Stem Cell Transplantation ARTICLE Association of uric acid levels before start of Ferrata Storti Foundation conditioning with mortality after allogeneic hematopoietic stem cell transplantation – a prospective, non-interventional study of the EBMT Transplant Complication Working Party Olaf Penack,1 Christophe Peczynski,2 Steffie van der Werf,3 Jürgen Finke,4 Arnold Ganser,5 Helene Schoemans,6 Jiri Pavlu,7 Riitta Niittyvuopio,8 Wilfried Schroyens,9 Leylagül Kaynar,10 Igor W. Blau,1 Walter van der Velden,11 Jorge Sierra,12 Agostino Cortelezzi,13 Gerald Wulf,14 Pascal Turlure,15 Haematologica 2020 Montserat Rovira,16 Zubeydenur Ozkurt,17 Maria J. Pascual-Cascon,18 Volume 105(7):1977-1983 Maria C. Moreira,19 Johannes Clausen,20 Hildegard Greinix,21 Rafael F. Duarte22# and Grzegorz W. Basak23# 1 Charité Universitätsmedizin Berlin, Berlin, Germany; 2EBMT Statistical Unit, Paris, France; 3EBMT Data Office, Leiden, the Netherlands; 4University of Freiburg, Freiburg, Germany; 5Hannover Medical School, Hannover, Germany; 6Department of Hematology, University Hospital Leuven and KU Leuven, Leuven, Belgium; 7Imperial College, London, UK; 8HUCH Comprehensive Cancer Center, Helsinki, Finland; 9Antwerp University Hospital, Antwerp, Belgium; 10Erciyes University Medical Faculty, Kayseri, Turkey; 11 Radboud University – Nijmegen Medical Centre, Nijmegen, the Netherlands; 12Hospital Santa Creu I Sant Pau, Barcelona, Spain; 13Fondazione IRCCS – Ca’Granda, Milan, Italy; 14 Universitätsklinikum Göttingen, Göttingen, Germany; 15CHRU Limoges, Limoges, France; 16Hospital Clinic, Barcelona, Spain; 17Gazi University Faculty of Medicine, Ankara, Turkey; 18Hospital Regional de Málaga, Malaga, Spain; 19Instituto National do Cancer, Rio de Janeiro, Brazil; 20Elisabethinen-Hospital, Linz, Austria; 21LKH – University Hospital Graz, Graz, Austria; 22Hospital Universitario Puerta de Hierro, Madrid, Spain and 23 Medical University of Warsaw, Warsaw, Poland # RFD and GWB contributed equally as co-senior authors. Correspondence: ABSTRACT OLAF PENACK olaf.penack@charite.de U ric acid is a danger signal contributing to inflammation. Its relevance to allogeneic stem cell transplantation (alloSCT) derives from preclin- Received: June 4, 2019. ical models where the depletion of uric acid led to improved survival Accepted: October 4, 2019. and reduced graft-versus-host disease (GvHD). In a clinical pilot trial, peri- transplant uric acid depletion reduced acute GvHD incidence. This prospec- Pre-published: October 10, 2019. tive international multicenter study aimed to investigate the association of uric acid serum levels before start of conditioning with alloSCT outcome. doi:10.3324/haematol.2019.228668 We included patients with acute leukemia, lymphoma or myelodysplastic syndrome receiving a first matched sibling alloSCT from peripheral blood, Check the online version for the most updated regardless of conditioning. We compared outcomes between patients with information on this article, online supplements, high and low uric acid levels with univariate- and multivariate analysis using and information on authorship & disclosures: www.haematologica.org/content/105/7/1977 a cause-specific Cox model. Twenty centers from 10 countries reported data on 366 alloSCT recipients. There were no significant differences in terms of baseline co-morbidity and disease stage between the high- and low uric acid ©2020 Ferrata Storti Foundation group. Patients with uric acid levels above median measured before start of Material published in Haematologica is covered by copyright. conditioning did not significantly differ from the remaining in terms of acute All rights are reserved to the Ferrata Storti Foundation. Use of GvHD grades II-IV incidence (Hazard ratio [HR] 1.5, 95% Confidence inter- published material is allowed under the following terms and conditions: val [CI]: 1.0-2.4, P=0.08). However, they had significantly shorter overall sur- https://creativecommons.org/licenses/by-nc/4.0/legalcode. vival (HR 2.8, 95% CI: 1.7-4.7, P<0.0001) and progression free survival (HR Copies of published material are allowed for personal or inter- 1.6, 95% CI: 1.1-2.4, P=0.025). Non-relapse mortality was significantly nal use. Sharing published material for non-commercial pur- poses is subject to the following conditions: increased in alloSCT recipients with high uric acid levels (HR 2.7, 95% CI: https://creativecommons.org/licenses/by-nc/4.0/legalcode, 1.4-5.0, P=0.003). Finally, the incidence of relapse after alloSCT was sect. 3. Reproducing and sharing published material for com- increased in patients with higher uric acid levels (HR 1.6, 95% CI: 1.0-2.5, mercial purposes is not allowed without permission in writing from the publisher. P=0.04). We conclude that high uric acid levels before the start of condition- ing correlate with increased mortality after alloSCT. haematologica | 2020; 105(7) 1977 O. Penack et al. Introduction Endpoints and statistical analyses Patient, disease, and transplant-related characteristics for the High treatment-associated complication rates remain an two cohorts (uric acid levels prior to alloSCT above median / uric important challenge to the success of allogeneic acid levels below median) were compared by using 2 statistics for hematopoietic stem cell transplantation (alloSCT). A com- categorical variables and the Mann-Whitney test for continuous mon mechanism of the major infectious- and non-infec- variables. The primary endpoint was the incidence of acute tious complications after alloSCT, such as graft-versus-host GvHD. Secondary endpoints were relapse incidence (RI), non- disease (GvHD) and sepsis, is severe inflammation. relapse mortality (NRM), overall survival (OS), progression free Critical to excess inflammation after alloSCT are host bio- survival (PFS) and the incidence of chronic GvHD. PFS was molecules termed ‘endogenous danger signals’, which can defined as survival with no evidence of relapse or progression. RI initiate and maintain non-infectious inflammatory was defined as the probability of having had a relapse during fol- responses by acting as pro-inflammatory mediators.1,2 One low-up time. Death without experiencing a relapse was a compet- such molecule is uric acid. During conditioning for ing event. NRM was defined as death without evidence of relapse alloSCT, uric acid is released from injured cells and or progression. OS was defined as the time from alloSCT to death, reduced renal clearance may contribute to high uric acid regardless of the cause. Acute GvHD was graded according to the serum levels.3,4 Uric acid acts as a danger signal by enhanc- modified Seattle-Glucksberg criteria9 and chronic GvHD accord- ing T-cell responses via activation of the NOD-like recep- ing to the revised Seattle criteria.10 Cumulative incidence was used tor protein (NLRP) 3 inflammasome.5,6 A pre-clinical study to estimate the endpoints of NRM, RI, acute and chronic GvHD to has provided basic evidence on the regulation of GvHD by accommodate for competing risks. To study acute and chronic uric acid via Nlrp3 inflammasome–mediated IL-1 produc- GvHD, we considered relapse and death to be competing events. tion.7 The significance of uric acid for alloSCT outcome Probabilities of OS and PFS were calculated using the Kaplan– has been underlined by results from a pilot study reporting Meier method. Univariate analyses were done using the Gray test reduced incidence of acute GvHD in patients undergoing for cumulative incidence functions and the log rank test for OS alloSCT after depletion of uric acid using urate oxidase.8 and PFS. A Cox proportional hazards model was used for multi- Based on these results, and because uric acid is a routine variate regression. All variables differing significantly between the laboratory parameter assessed in patients undergoing two groups or factors associated with one outcome in univariate alloSCT, there is a strong rationale to investigate the use analysis were included in the Cox model. The following variables of uric acid as a biomarker and possibly as a therapeutic entered the multivariate models as possible confounders: age, sex target during alloSCT. To study the role of uric acid levels mismatch between recipient and donor, diagnosis, disease status, in alloSCT, the Transplant Complications Working Party Karnofsky score, number of CD34 cells given, intensity of condi- (TCWP) of the European Society for Blood and Marrow tioning (EBMT definition: myeloablative conditioning (MAC) was Transplantation (EBMT) performed a prospective, multi- defined as total body irradiation (TBI) >6 Grey or oral busulfan >8 center, non-interventional study. We hypothesized that a mg/kg or intravenous busulfan >6.4 mg/kg), type of GvHD pro- high uric acid level prior to alloSCT is an independent risk phylaxis, ATG use, time from diagnosis to transplant, year of factor for increased mortality and for the development of transplant and CMV status. As the number of variables was too severe acute GvHD. high regarding the number of events, a stepwise selection using Akaike information criterion (AIC) was run for all the confound- Methods ing factors. The difference between the two cohorts was then assessed in the final selected model. Data source, study design and data collection Results were expressed as the hazard ratio (HR) with the 95% We asked EBMT centers performing more than 50 alloSCT per confidence interval (95% CI). Proportional hazards assumptions year if they were willing to participate in this prospective study. were checked systematically for all proposed models using the Twenty centers in ten countries agreed to participate. Data collec- Grambsch-Therneau residual-based test. All tests were two-sided. tion for the EBMT registry was approved by the institutional The type I error rate was fixed at 0.05 for the determination of fac- review board and/or Ethics Committee in all centers. Data were tors associated with time-to-event outcomes. Statistical analyses prospectively collected between August, 2014 and February, were performed in November 2018 with R 3.4.2 (R Core Team 2018. Consecutive alloSCT recipients with acute leukemia, lym- (2017). R: A language and environment for statistical computing. R phoma or myelodysplastic syndrome (MDS) receiving a first Foundation for Statistical Computing, Vienna, Austria. URL matched sibling alloSCT from peripheral blood, regardless of con- https://www.R-project.org/.) ditioning, were eligible, provided they had signed an informed consent document that permitted sharing of clinical data accord- ing to national rules. Basic data on patient and disease characteris- Results tics as well as longer term follow-up was taken from minimal essential data (MED-A) forms, which are submitted from all con- Patients, transplant characteristics and uric acid secutive patients to the central EBMT registry. In addition, we measurement designed registration and MED-B/C forms that were prospectively The entry criteria for analysis of primary and secondary collected and specific to this study (see the Online Supplementary endpoints were fulfilled in 386 patients. We used the last Materials and Methods). The MED-B/C form contained detailed uric acid serum level that was masured in the individual information on uric acid serum levels prior to alloSCT, patient patients before start of conditioning. The median time characteristics, infectious- as well as non-infectious complications, point of measurement was three days before start of con- GvHD staging, morbidity and mortality. Uric acid levels were ditioning. The range was between 0 days (morning before determined at the time of hospital admission for alloSCT directly start of conditioning) to 22 days before conditioning. The before the start of conditioning therapy. Treatment teams com- main patients and transplant characteristics that were pleted the specific forms at the time of registration and at day included in the analysis of OS are described in Table 1. +100 after alloSCT. Most parameters were balanced between the two cohorts. 1978 haematologica | 2020; 105(7) High uric acid levels and increased mortality after alloSCT Table 1. Patient characteristics. Uric acid ≤4.3 mg/dL (n=186) Uric acid >4.3 mg/dL (n=180) P Year of transplant median (range)[IQR] 2015 (2014-2018) [2015-2017] 2015 (2014-2018) [2014-2016] 0.10 Patient age (years) median(range) 52 (17-71) 55 (20-71) 0.11 Time from diagnosis to transplant 4 (1-61) 5 (1-71) 0.09 (months) median (range) Sex mismatch Female to male 30 (17%) 52 (29%) 0.005 Other combination 148 (83%) 125 (71%) Missing 8 3 Diagnosis Acute leukemia 130 (70%) 118 (66%) 0.7 Lymphoma 17 (9%) 19 (10%) MDS 39 (21%) 43 (24%) Disease status CR 120 (66%) 125 (72%) 0.2 Not in CR 61 (34%) 48 (28%) Missing 5 7 DRI 0.7 Low 8 ( 4% ) 7 ( 4% ) Intermediate 107 ( 60% ) 105 ( 63% ) High 51 ( 30% ) 49 ( 30% ) Very high 10 ( 6% ) 5 (3% ) Missing 10 14 ATG No 100 (54%) 106 (59%) 0.3 Yes 86 (46%) 74 (41%) Conditioning intensity MAC/Chemo 46 (25%) 37 (21%) 0.2 MAC/TBI 36 (20%) 26 (15%) RIC 102 (55%) 115 (64%) Missing 2 2 GvHD prophylaxis Calcineurin inhibitor + MMF 85 (48%) 84 (47%) 0.6 Calcineurin inhibitor + MTX 64 (36%) 71 (40%) Calcineurin inhibitor mono 27 (15%) 22 (12%) Missing 10 3 Donor/patient CMV 0.2 -/- 28 (16%) 38 (21%) Other 146 (83%) 136 (78%) Missing 12 6 Karnofsky score 0.14 10-80 38 (21%) 27 (15%) 90-100 140 (78%) 149 (84%) Missing 8 4 HT-CI = 0 64 ( 50% ) 56 ( 40% ) 0.19 HT-CI = 1 or 2 20 ( 15% ) 32 ( 23% ) HT-CI ≥3 45 ( 35% ) 51 ( 37% ) Missing 57 41 MDS: myelodysplastic syndrome; IQR: interquartile range; CR: complete remission; DRI: disease risk index; ATG: anti-thymocyte globulin; MAC: myeloablative conditioning; RIC: reduced intensity conditioning; TBI: total body irradiation; GvHD: graft-versus-host disease; MMF: Mycophenolate mofetil; MTX: Methotrexate; CMV: cytomegalovirus; HT-CI: hematopoietic cell transplantation cumulative incidence. haematologica | 2020; 105(7) 1979 O. Penack et al. A B Strata ≤ 4.3 mg/dL > 4.3 mg/dL OS PFS Strata ≤ 4.3 mg/dL > 4.3 mg/dL ≤ 4.3 mg/dL ≤ 4.3 mg/dL > 4.3 mg/dL PFS OS > 4.3 mg/dL Figure 1. Survival at one year after allogeneic stem cell transplantation. Overall survival (OS) (A) and progression free survival (PFS) (B) after allogeneic stem cell transplantation (alloSCT) in cohorts according to uric acid serum levels prior to alloSCT: with high uric acid serum levels (blue line), patients with low uric acid levels (red line). Table 2. Univariate global comparison of acute graft-versus-host dis- However, a higher percentage of sex mismatch transplants ease shown at day +100 after allogeneic stem cell transplantation. in the direction of female donor to male recipient was Group Acute GvHD Acute GvHD observed in the group of patients with uric acid above II-IV (95% CI) III-IV (95% CI) median before alloSCT. The frequency of patients with reduced-intensity conditioning (RIC) versus myeloablative Uric acid 4.3 mg/dL 26% (20-33) 14% (9-19) the high uric acid group and the low uric acid group P-value 0.39 0.08 (P=0.077). When looking exclusively at TBI (P=0.59) or by GvHD: graft-versus-host disease; CI: confidence interval. mixing conditioning intensity and TBI (MAC/Chemo vs. MAC/TBI vs. RIC: P=0.20), no statistically significant dif- ference could be observed. 74%. We found, that OS and PFS of alloSCT recipients with uric acid levels above cut-off, measured before start Incidence of acute and chronic GvHD of conditioning, were significantly shorter compared with In the present study, the incidence of acute GvHD the low uric acid cohort (Figure 1A, OS univariate HR 2.4, grades II-IV and grades III-IV in the whole population at 95% CI: 1.6-3.7, P<0.001; multivariate HR 2.8, 95% CI: 100 days were 25% and 11%, respectively. We observed 1.7-4.7, P<0.0001) (Figure 1B, PFS univariate HR 2.0, 95% no significant differences in the frequency of clinically sig- CI: 1.1-3.7, P=0.02; multivariate HR 2.7, 95% CI: 1.4-5.0, nificant acute GvHD grades II-IV in univariate (HR 1.2, P=0.003). Non-relapse mortality was significantly 95% CI: 0.8-1.9, P=0.4) and multivariate (HR 1.5, 95% CI: increased in alloSCT recipients with high uric acid levels 1.0-2.4, P=0.08) analysis comparing the high- versus the prior to start of conditioning (Figure 2A, univariate HR 2.0, low uric acid cohorts. (Table 2 and Table 3). We found no 95% CI: 1.1-3.7, P=0.02; multivariate HR 2.7, 95% CI: 1.4- significant differences in the frequency of acute GvHD 5.0, P=0.003). Multivariate as well as univariate analyses grades III-IV in univariate (HR=1.8, 95% CI: 0.9-3.4, are shown in Table 3 and Table 4. We conclude that serum P=0.08) and multivariate (HR 1.7, 95% CI: 0.9-3.2, P=0.11) uric acid levels prior to alloSCT are an independent risk analysis in alloSCT recipients with uric acid levels above factor for mortality. cut-off before transplantation versus those with low uric At the last follow-up 29.3% of the patients were dead acid levels (Table 2 and Table 3). (49.6% from relapse, 45.2% alloSCT-related causes and In the whole population, the incidence of chronic 2.6% from other causes). The causes of death in patients GvHD at one year and two years was 27% and 40%, without relapse were mainly due to infection. A descrip- respectively. The incidence of severe chronic GvHD at 1 tive analysis of the causes of death is given in Table 5. year and 2 years was 18% and 24%, respectively. We observed no differences in incidence and severity of Incidence of infections chronic GvHD between the two cohorts (Table 3 and We first looked at the incidence of infections in both Table 4). The chronic GvHD incidence was significantly cohorts and found no significant differences (multivariate lower in alloSCT recipients receiving anti-T-cell globulin HR 0.8, 95% CI: 0.6-1.1, P=0.4). We than looked at infec- (ATG) as part of the conditioning regimen (ATG, HR 0.25 tion-related mortality and found a significant higher num- 95% CI: 0.13-0.5, P<0.0001). ber of infection-related deaths in the high uric acid group (Figure 2C, multivariate HR 2.6, 95% CI: 1.2-5.2, P=0.01). Survival endpoints The median follow-up time was 15.2 months (95% CI: Incidence of relapse 13.1-17.0) and OS in the whole population at 1 year was Since uric acid is known to be released from malignant 1980 haematologica | 2020; 105(7) High uric acid levels and increased mortality after alloSCT A B ≤ 4.3 mg/dL ≤ 4.3 mg/dL > 4.3 mg/dL > 4.3 mg/dL C Figure 2. Non-relapse mortality, relapse incidence and infection-related mortality at one year after allogeneic stem cell transplantation. Incidence of non-relapse mortality (NRM) (A), relapse incidence (RI) (B) and infection-related mortality (C) after allogeneic stem cell transplanta- tion (alloSCT) in cohorts according to uric acid serum levels prior to alloSCT: patients with high uric acid serum levels (blue line), patients with ≤ 4.3 mg/dL low uric acid levels (red line). > 4.3 mg/dL Table 3. Multivariate global comparison. versus ≤4,3 mg/dL Uric acid >4.3 mg/dL HR 95% CI P Overall Survival 2,83 1,72 - 4,68 <0.0001 Progression Free Survival 1,59 1,06 - 2,39 0.03 Relapse Incidence 1,59 1,02 - 2,49 0.04 Non-Relapse Mortality 2,65 1,41 - 5,01 0.003 Chronic GvHD 0,96 0,63 -1,45 0.84 Extensive Chronic GvHD 1,08 0,62 - 1,88 0.79 Acute GvHD II-IV 1,52 0,95 - 2,43 0.08 Acute GvHD III-IV 1,68 0,88 - 3,19 0.11 GvHD: graft-versus-host disease; HR: hazard ratio; CI: confidence interval. tumours, we were interested in a possible connection of In animal experiments and in a clinical pilot study the uric acid levels and disease relapse. We found that the depletion of uric acid with rasburicase led to a reduced fre- relapse incidence in all patients at 1 year was 13%. The quency of severe GvHD.7,8 Therefore, our initial hypothe- relapse incidence was moderately increased in the cohort sis was to primarily find an association between uric acid with higher uric acid levels (Figure 2B, Table 3 and Table with acute GvHD and then secondarily with NRM. 4, univariate HR 1.6, 95% CI: 1.0-2.5; P=0.09; multivariate However, we found a strong positive association of uric HR 1.6, 95% CI: 1.0-2.5, P=0.04). acid levels with mortality after alloSCT whereas the asso- ciation between uric acid levels and acute GvHD inci- dence was not statistically significant. One of the reasons Discussion could be the low incidence of GvHD in our cohort due to the patient characteristics of human leukocyte antigen This prospective study identifies uric acid levels, taken (HLA)-identical sibling transplant with a high proportion before start of conditioning therapy, as a laboratory bio- of anti-thymocyte globulin (ATG) use. The use of in vivo marker to predict mortality after alloSCT. T-cell depletion with ATG in HLA-identical sibling haematologica | 2020; 105(7) 1981 O. Penack et al. Table 4. Univariate global comparison of overall survival, non-relapse mortality, relapse incidence as well as chronic graft-versus-host disease incidence and severity at one year after allogeneic stem cell transplantation. Group OS PFS RI NRM cGvHD Extensive cGvHD (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI) Uric acid 4.3 mg/dL (59-74) (48-64) (20-34) (12-24) (19-33) (12-25) P-value <0.001 0.001 0.09 0.02 0.43 0.85 OS: overall survival; PFS: progression free survival; NRM: non-relapse mortality; RI: relapse incidence: cGvHD: chronic graft-versus-host disease; CI: confidence intervall. Table 5. Causes of death in both cohorts. Uric acid ≤4.3 Uric acid >4.3 alloSCT in European centres increased recently.11 This trend is reflected by a higher frequency of ATG use in the present study (41-46%) than we had initially expected. (n=186) mg/dL mg/dL (n=180) Another likely reason is a hyperinflammatory status con- Alive at last follow up 155 (83.3%) 106 (58.9%) nected to high uric acid levels not restricted to acute Dead 31 (16.7%) 74 (41.1%) GvHD with a negative impact on different relevant clini- cal situations after alloSCT, such as sepsis12 and adult res- Dead due to relapse 13 (6.9%) 38 (21.1%) piratory distress syndrome (ARDS).13 This view is sup- Dead without relapse 18 (9.7%) 36 (20.0%) ported by our observation that death due to both infec- Dead due to infection 5 (3.2%) 20 (11.1%) tious and non-infectious complications after alloSCT was Dead due to GvHD 2 (1.0%) 10 (5.6%) increased in patients with high uric acid levels. Dead due to infection and GvHD 3 (1.6%) 5 (2.8%) Interestingly, an association of uric acid levels with decreased solid organ transplant survival has been Dead due to other causes or unknown 8 (4.3%) 1 (0.5%) described further underlining the significance of uric acid GvHD: graft-versus-host disease. for transplantation biology.14,15 Our study extends these findings to the setting of alloSCT. Our study contradicts results from a previous retrospec- population was restricted to alloSCT from HLA-identical tive single centre study, which did not show any signifi- sibling donors. We are therefore unable to draw definite cant association of uric acid levels prior to transplantation conclusions from these results regarding the association of to mortality after alloSCT.16 Differences in study designs uric acid levels with outcome in matched unrelated donor between the current larger, multicentre and prospective alloSCT or in haploidentical alloSCT, which are increas- study versus the already published smaller, single centre ingly used. Another limitation is the incomplete under- and retrospective study may be the reason for the differ- standing of the effects of uric acid on post alloSCT immu- ing results. nity. While previous studies showed that uric acid leads to A limitation of this clinical study is the lack of mecha- a inflammatory status in leukocytes and our results nistic insight on the role of uric acid in development of demonstrate impaired survival in alloSCT recipients with complications after alloSCT. The study design did not uric acid levels above median, it may be premature to con- allow the investigation of whether the impaired outcome clude that uric acid is a negative factor at any time around in the high urid acid cohort is due to an active role of uric alloSCT. There are clinical situations where immune acti- acid in causing inflammation or whether it is because uric vation is necessary and desired – e.g. in anti-infectious and acid primarily is a surrogate biomarker reflecting a hyper- anti-tumor immunity. The impact of depleting uric acid on inflammatory status and/or activity of the underlying immune activity in these situations remains to be deter- malignancy. However, we found no significant differences mined. In addition, it is possible that uric acid is involved regarding the disease status as well as disease risk index in immune reconstitution as it has been suggested recently (DRI) (complete remission [CR] vs. non-CR) in between that changes in uric acid serum levels can indicate incipi- the two groups (high uric acid vs.low uric acid). In addition ent or remaining immunological activity after SCT or there were no significant differences in the performance induction therapy in patients with hematologic malignan- status as well as the alloSCT-CI in between the two cies.17 groups. This data is no proof, but rather points in the In conclusion, this study supports the use of serum uric direction that high uric acid values were not merely a acid levels as biomarker for alloSCT outcome. reflection of advanced disease or increased co-morbidity in our study. Acknowledgments A further limitation of our results is that our patient JP acknowledges the support of the Imperial NIHR-BCR. 2. Zeiser R, Blazar BR. Acute graft-versus-host Marrow Transplant. 1992;10(4):337-339. References disease - biologic process, prevention, and 4. Joo SH, Park JK, Lee EE, et al. Changes in 1. Apostolova P, Zeiser R. The role of purine therapy. N Engl J Med. 2017;377(22):2167- serum uric acid levels after allogeneic metabolites as DAMPs in acute graft-ver- 2179. hematologic stem cell transplantation: A sus-host disease. Front Immunol. 2016; 3. Cannell PK, Herrmann RP. Urate metabolism retrospective cohort study. Blood Res. 7:439. during bone marrow transplantation. Bone 2016;51(3):200-203. 1982 haematologica | 2020; 105(7) High uric acid levels and increased mortality after alloSCT 5. Martinon F, Petrilli V, Mayor A, et al. Gout- GVHD grading. Bone Marrow Transplant. acid is an independent predictor of cardiac associated uric acid crystals activate the 1995;15(6):825-828. allograft vasculopathy after heart transplan- NALP3 inflammasome. Nature. 2006; 10. Lee SJ, Vogelsang G, Flowers ME. Chronic tation. J Heart Lung Transplant. 2018; 440(7081):237-241. graft-versus-host disease. Biol Blood 37(9):1083-1092. 6. Braga TT, Forni MF, Correa-Costa M, et al. Marrow Transplant. 2003;9(4):215-233. 15. Kim DG, Choi HY, Kim HY, et al. Soluble uric acid activates the NLRP3 11. Kroger N, Solano C, Wolschke C, Bandini Association between post-transplant serum inflammasome. Sci Rep. 2017;7:39884. G, et al. Antilymphocyte globulin for pre- uric acid levels and kidney transplantation 7. Jankovic D, Ganesan J, Bscheider M, et al. vention of chronic graft-versus-host dis- outcomes. PloS One. 2018; The Nlrp3 inflammasome regulates acute ease. N Engl J Med. 2016;374(1):43-53. 13(12):e0209156. graft-versus-host disease. J Exp Med. 12. Pehlivanlar-Kucuk M, Kucuk AO, et al. The 16. Ostendorf BN, Blau O, Uharek L, et al. 2013;210(10):1899-1910. association between serum uric acid level Association between low uric acid levels 8. Yeh AC, Brunner AM, Spitzer TR, et al. and prognosis in critically Ill patients, uric and acute graft-versus-host disease. Ann Phase I study of urate oxidase in the reduc- acid as a prognosis predictor. Clin Lab. Hematol. 2015;94(1):139-144. tion of acute graft-versus-host disease after 2018;64(9):1491-1500. 17. Haen SP, Eyb V, Mirza N, et al. Uric acid as myeloablative allogeneic stem cell trans- 13. Lee HW, Choi SM, Lee J, et al. Serum uric a novel biomarker for bone-marrow func- plantation. Biol Blood Marrow Transplant. acid level as a prognostic marker in patients tion and incipient hematopoietic reconsti- 2014;20(5):730-734. with acute respiratory distress syndrome. J tution after aplasia in patients with hema- 9. Przepiorka D, Weisdorf D, Martin P, et al. Intensive Care Med. 2019;34(5):404-410. tologic malignancies. J Cancer Res Clin 1994 Consensus conference on acute 14. Asleh R, Prasad M, Briasoulis A, et al. Uric Oncol. 2017;143(5):759-771. haematologica | 2020; 105(7) 1983