CD47 Levels Are Prognostic of Response to Magrolimab Plus Docetaxel in Metastatic NSCLC

Lung Cancer

Treatment with magrolimab plus docetaxel led to peripheral CD47 saturation in patients with metastatic non–small cell lung cancer (mNSCLC), which correlated with improved outcomes, according to findings from a biomarker analysis of the phase 2 ELEVATELung&UC trial (NCT04827576) that were presented at the 2024 SITC Annual Meeting.¹

Moreover, patients with NSCLC with confirmed partial responses (PRs) or complete responses (CRs) to magrolimab plus docetaxel had higher baseline levels of CD47 expression. Among patients with CD47 expression on more than 50% of tumor cells, the median progression-free survival (PFS) was 8.9 months (95% CI, 1.1-10.2) vs 2.6 months (95% CI, 0.8-5.4) among those with baseline CD47 expression below 50% (HR, 0.60; 95% CI, 0.24-1.54). The median PFS in the intention-to-treat (ITT) NSCLC population was 5.4 months.

“Despite a limited sample size, the correlation between CD47 expression and better outcome in both [ELEVATELung&UC] and in Tempus real-world data of patients treated with chemotherapy and/or [immunotherapy] suggests a positive prognostic role for CD47 in mNSCLC,” lead study author Nick van Buuren, of Gilead Sciences, and coauthors, wrote in a poster of the data.

The authors also noted that treatment with magrolimab triggered plasma myeloid cytokine response without signs of peripheral blood monocyte activation. Magrolimab did not induce interferon-related plasma cytokine production. Furthermore, circulating tumor DNA (ctDNA) data were found to be consistent with patient responses, and baseline macrophage gene signatures did not correlate with magrolimab response. Notably, 1 responder had tumor macrophage infiltration on day 1 of cycle 3.

Trial Design

The multicohort, open-label ELEVATE-Lung&UC trial investigated the safety, tolerability, and efficacy of magrolimab plus docetaxel in patients with mNSCLC, metastatic small cell lung cancer (mSCLC), and metastatic urothelial cancer. This trial included a safety run-in phase (n = 6) and three phase 2 cohorts. Phase 2 cohort 1a enrolled patients with mNSCLC in the second- or later-line setting (n = 31). Phase 2 cohort 1b enrolled patients with metastatic urothelial cancer in the second- or later-line setting (n = 29). Phase 2 cohort 1c enrolled patients with mSCLC in the second- or later-line setting (n = 46).

Patients with mNSCLC must have received 1 to 2 prior lines of platinum-based chemotherapy and/or a checkpoint inhibitor. Collection of archival or fresh baseline tumor biopsy was mandatory and occurred during screening. On-treatment biopsies were optional and were collected on day 1 of cycle 3. Biopsies were analyzed using CD47 immunohistochemistry (IHC) and RNA-Seq.

Patients received magrolimab at a priming dose of 1 mg/kg on cycle 1 day 1, followed by a weekly dose of 30 mg/kg from cycle 1 day 8 to cycle 2 day 15, followed by a maintenance dose of 60 mg/kg every 3 weeks from cycle 3 day 1 onward. Docetaxel was administered at 75 mg/m2 every 3 weeks.

The investigators assessed pharmacodynamic biomarkers including receptor occupancy by whole blood flow cytometry, 9-plex serum cytokine analysis by ELLA, myeloid cell whole blood immunophenotyping, and Guardant 360 ctDNA analysis. Biomarker analyses of the mNSCLC cohort (n = 29) evaluated peripheral blood and tumor samples to identify magrolimab’s receptor occupancy, predictive biomarkers, mechanism of action, and ctDNA dynamics.

Patient Baseline Characteristics

Among evaluable patients with NSCLC in the ITT population (n = 31), the median age was 65 years (range, 57-70). Most patients were male (77.4%), White (90.3%), had stage IV disease (96.8%), had an ECOG performance score of 1 at baseline (54.8%), and were a former smoker (58.1%). Best overall responses to magrolimab plus docetaxel in this population included CR or PR (19.4%), stable disease (SD; 38.7%), and progressive disease (PD; 16.1%).

The biomarker-evaluable population was a subset of 24 patients with NSCLC who had reportable baseline CD47 IHC results. The patient characteristics and best overall response rates reflected those observed in the ITT population. In this population, the median age was 64 years (range, 58-67). Most patients were male (75.0%), White (87.5%), had an ECOG performance score of 1 at baseline (62.5%), and were a former smoker (50.0%); all patients had stage IV disease. Best overall responses to magrolimab plus docetaxel in this population included CR or PR (20.8%), SD (37.5%), and PD (12.5%).

Additional Biomarker Findings

Mean variant allele frequency (VAF) in baseline ctDNA samples was prognostic; in patients with NSCLC, a lower mean VAF was associated with a longer PFS. Among patients with a mean VAF at or above the median of 2.8%, the median PFS was 3.0 months (95% CI, 1.9-not assessed [NA]), whereas among patients with a mean VAF below the median, the median PFS was 6.0 months (95% CI, 1.9-12.0). Among patients with ctDNA-low disease, defined as having insufficient molecular somatic variant counts to calculate baseline mean VAF, the median PFS was 10.1 months (95% CI, 4.2-NA). Furthermore, patients who achieved a response per RECIST 1.1 criteria were more likely to also achieve a molecular response at cycle 1 day 1.

The investigators found that abundance of baseline tumor macrophages, specifically SPP1 and C1QC TAM, which are prophagocytic and correlate with antitumor activity, were not predictive of response to magrolimab.

This study also included a differential gene expression analysis that was performed by evaluating gene expression in baseline samples from patients with confirmed PRs or CRs vs those with stable or progressive disease. This analysis corrected for the covariates of tumor content, age, histology, race, and gender. Investigators found that baseline macrophages did not significantly affect response, and that cell cycle regulation was the most significant pathway associated with poor response. CD8 infiltrate did not affect response rates, but CD47 gene expression was strongly prognostic for response.

In February 2024, Gilead announced a global pause in enrollment to studies investigating magrolimab in patients with solid tumors, including ELEVATELung&UC, to review the risk-benefit profile of magrolimab across the trials.² The FDA subsequently requested a partial clinical hold on these trials.

Gilead later decided to terminate the ELEVATELung&UC trial.³

References

1. van Buuren N, U’Ren L, Song K, et al. Biomarker analysis of magrolimab plus docetaxel in patients from ELEVATE-Lung and UC, a phase 2 multicohort study. Presented at: 2024 SITC Annual Meeting. November 6-10, 2024; Houston, Texas. Abstract 184.
2. Gilead statement on magrolimab studies in solid tumors. News release. Gilead Sciences, Inc. February 15, 2024. Accessed November 8, 2024. https://www.gilead.com/company/company-statements/2024/gilead-statement-on-magrolimab-studies-in-solid-tumors
3. Study of magrolimab in patients with solid tumors (ELEVATELung&UC). ClinicalTrials.gov. Updated October 22, 2024. Accessed November 8, 2024. https://clinicaltrials.gov/study/NCT04827576