Portal de Periódicos da CAPES

UK experience of unrelated cord blood transplantation in paediatric patients

2016; Wiley; Volume: 172; Issue: 3 Linguagem: Inglês

10.1111/bjh.13914

1365-2141

Paul Veys, Robert Danby, Ajay Vora, Mary Slatter, Robert Wynn, Sarah Lawson, Colin G. Steward, Brenda Gibson, Mike Potter, Josu de la Fuente, Geoff Shenton, Jacqueline Cornish, Andrew R. Gennery, John A. Snowden, Denise Bonney, Mark Velangi, Annalisa Ruggeri, Éliane Gluckman, Rachael Hough, Vanderson Rocha,

Neonatal Respiratory Health Research

Cord blood is a valuable alternative source of cells for haematopoietic stem cell transplantation when suitable human leucocyte antigen (HLA)-matched sibling or unrelated donors are unavailable. The use of unrelated cord blood transplantation (UCBT) has been widely adopted in North America (Kurtzberg et al, 2008), but uptake in the UK had been slow. Following recommendations for incorporating UCBT into standard transplant practice within the UK (Shaw et al, 2009), the use of UCBT in children has increased significantly. British Society of Blood and Marrow Transplantation and Eurocord data revealed that 335 children (0·1–17·9 years) underwent UCBT in the UK between 1998 and 2012 (Table 1). The median follow-up of survivors was 4·4 years (0·4–14·6). Amongst patients with malignant disease (MD) (n = 167), the median age was 6·5 years (0·3–17·7), most had acute leukaemia (n = 128) and the majority were cytomegalovirus (CMV) seronegative (n = 93). Most patients received myeloablative conditioning (n = 138), with a single cord blood unit (CBU) (n = 138) matched for 5/6 HLA antigens (n = 82). The median collected total nucleated cell (TNC) and CD34+ cell dose was 6·7 × 107/kg (range 1·4–62·3) and 2·7 × 105/kg (range 0·1–41·7), respectively. Amongst patients with non-malignant diseases (NMD), the median age was younger at 1 year (range 0·1–17·9), and the most frequent diagnosis was primary immune deficiency (PID) (n = 74). In contrast to MD, the majority of patients received reduced intensity conditioning (n = 92), a better HLA-matched CBU (6/6 n = 77; 5/6 n = 65) and higher TNC and CD34+ cell doses: 14·4 × 107/kg (2·9–65·1) and 5·3 × 105/kg (0·5–59·7). Engraftment was achieved in 149/167 (89%) patients with MD and 136/162 (84%) with NMD. The median time to neutrophil recovery >0·5 × 109/l was 23 days (7–77) and 19 days (9–80) in patients with MD NMD, respectively. Platelet recovery was also slower in MD patients at a median of 40 days (8–151) versus 33 days (13–152), presumably reflecting the lower TNC and CD34+ dose. The cumulative incidence of neutrophil and platelet recovery by day 100 was 90·9% [95% confidence interval (CI): 85·2–94·4] and 75·8% (95% CI: 68·1–82·0) for MD and 84·4% (95% CI: 77·8–89·3) and 72·9% (95% CI: 65·1–79·2) for NMD patients. The cumulative incidence of graft-versus-host disease (GvHD), non-relapse mortality and relapse are given in Appendix S1. Overall survival (OS) at 1 and 4 years was 64·6% (95% CI: 57·6–72·3) and 51·8% (95% CI: 44·4–60·4) for the group of patients with MD (Fig 1a). Univariate analysis (Table 1) showed improved OS for patients aged median (Fig 1b), and CBUs matched for 6/6 HLA antigens (Fig 1c). There was a trend to improved OS with patient CMV-negative serology and a single CBU. There was significantly worse survival in patients with lymphoproliferative disease. Surprisingly, there was no impact of disease status on outcome, so acute myeloid leukaemia (AML) and acute lymphoblastic leukaemia (ALL) patients were analysed separately. There remained no impact of disease status on outcome for AML patients (P = 0·52), but for ALL there was a trend to improved survival in CR1 (P = 0·07). For the NMD group, OS was 85·2% (95% CI: 79·9–90·8) at 100 days, 75·2% (95% CI: 68·8–82·1) at 1 year and 74·5% (68·1–81·6) at 4 years (Fig 1d). Univariate analysis showed a trend to improved OS with 6/6 HLA-matched CBU (Fig 1f), a significantly poorer outcome with a collected TNC dose in the lowest quartile (Fig 1e), and a trend to poorer outcome in haemophagocytic lymphohistiocytosis (HLH) (Table 1). Multivariate analyses are given in Appendix S2. This report provides a comprehensive picture of the use of UCBT in the UK paediatric population from 1998 to 2012, reflecting approximately 16% of all unrelated transplants performed during this time. Although similar numbers MD and NMD patients were transplanted throughout this period, use of UCBT significantly increased after 2009 in both groups. Despite the relative reluctance to use UCBT in the UK, the outcomes have been comparable to UCBT performed elsewhere. Eapen et al (2007) examined the outcomes of 503 children with leukaemia undergoing UCBT in the USA, determining 5-year probabilities of leukaemia-free survival of 60%, 45% and 33% after 6/6, 5/6 and 4/6 HLA-matched CBT, respectively. The equivalent figures from this UK study were similar, at 75%, 48% and 33%, respectively (Table 1). Treatment-related mortality (TRM) and relapse rates for leukaemia patients undergoing UCBT were also broadly similar between the two series: 37·8% and 23·5% respectively in the USA; 24% and 26·5%, respectively in the UK. Eurocord examined the outcome of 279 children with NMD undergoing UCBT (Rocha & Gluckman, 2009). There was better engraftment in this UK study compared to the Eurocord study, with a cumulative incidence of neutrophil and platelet recovery of 84% and 73% versus 69% and 50%, reflecting a larger proportion of children with marrow failure, lower cell doses infused and greater HLA-mismatching in the Eurocord cohort. Acute GvHD (II–IV) and chronic GvHD rates were very similar at 32% and 24% for Eurocord and 31% and 17% for the UK cohort. The 49% OS at 100 days in the Eurocord cohort was influenced by cell dose and the number of HLA mismatches. The improved survival of 85% in the UK cohort therefore reflects the significantly higher cell dose (median x2) as well as the use of better HLA-matched CBUs. Focusing on the largest group of patients, PID (n = 76), the 4-year OS was 81%, which compares favourably to previously reported outcomes of UCBT in PID [57% ± 6%] (Fernandes et al, 2012) and is equivalent, if not superior, to that of other alternative donor sources in the same disease groups (Gennery et al, 2010). This study of UCBT in the UK confirms the excellent outcome of 6/6 HLA-matched CBT for MD (Eapen et al, 2007), the lack of benefit of double CBT in the paediatric population (Wagner et al, 2014), and good outcome of UCBT in metabolic diseases (Boelens et al, 2013). Novel findings include excellent outcome of UCBT in PID, good outcome of myeloablative UCBT in patients with advanced AML and poor outcome of UCBT in patients with histiocytic diseases; these latter two require confirmation in a larger group of patients. This study found no apparent impact of anti-thymocyte globulin (ATG) in either the MD or NMD cohorts but perhaps survival could be further increased in both groups by individualized dosing and timing of ATG (Lindemans et al, 2014; Admiraal et al, 2015). We would like to acknowledge the contributions made by the following as clinical directors and/or treating physicians (in alphabetical order): P Amrolia, H Campbell, R Chiesa, M Cummins, M Ethell, A Ewins, P Hiwarkar, B James, G Lucchini, A Pagliuca, D Marks, K Patrick, F Pinto, K Rao, N Russell, J Silva, R Skinner, I Thakur. Robert Danby analysed the data. Paul Veys, Robert Danby, Vanderson Rocha wrote the paper. Ajay Vora, Mary Slatter, Robert Wynn, Sarah Lawson, Colin Steward, Brenda Gibson, Mike Potter, Josu de la Fuente, Geoff Shenton, Jacqueline Cornish, Andrew Gennery, John A Snowden, Denise Bonney, Mark Velangi, Annalisa Ruggeri, Eliane Gluckman, Rachael Hough contributed data and revised the manuscript critically. Appendix S1. The cumulative incidence of acute graft-versus-host disease (GvHD) (II–IV) at day 100 was 39% (95% CI: 31.2–46.6) and 30.5% (95% CI: 23.4–37.9) for malignant disease (MD) and non-malignant disease (NMD), respectively. The cumulative incidence of chronic GvHD in patients surviving beyond day 100 was 21.8% (95% CI: 15.1–29.4) for MD and 17.1% (95% CI 11.3–24.0) for NMD. Appendix S2. For patients with malignant disease, multivariate analysis showed TNC dose collected >median [Hazard Ratio (HR) 0.59 (95% CI: 0.35–0.98), P = 0.04] and HLA mismatch to be of prognostic significance: 1 mismatch [HR 2.20 (95% CI: 1.02–4.76); P = 0.04] 2 mismatches [HR 3.32 (95% CI: 1.44–7.65) P = 0.004]. For patients with non-malignant disease, multivariate analysis revealed a significantly worse outcome with TNC dose in the first quartile (HR 0.43 (95% CI: 0.21–0.87), P = 0.02) and a trend to worse outcome for HLH [HR 3.25 (95% CI: 0.8–13.2), P = 0.09]. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.