High levels of PTPN11 protein are prognostically adverse and independent of mutation status in acute myelogenous leukemia

Abstract INTRODUCTION: Mutations in Protein Tyrosine Phosphatase Non-Receptor Type 11 (PTPN11) occur in ~8% of Acute Myelogenous Leukemia (AML) cases, and confers an adverse prognosis. Almost all PTPN11 mutations result in constitutively active phosphatase activity, without changes in the catalytic and target-recognition sites of the protein. Thus, we hypothesized that overexpression of the wild-type (WT) protein could result in the same phenotype as the mutant (MUT) PTPN11 and confer a similarly adverse prognosis. METHODS: Levels of 434 proteins were measured with Reverse Phase Protein Arrays in 806 newly diagnosed, fresh, pre-treatment AML samples. Protein expression was normalized to non-G-CSF treated, normal bone marrow (NBM)-derived CD34+ cells. PTPN11 mutation status, treatment, outcome data were known for 521 cases, including 48 (9.2%) with a PTPN11 mutation and 473 that were WT. LogRank tests were used to compared outcomes; Fisher's Exact, Pearson's Chi-squared or Kruskal-Wallis' tests for comparing variables; Pearson's correlation for protein correlation (p0.01 and correlation coefficient 0.3); Wilcoxon tests adjusted by FDR for differential expression (p0.05 and LFC0.5); and Cox proportional hazards models (CoxPH) for Uni-(UV) and Multi-variate (MV) analysis. RESULTS: Patients were separated into tertiles based on PTPN11 protein expression: High (WT N=157 (33%), MUT N=10 (21%)), Med (WT N=158 (33%), MUT N=16 (33%)) and Low (WT N=158 (33%), MUT N=22 (46%)). High cases are older, have lower WBC count and percentage of blasts, and are associated with unfavorable cytogenetic risk, complex karyotype (CK), -5/5q-, -7/7q-, and TP53 MUT, but were less associated with t(9;11), and FLT3 and NPM1 MUT (P=0.02, 0.001, 0.001, 0.001, 0.001, 0.001, 0.001, 0.001, =0.04, 0.001, 0.001). High PTPN11 protein levels were adversely prognostic for overall survival (OS) in both MUT and WT cases and also for MUT-Med, while survival was similar between MUT-Low, WT-Med and WT-Low cases (5ys OS: MUT-High=10%, WT-High=18%, MUT-Med=6.7% vs WT-Med=25% vs MUT-Low=32%, WT-Low=29%; p=0.005). Notably, patients showed different responses to therapy according to PTPN11 levels: Low and Med cases performed better with Ara-C-based chemotherapy compared to Venetoclax plus Hypomethylating agent (VH), while High ones did worse with either therapy (5ys OS of AraC-treated: Low=44%, Med=37%, High=26% vs VH-treated: Low=0%, Med=0%, High=30%; p=0.006). PTPN11 MUT had a similarly bad prognosis independent of therapy and protein levels, except for those with Low expression treated with AraC, who performed as well as WT-Low cases that received the same therapy (5ys OS of MUT AraC-treated: Low=47%, Med=11%, High=14% vs MUT VH-treated: Low=0%, Med=0%, High=0% vs WT AraC-treated: Low=44%, Med=40%, High=27% vs WT VH-treated: Low=0%, Med=0%, High=32%; p0.001). In the CoxPH UV model of OS, High PTPN11 was prognostic (p =0.009), independent of mutation status, along with other clinical, cytogenetic and molecular features (e.g. age, CK, 2nd AML, FLT3-ITD mut, IDH1 and 2 mut, etc.). In the MV model, however, PTPN11 levels were not independently prognostic. Since WT-High patients showed a similar OS compared to the MUT-High and MUT-Med cases, we interrogated whether these WT cases have a 'PTPN11-MUT-like’ behavior. They have similar protein profiles with no differential expressed proteins, suggesting commonality. Next, we compared the correlation of PTPN11 with the other 433 in our database in WT-High vs MUT-High and MUT-Med to look for shared targetable signaling pathways, identifying 17 proteins positively associated with both groups. The shared proteins are related to apoptosis downregulation, and upregulation of PI3K/AKT/mTOR, VEGF, NRTK and CREB1 pathways. CONCLUSION: High PTPN11 levels were present in 33% of PTPN11-WT and 21% of MUT cases, and conferred independent adverse prognosis. OS of Low PTPN11 MUT cases was comparable to Low PTPN11 WT, suggesting that increased protein expression is the main driver of the pathophysiology in prognostically adverse cases, and not the genetic event. WT-High cases share commonalities with High PTPN11 MUT, that could be targeted to improve outcomes. Therefore, we postulate that therapies targeting PTPN11 MUT could be directed to a larger AML population (35% vs 8% of cases), greatly increasing their utility, and that proteomics could prospectively identify the WT cases that could benefit from it.