Partially Matched Related Donor Transplantation Can Achieve Outcomes Comparable with Unrelated Donor Transplantation for Patients with Hematologic Malignancies

Hematopoietic stem cell transplantations (HSCT) from human leukocyte antigen (HLA)-identical siblings offer the best results for hematologic diseases. Unfortunately, for patients without a HLA-matched sibling, the bone marrow or umbilical cord blood from an unrelated donor (1–4) is not always an option, due to unsuccessful donor searches or insufficient amounts of cord blood cells (5–7). One alternative is the use of hematopoietic stem cells from a HLA partially matched family donor (PMRD; refs. 8–12). In recent years, much progress has been made in PMRD HSCT, especially without the use of in vitro T cell depletion. Quick engraftment, a relatively low incidence of graft-versus-host disease (GVHD), and a good graft-versus-leukemia (GVL) effect can be achieved without graft failure (13, 14). Thus far, no clinical comparison of HSCT from PMRD and unrelated donors (URD), without in vitro T cell depletion, has been done. The purpose of this prospective, nonrandomized, single-center study was to comparatively analyze transplantation outcomes in a consecutive series of patients who underwent HSCT without T cell depletion from either PMRDs or URDs at our institute.

Patient eligibility. Consecutive patients (N = 297) receiving HSCT for hematologic malignancies from either URDs (n = 78) or PMRDs (n = 219) from January 2004 to December 2007 were enrolled. Eight of the 78 patients from the URD cohort and 39 of the 219 patients from the PMRD cohort were reported in 2006 (4, 13). A total of 108 of 139 acute leukemia patients and 44 of 61 chronic myeloid leukemia patients from the PMRD group were reported in 2008 (15, 16). These patients were enrolled and further followed in this study. The protocols were approved by the institutional review board at Peking University Institute of Hematology, and all patients and their donors signed consent forms. The characteristics of the patients and donors are summarized in Table 1.

Patients were classified as standard risk if they were in first or second complete remission (CR1 or CR2) without high-risk cytogenetics (17), such as t(4;11) or t(9;22), or if they had chronic-phase chronic myelogenous leukemia (CML) or myelodysplastic syndrome. Patients in CR3 or beyond, nonremission, or CR1 with high-risk cytogenetics, or those with CML beyond the chronic phase were classified as high risk.

Pretransplantation cytomegalovirus (CMV) serologic analysis showed that low-risk [recipient (R)−, donor (D)−], intermediate-risk (R−, D+), and high-risk (R+) patients for CMV reactivity after HCT were 6.2%, 7.0%, and 86.8%, respectively, in the PMRD group, and 4.5%, 4.5%, and 91.0%, respectively, in the URD group, with no significant difference between the two groups (P = 0.35).

Donor selection. Patients without a suitable closely HLA-matched unrelated donor, namely, with >8 of 10 matching HLA-A, -B, -C, -DR, and DQ loci, and 5 of 6 or 6 of 6 matching HLA-A, -B, and DR loci, were eligible for PMRD SCT if a matched sibling donor was unavailable as a first treatment option or if there was insufficient time for an unrelated donor search due to disease status.

Transplant procedure. Conditioning therapy consisted of cytarabine i.v. on days −10 to −9; busulfan (4 mg/kg/d) orally on days −8 to −6; cyclophosphamide (1.8 g/m²/d) i.v. on days −5 to −4; Me-CCNU (250 mg/m²) orally once on day −3; and ATG (thymoglobuline, 2.5 mg/kg/d; Sang Stat) i.v. on days −5 to −2. There was only one disparity between the two groups: patients in the PMRD group received cytarabine at 4 g/m²/d, whereas patients in the URD group received a lower dose of cytarabine (2 g/m²/d) on days −10 to −9. All patients received cyclosporine A, mycophenolate mofetil, and short-term methotrexate for GVHD prophylaxis (13). There was only one disparity between the two groups: mycophenolate mofetil was discontinued on day 30 after unrelated HSCT, whereas in PMRD patients, mycophenolate mofetil was tapered from 1 g/d to 0.5 g/d on day 30 and was discontinued on day 60.

Donor lymphocyte infusion and imatinib. When hematologic or cytogenetic relapse was diagnosed after transplantation, the relapse was treated with a trial phase of immunosuppressant withdrawal, followed by therapeutic donor lymphocyte infusion (DLI). Acute leukemia patients whose blast count in the bone marrow at the time of post-HSCT relapse was >20% received prior chemotherapy (18). DLI was given 48 h after the last chemotherapy dose. For patients whose blast count was <20%, DLI was given without chemotherapy, after immunosuppression had been discontinued for a minimum of 2 wk and there was no evidence of active GVHD. G-CSF-primed peripheral blood stem cells (GPBSC) were used instead of steady donor lymphocyte harvests (19). The median number of CD34+ cells infused was 2.69 (range, 0.82-9.69) × 10⁶/kg. The median number of CD3+ cells infused was 2.09 (range, 0.84-5.6) × 10⁸/kg. In very high-risk patients (patients with acute leukemia in the nonremission state before HSCT), patients with mixed chimerism, and patients with signs of minimal residual disease (MRD), prophylactic DLI was given without prior chemotherapy (20). The median number of CD34+ cells infused was 0.63 (range, 0.05-2.4) × 10⁶/kg.The median number of CD3+ cells infused was 0.49 (range, 0.2-1.4) × 10⁸/kg. All patients received short-term immunosuppressive agents for 2 to 4 wk for the prevention of DLI-associated GVHD.

Imatinib mesylate was given to CML or Philadelphia chromosome–positive acute lymphoblastic leukemia (Ph+-ALL) patients who experienced a cytogenetic or hematologic relapse after HSCT and to patients who showed rising levels (1 log increase) of bcr/abl RNA transcripts by real-time quantitative-PCR.

Definitions and assessments. Neutrophil engraftment was defined as absolute neutrophil count of ≥0.5 × 10⁹/L for 3 consecutive days and platelet engraftment as ≥20 × 10⁹/L for 7 consecutive days without transfusion. Primary engraftment failure was defined as the absence of donor-derived myeloid cells at day 60 in patients surviving beyond day 28 after transplantation, or as the need for a second allogeneic transplant or reconstitution with autologous cells. Acute and chronic GVHD were defined according to published criteria (21, 22). Real-time quantitative PCR for detection of CMV-DNA was used to detect CMV antigenemia. Relapse was defined by morphologic evidence of disease in peripheral blood, marrow, or extramedullary sites, or by the recurrence and sustained presence of pretransplantation chromosomal abnormalities. Patients showing MRD (e.g., the presence of bcr/abl RNA transcripts by PCR) were not classified as having relapsed. Leukemia-free survival was defined as survival in continuous CR.

Statistical analysis. Cumulative incidences were estimated for engraftment, CMV infection, GVHD, nonrelapse mortality, and relapse, to account for competing risks. Associations between graft type and outcome were evaluated using Cox proportional hazard regression models. In addition to the hematopoietic stem cell source, the following variables were considered as covariates: recipient age, recipient and donor sex, degree of ABO matching, degree of HLA matching, disease type, pretransplantation duration of the underlying disease, disease status at transplantation (standard or high risk), and dose of nucleated cells infused. The multivariate models were built using a stepwise model selection approach. When groups were compared according to continuous covariates, Mann-Whitney U-tests were used. A χ² test was used to compare categorical covariates. The 95% confidence interval (95% CI) was calculated to compute the standard error. SAS version 8.2 (SAS Institute) and S Plus 2000 (Mathsoft) were used for all analyses. End points were calculated at last contact, the date of the latest follow-up being December 31, 2008.

Engraftment. Analyses of chimerism indicated that all patients surviving beyond day 28 achieved full donor chimerism by day 30 after HSCT. All patients engrafted to absolute neutrophil counts >0.5 × 10⁹/L at a median time of 13 days (range, 9-24 days) in PMRD versus 12 days (range, 9-20 days) in unrelated HSCTs (P = 0.052). One patient in the unrelated group had secondary graft failure at day 63 after transplantation. This patient died of severe infection at day 220 after transplantation. No patients in the PMRD group had secondary graft failure. A total of 205 (94%) and 76 (97%) patients achieved platelet engraftments in the PMRD and URD groups, respectively. These platelet engraftments occurred at 15 days (range, 7-140 days) and 13 days (range, 4-85 days) for the PMRD and URD patients, respectively (P = 0.012).

Secondary platelet engraftment failure occurred in eight patients, with four failures occurring in the PMRD group and four occurring in the unrelated group. Among these eight, three patients in the PMRD and one patient in the unrelated group eventually received additional donor GPBSCs for secondary platelet engraftment failure. All four of these patients recovered from thrombocytopenia.

GVHD. At day 100 after transplantation, the cumulative incidences of grade II to IV acute GVHD in the PMRD and unrelated cohorts were 47% (95% CI, 39-62%) and 31% (95% CI, 20-42%), respectively (P = 0.033; Fig. 1A), with a relative risk of 1.72 (95% CI, 1.01-2.94), P = 0.046 (Table 2).

A total of 203 patients in the PMRD cohort and 72 patients in the unrelated cohort survived longer than 100 days after HSCT and were eligible for evaluation of chronic GVHD. The 2-year cumulative incidences of chronic GVHD in the PMRD and unrelated cohorts were 54% (95% CI, 47-61%) versus 40% (95% CI, 31-52%; P = 0.17; Fig. 1B), with relative risk of 1.15 (95% CI, 0.71-1.86; P = 0.58; Table 2).

CMV infection. At day 100 after transplantation, the cumulative incidence of CMV antigenemia was 66% (95% CI, 61-71%) versus 55% (95% CI, 46-64%; P = 0.26) in the PMRD and URD cohorts, respectively. The cumulative incidence of CMV-associated IPn or enteritis was the same at 8.9% (95% CI, 6-12%) in the PMRD group versus 9.7% (95% CI, 4-15%) in the URD group (P = 0.90).

DLI, imatinib, and relapse. In the PMRD group, three patients received prophylactic imatinib mesylate for rising levels (1 log increase) of the bcr/abl transcript. Nineteen patients received prophylactic DLI. In the URD group, three patients received prophylactic imatinib mesylate, and no patients received prophylactic DLI for the prevention of leukemia relapse.

The 2-year incidences of relapse for the PMRD and unrelated HSCT groups were 12% (8-16%) and 18% (10-27%), respectively (P = 0.12, Fig. 2A), with relative risk of 0.44 (95% CI, 0.22-1.10; P = 0.08, Table 2). Significant differences were found for standard risk patients, with the 2-year probability of relapse being 8% (4-13%) versus 19% (15-23%, P = 0.033) in the PMRD and URD cohorts, respectively.

At last follow-up, 26 patients (12%) in the haplo-identical group and 14 patients (18%) in the unrelated group experienced leukemia relapse. Fourteen patients received therapeutic DLI, which included 13 patients from the PMRD cohort and 1 from the URD cohort. Seven patients received imatinib mesylate for treatment of cytogenetic or hematologic relapse, including two in the PMRD group and five in the unrelated group. At the time of last follow-up, 16 patients had died of relapse, including 12 in the PMRD group (4 treated with DLI) and 4 in the URD group, with a median time to death of 11 months (range, 3-17) and 5.4 months (range, 2-7) after HSCT, respectively.

Nonrelapse mortality. Two patients (0.9%) died within 28 days of PMRD HSCT due to lung embolism or heart failure. There was one early death (1.3%) at day 20 due to infection in the URD group. Analyses of nonrelapse mortality are described in Table 3. The 2-year incidences of nonrelapse mortality for the PMRD versus unrelated HCT groups were 20% (95% CI, 15-26%) versus 18% (95% CI, 10-28%), respectively (P = 0.98, Fig. 2B), with relative risk of 1.01 (95% CI, 0.51-2.01; P = 0.98; Table 2).

Long-term follow-up and survival. The 4-year overall survival and leukemia-free survival after PMRD and unrelated HCT were 74% (95% CI, 68-80%) versus 74% (95% CI, 62-86%), with P = 0.98, and 67% (59-75%) versus 61% (95% CI, 47-74%), with P = 0.74 (Fig. 3A to B), respectively. There were no differences in survival probabilities according to disease status before transplantation between the two cohorts (data not shown). For comparison between father and mother donors, 4-year overall survival were 69.5% and 64.9% (P = 0.81, data not shown). The 4-year overall survival for those patients >20 years versus ≤20 years after PMRD and URD HSCT were 73.4% versus 73.5% and 75.0% versus 75.2%, respectively (P = 0.88, data not shown). Multivariate analyses showed that a diagnosis of CML was associated with higher overall survival (Table 2).

Engraftment is considered an important element of success for HSCT. Ottinger et al. (23) reported an incidence of 5.2% graft failure for unmanipulated HLA-matched unrelated HSCT. A similar result was observed in our previous report, which showed 4.7% graft failure in 63 unrelated patients (4). In the present study, engraftment was achieved in 99.1% of PMRD patients and 98.7% of unrelated patients (100% for patients surviving beyond 28 days after HSCT). These results reveal that our current protocol is safe, for both PMRD and unrelated HSCT, in terms of engraftment.

The current results show that the incidence of acute GVHD associated with PMRD HSCT was significantly higher than that associated with unrelated HSCT. Nearly identical protocols for GVHD prophylaxis were applied in both patient groups. The results seem to suggest that it is difficult to overcome the HLA barrier; however, the higher incidence of acute GVHD in the PMRD group is partly due to the relatively low incidence in the unrelated cohort, where only 10% of patients experienced grade III-IV acute GVHD. Satoshi Takahashi et al. (24) reported an incidence of grade III-IV acute GVHD of 27% in unrelated HCT. It is notable that, in the current study, 88% of patients from the unrelated group received PBSCs as the graft source. Our own comparative study of PBSCT versus bone marrow transplantation from unrelated donors showed that the cumulative incidence of grades II-IV acute GVHD in the PBSCT group was significantly lower than that in the BMT group (33% versus 67%; ref. 4). Previous studies have revealed that GPBSCs contain more CD34+ cells, CD14+ cells, and Th2 cells than nonprimed lymphocyte harvests, which may contribute to the decreased incidence of acute GVHD (25, 26). The results also confirm that G-CSF plays an important role in inducing immune tolerance. Finally, in our current study, the use of prophylactic DLI in some PMRD patients may also have contributed to the higher GVHD incidence in the PMRD cohort.

PMRD HSCT may potentially exert a strong GVL effect, but no comparative study testing this has been done. It has been speculated that the difference in the GVL effect between PMRD and unrelated HCT might be more prominent in high-risk patients than in standard-risk patients. In the present study, however, the results contradicted this. Explanations for this may include: First, it is generally more difficult to overcome the refractory nature of high-risk leukemia. Second, the population of high-risk patients in the unrelated group was relatively small (18 patients), making it more difficult to show significant differences when compared with PMRD patients.

Additionally, our results may be partly explained by the relatively low incidence of relapse in the PMRD cohort. In our previous studies, the 2-year incidence of relapse was 12% versus 38% for standard- versus high-risk patients undergoing PMRD HSCT before February 2005.¹³ In the present study, the overall incidence of relapse was 18% versus 10%, regarding HSCT done before or after February 2005. Significant factors affecting relapse rate in our multivariate analysis such as diagnosis, disease status, and time from diagnosis to HSCT were comparable between the two cohorts regarding PMRD HSCT done before or after February 2005. Alternatively, shorter follow-up time (P < 0.01), and the use of prophylactic DLI and imatinib may have contributed to decreased relapse rates seen in the more recent (after 2005) cohort of patients treated with PMRD transplants. For HSCT done before or after February 2005, four versus fifteen patients in the PMRD group received prophylactic DLI. For CML patients, the successful use of imatinib (27, 28) produced encouraging results, even in PMRD HSCT settings, with 2-year incidences of relapse of 3.0% versus 7.1% for HSCT done before or after February 2005. The incidences were 11% versus 7.5% and 31% versus 18% for standard- and high-risk patients, respectively (data not shown). Compared with the unrelated group, the lower incidence of relapse in standard-risk patients after PMRD HSCT suggests that PMRD HSCT does have a strong GVL effect.

The data show that outcomes in both PMRD and URD group are quite good, which maybe is partly due to the relatively small proportion of patients >35 years or with poor risk. Nevertheless, considering the comparable survival between PMRD and URD cohorts, PMRD HSCT seems to be one of the ready choices at experienced transplant centers and may be superior in term of the ready availability of donor cells and the convenience of donor-derived cellular therapy when needed (29).

In conclusion, this study shows that, for every major HSCT end point, including relapse, nonrelapse mortality, and survival, PMRD and unrelated HSCTs are not significantly different. However, a large, multi-center, randomized trial is needed to confirm these results.

No potential conflicts of interest were disclosed.

We thank American Journal Experts Team for critically reviewing this paper in English.