Mixed chimerism (MC) and minimal residual disease (MRD) strongly predict risk for relapse in children with acute lymphoblastic leukemia (ALL) following allogeneic stem cell transplantation (allo-SCT). Preemptive immunotherapy (IT), e.g. withdrawal of immunosuppression (WD-IS) or donor lymphocyte infusion (DLI) guided by chimerism and MRD monitoring can prevent impending relapse in allo-SCT recipients. In this study we retrospectively analyzed chimerism and MRD monitoring and the effect of preemptive IT in all pts undergoing allo-SCT for ALL in our institution.


Between January 2005 and July 2014, a total of 89 children and adolescents (median age 11.5; range 2.2-26.0 years) with ALL (pB-ALL, n=63; T-ALL, n=20; biphenotypic/bilinear ALL, n=6) were transplanted in our center after remission induction treatment. 47 pts were in first, 26 pts were in second, 15 pts were in third, and 1 pt was in fourth complete remission (CR) at time of transplant. 18 pts received grafts from matched sibling donors (MSD), 61 pts from matched unrelated donors (MUD), and 10 pts were grafted from a haploidentical parent. Pts who received their graft from a matched donor (n=79) all underwent a fully myeloablative conditioning regimen consisting of TBI (12 Gy) and etoposide. Pts who were grafted from a haploidentical donor were prepared with fludarabine, thiotepa and melphalane.


Serial and semi quantitative analyses of post-transplant hematopoietic chimerism in peripheral blood were performed weekly until day 200 and monthly thereafter. Bone marrow analyses were done at days +30, +60, +90, +180 and +365 post-transplant. Thereby, MRD was assessed in 56 pts, whereas no diagnostic material was available in 33 pts. Standardized quantification of MRD was performed according to the guidelines of the Euro-MRD-Group. All patients should receive preemptive IT if they develop MC after day +28.


64/89 pts (72%) showed complete chimerism (CC) in all follow-up analyses, and 25/89 pts (28%) developed MC although all pts received a fully myeloablative conditioning regimen. From 56 pts in whom MRD could be assessed, 40 pts remained MRD negative and 16 developed MRD positivity after transplantation. IT (WD-IS, n=12; DLI, n=14) was initiated based on chimerism analysis in 22/25 pts with MC, and was guided by MRD detection in 4 pts.

For the total cohort of pts, probability of event free (pEFS) and overall survival (pOS) were 0.67 and 0.77, respectively. Furthermore, cumulative incidence of treatment related mortality (CI-TRM) and relapse (CI-relapse) were 0.11 and 0.24 for all pts.


pEFS was 0.74 in CC-pts and 0.51 in MC-pts (p<0.032). 22/25 MC-pts received pre-emptive IT resulting in a pEFS of 0.58. Due to rapid progression of their disease IT was not initiated in 3/25 MC-pts. All MC-pts without IT relapsed, and in part received further treatment. While CI-TRM remained low (0.10 for CC resp. 0.12 for MC, p>0.68) in both groups, CI-relapse was 0.17 in CC- and 0.41 in MC-pts (p<0.021).


Besides MC, MRD level post-transplant emerged as poor prognostic factor. Pts were grouped according to their highest MRD value post-transplant in MRD negative, low positive (<10E-4), and high positive (>10E-4) pts. pEFS and pOS were 0.82 and 0.95 in MRD negative (n=40), 0.56 and 0.88 in low level MRD (n=8), and 0.25 and 0.25 in high level MRD (n=8) positive pts (p<0.001). CI-relapse was 0.14 in MRD negative, 0.36 in MRD low, and 0.75 in MRD high positive pts (p<0.001), while CI-TRM between these groups remained low (0.05 for MRD negative, 0.13 for MRD low and 0.00 for MRD high, p>0.55).

Other risk factors:

Furthermore, remission status at the time of transplant influenced relapse rate. CI-relapse was 0.16 in CR1, 0.26 in CR2, and 0.45 in CR>2 pts (p<0.025). Differences in pEFS, pOS and CI-TRM among CR1, CR2, and CR>2 pts were not statistically significant.

Multivariate analysis:

Multivariate analysis for pEFS also indicated that MC and high level MRD were independent poor prognostic factors (MC, p<0.049, RR 3.16 and high level MRD, p<0.022, RR 3.99), while remission status before transplantation, ALL lineage, donor type, graft source, T cell depletion or sex showed no significant influence.


Our results show that analysis of chimerism and MRD allow the prediction of impending relapse in virtually all children and adolescents transplanted for ALL and that preemptive IT can improve outcome in these high-risk pts.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.