Single Immunophenotypical Evaluation of Minimal Residual Disease before Autotransplantation of Non-Cryopreserved Bone Marrow Is Insufficient for Prediction of Outcome in High Risk Adult Acute Lymphoblastic Leukemia.

For high-risk adult ALL patients alloHCT is a preferable option. However, a significant proportion of those not having a suitable donor may be successfully treated with autotransplantation (autoHCT). Based on our experience this treatment ensures low transplant related mortality below 3% and a reasonable overall survival and disease free survival of 60% and 45% respectively. The status of the disease before transplantation is an important factor for long term results. In childhood ALL most studies suggest that the level of minimal residual disease (MRD) after induction evaluated immunophenotypically or with bio-molecular methods is predictive for outcome after different treatments including chemotherapy, alloHCT and autoHCT. The results in adult ALL are more controversial.

Patients selection. Among 1205 haematopoetic cell transplantations performed in our institution 224 (147 autologous, 77 allogeneic) were performed in 205 adults with ALL. For this study we selected an uniform group of 81 patients fulfilling following criteria’s: Ph (-) ALL, status CR1, evaluable MRD, strictly defined autoBMT procedure performed until the end of 2003.

Methods. MRD was tested before autoBMT (median interval 10 days) using 2 ore 3-color flow-cytometry, as appropriate. The atypical immunophenotypes were evaluated using the “quadrans” analysis in all cases and since 2002 also the “empty spaces” technique. The sensitivity equals at least 0.0001. For all autoHSCT bone marrow was used as a source of stem cells. The CAV conditioning regimen consisted of cyclophosphamide 60mg/kg on d. -3, -2, cytarabine 2 g/m2 d. -3, -2, -1, etoposide 800 mg/m2 d. -3, -2. Bone marrow was not cryo-preserved after collection but stored in 4⁰ C and re-transplanted after 72h.

Results. In 41 patients; age med. 26 y (15–53), F/M=12/29, the MRD level was <0,001: the MRD (−) group. In 40 patients; age med. 29 y (16–53), F/M=18/22, the MRD was detected at the level =/> 0,001; MRD+ group. The ALL-immunophenotypes of MRD−/MRD+ groups were as follows; proB 4/7, preB 2/6, Common 18/19, B 0/1, preT 5/2, T 12/1). The interval from DGN to BMT was similar in both groups.

The probability of LFS and OS at 10y calculated with median follow up time of 5y equaled; in the MRD(−) group 47% and 62% and in the MRD+ one 48% and 57% respectively (p=ns). The main reason of failure in both groups was a relapse which occurred after a median time of 277 days in the MRD(−) group and 134 days in MRD+ one (p=0.19).

Conclusion and comment. Based on this observation we conclude that a single evaluation stratifying patients before autoBMT according to MRD level below or above 0.001 is not predictive for DFS and OS, because it informs only about the current amount of the disease but not about its opportunistic nature. In this respect a repeatedly confirmed MRD positivity should be more significant.

Taking into consideration that the main reason of failures were relapses, this finding suggests also that in patients with chemotherapy-responsive ALL confirmed by stabile CR, the myeloablative CAV regimen is sufficiently strong to eliminate the residual disease at the level ranging 0.01–0.001. It may be speculated only that the 72h lasting incubation of bone marrow product before re-transplantation has also some kind of purging effect for leukemic blasts.