GAITHERSBURG, Md—The Food and Drug Administration’s Oncologic Drugs Advisory Committee (ODAC) has advised Zila, Inc., to take OraTest (tolonium chloride) back to the clinic for further testing. But members disagreed whether continuation of an ongoing clinical trial could provide the data necessary for the group to recommend approval of the test for detecting oral cancer, or if the company needs to undertake an entirely new study.
“I’m not sure that at the end of another 300, 400, or 500 patients, we’re going to have data that are going to be any better or any less confusing,” said ODAC consultant Stacy Nerenstone, MD, of the Helen and Harry Gray Cancer Center, Hartford, Conn.
Zila sought approval of OraTest (a ready-to-use kit whose active ingredient is a patented formulation of a dye commonly called toluidine blue) as “a diagnostic adjunct in patients with oral lesions suspected or known to be malignant, to help in detection of all sites of cancer, definition of borders or cancerous lesions, and selection of sites to be biopsied.”
According to the company, the dye in OraTest is taken up preferentially by cancer cells, which allows identification and definition of oral carcinomas and carcinomas in situ, often when the cancers are not found on visual examination.
Although toluidine blue is available for use in detecting oral cancer in dentists’ offices, it has been inconvenient for dentists to mix up the material on their own, Steven Corbin, DDS, chief of staff to the US Surgeon General, said on behalf of Oral Health America, a Chicago-based nonprofit organization.
Another speaker, Pier Cipriani, DDS, said that when ordered from a chemical supply house, toluidine blue arrives as a reagent grade powder. Preparing the solution stains fingers, clothing, and countertops, and the resulting liquid has a shelf-life of only 1 to 2 days. In addition, staining results may vary greatly from batch to batch due to impurities. “OraTest uses pure, pharmaceutical-grade ingredients; is ready to use, premixed, and flavored; and has a multiyear shelf life,” Dr. Cipirani said.
Zila provided the FDA staff with an interim analysis of 16 patients—from the first 367 patients enrolled in a trial of OraTest at 12 centers—in whom 17 oral cancers were detected by the test or by visual examination.
The study was designed to support a screening claim rather than the narrower adjunct claim that is the subject of the company’s new drug application (NDA). FDA reviewer Ken Kobayashi told ODAC members that the agency had not recommended that Zila seek approval based on the interim analysis. He and committee members questioned both the company’s interpretation of the data and how the data were gathered.
The company, for example, listed 10 of 17 oral cancers as identified solely by OraTest; the FDA review team put the number at five. The company claimed 100% sensitivity for OraTest and a specificity of 17%; the FDA listed sensitivity at 89% and specificity at 19%.
Among the problems with the study was a protocol violation at the University of Pittsburgh, which sent patients with suspicious lesions on to surgery, while enrolling them in the study. Zila Biomedical president Ralph Green, DDS, said that violation put the initial 131 patients enrolled at the Pittsburgh site off-protocol. “We have discovered it; we have monitored it, and for the rest of the cohort beyond the 131, they are now back in line,” Dr. Green said.
Committee members also expressed concern that the test might produce a high number of false positives.
Two supportive studies also submitted by Zila were criticized by the FDA review team as inadequate in design to support the company’s application.
Agreement on Value of Test
In spite of the criticisms of the data presented by Zila and flaws in the study’s execution, committee members expressed strong agreement that a test such as OraTest would have great value in the diagnosis of oral cancers.
“I don’t have the sense this is a bad product, and, in fact, I have the reverse feeling,” said Kathy S. Albain, MD, professor of medicine, Loyola University Medical Center, Maywood, Ill, who encouraged Zila executives to pursue data on OraTest. “We hope you take our comments as constructive, positive comments. We hope to hear about OraTest again,” she said.
The company said that 673 patients with 25 cancerous lesions have now been enrolled in the OraTest study. Originally, the protocol called for the study to end after 54 cancers were identified. Dr. Green said that following the ODAC meeting, Zila executives and key FDA personnel met to map a course of action to gain approval of OraTest. Zila will decide as soon as possible what to do about continuing the clinical trial.
