Introduction: Lemzoparlimab is a differentiated antibody targeting a distinct CD47 epitope to spare red blood cell binding while retaining potent anti-tumor activity1. Combination treatment with AZA is well tolerated, without requiring a priming dose, and exhibited clinically meaningful efficacy in newly diagnosed HR-MDS in an ongoing phase IIa trial (NCT04202003)2. Here we conducted pharmacodynamics, pharmacogenetics, and molecular biomarker analyses to investigate the potential underlying mechanism of the observed therapeutic effect of lemzoparlimab and AZA therapy in HR-MDS.
Methods: 53 patients with untreated IPSS-R intermediate or high-risk MDS were treated with lemzoparlimab at 30 mg/kg IV weekly and AZA at 75 mg/m2 SC on Days 1-7 in 28-day cycle. Clinical efficacy was assessed by IWG 2006 criteria per investigator. Patients achieved CR, PR, HI, or marrow-CR were defined as responders and those with SD or PD were defined as non-responders. Correlation of efficacy and biomarker data was evaluated in patients who received with ≥ 3 cycles of treatment.
Bone marrow aspirates obtained at baseline and after treatment (C2D1, C3D1, C4D1, C6D1) were subjected to deep immune profiling measured by flow cytometry, including expression of CD47 and calreticulin (CALR) in CD33+ blast cells, expression of CD91 (receptor for CALR) in macrophages, and immune cell infiltrates including CD8+ T cell, Treg and macrophages. Pharmacogenetics profiles of a panel of 267 MDS driver genes were generated from baseline bone marrow samples by next generation sequencing. In vitro phagocytosis was performed by co-culture of human macrophages with different AML cell lines in the presence of lemzoparlimab. Statistical analyses included paired and unpaired t test.
Results: As of March 31st, 2022, 28 efficacy-evaluable patients had received ≥ 3 cycles of treatment, provided bone marrow samples for biomarker analyses, and the overall response rate was 82.1% (23/28). Expression of CALR, a pro-phagocytotic signal, in bone marrow derived CD33+ blasts was measured in 28 patients which was increased and sustained throughout the treatment course, and peaked to 2-fold from baseline, in 23 responders (p<0.05 in all timepoints), and there was no increase in the remaining 5 non-responders. Expression of the anti-phagocytotic signal CD47 in CD33+ blasts was upregulated and peaked up to 1.3-fold from baseline in both responders and non-responders, with no significant difference between the two groups. The ratio of CD8+ effector T cells vs regulatory T cells (CD8/Treg) was increased in bone marrow aspirate throughout the treatment course in responders (n=23, p<0.05 at C6D1), but not in non-responders (n=5). Higher percentages of CD68+ total macrophages and CD91+ activated macrophages were observed at baseline in responders than those in non-responder (p<0.001), however the percentages of
Among the 28 efficacy-evaluable patients who received ≥ 3 cycles of treatment, patients harboring TP53 mutations (n=4) showed a trend of ( with 2 CR and 2 mCR) compared to those with wild-type TP53 (n=24, with 8 CR and 11 mCR), accompanied by increase in CALR expression in CD33+ blasts after treatment and more immune cell infiltrates at baseline, including macrophages, CD8+T cell, Treg and higher ratio of CD8/Treg (p<0.05 in all comparisons). Consistently, in vitro data showed that AML cell lines with TP53 mutations were more responsive to lemzoparlimab mediated phagocytosis than those with wild-type TP53.Conclusion: Here we show that an increased CALR expression in CD33+ blasts after lemzoparlimab and AZA combination treatment and higher immune infiltrates including total, CD91+ macrophages and CD8/Treg ratio in bone marrow at baseline is associated with better clinical response, suggesting the important role of activation of tumor derived pro-phagocytic signal and effector immune cells in the anti-tumor activity mediated by combination treatment. In addition, TP53 mutant patients with a higher increase in CALR expression and immune infiltrates in bone marrow showed a better clinical response than those with WT TP53. Our results pinpoint the potential mechanism of clinical benefits observed with lemzoparlimab and AZA treatment in HR-MDS and provide insight into future combination strategies.
Weng:F. Hoffmann-La Roche Ltd.: Other: All authors received support for third-party writing assistance, furnished by Bena Lim, PhD, of MediTech Media and funded by F. Hoffmann-La Roche Ltd, Basel, Switzerland., Research Funding.
Asterisk with author names denotes non-ASH members.