Be it efficacy or safety, Eli Lilly (LLY -0.64%) just can't win with its Alzheimer's disease program .

The latest drug to bite the dust is LY2886721, a beta secretase, or BACE, inhibitor. Eli Lilly stopped a phase 2 trial after some patients had abnormal liver biochemical tests. Companies routinely run liver tests on clinical trial patients because it's one of the most common safety issues that affect drugs. The liver's job, after all, is to filter toxins out of the blood stream.

Lilly told the Wall Street Journal that it doesn't believe the liver toxicity has to do with the mechanism of inhibiting BACE and is most likely an off-target effect. That's good news for Merck (MRK -0.11%), which has its own BACE inhibitor, MK-8931. The company has already moved the drug into a phase 2/3 trial. If there is an issue, it'll likely be discovered in the phase 2 portion that's enrolling 200 patients.

Assuming there's no class effect, LY2886721's demise is good news for Merck as long as there's no class effect. Merck had a lead, but it wasn't enough to make a difference commercially. It's also good news for AstraZeneca (AZN -0.25%), which has a BACE inhibitor, AZD3293, licensed from Astex Pharmaceuticals (NASDAQ: ASTX), that entered a phase 1 trial last December .

Assuming, of course, that the drug works. It's not clear yet if targeting BACE will slow or reduce the symptoms of Alzheimer's disease. In a phase 1 trial, MK-8931 reduced the amount of beta-amyloid in cerebral spinal fluid by more than 90%, but it isn't known whether that will reduce the amyloid plaques seen in Alzheimer's patients' brains, and, even if it does reduce the plaque formation, whether that will have a clinical effect.

Failures abound
Rather than because of safety, most Alzheimer's disease drug failures -- and there have been a lot of them -- have fallen short because of lack of efficacy.

Eli Lilly's solanezumab failed a couple of trials last year. The company is retesting the drug in a different population. I wouldn't hold my breath.

And there was Pfizer, Johnson & Johnson, and Elan (ELN) that proved that having more companies developing a drug doesn't make it any more likely to succeed. The company's Alzheimer's disease drug bapineuzumab failed four different clinical trials last year.

And I could go on. But let's not. It's depressing.

Give up?
I doubt companies are going to throw in the towel. As hard as it's been to develop a successful Alzheimer's drug, the rewards are great enough to justify the failures. From a company's perspective, even drugs that have a low likelihood of succeeding, like solanezumab, are probably worth rolling the dice on.

Investors, on the other hand, would be best off not assigning any value to Alzheimer's disease pipeline drugs. If a company does finally figure out the magical formula, investors can benefit from the unexpected pop.