Phosphorylated Akt Expression in Early-Stage NSCLC
Phosphorylated Akt Expression in Early-Stage NSCLC
We retrospectively reviewed 527 consecutive patients with stage IB primary NSCLC according to the American Joint Commission on Cancer 6th edition tumour, node, metastasis staging system, who underwent surgical resection between January 1990 and May 2008 at the Sydney Cancer Centre (Royal Prince Alfred Hospital/Concord Repatriation General Hospital). We excluded 56 patients who had segmental or wedge resections or who had received neoadjuvant or adjuvant treatments to keep the cohort homogenous. Pathology reports were reviewed, and pathological characteristics (tumour size, histopathological type, grade, visceral pleural, vessel and perineural invasion) were extracted. Two representative formalin-fixed paraffin-embedded tumour blocks from each case were retrieved from the pathology archives. They were sectioned and stained with H&E. An experienced pulmonary pathologist (WAC) reviewed the slides and marked representative areas. Histopathological type was classified according to the World Health Organisation Classification of Tumours 2004. Tissue microarrays (TMAs) were constructed using four 1 mm tumour cores taken from two separate tumour blocks in areas carefully selected by an experienced pulmonary pathologist (WAC) to account for morphological heterogeneity and potential loss of tumour cores during processing. Matched non-neoplastic lung tissue was also used from each case. Previous studies by Cooper et al and recent studies by Karlsson et al and Schmidt et al have shown that 3–4 tumour cores are adequate for TMA studies in lung cancer. Immunostaining on whole sections was not performed because the use of TMAs in NSCLC has been validated in previous pAkt immunohistochemistry studies. This study was approved by the Human Research Ethics Committee of Royal Prince Alfred (X10-0278; HREC/10/RPAH/491) and Concord Repatriation General Hospitals (CH62/6/2004-116).
Four-micrometre-thick sections were cut from the TMA blocks, mounted onto charged slides and sections were baked in an oven at 60°C for 2 h. Sections were de-paraffinised and rehydrated through graded alcohols. Antigen retrieval was performed by placing slides in 10 mmol/L citric acid buffer, pH 6.0 Dako Target Retrieval Solution, S1699 (Dako, Glostrup, Denmark) in a pressure cooker for 4 min at 125°C, followed by running cold water around the antigen retrieval box containing the slides for 15 min to cool down. Slides were removed from the retrieval solution, rinsed in distilled water for 1 min and then loaded onto the Dako Autostainer (Dako, Glostrup, Denmark). Slides were treated with 3% hydrogen peroxides endogenous block for 5 min, followed by a buffer rinse (Dako Wash Buffer, S3006). They were then incubated with 5% goat serum in TBST for 10 min to block non-specific background staining. The primary antibody, phosphorylated Akt (Ser473) rabbit monoclonal antibody (Cell Signaling Technology, Beverly, Massachusetts, USA) at a 1:50 dilution, was applied to the sections and incubated for 90 min. Slides were rinsed with Dako buffer and incubated with the Mach3 Rabbit Probe, M3R531H (Biocare Medical, Concord, California, USA) for 20 min. After two further buffer rinses, slides were incubated with the Mach3 Rabbit Polymer-HRP, M3R531H (Biocare Medical) for 20 min. Dako liquid DAB+ substrate chromogen system, K3468, was applied to slides and incubated for 10 min for visualisation. Sections were removed from the Dako Autostainer and rinsed in distilled water. They were then counterstained in haematoxylin, dehydrated and cleared in xylene before coverslipping. Sections were analysed by standard light microscopy. Prostate LnCAP cells were used as a positive control, and Rabbit IgG (Cell Signaling Technology) was used as a negative control.
The extent and pattern of pAkt-specific immunostaining within a tissue section was determined by the percentage (0%–100%) of positively staining cells and intensity of staining (0, absent; 1+, weak; 2+, moderate; and 3+, intense) within the cytoplasm (C) and the nucleus (N). Sections were assessed independently by two investigators (WAC and PYY), and a modified H-score (the product of percentage of positively staining cells and intensity) was calculated for each core. The final modified H-score of each case was the average from all available tumour cores. The concordance of pAkt score between the two investigators was high in both cytoplasmic (Spearman's correlation coefficient, r=0.93, p<0.001) and nuclear (r=0.89, p<0.001) compartments.
To examine the relationship between the level of pAkt expression and the clinicopathological factors, Pearson's χ test was performed. Kaplan–Meier and log-rank analyses were performed on the cytoplasmic and nuclear pAkt H-scores in a stepwise fashion (ie, using a cut-off of 10). The cutpoints that resulted in the most significant relation with overall survival were selected. The optimal modified H-score cutpoints for pAkt expression were 70 in the cytoplasmic (≤70, low vs >70, high) and 40 (≤40, low vs >40, high) in nuclear compartments, respectively. The cohort was grouped into four categories (high C/low N, low C/low N, high C/high N and low C/high N) according to the level of pAkt expression in these two subcellular compartments. We understand that this will overestimate the magnitude of the effect, but this raises hypotheses that warrant corroboration in an independent dataset before it can be considered for routine clinical practice. Although the median follow-up was 11.3 years, survival was censored at 10 years as lung cancer-related death was unlikely >10 years postresection. Overall 10-year survival was calculated from the time of resection until the time of last follow-up or death within 10 years from resection. Lung cancer-specific mortality could not be assessed due to inconsistency in the source documentation.
Survival analysis was performed using univariable and multivariable analyses in a Cox proportional hazards model for pAkt expression and known clinicopathological prognostic factors. Age, gender, extent of operation, tumour size, histopathology, histological grade and visceral pleural invasion were selected for inclusion in the multivariable model because they have previously proven prognostic in NSCLC, regardless of their statistical significance in this dataset. All p values were two-sided, and a p value <0.05 was considered significant. All statistical analyses were performed using PASW Statistics V.18.0 (SPSS Inc, Chicago, Illinois, USA).
Patients and Methods
We retrospectively reviewed 527 consecutive patients with stage IB primary NSCLC according to the American Joint Commission on Cancer 6th edition tumour, node, metastasis staging system, who underwent surgical resection between January 1990 and May 2008 at the Sydney Cancer Centre (Royal Prince Alfred Hospital/Concord Repatriation General Hospital). We excluded 56 patients who had segmental or wedge resections or who had received neoadjuvant or adjuvant treatments to keep the cohort homogenous. Pathology reports were reviewed, and pathological characteristics (tumour size, histopathological type, grade, visceral pleural, vessel and perineural invasion) were extracted. Two representative formalin-fixed paraffin-embedded tumour blocks from each case were retrieved from the pathology archives. They were sectioned and stained with H&E. An experienced pulmonary pathologist (WAC) reviewed the slides and marked representative areas. Histopathological type was classified according to the World Health Organisation Classification of Tumours 2004. Tissue microarrays (TMAs) were constructed using four 1 mm tumour cores taken from two separate tumour blocks in areas carefully selected by an experienced pulmonary pathologist (WAC) to account for morphological heterogeneity and potential loss of tumour cores during processing. Matched non-neoplastic lung tissue was also used from each case. Previous studies by Cooper et al and recent studies by Karlsson et al and Schmidt et al have shown that 3–4 tumour cores are adequate for TMA studies in lung cancer. Immunostaining on whole sections was not performed because the use of TMAs in NSCLC has been validated in previous pAkt immunohistochemistry studies. This study was approved by the Human Research Ethics Committee of Royal Prince Alfred (X10-0278; HREC/10/RPAH/491) and Concord Repatriation General Hospitals (CH62/6/2004-116).
Immunohistochemistry
Four-micrometre-thick sections were cut from the TMA blocks, mounted onto charged slides and sections were baked in an oven at 60°C for 2 h. Sections were de-paraffinised and rehydrated through graded alcohols. Antigen retrieval was performed by placing slides in 10 mmol/L citric acid buffer, pH 6.0 Dako Target Retrieval Solution, S1699 (Dako, Glostrup, Denmark) in a pressure cooker for 4 min at 125°C, followed by running cold water around the antigen retrieval box containing the slides for 15 min to cool down. Slides were removed from the retrieval solution, rinsed in distilled water for 1 min and then loaded onto the Dako Autostainer (Dako, Glostrup, Denmark). Slides were treated with 3% hydrogen peroxides endogenous block for 5 min, followed by a buffer rinse (Dako Wash Buffer, S3006). They were then incubated with 5% goat serum in TBST for 10 min to block non-specific background staining. The primary antibody, phosphorylated Akt (Ser473) rabbit monoclonal antibody (Cell Signaling Technology, Beverly, Massachusetts, USA) at a 1:50 dilution, was applied to the sections and incubated for 90 min. Slides were rinsed with Dako buffer and incubated with the Mach3 Rabbit Probe, M3R531H (Biocare Medical, Concord, California, USA) for 20 min. After two further buffer rinses, slides were incubated with the Mach3 Rabbit Polymer-HRP, M3R531H (Biocare Medical) for 20 min. Dako liquid DAB+ substrate chromogen system, K3468, was applied to slides and incubated for 10 min for visualisation. Sections were removed from the Dako Autostainer and rinsed in distilled water. They were then counterstained in haematoxylin, dehydrated and cleared in xylene before coverslipping. Sections were analysed by standard light microscopy. Prostate LnCAP cells were used as a positive control, and Rabbit IgG (Cell Signaling Technology) was used as a negative control.
Assessment of pAkt
The extent and pattern of pAkt-specific immunostaining within a tissue section was determined by the percentage (0%–100%) of positively staining cells and intensity of staining (0, absent; 1+, weak; 2+, moderate; and 3+, intense) within the cytoplasm (C) and the nucleus (N). Sections were assessed independently by two investigators (WAC and PYY), and a modified H-score (the product of percentage of positively staining cells and intensity) was calculated for each core. The final modified H-score of each case was the average from all available tumour cores. The concordance of pAkt score between the two investigators was high in both cytoplasmic (Spearman's correlation coefficient, r=0.93, p<0.001) and nuclear (r=0.89, p<0.001) compartments.
Statistical Analysis
To examine the relationship between the level of pAkt expression and the clinicopathological factors, Pearson's χ test was performed. Kaplan–Meier and log-rank analyses were performed on the cytoplasmic and nuclear pAkt H-scores in a stepwise fashion (ie, using a cut-off of 10). The cutpoints that resulted in the most significant relation with overall survival were selected. The optimal modified H-score cutpoints for pAkt expression were 70 in the cytoplasmic (≤70, low vs >70, high) and 40 (≤40, low vs >40, high) in nuclear compartments, respectively. The cohort was grouped into four categories (high C/low N, low C/low N, high C/high N and low C/high N) according to the level of pAkt expression in these two subcellular compartments. We understand that this will overestimate the magnitude of the effect, but this raises hypotheses that warrant corroboration in an independent dataset before it can be considered for routine clinical practice. Although the median follow-up was 11.3 years, survival was censored at 10 years as lung cancer-related death was unlikely >10 years postresection. Overall 10-year survival was calculated from the time of resection until the time of last follow-up or death within 10 years from resection. Lung cancer-specific mortality could not be assessed due to inconsistency in the source documentation.
Survival analysis was performed using univariable and multivariable analyses in a Cox proportional hazards model for pAkt expression and known clinicopathological prognostic factors. Age, gender, extent of operation, tumour size, histopathology, histological grade and visceral pleural invasion were selected for inclusion in the multivariable model because they have previously proven prognostic in NSCLC, regardless of their statistical significance in this dataset. All p values were two-sided, and a p value <0.05 was considered significant. All statistical analyses were performed using PASW Statistics V.18.0 (SPSS Inc, Chicago, Illinois, USA).
Source...