European Group For Blood And Marrow Transplantation Risk Score Research Articles

First Application Of The EBMT Risk Score In Double Umbilical Cord Blood Transplantation For Hematological Malignancies: Significant Impact On Different Outcomes

The European Group for Blood and Marrow Transplantation (EBMT) risk score provides a simple tool to assess chances and risks of hematopoietic stem cell transplantation (HSCT) for an individual patient from related and unrelated donors. To our knowledge, this score has never been applied in double umbilical cord blood transplantation (dUCBT). We recently published results of 136 patients who underwent dUCBT reported to the SFGM-TC registry from 23 centers between 2005 and 2007 (Labussiere et al., Exp. Hem. 2013). After having recorded the full data on CB units including HLA and sex, we decided to test the validity of the EBMT risk score and check its application in dUCBT. There were 88 males and 48 females with a median age of 41 years (range: 18 – 66), 76 patients with acute leukemias (42 AML, 27 ALL, 5 secondary AL and 2 Biphenotypic AL), 8 with chronic leukemias (5 lymphoid and 3 myeloid), 10 with myelodysplasic syndormes, 24 with Hodgkin and non Hodgkin lymphomas, 13 with multiple myelomas and 5 with other myeloproliferative diseases. Median interval between diagnosis and transplantation was 20.5 months (range: 3-385). At time of allogeneic transplantation, 41 patients were in first complete remission (CR1), 40 in ≥CR2 and 55 less than CR. Forty-six patients received myeloablative and 90 reduced intensity conditioning regimens. The EBMT risk score uses the 5 pre-transplant factors: age of the patient, disease status, time from diagnosis to transplant, kind of donor (related/unrelated), and donor-recipient sex combination with 0 to 1 or 2 points (pts) for each factor. Age was categorized as <20 years (0 pt; n= 6), 20 to 40 years (1 pt; n= 59) and >40 years (2 pts; n=71). For disease status we distinguished patients in CR1 at transplantation, (0 pt; n=41) from patients in CR >1 (1 pt; n=40) and patients not in CR (2 pts; n=55). Time from diagnosis to transplant was categorized into ≤12 months (0 pt; n=26) and >12 months (1pt; n=110). The adaptation of the new risk score concerned HLA and sex matching. Complete information concerning HLA typing was obtained from the database of the national network of cord blood banks on behalf of Eurocord. HLA typing was evaluated across 3 loci: HLA-A, -B and -DRB1; it was based on antigenic level for HLA-A and HLA-B, and on allelic level for HLA-DRB1. HLA matching showed patients with at least 4/6 HLA matching on HLA-A, -B and DRB1 without any mismatch on DRB1 alleles between the recipient and each UCB unit and the 2 units together (0 pt; n=77); patients with at least 4/6 HLA matching on HLA-A and -B between the recipient and each unit but without complete matching on DRB1 or with less than 4/6 HLA matching between the 2 units (1 pt; n=51), and finally patients with more than 2 mismatches on the three loci (2 pts; n=8). Donor-recipient sex combination separated all others (0 pt; n=73) from the male recipient with one female and one male cord blood unit (1pt; n=45) and male recipient with two female cord blood units (2 pts; n=18). Hence, we found 5 patients (4%) in score 1, 12 (9%) in score 2, 23 (17%) in score 3, 26 (19%) in score 4, 30 (22%) in score 5, 26 (19%) in score 6, 11 (8%) in score 7 and 3 (2%) in score 8. We pooled patients with scores 1-2; 3-4 and 5-8 as their outcomes were found to be similar. After a median follow-up of 49.5 months, the 3 years probabilities of overall and progression-free survivals for the whole population were 41% and 35% respectively. with a two-year cumulative relapse incidence of 28% (24.0 – 31.8). Interestingly patients with score 1-2 showed better overall survival rates than those with score 3-4 and 5-8 with a 3 years probability of 77%, 42% and 32% respectively (p=0.002); this was associated with a 3 years relapse incidence of 19%, 35% and 55% respectively (p=0.021), while the 3 years transplant related mortality (TRM) was 6%, 39% and 36% respectively (score 1-2 versus all others, p=0.02), Figure 1. [Display omitted] In conclusion, we found that the EBMT risk score can perfectly be applied to double UBCT with a significant impact on different outcomes; its use enables a better selection of patients who will benefit the more from this treatment strategy. Disclosures:No relevant conflicts of interest to declare.

European Group for Blood and Marrow Transplantation Risk Score Predicts the Outcome of Patients with Acute Leukemia Receiving Single Umbilical Cord Blood Transplantation

The European Group for Blood and Marrow Transplantation (EBMT) risk score has been implemented as an important tool to predict patient outcomes after allogeneic hematopoietic stem cell transplantation. However, to our knowledge, this score has never been applied in cases of single umbilical cord blood transplantation (sUCBT). We retrospectively analyzed 207 consecutive patients with acute leukemia who received sUCBT at our center between February 2011 and December 2015. The probabilities of 3-year overall survival (OS) and leukemia-free survival (LFS) of the entire cohort were 65.0% and 59.8%, respectively, whereas the cumulative incidences of 3-year nonrelapse mortality (NRM) and relapse rate were 19.5% and 20.3%, respectively. In the univariate analysis, a higher EBMT risk score was associated with worse OS and LFS and higher NRM and relapse rate, ranging from 81.7%, 75.9%, 7.3%, and 15.3%, respectively, for patients with a score of 1 to 43.8%, 44.3%, 31.7%, and 23.9%, respectively, for patients with scores of 4 to 6. Hazard ratios of OS, LFS, and NRM all steadily increased for each additional score point. Importantly, the prognostic value of the EBMT risk score on OS, LFS, NRM, and relapse was maintained in the multivariate analysis. Moreover, considering the univariate analysis results of donor–recipient gender and mismatched allele-level HLA-A, -B, -C, and -DRB1 loci on patient outcomes and the fairly strong interaction between time from diagnosis to sUCBT and disease status, we developed a modified sUCBT-EBMT risk score by using degrees of 8-allele HLA match instead of donor type, donor–recipient gender combination, and time from diagnosis to sUCBT, and found that the modified score could also be used as a predictor for patient outcomes after sUCBT. The EBMT risk score is a good predictor of outcomes of patients with leukemia after sUCBT. The modified sUCBT-EBMT risk score can also be used as a pretransplant risk assessment, but this metric still requires further evaluation with a larger cohort.

The Embt Risk Score Predicts the Outcome of Patients with Leukaemia Receiving Single Umbilical Cord Blood Transplantation

Aim The European Group for Blood and Marrow Transplantation (EBMT) risk score has been implemented as an important tool to predict patient outcomes following allogeneic haematopoietic stem cell transplantation (allo-HSCT). However, to our knowledge, this score has never been applied in cases of single umbilical cord blood transplantation (sUCBT). The aim of this study was to assess the capacity of the EBMT risk score in predicting the outcomes of patients with leukaemia receiving sUCBT. Methods We retrospectively analysed 218 consecutive patients with leukaemia who received sUCBT at our centre between February 2011 and December 2015. Sixty-eight of the 218 patients had AML, 137 had ALL, 10 had CML, 2 had mixed phenotype AL, and only one patient had plasma cell leukaemia. The median age was 12 years (range, 1-46 years), and the median weight was 40 kg (range, 10-100 kg). All of the patients received an intensified myeloablative conditioning regimen with cyclosporine A (CsA) and mycophenolate mofetil (MMF) as graft-versus-host disease (GVHD) prophylaxis. All of the patients were followed until 31 March 2016, with a median follow-up of 10.7 months (range, 0.5-58.9 months). Results The overall survival (OS) and leukaemia-free survival (LFS) of the entire cohort were 66.8% and 55.4%, respectively, whereas the cumulative incidences of relapse rate and non-relapse mortality (NRM) were 22.4% and 18.0%, respectively. In the univariate analysis, a higher EBMT risk score was associated with worse OS, worse LFS, lower NRM and higher relapse rate, ranging from 80.2%, 74.5%,7.5% and 18.0%, respectively, for patients with a score of 1 to 21.8%, 17.0%, 27.3% and 55.7%, respectively, for patients with a score of 5/6. Hazard ratios of the four indexes all steadily increased for each additional score point. Importantly, the prognostic value of the EBMT risk score on OS, LFS, NRM and relapse was maintained in the multivariate analysis. Moreover, considering the importance of allele-level HLA matching, we developed a modified sUCBT-EBMT risk score using HLA matching situations instead of donor type and found that the modified score could also be used as a predictor for patient outcomes following sUCBT. Conclusions The EBMT risk score is a good predictor of outcomes of patients with leukaemia following sUCBT. The modified sUCBT-EBMT risk score can also be used as a pretransplant risk assessment, but this metric still requires further evaluation with a larger cohort. DisclosuresNo relevant conflicts of interest to declare.

Prognostic Factors on the Graft-versus-Host Disease-Free and Relapse-Free Survival after Adult Allogeneic Hematopoietic Stem Cell Transplantation.

The cure of hematologic disorders by allogeneic hematopoietic stem cell transplantation (HSCT) is often associated with major complications resulting in poor outcome, including graft-versus-host disease (GVHD), relapse, and death. A novel composite endpoint of GVHD-free/relapse-free survival (GRFS) in which events include grades 3-4 acute GVHD, chronic GVHD requiring systemic therapy, relapse, or death is censored to completely characterize the survival without mortality or ongoing morbidity. In this regard, studies attempting to identify the prognostic factors of GRFS are quite scarce. Thus, we reviewed 377 adult patients undergoing allogeneic HSCT between 2003 and 2013. The 1- and 2-year GRFS were 40.8% and 36.5%, respectively, significantly worse than overall survival and disease-free survival (log-rank p < 0.001). European Group for Blood and Marrow Transplantation (EBMT) risk score > 2 (p < 0.001) and hematologic malignancy (p = 0.033) were poor prognostic factors for 1-year GRFS. For 2-year GRFS, EBMT risk score > 2 (p < 0.001), being male (p = 0.028), and hematologic malignancy (p = 0.010) were significant for poor outcome. The events between 1-year GRFS and 2-year GRFS predominantly increased in relapsed patients. With prognostic factors of GRFS, we could evaluate the probability of real recovery following HSCT without ongoing morbidity.

Incidence and risk factors of probable and proven invasive fungal infection in adult patients receiving allogeneic hematopoietic stem cell transplantation.

To investigate the incidence and risk factors for the occurrence of proven or probable invasive fungal infection (IFI) in adult patients receiving allogeneic hematopoietic stem cell transplantation (HSCT). We retrospectively analyzed 421 patients undergoing HSCT between 2002 and 2013 in our hospital. The risk factors for the occurrence of IFI were analyzed using Cox regression models. Thirty-one patients with the median age of 42 years (range, 19-60 years) developed IFI after HSCT. The post-HSCT IFI incidence was 7.4% and median time from HSCT to the diagnosis of IFI was 139 days (range, 2-1809days). Of the pretransplant factors, European Group for Blood and Marrow Transplantation (EBMT) risk score > 2 (p=0.001) and prior history of IFI (p=0.006) or type 2 diabetes mellitus (DM; p=0.042) were the significant predictors for post-HSCT IFI in univariate analyses. In multivariate analysis, EBMT risk score > 2 (p=0.015) and prior history of IFI (p=0.006) retained significance. Of the post-transplant factors, acute graft-versus-disease (GVHD) overall Grade III-IV (p<0.001), extensive chronic GVHD (p=0.002), development of post-transplant lymphoproliferative disorders (p=0.005), and the use of high-dose steroids (p<0.001) were statistically significant in univariate analyses. After multivariate analysis, high-dose steroids (p<0.001) and acute GVHD overall Grade III-IV (p=0.045) retained significance. These results suggest that risk group stratification prior to HSCT and monitoring of IFI in patients with severe GVHD receiving high-dose steroids is mandatory to reduce the mortality and morbidity of post-HSCT IFI, especially in those with prior history of IFI.

EBMT risk score can predict the outcome of leukaemia after unmanipulated haploidentical blood and marrow transplantation.

Systematic, standardised pretransplant risk assessment is an important tool for predicting patient outcomes following allogeneic haematopoietic SCT (HSCT). To assess the European Group for Blood and Marrow Transplantation (EBMT) risk score capacities for predicting patient outcomes following unmanipulated haploidentical blood and marrow transplantation (HBMT), we analysed 502 leukaemia patients who received transplants at our centre between 2008 and 2010. The cohort OS and leukaemia-free survival (LFS) were 72.1% and 68.1%, whereas the cumulative non-relapse mortality (NRM) and relapse incidences were 16.5% and 16.1%. According to univariate analysis, the values for OS, LFS and NRM were worse for an EBMT risk score of 6 (40.0, 40.0, 50.0%) than a score of 1 (83.1, 78.3, 8.4%). Hazard ratios steadily increased for each additional score point. Likewise, a higher EBMT risk score was associated with an increased relapse incidence. Importantly, the EBMT risk score prognostic value regarding OS, LFS, NRM and relapse was maintained in the multivariate analysis. Moreover, we also made a haploidentical EBMT (haplo-EBMT) risk score, which used number of HLA disparity instead of donor type, and the haplo-EBMT risk scores can also be used to predict patient outcomes following unmanipulated HBMT.

Validation of the European Group for Blood and Marrow Transplantation (EBMT) risk score in patients receiving allogeneic hematopoietic stem cell transplantation at a single center in Japan

We validated the European Group for Blood and Marrow Transplantation (EBMT) risk score in 273 consecutive adult patients receiving allogeneic hematopoietic stem cell transplantation between 2000 and 2010 at our center. The patients were divided into four groups according to the EBMT risk score: low risk (LR, score 0-2), intermediate risk-1 (IR-1, score 3), intermediate risk-2 (IR-2, score 4), and high risk (HR, score 5-7). The five-yr overall survival of the LR (n=65), IR-1 (n=67), IR-2 (n=70), and HR (n=71) groups was 72%, 57%, 41%, and 25%, respectively (p<0.001). The five-yr transplant-related mortality rates were 16%, 30%, 25%, and 36%, respectively (p=0.07). The five-yr cumulative incidence of relapse was 20%, 18%, 37%, and 41%, respectively (p<0.001). In the subgroup analysis, the prognostic value of the EBMT risk score was confirmed in patients undergoing myeloablative conditioning (MAC), but not in those undergoing reduced-intensity conditioning (RIC). The results suggest that the EBMT risk score is a useful tool to predict transplant outcome for patients undergoing MAC, but not for those undergoing RIC and may be beneficial for stratifying patients in clinical studies.

Randomized Trial On GvHD Prophylaxis with or without Anti T Cell Globulin ATG Fresenius (ATG-F) in Allogeneic Hematopoietic Cell Transplantation From Matched Unrelated Donors: Updated Results and Evaluation of the European Group for Blood and Marrow Transplantation (EBMT) Risk Score.

AbstractAbstract 3090 Background:Previously, in 201 adult patients with allogeneic hematopoietic cell transplantation from matched unrelated donors, we demonstrated that addition of ATG-F to standard cyclosporine, methotrexate GvHD prophylaxis (control) significantly reduces acute and chronic GvHD without negatively affecting non-relapse mortality (NRM), relapse, disease-free survival (DFS) and overall survival (OS) [1,2]. Methods:Now, we present updated results of extended follow-up (median 5.5, Q1 5.0, Q3 6.5 years). Furthermore, we investigated whether the European Group for Blood and Marrow Transplantation (EBMT) risk score (0 to 7 points) based on the sum of points for patient age (20–40y: 1 point, >40y: 2 points), disease stage (intermediate: 1 point, late: 2 points), time interval from diagnosis to transplant (>12m: 1 point), donor type (unrelated: 1 point), and donor-patient sex combination (donor female, patient male: 1 point) [3] was able to predict outcome with respect to OS in our trial. The explained variation R2 of the risk score was calculated by setting the prediction error (Integrated Brier Score) of the score in relation to the prediction error of an overall estimate of OS without dividing the patients into different risk groups. Results:The incidence of extensive chronic GvHD after 5 years was 13.6% in the ATG-F group vs 48.9% in the control group (p<0.0001). The 5-year rates with respect to outcome were: DFS 44.3% vs 37.2% (p=0.67), relapse 35.2% vs 29.9% (p=0.46), relapse mortality 29.7% vs 27.7% (p=0.99), NRM 20.5% vs 32.9% (p=0.15), and OS 49.8% vs 39.1% (p=0.28), ATG-F vs control respectively. Patients’ distribution to the EBMT risk score was 1 point 2, 2 points 43, 3 points 59, 4 points 43, 5 points 32, 6 points19, 7 points 1, unknown 2.Table 1Prognostic effect of the EBMT risk score, 5-year OS rates in the different risk groups in comparison to the respective rates reported for the EBMT patients [3]ATG-F-trialEBMT patients [3]EBMT scorePatients/deathsHazard Ratio 95% CI5-year OS rate 95%-CIEBMT score5-year OS rate 95%-CI1-245/191.000.55 [0.39,0.69]10.63 [0.60,0.67]10.54 [0.51,0.58]359/321.42 [0.80,2.50]0.45 [0.32,0.57]30.44 [0.40,0.48]443/231.51 [0.82,2.78]0.48 [0.32,0.62]40.33 [0.29,0.36]5-752/352.19 [1.25,3.83]0.32 [0.20,0.45]50.24 [0.18,0.31]The trend test for increasing risk with increasing risk score resulted in p=0.006. The risk score is able to separate patients with regard to OS, but in univariate and multivariate analyses of the factors building the score (age, disease stage, time from diagnosis, patient/donor sex) only disease stage showed a significant and relevant effect on OS (hazard ratio: intermediate vs early 1.54 95%-CI [0.92,2.56], late vs early 1.98 [1.30,3.02], trend test p=0.001). Thus, the addition of points to the risk score for factors that do not have a prognostic effect results in a dilution of its prognostic ability. In fact, the explained variation of the factor disease stage alone was higher than that of the risk score (R2=3.8% vs R2=3.4%).The main prognostic factor in our trial besides disease stage was donor age [4]. We therefore analyse the effect of a new risk score based on the sum of points for these two factors: donor age (>40y: 1 point) and disease stage (intermediate: 1 point, late: 2 points).Table 2Prognostic effect of new risk score, and 5-year OS rates.ATG-F-trialNew scorePatients/deathsHazard Ratio 95% CI5-year OS rate 95%-CI061/221.000.62 [0.48,0.73]169/412.12 [1.26,3.57]0.41 [0.29,0.52]245/312.69 [1.56,4.66]0.30 [0.18,0.44]313/113.86 [1.87,7.97]0.15 [0.02,0.39]The trend test for increasing risk with increasing score resulted in p<0.0001, and the explained variation was R2=8.1%. Conclusion:After a long-term follow-up of 5.5 years, it could be confirmed that ATG-F GvHD prophylaxis provides a sustained protective effect without compromising relapse and survival. The EBMT risk score is able to separate patients with regard to OS, but in our trial another risk score based on disease stage and donor age can be defined leading to a better prediction of OS. As this score was derived in the same dataset, the assessment of its predictive ability is overoptimistic and it has to be validated in an independent study. Disclosures:Finke:Fresenius Biotech GmbH: Honoraria, Research Funding. Bethge:Fresenius Biotech GmbH: Lecture remuneration Other. Bertz:Fresenius Biotech GmbH: Lecture remuneration Other.

EBMT Risk Score Predicts Outcome of Allogeneic Hematopoietic Stem Cell Transplantation in Patients Who Have Failed a Previous Transplantation Procedure

Increasing numbers of allogeneic hematopoietic stem cell transplantation (allo-SCT) are being performed for patients who have failed a previous allogeneic or autologous SCT. We investigated whether the EBMT risk score could predict outcome after a subsequent allo-SCT. We analyzed prognostic factors in 124 consecutive patients who underwent a second transplantation using an allogeneic donor at our institution. Patients with either a first autologous (N = 64) or first allogeneic (N = 60) SCT were included. Age, disease stage, time interval from diagnosis to transplantation, donor type, and donor-recipient sex combination were used to establish a score from 0 to 7 points, from which 3 groups were identified. The 5-year survival probability decreased from 51.7% for risk scores 0-3 (low, n = 25), to 29.3% for risk score 4 (intermediate, n = 42), and only 10.4% for risk scores 5-7 (high, n = 57), P = .001. We propose that the EBMT risk score can identify patients most likely to benefit from a second transplantation.

The EBMT risk score

The European Group for Blood and Marrow Transplantation (EBMT) risk score provides a simple tool to assess instantly chances and risks of hematopoietic SCT(HSCT) for an individual patient pre-transplant. Five factors, age of the patient, stage of the disease, time from diagnosis, donor type and donor recipient gender combination augment risk for an individual patient with increasing score from 0 as best to 7 as worst in an additive way. The score holds for all acquired hematological disorders, for allogeneic and autologous HSCT (score 0-5), is independent of the HSCT technology and is valid for standard or reduced intensity conditioning. Survival is uniformly worse for older patients, transplanted in advanced disease stage after a long-time interval and with a mismatched donor than for younger patients, transplanted soon in early stage with a well matched donor. Additional risk factors such as performance score, CMV serostatus or cytokine polymorphisms improve prediction but to different extents for low or high-risk patients. Comparative assessment of disease risk and global pre-transplant risk should guide decisions for each patient with his/her specific disease between HSCT and a non-transplant approach and replace the traditional 'donor vs no donor' with such a risk-adapted individualized strategy.

Allogeneic Hematopoietic Stem-Cell Transplantation for Myeloma: It's Time for the Appropriate Studies

TO THE EDITOR: The International Myeloma Working Group recently published a consensus statement regarding the use of allogeneic hematopoietic stem-cell transplantation (HSCT) in myeloma. 1 They conclude that new strategies are needed to make allogeneic HSCT safer and recommend reduced intensity conditioning in the context of clinical trials only. The group is to be congratulated for taking up the task for such a consensus statement, and the main conclusions are undoubtedly correct. It is essential to render allogeneic HSCT safer, and further clinical studies are urgently warranted. However, the report misses some important aspects. Some results with allogeneic HSCT for patients with myeloma could be looked at from a different point of view. The conclusions to be drawn and the way forward might then look completely different from the way they are presented in the consensus statement. The authors report that non–relapse mortality is excessively high after allogeneic HSCT for myeloma; they give no explanation for why this is the case. A recent analysis of the European Group for Blood and Marrow Transplantation (EBMT) sheds some light on this erroneous observation. Survival and non–relapse mortality of 1,351 patients with myeloma was not different from survival and non–relapse mortality of patients with any other acquired hematologic disease when risk was adjusted for by the EBMT risk score. 2 In contrast, risk distribution was strikingly different. The risk score includes age, disease stage, time interval from diagnosis, donor type (HLA-identical sibling donor v other), and donor-recipient sex combination (female donor for a male recipient v other). Median age of the patients with myeloma was 46 years, and 60% of them had received transplantations with an EBMT score of 4 or higher. Median age of the 16,000 patients with acute myeloid leukemia was 35 years, and only 24% who had received transplantations had a score of 4 or higher. The sense that there is a higher mortality rate with myeloma is correct. The cause is not the disease but the wrong selection. Recommending allogeneic HSCT to patients for whom autologous HSCT and modern drug therapy have failed will only perpetuate the assumption that allogeneic HSCT is associated with high mortality and is not beneficial. There is no need for novel conditioning regimens. There is an urgent need to investigate standard allogeneic HSCT for the few selected high-risk patients with myeloma who have a donor with a low EBMT risk score. Transplantations should be considered early in the disease course, as soon as remission is reached, when an HLA-identical sibling donor or a fully matched unrelated donor is available for a younger patient with a normal Karnofsky score, and a male recipient should not be given a female donor transplant. Risk of non–relapse mortality will not be higher than that for a comparable patient with leukemia. A randomized study comparing such early HSCT with the current conventional approach could rapidly provide the answer. There is no need to precede the allogeneic HSCT with an autologous transplantation. The tandem approach, autologous HSCT followed by a reduced-intensity conditioning allogeneic transplantation is believed to have improved outcome. This study was reported at a time of general improvement in outcome and provided the impression of early low mortality. 3 No study since has ever confirmed in any disease that separation of cytoreduction and allogeneic HSCT is indeed superior to a direct, full-conditioning allogeneic HSCT. It is an attractive theoretical concept but there are no facts. The authors simply present it as such. There is a more important missing aspect. The authors mention several potential approaches, including reduced-intensity allogeneic HSCT combined with maintenance with one of the novel antimyeloma agents. There are no data, and the authors do not mention costs. Long-term treatment with these novel agents will not be affordable for the vast majority of the global population. An early standard allogeneic HSCT for those few selected patients with high-risk disease and a low-risk transplantation score might provide the most cost-effective approach. Such considerations have been made for patients with chronic myeloid leukemia, and they relate as well to patients with myeloma. 4 Consensus statements should consider global aspects. 5 It is the perfect time to perform the appropriate studies evaluating early allogeneic HSCT for those patients with high-risk disease and a lowrisk donor. The International Myeloma Working Group should be challenged to take on this task.

Failure to Recover Thymopoiesis After Allogeneic Hematopoietic Stem Cell Transplantation Predicts for High Incidence of Opportunistic Infections and Non-Relapse Mortality

AbstractAbstract 832Recovery of thymopoiesis after allogeneic hematopoietic stem cell transplantation (alloSCT) is considered pivotal for full immune competence. Failure to recover thymopoiesis after alloSCT is suggested by the absence of newly developed T-cells that contain signal joint T-cell Receptor Excision Circles (sjTREC), which are produced during T-cell receptor rearrangement. Higher age, reduced pre-transplant thymic function, and graft versus host disease have all been associated with impaired thymic recovery after alloSCT. However, it is still unclear to what extent insufficient thymic recovery itself predicts for subsequent opportunistic infections and non-relapse mortality (NRM). A detailed, prospective survey of all post-engraftment infectious complications, NRM, overall survival (OS), and lymphocyte subsets and thymic recovery during long-term follow-up after alloSCT was performed in 83 high-risk recipients of T cell depleted related or unrelated donor grafts after myeloablative conditioning. A cumulative incidence of common toxicity criteria (CTC) grade 3 and 4 severe infections at 12 months after alloSCT was 66% with a median number of 1.6 severe infectious episodes per patient. The rate of severe post-engraftment infections was 0.64 per 100 patient days during the first year and 0.16 in the second year. After a median follow-up of 118 months, OS was 52%, as determined by a NRM of 25% and relapse mortality of 23%. Outcome was significantly predicted for by the European Group for Blood and Marrow Transplantation (EBMT)-risk score (Gratwohl et al, Cancer, 2009), based on age, gender, risk-status, donor-type and time to SCT (p=0.01). Lymphocyte recovery was slow with median CD4+ T cells exceeding 200/ul by 12 months post alloSCT. At that time, median sjTREC content measured 5.629/ml, which was significantly lower than the median sjTREC content in stem cell donors (19.044/ml, p=0.001). The recovery rate of overall lymphocyte counts as well as T cell subsets did not consistently predict for opportunistic infections and NRM. Patients without thymic recovery at 2, 6, 9, or 12 months were at 3- to 9-fold higher risk for severe infections, which remained significant following multivariable analysis (hazard ratio (HR)-6 months: 0.30 (95%CI 0.09–1.02, p=0.04), HR-12 months: 0.11 (0.01-0.93 p=0.02)). Impaired thymic recovery also translated into a higher risk for NRM (HR-6 months: 0.06 (0.01-0.47 p=0.008), HR-12 months: 0.00 (0.00-1.42 p=0.0005)) and outweighed pre-transplant risk factors including age, donor-type, risk-status when evaluated individually or collectively as the EBMT risk score. Impaired thymic recovery at 6 months was also predictive for lower OS (HR; 0.35 (0.12-1.00 p=0.05)). In conclusion, these results indicate that patients, who fail to recover thymopoiesis after alloSCT are at very high risk for severe infections and adverse clinical outcome, independent of other pre-transplant risk factors and despite intensive monitoring and prophylactic antimicrobial measures. These results indicate that monitoring thymopoiesis after alloSCT may be applied more routinely and strategies to reduce NRM should be directed at thymic regeneration. Disclosures:No relevant conflicts of interest to declare.

Clinical and Genetic Risk Assessment for Overall Survival in Haematopoietic Stem Cell Transplantation (HSCT).

Abstract 1189Poster Board I-211 IntroductionIn the last decade, several groups have demonstrated that non-HLA gene polymorphisms can impact on outcome after HSCT. So far, their role appears as a matter of debate; non HLA genotyping is not yet routinely used. In addition, the main clinical factors influencing outcome after HSCT have also been described. Most experience has been gained through analyses of patients undergoing HSCT for chronic myeloid leukemia (CML). Patient age, stage of the disease, time interval from diagnosis to transplant, histocompatibility and donor and patient gender combination were initially identified as key pre-transplant risk factors for CML in the European Group for Blood and Marrow Transplantation (EBMT) clinical risk score, recently confirmed as risk factors for all patients with haematological indications for HSCT. Patients and MethodsIn this study, we assessed the additional impact of polymorphisms within the tumour necrosis factor receptor II (TNFRSF1B), estrogen receptor (ESR1), vitamin D receptor (VDR), interleukin 6 (IL6), interleukin 1 receptor antagonist (ILRN), interferon gamma (IFNG) and interleukin 4 (IL4) genes on overall survival in a EUROBANK cohort of 915 ( median patient age 41 years, range 16 years to 67 years; 59 % male patients) HLA identical sibling (n=501; 55%) and matched unrelated donor (MUD) (n=414; 45%) transplants consisting of patients having either acute leukaemia (AL) (n=463; 51%), CML (N=187; 20%), plasma cell neoplasia (n=120; 13%) or lymphoma (n=145; 16%) from 8 transplant centres (Barcelona, Paris, Munich, Newcastle-upon-Tyne, Prague, Regensburg, Vienna and Rostock). The statistical analysis was performed for the full cohort and for the AL subgroup. Potential influential genetic factors were assessed using the log rank test (p value < 0.2). These candidate factors were further included in addition to the EBMT clinical risk score in a stepwise Cox regression procedure to select final genes. The power of a model to predict overall survival was assessed through prediction error curves. ResultsFor the full cohort, the SNP IL6−174 within the IL6 gene in donors improved the prediction of the outcome by EBMT clinical risk score. Specifically, the presence of allele C in the donor IL-6 genotype was associated with lower survival time. This SNP assessment only held true for the (combined) CML, lymphoma and plasma cell neoplasia subgroups (log rank, p=0.016), not for AL (log rank, p= 0.638). The AL subgroup, the largest, was further assessed separately. Absence of patient IL4 (any T) and presence of patient IL1RN (any C) were associated with lower survival time as shown by Kaplan Meier survival plots. When viewed together with the EBMT score, these two genotypes improved the predictive value. Patients with absence of patient IL4 (any T) and presence of patient IL1RN (any C) were assigned to the ‘high risk' group; remaining patients were assigned to the low risk group - Figure 1A and B illustrate the lower probability of survival and higher incidence of transplant related mortality (TRM) in the high risk polymorphism AL group compared with the low risk group. [Display omitted] ConclusionsThis study confirms the important role of non-HLA genotyping for risk assessment in allogeneic HSCT. Improvement of fit of the EBMT risk score presents a powerful novel tool to assess this impact of cytokine gene polymorphisms in a complex heterogeneous patient population such as HSCT. DisclosuresNo relevant conflicts of interest to declare.

Validation and extension of the EBMT Risk Score for patients with chronic myeloid leukaemia (CML) receiving allogeneic haematopoietic stem cell transplants

The European Group for Blood and Marrow Transplantation (EBMT) devised a scoring system to predict survival after allogeneic haematopoietic stem cell transplantation (HSCT) for chronic myeloid leukaemia (CML). The present International Bone Marrow Transplant Registry study of 3211 patients tested the EBMT Risk Score in a independent population, investigated the value of adding other variables, evaluated a new risk score specifically for chronic phase and compared the allograft risk scores with risk scores established by Sokal in 1984 and Hasford in 1998 for survival with non-transplant treatments. The primary outcome was 5-year survival after HSCT; survival curves, regression models and measurements of explained variation were used to compare scores. Using the EBMT scoring system, survival in the independent dataset was almost identical to those in the original EBMT publication, thus validating the EBMT Risk Score. Adding one extra variable, performance status, or designing a score specifically for early chronic phase by using the original five variables with different breakpoints gave results only slightly better than the original EBMT Score. Sokal and Hasford Scores did not predict survival after HSCT. We concluded that the EBMT Risk Score does not currently require modification.