(P 394) LEIOMYOSARCOMA INTRATUMORAL MACROPHAGE INFILTRATES ARE ASSOCIATED WITH DISTINCT CLINICAL AND GENOMIC PRESENTATIONS

Objective: Leiomyosarcomas (LMS) are aggressive malignancies of smooth-muscle origin encompassing heterogeneous clinical, pathologic, and molecular diseases. Currently, there exist few targeted molecular options for the treatment of LMS and immunotherapy (IO) has been disappointing. In this study, we examine the gene expression of LMS tumors with a focus on immune infiltrate, specifically macrophages. This cell type constitutes the majority of immune infiltrating cells in LMS and new macrophage targeting drugs are being developed. Our objective was to identify actionable targets for therapeutic intervention in specific subgroups of patients with LMS.

Methods: Tumor samples from a series of LMS patients were prospectively collected and banked for DNA and RNA sequencing as part of the MD Anderson Rare Tumor Initiative and the Patient Mosaic™ Project. Associated clinical and treatment data for the patients were collected and maintained. Differential gene expression (DGE) analysis was done using the DESeq2 package for R and the Benjamini-Hochberg procedure used to control for false discovery rate. Immune cell fractions were estimated using the immunedeconv package for R with the quantiseq method. Inferred prevalence of immune cells by type was then interrogated based on multiple tumor conditions, to include vascular tumors, presence of TP53 mutation, disease status at presentation (primary tumor, local recurrence or metastatic disease), and prior receipt of radiation therapy (RT), chemotherapy or IO before time of sample collection. Survival data was analyzed using the survival package for R.

Results: Analysis included 41 tumor samples from 34 patients (Table 1). Median age at diagnosis was 53 years and 79% of patients were female. Twenty-six tumors were soft tissue (STLMS) and 8 were uterine (ULMS). Of these, 11 (32%) were primary tumors at the time of sampling, 9 (26%) were locally recurrent and 14 (41%) were metastatic. Seventeen tumors were of vascular origin and 16 contained a TP53 gene mutation. Prior to sampling, 29 patients had received chemotherapy, 3 patients had received RT to the site sampled, and 4 patients had received IO.
DGE analysis and immune infiltrate based on RNA signature was completed for all samples (Figure 1) and revealed that immune infiltrates varied across samples despite LMS type (STLMS vs ULMS). Six paired samples from 3 patients demonstrate insight into dynamics of immune infiltrate in response to tumor status or treatment, to include synchronous primary tumor and metastasis, tumor pre- and post- RT and tumor pre- and post- TIL therapy.
Across all conditions studied, the presence of M1 macrophages varied between groups, whereas M2 macrophages were more homogeneous (Figure 2). Significant findings include higher M1 macrophage infiltrates in vascular LMS compared to non-vascular LMS (p 0.0029), lower M2 macrophage infiltrate in patient samples post-IO (p=0.01), and lower M2 macrophage infiltrates in tumors harboring TP53 mutations (p=0.021).
Survival analysis was then performed using available clinical data from the patient cohort to stratify risk by M1 macrophage infiltrate. Patients with M1 infiltrate quantified above the median had reduced overall survival (OS) compared to those that did not (5.25 vs. 9.75 years, P=0.3) (Figure 3). This finding was reproduced in patients with the top quartile of M1 infiltrate (2.32 vs. 12.68 years, P=0.2). DGE analysis was then repeated to stratify for M1 infiltrate, and expression of CD47 “don’t eat me” signal was noted to be increased in the tumors from the higher M1 infiltrate group.

Conclusion: M1 macrophage infiltrate is significantly increased in tumors of vascular origin. OS in tumors with high M1 infiltrate is inferior to tumors with low M1 infiltrate, although this did not reach statistical significance. Increased expression of CD47 within high-M1 infiltrate tumors suggests that macrophage infiltration could be used as a biomarker for therapies such as CD47/SIPRa inhibitors. Future directions include validating these findings in a larger cohort.