- Email: [email protected]
Diffuse alveolar hemorrhage in autologous bone marrow transplant recipients
Diffuse Alveolar Hemorrhage in Autologous Bone Marrow Transplant Recipients RICHARDA. ROBBINS, M.D., JAMESLINDER, M.D., MARLIN G. STAHL, M.D., AUSTIN B. THOMPSON111, M.D., WILLIAMHAIRE, M.D., ANNE KESSINGER,M.D., JAMES0. ARMITAGE,M.D., MARKARNESON, B.S., GAILWOODS, M.D., WILLIAM P. VAUGHAN, M.D., STEPHENI. RENNARD, M.D. omaha, Nebraska
purpose of our work was to evalucomplications in autologous bone marrow transplant recipients. PATIENTS AND METHOD!3 A total of 141 consecutive autologous bone marrow transplant recipients were evaluated. In 29 patients, a clinical syndrome characterized by progressive dyspnea, hypoxia, cough, diffuse consolidation on chest roentgenography, and characteristic bronchoalveolar lavage findings developed over one to seven days. RESULTS: In 29 patients, bronchoalveolar lavage performed by sequential instillation and aspiration of 20-ml aliquots of normal saline resulted in recovered lavage fluid that became progressively bloodier with each recovered aliquot. Autopsy and bronchoalveolar lavage in these patients revealed no pathogens that accounted for the clinical fmdings. Since the later aliquots sample predominantly alveolar material, this syndrome was termed diffuse alveolar hemorrhage (DAB). DAD was associated with a high inpatient mortality rate (23 of 29 died versus 14 of 112 without DAH, p
The
ate pubnonary
a major cause P of morbidity and mortality in patients who receive high-dose chemotherapy and bone marrow transplanulmonary
complications
represent
tation [1,2]. Although these complications may be secondary to infections, progressive malignancy, and/or drug or radiation toxicity, some patients do not have a readily identifiable source of their respiratory compromise [l-8]. The respiratory compromise in these patients has been termed idiopathic interstitial pneumonitis, and in one series was reported to account for 29% of the cases of pneumonia in patients with nonbacterial and nonfungal pneumonia after allogenic bone marrow transplantation for aplastic anemia or hematologic malignancy [a]. Recently, pulmonary hemorrhage has been recognized as a major entity in immunocompromised patients including bone marrow transplant recipients [9111. The relatively recent recognition of this entity was made possible predominantly by the use of bronchoscopy and bronchoalveolar lavage in these patients since hemoptysis is infrequent, roentgenographic findings are variable, and noninvasive tests for pulmonary hemorrhage are often difficult to perform in an immunocompromised patient population [g-11]. High-dose chemotherapy followed by autologous bone marrow transplantation (ARMT) or autologous peripheral stem cell transplantation (APSCT) has recently been increasingly utilized in patients with Hodgkin’s disease, lymphoma, and solid malignancies [12,13]. Therefore, we evaluated 141 consecutive patients undergoing ARMT or APSCT to determine the incidence, mortality, and risk factors associated with the development of pulmonary hemorrhage using bronchoalveolar lavage. A syndrome was identified in 29 of these patients that was associated with dyspnea, hypoxia, cough, and a progressively bloodier return of lavage fluid with sequential instillation and aspiration of 20-ml ahquots of normal saline in at least three subsegmental bronchi. Since these later aliquots represent predominantly alveolar material, the syndrome was termed diffuse alveolar hemorrhage (DAH). Furthermore, DAH could be associated with age over 40 years, treatment of solid malignancies, fever, severe mucositis, white blood cell recovery, and renal insufficiency. These findings suggest that DAH is a frequent occurrence in patients underlying ARMT or APSCT after high-dose chemotherapy. PATIENTS AND METHODS
From the Pulmonary and Critical Care Section and the Hematology and ;;;;zo@ Section, Department of Internal Medicine and the Department of University of Nebraska and Omaha Veterans Adminrstration Medical Y enters, Omaha, Nebraska. This work was supported in part by a grant from the Veterans Administration. Requests for reprints should be addressed to Richard A. Robbins. M.D., Pulmonary and Critical Care Section, Department of Internal Medicine, University of Nebraska Medical Center, 42nd and Dewey Avenue, Omaha, Nebraska 68105. Manuscript submitted March 9. 1989, and accepted in revised form July 31. 1989.
Patient Population
From November 1, 1983, to June 30,1987,141 patients undergoing ARMT or APSCT were evaluated for pulmonary complications using bronchoalveolar lavage. Underlying malignancies in these patients included Hodgkin’s disease (57 patients), non-Hodgkin’s lymphoma (48 patients), or solid malignancies
November
1989
The American
Journal
of Medicine
Volume
87
511
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBINS
(breast carcinoma, malignant melanoma, ovarian carcinoma, sarcoma, or glioblastoma; 36 patients). Clinical Evaluation Routine nursing care studies included vital signs and temperature at least every four hours. Daily weights and the severity of mucositis were recorded [l4]. Pulmonary function testing, including spirometry, lung volume, and diffusing capacity for carbon monoxide, were performed on each patient at the time of admission. Daily complete blood cell counts and platelet counts were assessed. In addition, serum creatinines, blood urea nitrogens, prothrombin times, partial thromboplastin times, chest roentgenograms, and spirometries were performed at least three times weekly. Bronchoscopy and Bronchoalveolar Lavage Patients underwent routine surveillance bronchoscopy and bronchoalveolar lavage after giving written informed consent at the time of admission and within 24 hours of ABMT or APSCT, or if radiographic evidence of consolidation developed. Bronchoscopy and bronchoalveolar lavage were performed as previously described [15-171. Briefly, bronchoscopy was performed by inserting an Olympus lT-R or PD-10 bronchoscope (Olympus Corporation of America, New Hyde Park, New York) through a membrane adapter fitted to the end of the endotracheal tube in intubated patients or transorally in nonintubated patients. Bronchoalveolar lavage was performed by advancing the bronchoscope into Ein area(s) of radiographic abnormality, if present, and sequentially instilling and gently aspirating five, 20-ml aliquots of normal saline in at least three separate subsegments. Since the aspirated bronchoalveolar lavage fluid from the first aliquot represents predominately bronchial material whereas the later aliquots represent predominately alveolar material, the bronchoalveolar lavage fluid from the first aliquots and from the later aliquots was collected and processed separately 1171.After bronchoalveolar lavage, the samples were immediately submitted for bacterial, fungal, mycobacterial, and viral cultures. The bronchoalveolar lavage fluids were filtered through nylon mesh and cell counts were performed on a hemacytometer. Cell cytocentrifuge and membrane filter preparations were prepared and evaluated by Gram stain, auramine-rhodamine stain, Gomori’s methenamine silver stain, Papanicolaou’s stain, Wright’s stain, and a direct fluorescent antibody stain for Legionella species. In addition, Perk+’ iron stain was performed and scored according to the methods of Linder et al [18].
ET AL
herpes simplex virus, varicella zoster virus, adenovirus, respiratory syncytial virus, and influenza virus. Pulmonary malignancy was verified by cytologic examination of bronchoalveolar lavage fluid and/or autopsy. Cardiogenic pulmonary edema was diagnosed by clinical findings, daily weights, and response to appropriate therapy. When all of the aforementioned were excluded, DAH was diagnosed by the following criteria: (1) diffuse consolidation on chest radiographs; (2) hypoxemia requiring supplemental oxygen with or without intubation; (3) bronchoalveolar lavage fluid that became progressively bloodier with each instilled aliquot from at least three separate lobes; and (4) progression or nonclearing of the pulmonary consolidation with correction of any underlying coagulopathy. Statistical Analysis Chi-square analysis was used to compare the age, underlying illness, differing conditioning regimens, and antimicrobial usage between the patients with DAH and those patients who did not develop DAH. Since the mean time of onset of DAH was Day 12 after transplantation, this day was used in those patients who did not develop DAH to compare clinical and laboratory values between the patients with DAH and the patients who did not develop DAH. Clinical (fever, mucositis) and laboratory variables (prothrombin time, partial thromboplastin time, platelet count, serum creatinine, blood urea nitrogen, and white blood cell count) as well as pulmonary function tests were compared by the two-tailed Student’s t test. Probability of survival was estimated by the Kaplan-Meier method [19].
RESULTS
Diagnostic Criteria The diagnoses of bacterial, fungal, or viral infections were established by standard stains and diagnostic criteria of bronchoalveolar lavage fluid, culture of bronchoalveolar lavage fluid, clinical findings, response to appropriate antimicrobial therapy, and/or autopsy. Conventional tube cell viral cultures were inoculated with 0.2 ml of specimen and incubated in 5% carbon dioxide at 37’C. Cell lines used include human foreskin fibroblasta (two tubes), A549 cells, primary monkey kidney cells, and Hep-2 cells. Cultures were examined twice weekly for four weeks for cytopathic effects characteristics of cytomegalovirus,
Illustrative Case Presentation A 47-year-old woman was admitted for treatment of adenocarcinoma of the breast. Ten months prior to admission, the patient presented with abdominal pain and was found to have a pelvic mass on laparoscopy. Examination of biopsy specimens of the pelvic mass and a left breast mass revealed adenocardinoma, and subsequently a modified left radial mastectomy was performed. Adenocarcinoma was also found in a bone marrow biopsy specimen. Therapy with tamoxifen was begun, and she underwent seven peripheral stem cell phereses over the next few months in anticipation of APSCT. The patient had a history of scleroderma with Raynaud’s phenomenon for 10 years and a history of bacterial pneumonia nine months prior to admission. However, there were no symptoms of lung disease and the patient had never smoked. Physical examination at the time of admission was noncontributory except for evidence of a left mastectomy, a well-healed abdominal scar, and skin findings compatible with scleroderma. The admission chest radiograph revealed interstitial prominence, and admission pulmonary function testing showed restrictive pulmonary physiology. Serum creatinine, blood urea nitrogen, white blood cell coutit, and platelet count were normal. A bronchoscopy performed at admission yielded normal results, and no bacteria, fungi, or viruses were identified on smears or culture. The patient received a preparative regimen of cis-
512
97
November
1989
The American
Journal
of Medicine
Volume
ALVEOLAR
HEMORRHAGE
platinum 125 mg/mz intravenously over 24 hours followed by cyclophosphamide 60 mg/kg of body weight intravenously daily for two days. Five total body irradiation treatments of 250 rads over four days were subsequently administered. On the 11th day after admission, the patient underwent APSCT without incident. The patient developed the expected neutropenia. A fever of 38.8”C occurred on the second day after APSCT. Antimicrobials consisting of piperacillin and ceftaxidime were begun, but multiple cultures of blood, urine, and sputum yielded negative results. Fever persisted despite the addition of nafcillin, trimethoprim, sulfamethoxazole, amphotericin B, and vancomycin. Increasingly severe stomatitis developed that eventually required intravenous morphine by continuous infusion. Dyspnea developed on the eighth day after APSCT, accompanied by bilateral basilar crackles on lung auscultation, bilateral pulmonary consolidation on chest radiographs, and progressive hypoxia. Bronchoscopy and bronchoalveolar lavage performed on the ninth day after APSCT revealed moderate to severe bronchitis, small blood clots in the right mainstem, and findings typical of DAH. No organisms were identified on multiple stains of the bronchoalveolar lavage fluid and the fluid was sterile. Numerous hemosiderin-laden macrophages were noted in the bronchoalveolar lavage fluid, as were 4% neutrophils in the bronchial bronchoalveolar lavage fluid and 6% in the alveolar bronchoalveolar lavage fluid despite a peripheral white blood celI count of 200 cells/mms. Multiple platelet transfusions were administered without improvement. The patient was electively endotracheally intubated on the ninth day after APSCT. A gradual rise in the blood urea nitrogen, serum creatinine, and the white blood cell count was noted. Despite aggressive supportive therapy, the patient became progressively hypoxic and hypotensive and died 12 days after APSCT. Autopsy revealed patchy areas of alveolar hemorrhage involving all lung lobes and scattered areas of hyaline membrane formation. No opportunistic organisms were identified. Mortality,
Demography,
and Underlying
TRANSPLANTATION
/ ROBBINS
ET AL
Figure 1. Probability of survival in patients with and without DAH. Probability of survival is on the vertical axis. Time is expressed on the horizontal axis in months after transplantation. Solid circles connected by the solid lines represent the patients with DAH, and the open circles connected by the dotted fines represent the patients without DAH. Small vertical lines on each horizontal line represent the surviving patients at the time of the last follow-up.
TABLE I Age, Underlying Disease, Clinical Characteristics, Pulmonary Function Tests, Radiatlon Therapy, and Laboratory Characteristics of Patients with and without DAH
Age (number of patients 240 years)
DAH Patients (n = 29)
Patients without DAH (n = 112)
16
17
Underlying disease (number of patients) Lymphoma 18 Hodgkin’s lymphoma 8 Non-Hodgkin’s lymphoma Solid malignancies !f Breast Malignant melanoma : Ovarian ; Sarcoma Others 0 Temperature (“C) Weight gain (kg compared with admission weight)
Illness
One hundred forty-one consecutive patients who underwent ABMT or APSCT were evaluated. The incidence of DAH was 21% (29 of 141). DAH was associated with a high mortality rate, accounting for 62% (23 of 37) of the in-hospital deaths occurring in the 141 patients, and was associated with a decreasedprobability of survival (p CO.001, Figure 1). DAH was seen more commonly in older patients (Table I). DAH occurred in 30% of patients over 40 years of age compared with 15% of those under 40 years. Furthermore, 55% of the patients who developed DAH were over 40 years of age. The patients were divided into three groups according to their underlying malignancy (Table I): lymphomas(n= 105), consisting of patients with Hodgkin’s disease(n = 57) or non-Hodgkin’s lymphoma (n = 48); solid malignancies (n = 31), consisting of patients with breast carcinoma (n = 14); malignant melanoma (n = 6), ovarian carcinoma (n = 6), and sarcoma (n = 5); and other malignancies not specified (n = 5). When each solid tumor was evaluated individually, no individual malignancy was associated with an increased
IN GONE MARROW
Mucositis score
p Value <0.05
ii ;: 10 : 3 5
39.3f0.2
38.4fO.l
o.ooQ2
3.2f4.3
1.7f3.9
NS
16.0f0.6
13.8~60.4
0.0009
Adrnsioo pulmonary function tests (perzeet4predicted) FEVt TLC
92f5 99f4 74f4
Ok0 Radiation (number of patients receiving total body or chest irradiation) ‘rothrombin
time (seconds)
‘artial thromboplastin (seconds)
time
‘latelet count (X lOs/mma)
83%3 83f3 94zt3 68*3
2 NS NS
16
39
13.3*0.3
12.9f0.3
NS
34.3k2.2
32.3& 1.3
NS
36f6
12f3
x0.05
N = not significant (p >0.05).
incidence of DAH. However, when grouped together, DAH occurred more commonly in patients with solid malignancies compared with those patients with Hodgkin’s diseaseor non-Hodgkin’s lymphoma (Table I).
November
1989
The American
Journal
of Medicine
Volume
87
513
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBlNS
Clinical Findings The median time of onset of DAH was the 12th day after transplantation and varied between seven days before transplantation to 40 days after transplantation. Interestingly, the two patients who developed DAH before transplantation (seven days prior to and one day prior to) experienced clinical improvement of their disease and neither required intubation. Since development of cardiogenic pulmonary edema might explain the clinical findings in patients with DAH, these patients were carefully evaluated for cardiac dysfunction. No patient developed clinical findings of cardiogenic pulmonary edema. Weight gain is a frequent occurrence after ABMT or APSCT. Weight gain occurred in both the patients with DAH and the controi subjects (Table I), but there was no statistical difference. Pulmonary artery catheters were infrequently inserted due to concerns of infection, prior placement of subclavian catheters, and the frequent occurrence of thrombocytopenia. Although dyspnea is a common complaint in patients undergoing bone marrow transplantation, it is noteworthy that all the patients who developed DAH had either dyspnea, tachypnea, and/or a nonproductive cough, but interestingly, none had hemoptysis prior to the diagnosis of DAH. The maximum mean oral or axillary temperature within 24 hours of developing DAH was 39.3 f 0.2”C. In contrast, the patients who did not develop DAH had maximal oral or axillary temperatures 12 days after bone marrow transulantation (the mean dav of developing DAH) that wh significantly lower (38:4 f O.l”C) (Table I). Pain, erythema, edema, and friability of the oral mucous membranes and tongue (mucositis) often occur after bone marrow transplantation. When the degree of mucositis was quantified [l4], the patients with DAH had more severe mucositis on the day of diagnosis compared with the patients who did not develop DAH 12 days after bone marrow transplantation (Table I).
ET AL
Conditioning Regimen The use of radiation before transplantation appears to influence the incidence of DAH. Patients who received chest radiation, either as total-body irradiation or as chest or mediastinal radiation for localized disease, just before (less than two weeks) ABMT or APSCT (Table I) constituted 52% of the patients with DAH comnared with 35% of the patients without DAH. Chemotherapeutic agents used prior to transplantation included cyclophosphamide, 1,3-bis(2-loroethyl)l-nitrosurea (BCNU), etoposide (VP-16), cytarabine &a-C), cis-platinum, thiotepa (N,N’N-triethylenethiophosphoramide), dacarbazine, doxorubicin, and melphalan. There was no difference in the usage of these chemotherapeutic agents between the patients who did and did not develop DAH (data not shown, p >0.05, all comparisons). Antimicrobial Usage Since certain antimicrobials have been associated with abnormal coagulation and pulmonary dysfunction, antimicrobial usage was examined. All patients received prophylactic acyclovir starting before transplantation. In addition, all patients received additional antimicrobials prior to the onset of DAH or Day 12 after bone marrow transplantation, including penicillin or penicillin derivatives, aminoglycosides, amphotericin B, trimethoprim/sulfamethoxazole, vancomytin, and cephalosporins. The only antimicrobial used with an increased frequency in the patients with DAH was trimethoprim/sulfamethoxaxole, which was utilized in 80% of the patients who developed DAH compared with 59% of the patients who did not develop DAH (x2 = 4.11, p <0.05). The patients with DAH had higher fevers of longer duration (Table I). Trimethoprim/Sulfamethoxazole was routinely added after 48 hours of fever in culture-negative patients whose fever did not respond to initial therapy with aminoglycosides and penicillin derivatives. Therefore, the increased usage of trimethoprim/sulfamethoxazole likely represents a physician response to continual high fevers in the patients with DAH.
Pulmonary Function Testing Pulmonary function testing was performed in all patients before ABMT or APSCT. There was no difference in pretransplantation forced vital capacity (FVC), forced expiratory volume in one second (FEVi), total lung capacity (TLC), or diffusing capacity for carbon monoxide (DL,, single breath) between the patients who developed DAH and those who did not (Table I, p >0.2 all comparisons). Spirometry was performed three times weekly in the majority of patients. However, spirometry was not performed in all patients due to some patients’ inability to cooperate due to severe mucositis. Nevertheless, spirometry was performed in all patients with DAH within two days prior to the onset of DAH or within six days prior to the 12th day after transplantation in those patients who did not develop DAH. Using the last spirometry performed prior to DAH or Day 12, after transplantation in the patients who did not develop DAH, there was a larger average decline in the FVC (38%) and FEVi (36%) in the DAH patients compared with the decline in FVC (18%) and FEVi (16%) in the patients who did not develop DAH (p
laboratory Evaluation There was no difference in the prothrombin time and partial thromboplastin time between the patients with DAH at the time of their diagnosis of DAH or the patients who did not develop DAH at 12 days after bone marrow transplantation (Table I, p >0.02 compared with patients who did not develop DAH). In contrast, the maximal platelet count was elevated in the DAH patients compared with the control subjects (Table I). This could be explained by an increase in the number of platelets transfused (36.2 f 6.0 units) on the day of diagnosis of DAH compared with 12 days after transplantation for the patients who did not develop DAH (11.9 f 0.8 units, p
514
87
Nmwnber
1989
The American
Journal
of Medicine
Volume
ALVEOLAR
Figure 2. Blood urea nitrogen and serum creatinine levels in the patients with DAH and the patients without DAH. Left, creatinine in mg/dL (mean f SD) is on the vertical axis. The days before or after the onset of DAH are on the horizontal axis. Day 0 is the day of onset of DAH or Day 12 after transplantation in the patients without DAH. Solid circles connected by the solid lines represent the creatinine level of the patients with DAH, and open circlesconnected by the dotted lines represent the values in the patients without DAH. Right, similar to left panel except blood urea nitrogen in mg/dL (mean f SD) is on the vertical axis.
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBINS
BEFORE
DAH
ET AL
I I4.--f-.
ued to have progressive respiratory compromise except for the two patients who developed DAH before AHMT or AF’SCT. Blood urea nitrogen was found to be elevated in the patients who developed DAH compared with those who did not develop DAH beginning about four days prior to the onset of DAH or Day 12 after bone marrow transplantation (Figure 2, left). Since blood in tissue sites can result in an elevated blood urea nitrogen, serum creatinine was examined and similar results were observed (Figure 2, right). An association between acute lung injury and a white blood cell influx into the lung has been demonstrated [20,21]. For this reason, the peripheral white blood cell count was examined to determine if white blood cell count recovery correlated with the onset of DAH. When initially examined, there was no relationship between peripheral white blood cell count and the onset of DAH (Figure 3). However, two patients developed DAH before transplantation, and the clinical course of these patients was much lesssevere than that of the other patients. Therefore, the results were reexamined eliminating these two patients and a gradual increase in the peripheral white blood cell count approximately at the onset of DAH was demonstrated (Figure 3). Bronchoalveolar Lavage Studies The bronchial mucosa was carefully inspected for mucosal ulceration at the time of bronchoscopy. No mucosal ulceration was identified, although the mucosa was often erythematous and edematous. Sequential instillation and aspiration of 20-ml aliquots of normal saline resulted in returned bronchoalveolar lavage fluid that was progressively bloodier with each aliquot (Figure 4). There were numerous hemosiderin-laden macrophagesin the majority of patients (Figure 5). Hemosiderin in the cells from the bronchoalveolar lavage fluids was scored according to the methods of Linder et al [18]. The results demonstrated that each specimen had a score of greater than 10 with one exception. Two patients had bronchoalveolar lavage findings similar to those of the patients with DAH. However, the patients differed in that both had aspergillus demonstrated by silver stain and culture, localized areas of consolidation, and bronchoalveolar lavage findings of a progressively bloodier return of bronchoalveolar lavage fluid limited to the areas of consolidation.
2.2 2.02 ,‘ = 0
1.8-
x co i 8
1.6 -
5
1.2-
2 o
l.O-
5 u 0
o.a-
8 2 m
1.4-
0.6-
DAYS
OR AFTER
Figure 3. White blood cell counts in the patients with DAH and the patients without DAH. The white blood cell count in cells X 103/pL (mean f SD) is on the vertical axis and the days before or after the onset of DAH are on the horizontal axis. Day 0 represents the day of onset of DAH or Day 12 after transplantation in the patients without DAH. Solid circles connected by the solid lines represent the values obtained from all patients with DAH, and the open circles connected by the dotted lines represent the values obtained from the patients without DAH. Solld squares connected by the solid line represent the values obtained from the patients with DAH excluding the two patients who developed alveolar hemorrhage at or before transplantation.
Neutrophils were seen on cytocentrifuge preparations of several bronchoalveolar lavages). Ten of 29 had more than 2% neutrophils in their first returned aliquot (bronchial lavage) and 12 of 29 had more than 2% neutrophils in their subsequent four returned aliquota (alveolar lavage). Interestingly, four of the 10 bronchial aliquots and four of the 12 alveolar aliquots were obtained from patients with marked leukopenia (peripheral white blood cell count was 200 cells/mm3 or less).
November
1989
The American
Journal
of Medicine
Volume
87
515
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
BRONCHOALVEOLAR
TRANSPLANTATION
LAVAGE
/ ROBBINS
ALIQUOT
ET AL
Figure 4. Representative returns from a bronchoalveolar lavage obtained from a patient with DAH. The return from each sequential 20-ml aliquot instilled was collected separately and approximately 2 ml was removed and placed in glass tubes.
NUMBER
.
Autopsy Studies Autopsies were performed on 15 of the 29 patients with DAH 12.1 f 3.9 days after the diagnosis was established by bronchoalveolar lavage (Table II). In general, the lungs were heavy (mean right lung, 1,020 g; normal, 360 to 570 g), (mean left lung, 920 g; normal, 320 to 480 g). Typically, fluid exuded from the cut surface of the lung and diffuse hemorrhage were grossly apparent. Inspection of the mucosa confirmed the findings at bronchoscopy. No ulcerations were identified that might account for the bronchoalveolar lavage fluid findings. Microscopically, the degree of alveolar hemorrhage ranged from mild and patchy, to confluent involving all alveoli (Figure 6). All 15 cases had histologic features typical of the proliferative phase of diffuse alveolar damage. Patchy distribution of hyalin membranes was present within alveoli. Proliferating type II cells with large nuclei and prominent nuclei 516
November
1989
The American
Journal
of Medicine
Volume
Figure 5. A, bronchoalveolar lavage in a patient with DAH showing numerous hemosiderin-laden macrophages. Compare this with B, the bronchoalveolar lavage of a normal non-smoker. Both specimens are stained with Perls’ iron stain (original magnification X 100; reduced by 30%).
lined the alveolar septa. In five cases, prominent connective tissue deposition was present within the alveoli septa along with a proliferation of fibroblasts. In only one case were microorganisms found on tissue silver or Gram stains.
COMMENTS Pulmonary complications are frequent in patients undergoing high-dose chemotherapy and bone marrow transplantation [l-11]. The current study demonstrates that a frequent cause of mortality in ABMT recipients is a syndrome characterized by dyspnea, nonproductive cough, hypoxemia, diffuse consolidation on chest radiographs, and characteristic bronchoalveolar lavage findings. This syndrome was recognized because of the application of bronchoalveolar lavage to this patient population and was termed DAH due to the bronchoalveolar lavage findings. 87
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
,’ ROBBINS
ET AL
TABLE II Autopsy Findings in DAH
Patient Number
Residual Pulmonary Malignancy
Alveolar* Hemorrhage
Diffuse* Alveolar Damage
Opportunistic* Pathogens
No
3+
i 6
No No
::
;I 3+
No No No
7
No
3
3
No
Yes
2+ I+
2+ 1+
No
:‘+
:::
1:
No
Other
Doxorubicin Cardiomyopathy Pulmonary Infarct emboli
No
No
Pulmonary Emboli
2
;:
::
No No No No
::
No
3+ I+
3+ 1+
No No
22
No
1+
3+
No
;:
No
:I
::
Pseudomonas aeruginosa No
I :;
No hemorrhage: l+ .= trace, spotty hemorrhage: 2+ = patchy. areas involving 20 or .S -Alveolar ,. ,... 1 vmuse alveolar aamage: I+ = focal areas 01 nyalln memorane tormatlon and reactive prominent reactive pneumocyter 3+ = diffuse hyaiin membrane formation and/or reactive *Opportunistic pathogens: includes microscopic examination and bacterial. fungal. and viral
more alveoli; pneumocytes; pneumocytes cultures.
Staphylococcus epidermidis Sepsis Pulmonary Myocardial
emboli infarction
3+_. = confluent alveolar hemorrhage z+ = scattered hyalln membrane with interstitial edema or proliferative
tormatlon change.
and/or
Figure 6. A, hyalin membrane formation in early diffuse damage. B, proliferative phase of diffuse alveolar damage showing hobnail-shaped type II pneumocytes with interstitial fibrosis and proliferation of interstitial fibroblasts (hematoxylin and eosin, original magnification X 200; reduced by 30%).
The data reported suggestthat no single characteristic identifies those patients who are prone to develop DAH after ABMT or APSCT; furthermore, they suggest that DAH occurs more frequently in older patients with solid malignancies about 12 days after transplantation. Moreover, these patients are more likely to have high fevers, severe mucositis, and renal insufficiency, and DAH correlated with the onset of peripheral white blood cell count recovery. Patients who develop respiratory compromise after allogeneic bone marrow transplantation without an infectious cause have been designated as having idiopathic interstitial pneumonitis [3-81. Idiopathic inter-
stitial pneumonitis has also been reported after ABMT [7]. Several lines of evidence suggestthat DAH may be a frequent subset of idiopathic interstitial pneumonitis after ABMT: (1) the incidence of DAH in the current seriesof 21% approximates the incidences of idiopathic interstitial pneumonia seen after allogeneic bone marrow transplantation (12%) and syngeneic bone marrow transplantation (11%) [8]; (2) the peak incidence of idiopathic interstitial pneumonia occurs within the first seven weeks of allogeneic or syngeneic transplantation, with the highest ratio at Week 2 after transplantation [8]; (3) idiopathic interstitial pneumonitis occurs more frequently with increasing age [8];
November
1989
The American
Journal
of Medicine
Volume
67
517
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
and (4) both DAH and idiopathic interstitial pneumonia are associated with a high mortality rate [8]. Bronchoalveolar lavage has previously been documented to be an effective technique for diagnosing occult pulmonary hemorrhage in the immunocompromised host [g-11]. The current study confirms this capacity of bronchoalveolar lavage in a selected subset of immunocompromised hosts, i.e., patients undergoing ABMT or APSCT. In contrast to previous investigations [ll], it was not possible to demonstrate a readily identifiable association between thrombocytopenia and aspergillus infections with pulmonary hemorrhage. Although thrombocytopenia was present in most patients with DAH before hemorrhage, the platelet counts did not differ between those patients who developed DAH and those who did not. Furthermore, platelet transfusion did not correct the pulmonary hemorrhage, and the majority of patients had progressive hemorrhage despite vigorous platelet transfusions. Aspergillus was not present on silver stain of bronchoalveolar lavage fluid or culture. Autopsies did not reveal Aspergillus species in our patients. Thus, it seems unlikely that either thrombocytopenia or aspergillus infections account for the findings in the majority of these patients. Both renal insufficiency and white blood cell recovery were associated with DAH. It has been suggested that renal failure can result in pulmonary edema in the absence of cardiac insufficiency [22]. Although pulmonary artery catheterization was performed in only a minority of these patients, the lack of significant weight gains compared with the patients without DAH and the absence of clinical signs of cardiac failure suggest that DAH was not due to cardiac insufficiency or fluid overload. Since renal insufficiency affects platelet functions [23], pulmonary hemorrhage could be exacerbated by concurrent renal insufficiency. The reappearance of white blood cells in the circulation at about the onset of respiratory compromise and the presence of neutrophils in the bronchoalveolar lavage of some DAH patients suggest that neutrophil influx into the lung could play a role in the lung injury of some patients with DAH. The neutrophil, by virtue of its broad armamentarium of proteases and oxygen radicals, is capable of initiating and maintaining lung damage [24]. Previous studies have documented that adult respiratory distress syndrome in non-leukopenic patients is associated with an influx of neutrophils into the lung and that this influx correlates with the extent of pulmonary dysfunction [20,21]. Although the neutrophil influx, when present, was modest in patients with DAH, it is possible that pulmonary structures previously damaged by chemotherapy and/or irradiation may be more susceptible to the damaging effects of neutrophils. Bone marrow aplasia has been the major dose-limiting toxicity of antineoplastic chemotherapy. This problem has been largely circumvented by ABMT and
518
November
1989
The American
Journal
of Medicine
Volume
/ ROBBINS
ET AL
thus has allowed the dosages of chemotherapy agents to be escalated. The current data suggest that pulmonary toxicity may be a limiting factor with these increasingly larger dosages of chemotherapy.
ACKNOWLEDGMENT We acknowledge the expert secretarial assistance of Ms. Sheila Peters, and Dr. Kashinath Patil and Ms. Jene Pierson for assistance with the statistical analysis,
REFERENCES 1. Cordonnier C. Bernaudin J-F, Bierling P, Huet Y, Vernant J-P: Pulmonary complications occurring after allogenic bone marrow transplantation. Cancer 1986; 58: 1047-1054. 2. Krowka MJ. Rosenow EC, Hoagland HC: Pulmonary complications of bone marrow transplantation. Chest 1985; 87: 237-246. 3. Neiman PE, Reeves W. Ray G, etafA prospective analysis of interstitial pneumonia and opportunistic viral infection among recipients of allogenic bone marrow grafts. J Infect Dis 1977; 136: 754-767. 4. Bortin MM. Kay HEM. Gale RP. Rimm AA: Factors associated with interstitial pneumonitis after bone-marrow transplantation for acute leukemia. Lancet 1982; I: 437-439. 5. Cardozo BL, Hagenbeck A: Interstitial pneumonitis following bone marrow transplantation: pathogenesis and therapeutic considerations. Eur J Cancer Clin Oncol 1985; 21: 43-51. 6. Appelbaum FR. Meyers JD. Fefer A, et a/: Nonbacterial nonfungal pneumonia following bone marrow transplantation in 100 identical twins. Transplantation 1982: 33: 265-268. 7. Pecego R. Hill R. Appelbaum FR. et at interstitial pneumonitis following autologous bone marrow transplantation. Transplantation 1986; 42: 515-517. 8. Meyers JD. Flournoy N. Thomas ED: Nonbacterial pneumonia after allogenic marrow transplantation: a review of ten years’ experience. Rev Infect Dis 1982; 4: 1119-1132. 9. Drew WL. Finley TN. Golde DW: Diagnostic lavage and occult pulmonary hemorrhage rn thrombocytopenic immunocompromised patients. Am Rev Respir Dis 1977; 116: 215-221. 10. Stover DE, Zaman MB. Hajdn SI. Lange M. Gold J, Armstrong D: Eronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the immunocompromised host. Ann Intern Med 1984; 101: l-7. 11. Kahn FW, Jones JM. England DM: Diagnosis of pulmonary hemorrhage in the immunocompromised host. Am Rev Respir Dis 1987; 136: 155-160. 12. International autologous bone marrow transplant registry: bone-marrow autotransplantation in man. Lancet 1986; II: 96D-962. 13. Kessinger A, Armitage JO, Landmark JD. Smith DM, Weisenburger DD: Autologous peripheral hematopoietic stem cell transplantation restores hematopoietic function following marrow ablative therapy. Blood 1988: 71: 723-727. 14. Eilers J. Berger AM. Petersen MC: Development, testing, and application of the oral assessment guide. Oncology Nursing Forum 1988; 15: 325-330. 15. Reynolds HY. Newball HH: Analysis of proteins and respiratory cells obtained from human lungs by bronchial lavage. J Lab Clin Med 1974; 84: 559-573. 16. Hunninghake GW, Gadek JE. Kawanami 0, Ferrans VJ. Crystal RG: Inflammatory and immune processes in the human lung in health and disease. Am J Pathol 1979; 97: 149-206. 17. Ghafouri MA, Robbins f?A, Rasmussen JK. et al: Use of sequential bronchoalveolar lavage to enrich for “bronchial” and “alveolar” material (abstr). Clin Res 1985; 33: 464A. 18. Linder J, Robbins R. Rennard SI: Cytologic criteria for pulmonary hemorrhage (abstr). Acta Cytol 1988; 32: 763. 19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457-481. 20. Weiland JE. Davis WB. Halter JF. Mohammed JR, Dorinsky PM, Gadek JE: Lung neutrophils in the adult respiratory distress syndrome: clinical and pathophysiologic significance. Am Rev Respir Dis 1986: 133: 218-225. 21. Robbins RA. RUSS WD. Rasmussen JK. Clayton MM: Activation of the comple ment System in the adult respiratory distress syndrome. Am Rev Respir Dis 1987; 135: 651658. 22. HoPPs HC. WfSSler RW: Uremic pneumonitis. Am J Pathol 1955; 31: 266-273. 23. Marcus Al: Platelet function. N Engl J Med 1969; 280: 1278-1280. 24. Martin WJ, Gadek JE. Hunninghake GW. Crystal RG: Oxidant-injury of lung parenchyma Cells. J Clin Invest 1981; 68: 1277-1288.
87
purpose of our work was to evalucomplications in autologous bone marrow transplant recipients. PATIENTS AND METHOD!3 A total of 141 consecutive autologous bone marrow transplant recipients were evaluated. In 29 patients, a clinical syndrome characterized by progressive dyspnea, hypoxia, cough, diffuse consolidation on chest roentgenography, and characteristic bronchoalveolar lavage findings developed over one to seven days. RESULTS: In 29 patients, bronchoalveolar lavage performed by sequential instillation and aspiration of 20-ml aliquots of normal saline resulted in recovered lavage fluid that became progressively bloodier with each recovered aliquot. Autopsy and bronchoalveolar lavage in these patients revealed no pathogens that accounted for the clinical fmdings. Since the later aliquots sample predominantly alveolar material, this syndrome was termed diffuse alveolar hemorrhage (DAB). DAD was associated with a high inpatient mortality rate (23 of 29 died versus 14 of 112 without DAH, p
The
ate pubnonary
a major cause P of morbidity and mortality in patients who receive high-dose chemotherapy and bone marrow transplanulmonary
complications
represent
tation [1,2]. Although these complications may be secondary to infections, progressive malignancy, and/or drug or radiation toxicity, some patients do not have a readily identifiable source of their respiratory compromise [l-8]. The respiratory compromise in these patients has been termed idiopathic interstitial pneumonitis, and in one series was reported to account for 29% of the cases of pneumonia in patients with nonbacterial and nonfungal pneumonia after allogenic bone marrow transplantation for aplastic anemia or hematologic malignancy [a]. Recently, pulmonary hemorrhage has been recognized as a major entity in immunocompromised patients including bone marrow transplant recipients [9111. The relatively recent recognition of this entity was made possible predominantly by the use of bronchoscopy and bronchoalveolar lavage in these patients since hemoptysis is infrequent, roentgenographic findings are variable, and noninvasive tests for pulmonary hemorrhage are often difficult to perform in an immunocompromised patient population [g-11]. High-dose chemotherapy followed by autologous bone marrow transplantation (ARMT) or autologous peripheral stem cell transplantation (APSCT) has recently been increasingly utilized in patients with Hodgkin’s disease, lymphoma, and solid malignancies [12,13]. Therefore, we evaluated 141 consecutive patients undergoing ARMT or APSCT to determine the incidence, mortality, and risk factors associated with the development of pulmonary hemorrhage using bronchoalveolar lavage. A syndrome was identified in 29 of these patients that was associated with dyspnea, hypoxia, cough, and a progressively bloodier return of lavage fluid with sequential instillation and aspiration of 20-ml ahquots of normal saline in at least three subsegmental bronchi. Since these later aliquots represent predominantly alveolar material, the syndrome was termed diffuse alveolar hemorrhage (DAH). Furthermore, DAH could be associated with age over 40 years, treatment of solid malignancies, fever, severe mucositis, white blood cell recovery, and renal insufficiency. These findings suggest that DAH is a frequent occurrence in patients underlying ARMT or APSCT after high-dose chemotherapy. PATIENTS AND METHODS
From the Pulmonary and Critical Care Section and the Hematology and ;;;;zo@ Section, Department of Internal Medicine and the Department of University of Nebraska and Omaha Veterans Adminrstration Medical Y enters, Omaha, Nebraska. This work was supported in part by a grant from the Veterans Administration. Requests for reprints should be addressed to Richard A. Robbins. M.D., Pulmonary and Critical Care Section, Department of Internal Medicine, University of Nebraska Medical Center, 42nd and Dewey Avenue, Omaha, Nebraska 68105. Manuscript submitted March 9. 1989, and accepted in revised form July 31. 1989.
Patient Population
From November 1, 1983, to June 30,1987,141 patients undergoing ARMT or APSCT were evaluated for pulmonary complications using bronchoalveolar lavage. Underlying malignancies in these patients included Hodgkin’s disease (57 patients), non-Hodgkin’s lymphoma (48 patients), or solid malignancies
November
1989
The American
Journal
of Medicine
Volume
87
511
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBINS
(breast carcinoma, malignant melanoma, ovarian carcinoma, sarcoma, or glioblastoma; 36 patients). Clinical Evaluation Routine nursing care studies included vital signs and temperature at least every four hours. Daily weights and the severity of mucositis were recorded [l4]. Pulmonary function testing, including spirometry, lung volume, and diffusing capacity for carbon monoxide, were performed on each patient at the time of admission. Daily complete blood cell counts and platelet counts were assessed. In addition, serum creatinines, blood urea nitrogens, prothrombin times, partial thromboplastin times, chest roentgenograms, and spirometries were performed at least three times weekly. Bronchoscopy and Bronchoalveolar Lavage Patients underwent routine surveillance bronchoscopy and bronchoalveolar lavage after giving written informed consent at the time of admission and within 24 hours of ABMT or APSCT, or if radiographic evidence of consolidation developed. Bronchoscopy and bronchoalveolar lavage were performed as previously described [15-171. Briefly, bronchoscopy was performed by inserting an Olympus lT-R or PD-10 bronchoscope (Olympus Corporation of America, New Hyde Park, New York) through a membrane adapter fitted to the end of the endotracheal tube in intubated patients or transorally in nonintubated patients. Bronchoalveolar lavage was performed by advancing the bronchoscope into Ein area(s) of radiographic abnormality, if present, and sequentially instilling and gently aspirating five, 20-ml aliquots of normal saline in at least three separate subsegments. Since the aspirated bronchoalveolar lavage fluid from the first aliquot represents predominately bronchial material whereas the later aliquots represent predominately alveolar material, the bronchoalveolar lavage fluid from the first aliquots and from the later aliquots was collected and processed separately 1171.After bronchoalveolar lavage, the samples were immediately submitted for bacterial, fungal, mycobacterial, and viral cultures. The bronchoalveolar lavage fluids were filtered through nylon mesh and cell counts were performed on a hemacytometer. Cell cytocentrifuge and membrane filter preparations were prepared and evaluated by Gram stain, auramine-rhodamine stain, Gomori’s methenamine silver stain, Papanicolaou’s stain, Wright’s stain, and a direct fluorescent antibody stain for Legionella species. In addition, Perk+’ iron stain was performed and scored according to the methods of Linder et al [18].
ET AL
herpes simplex virus, varicella zoster virus, adenovirus, respiratory syncytial virus, and influenza virus. Pulmonary malignancy was verified by cytologic examination of bronchoalveolar lavage fluid and/or autopsy. Cardiogenic pulmonary edema was diagnosed by clinical findings, daily weights, and response to appropriate therapy. When all of the aforementioned were excluded, DAH was diagnosed by the following criteria: (1) diffuse consolidation on chest radiographs; (2) hypoxemia requiring supplemental oxygen with or without intubation; (3) bronchoalveolar lavage fluid that became progressively bloodier with each instilled aliquot from at least three separate lobes; and (4) progression or nonclearing of the pulmonary consolidation with correction of any underlying coagulopathy. Statistical Analysis Chi-square analysis was used to compare the age, underlying illness, differing conditioning regimens, and antimicrobial usage between the patients with DAH and those patients who did not develop DAH. Since the mean time of onset of DAH was Day 12 after transplantation, this day was used in those patients who did not develop DAH to compare clinical and laboratory values between the patients with DAH and the patients who did not develop DAH. Clinical (fever, mucositis) and laboratory variables (prothrombin time, partial thromboplastin time, platelet count, serum creatinine, blood urea nitrogen, and white blood cell count) as well as pulmonary function tests were compared by the two-tailed Student’s t test. Probability of survival was estimated by the Kaplan-Meier method [19].
RESULTS
Diagnostic Criteria The diagnoses of bacterial, fungal, or viral infections were established by standard stains and diagnostic criteria of bronchoalveolar lavage fluid, culture of bronchoalveolar lavage fluid, clinical findings, response to appropriate antimicrobial therapy, and/or autopsy. Conventional tube cell viral cultures were inoculated with 0.2 ml of specimen and incubated in 5% carbon dioxide at 37’C. Cell lines used include human foreskin fibroblasta (two tubes), A549 cells, primary monkey kidney cells, and Hep-2 cells. Cultures were examined twice weekly for four weeks for cytopathic effects characteristics of cytomegalovirus,
Illustrative Case Presentation A 47-year-old woman was admitted for treatment of adenocarcinoma of the breast. Ten months prior to admission, the patient presented with abdominal pain and was found to have a pelvic mass on laparoscopy. Examination of biopsy specimens of the pelvic mass and a left breast mass revealed adenocardinoma, and subsequently a modified left radial mastectomy was performed. Adenocarcinoma was also found in a bone marrow biopsy specimen. Therapy with tamoxifen was begun, and she underwent seven peripheral stem cell phereses over the next few months in anticipation of APSCT. The patient had a history of scleroderma with Raynaud’s phenomenon for 10 years and a history of bacterial pneumonia nine months prior to admission. However, there were no symptoms of lung disease and the patient had never smoked. Physical examination at the time of admission was noncontributory except for evidence of a left mastectomy, a well-healed abdominal scar, and skin findings compatible with scleroderma. The admission chest radiograph revealed interstitial prominence, and admission pulmonary function testing showed restrictive pulmonary physiology. Serum creatinine, blood urea nitrogen, white blood cell coutit, and platelet count were normal. A bronchoscopy performed at admission yielded normal results, and no bacteria, fungi, or viruses were identified on smears or culture. The patient received a preparative regimen of cis-
512
97
November
1989
The American
Journal
of Medicine
Volume
ALVEOLAR
HEMORRHAGE
platinum 125 mg/mz intravenously over 24 hours followed by cyclophosphamide 60 mg/kg of body weight intravenously daily for two days. Five total body irradiation treatments of 250 rads over four days were subsequently administered. On the 11th day after admission, the patient underwent APSCT without incident. The patient developed the expected neutropenia. A fever of 38.8”C occurred on the second day after APSCT. Antimicrobials consisting of piperacillin and ceftaxidime were begun, but multiple cultures of blood, urine, and sputum yielded negative results. Fever persisted despite the addition of nafcillin, trimethoprim, sulfamethoxazole, amphotericin B, and vancomycin. Increasingly severe stomatitis developed that eventually required intravenous morphine by continuous infusion. Dyspnea developed on the eighth day after APSCT, accompanied by bilateral basilar crackles on lung auscultation, bilateral pulmonary consolidation on chest radiographs, and progressive hypoxia. Bronchoscopy and bronchoalveolar lavage performed on the ninth day after APSCT revealed moderate to severe bronchitis, small blood clots in the right mainstem, and findings typical of DAH. No organisms were identified on multiple stains of the bronchoalveolar lavage fluid and the fluid was sterile. Numerous hemosiderin-laden macrophages were noted in the bronchoalveolar lavage fluid, as were 4% neutrophils in the bronchial bronchoalveolar lavage fluid and 6% in the alveolar bronchoalveolar lavage fluid despite a peripheral white blood celI count of 200 cells/mms. Multiple platelet transfusions were administered without improvement. The patient was electively endotracheally intubated on the ninth day after APSCT. A gradual rise in the blood urea nitrogen, serum creatinine, and the white blood cell count was noted. Despite aggressive supportive therapy, the patient became progressively hypoxic and hypotensive and died 12 days after APSCT. Autopsy revealed patchy areas of alveolar hemorrhage involving all lung lobes and scattered areas of hyaline membrane formation. No opportunistic organisms were identified. Mortality,
Demography,
and Underlying
TRANSPLANTATION
/ ROBBINS
ET AL
Figure 1. Probability of survival in patients with and without DAH. Probability of survival is on the vertical axis. Time is expressed on the horizontal axis in months after transplantation. Solid circles connected by the solid lines represent the patients with DAH, and the open circles connected by the dotted fines represent the patients without DAH. Small vertical lines on each horizontal line represent the surviving patients at the time of the last follow-up.
TABLE I Age, Underlying Disease, Clinical Characteristics, Pulmonary Function Tests, Radiatlon Therapy, and Laboratory Characteristics of Patients with and without DAH
Age (number of patients 240 years)
DAH Patients (n = 29)
Patients without DAH (n = 112)
16
17
Underlying disease (number of patients) Lymphoma 18 Hodgkin’s lymphoma 8 Non-Hodgkin’s lymphoma Solid malignancies !f Breast Malignant melanoma : Ovarian ; Sarcoma Others 0 Temperature (“C) Weight gain (kg compared with admission weight)
Illness
One hundred forty-one consecutive patients who underwent ABMT or APSCT were evaluated. The incidence of DAH was 21% (29 of 141). DAH was associated with a high mortality rate, accounting for 62% (23 of 37) of the in-hospital deaths occurring in the 141 patients, and was associated with a decreasedprobability of survival (p CO.001, Figure 1). DAH was seen more commonly in older patients (Table I). DAH occurred in 30% of patients over 40 years of age compared with 15% of those under 40 years. Furthermore, 55% of the patients who developed DAH were over 40 years of age. The patients were divided into three groups according to their underlying malignancy (Table I): lymphomas(n= 105), consisting of patients with Hodgkin’s disease(n = 57) or non-Hodgkin’s lymphoma (n = 48); solid malignancies (n = 31), consisting of patients with breast carcinoma (n = 14); malignant melanoma (n = 6), ovarian carcinoma (n = 6), and sarcoma (n = 5); and other malignancies not specified (n = 5). When each solid tumor was evaluated individually, no individual malignancy was associated with an increased
IN GONE MARROW
Mucositis score
p Value <0.05
ii ;: 10 : 3 5
39.3f0.2
38.4fO.l
o.ooQ2
3.2f4.3
1.7f3.9
NS
16.0f0.6
13.8~60.4
0.0009
Adrnsioo pulmonary function tests (perzeet4predicted) FEVt TLC
92f5 99f4 74f4
Ok0 Radiation (number of patients receiving total body or chest irradiation) ‘rothrombin
time (seconds)
‘artial thromboplastin (seconds)
time
‘latelet count (X lOs/mma)
83%3 83f3 94zt3 68*3
2 NS NS
16
39
13.3*0.3
12.9f0.3
NS
34.3k2.2
32.3& 1.3
NS
36f6
12f3
x0.05
N = not significant (p >0.05).
incidence of DAH. However, when grouped together, DAH occurred more commonly in patients with solid malignancies compared with those patients with Hodgkin’s diseaseor non-Hodgkin’s lymphoma (Table I).
November
1989
The American
Journal
of Medicine
Volume
87
513
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBlNS
Clinical Findings The median time of onset of DAH was the 12th day after transplantation and varied between seven days before transplantation to 40 days after transplantation. Interestingly, the two patients who developed DAH before transplantation (seven days prior to and one day prior to) experienced clinical improvement of their disease and neither required intubation. Since development of cardiogenic pulmonary edema might explain the clinical findings in patients with DAH, these patients were carefully evaluated for cardiac dysfunction. No patient developed clinical findings of cardiogenic pulmonary edema. Weight gain is a frequent occurrence after ABMT or APSCT. Weight gain occurred in both the patients with DAH and the controi subjects (Table I), but there was no statistical difference. Pulmonary artery catheters were infrequently inserted due to concerns of infection, prior placement of subclavian catheters, and the frequent occurrence of thrombocytopenia. Although dyspnea is a common complaint in patients undergoing bone marrow transplantation, it is noteworthy that all the patients who developed DAH had either dyspnea, tachypnea, and/or a nonproductive cough, but interestingly, none had hemoptysis prior to the diagnosis of DAH. The maximum mean oral or axillary temperature within 24 hours of developing DAH was 39.3 f 0.2”C. In contrast, the patients who did not develop DAH had maximal oral or axillary temperatures 12 days after bone marrow transulantation (the mean dav of developing DAH) that wh significantly lower (38:4 f O.l”C) (Table I). Pain, erythema, edema, and friability of the oral mucous membranes and tongue (mucositis) often occur after bone marrow transplantation. When the degree of mucositis was quantified [l4], the patients with DAH had more severe mucositis on the day of diagnosis compared with the patients who did not develop DAH 12 days after bone marrow transplantation (Table I).
ET AL
Conditioning Regimen The use of radiation before transplantation appears to influence the incidence of DAH. Patients who received chest radiation, either as total-body irradiation or as chest or mediastinal radiation for localized disease, just before (less than two weeks) ABMT or APSCT (Table I) constituted 52% of the patients with DAH comnared with 35% of the patients without DAH. Chemotherapeutic agents used prior to transplantation included cyclophosphamide, 1,3-bis(2-loroethyl)l-nitrosurea (BCNU), etoposide (VP-16), cytarabine &a-C), cis-platinum, thiotepa (N,N’N-triethylenethiophosphoramide), dacarbazine, doxorubicin, and melphalan. There was no difference in the usage of these chemotherapeutic agents between the patients who did and did not develop DAH (data not shown, p >0.05, all comparisons). Antimicrobial Usage Since certain antimicrobials have been associated with abnormal coagulation and pulmonary dysfunction, antimicrobial usage was examined. All patients received prophylactic acyclovir starting before transplantation. In addition, all patients received additional antimicrobials prior to the onset of DAH or Day 12 after bone marrow transplantation, including penicillin or penicillin derivatives, aminoglycosides, amphotericin B, trimethoprim/sulfamethoxazole, vancomytin, and cephalosporins. The only antimicrobial used with an increased frequency in the patients with DAH was trimethoprim/sulfamethoxaxole, which was utilized in 80% of the patients who developed DAH compared with 59% of the patients who did not develop DAH (x2 = 4.11, p <0.05). The patients with DAH had higher fevers of longer duration (Table I). Trimethoprim/Sulfamethoxazole was routinely added after 48 hours of fever in culture-negative patients whose fever did not respond to initial therapy with aminoglycosides and penicillin derivatives. Therefore, the increased usage of trimethoprim/sulfamethoxazole likely represents a physician response to continual high fevers in the patients with DAH.
Pulmonary Function Testing Pulmonary function testing was performed in all patients before ABMT or APSCT. There was no difference in pretransplantation forced vital capacity (FVC), forced expiratory volume in one second (FEVi), total lung capacity (TLC), or diffusing capacity for carbon monoxide (DL,, single breath) between the patients who developed DAH and those who did not (Table I, p >0.2 all comparisons). Spirometry was performed three times weekly in the majority of patients. However, spirometry was not performed in all patients due to some patients’ inability to cooperate due to severe mucositis. Nevertheless, spirometry was performed in all patients with DAH within two days prior to the onset of DAH or within six days prior to the 12th day after transplantation in those patients who did not develop DAH. Using the last spirometry performed prior to DAH or Day 12, after transplantation in the patients who did not develop DAH, there was a larger average decline in the FVC (38%) and FEVi (36%) in the DAH patients compared with the decline in FVC (18%) and FEVi (16%) in the patients who did not develop DAH (p
laboratory Evaluation There was no difference in the prothrombin time and partial thromboplastin time between the patients with DAH at the time of their diagnosis of DAH or the patients who did not develop DAH at 12 days after bone marrow transplantation (Table I, p >0.02 compared with patients who did not develop DAH). In contrast, the maximal platelet count was elevated in the DAH patients compared with the control subjects (Table I). This could be explained by an increase in the number of platelets transfused (36.2 f 6.0 units) on the day of diagnosis of DAH compared with 12 days after transplantation for the patients who did not develop DAH (11.9 f 0.8 units, p
514
87
Nmwnber
1989
The American
Journal
of Medicine
Volume
ALVEOLAR
Figure 2. Blood urea nitrogen and serum creatinine levels in the patients with DAH and the patients without DAH. Left, creatinine in mg/dL (mean f SD) is on the vertical axis. The days before or after the onset of DAH are on the horizontal axis. Day 0 is the day of onset of DAH or Day 12 after transplantation in the patients without DAH. Solid circles connected by the solid lines represent the creatinine level of the patients with DAH, and open circlesconnected by the dotted lines represent the values in the patients without DAH. Right, similar to left panel except blood urea nitrogen in mg/dL (mean f SD) is on the vertical axis.
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
/ ROBBINS
BEFORE
DAH
ET AL
I I4.--f-.
ued to have progressive respiratory compromise except for the two patients who developed DAH before AHMT or AF’SCT. Blood urea nitrogen was found to be elevated in the patients who developed DAH compared with those who did not develop DAH beginning about four days prior to the onset of DAH or Day 12 after bone marrow transplantation (Figure 2, left). Since blood in tissue sites can result in an elevated blood urea nitrogen, serum creatinine was examined and similar results were observed (Figure 2, right). An association between acute lung injury and a white blood cell influx into the lung has been demonstrated [20,21]. For this reason, the peripheral white blood cell count was examined to determine if white blood cell count recovery correlated with the onset of DAH. When initially examined, there was no relationship between peripheral white blood cell count and the onset of DAH (Figure 3). However, two patients developed DAH before transplantation, and the clinical course of these patients was much lesssevere than that of the other patients. Therefore, the results were reexamined eliminating these two patients and a gradual increase in the peripheral white blood cell count approximately at the onset of DAH was demonstrated (Figure 3). Bronchoalveolar Lavage Studies The bronchial mucosa was carefully inspected for mucosal ulceration at the time of bronchoscopy. No mucosal ulceration was identified, although the mucosa was often erythematous and edematous. Sequential instillation and aspiration of 20-ml aliquots of normal saline resulted in returned bronchoalveolar lavage fluid that was progressively bloodier with each aliquot (Figure 4). There were numerous hemosiderin-laden macrophagesin the majority of patients (Figure 5). Hemosiderin in the cells from the bronchoalveolar lavage fluids was scored according to the methods of Linder et al [18]. The results demonstrated that each specimen had a score of greater than 10 with one exception. Two patients had bronchoalveolar lavage findings similar to those of the patients with DAH. However, the patients differed in that both had aspergillus demonstrated by silver stain and culture, localized areas of consolidation, and bronchoalveolar lavage findings of a progressively bloodier return of bronchoalveolar lavage fluid limited to the areas of consolidation.
2.2 2.02 ,‘ = 0
1.8-
x co i 8
1.6 -
5
1.2-
2 o
l.O-
5 u 0
o.a-
8 2 m
1.4-
0.6-
DAYS
OR AFTER
Figure 3. White blood cell counts in the patients with DAH and the patients without DAH. The white blood cell count in cells X 103/pL (mean f SD) is on the vertical axis and the days before or after the onset of DAH are on the horizontal axis. Day 0 represents the day of onset of DAH or Day 12 after transplantation in the patients without DAH. Solid circles connected by the solid lines represent the values obtained from all patients with DAH, and the open circles connected by the dotted lines represent the values obtained from the patients without DAH. Solld squares connected by the solid line represent the values obtained from the patients with DAH excluding the two patients who developed alveolar hemorrhage at or before transplantation.
Neutrophils were seen on cytocentrifuge preparations of several bronchoalveolar lavages). Ten of 29 had more than 2% neutrophils in their first returned aliquot (bronchial lavage) and 12 of 29 had more than 2% neutrophils in their subsequent four returned aliquota (alveolar lavage). Interestingly, four of the 10 bronchial aliquots and four of the 12 alveolar aliquots were obtained from patients with marked leukopenia (peripheral white blood cell count was 200 cells/mm3 or less).
November
1989
The American
Journal
of Medicine
Volume
87
515
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
BRONCHOALVEOLAR
TRANSPLANTATION
LAVAGE
/ ROBBINS
ALIQUOT
ET AL
Figure 4. Representative returns from a bronchoalveolar lavage obtained from a patient with DAH. The return from each sequential 20-ml aliquot instilled was collected separately and approximately 2 ml was removed and placed in glass tubes.
NUMBER
.
Autopsy Studies Autopsies were performed on 15 of the 29 patients with DAH 12.1 f 3.9 days after the diagnosis was established by bronchoalveolar lavage (Table II). In general, the lungs were heavy (mean right lung, 1,020 g; normal, 360 to 570 g), (mean left lung, 920 g; normal, 320 to 480 g). Typically, fluid exuded from the cut surface of the lung and diffuse hemorrhage were grossly apparent. Inspection of the mucosa confirmed the findings at bronchoscopy. No ulcerations were identified that might account for the bronchoalveolar lavage fluid findings. Microscopically, the degree of alveolar hemorrhage ranged from mild and patchy, to confluent involving all alveoli (Figure 6). All 15 cases had histologic features typical of the proliferative phase of diffuse alveolar damage. Patchy distribution of hyalin membranes was present within alveoli. Proliferating type II cells with large nuclei and prominent nuclei 516
November
1989
The American
Journal
of Medicine
Volume
Figure 5. A, bronchoalveolar lavage in a patient with DAH showing numerous hemosiderin-laden macrophages. Compare this with B, the bronchoalveolar lavage of a normal non-smoker. Both specimens are stained with Perls’ iron stain (original magnification X 100; reduced by 30%).
lined the alveolar septa. In five cases, prominent connective tissue deposition was present within the alveoli septa along with a proliferation of fibroblasts. In only one case were microorganisms found on tissue silver or Gram stains.
COMMENTS Pulmonary complications are frequent in patients undergoing high-dose chemotherapy and bone marrow transplantation [l-11]. The current study demonstrates that a frequent cause of mortality in ABMT recipients is a syndrome characterized by dyspnea, nonproductive cough, hypoxemia, diffuse consolidation on chest radiographs, and characteristic bronchoalveolar lavage findings. This syndrome was recognized because of the application of bronchoalveolar lavage to this patient population and was termed DAH due to the bronchoalveolar lavage findings. 87
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
,’ ROBBINS
ET AL
TABLE II Autopsy Findings in DAH
Patient Number
Residual Pulmonary Malignancy
Alveolar* Hemorrhage
Diffuse* Alveolar Damage
Opportunistic* Pathogens
No
3+
i 6
No No
::
;I 3+
No No No
7
No
3
3
No
Yes
2+ I+
2+ 1+
No
:‘+
:::
1:
No
Other
Doxorubicin Cardiomyopathy Pulmonary Infarct emboli
No
No
Pulmonary Emboli
2
;:
::
No No No No
::
No
3+ I+
3+ 1+
No No
22
No
1+
3+
No
;:
No
:I
::
Pseudomonas aeruginosa No
I :;
No hemorrhage: l+ .= trace, spotty hemorrhage: 2+ = patchy. areas involving 20 or .S -Alveolar ,. ,... 1 vmuse alveolar aamage: I+ = focal areas 01 nyalln memorane tormatlon and reactive prominent reactive pneumocyter 3+ = diffuse hyaiin membrane formation and/or reactive *Opportunistic pathogens: includes microscopic examination and bacterial. fungal. and viral
more alveoli; pneumocytes; pneumocytes cultures.
Staphylococcus epidermidis Sepsis Pulmonary Myocardial
emboli infarction
3+_. = confluent alveolar hemorrhage z+ = scattered hyalln membrane with interstitial edema or proliferative
tormatlon change.
and/or
Figure 6. A, hyalin membrane formation in early diffuse damage. B, proliferative phase of diffuse alveolar damage showing hobnail-shaped type II pneumocytes with interstitial fibrosis and proliferation of interstitial fibroblasts (hematoxylin and eosin, original magnification X 200; reduced by 30%).
The data reported suggestthat no single characteristic identifies those patients who are prone to develop DAH after ABMT or APSCT; furthermore, they suggest that DAH occurs more frequently in older patients with solid malignancies about 12 days after transplantation. Moreover, these patients are more likely to have high fevers, severe mucositis, and renal insufficiency, and DAH correlated with the onset of peripheral white blood cell count recovery. Patients who develop respiratory compromise after allogeneic bone marrow transplantation without an infectious cause have been designated as having idiopathic interstitial pneumonitis [3-81. Idiopathic inter-
stitial pneumonitis has also been reported after ABMT [7]. Several lines of evidence suggestthat DAH may be a frequent subset of idiopathic interstitial pneumonitis after ABMT: (1) the incidence of DAH in the current seriesof 21% approximates the incidences of idiopathic interstitial pneumonia seen after allogeneic bone marrow transplantation (12%) and syngeneic bone marrow transplantation (11%) [8]; (2) the peak incidence of idiopathic interstitial pneumonia occurs within the first seven weeks of allogeneic or syngeneic transplantation, with the highest ratio at Week 2 after transplantation [8]; (3) idiopathic interstitial pneumonitis occurs more frequently with increasing age [8];
November
1989
The American
Journal
of Medicine
Volume
67
517
ALVEOLAR
HEMORRHAGE
IN BONE MARROW
TRANSPLANTATION
and (4) both DAH and idiopathic interstitial pneumonia are associated with a high mortality rate [8]. Bronchoalveolar lavage has previously been documented to be an effective technique for diagnosing occult pulmonary hemorrhage in the immunocompromised host [g-11]. The current study confirms this capacity of bronchoalveolar lavage in a selected subset of immunocompromised hosts, i.e., patients undergoing ABMT or APSCT. In contrast to previous investigations [ll], it was not possible to demonstrate a readily identifiable association between thrombocytopenia and aspergillus infections with pulmonary hemorrhage. Although thrombocytopenia was present in most patients with DAH before hemorrhage, the platelet counts did not differ between those patients who developed DAH and those who did not. Furthermore, platelet transfusion did not correct the pulmonary hemorrhage, and the majority of patients had progressive hemorrhage despite vigorous platelet transfusions. Aspergillus was not present on silver stain of bronchoalveolar lavage fluid or culture. Autopsies did not reveal Aspergillus species in our patients. Thus, it seems unlikely that either thrombocytopenia or aspergillus infections account for the findings in the majority of these patients. Both renal insufficiency and white blood cell recovery were associated with DAH. It has been suggested that renal failure can result in pulmonary edema in the absence of cardiac insufficiency [22]. Although pulmonary artery catheterization was performed in only a minority of these patients, the lack of significant weight gains compared with the patients without DAH and the absence of clinical signs of cardiac failure suggest that DAH was not due to cardiac insufficiency or fluid overload. Since renal insufficiency affects platelet functions [23], pulmonary hemorrhage could be exacerbated by concurrent renal insufficiency. The reappearance of white blood cells in the circulation at about the onset of respiratory compromise and the presence of neutrophils in the bronchoalveolar lavage of some DAH patients suggest that neutrophil influx into the lung could play a role in the lung injury of some patients with DAH. The neutrophil, by virtue of its broad armamentarium of proteases and oxygen radicals, is capable of initiating and maintaining lung damage [24]. Previous studies have documented that adult respiratory distress syndrome in non-leukopenic patients is associated with an influx of neutrophils into the lung and that this influx correlates with the extent of pulmonary dysfunction [20,21]. Although the neutrophil influx, when present, was modest in patients with DAH, it is possible that pulmonary structures previously damaged by chemotherapy and/or irradiation may be more susceptible to the damaging effects of neutrophils. Bone marrow aplasia has been the major dose-limiting toxicity of antineoplastic chemotherapy. This problem has been largely circumvented by ABMT and
518
November
1989
The American
Journal
of Medicine
Volume
/ ROBBINS
ET AL
thus has allowed the dosages of chemotherapy agents to be escalated. The current data suggest that pulmonary toxicity may be a limiting factor with these increasingly larger dosages of chemotherapy.
ACKNOWLEDGMENT We acknowledge the expert secretarial assistance of Ms. Sheila Peters, and Dr. Kashinath Patil and Ms. Jene Pierson for assistance with the statistical analysis,
REFERENCES 1. Cordonnier C. Bernaudin J-F, Bierling P, Huet Y, Vernant J-P: Pulmonary complications occurring after allogenic bone marrow transplantation. Cancer 1986; 58: 1047-1054. 2. Krowka MJ. Rosenow EC, Hoagland HC: Pulmonary complications of bone marrow transplantation. Chest 1985; 87: 237-246. 3. Neiman PE, Reeves W. Ray G, etafA prospective analysis of interstitial pneumonia and opportunistic viral infection among recipients of allogenic bone marrow grafts. J Infect Dis 1977; 136: 754-767. 4. Bortin MM. Kay HEM. Gale RP. Rimm AA: Factors associated with interstitial pneumonitis after bone-marrow transplantation for acute leukemia. Lancet 1982; I: 437-439. 5. Cardozo BL, Hagenbeck A: Interstitial pneumonitis following bone marrow transplantation: pathogenesis and therapeutic considerations. Eur J Cancer Clin Oncol 1985; 21: 43-51. 6. Appelbaum FR. Meyers JD. Fefer A, et a/: Nonbacterial nonfungal pneumonia following bone marrow transplantation in 100 identical twins. Transplantation 1982: 33: 265-268. 7. Pecego R. Hill R. Appelbaum FR. et at interstitial pneumonitis following autologous bone marrow transplantation. Transplantation 1986; 42: 515-517. 8. Meyers JD. Flournoy N. Thomas ED: Nonbacterial pneumonia after allogenic marrow transplantation: a review of ten years’ experience. Rev Infect Dis 1982; 4: 1119-1132. 9. Drew WL. Finley TN. Golde DW: Diagnostic lavage and occult pulmonary hemorrhage rn thrombocytopenic immunocompromised patients. Am Rev Respir Dis 1977; 116: 215-221. 10. Stover DE, Zaman MB. Hajdn SI. Lange M. Gold J, Armstrong D: Eronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the immunocompromised host. Ann Intern Med 1984; 101: l-7. 11. Kahn FW, Jones JM. England DM: Diagnosis of pulmonary hemorrhage in the immunocompromised host. Am Rev Respir Dis 1987; 136: 155-160. 12. International autologous bone marrow transplant registry: bone-marrow autotransplantation in man. Lancet 1986; II: 96D-962. 13. Kessinger A, Armitage JO, Landmark JD. Smith DM, Weisenburger DD: Autologous peripheral hematopoietic stem cell transplantation restores hematopoietic function following marrow ablative therapy. Blood 1988: 71: 723-727. 14. Eilers J. Berger AM. Petersen MC: Development, testing, and application of the oral assessment guide. Oncology Nursing Forum 1988; 15: 325-330. 15. Reynolds HY. Newball HH: Analysis of proteins and respiratory cells obtained from human lungs by bronchial lavage. J Lab Clin Med 1974; 84: 559-573. 16. Hunninghake GW, Gadek JE. Kawanami 0, Ferrans VJ. Crystal RG: Inflammatory and immune processes in the human lung in health and disease. Am J Pathol 1979; 97: 149-206. 17. Ghafouri MA, Robbins f?A, Rasmussen JK. et al: Use of sequential bronchoalveolar lavage to enrich for “bronchial” and “alveolar” material (abstr). Clin Res 1985; 33: 464A. 18. Linder J, Robbins R. Rennard SI: Cytologic criteria for pulmonary hemorrhage (abstr). Acta Cytol 1988; 32: 763. 19. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457-481. 20. Weiland JE. Davis WB. Halter JF. Mohammed JR, Dorinsky PM, Gadek JE: Lung neutrophils in the adult respiratory distress syndrome: clinical and pathophysiologic significance. Am Rev Respir Dis 1986: 133: 218-225. 21. Robbins RA. RUSS WD. Rasmussen JK. Clayton MM: Activation of the comple ment System in the adult respiratory distress syndrome. Am Rev Respir Dis 1987; 135: 651658. 22. HoPPs HC. WfSSler RW: Uremic pneumonitis. Am J Pathol 1955; 31: 266-273. 23. Marcus Al: Platelet function. N Engl J Med 1969; 280: 1278-1280. 24. Martin WJ, Gadek JE. Hunninghake GW. Crystal RG: Oxidant-injury of lung parenchyma Cells. J Clin Invest 1981; 68: 1277-1288.
87