The Aurora family of serine/threonine protein kinases play a critical role in the regulation of chromosomal segregation and cytokinesis during mammalian cell division. There are several lines of evidence supporting Aurora kinase A as a cancer therapeutic target [11
]. First, AURKA
is amplified or overexpressed in many tumors [22
]. Second, the overexpression of AURKA results in the transformation of normal cells, supporting the hypothesis that Aurora A is an oncogene [25
]. Third, knockdown of AURKA using RNA interference technology results in mitotic spindle defects, mitotic delay, and apoptosis in human cells [26
]. Finally, two common polymorphisms in AURKA
(Phe31Ile and Val57Ile) have been shown to alter the enzyme’s kinase function, are associated with breast cancer susceptibility, [27
] and are preferentially amplified in some human cancers [28
]. To date, aurora kinase inhibitors have shown only limited clinical activity infrequently inducing objective responses in solid tumors, although showing more activity in leukemia patients [11
]. However, there are no data available to support AURKA as a therapeutic target in pediatric cancers.
MLN8237 is a small molecule reversible inhibitor of AURKA via competition with ATP binding that is being developed for the treatment of cancer. It is approximately 200-fold more selective for Aurora kinase A compared to Aurora kinase B, and is also relatively selective compared to other receptor and non-receptor kinases. Such an inhibitor would be expected to have potential application across a broad range of human tumors, given the central role of mitosis in the progression of virtually all malignancies, but perhaps with an increased therapeutic index when the gene is amplified and/or overexpressed as an acquired somatic event in cancer cells. MLN8237 has demonstrated activity against a broad range of adult preclinical tumor models, and is also expected to be toxic to proliferating normal tissues (such as bone marrow and GI epithelium) due to AURKA’s central role in mammalian cell division.
The PPTP is designed to prioritize agents being developed for cancer in general to the field of pediatric oncology. The results here show the potential power of this approach, as there remain no a priori
explanation for the broad and potent activity seen against both the ALL (including a CR in ALL-7 when tested at 0.5 × MTD) and neuroblastoma panels, as well as potent activity against some cell models in other histotypes. AURKA was not identified to date in any of the genomics efforts focused on leukemia or neuroblastoma as a potential therapeutic target. In addition, there appears to be no correlation between AURKA
copy number or expression with activity level in the xenografts studied here (data not shown, see [30
]). This suggests that oncogenicity of this kinase is determined by post-translational events and/or is cell-type specific. For example, Otto and colleagues discovered that AURKA binds MYCN and sequesters it from proteasomal degradation in a kinase independent manner [13
]. However, the neuroblastoma cell lines sensitive in this experiment were not only those with MYCN
amplification, and MLN8237 acts via inhibition of kinase function, suggesting that the MYCN (and perhaps MYC) binding function of AURKA potentially contributes to, but does not determine, cytotoxicity to MLN8237.
Development of MLN8237 for use in the pediatric cancer setting will require further work to identify MLN8237-based drug combinations with high-level activity. Preclinical testing of combinations that include an AURKA inhibitor have to date focused on adult cancer models. This early work suggests promising levels of activity when combining AURKA inhibition with microtubule-targeted agents such as taxanes and Vinca
The robust anti-tumor activity observed in this screen has been validated in a second stage of testing (manuscript under preparation) and led to the fast-tracking of this agent to the clinic. A pediatric phase 1/2 trial was designed and opened in the Children’s Oncology Group Phase 1 Consortium during the past year. This trial is exploring both once a day or twice daily dosing for 7 days, followed by 14 days of rest to recover from the anticipated myelosuppression. This is a substantially different dosing schedule than explored in the PPTP and was driven largely by the adult Phase 1 experience [34
]. As results from the clinical trial emerge, it will be important to correlate any observed anti-tumor activity with pharmacokinetic and pharmacodynamic measurements from both the human and murine experiments.