In total, 83 primary breast cancer patients treated with NAC at the First Hospital of China Medical University (Shenyang, China) between 2007 and 2010 were selected. Preoperative chemotherapy was performed as follows: 37 patients were administered docetaxel (75 mg/m²) with platinum (TP; 100 mg/m²) or cyclophosphamide (TC; 1.0 g) every three weeks for three to five cycles, while the other 46 patients received 5-fluorouracil (1.0 g), epirubicin (80 mg/m²) and cyclophosphamide (CEF; 1.0 g) every three weeks for three to four cycles. Patients were followed-up for a median of 45 months after their initial cancer surgery. Relevant clinical and pathological parameters are described in Table I. Archival formalin-fixed, paraffin-embedded breast tissues were collected from core biopsies (pre-NAC) and matched resection tissues (post-NAC). Six patients achieved pCR. All of the carcinomas had been histologically confirmed as invasive breast cancer according to the criteria of the World Health Organization (31) and the molecular subtypes of breast carcinoma were identified.

Immunohistochemical examination was performed on 4-μm formalin-fixed, paraffin-embedded sections. Briefly, following deparaffinization and rehydration, the endogenous peroxidase activity was blocked using 3% H₂O₂ (reagent A; UltraSensitive™ SP IHC kit; Maxim Biotech Inc., Fuzhou, China). Next, antigen retrieval was performed and normal serum was applied to the sections to block non-specific antibody binding (reagent B; UltraSensitive™ SP IHC kit; Maxim Biotech Inc.). Sections were then incubated overnight at 4°C with the primary antibodies. A goat anti-rabbit polyclonal antibody against mTOR (phospho S2448) was used for p-mTOR at a dilution of 1:500 (ab131538; Abcam, Cambridge, UK), and a goat anti-rabbit polyclonal antibody against eIF4EBP1 (phospho Thr36) was used for p-4EBP1 at a dilution of 1:200 (ab47365; Abcam). Following overnight incubation at 4°C, sections were incubated for 15 min with the secondary antibody solution (reagent C; UltraSensitive™ SP IHC kit; Maxim Biotech Inc.). The sections were then incubated with streptavidin-perosidase (reagent D; UltraSensitive™ SP IHC kit; Maxim Biotech Inc.) for 15 min and stained with 3,3-diaminobenzidine. Finally, sections were counterstained with hematoxylin for 5 min and mounted. Negative controls were processed with normal rabbit serum (Dako, Carpinteria, CA, USA) in place of the primary antibody. Positive controls were performed using breast cancer tissue sections that had shown strong staining for the respective protein during antibody optimization. This study was approved by the ethics committee of the First Affiliated Hospital (Shenyang, China) and written informed consemt was obtained from all patients.

The immunohistochemical staining results were evaluated and scored independently in a blinded manner by two pathologists. Cases of disagreement were reviewed jointly to obtain a consensus score. The score was the average of 10 distinct high-power fields observed under the 40× objective. The staining was considered positive when cytoplasmic and/or membranous staining was observed in the malignant cells, and the staining was evaluated using a semi-quantitative scoring system considering the extent and intensity. The percentage of cells stained were scored as follows: 0, no cells stained; 1, 1–10% of cells stained; 2, 11–50% of cells stained; 3, 51–80% of cells stained; and 4, >80% of cells stained. Staining intensity was scored as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. The two parameters were multiplied, resulting in an individual immunoreactivity score ranging between 0 and 12 for every case. The six patients who achieved pCR were regarded as negative post-NAC.

Statistical analyses were performed using SPSS v.19.0 (SPSS, Inc., Chicago, IL, USA). Wilcoxon signed-rank test was used to evaluate the independence between two groups of matched samples, and Mann-Whitney U test was used to assess the independence between two independent samples without any distribution assumption. Spearman’s correlation coefficients were used to reveal a correlation between two continuous variables. Receiver Operating Characteristic (ROC) curve analyses were used to select cut-off values (giving the highest combined sensitivity and specificity) to dichotomize pre- and post-NAC expression scores for the endpoint of DFS. DFS was recorded from the date of surgery to the date of relapse or last follow-up, and estimated using the Kaplan-Meier analyses. The statistical significance of the differential survival was assessed using the log-rank (score) test. Additionally, multivariate Cox regression analysis was performed taking into account the pre-NAC expression of p-mTOR and p-4E-BP1, and post-NAC expression of p-mTOR, as well as other clinicopathological features, including tumor grade, receptor status, pre-NAC tumor stage and axillary metastasis. All P-values presented are two-sided, and P≤0.05 was considered to indicate a statistically significant difference.

The expression of p-mTOR and p-4E-BP1 were examined in the tumors of 83 cases of breast cancer patients. The patients who had achieved pCR were not examined post-NAC. Tumors were obtained from diagnostic biopsies pre-NAC and surgical resections post-NAC. Table I shows the clinical and pathological features of the patient cohort. The staining of p-mTOR was predominantly cytoplasmic, and present in 71/83 cases (85.5%) and 58/83 cases (69.9%) pre-NAC (Fig. 1A) and post-NAC (Fig. 1B), respectively. The p-4E-BP1 was detected in the nucleus and cytoplasm, with positive rates of p-4E-BP1 in 69/83 cases (83.1%) and 58/83 cases (69.9%) pre-NAC (Fig. 1C) and post-NAC (Fig. 1D), respectively. The scores and their distributions are shown in Fig. 2. Scores were lower for p-mTOR (51/83 cases; 61.4%) and p-4E-BP1 (56/83 cases; 67.5%) in post-NAC samples than in the matched pre-NAC samples, indicating a decrease in their expression in response to NAC. Wilcoxon signed-rank test showed significant differences between pre- and post-NAC scores (P<0.001 for p-mTOR and p-4E-BP1). It was also examined whether a correlation exists between the expression of p-mTOR and p-4E-BP1 pre- and post-NAC, as well as for the expression change following treatment (pre-NAC minus post-NAC level). The expression of p-mTOR and p-4E-BP1 were found to significantly correlate with each other in pre-NAC samples (Spearman’s ρ analyses, P=0.004), which is consistent with previous studies (2,32,33). However, the two factors were not found to correlate with each other in post-NAC samples, indicating that chemotherapy may have changed the expression of the two factors to some degree. No significant correlations were found between the changes of p-mTOR and p-4E-BP1 following NAC. As would be expected, the pre- and post-NAC levels were significantly associated with the expression change of the respective factor.

Pre-NAC expression, post-NAC expression and the expression change following NAC of p-mTOR and p-4E-BP1 was evaluated to identify correlations with patient or tumor characteristics, including patient age at diagnosis, tumor grade and stage, estrogen receptor status, and HER2 status of tumors from core biopsy at diagnosis, as well as the tumor stage and presence of axillary metastases from resection pathology. Spearman’s ρ analyses were performed, and not only did the expression of p-mTOR and p-4E-BP1 in pre-NAC samples correlate with HER2 [ρ coefficient, 0.181 (P=0.05) for p-mTOR; ρ coefficient, 0.193 (P=0.04) for p-4E-BP1], which is consistent with the previous study, but the change of mTOR expression was also found to significantly correlate with HER2 (ρ coefficient, 0.275; P=0.006). Next, the changes of p-mTOR in HER2-positive and -negative groups were compared, and the expression of mTOR was found to decrease significantly following treatment with NAC in the HER2-positive group. The changes of p-mTOR expression had median values of 4 in the HER2-positive group and 1 in HER2-negative group. The Mann-Whitney U test was used to assess the differences between the two groups, and a P-value of 0.041 was obtained, suggesting some degree of cross-talk between the PI3K/Akt/mTOR-related pathways and HER2. The expression of p-mTOR and p-4E-BP1 pre- and post-NAC was also evaluated, as well as the expression change of the two factors to identify correlations with tumor size. Therefore, patients were classified into different groups according to tumor size; patients with tumors that had diminished by <1 cm, showed no change, or had increased were defined as group A, while patients whose tumor sizes had diminished by >1 cm, or patients who had achieved PCR were defined as group B. The only significant correlation was found between the expression change of mTOR and diminishing tumor size. The median values of expression change of mTOR were 2 in group A and 6 in group B (Fig. 3). The Mann-Whitney U test showed a significant difference between groups A and B (P=0.033). No significant correlations were found between the two factors and other clinicopathological features.

In order to assess the differential survival with respect to pre- and post-NAC expression, as well as the expression change following NAC for the two markers, Kaplan-Meier survival analyses were performed. ROC curve analyses were used to dichotomize the expression scores into high and low expression groups. The cut-off values were obtained from the highest combined sensitivity and specificity at the endpoint of DFS, and were as follows: p-mTOR, 8 and p-4E-BP1, 9 for pre-NAC; and p-mTOR, 7 and p-4E-BP1, 5 for post-NAC. A high expression of p-mTOR and p-4E-BP1 was found to significantly correlate with poor DFS pre-NAC (Fig. 4A and B, log-rank, P=0.013 for p-mTOR and P=0.025 for p-4E-BP1). The expression of p-4E-BP1 post-NAC was not significantly correlated with DFS (Fig. 4D). By contrast, the high expression of p-mTOR in post-NAC samples had a significant association with poor DFS compared with the pre-NAC samples (Fig. 4C, log-rank, P<0.001). It was also examined whether the expression changes of the factors correlate with DFS. Changes in expression of the two factors were dichotomized as up- or downregulated, but no significant correlation was found. In order to identify the predictive value of mTOR post-NAC, patients with high p-mTOR expression pre-NAC were defined as group A, while those with low p-mTOR expression pre-NAC were defined as group B. Next, the p-mTOR expression was compared between groups A and B post-NAC. High expression of p-mTOR post-NAC was found to correlate with poor DFS, regardless of whether the patients were in group A or B (P=0.043 for group A and P=0.006 for group B). Finally, multivariate Cox regression analysis was performed taking into account p-mTOR and p-4E-BP1 expression pre-NAC, p-mTOR expression post-NAC and other clinicopathological features, including tumor grade, receptor status, tumor stage pre-NAC and axillary metastasis. Post-NAC expression of p-mTOR was identified as the only significant factor, with its high expression associated with a hazard ratio of 3.073 (95% confidence interval, 1.4–6.8; P=0.006).