AACR 2020 | Patients with altered TP53 R/R AML have increased tumor immune-infiltration and show sensitivity to flotetuzumab

In some patients with acute myeloid leukemia (AML), immune cell infiltration, defined as presence of various immune cell populations is observed in tumor microenvironment, while in others immune depletion, absence of the infiltration, is observed. In a recent study, resistance to chemotherapy, the standard of care for the majority of patients with AML, and response to flotetuzumab was found to depend on the presence of immune infiltration.¹ However, the genetic events underlying immune infiltration in AML have not yet been established. Neither has the impact of TP53 mutations, often associated with chemoresistance and poor outcomes, on immune regulation.

During the AACR Virtual Annual Meeting I 2020, Sergio Rutella, from Nottingham Trent University, presented an abstract on the correlation of TP53 abnormalities with the composition and function of the immunological microenvironment in AML.¹ In the CP-MGD006-01 clinical study (NCT02152956), the authors also investigated whether the TP53 abnormalities could be used to predict sensitivity to flotetuzumab, a CD123 x CD3 bispecific dual-affinity host T-cell re-targeting antibody.

Methods

In silico analysis

RNA sequences from 147 patients with non-promyelocytic AML, including 14 patients with TP53 mutations (TP53-mut), from TCGA-AML were used to compute immune cell type and biological activity gene signatures. The signatures were then correlated with

• prognostic molecular lesions, including mutational status of
  • TP53
  • NPM1
  • FLT3-ITD
• clonal hematopoiesis of indeterminate potential defining mutations

• clinical outcomes

Primary AML blasts analysis

RNA from primary bone marrow samples of treatment-naïve patients with AML were used in immune microenvironment gene expression profiling, followed by network and gene ontology analysis

• 42 patients with TP53-mut
• 22 with non-mutated TP53 (TP53-wt)

Flotetuzumab-treated cohort analysis

RNA from bone marrow samples of 35 patients with relapsed/ refractory (R/R) AML treated with flotetuzumab, were used in immune microenvironment gene expression profiling and analyzed for predictors of response

• 11 patients with TP53-mut and/or genomic TP53 loss
• 24 patients with TP53-wildtype (TP53-wt)

Patients were classified as responders if they achieved complete response (CR)/ CR with partial hematologic recovery (CRh), partial response (PR) or other anti-leukemic activity (e.g. > 30% decrease in BM blasts compared to baseline). Patients with treatment failure presenting as stable or progressive disease were classified as not-responders.

Results

In silico analysis

Analysis of 44 immune genes signatures in 118 patients, including 14 with TP53-mut, demonstrated an association of TP53 mutations with an immune-infiltrated tumor microenvironment but not with immune-depleted

• All patients with TP53-mut belonged to the immune-infiltrated signature cluster and those with TP53-wt to the immune-depleted cluster
• Compared with intermediate and favorable molecular risk features, all patients with TP53-mut had higher tumor inflammation score (p = 0.0001), inflammatory chemokine score (p = 0.0001), INFγ score (p = 0.0008), and lymphoid score (p = 0.0001)
• Expression of mRNA encoding immune checkpoint and immune suppression markers: PD-1, FoxP3, and TIFIT was significantly higher in TP53-mut AML (p = 0.0001, 0.0002, and 0.0002, respectively)
• Overall survival (OS) of patients with TP53-mut AML was significantly shorter compared to patients with other risk defining molecular lesions (4.5 months vs 18.5 months; HR 3.43; 95% CI, 1.34–8.74; p < 0.0001)

Primary AML blasts analysis

• Majority of patients (84%) had missense TP53-mut often associated with loss of function
• Clustering analysis revealed similar results to the in silico analysis, a high to intermediate immune infiltration signature was present in the majority of patients with TP53-mut

• • 20 out of 23 (87%) cases with TP53-mut had high immune infiltration
  • 21 out of 28 (75%) cases with TP53-mut had an intermediate immune infiltration
  • 1 out of 13 (8%) of cases with TP53-mut had low immune infiltration

• Consistent with results of in silico analysis, patients with TP53-mut compared to TP53-wt had significantly increased levels of mRNA encoding genes associated with immune suppression and immune exhaustion
  • PD-L1 (p < 0.0001)
  • FoxP3 (p < 0.0001)
  • CD6A (p = 0.0013)
  • LAG3 (p = 0.0042)
  • INFγ (p < 0.0001)
  • GZMB (p < 0.0001)
• There were 34 differentially expressed genes between TP53-mut and TP53-wt AML (subsequently called the TP53-immune gene classifier)
  • None of these genes have previously been identified as associated with TP53 pathway
  • They showed enrichment for KEGG pathway related to
    • cytokine-cytokine receptor interactions (9 of 263)
    • rheumatoid arthritis (6 of 84)
    • IL-17 signaling (6 of 92)
    • NOD-like receptor signaling (6 of 166)
    • TNF signaling (5 of 108)
• The differentially expressed genes showed prognostic value in silico
• Patients with 19 upregulated immune genes in TP53-mut AML, as compared with ‘not-altered’ immune genes had significantly lower
  • relapse-free survival (RFS) (11.4 months vs 24.1 months; HR 1.86; 95% CI, 1.12–3.1; p = 0.0068)
  • OS (11.4 months vs 27.0 months; HR 1.88; 95% CI, 1.24–2.85; p = 0.0009)
• The 15 downregulated genes in TP53-mut AML, however, were not predictive of RFS or OS

Flotetuzumab-treated cohort analysis

• Patient characteristics are presented in Table 1

Table 1. Baseline patient characteristics

• 9 out of 11 samples with TP53-mut/loss AML were available for immune gene expression profiling
• • 3 patients had high immune-infiltration signature
  • 4 patients had intermediate immune-infiltration signature
  • 3 patients had low immune-infiltration signature
• 45.5% of patients with TP53-mut/loss had anti-leukemic activity following flotetuzumab
• On average a 42% blast reduction in bone marrow was observed in TP53-mut/loss patients in response to flotetuzumab treatment, with 2 patients subsequently undergoing transplantation
• At baseline, the responding vs non- responding patients with TP53-mut/loss had significantly higher
  • tumor infiltration score (p = 0.016)
  • inflammatory cytokine score (p = 0.016)
  • regulatory T lymphocyte score (p = 0.032)
  • INFγ score (p = 0.016)
• Median OS for patients treated with flotetuzumab who had TP53-mut/loss was 4 months (p = 0.016). This survival estimate compares favorably with survival predictions for patients with mutated TP53 in larger cohort studies such as HOVON

Conclusion

The results of the immune transcriptomic analysis suggest a correlation of TP53-mut/loss AML with high T-cell infiltration, overexpression of immune checkpoints, and increased INFγ signaling. The responses of TP53-mut/loss AML to flotetuzumab are promising, showing a 42% reduction in bone marrow blasts and anti-leukemic activity in 45.5% of patients. This highlights the association between response to T-cell engagers and an inflamed tumor microenvironment. Further studies are needed to confirm the efficacy and safety of flotetuzumab in this subset of patients.