It is well known that allogeneic hematopoietic stem cell transplantation (allo-HSCT) constitutes a major advance in the treatment of several malignant and non-malignant hematological pathologies [1–4]. However, donor selection remains a crucial challenge, particularly in the absence of an HLA-matched donor. Therefore, haploidentical transplantation has emerged in recent years as a promising alternative [5,6].The administration of high-dose cyclophosphamide within days of transplantation has proven its efficacy in selectively eliminating alloreactive T lymphocytes, responsible for adverse immune reactions, while sparing regulatory cells and hematopoietic stem cells. This approach has transformed haploidentical transplantation into a viable and accessible strategy for a wider patient population. However, several prognostic factors still influence the outcome of this procedure, notably the occurrence of graft-versus-host disease (GVHD). Understanding these factors will not only help improve patient and donor selection, but also refine therapeutic strategies, with the ultimate aim of optimizing the chances of success and minimizing transplant-related complications [7,8].

In this study, the influence of key factors on the development of acute and chronic GVHD were investigated in a cohort of Tunisian pediatric recipients undergoing haploidentical bone marrow allo-HSCT. Patient-specific variables (such as underlying diagnosis and donor-specific antibodies), donor characteristics (including age, sex, and weight), graft composition (mononuclear and CD34+ cell content, and ABO compatibility) and procedural aspects (particularly the conditioning regimen) were assessed with the goal of better understanding how these factors shape transplant outcomes and ultimately improve patient care.

Data, collected from the patient files, were processed using Microsoft Office Excel® 2016. Statistical analysis was carried out using MedCalc® 19.0.4 software. To better reflect the distribution of the data and facilitate their interpretation for statistical analysis, continuous variables were expressed as medians with interquartile ranges [Q1; Q3] and categorical variables were presented as counts (n) and percentage (%). Calculation of the median and the 25th and 75th quartiles was chosen to qualify graft volume and cellular richness of MNC and CD34+ cells. Univariate analysis of clinical and transplant-related variables was performed using contingency tables, with crude odds ratios (ORs) and 95% confidence intervals (95% CI) calculated to assess their association with the incidence of acute and chronic GVHD. To further identify independent predictors of GVHD, a multivariate logistic regression model was performed.

Forty-eight consecutive haploidentical bone marrow allo-HSCT were included in this study with a total of 47 patients (one patient was transplanted twice with two different donors). The characteristics of study population are presented in Table 1.

In this study, univariate analysis examining the occurrence of acute GVHD identified one statistically significant factor associated with an increased risk of acute GVHD: donor–recipient sex mismatch. Furthermore, in multivariate analysis, gender mismatch continues to be a statistically significant factor, while some factors did not reach statistical significance in the analysis. This may reflect either a limited role in the development of acute GVHD or constraints related to sample size. These findings do not rule out the potential relevance of some other factors and future studies with larger patient populations could provide more definitive insights.

Several studies have demonstrated that recipient minor histocompatibility antigens are targets of donor T cells, playing a central role in mediating both GVHD and the graft-versus-leukemia (GVL) effect. These minor H antigens originate from cellular proteins derived from polymorphic genes, which differ between donor and recipient. The corresponding peptides are then presented to CD8+ and CD4+ T lymphocytes via class 1 and 2 major histocompatibility complexes, respectively. During allo-HSCT from a female donor to a male recipient, donor T cells recognizing minor H antigens specific for the recipient Y chromosome genes may be present, resulting in both GVHD and GVL effects. The Y chromosome genes encoding these H antigens have already been identified and their roles in GVHD and GVL are currently under investigation. Bross et al. published the results of a study involving 136 patients. in which 71% of patients who received an opposite-sex donor transplant developed grade II-IV acute GVHD [12]. Similarly. Randolph et al. conducted a retrospective analysis of 3238 patients who underwent allo-HSCT with a genoidentical related donor. After receiving myeloablative conditioning a statistically significant OR was observed in male recipients who received female donor transplants for both acute grade II-IV GVHD and chronic extensive GVHD [13,14]. Since previous pregnancies or transfusions may influence the likelihood of donor-derived cells causing GVHD, Flowers et al. considered this variable in their study. Of the 136 genoidentical bone marrow donors, 30 were multiparous women, 30 nulliparous women and 76 were men. The cumulative incidence of grade II-IV acute GVHD was 57% for patients with multiparous female donors, 21% for those with nulliparous female donors and 46% for those with male donors. Multivariate analysis of the data confirmed that the risk of grade II IV acute GVHD was significantly higher in patients who received multiparous donor transplants compared with those who received nulliparous donor transplants (RR = 2.5; p = 0.02) [15].

The present study revealed no statistically significant factors associated with the development of chronic GVHD, likely reflecting limited sample size and statistical power. However, the complex and multifactorial nature of chronic GVHD likely also contributes to the lack of clear predictive patterns in this cohort.

Among the factors influencing the outcome of allo-HSCT, GVHD prophylaxis plays a key role particularly in preventing this dreaded complication. This field has seen many advances and remains at the heart of numerous scientific investigations and clinical trials exploring various protocols to optimize GVHD prevention and treatment. In the current study, GVHD prophylaxis was performed with the triad: CsA, MMF and post-transplant cyclophosphamide. In recent years, several strategies have been developed and refined to improve the results of haploidentical allo-HSCT, with a particular focus on reducing GVHD while maintaining a robust anti-leukemic effect for hematological malignancies. These innovative approaches include the use of post-transplant cyclophosphamide, as well as other methods, which continue to be the subject of in-depth study and clinical evaluation. The Baltimore group first described the use of post-transplant cyclophosphamide for haploidentical allogeneic HSCT. Luznik et al. proposed a novel approach based on the use of cyclophosphamide post-transplant, combined with non-myeloablative conditioning and the use of bone marrow as the cell source. Administration of high doses of cyclophosphamide equal to 50 mg/kg/day on the third and fourth days after haploidentical allo-HSCT enables selective elimination of alloreactive T cells, while preserving regulatory T cells and hematopoietic stem cells. Brunstein et al. and Luznik et al. demonstrated low rates of GVHD (6% for grade III-IV acute GVHD and 13% for moderate or severe chronic GVHD), as well as a non-relapse mortality rate of <15% [16,17]. The low toxicity of this strategy, combined with the non-myeloablative conditioning proposed by the Baltimore team, makes it a particularly attractive option for elderly patients, who are generally at greater risk of complications. Several studies have confirmed this low toxicity in this population, and single-center retrospective comparisons and registry data have revealed survival rates comparable to those of HLA-matched transplants [18–20]. However, due to the non-myeloablative nature of conditioning, this approach is limited by high relapse rates of up to 50% [16,17]. Several studies have shown that higher-intensity conditioning improves disease control, reducing relapse rates to between 20% and 40% [21–23]. Ultimately, there are many arguments in favor of this strategy: it is simple to implement, inexpensive and associated with low rates of graft failure, GVHD, infection and mortality not related to relapse [24]. Current research is focused on understanding the mechanisms of immune reconstitution after haploidentical allo-HSCT with high doses of post-transplant cyclophosphamide. Mastering these mechanisms would make it possible to exploit the beneficial alloreactivity linked to HLA incompatibility between donor and recipient. Several models also suggest that it is possible to specifically modulate the activity of immune cells, such as T or NK lymphocytes, to achieve an anti-tumor effect without the complications associated with GVHD. This approach redefines allogeneic immunotherapy. The establishment of a partially compatible allogeneic immune system, under conditions of reduced toxicity can provide the basis for targeted anti-tumor immune interventions. Currently, in vivo T-cell depletion strategies are tending to supplant those based on in vitro depletion, as evidenced by analysis of the European Society for Blood and Marrow Transplantation registry. The majority of centers specializing in haploidentical transplants have opted for the post-transplant cyclophosphamide approach, enabling the use of reduced-intensity conditioning and, consequently, the treatment of older patients [25,26].

Since 2010, the number of haploidentical HSC transplants has increased substantially worldwide gradually overtaking placental blood transplants and unrelated donor transplants with 9/10 compatibility, reversing the trend observed since 2012 in France [27]. In Europe, the use of haploidentical family donors is growing by around 25% each year [28]. Initially reserved for patients requiring allo-HSCT and for whom no compatible donor was available, this approach has rapidly become widespread. The encouraging results, the near-certainty of finding a donor within the family, the reduced cost and the shorter time it takes to identify a donor have all contributed to its growing popularity. Today, the place of this strategy within the therapeutic framework of allografts, as well as its position in the hierarchy of donor choices, is the subject of much debate within the scientific community.

In this cohort, donor–recipient sex mismatch was the main risk factor for acute GVHD, while no clear predictor was identified for chronic GVHD. A broader analysis of the patient cohort, together with an increase in sample size, will enable us to refine the predictive models and reinforce the robustness of the conclusions drawn. The aim will also be to better identify the impact of certain hitherto underestimated factors, and to continue exploring the complex interactions between different variables in order to optimize haploidentical transplant protocols. Continuing this work could open up new prospects for personalizing treatments and improving outcomes for transplant patients, while reducing the risk of complications.

Meriem Lameri: Writing – Original Draft, Methodology, Investigation, Data Curation, Formal analysis, Conceptualization, Visualization. Monia Ben Khaled: Methodology, Investigation, Resources. Eya Ghazouani: Data Curation, Formal analysis. Ilhem Ben Fraj: Methodology, Investigation, Resources. Mohamed Hichem Sellami: Formal analysis, Superviosion. Wafa Aissa: Investigation. Yesmine Boughzala: Investigation. Houda Kaabi: Resources, Investigation. Slama Hmida: Resources, Supervision. Monia Ouederni: Conceptualization, Methodology, Investigation, Resources, Writing – Review & Editing, Supervision. Manel Chaabane: Conceptualization, Methodology, Writing – Review & Editing, Resources, Project administration, Validation, Supervision.

There are no ethical issues.

This study did not receive any specific grant from funding agencies.

The data that support the findings of this study are available from the corresponding author upon reasonable request.