Long-term survivors demonstrate superior quality of life after haploidentical stem cell transplantation to matched sibling donor transplantation

This prospective study enrolled consecutive patients (n = 425) who had received HSCT between January 2018 and December 2019 at the HSCT Center, Blood Disease Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (CAMS&PUMC). Participants were informed of the objective of the study and asked to complete QoL questionnaires including the SF-36 and the FACT-BMT (version 4). In the present study, 44 patients were excluded due to declination to participate or inability to complete questionnaires independently. A total of 25 patients dropped off during follow-up. The total response rate and follow-up rate reached 89.6% and 93.4% respectively. A number of 279 patients received MSD HSCT (n = 182) or HID HSCT (n = 97). Patients who survived 1 year after transplant (n = 230) were included for the analysis of QoL. A detailed study flow chart is shown in Fig. 1.

All participants signed written informed-consent forms and completed questionnaires online at their earliest convenience. All procedures were in accordance with the ethical standards of the institutional and national research committee (Ethics committee of Blood Disease Hospital, Chinese Academy of Medical Sciences + KT2013019001-EC-1) and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The institutional review board approved all study procedures and forms.

For acute myelogenous leukemia (AML) patients, the myeloablative conditioning regimen consists of fludarabine (30 mg/m²/d, days − 6 to − 4), busulfan (3.2 mg/kg/d, days − 9 to − 7), cyclophosphamide (40 mg/kg/d, days− 3 to − 2) and cytarabine (2 g/m²/d, days − 9 to − 7). For myelodysplastic syndrome/myeloproliferative neoplasm (MDS/MPN) patients, the 5-day decitabine-included regimen was used including decitabine (20 mg/m²/d, days − 9 to − 5), fludarabine (30 mg/m²/d, days − 6 to − 4), busulfan (3.2 mg/kg/d, days − 9 to − 7), cyclophosphamide (40 mg/kg/d, days − 3 to − 2) and cytarabine (2 g/m²/d, days  9 to − 7). For acute lymphocyte leukemia (ALL) patients, total body irradiation (TBI) based myeloablative conditioning was administered: TBI (3.3 Gy/d, days − 9 to − 7), cyclophosphamide (40 mg/kg/d, days − 6 to − 5), fludarabine (30 mg/m²/d, days − 4 to − 2) and cytarabine (2 g/m²/d, days − 4 to − 2). The conditioning regimen for aplastic anemia (AA) patients includes FAC and BFAC. The FAC conditioning regimen was composed of fludarabine (30 mg/m2/d, days − 5 to − 1), cyclophosphamide (30 mg/kg/d or 37.5 mg/kg/d, days − 5 to − 2) and rabbit antilymphocyte globulin (rATG) (Thymoglobulin^®, Genzyme, Cambridge, MA) (2.5 mg/kg/d, days − 5 to − 1) or porcine antilymphocyte globulin (pALG) (Anti-lymphocyte Immunoglobulin®, Wuhan Institute of Biological Products Co.,Ltd., China) (25 mg/kg/d, days − 5 to − 1). The BFAC conditioning regimen included busulfan (3.2 mg/kg/d, days − 7 to − 6) on the basis of FAC regimen. All haplo-identical recipients received rATG (2.5 mg/kg/day, days − 5 to − 2) or pALG (20 mg/kg/day, days − 5 to − 2).

The GVHD prophylaxis was cyclosporin (CSA) or tacrolimus (FK506) (from day − 1) based [17‐19]. Methotrexate was used for 3 doses (days + 1, + 3 and + 6) in MSD-HSCT and 4 doses (days + 1, + 3, + 6 and + 11) for HID HSCT. Mycophenolate mofetil was added for HID-HSCT recipients which initiated at 500 mg twice daily on day-9 and was titrated slowly downward until day + 60 and discontinued on day + 90. Details of conditioning regimen, GVHD prophylaxis and treatment strategy, supportive care were referred to our previously published reports [17‐19].

Social-demographic variables were recorded including personal code, age, gender, clinical diagnoses, marital status, financial burden, insurance, employment status, and insurance payment. Clinical variables were recorded in the clinical records and analyzed upon completion of the study including primary disease, chromosome karyotype, conditioning regimen, date of transplant, HSCT type, stem cell source, minimal residual disease (MRD) before HSCT, Eastern Cooperative Oncology Group performance score (ECOG score), Hematopoietic Cell Transplant-Comorbidity Index (HCT-CI) score and infection The assessment of acute and chronic GVHD was conducted at real time by clinical teams. Acute graft versus host disease (aGVHD) and chronic graft versus host disease (cGVHD) were evaluated using international criteria [20, 21].

Kaplan–Meier and Laplace regression models were used to estimate and compare the median survival time (months) and 95% confidence intervals (CIs) according to different HSCT types. Cox proportional hazards regression models were used to estimate the hazard ratios (HRs) and 95% CIs for the incidence of death according to HSCT type. Follow-up time (months) was calculated as the time from study entry to death or the final examination. The linear mixed-effects model estimating the β-coefficients and 95% confidence intervals (CIs) was used to analyze the variation trend of different dimensions of both SF-36 and FACT-BMT with different follow-up times. The fixed effect included HSCT, follow-up time (month), and their interaction. Stratified analyses were performed to explore the association between HSCT and QoL according to incident aGVHD and cytomegalovirus (CMV) reactivation during the follow-up. All models were adjusted for age, sex, education, body mass index (BMI), main caregivers, diagnosis, and transplantation type, history of aGVHD and, cGVHD, and history of infection.

This study included 279 participants (median age: 36, ranging from 15 to 66, 49.1% female) who received HID (n = 97) or MSD (n = 182) HSCT at our center (Table 1). The majority of underlying diseases are hematological malignancies (93.1%) and all patients reached full chimerism post-HSCT. More than half of the patients (66.25%) were fully active before HSCT (ECOG score 0). Most patients (93.54%) were evaluated HCT-CI score ranging from 0 ~ 1 at HSCT. A total of 80 patients didn’t achieve hematologic response or retained untreated before HSCT. Most patients (96.42%) accepted school education for more than 6 years and all patients were able to complete the QoL questionnaires independently. While 72.04% of patients were married, spouses are main caregivers for 54.8% of patients. HID-HSCT recipients received more potent immunosuppressive regimens for the prophylaxis of GVHD. More patients in HID group were taken care of by their spouses (54.64%). No significant difference was documented between MSD and HID-HSCT groups in terms of HCT-CI score, ECOG score, and hematologic responses before HSCT. As we have mentioned in the method session, HID-HSCT recipients received more potent immunosuppressive regimens for the prophylaxis of GVHD, including the addition of ATG in the conditioning regimen, extension use of CSA/FK506/MTX and the addition of mycophenolate mofetil after transplant. The basic characteristics and its comparison between HID and MSD were shown in Table 1.

We firstly compared the overall outcome of the HID-HSCT and MSD-HSCT recipients. After a median follow-up of 24.33 months (95% CI 23.6–25.1 months), overall survival (OS) was comparable which was 82.42% and 86.46% at 3 years in all recipients of HID and MSD HSCT respectively (Fig. 2). Consistently, mortality rate in HID-HSCT was comparable to MSD-HSCT with HR (95% CI) at 0.77 (0.37–1.61) in a multi-adjusted COX model. Despite the similarity in survival, significantly higher proportion of HID-HSCT recipients had a history of acute GVHD (57.1% vs 34.0%, p = 0.001) and CMV reactivation (33.8% vs 21.6%, p = 0.045) which may affect the QoL (Table 1).

Additional factors that may contribute to survival or QoL were subsequently analyzed. Unsurprisingly, HID-HSCT recipients were significantly younger than recipients of MSD-HSCT however the difference was very small (34.25 [11.703] vs 37.71 [11.438], mean [SD], p = 0.032), which we didn’t explore further. Of note, more than half of the HID patients were taken care of by their spouses (54.55%) whereas more patients in the MSD group were taken care of by alternative relatives. Other investigated factors were comparable between these two groups including gender, family income, body weight index (BMI), underlying diseases, performance status, school years, marital status, insurance sources, history of bacterial/fungal infections and history of chronic GVHD (Table 1).

We then investigated the longitudinal recovery of QoL in the whole cohort during the follow-up. We have described the QoL values for the HID-HSCT and MSD-HSCT groups at each time point. According to the results, scores of most dimensions increased gradually after transplantation. For instance, the mean score of RP has been raised from 12.82 to 23.33 in the MSD-HSCT group, and from 16.25 to 20.00 in the HID-HSCT group (Additional file 1: Table S1). We performed the linear mixed-effects model to analyze the variation trend of different dimensions with different follow-up times. Using the SF-36 form, scores on most scales increased gradually over time (Fig. 3, Table 2), including physical functioning (PF), role physical (RP), bodily pain (BP), general health (GH), vitality (VT), social functioning (SF), and role emotional (RE). Of note, statistically significant difference was observed in RP (P < 0.001), SF (P < 0.001) and PF (P = 0.011) indicating significant recovery mainly on physical function. However, we didn’t observe significant improvement in mental health (MH) score over time (β = − 0.07, 95% CI: − 0.2 to 0.08). Consistently, most of the scales of FACT-BMT form also improved over time (Fig. 4, Table 2). Similarly, physical well-being (PWB), the dimensions evaluating recovery of physical function, also reached clinically relevant significance. Thus, both forms demonstrated longitudinal improvement of QoL post-HSCT, especially in physical function related scales.

We next ask if there was difference between recipients of HID- and MSD-HSCT. These two cohorts reported similar scores of QoL before transplantation (Table 2). Compared to MSD-HSCT recipients, HID-HSCT recipients exhibited better performance in physical dimensions using FACT-BMT form including global FACT score (FACT-G), physical well-being (PWB), social well-being (SWB), functional well-being (FWB) and FACT-BMT Trial Outcome Index (TOI). Using SF-36 form, the HID-HSCT recipients demonstrated superior advantage in GH, and mental dimensions including VT and MH (Table 1). In sum, HID-HSCT recipients demonstrated higher performance of QoL after HSCT in terms of physical and mental scales and the difference was more remarkable using FACT-BMT Form.

As more HID-HSCT recipients had a history of acute GVHD/CMV reactivation, we conducted stratified analysis to evaluate the effect of these complications on QoL recovery. In patients without a history of aGVHD, HID-HSCT recipients showed significantly greater recovery in both physical and mental scales especially for GH, SF, VT and MH items (Additional file 1: Table S2). For patients with a history of aGVHD, HID-HSCT maintained the advantages albeit at smaller difference in VT and physical dimensions (GH and BP) while demonstrated inferior scores for SF without statistical significance (β = − 3.19, 95% CI − 13.36 to 6.98). When FACT-BMT Form was used, HID-HSCT recipients with a history of aGVHD also showed significantly higher FWB subscale score (β = 2.51, 95% CI: 0.19 to 4.85) whereas lost the superiority for the rest of domains (PWB, SWB, FWB, FACT-G, TOI and FACT-BMT).

For patients without a history of CMV reactivation, recipients of HID-HSCT had significantly better scores on the PWB, SWB, FWB, FACT-G, TOI, and FACT-BMT scales using FACT-BMT. However, all these advantages were lost when there was a history of CMV reactivation. When SF36 Form was used, HID-HSCT recipients without CMV reactivation history demonstrated a trend toward worse QoL recovery including PF, RE, SF and VT in the context of CMV reactivation which further confirmed the detrimental effect of CMV infection on QoL (Additional file 1: Table S3). In sum, HID-HSCT recipients demonstrated superior QoL performance, but the advantages were compromised by the incidence of aGVHD or CMV reactivation.

As we have shown, either SF-36 or FACT-BMT is competent to describe the longitudinal QoL recovery in long-term survivors of HSCT. However, for the comparison of QoL between HID and MSD-HSCT cohorts, these two evaluation systems exhibited diverse point of focus. SF36 depicted the advantage of HID-HSCT in GH and mental subscales including VT and MH. Otherwise, FACT-BMT showed the higher performance in physical dimensions including global FACT score (FACT-G), physical well-being (PWB), social well-being (SWB), functional well-being (FWB) and FACT-BMT Trial Outcome Index (TOI) in HID-HSCT recipients. The difference for GH and mental domain scores were greater using SF36 whereas superiority in the recovery for functional/physical domains was more profound when FACT-BMT was used. Furthermore, both forms are able to show the effect of aGVHD and CMV reactivation on QoL recovery despite difference in performance.