LAUSR

Isolated central nervous system (CNS) relapse is frequent in relapsed childhood B-cell precursor acute lymphoblastic leukaemia (BCP-ALL) (Frishman-Levy & Izraeli, 2017), and is usually stratified as low or intermediate risk disease. Isolated CNS relapse is thought to be fuelled by cell clones in the bone marrow (BM). Indeed, most patients with CNS relapse harbour minimal residual disease (MRD) in the BM (BM-MRD) and high BM-MRD is associated with a poor prognosis (Hagedorn et al, 2007). Nevertheless, BM-MRD clearance after induction defines a favourable prognosis in intermediate risk relapsed BCP-ALL patients with BM involvement (Eckert et al, 2013a) and these patients can be spared the toxicities of allogeneic stem cell transplantation (SCT) (Eckert et al, 2013a,b). The establishment of patient-specific MRD assays is therefore crucial but requires sufficient amounts of leukaemic cells, which are rarely present at diagnosis in isolated CNS disease. We treated a 7-year-old girl with isolated early CNS relapse of BCP-ALL without available MRD and cytogenetic data from primary diagnosis guided by “mouse-MRD”. Due to insufficient DNA yield from CNS blasts, MRD targets were identified in a xenograft leukaemia derived from the cerebrospinal fluid (CSF). Leukaemic cells from CSF were freshly transplanted into two NOD.Cg-PrkdcIl2rg/SzJ mice. Human leukaemic engraftment was clinically apparent and confirmed by flow cytometry post mortem (Fig 1A). Cytogenetic analysis of these cells showed hyperdiploidy. DNA was extracted from 3 9 10 human leukaemic cells from the mouse spleen and MRD target screening was performed (Flohr et al, 2008), identifying two immunoglobulin gene rearrangements (Fig 1B). Marker specificity was confirmed in the CSF at relapse (MRD load 10 ) (Fig 1C). Upon leukaemia treatment, patient CSF samples showed no measurable MRD signal, however DNA yield in CSF is generally low and samples could not be analysed according to EuroMRD guidelines. Importantly, high BM-MRD (10 ) could be detected at relapse in the patient. Fortunately, BM became MRD-negative after induction and remained negative throughout treatment (Fig 1C). Taking into account the poor prognosis of isolated CNS relapse with initial MRD levels ≥10 4 in the BM (Hagedorn et al, 2007) and the favourable prognosis of MRD negativity after induction (Eckert et al, 2013a,b), our data reassures that this patient can safely be managed without SCT. Our data show that “mouse-MRD” is a strategy to establish MRD measurements in BM and/or CSF in isolated CNS leukaemia. “Mouse-MRD” can determine BM involvement, thereby identifying patients who may or may not profit from SCT. Similarly, the method may have implications in other situations of low cell yield, such as difficult BM aspirations, pretreated patients and, potentially, isolated testicular relapse.