Prognostic relevance of Aurora kinase A ( <i>AURKA</i> ) expression in prostate cancer (PCa).
Abstract
5061 Background: Amplification and overexpression of the AURKA gene characterize aggressive variants of prostate cancer, such as castration-resistant (CR) PCa and neuroendocrine PCa (NEPC), representing both a marker of progression and a promising therapeutic target. Methods: 7755 PCa specimens were sequenced for DNA and RNA at Caris Life Sciences and stratified into top (Q4) and bottom quartiles (Q1) based on AURKA expression. Castration status was defined as castrate (CS) if a specimen was androgen deprivation therapy (ADT)-naïve or collected within 90 days from the initiation of 1 st generation ADTs. Additionally, specimens that received 2 nd generation ADTs prior to collection were excluded from the CS cohort. All other ADT treated specimens (excluding the CS cohort) were considered CR. AR and NEPC scores were calculated as previously reported (Beltran, Nat Med 2016) and categorized into highest (H) and lowest (L) quartiles. Real-world survival was obtained from insurance claims data, and Kaplan–Meier estimates were calculated from specimen collection to last clinical contact for overall survival (OS) and from initiation to termination of specific ADTs to estimate time on treatment (TOT). Hazard ratios (HR) and p-values were calculated using the Cox model and log-rank test, with multiple testing corrections applied (q<0.05). Results: Compared to Q1, Q4 was associated with a higher median age (69 vs 67 years) and a higher proportion of non-Hispanic/Latinos (75 vs 70%), NEPC-H (42 vs 15%), metastatic (60 vs 22%), CR (46 vs 20%, all q<0.05) disease. Q4 was also associated with poor prognosis independent of race and ethnic backgrounds. Despite the significant enrichment of aggressive disease, Q4 was associated with poor prognosis independent of metastatic, NEPC-L or castration status. Further within NEPC-H specimens, Q4 was prognostic only among those that were also AR-L (Table 1). Relative to Q1, Q4 samples were enriched for mutations in TP53 (48 vs 23%), RB1 (10 vs 1%) and PTEN (11 vs 7%, all q<0.05). Interestingly, Q4 was associated with a longer Leuprolide-TOT (HR: 0.9(0.8-0.97), p<0.01) and a shorter Enzalutamide-TOT (HR: 1.2 (1-1.4), p<0.05). Conclusions: Analysis of a large dataset revealed that high AURKA expression correlates with poor prognosis across clinical and demographic subpopulations. AURKA inhibitors might enhance outcomes of metastatic PCa treated with AR pathway inhibitors by intensifying AR inhibition, increasing DNA-damage-related cell death, and/or preventing escape mechanisms like NEPC. Further studies are needed to identify contexts where AURKA inhibitors can improve metastatic PCa outcomes. HR comparing OS in Q4 vs Q1 (all p<0.0001). Conditions HR (95% CI) White 2.6 (2.3-2.9) Black/AA 1.9 (1.5-2.5) non-Hispanic/Latino 2.6 (2.3-2.9) Hispanic/Latino 2.3 (1.7-3.2) Primary 1.7 (1.5-2) Metastatic 2 (1.7-2.4) NEPC-L 2.3 (1.7-3.1) NEPC-H/AR-L 2.1 (1.6-2.8) CS 1.9 (1.6-2.2) CR 2.4 (2-3)
Article Details
Journal Info
Journal of Clinical Oncology
Lippincott Williams & Wilkins
Authors (10)
Maroun Bou Zerdan
3Emory University School of Medicine, Atlanta, United States
Nishant Gandhi
4Caris Life Sciences, Irving, United States
Andrew Elliott
Norm Smith
Caris Life Sciences, Irving, TX
Himisha Beltran
Justin Hwang
Masonic Cancer Center, University of Minnesota
Rana R. McKay
Department of Medicine, Urology, and Radiation Medicine and Applied Sciences University of California‐San Diego La Jolla California USA
Emmanuel S. Antonarakis
Masonic Cancer Center, University of Minnesota
Jindan Yu
Mehmet Asim Bilen
From the Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda (A.B.A., N.S., S.N., L.L., L.C.), the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore (J.H.-C.), and the Investigational Drug Branch, Cancer Therapy Evaluation Program, National Cancer Institute, National Institutes of Health, Rockville (H.S., E.S.) — all in Maryland; the Alliance Statistics and Data Management Center, Mayo Clinic, Rochester, MN (K.V.B., M.O., C.M., G.P.B.); AdventHealth Cancer Institute and the University of Central Florida, Orlando (G.S.); Dana–Farber/Harvard Cancer Center, Boston (S.B., B.M.); UNC Lineberger Comprehensive Cancer Center, Chapel Hill (W.Y.K.), and Duke University Medical Center and Duke Cancer Institute, Durham (J.H., S.H.) — both in North Carolina; the University of Kansas Cancer Center, Westwood (R.P.); Memorial Sloan Kettering Cancer Center, New York (M.Y.T., M.J.M., J.E.R.), and Roswell Park Comprehensive Cancer Center, Buffalo (G.C.) — both in...