Introduction

Since first performed in 1981, allogeneic stem cell transplantation (allo-SCT) is the only easily available curative treatment for transfusion dependent thalassemia (TDT) until the gene therapy could bring a paradigm shift [1]. Risk stratification of TDT classifies them into three groups (Lucarelli Classes I, II, and III) [2]. Lucarelli et al. demonstrated that thalassemia-free survival (TFS) was 94% in Class I, 77% in class II and 53% in class III [3, 4]. Subsequently, identification of factors impacting outcome by Mathews et al., led to the introduction of very high-risk subset of Class III i.e. Vellore high risk (HR) or Class IIIB [5, 6].

With identification of class IIIB, treosulfan (Treo) was introduced in the armamentarium because of its low hepatic toxicity profile [7, 8]. Various studies have reported Treo based conditioning in HR with TRM between 7 to 21% [9,10,11]. Conversely, intravenous Busulfan with therapeutic drug monitoring have been utilized to improve survival [12]. However, there is a need to address challenges in context to significant TRM. In addition, the importance of pre-allo-SCT preparation cannot be ignored and a clear strategy needs to be implemented [13, 14]. Henceforth, we are reporting our experience of allo-SCT for TDT conditioned with Flu/Bu/Cy/ rabbit ATG regimen.

Materials and Methods

All consecutive allo-SCT for TDT, with HLA matched sibling (MSD) or matched unrelated donor (MUD) at our center between October 2018 and April 2022 were included. This is a retrospective analysis of data collected from hospital records. The study was approved by the institution review board.

Conditioning Regimen and Graft Versus Host Disease (GVHD) Prophylaxis

All patients received iv Fludarabine @30 mg/sqm on day-18 to -13, rabbit ATG (brand name: thymoglobuline) @4.5 mg/kg over three days (day-12 to day-10), oral busulfan @14 mg/kg for boys and 12 mg/kg for girls on day-9 to day-6, cyclophosphamide @50 mg/kg/day for girls and 40 mg/kg/day for boys on day-5 to day-2. GVHD prophylaxis included oral cyclosporine and iv methotraxate @10 mg/sq on day + 1 and 7 mg/sqm on day + 3, and + 6.

Pretransplant Downstaging of Risk Class

All patients received adequate chelation and hydroxyurea pretransplant. Hydroxyurea was started at dose of 15–20 mg/kg/day with dose titration according to absolute neutrophil count (ANC). Liver and spleen size were documented every month. Chelation dose was optimized and double or triple chelation was done if required.

Bone Marrow Harvest

Bone marrow (BM) was used as stem cell source in all MSD. In case of MUD, we had received graft as BM for one patient and as peripheral blood stem cell (PBSC) for another two patients. BM was harvested under general anesthesia from the iliac crest and the target cell dose was 6–8 × 108 nucleated cells/kg body weight of the recipient. Harvest was collected in bags containing ACD (Acid-citrate-dextrose).

Bone Marrow Graft in Blood Group Mismatch

Recipients with ABO major or bidirectional mismatch and corresponding anti-A or anti-B titer of 1:32 or higher received aliquots of donor type red cell transfusion. We usually started with 5 ml on day 1 of Bu, then gradually increase it to 10–15 ml of day 2, 40–60 ml on day 3 and 4 while monitoring for any severe reaction. All patients received oral chlorpheniramine and acetaminophen as premedication. Patients also received iv hydration with forced diuresis. Patients were monitored for clinical symptoms and mild reactions were tolerated. Anti-A or anti-B titers were measured on day-5 or -4. For recipients with persistently elevated anti-A or anti-B titer of 1:32 or higher, an additional infusion of donor-type blood was given on day-1. If the patient was able to tolerate 50–60 ml of donor-type blood without clinically significant hemolytic reaction and the titers on day-1 were less than 1:32, BM was infused without manipulation on the day of transplant.

Definitions

Definition and severity of Veno-occlusive disease (VOD) was according to EBMT criteria, which included two of the following criteria: unexplained refractory thrombocytopenia, weight gain > 5%, hepatomegaly above baseline, rising bilirubin above baseline on 3 consecutive days [15]. Toxicity was graded according to Eastern Cooperative Oncology Group (ECOG) common toxicity criteria [16]. Hypertension was categorized using standard criteria [17]. Acute GVHD according to standard criteria and chronic GVHD according to National Institutes of Health guidelines [18, 19]. Transplant Associated Thrombotic microangiopathy (TA-TMA) was diagnosed by criteria by Jodele et al. [20]. Engraftment was defined as the first of three consecutive days with an ANC > 0.5 × 109/L and platelet count > 20 × 109/L unsupported for seven days. Mixed chimerism (MC) was defined as donor cells < 95%. MC was categorized in three levels according to percentage of donor cells (level I: 90% to 95%; level II:75% to 89%; and level III: 5% to 74%) [21] Graft failure was defined as either the absence of hematopoietic reconstitution of donor origin on day + 28 (primary graft rejection), or as confirmed loss of donor cells after transient engraftment, with a return to transfusion dependence (secondary graft rejection) [21].

Chimerism Analysis

Analysis of chimerism i.e. quantification of recipient and donor percentage was done by polymerase chain reaction amplification of informative Short Tandem Repeats (STR) in cases of same sex transplant. In case of sex mismatched transplants, fluorescein in situ hybridization (FISH) for XY chromosomes was performed. In CC, chimerism was assessed at day + 28, + 60, + 90, + 6 months and + 1 year. In MC, monitoring was done more frequently.

Supportive Care

All patients were transplanted in a positive pressure HEPA (high efficiency particulate air) filtered transplant unit. Prophylactic acyclovir was started on day-2 till one year and beyond if patient had GVHD and required additional immuno-suppression. Antifungal prophylaxis with fluconazole was administered until engraftment of neutrophils. Trimethoprim-sulphamethoxazole, and oral penicillin prophylaxis was initiated after stable engraftment. All patients who were off all immunosuppression and had no evidence of GVHD were vaccinated.

Statistical Analysis

Primary study endpoints were the incidences of graft rejection and TRM. Secondary endpoints were incidences of VOD, acute and chronic GVHD, MC and TA-TMA and other transplant-related complications. Incidences of rejection, TRM and GVHD were calculated using cumulative incidence estimates. For overall survival (OS) and TFS, Kaplan Meier curve was prepared by using SPSS version 21.

Results

Clinical Characteristics (Table 1)

Table 1 Baseline characteristics of patients
Full size table

Fifty-five patients {Males = 32(58.1), Females = 23(41.8%)} with TDT underwent allo-SCT {MSD (n = 52) or MUD (n = 3)} with Flu/Bu/Cy/rATG conditioning between October 2018 and April 2022. Median age was 7 (2–13) years. Median pretransplant ferritin level was 1775(234–3170) ng/ml. Median donor age was 7(2–33) years. Forty (72.7%) transplants had same blood group. Eight (14.5%) transplants had major, 5(9%) had minor and 2(3.6%) had bidirectional ABO mismatch. Sixteen (29%) transplants had no sex mismatch. Twenty-six (47.2%) transplants were from female to male and 13(23.6%) transplants were from male to female. Source of stem cells was granulocyte stimulating factor (G-CSF) mobilized BM in 53(96.4%) and PBSC in 2(3.6%) patients. CMV status was both donor and recipient reactive in 49(89%), donor reactive and recipient non-reactive in 1(1.8%) and donor non-reactive and recipient reactive in 5(9%) transplants.

Downstaging of Risk Class

On presentation, 18(32.7%) patients were in Lucarelli class I, 14(25.4%) patients were in class II and 23(41.8%) patients were in class III {Class IIIA = 14(25.4%), IIIB(Vellore HR) = 9(16.3%)}. After downstaging, there were 22(40%), 15(27.2%), and 18(32.7%) patients in Lucarelli Classes I, II, and III respectively. Out of class III patients, 15(27.2%) were in class IIIA and 3(5.4%) patients were in class IIIB (Vellore HR).

Cell Dose and Engraftment

Mean stem cell dose was 3.28(1.2–6.5) × 106/kg. Neutrophils and platelets engrafted at a median of 16 (12–32) and 17 (12–48) days respectively. Stable hemoglobin of more than 10 gm/dl without transfusion was achieved at the median of 17.5 (12–81) days.

Chimerism

Forty-five (81.8%) patients had CC whereas 10 (18.1%) patients had MC on day + 28. There was no primary graft rejection. Subsequently, 7 (12.7%) additional patients developed MC (secondary). Therefore, by day + 100, 17 (30.9%) patients had MC. Five (29.4%) patients had level I, 9(52.9%) had level II and 3 (17.6%) had level III. Median time of onset of MC was 31 (21–101) days. One (5.8%) patient had secondary graft rejection. For 6 (35.2%) patients, immunosuppression was weaned, whereas in 11 (64.7%) patients immunosuppression was augmented by adding mycophenolate mofetil. Patient who had secondary graft rejection was in immunosuppression tapering group. All patients with MC had persistent MC with stable hemoglobin level more than 9gm/dl without transfusion support.

Peri and Post-Transplant Complications (Table 2)

Table 2 Transplant related complications of entire cohort and Class IIIB patients
Full size table

Forty-five (81.8%) patients developed oral and/or abdominal mucositis {Grade1/2 = 34(75.5%), Grade 3 = 11(24.4%)}. Median day of onset of mucositis was 5(5–9) days. VOD was seen 12(21.8%) patients, out of which 9 (75%) had mild, 2(16.6%) had moderate and 1(8.3%) had severe VOD. Patient with severe VOD was amongst class IIIA. Median day of onset of VOD was 13 (6–19) days. All patients were managed with diuretics and fluid restriction protocol. Defibrotide was administered in 3(25%) patients. TA-TMA was observed in 3 (5.4%) patients. Hemorrhagic cystitis was observed in 3(5.4%) patients{Cyclophosphamide = 2, BK virus = 2}. One (1.8%) patient had subclinical seizure, detected on EEG performed for severe headache, due to busulfan toxicity.

Seventeen (30.9%) patients had CMV reactivation. Four (23.5%) patients required empirical treatment, neither had repeated reactivations nor the CMV disease. Fourteen (25.4%) patients had documented infections {bacterial = 11, COVID-19 = 1, fungal = 2}. Patients with bacterial infection had blood culture positive for Escherichia. coli (n = 3), Staphylococcus aureus (n = 3), Methicillin Resistant Staphylococcus aureus (n = 1), Salmonella typhi (n = 1), Carbepenam Resistant Klebsiella (n = 1) and Pseudomonas(n = 1). Patients with fungal infection had abdominal mucormycosis (n = 1) and fungal sinusitis(n = 1).

Autoimmune cytopenia was seen in 3(5.4%) patients. Hypertension secondary to cyclosporine was observed in 26 (47.2%) patients. Other complications noticed were pericardial effusion (n = 1) and cyclosporine induced pancreatitis (n = 3). There was no TRM in our cohort.

Graft Versus Host Disease

Four (7.2%) patients developed aGVHD{grade1/2 = 3(75%), grade 3 = 1(25%)}. Common sites were skin, followed by GIT. All patients (except one) had steroid responsive GVHD. cGVHD was observed in four (7.2%) patients{mild = 3(75%); moderate = 1(25%); severe = 0}. Common sites involved were skin, oral mucosa followed by liver. None of the patient died with GVHD.

Survival

Median duration of follow up was 20.7(1.8- 43.9) months. All the patients were alive and transfusion free (except the one who had secondary graft rejection). Median duration of OS and TFS was 20.7 (1.8–43.9) months and 19.5 (1.8- 43.9) months respectively. The 3-year estimated OS and TFS were 100% ± 0% and 98.1% + 2% respectively. (Fig. 1).

Fig. 1
figure 1

Thalassemia Free Survival of entire cohort is 98.1% ± 0%

Full size image

Discussion

After pioneering experience of thalassemia transplants with BuCy conditioning by Lucarelli et al., allo-SCT has become the most widely available curative option [2]. For patients in class I, three-year OS, event-free survival (EFS) and relapse were 94, 94, and 0% respectively whereas for class II patients the probabilities were 80, 77, and 9%. The worst outcome was for class III patients with probabilities of 61, 53, and 16% [2]. Subsequently, in an attempt to reduce TRM, dose of Cy was lowered to 160 mg/kg for HR, however this was associated with increased graft rejection [3]. To decrease rejection rate, fludarabine was introduced in addition to BuCy. Moreover, adequate chelation and hyper transfusion during preparation has led to significant improvement in TFS [22].

In 2007, Mathew et al. had reported 5-year EFS and OS in Class III as 70.3% and 78.3% respectively whereas in Class III HR, these were 24% and 39% respectively [5]. This led to many innovations such as the introduction of Treo and experience with Treo/Thiotepa/Fludarabine showed that OS and EFS was significantly better among HR (86 vs. 39% and 77 vs 32%) [11]. On the contrary, another Indian study concluded that Treo/Thio/ Flu demonstrated a trend towards higher TRM [10]. In a contribution to optimize BuCy conditioning, Faulkner et al. studied ATG/Bu/Cy versus Thio/Bu/Cy and showed that adding Thiotepa in BuCy regimen is effective with less TRM [23].

In our study, 15(27.2%) patients were in class IIIA and 3(5.4%) were in class IIIB. VOD has been reported to be as high as 78% with BuCy conditioning in class IIIB [11]. To overcome this risk, we attempted to downstage our patients pre-transplant from Class IIIB to lower class which led to lower rates of moderate to severe VOD (5.4%). Patients with HR are also more likely to have primary rejections. In a study of 179 patients transplanted with BuCy-based regimen, 17(9.5%) experienced graft failure [6]. We did not encounter any primary graft rejection. In addition, there has been a concern of MC when BM is used as graft, reported to be as high as 50% [11]. We experienced MC in 17(30.9%) patients. Additionally, it has been reported that early MC has high chances of secondary graft rejections, as close to 50% [24]. On the contrary, we encountered secondary graft rejection in only one patient.

In the above-mentioned study [6], aGVHD and cGVHD was 38% and 13% respectively with BuCy based regimen. In study by Choudhary et al., the incidence of aGVHD and cGVHD was 14.3% and 7% respectively in Treo/Thio/Flu compared with 16.6% and 8.3% in Bu/Cy/ATG [10]. In study by Mathew et al., patients in Treo/Thio/Flu regimen had Grade 2–4 aGVHD in 3 (23%), along with cGVHD in 12% cases [11]. We observed aGVHD and cGVHD in 7.2% and 7.2% patients respectively. Most likely, the use of BM graft reduced GVHD incidence in our study. Furthermore, we demonstrated safe and inexpensive way for the infusion of ABO mismatch (major and bidirectional) marrow grafts, attributing to our protocol of donor type blood infusion during conditioning [25].

TRM by Bernado et al. has been reported as 7% in Treo/Thio/Flu [9]. In contrast, Mathew et al. has compared TRM in class III HR and found it to be higher in BuCy (46% vs 13%, p =  < 0.005) [11]. However, we had no TRM, even in HR. Similarly, OS and TFS has been greatly impacted by risk stratification. Significant improvement in survival has been achieved over the decades, however, class III HR patients are still a concern. In a study by Sabloff et al., OS and TFS was 91% and 88% in class II and 64% and 62% in class III respectively with BuCy based conditioning [6]. Twenty years experience of Pesaro has elicited OS and TFS of 89.2% and 85.7% with BuCy based conditioning regimen [26]. Lawson et al. shared experience with BuCy with OS and TFS of 94.5% and 81.8% respectively [27]. In 2005, an Indian study by Chandy et al. reported OS of 68% with BuCy conditioning [7]. In Treo/Thio/Flu group, Mathew et al. had shown OS and TFS of 86.6% and 77.8% in HR group [11]. European Society for Blood and Bone Marrow Transplantation Hemoglobinopathy Registry had reported OS and EFS of 91% and 83% respectively [28]. Similarly, Yesilipek et al. had reported OS and TFS of 85 and 68% respectively [29], and Iravani et al. had showed OS and TFS of 80%, and 65% [30]. In our cohort, we could achieve 100% OS, and 98.1% TFS. (Table 3).

Table 3 Literature on stem cell transplant in patients with thalassemia major
Full size table

Conclusion

We were able to utilize conventional regimen of Flu/Bu/Cy/rATG, with dose modifications in comparison to standard Bu/Cy regimen [3] in TDT transplants even in HR with no TRM and primary graft rejection. Moreover, as we used BM as graft, our GVHD rates were very low. Our data supports that Flu/Bu/Cy/rATG still holds the promise of being an effective and well tolerated regimen even in twenty-first century. Also, it highlights the importance of pretransplant downstaging in improving the outcomes.