Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

707

Pneumonia Transplant

Robert

H. Wise, Jr.1 Myung S. Shin1 Jon P. Gockerman2 Sandra L. Zornes3 Eva Rubin1 P. H. Nath1

in Bone Marrow Patients

The radiographic, pathologic, and clinical features of 22 episodes of pneumonia in 18 bone marrow transplant recipients were analyzed retrospectively. These pneumonias could be divided into three diagnostic categories: (1) a transient form of nonspecific interstitial pneumonia resembling pulmonary edema in radiographic appearance occuffed in five patients during the first 2 weeks after transplantation; (2) a fatal, progressive form of interstitial pneumonia began from 2 weeks to several months after bone marrow transplantation in seven patients; and (3) airspace pneumonias occurred in nine patients within 2 months of transplantation and were uniformly fatal. While the temporal and radiographic presentation of transient early interstitial pneumonia is often charactenstic and may not require biopsy for diagnosis, the airspace and progressive interstitial pneumonias affecting bone marrow transplant recipients present variable patterns that are often radiographically indistinguishable. The direct pulmonary toxicity of high-dose total-body irradiation and chemotherapy may be contributory in producing both the early transient and later progressive forms of nonspecific interstitial pneumonia.

Allogeneic bone marrow transplantation has become a useful procedure in the treatment of lethal marrow diseases such as leukemia and aplastic anemia. The technique of transplantation involves eradication of both native and diseased marrow by administration of what would ordinarily be lethal doses of chemotherapy sometimes in association with high-dose, total-body irradiation [1 21. This preparatory or “conditioning” therapy may induce life-threatening, short- and long-term pulmonary sequelae by causing direct lung injury and by suppressing host immune response to infection [1-3]. A logical, although not entirely accurate, means of categorizing the pneumonias that occur frequently in these patients has been to designate them as either airspace or interstitial in nature [3-5]. In fact, it has been shown that so-called interstitial pneumonia can have a prominent airspace component that can render radiographic differentiation from bacterial and fungal pneumonias difficult [6]. Furthermore, the interstitial lung disease sometimes induced by many chemotherapeutic agents is often associated with fever, which makes distinction from infection difficult on clinical grounds [7]. It is important that the radiologist recognize the common radiographic presentations of pulmonary disease in bone marrow transplant recipients without overestimating their specificity. In this report we review the radiographic manifestations of pulmonary disease in 18 bone marrow transplant recipients and correlate these findings with pathologic and clinical information. ,

Received

February 28, 1984; accepted after re-

vision June 8, 1984. Presented at the annual meeting of the Society of Thoracic Radiology, Lake Buena Vista, FL, March 1984. Department of Diagnostic Radiology, University of Alabama School of Medicine, 619 S. 19th St., Birmingham, AL 35233. Address reprint requests to R. H. Wise. 2Departmt of Medicine, Division of Hematology and Oncology, University of Alabama School of Medicine, Birmingham, AL 35233. Present address: Duke University School of Medicine, Durham, NC 27706. Department of Pathology, University of Alabama School of Medicine, Birmingham, AL 35233. AJR 143:707-714, October 1984 0361 -803X/84/1 434-0707 © American Roentgen Ray Society

Materials

and

Methods

The chest radiographs of 22 patients who had received bone marrow transplantation between October 1980 and February 1983 were reviewed retrospectively and the findings correlated with clinical information and pathologic data obtained by biopsy or at autopsy when available. Except for one patient with mild interstitial lung disease, all chest radiographs were normal at the time of conditioning therapy before transplantation.

708

WISE

TABLE

1: Radiographic,

Pathologic,

and Clinical

Features

of Pneumonia

in Bone Marrow

Transient Interstitial Pneumonia

Etiology

?Direct

lung injury

by ra-

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

Radiographic

1-2 weeks

findings.

Pathologic

.

Diffuse

.

(up to 4 weeks)

interstitial

pattern

often with Kerley B lines. Airspace component may occur Interstitial mononuclear in-

findings

filtration,

fibrosis, and edema. Alveolar pneumocyte hyperplasia (type

Self-limited. Improvement or resolution within 8-17

Pneumonia

CMV, HSV, or Pneumocystis carinll; 40% idi-

60%

Gram-negative concurrent

opathic

bacteria,

fungi,

infections

Highly variable, usually 210 weeks, but may occur up to 7 months later

Within 2 months, when granulocytopenia is common*

Progressive mixed (combined or unclassifiable

Focal, multifocal, or diffuse airspace disease. Fungi often produce nodules and/or cavi-

pattems)

ties Same except greater interstitial fibrosis and alveolar filling with hyperplas-

filling with inflammatory cells and hemorrhage

Alveolar

tic type II pneumocytes

and hyaline membranes

II) and hyaline membranes Outcome

1984

Recipients

Progressive Interstitial Pneumonia

diation and/or chemotherapy Time of onset after transplantation

AJR:143, October

ET AL.

Fatal in our senest

Fatal*

days Note.-CMV

=

cytomegalovirus;

One case of focal airspace

t

Other

series

report some

The conditioning of arabinosylcytosine

HSV pneumonia

survivors

herpes

=

caused

simplex

virus.

by adenovirus

with permanent

interstitial

developed

2 years after transplantation

regimen for the patients with leukemia consisted (ara-C)

500 mg/m2

utive days followed by cyclophosphamide 2 days followed by total-body in’adiaiton

intravenously

for 5 consec-

60 mg/kg intravenously for at a dose of 850-1 000 rad

(8.5-1 0 Gy). One patient with leukemia received marrow transplantation at another institution and was pretreated with the same drugs although at a lower dosage. Conditioning therapy for the aplastic anemia group consisted of cyclophosphamide 50 mg/kg intravenously for 4 consecutive days. Marrow donors were selected on the basis of histocompatibility testing. Under general anesthesia, marrow was obtained from donors by repeated percutaneous needle aspirations from the posterior iliac crests and was infused intravenously into the host.

Patients were hospitalized for at least 1 month in positive-pressure rooms under aseptic conditions. Chest radiographs (posteroantenor and lateral projections) were obtained routinely on a weekly basis or more often (usually daily) during periods of pulmonary disease. A series of chest radiographs of each patient was reviewed independently by two chest radiologists (E. A. and P. H. N.) who knew both the patient population and the general nature of the project but had no other specific information about the patients. Pulmonary abnormalities were categorized as interstitial, airspace, or as mixed (combined interstitial and airspace) patterns. In instances in which there was a significant discrepancy in interpretation, a third opinion (R. G. F.) was obtained, and, when possible, the majority opinion was

accepted. pattern

If there was no agreement was

deemed

unclassifiabl&

and had a self-limited

course.

lung disease.

among the three observers the and included in the mixed”

category.

Results

Two patients had ne.. significant pulmonary disease after transplantation. Two other patients had radiographic evidence of pulmonary disease for which no diagnosis was established. At least one episode of pulmonary disease occurred in each of the other 18 patients for a totalof 22 separate episodes of

pulmonary disease for which diagnoses were established. Diagnoses were confirmed by autopsy, open lung biopsy, or by transbronchial biopsy in 20 episodes of pulmonary disease. The other two diagnoses were made on the basis of compelling clinical and radiographic evidence. For clarity of presentation, patients have been grouped into three diagnostic categones: transient interstitial pneumonia, progressive interstitial pneumonia,

Transient

and

Interstitial

airspace

pneumonia

(table

1).

Pneumonia

Transient interstitial pulmonary disease developed 6-12 days after transplantation in five of the 22 patients. Initial radiographic findings were entirely consistent with interstitial pulmonary edema (fig. 1). All five patients had diffuse, hazy, or ground-glass opacification of the lungs with loss of vascular sharpness, and three of the five had Kerley B lines. An airspace component developed later in one patient in a bilateral but asymmetric pattern and was questionably present in two other patients in a bilateral, symmetric distribution. Radiographic abnormalities resolved completely in three patients and improved markedly in the other two in 8-17 days. Clinicallyall patients had fever of 40#{176}-41 #{176}C, tachypnea, and hypoxemia beginning before or coincident with initial radiographic abnormalities. One patient had elective tracheal intubation for 4 days until oxygenation improved. All patients were in positive fluid balance at the time of initial radiographic abnormality and were promptly begun on vigorous diuretic therapy. With the exception of one patient who had a mildly decreased ejection fraction, none of the five had clinical evidence of left ventricular failure. High fever without evidence of an extrapulmonary source prompted transbronchial biopsy in three patients. Microscopic examination demonstrated widening of interalveolarsepta by

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

AJR:143, October 1984

PNEUMONIA

AFTER

BONE

MARROW

A

709

TRANSPLANT

B

C

D Fig. 1 -Transient

bone marrow pattern with interstitial

interstitial pneumonia. A, Chest film 10 days after second demonstrates diffuse, bilateral reticular interstitial Radiographic findings are indistinguishable from B, Magnified view showing Kerley B lines. Low-

transplantation Kerley B lines. pulmonary edema.

edema (two of three) and a predominantly interstitial mononuclear cell infiltrate (three of three). Diffuse alveolar damage was evidenced by intraalveolar hyaline membranes and hyperplasia of type II pneumocytes in all three (figs. 1 C and 1 D). Lung tissue sections and cultures revealed no organisms. Two of these patients subsequently developed acute infectious pneumonia and died. Autopsies 14 and 16 days after transbronchial biopsy demonstrated complete resolution of the interstitial and alveolar changes.

(C)and high- (D)power photornicrographs of transbronchial biopsy demonstrate predominantly interstitial inflammatory infiltrate with hyperplastic alveolar lining cells and hyaline membranes. Some alveolar infiltrate is present as well. (H and E, x150 and x600.)

Progressive

Interstitial

Pneumonia

Seven of the 22 patients had a progressive form of interstitial pneumonia presenting various initial radiographic patterns most compatible with a mixedinterstitial and airspace process. As might be expected these patterns were often difficult to classify,and there was disagreement regarding the pattern of lung disease (interstitial, mixed, or airspace) in three of the seven. This represented the only category in which

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

710

WISE

ET AL.

Fig. 2.-Progressive interstitial pneumonia from cytomegalovirus. A, 1 month after bone marrow transplantation. Diffuse, mixed interstitial and acinar pulmonary disease has developed. B, Photomicrograph at autopsy demonstrat-

A

AJR:143,

October

ing two enlarged, desquamated alveolar lining cells with intranuclear (arrows) typical of cytomegalovirus. (H and E, x600.)

1984

inclusions

B

Fig. 3.-Progressive interstitial pneumonia. A, Chest film 7 months after bone marrow transplantation demonstrates atelectasis and airspace pneumonia of right lower lobe. B, 1 week later. Atelectasis has improved, revealing coarse,

reticular, interstitial component

significant interpretive disagreements occurred. Only two of the seven patients had an initial radiographic picture interpreted as a diffuse interstitial process and, unlike the transient form of interstitial disease, Kerley B lines were not identified in these patients. The other five patients showed either airspace (one case) or mixed (four cases) patterns initially. Radiographic abnormalities began as a diffuse, bilateral process in three patients (fig. 2) but were initially focal or

patchy in four (fig.3). Progressive diffuse, mixed, or airspace disease eventually occurred in allpatients, producing respiratory failure and death within 8-25 days (median, 18 days) of the firstradiographic evidence of pulmonary disease. Fever and respiratory failuredominated the clinicalpicture in these patients. The diagnosis was established at autopsy or open lung biopsy in all cases. Histologically, the lungs were character-

accompanying alrspace process involving both lungs. At autopsy, viral inclusion cells indicative of cytomegalovirus or herpes simplex virus infection were identified though viral cultures were negative.

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

AJR:143, October

PNEUMONIA

1984

AFTER

BONE

MARROW

A Fig. 4.-Progressive

idiopathic

inter-

stitial pneumonia. Photomicrograph of autopsy specimen illustrates mixed interstitial and alveolar nature of process. Alveolar spaces contain hyaline membranes, hyperplastic alveolar type II pneumocytes, and macrophages. Interstitial fibrosis and inflammation are also present. The patient had received extensive multiagent chemotherapy before conditioning for transplantation. (H and E, x150.)

711

TRANSPLANT

B

Fig. 5.-A, Invasive pulmonary aspergillosis with bilateral, patchy airspace pulmonary consolidation. septated hyphae of Aspergillus are seen penetrating alveolar walls and small blood vessels. (H and E, x600.)

ized by interstitial edema, fibrosis, and a predominantly mononuclear iterstitial infiltrate. A prominent airspace component was also present and consisted of a combination of alveolar fibrosis, hyaline membranes, hemorrhage, and type II pneumocyte hyperplasia (fig. 4). Four of the seven cases of progressive interstitial pneumonia were of viral origin. Positive viral cultures and inclusion cells were present in three (two cytomegalovirus, one herpes simplex virus). Inclusion cells indicative of infection by either cytomegalovirus or herpes simplex virus were present in the lungs of one patient who had negative viral cultures. Although the etiology of the progressive interstitial pneumonia in the other three patients remains unproven, it may have been caused by drug toxicity and radiation. Two had received extensive chemotherapy before conditioning for transplantation, and the third underwent transplantation twice, receiving conditioning chemotherapy before each transplantation. Of the 22 patients reviewed, radiographic and/or pathologic evidence of either transient or progressive interstitial pulmonary disease was found in nine (75%) of 12 patients who had received both radiation and extensive preconditioning chemotherapy as compared with only two (29%) of seven patients who received neither radiation nor significant preconditioning chemotherapy. Of four patients who received either radiation or preconditioning chemotherapy, but not both, two devel-

oped interstitial the differences

Airspace

lung disease. Because the numbers were not statistically significant.

B,

Thin,

are small

Pneumonia

This heterogeneous category included nine of the 22 epiof pulmonary disease in which pneumonia was caused by bacteria, fungi, or by concurrent infections with two or more organisms. Six of the nine were characterized by focal or patchy airspace disease (fig. 5), which usually became more confluent and widespread. Two of five patients with invasive aspergillosis had multiple pulmonary nodules with cavitation in one. Exceptions to the typical airspace pattern produced by Gram-negative and fungal pneumonias occurred in two patients whose radiographic abnormalities were retrospectively thought to represent interstitial disease. Conversely, one patient who presented with a focal airspace process 2 years after transplantation was found at open lung biopsy to have pneumonia caused by adenovirus (fig. 6). Except for the latter case, these pneumonias occurred 061 days after transplantation and were invariably fatal.Each of four pneumonias occurring during a period of granulocytopenia was from bacteria, fungi, or a combination of both. Four of the five pneumonias caused by concurrent infections with two or more organisms involved Aspergillus and cytosodes

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

712

WISE

AJR:143, October 1984

ET AL.

Fig. 6.-Adenovirus

pneumonia.

A and B, Chest films 2

years

after bone

demonstrate focal alrspace process in left lower lobe. C, Open lung biopsy specimen shows acute intraalveolar inflammatory cell infiltrate, as well as enlarged, dark smudg& cells (straight arrows) and inclusion (curied arrow) typical of adenovirus pneumonia. (H and E, x400.) marrow

megalovirus. While histologic evidence of pulmonary infection cytomegalovirus was irrefutable in all four patients, an inflammatory response to the organism was minimal in three of the four. Most, if not all, macroscopic (and, therefore, radiologic) abnormalities in these four patients were attributable to invasive aspergillosis. with

Discussion

Transient

Interstitial

Pneumonia

Transient interstitial pneumonia occurs within 2-4 weeks of bone marrow transplantation in a considerable number of

transplantation

patients [3, 4]. These early interstitial changes developed within 2 weeks of transplantation in 23% of patients in our series and have been reported to occur in 11 o,/65% of patients [3, 4]. These patients present clinically with mild to severe dyspnea, tachypnea, and hypoxemia. Fever occurred in all of our patients but was not specifically mentioned in other series. The radiographic pattern is characteristic of interstitial pulmonary edema. There are diffuse, bilateral, ground-glass pulmonary opacities often accompanied by Kerley B lines, which are usually not identified in the progressive form of interstitial pneumonia. A diffuse, bilateral airspace component may occur

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

AJR:143,

October

PNEUMONIA

1984

AFTER

BONE

but often differs from the more severe patchy, focal, or acinar shadows of progressive interstitial pneumonia. Cardiomegaly and heart failure are generally absent, although fluid retention usually occurs clinically. Resolution occurs with diuretic therapy, although more slowly (8-17 days in our series) than is usual in uncomplicated pulmonary edema. Previous reports have attributed these transient, early interstitial changes to pulmonary edema [3, 4], but transbronchial lung biopsy in three of our patients demonstrated a round cell interstitial inflammatory infiltrate and edema as well as changes indicative of diffuse alveolar damage. Both the radiographic and histologic changes resolved in about 2 weeks.

Progressive

Interstitial

Pneumonia

Progressive interstitial pneumonia has been reported to occur in over one-fourth of all bone marrow transplant recipients and in up to 50% of patients who achieve functioning bone marrow grafts. It is perhaps the most important complication limiting the long-term survival of these patients [8]. Cytomegalovirus is the most common causative pathogen, being identified in 40%-50% of affected lungs, while in a smaller percentage of cases herpes simplex virus and Pneumocystis carinii have been implicated [8, 9]. Forty percent of cases are idiopathic [8]. There are apparently multiple risk factors that independently influence the likelihood for developing interstitial pneumonia in marrow transplant recipients [10]. The degree of immunosuppression induced in bone marrow transplantation has no parallel in human therapy and is perhaps the major factor allowing the establishment of fatal opportunistic interstitial pneumonia [1 1 ]. Our experience coincides with that of other investigators who have suggested that the incidence of both the transient and progressive forms of nonspecific interstitial pneumonia may be related to chemotherapy and radiation exposure [3, 6, 9, 10, 1 2, 1 3]. Synergistic pulmonary toxicities have been demonstrated with the combined use of radiation and some chemotherapeutic agents [3, 6]. Several investigators have also reported a significantly increased incidence of progressive interstitial pneumonia among those marrow transplant recipients with moderate to severe graftversus-host disease [8-1 1 ]. However, at least one other series, in addition to our own, fails to make such a correlation [3, 4]. Regardless of the etiology, radiologic and histologic findings are identical (except for specific signs such as inclusion cells) and are most characterisic of a mixed interstitial and alveolar process [3, 6, 8, 13, 15]. The term interstitial pneumonia has been applied [6, 15], but does not reflect the prominent airspace component recognized both pathologically and radiographically. Other reports [3, 4] indicate that the irreversible form of interstitial pneumonia occurs in the late posttransplant period (1-4 months after transplantation). However, in our series the disease first developed 2 and 3 weeks after transplantation in three patients with six of the seven cases appearing within 6 weeks of transplantation. One of the seven cases occurred 7 months after transplantation. In one series of 80 patients

MARROW

713

TRANSPLANT

[9] the time of onset was highly variable, dence was 60-70 days posttransplant.

Airspace

though

peak

mci-

Pneumonia

The common airspace pneumonias occurring in bone marrow transplant patients are those induced by Gram-negative bacteria, fungi, and various combinations of organisms [1 4, 5]. Our findings coincide with those in previous series in which infections caused by bacteria or fungi predominated during the first 2 months after transplantation when granulocytopenia is particularly common and severe [4, 5, 8]. Pneumonias caused by multiple organisms, particularly combined infections by Gram-negative bacteria and fungi, are reportedly common in bone marrow transplant recipients and other immunocompromised hosts [1 1 6]. Four of the five mixed pneumonias in our series were caused by the unusual cornbination of Aspergillus and cytornegalovirus. ,

,

Radiographic

Specificity

Radiographic patterns characteristic of a purely interstitial process have been described as typical of the progressive or irreversible interstitial pneumonia that develops in children who have received marrow transplants [5]. This does not coincide with our experience in adults, in whom the initial radiographic abnormalities are most often of a mixed airspace and interstitial pattern. This pattern is often very difficult to categorize radiographically. Regardless of the initial presentation, follow-up films typically demonstrate progressive and diffuse, mixed, or airspace disease. Bacterial, fungal, and mixed pneumonias appear as airspace disease in most cases. Exceptions are the nodular opacities, with or without cavitation, that sometimes develop in invasive aspergillosis, and the few patients with bacterial or fungal pneumonia who have radiographic patterns suggesting interstitial or combined interstitial and airspace disease. Conversely, viral pneumonia other than cytornegalovirus may occur as a well defined, focal airspace process. In summary, our findings coincide with those of other investigators who found the radiographic features of the airspace and progressive interstitial pneumonias to be of only limited specificity [1 1 1 5, 1 6]. Although there is considerable overlap in the radiographic presentation of most pneumonias that develop in bone marrow transplant recipients, the entity that produces the most specific radiographic picture is the transient interstitial pneumonia that appears in the early posttransplant period. In our opinion a patient who presents within the first 2 weeks of transplantation with a radiographic picture suggesting interstitial pulmonary edema with Kerley B lines should be regarded as most likely having a transient, nonspecific, and self-limited form of interstitial pneumonia that does not require biopsy. Diuretics may be of benefit. Elective, tracheal intubation may be necessary when hypoxemia is severe. In our patients and in those of a previous study, early, transient interstitial pneumonia after bone marrow transplantation does not predispose to the development of progressive, fatalinterstitial lung disease [3]. ,

714

WISE

ACKNOWLEDGMENTS

interstitial

Downloaded from www.ajronline.org by 37.44.207.157 on 01/17/17 from IP address 37.44.207.157. Copyright ARRS. For personal use only; all rights reserved

We thank Robert G. Fraser and Robert E. Koehler for assistance in manuscript preparation, Damella Johnson for secretarial help, and Susie Gray for preparation of photographic material.

2. 3.

4.

5.

MJ, Gale RP, Stieham

ER, et al. Bone

marrow

transplan-

tation in man. Ann Intern Med 1975;83:691 -708 Grouse LD, Young RK. Bone marrow transplantation. JAMA 1983;249:2528-2536 Khouri NF, Saral R, Armstrong EM, et al. Pulmonary interstitial changes following bone marrow transplantation. Radiology 1979;1 33:587-592 Pagani JJ, Kangarloo H, Gyepes MT, Feig SA, Falk PM. Radiographic manifestations of bone marrow transplantation in children. AJR 1979;132:883-890 Pagani JJ, Kangarloo H. Chest radiography in pediatric allogeneic bone marrow transplantation. Cancer 1980;46: 1741 -1 745

6. Gockerman JP. Drug-induced Chest Med 1982;3:521-536 7. Morrison DA, Goldman AL: duced lung disease. Radiology 8. Thomas ED, Storb R, Clift RA, (second of two parts). N EngI 9. Neiman PE, Reeves W, Ray

recipients

pneumonia and opportunistic of allogeneic bone marrow

viral infection among grafts. J Infect Dis

1977;1 36:754-767 10. Pino y Torres

JL, Bross

DS, Lam WC, Wharam

MD, Santos

GW,

Order SE. Risk factors in interstitial pneumonitis following allogenic bone marrow transplantation. Int J Radiat Oncol 810/ Phys 1982;8: 1301-1307 1 1 . Winston DJ, Gale RP, Meyer DV, Young LS. Infectious complications of human bone marrow transplantation. Medicine (Balti-

REFERENCES 1. Cline

AJR:143, October 1984

ET AL.

interstitial Radiographic

lung

diseases.

patterns

of

C/in drug-in-

1979;131 :299-304 et al. Bone marrow transplantation 1975;292:895-902 G, et al. A prospective analysis

J Med

of

more) 1979;58: 1-31 12. Meyers JD, Floumoy N, Neiman PE, Thomas ED. Recent advances in bone marrow transplantation. In: 12th annual UCLA symposium on bone marrow transplantation, suppl 7A. 1983: 49

13. Neiman P, Wasserman PB, Wentworth BB, et al. Interstitial pneumonia and cytomegalovirus infection as complications of human marrow transplantation. Transplantation 1973;1 5:478485 14. Beschorner WE, Saral R, Hutchins GM, Tutschka PJ, Santos GW. Lymphocytic bronchitis associated with graft-versus-host disease in recipients of bone marrow transplants. N EngI J Med

1978;299: 1030-1035 15. Meyers JD, Spencer HC, Watts JC, et al. Cytomegalovirus pneumonia after human marrow transplantation. Ann Intern Med 1975;82: 181 -1 88 16. William DM, Krick JA, Remington JS. Pulmonary infection in the compromised host. Parts I and II. In: Lung disease state of the art. New York: American Lung Association, 1977;131-201