Phospho-Akt immunoreactivity in prostate cancer: relationship to disease severity and outcome, Ki67 and phosphorylated EGFR expression

Abstract

Background: In the present study, we have investigated the prognostic usefulness of phosphorylated Akt immunoreactivity (pAkt-IR) in prostate cancer using a well-characterised tissue microarray from men who had undergone transurethral resection due to lower urinary tract symptoms.

Methodology/principal findings: pAkt-IR in prostate epithelial and tumour cells was assessed using a monoclonal anti-pAkt (Ser(473)) antibody. Immunoreactive intensity was determined for 282 (tumour) and 240 (non-malignant tissue) cases. Tumour pAkt-IR scores correlated with Gleason score, tumour Ki67-IR (a marker of cell proliferation) and tumour phosphorylated epidermal growth factor receptor (pEGFR)-IR. For cases followed with expectancy, a high tumour pAkt-IR was associated with a poor disease-specific survival, and the prognostic information provided by this biomarker was additive to that provided by either (but not both) tumour pEFGR-IR or Ki67-IR. Upon division of the cases with respect to their Gleason scores, the prognostic value of pAkt-IR was seen for patients with Gleason score 8-10, but not for patients with Gleason score 6-7.

Conclusions/significance: Tumour pAkt-IR is associated with both disease severity and disease-specific survival. However, its clinical use as a biomarker is limited, since it does not provide prognostic information in patients with Gleason scores 6-7.

Figure 1. Examples of staining intensities for pAkt-IR ranging from 1, 2, 3 and an area (indicated) with staining 4.

These photographs (20×) were used as standards by both evaluators throughout the scoring phase of the project. The top two cores are from tumours, whilst the bottom two cores are from non-malignant tissue.

Figure 2. Prognostic significance of tumour and non-malignant pAkt-IR for cases who were followed by expectancy.

Panels A and B are for tumour pAkt-IR (n = 204), C and D for non-malignant pAkt-IR (n = 194). In Panels A and C, Exp(B) (±95% confidence intervals), obtained from Cox proportional-hazards regression analyses are shown for different cut-offs. Exp(B) is defined as the increase in risk for death due to prostate cancer for a score above the cut-off value relative to a score below the cut-off value. When both confidence limits are above unity (filled symbols in the figure), the cut-off value provides significant prognostic information. Values with a significance level 0.052.5), 82 cases (40%) were ≤ the cut-off value and 122 cases (60%) above the cut-off value. In Panels B and D, Kaplan-Meier plots are shown for the cut-offs showing the highest significances. ^†Pca refers to the number of patients who died as a result of their prostate cancer during the follow-up period. The ķ² values are for the log-rank (Mantel-Cox) tests, with the P values shown: ***P<0.001, *P<0.05.

Figure 3. Prognostic significance of tumour pEGFR-IR and Ki67-IR for cases who were followed by expectancy: relationship with tumour pAkt-IR.

Panel A shows Exp(B) obtained from Cox proportional-hazards regression analyses are shown for different cut-offs of pEGFR-IR (n = 253). The cut-off value with the highest significance is shown as a red filled symbol. Values with a significance level 0.05†Pca refers to the number of patients who died as a result of their prostate cancer during the follow-up period. The ķ² values are for the log-rank (Mantel-Cox) tests, with the P values shown: ***P<0.001.

Figure 4. Prognostic significance of tumour and non-malignant pAkt-IR for cases who were followed by expectancy.

Panel A and B are for cases with Gleason scores 6–7 (n = 102); Panels C and D are for cases with Gleason scores 8–10 (n = 51). The Exp(B) values obtained from Cox proportional-hazards regression analyses (Panels A and C) and the Kaplan-Meier plots (Panels B and D) were determined as described in the legend to Fig. 2. The ķ² values are for the log-rank (Mantel-Cox) tests, with the P values shown: *P<0.05, ^NSP>0.8.