Thursday, Mar 8, 1990

New Trial Results Inconclusive for U.S. Practice; Role of Heparin in Gissi-2 Questioned

Florence, Italy -- March 8, 1990 --

FLORENCE, ITALY, March 8, 1990 -- Results of the GISSI-2 International t-PA/ streptokinase Mortality Trial released here today indicated that both agents were equally effective in saving lives of heart attack victims, but failed to prove which agent is best. The Italian-based study did not use the treatment regimen that is favored by 98% of the U.S. cardiologists using thrombolytic therapy, a fact which may compromise the study's value in regard to yielding relevant life-saving and safety data applicable to U.S. medical practice.

"All major U.S. trials with t-PA used intravenous heparin administered within four hours," said David Stump, M.D., director of clinical research for Genentech, Inc. The largest U.S. trial of this sort was TIMI-2, sponsored by the National Institutes of Health, in which over 3,200 patients received immediate intravenous heparin with t-PA, and resulted in an overall mortality rate of about 5 percent. The mortality rates for GISSI-2, which did not use early heparin, were 8.9 percent for t-PA and 8.5 percent for streptokinase, which is not a significant difference.

More than 98% of recently polled U.S. cardiologists who use thrombolytic therapy said that they use intravenous heparin concurrently with t-PA. In contrast, half of the patients in the GISSI-2 study didn't receive any heparin, and the remainder received it by subcutaneous injection 12 hours after t-PA administration. The GISSI-2 study utilized heparin in a manner common to medical practice in Italy.

Additionally, the GISSI-2 study was designed in late 1987, before the importance of concurrent early heparin treatment was fully understood. Two American studies presented in November 1989 provided unexpected information about the critical role of heparin. These studies demonstrated that without early heparinization, the initial high coronary artery patency (restored blood flow) rates associated with t-PA therapy may be lost.

"In retrospect, the GISSI-2 study design was not optimal for U.S. medical practice because the organizers did not have the advantage of knowing this important new information about heparin's role in preserving the very high reperfusion rates achieved with t-PA, which in turn should further reduce the infarct mortality rates," said Alan Ross, M.D. professor of medicine and director of cardiology at the George Washington University Medical Center and the principal investigator of one of the heparin studies.

"Genentech already is consulting key U.S. cardiologists about the need for another study that uses heparin as it currently is used in the United States," Stump said. "We think it's premature to make definitive comparisons of mortality reduction between agents based on the current GISSI-2 information base."

Statistically different safety findings included fewer allergic reactions for t-PA, at 0.2 percent versus 1.7 percent for streptokinase, and less hypotension, with 1.7 percent for t-PA and 3.8 percent for streptokinase. Also statistically significant were the rates of major bleeds, with 0.6 percent for t-PA and 0.9 percent for streptokinase; and the overall stroke rates of 1.3 percent for t-PA versus 1.0 percent for streptokinase. These stroke rates are within the 1%-1.5% range seen in other studies with thrombolytics, and are within the range seen in heart attack patients not receiving thrombolytic therapy.

For the 12,000 patients included in the Italian arm of the study (GISSI-2) where the data has been thoroughly analyzed, there is no statistically significant difference between the hemorrhagic stroke rate for t-PA, 0.3 percent, and streptokinase, 0.25 percent.

Genentech's Activase®, t-PA, currently holds approximately two thirds of the U.S. thrombolytic market for heart attack treatment. "It appears that this trial offers limited new information to the understanding of thrombolytic therapy," said G. Kirk Raab, president and chief executive officer of Genentech. "The situation is confusing. Since the results are inconclusive in regard to U.S. medical practice, the need for a new trial using the standard U.S. heparin regimen is indicated."

"Although we don't know how this study will ultimately impact our sales," Raab said, "we are confident that physicians will consider their personal experience in treating over 170,000 patients with t-PA in the past two years and the findings of studies conducted in the U.S. with Activase in evaluating the outcome of GISSI-2."

On February 2 Genentech announced that its Board of Directors had entered into a merger agreement with Roche Holding, Ltd. whereby Roche will obtain 60% equity in Genentech for a total investment in the order of $2.1 billion. In response to inquiries relative to the GISSI-2 trial, Roche said, "We were fully aware of the ongoing study. The agreement was made with a long term view of the value of Genentech represented by its science, product portfolio and people. Individual trials limited to certain aspects only would not influence such important investment decisions."

# # #