When Isis puts the finishing touches on its first NDA, Lotz will simply save the document in PDF, press it into one CD-ROM, and then ship it to the FDA. Of course, he won't get off without printing at least one paper copy for the agency's archives. But the time and cost savings over printing and hole-punching nine copies will be tremendous. Lotz estimates that it will take 30% less time (a full month less) to prepare and ship the NDA to the FDA than it would if the company weren't automated. That's substantial when you consider that every month saved could mean an additional million or so in the coffer.
Even more important, though, is the fact that the single CD-ROM will allow the FDA to review the application much faster. In the past, NDAs were divided into sections, which were dispersed to the appropriate departments throughout the agency. One master copy was always kept in the agency's library. That meant that if, for example, the FDA's toxicologists needed to review a section of the NDA that the chemists were holding, they had to request a copy from the librarian, who would photocopy the proper section and then deliver it to the reviewers. Satisfying the request could take as long as a week.
But with the electronic version, FDA reviewers can jump from section to section of the NDA via hyperlinks in the text. It's Lotz's job to anticipate the links that the FDA might want to use. For example, he might link raw data from an animal test to a toxicologist's analysis of the same test so that the reviewer can simply flip between the data and the write-up.
There have been unanticipated benefits to the electronic NDA as well. Lotz happily reports that it has actually helped soothe the often adversarial relationship between the pharmaceutical industry and the FDA. In order for him to anticipate all the proper links within the NDA for Isis 2922, he's had to work closely with FDA reviewers to find out exactly what they need and how they want it presented. "I now feel like I'm actually collaborating with the FDA," says Lotz.
The prognosis for Isis looks good. "Antisense has finally gained credibility as a method for drug treatment," says Crooke. Recent articles in the trade and popular press support that assessment. In May Science claimed that antisense drugs had "shed their troubled image" and were beginning to be shown clinically effective. And in June Chemical & Engineering News trumpeted "Antisense Drugs: Finally Fulfilling Their Promise."
The investment community seems to agree. Many analysts predict that Isis will be profitable by 1999. Right now, the company burns about $25 million in cash each year. Still, it has approximately $70 million in the bank, which should keep it operating comfortably until Isis 2922 hits the market.
Between now and then, Isis has lots of work to do: it must complete the clinical trials for Isis 2922 by year's end; analyze the results of its treatments for cancer, HIV, and Crohn's disease; work on its strategic partnership with Novartis; and bulk up its R&D department with an eye toward developing new generations of antisense molecules. "We are just at the end of the beginning," Crooke says.
Joshua Macht is an associate editor at Inc. magazine.
Safeguards: No Mistake About IT
Whether test marketing a fruit drink, beta-testing a software application, or clinically testing an antibiotic, most small companies face the same dilemma: how to gather and analyze data in a way that will enable them to get to market faster. A solid information system can be the linchpin in the process.
But for the system to be worth anything, the data must be entered correctly. To guard against errors, Isis's Stuart Nixon has devised this painstaking five-step program:
1. Manual review
Before a patient's information is even entered into the computer, a data-entry operator reads it and checks for obvious mistakes. For example, a patient's age should never be 150.
2. Double entry
Once the patient's information has been punched into the database, a second data-entry operator reenters the same information into a new file. The computer then automatically compares the two files and prints out a list of inconsistencies, which must be reconciled by the two operators.
3. Visual verification
A data manager compares the data that has been entered against the original handwritten notations.
4. Electronic verification
A data manager runs a series of relational edit-check programs that check the data in selected fields within the database to make sure it's logical and consistent.
5. Final audit
Once an entire clinical trial is completed, Nixon and his team conduct one final audit by checking a representative cross section of data against the original handwritten notations. --J.M.
