·CIBA FOUNDATION STUDY GROUPS
No. 1 PAIN AND ITCH
Nervous Mechanisms
No.2 STERIC COURSE OF MICROBIOLOGICAL
REACTIONS
No. 3 CANCER OF THE CERVIX
Diagnosis of Early Forms

·CIBA FOUNDATION
STUDY GROUP No. 4
Virus Virulence
and Pathogenicity
in honour of
Prof MULDER
Editors for the Ciba Foundanon
G E W POLSTEVIČOLNE, 0.8 KM ALB., MR.C.P
and
CECILIA M O CONNOR, o se
With 13 Illustrations
LONDON
J & A CHURCHILL LTD
10+ GLOUCESTER PLACE, W1
1980

·The purpose of the Cuba Foundation, which is to promote international co operation in medical and chemical research, is symbolized in the armorial bearings by five interlaced rings representing the continents, a sacrificial cock (emblem of Aesculaprus) holding a medical caduceus, and three regular hexagons for chemistry Its domicile in London is indicated by the sword of St Paul and the British lion, the witern and the crowier, symbols associated with Basle, refer to the sponsoring firm located in this ancient Swiss toron © & A CHURCHILL LTD 1960 ·PREFACE The Cina FOUNDATION, a unique international institution, oves its inception to the generosity of CIB4 Limited, Basle However, being established under British trust law, it enjoy complete independence in practice and policy Under the guidance of its distinguished Trustees, the Foundation offers accommodation to scientists from all over the world at its home in Portland Place foremost in its activities is the organization of small conferences, some lasting 3-7 days, others for one day only The proceedings of the former are published in book form, and the latter in the manner of the present booklet Each of these one-day study groups is being organized in honour and around the work of an eminent member of the Foundation's scientific advisory panci, on which there are 1-3 representatives for most of the countries of the world The Foundation conscnes many other informal discussions between research workers of different disciplines and different nationalines and each year invites an outstanding authority to deliver a special lecture An exchange programme between French and British postgraduates is conducted and a library service us available Furthermore, the Ciba Foundation attempts in every other way possible to and scientists, whether they be Nobel Laureates or young graduates making their first original contnbution to research The publication of the first four study groups 13 regarded as experimental Nevertheless with rapid production of the proceedings of both conferences and study groups, it is hoped that readers in more distant centres, away from easy reference to the latest literature, can feel a little more closely in touch with recent developments and may be stimulated to make their own topical contributions to international research

·THE CIBA FOUNDATION
for the Promotion of International Co-operation in Medical and Chemical
Research
41 PORTLAND PLACE, LONDON, WI
Trustees
THE RIGHT HON LORD ADRIAN, OM, FR.S
THE RT HON LORD BEVERIDGE, KCB, FBA
SIR RUSSELL BRAIN, BT
THE HON SIR GEORCE LLOYD JACOB
SIR RAYMOND NEEDHAM, QC, PSA
Executive Council
SIR RUSSELL BRAIN Chairman SIR GEORGE LLOYD JACOB
PROFESSOR A HADDOW FRS PROFESSOR DR DR hc R Meter
SIR ARTHUR HARVEY CBE, M.P PROFESSOR F G YOUNC FRS
Director, and Secretary to the Executive Council
DR GEW WOLSTENHOLAE O B.E
Deputy Director
DR HNH GENESE
Astistant Secretary
Miss N BLAND
Samtific Assistant
Editorial Assistant
Miss CECILIA A1 O CONNOR, B SC Miss MAEVE CONNOR BA
Librarian
MISS LYLIANE AR TREUIL

·CONTENTS PAGE 1 3 15 20 28 34 39 Chairman's opening remarks ŞIR MACFARLANE BURNET The definltion and measurement of virus virulence by C H STUART-H IARRTS Discussion ANDREWES, BURNET, Dicx, Goffe, Isaacs, KILDOURNE, MORGUN, PEREIRA, SYTI, STUART-TIARRIS, TYRRELL Host-cell factors and virus virulence by II R MORGAN Discussion ANDREWES, BURNET, DICK, ISAACS, KILBOURNE, MORGW, NIVEN, PEREIRA, SMITH, TYRRELL. The effect on virulence of changes la parasite and host by C I ANDREWES Drucussion ANDREWES, BURNET, Dick, HENDERSON, ISAACS, KILBOURNE, MUIDER, Novev, SMITH, TYRRELL Broad aspects of the problem of human virulence la lafluenza viruses by ) MULDER Discussion ANDREWES, BURNET, ILERS, KILBOURNE, MORGAN, MULDER, NIVEY, STUART-HARRIS The severity of Influenza as a reciprocal of host susceptibility by E D KILBOURNE Discusnon ANDREWES, BURNET, HENDERSON, KILBOURNE, MORGAN, PEREIRA, SMITH, STUART-HARRIS, TYRRELL The virulence for man of some respiratory viruses passed in tissue cultures by D AJ TYRRELL and F E BUCKLAND Duscussion ANDREWES, BUCKLAND, BURNET, DICK, GOFFE, ISAACS, KILBOURNE, MORGAN, MULDER, PEREIRA, STUART- HARRIS,'TYRRELL General Discussion AA DREWES, BURNET, Dick, GOFFE, Isac, Kil. BOURNE, ALORCAN, MULDER, PENEIRA, SMITH, STUART- HARRIS, TYRRELL 43 53 58 74 78 88 93 Index 111 YH ·List of those participating in or attending the Study Group on "Virus Virulence and Pathogenicity 15th June, 1959 T ANDERSON Dept of Infectious Diseases, Brownlee Laboratory, University of Glasgow CH ANDREWES National Inst for Medical Research, Mill Hifl, London FE BUCKLAND MRC Common Cold Research Unit, Harvard Hospital, Salisbury, Walts SIR MACFARLANE Walter and Eliza Hall Inst of Medical BURNET Research, Royal Melbourne Hospital, Melbourne GW A DICK Dept of Microbiology, Inst of Pathology, Queen's University, Belfast AP GOFFE Virus Dept, The Wellcome Research Laboratory, Beckenham, Kent DW HENDERSON Microbiological Research Establishment, Experimental Station, Porton, Wilts J F PH HERS Dept of Internal Medicine, Medical School, University of Leiden A Ismacs National Inst for Medical Research, Mill Hill, London ED KILBOURNE Division of Virus Research, Dept of Public Health and Preventive Medecine, Cornell University Medical College, New York HR MORGAN Dept of Bacteriology, School of Medicine and Dentistry, University of Rochester J MULDER Dept of Internal Medicine, Medical School, University of Leiden JANET S F NIVEN National Inst for Medical Research, Mill Hill, London HG PEREIRA National Inst for Medical Research, Milf Hill, London WILSON SMITH Dept of Bacteriology, University College Hospital Medical School, London CH STUART HARRIS Dept of Medicine, The Royal Hospital, Üniversity of Sheffield DAJ TYRRELL MRC Common Cold Research Unit, Harvard Hospital, Salisbury, Wilts ·CHAIRMAN'S OPENING REMARKS SIR MLACFARLANE BURNET It is an honour and a pleasure to be asked to chair a discussion in honour of Prof Mulder, particularly on a subject which he has made especially his own The central interest of our discussion is the pathogenicity and virulence of influenza virus, with other viruses included, mpre or less to throw light on that theme of Prof Mulder's Among the themes in microbiology which have human implications the virulence of viruses 13 probably outstanding The influenza pandemic of 1918 sas, apart from the classical bubonic plagues the greatest pesulence that has ever afflicted the human race and most of us were impressed recently by the picture of the age incidence of the 1957 Asian influenza in America It seems that we may have just barely escaped another 1918, because there was a Lveryone here will agree that we are up against a major difficulty 20 Calking about virulence It is not an absolute character It is wholly dependent on the relationship between host and parasite, and both will be considered by speakers bere My own special interest for the last ten years has been the genetic aspects of influenza virus There is no doubt that we can detect differences in virulence amongst strains acting on the same host, and that these differences in virulence are inheritable and hence genetic characters During the last ten years I have attempted to define in genetic terms what is meant by virulence in influenza virus at the expenmental level in fact this was the subject of a Csba Foundation Lecture which I gave a few years ago (1954, Lancel, 2, 559) But there 13 no simple answer There are no virulence genes as such, and my present feeling is that virulence must be considered as a sort of integration of all the characteristics ·2 SIR MACFARLANE BURNET of the virus which are related to its capacity to infect and multiply in the cell, and therefore must always be expressed in terms of the in mind that the genetics and the population dynamics of any pathogenic organism are determined essentially by the way in which the virus survives as a species and not by its human pathogenicity Virulence is, as it were, an epiphenomenon of the pro. cesses by which the virus survives in Nature in relation to the full totality of the environment, and where difficulties begin to arise in discussion of the present kind, we can often clarify them by using thus ecological approach ·THE DEFINITION AND MEASUREMENT OF VIRUS VIRULENCE C. H. STUART-HARRIS Department of Medicine, Untersity of Sheffield Virus virulence is almost indefinable and practically impossible The fact that we have a better word than virulence to describe the means of production of disease by viruses should surely also scan that it is not a useless symbol or abstract concept. The attribute of being poisonous which is the factual meaning of virulence becomes all the more real when substances, organisms or viruses are found wluch do not possess this property The fact that some viruses are non-virulent is probably the chief reason for attributing variations papers in this study group In truth virulence is an attribute of a micro-organism which can only be measured and defined by reference to a host system The definition of virulence It 19 first necessary to consider whether the word virulence should or should not be regarded as synonymous with pathogenicity. Most workers do in fact use the words in an alternative sense, but Miles pointed out in 1953 that it was useful to regard pathogenicity as an attnbute of a species, genus or some other grouping of parasites and to reserve virulence for the expression of the pathogenicity of a given stable line of micro-organism. There are good grounds for agreeing to this distinction in the case of viruses particularly because of the very wide variations in the behaviour of different strains of many anunal viruses. With this background it is possible to define pathogenicity as the power to 3 ·4 CH STUART-HARRIS produce pathological effects in a host, and virulence as the evidence of pathogenicity denved from observation of the symptoms and signs, degree of illness or death of the host. The measurement of virus virulence (a) The incidence of disease Attempts have been made to measure virulence with varying degrees of accuracy by studies of whole populations, of individual patients or by experimental study in the laboratory upon animals Let us first examine the facts obtained from a study of human ill ness in whole populations There are certain characters which can be measured in a population, such as the incidence and mortality from an infectious disease, and these are often used as indices of the degree of virulence of the infection It is, however, obvious that these indices cannot be divorced from the effects of age, for age introduces a variability into the equation and prevents a direct comparison between two populations unless these are idenscal in structure KT that immunity following childhood infection persists either by means of antibodies or by some change in the cellular resistance to attack When a population which has been previously unexposed to an infectious illness becomes infected the age incidence of disease is said to be quite different from that of the acute specific feters, children and the older adults may suffer less severely than young adults (Burnet, 1953) Thus, the age incidence does not give a clear picture of host resistance unaffected by the specific effect of previous contact with the infectious agent When, therefore, a disease changes its age incidence it may be difficult to say whether this change is due to a variation in the host or in the parasite Poliomyelitis may be chosen as the example Here 13 a striking example of altered attack rate and of age incidence in our own lifetime For many this is only reasonably explained by invoking an alteration in virulence of the poliovirus Others, however, point to serological differences between populations which have no epidemics of poliomyelitis and those in which the disease is epidemic The later development of specific antibodies in the ·DEFINITION AND MEASUREMENT OF VIRULENCE 5 children of the populations experiencing epidemics points 10 a lack of exposure to the virus in carly infancy which would bring about herd immunsty There is then no need to invoke an alterauon in the sirulence of the visus to explain the evolution of polio. myelitis at the present time Polsomyeliuss, however, illustrates another aspect of age which 19 of great significance This is that with increase in age there is increasing sesesity of sllaess and of mortality This is seen cicasly by reference to statistics from this country or from a populatiobs such as the Eskimos sho hase infrequent cpidemics (Peart, 1949) Sabin (1956) explains this by suggestsng that the matar neurone of the adult is more susceptible to attack than that of the child In typhus leser, too, mortality is not related to suscepubility to infection but is clearly correlated with age In the Naples epidemuc in 1943-44 the largest number of cases was in children and young adults in whom the mortality was about 5 per cent (Table I) In those arer 30 years of age mortality rose sharply even though the actual number of cases was less (van den Ende et al, 1946) Even in the Algerian Arabs among whom previous exposure to endemic typhus might kell kase modified severity, the same trend of monality with age occurred as in their European Ocighbours (Table II) One can perhaps etptan thus increased mortality from typhus in older persons as being due to the falure of the ageing circulatory system to resist the attack made upon it by the rickettsia So far as the respiratory tract 13 concerned it is a famuliar fact that young infants and the aged are subject to the highest mortality and young adults the least Exeo 3 an area where mlection has been absent for many years and is then reintroduced the age distribution of mortality may follow the same pattern Thu is uell shown by the measles epidemic of the Faroe Islands in 1846 which mturred suxty five years after the previous epidemie (Table 1) The incidence was the same for all ages, being 77 per cent of als those at risk who did not have the disease in 1781 But the mortality skas greatest (28 6 per cent) la infants under 1, next highest in adults oves 80 (16 3 per cent) and negligible in those from 1 to 30 years of age (Panum, 1940) Of course, this mortality must sepresent the combined effects of virus and nasopharyngeal bacter ia on the respiratory tract ·5 Table 1 THE NAPLES EPIDEMIC--EEX AND AGE AND MORTALITY DECEMBER 1943-FEBRUARY 1944 Death Females Deaths Total Απόσης Mortality Deaths per cent 18 Age groups Under 3 3-11 12-20 21-29 30-33 20 133 221 105 113 59 2 9 11 26 1 3 49 $ 30 10 18 28 39 15 3 3 19 21 32 57 37 22 166 108 127 116 56 27 8 1 718 38 224 387 213 238 175 99 37 10 2 1,423 CH STUART HARRIS 13 4 32 5 48-56 57-65 6674 75 and over Gross 29 18 2 2 1 10 2 1 500 100 705 94 2 199 fos 13 9 ·Age groups Mortal ty per cent 0-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80 and over Gross Tablet COMPARISON BETWEEN TYPHUS IN THE LONDON FEVER HOSPITAL 1848-1870 ALCIERS 1942 AND NAPLES 1943 1944 Algren 1942 London Naples 1848 1870 French Arabs 1943 1944 Mortal ty Mortal ty Mortal ty Aumbers per cent Aumbers per cene Numbers per cent Age group: Aumbers 1 430 4 20 100 76 6 6 0-11 262 5121 3 5 72 70 317 73 12 20 387 4 127 12 3 62 17 8 620 92 21 29 213 2 976 23 2 79 329 368 18 4 30-38 238 2 546 35 5 79 44 5 246 26 0 39-47 175 1 231 51 1 73 67 2 111 50 S 48-56 99 588 65 1 57-65 37 116 75 8 0 66 and 12 3 100 O O er 18 268 18 92 35 8 1 846 1 423 1 S 49 98 13 4 32 5 36 3 594 58 3 DEFINITION AND MEASUREMENT OF VIRULENCE 6 457 16 8 13 9 ·CH STUART STARRIS Tebic m FAROE ISLANDS BEASLES EPIDEMIC 1846 Age groups Under 1 year Death 4+ 3 Case forsity per cent 28 6 03 03 Papulation 199 1 410 1 525 1 470 842 791 723 480 272 118 7,86 30-19 20-29 30-39 40-49 50-59 60-69 70-79 80 and otet TOTAL Number attacked 154 1117 1 183 1 140 653 613 565 372 215 92 6 100 10 19 27 19 31 48 7) 90 16 3 370 Influenza is a disease phose age variatsons in different epidemics are regarded as classical examples of the effect of changes in the virulence of the causate virus Clinicians faced by the asse inspiring spectacle of the winter pandemics of 1918-19 Kere fara with the former innocence of influenza in young adults. They were therefore bound to explass the excessive mortality in young adults by postulating a change in the virulence of the infecting virus let Francis (1953) has sought to interpret the changed mortality in 1918 as an expression metely of the intensity of infection In his view the sparing of older persons from a hsgh attack rate was due to previous experience of the virus antigen which young adults lacked Francis suggested that the case mortality from influenzal pneumonia increased steadily with age in 1918 just as it d d in epidemses before or after the pandemic Even if we were able to accept that the 1918 pandemic caused an exceptionally hagh incidence of infection in young adults the secent experience of the Asians pandemic of 1957 should surely make us hesitate to follow Francis' opinion for sa 1957 the high attack rate in children and young adults experienced all over the world was not accompanied by any comparable incidence of pulmonary complications or mortality Thus the remarkable prevalence of pneumonsa 15 young adults in 1918 13 a phenomenon which cannot readily be explained except by reference to the biological propertses of the pandemic Virus No doubt members of this study group wel wish to refer again to this epidemic exper sence and sill give their own explanations ·DEFINITION AND MEASURESIENT OF VIRULENCE 9 (0) Observation of individual illnesses As it seems to be difficult to are at an estimate of virulence from the study of whole populations, let us non examine the evidence derived from study at a clinical level on individual patients This is the technique used by those who seek to estimate pathogenicity by the use of human Volunteers The latter are never $0 numerous as to constitute a population, and consist merely of a handful of individuals Something indeed can be learned by this method, as we shall hear later in this study group, but the technique is extremely limited in application Meanwhile, clinical studies have shown the importance of host factors Other than specific or immunological resistance which determine the outcome of infection Nutritional status is widely believed to be such a determinant particularly in the case of bacterial infections such as tuberculosis (Dubos, 1955) and Salmonella infections (Schneider and Zinder, 1956) Ict the findings in the case of virus diseases seem to be contrary to those of bacterial infections If it is true, as it seems to be that the bonniest children often fall victims to poliomy chius, is this an argument that those who are undernounshed are thereby protected from paralysis? Social circumstances are of such great importance in determining the spread of poliovirus and thus of aiding subclinical immuniza tion that one must beware of drawing hasty conclusions But the suggestion has certainly been made and there is no casy way either of refuting or of confirming it from human experience that some virus infections are more trivial in the under than in the over-nourished host Until quite recently the importance of the endocrince on the outcome of infections was minimized Now, however, cortisone has come along and it is possible to observe the effects of steroid administration on man as well as on animals with experimental infections Does the enhancement by cortisone of poliomyelitis in the hamster (Shwartzman 1950) or of Corsackie virus infection in mice (kolboume and Horsfall, 1951) have human analogies > One can point of course to the frequent severity of poliomyelitis in pregnancy, though this is not entirely undisputed Perhaps Flaggerty and Eley's observations (1956) on the excessive severity of varicella in children treated with steroids is analogous to the results with viruses in cortisone treated hamsters and mice Yet these are rare human happear and one would indeed be bold 원 ·10 CH STUART HARRIS suggest that endocrine changes are of much cancer in infuencing the outcome of the common virus infections Perhaps it is ex en more important to stress that hast tesistance 13 far more complex than this reference to nutntion and to metim bolism would suggest The clinicia now has it in tuis poner to amekorate or to prevent infection as when he uses gamma globulus to attenuate or to abort measles Unfortunately virus infections appear to involve a form of resistance which is not huronal or else children with agammaglobulinzerria irould perish from the common cold or influenza. The studies of Good and Zak (1956) and of AfacCallum (1959) on agammaglobulutaemia skot quite clearly that most virus infections and even vaccination take a normal course in children stripped of the barrier of specific immunological protection. So, like the studies in populations, clinical studies afford only a glimpse of the relative virulence of viruses and one turns with hope to the experimental study Jence m the laboratory, (2) The experimental measurement of virulence in the laboratory The first matter to be decideal by the virologist is the host species «kich will be employed. This is because the adaptation of a virus to a new host is linked with the acquisition of virulence for the Ticw species and perhaps wsth a loss or attenpation of virulence for the former host species Yet mere choice of species is not enough A1tles (1955) when referring to the measurement of virulence in micro-organisms stated that one can only determind the comparatie effect of two organisms or two strains of the organism in the samne set of lasts He inferred that there can be no such things a quantitative expression without a standard of srference Funhermore, the comparison of virulence is only possible if one can measure the dose or quantity of virus or other organism present in the preparations inoculated into the hosts This matter of quantitation 15, of course, the great stumblingblock for iologists No matter whether sve use 50 per cent afectious doses, or haemagglutinating units, or some other measure, it has to be admitted that it is far lesa precise than one would utsk. Moreover, with attenusted strains of virus one is dependent upon mess of estimating infectious units in some system other than the host ander study ·DEFINITION AND MEASUREMENT OF VIRULENCE The next requisite, when the host and the unit of measurement of the virus have been decided upon, 19 to define the set of condizons under which variation of the response of the host will be ninimized It may be necessary to define the genetic constitution of the host in addition to the species This was clearly shown by Dabın (1952) when he discovered that the Rockefeller Institute breed of mice are insensitive to the 17D strain of yellow fever virus egardless of dosage Swiss mice are, by contrast, uniformly sus cpuble to this strain of virus Sabın found that the resistance of he PRI mice also extended to other neurotropic viruses and erlained the lack of virulence of 17D virus as being due to the low evel of multiplication achieved by this virus in the mouse brain tle further showed that suckling mice of the PRI strata were suscptible to 17D virus up to four days of life, and age is certainly a actor of the greatest importance in host resistance in the laboratory As in man One thinks, of course, of the Coxsackie viruses as being the best example of limitation of pathogenicity by the age of the host, but an enhanced susceptibility of young animals is found with many viruses (Sigel 1952) Clearly, age does not matter when virulence 1. The route of inoculation is the third factor which requires standardization Route does not necessarily matter if one 13 studying viruses capable of a wide dissemination throughout the body but it 18 vital in many other instances because of particular tissue tropisms Influenza virus affords excellent illustrations of the importance of the route of inoculation in studying the virulence of different strains The need to bring the virus into direct contact with the cells of the respiratory tract or of the nervous system in the case of neurotropic strains is only obviated when using baby mice The fact that subcutaneous influenza virus does not initiate infec tion in adult mice prevented me for some time from recognizing that I was dealing with an instance of infection by the neurotropu WS virus in suckling mice I then found that the virus coul produce encephalitis with nervous system lesions and death up to ·12 CH STUART-HARRIS about the twelfth day of life Tyrrell and Cameron (1957) later showed, however, that lung adapted strains of influenza virus such as WS and MEL could also kill suckling mice after subcutaneous inoculation though lung and not brain lesions were then produced Clearly some form of barner develops with maturation of the host ferent ages (Sabin and Olitsky, 1937) Finally, the effects of the viruses on the host must be measured in a standard manner Mortality is obviously preferable to lesions unless the latter are sufficiently characteristic and can be graded quantitatively Unfortunately, lesions frequently give only crude indices of effect and estimates of the degree of virulence of attenuated viruses are correspondingly approximate Practical problems concerned with the measurement of the virulence of the poliovirus The importance of all these matters concerned with the attempted measurement of virus virulence can best be appreciated by reference to practical problems such as the use of virus vaccines for human immunization This will be illustrated here from the standpoint of poliomyelitis, which has in the recent past caused virologists to think a great deal about the definition and measurement of virus virulence The drive to produce a living attenuated poliovirus vaccine which can be used safely in man requires an estimation of virulence in two connexions First, it is necessary to find strains of poliovirus which cause no harm when fed to man via the alimentary route Secondly, it remains to be shown that the living viruses excreted by those to whom the vaccine strain 13 administered, are not hazardous to those in contact with the vaccinated subjects Both needs have been explored by means of comparison of the properties of the viruses in monkeys and chimpanzees inoculated by various routes Both of these practical problems have led to controversy Of the three variables mentioned already in the preliminary dis cussion, the choice of host has not been in doubt for most workers Rhesus or cynomolgus monkeys are the only practicable test species for determining the virulence of the viruses for the central nervous svstem Mice are only useful for Type II strains and chimpanzees ·DEFTYITION AND MEASUREMENT OF VIRULENCE 13 are too precious This means that most work on the measurement of the virulence of poliovirus has been simply a study of virulence for the monkey often on only small numbers of animals Now Sabin (1956) has pointed out that the virulence of polioviruses for the monkey is not an all-or-none character but that viruses densed from healthy children or after various laboratory manipulations can TE most virulent strains can reach the central nervous system after penpheral inoculation, others only produce an effect when large doses are inoculated directly into the brain, and the most attenuated viruses produce no paralytic effects after intracerebral injection but may in very large doses still paralyse the motor neurones of the spinal cord after intraspinal injection into the lumbar enlargement According to Sabın esen strains capable of causing some paralysis after intracerebral injection into monkeys are without effect after intraspinal injection into chimpanzees In his recent publication, Saben (1959) shows the relative effects of various Type I polioviruses after intraspinal inoculation into cynomolgus or rhesus monkeys If it is thus possible to measure virulence with such apparent precision, it should be equally possible for different observers to obtain good agreement between their results Unfortunately this is not true and Sabin considers that the discrepancy between different observers is explained by the fact that reproducible results demand scrupulous attention to technical detail particularly with intraspinal inoculations A second discrepancy confronts us, however, in regard to reported properties of the viruses excreted in the stools of voluateers fed attenuated viruses by mouth. In Belfast, the attenuated Types I and II viruses of Koprowskı gave rise to excreted viruses with power to paralyse monkeys after intracerebral inoculation (Dane et al, 1957, Dick et al, 1957) In Sheffield, Sabin's Type III attenuated virus was fed to children, some of whom became infected and excreted virus in the stools Comparison of the virul. ence of the excreted virus with the original vaccine strain led Clarke and her co-authors (1958) to conclude that the viruses had attained enhanced neurotropism for the monkey during passage through the alimentary canal Sabın himself (1957) has found the ·14 CH STUART HARRIS virulence of excreted vinises to be perceptibly different from that of the vaccine strains but considers that the quantitative change is not such that the viruses constitute a hazard for man He bases this deduction on the results of intraspinal inoculation into chimpanzees which suggest that the strains with residual paralytic properties in monkeys are harmless when given intraspinally to chimpanzees Nevertheless the fact that there is some sort of arly under the circumstances of infection of the human alımentary tract This over a period of weeks resembles a series of passages in a culture in the laboratory and thus affords the maxual oppor tunity either for genetic variation or for selection of strains with particular properties tonnenfonlayin eth 18 TI 1 because of this efforts have been made to estimate virulence in a host system capable of more rigid control such as is possible in tissue cultures This brings us to a discussion of virulence at a cellular level Virulence at a cellular level The present phase of virus work has given impetus to the study 1. h 1 - - systems to provide markers which can be equated or even sub stituted for tests in animals is refreshing It accounts for the dis covery that attenuated polioviruses may multiply to a different extent from virulent viruses in cell cultures held at a certain pH (Sabın 1956) Vogt Dulbecco and Wenner (1957) have indeed suggested that growth of virus in cells at different acidities may be a character which is genetically linked with that of the virulence of poliovirus for the monkey Kanda and Melnick (1959) have also ·DEF2\tTION AND MEASUREXTENT OF VIOLENCE 15 shown that violent and avrulent poliovirus strasns may grow to differing utres in different cell lines. Yet it annot be sard that tissue culture systems can be substituted for whole animal hosts in the measurement of virulence, for there is a wide gulf between the effects observed is tuiso and I titorith the same virus A given strain of influenza virus may have a high order of virula ence for the mouse lung in living mice but may not attack the cells from a mouse lung grown in tissue culture Again environmental factors of almost trivial importance appear to affect the destructive action of viruses on celis Thus, Frothingham (1959) has reoently showa that gronth of the egg-adapted Type II, MEF strain of polovirus in buman amnion cells as much affected by the age of the culture and by motson of the tubes The virus is cytopathics old, but not in young, cells and in rolled tubes but not in stationary cultures The profound effects of alteration in the brochemucal environ ment of tissue cultures on the resultant growth of viruses us known through the work of Prof H R Morgan Ile has shown how one can convert cells in 1533ue culture into ressstant hosts for pssssacosis virus by starving thetss of essential nutrients (Johnson and Morgan, 1956) It would be necessary to control all conceivable factors affecting cell growth 17 ctro if a stable reproduciblc host system were to be elaborated Peskapast nas knowledge of the modifying effects of environment on the host purus relationsług which led the virologists attending the Wisconsin symposium on Jatent virus Infections in 1957 to refer to the destructuye effects of viruses on cells in culture 2 "cytocdal". They resered 'vrulence" for observed pathological cffects in whole animal hosts Ought we to do the same? DISCUSSION Tyrrell I would add the suggestion that in a laboratory study one should always use, if possible & single genetic clone or group of nous particles, because it is probable that one of the vitiating factors 17% this work is that we are dealing with a mixture of different kinds of viruses He do one test and measure the properties of ons component, the next time ne do another test and measure the properties of another component of the mixture Burner I should like to underline that suggestion very strongly Morron Twas also true when you have the advantage of doing in fro expensest 1 nes of ·16 DISCUSSION -L you start ou with a dillerentiated kidney cell, by the end of a tety days it 19 rather different, and, as Frothingham (1959) showed, it goes on L JT - 1 yrrell No Kilbourne Lest we deprecate the intact host too much, we must remember its built in " homeostasis which we do not have in tissue culture Goffe There is also the disadvantage that the virus as excreted by the intact host is not a single clone If you want to examine the virus Ann polate from tissue cultures to the whole hosts 1111957 Li Lidproll iere and I studied the susceptibilty to myxoma virus of guinea pig human and other tissues in culture We found that cells of all kinds of spectes were susceptible At that time we found in the literature only one or two other examples of a non susceptible species whose cells would grow a virus in tissue culture Since then it has become Virus Smith One should also be cautious about accepting these so called clones as being relatively uniform How quickly do they change? From work we have done it seems that in spite of the use of what is hamaronen ove population I am quite sure that is not so YS do mal ·DISCUSSIOY 17 heterozygous Iara confident that many influenza virus particles are heterozygous I carrying two or more alternative genomes It has yet to be prored whether or not the poliovirus ts heterozygout That introduces a major difficulty into the question of producing and maintautung pure clone virus no . i. brought the disease home la homes where old people and voumg children had a chance of mixing, the old people got influenza just ** .. undehned Duk Is the actual quantity of virus multiplication known? We know that, in poliomyelius, the virus multiples better in small children than in adults given the same dose of vrug Does that ·HOST CELL FACTORS AND VIRUS VIRULENCE HERBERT R MORGAN Lous A Il ehle Virus Research Laboratory, Department of Bacteriology, University of Rochester School of Medicine and Dentistry, Rochester, Nero I ork In the interaction of a virus with its host cell, both members of this complex have an important role in determining the outcome which is a measure of the virulence of the virus, and each in turn 13 subject to factors in the microenvironment in which thus contact occurs In considering the role of host cell factors, it is desirable to minimize the role of environmental factors such as antibody and therefore a restriction of this discussion primarily to in entro systems has many advantages The in vitro system permits the study of a given virus with a defined cell system in which all of the cells are subject to direct observation and the extracellular environment is subject to control Also the in otro system makes it possible to simplify the definition of virulence which will be taken to mean any visible effect produced on the host cell which will include not only cell destruction (cytopathogenic effect, CPE) but cellular alterations associated with continued cell replication as produced by certain tumour viruses Our growing knowledge of these latter agents requires that they receive recognition as an important group of viruses that possess certain unusual characteristics, the most striking of which is their capacity to stimulate the unrestricted multiplication of the host cell 11 tivo Admittedly their effects on cells in vitro are not compatible with the more classical concept of cytopathogenicity based on cellular destruction, though under certain circumstances they may exhibit this action in tissue culture The interaction of the viris with its host cell may have several general forms (a) the virus may multiply and produce a cyto pathogenic effect (CPE), or (5) at may produce a CPE without the production of infectious virus though there may be evidence of the synthesis of viral components by the host cell In both of these 20 ·1 HOST CELL FACTORS AND VIRUS VIRULENCE 21 cases, the virus would be considered prulent in tytro On the other hand, the tus may (c) multiply without observable effects on the cell, or (1) fail to muloply and to produce an effect on the cell and be considered avrulent in ciber CASE The failure of completion of the infectious cycle by a virus an contact with a cell may occur at a number of different phases in thss cycle, 1c, adsorptsos, penetration eclipse, veral replication and release Cellular factors say influence most of these events in a decisive manner Hostcell factors io viraladsorption, penetration and celipse The susceptibility of freshly isolated cells to virus infection and subsequent destruction has been ettersively studied with polio. virus infection of monkey kidney cells Kaplan and Mielnick (1955) showed that kidney cells from the South American capuchun monkey supported the granth of Type 1 poliovirus without evidence of a CPE and thus the visus might have been thought to be avirulent for these cells However, it was shown that the cell population adsorbed only a small part of the virus inoculum (kap. lan 1955) and that only about 1 cells 2000 was suscepuble to visus infection These few infected cells accounted for the viral replıcafios Obviously the lack of a CPE was due to the fact that the des truction of a small number of cells in the large populatesn could not be detected Failure to adsorb virus rendered it avsrulent for the majority of cells However, the adsorptune capacity of the cell for poliovirus is not the only factor important in the susceptiblity of kidney cells from different monkey species, sunce paras and thesus monkey renal cells adsorb virus equally well, but a larger number of plaques appear on the patas cell monolayers This indicates a greater degree of susceptibility of patas cells which is due to sone event in the infectious cycle other than adsorption of virus (Ilstung and Melnick, 1958a) Other studies of a similar nature have been carried out using estabushed cell lines a continuous culture Usung Type 2 polio- Virus at tras found that resistant cells (rabbit kidney and strais L) adsorbed virus much less efficently than susceptible tells (monkey kidney and HeLx) and that the virus adsorbed was more firmly bound to the susceptsdle cells and appeared to enter the eclipse phase fairly candlu Silvarkers Wahet Ant Tabamata ini memang ·22 HERBERT R MORGAN loosely bound, since egg adapted mumps virus binds to L cells and fails to elute, though it produces no detectable effect on the cells and does not multiply (Bader and Morgan, 1959) A systematic study of this difference in susceptibility of various mammalian cells to poliovirus confirmed the fact that a number of susceptible strains of primate cells, including HeLa, strongly adsorbed poliovirus, while insusceptible non-primate cells did not and the virus that did adsorb to non-susceptible cells eluted spontaneously (McLaren, Holland and Syverton, 1959) Furthermore, the virus adsorbed to the non susceptible cells failed to enter the eclipse phase Additional studies (Holland and McLaren, 1959) revealed that the capacity of HeLa and possibly the other susceptible cells to adsorp, recere and eclipse polovirus was dependent on the presence in the cell wall of a specific cellular receptor which J. L TL 11 S non mi ha inand .h. At L-AL AM infectious virus L f 11. hy virus show a sharp reduction in infection rate when treated with the receptor-destroying enzyme (RDE) In the living chick egg, these receptors are regenerated with the passage of time and susceptibility to infection returns The discovery that the mucus secreted by respiratory epithelial cells contains a material sunılar to the receptor substance, which may bind virus and thus render it incapable of infection, provides evidence that two host cell factors have a rôle in the determination of viral virulence, 1 e a specific cell wall receptor and a cell secretion

·Host CELL FACTORS AND VIRUS VIRULENCE
23
Host cell factors in viral replication
A number of investigators have reported that following the
exposure of a large number of cells to a given virus which ordinarily
produces a CPE on these cells, a few may survive and multiply
even though virus may persist in the culture In the case of polio
L
od ) non
not due to differences in their capacity to adsorb polovirus but to
the subsequent probability of infection and cell killing by an
adsorbed virus particle
These experiments demonstrate that populations of continuous
cell lines are composed of members of a ferent na
die werelore resistant
In other investigations in which large cell populations have been
van de

·24 HERBERT R MORGAN that at a given time only a proportion of the cells were in the proper physiological state to undergo infection However, prolonged cultivation of the carrier cultures tended to select cells with in. creased resistance to Coxsackie A9 infection In studies with a persisting infection with NDV in MCN cells, Deinhardt and coworkers (1958) also came to the conclusion that, at any moment, 1n . CsL .La The author postulated that the presence of the serum produced an optimal physiological state of the cells in which they could carry the burden of viral infection without a CPE However, the possi bility that this was due to antibody in the serum could not be eliminated even though rabbit antibody did not produce a similar C.LOC with psittacosts virus demonstrated that the maintenance of chick . . namant nn and develonment of the eclipse pilase vul multiplied when the missing amino acid and horse sei um hele added Such omission of amino acids from the culture medium ·Host CELL FACTORS AND VIRUS VIRULENCE 25 has been shown to result in their disappearance from the free amino acid pool of the cell (Pict and Eagle, 1958) Studies with poliovirus infection in HeLa cells (Darnell and Eagle, 1958) showed that cells depleted of vitamin B, farled to support viral growth and that this was due to a disappearance of certain amino acids from the free amino acid pool from which the Virus was synthesized (Levintow and Darnell, 1959) Herpes virus infection was also altered in a similar manner since it infected L cells but failed to multiply when amino acids and water soluble Vitamins were absent from the culture medium (Pelmont and Morgan, 1959) Type 2 adenovirus produced a more rapid destructive effect on hB cells when the concentration of arginine in the medium was increased 10 times (Bonifas and Schlesinger 1959) An increase in virulence of the virus for these cells could be pro duced by raising the concentration of arginine It is of interest in this connexion that Mills (1958) obtained evidence that the requirement for certain amino acids of the cells of the chorioallantois of de-embryonated eggs 19 increased when they are infected with the Lee strain of influenza B virus An adequate nutritional state with reference to water soluble vitamins and amino acids is a necessary condition for the physiological competence of some cells to support the growth of certain viruses which have entered them and hence one type of intra cellular host factor determining virus virulence appears to be established A decrease in the concentration of bicarbonate in the medium v ntarforh.hr 1 assume that this was solely an effect of pH on the host virus interaction, but adjusting the pH of the medium with NaOH did not influence plaque development sith poliovirus (11siung and Mel nick, 19580) although virus growth is reduced at low pH Furthermore, in the investigations on the role of bicarbonate in growth of Corsackie B-1 and polioviruses in HeLa cells the bicarbonate defir enn mit het L ·26 HERBERT R MORGAN pH effect in phases of cellular metabolism necessary for poliovirus 1 proteins, presumably enzymes, which are necessary for virus synthes13 Perhaps the earliest observations on physiological factors and host susceptibility to virus infection in vitro are those of Enders and Pearson (1941) who reported that tissue cultures of chick embryo lung supported the growth of the Melbourne strain of influenza A virus at 37° but not at 41°, though other influenza virus strains grew at this temperature (Colville, Dunbar and Morgan, 1956) Growth of polio virus has also been shown to be reduced at a temperature of 39' (Lwoff and Lwoff, 1958) Wheeler (1958) showed that a change in the temperature of incubation of the host cell from 28° in successive steps to 40° had a significant effect on the multiplication of herpes virus in HeLa cells with a maximum virus yield at 35º Furthermore, cells incubated at 28° and showing little or no evidence of virus reproduction promptly produced increased amounts of virus when the incubation temperature was raised to 35° In other studies with influenza virus it was shown that at 25° the virus did not grow in the cells of the allantoic sac of the chick embryo, but soluble antigen and haemagglutinin accumulated in the cell and transfer of the infected eggs to 37° led to a rapid appearance and release of infectious virus (DeSanctis and Liu, 1959) The temperature of the host cell therefore is an important factor in viral replication, and if it is low, though virus may undergo incomplete replication, it does not attain the infectious form and therefore fails to manifest virulence ·HOST CELL FACTORS AND VIRUS VIRULENCE 27 limited replication of the virus which in most instances destroyed the cell However, all cells infected and undergoing this partial growth cycle were not destroyed A similar situation has been described with NDV infection of Ehrlich ascites tumour cells in which the virus penetrated the cell but failed to complete its growth cycle though the cell was destroyed (Prince and Ginsberg 1957) In these instances some host cell factor prevents completion of the growth cycle and thus influences viral virulence, for even if the cell 18 destroyed the viral components released are not infectious and therefore not virulent Other viruses such as mumps may destroy cells in eitro without detectable viral replication (Henle, Deinhardt and Gırardı, 1954) . Morgan, 1950) This viral property which may be related to virus virulence (Morgan, Soule and Marinettt, 1959) 19 dependent on host cell factors for its manifestation Host cell factors and tumour viruses Before concluding this discussion, it is important to consider host cell factors that influence the virulence of tumour viruses, for in this case virulence is not manifest in urto as a destructive action on host cells (with the exception of the haemorrhagic disease pro duced in chick embryos by Rous sarcoma virus) but rather as their conversion into a new host cell virus system which perpetuates itself and is associated with unrestrained multiplication of the cells In titro however the tumour viruses may demonstrate either the capacity to transform normal cells to a malignant state or to produce a CPE on the same cells An analysis of this pheno wan 10 subcultures could be made from these degenerating cultures and vigorously growing, apparently healthy cell colonies were obtained which continued to produce Rous sarcoma virus, but these in turn ·28 HERBERT R MORGAN degenerated None of the cells demonstrated an increased pro Jiferative capacity. It was later discovered (Temm and Rubin, 1958) that if the chick embryo fibroblasts transformed by Rous sarcoma virus were maintained under better nutritional conditions, the transformed cells did not show a CPE, but multiplied and could be grown for twenty generations Thus, the physiological state of the host cells determined their response to the virus, 1c + Ch. It was also discovered that at any one time only 10 per cent of the chick embryo cells were susceptible to infection by Rous sarcoma virus even if a clonal population was used, and the investigators believe that the physiological state of the host cell is of importance in deciding whether or not it is competent to be infected by Rous sarcoma virus (Temin and Rubin, 1958) In the case of Rous sarcoma virus, the physiological state of the host cell may determine whether or not the cell can be infected and which two diametrically opposite results se cell death or cell multiplication, will occur DISCUSSION Tesars One of the factors in cellular resistance may be the inducdue to the production of a substance whicu ley II inhibitory factor) That substance seems to be very similar in nature -tortls nfected 1 L multiplying? Aſorgan No Isaacs They produce virus without multiplying ? Morgan These cells so far, do not multiply in such a simple synthetic medium without the introduction of serum I have never

·DiscussIOY
29
yet been able to make the cells multiply Kichout the productson of
vitus These cells depleted of a certain single amino acid will remain
19
L
out & long time ago the case of pneumococcs and their pathogenicity
for the rabbit--that you could correlate this very neatly with
HAN
34
1

·30 DISCUSSION muth ar Chhh... Aſorgant If you select out the heat-resistant variant you will get a pathogenic effect also at 40°. Andrewes Dr Isaacs has suggested that in tissue cultures of LL the interferon theory used Kilbourne Is it clear that the virus is intracellular in those experiments ? Morgan It is probably multiplying at or near the cell surface Burnet Did those resistant cells maintain their resistance on further cloning? Morgan I think so Burnet This would be crucial for deciding whether it is a physiological or a genetic variation al

·No text on this page.

·32 DISCUSSION particle At any rate, under these circumstances found in the in. fected, depleted cell, for a period of three weeks nothing is happening in those cells except perhaps the replication of these large bodies Now if they were multiplying we ought to see them because these Han. A are easy to recognize in the electron micrnerana 1.11 La U40L WLLA carrying out over several years, and more recently in tissue culture material, I have noticed a sery constant relationship between the nucleolus and the nucleus during the stage preceding what is usually regarded as "cytopathic" change, nucleolar enlargement being the - along those lines At the submicroscopic level a vast amount of material would have to be exaruned before one could be certain whether or not significant changes exist in such altered cells Morgan Alterations have been reported in some of the carrier cells, with the poliomyelitis virus, in which you can produce the cell These cells are altered in that even though they carrier state ! 1 . naudade 10 A اه Tь , influenza visus Pereira When one considers the parallelism between virulence 1 ·DISCUSSION 33 pathogenicity for man Types 1 2 5 and 6 are the viruses which most frequently cause latent infections whereas Types 3 4 and 7 are the ones more commonly associated with epidemics in man These to groups formed on the basis of pathogenicity for man can also be grouped by several laboratory properties and an investigation of possible reasons for this grouping might lead to some interesting Pereira In that respect taking the adenoviruses the type of cytopathic effect of Types 2 5 and 6 for instance is remarkably dif L JT 3 2 ·THE EFFECT ON VIRULENCE OF CHANGES IN PARASITE AND HOST C. H ANDREWES National Institute for Medical Research, Mill Hill, London NOWADAYS we all accept Theobald Smith's dictum that the natural tendency of a host-parasite relationship is towards a mutual tolerance which will permit a survival of both partners True, the association is at times an uneasy one, liable to be disturbed by factors of various kinds Rather than dwell upon thus, however, let us consider the cases in which some degree of virulence favours survival of a virus, by facilitating its spread to fresh hosts Respiratory viruses Viruses which are spread from the respiratory tract would seem to have a better chance if the infected host coughs, sneezes or otherwise distributes excessive secretions Ameasles or a common cold virus which was completely avirulent and caused no coryza would be expected to spread very poorly This sounds entirely reasonable nevertheless it is not sure that this is always true In one experiment at Salisbury on transmission of the common cold by contact, the only cold amongst the recipient normal volunteers was in one exposed to an infected volunteer in the incubation period of his cold and as yet without any symptoms On the other hand, influenza spreads very successfully only when it is fairly virulent We know that inapparent immunizing infections occur, of anale low, We and sneezing almost certainly helps the spread of some respiratory viruses, even though others can spread at times without the aid of this expedient Enteric viruses The many viruses causing diarrhoea—in man, mice, pigs or cattle--are all, so far as we know,excreted in the faeces, they surely 34 ·EFFECT OF CHANGES IN PARASITE AND Host 35 bave a Bresser chance to spread because of the greater volume of excreted faecal matter which results from their vinalence for the Latestinal tract On the other hand, the enteroviruses properm the polso, Corsackse and Echo viruses seems to spread quite seadily, though cotturnonly causing no diarrhoes at all If we belong to a school of thought which believes in their spread from the respira. tory tract, it 13 then relevant that ctcept for the herpangina caused by some Coxsackieviruses, enteroviruses do not product symptoms at that end of the host esther The virulence they show for the nervous or muscular systeras seems to be accidental and quito intelevant to the virus's evolutionary necds These viruses are all tough, resistant to many physical and chemical agents, more so than the respiratory viruses (other than the adenoviruses). So perhaps they can persist in the environment long enough to have a reasonable chance of teaching a nen host they have not needed to 2cquire or retain the virulence necessary to aid thesr spread A parallel za afforded by kicks and trombreulids which can afford to walt hungrily for prey for weeks at a time, they do not need wings like the more delicate mosquito One sught forecast for the enteroviruses that, with upravements in hygiene, things will get flasdes for them and that they will have to evolve a more efficient mechan. ism for distribution, even that spread from the respiratory tract might prove to be operationally more worth while than reliance on a fatcal route Arborviruses The stuation with arborviruses 13 very different here, the mportant point is the height and duration of virseruagiving optimal chances of spread by an arthropod to another host Most zrborviruses probably inhabit a reservost host for which they are harmless this fact speaks against any evolutionary tendency towards higher virulence for the host to 22d spread Assuming that monkeys constitute the natural reservoir of jungle yellow feverin Africa, we may note that the situs, which is normally harmless to thern, perpetuates itself just as readily as it does in South America where the local monkeys tend to suffer from lethal epidetrics Nacnamara (1954) found that ciascal severity of yellow fever is man was not correlated with duration of viracma Virulent infec tions by the equine encephalomyelius viruses in horses and man are certainly of no value to the virus, they are blind-alley afections ·36 C. H, ANDREWES with a viraemia too short and too low to help the virus to spread One cannot conclude, however, that spilling over of arthropod viruses into strange hosts is never associated with a high viraemia leading to rapid epidemic spread in the new host. This is doubtless what has happened in the past when jungle yellow fever has led to an Aedes aegypts-Homo sapiens cycle, very favourable to the virus With certain apparently new epizootics caused by arborviruses, one may suspect that something similar is happening Before dismissing virulence as unimportant for the survival of arborviruses we must recall that it can be looked upon either from the point of view of the whole animal, as a factor leading to illness or death, or on a microscale as a matter between the virus and the parasitized cell-a sort of microvirulence If the cells are of an expendable sort this may matter little to the host At this cellular level, however, it may be that a more effective microvirulent virus may be one which leads to higher and longer viraemia and prospers in consequence. Myxomatosis We may speculate about the effect of natural selection on the virulence of other viruses, but about myxomatosis of rabbits we have actual knowledge. The facts are of much interest Natural selection appears to be acting to favour a virus, not the most or the least virulent, but one having an intermediate, optimal virulence In Australia, spread is almost wholly through the agency of mosquitoes, and these, to be efficient vectors, must pick up virus by biting through highly infectious skin lesions, not enough virus can be acquired from blood Since 1950–51 the virulence, as judged by mortality rate, has been falling Fenner and Marshall AN tant ananchine after a standard inoculum, they estimated this latter value, therefore, and could then classify their virus strains into any of 5 grades of virulence The evolutionary tendency towards a less virulent virus is explained by the usefulness to the virus of longer survival of rabbits in a state capable of transmitting infection to fresh hosts through mosquito bites The effectiveness of the selection is shown by the fact that highly virulent virus has been introduced into areas where the virus is less lethal than Australian farmers would like;

·EFFECT OF CHANGES N PARASITE AND HOST 37 Jet in a matter of months the older, milder strains have completely pushed out the more rapidly fatal ones The loss of virulence must not, however, be carried too far, strains whose virulence 19 too low fail to give rise to a good enough lesion or thus regresses too early, so that mosquitoes acqure virus less readily In Brican, also, attenuated variants have appeared but they have not been replacing the virulent strains as rapidly as in Australia It appears, in fact, that the ratio of virulent to attenuated strains may in places be increasing The differences may be due to the fact that the main vector in Batain in the rabbit ftea, not the mosquito Now the Australsan rabbit, infected with an attenuated VISUS, lives for many days and so can infect many mosquitoes which bice it when its tumours are full of virus, fly away and later bite and infect another rabbit The British rabbit flea, which is a rather sedentary species has, on the other hand no temptation to leave a mildly infected rabbit. If this dies after a chronic illness, there may be but little virus, or virus mixed with antibody, on the fica's proboscis, and if the rabbit recovers there may be no virus at all L-L. s. 11 1 strain A L. 1. survise to keep the infection going If less virulent virus is also present with it, numbers of rabbits will recover and their progeny, fully susceptible, will be on hand to help perpetuate either virulent or avrulent strains It 13 hard to forecast how natural selection might act upon such an intriguing and complicated situation, perhaps it might favour a virus which as regards its virulence was genetically unstable 1. - mortality was down after 4 or 5 years to 50 per cent, while after 7 ·38 CH ANDREWES annual epizootics only 30 per cent of inoculated wild rabbits died An interesting result of this change is foreseeable virus of optimal virulence for the unmodified rabbits may be too attenuated to have the best chance of survival in the rabbits of today Virus may therefore tend to attain not to a fixed level of virulence but rather to operationally similar behaviour this in practice will mean a more virulent strain assessed in terms of tests on unmodified domestic rabbits The level of virulence of other viruses at large today may well be the result of interplay of similar factors The low virulence of measles virus for Europeans as compared with that of virg n populations as in Fiji and the l'arões may be due partly to genetc influences and partly to factors concerned rather with immunology Rinderpest Another example of genetic differences in resistance to a virus is worth mentioning because it has such a practical bearing on efforts at control by immunization Rinderpest virus has been attenuated in the laboratory in three ways--by passage through goats through rabbits and through chick embryos in each instance the virus has gained virulence for the new host but has become more or less attenuated for cattle The precise level of attenuation which is best for practical use depends upon the breed of cattle concerned (Brotherston 1955 56) A vaccine which 13 excellent for many T highly attenuated for the more resistant zebu and the still more re sistant buffalo and will have little useful immunizing effect in them It seems not unlikely that the zebu and buffalo coming from countries where they have had to coexist with rinderpest for many years have acquired some genetic resistance just as have the Australian rabbits to myzomatosis Summary Multiplication and spread of a virus are favoured if it is not too virulent so that the invaded host survives for a reasonable time to furnish a source of infection of fresh hosts On the other hand a certain degree of virulence exther at the level of the whole host or of the cell may be necessary to provide ? tre in blood or سه ·EFFECT OF CHANGES IN PARASITE AND HOST 39 excretions, so that there is plenty of virus available for a new host The most successful virus will therefore have the optimum virulence for a particular bost But since the host's resistance may itself change through the operation of genetic or immunological factors, the virus may have to change too, and there results an unstable balance An excessive swing may lead to a temporary eclipse of the virus on the one hand or to an alarming epidemic on the other DISCUSSION Niren Dr Andrewes rarely speaks without producing some telling phrase, and the one which I liked on this occasion was 'expendable cells ·40 DISCUSSION A 11 verwecri rabo119. Bult, in his early work, thought that this was probably how it would spread in the field, and from about 1933 unul 1950 the standard strain was passed by cage contact between rabbits Fenner has always used that strain as control, and we can be quite 1.4-., L-inn- - L.LA 1. LALA large experiment -Latn-L hann.nhnarahwanid epizootics the whole situation has been dominated by what Fenner 1.Lat L affecting infectivity by causing a more severe or a less severe wedi ne affant nn tha mshit virus Dick Another important point, in relation to myxuinalus -ntot fram laboratory tests of the virul- Pry L. m- ԱՌԱԽԼԱՏ Burnet It is the integration of the whole ecological field which will termine what survives.

·DISCUSSION 41 Another point of some interest in regard to the Australian myxo matosis experience is the nature of the process by which a relatively less virulent strain can emerge, presumably as a mutant, from a population of highly virulent material Fenner and Day have dis cussed this, and their conclusion is that the average dose transferred virus you recover here 19 much more virulent than those in Australia It 18 surprising that seemingly, even if British Virus were assayed in British wild rabbits and Australian virus in Australian wild rabbits that in areas where the temperature is higher the mortality is lower, in arrondance v th Thomneana «L ILLUMN uf Anurewes idea that a certain amount of virulence 18 necessary for dissemination of virus is a fascinating one We have done some eyner monts recons! L sta interesting thing is that if one takes what is antt strain which is not mouse adapted, such ez dent man han ·42 DISCUSSION it LL. his strain of virus, and reproduced exactly the kind of cage that he used, but I just could not reproduce hus results under British conditions I never found out the reason for the differences Kilbourne We have succeeded thus far in all three experiments using the mouse-adapted strain Our test for the presence of infection in the cage mate the contact mouse-Is the presence of LL oba . Kilbourne Some of the utres obtained are 10 or more Henderson How many transfers have you obtained in this way? Kilbourne Our test for the presence of infection has been the palpable demonstration of virus in the contact mouse lung Henderson You have not tried to transmit this cross-infection to yet other normal mice? Kilbourne No, not yet Henderson In our experiments this can happen in a very small percentage of cases giving quite high virus titres, in the lung, but on nchant + know Henderson One possible explanation for the lack of sequential transfer is found in the type of experimental condition that we have been using The infection is induced first of all by very small particles - clouds-in the lung and we know that the particles that come out of the mouse whenever it sheds the virus are quite large and that the chances of getting to vulnerable sites are much reduced This also happens in guinea pigs with other infectious agents ·BROAD ASPECTS OF THE PROBLEM OF HUMAN VIRULENCE IN INFLUENZA VIRUSES J. MULDER Department of Internal Medicine, Medical School, Unit ersity of Laden IN 1953 Burnet described in his book "Natural History of Infectious Diseases" what he thought would probably occur should a new pandemic of influenza appear, noting that it was difficult to write about influenza without indulging in speculation Two years after the appearance of the 1957 pandemic, we can state that nearly every point which Burnet raised has been consistently proved correct Indeed, one of his remarks concerned the possibility that antibiotics might be relatively useless in saving many lives This point will be the main subject of the present paper because we beheve that an 1957 a member of patients died frutn pure influenza virus pneumonia T say, intrinsic bronchotropic virulence In mice, bronchotropic virulence (Straub, 1937) of straight egg-line virus can be demonstrated microscopically in the first or second passage without coexistent damage in the lung tissue (Fig 1, Table I) It is, however, probable that viral "toxtc" lesions can appear in the ciliated epithelium in mice when a high dosage of egg-line virus 19 inoculated (Table II, Mulder and Hers, 1959) Studies on the individual human dicasse the onda determined by (1) the degree of intrinsic bronchotropic viri 43 ·I MULDER 44 Table 1 DEVELOPMENT OF BRONCHOTROPIC AND PNEL NOTROPIC VIRULENCE IN MICE (STRAN A1 (1951)} (LIVERPOOL TYPE, P PHASE) (Passage 1 egg fluid amn 2-all 2-arn 2) Inoculum Lung + Aficroscopical Mfucroscopical amniotic antronic bronchiolar lung tissue Afortalutst furd passage cell damage damage (reathin FAD number (3 days) (3 days) 10 days) 32 0 014 32 ++++ 014 32 0/4 + + 32 ++++ 32 ++++ +++ • Specific viral cell lesions and lung tutuc damago (2 mice) t. No microscopical control F.A.D - douage of chicken cell haergagglutinating units 1/5 014 Table (Prelum nary erperiment) Toric (?) BRONCHIOTROPISM AND 'TOXIC PNEUVOTROPISY IN MICE (STRAIN A2 (1957) (P PHASE) (egg fiud amn 4)] Microscopical Microscopical bronchiolar lung tissue Mortalıtyt Inoculum cell damage damage (tctthin 10 days) EJ.D (2 5 days) (2 5 days) 100 107 100 105 104 10% 1/2 +++ 1/2 0/2 0/2 0/2 Specific vual cell lesions and lung tissue damage (1 mouse) I No microscop cal control. EID -egs infectious dose of the virus (11) the size of the infecting dosage, () the immunological resistance of the patient and the human herd, (iv) host factors of unknown nature (perhaps associated to some extent with age), and (v) unknown climatic factors (Andrewes, 1959) The apparently stepwise increase in extent and severity of epidemics of influenza Al observed in many countries in the period

·FIG 1 Section of mouse lung 25
days after intrunasal infection with
32 FAD of the first mouse egg
passage of lung non-adapred in
Auenza A1 (1921) virus (I rverpool
type P phase) Complete cytonecrosis
of bronchiolar epithebum Thc tung
tissue 19 normal and all control mice
surved (see Table I) oil immersion
w
(a)
Tic 29 and b Section of mouse lung 3 days after intranasal infection
With lung adapted influenza swine A Virus (allantoic furd) Broadening
of alcolar septa with interstitial exudate
Cytopathogenic changes in what are believed to be ah eolar cells wth
nuclear fragmentation and acuolization oil immersion
To fuce 4+

·(A)
(6)
nuclear fragmentation olimmersion

·FIUMAN VIRULENCE IN INFLUENZA VIRUSES 45 1946–1951, after the appearance of the antogenic vanant A1 (Stuart-Harris, 1953) may perhaps be explained by a periodic increase in intrinsic bronchotropic surulence of successful AI strains, and the same may have occurred after 1934 when the * - no ---- L. fr.L.C...ma Lene 11119 19 Lieally uerotisli dlou by the Uilleicht lesponses caused by the viruses of influenza A and B Careful comparative virological and histopathological studies within the groups of influenza A and B regarding their initial intrinsic virulence for the cilsated epithelium of the respiratory tract of ferrets and mice are, as far as we know, lacking Tutrated original infective human material will also be needed for experiments of this kind because €gg passage may lead to less pathogenic artifacts, as has been clearly demonstrated by Burnet and Bull (1943) in their work on ODphascs in former human A-strains and in vaccination experiments with intranasal application of straight egg-line virus (Isaacs and Roden, 1956, Isaacs, Negroni and Tyrrell, 1957) PQ-phase differences may also be related to the problem As far as we know, however, it is not certain that both phases actually occur in the human disease Q-phase egg-line virus is less antigenic hemisphere existing exclusively in the P-phase (Isaacs, Gledhill and Andrewes, 1952), and Smith, Westwood and Belyavin (1951) have demonstrated an abnormally virulent or "toric' behaviour . (Fukumi, 1959) Nothing is yet known about their intrinsic bronchotropic virulence in ferrets and mice in comparison with 1 1 ·46 J MULDER To summarize, we may conclude that differences in the degree of intrinsic human bronchotropic virulence in influenza viruses within the A- and B-group quite probably do exist, but more satisfactory standardized laboratory methods are needed to enable us to determine this quality more quantitatively, assuming that the ex perimental animal is a reliable test object Still more difficult is the situation concerning the problem of human pneumotropic virulence of influenza viruses Before 1957 We had only the vaguest idea about the possibility of the existence of human influenza-virus pneumonia, which emerged exclusively from the experience of pathologists during the catastrophe of 1918 and from the knowledge that most strains of influenza virus A can be made killers of mice, ferrets and hamsters, causiog the well known swollen, plum coloured consolidated lungs Naturally occurring swine influenza can also cause virus pneumonia in this animal On examination of 148 virologically proven fatal cases of Asian influenza in the Netherlands, Dr Hers and I came to the conclusion that about 20 per cent had died from a curious form of nonbacterial pncumonia which resembled closely some of the accurate descriptions of 1918 influenzal pneumonia given by Le Count riser 1095 Lannemally tinta 1959, Newcombe, Nixon and Thompson, 1958, Rock, Braude and Moran, 1958) In the penod 1941-1957, Hers (1954) examined 18 lagen of inin the trachea and, in the majority of cases, to extend down to the interlobular bronchioli but only very rarely to the respiratory bronchioli, and probably not to the alveolar ducts No evidence of Virus pneumonia was present All cases, except the one infected with influenza B, died from coexistent bacterial bronchopneumonia A similar type of legion was found in 66 fatal cases of secondarily infected Asian influenza as well as in 6 cases which died from other causes during the acute phase of this disease ·HCHAN VIRULENCE IN INFLUENZA VIRUSES 47 In 30 other fatal cases of Asian influenza without coexistent bacterial inflammation the epithelial lesions were found in the respiratory bronchioles and alveolar ducts, and heavy damage had been caused in the alveolar septa la the Netherlands, the incidence of this fatal type of pneumonia has been very low and can be roughly estimated to have been I in 4,000 cases of Asian influenza including cases with coexistent bacterial infection There 13 strong evidence that experimental influenza-virus pneumonia in mice, ferrets and swine is basically the same pathological process as in man, but it is more difficult to demonstrate capillary thrombosts in the animal disease. However, typical cytopathogenic changes in what is most probably alveolar epithelium with fragmentation of the nuclei can be clearly observed (Fig 29 and 6) It is a remarkable fact that we are leaming to understand experimental influenza-virus pneumonia from the autopsy findings in the human disease It is probable that the capillary thrombosis with interstitial focal leucocytic exudate, capillary bleeding and "Iyalne membrane" formation seen in human influenza-virus pneumorua are phenomena which are secondary to the necrosis of the alveolar lining, but the possibility that the virus can also damage the capillary endothelial cells cannot be refuted in the near future many new details of 19fluenza-virus pneumonia in both man and animals may be found by applying the technique of Auorescein-labelled antibody and by clectron microscopy. On indirect evidence, however, it is reasonably certain that we can define pneumotropism of the influenza virus as its potency to multiply in the alveolar lining with subsequent destruction of the alveolar cells. Through the work of Burnet (1954, 1955) we can postulate that, during the process of adaptation in the animal lung, the virus develops new genetically detemiped qualities to destroy the new type of cellular bosr We may conclude that human influenza virus pneumonia occurred in the 1918 and 1957 pandemics and that the lung damage In man is basically identical with the experimental virus pneumonia 11 suscepuble animals. The first question we must ask is whether the various strains of 101 -119e7h rod In -- -t L.-- IL ·48 J MULDER ann again under the influence of a new great pandemic, the same occurred but there is no certainty that diagnostic efforts were stimulated on a sufficient scale between 1919 and 1957 or that it was considered worth while in this period to publish findings identical with those seen in the pandemuc year of 1918 The negative findings of the 18 carefully studied cases in Leiden between 1941 and 1957 offer no convincing evidence in comparison with the 148 cases studied in 1957. There has been one case, published by Parker and co-workers (1946), in which there most probably occurred a pure influenza virus pneumonia This was a case of influenza A from the year 1943, caused by a PRS type of strain Perhaps there is one clue to an answer, relating to the fact that, in 1957, influenza-virus pneumonia often occurred in patients with chronic heart disease, including chronic rheumatic endocarda t19 In the Netherlands Dr Straub, a pathologist who has been a keen obsen er of fatal acute lung disease in Rotterdam, made a careful study of the literature on 1918 influenzal pneumonia and found to his surprise that early and late in this pandemic four publications by pathologists (Glaus and Fritsche 1919, Obern dorfer, 1918, de Vries, 1919, Wegelin, 1919) mention the relatively high frequency of coexistent chronic endocarditis, especi ally mitral stenosis (Straub, 1959) However, neither in hus nor in our experience has chronic endocarditis been a clearly defined underlying condition in influenzal pneumonia during the period 1937-1957 Semple (1959) has found no significant increase in the death rate from mitral stenosis or chronic endocarditis during the severe Liverpool epidemic of 1951 On the other hand, it should be emphasized that the fatal case of influenza-virus pneumonia described by Parker and co workers (1946) concerned a 39 yearold man who had myocardial fibrosis and was an iron worker, both conditions perhaps favouring the deselopment of this disease In conclusion, we can say tentatively i'rat in the period 1933– 1957 the human strains of influenza virus did, but only very rarely, cause virus pneumonia in contradistinction to the pandemic 1918 strain and to the Asian type {hat can experimental virology contabute ta the understanding ·HUMAN VIRULENCE N INFLUENZA VIKUSES 49 of the problem of human influenza virus pneumonia ? It has been shown by Shope (1935) that a laboratory strain of the swuc A-virus 19 pncumotropic in mice and ferrets in the first passage and remains so in the following passages This indicates that there might also be a chance that enhanced pneumotropic virulence of strains circulating in human beings can be demonstrated esperimentally although, at present, human host specificity cannot be etcluded The Asian virus isolated from sputum ot emulsion of lung tissue from fatal cases of virus pneumonia needed a certain number of passages to become fully adapted to the mouse lung, and did not differ in this respect from many former A-strains in 6 out of 7 cases tested, Asian virus from human sputum or lung tissue could be established directly in the lung of mice, but the same was observed by Francis and co-workers in 1937 when the A-PR8 virus was prevalent, and it should be emphasized that the Asian virus 13 non sensitive to the neutralızıng factor present in normal mouse serum Medill-Brown and Brody (1955) have shown that certain non-adapted egg-line strains which are made non-sensitive to this factor can more easily and more rapidly be adapted to the mouse lung than the orginal sensitive parent strains in our tests, using the crude method of passing 10 per cent emulsion of lung tssue, fatal virus pneumonia did not develop in mice before the fourth passage and there was no difference in this respect between strains isolated from sputum of uncomplicated cases and strains isolated from lung tissue of fatal cases of virus pneumonia Many isolates from cases of human virus pneumonia were Q phase strains in contradistinction to the observation that when Q-phase egg-line virus becomes adapted to the mouse lung, a change to P-phase strains invariably occurs. We have so far not carried out careful comparative work, but it should be noticed that is vestigations of this kind should mchude former A strains which are previously made nonben. . . - - - - 41 ASCII, le phenomenon of the egg-lines of the Asian virus being non-sensitive to this factor, first focused by Isaacs (1959) 13 avt- ·50 J MULDER adapted lines are for the most part non sensitive (Chu, 1951, Brans, Hertzberger and Binkhorst, 1953) We have thought for a while that, in general, non-sensitive A-strains might have a greater intrinsic human pneumotropic virulence than sensitive strains Human scrum, however, possibly does not contain the neutralizing factor (Chu, 1951) and the change from sensitivity to non sensitiv. ity during the adaptation in the mouse lung may also be caused by selection of non-pneumotropic, non sensitive mutants, causing a greater survival chance of the resistant variant in the mouse lung (Medall-Brown and Briody, 1955) A compler but very interesting phenomenon in the problem of experimental influenza-virus pneumonia is the so called "toxic" or "non-transferable" pneumonia which can appear in mice and hamsters when inoculated with live, high titre, complete egg fluids of lung non adapted virus (Anderson and Burnet, 1947, Burnet, 1955, Dudgeon et al, 1946, Sugg, 1949a and ) We found to our surprise that this pneumonia shows histopathologically very nearly the same picture as that caused by fully lung adapted strains (Fig 3a and b) In the light of recent experience it , 13 probable that this pneumonia is the result of viral destruction of the alveolar cells caused by a multiplication to incomplete virus T. P. iner Inneren shown that toxic pneumonia can be provoked in mice with a lower dosage of lung non-adapted A1 egg-line virus (allantoic fluids) that is previously made non-sensitive to the neutralızıng mouse serum factor What precisely is the underlying mechanism in this type of (cell) injury is unknown, but it should be emphasized that when a great quantity of lung non adapted influenza vitus reaches the alveoli, a fatal pneumonia may appear which so far cannot be dig tinguished microscopically from the pneumonia caused by fully lung-adapted strains Straight, high titred egg-line fluids of the Asian type also show the phenomenon, but we have not yet performed a careful quantitative comparison with former human A-strains which previously have been made non-sensitive to the neutralızıng serum factor. The problem is difficult because viral "toxicity" may be enhanced by a certain degree of genetically determined pneumotropic virulence It may be asked whether the lung damage from the 1957 ·51 HUMAN VIRULENCE IN INFLUENZA VIRUSES pandemic is perhaps related to this toxic pneumonia in experimental animals Kilbourne (1959) has shown that human lung tissue from fatal cases in this pandemic has yielded amniotic titres in a diluuion of 10-4, and we know that in fatal virus pneumonia the epithelium of many respiratory bronchioli and alveolar ducts are destroyed by the Situs "Thus it is concessable that mass aspisation of virus in the alveoli may occur This speculation of a possible relationship of 1957 human influenza-virus pneumonia to experimental toxic pneumonia 19 better suited to the observation of the low incidence of virus poeumonia in previously healthy people than the assumption of an increased degree of human intrinsic pneumotropic virulence of the Asian virus in the latter case we would rather have expected a mortality or morbidity of the 1918 type The problem of human influenza virus pneumonia cannot be dis. logical resistance in most individuals, thus causing the extensive and rapid spread It is quite possible that former human A-virus appearing after 1918, even when it reached the lung in sufficient quantities to cause lung damage, was often counteracted in the lung tissue by specific antibody provoked by dominant or shared minor antigen present in precursor strains This speculation 13 amenable to an experimental approach As in most clinical problems, host factors have played a distinctive role in human Asian influenza-virus pneumonia Whole series of clinical and fatal cases have been observed in North America (Louna et al., 1959, Rock, Braude and Moran, 1958), in England (Giles and Shuttleworth, 1957 Newcombe, Nixon and Thompson, 1958) and in the Netherlands (Hers, Mſasureſ and Mulder, 1958, Straub, 1959) in pre existent chronic congestion of the lungs 2250ciated with chronic heart disease (especially valvular defects), certain chronic lung disease and pregnancy The explanation of thus coincidence is not clear In chronic congestion of the lungs two factors might have been responsible First, the elimination of virus from the lungs might have been hampered by the chronic. congestion in blood and lymph vessels, thus causing an accumulation of a high concentration of the virus and, secondly, the hyperplasia of the alveolar cells commonly seen in chronic congestion ·52 J MULDER the lungs might have made these cells more susceptible to viral multiplication or to "touc" viral destruction The problem might perhaps be studied in ferrets or pigs, in which chronic congestion of the lungs has been established experimentally. In summary, we would tentatisely conclude that 1957 human, influenza-virus pneumonia was not caused by strains with a high degree of human, intrinsic pneumotropic virulence but that the lung damage was caused (1) by mass aspiration of virus, (11) by lack of protecting antibody, and (11) often by certain unknown, pneumotropism-enhancing host factors, especially chronic con gestion of the lungs If human serum or lung tissue should contain non-specific neutralızıng substance the non sensitivity to it of the Asian strain might have contributed to its pneumotropism We feel that the experience gained from the 1957 pandemic throws more light on the mortality in the 1918 pandemic. We may assume that in 1918 the human population was also deprived of The thus showing that influenza-virus lung lesions played a substantial rôle in causing mortality, which was also enhanced by secondary bacterial infection Coexistent chronic endocarditis was found in relatively high frequency by pathologists working in general hospitals The high mortality throughout the world population in the second wave suggests a certain degree of human intrinsic pneumotropic virulence of the causative virus, which may have been enhanced in October and November 1918 and by a host factor associated with the 20-50 age group Evidence now being accumulated (Davenport, Hennessey and Francis, 1953) makes highly probable the hypothesis of Laidlaw (1935) and Shope(1936, 1958) that the virus of swine influenza was the cause of the 1918 pandemic In this respect it would be very worth while to study freshly isolated field strains of swine A virus and to compare their degree of pneumotropic virulence with former human A-strains and the Asian virus, because these strains, which are antigenically closely related to the first isolated swine virus and can cause virus pneumonia in Wine, may show a higher degree of initial pneumotropic virulence in ferrets and mice ·HUYAN VIRULENCE IN INFLUENZA VIRUSES 53 The present paper, including more questions than established facts, has been born out of some disillusion which we have et which has contributed to the saving of thousands of human lives, it was not feasible in 1957 to predict on sound grounds the possibility of the appearance of a most dangerous complication of the discase, namely, human influenza-virus pneumonsa DISCUSSION Morgan You referred to the incidence of fatalities in patients with heart disease did you have evidence of an influenzal myocarditis in any of these patients? 17. NAT pneumonia Aforgan Do you accept any of the other descnbed cases of in fluenzal myocardius? Mulder tant nord- --- haen Whether in the case reported it may have been a pre existing event which in fact contributed to the development of the viral pneumonia Morgan What about the reported isolation of influenza virus from the myocardiurn? K Thorne T. el

·54
DISCUSSION
mvocarditie wae ant nroved to have been renant
Aha...
these patients
The mortality was seasonal, irrespective of the
epidemic These are two almost contrary experiences in the same
epidemic heart cases in some curious way seemed to be badly
hit, whereas the ordinary patient with chronic bronchutis and
emphysema was not so badly affected.
Alorgan Have you done any electrocardiographic studies in
patients whom you thought had influenzal pneumonia to see if there
was evidence in those that survived that they might have had
myocarditis ?
Stuart-Harns. Not serially
16-
TY
higher pressure curve in the lung circulation, a very high PCU, in the
peripheral blood and a right preponderance I think there is in all
Londola
Mot reproduced in the preferreduncil

·No text on this page.

·54 DISCUSSION myocarditis was not proved to have been present At the same time on my ward we had eight cases of influenza in patients with rheumatic heart disease of various kinds We have never had such numbers during any previous influenza epidemic but as none died there was 1 1 experience in the fairly large vaccination trial which we carned out in patients with chronic bronchitis during that winter we did not find that the Asian epidemic produced a peak in mortality amongst these patients The mortality was seasonal irrespective of the epidemic These are two almost contrary experiences in the same epidemic heart cases in some curious way seemed to be badly hit whereas the ordinary patient with chronic bronchitis and emphysema was not so badly affected Morgan Have you done any electrocardiographic studies in patients whom you thought had influenzal pneurnonia to see if there was evidence in those that survived that they might have had myocarditis ? Stuart-Harris Not serially L higher pressure curve in the lung circulation a very high PCU, in the peripheral blood and a right preponderance I think there is in all • Not reproduced in the present volume-Eds ·DISCUSSION 55 Ic 1.TINTA - 1 L before gross alterations appear in the nuclei, but have not so far carried out a sequential systematic investigation of the cytochemical changes L. TO -A stage but it is tempting to speculate on whether or not they may represent distinctive strain differences in an indicator cell system which might have relevance to a study of virulence Mulder Our problem is at present maunly topographical of course, most of the bronchial cells are degenerated or necrotic, 1 Hers The histopathological study of influenza virus pneumonia inoculation Kilbourne Dr Hers, have you correlated the existence of the socalled hyaline membrane with the presence of these changes in th ·56 DISCISSION alveolar cells? In other words, have you seen changes of this typen the alveolar cells in the absence of hyaline membranc? Flers Yes, I have The membranes are & late phenomenon Kilbourne This is a question of the specificity of the lesson because you can see it in hypoxic conditions in infants One wondert whether it is much to lean on in terms of assessing specific daryage to the cells by influenza virus Ilers I think the hyalıne membrane production 29 non specific and only a late phenomenon as a consequence of the de-epithelatesso! the alveolar wall Burnet I got the pression that you did not consider the ability to produce viral pneumonia (specific alveolitas) as a gcnenc character ustic of the 1957 strains but rather a sesult of complete lack of ant bodies in the persons concerned Is shat cosTECI, or do you think that the virus has a specific generic character, the ability to attack human ah cou? I am interested because a specific change in anu genic character took place in 1946-47, wluch was not associated with the appearance of alveolar os pneumonia It would seem that the 1918 and 1957 strains may have had a genetic character of being able produce this alveolar attack in a proportion of human cases The only two years for which there is evidence are 1918 and 1957 Do you think that there is a specific genetic difference »luch reparates those strains from the new Al strain ? Alulder This is still a matter for speculation With crude methods we do not find any difference with former A stran in relation to animal pneumotropism We might accept that many people aspirate virus during the acute disease Since we have had a mortality from influenza A -virus pneumonia of 1 in about 4,000 cases it is very difficult at present to accept a genetic enhanced pneumotropic virulence Therefore, we concluded that in some people mass Repstation might have taken place which might cause the same anatomical picture of pneumonitis But that is also a speculation On the other hand, in 1918 the epidemiological picture was quite different The mortality in some regions was 1 per cent It is diffi. cult to deny that the 1918 virus had no enhanced intrinsic pneumotropic virulence Therefore, I should like to see how freshly 130- lated &wine A-vious behaves I have the impression from the work of Shope that it goes sery rapidly in mice, so that these strains might have retained the original pneumotropic virulence But why did the former human A strains, which probably sre derived from A swine, farl to attack the lung? Is the lung protected by antibody, or 13 pneumotropic virulence lost by the continuous circulation in the bronchual epithelium? We know that when lung adapted strains are passed in eges, they may lose some virulence for the lung tissue Burnet There was some eysdence that the 1918 strains Sost virulence within the next few years (Vhen New Caledonia received the 1918 strain for the first time in 1920 it was much less virulent NIulder There may be several reasons First, the circulating strains in this period may not have been slike Secondly, straus may

·FIG 1 (Hers) fluo esce n labelled
ant body techn que app ed to colog
cal smears of nfected mouse lung,
Spec fic fluorescence n cytoplasm of
2 p obably al eol ce Is
mmers on
$
I 2 (Hers) The same cells as D F S 1 res a ned
wh haema ory n eos 0 One cel shows an ab
normal acuof zat on ol mers on
Tofa DS

·THE SEVERITY OF INFLUENZA AS A RECIPROCAL OF HOST SUSCEPTIBILITY* EDWIN D KILBOURNE Dirision of l'irus Research, Department of Public Health and Pra entre Medicine, Cornell Une ersity Medical College, New York Iv the wake of an influenza pandemic in which some victims died of infection with influenza sirus alone (Louria et al, 1959), the position of the present writer that the virus of the pandemic did not differ in virulence from the viruses of the interpandemic disease would appear paradoxical, if not untenable However, this paper will contend that persuasive evidence is lacking that this most mutagenic of human viruses has ever varied in its virulence for man, and that the undeniable occurrence of variation in the severity of influenza, the discase, is usually demonstrably linked to variation in the host or his environment. It is tautological, but worthy of emphasis, that the "virulence" or severity of infection is a reciprocal of host susceptibility Since in the lexicon of Webster (or Oxford) there is no such word as virulence, in the modern microbiological sense, we are all obliged to essay briefly into semantics—with, we trust, a minimum of т 1 which enable it to induce disease Thus, any virus which produces disease may be termed virulent, and the measure of virulence is the presence or severity of disease It is evident that comparative and quantitative measurement of virulence among virus strains de mands a comparison of their effects (disease) in similar hosts under similar conditions Thus, the measurement of virulence is reciprocally related to host susceptibility 58 ·INFLUENZA VIRULENCE AND Host SUSCEPTIBILITY 59 Susceptibility in the present contert is defined not as vulnerability to the acquisition of infection but rather as susceptibility to those effects of infection which lead to the development of disease By this view, the reciprocal of host susceptibility to the acquisition of infection is the property of viral communicability This capacity of virus to spread from host to host and initiate infectionimportant as it is-1s not necessanly relevant to the outcome of infection in the individual host but influences only the morbidity rate in the total population Before specific consideration of influenza virus and man it may be useful to consider in general the biological attnbutes of a virus which might be expected to contribute to its virulence These are (1) ability to make effective contact with host cells (to infect and survive in host cells) (2) ability to achieve high concentrations rapidly in host tissues (the resultant of synthesis and destruction of virus), (3) ability to effect a high ratio of damaging or infective particles to harmless or non infective particles (to preclude the limiting effects of auto interference) (4) ability to spread within (invade) the host to reach vulnerable Sites (5) ability to impau host function through damage to host cells (a) directly (primanly) or (6) indirectly (secondanly) as the result of inflammatory re- Sponse to primary cell damage (6) The ability of a virus to stimulate or pave the way for secondary bacterial infection should be viewed as a sort of pseudovirulence which initiates indirectly a worsening of disease This result is obviously not a real reflection of intrinsic virus virulence, as the worsening of disease will depend upon the nature and yagaries of the bacterial invader As we review those factors which contnbute to the susceptability of the host to the effects of virus infection we note that they constitute what might be termed the endogenous environment o fection These factors include (1) lack of prior experience with the virus (absence of immunity), (2) lack of natural (species) immunity for the v genetic resistance (the latter unproved in man), ·60 EDWIN D. KILBOURNE (3) impaired ability to reduce the infecting dose of virus by natural inhibitors, pH change, etc at the site of infection, (4) inability to contain or limit intercellular viral spread or invasion (as by maintenance of normal cell physiology, phagocytosis, ciliary activity, etc), (5) reduced capacity to resist cell damage and resulting physiological dysfunction These points will be elaborated further with specific reference to A anun --- duration of immunity even following known infection and the multiplicity of antigens to which the host may be exposed The final three factors may be influenced by underlying disease, pregnancy or the administration of drugs In the past, much consideration has been given to the exogenous environment in which epide occur—to such factors as weather, season temperature and crowding None of these save crowding is consistently associated with variation in the disease It is prob able that the mechanism by which crowding increases both morbidity and mortality rates is not related to the host but to the virus, in that confinement of infection to a limited area must increase the concentration of virus and hence the inoculum or dose of virus to which a subject is exposed Thus factor of dose is largely virus underltes the success of the primitive variolization methods T) م دادم و virulence of influenza virus as observed in man to its biological activity as observed in the laboratory? To study this question we must make a temporary judgment that influenza viruses do differ in virulence for man and may assume, for example, that pandemic yiruses or isolates from fatal human disease ·INFLLENZA VIRULENCE AND HOST SUSCEPTIBILITY 61 differ in virulence from those from patients with mild infection We may then compare these situses with respect to those properties which they manifest in the laboratory--with specific reference to the hypothetical attributes of virulent virus mentioned earlier Studies previously reported from this laboratory (Kilbourne, 1959) have indicated that among the pandemic strains of 1957-58 na difference was discemed among strains from fatal or non-fatal infections, even when isolates directly from the lungs of primary, abacterial pneumonia were examined Because definitively, severe or fatal infection with a respiratory virus is dependent upon its capacity to invade and induce dysfunction of the lower respiratory tract, it seems appropriate to examine the pneumotropicity" of the virus in laboratory manimals The results of many experiments in this laboratory may be summarized by stating that the capacity of first or second egg passage pandemic isolates to induce pneumonia in the mouse did not differ from that of earlier interpandemic viruses, nor was the number of chick embryo infectious doses necessary for murine infection significantly different from other strains not adapted to the mouse The ratio of chick embryo infective/mouse infective virus has been employed as a m.casure of the mouse virulence of influenza vnuses (huboume, unpublished data, Ginsberg 1953) In Table I are summarized the results of studies of other viral Table 1 THE COMPARATIVE OCCURRENCE OF CERTAIN BIOLOGICAL PROPERTIES IN PANDEVIC (A2) INFLLENZA VIAUSES AND EARLIER INTERPANDEMIC STRAINS Possessed by influenza truser Property A2 A1 A Murine pneumotoxicity + Murine neurotoxicity In tatro cytopathogenicity Filamentous morphology Sheep RBC agglutination at 30° Equivalent Romniotic and allanto.c SBC Fusceptibility 0 ? (10-11 day chick embryo) High neuraminidase activity 0 > • Laboratory stran NU S agglutinate sheep trythrocytes at this temperature, A1 A2 # the texently repormended nomenclature of the W10 Expert Com mttee on Reginton litur Diverses for the triajor we genic famibe of influenza A 7 + + + + + O++++ + r ·62 EDWIN D KILBOURNE properties in this laboratory and others It is evident that pandemic strains, whether from fatal or non-fatal human infection, did not differ significantly from one another nor from non pandemic strains with respect to cytopathic effect in tissue culture, neurotoxicity in mice (Seto, Hickey and Rasmussen, 1959) or morphology in early chick embryo passage However, the agglutination of sheep erythrocytes by the pandemic viruses at 30°, the case with which they may be isolated in the allantoic sac (Kilbourne, d የሰ፣ሰኔ TT 1 cult to correlate such traits with virulence, as egg-line influenza B strains (assumed to be less virulent than A) also possess sheep cell prompted the examination of pandemic and pre pandemic strains and" fatal"and"non-fatal "isolates with regard to their sensitivity to heat labile and -stable inhibitors in animal sera (Table II) All viruses examined had been passaged only in chick embryos and all but the FM-1 strain had had less than 10 passages In brief, it may be stated that sensitivity to neither heat-labile nor heat stable inhibitor could be correlated with the pandemnicityor interpandemicity of a virus, with the year of its isolation, nor with ts site of origin in the respiratory tract nor with its attendance upon mild or fatal disease In fact, the lung isolate from fatal case A/270(Louria et al, 1 sensitive and inhibitor-resistant, and that graded differences in the inhibitor sensitivity of various strains is explicable on the basis of the relative proportions of the two types of particles In view of husho anansant nh h tor vpsistance or *fie pristine nature of the isolate Thus, it would appear that we wust tum reluctantly from the laboratory for the definition of nfluenza viral virulence and return to a critical study of influenza ·Table I INH BITION BY ANIMAL SERA OF SEVERAL INTERPANDEM C AND PANDEM C INFLUENZA A (EGG PASSAGED) VIRUSES Loge tre of haemagg! mhb of 4-3 HA un ts calf guinea pig hoe Resistant Fatal to labile Year Plac Source case πος Α A not A A not A inhib tor FM 1 1947 NJ throat <1 2 1 1 IA 2 1951 NY throat 1 1 A/126 1953 La throat 1 A/150 1953 La throat 2 A/9 1956 throat 3 > 3 > 3 <1 V HAS OOOOO <1 <1 1 ++++o Conn با 3 <1 INFLUENZA VIRULENCE AND HOST SUSCEPTIBILITY 2 4 9 0 0 8 8 6 <1 2 <1 8 <1 2 0 + + <1 1 ܚ ܚܝܨܚܨܚܝ 1 Jap 305 A/2017 A/2011 A217 A.217 A270 A270 A/486 A/493 A/535 <1 51 1957 1957 1957 1957 1957 1958 1958 1959 1959 1959 Japan NY NY NY NY NY NY NY NY NY th oat th oat throat lung throat lung throat throat throat lung + 6 ܬܬ ܝܬܝܝܝܢ ܝ 2 <1 6 + OO <1 7 1 <1 4 <1 <1 8 4 <1 ܨܨܝܗܝܬ 1 1 7 3 1 + A 56 for 30 minutes HAUT-ha magg utinating units of virus • Derived from ongisa hub or- cuistant strun by mting dution pusage t Amniouc Van o panage. Allan OC 80 10-an pasage 63 ·64 EDWIN D KILBOURNE in man for pur assessment of Influenza virus virulence It may be that the proper study of influenza ss man How shall we judge the virulence of snfluenza virus for man? (1) By the severity of infection as manifested by the ratio of clinically apparent to apparent infections (What proportion of serologically diagnosed infections does not result in disease ?) (2) By the severity of the (uncomplicated) discase This may be assessed by such relaxsely objective crnenia as the degree and duration of feser and the incidence of signs of bronchiolar or pulmonary disease (3) By the case fatalsty rate of the disease Any interpretation of these aspects of infection in terms of viss vinilence per se must be done with full cognizince of the erogenous And endogenous (host) environment in which infection occurs, and with recognition of the limitations implicit in a coraparison of results obtained under field conditions by different investigators af different times with different methods Data on the medence of chnically inappssent serologically dugnosed infections are scanty but are sufficient for a tentausc comparison of epidemics associated with all three major antrgerac types of influenza A virus (Table III) The variability in the Table 1 L'ERCENTACE OF LIVICALLY INAPPARENT INFECTIONS (EENOLOGICALLY DIAGNOSED) 17721 VARIOUS STRAINS OF INFLUENZA AYIRLS N DIFFERENT EPIDEMICS A ALEASURE OF SEVERITY OF SWECTION tetek ant body must but no disease sus Epidemic Jear Population References A 1943 mitary camp 57 Al 32 A2 A2 1950 children a institution 1957 hosptal nord 1957 home 28 48 Conmission on Acute Resp ritory D sease (1948a} hibourne Anderson and Horsfa (1951) Blumenfeld a al (1959) Hennessey and Darts port (1959; Jensen Dunn and Robin son (1958) Stallones and Lenzerte (1959) AZ 25 1957 hugh school thuldren 1957 monta? institution AZ ·INFLUENZA VIRULENCE AND Host SUSCEPTIBILITY 65 - Andrews and McDonald (1955) with a similar population in a 1954 cpidemic in Warwickshire England and probably reflects the notoriously poor hygiene of such populations Despite the dis similarity of the other populations studied it must be concluded that severity of infection--as measured by the percentage of in fections which induced disease did not differ strikingly among the three virus types The severity of uncomplicated disease as a measure of virus virulence The classical description of influenza as a three day fever is fully confirmed by the data presented in Table IV from a variety of studies of infections with all major antigenic types of influenza A virus The somewhat higher maximum fever noted in two studies of patients infected with pandemic A2 virus may be related to the youth of these two groups or may reflect the activity of a more virulent virus The brevity of the total febrile course would seem to belie the latter interpretation The varying percentage of patients with auscultatory or roentgenographic evidence of lower respiratory tract involvement may well reflect the equivocal objectivity of this clinical parameter Nevertheless the data do not suggest significantly d ffering human pneumotropicity of the three viral types The clinical course of experimental tafection of man It would seem that study of the course of experimentally induced influenza in man would provide valuable comparative data in a situation in which homogeneity of virus and equivalence of dose may be assured However the relatively few experiments which have been undertaken have utilized viruses of markedly different animal passage history and analysis of clinical data has been less than ideal Thus reported differences in serological and clinical attack rates in these studies are more probably related to experi. mental conditions than to differences in the virulence of the original Virus prototypes It is of interest that the experiment of Bell and 3 ·66 Tablc IV FEVER AND SIGNS OF LOWER RESPIRATORY TRACT INVOLVEMENT AS OBJECTIVE CRITERIA FOR ASSESSING THE SEVERITY OF INFLUENZA ASSOCIATED WITH VARIOUS VIRUS STRAINS Virus Epidemic ear Population 1936-37 mixed Patients No Age A A 1939 mixed 138 A 1943 military 79 EDS IN D, KILBOURNE A1 1947 military 252 Ferer Pulmonary Ave duration Aue max signs in days 'F % Reference 2 6-4 4 101 2 13 0 Stuart-Harris, Andrewes and Smith (1938) 2 3 Horsfall, Hahr and Rickard (1940) 50% had 102 11 0 Commission on Acute Resp or inore Disease (19485) 24 101 3 37 Kilbourne and Loge (1950) 28 101 0 Sigel et al (1948) Sigel et al (1950) 20 102 1 75 Podosin and Felton (1958) <3 in 60% 5 4 Bell et al (1957) 2 in 53% 101-102 2-4 100 Blumenfeld et al (1959) Experimental Infection with nasul wuhings. 19 57 A1 A1 A2 A27 A2 A2 1947 1949 1957 1957 1957 1957 boys' school boys' school scout camp prisoners 616 23 13 4 21-57 girls' camp hospital ward 30 18-63 • By vueuitation or roentgenogram ·INFLUENZA VIRULENCE AND Host SUSCEPTIBILITY 67 co workers (1957) with human nasopharyngeal washings induced disease quite comparable to that of natural infection, and that the pandemic (A2) virus employed was similarly associated with some symptomless infections There is an interesting suggestion as one reviews the data on experimental infection that influenza strains isolated and passaged in mammals may retain human virulence in contrast to strains subjected to repeated egg-passage (Isaacs and Roden, 1956) The contribution of concomitant or underlying disease to the severity of influenza Although severe and fatal influenza has probably occurred in healthy young adults in the absence of demonstrable bacterial infection, the great majority of carefully studied cases of fatal influenza have been associated with underlying disease or attended by bacterial pneumonia The experience at the New York Hospital- Cornell Medical Center in 1957-58 emphasized the specific in portance of cardiopulmonary disease and, still more specifically, theumatıc heart disease with pulmonary hypertension, in the genesis of fatal influenza This experience is summarized in Table V In the New York Hospital studies the importance of coincident or secondary pulmonary bacterial infection was again recognized, and pneumococci and staphylococci were frequently isolated from severe or fatal cases of influenza However, the bacterial infections posed no important therapeutic problem when death occurred as it did in more than one third of 30 patients with pulmonary involvement—in but half the fatal cases were bacteria contributory and in these the concomitant role of the virus in the production of the diffuse haemorrhagic pneumonia was clear Nine of the 11 who died had serious antecedent disease in a society which nurses to adulthood the congenital and rheumatic cardiac behind a screen of antibacterial agents the relatively increasing threat to life of even the avirulent respiratory viruses is obvious Not only may pre exsting disease dramatically convert a benign infection into a malignant one, but more subtly it may augment the severity of influenza as measured by the reasonably objective criteria of fever or bronchrolitic chest signs In an analysis by Kilbourne and Loge (1950) of the symptomatology induced by the ·89 Table V THE ASSOCIATION OF UNDERLYING DISEASE WITH PULMONARY COTIPLICATIONS OF INFLUENZA THE IMPORTANCE OF ANTECEDENT CARDIAC DISEASE (From Blumenfeld et al, 1959 Reproduced by permission of the Editors, clin Intest) Incidence of underlying disease Deaths Total No enth No Total No of No with No with по CV terth Disease syndrome 710 patients underlying RIID patients disease RHD Tcho died disease who died Influenza complicated by bacterial pneumonia 15 87 3 3 1 1 Concomitant influenza 1/3 virus and bacterial pneumonia 77 5 3 S 3 1/3 Prmary influenza virus pneumonia 6 6 6 4 5 5 3/4 Totals 30 21 14 10 9 5/10 • CV, cardiovascular d acane, RHD Theumatic heart disease In one of the number the dircase we pregnancy 9 ·INFLUENZA VIRULENCE AND Host SUSCEPTIBILITY 69 1947 viris, it was noted that those patients in whom routine throat cultures were positive for B haemolytic streptococci er perienced higher maximum temperature elevation and more than twice the frequency of chest signs as those in whom streptococci were not detected A study by Podosin and Felton (1958) of Asian influenza in 616 Boy Scouts disclosed that 27 per cent of those patients with a history of asthma had abnormal chest roentgenograms in comparison with the 7 5 per cent abnormalities found in patients without Even the pattern of symptoms may be influenced by underlying disease or the administration of drugs In the New York Hospital studies (Blumenfeld et al, 1959) an unwarranted incidence of gastrointestinal symptoms in previously bed ridden patients was attributable to the administration of drugs potentially disturbing to gastrointestinal function It is significant that in earlier reports of fatal influenza in interpandemic periods the probable contribution of underlying disease 18 apparent, if not always emphasized In 7 of 20 cases proved by pulmonary virus isolation prior to 1957, 6 had coexisting disease, in 8 cases in which no clinical details were furnished all but one had evidence by culture of staphylococcalor pneumococcal pneumonia the 4 cases clearly without antecedent disease had bacterial pneumonia Thus in the purview of modern virology it may be said that not only is the fatal case of influenza rare, but it is the rare case in which death occurs in the absence of antecedent disease or coincident bacterial infection Death associated with influenza virus as the sole pathogenic micro organism is almost invariably restricted to those with antecedent cardiopulmonary disease On the other hand, the inroads of demonstrably virulent virus in high dosage may be sharply contained by the host which has experienced prior infection This point is epitomized in a single experiment which is summarized in Table VI In this experiment, mice of the same strain, age and sex were first infected with an avirulent (non mouse adapted) strain of FM-1 virus (A1 1947) or injected intranasally with control material Thirteen days after ·70 EDWIN D KILBOURNE Table V1 DIFFERING EFFECTS OF IDENTICAL INOCULA OF VIRULENT INFLUENZA VIRUS IN COMPARADLE HOSTS (MICE) WITII VARIED IMMUNOLOGICAL AND PHYSIOLOGICAL EXPERIENCE Experience Severity of disease induced Pret tous infection by virulent challenge kupus Injected Alouse truth ar rulent Elith Lesion Viru in group FA-1 tirus cortisone bitt scoret lungi 1 0 4/6 72 2 45 80 3 0 016 4 0/6 10 + 5 5 mg x 3 0/6 33 5 mg 5 mg Na pension infecte this initial experience (which all survived) certain of both control and infected mice were injected with cortisone in moderate or high dosage prior to challenge with a virulent FM-1 strain of virus The severity of the disease in these mice of differing immunological or physiological experience was assessed on the basis of mortality rate, the degree of grossly manifest pulmonary consolıdation and the persistence of virus in the lungs of survivors This experiment clearly shows not only the expected result that prior antigenic experience dramatically negates the severity of infection induced by intrinsically virulent virus, but that even this high order of resistance may be broken down by changes in the physiological experience (endogenous environment) of the host, as in this instance by administration of cortisone just prior to reinfection It 13 notable that the maximal dose of cortisone employed in this experiment is inadequate to suppress antibody formation to influenza virus in mice (Kilbourne, 1955) 1918-19 revisited Now that a modern aetiol to a Its has ve of virul he i + more ·INFLUENZA VIRULENCE AND HOST SUSCEPTIBILITY 71 1933, how does it compare with the scourge of 1918-19? Although the total morbidity of the World War I outbreak probably did not exceed that of the recent pandemic, the total mortality certainly did, and high case fatality rates in young adults contrasted with usual experience in other epidemics (including 1957–58) of mortality peaks only at both ends of the age spectrum The writing of still another agonizing reappraisal of the 1918 experience must be defended, as the virus of the outbreak (unless it be the swine virus) is sull not at hand But as it is the contention of this paper that study of the host and his environment are more crucial to the interpretation of virulence than laboratory study of the virus Itself, it may be of more than academic importance to ask whether the differences in our last two pandemics are explicable without invoking the bogey of changing virus virulence At the outset of this discussion it should be recapitulated that the best documented fact of 1918-19 is that all but a few who died of "influenza" died with, if not of, bacterial pneumonia (Mac- Callum, 1919, Wolbach, 1919, Opie et al 1921) The aetiological heterogeneity of these pneumonias was both recognized and em phasized at a time when Pfeiffer's bacıllus was the much disputed contender for the role of aetiological agent of influenza itself 'To be explained, then, is the undue prevalence of bacterial pneumonia in young adults and especially young men First, one asks if uncomplicated influenza (as a predecessor to the bacterial sequel) was more common in this age group It was not Rather the case fatality rate of those in the 20-40 age group was excessive (Jordan 1927) If we then infer that this age group suffered more secondary bacterial pneumonia, we must ask, why so ? As the host's endogenous environment contnbuted to the fatalities of 1957–58 then the exogenous environment of the 1918 training camp was similarly important in the earlier pandemic If it was the worst of times for the poor recruit, it was the best of times for the student of bacterial infection Meningococcal pneumonia and pericardius, Hemophilus influenzae pneumonia, streptococcal empyema-such bizarre manifestations of infection monopolize the journals of the period as chronic diseases do today Are we to say that the virus of measles was also more virulent at this time because 181 of 2,956 cases of measles at Camp Riley Kansas, con tracted "measles pneumonia" (streptococcal) and 83 died (Stone, Phillips and Bliss, 1918) ? (A pneumonia case fatality rate of 45 8 ·72 EDWIN D KILBOURNE per cent I) Pneumonic complications of measles were unusual in overseas (AEF) troops at a time when 4 5 per cent of those with measles in training camps des eloped pneumonia (Jordan, 1927) It may well be questioned whether such bacterial devastation would have followed the recent pandemic even in the absence of 22 20 19 10 17 16 IS 14 con 97 98-131 CA111/ 000 ATO STRENGTH/YEAR 12 197-930 10 9 Nie 1796 07:51 1956-OST JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN NONTH OF THE YEAR FIG 1 Comparative incidence of pneumonia (all forms) active duty US Army personnel (world wide) during 1956-57 (pre pandemuc) and 1957 58 (pandemic) year antimicrobial drugs A partial answer is provided by the data in Fig 1 which details the differing pneumonia incidence in the US Army in the year preceding and the year of the 1957–58 pandemic If a pneumonia case fatality rate of 25 per cent is projected from these data (in consideration of age, sex, military status and probable ·INFLCENZA VIRULENCE AND HOST SUSCEPTIBILITY 73 pneumococcal aetiology (Ileffron, 1939); the potential virulence of the recent pandemic may be partially assayed on the basis of mean increase of pneumonia incidence in the influenza year of 4-3 cases/1000 average troop strength, an increase in mortality of 25 per cent of 4 3, or 1 07 deaths/1000 would have occurred in the presulphonamide antibiotic era The total strength of the army was 932,162 in this penod, so that approximately 1000 Ines might have been lost in earlier times But we can never measure noreren guess at the reduction in secondary cases of bacterial infection effected by prompt antimicrobial treatment of the initial ones Finally, let us eramine the severity of the tencomplicated disease of 1918-19 as an index of the virulence of its virus To begin with, most contemporary students of the disease (Jordan, 1927, Lichty, 1919, Bloomfeld and Harrop 1919) recognized a brief, benign, uncomplicated illness distinct from the more protracted disease with pneumonic involvement Although clinical descriptions of the time were excellent, and even literary it is difficult to fiod flat quantitative statements concerning fever duration and the incidence of minor chest signs Even Lichry's fine study is confusing on this point llaving stated that the average duration of fever in soldiers wnhour" inflammation of the lungs" was 61 days, he later states that " when the temperature remained up longer than fire days, it could safely be concluded that lung involvement must be present" Still later in the same report the writer implies a mean fever duration of 3-4 days In his exhaustive review, Jordan (1927) states that 'the fever usually lasta 2–3 days in the simple'uncomplicated cases! The finding of Bloomfeld and Harrop (1919) that only 6 percent of 300 patients had pulmonary rales (but no roentgenographıc evidence of infiltrates) does not suggest any unusual pneumotropi city of the 1918 virus and supports the statement that' Roentgen rays (did) not afford a means of diagnosing the uncomplicated influenza" (Jordan, 1927) If one accepts the importance of bacterial complications in the fatal outcome of influenza then the greater "virulence" second wave in the winter of 1919 may be to the the year in which it occurred at respiratory bacterial pathogens and It is suggested that the d. resulted from the fortuitous 3 ·74 EDWIN D KILBOURNE Lahdollmach dhoi. dissemination and high dosage of virus but spread of bacterial pathogens to an unusual degree A note on reduced virus virulence between epidemics The foregoing remarks on virus virulence as a reciprocal of host susceptibility may be extended in explanation of the "disappear. ance" of influenza virus between influenza virus outbreaks and the postpandemic waning of disease severity As the occurrence and severity of clinically manfest disease is influenced by preexisting antibody levels, there seems no reason to doubt the probability that influenza virus infection continues in sporadic, muild and unrecognized guise as population antibody levels rise Influenza virus infection in the absence of disease has long been recognized (Table III), and it is nou established that virus may be recovered from the inapparently infected (Bell et al, 1957) The unaltered potential of the interepidemic virus to cause severe duscase is suggested by the occurrence of fatal cases of bacterial nearance in the line of patients arvell, 1 In summary, it is concluded that the host pamsite relanonship of man and influenza virus 19 a tenuous one subject to wide variation in the incidence and manifest severity of the resultant discase On the evidence that morbidity is conditioned by the presence or absence of immunological experience and mortality by abnormal ties of the host or his environment, it is concluded that variation in the intrinsic virulence of influenza virus for man has not demon strably occurred and need not be postulated to explain variation in the severity of the disease DISCUSSION

·No text on this page.

·76 DISCUSSION Morgan In regard to the data that you presented, Dr Kilbourne, on the ratio of clinical cases to individuals who show antibody rises, 13 it possible to analyse the situations and determine the degree of . saying this, one feels that probably there is just as much subclinical immunization in the latter situation. Then there is the problem of comparing different ages, because whereas the Asian influenza - Alanan these two variations of the infection It may be that influenza 18 باید مد به composition Kilbourne I fully agree Furthermore, one could, I think, explain persistence of influenza virus interepidemically by assuming that such mild infections would occur in populations with high antibody levels without postulating any animal vectors or intermediate hosts as reservoirs of infection between epidemics I was impressed with the paucity and inadequacy of the data which I tried to examine, but I have done so with the conviction that

·No text on this page.

·THE VIRULENCE FOR MAN OF SOME RESPIRATORY VIRUSES PASSED IN TISSUE CULTURES D A. J TYRRELL AND TE BUCKLAND MRC Common Cold Research Unit Salisbury THE facilities of the Common Cold Research Unit have been used in the past mainly for studies on experimental epidemiology attempts to cultivate the virus, and other aspects of the common cold (Andrenes 1958) In the past two years we have added to this work by testing the effects in volunteers of a variety of viruses which have been recovered from the human respiratory tract The viruses used were chosen because they were thought to be unlikely to produce more than a minor respiratory illness In some cases we wished to know whether the virus could produce symptoms like those called a common cold In other cases we hoped no symptoms would be produced but that there would be an antibody response variety of results, and they are presented here in such a way as to illustrate the types of clinical and laboratory changes observed These are all related we believe to the larger matter of the general concept and measurement of the virulence of viruses for man However these experiments were primarily designed to show whether certain viruses could cause colds and whether others had possibilities as live vaccines They were terminated when these questions were answered Although the data might have been added to in order to answer more academic questions on virulence this was not thought justifiable We isolated the volunteers and assessed their illnesses by methods previously developed at this unit (Andrewes 1948, Roden 1958) The observer who examined the volunteers and made the diagnosis was unaware whether volunteers had received virus or control material and usually did not know what virus was being used The laboratory methods for the experiments with 78 ·RESPIRATORY VIRUS VIRULENCE 79 infucaza were those in general use The methods employed with other viruses were generally similar to those used by previous workers wah the viruses concerned All volunteers were isolated for 2 or 3 days and then received the virus in i mi of balanced salt solution as nose drops Observation was continued for a further 7 of 6 days and specimens for virus isolation were collected on alternate days Paired sera were obtained from almost all volun. teers. Antibodies were measured by several techniques but for the purposes of this discussion antibody will mean neutralizing activity detected using about 100 TCD sa of virus In the case of influenza virus the only data available are hacmagglutination in hibition titres of cholera filtrate treated sera The results of these thals will be reported in full elsewhere ECHO viruses As their full name suggests (enteric cytopathogenic human orphan viruses) these viruses are most readily found in the faces of normal chuldren (Committee on Enteroviruses 1957) but we have studied two which were found in the respiratory tract and faeces of children with acute febrle linesses with marked respiratary symptoms (Fig 1) A strain of ECHO 11 ( virus of Phulipson and Wesslén 1958) was given in a dose of 10$ tissue culture infectious doses (TCD) The results are summanzed in Table 1 All volunteers who re ceived the virus became infected even though 4 of them had low levels of berum antibody before inoculation Although virus was isolated Table 1 EFFECT OF INOCULATING HUMAN VOLUNTEERS WITH ECHO TYPE 11 virus (U STRAIN) Turut culture passager Dose Isrtes Antbody finest of strain ven TCD rises Producedt 3 HEL' 100 000 919 919 819 3 HEL 100 000 8/8 7/8 08 +11 MIK Idney at Mk monkey kidney In thus and following tables, HEL - huna embryo lung HEK buman embryo la thu and following led at the newes such Fere Obrerved seturred in olunteers from who ina was recovered or showet in seabody shot both ·80 DAJ TYRRELL AND F E BUCKLAND recovered from the throat in almost all cases there were practically no respiratory symptoms Those volunteers who became all suffered from a general illness with headache, malaise and fever accompanied by disturbances of the alimentary tract-abdominal distension and discomfort and diarrhoea The only significant illnesses were found in those volunteers receiving virus passed 3 times in human embryo lung cultures Virus passed 11 times further in monkey kidney cultures produced no definite symptoms This may be related to the fact that it apparently multiplied less well in man than third pass virus, certainly virus isolations from the throat were less frequent and the mean titres of virus in throat swabs and in faeces were about tenfold less Nevertheless after infections with both kinds of virus we found a rise in serum antı body levels in all volunteers and the mean rise was about 100 fold in both groups ECHO 20 (Cramblett et al, 1958) was tested in the same general way and some of the results are summarized in Table II Table II EFFECT OF ECHO TYPE 20 VIRUS Dore TCD, Tissue culture passage of strain giten 0 0 0 fir44 isolated 2/4 3/4 Antbody 715e 014 2/4 2/2 Miness produced 0/4 1/4 3/4 5 20-30 300 3 HEK 100 000 3/3 1/3 0/0 0/0 3 HEL 100 000 3/3 3/3 In this case we were able to use the virus contained in swabs taken from sick children and sent to us from Bethesda by Dr L Rosen We also gave larger amounts of the virus after passage in tissue -1m leads to a oses OMS ·RESPIRATORY VIRUS VIRULENCE 81 ECHO I CONTROL ECHO 20 MALAISE HEADACHE BACKACHE CHILL PYREXIA MUCO Pus NASAL DESTRUC 34 HANKS / DAY SORE THAT COUGH ABDOM DISTENS ABDOM PAIN DIARRHOEA CONSTIPATION NAUSEA VOMITING o o SO % 50%. FIG 1 НА I PARALLENZA HA 3 PARANNULENZA CONTROL MALNSE HEADACHE BACHACHE CHLL PME MUCO PUS NASAL OBSTAC 4 MANMS /DAY SORE TOAT COUGH ABOOM DISTENS ABOOM PAIN QAFAROE CONSTIPATION NAUSEA VOMITING 50% FIC 2 Fios 1 and 2 The frequency of virus symptoms and signs in Infected with virus > 4 Hanks/Day means that the volunteers more than four paper handkerchiefs in one day, 1e that there significant increase in nua secretion ·82 DAJ TYRRELL AND F E BUCKLAND produced resembled those found with virus from throat swabs There was one case of coryzı and three of general influenza 1 ke symptoms Parainfluenza viruses The name parainfluenza has recently been coined to tidy up the classification of some recently isolated respiratory viruses (Andrewes et al 1959) Sendai and haemadsorption virus Type 2 are grouped together as strains of parainfluenza virus Type 1 Croup associated and ALTB viruses become parainfluenza virus a al 1959) We used strains of HA1 and HA2 viruses which had been isolated recently in Britain (Sutton Clarke and Tyrrell 1959) and in Denmark (Petersen and von Magnus 1958) The results ob tained are summarized in Tables III and IV and Fig 2 Table III EFFECT OF PARAINFLUENZA TYPE I VIRUS (COPENHAGEN 222 STRAIN OF HA2) Tursue culture passages of Dose เus Ant body Illness strain given TCD, Isolated rise produced 1 M. 15 4/5 3/5 3/5 2 MK 15 2/2 0/2 1/2 5 MK SMK 15 150 0/6 3/5 0/6 0/5 06 2/5 Viruses of several passage levels were used and the dose given varied widely Only volunteers who were infected became il The illnesses observed after inoculation of parainfluenza Type 1 were characterized by coryzz sore throat and slight malarse without fever or abdominal symptoms These results are generally similar to those of Reichelderfer and co workers (1958) There was no obvious clinical difference between the illnesses produced by large and by small doses of virus of any number of passages in tissue culture However antibody responses were detected in infections ·RESPIRATORY VIRUS VIRULENCE Table TV EFFECT OF PARAENELLENZA TYPE 3 VIRUS (Afoss AND K STRAINS OF HAt) Tume culture Pasrager of Dose Vở Antibody nest 17051 PLPW TCD Isses produced K 2 MK 03 0/3 K 2 AIK 1 500 212 212 Moss 5 X 35 0f3 0Z 0/3 Afost 5 MIK 1 300 1/4 Most S AIX 15 000 Moss SMK 150 000 314 3/4 Not the volunteer from whom nu *** ulated and who ** L. 15 ssolated 013 114 114 with the first pass of the parainfluenza Type 1 virus and not uith the second or fifth pass In general the clinical symptoms associated with Type 3 were identical with those produced by Type 1 infections and by a Pedigree common cold. However a sharp atrack of clinical influenza was produced in one volunteer by 1 500 TCDs of second pass h virus In addsson virus from the volunteers recesving Moss virus kas relatively stopathuc for monkey kidney cultures while that recovered from volunteers who were guyen X virus was non-cytopathic For ethical reasons the experiments with the vurulent strain were not extended All volunteers receiving parainfluenza Type 3 virus had detectable circulating antibody ar the time of moculation Antibody responses occurred in infections with both strains of parainfluenza Type 3 virus Comparauve titrations of viruses in man and in tissue cultures might be a useful way of studying virulence They are clearly in practicable but it can be seen in Table III that the infectious dose for man of the fifth passage virus of parainfuenza ) apparently Ises between 15 and 150 TCD, while the second passage virus is infectious when only 1 S TCD., 5guyen leexpected the ratio of the infectivity titres of these viruses in man and in tissue cultures to be correlated with the relatne virulence for tissue culture and for man and these results tertainly suggest that the adaptation of the virus to man was falling as the virus was passed in tissue culture However the results of similst mnoculations of parainfluenn Type 3 strains in man show that a correlation like this may not apply when different strains are compared ·84 DAJ TYRRELL AND F E BUCKLAND Adenoviruses It has been shown that infection of human volunteers with terenul passed several times in monkey kidney cells (Heubner et al 1955) In an attempt to produce an asymptomatic infection our col 10 antibodies when inoculated Most of those inoculated were infected and marked rises in neutralizing antibody occurred though there was no increase in complement fixing antibody The results are summarized in Table V Even 150 TCDs of virus Table V EFFECT OP ADENOVIRUS TYPE 7 Dose TCD, Virus isolated Tissue culture passages of straint giren 17 Mh+7 PK 17 MK +7 PK 17 MK +7 P. 15 150 1 500 0/2 1/2 1/2 Antibody MISES 0/2 1/2 1/2 Tulness produced 012 0/2 0/2 17 MK+17 PK 17 MK +17 PK 150 15 000 0/2 2/3 1/2 2/2 0/2 0/3 passed 7 or 17 times in pig kidney cultures produced infection in volunteers This represents a 1/3000 dilution of culture material Fever up to 99 80 occurred in 1 volunteer but no symptoms amounting to a definite illness were noted All the volunteers who resisted infection had detectable serum antibody when inoculated except for the one recerving 150 TCD50 of 17th pig kidney passage virus An attempt was made to detect changes in the virulence of viruses obtained from volunteers These viruses retained their pathogenicity for pig kidney cells and when virus contained in throat swab material (about 10 TCDs) was passed to 5 further volunteers only 2 became infected No illnesses characteristic of adenovirus infections were produced However, one infected and one uninfected volunteer developed the symptoms of a cold and this may well represent the activation of an ordinary cold virus ·RESPIRATORY VIRUS VI&ULENCE 85 the appearance of a common cold visus during an attempted serial passage of adenovirus has been observed at this unit (Pereira and Roden, unpublished data) Influenza viruses Russian workers have extensive experience of preparing antenuated strains of mfluenza virus for administration to mas (see Zhdanov, 1959) Earlier work at this usse suggested that, during adaptation to eggs strains of Influenza virus tended to pass capidly from a state in which they could infect sian and cause an artack of mad influenza to a state in which they fasled to fcct at all (Isaacs and Rodes, 1956, Isaacs, Negroni and Tyrrell, 1957) In collaboration with Dr I Fadeeva se troculated volunteers with to Russian vaccine strasns The results of the first experiments are shown in Table VI. No crmcal symptoms at all were noted but Table VI EYFECT OF EVELIENZA VIRUS A 1955 RUSSIAN VACCINE STATY Passages of Dore Antibady Minery Slydin pren EIDS trolated produced 1 Egg 1000 SIS 0/5 7 HEL 7 NIEL 1 000 000 5/5 015 [28 7 HEL 1 000 0/6 05 1 BB * The anabody fro was detected in the rolunteer from wborn rus va imluied. EID et infectious done virus was secovered from all the volunteers receiving the orginal vaccine, although most valunteers possessed detectable serum antibody In spite of this there were no rises in serum antibody levels after the infections were detected Sera were titrated at Salisbury and at the National Institute for Medical Research, alill 11:}), and 7 different strains of influenza A virus were used as antigeas in kaeinagglutination inhibitson tests (ranging from St. A (Assan}} Complement fixation tests using crude S antigens were Regative This vaccine virus was passed once la cess ac limiting dilutions, and iht egg flud was g313 io volunteers in doses Sorular to the original vaccine Only 1 out of 6 volunteers was infected, a ·86 DAJ TYRRELL AND TE BUCKLAND result which suggests that the virus had a decreased ability to initiate infection in man Five of the volunteers had initial serum HI antibody titres of 10 or less against the vaccine virus but others were infected in spite of a serum antibody titre of 80 Discussion In a straightforward acute infection the virus invades body cells More virus 19 produced in them and is shed to infect other cells The infected cells are damaged and so illness 19 produced and the shed virus stimulates the production of antibody If antibody is present when the virus reaches the body then infection does not begin or does not spread because virus particles are neutralized before they reach the susceptible cells Many of our experimental infections do not fit into this simple picture and we are forced to ask far more questions than we have answered If we assume that a virus adapted to tissue culture can multiply less well in human body cells—as seems to be true with ECHO 11—why did we not observe a relatively smaller antibody response when volunteers were infected with the tissue culture passed virus which produced no symptoms ? How can one have cnough cells infected by a virus to produce an illness and yet have no antibody response as was apparently the case in some para influenza Type 1 infections ? Perhaps the protein of adenovirus or of an ECHO 11 virus is a better antigen for man than the components of influenza and para influenza viruses Perhaps ECHO and adenoviruses infected or were absorbed from the gut and so gave an antigenic stimulus while in limited influenza and parainfluenza virus infections the varus was shed into the lumen of the nose and swallowed and destroyed in the stomach Whatever may be the correct answer we should recognize clearly that it is possible to have a limited infection with a respiratory virus which may cause enough symptoms to be called common cold but which may be initiated in spite of the presence of antibody in the serum and may end without producing any increase in circulating antibodies Infections of this sort may often occur in the colds which afflict adults in Nature on the other hand they may nes er occur in Nature but only in volunteers given rather large amounts of virus in an unnatural manner Why did we need more tissue culture doses of parainfluenza ·RESPIRATORY VIRUS VIRULENCE 87 Type 3 than of parainfluenza Type I to itrate infection in man ? It may be a reflection of the speed with which they multiply in and adapt to Lissue cultures of monkey kzdney cells, or a reflection of the higher titre of serum anubody against parainffuenza Type I in our volunteers On the other hand, if we used some other toute of infection we might have found that a much smaller dose of parumflucoza Type 3 would instiste infection, and this possibility Bill be explored in future etperiments In general it seems to us that were, in our experiments, no have seen apparent attenuation of the virus for man this has been the consequence of a limitation in the capacity of the viruses to multiply in man This may have been ducenther to the host being relatively resistant or 10muse, or to the virus taving lost some of its former abilty to multiply in human cells Our experiments suggest that the reduced amount of multiplicatson leads to a reduction in the intensity of symptoms produced without any qualitative change in the symptoms themselves However, this may not always be 80 The type of clinical illness produced by ECHO 20 virus was not Appzsertly modificd by passage of the virus in tissue culture of human cells On the other hand, the ECHO 11 virus was first isolated from children with respiratory disease and yet we did not ste any respiratory sytoptoms in the volunteers ghen tus virus Does this mean that ECHO 11 virus was modified by passage in tissue cultures, were the children's illnesses produced by another vaus which was not isolated, or as there a difference in the typt of symptoms which ECHO 11 produces in children in Sweden and in adults in England? Unfortunately, specimens from patients infected with this virus were not available and these possibilities could not be tested esperimentally Finally, although all the truses sore inoculated in the same way the type of symptoms produced varied with the biological type of the virus introduced It seems therefore, that strictly con trolled volunteer trials can tell us at least something about the wined pathogenx posentiality of viruses recovered from the capiraloty tract of mana Acknowledgements These experients sould have beca mpossible without the Weing and conscientious help of the volunteen And the staff of the Unut We wish i particular to thank Nius G Worthington for per mission to refer to her work on adenaytrues, Dr M L Bynot and ·89 DA). TYRRELL AND F. E. BUCKLAND . r. 11 1 Viruse DISCUSSION Dick What tests are used for purity, to exclude from your tissue culture material known pathogens or agents which we do not yet J.) F.L. .L...- .. response to the test virus is due to an infection with a contaminating Agent Tyrrell In general, we obtain a new virus from some other laboratory and culuvate it by the methods used in that laboratory, We check that its general properties are as they should be We check its identity by a neutralization test with specific immune 8059, we keep the cultures going for about three weeks to give any contaminating viruses a chance to grow up We have detected contacunating viruses on several occasions mixed up with a new virus, on these occasions we did not inoculate the original virus to volunteers Another safety test is to give the new virus to ourselves We have not nnr 11.1.slenr no ferm in 13 thawed and we had the impression that we were jusly u aveva.. in the process of doing that, but odd specimens had titres of up to 10 ot 10 infectious doses per ml qui nult in? nearly as much virus 7- 12 LI id to these post!VC utlane Burnett own

·aankha
पसर
क
६

·88 D A. J. TYRRELL AND F. E. BUCKLAND Miss Macdonald for help in the clinical work, Dr. C. H Andrewes viruses DISCUSSION Dick What tests are used for purity, to exclude from your tissue culture material known pathogens or agents which we do not yet know how to isolate? I am reminded of the fact that many people used Hela cells for a long time, without realizing that many strains are contaminated with PPLO up to 10 pes mi One wonders whether in the volunteers given the pig kidney tissue cultures, the symptoms might not be due to ECHO viruses and that no antibody response to the test virus is due to an infection with a contaminapng agent Tyrell in general, he obtain a new virus from some other laboratory and cultivate it by the methods used in that laboratory We check that its general properties are as they should be We check its identity by a neutralization test with specific immune sera; we keep the cultures going for about three weeks to give any contaminating viruses a chance to grow up We have detected contaminating viruses on several occasions mixed up with a new virus; on these occasions we did not inoculate the onginal virus to volunteers thaned and we had the impression that we were losing a lot of trus in the process of doing that, but odd specimens had titres of up to to ot 10% infectious doses per tal Kilbourne How many did you put in? Tyrell We put in 10%, so we were getting out nearly as much virus as we put in, some days after we had put it in Stuart-Harris Thinking of the background to these positive results with influenza, comparing Sir Macfarlane Burnei's own ·DisctsSIov 89 experiments of some years agoonegg-adapredstransand Dc basca's recent unaccessful attempts to infect with strains prunarly isolated an eggs, and camed in tggs, I wonder n hat happens to influenza YESUS, ben 11 is put into the egg, such makes it pon-infectious for human beings Burnel That is a problem back has interested me for many year and, as far as I know, until the advent of A2 strains in 1957, the only of LA 13 -.. your experience of that? Isaacs We tried the Asian virus, in particular, and after one amnions passage we found one typical influenza case out of three Voluntecri The other two showed nothing After two further virus passages in eggs at limutang dilution we found no chnical effects in six volunteers, o virus was recovered, and there was no antibody 43116 res, I suppose st would have been We have heard that Red Voreers produced with tair consistency e rise in antibody in thosc who had no, or very low, basse saubody The A results were not nearly #3 regular, but we got a fast number of nats, enough to rest it out on 20,000 people afterwards Mulder path the esg-lenges of Asian virus I got no entegetic Teaponse in humans, but as toon as the strains was passed in mice se gave in annibody response I thought from general experience, and Also from the experiments of Francis, that the mouse-adapted Lines could give responses and protection in Volunteen much better than ·88 DAI TYRRELL AND F. E BUCKLAND Miss Macdonald for help in the clinical work, D: C H Andrewes frohe Yiruses DISCUSSION Dick What tests are used for punty, to exclude from your tissue culture material known pathogens or agents which we do not yet L agent Type!! In general, we obtain a new virus from some other Laboratory and culuvate it by the methods used in that laboratory We check that is general properties are as they should be we check its identity by a neutralization test with specific immune sera we keep the cultures going for about three weeks to give any contaminatang viruses a chance to grow up We have detected contaminating viruses on seseral occasions mixed up with a new virus, on these occasions we did not inoculate the original virus to volunteers Another safety test is to give the new ytrus to ourselves We have not ...anu na nim in in senu-JU ILU 1" FL - Nomestion is related to Prof Dick's In connexion 1 swab was hem, these - frozep and thened and we had the uupiss lot of virus in the process of doing that, but odd specimens had betres of up to 10 ot 10' infectious doses per al Kulbourne How many did you put in? Tyrell We put in 10', so re kere getting out nearly as m #sve put in, some day after we had put it in Stutt Harris Thinking of the background to mulo with influenza, comparing Sir Alacfarlane ·DiscusSION 89 cxperiments of some years agoontes adapredstraas And Dr Isaacs's secent uszsuccessful attempts to infect with straini primanly isolated in eges, and carried in eggs, I wonder what happens to infuenza pirus, when it is put into the egg, which makes it non-infectious for human being Burmet That is a problem which has interested me for many years d, as far as I know, until the advent of A2 strasns in 1957, the only itive findings from human inoculation of influenza A in smail es were in the experiments which Afrss Foley and 1 did in 1940, kuch we got two undoubted typical influenzas and one hrgh-titre 1 antibody in three people who had no initiat antibody as tested he method then beug used All subsequent work, including our ş, undicated that a very large dose of egg-adapted material had to given intravenously to give any sort of ssfection at all Ar that to I thought that only O-phase virus had human pathogenicity in Aman A2 virus there appears to be no O-phase in the same sense, and the pathogeatesty seems to persist longer Dr Isaacs, what is your experience of that Isaacs We tried the Asian virus, in particular, and after one anaotic passagt we found one typical influenza case out of three volunteerst The acher two showed nothing After two further vnus passages in ¢&gs at bumuung dilution we found no chnical effects in six volunteers, no virus was recovered, and there was no antibody response It is not very convincing, but certainly the first passage matenal was the only one that showed anythung Tyrrell We then took some of that vurusand passed st ten times in monkey kidney cells, and it remained in the D phase It produced. cold-bác liness in one of the three people infected, but we did not work it out properly Permra understood that the anubody titres in these two experimenta were determined by haemagglutination thabitson and complement fixation testa Would it not be possible that you might have had some serum neutralızıng antibody rises ? Tyrrell Yes, I suppose se would have been We have heard that certain workers in Czechoslovakia have done experiments like this woh Russian A2 atrains, and got results almost identscal with our Burnet In the earliest work in this field, for what we were responsible, in 1940-43 the strain LEE B adapted to egg by Francis and given a good few passages in our laboratory, when given to volunteers produced wsch fair consistency a nse in antibody in those who had low, bask antibody The A sesults were not nearly as He got a fast sumber of rises, enough to test it out on to afterwards of Asiant terus I got to antigenic reaponic As the strain was passed in mice 16 guve en ht from general experience, and Iso from that the mouse adapted liber volunteers much better than uld g** ·89 DAJ TYRRELL AND F. E BUCKLAND Miss I. Macdonald for help in the clinical work, DECAndrewes DISCUSSION Dick What tests are used for punty, to exclude from your tissue culture material known pathogens or agents which we do not yet 1 1 L. .. agent Tyrrell In general, we obtain a new virus from some other laboratory and cultivate it by the methods used in that laboratory We check that its general properties are as they should be We check its identity by a neutralization test with specific immune sera, We 1 - fx.hn .sh. beint trannvrontarminat- AL A.. LA AN wolves We have not Tyrrell Dr Isaacs, of course, is very critical of these experiments from this point of view We used amniotic or allantoic inoculation to isolate the virus We took throat swabs, or throat swabs plus garglings and nasal washings, at daily intervals from 1 to 5 days thawed and we had the impression that we were losing a lot of Yuus in the process of doing that, but odd specimens had titres of up to 10* ot 10% infectious doses per ml Kilbourne How many did you put in? Tyrell We put in 10%, so we were getting out nearly as much virus as we put in, some days after we had put it in Stor-Harns Thinking of the background to these positive results with influenza, comparing Sir Macfarlane Burnet's own ·Discussion 89 experiments of some years agoonegg-adapted strains and Dr Isaaca'a recent unsuccessful attempts to infect with strains primarly isolated 241 eggs, and carried in eggs, I wonder what happens to ufuenza wut, when it put into the egg which makes it noft-infectious for human beings Burnet Thatss a problem which has interested me for many years and, as far as I know, until the advent of A2 strains in 1957, the only positive findingy from human inoculation of influenza A in small doses were in the experments which Miss Foley and I did in 1940, in which we got two undoubted typical Enfluenzas and one thigh-titre FISCLA anubody in three people who had no inicial entibody as tested by the method then being used Al subsequent work, including our own, indicated that a very large dose of egg-adapted matenal had to be given intravenously to give any sort of infection at all At that time I thought that only O-phase virus kad human pathogenscsty In Assan A2 virus there appears to be no O-phase in the same sense, and the pathogenicity seems to persist longer Dr Isaacs, what is your experience of that? Isaac lic tried the Asian virus, in particulas, and after one anotac passage we found one typical influenza case out of three volunteers! The other two showed nothing After two further virus passages in esgs as lemting dilutson we found no clasical effects in ex volunteers, no virus 39 recovered, and there was no entsbody response It is not very convincing, but certainly the first passage material as the only one that showed anything 7ywell We then cook some of that vinus and passed it ten tames in monkey kydney cells, and it remained the phase. It produced a cold-like liness in one of the three people infected, but we did not work it out properly Pereira I understood that the antibod, atres in these two experiments were determined by traernagglutination-inhibition and complement fixacion tests would 11 nor be possible that you might have had some scrum Reutralzing antibody sises 3 Tyrrell Yes, I suppose at would hase been We have heard that certas workers in Czechoslovakia have done experiments like this with Russian A2 strains, and got resulcs almost identical with ours Burne: In the carliest work in the field, for such ac were responsible, in 1940-43, the strain LET B adapted to tag by Franck and gwen a good few passages in our laboratory, when given to olunteers produced mth faux consistency a rise in antibody is those who had no, or sery lox, basic assbody The A results were not nearly as regular, but ne got a fair number of ses, enough to test it out on 20 000 people afterwards Mulder Wissh the 428 lunce of Assan virus got no en ugenis response u humans, but #* soon as the cran 35 passed so muce it Save an antibody response thought from general experience, and atsa from the crpcrunents of 1 rancis, thas the mouse-adapted inet could give responses and protecton in voluareers much better ·90 DISCUSSION Kilbourne I have just been reviewing some of the experiments hah 11 mammals Stuart-Harris Was the early American work not done with very large doses of yirus ? AL- 4. strains that you have recovered at an interval after infection where it 19 reasonable to suppose that multiplication has taken place, have you put that into volunteers either with or without further tissue culture passage ? Tyrrell Yes We did such an experiment in connexion with the adenovirus work because we were thinking of a live virus vaccine, . -L..LL-. - 11.-J1 L - In molt Kilbourne Dr Tyrrell, you raised another paradox that has to be explained In the case of parainfluenza 1 and the production of .La . human, it may be producing largely non infectious yirug, 80 1t 18 ·DISCUSSION 91 quantitatively not enough to be an antigenic stimulus, but it 19 producing in tico cytopathogenic effects which are sufficient to work to find this out Buckland We have worked with monkey kidney, and so have Andrewes In that connexion, the apparent attenuation of a virus which Tyrrell described has, of course, been achieved by passage in monkey kidney, and at the moment we are attempting to get a similar attenuation by passage in human kidney tissue Stuart-Harris In regard to the question of sal symptoms in relation to human infection, arguments have been put forward that they are based on an allergic reactios. As the parainfluenza viruses are probably agents which circulate widely in the populations, I wonder whether synntome due to them mav henrimarily due not to

·92
DISCUSSION
Burnet An interesting possibility there,
by analngy a ith influenza,
us that you might have an snfection which liberated a great deal of
incomplete non Infective virus which nevertheless had the allergi
This might make it extraordinanly difficult to demonstrate a virus
Andresses But, of the parainfluenza viruses were doing that, we
shouid be able to pick them up in assue culture
Morgon What about the effect of anthiszarumu
Burmet Has it been tested?
Alorgan I don't know whether sc has been tested in this disease
It has certainly been tested in respiratory infections that are not
defined and I do not think any of the rests offer much hope that thus
was the mechanism

·GENERAL DISCUSSION Burnet The important discussions of this meeting have centred around two topics, first the evidence for and against variation in No. - - vantation in influenza virus virulence over the years Andrewes I am interested in Dr hulbourne's view that the 1918 influenza virus was not particularly virulent and that virulence of associated bacteria can explain the facts of that pandemic I confess I don't see how that theory explains the curious age incidence of 119 RA L.L 1. infuenza virus, but these bacteria also, got up a certain momentum and spread together not only amongst the soldiery in the war area but also when they got into other countries such as India This is in a way a cheering point of view, because it means that we magnificent

·94
GEVERAL DISCUSSION
nathngene than on the higienfachonnan all of -
L 1 L J.
example, and a few others, almost always a bacterial pneumonic
,
.
Teh 19 11 lut pussture, Dr. Doutre, what we gawe Li H.
- Jam
ALL.
C.1A
may indeed produce variations but we are just not good enough at
reached ecologically in man over the years with this virus, and why
there should be any particular survival value in a virus being highly
virulent
Andrewes. It would be odd that influenza virus should be 80 very
stable as regards its virulence when it is the most labile virus we
know as regards its antigenic properties
Kilbourne Except that linkage between antigenic variation and
virulence 18 certainly not demonstrable.

·GENERAL DISCUSSION
95
Tyrrell. It seems to me that the mechanism whuch Dr Andrewes
suggested could be operating here The virus is basically variable
but 11 generally kept at such a les el of virulence that enough sputuin
sprised around during an infection for other hosts to be affected
1- AL
L

·96 GENERAL DISCUSSION Rar Comaka . 1 lines thus far tested Pereira Have you tried it in human embryonic cells? Kilbourne No Tyrrell Another of our safety precautions is that we never give viruses shich have been through any sort of transformed cell lines Isaacs You would agree, Dr Kilbourne, that Asian viruses can vary in virulence in mice, they can be adapted to mice Kilbourne Yes Isaacs I would take the stand that the Asian viruses could also vary in virulence for man Some of the components of virus virulence which have been mentioned today can, perhaps, be brought together One point that was mentioned was the rate of virus multi plication There is quite good evidence in the literature that in some strains, at any rate, the more rapidly muluplying strains are more virulent Antigenic novelty has also been mentioned, and a strain earlier His experiment gave the surprising result that a large dose L11 trea ·GENERAL DISCUSSION 97 are measuring when we talks about Yıtulence is almply the ability to reach a critical virus concentration. That would fit in very nicely LINEN f Andrettet Just as soon as the level of the saine A anubody in the human population has gone down low enough Mulder I think that there is a chance that it wil go down Burnet Why has it not escaped into children and young adults already? Mfuld Some children mught show shine antibody At the age of 20 years the incidence is about 30-40 per cent, and at the age of 4010 >> nearly 100 per cent If a child were to get swine A influenza at would be difficult for the virus to find a way through the popula ton Morgan There is a great deal of influenza ssolation in Iowa Kolbourne Dr Tyrrell, 1 recall that you found that a monkey. kidney adapted train was pathogenic, but pig-kidney adapted 10 . only a few tissue culture doses of parainfluenza virus passed once or ·98 GENERAL DISCUSSIOV Smith The question of virulence 19. closely tied up with the hormon other species I could be shown that adaptation I, Bay, 468 vius to the mouse is Associated with a change of the host component of the Virus, a start would have been made in the determination of a virus virulence factor Kilbourne Is there any evidence that, once influenza virus or any other virus 18 adapted to an alten host, to a certain state of virulence, Jt will change with passage in reasonably high díluson? Sur Mac farlane Burnet raised the point of the probability of a very small inoculum in the natural human disease and reminded us of his own experiments concerning the fact that the changeability of virus and the favouring of genetic change is greater with passage at low haben moderate virulence, but when introduced into South Arried uldu - highly Jethal effect on most of the antelopes there There is always doubt about what a virus will do when it is transferred from its natural host to an alien host It is very diffcult to generalize or to look for specific biochemical reasons for its behaviour Andrewes Prof Mulder has some justification for his fears about the Amencan reservoir, because if we take the swine A influenza at all seriously in relation to human influenza, we have to believe that it got from man to pig If it jumped one way it might jump the other ·GENERAL Discussiov 99 tropic for mice proves almost immediately to be pneumotropic for the ferret too and once kersa There is the fact that Shope's laboratory swine line is immediately pneumotropic in mice and ferrets I feel -- PT N. TI.. vaccines is that they spread, the second is that excreted vaccine viruses cannot be differentiated from some of the less virulent naturally occurring strains Both of these points are generally agreed upon A 1.L.L. community However, we cannot prove that excreted virus has com pletely different characteristire from naturally armurino avımılent ·100 GENERAL DISCUSSION imported virus, or was it a change in a naturally occurring avirulent strain which for some reason suddenly became epidemic? Isaacs Have you any information about the Russian results ? Dick The Russian passage experiments were done in a very artificial way The Russians took faecal vaccine virus and then put it 11 - TE 1 19 that in virulence tests I understand that in some tests monkeys .-LL wrote about this But is is quite unsound to draw a comparison between 17D and vaccinia, because the whole point is that 17D does not spread Andretces Ordinary yellow fever spreads and a modified variety called 17D does not Why can you not get a similar situation in polio? Tyrrell The adenovirus vaccine might be put in this group because the titres of virus in the throat are very low, it 19 quite difficult to reisolate the virus Burnet I gather that what you really want is a marker by which you - 1hol. Dick Exactly Stuart-Harris Prof Dick argues very strongly that one wants attenuated viruses which do not spread, and yet others are just as vehement that such a vaccine would be almost useless The very etation where you want an attenuated virus us in the presence of an LT ·GENERAL DISCUSSIOY 101 only conditions at the present monient under which one would like Incm , ut you use a neteralogous type you can sont things out, but if you use a homologous type you can never sort out which paralysic casts might hanc been due to your vaccine virus, and which had been due to the natural strain, unless you have an adequate marker Goffe Even so, the use of such a strasn mught well act only by an interference phenomenon One could argue that one might get exactly the same result of material containing interferon acre feu perhaps the right thing to do is a controlled experiment using adering the fact that Jeanerun vaccination is a dangerous procedure - thiết tha 4 hạt. - ---• Ireland Kelbourne Whar Prof Dick sants in the form of the non-spreading vinus mught cheoretically be available amply by giving raw nuclesc acid, if this 4 ever quanulatively possible Dick It mught pick up 113 host protein and corre out the way it tarted. Kibourne Thus would have to be one with no hurnan potential whatsoever Coffe Suc Macfarlane, whats the actual evidence for the availence of the Malca strain for the years preceding 1943 ?

·102 GENERAL DISCUSSION Dick Why should you do any damage in any immunuzing pro. cedure? Burnet I doubt whether any immunizing procedure is possible without some risk of damage Jennerian vaccination has definitely produced a certain proportion of damage I might say I am strongly in favour of sticking to Salk's! nhThaa PTM . somebody ALON dann af these instances adequate evidence of alteration of the genetic con- L.L Lehara netancre can be of the parasite ? Isaac: I am sure that both you and Dr Kilbourne are in an unassailable position because none of us can produce the 1918 virus for you, and if we could it could not possibly be tested out Burnet I would stress, with reference to the Malta situation that the cases of polio in the years preceding 1942-43 occurred in the ·103 GENERAL DISCUSSION L. 1917 19-dam. . he is truch greater Isaacs Could not this be explained on the basis of a different antigenic type ? Burnet I do not think so اد EL L a satisfactory one, namely the intracerebral pathogenicity of the - WW. :: strain of, say, good antigenicity with one that 19 manifestly avinilent (at least in a laboratory species) then in this condition where one mondon-m. I- L.----- transfer virulence and avrulence across to another antigenic site

·10+
GENERAL DISCUSSION
will emerge as a new dominant form General experience of virus
genetics suggests that there can be no such thing as a completely
stable virus
Kilbourne I was thinking more specifically on terms of translation
described
Sot +1 Anginf. em um balat
ALL -
and other viruses, but I cannot see that he would ever be satisfied,
because no matter how long a virus may have remained avirulent
under experimental conditions, as soon as it is introduced into the
human population it may be unstable and begin to throw off virulent
mutants
Dick I have thought about this in 17D quite a bit, although I have
not done an experiment to prove it I think it is quite possible that
-1
----
with regard to 17D
Burnet An alternative answer to Prof Smith's question might be
that if you gave the whole population the attenuated polio virus at
the same time, ie covered everybody, the chance of a mutant
Inneherary email and of vou do
Am mainnndamnarinn
Plouaule TISK UL not using it
Burnet Quite As in all human action!
Burnet This has been a very interesting discussion but I think we
all knew that no complete picture of virus virulence could emerge
hnon clanfield

·No text on this page.

·RETERENCES
ANDERSON, S G, and BURNET, F. M (1947) Aust I exp Biol med Sd, 25,
235
ANDREWFS CH (1948) roy Soc Arts 96, 200
ANDREWES, C H (1958) Bull Johns Hopk Hosp, 103, 1
ANDREWES, C, H. (1959) In Perspectives in Virology, P 184 New York
Wiley
ANDREWES, C H ,BANO, F B, CHANOCK, RM, and ZHDANOV, V M (1959)
Virology, in press
ANDREWS, B E, and McDONALD, J C. (1955) Monthly Bull Minist Hirh
Lab Serv, 14, 168
BADEO 1 P and Mnonsw HR110001 r.vn Aladine 617
BLUMENFELD HL, KILBOURNE, E D, LOURIA, D B, and ROGERS, DE
(1959) clin Invest , 38, 199
minem
Darang nammngton
G
tol,
uu
איי פי
UU
DEINHARDT F BERGS V V, HENLE, G, and HENLE, W (1958) Serp Med,
108, 573
DESANCTIS, A N and LOC (1959) Fed Proc , 18, 564
DICK GWA, DANF D S, FISHER O'D, CONNOLLY, J Hand McKEOWN
F (1957) Drit med 1, 1,65
Duros, R I (1955) J exp Med, 101, 59
106

·HORSTALL FL JE HAUN RG and RICKARD E R (1940) Clit Invest
STUNO GO and MELNICE I L. (1958) Immunof 80,
ISUNO O D and MELNICK I L. (19586) Immunol 80, 282
ISAACS A (1959) Personal communication
ISAAC A GUDIL, A W and ANDRI*ISCH (1952) Bull 1174 TIM
Ory 287
THAG A NEORON O and Truth,DAJ (1957) Landt 2, 656
ILACI A and RODEN AT (1956) Lancet 2, 697
JESEN KE, DUNN FL and ROBINSON R (1958) Progr med stol 1
t65
SOHNSON K. M and MONOAN UR (1936) exp Med 10%, 765
SOLDAN LO (1921) Epidemic Inuenze. Chicago American Medical
Association
KANDA Y and MILNCK ) L (1939) rp Mrd. 109
KAPLAN AS (1953) AM. N Y Acad Sci 61, 830
KATLAY A S., and MELNICK I L. (1935) Proc Soc exp Blor (VP) 90,
362.
KLOGH E Y FAA wts FA iad SITYISOV W J (1930)
Amer 11v 68,1
KUBURNT I D (1955) Prec Soc IP Rol (NY) *0,685
KILOLANE ED (1999) Isi 35, 266.

·108 REFERENCES LITWIN, J. and MOULDER,J W (1959) Bact Proc , p 85 LOW H.Y, GEY, G. O, and SHAPRAS, P. (1935) Bull Johns Hopk Hosp.97, 248. --- 17 T T 1 Mtu 1L MACCALLUM, W G (1921) Johns Hopk Hosp Rep. 20, 149 MCDONALD I F (1958) Proc roy Soc B, 51, 1016 MCLAREN, L. C., HOLLAND, JJ, and SYVERTON, J T.(1959) exp Med, 109, 473 MACNAMARA, F N (1994) Thesis, Cambridge MANSI, W (1959) Unpublished data. MARK, D D (1959) Amer Rev Tuberc , 79, 440 MARSHALL, ID, and FENNER, F. (1958) Hys (Lond) 56,288. MARTIN, C M. KUNIN, C. M. GOTTLIEB, L'S, and FINLAND, M (1959) Arch Intern Med , 103, 515, 332 MAXWELL E. S, WARD, TG, and VAN METRE, T. E. (1949) chin. Invest, 28, 307 W . n A ren 101 88, MORGAN, HR, SOULE, D. W., and MARINETT, G V (1959) Irans Ass - DL Ass NY PARKER, FJOLIFFE, L, S, BARNES, M W., and FINLAND, M (1946) Amer Path , 22, 797 PeART, A F. W (1949) Canad publ Alsh J, 40, 405 PELMONT, J, and MORGAN, H R. (1959) Ann Inst Pasteur, 96, 449 PETERSEN, K B, and MAGNUS PYON (1958) Dan med Bull, S. 157 PHULIPSON, L., and WESSLEN, T. (1958) Arch ges Virusforsch, 8, 27

·REFERENCES
109
Prez K. A., and Excu 3 (1954) I blol Chem. 230, 533
burray D 1 - Fet Witt (19491 Nr. F-W 70 774
f
11
RODEV A T (1958) Proc roy Soc Med. 31 271
Room AT and PEREIRA H G (1956) Lancet 2, 592
Sicet, MM, STAFFEX P W WIENER K. M LICHT AB and HENLE W
SELVZ RUTRO R. 1. HAREL, K. and TAKEMOTO K. K. (1959) Bact Proc p. 76
(M54)
Cir
STRANI M (1937) J Park Bart 45.75
Shani niata Nr. 116
Symposium on Latency and Masking in Vinil And Rkkettslal fafectlont.
(1958) Eds. Walter D L Hanson K and Evans AS Mancapoli
Burgst.
Tarm Vvir
. Bict 74,37
Virolory 4141
1

·110 REFERENCES WALKER O J (1919) J Lab cium Med, 3, 154 WALSI, J. BURCH G E., WHITE, A, MOGABGAB, W, and DIETLEIN L. (1958) Ann Intern Med 49, 502 ...nt.L TE rinin re The Patho- ·INDEX Adcoorruser, J2 33, 84-85 86 90 effects of 84 Age lacidence, in influenza 8 17, 18 75 76 93 th measles 45 19 1 1 -25, 29 1 Bacterial infection, contnbuting to the seventy of influenn 67 68 69. 73 74 93 Bacterist pneumonija, 68 69 71 Bicarbonate, in usceptibility of cells 25 Cell(*), Action of virus in 86 susceptibilty of anuno ads in, 24-25 29 bycarbonate in 25 effect of temperature 26 29 30 to Newcastle disease Vitus 23 24 to virus infertion 23 Vitamina in 25 virulence in 14-15 Common cold, 78 83 91 Coxsuche virus, spread of 35 Cytopathic effect, 32 33 Disease, contnbuting to seventy of influenza 67-70 incidence of 4-8 CC110 viruses, 79-82 86 87 91 effects of 79-82 spread of 35 1 fo 1. Tarde Islands, measlea epidemie in 5 8 19 Ilaemophilus inffumsar 74 Start Glease, complicating influenza 49 51 51 53 54 6S Font, changes in effecting Virulence 342 fenetration of virus into cell 11-22 utceptibulty of 102 to influenza virus 22 26 29 32 58 77 to poliomyeliti virus 27 28 29 ·112 INDET Host cell factors, and tumour viruses, 27-28 Host factors, contnbuting to susceptibility, 59-60 in adsorption of viruses, 21-22 in influenza virus pneumonia, 51-52 in toplication of viruses, 23-26 in virulence, 9, 10, 11, 20–33 Hilocsa, observations of single, 9 Infection, age incidence of, 4 due to nbonucleic acid, 25 2 .***, Iva -74, 75, 77, 93, 94, 48, 49, 51, 52, 54, 56, 61-70, . . 12 74 03 04 contribution of other disease, 67-70 with differing virus strains, 66 --- biological properties of, 61 differing effects of on mice, 70 effect of steroids, 29 effects of Russian vaccine strain, 85-86, 88-89 effect of temperature on, 26, 29 hurnan virulence of, 43-57 inhibition of by animal sera, 62, 63 multiplication of, 43 nutrition of 32 strain differences in, 95 variation in virulence, 95, 96 virulence of, criteria, 64 Influenza virus pneumonia, 46-51 Interferon, effect on virulence, 30 Malta, poliomyelitis epidemic in, 101, 102-103 Measles, age incidence in, 4, 5, 19 Myxomatosis, virulence of, 36-38, 39-40, 41, 98 cells to, 23, 24 Parainfluenza viruses, 82-83, 86, 87, 90, 92 ·INDEX 113 Paralaffuenza viruks, effects of 82 83 Puskegonielty, l'acamogla, incidence of 72 73 10 fuenza virus 46-51 heart disease in 48 51 52 53 54 Most factors in 51-52 l'oeumotropism, 43 44 46 47 61 62 03-99 l'oliomyelitis ARC IRC dence of 4-5 102 103 in Valta 101 102 103 mortality and age 4-5 sprrad of 100 101 Poliomyelitis vacclae, 12 14 100-101 102 104 Pollomyellda yirus effect of testperiture on 30 instabil ry of 94 99 measurement of vurulcace 12-14 spread of 35 susctpe at ty of cells to, 22 25 28 Population, study of d sease in + 'roteln, un denoviruses 33 Polttecosts virus, 24 28, 31 32 Heapuratary viruses. Se alto Parinfluenza virusca virulence for man 78-92 Tirolece of 34.41 Fibonucleic acid infect on due to, 25 Kinderpert, 38 98 Steroids, effect on influenza virut 29 Susceptiblity, defin ton of 59 to influenza virus 22 26 29 32 58 77 19 polomyel ts virus 22 25, 28 Swide virus, 49 52 56 96 97 98 Teroperaturs, effect on polomyel tos virus 30 neusceptiblity of cells 26 29 30 Tlstue culture, 14 15 16 Tespiory viruses passed in 78-41 Tumour viruses, host cell factors and. 27 28 Typhus [ever, compansor at mortality (a vingua ep demica 7 moralty and age's 13 Accine, pol omsel + 12 14 100-101 102 104 Sacclpla, 100 Hnal Adsorption, host factory in 21 Irulence, at cellular leve? 1413 brological attributes of virusen in 59 bronchotropic in influenza +3 44 45 46 defis Lyons of 34 39 40 59 fleet of change in pans te and host JHL effect of interferon 30 penet as of 1 2 host factors in 9-10 tt, 20-31 9,60 importance of andoxnine organs in, 9-10 in arhonch 11.16 ·NOTES

·No text on this page.

·NOTES

·No text on this page.