865

Dhurjaty, I.A.M.C., for much clinical assistance; MajorD’Monte, l.M.s., I.A.M.o., and Dr. V. Korke for laboratoryinvestigations ; and Brigadier H. K. Goadby for his adviceand criticism in the preparation of this article.

REFERENCES

Jones, H. E., Armstrong, T. G., Green, H. F., Chadwick, V.(1944) Lancet, i, 720.

Taylor, G. F., Chhutani, P. N. (1945) Brit. med. J. i, 800.

ATYPICAL PNEUMONIASIGNIFICANCE OF COLD AGGLUTININS

G. E. O. WILLIAMSM.D. Lpool, M.R.C.P.

ASSISTANT PHYSICIAN, BIRMINGHAM UNITED HOSPITAL ;LATE SQUADRON-LEADER R.A.F.V.R.

THE condition at present known as " atypicalpneumonia" or " virus pneumonia" may be a groupof diseases similar in clinical manifestations but ofdifferent aetiology. This view has been stated byDrew et al. (1943) and supported by many others. Thereis little doubt that most, if not all, of these cases areof virus origin (Longcope 1942, Weir and Horsfall1940, Eaton et al. 1941, 1944, Francis and Magill 1938,Commission on Acute Respiratory Diseases 1945).

In an attempt to discover some variations in theclinical picture which might be attributable to differencesin aetiology, the records of 56 patients admitted to aR.A.F. hospital in twenty-one months were examined.Clinical, pathological, and radiological findings were

reviewed, and particular attention was paid to the testsfor cold agglutinins which had been performed in mostinstances.The standards by which atypical pneumonia was

diagnosed have been described elsewhere (Drew et al.1943, Longcope 1940, 1942, Kneeland and Smetana1940, Ramsay and Scadding 1939). The present seriescorresponded closely to those of other workers in theincidence and duration of individual signs and symptoms,in values obtained for sedimentation-rates and leucocytecounts, and in radiological findings.

COLD AGGLUTININS

The primary object of this investigation was to assessthe significance, if any, of the cold-agglutinin titre inthese cases. Out of 51 cases in which this test was

performed, 26 had titres of 1 : 64 or higher. The incidenceof cold agglutinins at various stages of the disease isshown in table i. Maximal titres developed during thesecond and third weeks. This is consistent with thefindings of other workers (Commission on Acute Respira-tory Diseases 1945, Turner et al. 1943, Humphrey 1945).To enable the cold-agglutinin titres to be correlated

with other features of the disease the series was arbitrarilydivided into " cold-agglutinin positives," comprisingthe 26 with titres of 1 : 64 or more, and " cold-agglutininnegatives," which had titres below this figure. Toexclude errors arising from tests taken at irrelevantperiods in the course of the illness, only cases in which

TABLE I-INCIDENCE OF COLD-AGGLUTININ TITRES AT VARIOUSSTAGES OF ATYPICAL PNEUMONIA

cold agglutinins were tested during the second, third,or fourth week were included. The series was nextdivided into groups according to differences’ in certainindividual features, and the incidence of positives similarlycalculated in each group. The significance of thedifferences between the incidence of cold agglutinins inthese subgroups was then assessed by the usual statisticalmethod.

There was no significant correlation between theincidence of cold agglutinins and duration of the disease,duration of pyrexia, duration of radiological abnormalities,or incidence of pleural involvement.

Comparison of the incidence of cold-agglutinin positiveswith the various types of radiographic appearances,however, produced some interesting findings. The radio-grams were not available, but the radiologists’ reportsshowed three main types of radiological abnormalities.These are analysed in table 11 and related to the incidenceof cold-agglutinin positives.The standard error of the difference between the pro-

portion of positives in groups A (28’6%) and C (75%) being14-6, this difference appears to be statistically significant.Another clinical feature in 20 of the present cases of

atypical pneumonia was " migration " of the pneumonicprocess from one lobe to another. Of these migratorycases 18 were detected clinically and 15 presented radio-logical signs. Since this phenomenon was not especiallysought for, some cases -of migration may have beenmissed. The number of cold-agglutinin positives amongthe migratory cases, however, was 15, compared with7 in a group of 20 cases not showing migration. The

TABLE II-ANALYSIS OF RADIOLOGICAL FINDINGS IN ATYPICAL

PNEUMONIA IN RELATION TO THE INCIDENCE OF COLD

AGGLUTININS

difference between these two proportions being greaterthan double the standard error (14’4), this also appearsto be significant. ’

The 15 cases of migratory atypical pneumonia whichgave cold agglutinins are summarised in table III, withthe maximal titre of cold agglutinins and the radiologicalfindings. Of these 15 " migrators," 14 showed mottledlung shadows. The radiological reports of the 15thcase suggested that it might also have been includedin this category, but the information given was notsufficiently conclusive.The close association of cold agglutinins, migration,

and the mottled type of lung shadows suggests that thesecases may have a common origin, and table iv, showingthe parallel seasonal incidence of cold-agglutinin positivesand migration, lends further support to this possibility.

DISCUSSION

Opinions on the significance of cold agglutinins in

atypical pneumonia vary considerably. Peterson et al.

(1943) found them to be rare, and Stats and Wassermann(1943) consider the test to be

" no more than a laboratory

curiosity " apart from a few cases of hæmolytic anaemiaand Raynaud’s syndrome. Turner (1943) found themin 10 out of 22 cases of atypical pneumonia, Turner et al.(1943) in,28% of 132 cases, and Turner (1.945) in only

866

1 out of 39 tests. Viswanathan and Natarajan (1945),however, found titres of over 1 : 256 in 3 out of 5 cases,the Commission on Acute Respiratory Diseases (1944)found cold agglutinins in 31 % of cases, and McNeil (1945)found that the incidence of cold agglutinins differed indifferent outbreaks.

This wide variation can be explained by one of threepossibilities : (1) the test may be entirely irrelevant andthe positives due entirely to chance (the differencebetween their incidence in cases of atypical pneumoniacompared with controls, coupled with the demonstrablerise and fall in titre during the course of the illness,render this sl.-ipposition unlikely) ; (2) the discrepanciesmight have been due to errors or differences of techniqueor to some of the tests having been performed at thewrong stage in the illness ; and (3), as McNeil (1945)suggested, cold agglutinins may be present in significanttitres in only certain outbreaks of the disease. If thethird explanation is correct, the cases with cold agglu-tinins would be expected to have a common anddistinctive aetiology.

TABLE IV—SEASONAL INCIDENCE OF, (a) TOTAL OASES, (b)MIGRATING CASES, AND (c) COLD-AGGLUTININ POSITIVES INTHREE-MONTHLY PERIODS

Atypical pneumonia is almost certainly of virus origin,but there are probably several viruses capable of pro-ducing this clinical picture (Drew et al. 1943, Weir andHorsfall 1940, Francis and Magill 1938, Eaton et al. 1941).It is suggested that patchy radiological shadows in thelungs, migration of the pneumonic process from one lobeto another, and the presence of cold agglutinins in titres of1 : 64 or more may form a triad distinguishing one groupof cases from the remainder. That this distinction may

TABLE III—CLINICAL AND RADIOLOGICAL FINDINGS AND COLD-AGGLUTININ TITRES IN 15 CASES OFMIGRATORY ATYPICAL PNEUMONIA

* Radiological findings classified according to description in table 11.

867

be due to a specific aetiological factor is supported bythe,.seasonal incidence of this type of case and by thecloseness of the association between these features incases known to have a common origin.

SUMMARY

A survey was made of the clinical, pathological, andradiological findings in 51 cases of atypical pneumoniaadmitted to a R.A.F. hospital between September, 1943,and June, 1945.

In their general characteristics these cases closelyresembled each other and those described by otherworkers. Correlation of cold-agglutinin titres withother features of the disease, however, suggested thatsome of the cases, presenting the triad of patchy con-solidation, migration of the pneumonic process, and coldagglutinins, might have a common and distinct etiology.My thanks are due to the Director-General of Medical

Services, R.A.F., for permission to publish this paper ; to

Group-Captain W. A. S. Duck, O.B.E., for permission tocollect the necessary material ; and to Wing-CommandersF. H. Coleman, A. Nelson-Jones, and D. G. Ferriman for theircooperation in the investigation.

REFERENCES

Commission on Acute Respiratory Diseases (1944) Amer. J. med. Sci.208, 742.

— (1945) J. Amer. med. Ass. 127, 146.Drew, W. R. M., Samuel, E., Ball, M. (1943) Lancet, i, 761.Eaton, M. D., Beck, M. D., Pearson, H. E. (1941) J. exp. Med. 73,

641.

— Meiklejohn, G., van Herick, W. (1944) Ibid, 79, 649.Francis, T., Magill, T. P. (1938) Ibid, 68, 147.Humphrey, A. A. (1945) Nav. med. Bull., Wash. 43, 1117.Kneeland, Y. jun., Smetana, H. F. (1940) Bull. Johns Hopk. Hosp.

67, 229.Longcope, W. T. (1940) Ibid, p. 268.

— (1942) Practitioner, 148, 1.McNeil, C. (1945) Amer. J. med. Sci. 209, 48.Peterson, O. L., Ham, T. H., Finland, M. (1943) Science, 97, 167.Ramsay, H., Scadding, J. G. (1939) Quart. J. Med. 8, 79.Stats, D., Wassermann, L. R. (1943) Medicine, Baltimore, 22, 363.Turner, J. C. (1943) Nature, Lond. 151, 419.

— Nisnewitz, S., Jackson, E. B., Berney, R. (1943) Lancet, i, 765.Turner, R. W. D. (1945) Ibid, i, 493.Viswanathan, R., Natarajan, B. (1945) Ibid, p. 148.Weir, J. M., Horsfall, F. L. jun. (1940) J. exp. Med. 72, 595.

INTRADERMAL REACTION IN

TRICHOMONAS INFECTION

S. ADLERO.B.E., M.B. Leeds, M.R.C.P.

PROFESSOR OF PARASITO-

LOGY

A. SADOWSKYM.D. Odessa

LECTURER IN OBSTETRICS

AND GYNÆCOLOGY

THE HEBREW UNIVERSITY, JERUSALEM

IT is not difficult to prepare bacteria-free cultures ofTrichomonas vaginalis by exploiting the bacteriostaticaction of sulphonamides and penicillin.! With antigensprepared from such cultures we have investigated thequestion whether vaginal infection with T. vaginalisproduces any detectable immunological reaction in thehost.

Preparation of Antigen.—Rich cultures of liver slant over-laid with Trussel Johnson’s fluid medium adjusted to pH 6 areused. Inactivated calf serum 10% is used instead of humanserum..

The cultures are washed by centrifuging three times insaline, and the number of flagellates per c.cm. in a suspensionof saline is determined by counting in a hæmocytometer.

The suspension of flagellates is again centrifuged, anddistilled water is substituted for the saline in quantities-sufficient to produce 2,000,000 flagellates per c.cm. The

suspension is next allowed to stand overnight in the ice chestand is then centrifuged.The supernatant fluid is next added to an equal amount

of 1% solution of phenol in distilled water, so that the final

1. Adler, S., Pulvertaft, R. J. V. Ann. trop. Med. Parasit. 1944, 38,188.

product is an aqueous extract of 1,000,000 flagellates perc.cm. and 0-5% phenol. This is used for the intradermal tests.

Method of Testing.-The tests were carried out on

outpatients. For each intradermal test (on the forearm)0.1 c.cm. of the antigen is used, and 0-5% solution ofphenol is used as a control. The patients were observedfor half an hour after the injection, and in no case wasany difference noted then between the sites inoculatedwith antigen and the control sites. They were

re-examined after 24 hours and after 48 hours, and someof them subsequently.The site inoculated with the antigen was compared

with the control site, and a reaction was considered posi-tive if there was an area of very definite redness roundthe site of inoculation and the control was completelynegative.

Results.-The area of redness varied from 1 to 2 cm.,and in three refractory cases was 5 cm. in diameter. Thereaction in most cases began to fade after 48 hours, butin nine cases the reaction was still positive after 48 hours,and in one case after 72 hours. In one case a secondinoculation with antigen caused a severe reaction ; the

patient complained of malaise and was confined to bedfor 2 days.

In 59 cases in which T. vaginalis was not found onmicroscopical examination, 45 gave a negative, 7 a

positive, and 7 an indefinite skin reaction. Of the 7 which

gave a positive reaction one was reinvestigated and foundpositive on microscopical examination, and another wasagain found negative. The remaining 5 patients did notreturn for re-examination. Of this series (excluding thecase which was subsequently found positive on micro-scopical examination) 77% therefore gave a negativereaction.

In 43 cases proved positive by microscopical examina-tion, 35 gave a positive, 2 a negative, and 6 an indefiniteskin reaction. Thus 81% in which flagellates were foundon microscopical examination gave a positive skin test.

In 23 cases, 12 negative and 11 positive on micro-scopical examination, the reactions produced by an

aqueous extract of 2,000,000 flagellates per c.cm. werecompared with those produced by the extract of 1,000,000flagellates per c.cm. As a result of these tests 3 cases(2 negative and 1 positive on microscopical examination)which had not reacted to the extract of 1,000,000flagellates per c.cm. gave a positive reaction to the extractof 2,000,000 flagellates, per c.cm. ; 2 cases (1 positiveand 1 negative on microscopical examination) which hadgiven an indefinite reaction to the extract of 1,000,000flagellates per c.cm. gave’a positive reaction to the extractof 2,000,000 flagellates per c.cm. ; and 2 cases negativeon microscopical examination did not react to eitherextract.

In one case positive to both extracts the erythemadisappeared after 48 hours from the site inoculated withthe extract of 2,000,000 flagellates per c.cm. but persistedin the site inoculated with the extract of 1,000,000flagellates per c.cm.

In 15 of the cases (8 positive and 7 negative on micro-scopical examination) there was no difference in theresults obtained with the two extracts. In 7 other cases

negative on microscopical examination antigen was

inoculated simultaneously into two sites with identical(negative) results.

It seems that there is no important difference in theresults obtained with aqueous extracts of 1,000,000and of 2,000,000 flagellates per c.cm.

Discussion.—Obviously the use of the skin test for

diagnostic purposes is not to be recommended sincedirect microscopical examination gives better results.In only one case did we detect by the skin reaction aninfection which was not revealed by the first microscopicalexamination. But it is of interest to note that in about