The initial data, released earlier this month, sent Vertex shares soaring on hopes the company could have a multibillion-dollar franchise in the treatment of cystic fibrosis, a life-threatening genetic disorder that affects about 70,000 people worldwide.

"The corrected data appear weaker than the original data," ISI Group analyst Mark Schoenebaum said.

Vertex blamed the error on a misinterpretation of the data from a vendor involved in the statistical analysis.

Despite the revised data, Wall Street analysts said they were still optimistic that the combination would prove successful.

"This mistake is very disappointing," Vertex Chief Executive Officer Jeffrey Leiden told analysts on a conference call. "It's unacceptable to us."

Still, Leiden said, the company's conclusions "haven't changed" and that it expects to move forward with a pivotal clinical trial required for submitting a drug for regulatory approval.

Vertex expects final data on the study by mid-year, which will include more than twice as many patients.

Vertex shares were off 16.7 percent at $54.02 in late morning trading on Nasdaq. But they were still more than 40 percent above where the stock was trading before the initial release of the data earlier in the month.

Cystic fibrosis causes the thin layer of mucus that helps keep the lungs free of germs to become thick, clogging airways and leading to infections that damage the lungs.

The phase 2 study involves a combination of Vertex's new cystic fibrosis drug, Kalydeco, with an experimental drug called VX-809.

Kalydeco helps only about 4 percent of cystic fibrosis patients with a specific gene mutation. Vertex is testing combinations it hopes will eventually be able to address the larger CF population.

The study examined patient performance based on a measure of the maximum amount of air that can be exhaled in one second, known as FEV1.

On May 7, Vertex released data that it said showed that about 46 percent of patients experienced an improvement in lung function of 5 percentage points or more based on FEV1. On Tuesday, the company said that figure was actually about 35 percent.

The initial results also said about 30 percent of patients experienced an improvement of at least 10 percentage points. That number has been lowered to 19 percent.

The company also released data that showed patients on the combination saw an average absolute improvement in lung function of 8.5 percent compared to those taking a placebo.

That effect looked more pronounced in part because patients on a placebo saw a decline in lung function that was much greater than expected by Wall Street analysts.

Even with the placebo patients worsening, the improvement for patients on the Vertex treatments "is much higher than original Street expectations," Schoenebaum said in a research note.

Brian Abrahams, an analyst at Wells Fargo, said in a research note that the early drop in Vertex's shares seemed "excessive" and recommended buying the shares on weakness.

"While we are slightly disappointed by the less robust data, the Kalydeco/'809 combo still clearly appears active to us and we believe should still have multi-billion dollar potential."

(Reporting By Lewis Krauskopf in New York, additional reporting by Toni Clarke in Boston; editing by John Wallace and Maureen Bavdek)

• A Permanent Economic Emergency