Using 3D tumor organoids, researchers have added to existing evidence that inhibition of kinesin spindle protein (KSP) may represent a broadly relevant target for the treatment of the aggressive childhood tumor neuroblastoma.

A high-throughput drug screen in tumor organoids derived from multiple neuroblastomas expressing MYCN, an oncogene which is strongly associated with poor prognosis, uncovered several drugs with previously unexplored anti-neuroblastoma activity.

These included GSK-923295, pevonedistat, daporinad, AVN-944, selinexor, BAY-87-2243, UCN-01, rigosertib sodium, KX2-391 (tirbanibulin), CUDC-101 and filanesib (ARRY-520).

Of these, Array Biopharma Inc.'s (now part of Pfizer Inc.) KSP inhibitor filanesib, was selected for in depth study, because of a known association between expression of KIF11, (kinesin family member 11) the gene that encodes for KSP, and clinical outcomes.

Filanesib was shown to reduce tumor viability across a number of in vitro and in vivo neuroblastoma models.

This is significant because high-risk neuroblastomas exhibit few recurrent and potentially druggable targets. Relapsed tumors have been shown to be enriched for genes that activate oncogenic signaling pathways, prompting a search for personalized treatments. However, neuroblastomas that have progressed following first line treatment show substantial heterogeneity, both from one to another, and within individual tumors.

"There has been a lot of focus on personalized treatments targeting specific mutations. There have been a few promising cases, but the heterogeneity means each mutation is only found in a fraction of tumors. You need something more general," said Daniel Bexell of the Translational Cancer Center at Lund University, Sweden.

Showing that KSP inhibition was effective in multiple models indicates it could represent such a target, Bexell said. "It's good to have shown such reproducibility. A lot of [other drugs] are shown to work in one model and not in others," he told BioWorld Science.

Bexell is coauthor of a paper in the September 23, 2020, edition of Science Translational Medicine, describing the screening of 525 approved and clinical-stage oncology drugs for possible repurposing to treat neuroblastoma.

Challenging, and rewarding

The 3D tumor organoids established by Bexell and colleagues are proving better discovery tools than conventional cancer lines. They retain neuroblastoma-associated chromosomal aberrations and protein markers, and are tumorigenic and form metastases when implanted into mice.

"They are much more interesting and relevant, even if more difficult to work with," said Bexell.

KSP, also known as EG5, is a motor protein that plays an essential role in mitosis. KIF11 is known to be overexpressed in many different cancers, and it has been shown that inactivating this gene leads to apoptosis and cell death in a number of tumors.

An analysis of transcriptomics databases carried out by Bexell and colleagues, found neuroblastoma is head of this list: of 22 tumor types which express the gene, neuroblastoma has the highest KIF11 gene dependency. Meanwhile, across 27 tumor types neuroblastoma is among the most sensitive to KSP inhibition.

While based on public data, Bexell said this is the first time anyone has reported a link between KIF11 expression and sensitivity of neuroblastoma to KSP inhibition.

As further confirmation of the possible relevance of this correlation, among neuroblastoma patient samples, KIF11 showed higher expression in high-risk patients compared to those with low risk tumors. It also had significantly higher expression levels in stage four tumors, than in less advanced cancers.

Using filanesib, the researchers investigated the cellular level impact of inhibiting KSP in neuroblastoma. They found it results in the formation of monastral spindles, mitotic arrest, upregulation of genes associated with mitotic processes and apoptosis. KIF11 plays no role in resting, nondividing cells and filanesib had no inhibitory effect on healthy donor cells.

Taken overall, the research shows powerful effects of KSP inhibition in vitro and in vivo, with complete tumor regression in one high-risk MYCN-amplified patient-derived xenograft (PDX) model, and minimal off-target effects.

"For the first time, we have a candidate drug which can [regress] tumors. We've not seen such powerful effects in a PDX model before, and it's a more general mechanism, targeting cells that are dividing," said Bexell.

However, it has been shown in other cancers that tumors can circumvent KSP inhibition by activating another member of the kinesin family, or through other signaling pathways.

"We're excited that the in vivo data is so powerful. But you see renewed tumor growth and relapse in some models. So we now want to find combinations," Bexell said. He is now planning to conduct a screen combining KSP inhibitors with other drugs (Hansson, K. et al. Sci Transl Med 2020, 12(562): eaba4434).