Investigations into how best to target FGFR2 mutant endometrial cancer

Endometrial cancer (EC) is the fourth most common cancer in women in developed countries, such as North America, Europe and Australia. Patients with low-grade, early-stage disease usually have a favourable survival rate. However, patients that present at an advanced stage of disease have an average survival of only 12 months. Current treatments for these patients are radiation and chemotherapy, which offer limited clinical benefit. There is no efficient treatment for advanced EC. Improved therapeutic approaches are needed for the treatment of recurrent and metastatic endometrial cancer. Recent advances in cancer biology have resulted in the development of molecular targeted therapies. The Fibroblast Growth Factors Receptor (FGFR) family and their ligands (fibroblast growth factors, FGFs) regulate a broad spectrum of physiological processes as well as tissue patterning and organogenesis during embryogenesis. Abnormally activated FGFRs have been identified in various cancers and are emerging as potential therapeutic targets. The Pollock laboratory and other groups have demonstrated that 10-20% of endometrioid ECs carry FGFR2 mutations that may be a novel therapeutic target in endometrial carcinoma. Preclinical studies show that inhibition of FGFR can inhibit EC cell growth in vitro. However, FGFR inhibitors are not as efficient at inhibiting tumour growth in vivo. We aim to find a way to improve the efficacy of FGFR inhibition in cancer treatment. About 90% of EC patients harbour genetic aberrations in the components of the PI3K/AKT pathway which indicates this signalling pathway plays an important role in the development of EC. Work from our lab demonstrates that inhibition of FGFR results in abrogation of MAPK activation in sensitive EC cells, however, the PI3K/AKT signalling pathway remains unaffected. PI3K/AKT signalling plays a vital role in cancer cell proliferation and survival, furthermore crosstalk between the MAPK and PI3K/AKT signalling pathways is associated with resistance to targeted therapies. Thus, the first aim of this study was to examine whether combination of the FGFR inhibitor (BGJ398) with various different PI3K inhibitors was synergistic in FGFRi sensitive EC cells. We present data that the combination of the pan-FGFR inhibitor (BGJ398) with pan-PI3K inhibitors (GDC-0941, BKM120) or a p110α-selective PI3K inhibitor (BYL719) was synergistic in inhibiting cell growth. Significantly more cell death and inhibition of long-term cell survival was observed in the combination treatments compared to each of the single drug treatments. Importantly, these effects could also be observed at lower concentrations. This study is the first to indicate that partial inhibition of the PI3K signalling pathway could significantly increase cell death when combined with the FGFR inhibitor BGJ398 in FGFR2 mutant EC cells. These data provide evidence that sub-therapeutic doses of PI3K inhibitors could enhance the efficacy of anti-FGFR therapies and a combination therapy may represent a superior therapeutic treatment in FGFR2 mutant EC patients. The in vivo work (conducted by Dr Vanessa Bonazzi) shows that the combination of BGJ398 and GDC-0941 and BYL719 resulted in tumour regression, while single drug treatment only slowed tumour growth. Interestingly, BYL719 alone resulted in increased tumour growth in tumour xenografts of AN3CA but not JHUEM2. In the first results chapter we further investigated the mechanism of enhanced cell death from the combination of BGJ398 and PI3K inhibitors. The activation of ERK and AKT has been inhibited by the combination of BGJ398 and PI3K inhibitors. However, the combination of the MEK inhibitor trametinib and the PI3K inhibitors induced less cell death than inhibition of the FGFR and PI3K signalling pathways. BGJ398 but not trametinib or GDC-0941 inhibited the activity of PLCγ1. We have also found trametinib up-regulated PLCγ1 activity, which is a novel finding in the field. We next employed several pharmacological inhibitors to investigate whether PLCγ1 is involved in the cell death observed following the combination of BGJ398 and GDC-0941 treatment. As there is no PLCγ1 inhibitor available currently, we used two different pan-PLC inhibitors, manoalide and U73122. Co-inhibition of the MAPK, PI3K/AKT and PLC signalling recapitulated cell growth inhibition seen with the combination of FGFR and PI3K inhibitor in both cell lines. Cell death induced by the combination of PLC inhibitors with trametinib and GDC0941 was similar as the combination BGJ398 and GDC0941 in AN3CA, but significantly less than the combination BGJ398 and GDC0941 in JHUEM2. Unfortunately, Western blotting was unable to show inhibition of PLCγ1 bringing into question whether these PLC inhibitors inhibited PLC function sufficiently, and whether the phenotypic effects of manoalide and U73122 when added to the trametinib and GDC0941 combination are due to inhibition of PLCγ1. The second results chapter reports efforts to identify the mechanism of intrinsic resistance to FGFR inhibition in EC cell lines carrying FGFR2 activating mutations but showing intrinsic resistance to FGFR inhibition (EI, EN1078D, and MFE319) with comparisons to the two sensitive EC cell lines (JHUEM2 and AN3CA). We have observed sustained activation of ERK in the resistant cells after treatment with an FGFR inhibitor, while ERK was inhibited in the sensitive cells. Inhibition of the MAPK signalling pathway could not sensitise the resistant cells to FGFR inhibition. Although several other receptor tyrosine kinases (RTKs) were hyperactivated in these cells, pharmacological inhibition did not show they were reliant on these RTKs. Co-inhibition of these kinases did not sensitise these cells to BGJ398. Knockdown of FGFR2 by shRNA in the sensitive cells induced moderate cell death, but limited cell death in the resistant cells. Interestingly, co-inhibition of the MAPK, PI3K/AKT and PLC signalling pathways has induced markedly less cell growth inhibition in the resistant cells compared to the sensitive cells, suggesting the resistant cells are less dependent on these central signalling pathways than the sensitive cells. Western blotting results showed that FGFR2 expression was considerably lower in the resistant cells than in the sensitive cells. Based on these results we have concluded that FGFR2 mutation status is not the only factor that determines sensitivity to FGFR inhibition, high expression of mutant FGFR2 is also important. This is a novel finding in the field and one which could guide patient select criteria in future clinical trials. Lastly, we show that FGFR2 knockdown in medium containing 10% FBS has little impact on downstream ERK phosphorylation whereas pan FGFR inhibition with BGJ398 could totally abrogate ERK phosphorylation. In cells grown overnight in serum starved conditions, FGFR2 knockdown did reduce downstream ERK phosphorylation but not to the same extent as pan FGFR inhibition in full growth medium. These data suggest that inhibition of FGFR2 alone is insufficient and that inhibition of multiple FGFRs will be more effective as a cancer treatment