A critical evaluation of bronchoalveolar lavage. Criteria for identifying unsatisfactory specimens.

In a multicenter, city-wide study of the use of bronchoalveolar lavage for the evaluation of diffuse interstitial lung diseases, the occurrence of specimens unsuitable for analysis was evaluated. Using a standardized bronchoscopy technique, 26 physicians obtained 1,588 lavage specimens from 787 patients over a 52-month period. After transport to and processing in one laboratory using standardized procedures, all specimens were interpreted by one pathologist. Specimens were considered unsatisfactory if they contained: (1) a paucity of alveolar macrophages (i.e., less than ten alveolar macrophages/high-power field), (2) excessive numbers of airway-derived cells (i.e., more than the alveolar macrophages present), (3) a mucopurulent exudate, (4) cells altered by degeneration or (5) laboratory artifacts. Using these criteria, 30.4% of the specimens were considered unsuitable for analysis. There were no significant differences in the frequency of unsatisfactory specimens among participating physicians and institutions or between smoking and nonsmoking patients. Appraisal of alveolar inflammatory and immune effector cells in bronchoalveolar lavage specimens from patients with interstitial lung disease should include an assessment for contamination from airways proximal to the terminal bronchioles before conclusions are drawn about the activity of alveolar inflammation.

In vitro susceptibility of fungi to killing by neutrophil granulocytes discriminates between primary pathogenicity and opportunism.

Pathogenic fungi, according to their propensity to cause infection of apparently normal individuals, can be grouped into either primary pathogens (e.g., Coccidioides, Histoplasma, Paracoccidioides, Blastomyces, and Sporothrix) or opportunists (e.g., Candida, Mucoraceae, Aspergillus spp., Petriellidium, and Trichosporon). There is, however, no unifying concept explaining the difference between the virulence of the two fungal categories. Previously we have speculated that neutrophils are the common denominator of the high natural resistance to opportunistic fungi. Accordingly, we then compared the susceptibility to killing by neutrophil granulocytes of Histoplasma, Blastomyces, Paracoccidioides, and Sporothrix with that of 14 opportunistic fungi. We found the four virulent dimorphic yeasts, in contrast to opportunistic fungi, to be resistant to killing by neutrophils. Virulent dimorphic yeasts were ingested by neutrophils, and triggered a respiratory burst comparably to opportunists but were less susceptible to hydrogen peroxide, suggesting that differences in the susceptibility to microbicidal products of leukocytes may explain the difference in virulence.

Differentiation of Cryptococcus neoformans serotypes by isoenzyme electrophoresis.

Cryptococcus neoformans has been divided into four serotypes by specific agglutination in immune rabbit sera. Based on mating characteristics of the perfect state and epidemiologic and biochemical differences, the serotypes have been divided into two major pairs. In an attempt to characterize the serotypes further, the authors studied 22 strains of C. neoformans by the technic of horizontal starch-gel isoenzyme electrophoresis. The glucose-phosphate isomerase and phosphoglucomutase of serotypes A, C, D, and a subset of the serotype B strains migrated to distinguishable locations in this system. The activities of the remainder of the serotype B strains co-migrated with the serotype C strains. Thus, this technic distinguishes all the serotypes of C. neoformans except for a subset of serotype B and should be a useful adjunct for further elucidation of the epidemiologic and biochemical differences among serotypes.

Activation of complement by pathogenic and nonpathogenic Entamoeba histolytica.

Previous studies had demonstrated that strains of Entamoeba histolytica isolated from patients with colitis or amebic liver abscess were resistant to complement-mediated killing, whereas strains from asymptomatic patients were readily lysed by non-immune serum. Both serum-sensitive and serum-resistant strains of E. histolytica depleted complement rapidly as assessed by CH50, C3, and C7, and C5-9 hemolytic activities. Activation of the alternative pathway was important in lysis of nonpathogenic strains, as demonstrated by lysis by NHS (60.9 +/- 15.6%) and NHS + 5 mM EGTA (59.3 +/- 4.5%) as well as by C4-deficient guinea pig serum (72.8 +/- 7.1%) and C2-deficient human serum (64.4 +/- 11.1%), but not by NHS + 5 mM EDTA. Classical pathway activation also occurs as both pathogenic and nonpathogenic strains deplete greater than 98% of C4 activity, although it is not necessary for lysis. Pathogenic strains are not lysed by either the classical or the alternative pathway. These results suggest that pathogenic strains of E. histolytica activate complement but are able to evade an important host defense, complement-mediated lysis.

Lung inflammation in scleroderma: clinical, radiographic, physiologic and cytopathological features.

Twenty-five patients with systemic sclerosis were studied by chest radiography, lung function, esophageal motility, gallium-67 (67Ga) lung scanning and bronchoalveolar lavage (BAL). Alveolar inflammation, as defined by an elevation of proportional BAL lymphocyte or neutrophil counts, or increased thoracic uptake of 67Ga was found in 16 patients. An NIH gallium index greater than 65 index units identified a subgroup of patients with a significantly higher proportional BAL lymphocyte count (13.7 +/- 8.5 vs 5.6 +/- 3.1, p less than 0.0005). The presence of an abnormal chest radiograph correlated with physiologic evidence of lung restriction (p less than 0.01), and an elevation of proportional BAL lymphocyte count (15.5 +/- 8.2 vs 6.6 +/- 5.1, p less than 0.01). Eight patients receiving oral penicillamine therapy had significantly lower BAL lymphocyte counts compared to untreated patients (4.7 +/- 3.6 vs 11.3 +/- 7.7, p less than 0.05). We suggest that alveolar inflammation in scleroderma is characterized by lymphocyte accumulation and increased thoracic uptake of gallium.

Protection against gram-negative bacteremia and endotoxemia with human monoclonal IgM antibodies.

Hybridomas producing human monoclonal IgM antibodies (mAbs) against bacterial lipopolysaccharide (LPS) were generated by fusion of B lymphocytes from sensitized human spleen with heteromyeloma cells. The splenocytes were from patients undergoing splenectomy during staging for Hodgkin disease after vaccination with the J5 mutant of Escherichia coli, which is deficient in O antigenic side chains. This deficiency exposes the core oligosaccharide, common to LPS of all Gram-negative bacteria. The mAbs cross-reacted strongly with endotoxins from a wide range of unrelated species of Gram-negative bacteria. The mAbs also gave strong protection against LPS in the dermal Shwartzman reaction and against lethal Gram-negative bacteremia in mice. These findings indicate that monoclonal IgM against LPS endotoxin can neutralize its toxicity in vivo and might be valuable for treatment of patients with Gram-negative bacteremia. Analysis of one of the hybridoma clones, A6(H4C5), showed that the IgM mAb is directed against the covalently bound lipid A, which represents the most conservative and least variable structural element of LPS.

Assessment of drug disposition in the lung.

Airway disposition of drugs is assessed with either physiological changes in lung mechanics or nuclear scanning of the tagged medication. Several methods have been described for assessment of the pulmonary disposition of drugs delivered by routes other than the airways. These methods include tissue biopsy and sputum analysis of pooled secretions and tracheal washings. More recently, bronchoalveolar lavage fluid has been analysed for a variety of pharmacological agents and comparisons drawn between blood and lavage supernatant levels. Problems in correcting for dilution have been overcome by using a naturally occurring tracer substance, such as creatinine or albumin, which has a similar molecular weight to the test chemicals and which can be assayed readily in blood and lavage fluid. It has become apparent that neither naturally occurring not exogenous chemicals enter the lung in a concentration that is predictable from their levels in the blood. While the alpha 2-macroglobulin level in lavage fluid is approximately 25 times less than that in serum, a 1:1 relationship exists for alpha 1-antitrypsin. Cortisol achieves a concentration in lung fluid equal to that of blood, but lung fluid concentrations of methylprednisolone and prednisone are one-half, or at best one-third, of the blood concentration, respectively. Knowledge regarding the penetration of antibiotics into the lung is useful in determining the potential effectiveness of a given agent and its likely acinar MIC. It appears that the alveolar-capillary unit is not freely permeable to all agents, raising the possibility that a blood-lung barrier exists which is responsible for maintaining the alveolar environment. The knowledge that there is a differential permeability among drugs makes it important for clinicians to assess this characteristic of each agent before conclusions linking dose and response are drawn.

Pulmonary disposition of moxalactam.

Moxalactam is a new synthetic oxa-beta-lactam antibiotic with a broad spectrum of activity against Gram-positive and Gram-negative aerobic and anaerobic bacteria. It has proven clinical efficacy in pneumonia caused by a variety of infecting organisms. Therapeutic concentrations of moxalactam are achieved in most body tissues and fluids, including pleural fluid and sputum. However, assessment of the adequacy of lung tissue levels in pneumonia requires the sampling of material at an alveolar level. We performed bronchoalveolar lavage (BAL) in 13 patients one hour after they had been given moxalactam intravenously in doses ranging from 250 mg to 2 g. Absolute alveolar drug levels ranged from less than 1 to 6 micrograms/ml, and serum levels from 8 to 50 micrograms/ml. When expressed per micromole of creatinine, there was a significant relationship (r = 0.85; p less than 0.01) between serum and alveolar moxalactam levels in those patients in whom the drug concentration could be quantified accurately in BAL fluid.

An in vitro gallium-67 lung index for the evaluation of sarcoidosis.

In the evaluation of the active alveolitis of pulmonary sarcoidosis, both the proportional lymphocyte count obtained by bronchoalveolar lavage and state of activation of the alveolar macrophage by gallium scanning are required. We injected 6 mCi (200 MBq) of 67Ga intravenously on 24 occasions in 13 patients with biopsy-proved sarcoidosis. Forty-eight hours later, patients were scanned with a rectilinear scanner and the generated scintigrams were evaluated using the NIH index. Seventy-two hours after injection, bronchoalveolar lavage was performed, and venous blood was sampled. The harvested lavage fluid was analyzed for absolute and proportional cell counts, and radioactivity was measured in blood and BAL fluid. An in vitro 67Ga index was generated and expressed as counts/100,000 alveolar macrophages/ml blood (mean, 0.0146 +/- 0.0087 SD). There was a significant relationship between the in vitro index and proportional lymphocyte BAL counts (r = 0.79; p less than 0.002) that was comparable to that obtained using the NIH index (r = 0.74; p less than 0.005). These data suggest that the in vitro index might offer a more objective assessment of 67Ga uptake by the lung, but this would require validation against clinical parameters in a prospective study.

Specific cross-protective antigonococcal immunity in the murine genital tract.

Specific acquired immunity to gonococci was studied in systemically immunized mice, challenged with 10(7) gonococci by intrauterine inoculation. Protection after intraperitoneal immunization was monitored by vaginal cultures taken 24 h post-challenge, since events during the first 24 h postexposure to gonococci are crucial in determining the outcome of infection. Mice were protected against gonococcal challenge by two inoculations with either live or boiled gonococci given 4 weeks apart, whereas immunization with one inoculation did not protect against challenge 1 week later. Protection was correlated with high titers of IgG antibody in serum after two immunizations, but not with the high titers of serum IgM antibody found after the one immunization. IgG antibodies, but not IgM antibodies, were shown to pass into genital secretions. Protection could be passively transferred by serum with high titers of antibody. Of most practical importance was the finding that not only were heat-stable antigens protective, but also heterologous protection resulted after immunization with three strains differing in source (disseminated gonococcal infection versus gonorrhea), opacity-transparency characteristics, and serum sensitivity. The data indicate that IgG antibodies resulting from systemic immunization with heat-stable antigens may be able to provide cross-protection immunity against gonorrhea.

Killing of Aspergillus spores depends on the anatomical source of the macrophage.

To resolve the controversy over the capacity of macrophages to kill or inhibit germination of Aspergillus spores, we compared this function in peritoneal and alveolar macrophages. Alveolar macrophages from rabbits killed 82 to 90% and completely digested 72 to 82% of spores of Aspergillus fumigatus in 30 h. In contrast, peritoneal macrophages could not even inhibit the germination of ingested spores; more than 85% transformed into mycelia within 24 h. Killing by alveolar macrophages was delayed for 3 to 6 h after phagocytosis and was independent of oxidative killing mechanisms and immune activation. The ability of alveolar macrophages to kill Aspergillus spores without modulation by T lymphocytes or the generation of oxygen intermediates points out that concepts built on studies of peritoneal macrophages may be misleading and underscores the importance of studying the role of macrophages in immunity with cells from the appropriate anatomical site.

Prednisone and methylprednisolone disposition in the lung.

To compare the penetrability of methylprednisolone into lung tissue with that of prednisone, blood and bronchoalveolar lavage (BAL) fluid levels of methylprednisolone and prednisone were measured in 17 patients with a variety of lung diseases. To correct for variations in the quantity of lung fluid obtained by BAL, steroid levels were expressed in relation to creatinine concentrations. The level of methylprednisolone penetration in the pulmonary parenchyma, expressed by the slope of the relation between blood and BAL fluid, was 0.5 (r = 0.8; p less than 0.03). By contrast, despite serum levels of between 59 and 219 ng/ml, prednisone could not be detected in the BAL fluid in 3 patients; the overall correlation between blood and BAL fluid for all patients was r = 0.5 (slope = 0.3; p less than 0.1). Thus methylprednisolone is better able than prednisone to penetrate lung acini.