The details of a failed study involving Roche Holding AG's (RHHBY) Avastin showed the drug's effectiveness in preventing early-stage colon cancer diminished when patients stopped taking it, a disappointing result.

The data, presented at the annual meeting of the American Society of Clinical Oncology, were the first results involving Avastin as a treatment for such early-stage cancer, a potentially lucrative market. The outcome is seen as a setback for Avastin and other similar treatments, and bodes poorly for other early-stage trials of Avastin. The drug, already approved to treat advanced breast, lung, brain and colorectal cancer, had 2008 U.S. sales of $2.69 billion.

The study played an important role in the negotiations leading to Roche's $46.8 billion March buyout of Genentech, which invented the drug, its flagship product. The trial's failure became public in April, but no details were given until Saturday.

In a statement, Genentech said that Avastin's early effectiveness may support its role in helping prevent early-stage colon cancer from returning and spreading, but that additional research is needed.

The study, which included 2,710 patients, had a group that received only six months of standard chemotherapy, and another that got six months of chemotherapy combined with Avastin along with six additional months of Avastin.

Although the results ultimately failed, Avastin had a positive effect in the first year, showing a 40% reduction in the event rate, but then both study groups quickly reverted to having similar effectiveness.

"We came very close to reaching early stopping boundaries for efficacy," said Normal Wolmark, chair of the cooperative group that ran the trial, at a press conference Saturday. "But once the [Avastin] was stopped, the benefit disappeared."

The study seems to show that Avastin is stopping cancer from growing, but isn't eliminating it, physicians said.

The implication of the data is that Avastin should be used indefinitely in order to delay disease progression, said Leonard Saltz, of Memorial Sloan-Kettering Cancer Center.

Wolmark said his group hopes to soon begin a similar post-surgical trial that uses Avastin for a period of two years, but Saltz wasn't sure that such a trial would yield more insight - especially since data may not come until 2017.

Physicians also raised concerns that the cost of Avastin, which Genentech said was about $52,000 a year for patients in this setting, could make such prolonged, or indefinite, use prohibitive.

"It gets phenomenally expensive to think of years and years of therapy," Saltz said. Furthermore, long-term use of Avastin, which stops cancer growth by blocking the formation of new blood vessels, may raise safety questions because it isn't clear how the body will react to the prolonged blocking of this process.

Physicians and investors had high hopes for the study because its success would provide a new option to patients and add billions of dollars in new annual sales to Avastin.

Furthermore, the study's failure may shed light on the effectiveness of drugs similar to Avastin, including Pfizer Inc.'s (PFE) Sutent as well as Nexavar, sold by Onyx Pharmaceuticals Inc. (ONXX) and Bayer AG (BAYRY).

The study is a setback for Roche, because usage of Avastin's early usage remains a key long-term growth driver for the drug, and there are eight ongoing studies of Avastin in that mode, with the next data coming at some point in 2010.

The data suggests that those similar Avastin studies are more likely to fail, a notion that Dr. Wolmark confirmed in expressing a lack of confidence in the success of those trials.

Prior to the Roche deal, Wall Street's confidence in the study grew because of the Swiss company's persistence to swiftly close the Genentech deal, and Genentech's resistance to the takeover seemed to signal confidence in the study.

Roche has consistently played down the trial's role in the deal, saying it wanted Genentech for its research and not for the outcome of the single trial. Despite this assertion, the failure of the study has raised the question of whether Roche overpaid for Genentech.

Amit Roy, an analyst that follows Roche for Nomura Securities, expects that Wall Street will likely come to that conclusion, but he believes that Roche wasn't solely motivated by Avastin's potential expansion into earlier stages.

-By Thomas Gryta; Dow Jones Newswires; 201-938-2053; thomas.gryta@dowjones.com