Although the high mortality rate of pulmonary invasive aspergillosis (IA) in patients with prolonged chemotherapy-induced neutropenia (PCIN) can be reduced by timely diagnosis, a diagnostic test that reliably detects IA at an early stage is lacking. We hypothesized that an electronic nose (eNose) could fulfill this need. An eNose can discriminate various lung diseases through the analysis of exhaled volatile organic compounds (VOCs). An eNose is cheap and noninvasive and yields results within minutes. In a single-center prospective cohort study, we included patients who were treated with chemotherapy expected to result in PCIN. Based on standardized indications, a full diagnostic workup was performed to confirm invasive aspergillosis or to rule it out. Patients with no aspergillosis were considered controls, and patients with probable or proven aspergillosis were considered index cases. Exhaled breath was examined with a Cyranose 320 (Smith Detections, Pasadena, CA). The resulting data were analyzed using principal component reduction. The primary endpoint was cross-validated diagnostic accuracy, defined as the percentage of patients correctly classified using the leave-one-out method. Accuracy was validated by 100,000 random classifications. We included 46 subjects who underwent 16 diagnostic workups, resulting in 6 cases and 5 controls. The crossvalidated accuracy of the eNose in diagnosing IA was 90.9% (P ‫؍‬ 0.022; sensitivity, 100%; specificity, 83.3%). Receiver operating characteristic analysis showed an area under the curve of 0.93. These preliminary data indicate that PCIN patients with IA have a distinct exhaled VOC profile that can be detected with eNose technology. The diagnostic accuracy of the eNose for invasive aspergillosis warrants validation.

Long-term bone marrow cultures (LTBMC) from patients with multiple myeloma (MM) and normal donors were analyzed for immunophenotype and cytokine production. Both LTBMC adherent cells from myeloma and normal donor origin expressed CD10, CD13, the adhesion molecules CD44, CD54, vascular cell adhesion molecule 1, very late antigen 2 (VLA-2), and VLA- 5, and were positive for extracellular matrix components fibronectin, laminin, and collagen types 3 and 4. LTBMC from myeloma patients and normal donors spontaneously secreted interleukin-6 (IL-6). However, levels of IL-6 correlated with the stage of disease; highest levels of IL-6 were found in LTBMC from patients with active myeloma. To identify the origin of IL-6 production, LTBMC from MM patients and normal donors were cocultured with BM-derived myeloma cells and cells from myeloma cell lines. IL-6 was induced by plasma cell lines that adhered to LTBMC such as ARH-77 and RPMI-8226, but not by nonadhering cell lines U266 and FRAVEL. Myeloma cells strongly stimulated IL-6 secretion in cocultures with LTBMC adherent cells from normal donors and myeloma patients. When direct cellular contact between LTBMC and plasma cells was prevented by tissue-culture inserts, no IL-6 production was induced. This implies that intimate cell-cell contact is a prerequisite for IL-6 induction. Binding of purified myeloma cells to LTBMC adherent cells was partly inhibited by monoclonal antibodies against adhesion molecules VLA-4, CD44, and lymphocyte function-associated antigen 1 (LFA-1) present on the plasma cell. Antibodies against VLA-4, CD29, and LFA-1 also inhibited the induced IL-6 secretion in plasma cell-LTBMC cocultures. In situ hybridization studies performed before and after coculture with plasma cells indicated that LTBMC adherent cells produce the IL-6. These results suggest that the high levels of IL-6 found in LTBMC of MM patients with active disease are a reflection of their previous contact with tumor cells in vivo. These results provide a new perspective on tumor growth in MM and emphasize the importance of plasma cell-LTBMC interaction in the pathophysiology of MM.

We performed a detailed kinetic study on the in vivo effect of a single subcutaneous dose of granulocyte colony-stimulating factor (G-CSF; 300 micrograms) in four healthy individuals on the expression and function of neutrophil Fc gamma receptors (Fc gamma R). G-CSF did not induce Fc gamma RI (CD64) on circulating neutrophils. However, neutrophils newly formed in response to G-CSF were Fc gamma RI positive and were able to perform antibody-dependent cellular cytotoxicity in an Fc gamma RI- dependent way. Fc gamma RII (CD32) expression was not changed significantly. Fc gamma RIII (CD16, phosphatidylinositol-linked) expression, slightly increased immediately (30 minutes) postinjection, was found to be strongly decreased on the newly formed population. For comparison, we studied the expression of the PI-linked proteins leukocyte alkaline phosphatase (LAP) and CD14. Intracellular levels of LAP mirrored the biphasic expression pattern as membrane-bound Fc gamma RIII. In contrast, CD14 expression on neutrophils was initially constant, followed by high levels on the newly formed neutrophils. Soluble CD14 levels were found to be elevated transiently, whereas peak levels of soluble Fc gamma III were observed as late as 6 days postinjection. In conclusion, we have shown that G-CSF results in an immunophenotypically and functionally altered neutrophil population for an important part as a result of its effect on myeloid precursor cells.

To study the receptors involved in the interaction between extracellular matrix proteins and hematopoietic progenitor cells, we analyzed the expression of beta 1 integrins on CD34+ bone marrow cells by means of immunoflowcytometry. Alpha 4 beta 1 and alpha 5 beta 1 were expressed, whereas alpha 1 beta 1, alpha 2 beta 1, alpha 3 beta 1, alpha 6 beta 1, and alpha v beta 1 were virtually absent. Furthermore, we assessed the alpha 4 and alpha 5 expression on committed myeloid progenitor cells. These colony-forming cells were detected in the alpha 4 dull fraction and the alpha 5 dull fraction. During myeloid differentiation, both in vivo and in vitro, a differential expression of alpha 4 beta 1 and alpha 5 beta 1 was observed. alpha 5 beta 1 was found to be lost at the myelocytic-metamyelocytic stage, before the loss of alpha 4 beta 1, at the band stage. Functional studies showed no binding of erythroid progenitor-depleted, CD34+ bone marrow cells to fibronectin. However, protein kinase C activation strongly induced fibronectin binding (68% of the cells). Inhibition experiments with specific antibodies and peptides showed the binding to be mediated by both alpha 4 beta 1 and alpha 5 beta 1. Also, colony-forming cells of granulocytes and macrophages were demonstrated to adhere to fibronectin in an activation-dependent way. During granulocyte colony-stimulating factor-induced in vitro maturation, the activation-dependent fibronectin binding capacity is gradually lost. We conclude that: (1) CD34+ bone marrow cells express alpha 4 beta 1 and alpha 5 beta 1; (2) the expression of alpha 4 beta 1 and alpha 5 beta 1 is differentially expressed during myeloid differentiation; and (3) binding of CD34+ bone marrow cells to fibronectin is activation dependent.