- Email: [email protected]
FIBRINOLYTIC AND DEFIBRINATING EFFECT OF PHENFORMIN PLUS ETHYLŒSTRENOL IN VIVO
910
prevalent in man around the beginning of this century. The findings reported by Kasel et al. (1965) and Masurel and Mulder (1966), the antibody response to vaccination with A/Equi2/63 (Davenport et al. 1967, Masurel 1968a), and the results reported here provide conclusive evidence of an antigenic relationship between this virus and the A2 viruses. The relationship between Hong Kong virus and the A/Equi2 viruses is also supported by other work (Coleman et al. 1968, Davenport 1968, Masurel 1968b). The increase in antibody titres against A/Equi2 presumably due to infection with A2 strains was also seen in sera collected in 1958 and 1964 (Davenport et al. 1967). The distribution of antibody against A/Equi2/Richelieu/63 in sera collected in 1967 fits well with its antigenic relationship to the Hong Kong virus; this antibody pattern and that found in 1957 also confirm the suggestion that a virus related to both of these strains was circulating between 1890 and 1918, when persons now aged fifty-or morealive. The absence of antibody against the A/Turkey/ Massachusetts/65 strain in the human sera collected in 1956-57 suggests that this strain is not antigenically related to influenza strains which circulated in man in the Netherlands before 1957. It would be unwise to conclude that there is an exact correspondence between the antibodies found in sera and the timing and antigens of influenza epidemics in the past, but it seems reasonable to conclude that the 1957 influenza pandemic may have been similar to the pandemic of " 1889-90, and that the virus of the " Hong Kong 1968 influenza epidemic may be closely related to the virus which followed the " A2 pandemic " of the turn of the century. I thank A. J. Baars, E. M. Degger, Henny Frankema, P. Kuyt, J. Magdelijns, and J. J. Vingerling for technical assistance. were
REFERENCES M. W. Coleman, T., Dowdle, R., Pereira, H. G., Schild, G. C., Chang, W. K. (1968) Lancet, ii, 1384. Davenport, F. M. (1968) Personal communication. Hennessy, A. V., Minuse, E. (1967) J. exp. Med. 126, 1049. Francis, T., Jr. (1960) Proc. Am. Phil. 104, 572. Kasel, J. A., Alford, R. H., Knight, V. A., Wadell, G. H., Sigel, M. M. (1965) Ann. intern. Med. 62, 1308. Masurel, N. (1967) Ned. Tijdschr. Geneesk. 111, 1245. (1968a) Nature, Lond. 218, 100. (1968b) Ned. Tijdschr. Geneesk. 112, 1930. Mulder, J. (1962) Verh. Inst. prev. Geneesk. no. 52. — — (1966) Bull. Wld Hlth Org. 34, 885. Minuse, E., McQueen, J. L., Davenport, F. M., Francis, T., Jr. (1965) J. Immun. 94, 563. Mulder, J., Masurel, N. (1958) Lancet, i, 810. Schild, G. C., Stuart-Harris, C. H. (1965) J. Hyg., Camb. 63, 479. van der Veen, J., Mulder, J. (1950) Verh. Inst. prev. Geneesk. no. 6.
FIBRINOLYTIC AND DEFIBRINATING EFFECT OF PHENFORMIN PLUS ETHYLŒSTRENOL IN VIVO G. R. FEARNLEY
J.
R. CHAKRABARTI F. EVANS
FROM THE GLOUCESTERSHIRE ROYAL HOSPITAL, GLOUCESTER
The effects of
phenformin plus ethylfibrin-degradation products (F.D.P.) in serum, fibrinolytic activity, and plasmafibrinogen were investigated in ten patients with occlusive vascular disease, six patients with rheumatoid arthritis, and two healthy people. All showed an increase of F.D.P. during treatment, which greatly exceeded their pretreatSummary œstrenol
on
levels and the upper limit of normal found in eighteen healthy males. Phenformin plus ethylœstrenol increased fibrinolytic activity, and removal of fibrin in vivo was a consequence of this; but the possibility that the drugs were in some way directly influencing fibrinogenfibrin turnover has not been excluded. Increases of F.D.P. in the healthy volunteers and in some of the arteriosclerotic patients were temporary, and this may be ascribed to successful defibrination rather than to resistance since the increases were sustained in the patients with rheumatoid arthritis, a condition accompanied by continuing deposition of fibrin. The findings provide the first direct evidence that pharmacological enhancement of fibrinolysis as measured in vitro is accompanied by increased fibrinolysis in vivo; they also establish that the dilute-blood-clot lysis-time reflects the behaviour of a functioning fibrinolytic system in the body; and are consonant with a dynamic equilibrium between fibrinolysis and coagulation, imbalance of which may be of pathogenic importance. Phenformin plus ethylœstrenol seems to merit trial as a prophylactic measure in survivors of vascular occlusions and in patients after vascular surgery and the insertion of cardiac prostheses. ment
Introduction
-
-
-
-
"... educational considerations ... often lead good parents, who would like more children, to limit their families ... In the United Kingdom there is still a tradition of independent and often expensive private schooling, which in many instances, though in fewer perhaps than the parents realise, gives a better education than that provided in the state schools. The development of comprehensive education is, for a time at least, likely to accentuate these differences between the state and private education ... it is being found most difficult... to provide adequate sixth form education in comprehensive schools ... The eugenist must therefore welcome the disappearance of the good independent schools. They provide a temptation to good parents, who value education, to limit the size of their families. The number of parents influenced by such considerations is small but they probably include some genetically valuable ones."-C. 0. CARTER. Eugenics and Human Heredity: Essex Hall Lecture, April 14, 1969.
THERAPEUTIC defibrination with ’Arvin’ has given encouraging results in patients with acute vascular occlusions (Bell et al. 1968, Gilles et al. 1968, Sharp et al. 1968) but since arvin has to be injected and when given intramuscularly evokes resistance (Pitney et al. 1969) it is unlikely to be suitable for long-term prophylaxis. We have found that phenformin plus ethyloestrenol given twice daily by mouth produces a sustained increase in blood fibrinolytic activity, as judged by reduction of both the dilute-blood-clot and euglobulin lysis-times, and reduces plasma-fibrinogen and serum-cholesterol levels in arteriopathic patients (Fearnley, Chakrabarti, and Hocking 1967). This combination of drugs reduced platelet stickiness to glass in some 75% of patients (Chakrabarti and Fearnley 1967), and this effect was maintained with continued treatment (Fearnley and Chakrabarti 1968). Theoretically,
phenformin plus ethyloestrenol, by enhancing fibrinolysis, might be expected to exercise a defibrinating effect in patients with occlusive vascular disease and hence to merit trial as a prophylactic measure in survivors of vascular accidents. For spontaneous fibrinolysis to be demonstrated, blood has to be manipulated, either by dilution (Fearnley et al. 1957) or by precipitation of the euglobulin fraction (von Kaulla 1963), and so there is no guarantee that what is observed in vitro reflects events in vivo. Indeed this has been the greatest bar to acceptance that natural fibrinolytic
911 was allowed to clot at 37°C for 4 hours. The resultant serum was treated with 10 units of topical thrombin (Parke-Davis) to ensure that all the fibrinogen was clotted and after centrifugation was stored at -15°C. On the day before the test the serum was thawed at room temperature and then absorbed with sheep red cells for 24 hours at 4°C to remove non-specific antibodies. Tests were done once or twice weekly on 12-25 batches of sera. After being set up the tests were kept at 4°C for 4 hours before being read. The end-point for the sera being tested and for the standard was taken as the highest dilution to show complete inhibition of haemagglutination. The tests were read independently by the three of us; we did not know the source of the sera at the time. Agreed end-points were obtained in 60% of tests; in the remainder when two of the three readings agreed this was taken as the end-point; on the very few occasions when all three readings differed the test was set up again. Occasionally, incomplete absorption of non-specific antibodies made reading of the test difficult or impossible; when this happened the test was repeated after the serum had been further absorbed with sheep red blood-cells. Results are expressed as multiples of 1 25 ug. of F.D.P. per ml. of serum.
(aprotinin) were added
Results Normal Range
Of
eighteen healthy males aged 22-53 the F.D.P. level 1-25 ug. per ml. in three, 2-5 ug. per ml. in nine, and 5-0 ug. per ml. in six. Arteriosclerotic Patients Fig. 1 shows the results obtained in patient 1, a female inpatient aged 55 in whom treatment with phenformin plus ethyloestrenol was started 1 month after the onset of right-sided hemiplegia. The lysis-time fell on the 4th day of treatment, the plasma-fibrinogen had fallen by day 7, and the F.D.P. level rose from 1-25 to 5 ug. per ml. by was
ethylrestrenol
on F.D.P., plasmaFig. 1—Effect of phenformin plus fibrinogen, and dilute-blood-clot lysis-time in a woman with hemiplegia.
activity has any function in the body, and that its enhancement by drugs given orally could have therapeutic consequences. By measuring fibrin-degradation products (F.D.P.S) in the sera of patients treated with phenformin plus ethyloestrenol we have obtained evidence that increased fibrinolytic activity as measured in vitro is accompanied by increased fibrinolysis in vivo. Patients and Methods The effect of phenformin timed-release capsules 50 mg. twice daily plus ethylcestrenol 4 mg. twice daily on blood fibrinolytic activity, plasma-fibrinogen, and F.D.P. in serum were investigated in ten patients with various arteriosclerotic conditions and six patients with rheumatoid arthritis. Two healthy people who received phenformin capsules 50 mg. b.d. plus ethyloestrenol 2 mg. b.d. for 3 months were also studied. The normal range Of F.D.P. by the method used was obtained by single estimations in 18 healthy males. Blood-samples for all measurements were obtained between 10 A.M. and 11 A.M. Methods Blood fibrinolytic activity was measured by the dilute-bloodclot lysis-time (Fearnley et al. 1957) as modified by Fearnley
(1964). Plasma-fibrinogen was estimated gravimetrically by the method of Fearnley and Chakrabarti (1966). F.D.P. F.D.P. in serum were estimated by a modification of the tanned-red-cell hæmagglutination-inhibition immunoassay of Merskey et al. (1966). A microtitre kit (Flow Laboratories) was used. Formalin-treated sheep red blood-cells sensitised to human fibrinogen, purified human fibrinogen, and rabbit antifibrinogen serum were obtained from Burroughs Wellcome Ltd. 5-0 ml. of blood to which 100 units of’Trasylol’
Fig. 2-Effect of phenformin plus ethyloestrenol on F.D.P., plasmafibrinogen, and dilute-blood-clot lysis-time in a man with
hemiplegia.
-
912 TABLE I-EFFECT OF PHENFORMIN PLUS ETHYLCESTRENOL ON AND
(c)
F.D.P.
(ug./ml.)
(a)
LYSIS-TIME
(hr.), (b
PLASMA-FIBRINOGEN
(mg./1OO ml.),
IN EIGHT ARTERIOSCLEROTIC PATIENTS
after which it increased steadily to reach a peak of 2 weeks’ treatment, but in the rest this was not apparent 160 ug. per ml. by day 28. All these changes reverted to until the 3rd or 4th week of treatment. The increases of or towards their initial levels when placebo therapy was F.D.P. ranged from 40 to 320 ug. per ml. in different instituted from days 37 to 49. After discharge from patients. hospital this patientwas again given phenformin plus Rheumatic Patients ethyloestrenol and observed at weekly intervals. There Table 11 gives the results obtained in six outpatients was a sustained reduction of her lysis-times and plasmawith rheumatoid arthritis who have received phenformin fibrinogen which had risen during placebo treatment and plus ethyloestrenol for 3 months and are still continuing on a transient rise of F.D.P. to 20 ug. per ml. for 6 weeks after the drugs. After control readings blood-samples were which it fell to 2-5 ug. per ml. (i.e., close to the level taken at fortnightly and then at monthly intervals for obtained before treatment was started). measurements of lysis-time, plasma-fibrinogen, and F.D.P. 2 in a man 53 shows similar results aged Fig. patient 2, during treatment. The F.D.P. level rose significantly in all in whom treatment was started 1 month after the onset of patients and was still high at the 12th week of treatment. a right hemiplegia. Again the earliest change was reducDespite some variation in timing, fibrinolytic activity and tion of lysis-time, but in contrast to the findings illustrated F.D.P. level tended to change in parallel. The two patients in fig. 1, a sharp increase of F.D.P. from 2-5 to 320 ug. per (nos. 13 and 16) who did not respond clinically (indeed ml. preceded any appreciable change in plasma-fibrinogen they deteriorated during treatment) showed a sustained which did not start to fall until after the F.D.P. level had increase of F.D.P. despite the persistence of a raised reached its maximum. Again these changes were reversed fibrinogen level. In the remaining patients who improved when placebo therapy was instituted from days 31 to 54. clinically, increases of F.D.P. were accompanied by reductions of plasma-fibrinogen. Table i gives the lysis-times, plasma-fibrinogen levels, and F.D.P. levels of two inpatients (nos. 3 and 4) and six Healthy Controls outpatients (nos. 5-10) with various arteriosclerotic conFig. 3 shows the results obtained in J. F. E., aged 40. ditions who, after single control measurements, were given On the 4th day of treatment his lysis-time fell to 5 hours phenformin plus ethyloestrenol for 4 weeks followed by after which it remained at 4-5 hours with continued placebo capsules and tablets for 2 weeks, and from whom medication. Plasma-fibrinogen rose to 340 mg. per 100 ml. blood-samples were obtained at weekly intervals. Except on the 4th day of treatment, was then reduced and rose for patient 10, who was relatively unresponsive since his later to 330 mg. per 100 ml., which may have been due to fibrinolytic activity fluctuated during treatment, the lysis- coryza, and thereafter fell to 200-240 mg. per 100 ml. times of the remainder, only three of which (nos. 5, 7, F.D.P. rose from 1-25 to 5 ug. per ml. to reach a peak of 160 ug. per ml. on day 28, after which it fell gradually to and 9) were initially in the long range, were reduced by phenformin plus ethyloestrenol, and there were corres- its control level. Similar results were obtained in G. R. F., aged 53, ponding reductions of plasma-fibrinogen and increases of whose lysis-times, plasma-fibrinogen, and F.D.P. levels the levels of F.D.P. These changes reverted to or near to their pretreatment levels during placebo therapy. F.D.P. during the control period were 6-61/2 hours, 240 mg. per was most increased in two patients (nos. 7 and 8) after 100 ml., and 2-5 pg. per ml. respectively. In this subject
day 9,
913 TABLE II-EFFECT OF PHENFORMIN PLUS ETHYLCESTRENOL ON (a) LYSIS-TIME (hr.), (b) PLASMA-FIBRINOGEN (mg.l ml.), and IN SIX PATIENTS WITH RHEUMATOID ARTHRITIS (c) F.D.P.
(fLg.(ml.)
transitory despite sustained increases of fibrinolytic activity, patient 6 and, possibly, patients 7 and 8 (table I) showed a similar pattern after only 4 weeks’ treatment. In the patients with rheumatoid arthritis, on the other hand (table 11), F.D.P. was still elevated at the third month of treatment which suggests that the shorter response of the healthy and at least some of the arteriosclerotic patients in this respect was due not to the development of resistance but to equilibration between fibrin deposition and removal. In other words it is possible that when the F.D.P. returns to pretreatment levels during continued medication, intravascular defibrination has been achieved. On the contrary, the sustained elevation of F.D.P. in the patients with rheumatoid arthritis may be attributed to the effect of enhanced fibrinolytic activity on continuing
synovial deposition of fibrin.
gradually to reach 160 g. per ml. on day 35, after which it fell to 2-5 g. per ml. on day 42 and thereafter remained at this level. Plasma-fibrinogen levels remained unaltered, but the lysis-times were reduced to 4-5 hours after 4 days’ treatment and were maintained at this level. Phenformin plus ethyloestrenol, therefore, produced a sustained increase of fibrinolytic activity and a temporary increase of F.D.P. in both subjects. F.D.P. rose
Duration of Effect Besides J. F. E., G. R. F., and patient 1 during patient treatment, in whom the increases of F.D.P.
outwere
Fig. 3-Effect of phenformin plus ethylcestrenol on F.D.P., plasmafibrinogen, and dilute-blood-clot lysis-time in a healthy man.
Discussion These findings provide the first direct evidence that enhancement of fibrinolytic activity by phenformin plus ethyloestrenol, as judged by an in-vitro test, is accompanied by an increased breakdown of fibrin/fibrinogen in the body. They also confirm that the dilute-blood-clot lysis-time reflects the behaviour of a functioning fibrinolytic mechanism in vivo. The method of assay does not determine whether the F.D.P. arise from fibrin or from fibrinogen, or from both. Despite one report of the presence of free plasmin in gastric venous blood (Cox et al. 1967) available evidence indicates that fibrinogenolysis is encountered only in states of pathological fibrinolytic activity, and that activity of physiological degree, with which we are here concerned, is insufficient to overcome the antiplasmin defence and is thus incapable of lysing fibrinogen as opposed to fibrin. Admittedly, phenformin plus ethyloestrenol reduces raised levels of plasmafibrinogen (Fearnley, Chakrabarti, and Hocking 1967), but this may be due to depression of biosynthesis or to a feedback mechanism from removal of fibrin (Fearnley and Chakrabarti 1966) ; and it should be noted that the incidence and magnitude of increase of F.D.P. in our patients bore no relation to their levels of plasmafibrinogen before treatment was started. It seems likely therefore that the increase of F.D.P. by phenformin plus ethyloestrenol was due to lysis of fibrin. If this be accepted, more than one possibility needs considering. Should this combination of drugs promote deposition of fibrin, though there is no evidence that it does, then it would be conceivable that increased fibrinolytic activity and the rise of F.D.P. were homoeostatic consequences of such an effect. Were this so, however, the sequence expected would be reduction of plasmafibrinogen, followed by reduction of lysis-time and rise in the level of F.D.P. Whilst in some patients (nos. 3-16) the timing of the tests did not permit recognition of the order of change, in J. F. E., G. R. F., and patients 1 and 2, in whom more frequent measurements were made (figs. 1-3), reduction of lysis-time preceded both reduction of plasmafibrinogen and increase of F.D.P.; and observations in other patients, not reported here, confirm this order of change. It seems likely, therefore, as our previous findings suggest (Fearnley et al. 1967, Fearnley and Chakrabarti 1968) that phenformin plus ethyloestrenol primarily raises blood fibrinolytic activity and that the ensuing increase of F.D.P. reflects removal of deposited fibrin. In other words, the drugs achieve pharmacological defibrination. These findings do not of course establish the site of fibrin removal though it is probable that vascular endothelium in the
914
arteriosclerotic patients and synovial the patients with rheumatoid arthritis
membrane were
in
GROSS FRAGMENTATION OF CARDIAC
important
MYOFIBRILS AFTER THERAPEUTIC
sources.
At the same time the possibility remains that phenformin plus ethylcestrenol in some way directly influences the rate of fibrinogen-to-fibrin turnover. To what extent fibrinogen is converted to fibrin in vivo is at present uncertain, but studies with radioactive fibrinogen before and during treatment with this combination of drugs might throw light on this. In this connection our findings accord with the proposal first put forward by Nolf (1908) and subsequently espoused by others (Copley 1954, Astrup 1956, Fearnley 1961) that deposition and removal of fibrin is in continuous flux, in other words that coagulation and fibrinolysis are in a state of dynamic equilibrium. This concept has been extended to suggest that coagulation and fibrinolysis together form a whole, a physiological system of repair which uses fibrin as a cement; and that imbalance of either side of the system leading to increment and persistence of fibrin may be of pathogenic importance in vascular occlusion and other conditions
(Fearnley 1965). The behaviour of fibrinogen/fibrin may be of importin conditions other than occlusive vascular diseaserheumatoid arthritis (Fearnley and Chakrabarti 1966), e.g., cancer (O’Meara 1958, Cliffton 1966), and nephritis (Kincaid-Smith et al. 1968, Wardle 1969). Phenformin plus ethyloestrenol may have application both as a research tool and therapeutic agent in such situations. Meanwhile our findings lend weight to the suggestion (Fearnley et al. 1967, Chakrabarti et al. 1968, Fearnley and Chakrabarti 1968) that phenformin plus ethyloestrenol, which increases fibrinolytic activity, lowers plasma-fibrinogen and serumcholesterol levels, and reduces platelet stickiness to glassall of these effects being sustained with continued treatment-merits trial as a prophylactic measure in survivors of vascular occlusions, and also in other situations where deposition of platelets and fibrin are detrimental (e.g., after vascular surgery and the insertion of cardiac ance
prostheses). This work was supported by grants from the Nuffield Products Ltd., and Organon Laboratories Ltd.
Foundation,
Bayer
Request for reprints should be addressed
to
G. R. F.
STARVATION FOR OBESITY D. L. BARNARD
E. S. GARNETT
R. A. GOODBODY
J. FORD
M. A. WOODEHOUSE FROM THE WESSEX REGIONAL DEPARTMENT OF NUCLEAR MEDICINE AND THE DEPARTMENT OF PATHOLOGY, GENERAL HOSPITAL, SOUTHAMPTON SO9 4XY
Death from ventricular fibrillation consequent upon therapeutic starvation to her ideal weight is described in an obese but otherwise healthy 20-year-old girl. At the time of her death the plasma electrolytes, calcium, magnesium, pH, and blood-gas tensions were normal. The extracellular fluid volume was also normal although the lean body mass was reduced. The myocardial fibres were reduced to approximately half their normal diameter, and electron microscopy revealed gross loss and fragmentation of the myofibrils. It is concluded that prolonged total starvation is an unsafe
Sum ary
procedure. Introduction DEATH as a result of total starvation in obese but otherwise healthy patients has not been described. Cubberley et al. (1965) reported the death of a 44-yearold hypertensive diabetic woman after 3 weeks of total starvation. They attributed the death to idiopathic lactic acidosis and found focal stenosis of coronary arteries at necropsy. Spencer (1968) reported two deaths during
starvation, one in a 61-year-old woman with an old anteroseptal infarct and gross atheroma who had presented with acute left ventricular failure; the other in a 58-yearold severely hypertensive woman with right-bundlebranch block who had presented in gross congestive failure. These patients had starved for 3 and 8 weeks, respectively. We report here the death (after 30 weeks of total starvation) of an obese but otherwise healthy young woman at a time when she had achieved her ideal weight. total
REFERENCES
Astrup, T. (1956) Lancet, ii, 565. Bell, W. R., Pitney, W. R., Oakley, C. M., Goodwin, J. (1968) ibid. i, 490. Chakrabarti, R., Fearnley, G. R. (1967) ibid. ii, 1012. - Hocking, E. D., Fearnley, G. R., Mann, R. D., Attwell, T. N., Jackson, D. (1968) ibid. i, 987. Cliffton, E. E. (1966) Fedn Proc. 25, 89. Copley, A. L. (1954) Archs Int. Pharmacodyn. 99, 426. Cox, H. T., Poller, L., Thomson, J. M. (1967) Lancet, i, 1300. Fearnley, G. R. (1961) ibid. i, 992. (1964) J. clin. Path. 17, 307. (1965) Fibrinolysis; p. 174. London. Balmforth, G. V., Fearnley, E. (1957) Clin. Sci. 16, 645. Chakrabarti, R. (1966) Lancet, ii, 757. (1968) ibid. ii, 1004. Hocking, E. D. (1967) ibid. ii, 1008. Gilles, N. M., Reid, H. A., Odutola, A., Ransome-Kuti, O., Lesi, F., Ransome-Kuti, S. (1968) ibid. ii, 542. Kincaid-Smith, P., Saker, B. M., Fairley, K. F. (1968) ibid. p. 1360. Merskey, C., Kleiner, G. J., Johnson, A. J. (1966) Blood, 28, 1. Nolf, P. (1908) Archs int. Physiol. 6, 306. O’Meara, R. A. Q. (1958) Irish J. med. Sci. 394, 474. Pitney, W. R., Bray, C., Holt, P. J. L., Bolton, G. (1969) Lancet, i, 79. Sharp, A. A., Warren, B. A., Paxton, A. M., Allington, M. J. (1968) ibid. i, 493. von Kaulla, K. N. (1963) Chemistry of Thrombolysis: Human Fibrinolytic Enzymes; p. 79. Springfield, Illinois. Wardle, E. N. (1969) Lancet, i, 50. -
-
-
—
-
—
-
—
Fig. l-E.C.G. with voltage adjusted divisions.
(a) Hypokalaemia. (b) 1 month before death. (c) After cardiac arrest.
to 1 mV=10 small
vertical
prevalent in man around the beginning of this century. The findings reported by Kasel et al. (1965) and Masurel and Mulder (1966), the antibody response to vaccination with A/Equi2/63 (Davenport et al. 1967, Masurel 1968a), and the results reported here provide conclusive evidence of an antigenic relationship between this virus and the A2 viruses. The relationship between Hong Kong virus and the A/Equi2 viruses is also supported by other work (Coleman et al. 1968, Davenport 1968, Masurel 1968b). The increase in antibody titres against A/Equi2 presumably due to infection with A2 strains was also seen in sera collected in 1958 and 1964 (Davenport et al. 1967). The distribution of antibody against A/Equi2/Richelieu/63 in sera collected in 1967 fits well with its antigenic relationship to the Hong Kong virus; this antibody pattern and that found in 1957 also confirm the suggestion that a virus related to both of these strains was circulating between 1890 and 1918, when persons now aged fifty-or morealive. The absence of antibody against the A/Turkey/ Massachusetts/65 strain in the human sera collected in 1956-57 suggests that this strain is not antigenically related to influenza strains which circulated in man in the Netherlands before 1957. It would be unwise to conclude that there is an exact correspondence between the antibodies found in sera and the timing and antigens of influenza epidemics in the past, but it seems reasonable to conclude that the 1957 influenza pandemic may have been similar to the pandemic of " 1889-90, and that the virus of the " Hong Kong 1968 influenza epidemic may be closely related to the virus which followed the " A2 pandemic " of the turn of the century. I thank A. J. Baars, E. M. Degger, Henny Frankema, P. Kuyt, J. Magdelijns, and J. J. Vingerling for technical assistance. were
REFERENCES M. W. Coleman, T., Dowdle, R., Pereira, H. G., Schild, G. C., Chang, W. K. (1968) Lancet, ii, 1384. Davenport, F. M. (1968) Personal communication. Hennessy, A. V., Minuse, E. (1967) J. exp. Med. 126, 1049. Francis, T., Jr. (1960) Proc. Am. Phil. 104, 572. Kasel, J. A., Alford, R. H., Knight, V. A., Wadell, G. H., Sigel, M. M. (1965) Ann. intern. Med. 62, 1308. Masurel, N. (1967) Ned. Tijdschr. Geneesk. 111, 1245. (1968a) Nature, Lond. 218, 100. (1968b) Ned. Tijdschr. Geneesk. 112, 1930. Mulder, J. (1962) Verh. Inst. prev. Geneesk. no. 52. — — (1966) Bull. Wld Hlth Org. 34, 885. Minuse, E., McQueen, J. L., Davenport, F. M., Francis, T., Jr. (1965) J. Immun. 94, 563. Mulder, J., Masurel, N. (1958) Lancet, i, 810. Schild, G. C., Stuart-Harris, C. H. (1965) J. Hyg., Camb. 63, 479. van der Veen, J., Mulder, J. (1950) Verh. Inst. prev. Geneesk. no. 6.
FIBRINOLYTIC AND DEFIBRINATING EFFECT OF PHENFORMIN PLUS ETHYLŒSTRENOL IN VIVO G. R. FEARNLEY
J.
R. CHAKRABARTI F. EVANS
FROM THE GLOUCESTERSHIRE ROYAL HOSPITAL, GLOUCESTER
The effects of
phenformin plus ethylfibrin-degradation products (F.D.P.) in serum, fibrinolytic activity, and plasmafibrinogen were investigated in ten patients with occlusive vascular disease, six patients with rheumatoid arthritis, and two healthy people. All showed an increase of F.D.P. during treatment, which greatly exceeded their pretreatSummary œstrenol
on
levels and the upper limit of normal found in eighteen healthy males. Phenformin plus ethylœstrenol increased fibrinolytic activity, and removal of fibrin in vivo was a consequence of this; but the possibility that the drugs were in some way directly influencing fibrinogenfibrin turnover has not been excluded. Increases of F.D.P. in the healthy volunteers and in some of the arteriosclerotic patients were temporary, and this may be ascribed to successful defibrination rather than to resistance since the increases were sustained in the patients with rheumatoid arthritis, a condition accompanied by continuing deposition of fibrin. The findings provide the first direct evidence that pharmacological enhancement of fibrinolysis as measured in vitro is accompanied by increased fibrinolysis in vivo; they also establish that the dilute-blood-clot lysis-time reflects the behaviour of a functioning fibrinolytic system in the body; and are consonant with a dynamic equilibrium between fibrinolysis and coagulation, imbalance of which may be of pathogenic importance. Phenformin plus ethylœstrenol seems to merit trial as a prophylactic measure in survivors of vascular occlusions and in patients after vascular surgery and the insertion of cardiac prostheses. ment
Introduction
-
-
-
-
"... educational considerations ... often lead good parents, who would like more children, to limit their families ... In the United Kingdom there is still a tradition of independent and often expensive private schooling, which in many instances, though in fewer perhaps than the parents realise, gives a better education than that provided in the state schools. The development of comprehensive education is, for a time at least, likely to accentuate these differences between the state and private education ... it is being found most difficult... to provide adequate sixth form education in comprehensive schools ... The eugenist must therefore welcome the disappearance of the good independent schools. They provide a temptation to good parents, who value education, to limit the size of their families. The number of parents influenced by such considerations is small but they probably include some genetically valuable ones."-C. 0. CARTER. Eugenics and Human Heredity: Essex Hall Lecture, April 14, 1969.
THERAPEUTIC defibrination with ’Arvin’ has given encouraging results in patients with acute vascular occlusions (Bell et al. 1968, Gilles et al. 1968, Sharp et al. 1968) but since arvin has to be injected and when given intramuscularly evokes resistance (Pitney et al. 1969) it is unlikely to be suitable for long-term prophylaxis. We have found that phenformin plus ethyloestrenol given twice daily by mouth produces a sustained increase in blood fibrinolytic activity, as judged by reduction of both the dilute-blood-clot and euglobulin lysis-times, and reduces plasma-fibrinogen and serum-cholesterol levels in arteriopathic patients (Fearnley, Chakrabarti, and Hocking 1967). This combination of drugs reduced platelet stickiness to glass in some 75% of patients (Chakrabarti and Fearnley 1967), and this effect was maintained with continued treatment (Fearnley and Chakrabarti 1968). Theoretically,
phenformin plus ethyloestrenol, by enhancing fibrinolysis, might be expected to exercise a defibrinating effect in patients with occlusive vascular disease and hence to merit trial as a prophylactic measure in survivors of vascular accidents. For spontaneous fibrinolysis to be demonstrated, blood has to be manipulated, either by dilution (Fearnley et al. 1957) or by precipitation of the euglobulin fraction (von Kaulla 1963), and so there is no guarantee that what is observed in vitro reflects events in vivo. Indeed this has been the greatest bar to acceptance that natural fibrinolytic
911 was allowed to clot at 37°C for 4 hours. The resultant serum was treated with 10 units of topical thrombin (Parke-Davis) to ensure that all the fibrinogen was clotted and after centrifugation was stored at -15°C. On the day before the test the serum was thawed at room temperature and then absorbed with sheep red cells for 24 hours at 4°C to remove non-specific antibodies. Tests were done once or twice weekly on 12-25 batches of sera. After being set up the tests were kept at 4°C for 4 hours before being read. The end-point for the sera being tested and for the standard was taken as the highest dilution to show complete inhibition of haemagglutination. The tests were read independently by the three of us; we did not know the source of the sera at the time. Agreed end-points were obtained in 60% of tests; in the remainder when two of the three readings agreed this was taken as the end-point; on the very few occasions when all three readings differed the test was set up again. Occasionally, incomplete absorption of non-specific antibodies made reading of the test difficult or impossible; when this happened the test was repeated after the serum had been further absorbed with sheep red blood-cells. Results are expressed as multiples of 1 25 ug. of F.D.P. per ml. of serum.
(aprotinin) were added
Results Normal Range
Of
eighteen healthy males aged 22-53 the F.D.P. level 1-25 ug. per ml. in three, 2-5 ug. per ml. in nine, and 5-0 ug. per ml. in six. Arteriosclerotic Patients Fig. 1 shows the results obtained in patient 1, a female inpatient aged 55 in whom treatment with phenformin plus ethyloestrenol was started 1 month after the onset of right-sided hemiplegia. The lysis-time fell on the 4th day of treatment, the plasma-fibrinogen had fallen by day 7, and the F.D.P. level rose from 1-25 to 5 ug. per ml. by was
ethylrestrenol
on F.D.P., plasmaFig. 1—Effect of phenformin plus fibrinogen, and dilute-blood-clot lysis-time in a woman with hemiplegia.
activity has any function in the body, and that its enhancement by drugs given orally could have therapeutic consequences. By measuring fibrin-degradation products (F.D.P.S) in the sera of patients treated with phenformin plus ethyloestrenol we have obtained evidence that increased fibrinolytic activity as measured in vitro is accompanied by increased fibrinolysis in vivo. Patients and Methods The effect of phenformin timed-release capsules 50 mg. twice daily plus ethylcestrenol 4 mg. twice daily on blood fibrinolytic activity, plasma-fibrinogen, and F.D.P. in serum were investigated in ten patients with various arteriosclerotic conditions and six patients with rheumatoid arthritis. Two healthy people who received phenformin capsules 50 mg. b.d. plus ethyloestrenol 2 mg. b.d. for 3 months were also studied. The normal range Of F.D.P. by the method used was obtained by single estimations in 18 healthy males. Blood-samples for all measurements were obtained between 10 A.M. and 11 A.M. Methods Blood fibrinolytic activity was measured by the dilute-bloodclot lysis-time (Fearnley et al. 1957) as modified by Fearnley
(1964). Plasma-fibrinogen was estimated gravimetrically by the method of Fearnley and Chakrabarti (1966). F.D.P. F.D.P. in serum were estimated by a modification of the tanned-red-cell hæmagglutination-inhibition immunoassay of Merskey et al. (1966). A microtitre kit (Flow Laboratories) was used. Formalin-treated sheep red blood-cells sensitised to human fibrinogen, purified human fibrinogen, and rabbit antifibrinogen serum were obtained from Burroughs Wellcome Ltd. 5-0 ml. of blood to which 100 units of’Trasylol’
Fig. 2-Effect of phenformin plus ethyloestrenol on F.D.P., plasmafibrinogen, and dilute-blood-clot lysis-time in a man with
hemiplegia.
-
912 TABLE I-EFFECT OF PHENFORMIN PLUS ETHYLCESTRENOL ON AND
(c)
F.D.P.
(ug./ml.)
(a)
LYSIS-TIME
(hr.), (b
PLASMA-FIBRINOGEN
(mg./1OO ml.),
IN EIGHT ARTERIOSCLEROTIC PATIENTS
after which it increased steadily to reach a peak of 2 weeks’ treatment, but in the rest this was not apparent 160 ug. per ml. by day 28. All these changes reverted to until the 3rd or 4th week of treatment. The increases of or towards their initial levels when placebo therapy was F.D.P. ranged from 40 to 320 ug. per ml. in different instituted from days 37 to 49. After discharge from patients. hospital this patientwas again given phenformin plus Rheumatic Patients ethyloestrenol and observed at weekly intervals. There Table 11 gives the results obtained in six outpatients was a sustained reduction of her lysis-times and plasmawith rheumatoid arthritis who have received phenformin fibrinogen which had risen during placebo treatment and plus ethyloestrenol for 3 months and are still continuing on a transient rise of F.D.P. to 20 ug. per ml. for 6 weeks after the drugs. After control readings blood-samples were which it fell to 2-5 ug. per ml. (i.e., close to the level taken at fortnightly and then at monthly intervals for obtained before treatment was started). measurements of lysis-time, plasma-fibrinogen, and F.D.P. 2 in a man 53 shows similar results aged Fig. patient 2, during treatment. The F.D.P. level rose significantly in all in whom treatment was started 1 month after the onset of patients and was still high at the 12th week of treatment. a right hemiplegia. Again the earliest change was reducDespite some variation in timing, fibrinolytic activity and tion of lysis-time, but in contrast to the findings illustrated F.D.P. level tended to change in parallel. The two patients in fig. 1, a sharp increase of F.D.P. from 2-5 to 320 ug. per (nos. 13 and 16) who did not respond clinically (indeed ml. preceded any appreciable change in plasma-fibrinogen they deteriorated during treatment) showed a sustained which did not start to fall until after the F.D.P. level had increase of F.D.P. despite the persistence of a raised reached its maximum. Again these changes were reversed fibrinogen level. In the remaining patients who improved when placebo therapy was instituted from days 31 to 54. clinically, increases of F.D.P. were accompanied by reductions of plasma-fibrinogen. Table i gives the lysis-times, plasma-fibrinogen levels, and F.D.P. levels of two inpatients (nos. 3 and 4) and six Healthy Controls outpatients (nos. 5-10) with various arteriosclerotic conFig. 3 shows the results obtained in J. F. E., aged 40. ditions who, after single control measurements, were given On the 4th day of treatment his lysis-time fell to 5 hours phenformin plus ethyloestrenol for 4 weeks followed by after which it remained at 4-5 hours with continued placebo capsules and tablets for 2 weeks, and from whom medication. Plasma-fibrinogen rose to 340 mg. per 100 ml. blood-samples were obtained at weekly intervals. Except on the 4th day of treatment, was then reduced and rose for patient 10, who was relatively unresponsive since his later to 330 mg. per 100 ml., which may have been due to fibrinolytic activity fluctuated during treatment, the lysis- coryza, and thereafter fell to 200-240 mg. per 100 ml. times of the remainder, only three of which (nos. 5, 7, F.D.P. rose from 1-25 to 5 ug. per ml. to reach a peak of 160 ug. per ml. on day 28, after which it fell gradually to and 9) were initially in the long range, were reduced by phenformin plus ethyloestrenol, and there were corres- its control level. Similar results were obtained in G. R. F., aged 53, ponding reductions of plasma-fibrinogen and increases of whose lysis-times, plasma-fibrinogen, and F.D.P. levels the levels of F.D.P. These changes reverted to or near to their pretreatment levels during placebo therapy. F.D.P. during the control period were 6-61/2 hours, 240 mg. per was most increased in two patients (nos. 7 and 8) after 100 ml., and 2-5 pg. per ml. respectively. In this subject
day 9,
913 TABLE II-EFFECT OF PHENFORMIN PLUS ETHYLCESTRENOL ON (a) LYSIS-TIME (hr.), (b) PLASMA-FIBRINOGEN (mg.l ml.), and IN SIX PATIENTS WITH RHEUMATOID ARTHRITIS (c) F.D.P.
(fLg.(ml.)
transitory despite sustained increases of fibrinolytic activity, patient 6 and, possibly, patients 7 and 8 (table I) showed a similar pattern after only 4 weeks’ treatment. In the patients with rheumatoid arthritis, on the other hand (table 11), F.D.P. was still elevated at the third month of treatment which suggests that the shorter response of the healthy and at least some of the arteriosclerotic patients in this respect was due not to the development of resistance but to equilibration between fibrin deposition and removal. In other words it is possible that when the F.D.P. returns to pretreatment levels during continued medication, intravascular defibrination has been achieved. On the contrary, the sustained elevation of F.D.P. in the patients with rheumatoid arthritis may be attributed to the effect of enhanced fibrinolytic activity on continuing
synovial deposition of fibrin.
gradually to reach 160 g. per ml. on day 35, after which it fell to 2-5 g. per ml. on day 42 and thereafter remained at this level. Plasma-fibrinogen levels remained unaltered, but the lysis-times were reduced to 4-5 hours after 4 days’ treatment and were maintained at this level. Phenformin plus ethyloestrenol, therefore, produced a sustained increase of fibrinolytic activity and a temporary increase of F.D.P. in both subjects. F.D.P. rose
Duration of Effect Besides J. F. E., G. R. F., and patient 1 during patient treatment, in whom the increases of F.D.P.
outwere
Fig. 3-Effect of phenformin plus ethylcestrenol on F.D.P., plasmafibrinogen, and dilute-blood-clot lysis-time in a healthy man.
Discussion These findings provide the first direct evidence that enhancement of fibrinolytic activity by phenformin plus ethyloestrenol, as judged by an in-vitro test, is accompanied by an increased breakdown of fibrin/fibrinogen in the body. They also confirm that the dilute-blood-clot lysis-time reflects the behaviour of a functioning fibrinolytic mechanism in vivo. The method of assay does not determine whether the F.D.P. arise from fibrin or from fibrinogen, or from both. Despite one report of the presence of free plasmin in gastric venous blood (Cox et al. 1967) available evidence indicates that fibrinogenolysis is encountered only in states of pathological fibrinolytic activity, and that activity of physiological degree, with which we are here concerned, is insufficient to overcome the antiplasmin defence and is thus incapable of lysing fibrinogen as opposed to fibrin. Admittedly, phenformin plus ethyloestrenol reduces raised levels of plasmafibrinogen (Fearnley, Chakrabarti, and Hocking 1967), but this may be due to depression of biosynthesis or to a feedback mechanism from removal of fibrin (Fearnley and Chakrabarti 1966) ; and it should be noted that the incidence and magnitude of increase of F.D.P. in our patients bore no relation to their levels of plasmafibrinogen before treatment was started. It seems likely therefore that the increase of F.D.P. by phenformin plus ethyloestrenol was due to lysis of fibrin. If this be accepted, more than one possibility needs considering. Should this combination of drugs promote deposition of fibrin, though there is no evidence that it does, then it would be conceivable that increased fibrinolytic activity and the rise of F.D.P. were homoeostatic consequences of such an effect. Were this so, however, the sequence expected would be reduction of plasmafibrinogen, followed by reduction of lysis-time and rise in the level of F.D.P. Whilst in some patients (nos. 3-16) the timing of the tests did not permit recognition of the order of change, in J. F. E., G. R. F., and patients 1 and 2, in whom more frequent measurements were made (figs. 1-3), reduction of lysis-time preceded both reduction of plasmafibrinogen and increase of F.D.P.; and observations in other patients, not reported here, confirm this order of change. It seems likely, therefore, as our previous findings suggest (Fearnley et al. 1967, Fearnley and Chakrabarti 1968) that phenformin plus ethyloestrenol primarily raises blood fibrinolytic activity and that the ensuing increase of F.D.P. reflects removal of deposited fibrin. In other words, the drugs achieve pharmacological defibrination. These findings do not of course establish the site of fibrin removal though it is probable that vascular endothelium in the
914
arteriosclerotic patients and synovial the patients with rheumatoid arthritis
membrane were
in
GROSS FRAGMENTATION OF CARDIAC
important
MYOFIBRILS AFTER THERAPEUTIC
sources.
At the same time the possibility remains that phenformin plus ethylcestrenol in some way directly influences the rate of fibrinogen-to-fibrin turnover. To what extent fibrinogen is converted to fibrin in vivo is at present uncertain, but studies with radioactive fibrinogen before and during treatment with this combination of drugs might throw light on this. In this connection our findings accord with the proposal first put forward by Nolf (1908) and subsequently espoused by others (Copley 1954, Astrup 1956, Fearnley 1961) that deposition and removal of fibrin is in continuous flux, in other words that coagulation and fibrinolysis are in a state of dynamic equilibrium. This concept has been extended to suggest that coagulation and fibrinolysis together form a whole, a physiological system of repair which uses fibrin as a cement; and that imbalance of either side of the system leading to increment and persistence of fibrin may be of pathogenic importance in vascular occlusion and other conditions
(Fearnley 1965). The behaviour of fibrinogen/fibrin may be of importin conditions other than occlusive vascular diseaserheumatoid arthritis (Fearnley and Chakrabarti 1966), e.g., cancer (O’Meara 1958, Cliffton 1966), and nephritis (Kincaid-Smith et al. 1968, Wardle 1969). Phenformin plus ethyloestrenol may have application both as a research tool and therapeutic agent in such situations. Meanwhile our findings lend weight to the suggestion (Fearnley et al. 1967, Chakrabarti et al. 1968, Fearnley and Chakrabarti 1968) that phenformin plus ethyloestrenol, which increases fibrinolytic activity, lowers plasma-fibrinogen and serumcholesterol levels, and reduces platelet stickiness to glassall of these effects being sustained with continued treatment-merits trial as a prophylactic measure in survivors of vascular occlusions, and also in other situations where deposition of platelets and fibrin are detrimental (e.g., after vascular surgery and the insertion of cardiac ance
prostheses). This work was supported by grants from the Nuffield Products Ltd., and Organon Laboratories Ltd.
Foundation,
Bayer
Request for reprints should be addressed
to
G. R. F.
STARVATION FOR OBESITY D. L. BARNARD
E. S. GARNETT
R. A. GOODBODY
J. FORD
M. A. WOODEHOUSE FROM THE WESSEX REGIONAL DEPARTMENT OF NUCLEAR MEDICINE AND THE DEPARTMENT OF PATHOLOGY, GENERAL HOSPITAL, SOUTHAMPTON SO9 4XY
Death from ventricular fibrillation consequent upon therapeutic starvation to her ideal weight is described in an obese but otherwise healthy 20-year-old girl. At the time of her death the plasma electrolytes, calcium, magnesium, pH, and blood-gas tensions were normal. The extracellular fluid volume was also normal although the lean body mass was reduced. The myocardial fibres were reduced to approximately half their normal diameter, and electron microscopy revealed gross loss and fragmentation of the myofibrils. It is concluded that prolonged total starvation is an unsafe
Sum ary
procedure. Introduction DEATH as a result of total starvation in obese but otherwise healthy patients has not been described. Cubberley et al. (1965) reported the death of a 44-yearold hypertensive diabetic woman after 3 weeks of total starvation. They attributed the death to idiopathic lactic acidosis and found focal stenosis of coronary arteries at necropsy. Spencer (1968) reported two deaths during
starvation, one in a 61-year-old woman with an old anteroseptal infarct and gross atheroma who had presented with acute left ventricular failure; the other in a 58-yearold severely hypertensive woman with right-bundlebranch block who had presented in gross congestive failure. These patients had starved for 3 and 8 weeks, respectively. We report here the death (after 30 weeks of total starvation) of an obese but otherwise healthy young woman at a time when she had achieved her ideal weight. total
REFERENCES
Astrup, T. (1956) Lancet, ii, 565. Bell, W. R., Pitney, W. R., Oakley, C. M., Goodwin, J. (1968) ibid. i, 490. Chakrabarti, R., Fearnley, G. R. (1967) ibid. ii, 1012. - Hocking, E. D., Fearnley, G. R., Mann, R. D., Attwell, T. N., Jackson, D. (1968) ibid. i, 987. Cliffton, E. E. (1966) Fedn Proc. 25, 89. Copley, A. L. (1954) Archs Int. Pharmacodyn. 99, 426. Cox, H. T., Poller, L., Thomson, J. M. (1967) Lancet, i, 1300. Fearnley, G. R. (1961) ibid. i, 992. (1964) J. clin. Path. 17, 307. (1965) Fibrinolysis; p. 174. London. Balmforth, G. V., Fearnley, E. (1957) Clin. Sci. 16, 645. Chakrabarti, R. (1966) Lancet, ii, 757. (1968) ibid. ii, 1004. Hocking, E. D. (1967) ibid. ii, 1008. Gilles, N. M., Reid, H. A., Odutola, A., Ransome-Kuti, O., Lesi, F., Ransome-Kuti, S. (1968) ibid. ii, 542. Kincaid-Smith, P., Saker, B. M., Fairley, K. F. (1968) ibid. p. 1360. Merskey, C., Kleiner, G. J., Johnson, A. J. (1966) Blood, 28, 1. Nolf, P. (1908) Archs int. Physiol. 6, 306. O’Meara, R. A. Q. (1958) Irish J. med. Sci. 394, 474. Pitney, W. R., Bray, C., Holt, P. J. L., Bolton, G. (1969) Lancet, i, 79. Sharp, A. A., Warren, B. A., Paxton, A. M., Allington, M. J. (1968) ibid. i, 493. von Kaulla, K. N. (1963) Chemistry of Thrombolysis: Human Fibrinolytic Enzymes; p. 79. Springfield, Illinois. Wardle, E. N. (1969) Lancet, i, 50. -
-
-
—
-
—
-
—
Fig. l-E.C.G. with voltage adjusted divisions.
(a) Hypokalaemia. (b) 1 month before death. (c) After cardiac arrest.
to 1 mV=10 small
vertical