HER-2 gene amplification and/or protein overexpression has been described in 10–34 % of breast cancers and is both a prognostic and a predictive factor. HER-2-positive breast tumors are associated with pathologic and clinical characteristics such as high cell proliferation, cell motility, tumor invasiveness, high probability of progressive regional and distant metastases, accelerated angiogenesis, and reduced apoptosis. When compared to endocrine-responsive disease, they have a higher grade and are diagnosed more often with lymph node metastases. In 107 studies considering 39,730 patients, these cancers were a negative prognostic factor independently of other prognostic variables [54]. This was, however, before the advent of trastuzumab and subsequent anti-HER-2 therapies.

Since the first phase 2 clinical experience in 1996 [55], trastuzumab, an anti-HER2 recombinant humanized monoclonal antibody, has become, in combination, the mainstay of treatment of HER-2-positive breast cancer in the metastatic [56] and adjuvant settings [57–60]. However, despite the dramatic improvement in prognosis of patients treated with adjuvant trastuzumab, as many as 7–20 % of patients still present with a breast cancer relapse. Other anti-HER-2-targeted agents have been approved by the regulatory agencies for the treatment of metastatic HER-2-positive breast cancer: lapatinib, a tyrosine kinase inhibitor of HER-2 [61]; pertuzumab, an anti-HER2 humanized monoclonal antibody that inhibits receptor dimerization [62]; and T-DM1, an antibody–drug conjugate incorporating the HER2-targeted antitumor properties of trastuzumab with the cytotoxic activity of the microtubule inhibitory agent DM1 [63]. Pertuzumab has also received FDA approval in the neoadjuvant setting for locally advanced, inflammatory, or early stage breast cancer in combination with trastuzumab and chemotherapy. Other agents are in clinical testing such as neratinib [64] and afatinib [65]. Some tumors are still able to evade inhibition prompting the search for combinatorial approaches to reverse resistance.

Different mechanisms of resistance to anti-HER-2 therapies are hypothesized in the treatment of HER-2 positive breast cancer: increased expression of p95HER-2, a truncated HER-2 receptor with constitutive activity; increased HER-2 expression due to HSP90 (heat shock protein 90) overexpression, disrupted antibody–receptor interaction, and failure to elicit an immune response; increased signaling through other growth factor receptors (IGF-1R, VEGFR, MET); alterations in downstream molecules (PTEN downregulation, PIK3CA mutations, increased Akt signaling); and increased cell survival due to telomerase expression [66].

A large amount of preclinical and clinical data demonstrates the role of the PI3K-AKT-mTOR pathway in the resistance to anti-HER-2-targeted agents. This pathway is involved in both the de novo and acquired resistance mechanisms. In vitro, loss of PTEN as well as PIK3CA mutations were associated with resistance to trastuzumab [67] and lapatinib [68, 69]. Patients with PTEN loss or PIK3CA mutations treated with trastuzumab had a shorter PFS [70]; and in another study, patients with PTEN loss had a lower response rate [71]. Acquired resistance was seen in three in vitro trials: Akt-negative feedback loop that perpetuated HER-2 phosphorylation leads to a decreased response to trastuzumab [72], while modifications in the pathway and the balance between phosphorylated and non-phosphorylated PTEN after exposure to HER-2-targeted therapy lead to acquired resistance to trastuzumab [67, 73].

Robust preclinical experiments support the restoration of trastuzumab sensitivity with the combination of mTOR inhibitors and trastuzumab. Everolimus restored the sensitivity of trastuzumab in an in vitro model of breast cancer with PTEN loss, and the combination inhibited tumor growth in a mouse xenograft model more than either agent alone [74]. The same experiment yielded the same results in vivo and in vitro with the treatment of trastuzumab-resistant models with the combination of trastuzumab and rapamycin or everolimus [75].

Everolimus and trastuzumab is the combination that is most advanced so far in clinical development. A pooled analysis combined data from two phase 1b/2 clinical trials that enrolled women with HER-2+ metastatic breast cancer after progression on trastuzumab-based therapy. Forty-seven patients were treated with trastuzumab every three weeks and daily everolimus at two doses, 5 and 10 mg. Seven partial responses (15 %) were recorded, and persistent stable disease was seen in nine patients (19 %) for a clinical benefit rate of 34 % in patients deemed resistant to trastuzumab, 56 % of whom had relapsed within 1 year of completing adjuvant trastuzumab. The combination was found to be tolerable with 9 % of patients presenting with grade 3 stomatitis, 9 % with grade 3 diarrhea, 13 % with grade 3/4 hyperglycemia, and 9 % with grade 3 fatigue. No cardiac toxicity was recorded. Hematological toxicity included grade 3 neutropenia in 9 % of patients and grade 3 thrombocytopenia in 4 %. Dose reductions/delays occurred in 25 patients (53 %). Everolimus 10 mg a day was the recommended dosage [76].

Everolimus combined with weekly paclitaxel and trastuzumab showed very encouraging antitumor activity in a phase 1b trial enrolling 33 patients with HER-2+ metastatic breast cancer, 31 of whom were pretreated with taxanes and 32 were resistant to trastuzumab. Everolimus was tested at three dosages (5 mg/day, 10 mg/day, or 30 mg/week) in combination with paclitaxel 80 mg/m² on days 1, 8, and 15 every 4 weeks and trastuzumab 2 mg/kg weekly. Overall response rate was 44 %, and the disease was controlled for 6 months or more in 74 % of patients. Overall response rate in 11 patients resistant to both taxanes and trastuzumab was 55 %. There were three recorded DLTs: febrile neutropenia (5 mg/day), stomatitis (10 mg/day), and confusion (30 mg/week). Grade 3–4 neutropenia was observed in 17 patients (52 %), and everolimus 10 mg a day was the recommended dosage for further development [77]. Another phase 1 trial tested the tolerability of everolimus (5 mg/day, 20 mg/week, or 30 mg/week) combined with vinorelbine (25 mg/m² on days 1 and 8 every 3 weeks) and trastuzumab (2 mg/kg weekly) in 50 women with heavily pretreated HER-2+ metastatic breast cancer progressing after trastuzumab-based treatment. Encouraging antitumor activity was also seen in this setting with an overall response rate of 19 %, a disease control rate of 83 %, and a median PFS of 31 weeks. As with paclitaxel, the most common adverse event was grade 3/4 neutropenia, and other DLTs included febrile neutropenia, grade 3 stomatitis with concomitant fatigue, grade 2 stomatitis, grade 3 anorexia, and grade 2 acneiform dermatitis. Everolimus 5 mg/day and 30 mg/week were chosen as the recommended dosages [78].

On the basis of these results, phase 3 trials with these combinations were designed. The BOLERO-1 trial is a phase 3 trial randomizing patients with metastatic HER-2-positive breast cancer in the first-line setting to paclitaxel + trastuzumab + everolimus (10 mg daily)/placebo. It has been completed and results are awaited [79]. The BOLERO-3 trial enrolled HER-2+ metastatic breast cancer patients pretreated with taxane and trastuzumab resistant. A total of 569 patients were randomly assigned to receive either 5 mg everolimus daily with 25 mg/m² vinorelbine weekly and 2 mg/kg trastuzumab weekly (284 patients) or placebo plus the same vinorelbine and trastuzumab regimen (285 patients). A total of 27 % of patients in each group had received prior lapatinib. The everolimus group had a median PFS of 7.00 months compared with 5.78 months in the placebo group (HR 0.78, 95 % CI [0.65, 0.95]; p = 0.0067). The overall response rate (complete or partial response) was 40.8 % in the everolimus group and 37.2 % in the placebo group (p = 0.2108), and the clinical benefit rate (objective response or stable disease at 24 weeks or more) was also not significantly different between the groups (59.2 % everolimus vs. 53.3 % placebo; p = 0.0945). Quality-of-life measures also showed no differences between the treatment arms [80]. Class effect adverse events associated with mTOR inhibitors (e.g., stomatitis, rash, noninfectious pneumonitis, and hyperglycemia) occurred more frequently in the everolimus arm, and most were grade 1/2. The incidences of serious adverse events suspected to be treatment-related were 26.4 % in the everolimus arm and 6.4 % in the placebo arm. Grade 3 class effect adverse events in the everolimus arm each occurred in less than 15 % of patients: (stomatitis, 13 %; hyperglycemia, 4 %) [81]. The most common grade 3–4 adverse events were neutropenia (204 [73 %] of 280 patients in the everolimus group vs 175 [62 %] of 282 patients in the placebo group), leucopenia (106 [38 %] vs 82 [29 %]), anemia (53 [19 %] vs 17 [6 %]), febrile neutropenia (44 [16 %] vs 10 [4 %]), stomatitis (37 [13 %] vs 4 [1 %]), and fatigue (34 [12 %] vs 11 [4 %]). Serious adverse events were reported in 117 (42 %) patients in the everolimus group and 55 (20 %) in the placebo group [82]. Although the primary endpoint of the trial was met—with a statistically improved PFS for the everolimus arm—the magnitude of improvement is disappointing: 6 weeks only. It should be noted that the trial enrolled women with hormone receptor-negative as well as hormone receptor-positive disease: one may wonder whether, for the latter, the ER pathway did not provide an “escape” mechanism. Even though this trial confirmed the proof of concept established in earlier trials, the toxicity profile and the “limited” efficacy could be obstacles for regulatory approvals in comparison with the more favorable safety and efficacy data of agents such as pertuzumab and T-DM1.

Evidence also emerged from the neoadjuvant setting with the RADHER trial in which 82 patients with HER-2+ early breast cancer were randomized to a short course (6 weeks) of preoperative treatment with trastuzumab or the combination of trastuzumab and everolimus. The combination improved the clinical response rate (35 % versus 22.5 %) [83].

Everolimus and other mTOR inhibitors have been tested in phase 1 and 2 trials in combination with trastuzumab or with other anti-HER-2 agents (Table 4.3) [84–89].

Triple negative breast cancer (TNBC) refers to a subgroup of breast tumors that do not express ER, PR, and HER-2. Almost 15 % of breast cancers are classified as TNBC according to this definition, but recent experiments using gene expression profiling have led to further subtyping into six new groups: basal-like 1, basal-like 2, immunomodulatory, mesenchymal, mesenchymal stemlike, and luminal androgen receptor [90]. While the treatment of TNBC has been challenging because of the heterogeneity of the disease and the absence of well-defined molecular targets, this development is considered a landmark on the path to discovering new drug targets. Indeed, despite some sensitivity to systemic chemotherapy, mainly taxane and anthracycline-based regimens, patients with TNBC are generally at a greater risk of early systemic relapse and poorer survival than patients with ER+ or HER-2+ breast cancer [91, 92]. These cancers are characterized by a low correlation between the size of the primary and the metastatic potential, rapid growth, and frequent occurrence in young women, thus evading screening detection and higher likelihood of metastasizing to viscera (mainly the lung and the brain) [93]. New targeted agents are therefore an unmet need against a very lethal and heterogeneous subtype of breast cancer.

There are multiple candidate targets and pathways in TNBC [94]:

Phosphatase and tensin homologue (PTEN) is a protein that inhibits activation of the AKT/mTOR pathway, and PTEN losses have been observed in up to 30 % of TNBCs [95]. This aberration has been associated with activation of AKT in TNBC samples [96]. Furthermore, the Cancer Genome Atlas Network analyzed 825 primary breast cancers by genomic DNA copy number arrays, DNA methylation, exome sequencing, messenger RNA arrays, microRNA sequencing, and reverse phase protein arrays. In addition to identifying nearly all genes previously implicated in breast cancer, a number of novel significantly mutated genes were identified. The TNBC subtype was found to have the highest mutation rate, albeit that those mutations were more diverse in the other subtypes. TP53 mutations (80 %) were the most common mutation followed by PIK3CA mutations (9 %). The data showed, however, that PI(3)K pathway activity, whether from gene, protein, or high PI(3)K/AKT pathway activities, was highest in basal-like cancers: loss of PTEN and INPP4B and/or amplification of PIK3CA [97]. There is therefore a rationale to develop mTOR inhibition in patients with TNBC that show PTEN loss.

mTOR inhibitors may also play a role in rational combinations as well as chemotherapy sensitizers. EGFR inhibitors, once regarded as promising agents in the treatment of metastatic TNBC, have failed to impact outcome when administered as single agents [98]. TNBC cell lines and nude mice models were treated with co-inhibition of mTOR and EGFR using rapamycin and lapatinib, respectively. This combination was synergistic in decreasing cell survival and resulted in increased apoptosis in some TNBC cell lines and was associated with the downregulation of rapamycin-induced activation of Akt in vitro [99]. The authors concluded that mTOR inhibitors could improve the efficacy of EGFR-targeting agents in the treatment of some metastatic breast cancers.

Fifteen patients with metastatic breast cancer, including 15 % with TNBC, were enrolled in a phase 1b trial testing the combination of everolimus and erlotinib [100]. Unfortunately, all but one patient progressed at the first disease evaluation, and future development of this combination was abandoned.

Given that TNBCs are characterized by a deficiency in the DNA repair machinery, DNA alkylating chemotherapy is hypothesized to be particularly effective in this setting. Cisplatin monotherapy achieved a pathologic complete response in 22 % of TNBC patients treated in the neoadjuvant setting [101]. mTOR activation could be one mechanism of resistance to cisplatin, and the addition of everolimus to cisplatin increased its in vitro efficacy fivefold [102]. Fifty-five patients with heavily pretreated metastatic ER− breast cancer (62 % with a median of three prior lines) were treated in a phase 2 trial with weekly cisplatin (25 mg/m²), paclitaxel (80 mg/m²), and daily everolimus (5 mg), given on a 28-day cycle. Sixty-three percent of patients had triple-negative disease, and 81 % patients had visceral disease. Significant antitumor activity was recorded: 11 patients had a partial response and 21 had stable disease. The regimen was tolerable and toxicity was mainly hematological: 24 % grade 3/4 neutropenia and 13 % grade 3 anemia [103]. This combination is currently studied in a randomized neoadjuvant trial with paclitaxel and cisplatin with or without everolimus in stage 2 and 3 TNBC [104]. Another ongoing trial randomizes patients with residual disease after neoadjuvant anthracycline and taxane chemotherapy to everolimus or placebo. However, this trial enrolls breast cancer patients with any subtype [105].