Pneumocystis pneumonia REVIEW carinii pneumonia immunology pneumonia, alveolar interstitial plasma cell capillary block pentamidine isothionate Pneumocystis Carinii Pneumonitis A Review

Departments of Pediatrics and Microbiology, University of Florida, Gainesville, Fla. 32601, USA

Introduction In many instances, the desperate clinical condition of the affected individuals seems to have precluded The purpose of this essay is to organize the available those clinical studies necessary to understand the information about Pneumoqwtis carinii pneumonitis and distinctive and altered physiology of Pneumocystis ca- to rclatc the host factors and the pathophysiology with rinii pneumonitis. However, there are some reports, as the clinical disease. Since a successful chemotherapeu- well as unreported data available to the author, which tic approach has been rcported by several workers, correlate the pathological findings with the clinical and since the incidence of this disease seems to be in- picture of this disease. The review will try to emphasize creasing in a select group of patients, it is important these pathophysiologic features in order to present that the early ante nzortem diagnosis of this disease be those diagnostic hints which the physicians may readily established. The emphasis in this essay, therefore, will test. Rvcognition of these unusual clinical features be to relate the unique characteristics of the pathogen should aid early diagnosis of Pneumocystis carinii pneu- to the pathophysiological state characteristic of the monitis. pneumonitis induced by Pneumocystis carinii. It is hoped Of particular interest is the possibility that accurate that a logical analysis of the data with illustrative case analysis of the incidence of immunoglobulin deficien- material will permit a distinct clinical picture to cy syndromes in the general population may be in- emerge and to pose some new question about the pa- fluenced by Pneumocystis carinii infections. Statistics thogen and the unique host-parasite interaction which that project the incidence of immunoglobulin defi- may be subject to experimentation. ciency syndromes in the general population may dis- Several reviews have been extremely useful in the regard many cases of immunoglobulin deficiency dis- preparation of this manuscript and should be rrferred ease, especially hypogammaglobulinemic states of the to for a historical prospective as well as a morc com- congenital variety, which were previously classified as prehensive revirw of this literature [I-61. examples of the newborn human's unusual susceptibil- Pneunzo~ys/iscarinii pneumonitis is a disease induced ity to Pneunlocystis disease. The wide-spread use of more by an ubiquitous organism of low virulrnce and narrow precise diagnostic techniqurs for the immunoglobulin tissue spccificity in a susccptible host. In most instances, deficiency syndromes, as well as the increased atten- it is a latent infection which occurs in humans and tion of pathologists to this diagnosis, should reveal that animals at all stagrs of development. The distinction, many children with hypogammaglobulinemic states drawn by several workcrs, between the infantile and are afflicted with Pneumocystis carinii infection within adult form of the discasr may be due to the narrow age the first year of life. The difficulties in substantiating a range of patients obsrrvcd by individual authors. One diagnosis of hypogammaglobulinemia during this age aim of this review, therefore, will be to show that the period, however, must be appreciated. same host factors which permit infCction with Pneumo- cvstis carinii are operative in humans and animals of all ages. For instance, contagion, formerly thought to Case 2,figures IA-D be exclusively a feature of the infantile form of the disease has been shown to be an important factor in the E.T.F., an 8-year-old girl, was admitted to the Uni- transmission of thc disease to adults. vcrsity of Florida Teaching Hospital for the second Folds, Lisa Age. ll years Diognosis:Acquired Hypogammaglobulinemia

~~NUCOIL~CICD~~C~~SI~CELLULOSE ACETATE ELECTROPHORESIS

Tolal Protein - 6.29gm.r Albumin . 3.77 01 . 0.33 n2 - 0. 84 6. 1.01 7 . 0.13 SERUM IMMUNOGLOBllLlN LEVELS

CLINICAL SYMPTOMS: Irohemrgglutlnins: BlooO lypP A Anti-8 - 0 Freouent baclerial pneumonia chrbn~c011tir media wilh hearing deficlf Anli.d,pnther,a 0. 0015 ~.ii.iml. Pneumocyrlis C-I pneumon~llr bone marrow plasma cellr. OllDOO

Fig.1. A. Laboratory data. B. X-ray, acute phase of Thc vital capacity as determined with a Collins vitalo- disease. C. Onc month post pentamidine therapy. meter was 0.7 liter (normal 2.0 liter). X-ray study D. Three years post therapy. revealed minimal pulmonary scarring without evid- ence of acute disease. An appropriate work-up was con- time on July 31, 1964, with the chiefcomplaint offever sistent with a diagnosis of hypogammaglobulinemia. and dry cough, associated with increasing respiratory Treatmrnt was started with intramuscular injec- difficulty. The patient had had recurrent respiratory tions of pooled commercial Cohn fraction 11, 0.6 ml infections during her entirr life. Numerous bouts of per kilogram of body weight every month. She re- otitis media had necessitated multiple myringotomies. mained in relatively good health until 4 weeks before Severe attacks of bronchitis, with several episodes of admission, when a dry cough developed. Two weeks pnrumonia, were succcssfully treatcd with antibiotics. later she bccamr slightly febrile, with lethargy and Pneutnocystis carinii pneumonitis 133

Motz, Scott R. Age:3'/2 months ~iagnosis:~~~~~'~- hypogornrnoglobulinemto

IMhlUNOELECTROPHORESIS CELL1II.OSE ACETATE ELtClROPHOHISl5 ff A*& Total Protein . 2. Iiqm.?. Atburntn . 1.34 a1 . 0. 14 02 ' 0.10 &an R rn P . o. 19 1 . 0.09 SERUM IMMUNOGLOBULIN LEVELS I A

decreased appetite. Thr cough rcmaincd nonproduc- After one week of therapy definite clinical improve- tive, and there was no improvemrnt in thc clinical ment was observed. Decreased respiratory embarrass- condition dcspitc. tetracycline therapy. X-ray study of ment, as indicated by gradual diminution in cyanosis, the chest on the day of admission revealed diffuse bi- less dcpendcncy upon oxygen and a rcturn of appetite, lateral infiltrates most prominent at the hilar regions. was noticed during the 1st wcrk of thcrapy. Roentgeno- Physical examination showed a thin, chronically ill graphic evidence of improvement was detectable 3 girl in no acutc distress. Diminished chest expansion, weeks after treatment was started. Bone-marrow aspi- decreased breath sounds and hyporesonant percussion rations before therapy and on the 7th, 1 lth, and 14th were noted in the lung fields bilaterally. No advcnti- days of therapy showed the gradual de\~elopmcntof tious sounds werr heard, and clubbing was not present. crythroid hyperplasia (mycloid-erythroid ratio of 1 : 2), The temperature was 101.5"F (38.6"C), the pulse with marked megaloblastic changes. Occasional hyper- 130, and the rrspirations 44. The blood pressure was segmrnted ncutrophils and macroovalocytes wrre seen 95175. in the pcriphcral blood. Thr white-cell count was 2 1,000, with 87",, neutro- Clinical improvcment continued, and the patient phils, 6",, band forms, 3O,, eosinophils, 2 ",lymphocytes was discharged on the 48th hospital day with routine and 2 % monocytes. The hematorrit was 37",,; the rcd therapy of pooled gamma globulin at monthly intcr- cells and platelrts had normalmorphology. Urinalysis vals. Bone-marrow aspiration done the day before dis- was ncgative. Multiple cultures revcalcd no pathogcns. charge was within normal limits. One month later she Fungi and P. mrinii, sought with specific stains, were exhibitcd littlc respiratory embarrassment, with a not demonstratcd. High doses of penicillin, kanamycin weight gain of 2 kg, and was maintaining a full pro- and oxacillin wcre administerrd, but the patient's con- gram of activity without difficulty. The vital caparity dition deteriorated over the next 7 days. Breathing was 0.8 liter (normal, 2.0 liter) at that time, not signifi- became laborrd and shallow; thc respirations rose to cantly di!Yercnt from that obtained at a clinic visit 80. Attcmpts at removal from a moist oxygen-enriched before this episode (0.7 liter). environmrnt provoked severe cyanosis. Thoracotomy Continued sino-pulmonary infections have persisted under light gencral anesthrsia was performed to obtain despite antibiotic thcrapy. Her \pita1 capacity has risen a lung biopsy. A rrlatively avascular, yellowish, stiff to 1.5 liters. Currently, she enjoys a full range of activ- pulmonary parenchyma was noted. The pleural sur- ity. faces were grossly normal. The biopsy specimen ob- tained was characteristic in appearance for Pr~~u~troryslis cariuii pncumonitis. Immediately after the diagnosis Ca~e2,,Jigure~ PA arid B was established by means of the frozen section, hydro- xystilbamidine, 4 mg per kilogram of body weight, This four month old male was brought to the hospital was injected intravenously each day. Three days later with a chirfcomplaint of increasing respiratory distress, this was replaced by therapy with pentamidine iso- cyanosis, fever, and weight loss. Following a normal thionate, 4 mg per kilogram intramuscularly, which gestation and normal delivery, this infant developed was then given for 14 days. During this period, all other frequent severe upper rrspiratory infections until the chemotherapy was discontinued. age of threc and a half months. These wrrr succcssfi~lly treated with antibiotics. At age three and a halfmonths, specific organism was cultured at that time. Following he was admitted to a local hospital where he was discharge, he was placed on very large dosrs of Pred- treated with blood transfusions, antibiotics, and paren- nisone, because of large leukemic lymph nodes and a teral gamma globulin for severe upper respiratory in- marked rise in his white count. He responded to this fection. With increasing cyanosis and respiratory dis- treatment and subsequently was placed on a mainten- tress, he was transferred to the University of Florida ance dose of Prednisone of 15 mg per day. During Teaching Hospital. October of 1966, he was given two 2-week courses of Upon admission, he was noted to be a severely ill Chlorambucil. Complaints of cough, weakness and child with c~anosisand dyspnea. Scattered rales were shortness of breath and also a congestive feeling in heard throughout both lung fields. There was no pal- the chest were related. About ten days prior to ad- pable lymph node tissue and the liver and spleen were mission, fever was noted, and the symptoms men- not enlarged. White blood count was 12,000 with 700 tioned above seemed to be worsening. One week be- lymphocytes per mm3. Nose and throat and blood fore admission, penicillin therapy was given in con- cultures were essentially non informativr. Blood trans- junction with large doses of gamma globulin in the fusions, antibiotics, and oxygen were offered but the assumption that once again this was a pulmonary in- child expired 24 hours after admission. fection, but this treatment was of no value and his chest Autopsy verified the diagnosis of thymic aplasia with symptoms increased. hypogammaglobulinemia. No thymus was detected Physical examination on admission revealed his tem- after careful sectioning of the mediastinal contents. perature was 101°F. He was slightly pale, acutely and Diffuse Pneutnocystis carinii pneumonitis was present. chronically ill. There were large, nontender, rubbery There were no inclusion bodies seen. nodes in the anterior and posterior cervical region as well as in both axilla. The heart was normal, except for a mild tachycardia. A few rales were heard in the Case 3,,bgures 3A and B right lower lobe. The spleen was palpated three inches and the liver was one inch down below the costal This 57-year-old man was admitted on November 12, margin. Chest x-ray revealed an interstitial infiltrate, 1966. In 1963, a diagnosis of chronic lymphatic leuk- especially in the right lung. emia wasestablished, and in 1964 and 1965, the patient On admission, the hemoglobin was 11.1 and the had radiation to the mediastinum because of the in- white count 200,000. The urine was negative. The uric volved nodes having caused some venous compression. acid was 6.6. Blood and sputum cultures as well as A short course of cytoxan therapy was administered fungal skin tests, gastric aspirates for the culturing of and small doses of Chlorambucil were given in 1965. tuberculosis and fungi were all negative. In February of 1966, the patient had chest pain, which Antibiotic therapy was begun. Penicillin in high was attributed to arteriosclerotic heart disease. In doses was initially given, followed by Keflin, along March of 1966, he was admitted to the hospital for with streptomycin. These proved to be of no value and bronchopneumonia and he responded to massive doses the fever spiked to 103 during the first two weeks in of penicillin, Chloromycetin, Keflin and streptomycin, the hospital. Chest x-ray became progressively worse, along with very large doses of gamma globulin. No showing very extensive interstitial pulmonary infiltra- Pneumocystis carinii pneumonitis 135

IMMIJNOELECTROPHORESIS CfIIUIOSE ACETATE ELtClROPHORtStS

Tutal Prote~n. 4.34ym.B Atbum~n. 2.85 0 . 0.13 0: " 0.54 &LMN. a p . 0.54 1 , 0.27 SERUM IMMIJNOGLOBULIN LEVELS

IgA. 28mq.~ IgM . 34 mg.E lyG . 220 m9.r

CLINICAL SYMPTOMS: Irohtnagglut~nlnr; Blood Type A Cyanosts Ant)-0 - 0 Oyrpnra Tacypnea Anti-d~pliltteria lor~n: 0.012 I. lJ.lmt. Fallurv to Ihrivr. tion. On the 15th hospital day, an open lung biopsy versity of Florida Tcaching Hospital with a chief com- substantiated the suspected diagnosis of pneurnocystis plaint of increasing rcspiratory distress and difficulty carinii infection. Immediately aftcr surgery, Dihydro- in feeding of approximately two wceks duration. This stilbamidine therapy was given in a dose of 250 mg per was the eighth child of a 28-year-old mother whose day. This was discontinued after the fourth day be- gestation lasted for eight months. Delivery was normal causr of severe toxicity manifested as intense, lower but the child developed respiratory distress six hours abdominal cramping pain about four hours after the after birth which was treated by oxygen administration infusion. Pcntamidine was obtained with FDA ap- and supportive fluid therapy. The child improved and proval and this was givrn for ten days in a dose of 3.2 was sent home following several wceks of antimicrobial g over the ten-day-period total. During the period of therapy for a mild upper respiratory infection. He was treatment, his fever slowly came down and his respira- readmitted to the hospital at age one month with a tory symptoms decreased. Surgrry itself was very diffi- respiratory infection which was treated with anti- cult as there was a great degree of alveolar capillary microbial agents for two weeks. Because of the increas- block and the pulmonary compliance was very low. ing respiratory distress and the diffuse nature of the The anesthetist had great difficulty bag breathing the infiltrate of process and x-ray, he was admitted to our patient because of this. Postopcratively, thcrc was a service. The seven older siblings and parents were well great deal of respiratory distress, which was managed and healthy. Admission showed pulse rate of 120, res- with a Bcnnett respirator and continuous use ofoxygen. piratory rate of 84, and a blood pressure of 70 mm. The On the 30th hospital day, just after completion of child was extremely dyspneal with a weak cry and oc- Pentamidine treatment, the patient suddenly develop- casional ralcs were heard over the posterior portions ed coma. The clinical diagnosis was a small cerebro- of both lung fields. White blood count was 12,000 with vascular accident, involving the pontine area. Over the 4,000 to 5,000 lymphocytes per mm3. next three days, the patient slowly came out of coma No organisms were cultured from the nasopharynx and was left with a residual stupor and signs of pseudo- that were considered significant pathogens and re- bulbar involvement. He had difficulty swallowing and peated search for Pneumocystis carinii cysts were nega- speaking. tive. At open thoracotomy the lungs were slightly in- The patient was discharged aftcr inkrim treatment flammed and the pleural surfaces were clear. Lung of a subdiaphragmatic abscess, much improved. He biopsy showed an interstitial pneumonitis without any was beginning to cat and walk about. During his stay evidence of Pneumocystis carinii or cytomegalic inclusion in the hospital, his Prednisone therapy was disconti- disease. Supportive therapy, including high humidity, nued. oxygen and parenteral fluids were offered. After two The patient, at present, is continuing to convalesce weeks, the baby's condition improved as evidenced by satisfactorily and is on no medication at all. the decreased respiratory and pulse rates and weight gain. The child was discharged from the hospital three weeks after admission with increasing weight Case 4, Jigures 4 A and B gain and a diminution in the extent of the infiltrative process in the lung. His local physicians confirmed a con- A two-month-old male infant was brought to the Uni- tinued improvement in growth of the infant at home. Pancoast, David Age: 4 months

IMMUNOELErTROPHORESIS

L#K&@

SERUM IMMUNOGLOBULIN LEVELS

CLINICAL SYMPTOMS: Increasing respiratory di~lresr Cyanosis, dyspnea, tachypnea .A 1 Case 5,,figures 511 and B Table I. The worldwide distribution of ~neunrocyslisca- rinii is shown by the variety of countries with affected h three and onr half month old male infant was ad- individuals. mitted to the hospital with a chief complaint of in- Countries reporting Pneurnocyslis carinii pneumonitis creasing breathlessnrss and coughing. Maternal preg- nancy and delivery was normal and the birth weight Australia Italy was sevrn pounds. A normal three year old brother and Belgium Korea both parents were healthy. Normal development was Brazil New Zcaland arrested at age three months when signs and symptoms Canada Norway of an upper respiratory infection appeared. Antimicro- Chile Poland bial chemotherapy was administered without change Congo Spain in the progression of respiratory distress. Denmark Sweden A lung biopsy taken from the right middle lobe was Finland Switzerland originally interpreted as Pneunrocystis carinii pneumo- Germany Union of South Africa nitis. X-ray and serum protein analysis are shown. No Hungary United Kingdom effect of pentamidine isothionate therapy upon his Iran USA pulmonary disease was detected. The patient expired Ireland USSR following worsening of his cyanosis and dyspnea. Sec- Israel Yugoslavia tions rrmoved at post mortem examination as well as material takrn at thoracotomy were diagnosed as pul- Table II. 'Natural' and 'experimentally induced' in- monary alveolar proteinosis by Dr. A.A. Liebow, Yale fection may be found in a wide variety of animals. The University, New Haven, Connecticut, as well as the significance of 'natural' infection in animal to human , pathologists at San Diego, California and Gainesville, disease is not known. Florida. This material is presented to illustrate the difficulty in distinguishing thc signs and symptoms of Animal hosts for Pneumocystis cnrinii pulmonary alveolar protrinosis from those detected in 1. 'Natural' patients with Pneunlocy~ti~carilrii pneumonitis. \ a) rats b) dogs c) guinea pigs Ubiquig of Pneumocystis carinii d) man e) monkeys The drtection of the organism as a pathogen in human f) sheep disrase or in rxperimental wild and domestic animals appear to be world wide [7-131 (tables I and II). g) goats 2. 'Experimentally induced' Although formerly presumed to be an organism found a) suckling young and adult rats in temperatr climates, thr detection of Pneunlocystis b) suckling and adult rabbits rcrit:ii pneumonitis in the Congo, Central America, c) mice Middle East and Equatorial Pacific islands among Pneu?nocystis carinii pneumonitis 137 other examples, indicates that the organism is ubiqui- in nurseries and as well as other sourccs of contagion to11s [14-201. The mode of transmission is not well and animal vectors has bcen difficult because of the understood. Pandemics of the disease that occurred in inability to obtain a pure source of antigcn from in- nurserit-s in the middle European countries following fected lungs [lo]. Animal sources of Pneuntocystis cnrinii World \Var I1 have been explained as due to contact have been poor sources of serological reagents [I, 101. with carricrs of the organism. Serological analysis of It has been postulated that serological specific strains nursing prrsonnel and thcir patients in nurseries using of the parasite are related to the spccirs undcr study. a complement fixation tcst indicate that active immune Alternatively, the difficulties encountered may be due response of adult individuals may bc related to the con- to the reactions of lung tissue contaminants of the ex- tinued occurrence of Pneuinorv~tiscuriniiinfcction among perimental animal with human serum which must be children [2 1-24]. removed before the measurement of the reaction of If a similar increase in Pneun~ocysti~mrinii disease Pneumogstis carinii organisms with serum antibodies occurs in institutions concerned with the care of ncw- [I 0, 22-24, 36431. Finally, the mode of transmission borns undcr the conditions of deprivation that occur of the parasite, its extrapulmonary morphology, and following wars thcn prophylactic measures may be of its metabolic requirements arc unknown. Accurate value and experience with such tcchniqurs should be data regarding these important deficits in our know- reviewed. ledge await successful in rit~oculture of the parasite. The widesprrad distribution of the organism, its To datr, the only evidence that the 'cysts' represent morphological characteristics, and the infrequent viable organisms is their constant presence and uni- emergence of the parasite as a pathogen has prompted form morphologic characteristics in the lungs of pa- scveral investigators to compare P~zetcinorystiscclrinii to tients and cxpcrimental animals with this disease. The Toso/~lnst~ro~isgondii. Dircct drmonstration, using dc- susceptibility of the organisms to pentamidinc [44-521 fined biochemical characteristics of the organisms or and thcir morphology [5, 11, 13, 23, 321 is consistent serological cross reactivity data, however, is not avail- with classification as a protozoan organism, but this able 1131. data only constitutes indirect evidence.

If we assume that the cysts found in the intra-alveolar Tissue S/jeciJci!jl of Pner~~)rogstiscarinii spaces arc nrither fungal in origin nor drgradation With one exception recorded to datc [34], the parasite products of injured tissucs [25-331, thcn further ana- has not been detected in other than pulmonary tissue. lysis of P~~etltnog~tiscarinii must await in ritro cultivation The primary lcsion initially is in the pulmonary alveo- of the organisms. To datc, this has not been accomplish- lar spaces. Later, cysts are found in lesser numbers in ed. An explanation for this failure may be found in the surrounding supporting structures such as the alveolar pathological data obtained in animals and humans. septa. Small numbers of organisms are detectable in With few exceptions [5, 341 most studies have em- the exudate of bronchioles and larger airways. These phasized that the cysts arc found primarily within the extra-alveolar organisms arc mostly extracellular or intra-alveolar spaccs.Thc few organisms that have been within phagocytic cells and are irtfrequet~tlyfound in the detccted within the pulmonary parenchyma have bcen airways. This remarkable specificity is found in both shown to be within phagocytic cells. It is of interest to natural and experimental infections of animals as well note that at autopsy the cysts are rarely detccted in the as humans [5, 7, 9-13, 30-32, 54-63]. Careful search hilar lymphatic tissue despite the protracted course of of other organs in heavily infected individuals has the disease. These findings would indicate that the failed to demonstrate these organisms in extra pulrno- metabolic requirements for this organism arc restricted nary ti~sucs.It is inferred from the report of PAVLICA to conditions prevailing in the environment of the al- [62] that the organisms may have penetrated the pla- veolar spaces. The nutritional and biochemical envi- centa. In this report, good evidence is presented to ronment of the alveolus, well studied by scveral work- show an in utero infection can occur with formation of ers [35], might serve as a guide for selection and pre- the characteristic adult pulmonary lcsion. There was paration of the nutrient media which would support no evidence in this case that placental infection was growth of the organism. the cause of the in utrm pulmonary infection. Of interest The disadvantages of not having the organism in was the absence of Pneun~ocyslisorganisms in other relatively pure form arc numerous. Serological ana- tissues of the infected fetus. lyses of Pneumocystis carinii infections necrssary for the This narrow range of tissue infectivity may explain study of epidemiology and attempts at prophylaxsis two aspects of the problem presented to the physician and investigator by Pneunlocystis carinii. The first is the cells in the pulmonary infiltrate in individuals with apparent dissociation of signs and symptoms exhibited immunological deficiency syndromes was also con- by affected individuals. Patients with intense cyanosis sistent with their state of depressed antibody biosyn- (arterial oxygen saturation values of 70 % or less) and thesis. Analysis of the sera of such individuals revealed x-rays indicating bilateral diffuse disease may have no antibody to Pneumocystis antigen 143, 641. To ac- normal temperature and normal peripheral leukocyte count for the reaction of different hosts, Pneumocystis counts as well as few complaints of illness. This absence carinii pneumonitis would seem to be the best designa- of constitutional symptoms or reliable physical signs tion for this disease. may be related to the absence in significant number of The gross pathological findings are similar when Pneumocy.rtis from sites other than the pulmonary al- taken from lung biopsy material, experimentakanimal veolar spaces. Lung necrosis, significant extra pulmon- models, and autopsy tissue. Both lungs are similarly ary fluid accumulation, and other evidence of inflam- involved. There is little or no visceral or somatic pleural mation or dysfunction of this organ have not been rc- reaction. The parenchyma is stiff, relatively avascular ported. and may show small local areas of emphysematous The inability to cultivate this organism in ~~itromay change. This hyperinflation is presumably the result be explained in part by the observed tissue specificity of lung rupture secondary to air trapping. There is no of Pneumocy~tis.It is possible to explain the inability lung necrosis and the fibrotic changes are usually most to grow Pneumocystis carinii outside of the susceptible intense in the perihilar area. Hilar lymphadenopathy host by assuming that the unique metabolic character- is usually not present. istics of the alveolar space have not been duplicated The intra-alveolar exudate containing the Pneumo- in uitro. In addition, since one of the predisposing char- cystis cysts seems to be the earliest lesion 17, 9-1 1, 56, acteristics of the susceptible host is a diminished capa- 59, 61, 631. The exudate contains some polymorpho- city for antibody synthesis, the antibody content of the nuclear leukocytes, but the predominant cell type is the serum supplement, representing pooled sera, of the histiocyte, lymphocyte and plasma cell. The intra-al- culture media may act as an inhibitor. One source of veolar exudate is quite distinctive. At low magnifica- tissue culture nutrition without antibodies, precolostral tion, conventional staining shows a lace-like appear- calf or pig sera, may be a useful reagent in studying ance to the round balls of eosinophilic material which techniques to grow Pneumocystis carinii. is also positive to periodic acid Schiff base. This latter finding is consistent with the presence of glycoproteins Mor/~hological in the exudate. Presumably, fibrinogen is a major com- In a discussion of the pathological lesion of Pneutno- ponent of the exudate but this has not been demonstrat- cystis carinii pneumonitis, three factors should be em- ed with immunofluorescent reagents. MARSHALL[50] phasized. The first is that the disease is due to a slowly reported that the serum concentration of IgM mole- growing organism 110, 11, 54-56] which affects both cules was elevated during the acute phase of the dis- lungs rather evenly although there may be a more ease. The carbohydrate content (7 to 10 %) of IgM 1651 intense reaction surrounding the hilar areas. The sec- may offer another source of Pneumocystis inflammation- ond is the remarkable tissue specificity of the parasite. induced molecules that might account for the positive Initially, the pathological lesion is predominately an stains for carbohydrate. intra-alveolar exudate provoked by the parasite. The The cysts observed withseveralstainssuch as Giemsa fibrosis and consolidation is probably a secondary reac- and Gram Weigert are best demonstrated by a silver tion to the intra-alveolar exudate. The third is the impregnation technique 12, 66, 671. They are found nature of the infiltrative cells induced by the parasite. in the intra-alveolar spaces and occasionally in the The absence of tissue necrosis and the slow develop- septa walls within paghocytic cells. The bronchiolar ment of the lesion are paralleled by the cell types ob- and large airways are usually clear although they may served. Mononuclear cells including histiocytes, lym- contain scanty exudate with few Pneumocystis cysts. The phocytes, and plasma cells are the most frequent in- detection of bronchiolar cysts and exudates is not a flammatory cells observed. In the cases with an under- constant finding [30, 31, 60, 631. lying immunological deficiency, plasma cells are not Most of the respiratory pathophysiology can be re- observed [2, 5, 32, 43, 58, 61, 63, 641. Much of the related to the lesion of the alveolar septa. Although confusion regarding the term 'interstitial plasma cell little experimental data is available, a likely explana- ~neumonia'has been attributed to the careful obser- tion for this finding may be gleaned from the growth vation of investigators studying the same disease pro- characteristics observed from clinical material and voking different cellular reactions in various individ- experimental animals. Pneumocystis carinii would seem uals. This problem was resolved when the heterogene- to be best described asan organismof low virulence. The ity of the hosts was considered. The absence of plasma full pathological lesion induced by this parasite devel- Pneumocystis carin!ii pneumonitis 139

Fig. 6.4. A H and E section (440 x ) shows the progres- tion at this same location may offer some explanation sion of thc morphological lesion. Thc intra-alveolar for the pulmonary dissemination of Pneumocystis [30, exudate provokes a fibrotic reaction in the septa. Ob- 601. The pneumonitis is induced primarily by the local litcration of the alvcolar spaces and the thickened al- proliferation of Pneumocystis 0rganisw.s. The irritation veolar septa combine to produce an inadequatc func- and inflammation that occurs results in coughing. The tioning lung. cysts are expelled into the larger airways and fall back B. Direct demonstration of the parasite at its intra- into other airways where they are swept into the alveoli. alveolar site by silver impregnation stain (Gomori). There, ncw inflammation occurs. The predominance in the dorsal regions of the lungs in younger infants, noted by some workers [60], is consistent with this ops very slowly. At least three to six weeks are neccs- mode of intra-pulmonary spread of Pneumocystis in sary for the disease to involve most of the lung paren- patients constantly in the supine position. chyma [4, 6, 11, 21, 32, 54, 551. Since it is almost ex- As the disease progresses, the fibrotic reaction results clusively an intra-alveolar parasite, the surrounding in the fusion of thesepta with the intra-alveolar exudate pulmonary alveolar septa would be expected to show and the obliteration of the alveoli. The progression the most tissue reaction and fibrosis. The inflammation of these lesions is shown in figures 6A and B. The of the alveolar septum and the inability of the parasite thickened alveolar septa as well as the relatively clear to invade other tissues presumably causes the charac- airways are also well illustrated in this figure. As one teristic metabolic alteration, alveolar-capillary block. would predict, the increasing obliteration of alveolar The characteristic slow growth of the organism prob- spaces is accompanied by increasing cyanosis and even- ably permits a continued fibrotic degeneration of the tually causes death due to extreme hypoxia [4, 5, 8, 30, septa1 walls without causing death to the host until the 32, 57, 631. lesion has involved most of the lungs. The mechanism of spread of the organism through- X-ray Findings out the lung parenchyma is unknown. However, the There is no distinctive roentgenological character- increased number of organisms seen at the hilar areas istic to identify the interstitial pneumonitis induced by as well as the predominance of the most fibro~icreac- Pneumocystis carinii. As shown in- the case material, similar x-ray findings are found in individuals with Demonstration of the Organism interstitial pncumonitis of infancy [4, 6, 24, 27, 57, 59, Direct demonstration of Pneunzocystis cysts is neces- 68,801 and pulmonary alveolar proteinosis [24,78-831. sary for a precise clinical diagnosis. Ancillary labora- The altered transmission of the x-ray is due to the tory data may define an immunological deficiency thickcncd iliterstitial tissue and obliteration of the al- syndrome. They include immunoglobulin levels in veolar spaces. The pathological process exited by this serum and peripheral lymphocyte count. An estima- pathogen involves most, if not all, the lung parenchy- tion of the altered respiratory physiology as assessed ma. An understanding of the pathophysiology of this by x-ray examinations and pulmonary function stu- disease should provide a high index of suspicion to dies, together with the history and physical examina- physicians confronted with the characteristic x-ray tion, constitute indirect cvidence relating to the dia- findings early in the disease. gnosis. Although several workers have reported that An explanation for the observation that the inflam- demonstration of the cysts is possible in sputum, ante- mation and fibrosis is more prominent in the dorsal mortenl [37, 85, 89-96] even a careful search usually portions of the lungs of prcmatures may be inferred fails to reveal the organism in most instances [51, 841. from the pathophysiology [59]. This disease pro- Identification of Pneunzocyslis carinii in sbutum early in gresses slowly and causes much debilitation. For a the course of the disease, however, has been accom- variable but prolonged period, the affected individual plished in several cases. There is as yet no data on the is lying on his back. The cysts that are expelled into the percent of positive sputum analyses in procen cases of larger airways would be expected to fall back and re- Pneunzoystis carinii. Some ideas of the frequency with enter the dorsal parmchyma. This finding should be which the organisms are found in the sputum may be looked for to confirm its reliability as a diagnostic lead. derived from the study of Txircs and JANSSENS who re- The interstitial scarring occurs bilaterally and most corded the number of cases in which Pneumocystis carinii of the lung parenchyma is aff'ccted. The perihilar areas could be identified within the lumen of the bronchioles appear to be more opague than the rest of the parcn- among 15 autopsies performed upon infants. These chyma. This radiological finding has been correlated workers found intra-bronchiolar cysts [I81 in one third with gross pathological studies and is due to additive (5115) of the cases. In two instances, the diagnosis was effects of the increased scarring in this area and to the achieved by lung biopsy after intense search for the opacity of the major vessels and airways. organisms in sputum using specific stains had failed. Rupture of the ventral lung parenchyma with re- Lung aspiration by needle biopsy has proven success- sultant pncumomcdiastinum and/or pneumothorax ful in identification of the disease in several instances and subcutaneous emphysema [20, 67, 76, 841 is an [97,98]. In each recorded instance, lung tissue remov- unusual complication detected late in the course of ed by this method revealed the lesion in proven cases the disease, and then mostly in neonates and infants. of Pneutno~ystiscarinii pneumonitis. This success is prob- Without an associated disease, there is little hyper- ably best explained by the diffuse nature of the disease. plasia of the pulmonary lymphoid tissue. Hilar lymph- Of interest is the report, kindly made available to us adenopathy is not a constant feature of this discase. [92], describing a fatal pneumothorax following needle This lack of reactivity may be due to the defective aspiration of a lung infected with Pneumocystis carinii. immune status of these patients or to the low invasive- Attempts at decompression of the hemipneumothorax ness and absence of extracellular products of the or- were unsuccessful. At autopsy, the stiff and fibrotic ganism. lung showed no tendency to close over the needle tract. In summary, the overall x-ray finding of Pneumo- Other reports have indicated that direct visualization cystis carinii is not directly distinguished from other of the pulmonary parenchyma at thoracotomy results diffusc pneumonias characterized by scarring of the in a lower morbidity and mortality than needle biopsy interstitial septa and obliteration of the alveolar spaces [99]. The gross pathological finding of a stiff, non- [84-88]. The process is diffuse and bilateral at the time collapsing parenchyma suggests that tissue should be that the symptomatology provokes medical attention. obtained under direct visualization to avoid this com- The growth charactrristic, low level of virulence of plication. No ante-mortenz demonstration of Pneumo- Pneutno~stiscarit~ii and debilitation of the affected indi- qstis carinii cysts outside of the respiratory tract has vidual, may account for the more advanced inflamma- been reported. tion in the ventral portions of the lungs. The compen- satory rmphysematous changes occasionally result in Host-parasite Relationsh+ Characteris~icof adventitious air in the mediastinum, pericardium, and Pneumocystis carinii pneumonitis subcutaneous tissues. X-rays, characteristic of the discase, are illustrated The case material has been divided into three groups in thc srction devoted to case material. in order to identify more clearly the most common fac- Pneumocystis carinii pneumonitis 141

tors reported to predispose to Pneumocystis carinii pneu- terial or experimental animals have failed to find the monitis. Those cases were chosen for review provided parasite in other tissues. These three instances of fetal with clinical, laboratory, and morphological data in infection should provide the impetus for designing an addition to a positive identification of the organism. experimental model to study this process of trans- Intrauterine infection. The first group, though com- mission. Such an experimental model would permit a prising only three cases, nevertheless is most important. more accurate evaluation of the development of the In these cases, infection undoubtedly was transmitted pathological lesion and the viability of the parasite in to the individual during intrauterine life [20, 62, 1001. the various body fluids. Typical pathological characteristics of Pneumocystis Host mithout demonstrable predisposing lesion. The next carinii infection were detected in a full term macerated group of patients accounts for most of the cases of fetus [62] and a two-day-old infant with Down's Syn- Pneumocystis carinii infection. The author has distin- drome [20]. In the third report, a ten day old infant guished two clinical conditions which are important died with diffuse pneumonitis with the typical patho- in understanding the appearance and distribution of logical findings of Pneumocystis carinii infection [loo]. the disease. Data derived from studies of experimental models and The first factor, contagion, is the most important extrapolation of clinical symptomatology, lung biop- etiological mechanism operative in this group. The sies, and pathological studies provided strong evidence mode of spread of Pneumocystis infections in 'normal' that the cyst does not occur as a postmortem conta- host is presumed to be due to exposure of an individual minant [4, 9-1 1, 21, 22, 30, 31, 36-39, 42, 54-57, 61, to a high concentration of the parasites in the air. Such 62, 81, 86, 91, 981. Further, the lesion requires at least an environment is readily attained in newborn nur- two to three weeks for the typical pathology to develop. series. This may result in the unusual distribution of Therefore, we must presume that the infection was cases within the institutions ofone city (table 111) [104]. acquired in utero. No reports indicate maternal inges- Epidemics such as the one illustrated in table 111 have tion of drugs which have been associated with a pre- accounted for the death and possible permanent dis- disposition to Pneutnocystis carinii infection or of expo- ability of thousands of children. Nurseries designed to sure to infected individuals. Unfortunately, neither care for abandoned or disabled children must contend serological analysis of the maternal serum nor other with Pneumocystis carinii infection as a significant cause evidence of maternal infection have been presented. of perinatal death (table IV) [20, 1001. Our under- The most plausible explanation for the instances of standing of the factors initiating such epidemics has neonatal disease is that there was an intrauterine in- been limited by the few reliable techniques for identi- fection due to bloodborne transmission of the parasite. fying the organism. Using a serological analysis, a high Although there are several references to the demonstra- incidence of positive reactions has been detected in tion of Pneumocystis carinii cysts at extrapulmonary sites, workers and nurseries [24, 3C39, 41, 1011. There is most morphological studies using human autopsy ma- yet no other reliable method of identifying a carrier

Table 111. The unequal distribution of cases of Pneumo- Table IV. The importance of Pneumocystis carinii pneu- cystis carinii pneumonitis in newborn and premature monitis as a cause of death of infants admitted to a infants in a city is shown. The frequency of cases in hospital in Iran is shown. 3 institutions as compared to the remainder of Pneumo- cystis infections was not related to the type or number Cause of death of 40 infants in foundling home and of patients at each institution. newborn nursery

Early hospitalization history of 120 infants with Condition No. of Pneunloqy~tiscarinii in Hamburg patients

* Nursery homes for newborns ..... 62 % Septic infections, chiefly bronchopneumonia 13 Small clinics for infants ...... 24 % Gastro-enteritis...... 10 Cared for at home ...... 9 % Pneumocystis carinii pneumonia ...... 6 Various institutions ...... 5 % Interstitial viral pneumonia ...... 4 * 3 out of 11 hospitals accounted for 85 % of these Measles with complications ...... 3 cases Congenital malformations ...... 2 - - Birth trauma ...... 2 After: VON HARNAK:Mschr. Kinderheilkunde 108: 159 ( 1960) After: POSTet al.: Arch. Dis. Childh. 39: 35 (1964)

10 Pediat. Rcs., Vol. 1, No. 2 (1967) state or subclinical infection since the parasite has been 105-107, 109-1 12, 114-1 17, 121, 1221. In addition, reliably identified only in the lungs or sputum of data from the following sources were also used to study severely infected hosts. Indirect evidence of airborne newborn host-parasite relationships [122-1351. Most transmission of the parasite as the major cause of spread of the individuals with Pneumocystis carinii pneumonitis has been inferred from the effectiveness of sterilization without apparent previous disease have been newborns of the air with ionizing radiation in affected hospitals and prernatures. Some of the characteristics of these [22, 95, 103, 1041. Although it has been difficult to cases are shown in table V. For many of the cases, con- prove, a plausible explanation for the start of these tagion was the most likely explanation for the induction evidemics is that the initial infections in these nurseries of Pneumocystis infection. A history of intimate contact originated in prematures with poor nutrition (so-called with infected children, usually within a hospital or 'dysplastic' infants). The term 'dysplastic' describes nursery, was present in many cases [8, 15, 16, 201. any debilitated premature infant regardless of etiology. There are, however, documented incidenccs of infants Many instances of Pneumocy~tiscarinii infection occur with Pneumocystis infection detected as sporadic, isolated sporadically in malnourished premature infants [8, 16, instances. In these cases, there was no known contact 17, 21-23, 69-71, 77, 83, 99, 105-1171. In these cases, with infants suffering with Pneumocystis infection. A the incompletely developed immunological response of more careful scrutiny of the clinical history in these full term newborns or prematures and exposure to in- patients arouse the suspicion that the term 'without fected infants in a closed environment have been as- predisposing disease' may be misleading. The clinical sumed to interact synergistically to produce diseaes. history of many of these newborn cases reveals that Many instances of so-called infectious diseases unique antimicrobial chemotherapy was administered at least to newborn immunological immaturity have under- three weeks before the development of clinical symp- gone reclassification as a result of additional clinical toms of Pneumocystis infection (table V). Three weeks evidence. The predominance of severe and fatal in- was chosen as a critical period to identify these individ- fection with saprophytic bacteria has, in many in- uals because this period seems to be the minimal in- stances, been shown to have been due to an unusual cubation time for the clinical expression of Pneumo- exposure of prematures and newborns to these organ- cystis disease. Antimicrobial chemotherapy, per se, has isms of low virulence because of contaminated ventila- been shown to be sufficient experimental manipulation tion systems [118]. In the future, quantitative techni- to induce Pneumocystis infection in a small percentage ques for the demonstration of Pneumocystis carinii in in newborn and adult experimental animals [9-11,54, other than affected lungs may permit a more compre- 561. A similar effect of antibiotics in human hosts may hensive understanding of factors important in con- also occur. It is significant to note that many of these tagious spread of the disease in nurseries. In summary, cases were treated with antibiotics for respiratory in- the presence of one or more 'dystrophic' prematures fections [8, 1141. Another factor to consider whether with Pneumogds carinii infection has been postulated these newborns were indeed 'normal' may be extracted to be sufficient to initiate an epidemic in a nursery. from other clinical observations. The association of Factors sufficient for contagion, such as a high con- pulmonary infectious disease predisposing to Pneumo- centration of airborne parasites, may then result in a cystis carinii pneumonitis was first observed by CHAGAS spread of the infection to normal prematures and full [135]. Since then, several infectious diseases have been term newborns. associated with Pneumocystis infection [17-19, 11 31. The ' Of interest are recent reports of the contagious na- pathogenesis of this predisposition to Pneumocystis in- ture of the disease in normal full-term newborns and fection is not clear. One possibility is that chronic in- young adults [I 19-1211. Thus, with a certain level of jury to the pulmonary parenchyma may be sufficient exposure, an otherwise healthy newborn or adult may to permit suprainfection with ~neumocystis.The clinical contract the disease. This observation has more than symptoms that provoked care for these preceding in- theoretical importance because it should influence the fections were not typical of Pneumocystis pneumonitis judgment of physicians responsible for the care of im- and have been presumed, for the purpose of this report, mediate contacts of patients with Pneumocystis carinii to be different illnesses. pneumonitis. In summary, a critical evaluation of the host-parasite interaction operative in the induction of Pneumocystis carinii pneumonitis in newborns and prematures reveal Newborn and Premature Infants that many individuals are not truly 'normal'. Infected individuals are highly contagious for other infants in The followina- cases were analvzed because sufficient institutions or closed environments and may serve as a pathologic, clinical, and laboratory data were included reservoir of infection for otherwise healthy newborns. in the reports [3, 8, 15, 16, 20, 62, 69-71, 77-79, 99, Pre-existing respiratory infection treated with anti- Pneumocystis cart Inii pneumonitis 143 microbial agents may be sufficient to permit the latent Table V. Clinical characteristics of newborn and pre- Pneunzorystis infection to become life threatening. Ana- mature infants with histologically verified Ptzeumocystis lysis of sporadic cases of Pneumocystis infection in new- carinii pneumonitis are reviewed. borns reveal that many of these 'normal' individuals have received antimicrobial therapy for respiratory Clinical characteristics of newborn and premature infections at least three weeks prior to detection of their group with Pneumocystis carinii pneumonitis (74 cases) clinical disease. The possibility of these two factors, antibiotics plus a significant respiratory infection, as Age at death - 4.2 months (8 days - 18 months) sufficient conditions to activate a latent Pneumocystis Age at onset - 3.4 months raritzii infection in a few newborn and prematures, Duration of illness - 3-5 weeks should be considered. Previous disease - 28 Of interest is the occurrence of Ptzeutrzocystis carinii Previous treatment with antibiotics - 23 pneumonitis in adults without obvious predisposing Presenting symptoms (38 descriptions) disease or a significant history of chemotherapy pre- 1. dyspnea 36 crding the illness (table VI). There are four such re- 2. anorexia 2 3 ported cases in the literature [60, 120, 136, 1371. In 3. cough 20 the case reported by WANATABE[120], contagion 4. tachypnea 18 would seem the most likely causc of thc infection in the 5. cyanosis 18 so-called 'normal' adult. Perhaps a more careful search Clinical signs (38 descriptions) among pathologic material of adults who have died 1. dyspnea 36 of obscure lung disease may reveal that occurrence of 2. cyanosis 34 Ptzeun~orysti.rinfection is not as rare as formerly thought 3. tachypnea 34 in so-called 'normal' adults. 4. cough 10 5. fever 8 Itlfectious Diseases 6. rales 8 Several infectious discases have been associated with Laboratory data a high incidence of concomitant Ptzeurnocysfis carinii 1. WBC - > 10,000-10 infection. The most prevalent infectious disease asso- normal-2 ciated with Pneunzocystis carinii pneumonitis has been < 5,000-2 cytomegalic inclusion disease [5, 15, 69, 89, 99, 122, 2. X-ray - diffuse, bilateral infiltrates 25/25 129, 13%145]. In many cases, congenital cytomegalic adventitious air 4/25 inclusion diseasc has been the predisposing disease for Pneunzocystis caritzii infection in newborns. The diffuse nature of the cytomegalic virus infection and its ac- Table VI. The similarities between the clinical disease companying multi-system dysfunction result in a chro- of Pneurnog~stiscarinii pneumonitis in individuals with nically ill individual. This type of debilitation seems to primary immunological deficiency syndromes and be sufficient to induce the clinical expression of I'neu- 'normal' patients is shown. Note the high incidence of tnocystis. Cytomegalic inclusion disease is another preceding antibiotic treatment in this group. The time example of a latent infection which may be activated course of the progression of the disease in these individ- in the adult with a lymphoreticular malignancy and/or uals was similar to other groups of patients. chemotherapy. Cytotoxic chemotherapy administered for other therapeutic indications is also associated with Analysis of clinical characteristics of Pneumocystis carinii a high incidcnce of Ptzeurnocystis infection. Of interest is pneumonitis in paticnts with lo immunological defi- the occurrence of these two diseases in adults with ciency syndromes (31 cases) lymphoreticular malignancies and adults receiving immunosuppressive chemotherapy for renal trans- Age of patients 3 months-7 years plantation [140]. Congenital and disseminated tuber- Number receiving chemotherapy 18/22 cases culosis, pulmonary cryptococcus and some bacterial antibiotics 17/22 cases pneumonias havc also been reported to predispose to corticosteroids 3/22 cases Ptzeunzocystis infection [128, 145, 1461. As will be em- Duration of symptoms 3-6 weeks phasized further, a diffuse pulmonary disease in a patient with a pre-existing infection should be sufficient multi-organ systems, can predispose to Pneumocystis. grounds to excite the investigative energies of physi- Administration of cytotoxic, corticosteroid, and/or an- cians to search for Pneunzogstis infection. It would seem timicrobial agents to such patients increase the risk of that any chronic- infections, especially those involving activation of this latent infection.

10. Pediat. Res., Vol. 1, So. 2 (1967) Immunologic DeJiciency Stutes for prolonged periods. This should be a consideration Primary immunologic dejciency state^. That immunolog- in deciding whether a patirnt with an immunologic ical deficiency states could predispose to Pneumocystis deficiency syndrome should receive prophylactic anti- carinii infection was first reported by BURKEet al. [64]. microbial chemotherapy or replacement therapy with The following cases [48,50,5 1,53,64, 78,82, 109-1 11, gamma globulin. Parenteral administration of gamma 132, 134, 149-1551 are cited to study the clinical globulin, with its attendant difficulties, would seem to characteristics and host interaction mechanisms in be a therapeutic approach least likely to permit emer- patients with primary immunologic deficiency syn- gence of the latent Pneumocystis infcction. dromes with Pneumocy~tisinfection. Primary immuno- Another interesting point to be considered in dis- logic deficiency syndromes are defined as thosr states cussing patients with primary immunologic deficiency of clinically or laboratory defined immunologic defi- syndromes and Pneunfocystis infection is the possibility ciency which are unassociated with the administration that the true incidence of congenital immunologic de- of cytotoxic agents or are consequent to a known dis- ficiency states is masked by supervening Pneumocystis ease process. infection. Survey of the literature of patients in the Pneumocystis carinii pneumonia occurs in individuals newborn and young infant age group who have been with primary immunologic deficiencies. This associ- reported to have Przeutnocystis infection reveals that ation, first reported by BURKEel al., has been now many of them may have had immunologic deficiency well documented in the literature [64]. It would syndromes [70, 78, 98, 109, 110, 149, 1511. Autopsy seem that immunologic deficiency syndromes of all findings such as diminished lymph nodr tissue, absent typrs reported are associated with Pneumocystis carinii thymus, and low serum immunoglobulins arouse the infcction. Several cases of Pneumocystis infection re- suspicion that many so-called 'normal' newborn and ported as occurring in 'normal' individuals may have prematures or isolated cases of the sporadic variety may been, in fact, individuals with hypogammaglobulin- have been, indeed, examples of Ptzeumocyrtis infections emia [70, 1091. in patients with immunologic deficiency syndromes. The association between primary immunologic de- The finding of Pneumocystis infection at autopsy should ficiency syndromes and the absence of plasma cells in drive the pathologist and physician to seek evidcnce of the pulmonary infiltrates of' some patients with Ptleu- immunologic deficiency syndrome. Such evidcnce, of mocystis infection is well known 1431. The difficulty in course, has important considerations in genetic coun- detecting serum antibody to the parasites in patients seling as well as other items of interest for the investi- initially studied in this country with Pneumocystis infec- gator. tion and the absence ofplasma cells in autopsy material Secondary immunological deficiency states predispo~ing to of these patientswere directly correlated as the inability Pneumocystis cariniipneutnonitis. In this category, clinical of these individuals to produce antibodies to other in- and pathological data of patients with altered immune fectious agents. In these cases, the morphological lesion reactivity secondary to therapeutic agents are analyzed with its alveolar scarring is quite comparable to the [17, 33, 72, 75, 78, 80, 89, 96, 97, 139-141, 144, 146, histology observed in tissues from individuals with 150, 157-1741. Patients with lymphoreticular malig-

'normal' immunological function with the exception nancies and/or, patients- who have been treated with of the absence of plasma crlls in the infiltrate (see case immunosuppressive drugs and later developed Pneu- material). rnocystis carinii pneumonitis constitute the majority of Pneumocystis carinii pneumonitis and immunological cases. Other patients with a significant previous dis- deficiency can occur at any age during the coursc of a ease prompting immunosuppressive chemotherapy primary immunologic deficiency disease and is asso- prior to the development of Pneumocystis carinii infection ciated with deficiency states of all types [144]. are also included. In all these cases. there is a consistant It should be emphasized that a history of antimicro- pattern of predisposing disease and emergence of the bial chemotherapy can be found in the majority of Pneumocystis organism as a pathogen. As can be seen in patients with Pneumocystis infection and immunologic table VII, the list of chronic diseases that have termi- deficiencies (table VI). It is not clear if this is a direct nated with Pneutnocvs~iscnrinii infection is inclusive of association since one would expect that such individ- much of medical pathology. Not included in this list uals would be treated with antimicrobial agents quite are two cases of cyclic neutropenia in two young chil- often. Nevertheless, such an association is emphasized dren in which the diagnosis of Pneumocystis infection for two reasons: 1. Patients with chronic disease and was ascertained by lung biopsy and post mortem ana- induced secondary immunologic deficiencies also have lysis [172]. It is predictable that examples of most a similar association with Pneumocystis infection. 2. Ex- chronic diseases will be shown to have been associated perimentally, the disease may be induced in a small with an altered host reactivity so as to permit infection fraction of some animal species treated with antibiotics with Pneumocystis. The one characteristic common to Pneurnocystis carinii pneumonitis 145

these cases of Ptieutnocystis infection is that most. if not Table VZI. Clinical Characteristics of patients with all. of thc subjects received immunosuppressive and/or Pneuniocystis carinii pneumonitis and 2' immunological antimicrobial chemotherapy during the course of their Deficiency (68 cases) (ages: 3 months-78 years) illnrss prior to clinical detection of Pneumoc_l~stisinfec- tion . Of interest is the apparent high frequency of Etiology of 2" immunological deficiency cortisone and cyclophosphamide treatments and Pneu- Disease No. of tnoq~.\tisinfrction . Cyclophosphamide and cortisone arr patients particularly efficient in inducing Pneumocystis infection in rats [54.175] . Therelatively low toxicity ofthisformcr Rheumatoid arthritis ...... 1 drug. ascompared to other antimetabolites. may permit Alber's Shonberg ...... 1 longer survival of humans and animals and. thercfore. Hemophilia ...... 2 inducc a high incidence of Pneutnocystis infection . Bilateral nephrectomy ...... 2 Of 68 cases studied in thc literature. corticosteroids Wegener's granulomatosis ...... 1 were mcntioned as therapeutic agents in 53 patients. Thrombotic thrombocytopenia purpura . 1 The disease that prompted the corticosteroid therapy Pulmonary cryptococcosis ...... 1 rangcd from overzealous replacement therapy in adre- Hodgkin's disease ...... 5 nal cortico-virilizing syndrome to Hodgkin's disease . Henoch Scholein's disease ...... 1 The former had not been noted to be coincident with Rheumatic fever ...... 1 an immunological deficiency state when replacement Nephrosis ...... 2 thcrapy was accurately gauged to meet the patient's Hand-Schuller-Christen disease .... 1 requiremrnts although acute infectious disease may be Chronic lymphatic leukemia ...... 9 poorly tolerated . Hodgkin's disease and multiple mye- Hemoly tic anemia ...... 2 loma. however. have been shown to depress the de- Stem cell leukemia ...... 8 laycd skin reactivity and antibody formation respec- Lymphosarcoma ...... 3 tivcly to various antigens [173] . Although several in- Reticulum cell sarcoma ...... 1 vestigators havr searched autopsy material. document- Acute lymphocytic leukemia ...... 1 ed Prieutnocyslis carinii infection in patients with Hodg- Multiple myeloma ...... 3 kin's disease has not been identified prior to the use Chronic peritonitis ...... 1 of chemotherapy. both cytotoxic and/or antimicrobial Chronic sinusitis ...... 1 agents. although the disease has been a significant fac- Tuberculosis ...... 2 tor as a causc of newborn morbidity for at least fifty Congenital aplastic anemia ...... 2 years [I261. It would srcm. thrrrforr. that both an Eczema ...... 2 altered reactivity due to the lymphorcticular malig- Hypoglycemia ...... 1 nancy and the immunosuppressive and/or antibiotic Vitamin D intoxification ...... 1 agcnts are nccrssary for the activation of Pneumocy~ti~ Letter-Siwe ...... 1 with high frequency . Some idca of the incidence of Neuroblastoma ...... 1 Pneumoystis infection as a complication of lympho- Monocytic leukemia ...... 1 reticular malignancy may be surmised from the report Acute myelogenous leukemia ..... 1 of ROBINSONel al . [136] . Of two hundred and three Chronic myelogcnous leukemia ..... 3 autopsies of lymphorcticular malignancy examined. Waldenstrom's macroglobulinemia ... 1 two cases of Pneumocysti~infection wcrr discovered . One Adreno-genital syndrome ...... 1 would expect. that as supportive and anti-tumor meas- Follicular lymphoma ...... 1 ures increase in their therapeutic index. the incidence Thrombocytopenia ...... 1 of latent infections. such as Pneumo~slisrarinii. will Cytomegalic inclusion disease ..... 1 increase. A similar study (table VIII) rcveals a slightly higher incidence of Pncutnorystis infection in patients Irnmunosupprcssive agent with lymphoreticular malignancies . This higher in- Drug No . of cidence may reflect a rcsult of more eficicnt anti-tumor cases therapy with a longer survival . In contrast. ULTMANN receiving et al., reviewing clinical and pathological data of 60 autopsied casrs of lymphatic leukemia failed to detect Corticosteroids ...... 53 this diseasc 11741. Other factors influencing host re- Cytotoxic agents ...... 47 sistance to Ptieumoc_l~stismay not be evident as yet . Antibiotics ...... 28 Immunosuppressive chemotherapy. Per se. is suffi- X-irradiation ...... 4 cient to induce Pneumory~tiscarinii infection . Examina- Colloidal gold ...... 2 Table VIII. Incidence of Pneumocvstis carinii pneumoni- Thorough analysis of the respiratory deficiency in tis in patients with lymphoreticular malignancies Pneumocystis carinii pneumonitis is not available. A unique case, well studied, has been reported and the Total No. of Type of malignancy No. data is summarized in table X [60]. The subject was autopsies cases an adult male without a history of having received immunosupprcssive or antimicrobial chemotherapy. 323 12 Stem cell leukemia 6 Analysis of many tissues at the autopsy did not reveal Hodgkin's disease 2 other pathological processes. At the time of the physio- lymphosarcoma 2 logic studies, the x-rays showed a diffuse bilateral in- reticulum cell sarcoma 1 flammation characteristic of the interstitial pneumo- chronic lymphatic leukemia 1 nitis seen in Pneumocystis infection. The altered physiol- ogy is consistent with the syndrome called 'alveolar- After: ESTERLY,J.A. et 01.: Arch.Path. 80: 433 (1965) capillary block'. In this condition, the lungs are stiff (169) and noncompliant, but there is relatively little obstruc- tion to the flow of alr through the major airways. The tion of patients receiving renal allografts and receiving result of this altered physiology is that there is relatively concomitant immunosuppressive therapy reveal that little impairment of the excretion of the relatively sol- Pneumocystis infection comprises a significant cause of uble carbon dioxide which diffuses across the alveolar morbitity and mortality (table IX) [39,89]. Associated septa. In contrast, oxygen is only slightly soluble in the with Pneumocystis infection, as in the newborn and pre- pulmonary solvents and is poorly transported across mature group, is cytomegalic inclusion disease. In such the inflamed and thickened alveolar septa. Increas- cases, it has been possible to recover Pneumocystis cysts ing the oxygen tension of the inspired air increases the in a significant fraction of sputum samples of affected oxygenation of pulmonary capillary blood because patients [89]. there is relatively little obstructive element in this dis- Since a chemotherapeutic approach to this disease ease. Since there is no obstruction to the movement of has been shown to be successful in a patient with lym- gases in the large airways, there is relatively unimpair- phoreticular malignancy receiving cytotoxic agents, ed excretion of most of the venous carbon dioxide. diagnostic measures to demonstrate the lesion should Thus, there is effective regulation of normal serum pH made [17 11 (see case 3). until late in the course of the disease. Increasing the Of interest is the association of gold therapy and oxygen content of the inspired air has a marked clinical Pneumocy.rtis infection [145]. Recently, it has been re- effect as evidenced by a lessening of the cyanosis. Part ported that colloidal gold given for rheumatoid arthri- of the dissociation of the clinical symptoms (cyanosis tis may result in a deficiency of serum immunoglobulin and tachypnea) from the severe pulmonary findings formation [169]. It is probable that any drug with a exhibited by these patients may be due to the relatively depressant effect upon immunological function may unimpaired pulmonary regulation of the serum incite latent infections. This relationship should be pH. sought in the investigation of interstitial pneumonitis Similar but incomplete pulmonary function studies of obscure origin in patients receiving any chemothera- of infants with Ptzeumocystis carinii pneumonitis has been py known to have a depressant action upon host re- documented (table XI) [176]. A survey of the serum sistance. sodium, pH, and chloride levels induced by Pneumo- cystis disease with severe hypoxemia have shown the Clinical and Laboratorv Findings relatively normal PCO, and pH in the presence of cyanosis and severe arterial oxygen desaturation [177, As discussed in the previous section, the direct demon- 1781. Further analysis of serum chloride levels and pH stration of Pneumorvstis cysts in the sputum is quite as well as PCO, in various reports, which have been difficult. Obtaining a positive analysis of Pneumocystis used as a crude analysis of serum acid base regulation, in sputum is more infrequent in the younger age group, are consistent with the hypothesis that the distorted perhaps, reflecting the inability to obtain a proper pulmonary function induced by Pneun~ocystisis an alveo- specimen. Analysis of all the case material used in this lar-capillary block and is present in individuals of all text indicates that the most consistant pathophysiologic age groups [69, 75, 89, 97, 11 1, 121, 1401. This cor- state demonstrable in Pneumocystis carinii pneumonitis relates well with the primary pathologic lesion, that is the abnormal pulmonary respiratory physiology. of a thickened and scarred alveolar septa, character- The alteration in the pulmonary respiration is most istic of Pneumocystis carinii pneumonitis. Other diseases effectively described as being consistent with 'an al- that have alveolar-capillary block also have a similar \seolar-capillary block' syndrome. pathological lesion. Pneutnocystis carinii pneumonitis 147

Tnble IX. The significance of Pneutr,orysfis carinii infec- Table X. Analysis of the altered pulmonary physiol- tion as a cause of mortality and morbidity in patients ogy in an adult patient with Pneumoryslis carinii verified receiving imrnunosuppressivc chemotherapy following at post mortem examination is summarized. The ex- renal transplants is shown. treme degree of hypoxemia with relatively normal pCO, and pH is characteristic of impaired diffusion of Infectious pulmonary disease in patients receiving inspired air across alveolar septa1 membranes (alveo- immunosuppressive therapy for kidney transplantation lar-capillary block).

Transplant Etiologic agent No. of Pulmonary function studies of an adult male with ad- cases vanced Pneumocystis carinii pneumonitis

Kidney cytomegalovirus 1. Pulmonary compliance 0.02 I/cm H,O (normal pscudomonas aeruginosa 0.02 I/cm H,O) candida albicans 2. Nonelastic resistance 2.63 l/cm H,O,'second Ptvurnor~~~tiscclritcii (this is normal value) $taphylococcus aurcus 3. Tidal volume 435 ml (BTPS) asprrgillus 4. Respiratory rate 42/minute diplococcus pnrumoniae 5. Arterial 0, saturation 57.5 0/, (measured during norcardia asteroides IPP with 40 % 0, in inspired air) E. coli 6. Arterial CO, tension 26 mm Hg (normal 40 mm hl. tuberculosis Hg) Total 7. :herial pH 7.59 (normal 7.40)

- -- After: HILL,R. B. : New Engl. J. Med. 271: 102 1 (1964) After: LYONSet (I!.: jirch.intern. Med. 108: 173 (1961)

Table XI. Metabolic studies, conducted on normal newborns and infants with respiratory disease confirm the findings of al\~colar-capillaryblock syndrome. Patients with Pneumoystis carinii pneumonitis exhibit low arterial oxygrn saturation with relatively normal plasma pH values.

Metabolic studies of infants with Pneumoystis cnritrii pneumonitis

Diagnosis KO. of PCOZ Plasma Arterial cases (mm Hg) PH 0, Sat.

01 ,0

3-month malnourished prematures . 6 37.5 7.4 95 10.1 +0.02 12.38 Normal prcmatures ...... 12 45 7.25 89 k2.55 10.09 18.84 Anoxia newborn prematures . . . . 16 75.1 6.82 - (46-1 00) (6.45-7.25) Ptreuntoc~jcti~carinii ...... 7 55.2 7.25 29.5 (27-80) (7.10-7.40) (20.0-42 .O)

After: KERPEL-FRONIUS,E. et ill.: Arch. Dis. Childh. 39: 473 (1964)

Thr complication rcported by LEBEDA(731 is prob- :\nalysis of all cases, divided into newborn and ably relatcd to the marked increase in 'work' of premature groups, and those individuals with primary respiration, characteristic of the syndrome. 'Spontane- or secondary immunological deficiencies, has not re- ous' rib fractures may occur more frequently than this vealcd any consistent pattern of change of the periph- single report would indicate. Pain from such a fracture eral leukocyte count or distribution of cell types that could conceivably further compromise the depressed would aid in the diagnosis of Pneutnocystis infection. respiration and should be considered as a complication Fever is irregularly present. Immunoglobulin deter- to be watched for in infants. minations, absolute lymphocyte count, and immuno- logical reactivity towards antigens that are tested by lesion seen in humans [I I]. A careful search of experi- serum antibody or delayed-type hypersensitivity arc mental animals has failed to demonstrate cysts at any helpful in defining an immunological deficiency. How- other site other than the pulmonaryparenchyma.Thus, ever, the disease has been reported in all types of im- the nature of the induction in the pathologic process munological deficiencies described to date. Thus, there and the response to chemotherapeutic agents parallel is no good clue about immunity to Pneumocystis carinii these features observed in humans. Both animal and that can be extracted from these cases since both de- human data are consistent with the concept that Pneu- pression of serum antibody biosynthesis and 'cellular' mocystis carinii pneumonitis is a disease caused by an immunity may predispose the host to Pneumocystis ca- ubiquitous organism of low virulence and narrow rinii pneumonitis. Apparent eradication of the parasite tissue specificity in a susceptible host. by chcmothcrapy with pentamidine isothionate in one Of interest is the inability to show serological reac- patient has been followed by a recrudescence of in- tivity with cysts extracted from animal lungs as com- fcction with Pneumocystis [142]. A similiar finding has pared to those extracted from human material [42]. been reported in experimental animals [54]. One of the limitations of the serological test, however, is the nature of the antigenic material. The antigenic material is a crude extract taken from whole tissue. Cross species reactions with pulmonary tissue must be removed by absorption of either species serum with Pt2eurnocystis infection can be induced in a variety of the heterologous lung tissue. Such absorption proce- experimental animals, as well as occurring as a so- dures result in a loss of activity of one species antibody called 'natural' infection (tables I1 and XII) [7,9-14, to the antigen extracted from the other species lung. 42, 54, 55, 1791. Induction of the infection in normal It is entirely possible that thc extensive absorption animals in various experimental models has proven procedure itself may destroy the effectiveness of the to be a valuable adjunct to clinical and pathological test. It cannot be stressed too strongly that a source of data in humans because it has provided a greater pure antigen is needed to understand whether there understanding of the pathogenesis of the disease as is a species specificity in Pneumocystis carinii. The im- well as the useful experimental tool to study the effects portance of this observation is critical in an analysis of chemotherapeutic agents. In brief, the experimental of sources of organism in nature. If species specificity results may be summarized as follows. Pneumocystis in- does exist, then a search for this organism in the ex- fection can be induced in animals with a variety of perimental models already studied indicates that these chemotherapeutic agents without specific exposure to animals do not constitute a significant reservoir of in- a known source of Pneumocystis carinii. Prolonged treat- fection for humans. ment of normal animals. with immunosuppressive Thus, in the experimental model, Pneui~zocystiscarinii drugs such as corticosteroids alone, or in conjunction can be shown to be a latent infection. The cysts are with antimicrobial agents, is sufficient to result in present in experimental animals in too few number to Pneumocystis carinii pncumonitis. Two agents with be demonstrated by ordinary techniques. Suppres- different inhibitory activities towards host defense sion of the normal host's resistance by corticostcroids, mechanisms will induce a significant infection in most antimetabolites, or antimicrobial chemotherapeutic animals. Nonfatal infection as evidenced by the de- agents activates a Peeumocystis infection. The lesion is monstration of recovery and/or the presence of cysts confined to the pulmonary parenchyma and the pa- without severe pneumonitis in a significant proportion thologic process is very similar to that seen in huma- of animals may follow the use of one agent alone [I 1, ns. 561. Suckling, newborn, and adult animals are sus- ceptible to the induced induction of Pneumocystis infec- tion by these agents. The relative ease with which the Chemotherapy disease can be induced in the rat is shown in table XIII. Specie differences as well as different experimental Many attempts, using chemotherapeutic agents, designs used to study Pneumocystis carinii infection have been made to treat Pneumocystir carinii pneumoni- are shown in table XIV. tis [46-53, 92, 123, 124, 130, 142, 149, 171, 180-1831. The pathologic lesion induced by such manipulation The most successful among these has been the intro- has been shown to parallel very closely those seen in duction of pentamidine isothionate by IVADY[46,47]. humans. Illustrations (figs. 7A and B), kindly supplied Since introduction of the anti-trypanosome drug, by Dr. Walter Sheldon at the Johns Hopkins Univer- many other investigators have reported its usefulness sity, shows that thickened interstitial septa and intra both in human and experimental animals [54]. One alveolar exudate characteristic of the morphological of the considerations in evaluating such an experimen- Pneumocystis carinii pneumonitis 149

Tnble XII. The experimental manipulations necessary to induce Pneurnocyslis carinii pneumonitis in the rabbit are summarized. Note that human cysts, taken at autopsy do not serve as a good source of infecting material. Antimicrobials plus cortisone are sufficient to induce a Pneumocyslis infection without direct exposure of the animals to the cysts.

Induction of /~tteunlorystiscarinii pneumonitis in rabbits

Treatment Source Pathological findings of cysts Cysts present Pneumonitis No. detected Intensity No. detected Intensity

Cortisone ...... I Penicillin ...... I Streptomycin ...... 717 2+-4+ 717 3+-4+ Przeumoyslis ...... I human Penicillin ...... Streptomycin ...... I human 013 0 0,/3 0 Ptteun~ogsli~...... I Cortisone ...... Penicillin ...... rabbit 6 16 4f 616 4 + Streptomycin ...... I PIIPUI~IO~JI~~...... Penicillin ...... Streptomycin ...... I rabbit 616 2+ 616 3 + Pt~eurnotg~lis...... I Penicillin ...... Streptomycin ...1 none 414 2+-4+ 414 2-1--4$ Cortisone ...... I Nonc ...... none 10/13 1-t 7/13 1 i

After: SHELDON,\Y. : J. cxp. Mcd. 110: 147 (1959)

Table 'YIII. The relative ease with which Pneutrlocystis Table XIV. Experimental manipulations used to induce ccrritlii pneumonitis can be induced without an exogen- Pneu~nocysliscarinii pneumonitis in experimental ani- ous source of infectious material is shown. mals

Frrqucncy of Pneutno~stiscarinii pneumonitis in normal Species Manipulation Reference adult rats treated with cortisone (4.0 mg1100 gm body Rats cortisone RICKEN,D. : Virchows weight for 40 days) Arch. path. Anat. 331: Treatment No. of No. with 713 (1958) animals Ptzeun~ogsliscnrinii Rabbits antibiotics and SHELDON,J. : Exp. Med. infection cortisone 110: 147 (1959) Mice cortisone and BASTAMANTEel nl.: 1. Cortisone 40 28 cysts hIschr. Kinderheilk. 2. Cortisonr 103: 145 (1960) -1- Ptzeutnog~~lis 54 33 Rats cortisone, cyclo- FRENKEL,J. et 01.: 3. Ptleun~og~sti~ 14 33 phosphamide, Lab. Invest. 15: 1559 4. None cortisone and (1966) antimetabolites After: WELLER,R. : Z. Kinderheilk. 78: 166 (1956) X-ray irradiation (56) Fig.7 A and B. This photomicrograph, kindly made tal drug is the accurate determination of the clinical available to us through the courtesy of Dr. Walter course in this disease. This has been a difficult task. Sheldon, Department of Pathology, Johns Hopkins Thc two reports listed [103, 1841 would indicate that University College of Medicine, shows the morpholog- in some hands, recovery from the disease is a usual ical lesion of induced Pneutnocystis rarinii in the rabbit. outcome; whereas, in others, the disease seems to be The fibrosis of the alveolar interstitial tissue, the intra- almost universally fatal. The difficulty in assessing the alveolar exudate and the rclatively clear bronchioles outcome of a P~~eumocystiscarinii infection is compound- closely mimick the pathological process observed in ed by the nature of the previous dispossessing disease the human. and the difficulty in truly establishing whether or not the pneumonitis has resulted from a P~zeut~roq~stisin- Table XV. The effcctivencss of anti-P~reumocystischcmo- fection. therapy with pentamidine isothionate ovrr a four-year It can be seen from the illustrative case material at period is shown. The dramatic fall in the percent of the beginning of this report that many diseases have fatalities in affected prematures is consistant with the the typical clinical and x-ray appearance of Pneumo- few successfully treated cases in adults reported to date. cystis infection. Accurate diagnosis of this disease has Effectiveness of pentamidine isothionate in treatment been difficult because of the inability to positively of Pneumocystis raritrii pneumonitis in prematures identify the organism outside of lung tissue. Thus, evaluation of the effectiveness of drug therapy has been Treatmer~t Year No. of No. of Fatalities difficult. Nevertheless, in instances where the diagnosis cases deaths ( O/b) has been accurately established by pathologic material or demonstration of the cysts in the sputum, treatment None 1960 9 6 66 of the individual with the drug has proven to be an Pentamidinc 1961-64 44 4 9 effective therapeutic regime. A summary of a long ex- perience with the drug, pentamidine isothionate After: LORINCZI,K. et al.: Gyermekgyogyaszat 15: 270 (table XV), shows that the use of this drug would seem ( 1964) to be an effective therapeutic tool. Pneumocystis carinii pneumonitis 151

The drug, now available in the United States1, has ever, the high risk of thrombocytopenia due to the bern shown to be an effective agent for the prophylactic pyramethamine would seem to make pentamidine the treatment of African trypanosomasis [185]. A toxic current drug of choice. side effect noted in a case studied by the author [51] and later substantiated in humans and animals has been the appearance of an i~lcreascin megaloblasts in Cburse arid Proxnosi~ the bone marrow. These changes are consistent with a folic acid deficiency. Analysis of the serum of folic acid An analysis of the natural course of Ptzeu~rzocysfiscaritrii in our patient receiving this drug showed a normal infection is hindcrcd by several factors. The first, pre- level without displacemrnt of the bound folic acid. viously mrntioned, is the difficulty inarriving at precise Thereforr, our tentative conclusion is that pentamidine diagnosis of this disease. Direct demonstrationof Ptzcu- isothionate acts as a direct inhibitor of folic acid similar rtlocvstiscysts in sputum of affected patients is infrequent to that induced by aminopterin. In animal work, and obtaining tissue samples in a sick individual pre- Frankel reported that pretreatment with folinic acid sents a problem with a calculated risk. There is no did not disturb thc thcrapeutic effectiveness of penta- information which correlates the severity of clinical midine although it did prevent the mcgaloblastosis in symptoms and the frequency of positive identification thr rat [54]. While it iz still too early to conclude that of the cysts. folic acid is a useful adjunct to therapy with pentami- The second problem confronting the investigator is dine, it would srcm that this might be a beneficial the similarity between the symptoms and signs of proccdurc. The usefulness of the folic acid or its deri- Ptzeurtlog~.sfiscnrirtii pneumonitis and other diseases vativcs for treatment is based upon the observation characterized by pulmonary interstitial fibrosis [81, that folic acid cannot be directly utilized by micro- 85,88,99, 185, case 51. Case number five, diagnosed as organisms; thereforr, the interference with folic acid interstitial pneumonitis of infancy, was clinically in- presumably induced by pentamidine would sccm to distinguishable from Pt~eunzogstiscarinii pneumonitis. be an intraparasitic metabolic block. This assumption LVithout adequate morphological evidence, aspontanc- has not becn directly shown for Ptteurtrocyslis carirzii. The ous 'cause' would have been assumed in this instance. alteration in the bone marrow was readily rcversible Such discase entities may, in part, explain the diver- after cessation of the therapy. Thcrrfore, thcsr factors gent clinical findings of rrcovery from Ptretirnocystis should be considered when approaching the problrm disease [102, 1841. of the toxicity of pcntamidinc. The third problem is to decide if, in fact, recovery Dirrctly involved in consideration of the effectivc- from the Pneurttorystis infection has occurred in the ncss of a therapeutic rrgime is the long term effect of presence of another disease. This is a particularly the infection upon the host. Unfortunately, little data critical problem in the adult age group. Case 2 (D. H.) is available to answer this question. In our case [51] would suggcst that clinical cure can indeed be effected, (scr case rrport) two factors should be emphasized. as evidenced by a return of pulmonary function with- The first is thc persistence of confirmed sinopulmonary out limitation of activity. In the case reported by disease in this patirnt despite replacemrnt therapy WANTABE[120], a normal teenage child and her moth- with commercial immunoglobulin. Therapy was gaug- er presumably contracted the discase from an infected ed by analysis of serum IgG levrls prior to monthly father. The child and the mother developed clinical injection of gamma globulin. An attempt was made signs and symptoms strongly suggestive of Pneumocystis to maintain a serum level of at least 300 mg Oh prior disease within several months of each other. The child to reinjection. This inability to control thc prrsistent recovered without specific therapy or direct morpho- respiratory tract infections, reported by many others, logical diagnosis. Autopsy material obtained from the has promptrd the frequcnt use of various antimicrobial mother, who expired from a pneumonitis similiar to agents. To date, there has been no reacti\ration of that of the child and the father, confirmed the diagnosis Pneumocystis cnritlii infection during the three years of Prteurt~ocystis infection. One would have to assume following therapy in this patient despite the continued that both individuals contracted the same disease and chemotherapy. Therc is some indication that other that recovery without specific chemotherapy canoccur. drug regimes may be useful [54, 891. Anti-protozoan 'Spontaneous' recovery in an individual with an iden- agents, in grnrral, ha\ve bcrn studied. The therapeutic tifiable predisposing disease has not been reported. regime of pyramethaminc and sulfonamides, successful Our conclusion from these data is that'spontaneous' in the treatment of toxoplasmosis, seems to have a bene- recovery from Pneu~~zocystiscarinii pneumonitis has ficial effect in experimental Przeurtzocyslisinfection. How- occurred. With severe clinical symptoms, untreated individuals usually succumb to the disease. Similarly, ' Rhodia Inc., 600 Madison Avenue, New York, N.Y. patients with a significant predisposing disease would seem to have a highly predictable fatal prognosis with- data obtained from animal models have suggested that out chrmothcrapy. the differrnces in the disease thought to be due to age .4nothcr important aspect of the problem of analyz- and developmental status of the host are not as great ing the prognosis of affected individuals is the lack of as one had presumed. Similar pathogenetic and patho- information rrgarding the injured pulmonary paren- logical mechanisms are found in both the newborn and chyma. Information regarding long term follow- adult group. Chemothernpe~~ticagents play an im- through of patients who have recovered from Pneun~o- portant role in altering the normal physiology of the grti~caritiii pneumonitis is not available. Case number hosts of all age groups in normal and in diseased indi- one, an example of acquired hypogammaglobulinemia viduals. as predisposing causc, has bern studied for three years The narrow range of tissue sprcifirity and the un- following the initial infection. During this time, despite usual metabolic alterations induced by infection with rrpcatcd episodes of otitis media and pneumonia treat- this organism havr been emphasized. An attempt was ed with antibiotics, no evidence of reactivation of made to corrrlate the morphological lesion with clini- Pneun~opstirinfection has occurred. The patient has cal observation and laboratory data. A successful undergonr rarly sexual maturation, particpaties in 311 chemotherapeutic agent is now available for the trrat- activities in school and plays without restriction. How- ment of this disease. Since the few cases which have ever, a significant diminution in pulmonary function been studied have revealed a permanent disability due still persists as evidenced by a lowered vital capacity to the alvcolar septa fibrosis, and since a successful and diffusion gradient. The diminution in these func- chemotherapeutic agent is available, prompt diagnosis tions has remained unchanged through the three years and treatment have been advocated. The wide variety ofstudy. Although the restriction in vital capacity may of disease states that may prcrlispose to Ptieun~ocystis be attributed to the injury induced by her immunolog- carit~iipneumonia has been emphasized. It is hoped ical deficiency and resultant bacterial infections, the that future work, which may permit more careful bio- diminished alveolar function is best explained as a chemical, genetic and morphological analysis of the consequence of hcr Pneunzocystis infection. It may be organism, will allow precise identification of Prreumo- inferred that early diagnosis and treatment of this dis- cystis cnritrii outside of the diseased host and the distri- ease may rrsult in lcss permanent pulmonary injury. bution of the parasite in the environment. Interference with bone marrow metabolism, as X-ray films have returned to normal in about two evidencrd by the megaloblastosis observed in several to three months following therapy in the case studied individuals has not persisted following cessation of at our institution. therapy. Following institution of chemotherapy with pcnta- midinc, the progressive character of the disease ceases References atrd Notes within one week 146-53, 171, 1721. Fever, tachypnea, and cyanosis decrease in intensity and a sense of im- 1. CARINI,.A. und MACIEL,J.: fjber Pneumocystis provcment is reported by the patient. Roentgeno- carinii. Zbl.Bakt. 1: 46 (1915). graphic improvement may not be evident for two to 2. JIROVEC,0.: Pneumocystis carinii, pulvodect, zv. three weeks. Intramuscular administrationof4.0mg/kg interstitialnich plasmocelularnich pneumonii ko- intramuscularly for fourteen days has seemed to be an jencu. cs. Hyg. Epid. Mikrob. 1: 141 (1952). rffrctive schedule. 3. GERRARD,J.W.: Pneumocystis pneumonia. Sym- posium on recent clinical advances, May 1958. Pediat. Clin. N. Amer. pp. 327-335 (Saunders, Phi- ladelphia/London 1958). 4. GOETZ,0. : Klinik und Pathologie der interstitiel- Ptleunrory~tismrinii pneumonitis is best described as an len plasmazellularen Pneumonie. Archiv Kinder- example of an infection with an organism of low viru- heilk. 41: 1-73 (1960). Iencr. It is an ubiquitous parasite of undetermined 5. GAJDUSEK,D.C.: Pneumocystis carinii; etiologic classification because of the difficulty in examining the agent of interstitial plasma cell pneumonia of pre- morphology and physiology of the organism. The mature and young infants. Pediatrics 19: 543-565 cmergencc of thc organism in significant numbers rn (1957). the lung parenchyma under unusual conclitions would 6. FALKENBACH,K.H.; BACHMANN,K. D. and classify it best as an example of a latent infection in a O'LOUGHLIN,B. J. : Pneumocystis carinii pneu- predisposed host. The agents or pathological processes monia. Amer. J. Roentgenol. 85: 706-7 13 (196 1). which predispose the animal and human host have 7. PORTER,A.: The occurrence of Pneumocystis ca- been described. A carpful analysis of these cases and rinii in mice in England. Parasitology 8: 225 (19 15). Pneurnocystis carinii pneumonitis 153

8. VANEK,J.; JIROVEC,0. and LUKES,J.: Interstitial lar infant pneumonia. Mschr. Kinderheilk. 101: plasma cell pneumonia in infants. Ann. paediat. 147-149 (1953). Basel 180: 1-21 (1953). 26. BIELING,W. : Protozoan infections. Dtsch. med. 9. BRANDSTEIN,L. and CSILLAG,A.: Experimental Wschr. 78: 1801-1802 (1953). interstitial plasma cell pneumonia in sucking mice. 27. BAUCII,R. and LADSTATTER,L.: Pneumocystis Orv. Hetil. 95: 1003-1006 (1954). carinii and interstitial plasma cell pneumonia in 10. GOETZ,0. : Experimental contributions to thcetiol- premature infants. Klin. Wschr. 31: 37-38, 900- ogy of interstitial plasma cell pneumonia. Munch. 902 ( 1953). med. Wschr. 101: 141 1-1413 (1959). 28. SCHMID,K. 0. : Etiology of interstitial plasmocel- 11. SHELDON,W. H. : Experimental pulmonary Pneu- lular pneumonia (saccharomycosis) in infants; mocystis carinii infection in rabbits. J.exp. Med. identification of Pneumocystis carinii with Sac- 110: 147-160 (1959). charomyces fragilis. Frankfurt. 2. Path. 66: 426- 12. AKKER,S. and GOEKBLOED,E. : Pneumonia caused 448 (1955). by Pneumocystis carinii in a dog. Trop. geogr. Med. 29. STEINER,B. ; SZABON,J. and MOEIACSI,A. : Diagnos- 12: 54-58 (1960). tic value offungus studies in interstitial plasma cell 13. JIROVEC,0. : Toxopla~mosisand Pneumocystosis pneumonia. Helv.paediat. Acta 11: 501-505 (1956). as anthrapozoonoses. Zh. Mikrobiol. 32: 1890-1898 30. HAMPERL,H. : The pathological anatomy of inter- (1961). stitial pncumonia. Mschr. Kinderheilk. 108: 132 14. Fuc~s,U. Submicroscopic changes of the rat lung (1960). following the administration of a coumartin de- 31. NITSCHKE,.A. and HEYN,K.: On the problem of rivative. Frankfurt. Z. Path. 74: 555-564 (1965). the histological demonstration of Pneumocystis in 15. BUSTAMANTE,W. ; MORENO,L. and DIAZ,M. : In- interstitial pneumonia. Mschr. Kinderheilk. 108: terstitial plasma cell pneumonias with demonstra- 142-144 (1960). tion of Pneumocystis carinii in Chile. Mschr. Kin- 32. GADEKE,R.: Morphological antigen-antibody lo- derheilk. 108: 145-146 (1960). calization with fluorescent antibodies in lung tissue 16. BRITO,T. DE and ENGE,L. G.: Interstitial plas- in interstitial pneumonia. Mschr. Kinderheilk. 108: ma cell pneumonia caused by Pneumocystis 152-156 (1960). carinii. Rev. Inst. Med. trop. S. Paulo 4: 261-268 33. SCIIMID,K. 0.: Studies on Pneumocystis disease in (1962). man. 11. Developing Pneumocystis pneumonia in 17. RYAN,B. : Pneumocystis carinii infection in Mela- chronic lymphadcnosis of an adult. Frankfurt. Z. nesian children. J. Pediat. 60: 914-916 (1962). Path. 74: 146-161 (1964). 18. THYS,A. and JANSSENS,P.G.: Pneumocystosis in 34. ZANDANELL,E. : Finding of Pneumocystis outside Congolese infants. Trop. geogr. Med. 15: 158- 172 of the lungs in interstitial plasma cell pneumonia (1963). in infants and in premature. Zbl. allg. Path., path. 19. TIIYS,A. : Pneumocy~tosis in Congolese infants. Anat. 92: 75-80 (1954). Verh. \,laam. Akad. Geneesk. Belg. 25: 285-314 35. WILLMER,E. N. : Cells and tissues in culture; meth- (1963). ods, biology and physiology (Academic Press, New 20. POST,C.; DUTZ,W. and NASARIAN,I.: Endemic York/London 1965). * Pneumocystis carinii pneumonia in South Iran. 36. NAVRATIL,B. ; SMID,Z. and BARTA,K. : Serological Arch. Dis. Childh. 39: 35-40 (1964). determination of interstitial pneumocystic pneu- 2 1. JIROVEC, 0. : Interstitial pneumonia in infants monia in infants. Ann. paediat. (Basel) 183: 59-64 caused by Pneumocystis carinii. Mschr.Kinder- (1954). heilk. 102: 47-85 (1954). 37. VIVELL,0. : A new serodiagnostic method in inter- 22. REISETBAUER,E. and MORITSCH,H.: Epidemio- stitial plasma cell pneumonia in newborn and pre- logical and serological studies on interstitial plasma mature infants. Dtsch.med. Wschr. 79: 358-360 cell pneumonia in prematures. Mschr. Kinderheilk. (1954). 104: 41-46 (1956). 38. VIVELL,0.; BUHN,W. H. and LIPS,G.: Experi- 23. JIROVEC,0. : The problem of Pneumocystis pneu- ences with scrodiagnosis of interstitial plasma cell monia from the parasitological point of view. pneumonia of young and premature infants. Z. Mschr. Kinderheilk. 108: 136-142 (1960). Kinderheilk. 78: 653466 (1956). 24. PASCHLAU,G.: Epidemiological observations with 39. JAHN, E. and ROLLER-GUSINDE,R. E. : Serology the aid of the complement fixation reaction toward and clinical aspects of interstitial plasma cell pneu- the pathogen of interstitial pneumonia. Mschr. monia. Klin. Wschr. 35: 37-41 (1957). Kinderheilk. 108: 15 1-152 (1960). 40. GOETZ,0. : The antigen of interstitial pneumonia. 25. GIESE,H. : Etiology of the interstitial plasmacellu- Klin. Wschr. 36: 118-120 (1958). 41. VIVEI.L,0. : The serology ofinterstitial pneumonia. interstitial pneumonia of the newborn. Z. Kinder- Mschr. Kinderheilk. 108: 14G151 (1960). heilk. 78: 166176 (1956). 42. BARTA,K. and LYSEK,H. : Experimental pneumo- 57. SHELDON,W. H. : Pulmonary Pneumocystis carinii cytosis. I. Studics on complement fixation anti- infection. J. Pediat. 61: 780-791 (1962). bodies. cs. Epidcm. 11: 196-202 (1962). 58. PLIESS,G.: Interstitial plasma cell pneumonia of 43. GOETZ,0. : Serological findings in interstitial pneu- infants as a generalized disease. Frankfurt. Z. Path. monia from the U.S. Arch. Kinderheilk. 170: 60-66 68: 565-595 (1957). (1964). 59. MICHAILOV,G. : Morphological changes in lungs 44. GOETZ,0.: Fluorescence microscopic studies in in Pneumocystis pneumonia. Zbl. allg. Path. path intcrstitial plasma cell pneumonia. Msehr. Kinder- Anat. 96: 146-149 (1957). heilk. 113: 194-197 (1965). 60. LYONS,J. A.; VINIJCHAIKUL,K. and HENNIGAR, 45. BARTA,K. : On the problem of the antigen effective G. R. : Pneumocystis carinii pneumonia unassoci- components of the lung in pneumocystic pneu- ated with other disease. Clinical and pathological monia. Path. et Microbiol. Base1 29: 63-74 (1966). studies. Arch.intern.Med. 108: 929-936 (1961). 46. IVADY,G. and PALDY,L. : New method of treating 61. HAJDUK,A.: Certain aspects of the initial stages of interstitial plasma cell pneumonia in premature interstitial pneumocystis pneumonias. cs. Pediat. infants with 5-valent antimony and aromatic di- 15: 314-319 (1960). amidines. Mschr.Kindcrhei1k. 106: 10-14 (1958). 62. PAVLICA,F.: The first observation of congenital 47. IVADY,G.; PALDY,L. and UNGER,G.: Further ex- pneumocystic pneumonia in a fully developed still- periences in the treatment of interstitial plasma cell born child. Ann.paediat. (Basel) 198: 177-184 pneumonia with pentamidine. Mschr. Kinder- (1962). heilk. 111: 297-299 (1963). 63. AREAN,V. : Neumocy~tosi~pulmonar. Progr. Pat. 48. RODGER,T.S. and HAGGIE,M.H.K.: Pneumo- Clin. 12: 1 (1965). cystis carinii pneumonia associated with hypo- 64. BURKE,B. A. ; KROVETZ,L. J. and GOOD,R. A. : gammaglobulincmia responding to pentamidine. Occurrence of Pneumocystis carinii pneumonia in Lancet i: 1042 (1964). children with agammaglobulinemia. Pediatrics 28: 49. LORINCZI,K.: Our experiences with pentami- 196205 (1961). dine in the trcatment of interstitial plasma cell 65. MULLER-EBERHARD,H. J. and KUNKEL,H. G. : pneumonia. Gyermekgy6gyiszat 15: 207-2 12 Ultracentrifugal characteristics and carbohydrate (1964). content of macromolecular gammaglobulins. Clin. 50. MARSHALL,W. C. ; WESTON,H. J. and BODIAN,M. : chim. .4cta 4: 252-258 (1959). Pneumocystis carinii pneumonia and congenital 66. GOMORI,G.: Silver impregnation of reticulum in hypogammaglobulinemia. Arch. Dis. Childh. 39: paraffin section. Amer. J. Path. 13: 993 (1937). 18-25 (1964). 67. HAMPERL,H. : Demonstration of parasites in inter- 51. ROBBINS,J. B. ; MILLER,R. H. ; AREAN,V. M. and stitial plasmocellular pneumonia. Klin. Wschr. 30: PEARSON,H. A. : Successful treatment of Pneumo- 820-822 (1952). cystis carinii pneumonitis in a patient with con- 68. GARSCHE,R. : Clinical importance of roentgen di- genital hypogammaglobulinemia. New Engl. J. agnosis in interstitial pneumonia of premature in- Med. 272: 708-713 (1965). fants. Fortschr. Rontgenstr. 75: 125-138 (1951). 52. BLATTNER,R.J.: Pneumocystis carinii infection: 69. BAAR,H. S. : Interstitial plasma cellular pneumonia Treatment with pentamidine isothionate. J. Pediat. due to Pneumocystis carinii. J. clin. Path. 8: 19-24 67: 332-335 (1965). (1955). 53. PATTERSON,J.H. ; LINDSEY,I. L. ; EDWARDS,E. S. 70. GAGNE,F. and HOULD,F.: Interstitial plasma- and LoGAN,J~.,W. D. : Pneumocystis carinii pneu- cellular (parasitic) pneumonia in infants; report of monia and altered host resistance: Treatment of 3 cases. Canad. med. Ass. J. 74: 620-624 (1956). one patient with pentamidine isothionate. Pediat- 71. ROWE,C. W. : Pneurnocystis carinii pneumonia. rics 38: 388-397 (1966). Radiology 75: 257-261 (1960). 54. FRENKEL,J.K.; GOOD,J.T. and SHULTZ,J.A.: 72. RUBIN,E. and ZAK, F.G.: Pneumocystis carinii Latent Pneumocystis infection of rats, relapse, and pneumonia in the adult. New Engl. J. Med. 262: chemotherapy. Lab. Invest. 15: 1559-1 577 (1966). 1315-1317 (1960). 55. RICKEN,D. : Histologische Untersuchungen bei ex- 73. LEBEDA,D.: Spontaneous fractures of the ribs perimenteller Pneumocystis-Pneumonie. Virchows "from fatigue" in infants. Cs.~ediat.16: 667-670 Arch.path. Anat. 331: 713-728 (1958). (1961). 56. WELLER,R. : Further studies ofexperimental cystic 74. WHITE, W.F.; SAXTON,H. M. and DAWSON, disease of thr lung in rats with special reference to L. M. P. : Pneumocystis pneumonia. Report of Pneumocystis carinii pneumonitis 155

three cases in adults and one in a child with a dis- and SOLONUAR,W. : Diagnostic in-viuo de la pneu- cussion of the radiological appearances and pre- monie a pneumocystis. Arch.frang. PCdiat. 20:773 disposing factors. Brit. med. J. 5263: 1327-133 1 (1963). (1961). 92. FULGINITI,V. A. : Case report. Personal communi- 75. HENNIGAR,G. R. ; VINIJCHAIKUL,K. ;ROQUE, A.L. cation. and LYONS,H. A. : Pneumocystis carinii pneumonia 93. MULLER,H. and VIETOR,G.: Clinical and epi- in an adult. Report of a case. Amer.J.clin.Path. demiological aspects of interstitial plasma cell 35: 353-364 (1961). pneumonia in young children. Z. Kinderheilk. 79: 76. FEINBERG,S. B. ; LESTER,R. G. and BURKE,B. A. : 4048 (1957). 'The roentgen findings in Pneumocystis carinii 94. STOPKA,E.; WUNDERLICH,C. and CARLSON,S.: pneumonia. Radiology 76: 594-600 (1961). Morphological and cultural studies of Pneumo- 77. HOLLAND,P. D. and WARD,0. C. : Pneumocystis cystis in sputum and lung material. Z. Kinderheilk. carinii pneumonia. Irish J. med. Sci. 438: 258-261 79: 246-263 (1957). (1962). 95. BACHMAN,K. D. : Epidemiology and incubation of 78. HENDRY,W. S. and PATRICK,R. L.: Observations interstitial pneumonia. Z. Kinderheilk. 74: 133 on thirteen cases of Pneumocystis carinii pneumo- (1954). nia. Amer. J.clin. Path. 38: 401-405 (1962). 96. ANDERSON,C. D. and BARRIE,H. J. : Fatal pneumo- 79. ROBILLARD,G. ; BERTRAND,R. ; GREGOIRE,H. and cystis pneumonia in an adult. Report of a case. BERDNIKOFF,G. : Plasma cell pneumonia in infants. Amer. J. clin. Path. 34: 365-370 (1960). Review of 5 1 cases. J. Canad. Ass. Radiol. 16: 16 1- 97. NICASTRI,A.D.; HUTTER,R.V.P. and COLLINS, I68 (1965). H. S. : Pneumocystis carinii pneumonia in an adult; 80. MURRAY,J. F.; HAEGELIN,H. F. ; HEWITT,\Y. L.; emphasis on ante-mortem morphologic diagnosis. LATTA,H. ; MCVICKAR,D. ; Rasmusscn, A. F., Jr. N.Y. State J.Med. 65: 2149-2154 (1965). and RIGLER,L.G.: Opportunistic pulmonary in- 98. BENCROFT,D. M. and COSTELLO,J. M. : Pneumo- fections. i-\nn.intern. Med. G5: 566-594 (1966). cystis carinii pneumonia in siblings: diagnosis by 8 1. GRUENWALD,P. and JACOBI,M. : Mononuclear lung aspiration. N. Z. med. J. 6#:273-280 (1965). pneumonia in sudden death or rapidly fatal illnrss 99. GAENSLER,E. A.; MOISTER,V. B. and HAMM,J.: in infants. J. Pediat. 39: 650-662 (1951). Open-lung biopsy in diffuse pulmonary disease. 82. RUSSELL,J.G.: Pneumocystis pneumonia associ- New Engl. J. Med. 207: 1319 (1964). ated withagammaglobulinemia. Arch. Dis. Childh. 100. DONOHUE,W. : Interstitial plasma cell pneumo- 34: 33S341 (1959). nia. Labor. Invest. 5: 97-105 (1956). 83. CAHALANE,S. F. : Pneurnocystis carinii pneumonia. 10 1. LIM,S. K. and MOON,C. J. : Studies on Pneumo- Report of a case and review of the litcraturc. cystis carinii pneumonia : 11. Epidemiological and J. Irish mrd. Ass. 50: 133-136 (1962). clinical studies of 80 cases. Jonhap. Med. 6: 77 84. LUNSETH,J. H. ; KIRMSE,T.W. ; PREZYNA,A. P. and (1960). GERTA,R. E. : Interstitial plasma cell pneumonia. 102. PALITZSCII,D. VON and GOETZ,0. : Diaplacental J.Pediat. 46: 137-145 (1955). transfer of antibodies against interstitial plasma 85. FEINERMAN,B. and HARRIS,L. E. : Unusual inter- cell pneumonia. Arch. Kinderheilk. 173: 27-3 1 stitial pneumonitis; report of two cases occurring (1965). in children. Proc. Mayo Clin. 32: 637440 (1957). 103. TOBLER,W.: La prophylaxie de la pneumonie 86. WOLMAN,M. and GOLDBERG,M. G.: The micro- interstitielle, par I'irradiation non interrompue copical criteria of interstitial pneumonia. Arch. aux rayons ultra-violets des boxes. Ann. paediat. Path. 65: 272-278 (1958). (Basel) 187: 284 (1956). 87. ROSEN,H. S.; CASTLEMAN,B. and LIEBOW,A..4.: 104. VON HARNACK,G.A.: Organizational problems Pulmonary alveolar proteinosis. New Engl. J. Med. in the prevention of interstitial pneumonia. 258: 1123 (1958). Mschr. Kinderheilk. 108: 159-161 (1960). 88. DE SANCTIS,P. N. : Pulmonary alveolar proteinosis. 105. WALTHER,T. : Interstitial plasma cell pneumonia ; Boston med. Quart. 13: 19-25 (1962). report of two cases. Acta Paediat. 39: 545-553 89. RIFKIND,D.; FARIS,T. D. and HILL,R.B.: Pneu- (1950). mocystis carinii pneumonia. Ann. intern. Med. 65: 106. GORMSEN,H.: On interstitial plasma cell pneu- 943 (1966). monia in infants. Acta Paediat. Stockh. 39: 4-5, 90. ERCIIUL,J. W.; WILLIAMS,L. P. and MEIGHAN, 291-314 (1950). P. P. : Pneumocystis carinii in hypopharyngeal ma- 107. DEAMER,W. C. and ZOLLINGER,H. U. : Inter- terial. New Engl. J.Med. 267: 926-927 (1962). stitial plasma cell pneumonia of premature and 91. TAN-VINH,L. ; COCHARD,A. M. ; VU-TRIEU-DONG, young infants. Pediatrics 12: 11-22 (1953). 108. STERNBERG,S. D. and ROSENTHAL,J. H. : Inter- 124. DVORAK,J. ~~~JIROVEC,0. : Diagnosis and ther- stitial plasma cell pneumonia. J. Pediat. 46: 380- apy of parasitic atypical pneumonia in infant 393 (1955). caused by Pneumocystis carinii. Acta paediat. 109. GERRARD,J.W. and MOORE,D. F. : Interstitial (Stockh.) 42: 232-245 (1953). plasma cellular pneumonia due to pneumocystis 125. EBERLE,H. : Ost. Z. Kinderheilk. 9: 421 (1954). carinii. Canad. med. Ass. J. 76: 299-302 (1957). 126. FINGERLAND,A. and VORTEL,V. : Pneurnocystis 110. EBERLING,E.W. and COHEN,F.: Pneurnocystis pneumonia. Schweiz. Z. allg. Path. 17: 201-220 carinii pneumonia; report of a case from South- (1954). western U.S. Pediatrics 21: 345-352 (1958). 127. KOECEIER,P. H. : The problem of disposition to 11 1. HOWARD,R. M. and SHELDON,W. H.: Pneumo- early infantile interstitial pneumonia. Mschr. cystis pneumonia; report of a case. J. Dis. Child. Kinderheilk. 104: 322-325 (1956). 95: la24 (1958). 128. KOLACZEK,B. et al. : Congenital tuberculosis and 112. RUSSELL,H.T. and NELSON,B. M. : Pneumocystis pneumocystic infection. Cs.Pediat. 20: 629433 pneumonitis in American infants. Arner.J. clin. ( 1965). Path. 26: 1334-1340 (1956). 129. OFFEN~IAUER,R.: Epidemiological aspects and 113. BERLIN-HEIMENDAHL,S. VON und GOETZ,0. : Zur symptomatology of interstitial plasma cell pneu- Epidemiologic, Therapie, und Prophylaxe der in- monia. Z. Kinderheilk. 90: 251-262 (1964). terstitiellen Sauglingspneumonie. Z. Kinderhcilk. 130. SCHMOGER,R.: 502 Falle interstitieller Pneumo- 79: 689 (1957). nie eigener Beobachtung. Z. Kinderheilk. 79: 522 114. MCNEAL,J. E. and YAEGER,R. G. : Observations (1957). on a casc of pneumocystis pncumonia. Arch. Path. 131. WEISSE,K. and WEDLER,E. : Occurrence of the 70: 397-406 ( 1960). so-called Pneumocystis carinii. Klin. Wschr. 32: 1 15. STINSON,J. C. ; LEA,W.A., Jr. ; WOOD,H. L. and 11-12,270-271 (1954). CRAMM,C. J., Jr.: Coexisting heart disease, Scle- 132. WILSON,J. F. ; CRASS,G. and GARZA,B. L.: Pneu- rema Neonatorum, and possible portozoan infec- mocystis carinii pneumonia. Report of a case and tion. J. Arner. med. Ass. 181: 899-901 (1962). review of the recent literature. Pediatrics 25: 468- 116. DULITZKY,M. E. and REPETTI,L. P. : "Pneumo- 476 ( 1960). cystis Carinii" pneumonitis. Plasma cell pneu- 133. TOLDYOVA,V. : Apropos of interstitial pneumo- monia in infants. La Semana Medica 122: 265- cystic pneumonia in pediatric institutions. Bratisl. 270 (1963). lek. Listy 44: 230-236 (1964). 1 17. HUNEYCUTT,H. C. : Pneumocystis carinii pneu- 134. BIRD,T. and THOMSON,J.: Pneumocystis carinii monia. Case studies with electron microscope. pneumonia. Lancet 272: 59-64 (1957). Amer.J.clin. Path. 41: 41 1-418 (1964). 135. CIIAGAS,C. : Nova tripanozomiaza humana. Inst. 118. CABRERA,H. A. and DAVIS,G.H.D. : Epidemic Oswaldo Cruz, Rio 3: 2 19 (191 1). meningitis of the newborn caused by Flavobac- 136. ROBINSON,J.J. : Two cases of pneumocystis. Ob- teria. Amer. J. Dis. Child. 101: 289 (1961). servation in 203 adult autopsies. Arch. Path. 71: 119. LIM,S.K. and MOON,C. J. : Studies on Pneumo- 15G159 (1961). cystis carinii pneumonia. I. Epidemiological and 137. KAFTORI,J. K. ; BASSAN,H. ; GELLEI,B. and GRIF- clinical studies of 80 cases. Jonhap. Med. 6: 69 FEL,B. : Pneumocystis carinii pneumonia in the ( 1960). adult. Report of a primary case. Arch. intern. 120. WATANABE,J. M.; CHINCNINIAN,H.; WEITZ,C. Med. 109: 438-446 (1962). and MCILVANIE,S. K. : Pneumocystis carinii 138. TIMMEL,H. : Electronoptic studies on cytomegalia pncumonia in a family. J. amer. mcd. Ass. 193: nuclear inclusions. Frankfurt. Z. Path. 73: 514- 685-686 (1965). 519 (1964). 12 1. HYUN,B. H. ; VARGA,C. F. and THALHEIMER,139. SYMMERS,W. S. : Generalized cytomegalic inclu- L. J. : Pneumocystis carinii pneumonitis occurring sion-body disease associated with Pneumocystis in an adopted Korean infant. J. amer. med. Ass. pneumonia in adults. J.clin. Path. 13: 1-21 195: 784-786 (1966). (1960). 122. ZHUKOVA,E.K.: A case of pneumoceptic pneu- 140. RIFKIND,D. ; STARZL,T. E. ; MARCHIORO,T. L. : monia combined with generalized cy tomegaly in WADDELL,W. R.; ROWLANDS,J.T. and HILL, a child. Arkh.Pat. 24: 68-71 (1962). R. B. : Transplantation pneumonia. J. amer. med. 123. DRUBE,H. : The treatment ofearly infantile inter- Ass. 189: 808-8 12 ( 1964). stitial plasma cell pneumonia with moronaltacho- 141. NAKAMURA,R. M. : Coexistent cytomegalic in- liquin aerosols. Miinch.med.Wschr. 101: 1414- clusion disease and pneumocystis carinii infection 1417 (1959). in adults. Acta path.jap. 14: 45-59 (1964). Pneumocystis carinii pneumonitis 157

142. KRAMER,R. I.; CIRONE,V. C. and MOORE,H. : 158. MULLER,G. : Interstitial plasma cell pneumonia Interstitial pneumonia due to Pneumocystis ca- after corticostcroid therapy. Frankfurt. Z. Path. rinii, cytomegalic inclusion disease and hypo- 70: 657-675 (1960). gammaglobulinemia occurring simultaneously in 159. AHERNE,W. : A case of osteopetrosis (Albers an infant. Pediatrics 29: 816-827 (1962). Schonberg) with intercurrent pneumocystispneu- 143. HAMPERL,H.: Pneumocystis infection and cyto- monitis. Arch. Dis. Childh. 35: 495-502 (1960). megaly of the lungs in the newborn and adult. 160. KOSSEL,A. : Interstitial plasma cell pneumonia in Amer. J. Path. 32: 1-13 (1956). older children following prolonged corticosteroid 144. WILLIAMS,G.; STRETTON,T.B. and LEONARD, treatment. Report on 13 cases. Dtsch. med.Wschr. J. C. : Cytomegalic inclusion diseasc and pneumo- 87: 1133-1 139 (1962). cystis carinii infection in an adult. Lancet ii: 951- 161. GOETZ,0. and OIIRT,B. : Long-term cortisone 955 (1960). therapy and interstitial pneumonia. Mschr. Kin- 145. WEISSE,K.: Generalized cytomegaly and local- derheilk. 112: 163-164 (1964). ized pneumocystosis in a case of infantile inter- 162. GILBERT,C. F. ; FORDHAM,C. C. and BENSON, stitial plasma cell pneumonia. Z. Kinderheilk. 76: W.R. : Death resulting from Pneumocystis pneu- 27-32 (1955). monia in an adult. Arch. intern. Med. 112: 158- 146. WINSLOW,D. J. and HATHAWAY,B. M. : Pulmo- 163 (1963). nary pncumocystosis and cryptococcosis. Amer. J. 163. MIREJOVSKI,P.: Pneumocystic pneumonia with clin.Path. 31: 337-342 (1959). rupture of the right cardiac appendage after corti- 147. ASTOR,K. and SCHREIER,K. : On an epidemic of coid administration. cs. Pediat. 19: 73 1-735 interstitial and suppurative pneumonia and of (1964). mixed forms. Ann.peadiat. (Basel) 194: 1-10 164. MARKI,H.H. VON und WICK,A.: Transitorische ( 1960). Hypogammaglobulinamie nach Goldtherapie. 148. DAUZIER,G.; WILLIS,T. and BARNETT,R.N.: Schwciz. med. Wschr. 94: 131-133 (1964). Pneumocystis carinii pneumonia in an infant. 165. DYSZY-LAUBE,B.: Pneumocystis carinii in chil- Amer. J. clin. Path. 26: 787-793 (1956). dren with hematological diseases. Pediat.pol. 39: 149. MCKAY,E. and RICHARDSON,J.: Pncumocystis 1193-1 197 (1964). carinii pneumonia associated with hypogamma- 166. DA SILVA,G.; GOMES,M. C. and SANTOS,R. F. : globulinemia. Lancet ii: 713-7 15 (1959). Pneumocystis pneumonia in the adult. Report of 150. COIIEN,W. N. and MCALISTER,W. H. : Pneumo- case associated with corticosteroid therapy for cystis carinii pneumonia. Report of4 cases. Amer. rheumatoid arthritis. Rev. Inst. Med. trop. S.Pau- J.Roentgenol. 89: 1032-1037 (1963). lo 7: 31-38 (1965). 151. BOUTON,J. ; MAINWARING,D. and ~MITIIELLS,167. IRVING,W. R., Jr.; MARE,A. and MIALE,J.B. : R.W. : BCG dissemination in congenital hypo- Unusual infection in lymphoma and leukemia. gammaglobulinemia. Brit. med. J. i: 1512-1 5 15 Sth.med.J. 58: 891-895 (1965). (1963). 168. BOSMAN,C.: Hypervitaminosis D and pneumo- 152. COLLIER,F. C. : Pneumocystis carinii pncumoni- cystis pneumonia in an infant. Helv. paediat. Acta tis. Missouri Mcd. 61: 772-774 (1964). 20: 300-309 ( 1965). 153. ALLIDONE,E. C. ; GOLDIE,W. and MARMION, 169. ESTERLY,J. A. and WARNER,N. E. : Pneumocystis B. P. : Pneumocystis carinii pneumonia and pro- carinii pneumonia. Twelve cases in patients with gressive vaccine in siblings. Arch. Dis. Childh. 39: neoplastic lymphoreticular disease. Arch. Path. 26-34 (1964). 80: 433-441 (1965). 154. DOTSON,T. and KAPCAR,Jr., A. J. : Pneumocystis 170. CALLERAME,M. L. and NADEL,M. : Pneumocystis carinii pneumonia with hypogammaglobulin- carinii pneumonia in two adults with multiple rmia. W.Va. med. J. 61: 31-34 (1965). myeloma. Amer. J. clin. Path. 45: 258-963 (1966). 155. SCHALLER,J.; CHING,Y.; WILLIAMS,C. P. S. ; 17 1. Data kindly made available by Dr. EUGENESMITH, DAVIS,S.D.; LAGUNOFF,D. and WEDCWOOD, Closter, N. J. R. J. : Hypergammaglobulincmia, antibody defi- 172. Data kindly made available by Dr. EMANUEL ciency, autoimmune haemolytic anemia, and KAUDER,University of Cincinnati. nephritis in an infant with a familiar lymphopcnic 173. BERNIER,G.M. and PUTNAM,F.W.: Myeloma immune defect. Lancet ii: 00-000 (1966). proteins and macroglobulins: Hallmarks of dis- 156. ROB~INS,J. B. : Unpublished case. ease and models of antibodies; in Progress in 157. Roos, B. and KELLER,H.: Pneumocystis pncu- Hematology, cd. CARLV. MOOREand ELMERB. monia after corticoid therapy. Path. et Microbial. BROWN,~01.4, p. 160 (Grune and Stratton, New (Basel) 23: 277-283 (1960). York 1964). 174. ULTMANN,J.E.; FISH,W.; OSSERMAN,E. and 182. KALOUD,H. : Treatment of interstitial plasma cell GELLHORN,A. : The clinical implications of hypo- pneumonia with hydrated ergot alkaloids. Wien. gammaglobulinemia in patients with chronic med. Wschr. 108: 299-301 (1958). lymphocytic leukemia and lymphocystic lympho- 183. KALOUS,H.: On the treatment of interstitial sarcoma. Ann. intern. Med. 51: 501-516 (1959). plasma cell pneumonia with hydrogenated ergot 175. GOETZ,0. and RENTSCH,L.: Weitere Unter- alka!oids. Mschr.Kinderheilk. 108: 161-164 suchungen zur experimentellen Rattenpneumo- (1960). cystose. Z.Kinderheilk. 79: 578 (1957). 184. JOPPICH, G. : Unsere Erfahrungen mit der Chloro- 176. KERPEL-FRONIUS,F. ; VARGA,F. and BATA,G. : mycetin-Behandlung bei der interstitiellen Pneu- Blood gas and metabolic studies in plasma cell monie der Fruhgeborenen. Ann. paediat. (Basel) pneumonia and in newborn prematures with res- 180: 264 (1953). piratory distress. Arch. Dis. Childh. 39: 473-480 185. GOODMAN,L. S. and GILMAN,A.: The pharma- ( 1964). cological basis of therapeutics; 3rd ed., p. 1101 177. DROESE,W. : .Acid-base equilibrium of premature (Macmillan, London/Toronto 1965). infants; special characteristics in plasmacellular 186. Data kindly made available to us by Dr. P. CA- pneumonia. Helv. paediat. Acta 10: 97-103 TALBANO, San Diego, Calif., and Dr. I.LIEBOW, (1955). New Haven, Conn. 178. VARGA,F. : Hypoxia and metabolic disturbances 187. Consultation and help from DI'.VINCENTFULGI- in plasma cell interstitial pneumonia ofpremature NITI, Dr. EUGENESMITH, Dr. EMANUELKAUDER, and undernourished infants. cs. Pediat. 20: 31 7- Dr. PATRICIAMONTALBANO, Dr. RALPHWEDG- 319 (1965). WOOD, and Dr. ALVINROBINS are gratefully ac- 179. GOETZ,0. und RENTSCH,L.: Prufung der Wir- knowledged. kung von UV-Licht und medikamenten auf 188. The clinical use and evaluation of the drug, pent- Pneumocystis carinii im Tierversuch. Z. Kinder- amidine isothionate, was facilitated under the heilk. 79: 305 (1957). direction of Dr. FRANCISKELSEY. 180. MANEKE,M.: Therapy of interstitial plasma cell 189. The author is recipient of N.I.H. Career Develop- pneumonia in young and premature infants. ment Award No. 6 K3-HD-22 856-0141. This in- Mschr. Kinderheilk. 102: 48S487 (1954). vestigation was partially supported by a N.I.H. 181. LAURSEN,H.: Interstitial plasma cell pneumonia Research Grant No. AI-HD-08768-01 A.I.A. and in premature infants. Nord. Med. 54: 1285-1288 NO. NIH-DH-06786-02. (1955).