p-mTOR Expression in Luminal Breast Carcinoma
p-mTOR Expression in Luminal Breast Carcinoma
We analysed the expression of p-mTOR in 331 BC patients, and found an association between p-mTOR expression and tumour size and grade. Our results also demonstrated that p-mTOR expression is associated with a longer DFS and OS, in particular, in luminal BC.
Phospho-mTOR expression was considered positive in 43.8% of cases. Despite, different scoring methodologies and/or baseline population characteristics that limits comparisons, the prevalence reported has the same magnitude order of previously published studies. With the methodology used, our results do not confirm the previously described association between p-mTOR expression and lymph node metastasis, since no significant differences observed in lymph node status between p-mTOR-positive, and p-mTOR negative cases. Bakarakos et al described an association between p-mTOR expression and lymph node metastasis, a finding that led to the suggestion that activation of mTOR protein is related to a more aggressive phenotype. However, analysis of their results reveals that 81.9% of lymph node-negative BC cases have cytoplasmic expression of p-mTOR, contrasting with lymph node-positive BC cases with only 23% expressing p-mTOR. Moreover, in the same study, 62.8% of grade 1 carcinomas express p-mTOR (vs 31.9% of grade 2 and 23.4% of grade 3). Furthermore, it was recently demonstrated in patients with locoregional relapse that p-mTOR expression is associated with decreased risk of early relapse, and that high levels of p70S6K (a downstream target of mTORC1), are associated with prolonged OS in lymph node-positive patients. Others have demonstrated that p-mTOR-expression is more frequent in invasive BCs as compared to carcinoma in situ, or normal breast epithelium, yet, based on these results, one cannot assume that the percentage of p-mTOR-expressing tumours will be higher in advanced disease compared to localised invasive disease, or that higher mTOR expression is a surrogate for mTOR pathway activation. Taken together, these findings suggest that p-mTOR might be associated, at least in some BC subgroups, with a less aggressive disease. Mechanistically, it is well established that AKT in the PI3K-AKT-mTOR pathway is phosphorylated by the mTORC2 (which comprises mTOR) at Ser473, and that active AKT isoforms drive cell survival, proliferation and angiogenesis. Moreover, phosphorylation of S6K (by p70S6K) and 4EBP1 dependent on mTORC1 has similar effects;, thus, any mechanism that hampers the downstream effectors of phosphorylation by any of the mTORCs could eventually restrain oncogenic signalling, and lead to the accumulation of p-mTOR. It has also been shown that excessive mTOR signalling depletes stem cells and cancer stem cells, through the activation of fail-safe checkpoints, such as cellular senescence and apoptosis. Another hypothesis for the association of p-mTOR expression and favouble outcome could be the underliying existance of PI3KCA mutations. In fact, activating PI3KCA mutations are frequent in node-negative, ER-positive BC, and generally associated with favouble clinical outcome, however, no association has been found between exon 9 and exon 20 PI3KCA mutations and p-mTOR expression levels in treated ER-positive BC patients.
Despite the fair number of cases included in our analysis of the relationship between p-mTOR expression and associated outcomes, some important limitations remain. This was a retrospective, observational, single-centre-based study, which may limit the generalisability of our findings. We report p-mTOR expression association with outcome after controlling for some of the determinants of BC treatment (age, grade, size, lymph node status, hormone receptor status and HER2 status); however, we were incapable of determining if it was prognostic or predictive, since we were unable to study the possible recently suggested interaction between p-mTOR expression and chemotherapy. Note, that because p-mTOR expression was associated with smaller and lower-grade tumours, it was more likely to be detected in patients who had not undergone adjuvant chemotherapy; thus, even if data on adjuvant chemotherapy were available, it would be reasonable to expect a greater magnitude of the effect on outcome of p-mTOR expression. In particular, with regard to analysis of the luminal subgroup, we included HER2 overexpression as a covariate in the model, due to its predictor value, and because it was one of the defining criteria of the luminal B subgroup. Furthermore, we compared two groups of patients with some different baseline characteristics, but we considered that by the multivariable analysis performed we adjusted for these possible known confounders.
In summary, we show here that p-mTOR is associated with tumour size and grade, and p-mTOR expression in luminal breast tumours is associated with longer DFS and OS. To the best of our knowledge, this is the first study to examine the prognostic value of p-mTOR expression specifically in the luminal subgroup. Prospective studies are needed to confirm the independent association of p-mTOR with outcome, and to explore the role of p-mTOR as a prognostic marker in luminal BC.
Discussion
We analysed the expression of p-mTOR in 331 BC patients, and found an association between p-mTOR expression and tumour size and grade. Our results also demonstrated that p-mTOR expression is associated with a longer DFS and OS, in particular, in luminal BC.
Phospho-mTOR expression was considered positive in 43.8% of cases. Despite, different scoring methodologies and/or baseline population characteristics that limits comparisons, the prevalence reported has the same magnitude order of previously published studies. With the methodology used, our results do not confirm the previously described association between p-mTOR expression and lymph node metastasis, since no significant differences observed in lymph node status between p-mTOR-positive, and p-mTOR negative cases. Bakarakos et al described an association between p-mTOR expression and lymph node metastasis, a finding that led to the suggestion that activation of mTOR protein is related to a more aggressive phenotype. However, analysis of their results reveals that 81.9% of lymph node-negative BC cases have cytoplasmic expression of p-mTOR, contrasting with lymph node-positive BC cases with only 23% expressing p-mTOR. Moreover, in the same study, 62.8% of grade 1 carcinomas express p-mTOR (vs 31.9% of grade 2 and 23.4% of grade 3). Furthermore, it was recently demonstrated in patients with locoregional relapse that p-mTOR expression is associated with decreased risk of early relapse, and that high levels of p70S6K (a downstream target of mTORC1), are associated with prolonged OS in lymph node-positive patients. Others have demonstrated that p-mTOR-expression is more frequent in invasive BCs as compared to carcinoma in situ, or normal breast epithelium, yet, based on these results, one cannot assume that the percentage of p-mTOR-expressing tumours will be higher in advanced disease compared to localised invasive disease, or that higher mTOR expression is a surrogate for mTOR pathway activation. Taken together, these findings suggest that p-mTOR might be associated, at least in some BC subgroups, with a less aggressive disease. Mechanistically, it is well established that AKT in the PI3K-AKT-mTOR pathway is phosphorylated by the mTORC2 (which comprises mTOR) at Ser473, and that active AKT isoforms drive cell survival, proliferation and angiogenesis. Moreover, phosphorylation of S6K (by p70S6K) and 4EBP1 dependent on mTORC1 has similar effects;, thus, any mechanism that hampers the downstream effectors of phosphorylation by any of the mTORCs could eventually restrain oncogenic signalling, and lead to the accumulation of p-mTOR. It has also been shown that excessive mTOR signalling depletes stem cells and cancer stem cells, through the activation of fail-safe checkpoints, such as cellular senescence and apoptosis. Another hypothesis for the association of p-mTOR expression and favouble outcome could be the underliying existance of PI3KCA mutations. In fact, activating PI3KCA mutations are frequent in node-negative, ER-positive BC, and generally associated with favouble clinical outcome, however, no association has been found between exon 9 and exon 20 PI3KCA mutations and p-mTOR expression levels in treated ER-positive BC patients.
Despite the fair number of cases included in our analysis of the relationship between p-mTOR expression and associated outcomes, some important limitations remain. This was a retrospective, observational, single-centre-based study, which may limit the generalisability of our findings. We report p-mTOR expression association with outcome after controlling for some of the determinants of BC treatment (age, grade, size, lymph node status, hormone receptor status and HER2 status); however, we were incapable of determining if it was prognostic or predictive, since we were unable to study the possible recently suggested interaction between p-mTOR expression and chemotherapy. Note, that because p-mTOR expression was associated with smaller and lower-grade tumours, it was more likely to be detected in patients who had not undergone adjuvant chemotherapy; thus, even if data on adjuvant chemotherapy were available, it would be reasonable to expect a greater magnitude of the effect on outcome of p-mTOR expression. In particular, with regard to analysis of the luminal subgroup, we included HER2 overexpression as a covariate in the model, due to its predictor value, and because it was one of the defining criteria of the luminal B subgroup. Furthermore, we compared two groups of patients with some different baseline characteristics, but we considered that by the multivariable analysis performed we adjusted for these possible known confounders.
In summary, we show here that p-mTOR is associated with tumour size and grade, and p-mTOR expression in luminal breast tumours is associated with longer DFS and OS. To the best of our knowledge, this is the first study to examine the prognostic value of p-mTOR expression specifically in the luminal subgroup. Prospective studies are needed to confirm the independent association of p-mTOR with outcome, and to explore the role of p-mTOR as a prognostic marker in luminal BC.
