Risk class approach for patient and protocol selection

In the 1980s the influence of pre-transplant characteristics on the outcome of HSCT was analysed in 161 patients less than 17 years of age (Storb et al., 1977; Lucarelli et al., 1990). Multivariate analysis showed that (i) hepatomegaly of more than 2 cm below the costal margin, (ii) portal fibrosis and (iii) irregular chelation history were associated with a significantly reduced probability of survival.

On the basis of these risk factors, a prognostic scheme was developed by the Pesaro group: patients were categorised into three risk classes influencing the probability of survival.

Table 1

Expected 5 y probability of overall survival (OS) and thalassemia-free survival (TFS) after HSCT in β thalassaemia major.

Pre-transplant evaluation

Particular attention has to be paid to an appropriate pre-transplant work-up. In addition to classical pre-HSCT evaluations, this should include accurate iron studies based on magnetic resonance imaging (MRI) analysis to evaluate liver iron load and cardiac MRI to evaluate cardiac iron load and function. The key to reducing graft rejection is to reduce erythroid expansion prior to HSCT.

HSCT from HLA-matched sibling donors (MSD)

In the past 30 years more than 2,000 thalassaemia patients have undergone HLA-matched related donor HSCT, predominantly in the transplant centres of Pesaro and Rome (Lucarelli & Gaziev, 2008; Isgrò et al., 2010).

In a large EBMT survey of 1061 cases of MSD HSCT performed between 2000 and 2010 in 132 centres in 28 countries with a median patient age of 7 years, long-term overall survival (OS) and thalassaemia-free survival (TFS) were 91% and 83%, respectively (Baronciani et al., 2016).

Conditioning regimen for class 1 and class 2 patients

Preparatory regimens for HSCT must achieve two objectives: elimination of the disordered marrow and establishment of a tolerant environment that will permit the transplanted marrow to survive and thrive. There is considerable evidence and decades of experience for the use of busulfan (BU) and its derivatives for ablating marrow in patients undergoing HSCT for the treatment of non-malignant conditions (Parkman et al., 1978; Kapoor et al., 1981; Hobbs et al., 1986). Cyclophosphamide (CY) is a well-established agent for providing adequate immunosuppression for allogeneic engraftment (Thomas et al., 1972; Storb et al., 1991). The combination of 14–16 mg/kg BU and 200 mg/kg CY can eradicate thalassaemia and facilitate sustained allogeneic engraftment. Treosulfan-based conditioning has been introduced as a well-tolerated alternative (Mathews et al., 2013) and should be analysed in comparison to BU-based conditioning in a future randomised trial. When the allogeneic graft starts to proliferate in the recipient, an immunological reaction against the recipient (GvHD) may occur. Therefore, the prophylactic administration of ciclosporin (an immunosuppressive drug) is an important part of the pre-transplant and post-transplant treatment. Three doses of methotrexate (MTX) may also be given in addition to ciclosporin during the first 15 days after transplant.

Conditioning regimen for class 3 patients

To further decrease graft rejection in class 3 patients, the initial conditioning regimen was modified by the Pesaro group with the addition of thiotepa (TT) – a drug with potent myelosuppressive and immunosuppressive potential, but mild extra-haematological toxicity. This improvement in the treatment protocol resulted in a 5-year OS and TFS probability of 92% (95% CI, 77-97%) compared to 85% OS (95% CI, 64-94%) and 73% TFS (95% CI, 51-86%) achieved with the previous protocol (Gaziev et al., 2016).

HSCT from HLA-matched unrelated donors

A major obstacle to successful transplantation is the limited number of HLA-matched related donors within families. Approximately 60% of patients lack a suitable sibling donor. Some of these patients could benefit from HSCT from a matched unrelated donor (MUD). A number of studies with a few hundred patients transplanted worldwide have shown that MUD can cure a large proportion of patients with thalassaemia, provided that the donor is selected using high-resolution molecular typing for both HLA class I and II molecules. The risk of rejection can be reduced by selecting unrelated donors who do not have non-permissive mismatches at the HLA-DPB1 locus in the host-versus-graft direction (Fleischhauer et al., 2006). For example, a study of MUD HSCT in 68 patients with thalassaemia major who re-ceived BU/CY or BU/FLU (fludarabine) and/or thiotepa (TT) as a conditioning regimen re-ported 96.7% OS and 80% TFS for class 1 and 2 patients, whereas class 3 patients only had a OS of 65.5% and a TFS of 54.5% (La Nasa et al., 2005a). Outcome is affected by the risk clas-ses; so that in the group of 30 thalassaemic patients in risk classes 1 and 2, the probability of OS and DFS were 96.7% (CI 90-100%) and 80.0% (CI 65-94%), respectively, whereas in the 38 patients in class 3 OS was 65.2% (CI 49-80%) and DFS was 54.5% (CI 38-70%). In an-other study that included 21 children who received MUD transplants: the 2-year TFS was 71% compared with 82% for patients who received MSD transplants (Hongeng et al., 2007).

Thus when donor selection is based on stringent compatibility criteria and appropriate selection of class risk category in the recipient, the results of unrelated transplantation in thalassemia patients are comparable to those obtained when the donor is a compatible sibling.

HSCT from phenotypically matched related donors

In 2005, the Mediterranean Institute of Hematology (IME), Rome, adopted a new transplant approach for related donor HSCT in thalassaemia. This study was a prospective, single-centre investigation of the safety and efficacy of a novel preparative regimen for HSCT in patients with thalassaemia from related donors, who were not HLA-matched siblings.

Between 2005 and 2012, 16 patients with thalassaemia received their first HSCT from related donors who were phenotypically matched or 1-antigen–mismatched. The treatment protocol was based on BU/TT/CY/antithymocyte globulin (ATG) conditioning. Rejection incidence was 0% with a TFS probability of 94% and a transplant-related mortality of 6% (Gaziev et al., 2013).

Together, these data strongly suggest that improvements in donor selection and transplantation preparation improve the safety of unrelated and related donor HSCT for thalassaemia treatment (Gaziev et al., 2013).

Haploidentical HSCT

Haploidentical HSCT in principle may have the potential to extend the use of this treatment option to the 50%–60% of patients who lack a matched sibling donor or an HLA-identical unrelated donor. Initially high levels of graft failure of around 27% were seen (Sodani et al., 2010). A novel graft manipulation method that removes αβ+ T lymphocytes while retaining γδ+ T lymphocytes, natural killer (NK) cells, and other accessory cells was proposed (Chaleff et al., 2007) to reduce the graft failure rate and improve outcomes. Recent advances in graft engineering with effective ex vivo T-cell depletion through positive selection of CD34+ cells, depletion of CD3+/CD19+ cells or T-cell receptor αβ+ (TCRαβ+)/CD19+ depletion significantly improved the outcome (Aversa et al., 1998; Federmann et al., 2012; Diaz et al., 2016). Since June 2012, we have been using this new graft manipulation technique with selective depletion of TCRαβ+ and CD19+ lymphocytes. The use of TCRαβ+/CD19+-depleted grafts have been associated with significantly reduced graft failure (Gaziev et al., 2018). The haploidentical HSCT protocol was based on a BU/TT/CY/ATG conditioning regimen as previously published (Sodani et al. 2010). Patients received GvHD prophylaxis with cyclosporin and methylprednisolone or mycophenolate mofetil. The 5-year probabilities of OS and TFS were 84% and 69%, respectively (Gaziev et al. 2018). The incidence of graft failure was 14%. The incidence of grade 2 to 4 acute GvHD was 28% and 21% for extensive chronic GvHD.

While extensive T-cell depletion can significantly reduce the incidence of GvHD, it has been associated with delayed immune recovery and an increased risk of graft rejection, especially in non-malignant diseases (Willasch et al., 2006). Delayed immune reconstitution and associ-ated morbidity and mortality remain significant problems in this setting. Novel pharmacologic and cell therapy approaches to enhance immune reconstitution and improve outcome after T-cell–depleted haplo-HCT are warranted.

A recently developing alternative platform for haploidentical HSCT uses T-cell-replete grafts and post-transplant cyclophosphamide (PT-Cy). Bolanos-Meade et al. reported a high graft failure rate (46%) among 14 sickle cell disease (SCD) patients who underwent haplo-identical HSCT following nonmyeloablative conditioning regimen and PT-Cy (Luznik et al., 2008). An intensive preconditioning immunoablation followed by a myeloablative conditioning regimen and PT-Cy in the haploidentical setting for thalassaemia showed promising results with an OS and disease-free survival (DFS) of 95% and 94%, respectively (Anurathapan et al., 2016). In the most recent report (Anurathapan et al., 2020) of 83 consecutive transfusion-dependent patients with thalassaemia (median age, 12 years; range, 1 to 28 years) with a minimum follow-up of 6 months (median, 15 months; range, 7 to 53 months) the 3-year projected overall and event-free survival is over 96%, and there have been no secondary graft failures.

Mixed chimaerism following HSCT for thalassaemia

Mixed haematopoietic chimaerism (MC) is an interesting phenomenon that sometimes occurs after HSCT for thalassaemia. The incidence of MC in a study of 335 patients who received MSD HSCT for thalassaemia was 32.2% 2 months after transplantation (Lucarelli, Andreani & Angelucci, 2002). Of the 227 patients with complete donor chimaerism, none rejected their grafts, whereas graft loss occurred in 35 of 108 patients (32.4%) with MC, indicating that MC is a significant risk factor for graft rejection in thalassaemia patients. The percentage of residual host haematopoietic cells (RHCs) determined 2 months after transplant was predictive for graft rejection, with nearly all patients experiencing graft rejection when RHCs exceeded 25%. The risk of graft rejection was only 13% in patients with <10% RHCs and was 41% in patients with 10%–25% RHCs (Andreani et al., 1996, 2000). Of patients receiving HSCT for thalassaemia following myeloablative conditioning, 10% became persistent mixed chimaeras and became transfusion independent, suggesting that once donor-host tolerance is established, a limited number of engrafted donor cells might be sufficient to provide significant improvement of disease phenotype in patients with thalassaemia major.

Adult thalassaemia patients

Adult thalassaemia patients generally have more advanced disease, with both disease- and treatment-related organ complications that are mainly due to prolonged exposure to iron overload consequently, worse outcomes with transplanation have been seen in adults. From 1988 to 1996, 107 patients more than16 years of age received transplants from matched donors at the Pesaro Hospital. The median age for this population was 20 years, with a range of 17 to 35 years. The probabilities of overall survival, thalassaemia-free survival, rejection mortality and non-rejection mortality for this entire group were 66%, 62%, 4% and 37%, respectively (Lucarelli et al., 1992, 1999).

From April 1997, 15 high-risk adult patients were prepared for transplantation with a reduced total dose of 90 mg/kg cyclophosphamide. The probabilities of OS, TFS, rejection mortality and non-rejection mortality were 65%, 65%, 7% and 28%, respectively (Gaziev et al., 2005).

Encouraging results have been reported in adult thalassaemia patients who received bone marrow from matched unrelated donors, with rates of OS, TFS, transplant-related mortality (TRM) and rejection of 70%, 70%, 30% and 4%, respectively (La Nasa et al., 2005b).

Follow up after HSCT

Adequate post-transplant clinical follow-up is of particular importance. Within the first year, careful monitoring of haematological and engraftment parameters, infectious complications and GvHD is essential. Appropriate immunisation is necessary in the second year, if there is no GvHD. Long-term follow-up is of particular interest with respect to monitoring the evolution of multi-system problems (iron overload, pubertal development, growth and endocrine deficiencies) related to thalassaemia. A number of reports indicate that iron overload, chronic hepatitis, cardiac function and endocrine deficiencies can be managed more easily after transplantation, sometimes permitting the healing of severely damaged organs (Muretto, Angelucci & Lucarelli, 2002). It is also important to remove excess iron after transplantation by phlebotomy (venesection) or chelation therapy. All iron removal treatments should be started only once the graft is stabilised, and the patient free from any immunosuppressive treatment or prophylaxis, and in the absence of chronic GvHD. Endocrine dysfunction and infertility require specific expertise and follow up after HSCT.

Cost and cost effectiveness

Thalassaemia medical care is a complex, multidisciplinary and expensive process requiring dedicated and experienced units. An Italian study based on cost/benefit estimations from a societal perspective quantified tariffs, expenses and net earnings in 2006 for thalassaemia patients. The mean costs were €1242/patient/month, of which 55.5% was attributed to iron chelation therapy and 33.2% to transfusions (Scalone et al., 2008). These data compare to total overall median costs of HSCT from MSD of approximately 150,000 US $, which would translate to 1,900 US $ per expected life year after HSCT (Matthes-Martin et al., 2012). However, the cost of transplantation can vary significantly around the world. When considering the very significant combined costs of life-long blood transfusions, chelation and the management of complications for optimal thalassaemia care (which clearly exceed the healthcare resources available in most non-industrialised countries), transplantation is certainly a cost-effective option, if adequate expertise exists, even in developing countries.

Box

Summary and Recommendations.