UPDATE 3-Erbitux drug fails in UK study, hitting Merck KGaA

* UK study doesn’t show survival gain seen in earlier trial

* Merck KGaA shares down 3 pct, erasing earlier gains

* Merck oncology head says UK study result “surprising”

(Adds latest shares, further analyst comment)

By Ben Hirschler

BERLIN, Sept 23 (Reuters) - Colon cancer drug Erbitux produced conflicting results in clinical trials on Wednesday, as a keenly watched British clinical study failed to show the same survival benefit seen in an earlier international one.

The news knocked confidence in Erbitux -- sold by Merck KGaA MRCG.DE, Eli Lilly LLY.N and Bristol-Myers Squibb BMY.N -- which is battling Amgen's AMGN.O rival treatment Vectibix, and shares in Merck fell sharply.

British researchers said their study, known as COIN, found no overall survival benefit in combining Erbitux with chemotherapy as a first-line treatment for metastatic colon cancer patients with a particular genetic profile.

A few hours earlier, another international trial, called CRYSTAL, presented at the same ECCO-ESMO cancer congress in Berlin, had found just such a benefit.

Morgan Stanley analyst Andrew Baum said the “surprisingly inconsistent” results suggested sales expectations for Erbitux were too high, adding that regulators could theoretically further restrict the use of Erbitux because of COIN.

Shares in Merck, which had gained as much as 3 percent on the initial news, turned negative on the second finding and the stock was down 3 percent by 1400 GMT.

The median overall survival for Erbitux patients with the normal, or wild-type, version of a gene called KRAS was 23.5 months in CRYSTAL compared with 20 months for those on chemotherapy alone.

That success was a significant win for Erbitux, which is now the only targeted colon cancer drug to show an overall survival advantage when combined with modern chemotherapy.

A year ago, researchers had detected a similar survival gain in the same CRYSTAL trial, but at that stage the difference was not statistically significant because the number of patients whose tumours had been tested for KRAS status was too small.


With COIN also looking in detail at the genetic profile of patients, many experts thought it would show a positive outcome.

In fact, the British team found overall survival was not statistically significant in their study at 17.0 months in the Erbitux arm compared to 17.9 months for chemotherapy only.

Wolfgang Wein, head of oncology at German drugmaker Merck, said he was perplexed by the British trial result, but noted the population group study was particularly frail and there were also imbalances in chemotherapy given in the different study arms.

“This is a very surprising result indeed,” he told Reuters.

Erbitux currently dominates the market for EGFR drugs, which block a protein called epidermal growth factor involved in cancer cell growth. Its 2008 sales were $1.6 billion, 10 times more than Vectibix.

Amgen hopes to redress this imbalance by showing its drug is similarly effective, while offering dosing advantages.

Both drugs also compete with Roche's ROG.VX Avastin, which works in a different way by starving tumours of blood supply.

Data on Vectibix presented at the same congress in Berlin has shown encouraging results but cancer specialists don’t see the newer medicine as having any particular edge over Erbitux.

Richard Adams of Cardiff University, one of the investigators on the COIN study, said there appeared to be little difference between the two.

He also questioned whether it was really worth using EGFR drugs in first-line treatment, rather than reserving them for third- or fourth-line, after other options had failed.

Both drugs have recently been found to work only in the 60 to 65 percent of patients whose tumours contain wild-type KRAS -- a prime example of the arrival of personalised medicine, or tailoring treatment according to a patient’s genetic make-up.

By targeting this subset, drug companies are now able to identify which patients will benefit most from treatment, although it also shrinks the size of the potential market.

It is a trade-off that Merck’s Wein is willing to accept.

“It shows the importance of personalising cancer care, which is more health economic,” he told Reuters. “It’s a further endorsement of personalised medicine and this is clearly where the train is heading.” (Additional reporting by Kate Kelland, with Ludwig Burger in Frankfurt; editing by David Cowell and Hans Peters)