The association of elevated plasma homocyst(e)ine with progression of symptomatic peripheral arterial disease

The association of elevated plasma homocyst(e)ine with progression of symptomatic peripheral arterial disease Lloyd M. Taylor, Jr., M D , R o b e r t D. DeFrang, M D , E. John Harris, Jr., M D , and J o h n M. Porter, M D , Portland, Ore Plasma homocyst(e)ine (the sum of free and bound homocysteine, homocystine, and the mixed disulfide homocysteine-cysteine, expressed as homocysteine) levels were determined by high performance liquid chromatography in 214 patients with symptomatic (claudication, rest pain, gangrene, amputation) lower extremity arterial occlusive disease and/or symptomatic (stroke, cerebral transient ischemic attacks) cerebral vascular disease and in 103 control persons. Mean plasma homocyst(e)ine was significantly higher in patients than in controls (14.37 - 6.89 nmol/ml vs 10.10 -+ 2.16, p < 0.05). Thirtynine percent of patients (83 of 214) had plasma homocyst(e)ine values greater than control mean + 2 standard deviations. Plasma homocyst(e)ine values were contrasted to age, male sex, diabetes, hypertension, smoking, renal failure, and plasma cholesterol. No difference was found in the incidence and/or level of any of these risk factors when patients with normal plasma homocyst(e)ine were compared to those with elevated plasma homocyst(e)ine, both by univariate and multivariate analysis. Patients with elevated plasma homocyst(e)ine were more likely to demonstrate clinical progression of lower extremity disease and of coronary artery disease, but not of cerebral vascular disease than were patients with normal plasma homocyst(e)ine, and the rate of progression was more rapid (p = 0.002). Progression of lower extremity disease as assessed in the vascular laboratory was also more common in patients with elevated plasma homocyst(e)ine (p = 0.01). We conclude that elevated plasma homocyst(e)ine is an independent risk factor for symptomatic lower extremity disease or cerebral vascular disease or both. Symptomatic patients with lower extremity disease and with elevated plasma homocyst(e)ine also appear to have more rapid progression of disease. (J VAse SURG 1991;13:128-36.)

In Western countries atherosclerosis is sufficiently common in aged individuals to blur the distinction between disease state and normal aging. Despite the widespread prevalence of atherosclerotic changes, many individuals remain free of symptoms of atherosclerosis during their entire lives. Current research related to atherosclerosis has appropriately focused both on factors associated with development of the symptomatic state as well as factors associated with rapid disease progression. A large number of exper-

From the Division of Vascular Surgery.,Department of Surge~, Oregon Health SciencesUniversity,Portland. Supported in part by grant No. RR 00334 from the General Clinical Research Center Branches, Division of Research Resources, National Institutesof Health. Presented at the Forty-fourthAnnualMeeting of the Societytbr Vascular Surgery, Los Angeles,Calif., June 4-6, 1990. Reprint requests: LloydM. Taylor Jr., MD, AssociateProfessor of Surgery, Division of Vascular Surgeu OP-11, Oregon Health SciencesUniversity,3181 S. W. Sam JacksonPark Rd., Portland, Or 97201. 24/6/24913 128

imental, epidemiologic, and interventional studies have examined the relationship between lipid metabolism and atherosclerosis. Clearly, atherosclerosis, particularly coronary atherosclerosis, is more likely to occur in persons with elevated plasma lipids, ~ and recently an improved outcome has been confirmed in individuals in whom lipid values have been therapeutically lowered. 2 Despite the clear lipidatherosclerosis relationship, a large number of individuals with symptomatic atherosclerotic disease have no evidence o f abnormal lipid metabolism? Homocysteine is a thiol-containing amino acid formed from metabolism of methionine and is oxidized in plasma to the disulfide homocystine and to the mixed disulfide homocysteine-cysteine. These three compounds occur naturally in plasma in both free and protein bound forms and are hereafter referred to collectively as homocyst(e)ine (H(e)). Homocystinuria, an inborn error o f metabolism in which homocysteine accumulates abnormally in tissues and plasma and is excreted in large amounts in the urine causes rapidly progressive fatal atheroscle-

Volume 13 Number 1 January 1991

Plasma homo~st(e)ine and peripheral arterial disease progression 129

rosis in childhood. 4 Homocystinuria results from deficiency of one of several enzymes, the most frequent of which is deficiency of cystathionine b-synthase. Since homocystinuria results from an homozygous state, speculation has centered on whether individuals heterozygous for homocystinuria are more likely to develop symptomatic atherosclerosis than are healthy individuals, as well as whether elevation of plasma H(e) from other causes might predispose to accelerated atherosclerosis. Studies by Wilcken and Wilcken ~ and by Boers et al. 6 have demonstrated an increased incidence of coronary disease and peripheral arterial disease, respectively, in persons with an abnormal increase in plasma homocysteine in response to oral methionine loading. By culturing fibroblasts from affected persons and from controls Boers et al.6 demonstrated that the abnormal increase in plasma homocysteine occurred in individuals heterozygous for deficiency of cystathionine b-synthase.6 Recently Kang et al.7 have shown that most plasma homocysteine is bound to plasma proteins, and that elevation of protein bound homocysteine in the fasting state is associated with an increased incidence of coronary artery disease. Studies from our laboratory demonstrated a significantly increased incidence of elevated plasma H(e) in individuals with symptomatic peripheral arterial disease when compared with asymptomatic controls, s'9 Given this association, the obvious question is whether atherosclerosis in persons with elevated plasma H(e) progresses more rapidly than in persons with normal levels, and whether persons with elevated plasma H(e) are more likely to become clinically symptomatic. Lower extremity occlusive disease (LED), and extracranial carotid artery occlusive disease (CVD) are especially suited for studies relating progression of disease to risk factors for atherosclerosis because of the ease with which these vascular beds can be repeatedly quantitatively examined noninvasively in the vascular laboratory. ~°~3 The results of plasma H(e) testing in patients with peripheral arterial disease under long term follow-up in our vascular clinic, and the relationship of plasma H(e) levels to progression of disease as assessed clinically as well as objectively in the noninvasive vascular laboratory form the basis for this report. METHODS Plasma homocysteine testing

Plasma H(e) was determined by a modification of the method of Smolin and Schneider, ~4the details of which we have previously reported, s'ls In brief,

venous blood (5 ml) was collected in a vacuum tube containing a drop of K2-EDTA. Plasma was separated by centrifugation and stored frozen until assayed. The thawed samples were diluted with water and urea. Disulfide reduction was performed by incubation with NaBH4 at 50 ° C for 30 minutes. Proteins were separated by centrifugation. Automated high performance liquid chromatography by use of a mobile phase of 0.1 mol/L monochloroacetic acid and 1.8 mmol/L octylsulfate, pH 3.2, was performed. Total plasma H(e) values reported as homocysteine consisted of the sum of free and protein bound homocysteine, homocystine, and homocysteine-cysteine mixed disulfide. Noninvasive vascular laboratory cerebrovascular testing Cerebrovascular testing was performed by duplex scanning as described by Blackshear et al.,6 and Roedeter et al.17 Technologists performing the examination were unaware of the results of the plasma H(e) testing. The examinations were performed with either a Diasonics DRF-400 scanner (Diasonics, Milpitas, Calif.), or an Acuson 128 scanner (Acuson Inc., Mountain View, Calif.). The common carotid arteries, internal carotid arteries, and external carotid arteries were imaged and Doppler frequency spectra were recorded from proximal and distal sites in each artery as well as at the specific sites of any stenosis. Arteries were classified into one of six categories (normal, 1% to 16% stenosis, 17% to 49% stenosis, 50% to 79% stenosis, 80% to 99% stenosis or occluded) by use of published frequency and waveform criteria. ~6"17The accuracy of this testing in our laboratory has been established by blinded comparison of the results of scanning 534 arteries with angiograms, and exceeds 90% for all degrees of stenosis." For purposes of this study patients were classified as having vascular laboratory evidence of CVD if stenosis in at least one carotid artery exceeded 17% diameter reduction and atherosclerotic plaque was present at the site of the stenosis. Noninvasive vascular laboratory lower extremity testing Testing for the presence of LED was performed by palpating femoral, popliteal, and ankle pulses, and recording Doppler analog waveforms at high thigh, above-knee, below-knee, and ankle levels. Dopplerderived pressure ratios were determined at each level with the highest arm pressure used as the reference value? 8 Patients were classified as having vascular laboratory evidence of LED if the arterial pressure

130

Taylor et al.

Table I. Drugs assessed for correlation with elevated plasma H(e) Aspirin Digoxin ~3Adrenergicblocking agents c~adrenergicblockingagents Nitrates Calcium channel blocking agents Diuretics Multivitamins Angiotensin convertingenzymeinhibitors

ratio in one ankle was equal to or less than 0.85. The accuracy of this testing in our laboratory has been established at greater than 90% by blinded comparison to the results of over 300 arteriograms (unpublished data). Patient selection Clinical patients were selected for plasma H(e) testing who met the following criteria: (1) Presence of symptomatic CVD (stroke, cerebral transient ischcmic attacks) and/or presence of symptomatic LED (intermittent claudication, ischcmic ulceration, ischemic rest pain, ischemic gangrene, amputation for ischemia); (2) Vascular laboratory evidence of arterial disease at the symptomatic site; (3) Clinical symptoms for at least 1 year to permit assessment of clinical arterial disease progression. Consecutive patients meeting these criteria underwent plasma H(e) testing at the time of hospital admission or vascular clinic visit. Patients with serial vascular laboratory examinations recorded at the time of the study also underwent assessment of vascular laboratory progression of disease as described below. Control patients were healthy volunteers without symptoms of CVD, LED, or coronary athcrosclcrosis recruited from the Oregon Regional Primate Center staff and from the Portland Veterans Administration Hospital. Informed consent for plasma sampling was obtained from these individuals in accordance with protocols fully approved by the respective Institutional Review Boards. Plasma H(e) values of these control patients have been previously reported, s Risk factor assessment The presence of risk factors for atherosclerosis was determined by medical record review for each patient. All records were reviewed by one of the authors who was unaware of the results o f the plasma H(e) testing. Age, sex, race, and presence of diabetes, renal failure, hypertension and smoking were recorded for each patient. Plasma cholesterol levels

Journal of VASCULAR SURGERY

were recorded. Patients with angina pcctoris, myocardial infarction, congestive heart failure, or previous coronary bypass were said to have coronary artery disease (CAD). History of previous vascular surgery including coronary, carotid, and lower extremity revascularizations was recorded. Treatment with drugs commonly used in this patient group (Table I) was also recorded.

Definition o f progression o f disease Progression of disease was defined by clinical criteria for CVD, LED, and CAD, and by vascular laboratory criteria for both CVD and LED. Clinical criteria for progression of CVD included the occurrence of one or more of the following events during the period of clinical follow-up (defined as the period after the onset of symptomatic atherosclerosis and the treatment of initial symptoms). Thus for a patient who initially had transient ischemia attacks and was treated by carotid endartcrectomy, only new events occurring after the carotid endarterectomy were regarded as clinical progression events: (1) stroke, (2) onset of cerebral transient ischemic attacks, (3) performance of carotid cndartcrectomy for new symptoms or for progression of vascular laboratory determined asymptomatic disease occurring after initial assessment. Clinical criteria for progression of LED included the occurrence of one or more of the following: (1) onset of claudication, (2) appearance of ischemic ulceration or gangrene, (3) onset of ischemic rest pain, (4) need for lower extremity, amputation for new symptoms occurring after initial assessment (5) need for lower extremity, revascularization to correct new symptoms. Clinical progression of CAD was defined as the following (1) new occurrence of acute myocardial infarction, (2) occurrence of new symptoms of angina pectoris, (3) new occurrence of congestive heart failure. Performance of coronary bypass was considered progression of CAD if it was performed for newly occurring symptoms. The occurrence of these disease progression events was determined by medical records review for each patient by one of the authors who was unaware of the results of plasma H(e) testing. An indication of the rate of clinical progression of atherosclerotic disease was calculated by dividing the number of clinical progression events that occurred (CVD plus LED plus CAD) in each patient by the number of months the patient was under clinical follow-up. This value is referred to as the clinical progression index (CPI). Progression of CVD by vascular laboratory criteria was defined as an increase in carotid stenosis by one category, H or progression to carotid artery oc-

Volume 13 Number 1 lanuan, 1991

Plasma homo~st(e)ine and peripheral arterial diseaseprogression 131

Table II. Number and age of control persons and patients** Age (yr) *

Number Controls

Patients

Sex

<60 yr

>60 yr

Men Women Total

35 39 74

18 11 29

110 104 214

Controls

Patients

<60 ~

>60 yr

42 ± 10.4 40 ± 9 41 ± 9.6

66 ± 4.8 65 ± 2.1 66 ± 3.9

64 ___ 13.1 67 ± 11.1 65 ± 11.5

* Values are mean ± standard deviation. ** Data regarding control persons previously reported (ref. 8).

clusion. Progression of LED by vascular laboratory criteria was defined as a decrease in ankle/brachial pressure ratio in at least one extremity of at least 0.15. Data analysis Mean H(e) values were compared between patient and control groups by use of Student's t test. Risk factors present in patients with elevated H(e) (plasma H(e) > control mean + 2 standard deviations) were compared to those in patients with normal H(e) by means of chi-square analysis for categoric variables and t testing for continuous variables, as well as by multiple logistic regression analysis. The presence of progression of LED, CVD, and CAD by clinical criteria and by vascular laboratory criteria and the rate of progression were also tested individually by chi square, t testing, and Wilcoxon rank testing, as well as by multiple logistic regression analysis. RESULTS Symptomatic patients versus controls Two hundred fourteen patients with symptomatic peripheral arterial disease (men 110, 51%; women 104, 49%; mean age 65 years), and 103 asymptomatic (control patients) (men 53, 52%; women, 50, 48%; mean age 48 years) were studied. The sex distributions of these two groups were not significantly different. The difference in ages was significant (t test, p = 0.01). The control patients were therefore divided into older (> 60 years) and younger (<60 years) groups. The age and sex characteristics of the patients and control persons are seen in Table II. The mean plasma H(e) level in control patients was not significantly different when younger men or women were compared to their older counterparts as seen in Table III. Plasma H(e) values in control men were higher than in control women. These results of plasma H(e) testing in controls have been previously reported. 8 The difference between men and women was not present in patients with peripheral arterial disease. For both men and women,

Table III. Plasma H(e) values in patients and controls* Controls <60 yr

>60 yr

Patients

10.74 "_ 2.16 9.04 ~' 2 1 6 10.10 + 2.16

15.(J~3 ± 7.42 1~(~2 ± 6.22 14.37 ± 6.89

I

Men Women All

111.18 + 3.58 ' 8.~2 + 2.82 9.80 ± 3.44

i

Control values previously reported (ref. 8). *All values in nmol/ml, mean ± standard deviation. Brackets connect pairs with p < 0.05 by Student's t test.

the mean plasma H(e) values of patients with peripheral arterial disease were significantly higher than those of control persons as seen in Table III. Eightytwo of 214 patients (38%) had plasma H(e) values greater than 2 standard deviations above the control mean plasma H(e) value. For purposes of risk factor comparison, these patients are referred to as "elevated H(e)", whereas those with lower values are hereafter referred to as "normal H(e). Risk factors present in patients with elevated plasma H(e) versus those in patients with normal plasma H(e) The incidence of known risk factors for atherosclerosis including age, diabetes, smoking, male sex, hypertension, and renal failure was compared in elevated H(e) patients and those with normal H(e). The results are shown in Table IV. No significant difference was found in the incidence of any of these factors in either group, both when the factors were examined individually and when examined by logistic regression analysis. Cholesterol levels were contrasted with H(e) levels. No significant relationship was found to plasma H(e) (Table IV). Elevated plasma homocysteine was associated with use of diuretics when examined both by chi-square analysis (p = 0.003) as well as by logistic regression (p = 0.03, data not shown). No associ-

132

Journal of VASCULAR SURGERY

Taylor et al.

Table IV. Risk factors present in patients with elevated plasma H(e) (n = 82) versus patients with normal plasma H(e) (n = 132) Factor Male sex (%) Race (% white) Smoking (%) Diabetes (%) Hypertension (%) Renal failure (%) Plasma cholesterol m g / 1 0 0 ml, mean ± SD

Elevated homo~steiue

Normal homo~.steine

p Value*

p Vaduet

46% 96% 80% 28% 57% 9% 186 + 57

54% 97% 78% 22% 45% 4% 195 ± 51

ns ns ns ns ns ns ns$

ns ns ns ns ns ns ns

ns, Not significant. *Probability by chi square analysis. t Probability by logistic regression analysis. ¢ Probability by Student's t test.

ation was found between elevated plasma homocysteine and use of any of the other medications listed in Table I, nor with any combination of these medications (data not shown). Progression o f disease in patients w i t h elevated plasma H(e) versus patients w i t h normal

plasma H(e) The age at onset of symptoms of atherosclerotic disease did not differ significantly between the elevated H(e) (62 years _+ 11) and the normal H(e) (60 years + 13 years) groups (Student's t test, p = ns, data not shown). No significant differences were found in mean length of follow-up between the two groups, both for clinical follow-up (normal H(e)), 52 months + 51 months; elevated H(e), 55.9 months + 51 months, and for vascular laboratory follow-up (normal H(e), 30 months + 24 months; elevated H(e), 28 months + 25 months (Student's t test, p = ns)) (Tables V and VI). One or more events indicating clinical progression of disease were more likely to have occurred in patients with elevated H(e) (80 of 82, 98%) than in patients with normal H(e) (107 of 132, 81%), a significant difference (chi square, p = 0.001, data not shown). As is demonstrated in Table V, this difference was due to the more frequent occurrence of clinical progression of LED in the elevated H(e) group (73 of 82, 89% vs 94 of 131, 72% ;p = 0.003, chi square), as well as more frequent occurrence of clinical progression of CAD in the elevated H(e) group (32 of 82, 39% vs 24 of 132, 18%;p = 0.001, chi square). The number of clinical events occurring per patient was also significantly higher in the elevated H(e) group both for LED (3.14 _+ 2.94 vs 2.08 --- 2.48,p = 0.008 Wilcoxon rank test) and for CAD (0.72 + 1.08 vs 0.33 + 0.86, p = 0.0004

Wilcoxon rank test). No significant difference was found in the incidence of clinical progression of CVD between the two groups (elevated H(e), 26 of 82, 32% vs normal H(e), 31 of 132, 24%; p = ns, chi square) or in the number of CVD clinical progression events that occurred (0.57 __ 0.96 vs 0.44 ___ 0.93, p = ns, Wilcox rank test). The rate of progression of clinical disease was also significantly more rapid in the elevated H(e) patients when compared to the normal H(e) patients (clinical progression index = 0.14 +_ 0.15 events/month/patient for elevated H(e) vs 0.08 -+ 0.08 events/month/patient for normal H(e); p = 0.002, Student's t test). One hundred sixty-nine patients had serial vascular laboratory examinations for assessment of vascular laboratory progression of disease. Progression of LED as determined in the vascular laboratory occurred significantly more frequently in the elevated H(e) group (34 of 65, 52% vs 34 of 104, 33%; p = 0.01, chi square). No significant difference occurred in progression of CVD as determined in the vascular laboratory between the two groups (elevated H(e), 15 of 65, 23% vs normal H(e), 22 of 104, 21%; p = ns, chi square) (Table VI). Clinical progression of LED (p = 0.05), clinical progression of CAD (p = 0.005), rate of progression of clinical disease (p = 0.02), and vascular laboratory progression of lower extremity disease (p = 0.04) were also all significantly more common in the elevated H(e) group when these variables were examined by logistic regression analysis (Tables V and VI).

DISCUSSION Since the cause of atherosclerosis remains unknown, clinical research efforts have appropriately focused on identifiable risk factors associated with

Volume 13 Number 1 January 1991

Plasma homocyst(e)ine and peripheral arterial disease progression

133

Table V. Clinical progression of disease in patients with elevated H(e) (n = 82) versus patients with normal H(e) (n = 132) Catego~ Length of clinical follow-up (mean months ± SD) No. of LED progression events (mean events/patient ± SD) Presence of any LED progression (%) No. of CVD progression events (mean events / patient ± SD) Presence of any CVD progression (%) No. of CAD progression events (mean events/patient ± SD) Presence of any CAD progression Clinical progression index (mean cvents/ patient / month -+ SD)

Elevated homo~eine

Normal homocysteine

56 ± 51

52 ± 51

ns ~

3.14 + 2.94

2.08 - 2.48

0.008 **

0.02

73(89%)

94(72%)

0.003 *~°*

0.03

0.57 _+ 0.96

0.44 ± 0.93

ns ~

ns

26(32%)

31(24%)

ns ~

ns

0.72 _+ 1.08

0.33 -+ 0.86

0.0004 ~*

0.001

32(39%)

24(18%)

0.001 ~**~

0.002

0.14 ± 0.15

0.08 ± 0.08

0.002 *

0.002

p

p ns

LED, Lower extremi~ disease; CVD, carotid artery disease; CAD, coronary artery disease; ns, not significant (p > 0.05). Probabili~ by Student's t test. Probability, by logistic regression analysis. ~Probability by Wilcoxon rank test. *~* Probabiliq, by chi square.

Table VI. Vascular laboratory progression of disease in patients with elevated H(e) (n = 65) versus patients with normal H(e) (n = 104) Catego~ Length of vascular laboratory follow-up (mean months ± SD) Vascular laboratory LED progression (%) Vascular laboratory CVD progression (%)

Elevated homocysteine

Normal homo~steine

p~

p7~

28 ± 25

30 ± 24

ns

ns

34(52%)

34(33%)

0.01

0.03

15(23%)

22(21%)

ns

ns

SD, Standard deviation; LED, lower extremiw disease; CVD, carotid artery disease. Probability. by chi square. ? Probability. by logistic regression analysis.

disease. Definition of the disease state associated with atherosclerosis is difficult, since this process can be identified to some degree at autopsy in essentially all aged individuals in western countries. Certainly the presence of clinical symptoms, and the occurrence of rapid disease progression are the two most important characteristics distinguishing atherosclerotic disease from atherosclerosis as a normal aging process. The clear cut difference in plasma H(e) levels when symptomatic were compared to asymptomatic control persons confirms previous work from our group, 8"9"~'19 as well as many studies by others, s-7a° demonstrating a positive relationship between elevation of plasma H(e) and symptomatic atherosclerotic disease. The study population consisted of con-

secutive individuals referred to a clinical vascular surgery service. As expected the prevalence of the usual atherosclerotic risk factors was high. It is important to note that whereas the presence of atherosclerotic disease was objectively confirmed (by noninvasive vascular laboratory testing) in the symptomatic patients, similar testing was not performed upon the asymptomatic control patients. Thus this study does not establish elevated plasma H(e) as a risk factor for the presence of atherosclerosis. Rather the data herein confirm elevated plasma H(e) as a risk factor for symptomaticatherosclerosis. Risk factors for and the natural history of asymptomatic atherosclerosis remain largely unknown. The commonly accepted risk factors for athero-

134 Taylor et al.

sclerosis including plasma lipids, smoking, diabetes, hypertension, male sex, and age were not different in symptomatic patients with normal or elevated plasma H(e) levels. The incidence of chronic renal failure was also examined in the two groups since plasma H(e) values have been reported to be elevated in this patient group. 2~ No difference was found in the incidence of renal failure. These observations establish elevated plasma H(e) as a risk factor for symptomatic atherosclerosis that appears to be independent of the other factors analyzed in this study. Comparison of the progression data in this study indicated that individuals with symptomatic atherosclerosis were more likely to demonstrate clinical progression of LED and of CAD during the study if they had elevated plasma H(e). For LED this progression was observed both when assessed clinically by symptomatic events, and when assessed in the vascular laboratory. A similar significant difference was not observed for CVD. The reasons for this difference are presently unclear. Our study population was drawn from a vascular surgery service with a recognized interest in lower extremity ischemia. Obviously some preselection occurred for severe LED. It is unlikely that the difference is explained by the brevity of follow-up. The clinical follow-up (54 months) and vascular laboratory follow-up (29 months) are both well within the range in which significant numbers of patients demonstrate progression of CVD, both in our patient population and in the experience of others. ~°'H Recently, elevated plasma H(e) has been suggested as an independent risk factor for stroke when a population of patients participating in a large longitudinal stroke study was examined, z° The discrepancy between the incidence of clinical progression of LED (89%) and progression of LED as determined in the vascular laboratory (52%) may reflect the longer follow-up interval available for analysis of clinical events. Alternately, this may be a real reflection of the difference in disease progression resulting from subjective versus objective assessment parameters. A similar difference in incidence of LED between subjectively reported symptoms and more objective parameters was noted in the recently reported Multiple Risk Factor Intervention Trial. 2z The pathogenesis of the elevated H(e) levels present in our patients was not examined in this study. Previous work by others demonstrates multiple possible mechanisms that may be responsible. Relative deficiency of cystathionine b-synthase was demonstrated in similar patients by Boers et al. 6 Similar deficiency of methylenetetrahydrofolate reductase

Journalof VASCULAR SURGERY

may also produce elevated plasma H(e) as shown by Kang et al.23 Elevation of plasma H(e) may also result from deficiency of vitamins B6,24 B~22s or folate. 26 Chronic impairment of renal or hepatic function may also elevate plasma H(e). 21 We found an association between diuretic use and elevated plasma H(e). The association was not absolute. Approximately one half the patients using diuretics had elevated H(e); approximately one third of those with normal H(e) also used diuretics. Coull et al.20 recently found a similar unexplained association between elevated H(e) and uric acid elevation. More severe impairment in renal function in the elevated H(e) group (associated with increased use of diuretics) might explain this, but in our patients no difference was found in the incidence of renal failure between the elevated H(e) and normal H(e) groups (Table IV). Whatever the mechanism producing elevation of plasma H(e), accumulated evidence indicates a strong relationship between this finding and symptomatic vascular disease. Homocysteine given to baboons by intravenous infusion produces endothelial denudation, smooth muscle proliferation, and increased platelet consumptions Dysfunction and injury in cultured endothelial cells has been demonstrated both when homocysteine levels were artificially increased in the medium 28 and in cells from individuals with cystathione b-synthetase deficiency when given methionine. 28,29Homocysteine treatment of cultured endothelial cells results in increased activity of factor V as well as prothrombin activation? ° Ultimately, the reason fbr risk factor identification for atherosclerosis is the presumption that modification of these factors will produce clinical benefit consisting of decreased symptoms and/or decreased progression of disease. To date the primary atherosclerosis risk factors identified are not subject to modification (male sex, diabetes), or can be modified only by difficult changes in lifelong habits (smoking, diet) or through administration of medications with significant side effects (hypertension, elevated plasma lipids). Several workers have demonstrated that elevated plasma H(e) can be reduced to normal or near normal levels by administration of folatefl TM or by other similarly innocuous medications? 2 This reduction can be achieved independently of whether or not the individuals being treated have deficient levels of folate before treatment? ~ Whether reduction of elevated plasma H(e) in symptomatic patients with atherosclerosis will result in reduced symptoms and a reduced rate of progression of disease is unknown. In summary, this study confirms previous work demonstrating a significantly increased incidence of

Volume 13 Number 1 Januaq, 1991

Plasma homocyst(e)ine and peripheral arterial disease progression 135

elevated plasma H(e) in persons with symptomatic peripheral vascular disease, which is independent of other known risk factors. In addition patients with elevated H(e) appear to have a significantly higher incidence and rate of progression of clinical LED and CAD as well as a significantly higher incidence of vascular laboratory determined progression of LED. These findings as well as those of others discussed herein, in our opinion mandate prospective studies in which disease progression in patients with symptomatic atherosclerosis is related to plasma H(e) levels. If the findings of the present report are confirmed, then the logical next step is trials evaluating the effect of therapeutic reduction of elevated plasma H(e) on disease progression. If this treatment results in benefit, folic acid or other medications in folate resistant patients may be confirmed as the first effective treatment for symptomatic atherosclerosis that does not require surgery# drugs with significant side effects, or modification of lifestyle, addictions, or habits. The present study suggests that the results of such trials could be effectively determined by use of the noninvasive vascular laboratory for objective confirmation of results within a reasonable period of follow-up. The authors gratefully acknowledge expert assistance with statistical analysis from David Wilson, MS, and Gary, Sexton, PhD. The authors also gratefully acknowledge the assistance ofM. Rene Malinow, MD, Director of the Cardiovascular Research Laboratory, Oregon Regional Primate Center, Beaverton, Oregon, where the H(e) assays were performed. REFERENCES 1. Kannell WB, Castelli WP, Gordon T, McNamara PM. Serum cholesterol, lipoproteins and the risk of coronary, heart disease. Ann Int Med 1971;74:1-12. 2. Lipid research clinics program: the lipid research clinics coronary primary prevention trial results: II. The relationship of reduction in incidence of coronary heart disease to cholesterol lowering. JAMA 1984;251:351-74. 3. Gordon T, Garcia-Palrnieri MR, Dagan A, et al. Differences in coronary heart disease in Framingham, Honolulu, and Puerto Rico. J Chronic Dis 1974;27:329-44. 4. Mudd SH, Levy. HL. Disorders of transsulfuration. In: H Stanbury JB, Wyngarden JB, Fredrickson DS, Goldstein JL, Brown MS, eds. The metabolic basis of inherited disease. New York: McGraw-Hill, 1983;458-503. 5. Wilcken DEL, Wilcken B. The pathogenesis of coronary artery disease: a possible role for methionine metabolism. J Clin Invest 1976;57:1079-82. 6. Boers GHJ, Smals AGH, Trijbels FJM, et al. HeterozygosinT for homocystinuria in premature peripheral and cerebral occlusive arterial disease. N Engl J Med 1985;313:709-15. 7. Kang SS, Wong PWK, Cook HY, et al. Protein bound

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24.

25.

26.

homocyst(e)ine: a possible risk factor for coronary, artery, disease. J Clin Invest 1986;77:1482-6. Malinow MR, Kang SS, Taylor LM Jr, et al. Prevalence of hyperhomocyst(e)inemia in patients with peripheral arterial occlusive disease. Circulation 1989;79:1180-8. Harris EJ Jr, Taylor LM Jr, Malinow MR, Edwards JM, Porter JM. The association between elevated plasma homocyst(e)ine and symptomatic peripheral arterial disease. Surg Forum 1989;40:307-9. Roederer GO, Langlois YE, Jager KA et al. The natural history. of cartoid artery disease in asymptomatic patients with cervical bruits. Stroke 1984;15:605-13. Taylor LM Jr, Loboa L, Porter JM. The clinical course of carotid bifurcation stenosis as determined by duplex scanning. J VASCSURG 1988;8:255-61. Cronenwett JL, Warner KG, Zelenock GB, et al. Intermittent claudication: current results of nonoperative management. Arch Surg 1984;119:430-6. Rosenbloom MS, Flanigan P, Schuler JJ, et al. Risk factors affecting the natural history, of intermittent claudication. Arch Surg 1988;123:867-70. Smolin LA, Schneider JA. Measurement of total plasma cysteamine using high-performance liquid chromatography with electrochemical detection. Anal Biochem 1988;168:374-9. Malinow MR, Sexton G, Averbuch M, Grossman MD, Wilson D, Upson B. Homocyst(e)inemia in daily practice: levels in coronary, artery disease. Coronary Artery Disease 1990;1: 215=20. Blackshear WM Jr, Phillips DJ, Thiele BL, et al. Detection of carotid occlusive disease by untrasonic imaging and pulsed Doppler spectrum analysis. Surgen, 1979;86:698-704. Roederer GO, Langlois YE, Jager KA, et al. A simple spectral parameter for accurate classification of severe carotid disease. Bruit 1974;8:174-7. Baur GM, Zupan TL, Gates KH, Porter JM. Blood flow in the common femoral artery. Am J Surg 1983;145:585-8. Genest JJ, McNamara J, Upson B, et al. Prevalence of familial hyperhomocyst(e)inemia in men with premature coronary artery disease. J Am Coil Cardiol 1990;15:185A. Coull BM, Malinow MR, Beamer N, Sexton G, Nordt F, de Garmo P. Elevated plasma homocyst(e)ine concentration as a possible independent risk factor for stroke. Stroke 1990; 21:572-6. Wilcken DEL, Dudman NPB, Tyrrell PA, Robertson MR. Folk acid lowers elevated plasma homocysteine in chronic renal insufficiency: possible implications for prevention of vascular disease. Metabolism 1988;37:697-70i. Multiple risk factor intervention trial research group. Coronary heart disease death, nonfatal acute myocardial infarction and other clinical outcomes in the multiple risk factor intervention trial. Am J Cardiol 1986;58:1-13. Kang SS, Wong PWK, Zhou J, et al. Thermolabile methylenetetrahydrofolate reductase in patients with coronarv artery disease. Metabolism 1988;37:61 i-3. Smolin LA, Crenshaw TD, Kurtyca D, Benevenga NJ. Homocyst(e)ine accumulation in pigs fed diets deficient in vitamin B6: relationship to atherosclerosis. J Nutr 1983; 113: 2022-33. Brattstrom L, Israelsson B, Lindgarde F, Hultberg B. Higher total plasma homocysteine in vitamin B12 deficiency than in heterozygosiw for homocystinuria due to cystathionine bsynthetase deficiency. Metabolism 1988;37:175-8. Kang SS, Wong PWK, Norusis M. Homocysteinemia due to folate deficiency. Metabolism 1987;36:458-62.

136

Taylor et al.

Journal of VASCULAR SURGERY

27. Harker LA, Ross R, Schlicter SJ, Scott CR. Homocvstineinduced arteriosclerosis. The role of endothelial cell injury. and platelet response in its genesis. J Clin Invest 1976;58: 731-41. 28. Wall RT, Harlan JM, Harker LA, Stroker GE. Homocysteineinduced endothelial cell injury in vitro: a model for the study of vascular injury. Thromb Res 1980; 18:113-21. 29. Degroot PG, Willems C, Boers GHI, Gonsalves MD. Endothelial cell dysfunction in homocystinuria. Eur J Clin Invest 1986;13:405-10.

30. Rodgers GM, Kane WH. Activation of endogenous factor V by a homocysteine-induced vascular endothelial cell activator. J Clin Invest 1986;77:1909-16. 31. Brattstrom LE, Israelsson B, Jeppson J-O, Hultberg BL. Folio acid--an innocuous means to reduce plasma homocysteine. Scand J Clin Lab hwest 1988;48:215-21. 32. Malinow MR. Hyperhomoeyst(c)inemia, a common and easily reversible risk factor for occlusive atherosderosis. Circulation 1990 (In press).

DISCUSSION

plasma homocysteine in such patients, then here may be the first targeted medical therapy for perhaps as many as one third o f patients with symptomatic progressive atherosclerosis. Dr. John M a n n i c k (Boston, Mass.). If elevated homowsteine levels are in fact an indication o f rapidly progressing disease, why is it that the onset of disease was at exactly the same age in those who had it and in those who did not? Dr. Taylor. Dr. Abbott asked why we fbund a difference in coronary and lower extremity, disease and not in cerebral. Probably a selection bias exists in this study. All these patients were recruited from those under long-term follow-up in our vascular clinic. Our clinical orientation toward severe lower extremity ischemia is well known. A difference in cerebral disease was present, but it was not significant. Bruce Coull, a neurologist from our institution, published the results o f homocysteine testing o f patients in our stroke clinic in Stroke this summer. A significant difference was found in the incidence of elevated plasma homocysteine in the stroke patients. Dr. Abbott also noted the difference in the incidence o f progression o f disease in the lower extremities as assessed by clinical criteria, compared to the more objective method o f vascular laboratory assessment. There are two reasons for this. One reason is the period o f vascular laboratory follow-up was about one half the period o f clinical followup. The other reason is that whenever clinical assessment o f lower extremity disease is subjected to objective evaluation, the overall incidence decreases. This was especially illustrated in the results o f the Multiple Risk Factor Intervention Trial, which demonstrated a similar difference. Dr. Mannick, we do not know why the patients with elevated homocysteine levels, who appeared to have more rapid progression o f disease, had the same age at onset of symptoms as those patients who had normal homocysteine levels. Perhaps our upcoming prospective study will answer this question.

Dr. W i l l i a m A b b o t t (Boston, Mass.). The relationship between elevated plasma homocysteine and atherosclerosis is not new. The work today that we have heard is an extension o f that given at last year's Surgical Forum, and is new, and shows that elevated plasma homocysteine is an isolated risk factor for symptomatic coronary and peripheral vascular disease. More important, elevated plasma homocysteine seems to be associated with vascular disease progression, at least in the coronary and lower extremity circulations. This leads to my first question. Why do you suppose progression occurs in only two o f the three areas evaluated? The authors found an elevated plasma homowsteine in 39% o f patients and identified progression in lower extremity patients in 89% by clinical criteria, and progression by hemodynamic criteria in 43%. Does this mean that this represents a particularly virulent form o f atherosclerosis? If so, then we must relook at the controls where 72% progressed clinically and 28% o f the healthy individuals progressed by hemodynamic criteria. This suggests to me that both patient study groups may be skewed toward virulence in progression, and in a more general population the differences may disappear. Several well recognized pathophysiologic consequences o f increased homocysteine exist as demonstrated in the research laboratory: endothelial cell denudations, smooth muscle cell proliferation, platelet aggregation, and stimulation o f procoagulants from intact endothelium. Based on this knowledge, it might be tempting to assume a cause and effect relationship between elevated plasma homocysteine and accelerated atherosclerosis in man. The data presented today do not support such a conclusion, and thus this all may be only an interesting phenomenologic observation. However, this is the most important point o f all. If elevated plasma homocysteine is a cause o f vascular disease progression, and if folic acid can truly control elevated