Re: Cholesterol veers off-script

[Guest guest]

The pdf is availed of the paper.

Science 10 OCTOBER 2008 VOL 322, 220-3Cholesterol veers off-script–JENNIFER COUZIN

Recent drug trials have produced surprising results; along with genetics research, these findings have put in question some long-held beliefs.

CHOLESTEROL NUMBERS ARE A MANTRA OFmedicine, and millions of us regularly supplya vial of blood to measure this waxy substancethat circulates in the bloodstream. All cellsneed it to survive. But it also feeds plaques inthe arteries that can break open, causing aheart attack. Controlling cholesterol is gospelin cardiovascular medicine; it guides treatmentand sells billions of dollars’ worth ofdrugs. It has also been reinforced by a Hollywood-like story line: A villainous “bad” cholesterolclogs arteries, and a valiant “good”cholesterol clears them.The cholesterol hypothesis “is like religionfor some people,” says Harlan Krumholz, acardiologist at Yale University. “They’ve beentaught it in medical school. They’ve beentaught it forever.”But Krumholz and some others say thatafter many decades, the cholesterol story isturning out to be messier and more

nuancedthan previously believed. Scientists increasinglyrecognize that goodand bad cholesterol, thoughoften spoken of in the samebreath, are not equally wellunderstood.Hundreds of studieshave shown that an overabundance of bad cholesterol,known as LDL (low-density lipoprotein),is associated with heart attacks. Goodcholesterol, or HDL (high-density lipoprotein),is thought to be protective, but evidencefor HDL’s benefit is flimsier. Some scientistsare now asking whether HDL is even relevantto heart-disease risk at all. Other fundamentalquestions persist: Why do people with healthycholesterol levels still suffer heart attacks?Does the mechanism by which drugs tacklecholesterol matter to health? “You ask 20 people,you get 20 opinions,” says Fisher,director of preventive cardiology at New YorkUniversity (NYU) School of Medicine.That uncertainty is also

eroding confidencein long-practiced strategies for drug development.Twenty-one years ago, the U.S. Foodand Drug Administration (FDA) approvedthe first of the so-called statin drugs tolower LDL levels, based on the beliefthat lowering LDL also preventsheart disease. That proved wildlysuccessful. But subsequent effortsto go after heart disease with a differentclass of cholesterol drugsproduced muddled results. Inaddition, many people had hopedthat a complementary strategy—giving drugs to raise good cholesterol,HDL—would profferbenefits akin to those of statins.But it hasn’t worked out that way.A recent clinical trial that pushedup HDL failed to protect the heart.At the same time, new geneticstudies are yielding disparateresults that undermine assumptionsabout cholesterol. This has left scientistspuzzled, with some consideringwhether an altogether

different risk factor,inflammation, is a missing link. Nextmonth, a study of a cholesterol-loweringdrug that may also reduce inflammationwill report a benefit for the heart.The bad actorCholesterol was first tied to heart disease in1910, when a German chemist found thatpeople with atherosclerosis had a high concentrationof cholesterol in their aortas. Feedingrabbits cholesterol dissolved in sunfloweroil caused severe atherosclerosis, cementingthe connection.A fatty substance that’s both made by thebody and ingested in food, cholesterol helpsbuild cell membranes and form hormones, aswell as performing many other tasks. It doesn’tdissolve in the blood but instead is cartedfrom place to place by bulky complexes calledlipoproteins. Two of these travel oppositeroutes: LDL transports cholesterol from theliver to other tissues, and HDL is thought tocarry it from other

tissues, such as the arteries,back to the liver.For many years, scientists and physicianshave recognized that high LDL raises the riskof heart attacks and that lowering it saves lives.This remains the majority view. The first LDLloweringdrug, Merck’s lovastatin, wasapproved in 1987, heralding a new era of blockbustercholesterol drugs, most of which havereaped billions of dollars in sales every year.FDA was so confident of the link that, inapproving lovastatin and five other statins thatfollowed, it made an unusual departure:Instead of judging efficacy based on thedrug’s ability to improve health or increasesurvival, it relied on the therapy’s ability tolower LDL. That confidence proved justified:In the 1990s, a trial of 4444 patients in Scandinaviashowed that another statin drug, simvastatin,reduced deaths from heart disease by42%, a number that’s held up in

subsequentstudies. “I have the same position now that Ihad many years ago,” that lowering LDL saveslives, says Terje Pedersen, the cardiologist andclinical trialist at Ullevål University in Oslowho led that early study for Merck.How low to go?On the principle that low LDL is good,many favor driving it as far down as possible.What passes for an acceptableLDL level has steadily dropped overtime, says Steinberg of theUniversity of California, SanDiego. In the 1960s, he recalls,doctors didn’t worry about totalcholesterol levels until theytopped 280. Today, less than 200is desirable. For LDL alone,under 100 is considered healthy(and under 130 is “near optimal”).But is that good enough? HelenHobbs, a human geneticist at theUniversity of Texas SouthwesternMedical Center in Dallas, notes thatin rural China, LDL levels hover around60 or 70, and

heart disease is about 15 timeslower than in the United States.Hunter-gatherer populations living todayhave LDL levels of about 70. “That wouldsuggest that’s what our species evolved tohave,” says Lee, a Harvard physicianand the network president for PartnersHealthCare System in Boston, which setscholesterol and other guidelines for doctors.Steinberg also wishes that treatment withstatins started much earlier in life. Autopsieson young men killed in the Korean Warfound that even at 18 or 20 years old, theyalready had plaque buildup in their arteries.“We’ve not gone as far as we can go withLDL,” says Steinberg. “The disease startswhen you’re a kid,” and statins “should bestarted as early as 30.”But coaxing LDL down to 70 or so isn’teasy, and statins alone often can’t achievethis. Pharmaceutical companies, on the lookoutfor new cholesterol

strategies, especiallywith their statin drugs losing patent protection,in the mid-1990s renewed efforts to findnovel ways to lower LDL. Schering-Ploughcame up with a drug, ezetimibe, which acts inthe gut to prevent the body from absorbingcholesterol from food and bile. (Statins workdifferently, blocking the synthesis of cholesterolin the liver.) Researchers hoped thatadding a new cholesterol-control mechanismwould make the drug more powerful thanstatins alone. Like the statins, ezetimibesailed through FDA approval because it lowersLDL; it was marketed by Merck andSchering-Plough as Zetia and was combinedwith a statin to be marketed as Vytorin. In2007, Zetia and Vytorin had combined salesof more than $5 billion.But after nearly 4 years on the market,Vytorin unexpectedly opened a Pandora’sbox. In January 2008, Merck announcedpuzzling results from a closely

watchedVytorin trial called ENHANCE, in peoplegenetically predisposed to have very highLDL. In theory, the lower the LDL level, thebetter off the heart should be—butENHANCE suggested otherwise.The study, published in April in The NewEngland Journal of Medicine (NEJM), foundthat Vytorin performed no better than a traditionalstatin in slowing plaque formation incardiac arteries, a common measure of worseningatherosclerosis. The big shock, however,was that Vytorin’s middling artery protectionwas at odds with its superior LDLloweringperformance. The drug pulled LDLdown from 210 or more without treatment toa respectable average of 141, substantiallybetter than 192 for the statin alone. But thatdidn’t translate to healthier arteries.Vytorin was hammered in a second studycalled SEAS, which found that even thoughthe drug lowered LDL, it did not help aorticstenosis, a

disease that obstructs blood flowfrom the heart and can lead to heart failure.Results suggested it might also raise the riskof cancer. Aortic stenosis is a different diseasethan atherosclerosis, but there had beensome research suggesting that the two wereconnected. SEAS, led by Pedersen, was publishedin September, also in NEJM.Reaction to the disconnect between LDLbenefits and heart health in these studies wasswift and passionate. Some even questionedthe link between LDL and heart disease,wondering whether lowering the first reallydoes prevent the second. Most scientists stillagree that it does and that statin drugs reducethe risk of heart disease. But the broaderquestion—whether nonstatin drugs thatlower LDL have the same effect—is now upin the air. “For me, it’s an epiphany,” says Nissen, chair of cardiovascular medicineat the Cleveland Clinic in Ohio,

whoadmits to having been a longtime LDL cheerleader.“It says, ‘Let’s be careful when youcome up with a new class of drugs that lowersLDL by another mechanism.’ ”Others say the results have been ridiculouslyoverblown. “I’ve never seen suchherd mentality in my life,” especiallyamong cardiologists and the media, saysHobbs, who is convinced that LDL is a reliableindicator of heart-disease risk. TheENHANCE work is “just a terrible paperfrom beginning to end.” She criticizes thestudy for lasting only 2 years and for relyingon an indirect measure of heart disease.Hobbs points to recent genetic work inLDL that counters the clinical trial data andunderscores how potent LDL can be. About8 years ago, she set about studying the genesof people in the Dallas area with very lowLDL cholesterol. She discovered that 1 in 50African Americans has a genetic

mutationthat appears to lower LDL by about 30%,and heart-disease risk over 15 years bynearly 90%—far more than statins. This,says Hobbs, is because genetics, unlikedrugs begun at age 50, affects an individualover a lifetime.Hobbs once asked the leaders of the FraminghamHeart Study how many participantswith an LDL of below 70 throughout life hadsuffered a heart attack. The answer? One. “Itcan happen, but it’s very, very rare,” she says.The HDL puzzleIf LDL is weathering some controversy, HDLis in a much deeper puddle. Data backingHDL’s role in heart disease are much sparser,and HDL’s function is much more poorlyunderstood, as underscored by a recent clinicaltrial disaster.Unlike LDL, which has one major receptoron the liver and one primary function,HDL appears to have a hand in inflammation,infection, blood clotting, oxidizingmolecules, and more.

“Nobody has reallyshown,” even in animals, that HDL carriescholesterol out of arteries, as is generallybelieved, says Anne Tybjærg-Hansen, whostudies genetics and heart disease at CopenhagenUniversity Hospital in Denmark.Still, animal studies have consistentlyshown that raising HDL prevents heart disease.Researchers, physicians, and drug companieshoped that drugs to raise HDL couldbecome new arrows in their arsenal, making adifference to the many people who lower theirLDL levels but go on to have heart attacks.The first dramatic booster of HDL wasPfizer’s drug torcetrapib. That therapy—andother HDL modulators still in development—was inspired partly by a group of families inJapan who have very high HDL and very fewheart attacks. They carry mutations in a genecalled cholesteryl ester transfer protein(CETP). Torcetrapib goes after this target toincrease HDL

levels, and by all measures itdoes so remarkably well. But in the same waythat Vytorin lowered LDL but didn’t actuallyhelp people in ENHANCE and SEAS, torcetrapibraised HDL with no measurable healthbenefit, and a striking downside. “That drugkilled more people than it actually saved,” says Kastelein, a vascular medicine specialistat the Academic Medical Center at the Universityof Amsterdam in the Netherlands, whoparticipated in the torcetrapib trial.The 15,000-person trial found that peopleon torcetrapib were 60% morelikely to die than those taking thestatin Lipitor. In December 2006,after investing some $800 millionin torcetrapib, Pf izer hastilystopped the trial, abandoned thedrug, and recommended that trialparticipants stop taking it immediately.It’s still not known why torcetrapibhad the effects it did, some ofwhich were head-scratchers, suchas

deaths from sepsis. More important,no one knows yet whether those effectswere a fluke of this particular drug or an ominoushint that boosting HDL is dangerous.“HDL is a good epidemiologic marker forrisk,” says Rader, a preventive cardiologistat the University of Pennsylvania, meaningthat people with high HDL levels can restassured that their chance of suffering heartdisease is lower than average. But it’s “not thebest surrogate for drug development.”Genetics studies are trying to parse a linkbetween HDL and heart disease, examiningwhether people have heart-disease risks thattrack with HDL levels. Some have uncovereda connection, and others have not. Tybjærg-Hansen spent several years trying unsuccessfullyto link low HDL to atherosclerosis inhumans. An analysis of population studies inDenmark covering roughly 44,000 people,published by hergroup in June

in TheJournal of the AmericanMedical Association,found no increasein heart diseaseamong thosewith low HDL causedby a set of geneticmutations. Tybjærg-Hansen admits that her work gives HDLbelievers “high blood pressure.” “I thinkHDL is a bystander; it’s a lipid transporter inthe body somehow, but I don’t think it hasanything to do with risk” of heart disease,she says.Tybjærg-Hansen blames ties between academicsand drug companies that she says preventnew paradigms from bubbling up. Yale’sKrumholz agrees, noting that markers in theblood may have more complex interactionswith disease than thought or may not track asexpected if a drug has certain side effects.“This recent group of trials should fundamentallychange the way we think,” he says.“Knowing what [drugs] do to LDL and HDLdoesn’t tell you what they’ll do to

people.”Companies haven’t abandoned HDLdrugs, however. Rader, who is working on suchtherapies in concert with drug companies,admits that he and the companies putting upthe money are taking a risk, especially in lightof torcetrapib’s downfall. “But my view is, weneed to test this,” he says, to determine whethertorcetrapib is a fluke or whether it’s conveyinga broader message.A handful of large clinical trials are testingwhether targeting HDL can help; results won’tbe available for several more years. “Our onemajor attempt to increase HDL failed,” says Ridker, a cardiologist at Brigham & Women’s Hospital in Boston. It’s time, he says,to look beyond cholesterol when consideringhow to stop atherosclerosis.A new strategyRidker is focused on another factor that hethinks is involved in heart disease: c-reactiveprotein (CRP). High CRP has long been

considereda marker of inflammation, often foundin people with other risk factors, such as obesityand diabetes. But Ridker goes a step further,arguing that when CRP is lower, plaquesare less likely to break open and kill their hosts.And not only that: He and a growing numberof others argue that statin drugs can wrestledown CRP levels, just as they reduce LDL.The theory that statins do more than lowercholesterol is controversial. “It just amazes methat so many people have bought into this ideathat statins have all these magical properties,”says Rader, who is dubious that the drugs haveanti-inflammatory effects that prevent heartdisease. A survey published in 2005 of 19 trialswith more than 80,000 people found that itdidn’t matter how LDL was reduced; theeffect on heart disease was the same.Scientists will soon have more hard data tohelp them assess whether controlling

inflammationwith statins helps the heart. Ridker isrunning an unusual clinical trial namedJUPITER, which offers statins to people withan LDL of under 130—considered relativelyhealthy—but high levels of CRP. The trial isenormous, with more than 17,000 people in26 countries. By comparing a commonlyused statin, Crestor, made by the drug companyAstraZeneca, with a placebo, it is lookingto see if the high-CRP group reaps a benefit.Like nearly all cholesterol drug trials,this one is funded by the drug’s maker.Results haven’t been released, butAstraZeneca announced in March that it washalting the trial early because of “unequivocalevidence of benefit.” Ridker will elaborate onthe results at a cardiology meeting in November.An outspoken CRP proponent, he holds apatent on a method of measuring CRP in theblood. Hobbs, who disagrees with Ridker’sperspective, thinks he

is picking an easy targetby focusing on people with LDL values ofup to 130, because he “is talking about LDLsthat are still high” compared with “our ancestors,”she says. Pushing LDL to lower levelsin this group, she believes, may explain whyJUPITER succeeded: It cut down LDL, justas earlier statin trials have done.Meanwhile, cardiologists and others areconsidering the clouds in their crystal ball,hoping to reach beyond decades-old strategiesand more precisely predict an individual’srisk of heart disease. Even the biggestproponents of the cholesterol hypothesisadmit that there’s more to the equation thanlowering LDL. “What is causing the rest ofthat risk?” asks Fisher of NYU. It’s the questionto tackle next.

-- Aalt Pater