Targetting CD38 in T Cell Acute Lymphoblastic Leukemia

Targeted immunotherapy has become a promising emerging therapeutic strategy for many cancers. No safe and effective immunotherapies have been developed for T-cell acute lymphoblastic leukemia (T-ALL). CD38 is a transmembrane glycoprotein found on the cell surface of activated T cells, terminally differentiated B cells, but relatively low levels on normal lymphoid and myeloid cells. Daratumumab (Janssen Biotech, Inc.) is a human monoclonal antibody that binds to CD38 and is used in the treatment of refractory multiple myeloma (MM).

To ensure CD38 is a relevant target in T-ALL, we evaluated CD38 expression by flow cytometry from 21 patients with T-ALL (10 early T-cell precursor (ETP) and 11 non-ETP) at diagnosis and after induction chemotherapy. CD38 was expressed in all samples and surface expression did not change after induction (mean CD38 mean fluorescence intensity (MFI) at diagnosis vs end-of-induction: 3.27 log vs 3.19 (p = 0.25). Similarly, we evaluated CD38 expression of 15 primary T-ALL xenograft samples (7 ETP and 8 non-ETP T-ALL), and only 1 sample demonstrated low CD38 expression. In contrast, we evaluated CD38 expression in 10 patients with de novo B-ALL at diagnosis and found CD38 expression was significantly decreased after induction therapy (5351 vs. 723, p-0.002). Therefore we hypothesized that targeting CD38 with daratumumab would be effective against T-ALL.

We xenografted primary ALL blasts from 15 different patients, including 7 with ETP-ALL and 8 with non-ETP T-ALL. Mice were randomized to daratumumab (200 μg /mouse intraperitoneally weekly) vs isotype control (200 μg/mouse; 5 mice per arm for each sample) after they developed >1% peripheral blood (pb) blasts by FACS. Disease burden was assessed by FACS enumeration of pb blasts weekly and splenic blasts at sacrifice.

We demonstrate striking efficacy of daratumumab monotherapy in 6 of 7 ETP-ALL samples with significant reduction of pb and splenic blasts. Mice treated with high disease burden from 5 of the 8 non-ETP T-ALL samples were moribund immediately after daratumumab injection, presumably from tumor lysis. Autopsy performed on moribund compared to isotype-treated animals did not reveal embolism. We repeated the experiments in the 8 non-ETP T-ALL samples and the one ETP sample that did not respond when treated at high disease burden. For these experiments, we treated the mice after injection but before detectable pb blasts (termed low disease burden). Using this approach, daratumumab was effective in 7 of 8 non-ETP ALL samples and the one ETP sample that was non-responsive when treated at high disease burden. In summary, we found dara monotherapy was effective in patient-derived xneograft models developed from 14 of 15 patients with T-ALL.

The 1 T-ALL sample that did not respond to daratumumab, was the only sample with low CD38 expression. Therefore we hypothesize that the clinical benefit from daratumumab reflects the level of CD38 expression. In myeloid and MM cells, low dose all-trans retinoic acid (ATRA) significantly upregulates CD38 expression. We have similarly tested T-ALL cell lines across a broad range of ATRA doses, however ATRA has no effect on CD38 expression on T-ALL cells. Experiments are ongoing to evaluate alternative agents such as interferon to upregulate CD38 and CRISPR/Cas9 to knock-out CD38.

In summary, we found daratumumab is a highly effective novel monotherapy for T-ALL in preclinical models. Based on this work, we are planning to open a multi-institutional early phase clinical trial for children and young adults with relapsed and refractory T-ALL.