Subacute and chronic bird breeder hypersensitivity pneumonitis: sequential evaluation with CT and correlation with lung function tests and bronchoalveolar lavage.

PURPOSE: To evaluate lung involvement in the subacute (group 1) and chronic (group 2) stages of bird breeder hypersensitivity pneumonitis. MATERIALS AND METHODS: Computed tomographic (CT) findings in 45 patients were correlated with pulmonary function testing and bronchoalveolar lavage. Twenty-seven patients underwent sequential CT examination 0.3-4 years apart, with functional evaluation in 20 of them. RESULTS: In group 1, CT showed diffuse micronodules, ground-glass attenuation, focal air trapping or emphysema, and mild fibrotic changes, with normal lung volumes but impaired diffusing capacity and a predominant lymphocyte alveolitis. In group 2, two categories of chronic forms were identified at CT on the basis of presence or absence of honeycombing. After cessation of exposure, CT showed a return to normal or dramatic improvement in group 1 and considerable reduction in ground-glass attenuation and micronodules in group 2. CT depicted no fibrotic or emphysematous changes during follow-up. CONCLUSION: Subacute bird breeder hypersensitivity pneumonitis may share CT features with other types of hypersensitivity pneumonitis. CT may help in identification of emphysematous and fibrotic forms of chronic disease.

Depletion of OX-8 lymphocytes from the blood and airways using monoclonal antibodies enhances the late airway response in rats.

Recent evidence supports a role for T lymphocytes in allergic airway responses. We hypothesized that reducing blood T suppressor cells (Ts) might increase the late airway response (LR). Sprague-Dawley (SD) rats were sensitized with ovalbumin (OA). On days 8, 10, and 12, post-sensitization test SD (n = 14) received monoclonal antibody intravenously (OX-8; 1 mg) specific to rat Ts. Controls received saline (n = 7) or mouse ascites IgG (n = 7). On day 14, animals were challenged with OA aerosol (5% wt/vol) for 5 min, lung resistance was recorded for 8 h (n = 18) and bronchoalveolar lavage was performed. The LR was determined from the area under the lung resistance vs time curve from 75 to 480 min after challenge. In the remaining 10 rats, airway lymphocyte subsets were measured 8 h after OA aerosol challenge in minced and digested lungs. A decrease in percentage of blood and airway Ts, respectively, in test animals was observed vs controls (blood: 6.27 +/- 0.84 vs 32.95 +/- 1.94, P < 0.001); (airway: 5.05 +/- 0.66 vs 24.5 +/- 3.05, P < 0.02). Blood and airway helper T lymphocytes did not differ between test and control animals. The LR was significantly increased in test (22.89 +/- 3.92) vs controls (4.22 +/- 2.18, P < 0.001). Bronchoalveolar lavage macrophages, neutrophils and lymphocytes, and serum OA-specific IgE were also significantly elevated (P < 0.05) in test animals. We conclude that Ts play an important role in attenuating the LR in SD rats.

Impaired sensorineural function after allergen-induced mediator release.

We tested the hypothesis that allergen-induced mediator release augments the magnitude of isocapnic dry gas hyperpnea-induced bronchoconstriction in sensitized guinea pigs. Male Hartley guinea pigs were sensitized by spontaneous inhalation of ovalbumin (OA) aerosol on days 0 and 7 of the study. On day 14, sensitized animals again breathed OA aerosol and were prospectively divided into a group that exhibited labored breathing (LB), presumably reflecting OA-induced inflammatory mediator release, and a group that did not exhibit LB at this time. Control guinea pigs breathed saline aerosol on days 0, 7, and 14. Bronchoalveolar lavage on day 17 disclosed relative eosinophilia in OA+LB, but not in OA-LB, animals. On day 17, the bronchoconstrictor responses to increasing intravenous (i.v.) doses of acetylcholine (ACh), substance P (SP), neurokinin A (NKA), and capsaicin, as well as dry gas hyperpnea, were measured in vivo in animals from each group. Control and OA-LB guinea pigs exhibited similar responses, but OA+LB animals demonstrated augmented bronchoconstriction induced by i.v. administration of ACh, SP, or NKA. However, despite their augmented responsiveness to these exogenous constrictor agonists, OA+LB animals displayed no greater bronchoconstriction after dry gas hyperpnea or i.v. capsaicin administration. It is known that both dry gas hyperpnea and i.v. capsaicin cause bronchoconstriction in guinea pigs by releasing endogenous tachykinins from airway sensory C-fibers. Thus, our results suggest that allergen-induced mediator release impairs endogenous tachykinin release from airway sensory C-fibers in guinea pigs.

Prednisolone treatment in asthma is associated with modulation of bronchoalveolar lavage cell interleukin-4, interleukin-5, and interferon-gamma cytokine gene expression.

Although corticosteroids are effective in improving asthma symptoms and bronchial responsiveness, their mechanism of action is unknown. We examined whether changes in bronchial responsiveness with corticosteroid therapy of asthma are accompanied by a reduction in cytokine gene expression and eosinophil infiltration in the airways. Bronchoalveolar lavage (BAL) was performed in 18 patients with moderate asthma before and after 2 wk of treatment with prednisolone, 0.6 mg/kg/day, or matched placebo in a randomized double-blind parallel group study. Cells were counted in BAL cytocentrifuge preparations, and the numbers of cells expressing cytokine mRNA were assessed by in situ hybridization using 35S-labeled RNA probes. When the actively treated and placebo groups were compared, there was a decrease in airway methacholine responsiveness (p < 0.01) after prednisolone. This was accompanied by a decrease in bronchoalveolar lavage eosinophils (p < 0.05), a reduction in the numbers of BAL cells per 1,000 expressing mRNA for interleukin-4 (IL-4, p < 0.01) and interleukin-5 (IL-5, p < 0.005), and an increase in numbers of cells expressing mRNA for interferon-gamma (p < 0.005). These results are compatible with the hypothesis that the beneficial effects of corticosteroids in asthma may result from modulation of cytokine production, with consequent inhibition of local bronchial eosinophilia.

Enhanced production of IL-1 receptor antagonist by alveolar macrophages from patients with interstitial lung disease.

Alveolar macrophages (AM) produce various inflammatory and immunomodulatory cytokines. The objective of these experiments was to evaluate the production of IL-1ra, a specific receptor antagonist of IL-1, by AM from nonsmoking control subjects (n = 9), smoking control subjects (n = 6), and patients with interstitial lung disease (ILD) (n = 9). IL-1ra protein levels in cultured AM lysates and supernatants were determined by a specific ELISA; relative steady-state IL-1ra mRNA levels were measured using a specific cDNA probe. Before culture the isolated AM from all subject groups contained undetectable IL-1ra mRNA and no IL-1ra protein in the cell lysates as determined by ELISA. AM from nonsmoking control subjects spontaneously produced IL-1ra protein after a 20 h culture in medium, approximately 12 ng/ml with around half in cell lysates. AM from smoking control subjects produced levels of IL-1ra that were similar to the levels in AM from nonsmokers. In contrast, AM from nonsmoking ILD patients (n = 6) produced high levels of IL-1ra spontaneously (approximately 28 ng/ml), with no enhancement observed when cultured on adherent IgG. Interestingly, AM from smoking ILD patients (n = 3) produced lower levels of IL-1ra protein (approximately 11 ng/ml) that were comparable to levels noted in smoking control subjects. AM from all three types of subjects produced decreased amounts of IL-1ra in response to LPS and enhanced amounts in response to GM-CSF. In general, IL-1ra steady-state mRNA levels correlated with protein production.(ABSTRACT TRUNCATED AT 250 WORDS)

Fractional analysis of the 120-ml bronchoalveolar lavage. Determination of the best specimen for diagnosis of sarcoidosis.

STUDY DESIGN: Bronchoalveolar lavage with seven 20-ml aliquots was performed on 9 patients with sarcoidosis and 11 patients with idiopathic pulmonary fibrosis (IPF). The returns from the first 20-ml aliquot (volume 1), the next 100 ml of bronchoalveolar lavage fluid (volume 2), and the last 20 ml of lavage fluid (volume 3) were analyzed separately for cell content and differential cell counts. The cell content was calculated four different ways, including absolute number of cells per milliliter of return and as a differential cell count using three different denominators. The denominators used were (1) neutrophils, lymphocytes, eosinophils, macrophages, and epithelial cells; (2) neutrophils, lymphocytes, eosinophils, and macrophages; and (3) neutrophils, lymphocytes and eosinophils. The data were analyzed to see what volume and arithmetic expression of cell content provided the best means of distinguishing sarcoidosis from IPF. RESULTS: The results confirmed a proximal distribution of the first 20-ml aliquot with more bronchial epithelial cells and neutrophils than volumes 2 and 3. There were more macrophages and lymphocytes in volumes 2 and 3 than in volume 1. Volumes 2 and 3 had similar cell content. Sequential fractional analysis showed that the cells in volume 2 and 3 more clearly distinguished sarcoidosis from IPF than the cell content of volume 1. CONCLUSION: The presence of lymphocytes, and especially a predominance of lymphocytes with the relative absence of neutrophils, correlated best with sarcoidosis. The percentages of differential cell counts proved superior to both the absolute number of neutrophils and of lymphocytes per milliliter of return in distinguishing sarcoidosis from IPF. A percentage differential cell count with a denominator of neutrophils+lymphocytes+eosinophils provided the best means of distinguishing between sarcoidosis and IPF.

Bronchoalveolar lavage cell populations in the diagnosis of sarcoidosis.

STUDY DESIGN: Between February 1, 1984, and February 1, 1989, fiberoptic bronchoscopy was performed on 2,692 patients, 592 of whom had bronchoalveolar lavage (BAL). One hundred twenty-eight patients with 16 percent or more lymphocytes in BAL fluid (BALF) were selected for further study. The group included 27 patients with sarcoidosis, 28 with nonsarcoidosis interstitial lung disease (ILD), 22 with lung infection (organism isolated), 31 with inflammation (presumed infection, no organism isolated), 14 with neoplasm, and 6 with bronchial hyperreactivity. METHODS: The percentages of lymphocytes, B lymphocytes, and T lymphocytes, the CD4/CD8 ratio and the percentages of neutrophils and eosinophils were analyzed individually and in combination for discrimination between the sarcoidosis and nonsarcoidosis patients and compared with the diagnostic accuracy of multiple noncaseating granuloma (MNG) on a simultaneous transbronchial biopsy (Tbbx). RESULTS: Neither the percentages of lymphocytes, T lymphocytes, or B lymphocytes discriminated sarcoidosis from nonsarcoidosis patients. Sarcoidosis patients had higher CD4/CD8 ratios, fewer neutrophils, and 1 percent or less eosinophils in the BAL cell populations. An analysis of CD4/CD8 ratios, and percentages of neutrophils and eosinophils individually revealed that a CD4/CD8 ratio of 4:1 or greater had a positive predictive value of 94 percent in distinguishing sarcoidosis from other ILD but a sensitivity of only 59 percent. The positive predictive value of CD4/CD8 ratio of 4:1 or greater fell to 50 percent in separating sarcoidosis from all other diseases. A CD4/CD8 ratio of less than 1:1 has a 100 percent negative predictive value to exclude the diagnosis of sarcoidosis. Finding 1 percent or less neutrophils in BAL had an 80 percent positive predictive value in distinguishing sarcoidosis from nonsarcoidosis ILD and 51 percent for distinguishing sarcoidosis from all other disease groups. The CD4/CD8 ratio and the percentages of neutrophils and eosinophils also were combined and analyzed for the diagnosis of sarcoidosis. CONCLUSIONS: Results showed a BALF with a CD4/CD8 ratio of 2:1 or greater, 1 percent or less neutrophils, and 1 percent or less eosinophils has essentially the same specificity and positive predictive value as MNG on Tbbx in distinguishing sarcoidosis from nonsarcoidosis disease. The combination of finding MNG in a Tbbx specimen plus a BALF CD4/CD8 ratio of 4:1 or greater had a 100 percent positive predictive value in separating sarcoidosis from other ILD and an 81 percent value in separating sarcoidosis from all other disease. Finding MNG in a Tbbx specimen plus a BALF with a CD4/CD8 ratio of 2:1 or greater, 1 percent or less neutrophils, and 1 percent or less eosinophils had a 93 percent positive predictive value in distinguishing sarcoidosis from both nonsarcoidosis ILD and all other diseases.

Transient wheeze. Eosinophilic bronchobronchiolitis in acute eosinophilic pneumonia.

The clinicopathologic features of five patients with acute eosinophilic pneumonia who presented with transient wheeze as well as acute onset of high fever, severe hypoxemia, and diffuse pulmonary infiltrates are described. Eosinophilic pneumonia was diagnosed by bronchoalveolar lavage and transbronchial lung biopsy. The illness resolved rapidly with or without corticosteroid therapy. No relapse occurred. To characterize the transient wheeze, a transbronchoscopic bronchial biopsy and pulmonary function tests were performed. Specimens of bronchial wall revealed eosinophil infiltration into the bronchial mucosa. Pulmonary function tests demonstrated reduced diffusing capacity and small airway dysfunction. These findings suggested that eosinophil infiltration into the bronchial mucosa might temporarily cause transient wheeze, different from bronchial asthma, due to small airway dysfunction in acute eosinophilic pneumonia.

Cytologic diagnosis of metastatic epithelioid sarcoma. A cytologic mimic of squamous cell carcinoma.

A case of epithelioid sarcoma in a 39-year-old woman who had metastatic lesions in her lung from a primary arm tumor is presented. A review of the literature and analysis of this case did not identify any specific light microscopic or immunohistochemical way to separate definitively the lung cytologic smear in this case from squamous cell carcinoma. The authors believe cytologic smears from epithelioid sarcoma can closely mimic smears from squamous cell carcinoma. Pathologists should be alert to these similarities.

Pulmonary inflammatory cells in ventilated preterm infants: effect of surfactant treatment.

The aim of this study was to determine the effect of surfactant treatment on the number and distribution of inflammatory cells in bronchoalveolar lavage fluid (BALF) from mechanically ventilated preterm infants over the first week of life in relation to the subsequent development of chronic lung disease (CLD). The study included 25 babies who received surfactant on clinical grounds and 29 babies of similar severity who did not. BALF was collected on days 1, 3, 5, and 7 after birth. Cell counts were performed and differentials were calculated on 300 cells. CLD was equally common in both treatment groups. Of the 54 infants, 29 (53%) who developed CLD had a higher incidence of patent ductus arteriosus and air leak and needed a higher concentration of inspired oxygen on the fifth and seventh days of life. Babies who developed CLD had more polymorphonuclear leucocytes and fewer macrophages on days 5 and 7 than those who recovered. Surfactant treatment was associated with a higher total white cell count on day 3. Between days 3 and 7, macrophage numbers were higher in surfactant treated babies, whatever the pulmonary outcome. This data suggests that CLD was associated with persistence of high numbers of polymorphonuclear leucocytes in BALF at the end of the first week. Surfactant treatment caused a persistent increase in macrophage numbers. The association between persistent neutrophilia and CLD was unaffected by surfactant treatment.

Bronchoalveolar lavage in rheumatoid arthritis.

Bronchoalveolar lavage (BAL) was performed on 70 RA patients, 28 without extra-articular manifestations, nine with pulmonary involvement, 13 with sicca-syndrome, 20 with other extra-articular manifestations such as renal involvement, cutaneous vasculitis and rheumatoid nodules. Fifteen patients without rheumatic or pulmonary disease served as the control group. Compared with the control group RA patients showed a statistically significant increase of lymphocytes, especially of activated (DR+)T(CD3+)-helper (CD4+) cells, resulting in a significantly diminished percentage of alveolar macrophages, B(CD21+)-lymphocytes, T-suppressor (CD8+) cells and an increased CD4/CD8 ratio. This cell distribution pattern was more pronounced in RA patients with lung involvement with significant differences to the other RA patients with regard to lymphocytes, DR positive cells and CD4 positive/DR positive cells. It is concluded that these results indicate an altered balance of immunocompetent cells not only in the joints but also in the lung. The changes are more distinct if local manifestations can be diagnosed clinically.

Pathology of asthma and its clinical implications.

Although detailed histopathologic studies have described the inflammatory processes present in fatal asthma, until recently the pathology of less severe forms of the disease has been less well understood. Now a series of important studies has extended our understanding of the pathophysiology of mild asthma. Studies that examine sputum or fluid obtained by bronchoalveolar lavage from mildly asthmatic subjects revealed findings consistent with an active inflammatory process within the airways. The histopathologic examination of endobronchial biopsy specimens from stable asthmatic patients has shown that inflammatory cell infiltration of the mucosa is a distinctive feature of mildly asthmatic subjects requiring only intermittent inhaled beta-agonist therapy. These studies have also shown that marked tissue disruption may occur early in the natural history of mild asthma. These investigations have demonstrated that asthma is a disease characterized by acute and chronic inflammatory changes within the airways and that in many respects the histopathologic features of mild allergic asthma are similar to those observed in fatal asthma. The therapeutic implications of these findings are that management of mild asthma should be directed toward resolving this inflammatory process. There is now sufficient evidence to suggest that anti-inflammatory drugs such as cromolyn sodium, inhaled corticosteroids, and nedocromil sodium be used earlier in the course of the disease and that the use of these therapeutic agents should not be limited to patients with severe forms of asthma. This may be particularly important in view of the increasing awareness of the potential problems associated with the overreliance on beta-agonist therapy in patients with asthma.

Lung fibrosis in hypersensitivity pneumonitis. Association with CD4+ but not CD8+ cell dominant alveolitis and insidious onset.

Seventeen cases of hypersensitivity pneumonitis (HP) at a symptomatic phase were categorized into two groups based on computed tomographic (CT) findings and histologic features of transbronchial lung biopsy specimens, HP accompanied by lung fibrosis (fibrosis group), and HP unaccompanied by lung fibrosis (nonfibrosis group). The fibrosis group comprised bird fancier's lung and HP of unknown etiology, whereas the nonfibrosis group mainly comprised summer-type HP. Comparison of results of pulmonary function tests between these two groups confirmed a restrictive impairment in the fibrosis group. Analyses of cellular components of bronchoalveolar lavage (BAL) fluids revealed lymphocytes, especially CD8+ T lymphocytes, were significantly increased in the nonfibrosis group in comparison with the fibrosis group, whereas CD4+ T cells were increased to the same level in the both groups. Analyses of the onset of disease showed that acute onset was observed mainly in nonfibrosis group and strongly correlated with increased CD8+ T lymphocytes in BAL fluids, while insidious onset was related to lung fibrosis and relatively increased CD4+ T lymphocytes in BAL fluids. These findings raise the possibility that highly elevated CD8+ T cells might have a protective effect on pulmonary fibrosis or that relatively increased CD4+ T cells might play an important role in the pathogenesis of pulmonary fibrosis of HP at the chronic phase.

Relative release of interleukin-1 beta and interleukin-1 receptor antagonist by alveolar macrophages. A study in asbestos-induced lung disease, sarcoidosis, and idiopathic pulmonary fibrosis.

We examined the influence of untreated interstitial lung disease (ILD) on the in vitro release of interleukin-1 beta (IL-1 beta) and interleukin-1 receptor antagonist (IL-Ira) from alveolar macrophages (AM); AM were harvested from normal volunteers, ILD patients, and patients with asbestos-related pleural disease but no ILD. AM were cultured for 24 h and assays for IL-1 beta and IL-1ra were done using sensitive and specific enzyme-linked immunosorbent assay. A greater amount of IL-1 beta was detected in AM supernatants from asbestosis, sarcoidosis, and IPF patients than in those from normal subjects. The IL-1 beta:IL-1ra ratio (IL-1 beta activity index [IL-1AI]) was significantly lower in supernatants of normal macrophages compared with macrophage supernatants from individuals with ILD. The IL-1AI correlated with bronchoalveolar lavage cellularity, a marker of disease activity. Current smoking was associated with lower IL-1 beta and IL-1ra release in ILD. The IL-1AI is a convenient method for comparison of IL-1 beta activity between patient populations.

Amount of Pneumocystis carinii and degree of acute lung inflammation in HIV-associated P carinii pneumonia.

Correlations between semiquantitative amounts of Pneumocystis carinii (PC), the degree of inflammation, and the severity of pneumonia were analyzed in 58 patients with PC pneumonia (PCP). Material from both transbronchial biopsies (TBBs; n = 39) and bronchoalveolar lavage fluid (BALF; n = 57) was examined. In the TBB the amount of PC correlated strongly with overall inflammation in the interstitium (Kendall correlation coefficient [Kcc] = 0.59; p < 0.0001), type 2 pneumocyte proliferation, and edema formation. The amount of PC in the TBB also correlated with interstitial accumulation of neutrophils (Kcc = 0.54; p = 0.0001), lymphocytes, and macrophages. In BALF the amount of PC correlated with edema formation and type 2 pneumocyte proliferation in the TBB but not with the percentage of neutrophils, lymphocytes, or macrophages in BALF. The amount of PC in the BALF and the percentage of neutrophils in the BALF correlated significantly with Po2 and the serum lactate dehydrogenase (LDH) level. Neither short-term nor long-term survival was affected by the amount of PC, inflammatory markers in the TBB, inflammatory cells in BALF, Po2, or the serum LDH levels. In conclusion, the amount of PC is associated with the extent of the acute inflammatory reaction in the lung in PCP associated with human immunodeficiency virus (HIV).

Pulmonary involvement in human T-cell lymphotropic virus type-I uveitis: T-lymphocytosis and high proviral DNA load in bronchoalveolar lavage fluid.

The ocular manifestation of human T-cell lymphotropic virus type I (HTLV-I) infection has been recognized as a new clinical entity termed HTLV-I uveitis. In order to determine whether HTLV-I uveitis is associated with lymphocyte alveolitis, bronchoalveolar lavage (BAL) was carried out in 11 patients with HTLV-I uveitis, five asymptomatic HTLV-I carriers, 11 HTLV-I-negative patients with ocular sarcoidosis, and nine normal control subjects seronegative for HTLV-I. Six of the 11 patients with HTLV-I uveitis showed increased total cell counts, and T-lymphocytosis in BAL fluid. CD4+/CD8+ ratios of T-lymphocytes were decreased in three of these patients, and normal in three other patients. Such abnormalities of the lung were not found in asymptomatic HTLV-I carriers, and in normal control subjects. BAL findings in HTLV-I uveitis differed from those of patients with sarcoidosis in terms of the lymphocytic component. Interestingly, it was found that there was an increase of HTLV-I proviral deoxyribonucleic acid (DNA) load in peripheral blood mononuclear cells (PBMC) from seven patients with HTLV-I uveitis, and in the BAL cells from four patients with pulmonary involvement. These results provide further evidence in terms of HTLV-I tropism for the lung, and suggest that a systemic and local increase of HTLV-I proviral DNA load plays an important role in the pathogenesis of lymphocyte alveolitis in patients with HTLV-I uveitis.

Responses of horses affected with chronic obstructive pulmonary disease to inhalation challenges with mould antigens.

Eight control and 8 asymptomatic COPD-affected horses were given, on separate occasions, inhalation challenges with extracts of Micropolyspora faeni, Aspergillus fumigatus and Thermoactinomyces vulgaris. All horses were also given nebulised phosphate-buffered saline (PBS) challenges and 'natural challenges' (NCs), i.e. exposure to hay and straw, as control challenges. Responses were assessed by clinical, pulmonary mechanics, arterial blood gas tensions, arterial blood pH and bronchoalveolar lavage fluid cytological examinations. PBS challenges had no effect on control or COPD-affected horses, while NC induced COPD only in the COPD-affected horses. Pulmonary disease, similar to naturally occurring COPD, was induced, only in the COPD-affected horses, by M. faeni and A. fumigatus challenges, thus implicating these organisms in the aetiology of equine COPD. The role of T. vulgaris in the aetiology of equine COPD could not, however, be determined because the T. vulgaris challenges, in addition to inducing pulmonary disease in 4 COPD-affected horses, induced pulmonary disease in 2 control horses which had been unaffected by NC. The absence of pulmonary disease in control horses after M. faeni, A. fumigatus and NC challenges suggests that equine COPD is a pulmonary hypersensitivity, rather than a non-specific toxic response.

Evaluation of local endobronchial antigen challenges in the investigation of equine chronic obstructive pulmonary disease.

Local transendoscopic endobronchial antigen challenge, which has proved to be a valuable clinical and research technique in the study of human pulmonary hypersensitivity, was evaluated in control and asymptomatic COPD--affected horses. Transendoscopic endobronchial challenges with phosphate-buffered saline (PBS), Micropolyspora faeni extract at 60 and 600 micrograms/ml and mouldy hay extract elicited neutrophilic airway inflammatory responses in control (N = 5-7) and asymptomatic COPD-affected (N = 5-7) horses, as determined by cytological examinations of bronchoalveolar lavage fluid (BALF) harvested from the challenged lung segments. Endobronchial challenges with 600 micrograms M. faeni extract/ml induced a significant BALF neutrophilia only in horses with asymptomatic COPD, when compared with PBS challenges. However, as the BALF neutrophil ratios of COPD-affected horses after this M. faeni challenge did not differ significantly from those of control horses, this finding has little clinical diagnostic value. The BALF neutrophilia induced in control and asymptomatic COPD-affected horses by 60 micrograms M. faeni extract/ml and mouldy hay extract challenges was not significantly different from that induced by PBS challenge. Endoscopically visible bronchial changes were observed in some of the control and COPD-affected horses within 5 mins and at 5 h after PBS, M. faeni and mouldy hay extract challenges. We conclude that this technique is of no value in the investigation of equine COPD.