Biomarkers Speeding Cancer Drug Development

By Malorye Allison

Many types of companies are betting biomarkers will accelerate drug development, but the biggest frontier for their use is oncology. Now, a new wave of compounds, such as Exelixis’ XL880, Avalon’s AVN944, and SuperGen’s MP470, are starting to demonstrate when and how such markers are the most useful.

Exelixis’ MET and VEGFR2 Inhibitor

At Exelixis, "We do things a bit differently," says Michael Morrissey, President of Research and Development. Besides using biomarkers, the company’s scientists also have a screening approach they say delivers high-potency molecules earlier.

"We built a large compound library that helps us find molecules with low nanomolar activity right away," says Morrissey. "With a smaller library you get weaker potency and then spend a year to 18 months improving on that." This screening strategy lets them focus on optimizing the compounds at the in vivo model stage with respect to pharmacokinetics, pharmacodynamics, safety, and tolerability.

Animal studies confirmed Exelixis’ XL880 is a powerful inhibitor of both MET and VEGFR2 – two receptor tyrosine kinases (RTKs) that are key cancer targets. The compound was also long acting, requiring just a single oral dose in animals for effects to be seen. The scientists were also tracking key biomarkers during those early tests. "Next, we wanted to recapitulate the pharmacodynamics we saw preclinically in patients," says Morrissey. So, they looked for similar pharmacology and the same kinase inhibition spectrum in patients whose tumors are driven by one or both of XL880’s targets.

In a Phase I trial in various cancers, a couple of patients responded well to the drug, including one with melanoma. So the researchers focused in to see if the correct signaling pathway was being affected and if that was leading to the right phenotype.

A biopsy from the melanoma patient showed that the compound significantly inhibits MET autophosphorylation. "We also blocked signaling downstream of MET as judged by reduction in the level of pERK and pAKT – two important kinases involved in growth and survival signaling pathways," says Morrissey. The tumor sample also showed heightened apoptosis and reduced proliferation. Meanwhile, a powerful indicator of VEGFR2 inhibition in humans is a rise in blood pressure. "We saw a number of examples of that among the study subjects," Morrissey says.

MET is thought to play a significant role in certain papillary renal carcinomas. Some subjects with that type of cancer who took XL880 showed partial responses. According to Morrissey, the biomarker data and genetic signal of MET activation in those particular tumors is very striking, and the responses suggest reinforced that the compound is working the way they expect it to, and working well.

Picking out those targets that are real drivers of cancers is a difficult task. But thanks to these studies and others, Exelixis’ scientists believe they were first to validate MET’s role. XL880 is now in a Phase II trial involving patients with sporadic and hereditary papillary renal carcinoma, gastric and some other cancers.

Avalon Targets IMPDH

To best leverage their gene expression platform, Avalon has collected "An unprecedented amount of data from Phase I patients," says President & CEO Ken Carter. The company has two approaches – one looks at the patient’s entire genome, the other focuses on a set of about 30 genes that reflect different features such as apoptosis and DNA metabolism. These markers "Get to the central mechanism of how AVN944 works," he says.

AVN944 is Avalon’s inhibitor of Inosine Monophosphate Dehydrogenase (IMPDH) 1 and 2. IMPDH2 is necessary for many cancers to produce guanine nucleotides and grow. Avalon scientists used gene expression profiling and guanosine triphosphate (GTP) levels to first study AVN944’s effects in cancer cell lines. Next they did a Phase I trial and then used gene expression signatures to see if they could predict response to the drug.

That 30-plus gene set will probably be reduced to about 10-12 genes later, and will be used in upcoming trials. "Other people are dipping their toes in the water," says Carter. "Based on the Phase I trial, we are confident that we can do real patient stratification as early as Phase II." The company just initiated a Phase II trial of the drug in pancreatic cancer.

SuperGen Finds Markers for Combinations

SuperGen, meanwhile, has been studying a biomarker called Rad51. The company’s scientists believe this molecule will help them accelerate development of their multi-targeted tyrosine kinase inhibitor, MP470.

"There was initially no obvious connection between the activity of this compound and Rad51," explains SuperGen Chief Scientist David Bearss. "We came across this marker while we were studying combination therapies."

SuperGen aims to team MP470 with other treatments, such as radiation, platinum agents, and topoisomerase inhibitors. All these approaches attack cancer cells by causing single- or double-strand breaks in DNA. Most cells use DNA repair pathways to fix such breaks. Rad51 is a player in one of these pathways, and patients with lower expression levels of this gene appear to be most sensitive to combination therapies that include MP470.

The company has recently initiated a Phase I trial of MP470 that will include analysis of Rad51 levels. "If that works, we’ll use the marker in Phase II," he says. Bearss concedes that new challenges always arise in later clinical trials, "But here, we have proof of mechanism."

One goal of such studies is to remove as many questions as possible about whether a drug works, even if that means excluding some patients from receiving it. At the same time, biomarker proponents are hoping they can speed things up as well.

"Gleevec went from first in man to approved drug in 28 months," Carter says. "The idea is to accelerate the process by enriching patient populations." Avalon SVP of Product & Pharmaceutical Development David Bol says that payors will soon be forcing this issue: "If you are going to charge $100,000, you better know the drug is going to work in that patient."

But, as Bearss also points out, such studies can sometimes confuse as well as clarify. "What if you see a clinical response but no biomarker response? A lot of people are trying to wrap their heads around how to deal with that one."

Copyright 2007, Cambridge Healthtech Institute. All Rights Reserved.

7-11-07 - Biomarker Breakthroughs