A real-life overview of a hematopoietic cell transplant program throughout a four-year period, including prospective registry, exclusion causes and final donor selection.

Traceability of patients who are candidates for Hematopoietic cell transplant (HCT) is crucial to ensure HCT program quality. Continuous knowledge of both a detailed registry from a HCT program and final exclusion causes can contribute to promoting a real-life vision and optimizing patient and donor selection. We analyzed epidemiological data reported in a 4 year-monocentric prospective registry, which included all patients presented as candidates for autologous (Auto) and/or allogeneic (Allo) HCT. A total of 543 patients were considered for HCT: 252 (42.4%) for Allo and 291 (57.6%) for Auto. A total of 98 (38.9%) patients were excluded from AlloHCT due to basal disease progression more commonly (18.2%). Seventy-six (30.2%) patients had an HLA identical sibling, whereas 147 (58.3%) patients had only Haplo. UD research was performed in 106 (42%) cases, significantly more often in myeloid than lymphoid malignancies (57% vs 28.7%, p < 0.001) but 61.3% were finally canceled, due to donor or disease causes in 72.4%. With respect to Auto candidates, a total of 60 (20.6%) patients were finally excluded; progression was the most common cause (12%). Currently, Haplo is the most frequent donor type. The high cancellation rate of UD research should be revised to optimize further donor algorithms.

Efficacy of extended infusion of ß-lactam antibiotics for the treatment of febrile neutropenia in haematologic patients: protocol for a randomised, multicentre, open-label, superiority clinical trial (BEATLE).

BACKGROUND: Febrile neutropaenia (FN) is a very common complication in patients with haematological malignancies and is associated with considerable morbidity and mortality. Broad-spectrum antipseudomonal ß-lactam antibiotics (BLA) are routinely used for the treatment of cancer patients with FN. However, the clinical efficacy of BLA may be diminished in these patients because they present with pathophysiological variations that compromise the pharmacokinetic (PK) parameters of these antibiotics. Optimised administration of BLA in prolonged infusions has demonstrated better clinical outcomes in critically ill patients. However, there is a paucity of data on the usefulness of this strategy in patients with FN. The aim of this study is to test the hypothesis that the administration of BLA would be clinically more effective by extended infusion (EI) than by intermittent infusion (II) in haematological patients with FN. METHODS: A randomised, multicentre, open-label, superiority clinical trial will be performed. Patients with haematological malignancies undergoing chemotherapy or haematopoietic stem-cell transplant and who have FN and receive empirical antibiotic therapy with cefepime, piperacillin-tazobactam or meropenem will be randomised (1:1) to receive the antibiotic by EI (during half the time of the dosing interval) in the study group, or by II (30 min) in the control group. The primary endpoint will be clinical efficacy, defined as defervescence without modifying the antibiotic treatment administered within the first 5 days of therapy. The primary endpoint will be analysed in the intention-to-treat population. The secondary endpoints will be pharmacokinetic/pharmacodynamic (PK/PD) target achievement, bacteraemia clearance, decrease in C-reactive protein, overall (30-day) case-fatality rate, adverse events and development of a population PK model of the BLA studied. DISCUSSION: Data on the usefulness of BLA administration in patients with FN are scant. Only three clinical studies addressing this issue have been published thus far, with contradictory results. Moreover, these studies had some methodological flaws that limit the interpretation of their findings. If this randomised, multicentre, phase IV, open-label, superiority clinical trial validates the hypothesis that the administration of BLA is clinically more effective by EI than by II in haematological patients with FN, then the daily routine management of these high-risk patients could be changed to improve their outcomes. TRIAL REGISTRATION: European Clinical Trials Database: EudraCT 2018-001476-37. ClinicalTrials.gov, ID: NCT04233996.

Advantages of non-cryopreserved autologous hematopoietic stem cell transplantation against a cryopreserved strategy.

Autologous stem cell transplantation (auto-HSCT) is an effective treatment strategy for hematological malignancies. The standard mode of handling hematopoietic progenitors for the autologous procedure (CRYO) consists on its collection and freezing with dimethyl sulfoxide (DMSO) and its subsequent thawing and re-infusion. This process is toxic and expensive. Non-cryopreserved (non-CRYO) is a less expensive mode of auto-HSCT. We designed a comparative study between both strategies performed in two different centers to analyze the short-term complications. In total 111 auto-HSCT were performed from January/2015 to October/2016 (42 non-CRYO and 74 CRYO). There were 74 males and 69 (62%) patients had the underlying diagnosis of multiple myeloma. No differences were seen on the characteristics of the apheresis products and their viability. Engraftment was significantly faster in the non-CRYO group (p = 0.001). Febrile neutropenia and severe mucositis were lower in the non-CRYO group (40% vs 92% p = 0.0001 and 11% vs 64%, p = 0.001, respectively). In addition, length of hospitalization was 5 days shorter in the non-CRYO group (p = 0.0001). Overall responses and transplantation outcomes were similar. Our data demonstrate a clear advantage of the non-CRYO over CRYO auto-HSCT with faster engraftment, lower incidence of febrile neutropenia and shorter hospital stay after the transplantation procedure. These data are especially relevant for centers with high transplant activity or with limited resources.

Tacrolimus plus sirolimus with or without ATG as GVHD prophylaxis in HLA-mismatched unrelated donor allogeneic stem cell transplantation.

HLA-matched related or unrelated donors are not universally available. Consequently, patients can be offered hematopoietic stem cell transplantation (HSCT) from alternative donors, including mismatched unrelated donors (MMURD), known to cause a higher incidence of acute GVHD (aGVHD) and chronic GVHD. In vivo T-cell-depletion strategies, such as antithymocyte globulin (ATG) therapy, significantly decrease the risk of GVHD. We performed a multicenter, retrospective study comparing tacrolimus (TAC) and sirolimus (SIR) with or without ATG in 104 patients (TAC-SIR=45, TAC-SIR-ATG=59) who underwent MMURD HSCT. Use of ATG was associated with a lower incidence, albeit not statistically significant, of grades 2-4 aGVHD (46% vs 64%, P=0.09), no difference in grades 3-4 aGVHD (10% vs 15%, P=0.43), a trend for a lower incidence of moderate/severe chronic GVHD (16% vs 37%, P=0.09) and more frequent Epstein-Barr virus reactivation (54% vs 18%, P=0.0002). There were no statistically significant differences in 3-year overall survival (OS) (TAC-SIR-ATG=40% (95% confidence interval (CI)=24-56%) vs TAC-SIR=54% (95% CI=37-70%), P=0.43) or 3-year cumulative incidence of relapse/progression (TAC-SIR-ATG=40% (95% CI=28-58%) vs TAC-SIR=22% (95% CI=13-39%), P=0.92). An intermediate Center for International Blood & Marrow Transplant Research disease risk resulted in a significantly lower non-relapse mortality and better OS at 3 years. Our study suggests that addition of ATG to TAC-SIR in MMURD HSCT does not affect OS when compared with TAC-SIR alone.

Severe infections after single umbilical cord blood transplantation in adults with or without the co-infusion of CD34+ cells from a third-party donor: results of a multicenter study from the Grupo Español de Trasplante Hematopoyético (GETH).

BACKGROUND: Umbilical cord blood transplantation (CBT) is an established alternative source of stem cells in the setting of unrelated transplantation. When compared with other sources, single-unit CBT (sCBT) is associated with a delayed hematologic recovery, which may lead to a higher infection-related mortality (IRM). Co-infusion with the sCBT of CD34+ peripheral blood stem cells from a third-party donor (TPD) (sCBT + TPDCD34+) has been shown to markedly accelerate leukocyte recovery, potentially reducing the IRM. However, to our knowledge, no comparative studies have focused on severe infections and IRM with these 2 sCBT strategies. METHODS: A total of 148 consecutive sCBT (2000-2010, median follow-up 4.5 years) were included in a multicenter retrospective study to analyze the incidence and risk factors of IRM and severe viral and invasive fungal infections (IFIs). Neutrophil engraftment occurred in 90% of sCBT (n = 77) and 94% sCBT + TPDCD34+ (n = 71) recipients at a median of 23 and 12 days post transplantation, respectively (P < 0.01). RESULTS: The 4-year IRM was 24% and 20%, respectively (P = 0.7), with no differences at day +30 (5% and 4%, respectively) and day +100 (10% and 8%, respectively). In multivariate analysis early status of the underlying malignancy, cytomegalovirus (CMV)-seronegative recipient and high CD34+ cell content in the cord blood unit before cryostorage (≥1.4 × 10(5) /kg) were protective of IRM. Among the causes of IRM, bacterial infections and IFIs were more common in sCBT (15% vs. 4%), while CMV disease and parasitic infections were more common in the sCBT + TPDCD34+ cohort (5% vs. 16%). CONCLUSION: These data show that sCBT supported with TPDCD34(+) cells results in much shorter periods of post-transplant leukopenia, but the short- and long-term rates of IRM were comparable to those of sCBT, presumably because immune recovery is equally delayed in both graft types.

GVHD prophylaxis with sirolimus-tacrolimus may overcome the deleterious effect on survival of HLA mismatch after reduced-intensity conditioning allo-SCT.

Large studies, mostly based on series of patients receiving CSA/tacrolimus (TKR) plus MTX as immunoprophylaxis, have demonstrated a deleterious effect on survival of the presence of a single mismatch out of eight loci after allogeneic hematopoietic SCT (alloHSCT). We retrospectively analyzed a series of 159 adult patients who received sirolimus(SRL)/TKR prophylaxis after alloHSCT. We compared overall outcomes according to HLA compatibility in A, B, C and DRB1 loci at the allele level: 7/8 (n=20) vs 8/8 (n=139). Donor type was unrelated in 95% vs 70% among 7/8 vs 8/8 pairs, respectively (P=0.01). No significant differences were observed in 3-year OS (68 vs 62%), 3-year EFS (53 vs 49%) and 1-year non-relapse mortality (9 vs 13%). Cumulative incidence of grades II-IV acute GVHD (aGVHD) was significantly higher in 7/8 alloHSCT (68% vs 42%, P<0.001) but no significant differences were found for III-IV aGVHD (4.5% vs 11%), overall (35% vs 53%) and extensive (20% vs 35%) chronic GHVD in 7/8 vs 8/8 subgroups, respectively. In summary, the present study indicates favorable outcomes after alloHSCT using the combination of SRL/TKR combination as GVHD prophylaxis with OS in the range of 55-70%, and non-significant differences in overall outcomes, irrespective of the presence of any mismatches at obligatory loci.

Fungal and viral infections after allogeneic hematopoietic transplantation from unrelated donors in adults: improving outcomes over time.

Umbilical cord blood (CB) is increasingly used as an alternative source of stem cells in adult unrelated transplantation. Although registry studies report similar overall outcomes in comparison with BM/PB, comparative studies focusing on severe infections and infection-RM (IRM) with a long follow-up are scarce. A total of 434 consecutive unrelated transplants (1997-2009) were retrospectively analyzed to compare overall outcomes, incidence and risk factors of severe viral and invasive fungal infections in CB (n=65) vs BM/PB recipients (n=369). The 5-year OS was 38 vs 43%, respectively (P=0.2). CB transplantation (CBT) was associated with a higher risk of invasive aspergillosis (100-days-cumulative incidence 16 vs 6%, P=0.04) and CMV infection without differences in RM. No statistically significant differences were found regarding NRM (NRM of 38% in CB vs 37% in BM/PB at 1 year) nor IRM (30% in CB vs 27% in BM/PB at 1 year). In the overall population, NRM and IRM improved in more recent years. In adults who receive a single CBT, the risk of severe infections is increased when compared with unrelated BM/PB recipients, but mortality from infections is similar, leading to similar NRM and survival.

Improvement over the years of long-term survival in high-risk lymphoma patients treated with hematopoietic stem cell transplantation as consolidation or salvage therapy.

The role of hemopoietic stem cell transplantation (HSCT) is not well established in certain types of lymphoma, such as those with a high relapse risk or relapsing after initial therapy. New chemotherapeutic schemes and immunotherapy have improved survival of these patients. Nevertheless, there is not enough evidence regarding whether transplantation is the best therapeutic approach. Moreover, published data on long-term follow-up of high-risk lymphoma patients treated with HSCT are scarce. We analyzed 177 consecutive patients diagnosed with a high risk of relapse or with relapsed lymphoma who underwent HSCT after induction with standard chemotherapy in a tertiary academic center from 1989 to 2013. The median age was 40 years. Diagnoses were Hodgkin disease (n = 56), diffuse large B-cell lymphoma (n = 44), follicular lymphoma (n = 29), mantle cell lymphoma (n = 15), T-cell lymphoma (n = 18), and others (n = 15). Patients received either an autologous graft (n = 154) in first complete remission (1CR; n = 59) or more advanced stages (AS; n = 95), or an allogeneic graft (n = 23) in 1CR (n = 4) or AS (n = 19). In the autologous group, overall survival (OS) at 5 years was 57% and 75% in the periods 1989-2001 and 2002-2013, respectively (P = .05). Patients receiving an allogeneic graft presented an OS of 25% and 43% in the 2 periods. With a mean follow-up of 5 years (95% confidence interval 3.5-6.6), for patients receiving a transplant in 1CR, OS at 5 years was 80%, and for those receiving a transplant in AS it was 59% (P = .003). Nonrelapse mortality (NRM) at 5 years was 3.1% in the autologous group and 27.9% in the allogeneic group (P < .001). The main cause of NRM was infection (44%) in the whole cohort. All this leads to the conclusion that transplantation, as a therapeutic strategy, has shown a high long-term OS in this subgroup of patients with such a poor prognosis. OS improved over the years and reaching 1CR was a good prognostic feature. Infections were the main cause of NRM.

Clinical efficacy and safety of primary antifungal prophylaxis with posaconazole vs itraconazole in allogeneic blood and marrow transplantation.

Posaconazole has been recently approved for primary antifungal prophylaxis in patients with prolonged neutropenia after AML induction chemotherapy and patients with GVHD. We now present the first experience of the efficacy and safety of posaconazole during the early phase of post-allogeneic BMT (n=33; from June 2007), in comparison with itraconazole primary prophylaxis (n=16; up to May 2007). More patients receiving posaconazole were T-cell depleted (P=0.003). Groups were otherwise comparable in terms of age, sex, disease, neutrophil engraftment, incidence of GVHD, use of unrelated donors and type of conditioning. Safety data as well as the incidence of fever (84%) and persistent fever (27%) during the 100-day treatment period were comparable for both antifungal agents. Patients receiving posaconazole had a lower cumulative incidence of proven or probable invasive fungal disease, as defined by the European Organization for Research and Treatment of Cancer criteria (0 vs 12%; P=0.04), which associated with a higher probability of fungal-free survival (91 vs 56%; P=0.003) and an improved probability of OS (91 vs 63%; P=0.011) compared with patients receiving itraconazole. Our single-centre experience suggests that antifungal prophylaxis with posaconazole may lead to a better outcome than itraconazole for patients in the early high-risk neutropenic period after allogeneic BMT.

Empirical antifungal therapy in selected patients with persistent febrile neutropenia.

Universal empirical antifungal therapy (EAT) in patients with unexplained persistent febrile neutropenia (PFN) is the standard of care, but EAT could be applied in selected patients on the basis of clinical criteria and risk factors. A prospective interventional study was carried out to analyse the incidence and related mortality of invasive fungal infection (IFI) in patients with PFN according to whether or not EAT was indicated. EAT was indicated according to the following criteria: (a) severe sepsis or septic shock; (b) focused infection: lung, central nervous system, sinus, abdominal or skin; (c) individualized clinical decision in patients at high risk. Sixty-six (19%) of 347 episodes of febrile neutropenia fulfilled PFN criteria, 97% with a haematological malignancy. Just 26 (39.4%) were treated with EAT. The overall IFI incidence was 4.5%. In the group that received EAT, three patients developed IFI (11.5%), in comparison with none in the group that did not receive it (P=0.04, RR 2.7:1.9-3.8). IFI-related mortality was null in the group that did not receive EAT and 8% (two of 26 patients) in the group that received EAT. These data suggest that in patients with PFN, EAT in selected patients may be safe and avoid unnecessary antifungal therapy.

Lower respiratory tract respiratory virus infections increase the risk of invasive aspergillosis after a reduced-intensity allogeneic hematopoietic SCT.

We have analyzed the incidence and risk factors for the occurrence of invasive aspergillosis (IA) among 219 consecutive recipients of an allogeneic hematopoietic SCT after a reduced-intensity conditioning regimen (Allo-RIC). Twenty-seven patients developed an IA at a median of 218 days (range 24-2051) post-Allo-RIC, for a 4-year incidence of 13% (95% confidence interval 4-24%). In multivariate analysis, risk factors for developing IA were steroid therapy for moderate-to-severe graft vs host disease (GVHD) (Hazard Ratio (HR) 2.9, P=0.03), occurrence of a lower respiratory tract infection (LRTI) by a respiratory virus (RV) (HR 4.3, P<0.01) and CMV disease (HR 2.8, P=0.03). Variables that decreased survival after Allo-RIC were advanced disease phase (HR 1.9, P=0.02), steroid therapy for moderate-to-severe GVHD (HR 2.2, P<0.01), not developing chronic GVHD (HR 4.3, P<0.01), occurrence of LRTI by an RV (HR 3.4, P<0.01) and CMV disease (HR 2, P=0.01), whereas occurrence of IA had no effect on survival (P=0.5). Our results show that IA is a common infectious complication after an Allo-RIC, which occurs late post-transplant and may not have a strong effect on survival. An important observation is the possible role of LRTI by conventional RVs as risk factors for IA.

Correlation between high-resolution computed tomography and galactomannan antigenemia in adult hematologic patients at risk for invasive aspergillosis.

OBJECTIVES: To analyse the predominant radiological pattern of pulmonary lesions in adult hematologic patients at risk for invasive aspergillosis (IA) together with the results of serial serum Aspergillus galactomannan antigen testing (GM). MATERIAL AND METHODS: In a prospective study for patients at high risk of aspergillus pulmonary infection, serum GM were performed 2-3 times per week during the periods of high risk for IA and high-resolution CT (HRCT) was performed in case of abnormal chest X-ray (CXR) and/or persistent fever after 5 days of antibiotic treatment. Changes on HRCT scan were classified as airway IA and angioinvasive IA. IA was classified as proven or probable in accordance with the definitions stated by the European Organization for Research and Treatment of Cancer/Mycosis Study Group (EORTC-MS). Positive GM testing was not considered as microbiological criterion. RESULTS: 38 hematological patients were diagnosed of probable (n=28) or proven (n=10) IA. 55% patients had a neutrophil count less than 500 mm(-3) (n=21), and 37% patients > or =2 risk factors for IA. All probable IA were diagnosed by bronchoalveolar lavage (BAL). Proven IA was reached by positive histopathologic and culture results of samples obtained by autopsy (n=4), percutaneous (n=3) or transbronchial biopsy (n=3). 18 patients had airway IA, and 60% had a GM level > or =1.5. 20 patients were diagnosed of angioinvasive IA from which 80% had a GM level > or =1.5. CONCLUSION: Serum GM levels may be lower in patients with airway IA than in those with an angioinvasive form. HRCT and serum GM are complementary tests in the diagnosis of IA.

High rate of long-term survival for high-risk lymphoma patients treated with hematopoietic stem cell transplantation as consolidation or salvage therapy.

Patients with high-relapse-risk lymphomas or those relapsing after initial therapy have a limited probability of cure with conventional treatment. There is recent inconclusive evidence that, in such cases, intensification or salvage treatment with high-dose chemotherapy followed by hematopoietic stem cell transplantation (HSCT) increases the response rate and may improve survival. Nevertheless, published data on long-term follow-up of high-risk lymphoma patients treated with HSCT are scarce. We analyzed 101 consecutive patients receiving high-dose chemotherapy followed by HSCT after induction with standard chemotherapy. The median age was 38 years (range, 12-63 years). The diagnoses were Hodgkin's disease (n = 32), follicular lymphoma (n = 33), diffuse large B-cell lymphoma (n = 12), mantle cell lymphoma (n = 7), T-cell lymphoma (n = 14), and others (n = 3). Patients received either an autologous graft (n = 72) in first complete remission (1CR; n = 23) or in advanced stages (AS; n = 49), or an allogeneic graft (n = 29) in 1CR (n = 7) or in AS (n = 22). We concluded that transplant-related mortality was 2.7% for patients receiving an autologous HSCT and 27% for patients receiving an allogeneic HSCT. The main etiologies were graft-versus-host disease and infection in the allogeneic setting, and infection in the autologous setting. The probability of long-term (12-year) overall survival was 71%, higher than that described for high-relapse-risk lymphoma patients treated without HSCT and significantly better (P < .05) for patients who received the transplant in 1CR (89%) than in AS (65%). Finally, the probability of long-term survival was significantly better for patients treated with HSCT during the period from 2000-2007 (85%) compared with the period from 1989-1999 (72%).

Upper and lower respiratory tract infections by human enterovirus and rhinovirus in adult patients with hematological malignancies.

The impact of human enterovirus (HEV) and human rhinovirus (HRV) respiratory tract infections in adult patients with hematological malignancies has been infrequently reported. We retrospectively studied 31 patients with an upper or lower respiratory tract infection (URTI/LRTI) by HEV (n = 18) or HRV (n = 15). At onset, a LRTI was present in 6 (33%) and 2 (13%) episodes of HEV and HRV infections, respectively, with or without an URTI. Progression to LRTI (pneumonia) from prior URTI was seen in 1 (6%) and 2 (13%) HEV and HRV infections, respectively. The presence of lymphocytopenia (<0.5 x 10(9)/l) was higher in LRTI by HEV: 4/5 (80%) versus 2/10 (20%) by HRV. Eight of 18 (44%) patients with immunosuppression versus 3/14 (21%) patients with no immunosuppression at the onset of respiratory infection developed a LRTI. Thirteen per cent of patients had associated respiratory infections from bacteria, aspergillus, or CMV. Pulmonary aspergillosis was diagnosed in 20% of HRV infections. Three of 11 patients (27%) with a LRTI died, but pulmonary copathogens were also involved in all cases. In conclusion, HEV and HRV can be associated with LRTI in immunocompromised patients, although their direct impact on mortality is uncertain.

Evaluation of readmissions in hematopoietic stem cell transplant recipients.

BACKGROUND: There is a lack of information on health expenses caused by readmissions among hematopoietic stem cell transplant (HSCT) recipients. We analyzed the rate, causes, and evolution of hospitalization after HSCT. METHODS: We retrospectively studied 140 consecutive patients who received an autologous HSCT (n = 107; 76.4%) or an allogeneic HSCT (n = 33; 23.6%) in our institution from May 2001 through September 2004. RESULTS: There were 45 readmissions in 28 patients (20%): three (10%) in the autologous and 25 (90%), in the allogeneic HSCT cohorts. The overall median age was 35.3 +/- 13.5 years and 54% were women. Hematologic diseases were: multiple myeloma (n = 1, 4%), myelodysplastic syndrome (n = 2, 7%), acute lymphoblastic leukemia (n = 2, 7%), aplastic anemia (n = 2, 7%), chronic myeloid leukemia (n = 3, 11%), non-Hodgkin's lymphoma (n = 4, 14%), Hodgkin's disease (n = 4, 14%) and acute nonlymphoblastic leukemia (n = 10, 38%). The length of stay for each readmission was 25 +/- 21 days. The median day of readmission was +62.5 (range = +19 to +987); however, 75% occurred between days +30 and +70. The causes of hospitalization were: infections (n = 24, 54%), due to the graft (n = 14, 31%), graft failure (n = 4, 9%), coagulation disorders (n = 2, 4%), and second neoplasm (n = 1, 2%). Mortality due to the transplant was 10 patients (14%) including: graft-versus-host disease (n = 3), sepsis (n = 3), thrombotic thrombocytopenic purpura (n = 1), and relapse (n = 3). CONCLUSIONS: Although there was a frequent use of hospital resources (20%) after HSCT with patients hospitalized for a median of 25 days, it was beneficial since there were 86% survivors at 36 months follow-up.

Relapse of chronic myeloid leukemia after allogeneic stem cell transplantation: outcome and prognostic factors: the Chronic Myeloid Leukemia Subcommittee of the GETH (Grupo Español de Trasplante Hemopoyético).

In order to analyze the outcome of patients with chronic myeloid leukemia (CML) who relapse after allogeneic stem cell transplantation (SCT), we investigated data from 107 patients reported to the Spanish Registry, GETH. In all, 93 (87%) patients were treated after relapse; 36 out of 49 that failed to achieve a response received a second relapse-treatment, and seven a third one. At the last follow-up, the number of patients in molecular or cytogenetic remission was 29 and 13, respectively. Overall survival and progression-free survival after relapse were 53.6% (95% CI: 42.9--64.2) and 52% (95% CI: 41-63) at 5 years, respectively. In multivariate analysis, survival was significantly related to CML phase at relapse (cytogenetic or chronic phase vs advanced phases) and time from transplant to relapse (<1 vs >or=1 year). Patients with no adverse factors had a better survival compared with patients with one or two adverse features (65 vs 35 vs 0%, respectively). We conclude that a significant proportion of CML patients that relapse after transplantation can regain complete and long-lasting remissions with one or more salvage therapies. Disease stage at relapse and time from transplant to relapse should be taken into account when comparing results of different salvage treatments.