Second Allogeneic Hematopoietic Stem Cell Transplantation for Post-Transplant Relapsed Acute Leukemia in Children - an EBMT PDWP Retrospective Study

Introduction

Using the EBMT registry, we retrospectively analyzed outcomes for 373 pediatric patients who underwent second allogeneic transplant for relapsed acute leukemia at 120 centers in 32 countries, between the years 2004 and 2013, in an attempt to assess relapse, survival, GVHD and other outcomes, as well as identify factors correlating with prognosis in this cohort of patients. To our knowledge, this is the largest analysis of pediatric patients undergoing second allogeneic HSCT for relapsed acute leukemia to date. This allowed for an independent analysis of each disease, including 214 patients with ALL and 159 with AML.

Patients and Methods

Centers received a questionnaire completing data already available in the ProMISe database on patients between 0-18 years of age treated between 2004 and 2013.

Results

A total of 387 patients received a second SCT after relapse. 373 have been included in the analysis, 214 for ALL and 159 for AML. Detailed data were available for 201 patients from 48 centers; for the remainder, analysis was based on the registry.

For the entire cohort overall survival (OS) at 2 and 5 years were 38% and 29%, and leukemia free survival (LFS) 30% and 25% respectively.

ALL:

With a median follow up from 2^nd SCT of 36.4 months, OS at 1 and 5 years were 47% and 28% respectively. LFS was 39% and 28% respectively. NRM at 2 years was 22%.

In multivariate analyses favorable prognostic factors for both OS and LFS were: CR prior to 2^nd SCT (p=0.0001), interval > 12 months between transplants (p=0.0007), use of myeloablative conditioning (p=0.039) and the presence of cGvHD after the first SCT (p=0.0001). Good prognostic factor for low NRM was interval of more than 12 months between transplants (p=0.0002).

AML:

With a median follow up from 2^nd SCT of 50 months, OS at 1 and 5 years were 44% and 15% respectively. LFS was 28% and 15% respectively. NRM at 2 years was 18%.

In multivariate analyses, favorable prognostic factors for OS as well as LFS were: CR prior to 2^nd SCT (p=0.031;0.044 respectively), interval > 6 months between transplants (p=0.0003;0.0001 respectively), and having cGvHD after the first SCT (p=0.0001).

Most patients experience disease relapse or NRM within the first year after their second transplant. This observation seems to be more consistent in patients transplanted for ALL, with more changes over time in patients with AML. For ALL in particular, the 2-year incidences of relapse, NRM and LFS were not different from those at 5-years.

Even in the relapse setting, survival rates for patients with ALL remain superior to patients with AML, consistent with the prognostic differences at diagnosis. Our findings, consistent for the AML and ALL subgroups, suggest that cGHVD prior to second HSCT is associated with better outcome. The identification of cGHVD prior to second transplant has not been heretofore described as a favorable prognostic factor. This strong correlation merits further study, specifically as to the underlying biology for this association.

Conclusion

Children with relapsed acute leukemias have a substantial chance to become long term survivors following a second SCT. CR prior to second SCT, longer interval between transplants and the presence of cGvHD after the first transplant, are favorable prognostic factors for ALL and AML. Our findings may help physicians in discussing the risk-benefit of a second transplant. These results are particularly relevant in an era where an explosion of new therapies, specifically targeted therapies and those that modulate the immune response, behoove us to carefully identify subpopulations of patients for whom specific therapies are appropriate.

Novel approaches are needed to minimize relapse risk as well as short and long term morbidity in these pediatric patients while considering a second SCT for relapsed acute leukemia.