Deficient Apoptosis Signaling in Primary Pediatric BCP-ALL Xenograft Cells Determines the Kinetic of Engraftment in Vivo in a NOD/SCID Model and Patient Outcome

We previously identified the importance of intact apoptosis signaling for treatment response in primary pediatric ALL and AML by analysing two key apoptogenic events, caspase-3 activation and cytochrome c release. A correlation of both events was exclusively found in the groups of patients with good response to treatment, showing that intact apoptosis signaling is a critical feature for favorable outcome. Using a NOD/SCID mouse model for pediatric B cell precursor ALL we found that rapid in vivo leukemia growth (Time to leukemia (TTL) short) in transplanted recipients determines poor patient outcome. In this study we now investigated the importance of deficient apoptosis signaling for leukemia engraftment in this model. We analyzed a total of 17 BCP-ALL xenografts, 5 of which had a short time (TTLshort) to clinical manifestation of the disease (< 10 weeks) and 12 with a significantly longer time (TTLlong) to leukemia (> 10 weeks). The apoptosis signaling parameters cytochrome c release and caspase-3 activation were evaluated flowcytometrically in directly isolated cells from the leukemia bearing mice before and after induction of apoptosis by factor deprivation in culture and related to TTL. Comparing TTLshort and TTLlong a significant correlation of both apoptosis parameters was found in the TTLlong group only (P= .017). Importantly, patients whose blasts engrafted in NOD/SCID-mice with TTLlong showed a superior outcome compared to the TTLshort group. Additionally, proficient apoptosis signaling as indicated by the correlation of cytochrome c release and caspase-3 activation, was only found in blast cells from recipients where the corresponding patient had favorable prognostic factors such as TEL/AML1-positivity (P= .023), absence of hyperleucocytosis at diagnosis (P= .012), absence of relapse (P= .038) and no early relapse (P= .017). In contrast, the correlation of cytochrome c release and caspase-3 activation was absent in the groups with poor prognosis. We also found a direct positive correlation between TTL and cytochrome c (P= .022) and TTLlong and caspase-3 activation (P= .048), indicating that a longer engraftment time may be due to proficient apoptosis signaling in the leukemic blasts. Our finding in the NOD/SCID human-ALL model matches our retrospective results in pediatric ALL and AML to conclude that the functional integrity of a downstream apoptotic checkpoint (cytochrome-c related caspase activation) is an important feature regulating leukemia biology. Thus, deficient apoptosis signaling appears to determine rapid engraftment of leukemia cells in the NOD/SCID model in vivo and consequently poor patient outcome.