INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The management of breast carcinoma is partly based on radiation therapy. Because the lung is adjacent to the target of the radiation therapy, it is susceptible to being injured by the treatment (1). Radiation-induced lung injury may result in pneumonitis and fibrosis. Radiation-induced pneumonitis occurs classically 4 to 12 wk after completion of radiotherapy and is characterized by fever, dry cough, dyspnea, and alveolar opacities confined to the treatment port (2). However, bilateral lymphocytic alveolitis develops in both lung fields in most patients after strictly unilateral breast irradiation (3).

Bronchiolitis obliterans organizing pneumonia (BOOP) is a distinct clinicopathologic entity characterized clinically by a flulike illness with cough, crackles, and patchy and often migratory peripheral air-space infiltrates on chest radiographs (4). BOOP is defined histopathologically as granulation tissue plugs within the lumen of distal air spaces (bronchioles, alveolar ducts, and alveoli) (5). Many cases are idiopathic, but BOOP may also occur after infectious pneumonia or may be associated with a variety of systemic disorders and clinical settings (5).

We reported previously the cases of three women with migratory and relapsing organizing pneumonia occurring a few months after radiation therapy for breast cancer, suggesting that radiation to the lung could participate in the development of a BOOP syndrome (6, 7). We now report a series of 15 cases demonstrating that radiation-primed BOOP syndrome merits being individualized as a distinct entity.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Collection of Cases

Following our observations of BOOP syndrome after radiation therapy (6, 7), we asked our colleagues through the Groupe d'Etudes et de Recherche sur les Maladies "Orphelines" Pulmonaires (GERM"O"P) to report the similar cases they had observed in their clinical practice.

The GERM"O"P is a French collaborative network founded in 1993 with the purpose of holding a registry of rare and unusual (so called "orphan") diseases of the lung to further realize clinical studies (retrospective and prospective) on these disorders. At the present time, chest physicians have reported more than 900 cases of orphan pulmonary disorders to the registry. Forty different groups, mainly from French university hospitals, participate in the GERM"O"P.

Study Population

Seventeen cases were reported to the GERM"O"P registry between June 1995 and December 1996. We included in the study only cases fulfilling the following criteria: (1) radiation therapy to the breast for carcinoma within 12 mo, (2) general and/or respiratory symptoms lasting for at least 2 wk, (3) radiographic lung infiltrates outside the radiation port, and (4) no evidence of a specific cause. Two cases reported on the GERM"O"P registry were excluded from the current study because the delay between radiation therapy and respiratory symptoms exceeded 12 mo in one (2 yr) and because of insufficient data in the other. Finally, 15 cases were included in this series.

Breast Cancer

Localized breast carcinoma had been diagnosed in all patients included in this study, right-sided in eight of them and left-sided in seven. All of the patients had undergone breast surgery, which consisted of limited tumorectomy (13 patients) or total mastectomy (two patients), plus axillary dissection for lymph nodes pathologic analysis (14 patients). Two patients received thereafter chemotherapy consisting of epirubicin, cyclophosphamide, and 5-fluorouracile that was administered concomitantly with radiotherapy in one patient and before radiotherapy in the other. Nine patients were receiving tamoxifen at the time of lung disease diagnosis.

Characteristics of Radiotherapy

Detailed data were available for 14 patients: 45 to 55 Gy were given to the affected breast using tangential fields in all cases. A boost to the primary tumor site of 10 to 20 Gy was added in five patients using tangential fields. Ipsilateral radiotherapy was administered to the axilla in three patients, to the supraclavicular region in seven patients, and to the internal mammary chain in nine patients.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

All the patients were women 47 to 71 yr of age (mean ± SD = 60 ± 6). Two were former smokers and 13 were nonsmokers. None of the patients was exposed to inhaled antigens or toxics. There was no clinical or biologic evidence of any collagen vascular disease either at initial presentation or at follow-up. The coexistent diseases and the drugs used in the previous 6 mo before diagnosis are shown in Table 1.

Clinical Presentation

All the patients were symptomatic. Symptoms appeared between 3 and 47 wk (14 ± 13 wk) after completion of radiotherapy. Fever (> 37.5° C) was always present, and asthenia (13 patients) and weight loss (six patients) were common. A persistent cough was noted in 13 patients, and nine had dyspnea (Class I or II of the NYHA, eight patients; Class III, one patient). Wheezing was noted in one patient. Localized crackles were heard in nine patients. Extrapulmonary symptoms were absent except for one patient who complained of transient arthralgias. The mean time from symptoms to diagnosis was 18 wk (two to 48 wk). Nine patients were treated with antibiotics before diagnosis without any clinical improvement.

Chest Radiography

At initial examination the chest radiographs showed peripheral air-space opacities (14 patients) or diffuse infiltrative opacities (one patient) (Figure 1). At initial examination the opacities were unilateral and limited to the irradiated side in eight patients, and the contralateral unirradiated lung was involved in seven. A migratory pattern of the lung infiltrates was noted in 11 patients before corticosteroid treatment, with the infiltrates beginning in the irradiated area, then spreading to the unirradiated areas of the ipsilateral lung then to the contralateral lung. In three patients, the migratory pattern of infiltrates was evidenced only after corticosteroid treatment when relapses occurred in previously unaffected areas, involving both lungs. In one case the infiltrates were bilateral at initial examination and did not migrate.

View larger version (143K): [in this window] [in a new window]   View larger version (148K): [in this window] [in a new window]   View larger version (139K): [in this window] [in a new window]   View larger version (153K): [in this window] [in a new window]   View larger version (141K): [in this window] [in a new window]   Figure 1.   Chest radiographic evolution in Patient 11. The patient was treated with radiotherapy for a carcinoma of the left breast: 65 Gy to the left breast and 45 GY to the ipsilateral axillar, supraclavicular and internal mammary chain lymph nodes. Ten weeks after completion of radiotherapy, alveolar infiltrates were noted in the left upper and the left lower lobes (panel A). Clinical and radiographic improvement was obtained with corticosteroids. A first relapse occurred while tapering steroids (prednisone 5 mg/d), with a new infiltrate in the lingula (panel B) that resolved with an increase in steroids dosage (panel C ). A second relapse occurred 1 wk after stopping steroids, with an infiltrate in a previously uninvolved area (panel D). A complete remission was obtained while reassuming corticosteroids (prednisone 10 mg/d). Corticosteroids were definitively stopped 8 mo later without sequelae (panel E ).

Computed Tomography of the Chest

Computed tomography (CT) of the chest was performed in 13 patients (Figure 2). It was performed at the initial presentation in 10 patients and again at the time of relapse in two. In three patients, CT was done only at the time of relapse. Peripheral air-space opacities with air bronchograms were observed in all patients, and further ground-glass opacities were detected in five. A limited pleural effusion was observed in one patient. Mediastinal lymph nodes were detected in one patient at initial examination and resolved with steroid therapy. In two patients, CT showed bilateral abnormalities, whereas abnormalities appeared only unilateral on chest radiographs.

View larger version (126K): [in this window] [in a new window]   View larger version (140K): [in this window] [in a new window]   View larger version (136K): [in this window] [in a new window]   View larger version (150K): [in this window] [in a new window]   Figure 2.   CT of the chest in Patient 3. At the time of diagnosis, 9 mo after radiotherapy to the left breast, bilateral lung infiltrates with air bronchograms were evident (panels A and B) resolving without sequelae with corticosteroids (panels C and D).

Pulmonary Function Tests

Pulmonary function tests performed in 10 patients at the time of diagnosis (Table 2) showed a mild restrictive ventilatory defect in two. A reduction of the transfer coefficient for carbon monoxide (< 80% of predicted) was the most common abnormality (six of eight patients tested). Mild hypoxemia was noted in four of 10 patients.

Bronchoalveolar Lavage

Bronchoalveolar lavage (BAL) was done in 10 patients at the time of diagnosis. One patient had BAL both in the irradiated lung and in the unirradiated lung. Thus, we analyzed 11 BALs performed in 10 patients. BAL cell count (Figure 3) was abnormal in all patients, with lymphocytosis greater than 20% at differential cell count present in all of them. Increased total cellularity (> 250,000 cells/ml) in seven patients, neutrophilia (> 5%) in eight patients, and eosinophilia (> 5%) in five patients were less frequent. A mild mastocytosis was noted in three patients (range, 1 to 8%). The search for bacterial, mycobacterial, fungal, and viral pathogens was always negative when performed (Table 1). There was no evidence of a viral cytopathic effect on BAL cells.

View larger version (14K): [in this window] [in a new window]   Figure 3.   Bronchoalveolar lavage results at the time of diagnosis. Data are from 11 bronchoalveolar lavages performed in 10 patients (in one patient BAL was performed in the right and left lungs).

Biologic Data

The erythrocyte sedimentation rate was increased (> 20 mm) in 14 patients (range, 45 to 140 mm) and normal in one patient (16 mm; Patient 10). A mild elevation of serum aspartate aminotransferase (ASAT) and alanine aminotransferase (ALAT) was noted in two patients of the 11 tested. Serum creatinine level was normal in all patients. Antinuclear antibodies (seven patients), antiDNA antibodies (four patients), rheumatoid factor (three patients), and antineutrophil cytoplasmic antibodies (five patients) were negative when tested. Serologic tests for Mycoplasma, Chlamydia, Legionella, and Rickettsiae were performed in Patient 12 and were negative. Respiratory syncitial virus serodiagnosis was negative in Patient 4.

Lung Biopsy

Lung biopsies performed in seven patients consisted of transbronchial biopsy in four and video-assisted lung biopsy in three. The biopsies were performed in radiographically abnormal areas, either homolateral to the irradiated breast (four patients) or contralateral (three patients).

All the samples were reviewed by an experienced pathologist (R. Loire, Lyon, France). In one patient, transbronchial biopsies gave no alveolar tissue, and the patient refused a surgical lung biopsy. In another patient (Patient 13), the transbronchial biopsies were compatible with, but not characteristic of, BOOP. A video-assisted lung biopsy was not considered. In five patients (video-assisted lung biopsy in three, and transbronchial biopsy in two), biopsies gave a histologic pattern typical of BOOP. In Patients 9 and 12, the lung biopsies were performed in the unirradiated lung, thus excluding any direct effect of radiation therapy. No granulomatous lesion or infectious agent was detected. There was no indirect evidence of viral infection on lung samples.

Disease Evolution

Mean follow-up was 15 ± 7 mo after the diagnosis (range, 4 to 29 mo). No patient required mechanical ventilation and no patient died.

All the patients were treated with corticosteroids. Steroid therapy induced a complete resolution of symptoms (usually within 1 wk) and of lung infiltrates (usually within 2 wk). However, relapses were very common. Twenty-one relapses occurred in 12 patients (one relapse in five patients, two relapses in five patients, three relapses in two patients). Three patients had no relapse at all. Most relapses occurred 1 to 6 wk after steroid withdrawal (17 relapses in 10 patients). Four relapses occurred while tapering steroid dosage (four patients), with daily doses ranging between 5 and 10 mg of prednisone at the time of relapse. Relapses were characterized by the recurrence of fever and new lung infiltrates, usually in previously unaffected areas. The definitive arrest of steroids was obtained in 12 patients. Mean total duration of steroid treatment was 49 wk (range, 17 to 100 wk). Three patients were still receiving corticosteroid treatment at the time of this study (since 91, 95, and 104 wk of treatment, respectively). No patient required immunosuppressant drugs.

In 14 patients, evaluation of the chest radiograph after treatment revealed minimal residual opacities at the site of the radiation therapy (four patients), limited linear densities in previously damaged areas (two patients), and normal chest radiographs in eight patients. Data were not available in one patient.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We identified 15 women with a characteristic clinical and radiographic syndrome that appeared within 12 mo of radiotherapy to the breast and was highly responsive to corticosteroid therapy. In five patients, BOOP was present at lung pathologic analysis. However, all our patients presented with the most typical clinical and imaging signs of BOOP. They had fever, nonproductive cough, mild dyspnea, and peripheral alveolar opacities on chest radiography and chest CT with a characteristic migratory pattern. When performed, bronchoalveolar lavage showed a mixed cellular pattern (with increased lymphocytes, neutrophils, and eosinophils). In all of the patients, dramatic improvement was obtained with corticosteroid therapy, but relapses were observed in 12 patients while tapering or after stopping corticosteroid therapy. Because pathologic diagnosis was not available in all cases, we preferred the terminology of BOOP syndrome by analogy with the terminology of bronchiolitis obliterans syndrome used to describe the dysfunction occurring in lung allografts, recognizing that there may or may not be pathologic evidence of bronchiolitis obliterans present (8).

Several factors suggest that radiation therapy was responsible for the occurrence of the BOOP syndrome observed in our patients.

First of all, no other cause was identified. Because of the retrospective nature of this work, exhaustive data were not obtained for every patient, particularly concerning the search for infectious pathogens. However, when performed, the search for infectious pathogens (bacteria, fungi, mycobacteria, viruses) was always negative. Medications used by the patients included tamoxifen in nine of them. To the best of our knowledge, despite the wide use of this drug, only one case of tamoxifen-associated lung disease has been reported in the literature dealing with the diagnostic significance of increased bronchoalveolar lavage fluid eosinophils (9). The patient had been receiving that medication as adjunctive therapy for breast cancer for 4 mo when she developed subacute onset of cough and a left lower lobe infiltrate. Whether she was given radiotherapy or not is not known. Tamoxifen has been shown to enhance radiotherapy-related lung fibrosis (10), an effect that could evolve through the increased release of transforming growth factor- (TGF-) that has been demonstrated both in vitro (11) and in vivo (12). However, we believe that tamoxifen did not play a significant role in the occurrence of lung disease in our patients since the disease evolved favorably in all of them even though tamoxifen was maintained.

Second, the mean interval between the end of radiation therapy and the onset of the first symptoms (14 wk) is consistent with the delay in the onset of a radiation-triggered alveolitis after the period of irradiation (3). Moreover, in 11 patients, serial radiographs undoubtedly showed that infiltrates began in the irradiated area, then spread to nonirradiated areas of the ipsilateral lung, then to the contralateral lung. Altogether these elements support the causative role of radiation therapy in the occurrence of the respiratory syndrome.

One of the most classic features of radiation-induced injury to the lung is that the pathologic and radiologic changes are confined to the outline of the fields of irradiation. Two distinct clinical stages of radiation pneumonitis have been recognized: an early transient radiation pneumonitis, which occurs about 4 to 12 wk after completion of radiotherapy, and a later chronic radiation fibrosis occurring usually more than 9 mo after treatment. The underlying pathologic events are well-characterized and consist successively of an acute exsudative phase, an organizing phase with interstitial infiltration with mononuclear cells and other inflammatory cells (2, 13), and a chronic or fibrotic phase (2). Some further findings such as hyaline membranes, marked cytologic atypia within hyperplastic pneumocytes, and prominent vascular changes (with thrombosis and vascular sclerosis) are also present in radiation-induced lung injury (16). However, a histologic pattern of BOOP is not a common finding after radiation exposure. Before our initial observation of three different patients with a BOOP syndrome after radiation therapy to the breast (6, 7) a histologic pattern of BOOP, strictly limited to the irradiated area, had been reported in only two cases in the literature (17, 18). Recently, van Laar and colleagues (19) described two patients with a syndrome very similar to the BOOP syndrome that we describe. Therefore, we believe that the syndrome reported here is clearly distinct from usual radiation pneumonitis, especially because of the migratory character of the alveolar opacities and the lack of significant sequelae.

Strictly unilateral pulmonary irradiation may occasionally result in bilateral pneumonitis (20). Gallium lung scan studies have shown further an increased gallium uptake in both lungs in patients with acute radiation pneumonitis after unilateral irradiation of the thorax (3, 24). Bronchoalveolar lavage studies evidenced a lymphocytic alveolitis in patients with radiation-induced pneumonitis (25, 26). More importantly, a prominent lymphocytic alveolitis, more pronounced in patients developing clinical pneumonitis, has been shown to develop in both lungs after strictly unilateral breast irradiation (3). Lymphocytes were predominantly activated CD4+ T-cells (3). This clearly indicates that a focal lung injury induced by localized irradiation results in a lymphocytic alveolitis involving both lungs. At this time the underlying mechanism is unknown. However, recent studies have shown that irradiation induces gene transcription, and results in the induction and release of proinflammatory cytokines and fibroblast mitogens (27). Also, radiation might induce the activation of pulmonary lymphocytes that recognize autoantigens released and unmasked after the initial lung tissue damage (3).

Whereas lymphocytic alveolitis appears to be very common after radiation therapy to the breast, the BOOP syndrome that we describe appears to be infrequent. This suggests that radiation alone is not sufficient, and that other factors are probably needed for BOOP to develop such as a genetically susceptible host and/or an as-yet unrecognized triggering factor acting on "radiation-primed" lymphocytes to produce a pneumonitis. This is particularly illustrated by the high incidence (50%) of severe or fatal radiation pneumonitis observed in a pilot study of human recombinant interferon gamma and accelerated hyperfractionated thoracic radiation therapy in patients with unresectable Stage IIIA/B non-small-cell lung cancer (30). Environmental factors may also be important since it has been shown that tobacco smoke suppresses radiation-induced inflammation in the lung in the rat and humans in vivo (31, 32). It must be noted that among the patients that we evaluated none was a current smoker.

At this time no case of BOOP associated with lung cancer radiation therapy has been reported to the GERM"O"P's registry despite the pulmonologists in this collaborative group regularly treat patients with lung cancer. The specific occurrence of this syndrome in the course of breast irradiation is indeed intriguing. We suggest that the radiation methods used may explain part of this specificity. Tangential fields used for radiation to the breast have been developed to limit the dose given to the lung in order to reduce the incidence of radiation pneumonitis. Tangential fields induce mainly an irradiation of the subpleural regions of the lung. Whether the subpleural location of the initial insult with the low doses received by the lung are keys for the understanding of the development of BOOP is currently unknown.

We suggest that, in addition to the classic radiation pneumonitis localized in the irradiated area, unilateral lung irradiation may "prime" the development of a characteristic BOOP syndrome quite similar to idiopathic BOOP.