There is currently no standard first line chemotherapy in advanced or recurrent disease [1, 8, 10–15, 28, 29, 33]. Although Phase II trials have demonstrated activity with doxorubicin, paclitaxel and platinum compounds, only paclitaxel-containing regimens have consistently shown a response rate (RR) >20 % in advanced disease [1, 45, 46]. Consideration of alternative schedules of chemotherapy may help to improve response rates, with weekly carboplatin-paclitaxel being the most promising with RR up to 71 % in one study and an acceptable toxicity profile [47–50]. Combination chemotherapeutic regimens are generally more active in endometrial cancer than monotherapies, but this is at the expense of greater toxicity. One systematic review of eight randomized controlled trials (RCT) showed that treatment consisting of ‘more’ chemotherapy was associated with longer OS and progression free survival (PFS) [29] although no definitive recommendation could be made for a specific regimen [8, 29]. The GOG177 trial showed that combination of cisplatin/paclitaxel/doxorubicin had significantly greater RR (57 % vs 34 %) and OS (15.3 months vs 12.3 months) compared to doxorubicin/cisplatin but the utility of this regimen is compromised by significant toxicity [51]. Three meta-analyses conclude there is currently no statistically significant evidence to suggest one particular doublet chemotherapy over any other doublet, or a particular single-agent chemotherapy over another [8, 28, 29]. However, data for making comparisons is limited [8, 28, 29]. In view of the substantial side-effect profile of three drug regimens, recurrent endometrial cancer is most commonly treated with a combination of carboplatin/paclitaxel and less commonly, a doxorubicin-containing doublet [40].

A systematic review of the use of endocrine therapy for the management of endometrial cancer concluded that there is insufficient evidence that hormonal treatment improves the survival of patients with advanced or recurrent disease [9]. However, the authors concede the conclusions are limited by the small patient numbers on RCT which limit the ability to prove a significant benefit [9]. GOG153, a Phase II trial evaluating patients receiving MPA (medroxyprogesterone acetate) alternating with tamoxifen, reported meaningful activity where RR was observed to be 27 % and OS 14 months [52].

Preliminary data for several molecular targeted agents in endometrial cancer are emerging [14, 15, 19, 21, 33, 40, 41]. Although a number of established drug targets are expressed in endometrial cancer, therapeutic targeting of several of these in an unselected endometrial cancer population has not yielded much activity. The epidermal growth factor receptor (EGFR) is frequently expressed in normal endometrium as well as endometrial cancer [53] but use of the EGFR inhibitor erlotinib was associated with a RR of only 13 % [54]. Similarly, although HER-2 is overexpressed or amplified in a proportion of endometrial cancers [55], there were no objective responses to trastuzumab in a phase II trial of a selected population of HER-2 positive endometrial cancer patients [56]. Evidence suggests that angiogenesis and vascular endothelial growth factor (VEGF) signaling have a key role in endometrial cancer progression [57]. Two Phase II clinical trials have shown very encouraging activity of bevacizumab monotherapy in advanced or recurrent endometrial cancer [58, 59] making this one of the most promising novel targets for monotherapy in endometrial cancer. Several trials with bevacizumab in endometrial cancer are continuing [40]. A Phase II trial of the antiangiogenic agent sorafenib, an oral, multitargeted tyrosine kinase inhibitor showed disappointing activity in the endometrial population [60]. More recent trials are tending to select an appropriate patient population based on prospective mutational status of a molecular target, e.g. FGFR2 (dovitinib) or PI3K (PF-04691502) when targeting specific pathways (Table 31.2), a practice that is likely to increase.

Of all drugs targeting the PI3K pathway, the first generation mTor inhibitors (rapamycin analogues) have been the most extensively evaluated in endometrial cancer. Second generation compounds targeting mTor catalytic function through the mTor kinase domain, inhibit mTorc1/mTorc2 simultaneously and demonstrate greater potency [87–89]. Dual targeting of mTorc1 and mTorc2 circumvents the unwanted positive feedback loop where uninhibited mTorc2 stimulates AKT activation as can occur with the rapalogs [68, 69].

Temsirolimus [90], ridaforolimus [44, 91, 92] and everolimus [43] have been evaluated as monotherapy in unselected patients with inoperable advanced or metastatic endometrial cancer. Objective response rates were 14 % (temsirolimus [90]) and 4.4–7.7 % (ridaforolimus [91, 92]) with no objective response obtained with everolimus [43]. Many patients within these Phase II studies, however, benefited from long term disease stabilisation and those with no prior therapy for advanced disease benefited more [43, 91, 92]. A randomized Phase II ridaforolimus trial demonstrated a statistically significant benefit in progression free survival for ridaforolimus over more standard treatment [44]. Common adverse events were fatigue, rash, mucositis, hyperglycaemia (defined as on-target toxicity) and pneumonitis with hypertriglyceridaemia being less common [43, 44, 90–92]. Neither PTEN nor PIK3CA status correlated with response to temsirolimus [90] and analysis is ongoing for ridaforolimus [92]. Studies examining the relationship of PTEN and PIK3CA mutation as predictors of response to everolimus also found no correlation with outcome [93, 94].

The mechanism of the greater efficacy and potency of second generation mTOR catalytic site inhibitors may be due to their ability to inhibit the rapamycin-resistant mTorc1 functions rather than the additional inhibition of mTorc2 [95, 96]. It was established, however, that inhibition of mTorc2 prevented phosphorylation of Akt [96]. However, these agents had minimal effects on the phosphorylation state of several Akt substrates despite effectively inhibiting Akt S473 phosphorylation, suggesting they may not disable all components of Akt signalling [96, 97]. Other pathways of activation may remain unblocked allowing feedback activation still to occur which may result in hyperactivation of Akt-independent effectors of PI3K signalling [96–98]. mTor kinase inhibitors in early phase trial for solid tumours that include endometrial cancer are AZD2014, OSI-027 and INK128 (clinicaltrials.gov).

PI3K inhibitors can be divided into isoform-specific inhibitors or pan-PI3K inhibitors which target all four Class I PI3Ks. Isoform-specific inhibitors are potentially more tumour-specific and have a better side effect profile due to their specificity. For example, p110α[alpha]-specific inhibitors may effectively shut off PI3K–Akt signalling in cancers with PIK3CA mutations [98]. A p110β[beta]-specific inhibitor has shown promise in preventing PI3K signalling in PTEN-deficient cancers [99–102]. However, even in cancers that seem to be specifically reliant on either p110α[alpha] or p110β[beta], there is the concern that other non-targeted p110 isoforms might eventually compensate for decreased activity of the targeted isoform [98]. Given the prevalence of PTEN loss and PIK3CA mutation in endometrial cancers, the isoform-specific inhibitors, e.g. INK1117, NVP-BYL719 and GSK2636771 may be beneficial in this patient population. However, this depends on these genetic mutations predicting for response which has not been the case for mTor inhibition [90, 93, 94].

Phase 2 clinical trials are ongoing with the pan-PI3K inhibitors NVP-BKM120 and XL147 in endometrial cancer (clinicaltrials.gov). The side effect profile of these agents has been tolerable in Phase I studies mainly with nausea, vomiting and g-i upset and manageable hyperglycaemia [103, 104]. The results from the endometrial cancer trials are awaited with interest. Consistent with its mechanism of action, NVP-BKM120 decreases the cellular levels of p-Akt and preferentially inhibits tumour cells bearing PIK3CA mutations, in contrast to either KRAS or PTEN mutant models [105]. NVP-BKM120 behaves synergistically when combined with either targeted agents such as MEK or HER2 inhibitors or with cytotoxic agents such as docetaxel or temozolomide [105]. These findings may impact future development of this class of drugs.

Most of the dual PI3K-mTOR inhibitors target the Class I PI3K isoforms (p110α[alpha], p110β[beta] and p110δ[delta] predominantly) as well as mTORC1 and mTORC2 in order to maximally inhibit PI3K pathway signaling. This approach may also minimize the feedback activation of PI3K signaling observed with mTorc1 inhibitors and generate greater therapeutic benefit [68, 106]. This class of inhibitors might also be effective in cancers with Akt mutations or amplifications [98]. The potential advantages of these compounds over other agents targeting the PI3K cascade are being established as the drugs progress through clinical development.

PF-04691502 and PF-05212384 are dual PI3K/mTor inhibitors being investigated in a second line Phase 2 trial in endometrial cancer. In PIK3CA-mutant and PTEN-deleted cancer cell lines, PF-04691502 reduced phosphorylation of AKT inhibited cell proliferation via mTORC1 activity inhibition [107]. In the first in human, Phase I studies both PF-04691502 and PF-05212384 had been well tolerated with the most common treatment-related adverse events being fatigue, nausea, vomiting, decreased appetite, rash, and hyperglycaemia, similar to pan-PI3K inhibitors [108, 109]. GDC-0980 is another dual PI3K/mTor inhibitor [110] found to be tolerable in Phase I [111], also being evaluated in endometrial cancer with results awaited (Table 31.2).

Although Akt mutation is uncommon in endometrial cancer, this target is being pursued in this indication with MK2206 (Table 31.2). Cancers with AKT1 mutations and AKT1 and AKT2 amplifications might be expected to be more sensitive to Akt inhibitors [98]. MK2206 is a potent inhibitor of Akt 1, 2 and 3, with broad preclinical antitumor activity and tolerability as well as clinical activity in Phase I trial [112, 113]. Common drug-related toxicities include hyperglycemia, skin rash, nausea, fatigue, and diarrhoea [112].

In an attempt to reduce resistance and enhance efficacy of the various drugs targeting the PI3K pathway, trials investigating combinations with mechanistically diverse agents are underway (clinicaltrials.gov, Table 31.2). The main limitation for this approach is increased toxicity. The role of overactive PI3K signaling in chemoresistance supports the combination of PI3K pathway targeting agents with chemotherapeutics (temsirolimus, carboplatin, paclitaxel) as well as endocrine therapy (letrozole and everolimus) [19]. In addition, combinations of mechanistically complementary agents both with activity in endometrial cancer for example bevacizumab and temsirolimus, are being evaluated.

Despite a high incidence of molecular disruption in the PTEN/PI3K/AKT/mTor axis in endometrial cancers, only a small percentage of patients have responded to this class of drugs. One possible reason for this is the lack of patient selection and a tendency towards including patients with mutations in this pathway in clinical trials of PI3K/AKT/mTOR inhibitors has gained some momentum [115]. However, even though these are targeted agents, it is not clear which patients are most likely to benefit as the presence of PIK3CA mutation and PTEN loss have not so far been predictive of clinical benefit [90, 93, 94]. One recent report suggests a subtype of PIK3CA mutation, H1047R, may be predictive of response to drugs targeting this pathway but warrants further investigation [116]. This is likely to be a reflection of the complexity of the signalling networks and the multiplicity of feedback loops within and around the pathway [98]. Possibly, these drugs should be investigated in the first line/adjuvant setting, before resistance is established.

Further understanding of the molecular events and drivers in endometrial cancer through initiatives such as The Cancer Genome Project (TCGA) will help to inform biomarker development in order that patients can be appropriately selected for treatment [17, 18]. Validating assays for biomarker assessment and finding cost-effective measures to roll-out testing of samples in a directed way is becoming more important.

The specific toxicity of these drugs has limitations within this patient population. Whilst on the whole they are fairly well-tolerated and orally administered, the potential for hyperglycaemia will be problematic for a proportion of endometrial cancer patients due to pre-existing insulin resistance associated with obesity. Optimal absorption of oral preparations may be impeded by intra-abdominal disease. Other considerations are general ill-health of patients presenting with advanced disease due to other co-morbidities [32]. If used in combination regimens with other drugs toxicity may be enhanced.