In contrast to the previous cases, all of our patients presented with ill-defined roentgenographic densities that were localized to one or two lobes of the lung. Since these roentgenographic findings were associated with fever and cough, the overall clinical picture suggested bacterial pneumonia. We do not know whether our patients with localized infiltrates repre­sent a limited or indolent form of leukemic pulmonary infiltration or whether their conditions were simply diagnosed at an earlier stage than the cases reported previously. However, because the infiltrates tended to grow worse before chemotherapy took effect, it is possible that the lung involvement would have pro­gressed to diffuse pulmonary infiltrates and severe respiratory distress had the treatment been delayed.

As can be seen from Table 1, the diagnosis of acute leukemic pulmonary infiltrates has been made in the past by open lung biopsy specimen or at autopsy with two previously reported exceptions. Rossi and colleagues made a diagnosis of pulmonary leukemic infiltrates in a patient with acute myelomonocytic leukemia after cytologic examination of bronchoalve- olar lavage fluid revealed a high proportion of leukemic cells. Young et al also reported recovery of leukemic cells from the lung by bronchoalveolar lavage, al­though no details were given. Similarly, bronchial lavage specimens from two of our three patients contained leukemic cells in sufficient numbers to suggest leukemic pulmonary infiltrates. These cases suggest that bronchoalveolar lavage may be useful for the diagnosis of leukemic pulmonary infiltrates in some instances. Indeed, bronchoalveolar lavage may be the preferred approach for diagnosis by bronchos­copy when the platelet count is unacceptably low for transbronchial biopsy. canada viagra online

Care must be taken, however, to avoid a false- positive diagnosis of pulmonary leukemic involvement on the basis of lavage specimens that contain an appreciable quantity of blood, especially in patients with high circulating blast counts. Moreover, the characteristic interstitial distribution of leukemic cells seen on histologic samples suggests that bronchoal- veolar lavage may not be diagnostic in every case (see case 3). On the other hand, leukemic infiltrates within alveolar septa and peribronchiolar spaces should be readily accessible to transbronchial biopsy. In fact, in each of our patients, multiple transbronchial biopsy specimens showed dense accumulations of leukemic cells within the alveolar septa. The diagnosis of leukemic pulmonary infiltrates was confirmed in all three patients when respiratory symptoms and roent­genographic abnormalities resolved simultaneously with the onset of chemotherapy-induced neutropenia. No complications of the procedures were observed other than one self-limited episode of epistaxis.

Although our experience with transbronchial biopsy was favorable in three patients, we suggest that great care be taken in establishing a diagnosis of leukemic pulmonary infiltrates by bronchoscopy. On small tissue samples, leukemic infiltration may be equivocal if blasts are not uniformly distributed throughout alve­olar septa and around blood vessels. Also, the blasts may be difficult to distinguish from reactive lymphoid infiltrates in some instances. Comparison with blasts from bone marrow may then be helpful (Fig 1). In addi­tion, care must be taken not to erroneously attribute pulmonary symptoms or roentgenographic abnormal­ities to leukemic infiltrates seen on transbronchial biopsy specimens, since autopsy studies have shown that pulmonary leukemic infiltrates frequently coexist with lung infections or other acute processes.

To avoid these pitfalls, we suggest that a diagnosis of symptomatic leukemic pulmonary infiltrates be made by transbronchial biopsy specimen only when several criteria are met: (1) the patient has uncon­trolled leukemia with a peripheral blast count exceed­ing 40 percent of circulating leukocytes (or an absolute blast count of 6,000); (2) widespread aggregates of leukemic cells are seen in septal, perivascular, or peribronchial locations; (3) neither biopsy specimen nor culture provides evidence of infection; and (4) fluid overload is excluded on clinical grounds or by lack of response to a diuretic. Even when these criteria are met, administration of broad-spectrum antibiotics may be appropriate prior to or in addition to chemo­therapy, especially for patients with cough, fever, and localized roentgenographic infiltrates. eriacta

In summary, although symptomatic involvement of the lung frequently presents with diffuse bilateral pulmonary infiltrates, this unusual complication of acute leukemia can occasionally also mimic a focal bacterial pneumonia. Pulmonary leukemic infiltrates should be considered, particularly in patients with acute leukemia of a myelogenous cell type who have a high proportion (>40 percent) of blast cells in the blood. An invasive diagnostic procedure should be employed early in such patients because leukemic involvement of the lung may progress rapidly to respiratory failure. The diagnosis can be made by transbronchial biopsy specimen and in some instances by bronchoalveolar lavage, but great care should be taken to exclude concurrent respiratory infection before chemotherapy is begun.