Some tumors, though, survive radiation therapy because they are hypoxic--starved of oxygen. Abraham knew that one of the body's responses to this condition is to create 2,3-diphosphoglycerate. This molecule chemically decreases the oxygen binding affinity of hemoglobin, in turn increasing the unloading of oxygen from hemoglobin.

2,3-DPG's drawback is that it can't be administered therapeutically because it is unable to cross the red blood cell membrane. Abraham's molecule, however, easily crosses the cell membrane and is actually better than 2,3-DPG at unloading oxygen from hemoglobin, thus increasing the susceptibility of tumors to radiation damage.

Formed with the goal of commercializing this molecule, Allos chose the virtual start-up company model and has never had laboratories or manufacturing capability. Indeed, for the firm's first three years, Stephen J. Hoffman, Allos' former chief executive officer and now its chairman, was the only Ph.D. chemist on its staff. Douglas G. Johnson, vice president for manufacturing, became the second when he joined in 1997. Today, Allos employs four chemists with advanced degrees to manage its outsourced projects.

In its early days, Allos contracted with two custom synthesis labs to make developmental quantities of RSR13 under current Good Manufacturing Practices (cGMP). By 1997, Johnson says, Allos had advanced the compound far enough that it was ready to conduct larger scale clinical trials that would require significant amounts of RSR13. With the help of a consultant, the firm canvassed the fine chemicals marketplace for a partner that could manufacture commercial quantities of the compound under cGMP conditions.

The initial synthesis of RSR13 wasn't particularly difficult, but purity would be critical during scale-up because the drug is administered intravenously and at high doses. Allos was attracted to Hovione, Johnson says, in large part because the Portuguese firm already manufactured iopamidol, an X-ray contrast media agent that is also injected at high doses. "We commissioned Hovione and one other company to make a test batch," he says. "Hovione's was of higher quality. We liked what we saw and worked out a development and supply agreement."

	HOW DRY I AM This injectable-grade drug dryer at Hovione's Lisbon headquarters is used to prepare products such as RSR13. HOVIONE PHOTO

According to Santos, Hovione was making trial quantities of the compound as early as April 1997. But an unanticipated problem arose in 1999 as the company started scaling the laboratory synthesis into a commercially viable process. The problem was related to an impurity profile and could be remedied only with equipment changes.

"Hovione was committed to this product, so we brought in the needed equipment," Santos says. "We fully qualified it and restarted within a month." One side benefit: Allos and Hovione are now patenting technology that arose during the troubleshooting process.

Johnson says he was impressed by Hovione's determination to solve the impurity problem, although at the time he was somewhat disturbed by the company's hesitation to provide detailed updates on its progress. He's learned since then that the Hovione team was only doing what many of its customers want--keeping them blissfully ignorant of the nitty-gritty of the scale-up process.

"I asked them at dinner one time if we were different from other clients," Johnson recalls. "The response of one of their chemical engineers was that they hadn't ever worked with a company that wanted to know so much about what they were doing."

Hovione completed manufacturing validation for RSR13 in 2001. Last year, Santos says, the company transferred the production technology to its Macau plant to provide backup capacity. Hovione also transferred the technology to Degussa's Raylo subsidiary in Edmonton, Alberta, which will serve as a third supply location.

That supply could soon be needed. In February, Allos completed submission of a three-part New Drug Application for RSR13 for the treatment of brain metastases originating from breast cancer. FDA is evaluating the compound under its priority review program and has established a target date of June 4 to take action. RSR13 will be on the agenda of the May 3 Oncological Drugs Advisory Committee meeting.

If they do, they will be spending more time with Santos. In addition to acting as manager of the RSR13 project, he is director of operations at Hovione's one-year-old technology-transfer facility in East Windsor, N.J., where at least the early phases of PDX scale-up would likely be conducted.

That would be fine with Johnson. With eight years of experience now with Hovione, he says Allos knows it can rely on the fine chemicals firm. "We know the Hovione culture and people and are comfortable with them," he says.