In patients with neutropenia, CT halo sign is a highly effective modality for IPA diagnosis. The duration of the halo sign is short, and it demonstrates the value of early CT. The increase of the aspergillosis size on CT in the first days after IPA diagnosis is not correlated with a pejorative immediate outcome when using a combined medical-surgical approach.

We analyzed prognostic factors of response, response duration, and possible impact on survival of epoetin ␣, epoetin ␤, or darbepoetin ␣ (DAR) with or without granulocyte colony-stimulating factor in 403 myelodysplastic syndrome (MDS) patients. Sixty-two percent (40% major and 22% minor) and 50% erythroid responses were seen, and median response dura-

The blastic variant (BV) form of mantle cell lymphoma (MCL) is considered to be a very aggressive subtype of non-Hodgkin's lymphoma (NHL). In order to determine its clinico-biological features and response to therapy we studied 33 patients (17%) out of 187 suffering from MCL who were diagnosed with a BV of MCL. Blastic variant was diagnosed according to histopathological patterns, immunophenotyping, and bcl1 gene rearrangement and/or cyclin D1 overexpression. Three patients initially diagnosed with large cell NHL were classified as BV. Patients received front-line therapy including CHOP-like regimen or CVP (n = 29), or chlorambucil (n = 4) and CHOP or ESAP as second-line therapy. High-dose intensification with stem cell transplantation (SCT) was performed in 11 cases (autoSCT, n = 8; alloSCT, n = 3). All but two patients were in complete remission (CR) at the time of transplant (CR1, n = 5; CR2, n = 4). Clinical and biological characteristics did not differ from those of the common form of MCL. The median age was 62 years (29-80), with a sex ratio (M/F) of 2.6:1. Of the 33 patients, 66% had extranodal site involvement, 85% had an Ann Arbor stage IV, and 82% had peripheral lymphadenopathy. Circulating lymphomatous cells were seen in 48% of cases. Twelve patients (36%) entered a CR1 with a median duration of 11 months. Fifteen patients (46%) failed to respond and rapidly died of progressive disease. Second-line therapy led to a 26% (6/23) CR2 rate. Nine patients relapsed after high-dose therapy. Twenty-two of the 33 patients (66%) died of refractory or progressive disease. Median overall survival (OS) time was 14.5 months for the 33 BV patients as compared to 53 months for the 154 patients with a common form of MCL, P Ͻ0.0001. In the univariate analysis, OS was influenced by age, extranodal site involvement, circulating lymphomatous cells, and international prognosis index (IPI). In the multivariate analysis, only IPI affected OS: patients with IPI у2 had 8 months median OS as compared to 36 months median OS for patients with IPI Ͻ2, P = 0.003. Blastic variant is one of the worst forms of NHL. An improved recognition of BV of MCL is required, particularly in high-grade CD5 + NHL using immunophenotyping and bcl1 molecular study. Standard therapy using anthracycline or even high-dose intensification produce poor results and an alternative treatment should be proposed to such patients. Leukemia (2001) 15, 1785-1791.

We report outcomes after single (s) and double (d) umbilical cord blood transplantation (UCBT) after myeloablative conditioning (MAC) regimen for 239 patients transplanted for acute leukemia in first complete remission (CR1). All sUCBT patients received a total nucleated cell dose >2.5 × 10(7)/kg. Conditioning regimen for sUCBT was total body irradiation (TBI)12 Gy- or busulfan (BU)-based ± fludarabine (Flu) (n=68, group 1), thiotepa+BU+Flu (TBF) (n=88, group 2), and for dUCBT it was TBI12 Gy+cyclophosphamide ± Flu (n=83, group 3). dUCBT recipients were younger, received higher cell dose and less frequently antithymocyte globulin. In multivariate analysis, we found similar neutrophil recovery among the three groups; however, acute graft-versus-host disease II-IV was higher in dUCBT compared with others. Non-relapse mortality and relapse incidence were not statistically different among the three groups. Leukemia-free survival was 30% for sUCBT using TBI- or BU-based MAC compared with 48% for sUCBT TBF and 48% for dUCBT (P=0.02 and P=0.03, respectively), and it was not statistically different between sUCBT with TBF and dUCBT. In conclusion, use of sUCBT with adequate cell dose (>2.5 × 10(7)/kg) and a specific conditioning regimen in the MAC setting results in similar outcomes as dUCBT. The choice of TBF conditioning regimen for sUCBT may improve results, and whether this regimen may be effective in dUCBT should be further analyzed.