X4
`
`fl
`
`Circulation
`
`JOURNAL OF THE AMERICAN HEART ASSOCIATION
`
`Preventing Coronary Heart Disease : B Vitamins and Homocysteine
`Gilbert S. Omcnn, Shirley A. A. Beresford and Arno G. Motulsky
`
`Circulation 1998, 97:42!-424
`doi: 10.1 I61/01 .CIR.97.5.42l
`Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514
`Copyright 0 1998 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online
`ISSN: 1524-4539
`
`The online version of this article, along with updated information and services, is
`located on the World Wide Web at:
`
`http://circ.ahajouma|s.org/content/97/S/421
`
`Subscriptions: Information about subscribing to Circulation is online at
`http://circ.ahajoumalsorgllsubscriptionsl
`
`httpzl/www.|ww.com/reprints
`
`Pennissions: Permissions &. Ri hts Desk, Lippineott Williams & Wilkins, a division of Wolters
`Kluwer Health, 35] West Cam en Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax:
`410-528-8550. E-mail:
`joumalpcrmissions@lww.com
`
`Reprints: lnfonnation about reprints can be found online at
`
`Downloaded from hup://circ.ah:ljoumals.org/ by guest on November 7, 20] I
`
`Lilly Ex. 2082
`Sandoz V. Lilly IPR2016-00318
`
`Lilly Ex. 2082
`Sandoz v. Lilly IPR2016-00318
`
`

`
`Preventing Coronary Heart Disease
`B Vitanains and Homocysteine
`
`Gilbert S. Omenn, MD, PhD; Shirley A.A. Beresford, PhD; Arno G. Motulsky, MD
`
`The list of preventable and reversible risk factors for
`
`atherosclerotic cardiovascular disease continues to grow.
`Cigarette smoking, high blood pressure. physical inac-
`tivity, elevated cholesterol, underlying Iipoprotein abnonnali-
`ties, lipoprotein(a), diabetes, overweight, male gender, and age
`are well-established risk factors. During the 1990:, there have
`been many reports associating elevated plasma homocysteine
`levels with arteriosclerotic cardiovascular disease and consistent
`
`evidence that dietary and supplemental folic acid can reduce
`homocysteine levels."
`
`See p 437
`
`The article by Robinson and colleagues’ in this issue of
`Cinrulalion presents further evidence of the importance of
`homocysteine and suggestive evidence that plasma folate and
`plasma pyrixodal-L-phosphate (vitamin B6) are protective
`factors. Their study is part of the European Concerted Action
`Project,‘ which examined 750 patients younger than age 60
`with diagnoses within the previous 12 months of coronary,
`cerebrovascular, or peripheral vascular disease and 800 healthy
`control subjects. The patient groups were young (47 years for
`cases and 44 years for control subjects) and heterogeneous,
`with nonfatal clinical events or symptoms of ameriosclerotic
`cardiovascular disease supported by ECG. angiographic. or
`Doppler evidence;
`the study involved 19 centers in nine
`European countries. Men in the highest quintile for fasting
`total homocysteine (tHcy), compared with the remainder of
`the population, had an estimated relative risk of 2.2 (95%
`confidence interval [CI]. 1.6 to 2.9), with a striking dose-
`response relationship and a more-than-multiplicative interac-
`tion with cigarette smoking and high blood pressure on
`vascular disease risk‘; the corresponding esdmated relative risk
`for coronary heart disease was similar (2.0; 95% CI 1.6 to
`2.8). (tHcy is the sum of homocysteine and homocysteinyl
`moieties of oxidized disulfides, homocystine, and cysteine-
`homocysteine.)
`Robinson and colleagues’ examined three B vitamins in
`detail to detennine their effects on fasting and post-methio-
`nine-loading tHcy levels and any independent effecu on
`cardiovascular disease risk. The results should be considered
`
`preliminary. Low folate and low vitamin B6 levels were
`
`The opinions expressed in this editorial are not necessarily those of the
`editors or of the American Heart Association.
`From the School of Public Health and Community Medicine (G.S.O.),
`Departments of Epidemiology (S.A.A.B.) and of Medicine and Genetics
`(A.G.M.), University of Washington. Seattle, Wash.
`Reprint requests to Gilbert S. Omenn. MD, PhD. Executive Vice
`President for Medical Affairs. The University ofMichigan, Ann Arbor. MI
`48109-0624.
`E-mail gomenn@umich.edu
`(Cirailarien. 1998;97:421-424.)
`0 1998 American Heart Association, Inc.
`
`statistically significantly more frequent among patients than
`among control subjects; a similar tendency for plasma B12 was
`not statistically significant. The inverse association with disease
`for folate was in part accounted for by increased tHcy levels,
`but the association for vitamin B6 was not. Relative risks for
`
`the top quintile of fasting tHcy and for high postload tHcy
`were 1.69 (CI, 1.26 to 2.26) and 1.62 (CI, 1.22 to 2.16),
`respectively. compared with all other quintiles, after adjust-
`ment for the vitamin effects. These relative risks are similar to
`
`the weighted means of other studies; it should be noted that
`one must scrutinize reported results for the choice of compar-
`ison groups and the units of change in homocysteine when
`comparing estimated relative risks between studies.’
`The findings regarding folate and B6 by Robinson et al’
`would be stronger if they better matched the metabolic roles
`of the vitamins in homocysteine metabolism. Homocysteine
`is formed from the sulfur-containing essential amino acid
`methionine. Homocysteine can be transsulfurated to cys-
`teine via two B6-dependent reactions or remethylated to
`methionine via 812- and folate-dependent reactions. Be-
`cause fasting levels are more influenced by remethylation
`and post-methionine-load levels are more influenced by
`transsulfuration, one would expect folate (and B12) to be
`acting primarily on fasting levels and B6 to be primarily
`acting on postload levels. Both measures of tl-lcy were
`investigated in the 1550 study participants, but no such
`differential effects of folate and B6 were found.
`
`In one of several reports since the meta-analyses by our
`youp,” Verhoef et a1‘ found relative risks of 1.3 (CI, 1.0 to
`1.6) for each 1 SD increase (5 trmol/L) in fasting tHcy in a
`comparison of 131 patients with 88 less severely affected
`patients with coronary artery disease and 101 population
`control subjects. Within patient: and within control subjects.
`there was the expected inverse relationship between each of
`the three B vitamins and ti-lcy levels; but, contrary to expec-
`tations, pyridoxal-5-phosphate and folate levels were not lower
`in patients compared with the combined control groups.
`Among men who had received routine examinations in
`London, tHcy was strongly associated with death from ische-
`mic heart disease (estimated relative risk, 2.9; Cl, 2.0 to 4.1.
`after adjustment for apolipoprotein B and blood pressure).’
`Nygarcl et al' reported mortality results for 587 patients with
`angiographically confirmed coronary artery disease. After a
`median follow-up of 4.6 years. 64 had died (50 from cardio-
`vascular causes). There was a striking graded relationship
`between plasma tHcy and overall mortality (eg, 25% of those
`with tHcy levels of 215 umol/L had died compared with 4%
`of those with levels of <9 p.mol/L). ti-icy levels were strongly
`related to history of myocardial
`infarction.
`left ventricular
`
`Downloaded from hnp:/lcirc.ahajoumnl42>1-y by guest on November 7, 201 1
`
`Lilly Ex. 2082
`Sandoz V. Lilly IPR2016-00318
`
`Lilly Ex. 2082
`Sandoz v. Lilly IPR2016-00318
`
`

`
`422
`
`Preventing CI-ID: B Vitamins and I-Iomocysteine
`
`ejection fraction, and serum creatinine level but much less
`related to extent of coronary artery disease on angiography.
`Not all new studies are consistent regarding the risk of tHcy.
`An updated analysis of Physicians’ Health Study data yielded a
`relative risk for elevated tHcy of only 1.3 (CI. 0.5 to 3.1).’ An
`analysis of the MRFIT cohort showed no effect of ti-Icy after
`adjustments for other variables (relative risk, 0.94; CI, 0.56 to
`1.56).” The ARIC study showed no association between the
`incidence of coronary heart disease and tHcy (A Folsom,
`unpublished data, 1997).
`Thus, many questions remain regarding the relationship of
`folate, vitamin B12. and vitamin B6 to levels of tHcy; the
`relationship of homocysteine to cardiovascular risk; and the
`best ways to demonstrate and recommend risk reduction for
`individual patients and for populations.
`
`What Is the Best Estimate for the Increased
`
`Risk of Coronary Heart Disease Mortality
`Associated With Elevated tHcy Levels?
`On balance, based on the references above and the meta-
`
`analysis,‘ we believe the best estimate is a relative risk of 1.4 for
`the diflerence between tHcy levels of >15 umol/L compared
`with levels of <10 umol/L after adjustment for other cardio-
`vascular
`risk factors. This efcct
`is similar to the impact
`expected from a reduction in total serum cholesterol fiom 7.1
`to 4.9 ttmol/L (275 versus 189 mg/dL).'
`
`How Much Supplementation With Folate (or
`Other B Vitamins) Is Desirable and Safe?
`We believe it is desirable to bring tHcy levels down to the
`range of 9 to 10 pmol/L. Diet alone is unlikely to be sufficient
`to increase circulating folate levels and decrease tHcy levels.’
`Also, the bioavailability of folic acid from typical conjugated
`folates in the diet is one half that from supplemenm. Feeding
`studies are needed;
`they must
`incorporate information on
`methylenetetrahydrofolate reductase (MTHFR) genotypes
`(see below) to account for marked variation in response.
`Fourteen intervention studies’ showed substantial decreases
`in average tHcy levels after the administration of 650 to 10 000
`ttg of supplemental folic acid.
`ti-lcy concentrations do not
`appear to reach a plateau until folate intakes approach 400
`pg/d and semm folate reaches 215 ttmol/L." Other studies
`have shown significantly lower tHcy levels in persons taking
`supplements containing folic acid than in those relying on diet.
`Thus, a supplement containing 400 pg of iblic acid is expected
`to produce an average reduction of 5 p.mol/L tHcy in
`nonusers ofsupplements.” A higher dose offolic acid might be
`necessary in some persons (see below).
`Clinical trials with folic acid in at least four countries have
`shown that the incidence of children born with neural tube
`
`closure defects (spina bifida. meningomyelocoele, anenceph-
`aly)
`to mothers with a prior neural
`tube defect—afl'ected
`pregnancy can be reduced by 50% to 75%, leading to an official
`Centers for Disease Control and Prevention recommendation
`
`that all women of childbearing age consume 2400 pg/d
`folate." This public health intervention may be one of the
`most important steps to prevent serious birth defects. Because
`surveys showed that only 7% to 12% of women were taking
`folic acid supplements at the time they became pregnant. in
`
`March 1996 the Food and Drug Administration responded to
`advice and petitions by requiring that cereal grain products be
`fortified at the level of 140 pg of folic acid/100 g ofproduct,
`beginning in January 1998.
`leads to exposure for the total
`Fortification, of course,
`population, which stimulated urgent consideration of potential
`health risks, especially for older adults and any persons with a
`B12 deficiency. One of us recalls that Dr William B. Castle
`informed the first-year Harvard Medical School class that one
`should "never give folic acid without giving B12 first" to avoid
`potential exacerbation of possibly irreversible neurological
`impairments.
`It
`is hard to End documentation for serious
`neurological complications in previously undiagnosed per-
`sons." Nevertheless. to prevent such adverse outcomes even in
`individuals with lack of intrinsic factor for B12 absorption, 200
`to 1000 pg of cobalarnin (B12) could be included in supple-
`ments that contain 400 pg of folic acid.’°"
`
`Is There Evidence of Genetic Variation in
`
`Folate and Homocysteine Levels?
`The original proposal that hyperhomocysteinemia might be
`important in atherosclerosis came from work by McCully“ on
`the vascular pathology of the inborn error of metabolism,
`homocystinuria. However, carriers for homocystinuria (1 in
`400 population). who have hall'—deficiency of cystathionine
`B-synthase, rarely account for the homocysteine elevations
`observed in vascular disease. '5 A common thermolabile variant
`
`of the enzyme MTHFR is homozygous in 10% to 13% of the
`Caucasian population (77 genotype). Such individuals, partic-
`ularly in the presence of suboptimal folate nutrition. tend to
`have slightly elevated homocysteine levels." In two diflerent
`European studies, 35% of workin§ men in the top decile of
`tHcy levels had the TT genotype," as did almost everyone
`with tl-Icy levels of >20 umol/L." Despite several reports that
`the 'I'I"genotype is increased among patients with premature
`vascular disease." no such association could be demonstrated in
`2029 patients with coronary heart disease when compared with
`1639 control subjects across seven difi'erent independent stud-
`ies (A.G. Motulsky_ unpublished dam, 1997). It is noteworthy
`that homozygotes for the TI‘ variant had a 21% reduction in
`tl-Icy’ levels after supplementation with 1000 ug of folic acid
`compared with lesser reductions among heterozygotes (13%)
`and the more common homozygotes for the CC variant
`0%)." It is likely, therefore, that genetic variants in MTHFR
`and other enzymes related to folic acid metabolism (cg.
`methionine synthase) will require individuals to have different
`nutritional and supplement needs." Nutritional needs and
`intervention dosages must be tailored to the underlying patho-
`physiology, a general challenge we still
`face in national
`guidelines for screening and treatment of elevated serum
`cholesterol values.
`
`What Kinds of Prevention Trials Are Needed
`
`and May Be Feasible?
`Cardiovascular researchers have led the way with large-scale
`randomized trials of interventions for patients with specific
`clinical conditions (secondary prevention) and for healthy
`populations with risk factors for developing cardiovascular end
`points (primary prevention trials). The homocysteine hypoth-
`
`Downloaded from hnp://cirr.ahajoumals.orgJ by guest on November 7, 201 I
`
`Lilly Ex. 2082
`Sandoz V. Lilly IPR2016-00318
`
`Lilly Ex. 2082
`Sandoz v. Lilly IPR2016-00318
`
`

`
`Omenn er al
`
`423
`
`esis should be well suited to a direct test with folic acid as the
`intervention. However. as with all trials. the choice of the
`
`study population, choice of the agent or combination of
`agents, determination of an adequate and safe dose. and
`parameters of the design (incidence rates for the end points,
`size ofe&'ect expected. duration of intervention and follow-up.
`and allowance for nonadherence and for competing causes of
`death) must all be taken into account. Potential study popu-
`lations include cardiac patients and healthy populations, and
`genotyped and high tl-lcy subgroups of each. Many investiga-
`tors around the world are considering such trials; pilot trials
`will be needed. The lesson learned from the randomized trials
`in Finland and the United States that tested the seemingly
`compelling hypothesis that B-carotene would reduce lung
`cancer and coronary heart disease incidence and found that this
`vitamin/chemical instead increased lung cancer incidence and
`cardiovascular mortality" is that statistical associations do not
`prove cause-and-effect relationships and do not
`rule out
`adverse effects. Associations should not be described as
`“efi'ects."
`
`Potential trials are complicated by the introduction of folate
`fortification of grains and by increasing recommendations for
`the use of folic acid supplements in the general population and
`in cardiac patients. The Beta-Carotene and Retinol Eflicacy
`Trial (CARET) faced a similar dilemma when B-carotene was
`being added to cereals and multivitamins and was highly
`promoted before the trial results were obtained. It may prove
`impossible to mount a sufficiently powerful u-ial. In that case.
`the stronger the biochemical, pathophysiological, nutritional.
`and genetic information about the cascade from dietary intake
`and genetic variation to circulating levels of folate and tl-lcy,
`the more persuasive will be the current inference of benefit. In
`CARET. we are analyzing the full cascade, liom food fre-
`quency questionnaire estimates of folate intake and polymerase
`chain reaction analyses of genetic variation in the MTHFR
`enzyme to serum folate and B12 levels, tl-lcy concentrations,
`and observed fatal cardiovascular end points (G.S. Omenn,
`M.R. Malinow et al. unpublished data).
`The Vitamins in Stroke Prevention (VISP) trial is recruiting
`3600 patients with nondisabling strokes to receive a multivi-
`tamin combination containing 2.5 mg of folic acid, 25 mg of
`B6, and 0.4 mg of B12 versus a multivitamin with 20 mg of
`folic acid, 0.2 mg of B6, and 6 mg of B12, with a primary end
`point of recurrent stroke and secondary end points of death
`from cardiovascular disease or myocardial infarction. The trial
`is based on a pilot study of homocysteine lowering in patients
`with acute stroke." In addition, a protocol has been an-
`nounced for a nested case-control study among 30 000 patients
`receiving drugs for heart disease or high blood pressure in
`general practices in Norway.”
`
`What Should Clinicians Do With
`
`Present Knowledge?
`Because we interpret the totality of the current evidence
`linking folic acid, homocysteine, and cardiovascular disease risk
`as remaining strong with respect to the potential benefits of
`increasing folic acid intake on a population-wide basis. we
`recommend that everyone consume 2400 pg of folic acid/d.
`Potentially pregnant women should take more to maximize
`
`tube closure defects.
`the protective effect against neural
`Screening for tHcy levels would be useful for individual risk
`profiles and for targeting eflbrts at adherence or recommenda-
`tions for higher doses. Common multivitamins contain 2 to 3
`mg of B6 and 6 to 9 ug of B12. We have no recommendation
`on B6 because definitive evidence of an inverse association
`
`with tHcy levels and of an optimal dose does not exist. As
`noted above, we recommended’ and urged the Food and Drug
`Administration to mandate inclusion of suflicient B12 in folic
`
`acid capsules (200 to 1000 rig) to ensure adequate absorption
`by passive mechanisms even in the absence of intrinsic factor.
`Inclusion of B12 in the fortified grains deserves consideration,
`as well;
`if it
`is not
`included, B12 should be prescribed,
`especially to protect older individuals with various degrees of
`B12 deficiency."
`
`References
`I. Malinow MR. l-lotnocyst(e)inemia: a common and easily reversible risk
`Ector for occlusive atherosclerosis. Circulation. ‘l990;8l :2004 -2006.
`2. Boushey C]. Bereaford SAA. Omenn GS, Motulsky AG. A quantitative
`asessment ofplasma hornocyrteine as a fietor for vasculardisease: probable
`benefits of increasing folic acid intakes. j/IMA. l995;472:l049-I057.
`3. Robinson K, Arhean K. Reéum H. Brattstriirn L. Boers G. Ueland P,
`Rubba P. Palma-Rr.-is R, Meleady R, Daly l., Wirteman J, Graham I. The
`European Coma: Group. Low circulating folate and vitanrin B. concen-
`trations: risk factors for stroke. peripheral vscular disease. and coronary
`artery disease. Circulation. l993:97:437-443.
`4. Graham IM, Daly LE. Refium HM. Robinson K. Bnmtriirn L. Ueland P,
`Palma-Reis R. Boers G. Shealran R. ltraelsson B. Uiterwaal CS, Meleady
`R, McCaster D. Verhoef P, Wirternan J. Rubba P. Belle! H, Wautrecht
`JC, de Valk HW, Sales Luis AC. Parrot-Rouland FM. Tan KS, l-ligjns l.
`Garcon D, Medrano]S, Condito M, Evans AE. Andria G. Plasma homo-
`eysteine as a risk factor for vascular disease: the European Concerted Action
`Project. JAMA. 1997;277:1775-1731.
`5. Beresford SAA, Boushey C]. Homocysteine. folic acid, and cardiovascular
`disease risk. In Bendich A. Decltelbaum R], eds. Prewnrive Nutrition: 77!:
`Conrprrlrtrrsivr Guidefrr Harlrlr Professionals. Totowa. NJ: Humana Press;
`1997393-224.
`6. Verhoef P. Kok F], Kruysren ACM. Schouten EG. Wittermn JCM.
`Grobbee DE. Ueland PM. Refsum H. Plasma total homocysteine. B
`vitamins, and risk of coronary atherosclerosis. Arteriasder Tltlotnb Vase Biol.
`l997;l7:989-995.
`7. Wald NU]. Watt HC. Law MR. Homoeysteine and isclraemic heart
`disease: resuh ofa prospective study with implicatioru on prevention. Ardt
`Int Med. In press.
`8. Nygard O. Nrdrehaugjfi. Reéurn H, Ueland PM, Faratad M. Vollset SE.
`Plasma homocysteine levels and mortality in patients with coronary artery
`disease. N Eng!) Med. l99'7:337:230-236.
`9. Chasan-Taber KL. Sclllub _l. Rosenberg IH, Malinow MR. Terry P,
`Tishler PV. Willett W, l-lennekens CH. Stampfer M]. A prospective study
`offolate and vitamin B. and risk ofmyocardial infirction in US physician.
`J Art Coll Nurr. 1996;15:136-143.
`I0. Evans R\V. Shaten B]. I-lempel JD. Cutlrrjh. Kuller LH. I-lomocyst(e)ine
`and risk of cardiovascular disease in the Multiple Risk Factor Intervention
`Trial. Almioatlel Tlrrorrtlr Vast Biol. l997;l7:l947-I953.
`ll. Selhub].]acques PF, Wilson PWF, Rush D. Rosenberg ll-l. Vitamin status
`and intake as primary determinants of hornocytteinemia in an elderly
`population. JAMA. 1993;170:2693-2698.
`I2. Morbidity and Mortality Weekly Report. Recommendations for the use of
`folic acid to reduce the number of cases ofspina bifida and other neural
`tube defects. MMIVR. l992;4l:l-7.
`l3. Savag DC. Lindenbaum J. Folateqsobahmin interactions. In: Bailey LB.
`ed. Failure in Health and Disease. New Yorlt, NY: Marcel Deltker. I995:
`237-285.
`'
`I4. McCully KS. Vascuhr pathology of homocysteinemia. Ant j Palhol. I969:
`56:1] 1-1 28.
`I5. Motulslry AG. Nutritional ecogenetics: lromocysteine-related arterioscle-
`rotic vascular disease. neurfl tube defects. and folic acid. Arrrj Hum Genet.
`l996:58:l7-20.
`
`Downloaded from http://cirt.ahajoumaIs.org/ by guest on November 7, 20! I
`
`Lilly Ex. 2082
`Sandoz V. Lilly IPR2016-00318
`
`Lilly Ex. 2082
`Sandoz v. Lilly IPR2016-00318
`
`

`
`424
`
`Preventing CI-ID: B Vitamins and Homocysteine
`
`17.
`
`18.
`
`I6. Malinow MR. Nreto F]. Kruger WD. Duell PB. Hess DL. Gluclunan RA.
`Block PC. Holzgaug CR. Anderson PH. Seltzer D. Upton B. Lin QR.
`The efleets offolic acid supplernenlation on plain: total hornocysteine are
`modulated by multivitamin use and metlrylenenernlrytlrofolate reductase
`genotypes. Armiotder Tlunrnb Vat Eel. l9W:l7:I I57-‘H62.
`Harmon DL. Woodaide JV. Yarnell JWG. MeMa.tter D. Young IS.
`McCrum EE. Gey KF. Whitehead AS. Evans AE. The common ‘thermo-
`IaI>ile' variant of methylene tell-ahydrofolate reductate is a major deter-
`minant of mild liyperlromoeytteinaemia. QJ Med. 1996;89:571-577.
`Guttormren AB. Ueland PM, Nestlrus l, Nyprd O. Schneede J. Vollser
`SE. Refsum H. Determinants and vitamin responsiveness of intermediate
`lryperhornocysreinemia (>40 umol/L):
`the Hordaland I-lomocysteine
`J Clin luverr. I996:98:2l74-2183.
`Molloy A. Sean D. Mills]L. Kirk: PN. Whitehead AS. Rarrubottotn D.
`Conley MR, Weir DG. Scott ]H. Therttlolabile variant of 5.I0-rnetlry-
`
`I9.
`
`lennetrahydrofolate rt.-ducrase associated with low red-cell folatesz impli-
`cations for folare intake recornrneudarions. lnnru I997;349:l59l-I593.
`Ornenn GS. Goodman GE. 'l'1Iomqui.1t MD. Balmes J. Cullen MR, Glass
`A. Keogh JP. Meyskeru Fl. Jr. Valanis 8. Williams. JH Jr. Bamlurt S.
`Huturtar S. Efecu of a combination of beta-carotene and vitamin A on
`lung cancer incidence.
`total mortality. and cardiovascular mortality in
`smokers and asbestos-exposed workers. N Eng! J Med.
`l996;334:
`l I50-I 155.
`. Howard V]. Clrarnblen LE. Malinow MR. Lelkowilz D. Toole JF. Results
`of a hornocysr(e)ine lowering pilot study in acute stroke patients. Strobe.
`1997331234.
`Aurtnes 1. Protocol for a nested care-control study with folie acid in
`llyperltorrtocysteinertria. Can] Carliol. l997;l3(suppl B):3l5B. Abstract.
`
`22.
`
`Kn Worms: Editorials I heart disease I vitamins I hornocysteine
`
`Downloaded from lmp://circ.ahajoumals.org./ by guest on November 7, 201 I
`
`Lilly Ex. 2082
`Sandoz V. Lilly IPR2016-00318
`
`Lilly Ex. 2082
`Sandoz v. Lilly IPR2016-00318