Genentech R&D team fixes a flawed approach to Alzheimer's

Roche ($RHHBY) never publicly described what went wrong with its early-stage BACE drug for Alzheimer's when it killed the program. But a new study from the pharma giant's Genentech unit outlines the potential threat one of their BACE efforts posed to patients--and how they've managed to adjust for it in animal studies that may well open the door to new human trials.

As described by Nature, investigators at Genentech have been working for several years on using the protein transferrin to smuggle drugs into the brains of patients. The body uses transferrin to carry iron past the notoriously impenetrable blood-brain barrier--a natural barrier to dangerous invaders--and the researchers developed an antibody that attached to transferrin. Another end of the bispecific antibody attached to beta secretase, or BACE, which helped eliminate amyloid beta in patients. But they tell Nature that it also linked to red blood cells that had transferrin on the surface, killing blood cells while attacking the toxic protein many--though not all--believe cause the memory-wasting disease.

So the Genentech team made a switch. By modulating the dosage, the investigators say that they have demonstrated that they can attack amyloid beta without evidently damaging the red blood cells. And now that they believe they can get into the brain safely, they have the proof-of-concept data in hand from a new study with monkeys demonstrating that they may be able to attach other drugs to the transferrin model--including their drug crenezumab.

Robert Vassar

"It's a very elegant, clever approach," Alzheimer's investigator Robert Vassar tells Nature. In particular, Genentech believes it also holds the promise of going after BACE without raising the same tox issues that have afflicted other molecules with the same target.

BACE is one of the most popular targets in Alzheimer's research, a controversial field that has seen a slew of failures, including at least one from Eli Lilly that harmed patients.

Off-target toxicity was the likely culprit behind the death of Eli Lilly's ($LLY) BACE program for LY2886721. That failure, one in a series of setbacks for the pharma giant in Alzheimer's, pushed Lilly to ink a $500 million deal--$50 million upfront--with AstraZeneca ($AZN) to take charge of its BACE drug--AZD3293. Astellas recently backed out of its $760 million deal with CoMentis on another early-stage BACE drug. Merck ($MRK), meanwhile, has emerged as the leader in the BACE race, accelerating a late-stage study of its own BACE drug--MK-8931--after clearing a safety hurdle in the clinic. Biogen Idec recently partnered with Eisai on E2609. The newly public Vitae reported after the market closed on Thursday that the oral BI1181181/VTP-37948 hit the goalposts on a key biomarker for BACE, flushing a prime suspect behind the disease without triggering any red safety flags in a pair of Phase I studies among a small group of healthy volunteers. And Novartis ($NVS) is positioning its own BACE inhibitor for the clinic, targeting presymptomatic patients.

Genentech's work in the field helps underscore how improvements in antibody technology have opened up new avenues for Alzheimer's investigators. But they have also underlined just how easy it is for these drugs to pose a threat to patients. And their delivery research has implications that extend to a variety of neurodegenerative ailments, including ALS, Parkinson's and Huntington's disease.

- here's the report from Nature