Preclinical evidence for anaplastic lymphoma kinase inhibitors as novel therapeutic treatments for cholangiocarcinoma

Introduction

Bile duct cancer (cholangiocarcinoma, CCA) has a poor prognosis for patients, and despite recent advances in targeted therapies for other cancer types, it is still treated with standard chemotherapy. Anaplastic lymphoma kinase (ALK) has been shown to be a primary driver of disease progression in lung cancer, and ALK inhibitors are effective therapeutics in aberrant ALK-expressing tumors. Aberrant ALK expression has been documented in CCA, but the use of ALK inhibitors has not been investigated. Using CCA cell lines and close-to-patient primary cholangiocarcinoma cells, we investigated the potential for ALK inhibitors in CCA.

Methods

ALK, cMET, and ROS1 expression was determined in CCA patient tissue by immunohistochemistry and digital droplet polymerase chain reaction, and that in cell lines was determined by immunoblot and immunofluorescence. The effect on cell viability and mechanism of action of ALK, cMet, and ROS1 inhibitors was determined in CCA cell lines. To determine whether ceritinib could affect primary CCA cells, tissue was taken from four patients with biliary tract cancer, without ALK rearrangement, mutation, or overexpression, and grown in three-dimensional tumor growth assays in the presence or absence of humanized mesenchymal cells.

Results

ALK and cMet but not ROS were both upregulated in CCA tissues and cell lines. Cell survival was inhibited by crizotinib, a c-met/ALK/ROS inhibitor. To determine the mechanism of this effect, we tested c-Met-specific and ALK/ROS-specific inhibitors, capmatinib and ceritinib, respectively. Whereas capmatinib did not affect cell survival, ceritinib dose-dependently inhibited survival in all cell lines, with IC50 ranging from 1 to 9 µM and co-treatments with gemcitabine and cisplatin further sensitized cells, with IC50 ranging from IC50 0.60 to 2.32 µM. Ceritinib did not inhibit cMet phosphorylation but did inhibit ALK phosphorylation. ALK was not mutated in any of these cell lines. Only ceritinib inhibited 3D growth of all four patient samples below mean peak serum concentration, in the presence and absence of mesenchymal cells, whereas crizotinib and capmatinib failed to do this. Ceritinib appeared to exert its effect more through autophagy than apoptosis.

Discussion

These results indicate that ceritinib or other ALK/ROS inhibitors could be therapeutically useful in cholangiocarcinoma even in the absence of aberrant ALK/ROS1 expression.

Looking Under the Lamppost: The Search for New Cancer Targets in the Human Kinome

The number of cancer drugs is increasing as new chemical entities are developed to target molecules, often protein kinases, driving cancer progression. In 2009, Fedorov et al. identified that of the protein kinases in the human kinome, most of the focus has been on a small subset. They highlighted that many poorly investigated protein kinases were cancer drivers, but there was no relationship between publications and involvement in cancer development or progression. Since 2009, there has been a doubling in the number of publications, patents, and drugs targeting the kinome. To determine whether this was an expansion in knowledge of well-studied targets-searching in the light under the lamppost-or an explosion of investigations into previously poorly investigated targets, we searched the literature for publications on each kinase, updating Federov et al.'s assessment of the druggable kinome. The proportion of papers focusing on the 50 most-studied kinases had not changed, and the makeup of those 50 had barely changed. The majority of new drugs (80%) were against the same group of 50 kinases identified as targets 10 years ago, and the proportion of studies investigating previously poorly investigated kinases (<1%) was unchanged. With three exceptions [p38 mitogenactivated protein kinase (p38a), AMP-activated protein kinase catalytic α-subunit 1,2, and B-Raf proto-oncogene (BRAF) serine/threonine kinase], >95% of publications addressing kinases still focused on a relatively small proportion (<50%) of the human kinome independently of their involvement as cancer drivers. There is, therefore, still extensive scope for discovery of therapeutics targeting different protein kinases in cancer and still a bias toward well-characterized targets over the innovative searchlight into the unknown. SIGNIFICANCE STATEMENT: This study presents evidence that drug discovery efforts in cancer are still to some extent focused on a narrow group of well-studied kinases 10 years after the identification of multiple novel cancer targets in the human kinome. This suggests that there is still room for researchers in academia, industry, and the not-for-profit sector to develop new and diverse therapies targeting kinases for cancer.