NEW YORK – After a recent meeting with the US Food and Drug Administration, the Denmark-headquartered biotech Oncology Venture said it plans to file a new drug application for dovitinib as a third-line option for patients with renal cell carcinoma (RCC) who are deemed likely to benefit based on the company's proprietary drug response predictor, or DRP, tool.
Dovitinib, originally a product of Novartis, was once regarded as a failed therapeutic based on its inability to outperform sorafenib (Bayer's Nexavar) in clinical trials. But this time around, Oncology Venture is hoping that its predictive DRP tool can turn the drug's "failure" on its head.
"If and when we're successful, we'll be the first company able to utilize an algorithm-based AI tool as a diagnostic predictor in combination with an approved therapeutic that could treat a pretty challenging disease," said Oncology Venture CEO Steve Carchedi. The company plans to submit data on the drug and predictive tool at the same time, enabling the FDA to review dovitinib and the DRP simultaneously, according to Carchedi.
A recent meeting with the FDA was encouraging for Oncology Venture, though the company remained cautious in its statement about the interaction: "The FDA’s feedback provides guidance only and the review process is unpredictable and may or may not lead to a formal approval."
'Reinvigorating' dovitinib
When Oncology Venture bought the pan-tyrosine kinase inhibitor dovitinib from Novartis in 2018, the company was just beginning to transition into therapeutics from its previous standing as a pure diagnostic development company.
"Around 2017, 2018, we went on what I'll call a 'shopping spree' to identify compounds that had demonstrated levels of effectiveness for treating various cancers, but for whatever reason fell just short by way of statistical significance for their indications," said Carchedi.
Dovitinib, having demonstrated a median progression-free survival of 3.7 months versus 3.6 months for sorafenib, had not met its primary endpoint in its Phase III trial. But Oncology Ventures saw an opportunity in that more-or-less equivalent PFS result. While the drug hadn't proven itself any better than sorafenib, it had clinically demonstrated that it was safe, and that it did have activity in RCC. Oncology Venture, accordingly, bought dovitinib from Novartis with the goal of using its DRP to determine the patients for whom the drug would show a higher response rate.
Oncology Ventures "had confidence in our ability to reinvigorate drugs," said Carchedi. "We developed a DRP for the molecule, and what we found was that we could identify those patients that would likely respond with upwards of 80 percent accuracy."
'More than a diagnostic'
Oncology Venture's DRP tool, which Carchedi said will go "hand-in-glove" with dovitinib if approved, uses a proprietary algorithm-based AI technology to pinpoint which patients have a high likelihood of responding to the drug.
To develop the algorithm underlying the DRP tool for dovitinib, the company used cell lines treated with dovitinib, then trained the AI to pick up specific tumor biology and genomic tumor features associated with treatment response or resistance. The company did not disclose what, exactly, those specific tumor biology and genomic tumor features were.
In developing the DRP, Carchedi said, other information was used as well, such as gene expression from a large number of human tumors in their microenvironments. When applying DRP's AI algorithm to analyze tumor samples, the tool uses the expression of 58 genes that have been gleaned using DNA arrays.
To validate the DRP for dovitinib, Oncology Venture used anonymized biopsy data from Novartis' Phase III trial.
"It really is more than a diagnostic," said Carchedi. "I use the term diagnostic loosely, but it's really a drug-specific diagnostic platform."
If approved, Carchedi said, included on the label for dovitinib would be the fact that the drug is approved specifically for patients with renal cell carcinoma who are shown likely to respond based on the DRP screening tool. In practice, Carchedi explained, the DRP would be used on a patient's tumor sample and then roughly one week later, the requesting oncologist would receive a report containing a prediction as to whether a patient is likely to respond to dovitinib.
"At the end of the day, the FDA wants to narrow the patient population … and we're doing it with a tool that is not a basic tool for patients," Carchedi said.
Using data from more than 2,500 patients, Oncology Venture claims to have ultimately shown that the dovitinib DRP was able to identify responders, resulting in a two- to fourfold higher response rate among those deemed likely to respond by the DRP tool than among the group that had not been run through the DRP.
Approval on the basis of non-inferiority
In accordance with guidance from the FDA, Oncology Venture plans to package the DRP tool together with the data-mining results based on Novartis' Phase I, II, and III clinical trials when it files its NDA seeking approval for dovitinib based on non-inferiority compared with sorafenib.
"When we acquired the asset, [Novartis] had done all the heavy lifting that a company would have to go through in order to get a drug approved," said Carchedi. "We inherited all of that, so our clinical studies are done. Now, we are working on packaging the data on our DRP and the data on its activity with dovitinib to create a very robust package to file to the FDA."
Because the clinical data that the company will package within its NDA comes from the original trial comparing dovitinib and sorafenib, and that trial essentially showed that dovitinib was non-inferior to the comparator, this will become the basis on which Oncology Venture seeks approval. Although "non-inferiority" was not originally written into the protocol for the Novartis trial, Oncology Venture discussed with the FDA using PFS as an endpoint to seek approval based on non-inferiority. The FDA indicated that the NDA would be referred to the Oncologic Drugs Advisory Committee for review once submitted, and that it would not be necessary for the company to conduct any additional pre-clinical, safety, or clinical testing, for dovitinib.
According to Oncology Venture, dovitinib would address a significant unmet need for improved therapies in RCC. Globally, there are more than 342,000 new RCC diagnoses each year, and metastatic kidney cancer has a five-year survival rate of just 12 percent. Currently, sorafenib — possibly together with a PD-1 inhibitor — is standard of care for patients with stage III or IV RCC. Unlike other TKIs that exist as standard of care for advanced kidney cancer, including sorafenib, one of the targets that dovitinib inhibits is FGFR, shown to be associated with RCC progression.
Going forward, Oncology Venture plans to develop its DRP for use with other assets in its pipeline, too, including a PARP inhibitor, 2X-121, which is currently in Phase II clinical trials for patients with advanced ovarian cancer. Similar to how the company is packaging the DRP with dovitinib, Oncology Venture plans to refine its development of a specific 2X-121 DRP to predict patients who are highly likely to respond to the PARP inhibitor.