View Single Post
Old 05-25-2014, 10:56 AM   #3
'lizbeth
Senior Member
 
'lizbeth's Avatar
 
Join Date: Apr 2008
Location: Sunny San Diego
Posts: 2,214
Re: Phosphorylated p-70S6K Predicts Tamoxifen Resistance in Postmenopausal Breast Can

Results

Association of Downstream Activated Proteins in the PI3K and/or MAPK Pathways with Clinico-Pathological Factors

The inter-observer variability analyzed using the (weighted) Cohen's kappa coefficient is depicted in Table S4 in Additional file 1. Figure S2 in Additional file 1 shows the number of evaluable cases for each downstream protein. We did not find a significant difference in median phospho-protein expression and relative age of tumor samples (Figure S3a,b,c,d,e in Additional file 1). Median phospho-protein expression was not significantly different among inclusion centers (Figure S4a,b,c,d,e in Additional file 1).
We did not find a significant association between p-p70S6K and any of the known clinico-pathological variables (Table 1). Significant associations were found between p-p70S6K and all other downstream activated proteins (Table 1). The association between other downstream activated proteins and known prognostic variables is shown in Table S5 in Additional file 1. High p-mTOR and positive p-ERK1/2 staining were both associated with a positive PgR status and low tumor grade. In addition, high expression of p-AKT(Ser473) was associated with a positive PgR status.
Downstream Activated Proteins in the PI3K and/or MAPK Pathways Predict Resistance to Tamoxifen

In the total group of ERα-positive patients (n = 563), a total of 132 recurrence-free interval events occurred (Table S6 in Additional file 1). The number of patients in each treatment arm pre and post interim analysis is shown in Figure S5 in Additional file 1. The median follow-up of patients without a recurrence event is 7.8 years. When stratified by nodal status, the hazard ratio (HR) for tamoxifen versus control in this cohort was 0.54 (95% confidence interval (CI) = 0.36 to 0.83, P = 0.004). In our primary analysis, using the median expression levels as the cutoff value for a binary factor, we did not find a significant interaction between either p-AKT(Ser473) or p-mTOR and tamoxifen. In addition, the interactions between both p-AKT(Thr308) and p-ERK1/2 and tamoxifen were nonsignificant (adjusted P for interaction 0.09 and 0.06 respectively) (Table 2). However, we observed a significant interaction for p-p70S6K with tamoxifen (P = 0.004). Patients whose tumor did not express p-p70S6K derived significant benefit from tamoxifen (HR = 0.24, 95% CI = 0.12 to 0.47, P < 0.0001), while patients whose tumor did express p-p70S6K had no benefit (multivariate HR = 1.02, 95% CI = 0.48 to 2.21, P = 0.95) (Figure 1A; Table S7 in Additional file 1). Performing the same analysis using a 1% cutoff value for ERα positivity added four extra patients and did not substantially change these results (data not shown).

(Enlarge Image)

Figure 1.
Kaplan–Meier survival analysis according to tamoxifen treatment and PI3K and/or MAPK pathway marker. (A) Kaplan–Meier survival analysis according to tamoxifen treatment and p-p70S6K expression: recurrence-free interval according to tamoxifen treatment in patients whose tumors do not express p-p70S6K (left) and patients whose tumors do express p-p70S6K (right). (B) Kaplan–Meier survival analysis according to tamoxifen treatment and p-mTOR and p-ERK1/2 expression: recurrence-free interval according to tamoxifen treatment in patients whose tumors do express low p-mTOR and are p-ERK1/2 negative (left) and in patients whose tumors do express p-ERK1/2 and/or express high p-mTOR (right). Contr, control; ERK, extracellular signal-regulated kinase; HR, hazard ratio; MAPK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycine; PI3K, phosphatidylinositol-3-kinase; p, phosphorylated; TAM, tamoxifen.


In our exploratory analyses, analyzing the expression of downstream activated proteins as a continuous variable, we observed an interaction with tamoxifen for p-AKT(Thr308) (adjusted P = 0.03), p-mTOR (adjusted P = 0.03) as well as p-p70S6K (adjusted P = 0.006) (Table S8 in Additional file 1). Examining Akaike's information criteria values indicated that a dichotomization of p-mTOR submembranous staining into <60% and ≥60% provided the best fit to the data (Table S9 and Figure S6 in Additional file 1). For p-AKT(Thr308), p-ERK1/2 and p-p70S6K, a dichotomization in any positive versus negative staining (which corresponded with dichotomization with median expression levels as the cutoff values) provided the best fit to the data (Tables S10,S11 and Figures S7,S8 in Additional file 1). Tables S12,S13,S14,S15 in Additional file 1 present the distribution of known prognostic factors over the treatment arms both for patients who were positive and for patients who were negative for these markers.
Based on preclinical knowledge that p-ERK1/2 and p-mTOR represent activation of, respectively, the MAPK pathway and the PI3K pathway, we explored the interaction with tamoxifen for the combination of these markers. Patients with a tumor that expressed either high p-mTOR or positive p-ERK1/2 protein did not benefit from tamoxifen (adjusted HR = 1.00, 95% CI = 0.48 to 2.08, P = 1.00), while patients whose tumor was negative for these two markers did derive significant benefit (adjusted HR = 0.25, 95% CI = 0.13 to 0.48, P < 0.0001) (P for interaction = 0.004) (Figure 1B; Table S16 in Additional file 1).
Downstream Activated Proteins in the PI3K and/or MAPK Pathways are Associated with Good Prognosis in the Absence of Adjuvant Systemic Treatment

When systemically untreated breast cancer patients were dichotomized according to p-AKT(Ser473), p-AKT(Thr308), p-mTOR or p-p70S6K status, the group that was positive for the marker (dichotomized according to Akaike's information criteria) exhibited a decreased risk for recurrence compared with patients whose tumor had no or low expression of the marker (Figure 2 and Table 3; Tables S17,S18,S19,S20 in Additional file 1). A similar trend was seen for patients whose tumor did express p-ERK1/2 (P = 0.14). Hierarchical clustering of the downstream activated proteins in the PI3K and/or MAPK pathway is shown in Figure 3.

(Enlarge Image)

Figure 2.
Kaplan–Meier survival analyses in control patients according to p-p70S6K and p-mTOR. Kaplan–Meier survival analysis for recurrence-free interval according to (A) p-p70S6K and (B) p-mTOR. HR, hazard ratio; mTOR, mammalian target of rapamycine; p, phosphorylated.



(Enlarge Image)

Figure 3.
Unsupervised hierarchical clustering of tumor samples and corresponding expression of PI3K and/or MAPK pathway proteins. Heat map representing unsupervised hierarchical clustering of tumor samples and corresponding expression of downstream activated proteins in the PI3K and/or MAPK pathways from patients for whom the status of all five proteins were known (n = 350). Patients are represented horizontally. Phosphorylated proteins are indicated vertically. Red, high/any expression of phosphorylated protein; green, no/low expression of phosphorylated protein (dichotomization was performed according to Akaike's information criteria). In addition the presence (red) or absence (green) of different clinico-pathological factors is shown. ERK, extracellular signal-regulated kinase; HER2, human epidermal growth factor receptor 2; MAPK, mitogen-activated protein kinase; mTOR, mammalian target of rapamycine; PI3K, phosphatidylinositol-3-kinase; p, phosphorylated.
'lizbeth is offline   Reply With Quote