Asco – Clovis impresses then disappoints with early lung cancer data


The decision to hold back data from Clovis Oncology’s early-stage second-line lung cancer trial of CO-1686 until the Asco cancer meeting turned out to be both wise and unwise. The data in T790M mutation patients look promising, with early estimates suggesting progression-free survival of 12 months or more, but the company's party was spoiled as it was forced to disclose to the watching oncology world a hyperglycaemia side effect that has led to some patients having to take insulin.

Without the revelation that this hyperglycaemia can become resistant to oral antidiabetic therapy, sentiment might have swung towards it – and away from AstraZeneca’s AZD9291 – based on the early PFS curves. Still, some leading oncologists believe that the survival data could still carry the day as more mature data emerges from the competing agents, particularly if investigators can figure out how to resolve hypoglycaemia with metformin alone.

“Against [the side-effect] we’ve seen outstanding clinical activity,” said Thomas Lynch, director of the Yale Cancer Centre. “If this holds, it won’t matter if you have to take metformin.”


AstraZeneca and Clovis, now joined by a Korean upstart, Hanmi Pharmaceuticals, are leading the race to get a new agent into non-small cell lung cancer for patients who have progressed after treatment with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors – Roche’s Tarceva and Astra’s Iressa (Asco preview – Astra fight against Clovis gets prominent billing, May 15, 2014).

The territory being explored by these agents centres on patients who develop a secondary mutation called T790M that allows tumour cells to resist treatment with first-line EGFR agents.

Astra’s data had been pre-released by the oncologists’ organisation, so its presentation was primarily an update. AZD9291 achieved an overall response rate of 83% in patients with a T790M mutation at a 240mg once-daily dose. The lowest dose, 20mg once daily, achieved a 50% response rate.

The presentation by Pasi Jänne, a Harvard Medical School professor, did not disclose a specific a PFS estimate. Of the patients showing a response, 94% were still under treatment at April 2; the longest duration of response was seven and a half months.

Which is why the data reveal on the Clovis molecule, CO-1686/AVL-301, was big news. A number for median progression-free survival has not been reached, and based on the current survival curve the figure is estimated to be at least 12 months.

A third agent that had not been spotlighted is Hanmi’s HM61713, which hit a 19-week median PFS mark in the T790M-positive population. Like the Astra project, patients taking it experienced elevated levels of rash, but neither it not AZD9291 had the blood sugar and cardiovascular effects of Clovis’s candidate.

Emerging competitive landscape

But with the Clovis disclosure that “three or four” patients have started insulin therapy after becoming resistant to metformin, the road has got rougher for CO-1686, thanks in large part to the way in which it was disclosed.

It did not come at an Asco data presentation but at an off-site analyst meeting, which should spark some investor questions about company transparency. In early afternoon trading Clovis shares were down 8%.

Both CO-1686 and AZD9291 have received breakthrough therapy designation from the FDA, with Clovis receiving word just two weeks ago, so speed in development will be of the essence – the agency only awards the regulatory advantages of designation to the first in an indication to submit a new drug application.

According to EvaluatePharma’s consensus, the Clovis candidate is forecast to sell $791m and AZD9291 $444m in 2020.

Neither seems noticeably ahead of the other in terms of developmental stage. It does not help Clovis that it also has a cardiovascular side effect in the form of a prolongation of the QT interval, something regulators will look at closely as this is a predictor of sudden death. Both the blood-sugar and cardiovascular effects will also need to be fully explored by the time the application lands on desks at the FDA and the European Medicines Agency.

The regulators’ data demands may very well create a new layer of safety-related trials at the phase I and early phase II levels, something that will not speed the drug's development.

For all the excitement over the much-touted wide applicability of immuno-oncology agents like Yervoy and nivolumab, targeted agents are still showing they should have a continuing role in cancer care. Given how difficult it has been to find effective agents in advanced non-small cell lung cancer, it is not surprising to see the targeted approach continue to be relevant in this setting.

To contact the writer of this story email Jonathan Gardner in Chicago at or follow  @JonEPVantage on Twitter

Share This Article