Monitoring of Minimal Residual Disease before and after Allogeneic Stem Cell Transplantation Childhood ALL - a Retrospective Assessment on Behalf of the PDWP of the EBMT, the COG, PBMTC, the I-BFM and the Westhafen-Intercontinental-Group

Objectives: A small number of previous studies have demonstrated prognostic significance of MRD detection before and after allogeneic transplantation in children ALL. A major challenge of studies has been low patient numbers that were insufficient for multivariate analysis and assessment of specific post-transplant time points. To address this, the Westhafen Intercontinental Group (WIG) performed a retrospective analysis to investigate the impact of pre- and post-transplant MRD assessment in children with ALL.

Patients and Methods:

Patients (pts) treated in Europe (France, Germany, Italy, and Netherlands), North America (COG/PBMTC, Seattle, Minnesota, and Johns Hopkins) and Australia were included. All pts were registered either in local or national trials. Pts were transplantations between 09/99 and 05/16; analysis was performed in 06/16. Altogether 747 pts were included on whom 2960 MRD measurements were performed in bone marrow. MRD prior to transplant was available in 650 pts. MRD was assessed in bone marrow prior to transplant and on days +30, +60, +100, +180, +365 and beyond day +365 post-transplant either by flow cytometry or by RQ-PCR of TCR and IG gene rearrangements.

Of these 747 pts, 466 (62%) were male and 281 (38%) were female. Pts received their transplant in CR1 (n=275, 37% ), CR2 (n=410, 55%), >CR3 (n=55, 7%), or non-remission (NR) (n=7, 1%). 586 patients (78%) had pre-B ALL, 145 pts (20%) had T-cell phenotype and 16 pts (2%) had bi-lineage or bi-phenotype AL. Grafts were matched sibling donors (MSD; n=227, 30%), matched unrelated donors (MUD; n=314, 42%), from mismatched donors (MMD, n=75, 10%), cord blood (n=128, 17%) or no data (n=3, 1%). 31 patients (4%) were younger than 2, 365 pts (49%) were ≥2 and ≤ 10, and 351 pts (47%) were older than 10 years at the time of transplant.

After MRD assessment, pts were grouped for analysis according to the level of MRD in three groups: pts in whom no MRD was detectable were categorized "MRD negative", pts with MRD <10E-4 (0.01% by flow) were "MRD low positive", and finally pts with MRD ≥10E-4(≥0.01% by flow) as "MRD high positive".

Results:

The total group of pts showed a 4y-pEFS and a 4y-pOS of 55.0% and 60%; cumulative incidence (CI) of relapse (CIR) and CI of treatment related mortality (CI TRM) were 31% and 13%. The following pre transplant factors did not influence outcome: age, gender, relapse site, cytogenetics, donor, and stem cell source.

MRD prior to transplant was available in 648 pts. MRD high, low and negative, patients showed a 4y-pEFS of 37%, 67%, and 62% (p<0.0005). Pts with a high MRD load prior to transplant had a 2.47 fold increased hazard ratio to develop relapse and a 1.67 increased risk to develop TRM (p<0.0005). Multivariate Cox regression analysis revealed that level of pre transplant MRD and remission status at transplant also significantly influenced survival.

MRD values post-transplant were analysed as time-dependent covariates. 4y-pEFS at day +30 for MRD high, MRD low and MRD negative pts were: 43%, 65%, and 59%, respectively (p=0.002). At Day +60 for MRD high, low and MRD negative pts: 36%, 47%, and 64%, respectively (p<0.008). At day +90 for MRD high, MRD low and MRD negative pts: 44%, 69%, and 65%, respectively (p<0.003). At day +180 for MRD high, MRD low and MRD negative pts: 12%, 40%, and 79%, respectively. And at day +365 for MRD high, low and MRD negative patients: 25%, 36, and 87% respectively. High MRD at all time points post-transplant led to a higher relapse and an increased TRM rate.

There was an interaction between MRD and aGVHD at all time-points investigated after transplantation. At day +30, patients with high MRD had a 2.54 fold increased risk to develop an event which is defined as relapse or TRM, where the risk for high MRD pts with aGVHD was 0.64. This was due to a lower relapse risk compared to MRD positive patients without aGVHD.

Summary and Conclusion:

This world wide study confirmed in the largest patient population investigated so far, MRD prior to transplant is a powerful predictor for survival after transplant. When analysed either for the different time points or for the highest MRD value post- transplant, MRD results were always significantly associated with pEFS. All pts who showed an MRD load of >10E-4 (0.01% by flow) at any time after transplant had an increased risk. Occurrence of aGVHD in MRD positive pts reduced the risk of relapse. Serial MRD detection after transplantation allows for the prediction of relapse.