Dysentery: Some personal experiences and observations

(The previous number of these Transactions, Vol. 51, No. 5, was published on 23rd September, 1957).

OPENING MEETING OF THE FIFTY-FIRST SESSION of the ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE,

held at Manson House, 26, Portland Place, London, W., on Thursday, 17th October, 1957, at 7.30 p.m. The President : Brigadier J. S. K. BOYD,O.B.E., LL.D., M.D., F.R.C.P., F.R.S., in the Chair.

PRESIDENTIAL ADDRESS

D Y S E N T E R Y : SOME P E R S O N A L EXPERIENCES A N D OBSERVATIONS BY

J. s. K. BOYD During the last ten years, my association with tropical medicine has been mainly of an administrative nature and the research work I have been doing does not provide material suitable for an address such as I have to give this evening. It has, therefore, been necessary for me to dig into the past to find a subject. During my overseas service in the Army I had the opportunity of seeing a great deal of dysentery of all kinds, epidemic and endemic, and had the good fortune to take part in certain interesting investigations. Looking back on these experiences, it has seemed to me to be worth while to review them briefly, partly because there have been some important changes in these forty years and the story I have to tell of the earlier days may be of interest to the younger generation, and partly because some of the lessons learnt in the past seem to be in danger of being overlooked or forgotten. My address this evening will therefore be, not a scientific exposition, but a somewhat rambling, semi-historical narrative. I have been looking through the records of the Society and I find that, since meetings were resumed after the war, four evenings have been devoted to amoebiasis and amoebic dysentery and none at all to bacillary dysentery. The reason for this is obvious. Amoebic dysentery, unless very thoroughly treated, tends to persist as a subacute or chronic condition, and is frequently seen in this country in patients who have acquired the infection in the tropics. Bacillary dysentery, on the other hand, rarely becomes chronic, and in any event is easily cured by modern drugs and antibiotics. Nevertheless, those with limited experience of the tropics might conclude, from the attention which it is receiving, that amoebic dysentery is a more important disease than bacillary. As far as Europeans resident in the tropics

472

DYSENTERY

are concerned and, in particular, the British A r m y - especially when under field service conditions - - this is not so. Amoebic dysentery - - and by that I mean dysenteric symptoms produced by Entamoeba histolytica and not merely symptomless infection with this protozoon - - i s much less common than bacillary dysentery. It occurs scattered at random through the community and, while in certain circumstances a few associated cases may be reported, there is no such thing as epidemic amoebic dysentery comparable to epidemic bacillary dysentery. There are, of course, localities in which the incidence of amoebic dysentery is high and the type of infection severe, as, for example, among certain native tribes in South Africa. Again, in the last war, many cases from the Burma front were of great severity, d u e - - a s was shown by HARGREAVES ( 1 9 4 5 ) - - t o heavy secondary infection. These, however, cannot be regarded as epidemics. On the other hand, bacillary dysentery has always a high nuisance value. In the absence of specific treatment, which fortunately is now available, cases occurring even in non-epidemic periods may occasionally be fatal, and in infants must always cause the greatest concern, while under the adverse conditions often associated with active service the disease may develop epidemic proportions with a relatively high mortality rate. Despite the fact that it is usually an easy matter to differentiate amoebic and bacillary dysentery, it is surprising how often these two conditions have been confused and how frequently, on quite unsatisfactory laboratory evidence, bacillary dysentery has been mis-diagnosed as amoebic and treated accordingly. I shall give some instances of this in the course of my narrative. I have mentioned that we now have specific treatment for bacillary dysentery. But I would remind you that bacteria can become resistant to sulphonamides and to antibiotics, and it is by no means impossible that the emergence of resistant strains of dysentery bacilli may jeopardize the present satisfactory position. I have recently made some enquiries on this point through the Consulting Physician to the Army. He informs me that resistance to sulphonamides has not so far been a problem but that there is evidence suggesting that strains having this property may be developing in some localities. Sonne's bacillus is, of course, resistant to sulphonamides but does not produce the type of dysentery which I am discussing now. DYSENTERY IN MACEDONIA, 1916 TO 1918

I first encountered the dysentery problem in Macedonia in 1916. It is always well, when studying a subject, to begin at the beginning, and it was my fortune to do so - - quite inadvertently. I began at what I believe our commercial colleagues would call the " consumer end," during a trek across the Chalcidean peninsula, when I was medical officer to a company of engineers. Fortunately it was not a very virulent organism which attacked me, for I recovered without going into hospital, and was none the worse for it ; but it enabled me to learn, from personal experience, the signs and symptoms of bacillary dysentery. It is perhaps worth recording that from that day to t h i s - although I have undoubtedly been exposed on many occasions to infection - - I have never been re-infected. My attack occurred early in the first of three waves of epidemic dysentery which swept through the Salonika Army in the three succeeding years, the figures for which are shown in Table I. In 1916 and 1917, there were just under 6,000 reported cases and in 1918 about 9,000. The majority of these were acute, and a number were fatal. Exact death rates are not available, but according to the Official History of the War were somewhere between 2 per cent. and 3 per cent. Many infections became chronic, and the patients

J. s. K. BOYD TABLE I.

473

Incidence of dysentery (both bacillary and amoebic). 1916

Total cases

Admissions to Medical Units.

1917

Ratio per 1,000

Total cases

1918

Ratio per 1,000

Total cases

Ratio per 1,000

Sa!onika

5,987

63.89

5,842

28.89

9,318

58.23

Egypt

5,599

31.19

4,341

23.13

4,906

21.80

Mesopotamia

1,839

50.94

4,960

60.34

5,455

51.12

(Adapted from a Table in the Official History of the War (1914-18) Medical Services, Diseases of the War, Volume 1). had to be invalided from the Command to Malta and to the United Kingdom. T h e length of time during which some of these cases remained under treatment is rather startling. Figures compiled by the Medical Research Council* from index cards and admission and discharge books show that 330 cases from Salonika averaged 250.5 days before the subjects were fit to return to duty, while 2,319 cases from Gallipoli, of all grades of severity, averaged 75.6 days. In addition to the reported cases shown in the Table, there were many others which never reached medical units. In fact it would not be far wrong to say that no one who served in the Salonika Army escaped infection. At the time, dysentery took second place to the malaria epidemic which decimated the army, but it was nevertheless a serious drain on the efficiency of the fighting forces. T h e loss of manpower was considerable, and the discomfort and misery suffered by the men had a bad effect on their morale. In civilized surroundings, even a mild attack of dysentery is an unpleasant experience : on active service, with no amenities and no proper diet, the term " unpleasant " is a gross understatement. T h e explanation of these epidemic waves was obvious. Those were the days of horse and mule transport. In consequence, manure lay about everywhere, and M u s c a domestica enjoyed an era of uninhibited fertility. In fairness, it must be said that, within the lines of established units, a reasonably high state of hygiene prevailed, but elsewhere there were extensive areas which came under the supervision of sanitary squadS, too few in n u m b e r to tackle the problem. T h e result was a veritable plague of flies, which swarmed in unbelievable numbers. I recall in particular one occasion in the summer of 1916 when, after a night march, we came at dawn to a seemingly clean and pleasant transit camp. T h e first few hours were passed in comfort, but as the sun grew hotter the flies came to life and covered us, the mess table, the f o o d - everything. T h e r e was no escape from them, and it was well nigh impossible to eat without getting them into the mouth. T h e i r habits are well known - - I believe it was Sir Andrew Balfour who coined the alliterative phrase " the filthy feet of faecal feeding flies." Although experiments were carried out which showed that flies did in fact carry dysentery bacilli, both on their feet and in their gut, these did no more than prove the obvious. One example will illustrate the sort of problem with which we were faced. In the summer of 1918, the General Hospital, to which I was pathologist, was situated on the Kalamaria Peninsula, a few miles south of Salonika. This should have been a salubrious Quoted in the official History of the War,

474

DYSENTERY

ske, but unfortunately we had next to us a French veterinary hospital, whose staff had rather primitive ideas about the disposal of manure. As a result, when the warmer weather came, our hospital was invaded by countless flies. Cases of bacillary dysentery soon appeared among both staff and patients, and in a week or two the outbreak developed to such an extent that the patients had to be sent elsewhere and the hospital closed. Meantime representations were made at a high level, and the veterinary unit was moved. The skeleton staff left in the hospital was given the task of cleansing the Augean stable. In the compound of the veterinary unit this was a comparatively simple matter, and consisted in raking together the wide-spread collection of manure and burning it on improvised incinerators. But this proved to be only a small part of the problem. The camps were adjacent to the beach, and indeed only separated from it by sandy cliffs some 30 to 50 feet high. We found that enormous quantities of manure had been thrown on to the shore in the hope that it would be borne off by the tide. However, the tide in the Gulf of Salonika rises and falls only a few inches, and so there was a mound of manure several hundred yards long, in which the flies, undeterred by the salinity of their surroundings, were breeding freely. The task of incinerating this sodden mass took several weeks, and it was a long time before the hospital site was declared fit for reoccupation. By the spring of 1917 I had ceased to be a regimental medical officer and was in charge of a Mobile Laboratory attached to a Casualty Clearing Station through which the casualties of two Divisions passed on their way to base hospitals in Salonika. Here, as elsewhere along the front, malaria dominated the picture, but dysentery was also rife. The conditions for carrying out bacteriological tests were excellent. The wards were near the laboratory, the ward orderlies were very co-operative in bringing fresh specimens in bedpans, and as we were to all intents and purposes permanently on duty, there was never any delay in getting specimens plated. We were consequently successful in isolating dysentery bacilli from a high percentage of the stools, while at the same time microscopic examinations for amoebae were consistently negative. Thus it was possible to say with confidence that the dysentery was bacillary in type and that amoebic dysentery was practically non-existent. While the mucus or muco-pus from the stools was being searched for amoebae, the very characteristic cytological picture which it p r e s e n t e d - unexpected cellularity, a preponderance of pus cells, varying numbers of red blood cells scattered among them, and occasional shed epithelial cells and large refractile m a c r o p h a g e s - attracted attention. When our late President, Dr. C. M. Wenyon, came from Egypt to Salonika later in the year and worked in my laboratory I discussed this with him and was greatly interested to find that he agreed with me that such a picture was more or less diagnostic of bacillary dysentery. WILLMORE and SHEARMAN (1918) record a similar observation in the Lancet, this being the first time it appears in print, for it is not mentioned in the 6th edition of Manson's Tropical Diseases, published in 1917, though it appears in the 7th edition in 1921. Over the years, my opinion on the significance of this cytological picture has not changed. No other acute diarrhoeal condition of sudden onset produces muco-pus with these characters, and when both this and the clinical signs and symptoms are present, a firm diagnosis of bacillary dysentery can be made with confidence. Obviously the cytology gives no clue to the particular type of infecting organism. Owing to a shortening of our front, my mobile laboratory was closed down in autumn, 1917, and I spent the rest of my time in Salonika as pathologist to a general hospital. Here for some months I worked two mornings a week side by side with the army protozoologist, looking for amoebae in the stools of subjects of chronic dysentery. Unfortunately I have

J. $. K. BOYD

475

no record of the total number of cases examined nor of the number of amoebic infections we found, but there were certainly not more than five positives. I mention this in confirmation of our findings during the epidemic season and to emphasize the fact that E. histolytica played a trivial part in the dysentery which occurred in the Salonika army. This personal experience is confirmed in the Official History of the War, where it is stated that amoebic dysentery comprised about 3 per cent. of all the cases of dysentery admitted to hospital during the campaign. DYSENTERY IN GALLIPOLI,EGYPT, AND MESOPOTAMIA

It is of interest to review briefly the incidence of dysentery during these years in the adjoining theatres of war, and for this purpose I have made use of the Official History of the War, from which all the facts and figures which follow- - unless otherwise stated - - have been abstracted. As regards the Gallipoli campaign, the position is somewhat confused. There are no exact statistics, but it is stated that in the 3 months from August to October, 1915, a high proportion of the 120,000 casualties evacuated from the Peninsula was caused by dysentery, and it is claimed that 65 per cent. of these cases were due to amoebic infection (ARCHIBALD, HADFIELD, LOGANand CAMPBELL, 1916) : but in November and December, 1915, and January, 1916, amoebic dysentery disappeared and the incidence of the bacillary type increased. These findings were received at home with considerable scepticism, which was justified by subsequent investigation of the convalescents carried out by Lt.-Col. C. M. Wenyon, as he then was, and Captain F. W. O'Connor, who were sent for this purpose. They found that, of 246 men who had been evacuated from Gallipoli with a history of dysentery, only 6.5 per cent. were carriers of E. histolytlca cysts, while among 1,137 men with no history of dysentery, 4.5 per cent. were carriers (WEN¥ON and O'CONNOR, 1917). Various other investigations in the Eastern Mediterranean area produced results in conformity with these figures, while serological tests carried out at home on the blood serum of convalescents gave evidence that the majority had suffered from bacillary dysentery. The considered opinion expressed in the Official History is that, while amoebic dysentery undoubtedly occurred in Gallipoli, most of the cases, as elsewhere, were bacillary. In Egypt, the incidence of dysentery of all types was considerably lower than in Salonika, as can be seen from Table I, while the admissions to hospital of cases of amoebic infection varied, in different years, from 2 per cent. to 7 per cent. of all dysentery cases. Thus a careful examination of 659 active cases of dysentery and diarrhoea by the protozoologist to the Mediterranean Expeditionary Force (WooDCOCK, 1917) revealed that 7.7 per cent. were amoebic. However, when this group was broken down into British troops and Indian troops, only 1.9 per cent. of the British cases were amoebic, while 15.7 per cent. of the Indian cases were due to this infection. Examination of the native population of Egypt revealed that a high percentage--13.5 per cent. of 524 healthy natives--were passing cysts, and from this it was concluded that amoebic dysentery was of more frequent occurrence among them than among British troops. In the light of recent work which suggests the existence of non-pathogenic strains of E. histolytica it is by no means certain that this is a valid conclusion. In Palestine, figures ran at approximately the same level, some 7 per cent. of all cases of dysentery and diarrhoea being amoebic in origin. In Mesopotamia, the incidence of amoebic dysentery was higher than elsewhere, and in one series of 309 patients suffering from bacillary dysentery, passing blood and mucus,

476

DYSENTERY

26 per cent. showed active amoebae containing red blood corpuscles in their stools (BONEY, CROSSMAN and BOULENGER, 1918). These, however, were cases which had reached a base hospital. In the forward areas, the incidence of amoebic dysentery was low, though somewhat higher in Indian than in British troops, and there was clear evidence that, in the spring and autumn outbreaks, the dysentery was bacillary in type in both categories. The Medical Advisory Committee reported that " bacillary dysentery is the predominant type in Mesopotamia and there is little doubt that, had the cases investigated been in the main local admissions instead of transfers from up-river, the proportion of bacteriologically proved cases would have been still higher." Amoebic dysentery was, however, responsible for a considerable proportion of the steady admissions all the year round (LEDINGHAM, 1920). Thus, the records from the different fronts all tell a similar story. Amoebic dysentery had a random distribution, and admissions to hospital ran at a relatively constant level throughout the year. In the chapter on Amoebic Dysentery in the Official History of the War (Pathology), which was written by Dobell and Harvey, the authors, after a critical review of all the evidence, find themselves able to state categorically that " epidemics of amoebic dysentery do not occur." In contrast, bacillary dysentery occurred in epidemic waves, mainly in late summer and autumn, but to a lesser extent in spring, and was everywhere the predominant type. THE SYMPTOMLESS CARRIER OF Entamoeba

histolytica

During and immediately following the 1914-1918 war, numerous experienced protozoologists, prominent among whom was the late Clifford Dobell, did a vast amount of work on the problems of amoebiasis, and in particular on the significance of the symptomless carrier of E. histolytica. As the result of this work, it came to be realized for the first time that E. histolytica is a common parasite in Britain and in other temperate regions as well as in the tropics, it being estimated that several million of the inhabitants of Britain, who had never been overseas, were infected. Nevertheless, disease caused by this parasite is practically unknown except in those who live, or have lived, in subtropical and tropical countries. The concept that some races of the amoeba are invasive, while others are not, was not generally accepted. Thus DOaELL (1919) expressed the opinion that it was certain, from what was known of the life history of the parasite, that erosion of the lining of the large bowel must occur even in the healthy carrier, in whom a state of balance is struck between the regenerative powers Of the host and the destructive powers of the parasite. Some years later, it should be noted, he revised his views because of results obtained in artificial culture and in experimental infections of monkeys (DoBELL, 1928, 1931), and admitted that E. histolytica could live i n man as a harmless commensal. However, in 1918 the earlier dictum was generally accepted, but in spite of this it was decided, as an official policy, that the laborious detection, segregation, and treatment of amoebic carriers was, from the military standpoint, inexpedient, and that, though care should be exercised in the case of those who had shown special susceptibility to the parasite, carriers could be set free in a community without exposing either themselves or the community to any grave risk. The soundness of this decision is borne out by the fact that there was no outbreak of amoebic dysentery after either the first or the second World War. According to HOARE (1950), the total number of clinical cases recorded in Great Britain in the course of thirty-five years among those who have never been abroad does not exceed three dozen.

J, S. K.

BOYD

477

PROGRESS IN THE CLASSIFICATION OF DYSENTERY BACILLI

The first World War saw considerable progress in the identification and classification of the dysentery bacilli. At the beginning, Shiga's b a c i l l u s - a non-mannitol fermentor with a distinctive antigenic structure and a potent e x o t o x i n - was clearly described and firmly established. Mannitol-fermenting types of related but not identical antigenic structure had been discovered by Flexner and Strong, and it was the practice to group them together as " Flexner-Y." Experience soon showed that Flexner-Y was a larger group than was originally thought, and many bacilli were isolated in the different theatres of war which were obviously closely related to the described types but were of different antigenic structure, being inagglutinable by specific antiserum. Numerous a t t e m p t s - none very s u c c e s s f u l were made to characterize and classify these inagglutinable strains, until ANDREWES and INMAN (1919) made a detailed study of a collection of 200 strains, and divided them into five types. This was an important advance, though not the final solution of the problem. AMOEBIC AND BACILLARY DYSENTERY IN INDIA

In the years between the wars, I was chiefly concerned with dysentery as it occurred in India. While most bacteriologists who worked in the Mediterranean area during the First World War were left in no doubt that bacillary dysentery was much more common than amoebic dysentery, this idea was by no means acceptable to their colleagues in India. One the whole, this is not surprising. Under field conditions in war, the dysentery wards and the laboratories were usually in close proximity, so that stools when received in the laboratory were freshly passed, and in consequence there was a high percentage of isolations of dysentery bacilli. In India, however, one laboratory had to serve several hospitals, and fresh specimens were the exception. It must be admitted, too, that the laboratories were not very welt equipped for this particular kind of work, nor did their staffs show a great deal of enthusiasm. The result was that dysentery bacilli were rarely found. To make matters worse, a significant proportion of such bacilli as were isolated, although conforming to the standard types in biochemical reactions, were inagglutinable with the available diagnostic antisera, and consequently open to suspicion. On the other hand, the identification of E. histolytica lay with the microscopist. No rigid tests had to be passed, and there can be no doubt that the criteria of identification were often laxly applied, yet a positive report could not be ignored. Furthermore, an inconclusive report on the microscopic findings in a specimen from which no dysentery bacilli had been isolated left the clinician to exercise his own judgment. In fact, the "amoebic school" held all the cards, and were inclined to scoff at those who disagreed with their views. This was the unsatisfactory state of affairs which existed when an active crusader, Major (now Major-General, retired) J. A. Manifold, arrived in India in 1924 and launched the attack which was to put an end to it. His work has not, in my opinion, received the credit it deserves, for there is no doubt that he, more than anyone else, was responsible for convincing the sceptics that bacillary dysentery was the prevalent type in India as elsewhere, first, by his own extensive investigations (MANIFOLD and DEMONTE, 1928), and later, by the active part he played in re-organizing and re-equipping military laboratories throughout the Army and in arranging for efficient training of the staff. His enthusiasm and persistence had the desired effect, and in the course of a few years the standard of the bacteriological work carried out in the laboratories reached a high level and the returns from the hospitals

478

DYSENTERY

came to present a reliable picture of the type of dysentery occurring in the Army in India - a picture not very different from that seen in the Mediterranean Commands during the war. This is well illustrated by some figures given in the Annual Report of the Public Health Commissioner with the Government of India, 1935, where the incidence of dysentery, diarrhoea, and colitis for the years 1920 to 1925 is compared with that for the years 1930 to 1935 (Table II). For the benefit of those unversed in army nomenclature, I should perhaps TABLE II.

Incidence of dysentery in the Army in India.

British Other Ranks

Average strength

Indian Other Ranks

1920-1925

1930-1935

1920-1925

1930-1935

59,218

55,613

120,771

120,771

PercentPercentPercentage (exage (exage (exclusive of Number clusive of Number clusive of diarof diarof diarrhoea) cases rhoea) cases rhoea)

Percentage (exNumber clusive of of diarcases rhoea) Bacillary dysentery

~

5,470

63.7

Amoebic dysentery

3,176 ~

1,079

12.4

3,272

497 ~

1,987

22.8

4,281

Colitis

1,420 ~

187

2.1

15,485

Diarrhoea

5,878

--

14,627

Clinical dysentery

Total

11,373

4,960

241 ~

37,906

18.4

--

7,892

67.8

814

7.0

2,767

23.8

167

1.4

5,065

--

16,705

(Adapted from a Table in the Annual Report of the Public Health Commissioner with the Government of India, Part II, 1935). explain that clinical dysentery is diarrhoea in which the stools contain mucus of indefinite character, but in which neither entamoebae nor dysentery bacilli are found. In " simple " diarrhoea, the stools contain no mucus. Colitis i s - or should b e - a non-dysenteric inflammatory condition of the mucous membrane of the colon. T h e striking features in this Table are the greater precision of diagnosis and the switch from amoebic to bacillary infection. I f the percentage of clinical dysentery appears to be high, it must be remembered, first, that the very selective desoxycholate m e d i u m - now used for isolating dysentery bacilli - - had not then been devised, and second, that a considerable proportion of the cases occurred in outstations from which specimens for laboratory tests had to be sent by post. T h e relative incidence of amoebic dysentery in British troops and in Indian troops is of interest. T h e figures for the years 1930 to 1935 are seen in Table II. T h e rather surprising feature is that amoebic dysentery was significantly more common in British than in Indian troops, a complete reversal of the state of affairs found in Mesopotamia during the war. T h e figures relating to Indian troops are of particular interest because they show that active amoebic dysentery is not a disease of much importance in the indigenous population of a country where pathogenic races of the amoeba are of common occurrence. Unfortunately these figures are not correlated with the corresponding incidence of symptomless carriers.

J. S. K. BOYD

479

Lest it be said that the diagnosis of amoebic dysentery was more realistic under the earlier r~gime and that in later years cases were being missed, I will give one further set of figures--those dealing with hepatitis (all causes) and liver abscess (Table III). The TABLE III.

Incidence of hepatitis and liver abscess in the Army in India. Hepatitis, all causes (including amoebic)

Liver abscess (amoebic)

1920-1925

1930-1935

1920-1925

1930-1935

British troops

948

417

184

46

Indian troops

681

358

84

37

(Adapted from a Table in the Annual Report of the Public Health Commissioner with the Government of India, Part II, 1935). significant fall in the incidence of liver abscess affords conclusive evidence that there has been no increase in missed cases of amoebic infection. In terms of the numbers at risk, amoebic abscess is more than twice as common in British troops as it is in Indian troops, a figure which is in general agreement with the incidence of amoebic dysentery in the two races. In general, the dysentery which occurred during my years in India (where, as I filled administrative appointments, I had an opportunity of seeing the over-all picture) was endemic in nature. A few small outbreaks occurred, attributable to one particular type of dysentery bacillus, but this was quite exceptional. On no occasion did amoebic dysentery arise, other than in the form of sporadic cases. At all times, bacillary dysentery predominated. THE INAGGLUTINABLE FLEXNER DYSENTERY BACILLUS

When I returned to India in 1929 for a second tour of duty, the problem of the inagglutinable bacilli of the Flexner group had still to be solved, and I was given the opportunity of working on it. I have previously read a paper to this Society on this subject (BoYD, 1940), and need not repeat the findings in detaik As so often happens, the key to the problem was found quite accidentally, when I observed two types of colony on a plate made from a culture of one of the " inagglutinable " types. One of the colonies, which was smooth, w a s - like the parent s t r a i n - composed of inagglutinable bacteria. The organisms in the other, which was semi-rough in appearance, were highly agglutinable. After a good deal of experimentation it became clear that this strain had undergone degradation , and was producing two different types of bacteria. One had the characters of the parent organism, producing smooth colonies, and possessing a specific surface antigen which rendered it inagglutinable by the antisera of the known strains. The second type, which produced rough colonies, had lost this surface antigen, and as a result, an underlying group antigen had been exposed, with components common to the other members of the Flexner series, so that it had become agglutinable by their antisera. By techniques familiar to all serologists it became possible to recognize the different members of the group by their specific surface antigen and in this way to identify three new types which comprised most of the hitherto inagglutinable Flexner strains. A few remained, each with a specific antigen but lacking the common group antigen, and these were classified as a separate species. Experience

480

DYSENTERY

has shown that most of the latter are pathogenic and capable of producing symptoms of dysentery, but only three of them are sufficiently common to be of any importance.

DYSENTERY IN MIDDLEEAST, 1940-1943 The final chapters of my story have their setting in Middle East during the second World War, where my introduction to the dysentery problem was abrupt and disturbing. Together with a group of consultants going to form Army headquarters, I reached Cairo in mid-August, 1940. A fortnight later I accompanied the Consulting Surgeon on a visit to the hospitals in Alexandria. On the outward journey he complained that he felt unwell and, when returning in tile evening, he had some diarrhoea and abdominal pain, a common complaint in those recently arrived in Egypt. Without being seriously ill, he showed no improvement in the next 24 hours, and was admitted to hospital, where Shiga's bacillus was isolated from his stools. Despite the most careful nursing and the best treatment as we knew it then, including massive doses of Shiga antitoxin, he grew slowly but steadily worse and died in a little over a fortnight. About this same time, the medical specialist of a newly-arrived hospital was similarly infected and, after a more protracted illness lasting just over 8 weeks, he too died. This type of illness - - slow in onset and progress, but quite unamenable to treatment - was new to me. In Salonika in the first war, the fatal cases were mainly fulminating in nature, the patient being violently ill from the first, and usually dying in a few days. In India, Shiga infections were by no means uncommon but, although always regarded as more serious than Flexner infections, if taken early they responded to treatment. We soon found that in Middle East the majority of cases ran this milder course, but from time to time others occurred, such as the two I have just described, which were solidly resistant to treatment, and progressed steadily until the patient died. Most of the more serious cases were caused by Shiga's bacillus, but a few Flexner infections also ran a very protracted course and were extremely difficult to cure.

TREATMENT OF BACILLARYDYSENTERY These cases,and in particular the Shiga cases, gaveus a feeling of completehelplessness, for at this time we had no adequate treatment. I have not so far discussed treatment, for I have had nothing important to say about it. Longbefore the first World War the recommended procedure was repeated doses of sodium or magnesium sulphate, and this was still standard in 1940. It resulted in more copious stools and probably lessened the discomfort and tenesmus engendered by attempts to void small quantities of sticky mucus, but there is no doubt that when used indiscriminately--by rule of t h u m b - it prolonged the course of a mild attack. It certainly had no specific action and had no curative effect on these deadly Shiga infections. T h e only specific treatment then available was antiserum for the Flexner group of infections and antitoxin for Shiga. The value of the former was very doubtful, partly because Flexner infections were rarely acute enough to warrant its use, a n d partly because the antiserum often did more harm than good by producing serum sickness 10 days after administration. Nevertheless, we gave careful consideration to a suggestion that it should be used in all severe cases, in the faint hope that it might have some antibacterial action against all dysentery group organisms. But this policy was not put into action.

I. s. K. BOX'D

481

Shiga antitoxin falls into a different category. It is a definite entity, capable of quantitative titration. About the middle of 1940 we received supplies of a highly purified antitoxin containing 20,000 units per ml. which could be given intravenously in doses up to 10 ml., i.e. in quantities of 200,000 units. This antitoxin was repeatedly administered in the two fatal cases already mentioned. Its action was unmistakeable, but unfortunately transient. The patients felt better, the temperature fell, and abdominal colic was lessened, but in 24 to 48 hours, absorption of fresh toxin from the bowel lesions had restored the status quo. In a word, this preparation was a good antitoxin but, as it had neither bacteriostatic nor bactericidal action, it failed to effect cure in established cases. We had the impression that when given in the early stages of the illness it tipped the balance in the patient's favour and gave his defensive mechanism a better chance to overcome the infection. But, as many cases recovered just as quickly without antitoxin, it is impossible to say with certainty that there was any such action.

INTRODUCTION OF 8ULPHONAMIDE TREATMENT

This unsatisfactory state of affairs continued throughout 1940 until, in a dramatic way, the situation was completely altered by the introduction of sulphaguanidine. At this time I was working in close co-operation with the Consulting Physician for Tropical Diseases, who was on loan from the Australian Army and who was none other than our ex-President, Sir Neil Hamilton Fairley. We had access to the cases in a large General Hospital which at that time took in most of the dysentery occurring in British troops stationed in the Cairo area. These patients were treated in dysentery wards under the charge of Captain R. P. Hendry, R.A.M.C., to whose careful notes I am indebted for many of the details which follow. In the years before the war, Dr. G. A. H. Buttle, now Welleome Professor of Pharmacology in the School of Pharmacy of the University of London, was engaged in testing new sulphonamide compounds and had passed on samples to Dr. E. K. Marshall at Johns Hopkins Hospital. One of these, ultimately and correctly described as sulphaguanidine, was found by Marshall to be slowly absorbed from the intestine, and because of this he thought that it might be useful in intestinal infections. He accordingly used it to treat some mild cases of dysentery and enteritis in children, with so much success that he had a large batch prepared and sent it to Buttle, who was then in Cairo. Buttle in turn handed it over to Colonel Fairley and me. We were only too pleased to have something new to try in the more serious acute cases, as well as in refl'actory and long-standing infections, of which there were at that time a number in the dysentery ward (FAIRLEY and BOYD, 1943). The results of treatment with sulphaguanidine in the acute cases were at once obvious. There was early improvement, manifested by relief of abdominal pain, by a decrease in the number of stools passed per day, and in particular by the disappearance of the feeling of all-in misery which is a constant symptom in most cases of dysentery. The stools became normal in a few days, appetite returned, and the patient was fit for discharge from hospital in a much shorter time than with any other form of treatment. Most important of all, the occasional resistant and ultimately fatal case of Shiga infection disappeared. Needless to say, some patients reacted better than others, and in a small number of cases results were disappointing for one reason or another ; but we were left in no doubt that we had a specific remedy of great potency. In this first series, however, the most convincing evidence of the value of sulphaguanidine

482

DYSENTERY

came from subacute cases in patients who had been ill for prolonged judged from the following brief account of two of them:

periods, as can be"

The first was a Shiga infection of 80 days' duration, which had proved completely resistant to every form of treatment we had tried ; antitoxin, in this as in other similar cases, had produced only a temporary recession of symptoms. Stools, of which he passed six to I0 daily, were liquid or liquid and semi-solid, always containing mucus or muco-pus, or even frank pus. His pulse and temperature were a|most constantly raised and his general condition was poor. He had no appetite and his weight had gone down to 7½ stones. The sigmoidoscope revealed well-marked generalized inflammation of the bowel mucosa, with a granular surface which bled readily on instrumentation; the lumen was contracted and the bowel difficult to distend. The patient was started on a course of sulphaguanidine on 19th January, 1941, being given 0.5 gramme per kilo of body weight 4-hourly for 48 hours and thereafter 8-hourly for 50 days. The response was immediate. The fever subsided, abdominal colic decreased and disappeared, and in a day or two the patient was demanding food. Four days after treatment was started he passed a semi-solid stool -- the first since the onset of his illness -- and in a fortnight his stools became solid and remained so. A second course of 7 days' treatment -- starting on 6th Februarymade assurance doubly sure, and he had an uninterrupted convalescence. On ]9th February, a month after starting treatment, his weight had risen by just under two stones to nine stones five pounds. Sigmoidoscopy on that day showed a red but relatively normal mucous membrane. A few scattered pitted areas, the site of former ulcers, could be seen, but there was no bleeding and the bowel distended satisfactorily under pressure. The bowel wall had in fact practically healed. The second patient was infected with one of the Flexner group (Flexner If, or "W" according to the old nomenclature). His symptoms were only moderately severe when he was admitted to hospital 8 days after onset, but in spite of treatment he continued each day to pass three or four stools containing blood and mucus for 3 months, and thereafter more irregularly for a further 2 months. During this time the following forms of treatment had been tried -- bismuth subgallate, emetine hydrochloride, quinoxyl enemata, stovarso], eusol enemata, and sedatives such as luminal and bromides, The antiamoebic treatment was given as a forlorn hope, for at no time was there microscopic evidence of amoebic infection. The sigmoidoscope revealed the presence of numerous small aphthous ulcers on an otherwise normal mucosa. A course of sulphaguanidine was started on the 149th day of his illness, and 4 days later the stools were normal. On the 9th day after treatment was started, the ulcers had disappeared and a few small red patches, which did not bleed when swabbed, were the only abnormal features to be seen on sigmoidoscopy. This patient had never been acutely ill, but his continuing symptoms precluded his discharge from hospital. The response to sulphaguanidine was immediate. Having established the efficacy of sulphaguanidine, of using our limited supply to the best purpose, while

we were faced with the problem awaiting the manufacture of bulk

quantities, which we could not hope to receive for many months. As the vast majority of severe cases were caused by Shiga's bacillus, the obvious course was to restrict its use as far as possible to the treatment of patients infected with this more pathogenic organism, and this was accordingly done. This arrangement had the disadvantage of delaying the start of treatment until the causative bacillus had been identified, but to this there was no alternative. Meantime, attempts were made to use other sulphonamide drugs. Sulphanilamide--of which we had ample supplies--was not particularly successful, but sulphapyridine proved reasonably active, though with very unpleasant side-effects. At a later date, sulphathiazole was tried but was found to compare badly with sulphaguanidine. Sulphadiazine was good, as was sulphamethazine (not sulphamezathine), but the latter was never produced in bulk. Our American allies made sulphadiazine their choice, but we could not agree with them about this : we found sulphaguanidine just as satisfactory in producing cure and preferable because it had no tendency to crystallize in the urine and cause blockage of the renal tubules, a definite danger in a climate where there is intense sweating and where, in consequence, the urine is concentrated. As increasing supplies of sulphaguanidine became available, bacillary dysentery ceased to be a source of worry, for when treatment with this drug was given from the earliest stages, no severe cases developed.

I" $" K. BOYD TREATMENT

WITH

483

BACTERIOPHAGE

I must say a little about a rival form of treatment which enjoyed great popularity among the civilian population of Egypt. It will be remembered that it was in Alexandria that d'Herelle discovered bacteriophage and explored its use as a therapeutic agent, recommending it especially for the treatment of bacillary dysentery. A dysentery phage prepared by a Parisian firm had long been on the market, and considerable quantities of this were still in stock in the local chemists' shops in Egypt in 1940. Strong pressure was brought to bear on us to use it, and we were nothing loath to give it a trial. To stifle in advance one criticism certain to be made in the event of failure, our supplies were obtained through one of the ardent advocates of this form of treatment, and phage was administered in accordance with his instructions to a small series of patients under the direct supervision of Captain Hendry. Only cases of less than 72 hours' duration were treated, and alternate cases were placed on bacteriophage and on the standard sodium sulphate treatment. Apart from the fact that, as might be expected, the stools of the control cases on sodium sulphate remained fluid for a slightly longer period than those of the bacteriophage-treated cases, there was no significant difference between the two series and certainly no dramatic cure in the case of the phage-treated patients. An unsponsored and uncensored experiment, which is no doubt open to criticism, but which nevertheless gave very interesting results, was carried out without our prior knowledge. We discovered, quite accidentally, that the inmates of a detention camp for Italian internees, who enjoyed considerable amenities, h a d - true to the local f a i t h - been importing and consuming, both for prophylactic and therapeutic purposes, large quantities of bacteriophage, while in an adjoining prisoner-of-war camp with Italian inmates, run under the same supervision and with the same standards of hygiene, bacteriophage was unknown. The rates of admission to hospital per 1,000 for clinical dysentery are shown in Table IV. Here again, there is no evidence of a bacteriophage miracle. TABLE IV. Rate per 1,000 of admissions to hospital for clinical dysentery. Internee Camp (Phage used)

Prisoner of War Camp (No phage)

May

3.52

3.19

June

8.23

2.47

July

4.88

1.60

August

3.32

1.42

(Reproduced from the Transactions of the Royal Society of Tropical Medicine and Hygiene, 1944, 37, 247). Finally, using therapeutic bacteriophage of German manufacture, captured with other German stores at the Battle of Alamein, we found it possible to organize a large-scale and well-controlled experiment in prophylaxis and therapy (BoYD and PORTNOY, 1944). I need not worry you with details, but once again there was no evidence of any benefit resulting from the bacteriophage treatment. These results are in accordance with expectations based on in vitro experiment. Under the conditions of the test-tube - - ideal for the bacteriophage - - primary destruction of the

484

DYSENTERY

majority of susceptible bacteria is quickly followed by secondary growth of more resistant types which have escaped. In vivo, the facilities for survival are much more favourable than they could ever be in vitro, and although in theory the concentration of bacteriophage might get progressively greater until the few remaining bacteria are overwhelmed, in practice a state of balance is quickly reached. From the stools of the German patients treated with phage, colonies of dysentery bacilli, contaminated with phage but still growing vigorously, were recovered on several occasions. There is an interesting postscript to this story. While in Edinburgh in 1945, I was approached by a young German medical officer who had been taken prisoner and was working in my laboratory while awaiting repatriation. He told me that although the Afrika Korps carried bacteriophage, their medical officers were quite convinced that it was useless. More interesting than this, however, was a story he had to tell about the parachute troops who dropped into Crete in 1941. Each man carried a package of sulphonamide t a b l e t s - I do not know which particular preparation it was. They were told to s a y - if they were captured - - that the tablets were to be taken if they were wounded, to keep the wound sterile. In fact, however, they were instructed not to use them for this purpose but to keep them for treating dysentery, should they develop this disease. AN ABORTIVE EXPERIMENT I N IMMUNIZATION

Prevention is always better than cure, and we gave careful consideration to the possibility of using a dysentery vaccine. Assuming that such a vaccine might be effective, there were two major obstacles to preparing i t . The first was the marked toxicity of Shiga's bacillus, and we decided to circumvent this by using, not the bacteria themselves, but a formalized toxoid of the kind which had proved so effective in immunization against diphtheria and tetanus. The second was the multiplicity of types of Flexner bacilli, for it was not feasible to prepare a polyvalent vaccine containing adequate quantities of each type. Ultimately we decided to use only the most common type, Flexner II (previously "W") in the hope that this would create cross-immunity sufficient to cover the others. As we could not prepare Shiga toxoid locally, a cable was sent to the War Office asking for a litre for a pilot trial. In due course this arrived, a carton containing 20 x 50 ml. bottles, but apparently no covering letter. After waiting a few weeks and still getting no letter with particulars of the toxoid, it was decided to go on with the trial, and two guinea-pigs were each injected intraperitoneally with 1 ml. As they showed no reaction of any kind in 48 hours (nor indeed did they at any time subsequently) I next injected 0.5 ml. into my left upper arm. Twenty-four hours later there was only a trifling local reaction, which conformed to my experience with tetanus toxoid, so two volunteers on the laboratory staff were each given 0.5 ml. Twenty-four hours later - - i.e. 48 hours after my dose and 24 hours after the two volunteers had theirs - none of us had any significant reaction, and accordingly two further volunteers were given doses of 1.0 ml. I may say - - as an aside - - that this was obviously one of Colonel Hamilton Fairley's lucky days. He was to have been one of the volunteers but was prevented from coming at the time arranged, so that his place was taken by another of the laboratory technicians. The following day, the third after my inoculation, I noticed my arm becoming really painful and stiff, and at bedtime I had a slight rigor. The two men who had 0.5 ml. 48 hours previously were also beginning to develop a local reaction. The next day all five of us had local reactions of varying severity, but not sufficient to be taken very seriously. The two who had 0.5 ml. 72 hours previously were excused duty, but were well enough to

j. s. K. BoYD

485

go out into Cairo on pass. That evening, however, things began to happen. First, I had occasion to turn out the carton containing the bottles of toxoid and found, tucked in the bottom, the missing covering letter, which stated that the material was not completely toxoided, and, although used in large doses for horses, did generally give local reactions. Next, Colonel Fairley's substitute developed a temperature of 103 °, with symptoms of dysentery, and was admitted to hospital. The second volunteer who got 1.0 ml. soon followed suit, also with a high temperature and symptoms of dysentery, and next day the remaining two were taken into hospital because of the local reaction. I remained at work for another day but after that had no alternative but to get myself admitted with a very swollen and painful arm, which looked as if it would break down, but in fact did not. These results somewhat damped our enthusiasm, and as it was clearly impossible to have a safe toxoid ready for use that year, we dropped the experiment for the time being. Because of the success of sulphaguanidine treatment, the mounting supplies of this drug which became available, and our preoccupation with other and more important matters, it was never taken up again. The one lesson we learnt was that partially toxoided Shiga toxin is very slow in producing a local reaction in man, but when the reaction does occur it is apt to be severe. Before leaving the question of immunization against bacillary dysentery, there is one point t would like to make. New divisions arriving in Middle East were always stationed in reception camps for a few weeks to be equipped for desert warfare. Among these " unsalted " troops there was almost invariably an outbreak of bacillary dysentery of varying degrees of severity. Thereafter, in the now " salted " troops, infection was never more than sporadic, and there can be little doubt that some degree of immunity was conferred by this primary attack. There would be nothing very unusual about this, were it not for the already mentioned fact, that many different types of dysentery bacilli are found in Middle East, and it would therefore appear that infection with any one of them gives some degree of protection against the others. If this is so, the immunizing agent cannot be the specific O antigen, but must be some more subtle property which is common to them all. It was this line of thought which led me to hope that immunization with a vaccine containing a single type might possibly have some measure of success. AMOEBIC DYSENTERY IN MIDDLE EAST

Finally let me say a few words about the incidence of amoebic dysentery in Middle East during the last war. The Director of Army Health has very kindly supplied me with some figures which show that the annual admission rate for all types of dysentery from 1941 to 1943 ran at a steady level of about 33 per 1,000 of the total strength. Details of the relative proportions of bacillary and amoebic dysentery can be extracted from the monthly reports rendered by laboratories. From August 1940 until June 1943, specimens from just under 65,000 cases of clinical dysentery were examined. Of these, 5.3 per cent. were found to contain trophozoites of E. histolytica. This is much the same as in the 1914-1918 war, when the figure varied from 2 per cent. to 7 per cent. in different years. In contrast to these findings, made by competent pathologists in our Services, one or two instances of the older school of thought were encountered elsewhere. Following the occupation of Lebanon and Syria in 1941, an Australian unit took over a Military Hospital in Syria. Visiting it shortly afterwards, we found graphs prepared by its previous occupants showing the incidence by months and years of different d i s e a s e s - malaria, enteric group of fevers, sandfly fever, etc. One was labelled amoebic dysentery and showed a curve

486

DYSENTERY

rising steadily during the summer months to a peak of considerable height. There was no graph for bacillary d y s e n t e r y - presumably because the number of reported cases was too small. In 1942 and 1943, bacillary and amoebic dysentery occurred in the Allied troops stationed in this area in the same proportions as elsewhere, with the usual preponderance of bacillary infections. I can see no reason for thinking that some radical change in the epidemiology of these two diseases had occurred, and can only conclude that there had been confusion in diagnosis. In 1941, after the capture of Ethiopia and Eritrea, considerable numbers of Eritrean prisoners were brought into Sudan, and the sick from the P.O.W. camps were treated in an expanded section of an Indian General Hospital, staffed mainly by captured Italian medical officers. T h e D.D.M.S. was surprised at the large quantities of emetine which were being requisitioned, and asked me to look into this. I found that the hospital laboratory was under the charge of a keen young pathologist, who at my request demonstrated the microscopical picture on which he made his diagnosis of amoebic dysentery, of which there were said to be many cases. A fresh specimen was procured from one of them and a portion of mucus neatly mounted and placed on the stage. After a brief search he stood back in triumph and waved me to the microscope. The picture was typical of bacillary infection - - large numbers of pus cells, a few red cells, and in the centre a globular non-motile refractile object with a diameter of roughly three times that of a pus c e l l that is, over 30~z - - which he said was a cyst. But, of course, it was neither cyst nor trophozoite, but just one of the large macrophages which are so characteristic a feature of bacillary exudate. Such mistakes are, alas, common and are the outcome of lack of training and lack of experience. THE DIAGNOSIS OF ACTIVE AMOEBIC DYSENTERY

It is interesting to recall that Dobell wrote in 1917 : " Stool examinations made by persons who have not served their apprenticeship to the actual work itself possess no scientific value w h a t s o e v e r - and for the average worker, a practical training of not less than four to six weeks is, even under the most favourable conditions, requisite." His remark referred especially to the examination of faeces for cysts, but in my experience the same applies to the examination of muco-pus for amoebae. The golden rule is that a diagnosis of amoebic dysentery should not be made unless active amoebae containing red blood corpuscles have been demonstrated in the patient's stools. Disregard of this rule will inevitably lead to mis-diagnosis and may create an entirely false conception of the prevalence and importance of amoebic dysentery in a community. CONCLUSION

This brings me to the end of a story which cannot be summarized. The main points which I would like to emphasize are that bacillary dysentery far exceeded amoebic dysentery in importance until the efficacy of sulphonamide treatment was discovered : that, with the possibility of strains emerging which are resistant to sulphonamides (and antibiotics), it remains a lurking danger, for it is only controlled, and not banished : that active amoebic dysentery should be diagnosed only when the golden rule is satisfied : and that we should remember that, as far as this country is concerned, the symptomless cyst-passer, no matter where he acquired his infection, has been proved, by the experience of the aftermath of two wars, to be of no epidemiological importance. Nevertheless, this same cyst-passer,

J. s. K. BOYD

487

and the a m o e b a e with which he is infected, p r e s e n t so m a n y p r o b l e m s that I a m confident he will p r o v i d e material for discussion and a r g u m e n t at future meetings of this Society. REFERENCES ANDm~WES, F. W. & INMAN, A. C. (1919). Spec. Rep. Set. med. Res. Coun., Lon& No. 42. Annual Report of the Public Health Commissioner with the Government of India for 1935. Government. of India Press. New Delhi. ARCHIBALD, R. G., HADFIELD, G., LOGAN, W. & CAMPBELL, W. (1916). J . R . Army reed. Cps., 26, 695. BONEY, T. K., CROSSMAN, L. G. & BOULENGER~ C. L. (1918). Ibid., 30, 409. BOYD, J. S. K. (1940). Trans. R. Soc. trop. Med. Hyg., 33, 553. - & PORTNOY, B. (1944). Ibid., 37, 243. DOBELL, C. (1917). Spec. Rep. Ser. med. Res. Coun., Lond. No. 4. (1919). The Amoebae Living in Man. London : John Bale, Sons & Danielsson. (1928). Parasitology, 20, 357. - (1931). Ibid., 23, 1. FAIRLEY, N. HAMILTON & BOYD, J. S. K. (1943). Trans. R. Soc. trop. Med. Hyg., 36, 253. HARGREAVES,W. H. (1945). Ibid., 38, 244. HOARE, C. A. (1950). Brit. med. J., 2, 238. LEDINGHAM, J. C. G. (1920). J . R . Army med. Cps., 34, 189 & 306. MANIFOLD, J. A. 8c DEMONTE, A. J. (1928). Indian ft. med. Res., 15, 601, Official History of the War. Medical Services. Diseases of the War., Vol. 1, 1922. London : H.M. Stationery Office. Official History of the War. Medical Services. Pathology, 1923. London : H.M. Stationery Office. WENYON, C. M. & O'CONNOR, F. W. (1917). J . R . Army med. Cps., 28, 1. WILLMORE, J. G. & SHEARMAN, C. H. (1918). Lancet, 2, 200. WooDcocK, H. M. (1917). J . R . Army reed. Cps.. 29, 290. -

-

-

-