Minimal Residual Disease (MRD) As Exploratory Endpoint in a Phase 1 Study of the Anti-CD123 Mab CSL362 Given As Post-Remission Therapy in Adult Acute Myeloid Leukemia (AML)

Background

AML patients (pts) who achieve a morphologic complete remission (CR) may have MRD in bone marrow (BM) and/or peripheral blood detected as patient-specific leukemia-aberrant immunophenotype by multiparameter flow cytometry (MFC) or as abnormal transcripts of AML gene mutations/translocations by quantitative RT-PCR. Presence of MRDafter induction and/or consolidation therapy identifies pts at high risk of relapse, with increasing MRD levels indicating impendingrelapse in individual pts. Although standardized testing and validated endpoints for MRD determination across institutions are lacking, major cancer treatment centers can diagnose and track MRD reliablywhen assessed in serial BM samples.

CD123 (IL3-Rα) is overexpressed by AML blasts and leukemia stem cells (LSC), suggesting that agents designed to target and kill CD123+ AML LSC and blasts may eliminate MRD and delay or prevent relapse. Retrospective clinical data suggest that conversion of AML pts from MRD + to MRD negative (-) status may confer clinical benefit; however this has not been fully characterized in prospective clinical trials.

Methods

CSL362 is a fully humanized, anti-CD123 monoclonal antibody, Fc domain-engineered to bind with high affinity to CD16 on natural killer (NK) cells. CSL362 activates antibody-dependent cell-mediated cytotoxicity (ADCC) of CD123+ cells. The first-in-human Phase 1 study of CSL362 assessed safety, PK and targeted bioactivity. An exploratory objective was to track changes in MRD and to correlate changes with maintenance of CR.

Eligible pts had CD123+ de novo or secondary AML in 1^st or 2^nd CR or CRp who recovered from induction with or without consolidation therapy, had risk factors prognostic for early relapse, and were not candidates for hematopoietic stem cell transplantation. CSL362 was infused IV over 3 hours once every 14 days for up to 6 doses. Sequential pt cohorts were treated at ascending dose levels from 0.3 to 12.0 mg/kg. The final safety evaluation visit was at week (wk) 16, and each center reported remission status at wk 24 from first dose. BMwas assayed for MRD at baseline as well as in wk 5 and wk 12 during therapy according to individual site's laboratory standards using MFC and/or PCR. Changes in MRD were compared to the baseline result for each pt.

Results

Enrollment completed with 30 pts total, in 6 dose level cohorts, treated at 4 hospitals. Safety, PK and biomarker results showing an acceptable safety profile, and targeted bioactivity of CSL362 were reported at the ASH 2014 Annual Meeting [Blood 124 (21); 120]. At baseline, 12 pts were MRD+ (5 by MFC only, 4 by PCR only and 3 by both methods); 11 of 12 had a follow up evaluation. Molecular mutations in post-remission BM identified by PCR were FLT3 (3 pts), NPM1 (3 pts) and MLL (1 pt).

Of 4 pts MRD+ by MFC only, 1 converted to MFC- and maintained CR at wk 24. Of 4 pts MRD+ by PCR only, 1 converted to PCR- and maintained CR at wk 24; a 2^nd pt, who remained PCR+ on study also maintained CR at wk 24. Of 3 pts MRD+ by both MFC and PCR, 1 converted to MFC-, PCR follow-up was not done but the pt maintained CR at wk 24. The 7 pts MRD+ who did not convert to MRD- relapsed prior to wk 24. [See table].

Discussion

Conversion from MRD + to  MRD - in 4 of 11 pts, with CR maintained at wk 24, is hypothesis generating; MRD conversion suggests possible eradication of residual leukemia cells while on study. Maintenance of morphologic CR was more likely for patients who converted to a MRD-state than those who remained MRD+ during therapy. Multiple factors may affect AML response to CSL362 immunotherapy, including cytogenetic abnormalities, specific molecular mutations, and each pt's immune system status. Preclinical data suggest the activity of CSL362 requires adequate numbers and function of NK cells and expression of the target receptor (CD123) on LSC and MRD cells. This observation would provide validation for the use of MRD response as efficacy endpoint and indicator of benefit for CSL362 and, perhaps, other AML therapies. With further efforts to standardize AML MRD assessments across laboratories, MRD response may become useful as a surrogate endpoint for clinical drug development in AML.