LONDON - British Biotech plc asked investors "not to jump to conclusions" Monday when the company disclosed that its oral anticancer drug, marimastat, had failed its first pivotal trial, investigating the compound in pancreatic cancer.

The company's shares initially fell by 7 pence to 19 pence, but recovered to close at 22 pence.

Elliot Goldstein, CEO of British Biotech, told BioWorld International the study is one of nine under way, "and pancreatic cancer is the hardest to treat. We expect to report the results of three more trials in the next 12 months."

The pancreatic-cancer study was a randomized, controlled study of 400 patients with advanced disease, and was conducted in the U.S. and Europe. Its primary objective was to compare marimastat, a matrix metalloproteinase inhibitor, with the cytotoxic agent gemcitabine as a first-line therapy for survival. Patients were randomly allocated to four groups, receiving 5 milligrams of marimastat twice daily, 10 milligrams of marimastat twice daily, 25 milligrams of marimastat twice daily, or gemcitabine. Enrollment commenced in June 1996 and the study ended in October 1998, when 90 percent of the patients receiving gemcitabine had died.

The study failed to reach its primary endpoint of a reduction in mortality of 16 percent or more in the patients receiving 10 milligrams and 25 milligrams of marimastat twice a day. There was no significant difference between the survival curves for gemcitabine and the three doses of marimastat.

In a secondary analysis, the 25-milligram marimastat group and the gemcitabine group were not significantly different from each other with respect to survival, and both 25 milligrams of marimastat and gemcitabine appeared to be better than either the 10-milligram or the 5-milligram marimastat group.

Goldstein said it would be "speculation" to interpret the secondary analysis as demonstrating 25 milligrams of marimastat twice per day is as effective as the standard cytotoxic therapy. "There is a suggestion that the drug is showing activity, but the study was not designed to show equivalence; it was powered up to show a difference," he said, adding that the results could be interpreted as a signal marimastat may be active at higher doses.

A further nine trials are in progress, from which the data will be available progressively over the next three years. In these trials doses of either 10 milligrams twice a day or (in two studies) 20 milligrams of marimastat twice a day are being compared either alone or in combination with cytotoxic drugs to placebo. None of the trials has reached its endpoint, so British Biotech says it cannot be precise about when further clinical data will be available.

The results of the pancreatic study, known as Study 128, raise the question for British Biotech of whether it should undertake a further study in an attempt to show efficacy at a higher dose in patients with a less aggressive cancer. Advanced pancreatic cancer is a very difficult one in which to show efficacy; in this trial, more than 80 percent of the patients died within a year. "The question is, 'If the tumor is less aggressive, will marimastat work chronically?'" Goldstein said.

Comparison of safety data between the four groups revealed no marked difference, other than the expected side effects of musculoskeletal events with marimastat and hematological events with gemcitabine.

British Biotech said it is "assessing the full implications of the results of Study 128 for the overall marimastat development program." The company said it believes the results of further studies are needed before a complete assessment of efficacy, tolerability and dose regimen can be made. Discussions will now be held with external experts, including regulatory agencies.

Goldstein said he could not say if these results will aid British Biotech in its search for partners for marimastat. "My sense is that a partner would look at this data as quite interesting in that the drug may be showing activity in this difficult cancer," he said. "They will, at the very least, be intrigued."

The marimastat clinical trial program was reviewed between August 1998 and December 1998, following the appointment of Goldstein to replace previous CEO Keith McCullagh. As a result, one of the trials, a European Phase III monotherapy trial in ovarian cancer, was dropped because it was badly designed and would not have generated registerable data, the company said.

It was also decided that another trial in small-cell lung cancer would need to be increased in size, and British Biotech said Monday that Canada's National Cancer Institute had agreed to this.

British Biotech, based in Oxford, also disclosed results for the third quarter and nine months ended Jan. 31, showing a cumulative loss of #24.9 million (US$40.6 million), down #6 million compared with the year 1997 to 1998. Spending for the nine months was #27.3 million, down #12.7 million compared to the previous year.

The loss for the third quarter was #8.3 million, down from #11 million for the same period in 1997 to 1998 as a result of lower expenditure. Research and development expenditure was cut to #8 million from #11.4 million.

Cash on hand as of Jan. 31 was #105.3 million.