ALBERT

Description: There are no validated criteria to measure skin response in chronic GVHD. In a prospectively assembled, multicenter cohort of patients with chronic GVHD (N = 458), we looked for correlation of change in several different scales recommended by the National Institutes of Health (NIH) Consensus with clinician and patient perception of change and overall survival. Of the clinician scales, the NIH composite 0-3 skin score was the only one that correlated with both clinician and patient perception of improvement or worsening. Of the patient-reported scales, the skin subscale of the Lee Symptom Scale was the only one that correlated with both clinician and patient perception of improvement or worsening. At study entry, NIH skin score 3 and Lee skin symptom score 〉 15 were both associated with worse overall survival. Worsening of NIH skin score at 6 months was associated with worse overall survival. Improvement in the Lee skin symptom score at 6 months was associated with improved overall survival. Our findings support the use of the NIH composite 0-3 skin score and the Lee skin symptom score as simple and sensitive measures to evaluate skin involvement in clinical trials as well as in the clinical monitoring of patients with cutaneous chronic GVHD.

Description: There exists controversy regarding the benefit of HSCT in infant ALL. At our institution, the standard treatment for high-risk infant ALL has been intensive chemotherapy per POG 9407 followed by HSCT in CR1. We retrospectively reviewed the outcome of the 15 patients (9 males, 6 females) with infant ALL that underwent allogeneic HSCT since 1992. We defined “very high-risk” as an infant with one or more of the following criteria: age 6 months or less at diagnosis, MLL+ or t(4:11), or WBC〉100,000 at diagnosis. Fourteen of fifteen patients are considered “very high-risk.” All except one were under 12 months at diagnosis (range 2–14 months, median 7). The fourteen month old was included as she displayed infant ALL biology, with MLL+ and WBC〉500,000. Seven were 6 months or under at diagnosis. Cytogenetics at diagnosis showed MLL+ or t(4:11) in 9, normal in 4, and indeterminate in 2 patients. WBC count at diagnosis was 4,300–1,000,000 (median 360,000). Twelve of the patients were CALLA-negative and four had CNS involvement at diagnosis. Time from diagnosis to transplant was 2.3–13.6 months (median 4.2). All patients were in morphologic CR1 at time of HSCT but one patient had persistent t(4:11). Cytoreduction: Patients received TBI 150cGy x 8 (d-10 to-7); VP-16 1g/m2 CI (d-5 to-4); cyclophosphamide 60 mg/kg/day x 3(d-4 to-2). GVHD prophylaxis: CSA, short-course MTX, and ATG d +1, +3, +5, +7 (for unrelated cord blood (UCB)). Grafts were not T-cell depleted. Stem cell sources: 8 UCB, 6 HLA-identical sibling bone marrow, 1 HLA-identical sibling PBSC. Days to neutrophil engraftment was 11–25 days (median 16) and to platelet engraftment was 17–82 days (median 37). CNS prophylaxis consisted of intrathecal methotrexate and Ara-C monthly x 6 post-HSCT. Acute GVHD: 2 grade II and 1 grade III. Chronic GVHD: 3 limited, 1 extensive. Mortality: 4 patients have died. 2 (13%) of transplant-related mortality (TRM) within 100 days of HSCT and 2 of relapse (15% of evaluable patients), which included the patient with t(4:11) at time of HSCT. Range of follow-up is 0.1–11.8 years with a 73% event-free survival (EFS) and overall-survival at a median follow-up of 3.5 years. We conclude that patients with infant ALL who attain a CR1 (morphologic and cytogenetic) have good EFS with acceptable TRM following HSCT using either UCB or HSC from a matched-sibling. Long-term follow-up, including developmental assessments, are being pursued to document the impact of this intensive therapy in young children. Figure Figure

Description: Pentostatin, one of the purine nucleoside analogues, is known to decrease lymphocyte number and function. We are currently investigating pentostatin for refractory chronic graft-vs.-host disease (cGVHD). Within an adult/pediatric phase II study, 16 patients under age 20 are evaluable. All children presented here failed at least two immunosuppressive regimens (including steroids at a dose equivalent to at least 1 mg/kg/day prednisone for one month) before enrollment. The treatment protocol consists of adding pentostatin 4 mg/m2/dose IV every 2 weeks for 6 months. Patients with improving disease are permitted to continue pentostatin therapy at a 3 to 4 week interval. To reduce the risk of infection, steroids are tapered early in all patients. At the end of 3 months, patients with stable or improving cGVHD are weaned off their other medications but maintained on calcineurin inhibitor. Pentostatin is stopped at 6 months if complete response (CR) is achieved or continued if partial response (PR). Patients are followed for improvement in the skin, mouth, and liver. The severity of GVHD is scored for each system on a scale from 0 to 4. Major response is defined as an improvement in symptom score of 2 points or more without worsening in any system. Minor response is a 1 point improvement. Mixed response is improvement in 1 system but worsening in another. Patients have received a median of 15 doses (range 5–30). Median age of the pediatric cohort is 14.5 years (range 5 to 19). Diagnoses include AML/MDS (6), ALL (5), hemoglobinopathy (2), T-cell leukemia (1), aplastic anemia (1), and lymphohistiocytosis (1). Ten had HLA-identical sibling donors (7 BM/3 PBSC) and 6 had HLA-matched unrelated donors (4 BM/2 PBSC). Median time since transplant to start of Pentostatin is 20.5 months (range 8–72). Most patients were on calcineurin inhibitor, eleven on prednisone, seven on MMF, one on rapamycin, and one was receiving ECP at entry on the study. Five patients attained a CR, eight a major response, and three patients have progressed. The overall response rate is 81%. Since some patients had multiple organ involvement, response rate by organ (PR or CR for individual organ) is below. Therapy has been well-tolerated. There has been only one toxicity likely due to pentostatin (creatinine elevation to 4.8 in a patient with a single kidney), and this required the patient to withdraw from the study. There have been no severe infections. There have been 2 deaths, both cGVHD-related, in patients who did not respond. Of the 11 patients on prednisone, 8 were weaned completely off steroids or to

Description: RIC for HSCT has been used to treat a variety of diseases in adults. Acute and chronic graft vs host disease (GVHD) continue to cause morbidity and mortality. ECP is an immunomodulatory cell therapy being investigated as treatment for various immune-mediated disorders, including GVHD. ECP is an established therapy for chronic GVHD, but has not been routinely used as prophylaxis. We hypothesized that combining ECP with RIC prior to HSCT would reduce the incidence of GVHD and 100-day transplant related mortality (TRM) without affecting time to engraftment. Eight children have been enrolled; ranging in age from 10 – 16 years (mean 13.6), with CML (n=2), severe aplastic anemia (SAA) (n=3), ALL in CR2 (n=1), or relapsed Hodgkin’s disease after autologous transplant (n=2). Stem cell sources included peripheral blood from matched related (n=6), mis-matched related (n=1), and matched unrelated (n=1) donors. Conditioning consisted of 2 weekly treatments of ECP (treatment=2 consecutive days) beginning two weeks prior to transplant, followed by fludarabine (days -6 to -2) and targeted dose busulfan to reach a daily AUC of 4000 μmol*min (days -5, -4); GVHD prophylaxis was: mycophenolate mofetil (MMF), cyclosporine, and weekly treatments of ECP starting once leukocyte chimerism reached 50% through day +100. Full engraftment (〉95% total donor chimerism) was achieved by day 12 – 81 (median 29 days) in all patients. Two patients required immune modulation for falling chimerism; one responded to removal of immune supression, the other required donor lymphocyte infusion (DLI). Both subsequently developed grade 3 aGVHD. Patients with CML and SAA are fully engrafted and in remission. Four had central line infections, one had adenovirus antigenemia, and all 5 at risk for CMV had reactivation. They underwent pre-emptive treatment without development of disease. ECP was well tolerated. There were no toxic deaths, episodes of veno-occlusive disease or mucositis; none required parenteral nutrition. Patients were admitted to the hospital for 2 days during the preparative regimen (IV busulfan); between day 0 and 100 patients were admitted for a median of 11 days (range 0 – 25). We conclude that in children using ECP as part of RIC and GVHD prevention is feasible, time to engraftment is acceptable, and complications are minimal as evidenced by short inpatient stays. However, immune supression appears to be excessive, given the rates of CMV reactivation and the declining chimerism of two patients while on ECP. ECP appears to be effective in preventing aGVHD, as the only patients who developed aGVHD were those requiring immune modulation. Patient Characteristics Patient # Diagnosis Donor Chimerism 〉 50% (days) Chimerism 〉 95% (days) aGVHD cGVHD Disease Status *Interventions due to falling chimerism 1 CML Matched Sib 13 29 None None NED following DLI & Imatinib for rising BCR/ABL, Day +384 2 SAA Matched Sib 8 26 None Extensive NED, Day +328 3 CML Matched Sib 26 63 (following DLI)* Grade 3 Extensive NED, Day +278 4 SAA Mismatched Related 12 81 (following removal of immune supression)* Grade 3 Limited NED, Day +222 5 ALL (CR2) Matched Sib 12 28 None None Relapse, Day +100 6 HD (CR3) MUD 20 33 None None Relapse, Day +100 7 SAA Matched Sib 7 12 None None NED, Day +125 8 HD (CR2) Matched Sib 28 47 None N/A NED, Day +62

Description: Background Allogeneic Hematopoietic Stem Cell Transplantation (HSCT) constitutes a curative treatment for children with malignant and non-malignant disorders. HLA-partially matched haploidentical (haplo) donors represent a viable alternative option for those children who lack an HLA-compatible donor. T-cell depletion approaches with positive (CD34 selection) or negative selection (alpha/beta T-cell and CD19+ B-cell-depletion) may allow engraftment of donor cells with a low risk of GvHD; however, success is limited by delayed immune recovery, increasing the risk of fatal infections. Infusion of unmanipulated donor T cells (DLI) to accelerate immune recovery is associated with high risk of fatal GVHD. In contrast, adoptive transfer of donor T cells genetically manipulated to include a safety switch can be a suitable strategy to render DLI safer and more widely applicable. BPX-501 is an allogeneic product consisting of T cells modified to express the inducible caspase-9 (iC9) safety switch. The polyclonal natural of these modified T cells can provide viral immunity following stem cell transplant, with the unique ability to promptly and durably resolve GvHD symptoms following the administration of rimiducid, an inert, lipid-permeable compound that rapidly induces dimerization and activation of iC9, inducing apoptosis of the gene modified T cells. Aims To evaluate the safety and efficacy of rimiducid in the treatment of GvHD following administration of BPX-501 T cells in pediatric patients with malignant or non-malignant disorders given an αβ T-cell receptor and B-cell depleted haplo-HSCT. A key objective of this study is to assess the activity of rimiducid infusion following onset of GvHD which is refractory to standard of care therapies. Methods Two multicenter (US [NCT03301168] and EU [NCT02065869]), prospective trials utilized αβ-T-cell and B-cell-depleted haplo-HSCT followed by infusion of a titrated number of donor lymphocytes genetically modified with iC9 (BPX-501 T cells) in patients with malignant or non-malignant disorders. BPX-501 T-cells were planned to be infused on day14+/-4 after the allograft. No post-transplant pharmacological GvHD prophylaxis was employed. Patients who develop GvHD resistant to conventional steroid therapy could receive ≥1 dose of dimerizing rimiducid to activate the iC9 safety gene. The efficacy evaluable population is defined as any patient who received ≥ 1 dose of rimiducid for the treatment of GvHD and had a follow up response assessment. Results At the time of clinical cut-off (June 30th, 2018) 249 patients with a malignant (41.4%) or non-malignant disorder (58.6%) were enrolled. The conditioning regimens varied according to the original disease and were Treosulfan-based (14.9%), Busulfan-based (28.9%), TBI-based (34.9%) or other (19.3%). The donor was a parent in 229 children (92 %), a sibling in 17 (6.8 %), and a half-sibling in the remaining 3 (1.2 %). The median time to BPX-501 infusion was 18 days (10 - 66 days). Fifty-two patients developed Grade I-IV aGvHD (cumulative incidence [CI] 21.9 % [95% confidence interval (CoI): 16.7 - 27.2]). Twenty-six patients developed Grade II-IV aGvHD (CI 10.9 %). Five patients developed Grade III-IV aGvHD (CI 2.1 %). Eight patients developed cGvHD (CI of 4.6% [95% CoI: 1.3 - 7.8]). Twenty-one patients met the rimiducid efficacy evaluable population definition. An overall clinical response rate of 86% was observed. A CR or PR to rimiducid was observed in 12 and 6 patients, respectively. Median time to initial response was 2 days (1-61 days). Median number of doses received was 1 (1 - 2). At a median follow-up of 7.8 months (2.3 - 30.8 months), 77% of the initial responders were still in either complete (n=8) or partial response (n=6). Conclusion These data suggest that administration of rimiducid for treatment of steroid-refractory GvHD represents a novel and highly effective treatment approach in pediatric patients with non-malignant or malignant disorders who received a αβ-T-cell and B-cell depleted haplo-HSCT followed by infusion of BPX-501 cells. The administration of rimiducid in children given BPX-501 T cells allows for effective control of GvHD occurring after the adoptive transfer of genetically modified T cells. Disclosures Slatter: Medac: Other: Travel assistance. Merli:Neovii Biotech: Honoraria; AMGEN: Honoraria. Aldinger:Bellicum Pharmaceuticals, Inc.: Employment. Locatelli:Bellicum: Consultancy, Membership on an entity's Board of Directors or advisory committees; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees; Miltenyi: Honoraria; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; bluebird bio: Consultancy.

Description: Despite improvements, mortality after allogeneic hematopoietic cell transplantation (HCT) for nonmalignant diseases remains a significant problem. We evaluated whether pre-HCT conditions defined by the HCT Comorbidity Index (HCT-CI) predict probability of posttransplant survival. Using the Center for International Blood and Marrow Transplant Research database, we identified 4083 patients with nonmalignant diseases transplanted between 2007 and 2014. Primary outcome was overall survival (OS) using the Kaplan-Meier method. Hazard ratios (HRs) were estimated by multivariable Cox regression models. Increasing HCT-CI scores translated to decreased 2-year OS of 82.7%, 80.3%, 74%, and 55.8% for patients with HCT-CI scores of 0, 1 to 2, 3 to 4, and ≥5, respectively, regardless of conditioning intensity. HCT-CI scores of 1 to 2 did not differ relative to scores of 0 (HR, 1.12 [95% CI, 0.93-1.34]), but HCT-CI of 3 to 4 and ≥5 posed significantly greater risks of mortality (HR, 1.33 [95% CI, 1.09-1.63]; and HR, 2.31 [95% CI, 1.79-2.96], respectively). The effect of HCT-CI differed by disease indication. Patients with acquired aplastic anemia, primary immune deficiencies, and congenital bone marrow failure syndromes with scores ≥3 had increased risk of death after HCT. However, higher HCT-CI scores among hemoglobinopathy patients did not increase mortality risk. In conclusion, this is the largest study to date reporting on patients with nonmalignant diseases demonstrating HCT-CI scores ≥3 that had inferior survival after HCT, except for patients with hemoglobinopathies. Our findings suggest that using the HCT-CI score, in addition to disease-specific factors, could be useful when developing treatment plans for nonmalignant diseases.

Description: Several risk factors are associated with increased mortality in patients with chronic graft-versus-host disease (cGVHD), but there is considerable variability in the reported factors. Therefore, we evaluated patient, transplantation, and cGVHD characteristics to develop a risk score in 5343 patients with cGVHD. Ten variables were identified as being significant in multivariate analysis of overall survival and nonrelapse mortality (NRM): age, prior acute GVHD, time from transplantation to cGVHD, donor type, disease status at transplantation, GVHD prophylaxis, gender mismatch, serum bilirubin, Karnofsky score, and platelet count. These 10 variables were used to build a cGVHD risk score, and 6 risk groups (RGs) were identified. The 5-year NRM was 5% (1%-9%) in RG1, 20% (19%-23%) in RG2, 33% (29%-37%) in RG3, 43% (40%-46%) in RG4, 63% (53%-74%) in RG5, and 72% (59%-85%) in RG6. The 5-year overall survival was highest at 91% (95% confidence interval [CI]:85%-97%) in RG1, followed by 67% (65%-69%) in RG2, 51% (46%-55%) in RG3, 40% (37%-43%) in RG4, 21% (12%-30%) in RG5, and 4% (0%-9%) in RG6 (all P 〈 .01). This analysis demonstrates the usefulness of data from a large registry to develop risk-score categories for major transplantation outcomes. Validation of this cGVHD risk score is needed in a different population to ensure its broad applicability.

Description: Background Allogeneic hematopoietic stem cell transplantation (HSCT) is a well-established treatment for children with Acute Leukemia (AL). For patients lacking a compatible matched related or unrelated donor, HLA-haploidentical HSCT (haplo-HSCT) from a relative represents a viable alternative. Promising results were reported with a novel method of selective depletion of αβ T and B cells (Locatelli, Blood 2017). This approach is associated with limitations such as suboptimal adaptive immune reconstitution, increased risk of infection and disease relapse. BPX-501 is an allogeneic product consisting of T cells modified to express the inducible caspase-9 (iC9) safety switch and truncated CD19 to allow monitoring and expansion of BPX-501 following transplant. The polyclonal nature of the BPX-501 provides broad virus and tumor-specific immunity, while the safety switch provides the unique ability to promptly and durably resolve graft-versus-host disease (GvHD) symptoms following the administration of rimiducid which induces dimerization and activation of iC9, inducing apoptosis of BPX-501. Aims To evaluate the safety and efficacy of BPX-501 administered after a αβ T and B-cell depleted haplo-HSCT in pediatric patients with AL in morphological complete remission (CR). The objective was to determine whether BPX-501 infusion can increase relapse-free survival (RFS) and overall survival (OS) through an enhanced graft-versus-leukemic (GvL) effect, while maintaining a low risk of GvHD. Methods This multicenter US (NCT03301168) and EU (NCT02065869), prospective trial utilizes αβ-T and B-cell-depleted haplo-HSCT followed by infusion of donor lymphocytes genetically modified with iC9 safety switch (BPX-501) in patients with malignant or non-malignant disorders. A subset of patients had acute high-risk leukemias (AML and ALL). BPX-501 was planned to be infused on day14±4 after the allograft. No post-transplant pharmacological GvHD prophylaxis was employed. Patients who develop GvHD resistant to conventional steroid therapy could receive ≥1 dose of rimiducid to activate iC9. The efficacy-evaluable population (EEP) was defined as any patient with AL who received HSCT, BPX-501 infusion, and at least one follow-up assessment. Results At clinical cut-off (June 30, 2018), 100 patients (EU: 75, US: 25) with AL met the EEP definition. Median follow-up was 14.7 mos (1 - 40.6 mos). Key baseline characteristics are shown in Table 1. The median time for neutrophil and platelet engraftment was 16 (15 - 17) and 12 (11 - 12) days, respectively. Four patients (4.1% [95% CI: 0 - 8%]) experienced primary graft failure. Of 96 evaluable patients, 21 patients developed Grade I-IV aGvHD (21.7% [95% CI: 13.5 - 29.8%]). Five patients developed Grade III-IV aGvHD (3.1% [95% CI: 0 - 6.5%]). Of 82 evaluable patients, 12 patients developed cGvHD (18.1% [95% CI: 8.2 - 22%]), with only three cases being moderate-severe. Rimiducid was administered to 10 patients with steroid-resistant acute GvHD. Best overall clinical response of CR or PR post-rimiducid administration was seen in 8 patients (80%). Among responding patients, 7 patients (87.5%) had a CR. Six patients died after transplantation (6.6% [95% CI: 1.4 - 11.7%]). Relapse Free Survival (RFS) was 82.2% (95% CI: 74.5 - 89.7%). Overall Survival (OS) was 90.1% (95% CI: 83.9 - 96.3%). Efficacy outcomes (TRM, RFS and OS) in AL subsets (AML and ALL) are shown in Table 2. CD3+ and CD3+CD4+ T cells above 500 cells/ml were achieved by 180 and 270 days, respectively. IgA and IgM levels achieved normal values by 180 days. The percentage of circulating and median absolute BPX-501 cells at Day 100 were 9.96% ± 11.6% (0 - 54.9%) and 85.58 ± 165.57 cells/ul (0 - 1001 cells/ml), respectively. Conclusion The adoptive transfer of BPX-501 following αβ-T and B-cell depleted haplo-HSCT represents a novel and highly effective transplantation strategy for pediatric patients with AL. Compared to data from children receiving only αβ T and B-cell depleted haplo-HSCT or matched unrelated donor HSCT (https://bloodcell.transplant.hrsa.gov/research/transplant_data/us_tx_data/survival_data/survival.aspx), this novel approach resulted in a comparable risk of transplant-related mortality and a lower risk of recurrence. Rimiducid was also an effective treatment for patients who developed steroid-resistant GvHD. Disclosures Locatelli: bluebird bio: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Miltenyi: Honoraria; Bellicum: Consultancy, Membership on an entity's Board of Directors or advisory committees; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees. Qasim:Orchard: Equity Ownership; Autolus: Equity Ownership; Servier: Research Funding; Bellicum: Research Funding. Nemecek:Novartis Pharmaceuticals Corporation: Other: advisory boards.

Description: RIC regimens are increasingly used for the treatment of adults with malignancies, relying predominantly on graft-vs-tumor cellular therapy. RIC reduces the risk for regimen related morbidity (RRM) and mortality in adults. Pediatric experience with RIC protocols has been limited despite the potential benefit for reduced long-term morbidity. We report the experience of 35 RIC HSCT for malignancies in pediatric patients (pts), 31 were at very high-risk for complications of full-intensity conditioning due to disease state, prior HSCT (7 allo, 8 auto) and/or co-morbid medical conditions. The RIC regimen was comprised of fludarabine, 30 mg/m2 for 6 consecutive days (days -10 through -5), followed by intravenous busulfan (days -5 and -4), 0.8 - 1 mg/kg for 8 doses (n=15) or targeted AUC 4000 microMol*min for 2 doses (n=20) and ATG days -4 through -1, (equine 40 mg/kg/d or rabbit, 2 mg/kg/d). RIC was followed by allogeneic HSCT. Diagnoses included: recurrent ALL (15, ≥CR3 in 8), AML (5, refractory in 4), refractory neuroblastoma (5), recurrent Non-Hodgkin’s Lymphoma (4), CML (3) or treatment-related myelodysplastic syndrome (3). There were 21 males, 14 females, ages 2 to 17 yrs, median 9.14. Stem cell sources included 22 unrelated donors (URD), 11 matched sibs and 2 mis-matched relatives. 31 of 35 stem cell donations were G-CSF mobilized peripheral blood. The median cell doses infused were 5.05 × 108 MNC/kg and 4.6 × 106 CD34+ cells/kg. Graft-versus-host disease (GVHD) prophylaxis was cyclosporin A (CsA) alone in the first 12 pts, CsA and mycophenolate mofetil (MMF) in the subsequent 23 pts. Significant (〉80%) donor chimerism was achieved in 32 pts with a median time to reach an ANC 〉500/μl of 18 days (range 9–62). Three pts with URD failed to engraft and another 3 had graft loss. An unsupported platelet count 〉 20,000/μl was achieved in 29 of 35 pts at a median of 15 days (7–228). Six pts developed Gr III–IV acute GVHD, 7 of the 25 pts surviving more than 100 days developed chronic GVHD (4 limited, 3 extensive). Since addition of MMF, only one pt developed chronic GVHD (limited). While there was no immediate RRM, 10 pts (8 URD, 4 prior HSCT) died of infection (4), infection + GVHD (3), GVHD (2) or veno-occlusive disease after a subsequent HSCT (day + 146 post RIC HSCT). Infections included 4 probable fungal, 2 viral and one bacterial infection. Sixteen pts have relapsed from 25 to 427 days post HSCT for an event-free-survival of 22%. Because RIC HSCT creates a platform for cellular immuno-therapy, early post-HSCT relapse or persistent disease was treated by withdrawal of GVHD prophylaxis (n= 11) and/or donor leukocyte infusions (n=3). 4 pts responded to immune modulation and 3 remain in remission. Thus, the projected three-year overall survival in this very high-risk group of pts is 44%. Our results indicate RIC allogeneic HSCT has a role for pediatric pts at high-risk for complications after conventional HSCT.

Description: Abstract 898 Chronic graft-versus-host disease (cGvHD) remains a major barrier to allogeneic (allo) hematopoietic cell transplant (HCT). Previous studies have identified several variables associated with high mortality in persons with cGvHD. The most consistent risk-factors for mortality are thrombocytopenia and the progressive type of cGvHD onset. We evaluated subject-, disease-, and transplant related variables at diagnosis of cGvHD to develop a risk score in 5343 patients with cGvHD. All patients received first allo-transplant for acute myeloid leukemia (AML), acute lymphoid leukemia (ALL), chronic myeloid leukemia (CML) and myelodysplastic syndrome (MDS) between 1995 and 2004 and reported to CIBMTR. 10 variables were significantly correlated with cGvHD in multivariate analysis of overall survival and non-relapse mortality (NRM) and were used to build the risk score (Table 1). Variable-specific risk scores (VSRC) were constructed for each factor based on the relative risk (RR) of overall or NRM associated with each factor category. Scores were summed for each subject to assign an overall risk score (ORS). Risk groups (RG) of 1–6 were assigned based on the overall risk score. RG1: ORS 0–2, RG2: ORS 3–6, RG3: ORS 7–8, RG4: ORS 9–10, RG5: ORS: 11, RG6: ORS ≥ 12. Survival and NRM were significantly different between RGs (Figure 1 and Figure 2). Although this cGvHD risk score requires independent validation, knowledge of the above HCT factors and cGvHD-specific factors at the time of initial diagnosis provides important prognostic information for clinical care and clinical trial planning. Future analysis will validate this risk score in a more recent allo-transplant cohort assembled after 2005. Table 1: Variable Overall Survival NRM VSRC RR P RR P Recipient age (baseline: