HERPES SIMPLEX VIRUS TYPES 1 AND 2 AND MULTIPLE SCLEROSIS

777 TABLE II-CORONARY HEART DISEASE MORTALITY RATIO RELATIVE TO NON-SMOKERS

(NUMBER

OF

DEATHS)

IN FOUR PROSPECTIVE

STUDIES ACCORDING TO SMOKING CATEGORY IN MEN AGED YEARS

35-64

is correlated with the CO aetiological agent.

yield might

be the

principal

We thank the Medical Research Council for contributing to the financial support of the Oxford study of BUPA men. The cotinine assays were done with support from the National Institute on Drug Abuse DA 2507, DA 0007. H. V. V. is the recipient of a U.S. Public Health Service research career award 5K6-A12372. J. B. is a Lamg research fellow in preventive medicine.

Requests for reprints should be

addressed to N.

J. W.

REFERENCES 1. Hammond EC. Smoking in relation to the death rates of one million men and women. Nat Cancer Inst Monogr 1966, no. 19: 127. 2 Kahn HA. The Dorn study of smoking and mortality among U.S. veterans: Report on eight and one-half years of observation Nat Cancer Inst Monogr 1966; no 19: 1. 3 Best EWR. A Canadian study of smoking and health. Department of National Health and Welfare, Ottawa, 1966. 4. Doll R, Peto R. Personal communication based on data in: Mortality in relation to smoking: 20 years’ observations on male British doctors. Br Med J 1976; ii:

*All ages.

1525-36 TABLE III-MEAN TOBACCO

CONSUMPTION, SERUM COTININE, AND CARBOXYHAEMOGLOBIN (COHb) LEVELS IN MEN WHO SMOKED CIGARS

5.

ONLY AND PIPES ONLY ACCORDING TO WHETHER THEY HAD

6.

PREVIOUSLY SMOKED CIGARETTES

7.

8. 9.

Langone JJ, Gjika HB, Van Vunakis H. Nicotine and its metabolites: Radioimmunoassays for nicotine and cotinine. Biochemistry 1973; 12: 5025-30. Wald N, Idle M, Bailey, A. Carboxyhaemoglobin levels and inhaling habits in cigarette smokers. Thorax 1978; 33: 201-06. Armitage AK, Turner DM. Absorption of nicotine in cigarette and cigar smoke through the oral mucosa. Nature 1970; 226: 1231-32. Elson LA, Betts TE, Passey RD The sugar content and the pH of the smoke of cigarette, cigar and pipe tobaccos in relation to lung cancer. Int J Cancer 1972; 9: 666-75. Turner JAMcM, Sillett RW, McNichol MW The inhaling habits of pipe smokers. Br J Dis Chest 1981; 75: 71-76.

Hypothesis HERPES SIMPLEX VIRUS TYPES 1 AND 2 AND MULTIPLE SCLEROSIS

JOHN R. MARTIN Laboratory of Neuropathology and Neuroanatomical Sciences, National Institute of Neurological and Communicative Disorders and Stroke, Bldg. 36, Rm. 4817, National Institutes of Health, For each of the four variables the difference between "never smoked cigarettes" and "previously smoked cigarettes" is significant (p<0-01) in the cigar smokers but not significant in the pipe smokers (Wilcoxon rank sum test).

smokers with those relating to the mortality experience observed in the four large prospective studies. This was not the case for the cigar smokers, whose tobacco consumption, cotinine levels, and COHb levels were all very different in the two

categories.

The observation that pipe smokers have a relatively high nicotine intake but little if any excess risk of death from coronary heart disease strongly suggests that nicotine is not the major cause of the excess risk found in cigarette smokers. Indeed it cannot be, unless nicotine absorbed through the pulmonary alveoli is more toxic than when absorbed through the buccal mucosa. The chewing of nicotine gum is a proposed method of helping people to stop smoking. While we have not studied nicotine gum in relation to all diseases, our results indicate that it is unlikely to cause an excess risk of coronary heart disease, since the method of absorption of nicotine from the gum is likely to be similar to that associated with pipe

smoking. The observation that COHb levels, in pipe and cigar smokers, fall between those in cigarette smokers and nonsmokers is consistent with CO being responsible for the excess risk of coronary heart disease in smokers. However, it is not strong evidence in favour of this hypothesis; it is possible that some other constituent of cigarette smoke which

Bethesda, Maryland 20205, U.S.A. A

hypothesis is proposed which states that multiple sclerosis (MS) is caused by herpes simplex virus type 2 (HSV-2) in persons lacking herpes simplex virus type 1(HSV-1) immunity. It is examined by comparing the epidemiology of these viruses with that of MS. Summary

The age at which HSV-2 infections are first detected, as well as the age of peak incidence are similar to those for MS. If agespecific HSV-1 immunity data from various places is compared, an inverse relationship of immunity levels with latitude is apparent, and may be related to the effects of ultraviolet irradiation on HSV-1 activation and transmission rates. Prior HSV-1 immunity, if related to subsequent risk of MS, would be protective. This hypothesis also appears to be consistent with the generalisations concerning effects of migration on subsequent rates of MS, and with certain other features of MS epidemiology. INTRODUCTION

MANY investigators have suggested that a virus may cause multiple sclerosis (MS), yet no specific aetiological hypothesis has been generally accepted. An important test of any such hypothesis is its ability to explain the central epidemiological features of the disease. As recent reviews have indicated1,2 these include ages of onset and peak incidence, geographical variations in disease rates, migration effects, socioeconomic and other factors, and temporospatial clustering of cases. The difficulties of fitting the

778

epidemiological patterns of MS to any of the several forms of the viral hypothesis have been emphasised.2 A herpesvirus has been considered by some to be a candidate agent.3,4 In 1964, Gudnadottir et al. reported isolation of a herpes-like virus from the midbrain of a patient with MS.5 When the differences between herpes simplex virus types 1 and 2 (HSV-1 and 2) were subsequently recognised, this isolate was identified as HSV-2.6 No other isolations of herpesviruses from MS brains have been described. Early serological studies did not distinguish between the two virus types, and more recent measurements of HSV antibodies in MS patients have provided mixed

of maternal following the disappearance i and declines." antibody, subsequently HSV-2 infections are first seen at about the time of puberty. Seroepidemiological surveys demonstrate that the prevalence of HSV-2 antibody rises thereafter, with a peak incidence of HSV-2 infection in the 25-34 age group, and a decline in new infections in succeeding decades11,24 (fig. 1, top). MS incidence follows a similar course. First detected in the latter half of the second decade, the peak incidence is remarkably similar in many studies, occurring between 25 and 35 years, with a decline thereafterl,26 (fig. 1, bottom).

first 6 months of life,

results.4,7 However,

recent studies of mice infected intracerebrally with Gudnadottir’s MS strain of HSV-2 have demonstrated that, under some conditions, multifocal primary demyelination is a major component of the pathology.8This observation has prompted the examination of herpes simplex virus epidemiology for a plausible relationship to that of MS. A hypothesis-that MS develops in a small proportion of those individuals infected with HSV-2 who lack previous HSV-1 immunity-is proposed. The predictions of this hypothesis, which are more complex than if only a single agent were considered, are compared with well documented features of MS epidemiology.

HSV-1 AND 2 AND THEIR RELATIONSHIPS

To assess this hypothesis, the biology of HSV-and HSV-2 and the infections they produce must be examined. Useful reviews of the vast literature are available.9-12 HSV-1 infections frequently occur before puberty, are transmitted by oral contact, and establish persistent infections in trigeminal13 and autonomic ganglia.14 Sunburn, 15 fever, 16 trauma and stress,12 are among the frequent activators of latent infection, with formation of recurrent ulcerations, especially in and around the mouth. A small proportion of infected individuals develop acute -

encephalitis. HSV-2 typically is acquired after puberty, is transmitted venereally, and establishes latent infection in sacral ganglia, 17 with subsequent recurrent shedding of virus, with or without symptoms. Neonates may develop encephalitis following infected birth canal. In adults, central nervous invasion has been recognised and manifests as system (CNS) exposure to

an

aseptic meningitis.’ 8,19 Evidence for a close relation between these two viruses has been obtained from DNA hybridisation studies, serological data, and animal experiments. DNA hybridisation studies have demonstrated approximately 50% base pair homology between the two agents.20 HSV-1and HSV-2 also show a high degree of serological cross-reactivity to each other. 21 That immunological cross-reactivity is of biological importance has been convincingly demonstrated in mice. Previous HSV-1immunity, while producing only a limited reduction in HSV-2 infection rates, does provide major protection against severe disease due to HSV-2.22,23 Clinical observations also suggest that primary HSV-2 infections are accompanied by milder disease in individuals with previous HSV-1immunity than in those who lack it.24,25 AGE OF ONSET

In all studies, most primary HSV-1infections occur before The incidence rises to a peak very shortly after the

puberty.

Fig. I-Primary HSV-2 infection and MS incidence. Top: Age distribution of primary HSV-2 infection. After Nahmias (1970). Bottom: Incidence of multiple sclerosis with age. After Kurland (1951).26

Any attempt directly to relate HSV-1 infection to MS must account

for the 2

to

3 decade hiatus between their incidence

peaks. If, however, HSV-2 is thus considered, postulation ofa prolonged incubation interval is not required. GEOGRAPHIC DISTRIBUTION

The observed direct variation of MS prevalence rates with latitude, along with certain exceptions to this generalisation, have perhaps been more enigmatic than any other feature of MS epidemiology.27 In light of the present hypothesis, variation of either HSV-1 or HSV-2 prevalence with latitude would be of interest. There is a striking gradient of decreasing ultraviolet irradiation from the equator to higher latitudes, symmetrical in both hemispheres. The curves are linear, except near the equator and poles, where small deviations occur.28 This, together with the known solar activation of perioral herpetic lesions,15 and the demonstrated efficacy of ultraviolet light in activation of HSV-1 lesions experimentally,29 provides a rationale for the examination of HSV-1 epidemiological literature for a relationship of age-specific

immunity levels to latitude. The results of a comparison between HSV immunity data and latitude are shown in fig. 2. The nine studies shown include all those in which data for 4

or

5 year age groups between 10 and 16 years of

779

the

country’s

15million Coloured

population, and MS had

been identified in the 11 million Bantu. The inverse nature of these two bodies of data is tantalising. It may have an never

explanation. The fit, however, is as one would predict if HSV-1 immunity was protective against subsequent acquisition of MS. Japanese rates for MS are unexpectedly low for latitude, when compared with those for the U.S. and Europe. Within the present hypothesis, HSV-1immunity might be expected to be unusally high. The limited data available, however, do not support this.39-41 On the other hand, in Japanese women over 30 years of age, HSV-2 antibody prevalence rates of 15% have been found.42 In a study from the U.S. which used similar methods, immunity levels of 35% have been reported in women above 30 years of age. 43 Other U.S. studies, using somewhat different methods, have likewise shown higher rates in U.S. women above age 35.24,44 If these figures are representative for these populations, it is then plausible that differences between MS rates in Japan and Western countries might, at least in part, be accounted for by differences in HSV-2 infection rates. Further, as these two examples illustrate, factors modulating either HSV-1 or HSV-2 infection rates may be invoked to explain variations in MS alternate

2-Inverse relation of HSV-1 immunity latitude (n = 9, r= 0.11; p<0.05).

Fig.

at

age 13±3 years with

rates.

(See text.)

age (10-14, 12-16, 13-16) are available.24,30-37 As distinctions between HSV-1 and HSV-2 immunity were not made in most of these studies, this age range was chosen so as to determine immunity levels as near as possible to puberty but minimise HSV-2 effects. If differences between studies, such as type of serological test employed, group size and age range are ignored, and each point is given equal weight, a statistically significant relationship is seen

(n=9,r=-0-71,p<0-05). For HSV-2, a similar or alternate rationale is not readily evident, for, while clinical experiments have demonstrated ultraviolet

activation of cutaneous HSV-2 lesions,15 this mechanism seems unlikely to be of much importance in enhancing natural spread of infection. Further, HSV-2 antibody prevalence surveys are fewer and even more difficult to compare than those for HSV-1. Thus the possibility of a latitude related effect in HSV-2 infection rates can probably not be assessed at this time. If the HSV-1 relationship to latitude is of importance in MS epidemiology, the relationship of infection to disease is inverse, suggesting that HSV-1immunity is protective. This would fit with the previously noted evidence that antecedent HSV-1 immunity reduces the severity of HSV-2 infection, and be consistent with a direct role for HSV-2. This, then, is the basis for formulation of the

hypothesis proposed here. SOCIOECONOMIC AND OTHER FACTORS

Rates of infection with both HSV-1and HSV-2 may vary with socioeconomic, racial, or cultural factors. For HSV 1, South Africa furnishes a dramatic example." Becker reports that in the non-White population (Bantu and Coloured), HSV-1 infection was acquired much earlier than in Whites. Non-Whites achieved 50% antibody prevalence by age 1 year, while this level was not reached until about 5 years by Whites. Further, 100% of Bantu and 98% of Coloured groups had HSV-1 antibody by age 15, whereas in the White population, the prevalence was lower, or about 60%. By comparison, Dean, reporting on MS in South Africa,38 found an overall prevalence rate in Whites of 9.per 100 000, whereas the rate was very low (0 - 2 per 100 000) in

MIGRATION EFFECTS

Population studies of the effects of migration on rates of acquisition of HSV-1 and HSV-2 infections and immunity have not been published. A comparison may be made, however, between predictions of the hypothesis and observed effects of migration on MS rates. For this purpose, it is assumed that age-specific HSV-1 immunity is inversely related to latitude, that age-specific HSV-2 immunity is the same at all latitudes, and that HSV-1 immunity is protective against subsequent risk of MS. It is also assumed that the rate of acquisition of HSV-1immunity in a migrant population tends to be modified to resemble that of the population at destination, and that the earlier the age at which migration occurs, the greater this

modification. The predicted effects of migration on MS rates would then depend both upon age at migration and upon the magnitude (not discussed here) and direction of change in latitude. These are depicted graphically in fig. 3, in which it should be noted that the curves for HSV-1 and HSV-2 are not drawn directly from published data, and that the MS incidence curves show predicted effects only in

a

relative way

as

a

conceptual aid,

and

are

not

derived

quantitatively. The greatest effects of migration on subsequent MS risk would be in the comparison of MS incidence in populations migrating in early childhood with that of their country of origin. For migration at later ages, differences would diminish throughout childhood, perhaps becoming undetectable as the HSV-2 incidence peak is approached. Direction of migration during childhood would, on the other hand, be expected to determine the nature of the migration effect. For migration from high to low latitudes, the effect of migration would be protective against subsequent MS risk, whereas migration from low to high latitudes during childhood would result in a higher MS risk. By comparison, observed MS incidence rates in migrant populations appear to be related both to age at migration and whether MS incidence of the resident population at the destination is higher or lower than that in the migrants’ country of origin. From some of the more extensive and interesting studies, not summarised here, epidemiologists have drawn several generalisations.2,38,45-48 First, migration after about age 15 has an associated MS risk similar to that of the country of origin. Second, migration before adolescence results in a risk closer to that of the new environment. Third, the effect of migration on MS risk can be conferred seen

780

Fig. 3-Graphical representation of hypothesis specific immunity and relative MS incidence. HSV-2 infection rates assumed constant. (See text.)

throughout the childhood period. A fourth and perhaps less generally accepted conclusion is that the effect of migration from regions of high to low MS risk before age 15 is protective in nature. A comparison of each of these generalisations with the predictions of this hypothesis as represented in fig. 3 shows apparent congruity. TEMPOROSPATIAL CLUSTERS

Perhaps the clearest example of the occurrence of clusters cases is that observed in the Faroe Islands.49 In the 18 year period between 1943 and 1960, 24 new cases ofMS were diagnosed, whereas in the preceding 15 years, none was seen, and only 1 case was identified in the 15 years after 1960. British troops occupied the islands from 1940 to 1945. The possible relationship between these two events has been of much interest. No data on HSV immunity is available for these islands. Military troops could have introduced HSV-2 into the island population. If this hypothesis is correct, it predicts that little or no HSV-2 virus circulated in the population before troop arrival, and that it gradually diminished after their departure.

ofMS

to

show

predicted

effects of

migration

on

HSV-1 age

necessary conditions for development of MS, the low case to infection ratio would indicate that other factors, such as host and virus genetics, may also operate. The explicit nature of this hypothesis lends itself to direct examination. Case-control studies of HSV-1 and HSV-2 immunity in patients with initial attacks of optic neuritis may be the single most important seroepidemiological test. These patients are predicted to have no HSV-1 immunity and to show rising or raised HSV-2 antibodies. Further, careful histories may provide evidence of antecedent HSV-2 infection. An earlier study does show significantly higher herpesvirus antibody levels in MS patients than in ones with acute optic neuritis. 50 Such tests should provide a basis for

further evaluation of this

hypothesis.

I thank Marion Webster for help in preparing fig. 2 and Patricia Weir for secretarial assistance. Dr Henry Webster and Dr Neal Nathanson provided encouragement and useful comments. This work was begun at Johns Hopkms School of Medicine and was supported in part by Public Health Service grant EY07047.

REFERENCES

CONCLUSION

The evidence reviewed here suggests that MS could result from HSV-2 infection in individuals lacking antecedent HSV-1immunity. As summarised in the table, either or both PROPOSED ROLES FOR HSV-1 AND HSV-2 IN DETERMINING EPIDEMIOLOGICAL FEATURES OF MS

1. Acheson ED. Epidemiology of multiple sclerosis. Br Med Bull 1977; 33: 9-14. 2. Nathanson N, Miller A. Epidemiology of multiple sclerosis: Critique of the evidence for a viral etiology. Am J Epidemiol 1978; 107: 451-61. 3. Elhson GW. Multiple sclerosis: a fever blister on the brain? Lancet 1974; ii: 664. 4. Koprowski H, Warren KG. Can a defective herpes simplex virus cause multiple sclerosis? Perspect Biol Med 1978; 22: 10-18. 5. Gudnadottir M, Helgadottir H, Bjarnason O, Jonsdottir K Virus isolated from the brain of a patient with multiple sclerosis. Exp Neural 1964; 9: 85-95. 6. Nahmias AJ, Dowdle WR. Antigenic and biologic differences in herpesvirus hominis. Progr Med Virol 1968; 10: 110-59. 7. Norrby E. Viral antibodies in multiple sclerosis. Progr Med Virol 1978; 24: 1-39. 8. Martin JR. Multifocal CNS demyelination in mice infected with herpes simplex virus

type 2. Unpublished. 9. Rawls WE. Herpes simplex virus. In: Academic Press, 1973: 291-325.

of these conditions may be invoked to account for major features of MS epidemiology. While these may prove to be

Kaplan AS, ed. The Herpesviruses. New York:

10. Nahmias AJ, Roizman B. Infection with herpes simplex viruses 1 and 2. N Engl J Med 1973; 289: 667-75, 719-25, 781-89. 11.Nahmias AJ, Josey WE. Epidemiology of herpes simplex viruses 1 and 2. In: Evans AS, ed. Viral infections of humans: Epidemiology and control. New York: Plenum, 1976 253-71. 12. Baringer JR. Herpes simplex virus infection of nervous tissue in animals and man Progr Med Virol 1975; 20: 1-26.

781 13. 14.

15 16 17.

18

19

Baringer JR, Swoveland P. Recovery of herpes simplex virus from human trigeminal ganglions. N Engl J Med 1973; 288: 648-50. Warren KG, Brown SM, Wroblewska Z, Gilden D, Koprowski H, Subek-Sharp J. Isolation of latent herpes simplex virus from superior cervical and vagus ganglions of human beings. N Engl J Med 1978; 298: 1068-69. Wheeler CE Pathogenesis of recurrent herpes simplex infections. J Invest Dermatol 1975; 65: 341-46. Warren SL, Carpenter CM, Boak RA. Symptomatic herpes, a sequella of artificially induced fever. J Exp Med 1940; 71: 155-67. Baringer JR Recovery of herpes simplex virus from human sacral ganglions. N Engl J Med 1974; 291: 828-30. Craig CP, Nahmias AJ. Different patterns of neurological involvement with herpes simplex virus types 1 and 2: Isolation of herpes simplex virus type 2 from the buffy coat of two adults with meningitis. J Inf Dis 1973; 127: 365-72. Stalder H, Oxman MN, Dawson DM, Levin MJ. Herpes simplex meningitis: Isolation of herpes simplex virus type 2 from cerebrospinal fluid. N Engl J Med 1973; 289:

Methods and Devices AN IMPROVED SYSTEM FOR EXTERNAL BILIARY DRAINAGE

J. I. BLENKHARN

G. A. D. MCPHERSON L. H. BLUMGART

Departments of Bacteriology and Surgery, Royal Postgraduate Medical School, London

1296-98. 20

21 22

23

24.

Kieff E, Hoyer B, Bachenheimer S, Roizman B Genetic relatedness oftype 1 and type 2 herpes simplex viruses. J Virol 1972; 9: 738-45. Plummer G. A review of the identification and titration of antibodies to herpes simplex viruses type 1 and type 2 in human sera. Cancer Res 1973; 33: 1469-76. McKendall RR Efficacy of herpes simplex virus type 1 immunization in protecting against acute and latent infection by herpes simplex virus type 2 in mice. Inject Immun 1977; 16: 717-19. Sturn B, Schneweis KE. Protective effect of an oral infection with herpes simplex virus type 1 against subsequent genital infection with herpes simplex virus type 2. Med Microbiol Immunol 1978; 165: 119-27. Nahmias AJ, Josey WE, Naib ZM et al. Antibodies to herpesvirus hominis types 1 and 2 in humans: I Patients with genital herpetic infections. Am J Epidemiol 1970; 91: 539-46

WE, Garnder HL, Flanders RW, Lowry SP, Kaufman RH, Melmck JL Genital herpes in two social groups. Am J Obstet Gynecol 1971; 110: 682-89. 26. Kurland LT The frequency and geographic distribution of multiple sclerosis as indicated by mortality statistics and morbidity surveys in the United States and Canada Am J Hyg 1951; 55: 457-76. 27 Acheson ED, Bachrach CA, Wright FM. Some comments on the relationship of the distribution of multiple sclerosis to latitude, solar radiation, and other variables. Act Psych Neurol Scand 1960; 35 (suppl 147): 132-47. 28 Schulze R, Grafe K. Consideration of sky ultraviolet radiation in the measurement of solar untraviolet radiation. In: Urbach F, ed. The biologic effects of ultraviolet radiation Oxford: Pergamon, 1969; 359-73. 29. Blyth WA, Hill TJ, Field HJ, Harbour DA. Reactivation of herpes simplex virus infection by ultraviolet light and possible involvement by prostaglandins. J Gen Virol 1976; 33: 547-50. 30. Becker WB. The epidemiology of herpesvirus infection in three racial communities in Cape Town. S Afr Med J 1966; 40: 109-11 31. Smith IW, Peutherer JF, MacCallum FO. The incidence of herpesvirus hominis antibody in the population. J Hyg (Camb) 1967; 65: 395-408. 32 Rawls WE, Tompkins WAF, Melnick JL. The association of herpesvirus type 2 and carcinoma of the uterine cervix. Am J Epidemiol 1969; 89: 547-54. 33. Wentworth BB, Alexander ER Seroepidemiology of infections due to members of the herpesvirus group. Am J Epidemiol 1973; 94: 496-507. 34. Terzin AI, Masic MG. Age specific incidence of neutralizing antibodies of herpes simplex virus. J Hyg (Camb) 1976; 77: 155-60. 35 Doerr HW, Lehmair H, Schmitz H. et al. Simple mathematical deductions in the seroepidemiology of viral infections: I Herpesvirus group (herpes hominis, varicella-zoster virus, cytomegalovirus, Epstein-Barr virus). Zbl Bakt Hyg I Abt Orig A 1977; 238: 149-64. 36. McMillan BC, Golubjatnikov R, Hanson RP, Sinha SK. A study of cytomegalovirus, Epstein-Barr virus and herpes hominis (types 1 and 2) antibody in institutionalized and non-institutionalized children. Health Lab Sci 1977; 14: 261-68 37. Sogbetun AO, Montefiore D, Anong CN. Herpesvirus hominis antibodies among children and young adults in Ibadan. Br J Vener Dis 1979; 55: 44-47. 38 Dean G. Annual incidence, prevalence and mortality of multiple sclerosis in white South African-born and in white immigrants to South Africa. Br Med J 1967; ii: 724-30. 39. Tateno I, Yokoyama T, Suzuki S et al. Age distribution of the neutralizing antibody to herpes simplex virus. Japan J Exp Med 1958, 28: 375-80. 40 Yoshino K, Taniguchi S, Furuse R. A serological survey for antibodies against herpes simplex virus with special reference to comparatively heat-labile complement-fixing antibodies. Jap J Med Sci Biol 1962; 15: 235-47. 41. Hondo R. A seroepidemiological study of herpes simplex virus Japan J Med Sci Biol 1974; 27: 205-13. 42 Ishiguro T, Osaki Y. Sexual behavior and antibodies to herpes simplex virus in healthy Japanese women. J Am Med Women’s Assoc 1978; 33: 271-74. 43 Rawls WE, Adam E, Melnick JL. Geographic variation in the association of antibodies to herpesvirus type 2 and carcinoma ofthe cervix. In: Biggs PM, ed. Oncogensis and herpesviruses. Lyon: IARC, 1972: 424-27. 44. Royston I, Aurelian L. The association of genital herpesvirus with cervical atypia and carcinoma m situ. Am J Epidemiol 1970; 91: 531-38. 45 Dean G, Kurtzke JF. A critical age for the acquisition of multiple sclerosis. Trans Am Neurol Assoc 1970; 95: 232-33. 46. Kurtzke JF, Dean G, Botha DPJ A method for estimating the age at immigration of white immigrants to South Africa, with an example of its importance. S Afr Med J 1970; 44: 663-69. 47 Dean G, Kurtzke JF. On the risk of multiple sclerosis according to age at immigration to South Africa. Br Med J 1971; iii: 725-29. 48 Alter M, Kahana E, Loewenson R. Migration and risk of multiple sclerosis. Neurology 1978; 28: 1089-93. 49 Kurtzke JF, Hyllested K Multiple sclerosis in the Faroe Islands: I. Clinical and epidemiological features. Ann Neurol 1979; 5: 6-21. 50. Ross CAC, Lenman JAR, Melville ID. Virus antibody levels in multiple sclerosis Br Med J 1969; iii: 512-13. 25. Rawls

Introduction Percutaneous transhepatic biliary drainage offers a new approach of the high-risk patient who is a candidate for surgery.1 The benefits of preliminary relief of jaundice, such as improvement of hepatic and renal function, are prejudiced when the drainage system becomes infected.2Studies of 33 patients receiving percutaneous transhepatic biliary drainage showed that the incidence of bacterial colonisation of the drainage system increased progressively after the first four days. Furthermore there was clear evidence of exogenous acquisition of bacteria to the drainage to treatment

system. Patients

undergoing percutaneous transhepatic biliary drainage require a system that is simple and easy to manage and allows mobility. Moreover, the system must be closed to bacterial contamination, avoid environmental contamination with potentially infected bile, allow saline flushing of the transhepatic catheter to prevent blockage, and yet permit aseptic sampling of catheter bile for laboratory examination. Currently available systems do not meet these requirements. We describe a new system for external biliary drainage which offers a closed fluid pathway with safe access for saline flushing and specimen collection. The System The Hammersmith

figure) consists of a

15

biliary drainage system (see accompanying cm polyvinylchloride tube with an integral