Toward a definition of immunological risk for patients with anti-HLA antibodies before stem cell transplantation

In the setting of allogeneic hematopoietic stem cell transplantation (HSCT), graft failure (GF) is an unfrequent but life-threatening complication, carrying a high mortality rate^1,2. The occurrence of this event varies according to the donor type, ranging from less than 5% after HLA-matched and up to 20% after cord blood (CB) transplantation³, since the HLA disparities between patient and donor enhance the risk of engraftment failure. Poor graft function (PGF), a condition when one or more citopenias occur in the presence of full donor chimerism, is associated with higher morbidity and mortality as well as relevant transfusion support, similarly to GF.

In recent years, the presence of antibodies in patients' serum directed against HLA epitopes in their respective stem cell donors (i.e., donor-specific antibodies, DSAs) has emerged as one of the causes of GF or PGF after HSCT from HLA-mismatched donors, mainly CB or haploidentical family donors⁴. Although historically less considered with respect to cellular immunity, humoral immunity has proven to be responsible of severe complications after HSCT and now the screening for anti-HLA antibodies deserves to be part of pre-transplant patient work-up^5,6. Most knowledge on anti-HLA antibodies comes from the context of solid organ transplantation where they had been extensively studied and where HLA laboratories appropriately support the clinical programs.

Despite the great amount of clinical studies on DSAs in the context of HSCT, the overall immunological risk of engraftment failure (namely GF of PGF) of a defined patient-donor pair is still based solely on laboratory assays, either solid-phase or eventually cell-based, detecting the presence and providing the identification of anti-HLA antibodies, expressed as mean fluorescence intensity (MFI, a semi-quantitative measure of the amount of antibodies) and distinguishing complement fixing from non-complement fixing antibodies, through the C1q testing and the crossmatch (XM), this latter using both patient’s and donor’s samples⁵. There is consensus about the need for desensitization in the presence of DSAs with MFI values >5,000^5,7, whereas it is still matter of debate for lower amounts. When DSAs with MFI >20,000 are detected, desensitization strategies may not be successful⁸.

Nevertheless, the complexity of HSCT applies to a wide spectrum of hematological and non-hematological disorders, a wide range of conditioning regimen intensities (from the minimal non-myeloablative to the maximal myeloablative)⁹, graft-versus-host disease (GvHD) prophylaxis combinations, donor types and stem cell contents in the graft. Moreover, an accurate reproducibility of MFI measurement and XM techniques across HLA laboratories still lacks, although huge efforts are ongoing. For these reasons, MFI values and XM or C1q testing results might only partially account for the multifaceted, composite immunological risk of engraftment complications in patients with DSAs. As a consequence, a wider definition of this risk based on the above cited multiple elements would be worthy of use, in an attempt to go beyond the current determination of the immunological risk on a case-by-case basis⁷. Besides this, the desensitization treatment could enhance the relapse risk and/or the infectious burden¹⁰, by increasing the immunosuppression status of patients who are already immunosuppressed by prior therapies, transplant conditioning and GvHD prophylaxis. Ideally, desensitization should remove the dangerous DSAs without exceeding the risk of disease relapse or the infectious morbidity.

La Rocca et al. nicely present here such an approach¹¹, detailing a modulation of their desensitization strategy according to the expected immunological risk in a series of 126 HSCTs prospectively studied. Desensitizing treatment was well tolerated and all patients who were considered at risk fully engrafted; overall, these data support the efficacy of this approach and provide further evidence for the need of pre-transplant screening for DSA in patients undergoing HSCT. As authors stated, multicenter collaborative studies are warranted and one is ongoing¹², with awaited results.

Nonetheless, some studies showed that full engraftment can occur in DSA+ patients even in the absence of any desensitization^10,13–15, challenging the above cited findings or at least warranting a more comprehensive approach to the question. Also, a recent study on 303 unrelated HSCTs failed to show survival impairment among DSA+ patients¹⁶. One possible explanation is that, conversely to what happens after solid organ transplantation, the HSCT recipient’s immune system is replaced by that of the donor and this may prevent the antibodies to be harmful against the donor stem cells; this phenomenon is enhanced with the use of myeloablative conditioning regimens and with the infusion of adequate stem cell contents, both elements favouring engraftment. Yet, differences between patients exist depending on the immunosuppression status associated with the underlying disease, i.e., patients with acute leukemia or lymphoma vs myelodysplastic syndrome, this latter showing to be more frequently associated with the presence of DSA before HSCT¹⁷. Moreover, the patient-donor relationship matters when DSAs are detected, and the child-to-mother transplantation is paradigmatic for this perspective, where the mother (patient) usually presents with anti-HLA antibodies directed against paternal antigens in offspring (donor). Such an immunological profile is mostly associated with multiple DSAs and sometimes very high MFI, carrying an elevated risk of engraftment complications¹⁸.

All these elements support the inclusion of disease, conditioning regimen and GvHD prophylaxis, donor type (including stem cell source) and patient-donor relationship within the definition of the immunological risk of a specified patient-donor pair, together with the MFI values and XM or C1q testing results. That will be expected to account for the risk of engraftment complications and will drive clinicians towards the more appropriate desensitization strategy (Figure 1). Then, protocols should be modulated according to this risk, aiming at balancing the risk of engraftment complications with that of relapse or infectious complications^5,8,10,19,20.

Figure 1.

Indications for desensitization according to DSAs before allogeneic hematopoietic stem cell transplantation

MFI: mean fluorescence intensity; FC-XM : flow cytometric crossmatch; CDC-XM: complement-dependent cytotoxicity crossmatch.

Active collaboration between transplant physicians, immunogenetic and immunology experts will make it possible to define a comprehensive risk stratification, ref lecting the complexity of the transplantation setting and the multifaceted elements to be taken into account. By creating multidisciplinary working groups, designing multicenter collaborative studies and prospective interventional protocols, a shared desensitization policy across centers could be feasible, hopefully as homogeneous as possible in order to collect meaningful data and validate the most effective treatments.