NATIONAL HARBOR, MD. – New research has shown increased immune infiltration in patients with TP53-mutated acute myeloid leukemia (AML).
Patients with TP53-mutated AML had higher levels of T-cell infiltration, immune checkpoint molecules, and interferon (IFN)–gamma signaling than patients with wild-type TP53.
These findings may indicate that patients with TP53-mutated AML will respond to T-cell targeting immunotherapies, but more investigation is needed, according to Sergio Rutella, MD, PhD, of Nottingham (England) Trent University.
Dr. Rutella described the findings at the annual meeting of the Society for Immunotherapy of Cancer.
He and his colleagues recently identified subgroups of AML, called “immune infiltrated” and “immune depleted,” that can predict chemotherapy resistance and response to flotetuzumab (ASH 2019, Abstract 460). However, the team has not determined the genetic drivers of immune infiltration in AML.*
With the current study, Dr. Rutella and his colleagues wanted to determine if TP53 mutations are associated with the AML immune milieu and see if TP53-mutated patients might benefit from immunotherapy.
Discovery cohort
The researchers first analyzed 147 patients with non-promyelocytic AML from the Cancer Genome Atlas. In total, 9% of these patients (n = 13) had TP53-mutated AML. The researchers assessed how 45 immune gene and biological activity signatures correlated with prognostic molecular lesions (TP53 mutations, FLT3-ITD, etc.) and clinical outcomes in this cohort.
The data showed that immune subtypes were associated with overall survival (OS). The median OS was 11.8 months in patients with immune-infiltrated AML, 16.4 months in patients with intermediate AML, and 25.8 months in patients with immune-depleted AML.
The inflammatory chemokine score (P = .011), IDO1 score (P = .027), IFN-gamma score (P = .036), and B7H3 score (P = .045) were all significantly associated with OS. In fact, these factors were all better predictors of OS than cytogenetic risk score (P = .049).
The IFN-gamma score, inflammatory chemokine score, and lymphoid score were all significantly higher in TP53-mutated patients than in patients with RUNX1 mutations, NPM1 mutations, FLT3-ITD (with or without NPM1 mutations), and TET2/DNMT3A/ASXL1 mutations (P values ranging from less than .0001 to .05).
Likewise, the tumor inflammation signature score was significantly higher among TP53-mutated patients than among patients with NPM1 mutations, FLT3-ITD (with or without NPM1 mutations), and TET2/DNMT3A/ASXL1 mutations (P values ranging from less than .0001 to .01).
Validation cohort and bone marrow samples
The researchers also looked at data from a validation cohort, which consisted of 140 patients with non-promyelocytic AML in the Beat AML Master Trial. Twelve percent of these patients (n = 17) had TP53 mutations.
Data in this cohort showed that CD3G messenger RNA (mRNA) was significantly higher in TP53-mutated AML than in TP53-wild-type AML (P = .04). The same was true for CD8A mRNA (P = .0002) and GZMB mRNA (P = .0005).
Likewise, IFN-gamma mRNA (P = .0052), IFIT2 mRNA (P = .0064), and IFIT3 mRNA (P = .003) were all significantly higher in patients with TP53-mutated AML.
Lastly, the researchers analyzed gene expression profiles of bone marrow samples from patients with AML, 36 with mutated TP53 and 24 with wild-type TP53.
The team found that IFN-gamma–induced genes (IFNG and IRF1), markers of T-cell infiltration (CD8A and CD3G) and senescence (EOMES, KLRD1, and HRAS), immune checkpoint molecules (IDO1, LAG3, PDL1, and VISTA), effector function molecules (GZMB, GZMK, and GZMM), and proinflammatory cytokines (IL17A and TNF) were all significantly overexpressed in TP53-mutated AML.
Among the top overexpressed genes in TP53-mutated AML were genes associated with IFN signaling and inflammation pathways – IL-33, IL-6, IFN-gamma, OASL, RIPK2, TNFAIP3, CSF1, and PTGER4. The IL-17 and TNF signaling pathways were the most enriched pathways in TP53-mutated AML.
“Our analysis of primary bone marrow samples showed that TP53-mutated samples are enriched in IL-17, TNF, and IFN signaling molecules, and show higher levels of T-cell infiltrations and immune checkpoints relative to their wild-type counterparts,” Dr. Rutella said.
“The in silico analysis indicated that TP53-mutated cases will show higher levels of T-cell infiltration, immune checkpoints, and IFN-gamma signaling, compared with AML subgroups without risk-defining molecular lesions,” he added. “This is speculative. Whether TP53-mutated AML can be amenable to respond to T-cell targeting immunotherapies is still to be determined.”
Dr. Rutella reported research support from NanoString Technologies, MacroGenics, and Kura Oncology.
SOURCE: Rutella S et al. SITC 2019. Abstract O3.
*This article was updated on 11/19/2019.