Page 38 - Read Online

Lee et al. Cancer Drug Resist 2020;3:980-91 I http://dx.doi.org/10.20517/cdr.2020.73 Page 989

the acquired resistance to prexasertib in a rhabdomyosarcoma xenograft and innate resistance in Ewing’s
sarcoma and osteosarcoma cell lines. Acquired resistance was characterized by the activation of the PI3K
and MAPK pathways, but there was no characterization of the pathways involved in innate resistance.
[25]
More recently, Nair et al. observed acquired resistance in BRCA wild-type high-grade serous ovarian
cancers and attributed resistance to downregulation of CDK1/cyclin B1 prolonging G2 and delaying mitotic
catastrophe. Contributions of the PI3K and MAPK pathways were not examined.

While acquired resistance may contribute to the modest clinical response observed with prexasertib in
TNBC, innate resistance in TNBC tumors may also be a driving factor. We currently lack biomarkers
for intrinsic resistance to prexasertib in TNBC and other cancers. Examining a panel of BRCA wild-
type and BRCA mutated TNBC cell lines, we noted innate resistance to prexasertib in the MDA-468 cell
line [Figure 1]. Immunoblotting showed highly activated MAPK pathways in this cell line, similar to the
[8]
results reported by Lowery et al. for acquired resistance [Figure 2]. We also observed AKT activation, but
both the MDA-468 and MX-1 showed highly activated AKT, and they correspond with the most and least
resistance to prexasertib, respectively [Figures 1 and 2]. Critically, we also noted overexpression of EGFR
in the MDA-468 cell line, which is an upstream effector of the MAPK pathway and, to a lesser extent,
[8]
the PI3K pathway. Given that Lowery et al. observed no reversal of prexasertib resistance with MAPK
inhibitors, we examined EGFR’s role in innate prexasertib resistance.

Stimulation of EGFR by EGF in the MDA-231 cells resulted in increased resistance to prexasertib [Figure 3].
Inhibition of EGFR with erlotinib showed synergistic activity with prexasertib in MDA-231 and MDA-468
[Figure 4]. Immunoblotting of combination therapy showed that phosphorylation of BAD was
downregulated after erlotinib treatment, promoting the release of BAD from 14-3-3 and signaling of
[24]
apoptosis [Figure 5]. The importance of phosphorylated BAD was further confirmed in EGF stimulated
MDA-231 cells, where EGF increased phosphorylated BAD promoting resistance to prexasertib [Figure 6].

Tumor xenografts of MDA-468 confirm the synergistic activity of prexasertib and erlotinib in vivo,
although there was no apparent benefit to combination treatment in MDA-231 xenografts [Figure 7]. While
synergy between prexasertib and erlotinib was observed for MDA-231 in 3D culture [Figure 4], the lack of
synergy in the in vivo experiments is likely due to higher dosing of prexasertib and erlotinib in the in vivo
experiments. Erlotinib dosing for the xenograft models was selected based on existing literature and known
sensitivity of TNBC cell lines [22,23] . Prexasertib dosing was also selected based on existing literature . The
[8,9]
synergy of the 3D experiments and the high tumor reduction observed in our animal experiments suggest
future dosing experiments should be conducted to optimize the prexasertib to erlotinib ratios to minimize
side effects, while maintaining tumor reduction. Myelosuppression is commonly a dose-limiting factor for
CHK1 inhibitors in the clinical setting . We did not see significant toxicity or weight loss with prexasertib,
[1]
erlotinib, or combination treatment in either xenograft model [Supplementary Figure 2]. However,
further dose optimization of the combination treatment could reduce adverse effects while maintaining
[26]
tumor reduction. Our results are also consistent with those of Zeng et al. , who reported enhanced cell
killing and tumor targeting with cetuximab and prexasertib in head and neck squamous cell carcinoma.
Prexasertib resistance was not reported in their xenograft models.

EGFR overexpression has been noted in at least 50% of TNBCs, which is higher than other breast cancer
subtypes [17-19,27] . Our data indicate that activation or overexpression of EGFR contributes to innate resistance
to prexasertib in TNBC and may contribute to the modest clinical efficacy observed in phase I and II
trials. Stratification of prexasertib clinical data based on EGFR expression status could offer new insight
into the clinical use of prexasertib as a monotherapy for TNBC and other cancers. It should be noted that
EGFR overexpression is also observed in HCT116 and PANC-1 cell lines, which have an innate resistance
to prexasertib [6,28,29] . Ewing’s sarcoma and osteosarcoma also have a significant percentage of EGFR