* Recentin fails vs Roche's Avastin in late stage study
* Data on second late stage trial due in first half of 2010
* Company re-confirms its financial guidance for 2010
* Shares down more than 1.0 percent
(Updates with shares down, more quotes, forecasts)
By Kate Kelland
LONDON, March 8 (Reuters) - AstraZeneca's cancer drug Recentin failed in a head to head late stage trial with Roche's Avastin in colorectal cancer patients, hitting the British drugmaker's shares in early Monday trade.
The company said the data, which had been widely expected to be negative for Recentin, showed the experimental drug did not meet primary endpoint in the Horizon phase III study.
AstraZeneca shares lost more than 1.0 percent and were trading around 29.63 at 0850 GMT. Roche shares were down 0.33 percent.
"While we recognised that challenging Avastin would be a high hurdle, it is still disappointing, despite evidence of clinical activity with Recentin, not to have met the primary endpoint in this study," Astra's head of oncology Alan Barge said in a statement.
Recentin, a pill, was developed to challenge Roche's injectable treatment Avastin. Both are targeted therapies designed to starve tumours by stopping them from building blood vessels, a process called anti-angiogenesis.
Analysts see Recentin reaching sales of around $329 million by 2013, according to Thomson Pharma consensus forecasts.
Astra, which reaffirmed its financial guidance for 2010, said this was the first of two pivotal studies of Recentin, known generically as cediranib, in colorectal cancer.
The other study, Horizon II, is testing Recentin combined with chemotherapy against chemotherapy alone, and data are expected in the coming months.
Barge said the results of this trial would offer "further information on whether Recentin may provide benefit for patients with colorectal cancer and will inform any decision about possible regulatory filings".
The firm also said results of a separate late stage trial with Recentin in treating recurrent glioblastoma brain tumours are also expected in the first half of this year.
Initial data on the treatment of this type of brain cancer showed it helped stop the fluid build-up caused by glioblastoma and helped mice with the brain tumours live longer.
A trial of Recentin in lung cancer was stopped in 2008 after the drug failed to help enough patients. (Editing by Kate Holton and Hans Peters)