911LEADING ARTICLES

. THE LANCETLONDON: : SATURDAY, OCT. 29, 1955

Gastro-enteritis in InfantsIN 1922 ADAM 1 published his first paper on dyspep-

siekoli, a group of Escherichia coli which he believedgave rise to infantile gastro-enteritis. His differentia-tion was based on biochemical reactions, and fewworkers followed up his findings. Twenty-three yearslater BRAY 2 suggested that a specific serologicaltype of Esch. coli was responsible for an epidemic ofinfantile gastro-enteritis. Very soon other workershad confirmed and extended BRAY’s findings 3-5 ;and in the course of this work many further types ofEsch. coli which cause enteritis in babies were

identified. If new admissions to a babies’ unit canbe kept in a separate part of the hospital until theyare known not to be excreting any of these specificserological types of Esch. coli, outbreaks of enteritiscan be largely avoided. But this system will fail whenan infant is the carrier of an unidentified new type ;and many workers now hold that if an epidemiccannot be traced to any recognised organism theaffected babies and their contacts should be investi-gated for the presence’of a new type of Esch. coli.In England many hospital bacteriologists have

reason to be grateful for the help that is so readilyand quickly given by the reference laboratories ofthe Public Health Laboratory Service in investigatingthese outbreaks. In the U.S.A. there is obviouslythe same close liaison between clinicians and laboratoryworkers. Last year STULBERG and his colleagues 6in Michigan described an epidemic in a premature-baby unit due to Esch. coli type 0.111 (B 4). Thesame group of workers have now reported anotherepidemic made up of six connected series of casesspread over a period of four months ; of 60 affectedbabies 5 died. In the six preceding months therehad been several sporadic cases of diarrhoea in theunit, and although no setiological agent could be

identified, strains of Esch. coli from each ill baby weresaved. In the actual epidemic each epidemic wavesubsided when neomycin was administered to theaffected babies, but fresh waves of cases continuedto arise until the neomycin was given to all the infantsin the unit. Routine bacteriological investigationsdid not reveal the presence of salmonellse or shigellae,- ttnd because of the uniform colonial appearance of theEsch. coli it seemed obvious that a specific serologicaltype of this organism was the pathogen. Later,EWING et al. 8 identified as type 0-127 (B 8) theEsch. coli isolated both from the earlier sporadiccases and in the epidemic. In this outbreak it wasfound that the longer a baby had been in hospitalthe greater was the risk that it would become infected1. Adam, A. Jb. Kinderheilk. 1922, 99, 86.2. Bray, J. J. Path. Bact. 1945, 57, 239.3. Giles, C., Sangster, G. J. Hyg., Camb. 1948, 46, 1.4. Taylor, J., Powell, B. W., Wright, J. Brit. med. J. 1949, ii, 117.5. Smith, J. J. Hyg., Camb. 1949, 47, 221.6. Stulberg, C. S., Zuelzer, W. W., Nolke, A. C. Pediatrics, 1954,

14, 133.7. Stulberg, C. S., Zuelzer, W. W., Nolke, A. C., Thompson, A. L.

Amer. J. Dis. Child. 1955, 90, 125.8. Ewing, W. H., Tanner, K. E., Tatum, H. W. Publ. Hlth

Rep., Wash. 1955, 70, 107.

-a feature well demonstrated in Glasgow byANDERSON and his co-workers.9

Controlled trials showed that the attack-rate of

gastro-enteritis in carriers of type-specific Esch. coli wasreduced by prophylactic administration of chloram-phenicol or sLilphadiazine.10 American workershave shown that in treatment neomycin is valuable. 6711-STULBERG found that administration of neomycinhad reduced the attack-rate of gastro-enteritis inthe carriers ; and he depended on bacteriologicalscreening and chemoprophylaxis with neomycinto prevent the other infants in the premature-babyunit from becoming infected by Esch. coli type 0127.Despite neomycin therapy the mortality was 8%,and there was a very high bacteriological and relativelyhigh clinical relapse-rate. The schemes describedin England 12 13 were unlike this since they combinedbacteriological screening with segregation of theexcreters. STULBERG’s practice apparently reducedthe morbidity ; but it was more than four monthsbefore the nursery was free of the Esch. coli type 0127.The view that type-specific Esch. coli cause most

of the epidemics of infantile gastro-enteritis ininstitutions is based mainly on epidemiological investi-gations and feeding experiments. Direct proof frompathological findings has been lacking ; the post- .

mortem appearances are not in any way distinctive,septicaemia is extremely rare, and antibody productionis difficult to demonstrate by ordinary routine methods,which show irregular or no rise in titre. NETER 14has evolved a method of demonstrating hsemag-glutinins which promises to provide more informativeresults. THOMSON 15 has attempted to provide directproof of the pathogenicity of these organisms bya careful quantitative bacteriological study of duo-denal juice aspirated during life and of post-mortemmaterial from affected babies. - This important worklends great weight to earlier qualitative findings.16THOMSON found that the specific serological types ofEsch. coli were present in very large numbers from theduodenum to the rectum, and in the bile. (Ascentof bacteria into the upper small intestine has alsobeen demonstrated in cases of sprue.17) Culture of thevomit of babies with enteritis due to Esch. coli often

yields a pure growth of the specific organism, andROGERS 18 found that these vomiting babies can bedangerous to nearby infants.Thus direct evidence is confirming the pathogenicity

of specific serological types of Esch. coli, and neomycinis proving useful and reliable in treatment. But thereis still no generally accepted way of handling anepidemic. In Britain many children’s hospitalshave had relatively few cases of this form of enteritisin the present year ; so perhaps now is the timeto plan for speedy localisation of any epidemic.STULBERG’s experience reinforces a suggestion byROGERS 19 that, if there is an epidemic in a babies’9. Anderson, T., Crockatt, H., Ross, C. A. C. J. Path. Bact. 1954,

68, 1.10. Todd, R. McC., Hall, E. G. Brit. med. J. 1953, i, 1359.11. Wheeler, W. E., Wainerman, B. Pediatrics, 1954, 14, 357.12. Jameson, J. E., Mann, T. P., Rothfield, N. J. Lancet, 1954, ii,

459.13. Brit. med. J. 1953, ii, 219.14. Neter, E., Bertram, L. F., Arbesman, C. E. Proc. Soc. exp.

Biol., N.Y. 1952, 79, 255. Neter, E., Zalewski, N. J.,Fergusson, W. W. Ibid, 1953, 82, 215.

15. Thomson, S. J. Hyg., Camb. 1955, 53, 357.16. Taylor, J., Charter, R. E. J. Path. Bact. 1952, 64, 715.17. Frazer, A. C. Brit. med. J. 1949, ii, 731.18. Rogers, K. B. Proc. R. Soc. Med. 1951, 44, 519.19. Rogers, K. B. Med. ill. 1954, 8, 239.

912

ward or maternity unit, admissions to the ward shouldbe stopped, and the ward emptied of patients andfumigated before being reopened. This may seem adrastic course, but half-measures usually result inthe establishment of an endemic infection in theunit 20 : no antibiotic has been discovered whichcan be used for chemoprophylaxis without subsequentbacteriological relapses. In the larger cities hospitalsmight arrange to help each other if an epidemicbecame very extensive.

Poliomyelitis ProblemsTHIS year poliomyelitis has been much in everyone’s s

thoughts. It is now established that some batches ofpoliomyelitis vaccine used in the U.S.A. last springcontained live poliomyelitis virus.21 It is also clear fromthe technical report by the United States PublicHealth Service 22 that the manufacturing laboratories,including the laboratory whose vaccine was found tocontain live virus, had conformed with detailed direc-tives which they had been given. Study of the manu-facturers’ experience shows that when in preparingvaccine they treated the virus with formaldehyde,this often did not result in satisfactory inactivation ;so certain lots of vaccine had to be discarded. Thusinactivation by formaldehyde is not so certain as

SALK 23 has supposed, and stricter criteria of virusinactivation are now being applied. The importantquestion whether, in satisfying the new criteria ofsafety, the antigenicity of the vaccine will be retainedhas been exhaustively discussed, and time will showwhether a formula can be devised for preparing aformaldehyde-inactivated vaccine which will combinesafety with antigenicity at least as good as that

already described.24The incidence of poliomyelitis in this country has

been high this summer. One interesting feature hasbeen the rather high proportion of non-paralytic cases-for example, in the Willesden area-and this hasaroused doubts whether some of the cases could bedue to another virus. Dr. McMATH, the superintendentof Neasden Hospital, whose letter we publish onp. 924, does not feel that there is any reason to suspecta different aetiology. Dr. BREEN (p. 923) finds that theproportion of non-paralytic to paralytic poliomyelitisin the South Middlesex Hospital is very much lowerthan that in Neasden Hospital. This difference is

unlikely to be due entirely to differences in reporting ;and it is more probable that differences in polio-myelitis-virus type or virulence are responsible. InLowestoft an outbreak of an infectious illness occurredwhich Dr. A. C. GEE, the medical officer of health,believes was not poliomyelitis.25 5 Infected childrenhad neck rigidity, fairly high fever, and vomiting ;they recovered fully, usually within a week. This

may have been an outbreak of non-paralytic polio-myelitis due to a virus of low virulence, or it mayhave been caused by an unrelated virus ; laboratoryinvestigations will be necessary to decide this point.Laboratory investigations will be needed, too, toestablish the cause of the outbreak this summer at the

Royal Free Hospital. The clinical picture 26 was remi-20. Singer, E. Med. J. Aust. 1954, ii, 129.21. Peterson, L. J., Benson, W. W. J. Amer. med. Ass. 1955, 159, 241.22. See Publ. Hlth Rep., Wash. 1955, 70, 738.23. Salk, J. E. J. Amer. med. Ass. 1955, 158, 1239.24. See Lancet, 1955, i, 851.25. Ibid, Oct. 8, 1955, p. 777.26. Ibid, Aug. 13, 1955, p. 351.

niscent of that in outbreaks of poliomyelitis-like infec-tions in Iceland 27 and later in this country,28 fromwhich no causal virus has yet been recovered.27-29

There have been many recent improvements inlaboratory techniques for isolating and identifyingviruses, and there is a natural desire to investigateoutbreaks of poliomyelitis and related infections asfully as possible. In his letter Dr. BREEN suggeststhat the facilities for routine virus investigations aredeficient. This is a particularly difficult problem.Developments in the production of poliomyelitisvaccines will inevitably strain our resources in viro--logists and technicians ; many will be needed to

prepare poliomyelitis vaccine commercially, and to,

supervise the elaborate safety tests ; and if polio-myelitis vaccines are tested in this country there willbe further deman.ds on man-power for assessing theirefficacy. Inevitably these needs will be met first, andinvestigations of outbreaks of poliomyelitis or polio-myelitis-like infections will suffer. Dr. BREEN’S pleafor further " routine " virus laboratories is well

justified ; but the need can be met only by augmentingour small band of virologists ; and this will take time.

Second Thoughts on GlucagonGLUCAGON-the hyperglycsemic-glycogenolytic fac-

tor (H.G.F.) of the pancreas-raises the blood-sugar inman and animals by stimulating glycogenolysis inthe liver.30-32 So much has been proved. But theadditional hypotheses that it is derived from the &agr;-cellsof the pancreas,33 and is a hormone which acts as a" powerful insulin antagonist," 34 are being challenged.GOLDNER et al. 35 were able to extract glucagon from

the pancreas of dogs treated with alloxan to destroythe &bgr;-cells and with cobalt to destroy the a-cells, andthey now report 36 repetition of these experiments indogs whose pancreatic ducts had been ligated three tofive months before the treatment by alloxan andcobalt. These animals showed considerable acinar

fibrosis ; and, though extracts of their pancreas stillhad hyperglycaemic activity, this was reduced in

degree. The investigators therefore raise the questionwhether the hyperglycaemic factor does not come fromthe excretory parenchyma of the pancreas. Glucagoncan also be extracted from the gastric mucosa and theduodenum. 37

The experiments of VAN ITALLIE et al. 38 suggestthat glucagon does not inhibit the peripheral utilisa-tion of glucose, and therefore cannot be regarded as inany essential sense an antagonist of insulin. By astudy of the glucose levels in capillary and vein afteran infusion of glucagon they demonstrated that theblood-sugar values in the falling limb of the curvewere appreciably lower in venous than in capillary27. Sigurdsson, B., Sigurjonsson, J., Thorbelsson, J., Gudmunsson,

K. R. Amer. J. Hyg. 1950, 52, 222.28. Lancet, 1954, ii, 1060.29. Pellew, R. A. A., Miles, J. A. R. Med. J. Aust. 1955, ii, 480.30. Sutherland, E. W., Cori, C. F. J. biol. Chem. 1951, 188, 531.31. Myers, J. D., Kibler, R. F., Taylor, W. J. Fed. Proc. 1952, 11,

111.32. Kirtley, W. R., Waife, S. O., Helmer, O. M., Peck, F. B. Diabetes,

1953, 2, 345.33. Vuylsteke, C. A., Cornelis, G., de Duve, C. Arch. int. Physiol.

1952, 61, 273.34. de Duve, C. Lancet, 1953, ii, 99.35. Goldner, M. G., Volk, B. W., Lazarus, S. S. J. clin. Endocrin.

1954, 14, 184.36. Lazarus, S. S., Volk, B. W., Goldner, M. G. Metabolism, 1954,

3, 449.37. Sutherland, E. W., de Duve, C.- J. biol: Chem. 1948, 195, 663.38. Van Itallie, T. B., Morgan, M. C., Dotti, L. B. J. clin. Endocrin.

1955, 15, 28.