Phospho-AKT Expression and Prognosis of Locally Advanced Prostate Cancer: A Study Based on RTOG 8610
G Zhai,A Ho,E Hammond,J Fontanesi,M Rotman,M Pilepich,W Shipley,H Sandler,A Pollack,M Zhan-,A Chakravarti
DOI: https://doi.org/10.1016/j.ijrobp.2005.07.546
2005-01-01
Abstract:Purpose/Objective: Preclinical models suggest that activation of PI3K/serine-threonine kinase AKT pathway promotes cell survival and potentially treatment resistance in several tumor types. To our knowledge, no study has previously examined the prognostic value of phospho-AKT expression in men with prostate cancer who have been treated with RT ± AD. We examined the association between phospho-AKT expression and overall survival, prostate cancer survival, distant metastasis, and local progression in men treated with RT alone or RT + short term androgen deprivation (STAD) who were enrolled in Radiation Therapy Oncology Group (RTOG) protocol 8610.Materials/Methods: Of the 456 eligible and analyzable cases (parent cohort), 80 patients had adequate and suitably-stained tumor material for phospho-AKT analysis (AKT cohort). Of these, 47 were randomized to RT alone and 33 to RT+STAD. Pretreatment characteristics of the parent and AKT cohorts were not statistically different. Expression levels of phospho-AKT were determined by immunohistochemical staining (DAKO Envision+, DakoCytomation, Carpinteria, CA) using anti-phospho-AKT antibody (Cell Signaling, Beverly, MA), and reviewed under light microscopy for measuring the percentage of cells with positive staining and the intensity of the staining. The five-year actuarial estimates for overall and prostate cancer survival were calculated using the Kaplan-Meier method and the cumulative incidence method was used to estimate the five-year local progression and distant metastasis failure rates. The univariate analyses were done using phospho-AKT Mean Index % as a continuous variable and four additional cutpoints of <5 vs. ≥5, >25 vs. ≥25, <50 vs. ≥50, and a median ≤70 vs. >70, or using phospho-AKT Intensity Score as a continuous variable and an additional cutpoint of 0 vs. ≥1. Moreover, phospho-AKT Mean Index % was modeled in multivariate analysis using Cox proportional hazards models to identify the impact of AKT expression on the four endpoints in the presence of other covariates such as Gleason score, clinical stage, and assigned treatment.Results: None of the analytic results from the univariate models of the four endpoints demonstrated a statistically significant association with phospho-AKT score. Although the relative risk (RR) in local progression increased as phospho-AKT Mean Index % increased (RR = 1.04, 1.10, 1.26, and 1.56 for cutpoints 5, 25, 50, and 70, respectively), all the associations are non-significant. Of all the cutpoints of phospho-AKT Mean Index %, cutpoint of median (≤70 vs. >70) showed consistently increased relative risks of at least 20% for all the 4 endpoints as phospho-AKT Mean Index % increased. In multivariate analysis using the same cutpoint of a median (≤70 vs. >70), increased phospho-AKT Mean Index % still had 32 to 57% increases of relative risks for all the 4 endpoints when other factors such as Gleason score, clinical stage, assigned treatment were considered in the model. However, these increased risks were not statistically significant.Conclusions: Expression of phospho-AKT was not found to be a significant, independent prognostic marker in our relatively small cohort of patients with locally advanced prostate cancer who were treated primarily with radiation therapy ± androgen deprivation. Purpose/Objective: Preclinical models suggest that activation of PI3K/serine-threonine kinase AKT pathway promotes cell survival and potentially treatment resistance in several tumor types. To our knowledge, no study has previously examined the prognostic value of phospho-AKT expression in men with prostate cancer who have been treated with RT ± AD. We examined the association between phospho-AKT expression and overall survival, prostate cancer survival, distant metastasis, and local progression in men treated with RT alone or RT + short term androgen deprivation (STAD) who were enrolled in Radiation Therapy Oncology Group (RTOG) protocol 8610. Materials/Methods: Of the 456 eligible and analyzable cases (parent cohort), 80 patients had adequate and suitably-stained tumor material for phospho-AKT analysis (AKT cohort). Of these, 47 were randomized to RT alone and 33 to RT+STAD. Pretreatment characteristics of the parent and AKT cohorts were not statistically different. Expression levels of phospho-AKT were determined by immunohistochemical staining (DAKO Envision+, DakoCytomation, Carpinteria, CA) using anti-phospho-AKT antibody (Cell Signaling, Beverly, MA), and reviewed under light microscopy for measuring the percentage of cells with positive staining and the intensity of the staining. The five-year actuarial estimates for overall and prostate cancer survival were calculated using the Kaplan-Meier method and the cumulative incidence method was used to estimate the five-year local progression and distant metastasis failure rates. The univariate analyses were done using phospho-AKT Mean Index % as a continuous variable and four additional cutpoints of <5 vs. ≥5, >25 vs. ≥25, <50 vs. ≥50, and a median ≤70 vs. >70, or using phospho-AKT Intensity Score as a continuous variable and an additional cutpoint of 0 vs. ≥1. Moreover, phospho-AKT Mean Index % was modeled in multivariate analysis using Cox proportional hazards models to identify the impact of AKT expression on the four endpoints in the presence of other covariates such as Gleason score, clinical stage, and assigned treatment. Results: None of the analytic results from the univariate models of the four endpoints demonstrated a statistically significant association with phospho-AKT score. Although the relative risk (RR) in local progression increased as phospho-AKT Mean Index % increased (RR = 1.04, 1.10, 1.26, and 1.56 for cutpoints 5, 25, 50, and 70, respectively), all the associations are non-significant. Of all the cutpoints of phospho-AKT Mean Index %, cutpoint of median (≤70 vs. >70) showed consistently increased relative risks of at least 20% for all the 4 endpoints as phospho-AKT Mean Index % increased. In multivariate analysis using the same cutpoint of a median (≤70 vs. >70), increased phospho-AKT Mean Index % still had 32 to 57% increases of relative risks for all the 4 endpoints when other factors such as Gleason score, clinical stage, assigned treatment were considered in the model. However, these increased risks were not statistically significant. Conclusions: Expression of phospho-AKT was not found to be a significant, independent prognostic marker in our relatively small cohort of patients with locally advanced prostate cancer who were treated primarily with radiation therapy ± androgen deprivation.