Studies on Primary Atypical Pneumonia. I. Clinical Features and Results of Laboratory Investigations

STUDIES ON PRIMARY ATYPICAL PNEUMONIA. I. CLINICAL FEATURES AND RESULTS OF LABORATORY INVESTIGATIONS

Edward C. Curnen, … , Lewis Thomas, Frank L. Horsfall Jr.

J Clin Invest. 1945;24(2):209-226. https://doi.org/10.1172/JCI101598.

Research Article

Find the latest version: https://jci.me/101598/pdf STUDIES ON PRIMARY ATYPICAL PNEUMONIA. I. CLINICAL FEATURES AND RESULTS OF LABORATORY INVESTIGATIONS1 2 By EDWARD C. CURNEN,3 GEORGE S. MIRICK,3 JAMES E. ZIEGLER, JR.4 LEWIS THOMAS,' AND FRANK L. HORSFALL, JR.' (From the United States Navy Research Unit at the Hospita of The RockefeUer Institute for Medical Research, New York City) (Received for publication August 14, 1944) "Primary atypical pneumonia, etiology un- forces, however, it has appeared to be more pre- known" (1) is one of many terms used to desig- valent than the pneumococcal pneumonias and nate an acute respiratory infection which has the total number of man days lost as a result has been widely prevalent in recent years. The ill- been considerable. ness is characteristically gradual in onset, with The term "primary atypical pneumonia" has constitutional as well as respiratory symptoms, been useful in distinguishing the syndrome under and pulmonary changes more manifest in roent- discussion from clinically similar illnesses due to genograms than by physical examination. The known infectious agents (10) such as Rickettsia course of illness varies considerably in duration diaporica (11) and members of the psittacosis and severity. Complications are uncommon and group of viruses (12, 13). although convalescence is frequently protracted Although primary atypical pneumonia has the illness almost invariably terminates with been frequently referred to as "virus pneumonia" complete recovery. Therapy with sulfonamide and most investigators have considered it to be drugs is not effective. of non-bacterial origin, the causative agent, or Contemporary interest in this form of pneu- agents, in the majority of cases has remained monia was aroused by several fairly recent re- obscure. The recovery of a number of different ports (2 to 4), although one author (5) had ob- viruses (14 to 18) from certain patients with served patients with a similar illness several primary atypical pneumonia has been reported years earlier. Numerous other descriptions and but the importance of any of these agents in the investigations of primary atypical pneumonia pathogenesis of the disease has not been es- have since been published, and have been re- tablished. viewed recently (6 to 9). In some of these re- An investigation of primary atypical pneu- ports, unusual outbreaks of acute respiratory monia has been carried out continuously at the illnesses described during the past century were Hospital of The Rockefeller Institute since cited to show that primary atypical pneumonia is February, 1942. Because of circumstances which probably not a new disease. Nevertheless, it is afforded extensive clinical as well as laboratory apparent from these reports that, in recent years, facilities, it was possible while investigating the the incidence of the disease has actually in- pathogenesis of primary atypical pneumonia to creased. The rate of its occurrence in the civil- conduct also an appraisal of its clinical aspects. ian population is not known. In the armed An effort was made to correlate these two lines of investigation and to study each patient in as I The Bureau of Medicine and Surgery does not necessarily undertake to endorse the views or opinions which comprehensive a manner as possible are expressed in this paper. Recently, a non-hemolytic streptococcus, iso- ' The work described in this paper was done under a lated from the lungs of 2 patients who died of contract, recommended by the Committee on Medical primary atypical pneumonia, was found to react Research, between the Office of Scientific Research and with sera obtained, during convalescence, from Development and The Rockefeller Institute for Medical not react Research. other patients with this illness but did 3Lieutenant Commander (MC), U.S.N.R. with sera obtained during the acute stage of the 4Lieutenant (MC), U.S.N.R. disease (19). For convenience in reference, this 209 210 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR. microorganism has been designated streptococ- laboratory evidence that the pneumonia was not cus MG. Because of the possibility that strep- due to pneumococcus or other microorganisms tococcus MG might have more than an inci- of recognized pathogenicity. dental association with primary atypical pneu- Ninety-five of the patients were male and 11 monia, the properties of this microorganism and were female. All patients were young adults. its relation to the disease were investigated. These patients came from widely separated The bacteriological and serological characteris- areas and had been exposed to multiple and tics of streptococcus MG are described in diverse sources of infection. Nevertheless, as separate communications (20, 21). The special a group, they exhibited a remarkable degree of techniques for isolating streptococcus MG from clinical homogeneity. the respiratory tract of human beings and an extension of the original observations concerning CLINICAL FEATURES the relation of this microorganism to primary The clinical features of primary atypical atypical pneumonia are described in the accom- pneumonia, as observed in these 106 patients, panying paper (22). were on the whole similar to those which have It is the purpose of the present paper to de- been described in other reports. The onset of scribe the clinical features of primary atypical illness was gradual in 73 per cent of the patients pneumonia as observed in patients studied in the and, in contrast to lobar pneumonia, was usually Hospital of The Rockefeller Institute, to present ill-defined. Constitutional symptoms generally the results of various investigations pertaining preceded those referable to the respiratory tract. to the disease, and to correlate some of these In many instances, the initial complaint, vari- findings with the studies concerning streptococ- ously described by the patient, suggested a cus MG which have been referred to above (20 combination of lassitude, weakness, and fatigue. to 22). From August 1, 1942, to March 2, 1944, 180 TABLE I patients were admitted to this hospital with Symptoms in 106 patients with primary atypical pneumonia acute respiratory infections suspected to be Patients Symptoms Number Per cent primary atypical pneumonia. In each case, a General Headache 69 65 careful and detailed record was made of the Malaise 65 61 Chills or chilliness 63 59 clinical features. Roentgenograms of the chest Generalized aches 30 28 were made at frequent intervals and numerous Anorexia 37 35 Nausea 30 28 clinical laboratory studies were carried out. In Vomiting 25 24 addition, samples of serum were obtained re- Respiratory Cough 104 98 peatedly throughout the illness and were stored Sputum 87 82 at 40 C. for use in serological tests. Specimens Bloody sputum 27 25 Sore throat 32 30 of plasma, sputum, and nasopharyngeal wash- Chest pain ings, as well as specimens of pleural or cerebro- Substernal 25 24 Lateral 16 15 spinal fluid, were examined immediately after Epistaxis 16 15 they were obtained and portions of each speci- Dyspnoea 5 5 men were preserved at - 700 C. for use in subsequent investigations. In Table I, the commoner symptoms have been From the completed records of these 180 pa- listed, together with the frequency of their oc- tients, 106 were selected on the basis of clinical currence. Chills, fever, headache, and cough and laboratory criteria as representative ex- developed early and occurred frequently. Shak- amples of primary atypical pneumonia. The ing chills were rare but sensations of chilliness, criteria used for selection permitted the inclusion sometimes recurring, were experienced by 59 of only those patients who showed each of the per cent of the patients. Headache, usually following findings: (1) acute febrile respiratory generalized but sometimes predominantly fron- disease; (2) pulmonary consolidation demon- tal, occasionally associated with photophobia, strable by x-ray; and (3) adequate clinical and and ranging from mild to extremely severe, oc- ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 211 curred in 65 per cent of the patients. Aching TABLE II of the trunk and extremities was present in only Abnormal physical signs in 106 pients with one-fourth of the patients and as a rule was mild. primary atypic pneumonsa Patients About a third of the patients exhibited an ap- Physical signs NiNumber Per cent preciable degree of anorexia and nausea. Vomit- Fever 101 95 ing occurred at some stage of illness in one- Nasal congestion 60 57 Pharyngitis 73 69 fourth of the patients but this symptom was Cervical adenopathy 25 24 never a conspicuous or severe manifestation of Pulmonary signs 104 99 Dulness 57 54 the disease. Altered breath sounds 74 70 Cough, one of the most constant and charac- Rftles 99 93 Friction rub 7 7 teristic features of primary atypical pneumonia, Fluid 1 1 was present in all but 2 of the 106 patients. In Bradycardia 72 68 Tachycardia 8 8 some instances, this was the first symptom to Tachypnoea 14 13 appear; in others, it developed gradually after a Cyanosis 12 11 Signs referable to the central nervous 6 6 lapse of several days. Frequently, it becamne system paroxysmal and exhausting. Coughing, al- Palpable spleen 1 1 though dry at first, was ultimately productive of sputum in 82 per cent of the patients. The cant tachycardia was infrequent and, when sputum of one-fourth of the patients contained present, usually developed during convalescence. blood, usually, however, only as streaks or Nasal congestion and mild pharyngitis were traces. Grossly bloody or "rusty" sputum was noted respectively in 57 and 69 per cent of the uncommon. Sore throat when present was patients. Cervical lymph nodes were enlarged rarely severe, and, in many instances, appeared in one-fourth of the patients, but usually were not to result from the irritation of coughing. prominent or tender. In association with the cough, about one- Physical signs of pulmonary disease, although fourth of the patients experienced dull, burning, much less definite than in lobar pneumonia, were or oppressive substernal discomfort. Less fre- nevertheless detected at some stage of illness in quently, sharper pains developed which were all but 2 of the 106 patients. Dulness to per- referred to the thorax laterally. cussion, usually of slight degree, was noted over Epistaxis occurred in 16 patients and syn- the affected area in more than half of the pa- cope in 3. Several patients complained of tients. Alterations in the breath sounds, usually transient numbness and tingling of the extremi- harshness or diminished transmission, occurred ties. Abdominal pain was only an occasional in 70 per cent of the patients. On the other and then unimpressive complaint. hand, breath sounds of tubular quality and pec- The results of physical examinations of these toriloquy were rarely heard, Rales, usually fine patients are summarized in Table II. On ad- and subcrepitant during the early stages of ill- mission, most of the patients did not appear to ness, later coarse and more moist, were detected be very ill. An elevation of temperature was in 93 per cent of the patients. Often rales noted during hospitalization in all but 5 in- could be heard only after coughing or at the end stances. Cyanosis was rarely present, and of deep inspiration. A pleural friction rub was herpetic lesions were observed in only 3 in- heard in 7 instances and, in 1 patient, signs of stances. Respirations were usually normal in pleural effusion were detected. A few patients rate and seldom labored. Tachypnea was seen with severe headache had slight stiffness of the in only 14 of the more severely ill patients. Re- neck. One patient briefly lost the power of lative bradycardia, however, observed in over ocular accommodation. The spleen was pal- two-thirds of the patients, was a frequent finding pable in only 1 instance. Abdominal distension, during the early febrile phase of illness, and a so common in lobar pneumonia, was conspicu- feature of considerable diagnostic assistance ously absent. No other relevant abnormalities in distinguishing clinically between this syn- were noted on physical examination. drome and pneumococcal pneumonia. Signifi- The clinical course of illness in the present 212 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR. group of patients was extremely variable. All before a response could reasonably be expected. recovered, but it was not possible to predict These findings are in agreement with previous accurately either the- duration or the severity of observations that sulfonamides do not obviously illness in any given case. In the total group of benefit patients with this disease. patients, elevations of temperature to 1000 F. or more were present for an average of 10.1 days, X-RAY FINDINGS ranging in individual patients from 0 to 41 days. X-ray films of the lungs, taken at frequent Physical signs of pneumonia were present for an intervals throughout the course of illness, showed average of 12.8 days, but also ranged from 0 to 41 evidence of more extensive pulmonary lesions days. The maximum recorded temperature than the findings on physical examination- would averaged 103.50 F. with a range of 99 to 1060 F. have led one to expect. Data concerning the The charted temperature curves of individual site, extent, and duration of pulmonary con- patients also showed great variation. The com- solidation as determined by multiple serial monest type of curve was moderatelyremittent, roentgenograms on each patient are presented and rarely was either high persistent fever or in Table III. In this study, the word "con- markedly intermittent fever seen. Deferves- TABLE III cence was usually by lysis over a period of more Distribution, extent, and duration of pulmonary consolidation than one day. Hospitalization was required for as determined by serial x-ray fdms in 106 an average of 33 days with a range of 10 to 73 patients with primary atypical pneumonia Patients days. Number Per cent Complications were rare and usually not of Lobe involved Left lower 66 (18)* 62.2 (16.9) great significance. Pleuritis occurred in 6 pa- Left upper 19 ( 2) 17.9 ( 1.9) tients. In only 1, however, was the accumula- Right lower 64 (20) 60.3 (18.9) Right middle 22 ( 3) 20.7 ( 2.8) tion of fluid sufficient to warrant thoracentesis. Right upper 25 ( 7) 23.5 ( 6.6) In the others, the amount of fluid was evidently Number of lobes involved 1 50 47.1 small and, in several, it was located principally 2 38 35.8 in the interior fissures. Acute sinusitis de- 3 8 7.5 4 4 3.8 veloped in 8 patients and non-suppurative otitis 5 6 5.7 media occurred twice. One patient had severe Duration of consolidation 1 week 9 8.4 stomatitis and gingivitis. Another patient, dur- 2 weeks 51 48.1 ing convalescence from primary atypical pneu- 3 weeks 25 23.4 4 weeks 11 10.3 monia, contracted lobar pneumonia due to 5 weeks 8 7.5 pneumococcus Type XIV, evidently by contact, 6 weeks 2 1.9 and in both infections, the left lower lobe was 7 weeks 0 * Numbers in parentheses indicate patients in whom the the site of consolidation. designated lobe was the only site of pulmonary consolida- Although no systematic investigation of the tion. therapeutic value of sulfonamides in primary atypical pneumonia was carried out, 32 patients solidation" has been used in a broad sense to in this study received sulfathiazole or sulfa- designate findings thought to represent pneu- diazine during the course of illness. Of these, monic involvement of the lung parenchyma. 16 received an adequate therapeutic trial over a In referring to x-ray films of the lung, this term period of 4 to 9 days, with total dosages of 21 to is used irrespective of the extent or density of the 50 grams. Ten patients showed no evidence of pneumonic shadow. a therapeutic response. In 6 patients, the tem- As shown in Table III, the site of pulmonary perature returned to normal within 72 hours of consolidation was most frequently in the lower the initiation of treatment, but in all 6, deferves- lobes and the incidence of involvement was al- cence occurred between the sixth and tenth days most identical on the 2 sides. Consolidation of illness. Moreover, in 2 of these patients, the was present in the left and right lower lobes in fall in temperature occurred within 24 hours, 62.2 and 60.3 per cent of patients, respectively. before an adequate dose had been given and Each of these lobes was the only site of consolida- ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 213 tion, however, in 16.9 and 18.9 per cent of ill-defined, and not always confined by interlobar patients, respectively. Consolidation was pres- boundaries. Sometimes the shadow was present ent in the left and right upper lobes and the only as a diffuse haze; in other instances, it was right middle lobe in 17.9, 23.5, and 20.7 per cent quite dense. Only rarely, however, was the of patients, respectively. On the other hand, density or extent of consolidation in a given each of these lobes was the only site of consolida- lobe comparable to that seen in lobar pneumonia. tion in 1.9, 6.6, and 2.8 per cent of patients, In other instances, consolidation was mottled respectively. and sometimes the appearance of small abscess Pulmonary consolidation as observed in roent- cavities was suggested. Infiltrations radiating genograms was confined to a single lobe in 50 from the hilum and involvement of the upper patients, located in one of the lower lobes in 38 lobe resembled tuberculosis- even more fre- patients, in one of the upper lobes in 9 patients, quently, and other lesions, especially in the and in the-middle lobe in 3 patients. Pulmon- lower lobes, simulated bronchiectasis. The rela- ary consolidation occurred in more than one tive rapidity with which the lesions cleared in lobe in 56 patients. Two lobes were involved primary atypical pneumonia assisted in exclud- in 38 patients, 3 lobes in 8 patients, 4 lobes in 4 ing a diagnosis of chronic pulmonary disease. patients, and all 5 lobes in 6 patients. The Other forms of pneumonia could not be dis- average number of lobes involved for all pa- tinguished from primary atypical pneumonia in tients was 1.85. a given case on the basis of roentgenograms In some instances, multiple lobes were in- alone. volved simultaneously, in others consolidation Resolution of the consolidated areas was ob- extended from one lobe to another or new shad- served to occur in several ways. In most in- ows appeared in one or more lobes during the stances, clearing proceeded from the periphery course of illness. Occasionally, fresh areas of of the lung toward the hilum but occasionally consolidation appeared while other areas were resolution progressed in the reverse direction. undergoing resolution, and consolidation was Clearing was also observed to occur by progres- observed to recur in the same site or to reappear sive reduction in the size of the affected area in another area after the initial lesion had cleared. and by a diminution of density until no evidence Thickening of the interlobar septa, occasional of consolidation remained. In many instances, small collections of fluid in the interlobar fis- during the process of resolution, the affected area sures or in the costophrenic angles, and, in assumed a diffusely mottled appearance. Fre- films of one patient, definite pleural effusion were quently, after all parenchymal involvement was also observed. thought to have disappeared, the bronchovascu- The individual pulmonary lesions as viewed in lar markings remained prominent. roentgenograms showed extraordinary variation In this group of patients, the duration of con- in density as well as in distribution. Usually, solidation in roentgenograms was arbitrarily an area of pneumonic consolidation was present calculated as the interval from the onset of illness in the initial film of the chest taken at the time to the last day on which evidence of parenchymal of admission, even as early as the first or second involvement was still visible. As recorded in day of illness. In a few patients, evidence of Table III, pulmonary consolidation had dis- pneumonia was not present on admission but appeared by the end of the second week of illness appeared subsequently. In some of these pain more than half the patients and by the end tients, consolidation first appeared as an iso- of the third week in 80 per cent of the total lated area well out in the lung parenchyma with- number. The average duration of consolidation out obvious connection with the hilar shadow. as determined by serial roentgenograms was 14.2 In most instances, however, the initial consolida- days with a range of 3 to 41 days. tion made its appearance and was most dense at Although in the present study, roentgeno- the hilum, spreading and becoming less dense logical evidence of pulmonary consolidation was toward the periphery of the lung. The borders established as one of the essential criteria for a of the consolidated areas were usually irregular, diagnosis of primary atypical pneumonia, it 2414 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR. should be emphasized that the x-ray shadows than 15,000 occurred during the first 3 weeks of observed were not distinctive of this disease illness but were uncommon. The highest count syndrome exclusively and a diagnosis of primary noted was 25,800 and occurred in the second atypical pneumonia could not be made on the week of illness. Sudden and unexplained shifts basis of roentgenological evidence except in in the total counts sometimes occurred. Usu- conjunction with clinical and laboratory findings. ally, however, in individual patients, neither high nor low counts were sustained on repeated exam- CLINICAL LABORATORY FINDINGS inations. The differential leukocyte pattern as The total leukocyte count on admission and at observed in these patients was within normal various intervals thereafter was usually normal limits. or only slightly elevated. This was another The erythrocyte sedimentation rate was al- helpful feature in distinguishing primary atypical most always increased during the acute phase of pneumonia from pneumonia due to pneumococci illness, and the accelerated rate usually persisted or other bacterial agents in which leukocytosis for some weeks during convalescence. A 5.0 ml. characteristically occurs. Data concerning the specimen of blood was collected in a tube con- total leukocyte counts in the present group of taining dried potassium and ammonium oxylate patients are summarized in Table IV. and, within an hour, was transferred to a The total leukocyte counts recorded upon the calibrated 100 X 3 mm. tube for the test. 106 patients with primary atypical pneumonia Readings were made at the end of 1 hour without during each week of illness were grouped in 4 correction for hematocrit, inasmuch as significant categories on the basis of their magnitude. anemia was rarely present. The maximum rate These included total counts of less than 5000, was less than 10 mm. per hour in 4 per cent of those of 5000 to 10,000 considered to represent patients, 10 to 19 in 2 per cent, 20 to 29 in 13 the normal range, those from 10,000 to 15,000, per cent, 30 to 39 in 22 per cent, 40 to 49 in 49 and those of more than 15,000. per cent, 50 to 59 in 9 per cent, and more than It is apparent that throughout the course of 60 in 1 per cent. A sedimentation rate of more illness as measured in weeks, considerably more than 15 mm. in 1 hour persisted on the average than half the total leukocyte counts were within for 28.5 days with a range of 9 to 73 days. In a the normal range, and two-thirds or more of the few instances, the sedimentation rate remained counts were less than 10,000. Leukopenia was or became normal while the patient was still relatively infrequent and never extreme. The acutely ill. This observation was amply con- lowest recorded count of 3300 was observed in firmed but not satisfactorily explained. the second week of illness. Total leukocyte The urine was usually normal. Slight or counts above normal were in most instances moderate albuminuria was noted during the between 10,000 and 15,000. Counts of more acute phase of illness in 22 per cent of the pa-

LE IV Total leukocyte counts in 106 patients with primary atypical pneumonia

Percentage of leukocyte counts* per week of illness

Total leukocyte count Week of illness

1 2 3 4 5 6 7 8 9 10 .<5,000 5.9 8.0 1.9 9.7 5.9 13.3 5,000 to 10,000 66.7 57.3 66.0 74.2 88.2 80.0 77.0 50.0 100.0 100.0 10,000 to 15,000 26.2 29.3 22.6 16.1 5.9 6.7 23.0 50.0 > 15,000 1.2 5.3 9.4 Number of patients 84 75 53 31 17 15 13 4 1 2

* Only one count per patient per week was included in this analysis. ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 215 tients. Examination of the urinary sediment BACTERIOLOGICAL FINDINGS revealed granular or hyaline casts and transient occult hematuria in 10 and 8 per cent of the pa- The bacteriological flora of the nasopharynx tients, respectively. Three of the 9 patients with and sputum was studied in each patient and an microscopic hematuria had previously received effort was made to obtain both qualitative and sulfonamide therapy. quantitative data. A culture of the blood was Lumbar puncture was done only when severe obtained from every patient at the time of headache and meningismus were present. Speci- admission and at intervals thereafter when mens of cerebrospinal fluid from 6 patients were indicated. examined. Two contained lymphocytes (less The cultures of the nose and throat in these than 50 per c. mm.) and gave weakly positive patients yielded a variety of microorganisms. tests for protein. The other 4 fluids were In cultures from occasional patients, one or normal. another bacterial species appeared to pre- dominate; but in the group as a whole, the bac- ELECTROCARDIOGRAMS terial flora appeared to be quantitatively and Serial electrocardiograms 6 were taken on 50 qualitatively similar to that commonly found in of the patients with primary atypical pneumonia, the upper respiratory passages of normal persons. none of whom gave a history of previous cardio- The sputum of these patients also contained a vascular disease. No significant deviations from variety of microorganisms, but, by ordinary normal were detected during any stage of the methods of culture, no one species was found to illness from the third through the thirtieth day be conspicuously or consistently present in after onset. Specifically, there was no increase specimens from the total group. in the P-R interval, no elevation or depression When pneumococci, streptococci, or H. in- of the S-T segment, and no T-wave change. fluenzaewere isolated from anysource, an attempt was made to obtain specific identification by BIOCHEMICAL STUDIES serological methods. The results are summarized in Table V. Biochemical assays6 were carried out upon Pneumococci were isolated from 54 per cent specimens of blood and urine from a number of the patients but in only 7 per cent was it was found o-f representative patients. It that the possible to identify the organisms in the sputum icteric indices, prothrombin, serum C02, serum directly by the Neufeld quellung technique. chloride, and plasma protein levels were within In most instances, pneumococci were isolated normal limits. Plasma carotene and vitamin A and identified following the inoculation of levels were reduced but only to a degree com- sputum into mice. Twenty-six different types acute infectious monly encountered in other of pneumococci were obtained from 57 of the 106 diseases. As previously reported (23), chloride patients. It is noteworthy that pneumococci of there was no balance studies indicated that types I and II, those most commonly associated striking disturbance of chloride metabolism, with lobar pneumonia, were not found in any and plasma a-amino acid levels were found to of the patients in this group. The distribution be within the normal range. These latter two this of pneumococcal types, as well as the number of findings serve additionally to distinguish pneumococci present in specimens from indi- syndrome from pneumococcal pneumonia, in disturbed vidual patients, was similar to the distribution of which chloride metabolism is usually these organisms in the upper respiratory passages and the plasma a-amino acid levels are sig- of normal persons. nificantly reduced (24). Beta hemolytic streptococci were isolated from ' Lt. C. G. Neumann (MC), U.S.N.R., kindly analyzed 10 patients, each of whom gave a history of the electrocardiograms. recent contact with cases of streptococcal infec- ' Lt. Cdr. K. Emerson, Jr. (MC), U.S.N.R., Dr. C. L. tion. These to at least Hoagland, Lt. Cdr. R. A. Phillips (MC), U.S.N.R., and streptococci belonged Lt. Cdr. R. E. Shank (MC), U.S.N.R., kindly performed 9 distinct serological varieties. Five different these analyses. types and 2 strains of as yet unclassified types 216 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR.

TABLE V Bacteria isolated from 106 patients with primary atypical pneumonia

Pneumococcus Beta hemolytic streptococcus H. inhtuenzae Streptococcus

Number of of Number of Number of TypeTypepatients* GroupuGrupType mNuberpatients* Type patients patients 6 7 A 14, 15, 17 1 Could not 34 45 3 6 19, C-94 be identified 29 5 ?2 9,11,16,18,20,23 3 4, 7, 8,19,28,32,37 2 5, 10, 12, 17, 24, 25, 33, 34, 35, 36 1 B, C, H 1 Total 26 57 10 34 45 Percentage of total patients 53.7 9.4 32.0 42.0 * Number of patients from whom each designated type was isolated. of group A streptococci were represented. In no instance was evidence obtained which Strains of streptococci belonging to groups B, C, indicated that any one of the bacterial species and H were each isolated from individual isolated from these patients by routine methods patients. was causally related to the disease. Strains of H. influenzae were present in cul- A non-hemolytic streptococcus, designated as tures from 34 of the 106 patients but in no in- streptococcus MG, was isolated from sputum or stance could the organisms isolated be identified nasopharyngeal cultures of 45 of the 106 patients. by capsular swelling in immune rabbit sera In some instances, this streptococcus was cul- against any of the established types. tured directly from the throat or from fresh B. friedldnderi, type A, was isolated from the sputum. It was also cultured, however, from nasal passages of 1 patient, and N. catarrhalis in source materials which had been stored at considerable numbers from 6 patients. - 700 C. for periods varying from a few weeks Staphylococci were frequently noted in cul- to more than 1 year. The special techniques for tures of the respiratory passages. They showed isolating streptococcus MG and the results of varying pigmentation and capacity to produce studies on this microorganism in primary atyp- hemolysis on blood agar but only rarely did they ical pneumonia are described in the accompany- appear to be the predominant organism in ing paper (22). cultures from the patients in this study. Colonies of organisms having the character- VIRUS STUDIES istics of alpha or gamma streptococci were fre- A continuous effort was made throughout this quently seen on blood agar plates inoculated study to recover from specimens obtained from with sputum or material from the nasopharynx. the patients in this series, as well as from speci- Appreciable numbers of organisms in these gen- mens received from similar patients elsewhere, eral categories were seen on blood agar cultures infectious agents other than bacteria which might from 33 of the 106 patients. None of the many be etiologically related to primary atypical strains of alpha and gamma streptococci iso- pneumonia. A large number of specimens ob- lated by this method and tested by appropriate tained from patients with this illness were in- serological techniques could be identified as oculated in various animal species by several streptococcus MG. routes and, in many instances, serial passages Cultures of the blood from all patients showed were carried out. The specimens inoculated in- no bacterial growth. Cultures of each available cluded sputa, nasopharyngeal washings, plasma, specimen of pleural and cerebrospinal fluid re- pleural, and cerebrospinal fluids, and suspensions mained sterile. of consolidated human lung from 9 fatal cases. ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 217 The animals used included mice, cotton rats, served (27) that cotton rats and rabbits inocu- hooded rats, white rats, and hamsters which lated with certain specimens from patients or were inoculated by the intranasal, intracerebral, with passage materials derived from them de- intraperitoneal, or subcutaneous route; guinea- veloped antibodies capable of neutralizing a pigs which were inoculated intraperitoneally and heterologous virus, P.V.M., whereas animals subcutaneously; rabbits which were inoculated inoculated with control materials did not. This intravenously, intraperitoneally, and subcutane- evidence was at that time thougjit to indicate the ously; as well as monkeys, cotton rats, and ham- presence, in these specimens from patients with sters, which were inoculated intratracheally and primary atypical pneumonia, of a virus which intranasally. Chick embryos at various stages possessed minor antigenic components in com- of maturation were also inoculated by several mon with P.V.M. techniques, including inoculation into the allan- Subsequent studies have necessitated reinter- toic, amniotic, or yolk sacs, as well as directly pretation of the original data. It has been found upon the chorioallantoic membrane. Chick- that in cotton rats and rabbits, as well as in embryo passage materials were inoculated in hamsters, the formation of neutralizing anti- each of the animal species mentioned above as bodies against P.V.M. sometimes could be well as in chimpanzees. stimulated by the inoculation of non-infectious With sputa obtained from certain patients, materials. Thus it appears that each of these pulmonary consolidation was repeatedly in- animal species at times and under circumstances duced following primary intranasal inoculation not as yet predictable may harbor latent agents in cotton rats. Chick-embryo passage materials which in response to non-specific stimulation were also found to be capable of inducing pul- induce the formation of specific antibodies monary consolidation in either cotton rats or against P.V.M. Under these circumstances it is hamsters. Similar consolidations, however, were apparent that an antibody response against found to occur in both species following the P.V.M. in any of these species cannot be at- intranasal inoculation of various control ma- tributed to specific stimulation by a virus in terials. The pulmonary lesions which were noted material of human origin. following inoculation with potentially infective specimens could not be distinguished from those SEROLOGICAL STUDIES which followed inoculation with control ma- Sera obtained from patients with primary terials, and none of the lesions could be repro- atypical pneumonia during the acute and con- duced regularly on serial passage in either valescent phases of the illness were tested for the species. No evidence was obtained to indicate presence of antibodies against a number of the presence, in any of the human source ma- viruses known or thought to be capable of terials studied, of a non-bacterial infectious inducing acute respiratory disease in human agent capable of producing signs of infection beings. Tests of this nature were- carried out which could be reproduced in series. upon sera, not only from the 106 patients ad- Infectious agents, apparently harbored by mitted to the Rockefeller Hospital and described animals used in these experiments, were en- in this study, but also from a number of patients countered repeatedly and seriously complicated with primary atypical pneumonia treated else- attempts to isolate an infectious agent from the where. human materials. Lymphocytic choriomenin- The agents used in these tests were influenza A gitis virus, the pneumonia virus of mice (25), virus, influenza B virus, swine influenza virus, hereinafter referred to as P.V.M., as well as psittacosis virus, and lymphocytic choriomenin- another latent virus of mice belonging to the gitis virus. The results of these tests are shown psittacosis group of viruses (26), were recovered in Table VI. and identified. It was possible to show in each With but one exception, no significant increase instance that these agents were derived from in specific antibodies against any of these viruses the mice used and not from the human materials. was demonstrable in the serum obtained during Early in the course of this study, it was ob- convalescence as compared with the serum ob- 218 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFA-LL, JR.

TABLE VI Resuls of serological tests against certain pneumotropic viruses in patiets with primary atypical pneumonia

Number of patients tested Results of tests

Vlrus Method I.H.* OutYidet Total Significant No increase in R.I.H.*Outsidet Total increase in anioes patients patients patients antioofbients| No. of patients Influenza A R.B.C.t agglutination 47 49 96 1** 95 Influenza B R.B.C. agglutination 42 47 89 0 89 Swine influenza R.B.C. agglutination 8 0 8 0 8 Psittacosis Complement fixation 25 37 62 0 62 Lymphocytic choriomeningitis Complement fixation 16 12 28 0 28 * R.I.H. patients = Patients studied in the Rockefeller Institute Hospital. t Outside patients = Patients studied elsewhere. t R.B.C. agglutination = agglutination of chick erythrocytes (Hirst technique). ** R.I.H. patient studied during the 1943-44 influenza A epidemic. taned from the same patient during the acute tion tests in primary atypical pneumonia. In phase of illness. A definite antibody response the light of these observations, it seems necessary against influenza A virus occurred in the con- to re-evaluate the significance of serological valescent sera of one patient who had primary studies in this disease which were dependent atypical pneumonia at the Rockefeller Institute upon results obtained with the complement- Hospital during the outbreak of influenza A in fixation technique when relatively crude tissue the winter of 1943-44. The evidence obtained antigens were employed. from serological studies strongly suggests that Sera from 18 of the patients with primary none of the viruses tested was causally related atypical pneumonia were also tested for the to the illness in the group of patients included presence of the so-called "C-reactive protein" in this investigation. (30). It will be recalled that notably in pneumo- Tests for the presence of antibodies against coccal pneumonia, but also in other acute infec- Rickettsia diaporica were not carried out with tious diseases, there is present in the serum sera from these patients. However, various during the acute phase of illness an abnormal specimens obtained from most of the patients protein which has the peculiar property of react- were inoculated into susceptible animals and in ing with the C or somatic polysaccharide of the no instance were rickettsiae recovered. The pneumococcus. It was found that this protein attempts of others (28) to isolate rickettsiae was present in the acute-phase serum of all but were unsuccessful also and antibodies against the one of the patients with primary atypical rickettsia of Q fever were not demonstrable in the pneumonia whose sera were tested, and that it sera of any of their patients with primary atypical disappeared from their sera during convalescence, pneumonia. as it does in other acute infectious diseases. In the course of testing sera from the patients Sera obtained from 64 of these patients in this study for complement-fixing antibodies, throughout the course of primary atypical pneu- a peculiar and unexpected phenomenon was monia were tested for the presence of cold hemag- encountered (29). The convalescent sera from glutinins against human Group 0 erythrocytes 14 of 35 patients tested showed the capacity, (31, 32). It was found that sera of 12, or 18.5 not present in acute-phase sera, to fix comple- per cent, showed this property. ment with a variety of apparently unrelated Reports from other laboratories (31 to 36) have antigens prepared from normal and infected recorded a variable although generally higher organs of several different species. Similar re- incidence -of cold hemagglutinins in sera of actions were not observed in comparable sera patients with this disease. The lower percentage from 23 patients with other acute infectious of positive reactions and the relatively low titers diseases. This phenomenon obviously com- with sera of patients in this group may be at- plicates the interpretation of complement-fixa- tributable to the method used in collecting some ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 219 of the sera, to prolonged storage of certain of 106 patients were tested for the presence of the sera before testing, and possibly to the fact agglutinins against the non-hemolytic strepto- that many of the patients in this series were only coccus designated as streptococcus MG (22). mildly ill. It was found that 68, or 64 per cent, of the 106 Prior to the observation of cold hemagglu- patients developed in their sera during convales- tinins in this disease, some of the specimens of cence agglutinins against this microorganism. blood obtained from patients in this series were The incidence of positive reactions in agglu- allowed to clot at 40 C. before the serum was tination tests with streptococcus MG and serial separated. Subsequently, it was found that specimens of serum from these 68 patients (MG hemagglutinins present in high titer could be positive group) is shown in Figure 1. Although completely absorbed from a specimen of serum by contact with a suspension of human ery- Number of tested throcytes at a temperature of 40 C. It seemed 46 36 36 35 «era29 19 10 13. advisable therefore to allow blood specimens to 100 clot at room temperature and to separate the serum prior to refrigeration. It has been noted previously that in sera ob- *80- tained from patients with primary atypical pneumonia and stored for long periods of time, the capacity for cold hemagglutination appeared ~60 to be less than in sera recently obtained from similar patients (31, 33). One worker (35) re- tested 20 specimens of serum after storage at ,40O 40 C. for periods ranging from 2 to 5 months and found that, in all but 1, the titer of cold agglutinins had diminished. Seven specimens of 20 serum from patients in the present series were likewise tested on 2 occasions with similar results. At the time of the initial tests, the sera had been stored for periods varying from 2 weeks to 6 12 3 4 5 6 7 8 months but all showed cold agglutination, in dilu- Week of illnems tions ranging from 1:40 to 1: 160. When tested again 10 or 11 months later, 2 sera agglutinated FIG. 1. TIME OF OCCURRENCE OF POSITIVE REACTIONS erythrocytes with an undiminished titer of 1-80, IN AGGLUTINATION TESTS WITH STREPTOCOCCUS MG AND a and 3 SERA FROM 68 SELECTED PATIENTS WITH PRIMARY ATYPI- 2 agglutinated only in dilution of 1-10, CAL PNEUMONIA (MG-POSITIVE GROUP) had lost completely the property of cold ag- Each patient in this group developed in his serum during glutination. convalescence agglutinins against streptococcus MG. Tests for cold hemagglutination were carried Only one specimen of serum per patient per week was out upon sera of patients in the present series included in this analysis. using 2.0 per cent suspensions of erythrocytes as originally recommended. Ten of the 12 pa- positive reactions were noted in 15, 44, and 67 tients in this series whose sera contained cold per cent of the sera obtained from these patients hemagglutinins were ill for periods longer than during the first, second, and third weeks of ill- the average duration of illness for the total ness, respectively, they were found in 86 and 90 group. Other observers have also noted a per cent of the sera obtained during the fourth positive correlation between either the incidence and fifth weeks of illness. In 58 (85 per cent) or the titer of cold hemagglutinins and the patients, agglutinins were first detected in severity of illness in patients with primary specimens of serum obtained on the day of atypical pneumonia (31, 36). defervescence or later in convalescence. In the Acute and convalescent sera from each of the sera of 1.0 patients (15 per cent), agglutinins were 220 CURNEN, MIRICK, ZIEGLER., JR., THOMAS, AND HORSFALL, JR. noted from 2 to 19 days before defervescence, but ization, were considerably longer in the MG- the maximum titer of the sera of all but 3 of these positive group than in the MG-negative group. patients was not attained until after the tem- Comparable data for the 2 groups are illus- perature had returned to normal. It may be trated graphically in Figure 2. Fever persisted said, therefore, that agglutinins against streptococcus MG are most frequently detectable in sera of patients with primary atypical pneumonia 4 to 5 weeks after the onset of illness and, in most instances, make their first appearance following defervescence. In view of the frequency with which a serological response against streptococcus MG was observed in these patients, it seemed of interest to determine in what other respects the 68 patients who developed antibodies against this microorganism (MG positive group) could be distinguished from the 38 patients in whose sera 60 such antibodies were not demonstrable (MG- 50 I,:4 II negative group). 40 .4 I On the basis of clinical impressions, the MG- 0/1 Duvtion of positive group seemed to include patients who -v i physical si9ns were more ill than most patients in the MG-nega- *e2 tive group but no satisfactory quantitative cri- i' conso1idctioIi of teria except those relative to the duration of the vi1 illness could be found for estimating the relative severity of illness in individual patients of the 2 ~60 0 groups. From the standpoint of severity, the A I , symptoms, physical signs, and clinical-labora- 50 UsiU'0 tory findings of many individual patients in the 40 I Duration of 2 groups were indistinguishable. The average 30 X-ray evdence number of lobes involved was by physical ex- of conolidcitjox amination 1.64 and 1.39 and by x-ray 2.01 and 20 1.55 in the MG-positive and MG-negative 10 ~N groups, respectively. On the basis of x-ray I_ ~- I evidence, involvement of more than one lobe occurred in 56 per cent of the MG-positive group 40 and in 47 per cent of the MG-negative group. 30 _ Pe,oF-e Duvation of %t hospitaliz:ationl The average and range of the total leukocyte 20 0 counts, the degree and frequency of E.S.R. ac- celeration, the distribution of bacteria other than 10 streptococcus MG, and the negative results of serological tests against certain pneumotropic I X 3 e 5 6 7 8 9 30 11 viruses were essentially the same in both groups. Week of ilnes Nevertheless, it was possible to demonstrate FIG. 2. COMPARISON OF QUANTITATIVE DATA RELA- certain quantitative differences between patients TIVE TO THE DURATION OF ILLNESS IN THE MG-POSITIVE in the 2 groups taken collectively if data relating AND MG-NEGATIVE GROUPS OF PATIENTS WITH PRIMARY to the duration of illness were compared. It was ArYPIcAL PNEUMONIA found that the duration of of MG-positive group comprises those patients who de- fever, abnormal veloped in their serum agglutinmrs against streptococcus physical signs, and of roentgenological evidence MG. MG-negative group is composed of those patients of pneumonia, as well as the duration of hospital- in whose serum such agglutinins were not demonstrable. ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 221 for more than 2 weeks in 21 patients (30 per physical signs of pneumonia for 8.9 days, evi- cent) of the MG-positive group and in none of dence of pneumonia by x-ray for 11.4 days, and the patients in the MG-negative group. Phys- hospitalization for 26.2 days. In the groups of ical signs of pneumonia were present for more patients with maximum agglutinin titers in their than 2 weeks in 34 patients (50 per cent) of the sera of 1-10, 1-20, and 1-40 or more, respectively, MG-positive group and in only 3 patients (7 per the average duration of illness was progressively cent) of the MG-negative group. Roentgeno- longer in each of the 4 categories mentioned logical evidence of pneumonia was detectable above. Thus, the 20 patients in whose sera the for more than 3 weeks in 20 patients (29 per cent) maximum titer of agglutinins (against strepto- of the MG-positive group and in only 1 patient coccus MG) was 1-40 or more the average dura- of the MG-negative group. Similarly, 23 pa- tion of fever was 17.3 days, signs of pneumonia tients (34 per cent) of the MG-positive group were present for an average of 20.4 days, and but only 1 patient of the MG-negative group re- evidence of pneumonia by x-ray for 23 days, quired hospitalization for more than 6 weeks. while the average period of hospitalization was Further analysis revealed that the average 46.3 days. It should be pointed out that within duration of illness was longer, not only in the the groups of patients divided according to the MG-positive group as compared with the MG- titer of their sera against streptococcus MG, the negative group, but also within the MG-positive duration of illness in individual patients varied group in direct proportion to the maximum titer over a wide range. Moreover, as might be ex- of agglutinins against streptococcus MG. This pected, there was considerable overlapping in is shown in Table VII. The duration in days of the range of the values among the several titer groups. Nevertheless, these observations showed TABLE VII in patients of this series a positive correlation be- Relation of duration of illness to the titer of agglutinins against streptococcus MG in the sera of patients tween the duration of illness and the titer of with primary atypical pneumonia agglutinins against streptococcus MG. The MG-positive and MG-negative gtoups of Duration of illness patients also differed markedly with respect to - ~~~~Dura- tion of the occurrence in their sera of cold hemagglu- Maximum titer Physical X-ray hospital- of agglutinins Number Fever signs of evidence ization tinins. The results of serological tests for both against strep- at pneu- of pneu- tococcus MG patients monia monia cold hemagglutination and agglutination of Average Average Average Average number number number number streptococcus MG are shown in Table VIII. of days of days of days of days TABLE VIII 0 38 6.6 8.9 11.4 26.2 1-10 28 9.4 11.6 14.1 30.5 Results of testsfor cold hemagglutination andfor agglutination 1-20 20 10.8 14.7 15.8 35.4 of streptococcus MG with sera of patients 1-40 and over 20 17.3 20.4 23.0 46.3 with primary atypical pneumonia Total group 106 10.1 12.8 14.2 33.0 Agglutination of streptococcus MG Cold Total number hemagglutination Positive- Negative- of patients fever, of physical signs of pneumonia, of x-ray Number of Number of evidence of pneumonia, and the period of hos- patients patients pitalization were arbitrarily estimated for each Positive 13 3 16 Negative 19 36 55 patient from the onset of illness. The average Total 32 39 71 duration in each of these 4 categories was then calculated in groups of patients arranged according to the maximum titer of agglutinins against Sera from a total of 71 patients, including 64 pa- streptococcus MG observed in their sera through- tients in the present series and 7 patients studied out the course of illness. elsewhere,7 were tested for both types of agglu- As shown in Table VII, 38 patients with no 7 Dr. Warfield T. Longcope of the Johns Hopkins demonstrable agglutinins against streptococcus Hospital, Baltimore, Md., kindly provided the specimens MG had, on the average, fever for 6.6 days, of serum from these 7 patients. 222 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR. tinins. Convalescent sera from 32 of these 71 dence of each of these serological properties and patients agglutinated streptococcus MG. Of the duration of the illness. In the accompanying these 32 patients, the sera of 13 also.agglutina- paper (22), evidence is presented which shows ted human Group 0 erythrocytes in the cold. that, despite these similarities, the 3 phenomena The sera of 36 patients possessed neither agglu- are distinct and can be clearly differentiated. tinins against streptococcus MG nor cold hemagglutinins. The sera of 3 patients possessed DISCUSSION cold hemagglutinins but not agglutinins against During the past 10 years, the illness which has streptococcus MG. Thus it is evident that cold been designated by the term "primary atypical hemagglutinins occurred predominantly but not pneumonia" has become a familiar clinical entity exclusively in the sera of patients who devel- which can be recognized on the basis of certain oped agglutinins against streptococcus MG. diagnostic criteria. It is possible that this dis- In the 12 patients of the present series whose ease may have exisfed at the time of the last war sera contained cold hemagglutinins, the average and even earlier. However, as the validity of duration of fever was 16.4 days, pneumonia was diagnosis is dependent upon the careful exclu- detected by physical signs for an average of 20.2 sion of other diseases and to a considerable de- days, and by x-rays for an average of 20.7 days; gree upon the results of roentgenological exami- the average duration of hospitalization was 44.8 nations, there is a limit beyond which retroactive days. It is evident that the average duration of diagnosis becomes merely speculation. illness in the group of patients with cold hemag- The occurrence of primary atypical pneu- glutinins corresponded closely to that observed monia in certain members of a community in in the group of patients in whose sera the titer which undifferentiated forms of acute respira- of agglutinins against streptococcus MG was tory infection are prevalent has been frequently 1-40 or more. observed and suggests that this form of pneu- It was pointed out previously that the con- monia may be but one manifestation of a wide- valescent sera from 14 of the patients in the spread respiratory disease. This view is sup- present series were capable of fixing complement ported by epidemiological observations (37). with a variety of apparently unrelated tissue In making the diagnosis of primary atypical antigens. All but 1 of the patients in whose pneumonia, certain clinical features are particu- sera this property was demonstrated were in the larly helpful. The gradual onset of malaise, MG-positive group. The sera of five also con- cough, headache, and fever, the lack of respira- tained cold hemagglutinins. The average dura- tory distress, the relative bradycardia, the tion of fever in these 14 patients was 13.2 days. normal leukocyte count, and the paucity of Pneumonia was detectable by physical signs for pulmonary signs on physical examination in the an average of 18.5 days and by x-ray for 17.3 presence of x-ray evidence of pneumonia are days. The average period of hospitalization sufficiently frequent and characteristic findings was 37.7 days. Thus, the average duration of of this disease syndrome to warrant a presump- illness in these patients was considerably longer tive diagnosis. than the average duration of illness for all the X-ray evidence of pneumonia may not neces- patients and was similar to that observed, not sarily be present at the initial examination but only in patients who had higher titers of agglu- usually appears early in the course of illness and tinins against streptococcus MG, but also in the clears completely within a few weeks. The group of patients with cold hemagglutinins. x-ray shadow frequently is diffuse and radiates These findings indicate that agglutinins against from the hilum but may assume a variety of streptococcus MG, cold hemagglutinins, and the patterns and, at times, is indistinguishable from property of non-specific complement-fixation that observed in other diseases, particularly other may coexist in specimens of serum from indi- varieties of -pneumonia, bronchiectasis, and vidual patients with primary atypical pneu- tuberculosis. The duration of illness should be monia. Moreover they strongly suggest that considered in interpretating x-ray films of the there is a positive correlation between the inci- chest in patients with this syndrome and the ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 223 rate of clearing is of particular value in excluding though the significance of the phenomenon has a diagnosis of chronic pulmonary disease. yet to be ascertained, its value as a practical By appropriate techniques, pneumococci can laboratory aid to diagnosis has been disap- frequently be recovered from patients with the pointing. characteristic features of primary atypical pneu- Another phenomenon which has been observed monia. It is unusual however to find them in with the convalescent sera of certain patients this disease by direct examination of sputum, with primary atypical pneumonia is the capacity employing the quellung technique. Moreover, to fix complement with unrelated tissue antigens the pneumococcal types isolated are rarely those (29, 38). This property also does not occur with most frequently associated with lobar pneu- sufficient frequency to be reliable as a diagnostic monia. It should be pointed out, however, technique. Nevertheless, its recognition has that during the administration of sulfonamide served to demonstrate further the frequency drugs, a considerable reduction in the number of with which non-specific reactions occur in the susceptible microorganisms in the nasopharynx sera of patients with primary atypical pneu- and sputum may take place. As a consequence, monia. It also has made necessary a more the presence of certain pathogenic micro- critical evaluation of serological data based upon organisms, particularly pneumococci or beta the results of complement-fixation tests in hemolytic streptococci, may not be detected. studies of this disease. It is therefore important in distinguishing this It has been pointed out that the clinical pic- syndrome from other forms of pneumonia, that ture of primary atypical pneumonia may repre- the bacteriological examination be carried out sent instances of infection by different etio- whenever possible before sulfonamide therapy is logical agents. In certain patients, Rickettsia instituted. diaporica (11) and the psittacosis group of viruses The sulfonamide drugs have not been found (12, 13) have been shown to be responsible. to be effective in primary atypical pneumonia These agents, however, have been isolated from and are not recommended for treatment. In or otherwise identified with relatively few cases. certain instances, however, it may be impossible In the present series of patients, it was found to determine whether or not pneumococci or that neither of these agents had caused the hemolytic streptococci isolated from a patient disease and ample evidence is now available to are implicated in the pathogenesis of the illness. indicate that they were not responsible for the Under such circumstances, an adequate trial of great majority of cases which have been studied sulfonamide therapy is justified. Prompt defer- during recent years. vescence following the administration of a sul- The recovery of several infectious agents from fonamide drug may be an indication of effective patients with primary atypical pneumonia has therapy but may occur also as a coincidence. been reported. A virus obtained by one group Consequently, an apparent response to sul- of workers (39) appeared to be infectious for fonamide therapy does not necessarily exclude mice, guinea pigs, and ferrets, but was lost after a diagnosis of primary atypical pneumonia. a few passages. Others (14) described the Early reports on the occurrence of cold hemag- recovery of a virus which was infectious for the glutinins in the sera of patients with primary mongoose, and presented evidence indicating atypical pneumonia (31 to 33) indicated that this that the agent could be neutralized by the sera phenomenon might provide a technique useful of patients convalescent from primary atypical in diagnosis. It is now apparent, however, that pneumonia. Unfortunately, this species was many patients with primary atypical pneumonia not available for an attempt to repeat these do not develop in their sera the capacity to observations with materials from patients in the agglutinate erythrocytes in the cold. Further- present series. more, it has been found that this serological Two viruses recovered from cats have been property occurs not only in patients with primary described in relation to primary atypical pneu- atypical pneumonia but also in an appreciable monia in man. Epidemiological evidence has number of patients with other conditions. Al- been presented (16) to indicate that a few cases 224 CURNEN, MIRICK, ZIEGLER, JR., THOMAS, AND HORSFALL, JR. of the human disease in one family might have pigs, which produced pulmonary consolidation been related to a respiratory infection of cats in cotton rats and hamsters and resembled an which was caused by a virus apparently infec- agent obtained from normal hamsters. tious for this species but not for mice. This The significance of any of these viruses in the agent is different from the other feline virus, pathogenesis of primary atypical pneumonia as recently isolated (40), which has been shown observed in the majority of patients remains to to belong to the psittacosis-lymphogranuloma be determined. group of viruses (41). As reported previously (19), a non-hemolytic Certain investigators (15) observed that pul- streptococcus, isolated originally from the lungs monary consolidation developed in cotton rats of patients who had died of primary atypical following primary intranasal inoculation of pneumonia, was found to react with the sera of sputum from certain patients with primary many patients convalescent from this disease. atypical pneumonia and presented evidence for The characteristics of this microorganism, which the existence of a transmissible agent. Re- has been designated streptococcus MG, have cently (18), these workers have reported the been investigated intensively. The results of recovery of a filterable agent from certain pa- these studies are presented in separate com- tients with atypical pneumonia. They also munications (20, 21). In the accompanying reported that this agent was transmissible in paper (22), studies concerning the relation of chick embryos and was capable of inducing streptococcus MG to primary atypical pneu- pulmonary lesions after intranasal inoculation monia are presented in detail and the possible in either cotton rats or hamsters. Moreover, r8le of this microorganism in the pathogenesis evidence was presented to show that the agent of the disease is discussed. could be neutralized by the convalescent serum but not the acute-phase serum from patients SUMMARY with primary atypical pneumonia. They also A comprehensive study of 106 patients- with reported that pulmonary lesions were induced primary atypical pneumonia is presented. The in both cotton rats and hamsters by the intra- study includes observations on clinical charac- nasal inoculation of non-infectious materials and teristics, roentgenological, electrocardiographic, that these lesions were similar to those induced and clinical-laboratory findings, as well as the by the agent. Furthermore, they found that results of biochemical, bacteriological, viral, and latent agents present in both species were en- serological investigations. The observations countered frequently when serial passages were made upon patients in this series are correlated carried out. with the findings reported in other studies on As has been pointed out, the agent previously this disease. described in a report from this laboratory (27) was found on further study to be indistinguish- BIBLIOGRAPHY able from latent agents antigenically related to 1. Official Statement: Primary atypical pneumonia, the pneumonia virus of mice (25). In the course etiology unknown. War Med., 1942, 2, 330. of the present study, evidence has been obtained 2. Gallagher, J. R., Bronchopneumonia in adolescence. which indicates that such latent agents are Yale J. Biol. and Med., 1934, 7, 23. present in cotton rats, hamsters, and rabbits. 3. Bowen, A., Acute influenza pneumonitis. Am. J. A filterable virus recovered from patients Roentgenol., 1935, 34, 168. 4. Allen, W. H., Acute pneumonitis. Ann. Int. Med., with primary atypical pneumonia (17) was found 1936, 10, 441. to be transmissible on intranasal inoculation in 5. Arrasmith, T. M., Jr., Influenzal pneumonia-a young guinea-pigs and cotton rats. The results clinical report, with special reference to diagnosis. of neutralization tests with human sera were U. S. Nav. M. Bull., 1930, 28, 769. inconclusive. It was found (18) that guinea- 6. Dingle, J. H., and Finland, M., Virus pneumonias: pigs frequently developed lung lesions after II. Primary atypical pneumonias of unknown etiology. New England J. Med., 1942, 227, 378. intranasal inoculation of broth and one agent 7. MacLeod, C. M., Primary atypical pneumonia. was recovered, by intranasal passage in guinea- M. Clin. North America, 1943, 27, 670. ATYPICAL PNEUMONIA, CLINICAL AND LABORATORY STUDIES 225

8. Owen, C. A., Primary atypical pneumonia: an analysis 22. Thomas, L., Mirick, G. S., Curnen, E. C., Ziegler, of seven hundred and thirty-eight cases occurring J. E., Jr., and Horsfall, F. L., Jr., Studies on during 1942 at Scott Field, Ill. Arch. Int. Med., primary atypical pneumonia. II. Observations 1944, 73, 217. concerning the relationship of a nonhemolytic 9. Commission on Acute Respiratory Diseases, Fort streptococcus to the Disease. J. Clin. Invest., Bragg, N. C., Primary atypical pneumonia. Am. 1945, 24, 227. J. Pub. Health, 1944, 34, 347. 23. Emerson, K., Jr., Curnen, E. C., Mirick, G. S., and 10. Finland, M., and Dingle, J. H., Virus pneumonias: Ziegler, J. E., Jr., Chloride metabolism and plasma I. Pneumonias associated with known nonbacterial amino acid levels in primary atypical pneumonia. agents; influenza, psittacosis and Q fever. New J. Clin. Invest., 1943, 22, 695. England J. Med., 1942, 227, 342. 24. Farr, L. E., MacLeod, C. M., Futcher, P. H., Emerson, 11. Dyer, R. E., Topping, N. H., and Bengston, I. A., An K., Jr., Mirick, G. S., and Curnen, E. C., Hypo- institutional outbreak of pneumonitis. II. Isola- aminoacidemia in patients with pneumococcal tion and identification of causative agent. Pub. pneumonia. Proc. Soc. Exper. Biol. and Med., Health Rep., 1940, 55, 1945. 1940, 44, 290. 12. Eaton, M. D., Beck, M. D., and Pearson, H. E., A 25. Horsfall, F. L., Jr., and Hahn, R. G., A latent virus virus from cases of atypical pneumonia: relation to in normal mice capable of producing pneumonia in the viruses of meningopneumonitis and psittacosis. its natural host. J. Exper. Med., 1940, 71, 391. J. Exper. Med., 1941, 73, 641. 26. Thomas, L., and Kolb, E. M., Activation of latent 13. Smadel, J. E., Atypical pneumonia and psittacosis. mouse pneumonitis virus by human serum. Proc. J. Clin. Invest., 1943, 22, 57. Soc. Exper. Biol. and Med., 1944, 55, 1. 14. Weir, J. M., and Horsfall, F. L., Jr., The recovery from patients with acute pneumonitis of a virus 27. Horsfall, F. L., Jr., Curnen, E. C., Mirick, G. S., causing pneumonia in the mongoose. J. Exper. Thomas, L., and Ziegler, J. E., Jr., A virus recovered Med., 1940, 72, 595. from patients with primary atypical pneumonia. 15. Eaton, M. D., Meiklejohn, G., van Herick, W., and Science, 1943, 97, 289. Talbot, J. C., An infectious agent from cases of 28. Dingle, J. H., Abernethy, T. J., Badger, G. F., atypical pneumonia apparently transmissible to Buddingh, G. J., Feller, A. E., Langmuir, A. D., cotton rats. Science, 1942, 96, 518. Ruegsegger, J. M., and Wood, W. B., Jr., Primary 16. Blake, F. G., Howard, M. E., and Tatlock, H., Feline atypical pneumonia, etiology unknown. War Med., virus pneumonia and its possible relation to some 1943, 3, 223. cases of primary atypical pneumonia in man. 29. Thomas, L., Curnen, E. C., Mirick, G. S., Ziegler, J. Yale J. Biol. and Med., 1942, 15, 139. E., Jr., and Horsfall, F. L., Jr., Complement 17. Rose, H. M., and Molloy, E., Observations concerning fixation with dissimilar antigens in primary atypical the etiology of primary atypical pneumonia. pneumonia. Proc. Soc. Exper. Biol. and Med., Science, 1943, 98, 112. 1943, 52, 121. 18. Eaton, M. D., Meiklejohn, G., and van Herick, W., 30. Tillett, W. S., and Francis, T., Jr., Serological reactions Studies on the etiology of primary atypical pneu- in pneumonia with a non-protein somatic fraction monia: a filterable agent transmissible to cotton of pneumococcus. J. Exper. Med., 1930, 52, 561. rats, hamsters, and chick embryos. J. Exper. 31. Peterson, 0. L., Ham, T. H., and Finland, M,, Cold Med., 1944, 79, 649. agglutinins (autohemagglutinins) in primary atypi- 19. Thomas, L., Mirick, G. S., Curnen, E. C., Ziegler, cal pneumonias. Science, 1943, 97, 167. J. E., Jr., and Horsfall, F. L., Jr., Serological 32. Turner, J. C., Development of cold agglutinins in reactions with an indifferent streptococcus in atypical pneumonia. Nature, 1943, 151, 419. primary atypical pneumonia. Science, 1943, 98, 33. Horstmann, D. M., and Tatlock, H., Cold agglutinins: 566. a diagnostic aid in certain types of primary 20. Mirick, G. S., Thomas, L., Curnen, E. C., and Horsfall, atypical pneumonia. J.A.M.A., 122, 369 (June 5) F. L., Jr., Studies on a nonhemolytic streptococcus 1943. isolated from the respiratory tract of human beings. 34. Turner, J. C., Nisnewitz, S., Jackson, E. B., and I. Biological characteristics of streptococcus MG. Berney, R., Relation of cold agglutinins to atypical J. Exper. Med., 1944, 80, 391. pneumonia. Lancet, 1943, 1, 765. 21. Mirick, G. S., Thomas, L., Curnen, E. C., and Horsfall, 35. Meikeljohn, G., The cold agglutination test in the F. L., Jr., Studies on a nonhemolytic streptococcus diagnosis of primary atypical pneumonia. Proc. isolated from the respiratory tract of human beings. Soc. Exper. Biol. and Med., 1943, 54, 181. II. Immunological characteristics of streptococcus 36. Commission on Acute Respiratory Diseases, Station MG. J. Exper. Med., 1944, 80, 407. III. Im- Hospital No. 2, Fort Bragg, N. C., Cold hemagglu- munological relationship of streptococcus MG to tinins in primary atypical pneumonia and other Streptococcus salivarius Type I. J. Exper. Med., respiratory infections. J. Bact., 1944, 47, 460. 1944, 80, 431. 37. Commission on Acute Respiratory Diseases, Fort 226 CURNEN, MIRICK, ZIEGLER; JR., THOMAS, AND HORSFALL, JR.

Bragg, N. C., Epidemiology of atypical pneumonia infections. Tr. and Stud., Coll. Physicians, Phila- and acute respiratory disease at Fort Bragg, North delphia, 1939, 6, 329. Carolina. Am. J. Pub. Health, 1944, 34, 335. 40. Baker, J. A., A virus obtained from a pneumonia of 38. Eaton, M. D., and Corey, M., Complement fixation cats and its possible relation to the cause of atypical in human pneumonitis with group-reactive virus pneumonia in man. Science, 1942, 96, 475. antigens. Proc. Soc. Exper. Biol. and Med., 1942, 41. Thomas, L., and Kolb, E. M., Relationship of the 51, 165. virus of cat pneumonia (Baker) to the psittacosis- 39. Stokes, J., Jr., Kenney, A. S., and Shaw, D. R., A lymphogranuloma group of agents. Proc. Soc. new filtrable agent associated with respiratory Exper. Biol. and Med., ,943, 54, 172.