神田 善伸

BACKGROUND: We retrospectively compared the impact of the areas over and under the lymphocyte curve (L_AOC vs L_AUC) on cytomegalovirus (CMV) reactivation after allogeneic hematopoietic stem cell transplantation (HSCT). METHODS: Among 394 consecutive patients who underwent their first allogeneic HSCT at our center between 2007 and 2018, 301 patients met the inclusion criteria. L_AOC was calculated as the area over the lymphocyte curve during lymphopenia (absolute lymphocyte count [ALC] <700/μL). We calculated L_AOC and L_AUC from day 0 to day 15 (L_AOC15, L_AUC15) and from day 0 to day 30 (L_AOC30, L_AUC30). RESULTS: CMV antigenemia was defined as more than 3 cells/2 slides by the C10/11 method and detected in 204 cases (CMV reactivation) at a median of 39 days after HSCT. Although there were significant differences in L_AOC15, L_AOC30, L_AUC15, and L_AUC30 between patients with and without CMV reactivation, there was no difference in accuracy for predicting CMV reactivation between L_AOC and L_AUC. In a multivariate analysis, L_AOC15 and L_AUC15 were each identified as independent predictive factors for CMV reactivation, along with advanced age and CMV serostatus. However, ALC at day 14 or day 21 was as accurate as area-based indexes such as L_AOC15 and L_AUC15. L_AOC15 and L_AUC15 were significantly associated with longer duration of anti-CMV antiviral therapy while ALC was not. CONCLUSIONS: L_AOC15 and L_AUC15 had similar impacts on CMV reactivation. Although these area-based indexes were not superior to ALC for predicting CMV reactivation, they might predict patients who need longer duration of antiviral therapy more accurately.

Although measurable residual disease (MRD) at the time of allogeneic hematopoietic cell transplantation (allo-HCT) has been reported to be an important prognostic factor for Philadelphia chromosome (Ph)-positive acute lymphoblastic leukemia (ALL) during first complete remission (CR1), the prognostic impact of MRD is unclear during second CR (CR2). To clarify the impact of MRD for both CR1 and CR2, we analyzed data from a registry database including 1625 adult patients with Ph+ ALL who underwent first allo-HCT during either CR1 or CR2 between 2002 and 2017. Adjusted overall and leukemia-free survival rates at 4 years were 71% and 64%, respectively, for patients undergoing allo-HCT during CR1 with MRD-, 55% and 43% during CR1 with MRD+, 51% and 49% during CR2 with MRD-, and 38% and 29% during CR2 with MRD+. Although survival rates were significantly better among patients with CR1 MRD- than among patients with CR2 MRD-, no significant difference was observed in survival rate between patients with CR1 MRD+ and CR2 MRD-. Relapse rates after 4 years were 16% in patients with CR1 MRD-, 29% in CR1 MRD+, 21% in patients with CR2 MRD-, and 46% in patients with CR2 MRD+. No significant difference was identified in relapse rate between patients with CR1 MRD- and CR2 MRD-. CR2 MRD- was not a significant risk factor for relapse in multivariate analysis (hazard ratio, 1.26; 95% confidence interval, 0.69-2.29; P = .45 vs CR1 MRD-). MRD at time of allo-HCT was an important risk factor in patients with Ph+ ALL during both CR1 and CR2.

The safety and efficacy of allogeneic hematopoietic stem cell transplantation (HSCT) for paroxysmal nocturnal hemoglobinuria (PNH) remain unclear. Therefore, we retrospectively analyzed the outcomes of 42 adult patients with PNH who underwent allogeneic HSCT using the registry database of the Japan Society for Hematopoietic Cell Transplantation. The median patient age was 32.5 years. The number of packed red cell (PRC) transfusions was < 20 times in 19 patients and ≥ 20 times in 16; 7 patients had missing data. Stem cell sources were bone marrow (N = 15) or peripheral blood (N = 13) from a related donor or bone marrow (N = 11) and cord blood (N = 3) from an unrelated donor. The cumulative incidence of neutrophil engraftment at day 40 was 81%. Six patients died before engraftment, and the 6-year overall survival (OS) was 74%. The OS of patients with < 20 pretransplant PRC transfusions was significantly higher than that of patients with ≥ 20 pretransplant PRC transfusions (95% vs. 63%; P < 0.05). Moreover, the OS of patients aged < 30 years was significantly higher than that of patients aged ≥ 30 years (90% vs. 59%; P < 0.05). Allogeneic HSCT for PNH could provide favorable survival; however, pretransplant transfusion burden and patient age should be considered when deciding the timing of allogeneic HSCT.

A multicenter retrospective study was conducted to evaluate the clinical significance of preconditioning intervention (PCI) before allogeneic hematopoietic cell transplantation (HCT) in patients with acute myelogenous leukemia (AML) not in remission. The study cohort consisted of 519 patients classified according to the intensity (intensive/moderate) of PCI and their response to PCI. The group treated with PCI had higher blast counts in the peripheral blood (PB) and had a lower overall survival (OS) rate (P < .001) and higher nonrelapse mortality (NRM) rate (P = .035) compared with those without PCI (no PCI group). Approximately 40% of the patients (68 of 236) achieved a good response to PCI (good PCI group), and those patients had lower blast counts in the PB compared with the group with poor response to PCI (poor PCI group). OS in the good PCI group was comparable to that in the no PCI group and significantly better than that in the poor PCI group (hazard ratio, .54; 95% confidence interval, .39 to .77; P < .001). However, OS was significantly lower in patients with intensive/moderate PCI compared with the no PCI group. These results suggest that PCI increases NRM without decreasing the post-transplantation relapse rate, but may be beneficial for patients with lower blast counts in PB irrespective of its intensity.

Graft-versus-host disease is a major complication after allogeneic hematopoietic stem cell transplantation for hematological malignancies. Immunosuppressive drugs, such as anti-thymocyte globulin, alemtuzumab, and post-transplant cyclophosphamide, have been used to prevent graft-versus-host disease in HLA-mismatched haploidentical hematopoietic stem cell transplantation. Here, we investigated whether these drugs could ameliorate graft-versus-host disease without diminishing the graft-versus-leukemia effect by using a xenogeneic transplanted graft-versus-host disease/graft-versus-leukemia model. Anti-thymocyte globulin treatment diminished graft-versus-host disease symptoms, completely depleted the infiltration of inflammatory cells in the liver and intestine, and led to prolonged survival. By contrast, improvement after post-transplant cyclophosphamide treatment remained minimal. Alemtuzumab treatment modestly prolonged survival despite an apparent decrease of Tregs. In the graft-versus-leukemia model, 1.5 to 2.0 mg/kg of anti-thymocyte globulin and 0.6 to 0.9 mg/kg of alemtuzumab reduced graft-versus-host disease with minimal loss of graft-versus-leukemia effect. Mice treated with 400 mg/kg of post-transplant cyclophosphamide did not develop graft-versus-host disease or leukemia, but it was difficult to evaluate the graft-versus-leukemia effect due to the sensitivity of A20 cells to cyclophosphamide. Although the current settings provide narrow optimal therapeutic windows, further studies are warranted to maximize the benefits of each immunosuppressant.

BACKGROUND: Despite a strong association between acute graft-versus-host disease (GVHD) and cytomegalovirus reactivation (CMVR), the joint effect of acute GVHD and CMVR on nonrelapse mortality (NRM) has not been well studied. METHODS: We evaluated the impact of CMVR on NRM stratified according to the development of acute GVHD using a landmark method. This study included 6078 patients who received their first allogeneic hematopoietic cell transplantation (HCT) with a pre-emptive strategy for CMVR between 2008 and 2017. RESULTS: The cumulative incidences of grade II-IV acute GVHD (G24GVHD), CMVR by day 100, and CMV disease by day 365 were 37.3%, 52.1%, and 2.9%, respectively. Patients with G24GVHD were associated with the subsequent development of CMVR, and the presence of CMVR also increased the risk of G24GVHD. In a landmark analysis at day 65, the cumulative incidence of NRM at 1 year was 5.4%, 10.0%, 13.9%, and 19.7% in patients with G24GVHD-/CMVR-, G24GVHD-/CMVR+, G24GVHD+/CMVR-, and G24GVHD+/CMVR+, respectively. In a multivariate analysis, CMVR was respectively associated with an increased risk of NRM by day 365 in patients without G24GVHD (HR [hazard ratio], 1.59, 95% CI, 1.24-2.05, P < 0.001) and with G24GVHD (HR, 1.34, 95% CI, 1.06-1.70, P = 0.014), but the interaction between G24GVHD and CMVR was not significant (P = 0.326). Subgroup analyses suggested that the joint effect of acute GVHD and CMVR might vary according to the baseline characteristics. CONCLUSIONS: These data regarding the close relationship between acute GVHD and CMVR should provide important implications for the treatment strategy after HCT.

This study aimed to analyze the factors associated with outcomes of bone marrow transplantation (UR-BMT) or cord blood stem cell transplantation from unrelated donors (UR-CBT). We assessed the time from diagnosis to transplantation among acute myeloid leukemia (AML) patients with intermediate- or poor-risk cytogenetics to identify the potential clinical efficacy of transplantation. We retrospectively analyzed 5331 patients who received UR-BMT or UR-CBT between 2008 and 2017. Patients were divided into four groups according to time from diagnosis to transplantation: (1) UR-BMT and > 5 months (n = 2353), (2) UR-BMT and ≤ 5 months (n = 379), (3) UR-CBT and > 5 months (n = 1494), and (4) UR-CBT and ≤ 5 months (n = 1106). There was no difference in overall survival (OS) for transplantation at ≤5 months and > 5 months in patients with first complete remission for both UR-BMT and UR-CBT, but OS in patients with primary induction failure (PIF) and transplantation at ≤ 5 months was significantly higher in the UR-CBT group compared with that at >5 months (P < 0.001). Multivariate Cox regression analysis also showed that transplantation at >5 months in patients with PIF was an independent predictor of poorer OS. Therefore, UR-CBT at ≤ 5 months after diagnosis is an alternative option for AML patients with PIF.

BACKGROUND: We retrospectively evaluated the association between the D-index, which reflects both the depth and duration of neutropenia, and proven/probable invasive fungal disease (IFD) early after allogeneic hematopoietic stem cell transplantation (HSCT) at our center (n = 394). METHODS: The D-index was defined as the area over the neutrophil curve during neutropenia. The cumulative D-index from the start of neutropenia until the development of infection (c-D-index) was also evaluated as a real-time assessment of neutropenia. RESULTS: There were 19 cases of early proven/probable IFD before and within 1 week after engraftment. Fifteen cases (78.9%) were seen as breakthrough infection while on empiric (n = 7), preemptive (n = 4) or prophylactic (n = 4) antifungal administration with mold-active agents. The c-D-index and lower performance status were identified as independent significant predictive factors for IFD. A receiver operating characteristic (ROC) curve analysis showed that the D-index and c-D-index were more accurate than the simple duration of neutropenia and as accurate as the duration of profound neutropenia for predicting IFD. The sensitivity, specificity, and positive and negative predictive values of the c-D-index using an appropriate cutoff (CO) value (10 644) determined by ROC curve analysis were 73.1%, 63.2%, 9.1%, and 97.9%, respectively. The advantage of the c-D-index to cumulative days of neutropenia in terms of positive and negative predictive values seemed to be small. CONCLUSIONS: The appropriate CO value for the c-D-index for predicting IFD was as high as 10 644 in allogeneic HSCT with a more frequent use of empiric antifungal therapy. The c-D-index is useful for assessing the risk of breakthrough IFD.

The pulmonary function test (PFT) is an important test for risk stratification before allogeneic transplantation (allo-HCT). However, it might be preferable to avoid PFT as much as possible in the recent era of coronavirus disease 2019 (COVID-19), because PFT requires forced expirations and might produce aerosols, increasing the risk of COVID-19 transmission. Therefore, we tried to predict normal PFT results before allo-HCT based on computed tomography (CT) findings. This study included 390 allo-HCT recipients at our center for whom lung CT images and PFT results before allo-HCT were available. Abnormal CT findings were less likely to be observed in the normal PFT group (47.0% versus 67.4%, P = .015), with a high negative predictive value of 92.9%. In a multivariate analysis, normal CT was significantly associated with normal PFT (odds ratio, 2.47; 95% confidence interval, 1.22 to 4.97; P = .012). A model for predicting normal PFT was constructed based on the results of a multivariate analysis, and the area under the curve of the receiver operating characteristic analysis was 0.656, which gave a sensitivity of 45.5% and a specificity of 86.0%. The relatively high specificity of the model suggested that PFT can be omitted in patients with normal CT findings before allo-HCT.

The long-term outcomes of allogeneic hematopoietic cell transplantation (HCT) with reduced-intensity conditioning (RIC) remain inconclusive. To address this issue, we conducted a nationwide registry-based study of patients with acute myeloid leukemia (AML) age 50 years or older who underwent allogeneic HCT in complete remission using RIC (n = 284) or myeloablative conditioning (MAC, n = 190) between 2002 and 2007. The median follow-up period for surviving patients was 10.1 years for RIC recipients and 10.4 years for MAC recipients. The 10-year probabilities of overall survival, relapse, and non-relapse mortality were 36.4%, 30.0%, and 35.7% for RIC recipients, and 39.8%, 26.3%, and 35.5% for MAC recipients, respectively. Multivariate analysis revealed that the conditioning intensity did not affect overall mortality (P = 0.184), relapse (P = 0.904), or non-relapse mortality (P = 0.387). For the 218 patients qualifying for propensity score-matched pairing (109 pairs), RIC was found to be associated with similar survival (P = 0.095) and relapse (P = 0.467), and significantly lower non-relapse mortality (P = 0.046) compared with MAC. Our results confirm the long-term efficacy of RIC allogeneic HCT for older patients with AML and mitigate concerns over an increase in late relapse.

Graft failure (GF) is a life-threatening complication after allogeneic stem cell transplantation (SCT). Although salvage SCTs can be performed with haploidentical donor (HID) or cord blood (CB), no study has compared the performances of these two sources. Using nationwide registration data, we compared the transplant outcomes of patients who developed GF and underwent salvage transplantation from HID (n = 129) and CB (n = 570) from 2007 to 2016. The HID group demonstrated better neutrophil recovery (79.7 vs. 52.5% at 30 days, P < 0.001). With a median follow-up of 3 years, both groups demonstrated similar overall survival (OS) and nonrelapse mortality (NRM; 1-year OS, 33.1 vs. 34.6% and 1-year NRM, 45.1 vs. 49.8% for the HID and CB groups). After adjustments for other covariates, OS did not differ in both groups. However, HID was associated with a lower NRM (hazard ratio, 0.71; P = 0.038) than CB. The incidence of acute graft-versus-host disease (GVHD)-related deaths was significantly higher in the HID group, although infection-related deaths were observed more frequently in the CB group. HID may be a promising salvage SCT option after GF due to its faster engraftment and low NRM.

We conducted a nationwide questionnaire-based survey in 2019 following 2001, 2007 and 2013 surveys to clarify the real-world management of infection during chemotherapy for acute leukemia in Japan. An online questionnaire was sent through SurveyMonkey® to member institutions of the Japan Adult Leukemia Study Group in June 2019. The questionnaire consisted of 52 multiple-choice questions covering prophylactic measures, screening and diagnostic tests, empirical antibiotic therapy, antifungal management, the usage of granulocyte-colony stimulating factor, and vaccinations against influenza and pneumococcus during intensive chemotherapy for acute leukemia. Questions associated with antimicrobial stewardship were also included. Usable responses were received from 163 of 218 (74.8%) institutions. Approximately, half (52.2%) of the institutes did not have infectious disease department. As antibiotic prophylaxis, fluoroquinolones (62%) were most commonly used in induction chemotherapy for acute myeloid leukemia. No prophylaxis accounted for 19% of the institutions, which has gradually increased compared to previous surveys. In empirical antibiotic therapy for febrile neutropenia, monotherapy with β-lactam antibiotics was the most commonly used first-line therapy. De-escalation was not considered in 42.2% of the institutions. In conclusion, this study clarified the real-world management of infection during intensive chemotherapy for acute leukemia in 2019 and raised future issues in Japan.

Hepatitis B virus (HBV) reactivation reportedly occurs frequently after hematopoietic stem cell transplantation (HSCT) in resolved HBV-infected patients. Here, 50 patients with resolved HBV infections and scheduled to undergo HSCT were enrolled; all subjects were vaccinated with three doses of hepatitis B vaccine 12 months after HSCT and the incidence of HBV reactivation was monitored. The patients' characteristics were: median age, 61 (34-72) years; male/female, 27/19; allogeneic/autologous, 40/6; bone marrow/peripheral blood stem cells/cord blood, 26/16/4. Of the 46 patients who underwent HSCT, 19 were excluded and did not make it to vaccination due to relapse of underlying disease, HBV reactivation within 12 months of HSCT, or transfer of patients. The remaining 27 were vaccinated 12 months after HSCT and monitored for 2 years. Six showed HBV reactivation, with a 2-year cumulative reactivation incidence of 22.2%; the same incidence was 27.3% only in allogeneic HSCT patients. Factors associated with HBV reactivation included the discontinuation of immunosuppressants (P = 0.0379) and baseline titers of antibody against hepatitis B surface antigen (P = 0.004). HBV reactivation with vaccination following HSCT could occur despite maintenance of serum anti-HBs at more than protective levels.

Posttransplant lymphoproliferative disorder (PTLD) is a serious complication after hematopoietic stem cell transplantation (HSCT). Several studies of risk factors for PTLD have been reported; however, the probability of, and risk factors for, PTLD in patients with lymphoma is unknown. Japanese nationwide transplant registry data from 5270 patients with lymphoma after allogeneic HSCT were analyzed. Mature B-cell, T/NK-cell, and T-cell lymphoblastic subtypes accounted for 49%, 26%, and 9.6% of lymphoma cases, respectively. Rituximab was used in 1678 lymphoma patients, most of whom (89%) received HSCT for mature B-cell lymphoma. Thirty-one patients with lymphoma developed PTLD, representing a probability of 0.77% at 2 years post-HSCT, which did not differ significantly from that in patients with other diseases (P = .98). Year of HSCT after 2010 (hazard ratio [HR] = 5.6, 95% confidence interval [CI], 1.48-21.3), antithymocyte globulin (ATG) use in the conditioning regimen (HR = 4.5, 95% CI, 1.61-12.5), and no rituximab use before HSCT (HR = 3.2, 95% CI, 1.26-7.90) were identified as risk factors for PTLD. Probabilities of PTLD at 1 year post-HSCT according to rituximab and ATG use were 0.23% (rituximab+, ATG-), 0.75% (rituximab-, ATG-), 1.25% (rituximab+, ATG+), and 3.53% (rituximab-, ATG+). Regarding lymphoma subtypes, patients with mature B-cell lymphoma had the lowest incidence of PTLD (0.35% at 2 years). Among high-risk patients receiving ATG, the mortality rate due to infection was elevated in those previously treated with rituximab (22%) relative to those without (14%); however, the difference was not significant (P = .10). Rituximab use before HSCT significantly reduces the risk of PTLD. Adding rituximab to the conditioning regimen is potentially a good strategy to prevent the development of PTLD in high-risk patients.

The major limitation of cord blood transplantation (CBT) for adults remains the delayed hematopoietic recovery and higher incidence of graft failure, which result in a higher risk of early mortality in CBT. We evaluated early overall survival (OS), non-relapse mortality (NRM), neutrophil engraftment, acute graft-vs-host disease, and cause of early death among 9678 adult patients who received single-unit CBT in Japan between 1998 and 2017. The probability of OS at 100 days was 64.4%, 71.7%, and 78.9% for the periods 1998 to 2007, 2008 to 2012, and 2013 to 2017, respectively (P < .001). The cumulative incidences of NRM at 100 days during the same period were 28.3%, 20.8%, and 14.6%, respectively (P < .001). The cumulative incidences of neutrophil engraftment were also improved during the same period (P < .001). The most common cause of death within 100 days after CBT was bacterial infection in 1998 to 2007 and primary disease in the latter two time periods. Across the three time periods, the proportions of deaths from bacterial and fungal infection, graft failure, hemorrhage, sinusoidal obstructive syndrome, and organ failure decreased in a stepwise fashion. Landmark analysis of OS and NRM after 100 days showed that OS did not change over time in the multivariate analysis. Our registry-based data demonstrated a significant improvement of early OS after CBT for adults over the past 20 years. The landmark analysis suggested that improvement of early mortality could lead to an improvement of long-term OS after CBT.

Pulmonary complications are fatal adverse events after allogeneic hematopoietic cell transplantation (allo-HCT). On the other hand, smoking is a well-known risk factor for various pulmonary diseases and also increases the incidence of pulmonary complications and overall mortality in allo-HCT recipients. In this study, we retrospectively assessed the impact of smoking intensity on survival outcomes. This study included consecutive allo-HCT recipients at our center between June 2007 and May 2019 whose smoking profiles were available (n = 408); they were divided into high (pack-years >10, n = 171) and low (pack-years ≤10, n = 231) pack-years groups. In univariate analyses, nonrelapse mortality (NRM) and overall survival (OS) were significantly inferior in the high pack-years group (1-year NRM 26.6% versus 13.9%, P < .001; 1-year OS 58.4% versus 70.1%, P = .0067). However, this association was not observed in multivariate analyses. In subgroup analyses according to sex, the survival outcomes in the high pack-years group were significantly inferior in males (NRM hazard ratio [HR], 2.24 [95% confidence interval (CI), 1.23 to 4.07], P = .0082; OS HR, 1.54 [95% CI, 1.04 to 2.28], P = .031), but not in females (NRM HR, 0.587 [95% CI, 0.241 to 1.43], P = .24; OS HR, 0.689 [95% CI, 0.400 to 1.19], P = .18). In summary, high pack-years were associated with inferior survival of allo-HCT recipients, especially in males.

The optimal pre-transplant conditioning for aplastic anemia (AA) remains unclear. We performed a prospective study on allogeneic transplantation from a related or unrelated donor for adult patients with AA. We assessed whether reduced-dose cyclophosphamide (CY) could decrease toxicity while maintaining engraftment, and low-dose thymoglobulin could safely prevent graft-vs-host disease (GVHD). The pre-transplant conditioning regimen consisted of fludarabine 120 mg/m2 , CY 100 mg/kg, and thymoglobulin 2.5 mg/kg with or without 2 Gy of total body irradiation. Twenty-seven patients with a median age of 36 years were analyzed. Sixteen patients received graft from related donors. The stem cell source was bone marrow in 26 patients. All of the patients but one, who died early, achieved neutrophil engraftment at a median of 19 days. Mixed chimerism was observed in six and five patients at days 30 and 90, respectively. Only one patient experienced secondary engraftment failure with complete donor-type chimerism. None of the patients developed severe acute GVHD. The cumulative incidence of chronic GVHD was 37.7% at 1 year. The overall survival rate was 96.3% at 1 year and 3 years. A high EB virus-DNA load was detected in one patient at days 60. No one developed EBV-lymphoproliferative disorder within a year. The results suggest that the conditioning regimen in this study was safe and effective. However, relatively high incidence of chronic GVHD needs further improvement.

Natural killer (NK) cells expressing killer cell immunoglobulin-like receptors (KIRs) can recognize specific HLA class I molecules as their ligands. By studying a large Japanese transplant registry, we compared transplant outcomes between patients heterozygous for HLA-CAsn80/CLys80 (HLA-C1/C2) and those homozygous for HLA-C1 (HLA-C1/C1) among patients who had undergone HLA-matched hematopoietic stem cell transplantation (HSCT). A high frequency of KIR2DL1 with strong HLA-C2 binding capacity and a low frequency of HLA-C2 and KIR haplotype B are characteristic of the Japanese population. In our previous report, HLA-C1/C1 patients with myeloid leukemia were less likely to relapse than HLA-C1/C2 patients. We newly assessed 2884 patients with acute lymphoblastic leukemia (ALL) who received HLA-matched allogeneic HSCT and analyzed their leukemia relapses by using adjusted competing-risk methods. HLA-C1/C1 patients with ALL experienced significantly higher relapse rates than HLA-C1/C2 patients (hazard ratio [HR] = 1.55, P = .003), contrary to our results in patients with myeloid leukemia. We allocated patients with ALL to several subgroups and found a higher frequency of relapse (HR >1.8) in the HLA-C1/C1 group than in the HLA-C1/C2 group among patients with Ph-negative ALL, those who had no cytomegalovirus reactivation, those who received transplants from donors who were aged 41 years or older, and those who experienced acute graft-versus-host disease, especially if it required systemic treatment. One interpretation of our results is that KIR2DL1-positive NK cells disrupt T cells, antigen-presenting cells, or both from working efficiently in transplant immunity in HLA-C1/C1 patients with ALL. Another is that KIR2DS1-positive NK cells directly attack HLA-C2-positive ALL blasts in HLA-C1/C2 patients. Whether HLA-C2 can cause recurrence to decrease or increase in patients depending on the disease (ALL or myeloid leukemia) will be a very important finding. We hope that our results will provide clues to the real mechanisms behind relapse after transplantation in patients with different HLA profiles.

Some studies support the hypothesis that HLA genes and haplotypes evolved by natural selection through their protective abilities against specific infectious pathogens. However, very little is known regarding the impact of high-frequency HLA haplotypes on the risk of relevant infectious diseases among a given ethnic group. We evaluated the impact of high-frequency HLA haplotypes on cytomegalovirus (CMV) reactivation and infection in allogeneic hematopoietic stem cell transplantation (allo-HSCT) in a Japanese population as a model of infectious disease that has coexisted with humans. A total of 21,127 donor-patient pairs were analyzed. HLA-A-B-DRB1 haplotypes were estimated using the maximum probability algorithm. Seven haplotypes with >1% frequency were defined as high-frequency haplotypes (HfHPs). Homozygotes of HfHP and heterozygotes had significantly lower risk of CMV reactivation and infection (hazard ratio [HR] = 0.88, P = .009 and HR = 0.93, P = .003, respectively) than homozygotes of low-frequency HLA haplotypes (LfHPs). In subgroup analyses of a different donor source, these associations were statistically significant in unrelated donor transplants. Finally, CMV risk for homozygotes and heterozygotes of each HfHP was compared with that of homozygotes of LfHPs. The 2 most predominant HfHP groups (A*24:02-B*52:01-DRB1*15:02 group and A*24:02-B*07:02-DRB1*01:01 group) had a significantly lower risk of CMV reactivation and infection (HR = 0.86, P < .001 and HR = 0.91, P = .033, respectively). Our findings suggest that HfHPs may be protective against CMV reactivation and infection and that increased care regarding CMV reactivation and infection may be necessary for patients with LfHP after allo-HSCT.

Acute graft-versus-host disease (aGVHD) is 1 of the critical complications that often occurs following allogeneic hematopoietic stem cell transplantation (HSCT). Thus far, various types of prediction scores have been created using statistical calculations. The primary objective of this study was to establish and validate the machine learning-dependent index for predicting aGVHD. This was a retrospective cohort study that involved analyzing databases of adult HSCT patients in Japan. The alternating decision tree (ADTree) machine learning algorithm was applied to develop models using the training cohort (70%). The ADTree algorithm was confirmed using the hazard model on data from the validation cohort (30%). Data from 26 695 HSCT patients transplanted from allogeneic donors between 1992 and 2016 were included in this study. The cumulative incidence of aGVHD was 42.8%. Of >40 variables considered, 15 were adapted into a model for aGVHD prediction. The model was tested in the validation cohort, and the incidence of aGVHD was clearly stratified according to the categorized ADTree scores; the cumulative incidence of aGVHD was 29.0% for low risk and 58.7% for high risk (hazard ratio, 2.57). Predicting scores for aGVHD also demonstrated the link between the risk of development aGVHD and overall survival after HSCT. The machine learning algorithms produced clinically reasonable and robust risk stratification scores. The relatively high reproducibility and low impacts from the interactions among the variables indicate that the ADTree algorithm, along with the other data-mining approaches, may provide tools for establishing risk score.

Macrophages play a crucial role in the pathogenesis of chronic graft-versus-host disease (cGVHD). We hypothesized that galectin-3, Mac-2 binding protein (M2BP), or Wisteria floribunda agglutinin (WFA)+-M2BP, called M2BP glycan isomer (M2BPGi), might contribute to macrophage activation, and fibrosis would be associated with cGVHD and nonrelapse mortality (NRM) in hematopoietic stem cell transplant (HSCT) recipients. Patients who underwent their first allogeneic HSCT and survived for >180 days without relapse were included. The predictive potential of the 3 markers for NRM was assessed using the discovery cohort (n = 55) and validation cohort 1 (n = 55). When we used the threshold determined by a receiver operating characteristics curve analysis in the discovery cohort, only M2BPGi at day +180 was significantly associated with a higher NRM in the discovery cohort (15.0% vs 0.0% at 5 years, P = .001) and in validation cohort 1 (34.0% vs 8.4% at 5 years, P = .014). This result was confirmed in validation cohort 2 (n = 50). M2BPGi was not increased in healthy individuals or in patients who received autologous HSCT. In the entire cohort (N = 110), M2BPGi was significantly related to liver cGVHD but not to other organ involvement. In multivariate analyses, M2BPGi was an independent risk factor for NRM. In immunofluorescence staining of autopsy cases, WFA+-M2BP-positive macrophages were found only in the liver sections with cGVHD. In conclusion, M2BPGi could be a promising predictor of late NRM after HSCT and was associated with liver involvement.

The prognostic significance of FLT3-tyrosine kinase domain (TKD) mutations remains unknown. To investigate the prognostic impact of FLT3-TKD, 676 de novo acute myeloid leukemia (AML), we retrospectively analyzed cases and conducted a review of the literature. Of the 676 de novo AML cases, 34 (5.0%) were FLT3-TKD-positive; both FLT3-TKD and FLT3-ITD were noted in only two cases (0.3%). Although no significant differences in relapse-free survival (RFS) were noted, FLT3-TKD-positive cases showed better prognoses than FLT3-ITD-positive cases (FLT3-TKD versus FLT3-ITD, p = 0.152). For overall survival (OS), although FLT3-TKD-positive cases showed prognoses similar to those for FLT3-WT cases, their prognoses were significantly better than those of FLT3-ITD-positive cases (FLT3-TKD versus FLT3-ITD, p = 0.032). Moreover, the 5-year OS for FLT3-TKD-positive cases was 46.1%, indicating that this as an intermediate prognosis group. Although no reports from Asia have indicated a frequency of FLT3-TKD-positive cases > 10%, several reports from Europe and the United States have indicated frequencies > 10%. This suggests the possibility that FLT3-TKD-positive cases are less common in Asia than in Europe and the United States. We anticipate that in the future, the appearance of targeting agents, such as FLT3 inhibitors, will improve the prognosis of FLT3-TKD-positive AML relative to that of FLT3-WT AML.

To investigate optimal unit selection for umbilical cord blood transplantation (UCBT), we conducted a registry-based study of 1355 adults with acute myeloid leukemia in first or second complete remission who underwent single-unit UCBT. To be eligible for analysis, UCB units had to contain a total nucleated cell (TNC) dose of 2.0 × 107/kg or higher and present at least a 4/6-match for HLA-A, -B, and -DR antigens in line with clinical practice in Japan, both of which are less stringent criteria than those used in Western countries. Neither TNC dose nor the degree of HLA matching affected survival (P = 0.138 and P = 0.696, respectively). As for HLA-A, -B antigens and -DRB1 allele, better HLA matching was associated with lower non-relapse mortality (P = 0.011) but higher relapse (P = 0.046), resulting in no improvement in survival (P = 0.680). Taking the allele level for each HLA-A, -B, and -DRB1 into consideration was less useful for predicting non-relapse mortality (P = 0.198). These findings suggest that the less stringent criteria for UCB unit selection are acceptable for Japanese patient population and perhaps even more beneficial in terms of providing a better chance to find a suitable UCB unit.

The choice of alternative donor is a major issue in allogeneic hematopoietic stem cell transplantation (HSCT) for patients with primary myelofibrosis (PMF) without an HLA-matched related donor. We conducted this retrospective study using the Japanese national registry data for 224 PMF patients to compare the outcomes of first allogeneic HSCT from HLA-matched related donor bone marrow (Rtd-BM), HLA-matched related donor peripheral blood stem cells (Rtd-PB), HLA-matched unrelated donor bone marrow (UR-BM), unrelated umbilical cord blood (UR-UCB), and other hematopoietic stem cell grafts. Nonrelapse mortality (NRM) rates at 1 year after Rtd-BM, Rtd-PB, UR-BM, UR-UCB, and other transplantations were 16%, 36%, 30%, 41%, and 48%, respectively. Multivariate analysis identified UR-UCB transplantation, other transplantation, frequent RBC transfusion before transplantation, and frequent platelet (PLT) transfusion before transplantation as predictive of higher NRM. Relapse rates at 1 year after Rtd-BM, Rtd-PB, UR-BM, UR-UCB, and other transplantation were 14%, 17%, 11%, 14%, and 15%, respectively. No specific factor was associated with the incidence of relapse. Overall survival (OS) at 1 and 4 years after Rtd-BM, Rtd-PB, UR-BM, UR-UCB, and other transplantation were 81% and 71%, 58% and 52%, 61% and 46%, 48% and 27%, and 48% and 41%, respectively. Multivariate analysis identified older patient age, frequent RBC transfusion before transplantation, and frequent PLT transfusion before transplantation as predictive of lower OS. In conclusion, UR-UCB transplantation, as well as UR-BM transplantation, can be selected for PMF patients without an HLA-identical related donor. However, careful management is required for patients after UR-UCB transplantation because of the high NRM. Further studies including more patients after HLA-haploidentical related donor and HLA-mismatched unrelated donor transplantation would provide more valuable information for patients with PMF when making decisions regarding the choice of alternative donor.

Outcomes of allogeneic hematopoietic stem cell transplantation (HSCT) for hepatitis-associated aplastic anemia have not been fully evaluated. In the present study, the outcomes of 37 adult patients with hepatitis-associated aplastic anemia who underwent allogeneic HSCT were retrospectively analyzed using the registry database of Japan Society for Hematopoietic Cell Transplantation. The median age of the patients was 24 years (range, 16-61). The median period between diagnosis of hepatitis-associated aplastic anemia and HSCT was 6.0 months (range, 0.5-430.8). Stem cell sources were bone marrow (N = 19) or peripheral blood stem cells (N = 5) from an HLA-identical sibling or bone marrow (N = 11) and cord blood (N = 2) from an unrelated donor. The majority of conditioning regimens were fludarabine-based or high-dose cyclophosphamide-based. In all but 2 cases of early death, neutrophil engraftment was achieved. At the time of analysis, 32 patients were alive, with a median follow-up of 54.1 months. Five-year overall and failure-free survival rates were 86.0% (95% CI, 69.4-93.9%) and 75.0% (95% CI, 57.4-86.2%), respectively. Despite the heterogeneity in transplant procedures in a small number of patients, these results suggest that allogeneic HSCT is safe for use in hepatitis-associated aplastic anemia with a low rate of transplant-related mortality.

It remains unclear how specific innovations in allogeneic hematopoietic cell transplantation (HCT) attained over the past decades have contributed to improvement in transplantation outcomes. To address this question, we conducted a registry-based study of adults with acute myeloid leukemia in first or second complete remission who underwent allogeneic HCT between 1994 and 2013 from a sibling (N = 1600) or unrelated (N = 2113) donor matched at the antigen level for HLA-A, -B, and -DR. Preliminary analysis led us to focus on comparisons between the 1994-2006 and 2007-2013 periods. Significant improvement in survival was observed in the later cohort compared to the earlier cohort for unrelated HCT (P = 0.004), but not for related HCT (P = 0.767). The improvement in unrelated HCT was solely due to diminished non-relapse mortality (P = 0.001), while incidence of relapse did not change over time (P = 0.934). The percentage of patients receiving transplants from 8/8-matched unrelated donors was significantly higher in the later cohort (P < 0.001), and their survival was significantly better than that of those undergoing mismatched unrelated HCT (P = 0.022). These findings suggest that advances in HLA-typing technology have been vital for improvement in transplantation outcomes.

We compared the effect of HLA single-antigen and single-allele mismatched unrelated bone marrow transplantation (UBMT) without in vivo/ex vivo T cell depletion. Becasue a single DRB1 mismatch is preferred among 1-allele or 1-antigen mismatched donors, we performed mismatched allele- or antigen-specific analyses with a single DRB1 mismatch as the reference. In adjusted comparison by multivariate analyses, an HLA-DRB1 single-allele mismatch resulted in a decreased risk of nonrelapse mortality (NRM; relative risk [RR], 1.33; 95% confidence interval [CI], 1.08 to 1.63, P = .006) compared with an HLA-DR single-antigen mismatch and conferred a decreased risk of NRM (RR, 1.25; 95% CI, 1.01 to 1.57; P = .025) and overall mortality (RR, 1.16; 95% CI, 1.00 to 1.37; P = .046) compared with an HLA-C single-antigen mismatch. Relative to an HLA-DRB1 single-allele mismatch, 2-mismatch transplants, including those with 1 or more antigen mismatches, resulted in a significantly increased risk of NRM (1-antigen/1-allele mismatch: RR, 1.68; 95% CI, 1.03 to 2.05; P < .001; 2-antigen mismatch: RR, 1.58; 95% CI, 1.04 to 2.02; P = .001) and overall mortality (1-antigen/1-allele mismatch: RR, 1.27; 95% CI, 1.09 to 1.47; P = .002; 2-antigen mismatch: RR, 1.27; 95% CI, 1.03 to 1.57; P = .02). NRM correlated with the combined number of mismatches and allele or antigen mismatches, with rates of 22%, 27%, 32%, 31%, and 38% at 4years for full match, single-allele mismatch, single-antigen mismatch, 2-allele mismatch, and 2 mismatches that included an antigen mismatch, respectively. Our results support the preference for an allele mismatch rather than an antigen mismatch in unrelated bone marrow donors with 1 DR mismatch or 2 mismatches for T cell-replete UBMT.

To reduce post-transplant relapse, acute myeloid leukemia (AML) type remission induction chemotherapy has been attempted to reduce disease burden before allogeneic hematopoietic cell transplantation (HCT) in patients with advanced myelodysplastic syndrome (MDS). However, the efficacy of induction chemotherapy before HCT is unclear. We retrospectively analyzed the Japanese registration data of 605 adult patients, who had received allogeneic HCT for advanced MDS between 2001 and 2016, to compare the post-transplant relapse between patients who received induction chemotherapy followed by allogeneic HCT and those who received upfront HCT. Propensity score matching identified 230 patients from each cohort. There were no significant differences in overall survival and non-relapse mortality between the two groups. The cumulative incidence of relapse was significantly higher in patients who received induction chemotherapy than those who received upfront HCT. In the subgroup analyses, upfront HCT had a significantly reduced relapse incidence among patients with poor cytogenetics, those with higher international prognostic scoring system at diagnosis, and those who received reduced-intensity conditioning. Our results suggested that AML type remission induction chemotherapy before HCT did not improve post-transplant relapse and survival for adult patients with advanced MDS. Upfront HCT is preferable for patients with a poor karyotype.

Antithymocyte globulin (ATG) is widely used to reduce acute graft-versus-host disease (aGVHD) and chronic GVHD (cGVHD). To clarify the different impacts of ATG for conditioning across different donor types, we retrospectively analyzed patients with acute leukemia (n = 6617) who underwent hematopoietic stem cell transplantation between 2008 and 2015 with ATG (n = 279) or without ATG (n = 6338). Because thymoglobulin is the only ATG drug approved for GVHD prophylaxis in Japan since September 2008, we included thymoglobulin alone in the present analysis. The survivors' median follow-up time was 1081 days. Patients were categorized into 5 groups: cord blood (CB; n = 1915), matched related donor (n = 1772), 1-antigen mismatched related donor (1-MMRD; n = 225), matched unrelated donor (MUD; n = 1742), and 1-allele mismatched unrelated donor (1-MMUD; n = 963). In multivariate analysis, ATG decreased overall survival (hazard ratio [HR], 1.403; P = .054) and GVHD-free/relapse-free survival (GRFS) (HR, 1.458; P = .053) in association with increased nonrelapse mortality (NRM) (HR, 1.608; P =03) with CB, whereas it improved GRFS (HR, 0.515; P < .01) and decreased grades II to IV aGVHD (HR, 0.576; P < .01), extensive cGVHD (HR, 0.460; P = .02), and NRM (HR, 0.545; P = .03) with 1-MMUD. ATG did not impact survival with 1-MMRD and MUD. The use of ATG in conditioning is beneficial due to the reduction in acute/chronic GVHD without increasing NRM or disease relapse only in 1-MMUD transplantation. On the other hand, ATG is not recommended for CB transplantation.

Blastic plasmacytoid dendritic cell neoplasm (BPDCN) is a rare skin-tropic hematological malignancy of uncertain pathogenesis and poor prognosis. We examined 118 BPDCN cases for cytomorphology, MYC locus rearrangement, and MYC expression. Sixty-two (53%) and 41 (35%) cases showed the classic and immunoblastoid cytomorphology, respectively. Forty-one (38%) MYC+BPDCN (positive for rearrangement and expression) and 59 (54%) MYC-BPDCN (both negative) cases were identified. Immunoblastoid cytomorphology was significantly associated with MYC+BPDCN. All examined MYC+BPDCNs were negative for MYB/MYBL1 rearrangement (0/36). Clinically, MYC+BPDCN showed older onset, poorer outcome, and localized skin tumors more commonly than MYC-BPDCN. MYC was demonstrated by expression profiling as one of the clearest discriminators between CAL-1 (MYC+BPDCN) and PMDC05 (MYC-BPDCN) cell lines, and its shRNA knockdown suppressed CAL-1 viability. Inhibitors for bromodomain and extra-terminal protein (BETis), and aurora kinases (AKis) inhibited CAL-1 growth more effectively than PMDC05. We further showed that a BCL2 inhibitor was effective in both CAL-1 and PMDC05, indicating that this inhibitor can be used to treat MYC-BPDCN, to which BETis and AKis are probably less effective. Our data will provide a rationale for the development of new treatment strategies for patients with BPDCN, in accordance with precision medicine.

An 11q23 abnormality presents in approximately 5% of adults with acute myeloid leukemia (AML) and is associated with adverse outcomes even after allogeneic hematopoietic cell transplantation (allo-HCT). To evaluate the outcomes and prognostic factors following allo-HCT for adult AML with 11q23 abnormality, we retrospectively analyzed the Japanese registration data of 322 adult AML patients with 11q23 abnormality who had received allo-HCT between 1990 and 2014. In total, the disease status at HCT was first complete remission (CR1) in 159 (49%) patients. The probability of overall survival and the cumulative incidence of relapse at 3 years were 44 and 44%, respectively. In the multivariate analysis, disease status beyond CR1 at the time of HCT was significantly associated with a higher overall mortality and relapse. The 11q23 fusion partner did not have a significant impact on survival. We also evaluated the prognostic value of minimal residual disease (MRD) status at HCT on transplant outcomes among hematological CR patients. MRD status at HCT was the significant prognostic indicator for hematological relapse and survival. These data suggested that allo-HCT offered a curative option for adult AML with 11q23 abnormality. Pretransplant MRD status was the significant prognostic indicator for relapse and survival in CR patients.

In the opinion of the European LeukemiaNet (ELN), nucleophosmin member 1 gene mutation (NPM1 mut)-positive acute myeloid leukemia (AML) with an fms-like kinase 3-internal tandem duplication (FLT3-ITD) allele ratio (AR) <0.5 (low AR) has a favorable prognosis, and allogeneic hematopoietic stem cell transplant (allo-HSCT) in the first complete remission (CR1) period is not actively recommended. We studied 147 patients with FLT3-ITD gene mutation-positive AML, dividing them into those with low AR and those with AR of ≥0.5 (high AR), and examined the prognostic impact according to allo-HSCT in CR1. Although FLT3-ITD AR and NPM1 mut are used in the prognostic stratification, we found that NPM1 mut-positive AML with FLT3-ITD low AR was not associated with favorable outcome (overall survival [OS], 41.3%). Moreover, patients in this group who underwent allo-HSCT in CR1 had a significantly more favorable outcome than those who did not (relapse-free survival [RFS] P = .013; OS P = .003). Multivariate analysis identified allo-HSCT in CR1 as the sole favorable prognostic factor (RFS P < .001; OS P < .001). The present study found that prognosis was unfavorable in NPM1 mut-positive AML with FLT3-ITD low AR when allo-HSCT was not carried out in CR1.

High-dose melphalan followed by autologous hematopoietic stem cell transplantation (ASCT) is a standard treatment for younger myeloma patients. However, the correlation between its toxicity and renal impairment is not clear. We analyzed this relationship, focusing on estimated glomerular filtration rate (eGFR) as an index of renal function. We evaluated 78 multiple myeloma patients who underwent ASCT following high-dose melphalan at our center. Patients were divided into a higher eGFR group (eGFR ≥ 60) and a lower eGFR group (eGFR < 60). Multivariate analyses revealed that lower eGFR was independently associated with alkaline phosphatase elevation (OR 10.2, P = 0.038), mucositis (OR 10.5, P = 0.032), grade 2-4 co-elevation of both aspartate aminotransferase and alanine aminotransferase (OR 21.3, P = 0.016), delay of reticulocyte engraftment (HR 0.524, P = 0.034), and delay of platelet engraftment (HR 0.535, P = 0.0016). However, lower eGFR was not correlated with overall survival or time-to-next treatment. In summary, renal dysfunction secondary to administration of high-dose melphalan was associated with increased hepatic and mucosal toxicity and delay of hematological recovery, but did not affect survival outcomes.

The optimal treatment strategy for gastrointestinal graft-versus-host disease (GI-GVHD) after allogeneic hematopoietic cell transplantation remains to be established. We retrospectively analyzed 68 cases of GI-GVHD at our institution between 2007 and 2017. The survival outcomes were significantly inferior in patients who did not respond to the first-line treatment (1-year overall survival 27.3 vs 69.2%, P = 0.0017; non-relapse mortality 50.0 vs 18.6%, P = 0.026). After subsequent treatments, 18 patients were refractory to all steroid-based treatments such as steroid pulse therapy and oral beclomethasone dipropionate (BDP). However, these steroid-refractory cases showed a gradual increase in the response rate after the initial diagnosis of steroid refractoriness. This result may be explained by the problem of evaluating the response based solely on the volume of diarrhea, i.e., severe mucosal damage due to refractory GI-GVHD may require a long recovery and sometimes be complicated with other diseases. In conclusion, patients with GI-GVHD who failed to respond to the first-line treatment had inferior survival. However, later improvement may be observed without additional immunosuppressant other than steroid among patients who initially do not respond to steroid therapy. It is important to repeat colonoscopy in patients with refractory GI-GVHD to monitor the activity of GVHD.

Autologous stem cell transplantation (ASCT) is a treatment option for HIV-positive patients with non-Hodgkin lymphoma (NHL) and multiple myeloma (MM). However, the prognosis after ASCT in HIV-positive Japanese patients remains unclear. The aim of this study was to evaluate the impact of HIV infection on transplant outcomes after ASCT in Japan. Using the national database of the Japan Society for Hematopoietic Cell Transplantation, we retrospectively evaluated patients with NHL (n = 3862) and MM (n = 2670) who underwent their first ASCT between 2001 and 2014. The presence of HIV antibody was used to diagnose HIV infection. Fifty-six patients with NHL (1.4%) and 23 with MM (.8%) were positive for HIV antibody. Among patients with NHL overall survival (OS) was lower in HIV-positive patients than in HIV-negative patients (5-year OS: HIV-positive patients, 44% versus HIV-negative patients, 65%; P < .001). In a multivariate analysis HIV infection was significantly associated with an increased risk of overall mortality (hazard ratio, 2.30; P < .001). The incidence of relapse was higher in HIV-positive patients (P = .036), whereas there was a similar incidence of nonrelapse mortality (P = .879). OS in patients with MM was similar between those with/without HIV infection (5-year OS: HIV-positive patients, 61% versus HIV-negative patients, 63%; P = .988). HIV infection was associated with a higher risk of overall mortality and relapse after ASCT for NHL in a Japanese population.