Comparison of efficacies of haploidentical transplantation and matched sibling donor transplantation in treating T‐cell lymphoblastic lymphoma

Abstract Objective To investigate the differences in efficacy and safety between haploidentical donor hematopoietic stem cell transplantation (HID‐HSCT) and matched sibling donor HSCT (MSD‐HSCT) in patients with T‐cell lymphoblastic lymphoma (T‐LBL). Methods In this retrospective analysis, we enrolled 38 patients who had undergone allogeneic HSCT at our institution between 2013 and 2021. The study participants included 28 patients who underwent HID‐HSCT and 10 patients who underwent MSD‐HSCT. We compared the patient characteristics and treatment effectiveness and safety between the two groups and evaluated potential prognostic variables for patients with T‐LBL. Results The median follow‐up durations in the HID‐HSCT and MSD‐HSCT groups were 23.5 (range: 4–111) and 28.5 (range: 13–56) months, respectively. All patients showed full‐donor chimerism after hematopoietic stem cell transplantation (HSCT). Except for two patients in the HID‐HSCT cohort who developed poor graft function, all patients showed neutrophil and platelet engraftments after HSCT. The cumulative incidences of grades III–IV acute graft‐versus‐host disease were 37.5% and 28.57% in the HID‐HSCT and MSD‐HSCT groups, respectively (p = 0.84). The cumulative incidences of limited (34.13% vs. 28.57%, p = 0.82) and extensive (31.22% vs. 37.50%, p = 0.53) chronic graft‐versus‐host disease did not differ between the two cohorts. In the HID‐HSCT and MSD‐HSCT cohorts, the estimated 2‐year overall survival rates were 70.3% (95% confidence interval [CI]: 54.9%–90.0%) and 56.2% (95% CI: 31.6%–100%), respectively (p = 1.00), and the estimated 2‐year progression‐free survival (PFS) rates were 48.5% (95% CI: 32.8%–71.6%) and 48.0% (95% CI: 24.6%–93.8%), respectively (p = 0.94). Furthermore, the Cox proportional‐hazards model showed that a positive positron emission tomography/computed tomography (PET/CT) status before HSCT in patients who had completed chemotherapy was an independent risk factor for PFS in the multivariate analysis (p = 0.0367). Conclusion This study showed that HID‐HSCT had comparable effectiveness and safety to MSD‐HSCT in treating T‐LBL. HID‐HSCT could serve as an alternate treatment option for T‐LBL in patients without an eligible identical donor. Achievement of the PET/CT‐negative status before HSCT may contribute to better survival.


| INTRODUCTION
T-cell lymphoblastic lymphoma (T-LBL) is an extremely aggressive type of non-Hodgkin's lymphoma. Its incidence rate is 0.1 per 100,000 adolescents and young adults and 0.4 per 100,000 children <15 years. 1 T-LBL and T-cell acute lymphoblastic leukemia (T-ALL) have similar morphologic and immunologic features and have been called "T lymphoblastic lymphoma/leukemia (T-LBL/ALL)" by the World Health Organization (WHO) since 2008. 2 The key distinction between them is the degree of bone marrow involvement: T-LBL, <25% (or 20% according to WHO); and T-ALL, ≥25%. 3 The prognosis of adults with T-LBL is poor, with a low complete remission (CR) rate and overall survival (OS) ranging from 11 to 17 months. [4][5][6][7] Currently, T-LBL has no standard treatment. Although high CR rates may be achieved with ALL-like therapy and the modified Berlin-Frankfurt-Münster-90 (BFM-90) regimen, approximately 30% of patients develop relapse. [8][9][10] Hematopoietic stem cell transplantation (HSCT) may improve the prognosis of patients with T-LBL; however, whether allogeneic HSCT (allo-HSCT) or autologous HSCT (auto-HSCT) is more favorable remains debatable. Allo-HSCT may yield similar survival outcomes as auto-HSCT. 11,12 Allo-HSCT is associated with a relatively lower relapse rate but a higher transplantation-relative mortality (TRM). While the safety of auto-HSCT is better, the relapse rate is higher. A conclusion regarding which one is better for T-LBL has not been established. Allo-HSCT has been shown to reduce the risk of relapse in T-cell non-Hodgkin lymphoma. 13,14 However, only 30% of patients have human leukocyte antigen (HLA)-identical sibling donors, 15 most patients need alternative donors. Currently, haploidentical donor HSCT (HID-HSCT) is widely used to treat hematological malignancies and has been undergoing rapid advancements in China. 16 HID-HSCT and matched sibling donor (MSD-HSCT) have comparable therapeutic effects in treating ALL. [17][18][19] However, although HID-HSCT treated T-LBL with good outcomes, no studies on the comparison between HID-HSCT and MSD-HSCT have been performed. 20,21 Their advantages and disadvantages are yet unknown. Therefore, the aim of this study was to compare the effectiveness and safety of MSD-HSCT and HID-HSCT and investigate prognostic factors for patients with T-LBL who undergo HSCT.

| Study population
This was a single-center, retrospective analysis of 38 patients with T-LBL who had undergone allo-HSCT between May 2013 and September 2021. T-LBL was diagnosed according to the 2016 criteria by WHO. 22 The examinations mainly included bone marrow biopsy and morphologic assessment, flow cytometry of bone marrow aspirates, immunohistochemistry, and positron emission tomography/computed tomography (PET/CT). All patients underwent HSCT after showing CR, partial remission, or stable disease with the BFM-90 or hyper-CVAD chemotherapy regimen. The study protocol was approved by the ethics committee of the Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. All patients provided informed consent. The study was performed in accordance with the tenets of the Declaration of Helsinki.
typing for the DRB1 and DQB1 loci. A high-resolution molecular approach was used to verify the validities of HLA-A, HLA-B, HLA-C, HLA-DRB1, and HLA-DQB1.

| Stem cell harvest
All donors were subcutaneously administered with 8-10 μg/kg of recombinant human granulocyte colonystimulating factor (rhG-CSF) daily to mobilize stem cells from 3 days before HSCT to the last day of HSCT. Basing on the target count for CD34 + cells set at above 3 × 10 6 / kg of the recipient weight, rhG-CSF was applied for an average of 5 days. rhG-CSF-primed peripheral blood stem cells and/or bone marrow stem cells were collected on the fourth to fifth day after rhG-CSF administration and transfused to the recipients on the same day without manipulation. A one-day extension of mobilization and transfusion period would be required if the count of CD34 + cells did not reach the target level.

| Regimen-related toxicities
Regimen-related toxicities were assessed according to the Bearman toxicity criteria. 24 Organ damage resulting from GVHD and/or infection was excluded from the evaluation.

| Therapeutic evaluation and study definition
Eligible patients had undergone at least three PET/ CT examinations: at diagnosis and during and after chemotherapy (before HSCT). Response was assessed as CR, partial remission, stable disease, or progressive disease according to the 2014 Lugano Classification. 25 The Deauville scale was used to determine the PETnegative (Deauville ≤3)/PETpositive (Deauville >3) status. Hematopoietic reconstitution and chimerism were evaluated as previously described. 26 ABO mismatch was classified into three types, depending on the transfusion strategy. 27 Primary poor graft function (PGF) was defined as Sun et al. described earlier. 28 The diagnosis and the grade of acute GVHD (aGVHD) and chronic GVHD (cGVHD) were assessed according to the Glucksberg-Seattle criteria and the National Institutes of Health consensus. [29][30][31][32] Progression-free survival (PFS) was defined as the duration between diagnosis and relapse, death from any cause, or the final follow-up day. OS was defined as the duration between diagnosis and death from any cause or the last follow-up day. Death without relapse or progressive disease after allo-HSCT was defined as TRM. OS, PFS, and cumulative incidence of relapse (CIR) were the primary endpoints of the study. TRM and cumulative incidence of GVHD were the secondary endpoints.

| Statistical analysis
Descriptive statistics were performed for demographic and clinical patient characteristics. Continuous variables were compared using the Mann-Whitney U test. Classification variables were assessed using the chi-square or Fisher's exact test. The Kaplan-Meier estimator with the log-rank test was used to estimate PFS and OS. Potential factors affecting OS and PFS were explored using univariate and multivariate analyses of the Cox proportional-hazards model. Cumulative incidence curves with Fine & Gray model were used to estimate the probabilities of TRM, GVHD, and relapse. TRM was considered as a competing risk of relapse. Relapse and death without GVHD were competing events for aGVHD and cGVHD. All p-values were two-sided. A p-value <0.05 was considered to indicate statistical significance. All statistical analyses were performed using SPSS version 25.0 (SPSS Inc.) and R Studio (version 1.4.1106). Table 1 shows detailed patient information. The median follow-up durations in the HID-HSCT and MSD-HSCT groups were 23.5 (range: 4-111) and 28.5 (range: 13-56) months, respectively. Table 2 shows the clinical outcomes of HSCT.

| Hematopoietic reconstitution
All patients showed neutrophil engraftment. The median time (11 vs. 11 days, p = 0.832) of neutrophil engraftment did not differ between HID-HSCT and MSD-HSCT groups. Except for two patients who developed PGF in the HID-HSCT cohort, all patients showed platelet engraftment. The rate (92.9% vs. 100.0%, p = 1.000) and the median time (11 vs. 10 days, p = 0.166) of platelet engraftment did not differ significantly between the two groups. Except for two patients who died from TRM, all patients showed full-donor chimerism 30 days after HSCT.

| Infection and regimen-related toxicities
The incidence of cytomegalovirus (CMV) reactivation was higher in the HID-HSCT group than in the MSD-HSCT group but without statistical significance (32.1% vs. 20.0%, p = 0.690). The incidence of Epstein-Barr virus reactivation was higher in the MSD-HSCT group than in the HID-HSCT group but without statistical significance (17.9% vs. 20.0%, p = 1.000). Ten patients in the HID-HSCT group and two patients in the MSD-HSCT group had pneumonia within 3 months after HSCT (35.7% vs. 20%, p = 0.453). Twelve and four patients in the HID-HSCT and MSD-HSCT groups had mild oropharyngeal mucositis, respectively (42.9% vs. 40%, p = 1.000). Moderate hemorrhagic cystitis only occurred in the HID-HSCT group (10.7% vs. 0%, p = 0.552). No case of severe toxicity caused by regimens was found.   Figure 2B), respectively. Since MSD-HSCT group had a smaller number of patients, the TRM incidence was slightly higher than that of HID-HSCT group. In a randomized multicenter study comparing MSD-HSCT with HID-HSCT in adults with Ph-ALL in CR1, the 3-year cumulative TRM were 13% and 11% (p = 0.84) in respective group. 17 According to a retrospective study of allo-HSCT outcomes in adults with standard-risk ALL in CR1, the 5-year cumulative TRM were 16.4% and 11.6% for HID-HSCT and MSD-HSCT, respectively. 18 Our results were comparable with parallel studies. Eleven (39.29%) patients in the HID-HSCT group and four (40%) patients in the MSD-HSCT group died, with causes including infection (four cases in HID-HSCT and two cases in MSD-HSCT), respiratory/circulatory failure (four cases in HID-HSCT and one case in MSD-HSCT), and relapse (three cases in HID-HSCT and one case in MSD-HSCT). The estimated 2-year OS rates was 70.3% (95% CI: 54.9%-90.0%) in the HID-HSCT group and 56.2% (95% CI: 31.6%-100%) in the MSD-HSCT group (p = 1.00, Figure 3A). The estimated 2-year PFS rates were 48.5% (95% CI: 32.8%-71.6%) and 48.0% (95% CI: 24.6%-93.8%) in the HID-HSCT and MSD-HSCT groups, respectively, showing no significant difference (p = 0.94, Figure 3B), with a median PFS of 18 months after diagnosis.

| Univariate and multivariate analyses
In order to identify potential influencing factors, basic information and clinical outcome information in Tables 1  and 2 were included in the Cox analysis. Continuous variables were converted into categorical variables by using the median as the dividing line such as age, PLT engraftment time and neutrophil engraftment time. The univariate analysis showed that B symptoms, Eastern Cooperative Oncology Group score 2, International Prognostic Index score 3-4, mediastinal mass at diagnosis, PGF, cGVHD, and relapse affected OS in patients with T-LBL who underwent allo-HSCT. Nevertheless, these variables did not differ significantly in the multivariate analysis. Risk factors significantly affecting PFS included B symptoms, cGVHD, PGF, platelet engraftment time >11 days, and a PET/CT-positive status before HSCT in the univariate analysis. However, only the PET/CT-positive status before HSCT (p = 0.0367, HR = 2.99) was an independent risk factor for PFS in the multivariate analysis (Table 4).

| DISCUSSION
The prognosis of T-LBL is poor. Although no standard therapy for T-LBL exists, chemotherapy combined with transplantation provides survival advantages and, occasionally, improves the prognosis. 12,33,34 Among transplantation approaches, allo-HSCT lowers the risk of relapse significantly greater than auto-HSCT, improving PFS and survival. 11,35 HID-HSCT is occasionally used to treat T-LBL/ALL, and studies on selection criteria and comparison of HID-HSCT and MSD-HSCT are limited. This retrospective study offered novel evidence of their effectiveness and safety being comparable. Patients with T-LBL treated with allo-HSCT have a wide range of survival results, which may be attributed to differences in patient characteristics. In a multicenter study, PFS of 75% at 3 years, not significantly differing from auto-HSCT. 12 Another multicenter study enrolled 21 patients with T-LBL who received HSCT (16 of them underwent allo-HSCT), the 5-year OS were 56.3% in the whole cohort. 36 A retrospective analysis showed an OS of 70% and a PFS of 65% at 3 years in 25 patients with T-LBL who underwent T-cell-replete HID-HSCT. 20 An American study involving 28 patients who underwent HSCT, including auto-HSCT and allo-HSCT, showed an OS of 49% and a PFS of 46% at 3 years. 37 In a single-center study, the 3-year OS and PFS were both 56.2% of 16 patients undergoing allo-HSCT. 38 In this study, the survival outcome showed similar estimated 2-year OS (70.3% vs. 56.2%, p = 1.000) and PFS (48.5% vs. 48%, p = 0.94) and the same median PFS time (18 months). Our results were consistent with those of previous studies. These two procedures are also similar in safety. In this study, hemorrhagic cystitis occurred only in the HID-HSCT group, because of the conditioning regimens, but all of the cases were mild. CMV and Epstein-Barr virus reactivation rates did not differ between the two groups; however, the HID-HSCT group had a slightly greater CMV activation rate compared to the MSD-HSCT group. Atibordee et al. summarized several risk factors for CMV infection in transplantation recipients, such as ATG, ALG, high-dose MMF, and other immunosuppressive drugs. 39 Compared to the conditioning regimen and prevention of GVHD in patients with MSD-HSCT, immunosuppressive strategies for patients with HID-HSCT are stronger.
In clinical practice, care should be taken to monitor the virus reactivation status of patients, and antiviral medication should be administered promptly if virus reactivation occurs.
The results of hematopoietic reconstitution were statistically equivalent in the two groups. Two patients in the HID-HSCT group experienced PGF, one of whom died from severe infection with respiratory failure 33 days after HSCT. This patient had received <5 × 10 6 /kg of CD34 + infusion. Risk factors for primary PGF include a low dosage of CD34 + cell infusion, donor-specific antibodies, GVHD, CMV infection, iron overload, and splenomegaly. 40,41 Patients with PGF have worse OS and PFS. 28,[41][42][43] In some clinical trials of ALL, HID-HSCT showed marginally higher TRM or non-relapse mortality but comparable and even slightly superior PFS or disease-free survival compared to MSD-HSCT. 17,18 In our multivariate analysis, PGF was not independent risk factor. Although a limited  number of patients developing PGF or TRM did not affect the overall statistical outcome of patients, these events should be paid attention to during transplantation to reduce the risk of complications.
We identified multiple univariate variables associated with the prognosis in the Cox analysis of OS and PFS (Tables 3 and 4). A PET/CT-positive status before HSCT was an independent hazard factor for PFS (multivariate pvalue = 0.0367; hazard ratio = 2.99). This finding suggests that patients with T-LBL who attain a PET/CT-negative status before transplantation could experience better outcomes ( Figure 4).
However, the role of PET/CT in T-LBL is debatable. Dai et al. discovered that PET/CT after allo-HSCT was a powerful predictor of PFS but not of OS in 67 patients with LBL. 44 Sun et al. demonstrated that PET/CT before HSCT could not predict PFS or OS but that the post-HSCT PET/CT-positive group had a greater risk compared to the PET/CT-negative group among 21 consecutive patients with relapsed or refractory T-LBL who underwent HSCT. 45 This paradoxical outcome may be attributable to differing time points of PET/CT evaluation before HSCT. We performed PET/CT assessment after the completion of chemotherapy, while Sun et al. performed PET/CT assessments 2-10 weeks before transplantation, thereby possibly including patients who were still on chemotherapy without achieving remission. Due to a paucity of relevant research, the importance of PET/CT before and after   In summary, we demonstrated the effectiveness of allo-HSCT in treating T-LBL. HID-HSCT is a viable alternative that can yield equivalent survival outcomes compared to MSD-HSCT. Achievement of a PET/CT-negative status before allo-HSCT may contribute to better survival of patients with T-LBL. A bigger, prospective study is required to further validate our findings.

Characteristics
T A B L E 4 Univariate and multivariate analysis of progression-free survival in allo-HSCT.