When a clinical trial gives you lemons, try to make lemonade.

That's what Atherogenics ( AGIX) did Tuesday with results from a phase III trial of its heart drug AGI-1067. The company downplayed that the study failed to meet its primary endpoint in a spectacular fashion; instead, it focused on a hodgepodge of positive secondary endpoints.

But in this case, Atherogenics' glass of lemonade might be the last one the company squeezes since the phase III trial turned up some potentially serious side effects associated with AGI-1067, most notably liver toxicity, which has meant the end for drugs in the past.

Atherogenics traded in the green most of the day, but concerns over AGI-1067's safety forced the stock lower in the afternoon session. The stock closed down 34 cents, or 9.5%, to $3.24. After hours, shares fell another 4% to $3.11.

The phase III data from the ARISE trial of AGI-1067 was presented Tuesday at the annual meeting of the American College of Cardiology.

AGI-1067 is a pill designed to break down and reduce the level of fatty plaque deposits in a person's arteries. Clearer arteries are healthier arteries and should lower the risk of cardiovascular illnesses such as heart attacks and strokes and reduce the need for additional surgery to reopen arteries.

The ARISE trial enrolled more than 6,100 patients with heart disease and tested the effect of AGI-1067 vs. placebo on the time to first incidence of a composite of major cardiovascular events, such as cardiovascular death, heart attack, stroke, need for coronary revascularization and unstable angina.

Typically, efficacy gets a front-row seat in discussions of phase III trials, but let's focus first on safety.

Eight patients receiving AGI-1067 during the study reported abnormal liver function, measured by an increase in a key liver enzyme as well as an increase in bilirubin. These markers of liver toxicity are a big red flag for severe liver injury, and they occurred at twice the rate in AGI-1067 patients as they did in patients receiving a placebo.

A single patient taking AGI-1067 in the study went into liver failure, while no patients taking the placebo suffered liver failure.

Liver toxicity is a big concern for drug regulators. In 2000, the diabetes drug Rezulin was pulled off the market because it was linked to severe liver injury and death. More recently, in February 2006, AstraZeneca ( AZN) was forced to withdraw its anticoagulant drug Exanta from the market because of increased reports of severe liver injury.

AstraZeneca is Atherogenics' partner in the AGI-1067 program. AstraZeneca executives said today that the company is reviewing all of the AGI-1067 data before deciding whether it wants to proceed with the drug's development. You can be sure that the liver toxicity data made public today will weigh heavily in that decision.

During a conference call Tuesday afternoon, an Atherogenics executive said liver toxicity "was not a major source of concern" because the level of toxicity is within historical limits for other cardiovascular drugs.

AGI-1067 also lowers good cholesterol, or HDL, and raises bad cholesterol, or LDL, even when patients are also taking cholesterol-lowering statins. This is a repeat finding from AGI-1067's phase II trial and is another source of concern about the drug's safety profile.

Moving to efficacy results, AGI-1067 showed no benefit whatsoever compared with the placebo on the study's primary endpoint. The hazard ratio -- a measure of risk reduction -- was 1, which means AGI-1067 and placebo had the same effect.

The p value on the primary endpoint -- the odds that the results happened by chance -- was 0.985, well above the 0.05 level required to reach statistical significance demonstrating the drug's effect.

But the efficacy of AGI-1067 did look better when Atherogenics examined some secondary efficacy endpoints. Here the drug caused a 19% reduction compared with the placebo in certain "hard" clinical endpoints, such as cardiovascular death, heart attack and stroke.

But these hard endpoints seem cherry-picked. For instance, revascularization and unstable angina -- two other hard clinical endpoints -- were excluded from the secondary analysis. The former is a particularly strange endpoint to exclude because the original rationale for AGI-1067 was as a drug that breaks down fatty deposits in arteries. If that's the way the drug works, then it should reduce the need for patients to undergo artery-opening revascularization procedures.

Atherogenics executives also were pushing the ability of AGI-1067 to reduce by 64% the time-to-development of new onset diabetes. But compared with existing diabetes drugs and some in the latter stages of development, AGI-1067's impact on diabetes appears limited. And that's before factoring in the potential toxicity issues.

In short, AGI-1067 is no tasty drink. Investors will be well-served to quench their thirst elsewhere.

Adam Feuerstein writes regularly for RealMoney.com. In keeping with TSC's editorial policy, he doesn't own or short individual stocks, although he owns stock in TheStreet.com. He also doesn't invest in hedge funds or other private investment partnerships. Feuerstein appreciates your feedback; click here to send him an email.

More from Stocks

How Small-Cap Stocks Can Protect Your Portfolio From a Trade War

How Small-Cap Stocks Can Protect Your Portfolio From a Trade War

Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

3 Great Stock Market Sectors Millennials Should Invest In

3 Great Stock Market Sectors Millennials Should Invest In

Why Millennials Are Ditching Stocks for ETFs

Why Millennials Are Ditching Stocks for ETFs

Trump's 'Space Force' Could Launch a $1 Trillion Industry, Morgan Stanley Says

Trump's 'Space Force' Could Launch a $1 Trillion Industry, Morgan Stanley Says