Long-Term Outcomes of TBI-Based Myeloablative SCT In 292 Consecutive Adults with Ph-Negative ALL: Similar Outcomes for Transplantation Using Related Donor, Well-Matched Unrelated Donor, or Partially Matched Unrelated Donor Sources

The graft-versus-leukemia effect in adult acute lymphoblastic leukemia (ALL) has now been definitely confirmed from the ‘matched related donor (RD) versus no donor' comparisons. However, no definite conclusions can be extracted from these data as to whether or not there is a survival advantage to RD-stem cell transplantation (SCT) over other therapeutic modalities for both high-risk and standard-risk patients with Philadelphia (Ph)-negative ALL. In addition, ‘RD versus no donor' approach is becoming outmoded, as in many studies those previously in a ‘no donor' category are now undergoing unrelated donor (URD)-SCT.

We report long-term outcomes of total body irradiation-based myeloablative SCT in 292 consecutive adults with Ph-negative ALL who received transplants at our center between 1995 and 2008 (median follow-up of survivors, 85 months). This study focused on following questions: (1) How different are the outcomes of SCT according to the donor sources? (2) Which factors are important to determination of transplantation outcome? (3) Which URD should be chosen? (4) Is there a role of autologous (AUTO)-SCT plus maintenance chemotherapy?

Median age was 25 years (range, 15–63 years). Overall, 237 (81.2%) of 292 patients had one or more high-risk features, including adverse cytogenetics [t(4;11), t(8;14), complex (>=5 abnormalities), Ho-Tr], older age (>=35 years), high leukocyte counts (>=30×10⁹/L for B-ALL, >=100×10⁹/L for T-ALL), or delayed first complete remission (CR1; >28 days). Two hundred and forty-one patients (82.5%) were transplanted in CR1; 22 (7.6%) in CR2; and 29 (9.9%) in advanced status. URD sources were classified as well-matched (WM), partially matched (PM), and mismatched (MM) based on a new proposed guideline from the NMDP-CIBMTR. Donor sources were RD (n=132), WM-URD (n=30), PM-URD (n=19), MM-URD (n=19), and AUTO (n=92). All patients and donors provided written informed consent, and the treatment protocol was approved by the institutional review board of The Catholic University of Korea.

Cumulative incidence of relapse at 5 years was 48.5% for AUTO versus 32.6% for RD, 19.4% for WM-URD, 32.3% for PM-URD, and 51.0% for MM-URD (RR, 2.70; 95% CI, 1.65 to 4.42; p<0.001). In multivariate analyses, other factors associated with higher relapse risk included transplantation in CR2 or later (p<0.001), T-lineage ALL (p=0.020), and adverse cytogenetics (p=0.038). Cumulative incidence of non-relapse mortality (NRM) at 5 years was 40.5% for MM-URD versus 19.6% for SD, 20.3% for WM-URD, 15.8% for PM-URD, and 9.8% for AUTO (RR, 3.09; 95% CI, 1.32 to 7.25; p=0.010). Patients older than 35 years had higher NRM (p=0.007). As a result, disease-free survival (DFS) at 5 years was inferior using AUTO (46.1%; RR, 1.69; 95% CI, 1.14 to 2.51; p=0.010) or MM-URD (26.3%; RR, 2.03; 95% CI, 1.05 to 3.95; p=0.036), compared to RD sources, while DFS from all other donor sources was approximately equivalent (53.5% for RD, 63.3% for WM-URD, and 57.0% for PM-URD). Transplantation in CR2 or later (p<0.001), older age (p=0.020), and adverse cytogenetics (p=0.041) were associated with poorer DFS. In a pairwise comparison of outcomes between RD-SCT and AUTO-SCT for patients in CR1, the inferiority of AUTO-SCT was observed, particularly in high-risk patients. Conversely, in standard-risk patients, AUTO-SCT yielded comparable outcomes to RD-SCT.

Our long-term data suggest that outcomes are similar for transplantation using SD, WM-URD, or PM-URD sources, and these may be considered the best donor sources for adults with Ph-negative ALL, especially for those with high-risk features.

No relevant conflicts of interest to declare.