound. INTRODUCTION
Acute hypoglycaemia in humans provokes stimulation of the autonomic nervous system, including sympathoadrenal activation and the release of catecholamines [ 11. The effects of acute hypoglycaemia on heart rate and blood pressure are well defined: hypoglycaemia causes an increase in heart rate, a rise in systolic and a fall in diastolic blood pressure, with no change in mean arterial blood pressure [2]. Cardiac function in response to hypoglycaemia has been examined using non-invasive techniques: cardiac output [3] and myocardial contractility, Correspondence: Dr B. M. Fisher, Wards 4/5, Royal Infirmary, Glasgow G4 OSF, Scotland,U.K.
estimated by left ventricular ejection fraction [4], both increase significantly. These profound haemodynamic changes are thought to occur secondary to sympathoadrenal activation [3]. Knowledge of the physiological effects of hypoglycaemia on other vascular systems and perfusion of organs is, however, fragmentary. T h e aim of the present study was to examine the mechanisms underlying the responses of the spleen to acute hypoglycaemia. The role of autonomic mechanisms was assessed by pharmacological blockade: (a) with propranolol, pro.ducing non-selective blockade of p-adrenoreceptors, (b) with phentolamine, producing non-selective blockade of a-adrenoreceptors, and (c) with atropine, producing blockade of parasympathetic and sympathetic cholinergic neurotransmission. METHODS Subjects
The study was approved by the local medical ethical advisory committee and informed consent was obtained from all subjects. Each study group comprised six normal healthy male subjects, aged 21-30 years, all of whom had a normal body mass index, and none of whom was taking medications. A total of 30 experimental studies were performed in 24 subjects, with the six subjects in the control group being studied twice. As the changes in splenic radioactivity were greater than the changes in splenic area, the role of autonomic mechanisms was assessed by pharmacological blockade with radioisotopic imaging, and changes in splenic area were not measured during pharmacological blockade to avoid unnecessary exposure to hypoglycaemia. Protocol
T h e studies were divided into four groups as follows: (1) Control studies (n= 6): acute hypoglycaemia with no pharmacological blockade. The cardiac effects of
470
B. M. Fisher et al.
acute hypoglycaemia on the left ventricular ejection fraction in this control group have been published previously [4]. Each control subject was studied twice, on two separate occasions. On the first day, radioisotopic imaging of the spleen was performed. On the second day, ultrasonic imaging of the splenic area was performed with no pharmacological blockade. (2) a-Adrenergic blockade (n= 6): acute hypoglycaemia with radioisotopic studies during a-adrenergic blockade with 5 mg of phentolamine administered as an intravenous bolus injection, followed by 30 mg of phentolamine/h as an intravenous infusion [5]. (3) B-Adrenergic blockade (n= 6 ) : acute hypoglycaemia with radioisotopic studies during non-selective /3-adrenergic blockade with 10 mg of propranolol administered as an intravenous bolus injection, followed by 3 mg of propanolol/h as an intravenous infusion [6-81. (4) Cholinergic-blockade (n= 6): acute hypoglycaemia with radioisotopic studies during cholinergic blockade with 1.2 mg of atropine administered as an intravenous injection 30 min before insulin, and followed by 0.6 mg of atropine 1 h after the initial dose [9]. Subjects were fasted overnight and studied after remaining in a recumbent position for 1 h. The consumption of alcohol, tobacco, tea and coffee was avoided for 12 h before each study. At 08.30 hours, basal venous blood samples were withdrawn from an indwelling Teflon cannula which had been inserted into an antecubital vein. Pharmacological blockade was commenced after basal blood sampling, and intravenous soluble insulin (Human Actrapid; Novo, Basingstoke, Hants, U.K.) was administered as a single bolus at 09.00 hours in a standard dose of 0.15 unit/kg body weight. Serial estimation of haemodynamic variables and blood glucose were made every 5 min until the onset of the acute autonomic reaction (‘R), which was determined by the sudden increase in heart rate and the development of autonomic symptoms, including sweating, and coincided with the nadir of blood glucose. To eliminate the individual variability in the time taken to reach the nadir of blood glucose after the intravenous injection of insulin, subsequent blood sampling and cardiovascular measurements were timed from ‘R. Sampling was continued until ‘ R + 90 min when the study was terminated. The heart rate was measured continuously using precordial electrodes with the heart rate displayed on an oscilloscope (Life Trace 12; Albury Instruments Ltd, London). Blood pressure was measured using a mercury sphygmomanometer by a single observer with diastolic pressure being measured by Korotkoff’s 5th sound. Blood glucose was measured with the Cobas Bio Centrifugal analyser (Roche Diagnostica, Basel, Switzerland) using a hexokinase method. Splenic radionuclide imaging was performed using the multiple-gated technique with yymT (800 MBq). Erythrocytes were labelled iri vivo by the injection of isotope 30 min after the intravenous injection of stannous pyrophosphate. Subjects were studied supine using the Siemens LEM mobile gamma camera (Siemens Ltd, Des Plaines, IL, U.S.A.) in the left anterior oblique position.
Data were obtained at a rate of 20 frames/cycle with acquisition of data during 5 min. To determine the splenic radioactivity, a region of interest was drawn around the spleen. Splenic radioactivity was determined by a single observer who was unaware of the timing of the measurements. The basal splenic radioactivity was designated as 10Oo/~,and subsequent measurements were compared with the basal value. The intra-observer coefficient of variation for this technique is 6%. Ultrasonic imaging of the spleen was performed using a ATL real-time B Scanner using a 3.5 mHz medium focus probe (Squibb Medical Systems Ltd, Hemel Hempstead, Herts, U.K.). Longitudinal scans were made in the left axilla with the subject lying supine, using a fixed depth of focus. The splenic perimeter was outlined using electronic cursors and the maximum area of the longitudinal section was calculated. The intra-observer coefficient of variation for this technique is 8%. All ultrasonic measurements were made by a single observer (E.B.), who was unaware of the timing of the measurements with respect to hypoglycaemia. Statistics Statistical analyses were performed using the Statview 5 12 package (Brainpower Inc, Calabasas, CA, U.S.A.) on an Apple Macintosh Plus Computer (Apple Computer UK Ltd, Hemel Hempstead, H e m , U.K.). Initial comparisons were made using analysis of variance with post hoc comparisons of means using Student’s t-test for unpaired data. The results are presented as means k SEM. P values less than 0.05 were considered significant. RESULTS
During phentolamine infusion subjects developed nasal congestion and flushing of the extremities. No particular symptoms were reported after the administration of propranolol. After the injection of atropine all subjects developed a dry mouth and blurring of vision. The subjects all experienced an acute hypoglycaemic reaction (‘R) associated with the onset of autonomic and neuroglycopenic symptom, which coincided with the nadir of blood glucose. Sweating was observed at the time of the acute autonomic reaction, except in subjects exposed to pharmacological blockade with atropine. An increase in heart rate was noticed at ‘R,except during the propranolol studies. The typical blood pressure changes of acute hypoglycaemia were observed in the control studies. During a-adrenergic blockade the basal blood pressures were similar to the control group, but after hypoglycaemia a fall in diastolic blood pressure was observed, with lesser falls in systolic and mean arterial pressures, and widening of the pulse pressure. During P-adrenergic blockade, the blood pressure changes after hypoglycaemia in the present study were concurrent increases in systolic, mean and diastolic blood pressures, as described previously by others [6]. During blockade with atropine, the blood pressure changes after hypoglycaemia were similar to changes observed in the
Splenic responses to hypoglycaemia 140
47 1
1
-m
120-
2
100-
v)
c
8 .-21
-
v
.>, a
80-
60-
mV
.Y C
cn" 0'
200-
I
Basal
40-
15
'R'
'R'+ 15 'R'+30'R'+45 'R' + 60 'R' t 75 'R'+90
Basal
+5
+15
'R' 'R'+15'R'+30'R'+45'R'+60'R'+75'R+90
Time (min)
Time (min)
Fig. 1. Changes in blood glucose ( m e a n k s m ) in response to insulin-induced hypoglycemia in the control study (0)and during treatment with phentolamine ( A ), propranolol (0)or atropine (0).Insulin was given immediately after basal sampling, and 'R' denotes the acute autonomic reaction.
Fig. 3. Changes in splenic activity (mean ~ S E M ) in response to insulin-induced hypoglycaemia in the control study (0)and during treatment with phentolamine ( A ) , propranolol (0)or atropine (0). Insulin was given immediately after basal sampling, and 'R' denotes the acute autonomic reaction. No significant differences in blood glucose recovery were observed between the groups treated with either phentolamine o r atropine when compared with the control values.
Splenic activity
Fig. 2. Changes in the splenic radionuclide image in response to insulin-induced hypoglycaemia in one control study. Images were taken after basal sampling ( a ) , at 'R' ( b ) ,at 'R'+15 min ( c )and at 'R' + 60 min (d). control studies. No differences were observed between the study protocols in the mean time taken to reach 'R' after the administration of the bolus of insulin.
In the control studies splenic radioactivity declined suddenly at 'R', and the splenic image completely disappeared during hypoglycaemia in all six subjects (Fig. 2). The disappearance of the splenic image occurred at 'R' in four subjects and at ' R + 15 min in two subjects. The mean splenic activity fell to 16.8 k 10.5% of basal radioactivity at 'R, and to 10.8k6.7% of basal activity at 'R' + 15 min (Fig. 3). T h e acute changes of splenic radioactivity were abolished during pharmacological blockade with phentolamine, but were unaffected by blockade with either propranolol or atropine (Fig. 3). Ultrasonographyof spleen The change in splenic size on ultrasound in one of the control studies is shown in Fig. 4. The mean splenic area fell to a nadir of 6 2 k 6 % of basal value at 'R'+15 min (Fig. 5). DISCUSSION
Blood glucose (Fig. 1) In the control radioisotopic study, the mean blood glucose fell significantly from 4.6 kO.l to 2.1 *0.2 mmol/l at 0 + 15 min, and reached a nadir of 1.O k 0.2 mmol/l at 'R' (IW0.05).In the group treated with propranolol, the blood glucose nadir was similar at 1.0 0.1 mmol/l, but thereafter recovery was slightly slower than the other groups, and the blood glucose was lower at all subsequent times of measurement (one-way analysis of variance, P < 0.05), as described previously by others [7].
*
In humans, acute insulin-induced hypoglycaemia stimulates hypothalamic centres to promote activation of the autonomic nervous system. Hypoglycaemia is one of the most potent stimuli for the release of adrenaline in humans [ 101 and, in conjunction with the concurrent activation of parts of the sympathetic nervous system, is believed to be primarily responsible for the major cardiovascular changes which occur in response to the hypoglycaemia [3]. The effects of hypoglycaemia on the blood flow of many major organs and vascular systems has not
472
B. M. Fisher et al.
Fig. 4. Changes in the ,ultrasonic image of the spleen in response to insulin-induced hypoglycaemia under control conditions in one subject. Images were taken after basal sampling (a),at 'R' (b),at 'R' + 15 min (c)and at 'R' + 6 0 min (d).
received cietailed examination in humans. An increase in the total blood flow to the limbs has been described [I, 111, comprising increments in both skin [ 121 and muscle [13] blood flow. Modest increases in cerebral [14] and total hepatic [ 151 blood flow have been demonstrated in humans. A fall in the overall peripheral vascular resistance occurs during hypoglycaemia, with no change in splanchnic vascular resistance [3].In the present study a sudden and dramatic decline in splenic radioactivity was observed in response to hypoglycaemia. The radioactivity was associated with erythrocytes, which were labelled by the injection of isotope into the peripheral vascular system, but it was not possible in the present study to determine exactly how much of the splenic radioactivity was related to the arterial blood supply and how much was related to erythrocytes stored in the spleen. The liver is of central importance to glucose homoeostasis, being vital for the rapid production of glucose in response to hypoglycaemia, and a rapid reduction in splenic blood flow could promote preferential diversion of blood to the liver by the hepatic artery. A slight increase in hepatic blood flow after hypoglycaemia could accelerate the delivery of substrates to the liver for gluconeogenesis, and the outflow of glucose via the hepatic veins to the systemic circulation. This diversion of blood to the liver by diminishing splenic perfusion may represent a further minor homoeostatic mechanism to promote blood glucose recovery and protect the individual from hypoglycaemia if other glucose counter-regulatory mechanisms were impaired.
In the present study an apparent dissociation was demonstrated between the effects of hypoglycaemia on the splenic radioactivity and on splenic size using different but complementary techniques. Allowing for the severe limitations of ultrasonography, which measured the two-dimensional area of the spleen and not its actual volume, and could include parenchymal tissue in addition to erythrocytes, the profound reduction in splenic radioactivity was associated with a modest reduction in splenic size to about half of the resting volume. A direct contraction of the spleen in response to autonomic neural stimulation is unlikely in humans, as the capsule of the human spleen consists mainly of fibroelastic tissue, and contains only a small muscular component; contraction and distension of the spleen are thought to depend on constriction or relaxation of the splenic vasculature which alter the volume of blood within the organ [16]. This suggests that the primary physiological mechanism after hypoglycaemia was a reduction in the volume of blood entering the spleen with a consequential reduction in splenic size as a secondary effect. Although a comparison with a euglycaemic-hyperinsulinaemic clamp was not performed in the present study, it seems unlikely that insulin had a direct effect on the spleen as a bolus of insulin was given which would be cleared rapidly from the circulation. Any direct effect would be expected to occur in the first 5-10 min after injection, when the direct effects of insulin on cardiac contractility were noted [4], and this was not the finding in the present study. Similarly,
Splenic responses to hypoglycaemia
OJ
, Basal
I
'R'
I
'R'+15 'R'+30 'R'+45 'R'+60 'R'+75 'R'+90
Time (min)
Fig. 5. Changes in splenic activiiy ( 0 ) and& the ultrasonic area of the spleen )(. (means f SEM) in the control study in response to insulin-induced hypoglycaemia. Insulin was given immediately after basal sampling, and ' R denotes the acute autonomic reaction.
it would be unlikely that autonomic blockade would affect the response if this were caused directly by insulin, but a minor direct effect of insulin cannot be completely excluded. Various clinico-pathological conditions have been utilized as experimental models to attempt elucidation of the autonomic mechanisms which modulate changes in the blood flow in different vascular systems in response to hypoglycaemia and other stimuli. These have included studies of patients with a pre-ganglionic sympathectomy caused by traumatic transection of the cervical spinal cord [17], patients with a therapeutic splanchnic sympathectomy [I], patients with bilateral adrenalectomy [ 181 and patients with unilateral sympathectomy for localized vascular disease [ 191. By using pharmacological blocking agents we have previously demonstrated that the effects of hypoglycaemia on the heart are mediated by padrenergic mechanisms [20]. In the present study the changes in the splenic radioactivity after hypoglycaemia have been shown to be mediated by an a-adrenergic mechanism. It is not possible, however, to determine whether this represented a response to circulating adrenaline or to the local release of noradrenaline as a neurotransmitter. The present results in intact humans are inconsistent with the elegant studies of Ayers et al. [21], who examined the splenic vascular resistance and splenic volume in vitro in human spleens which had been removed surgically. In that study the splenic vascular responses to sympathetic neural stimulation and to the infusion of catecholamines, polypeptides and acetylcholine were examined and were compared with the responses of isolated canine spleens. In the human spleen a pronounced increase in splenic vascular resistance occurred in response to sympathetic nerve stimulation which caused vasoconstriction of splenic blood vessels, but the change in splenic volume was minimal. The increase in splenic vascular resistance was simulated by the infusion both of adrenaline and noradrenaline, and the responses to either of three
473
catecholamines and to sympathetic nerve stimulation were abolished completely by a-adrenergic blockade with phenoxybenzamine. By contrast, the increase in splenic vascular resistance in the canine spleen was associated with a marked reduction in splenic volume, indicating a definite interspecies variation in response [2 11. The responses of the intact human spleen in the present study, with a marked reduction in splenic radioactivity and minimal changes in splenic volume, were similar to the responses of the isolated human spleen described previously [21], and supports the role of an a-adrenergic mechanism in producing this effect. The absence of an effect by atropine on splenic blood flow precludes a role for any putative sympathetic cholinergic innervation of the spleen. Similarly, Ayers et al. [21] could not demonstrate any effect of infused acetylcholine on the isolated perfused spleen. Acute hypoglycaemia has previously been associated with changes in peripheral blood cell counts [22]. A n increase in the packed cell volume was observed, with an early rise in the lymphocyte count and a later rise in granulocyte count. Similar increases in the packed cell volume and leucocyte counts were observed in splenectomized subjects, and it was concluded that the peripheral blood cell changes did not involve splenic contraction, although no imaging of the spleen was performed during that study [22]. We recently demonstrated that during hypoglycaemia with a-adrenergic blockade the increases in packed cell volume and lymphocyte counts after hypoglycaemia were abolished [23]. The results of the present study indicate that the responses of these cells and of the spleen to hypoglcycaemia may share similar autonomic mechanisms. Alternatively, in normal subjects contraction of the spleen may indeed be contributing to the rise in blood cells which occurs after hypoglycaemia.
ACKNOWLEDGMENTS
Sincere thanks are given to Ms E. Henderson and Mr J. Wilson for expert technical assistance with the radionuclide studies, and to the staff of the Department of Pathological Biochemistry, Gartnavel General Hospital, Glasgow, for practical assistance with the studies.
REFERENCES 1. Cannon, W.B., Mclver, M.A. & Bliss, S.W. Studies on the conditions of activity in endocrine glands. XIII. A sympathetic and adrenal mechanism for mobilizing sugar in hypoglycemia. Am. J. Physiol. 1924; 69,46-66. 2. French, E.B. & Kilpatrick, R. The role of adrenaline in hypoglycaemia reactions in man. Clin. Sci. 1955; 14, 639-51. 3. Hilsted, J., Bonde-Petersen, F., Nmgaard, M.-B. et al. Haemodynamic changes in insulin-induced hypoglycaemia in normal man. Diabetologia 1984; 26,328-32. 4. Fisher, B.M., Gillen, G., Dargie, H.J., Inglis, G.C. & Frier, B.M. The effects of insulin-induced hypoglycaemia on cardiovascular function in normal man: studies using radionuclide ventriculography. Diabetologia 1988; 30, 841-45.
474
B. M. Fisher et al.
5. Clarke, W.L., Santiago, J.V., Thomas, L., Ben-Galim, E., Haymond, M.W. & Cryer, P.E. Adrenergic mechanisms in recovery from hypoglycemia in man: adrenergic blockade. Am. J. Physiol. 1979; 236, E147-52. 6. Davidson. N. McD., Corrall, R.J.M., Shaw,T.R.D. & French, E.B. Observations in man of hypoglycaemia during selective and non-selective beta-blockade. Scot. Med. J. 1977; 22, 69-72. 7. Corrall, R.J.M., Frier, B.M., Davidson, N.McD. & French, E.B. Hormonal and substrate responses during recovery from hypoglyceamia in man during betal-selective and nonselective beta-adrenergic blockade. Eur. J. Clin. Invest. I98 I ; 1I , 279-83. 8. Frier, B.M., Corrall, R.J.M., O'Brien, I.A.D., Lewin, I.G., Hay, I.D. & Roland, J. Hypoglycemia during adrenergic beta-blockade: evidence against mediation via a deficiency of lactate for gluconeogenesis. Metabolism 1985; 34, 1039-43. 9. Corrall, R.J.M., Frier, B.M., Davidson, N.McD., Hopkins, W.M. & French, E.B. Cholinergic manifestations of the acute autonomic reaction to hypoglycaemia in man. Clin. Sci. 1983; 64,49-53. 10. Cryer, P.E. Physiology and pathophysiology of the human symprtthoadrenal neuroendocrine system. N. Engl. J. Med. 1980; 303,436-44. 1 I . Abramson, D.L., Schkloven, N., Margolis, M.N. & Mirsky, LA. Influence of massive doses of insulin on peripheral blood flow in man. Am. J. Physiol. 1939; 128, 124-32. 12. Allwood, M.J., Birchall, 1. & Staffurth, J.S. Circulatory changes in the forearm during insulin hypoglycaemia studied by regional ?"a clearance and by plethysmography. J. Physiol. (London) 1958; 143,332-42. 13. Allwood, M.J., Hensel, H. & Papenberg, J. Muscle and skin blood flow in the human forearm during insulin hypoglycaemia. J. Physiol. (London) 1959; 147,269-73.
14. Neil, H.A.W., Gale, E.A.M., Hamilton, S.J.C., LopezEspinoza, I., Kaura, R. & McCarthy, S.T. Cerebral blood flow increases during insulin-induced hypoglycaemia in Type 1 (insulin-dependent) diabetic patients and control subjects. Diabetologia 1987; 30,305-9. 15. Bearn, A.G., Billing, B.H. & Sherlock, S. The response of the liver to insulin in normal subjects and diabetes mellitus: hepatic vein catheterization studies. Clin. Sci. 1952; 1 1 , 151-65. 16. Warwick, R. & Williams, P.L. In: Warwick, R. & Williams, P.L., eds., Gray's anatomy, 35th edn. Edinburgh: Longmans, 1973: 71 8-22. 17. Mathias, C.J., Frankel, H.L., Turner, R.C. & Christensen, N.J. Physiological responses to insulin hypoglycaemia in spinal man. Paraplegia 1979; 17,319-26. 18. Ginsburg, J. & Paton, A. Effects of insulin after adrenalectomy. Lancet 1956; ii, 491-4. 19. Middleton, W.G. & French, E.B. Studies of the peripheral vasodilator response to acute insulin-induced hypoglycaemia in man. Clin. Sci. Mol. Med. 1974; 47,46 1-70. 20. Fisher, B.M., Gillen, G., Hepburn, D.A., Dargie, H.J. & Frier, B.M. Cardiac responses to acute insulin-induced hypoglycemia in humans. Am. J. Physiol. 1990 (In press). 21. Ayers, A.B., Davies, B.N. & Withrington, P.G. Responses of the isolated, perfused human spleen to sympathetic nerve stimulation, catecholamines and polypeptides. Br. J. Pharmacol. 1972; 44, 17-30. 22. Frier, B.M., Corrall, R.J.M., Davidson, N.McD., Webber, R.G., Dewar, A. & French, E.B. Peripheral blood cell changes in response to acute hypoglycaemia in man. Eur. J. Clin. Invest. 1983; 13,33-9. 23. Fisher, B.M., Hepburn, D.A., Smith, J.G. & Frier, B.M. The effect of alpha-adrenergic blockade on responses of peripheral blood cells to acute insulin-induced hypoglycaemia in humans. Eur. J. Clin. Invest. 1990 (In press).
