Serum Lipids in Angiographically Assessed Coronary Atherosclerosis

Serum Lipids in Angiographically Assessed Coronary Atherosclerosis* M. Heikki Frick, M.D., F.C.C.P.; Costa Dahlen, M.D.; Kare Berg, M.D.; Matti Valle, M.D.; and Pauli Hekali, M.D.

Cholesterol, triglycerides, and Lp(a)/pre-beta1 lipoprotein were analyzed in 153 patients typed for liproprotein patterns. Coronary atherosclerosis was determined by selective coronary angiography and graded by a system taking into account proximal, middle and distal segments. Smoking habits, family history and hypertension were also recorded. Normal coronary arteries were encountered in 45, moderate coronary atherosclerosis (< median score) in SO, and severe changes (> median score) in 58 patients. Cholesterol (P
Lp (a)/pre-beta1 lipoprotein (P
There is ample epidemiologic evidence that elevated serum lipids constitute a risk factor for coronary heart disease. Since the lipids are pathogenetically related to coronary atherosclerosis and not necessarily to the clinical manifestations of the disease, both postmortem and angiographic studies have been conducted to demonstrate the above relationship. Both approaches carry a strong element of selection, presumably introducing a stronger disparity in autopsy series due to the large number of sudden deaths caused more by electrical instability than by extensive coronary atherosclerosis. The results of angiographic studies have ranged from entirely negative correlations to significant relationships between coronary atherosclerosis and one or several lipid components of the serum. I.s A number of studies are limited by the absence of relevant grading of the lesions, and similarly only a few studies have taken either Lp (a) lioprotein9 or prebeta-1lipoprotein10 into account. 11• 13 We report a comprehensive lipid analysis in a series of patients with angiographic assessment of coronary atherosclerosis which was subjected to segmental grading.

were women aged 10-62, mean 44.5 years. The youngest patient was a ten-year-old girl with homozygotic familial hypercholesterolemia and severe triple vessel disease successfully bypassed by insertion of venous grafts. The indications for coronary angiography were : (I) preoperative evaluation of coronary arterial anatomy for subsequent bypass surgery, and (2) evaluation of angina-like chest pain combined with abnormal resting and/ or exercise electrocardiogram. None of the patients had significant renal, hepatic or thyroid disease. The presence of the following factors was determined : (I) smoking at the time of the angiography or at the time of the appearance of the first symptoms; ( 2) hypertension indicated by blood pressure readings of ~ I60/IOO mm Hg in repeated measurements, and patients under treatment with antihypertensive drugs; (3) diabetes requiring dietary or medical treatment; and ( 4) a fam!ly history of coronary heart disease in firstdegree relatives. Selective coronary angiography was performed using Judkins' technique in multiple projections as described in detail earlier.H,t5 The degree of intraluminal obstruction was derived by comparing the most diseased portion with the normal portion proximally and expressed in percentage. The following grading was applied: 0 =no intraluminal irregularities; 1 <25 percent obstruction; 2 = 25 <50 percent obstruction; 3 = 50 <75 percent obstruction; 4 = subtotal obstruction; and 5 = total occlusion. The changes were analyzed in the left main stem and in the three major epicardial arteries which were divided into proximal, middle, and distal segments as described earlier.1 5 Thus, total occlusions in the left main stem and in all segments would give a score of 50 for this hypothetic patient. The blood samples were dmwn at 8-10 AM after an overnight fast at Helsinki University Central Hospital and shipped frozen by air to Boden, Sweden, and Oslo, Norway. Cholesterol 1 o and triglyceridet7 determinations were performed with a Technicon autoanalyzer. Electrophoresis was performed in 0.5 percent agarose and stained with Sudan black. 1 8 The Lp testing was conducted as described in detail earlier.9 In the electrophoretic analysis, any serum sample exhibiting a lipoprotein zone between the beta and ordi-

PATIENTS AND METHODS

The study group comprised 153 patients. One hundred fourteen were men aged 23-60, mean 44.6 years, and 39 °From the First Department of Medicine, University Central Hospital, Helsinki, Finland; and the Laboratory of Clinical Chemistry, County Hospital, Boden, Sweden; and the Institute of Medical Genetics, University of Oslo, Oslo, Norway. Manuscript received March 9; revision accepted May 24. Reprint requests: Dr. Frick, Cardiovascular LaboratOf'!J, University Central Hospital, 00290 Helsinki 29, Finland

62 FRICK ET AL

=

CHEST, 73: 1, JANUARY, 1978

Table 2-Data on Upid Let7el•

nary pre-beta lipoprotein bands was scored as containing pre-beta, l:poproteln.tO The close relationship between Lp

( a) lipoprotein and pre-beta 1 lipoprotein has been demonstrated in several studies. 1 9-22 To facilitate · the analysis in the present study, their presence in the serum is reported as Lp(a)/prebeta 1 lipoprotein disregarding the weak Lp(a) lipoprotein positivity in patients without demonstrable prebeta, lipoprotein due to the greater sensitivity of the immunologic Lp(a) lipoprotein determination. Testing with proper antisera has demonstrated the identity of Lp(a) lipoprotein and the "sinking" pre-beta lipoprotein.za The lipoprotein typing was done according to Fredrickson.z' This was, however, done on frozen samples known to influence the results. Moreover, beta-cholesterol was not determined separately after ultracentrifugation. Thus, the results of this typing are merely given as a rough illustraeon of the series in this respect. All the lipid analyses were conducted without knowledge of the angiographic data. The chi square test with Yate's correction was applied to test the significance of the differences in the prevalence of the variables analyzed, and Student's t-test in analyzing the parametric data. REsULTS

Entirely normal coronary arteries ( NCA) were found in 45 patients. Using the median coronary score, the 108 patients with coronary obstructions were divided into those with moderate atherosclerosis (MCA, scores 1-14) and those with severe atherosclerosis ( SCA, scores 15-29). The characteristics of these groups are shown in Table 1. The NCA group had a slight preponderance of women and correspondingly a smaller height and weight average than the other groups, which had a strong prevalence of men. There were only seven diabetic patients in the whole series, and hence, diabetes was not analyzed separately. Table 1--Some Clr.arocteri•lic•

o/ alae Serie•

NCA 45 pts

MCA 50 pts

SCA 58 pts

Males/females

19/26

44/6

51/7

Age, years

42.1 t 23-62

44.5 23-00

46.4 10-62

Height, em

166.4t 151-192

171.9 143-189

171.4 138-187

Weight, kg

67.5t 40-111

73.9 44-96

73.7 29-116

No of smokers

14/45 (3%)

33/50 (66%)

27/58 (47%)

No with hypertension

5/45 (11%)

14/50 (28%)

14/58 (24%)

No with a positive family history

22/43 (51%)

32/49 (65%)

43/55 (78%)

NCA =normal coronary arteries, MCA =moderate coronary atherosclerosis, SCA ""severe coronary atherosclerosis, t "" mean and range

CHEST, 73: 1, JANUARY, 1978

Lp(a)/prebeta,-lipoprotein posCholesterol Lipoprotein typing itivity, TG, mg% mmol/L no. and% N IIA liB III IV V NCA- 231.51 t 1.86 group ±49.51 ±1.06 45 pts

11, 24%

30

3

2 0

MCAgroup 50 pts

8 2

247.44

1.93 ±1.14

21, 42%

31

3

4

0

11 1

SCA278.24 group ±116.12 58 pts

2.11 ±1.03

33, 57%

21 10

10

0

15 2

± 42.25

tmean ± standard deviation

The prevalence of smokers differed significantly between the NCA and MCA groups ( 'l = 10.179, P<0.01) and between NCA and MCA and SCA combined ( x2 = 6.652, P < 0.01), but not between the NCA and SCA groups (x2 = 1.918, ns). The difference in the prevalence of hypertension did not reach statistical significance in any comparisons between the groups. The positivity of a family history was already high in the NCA group and did not differ from the prevalence in the MCA group (t = 1.351, ns), but did reach statistical significance between the NCA and SCA groups (x2 = 6.724, P
=

Table 3--Factor• Ditferendatin« Patient• with Normal Coronary Arterie• (NCA) from Padentl with Coronary Adaer01eleroli• (MCA + SCA) p

Accentuated in NCA/SCA comparison

Prevalence of smoking

0.01

no

Positive family history

0.05

yes

Cholesterol

0.05

yes

Positive Lp(a) I pre-beta1-lipoprotein

0.01

yes

Variable

SERUM UPIDS IN CORONARY ATHEROSCLEROSIS 63

Table 4--Coronary Score• in Relation to Different ~ariable.

Variable

Coronary score

t

p

Smoking+ Smoking-

13.4±5.9t 16.5 ±5.9

2.722

<0.02

Family hist. + Family hist. -

14.9±5.8 14.1 ±6.6

0.580

ns

Elevated cholesterol* with any TG

16.0±5.3

2.032

<0.05

Normal lipids

13.0±6.9

Elevated TG • • with any cholesterol

15.4±5.7

1.755

ns

Normal lipids

13.0±6.9

Lp(a) /pre-beta~lipoprotein +

15.9±5.7

2.113

<0.05

Lp(a) /pre-betatlipoprotein -

13.4 ±6.4

• ii1:280 mg%, •• ii1:2.0 mmol/L tmean, ± standard deviation

but not between the NCA and MCA groups of patients ( x2 = 2.878, ns) . A compilation of these data (Table 3) shows that the differentiating factors usually reach higher statistical significance when the NCA group is compared with patients with severe atherosclerosis ( SCA), the only exception being the prevalence of smoking. These relationships were studied more closely by analyzing the coronary scores in patients with lesions in relation to the variables (Table 4). Elevated cholesterol ( >280 mg percent) with any value of triglycerides and positivity of the Lp( a) lpre-betat lipoprotein were related to significantly higher coronary scores. The analysis of the prevalence of smoking revealed that the smokers had a lower mean coronary score than the nonsmokers. An analysis of the interrelationships of the significant variables showed that smoking was strongly associated with the presence of Lp (a) I pre-beta1 lipoprotein ( x2 = 7.116, P
The data on the association between either cholesterol or triglycerides and angiographically assessed coronary atherosclerosis are at variance. In 147 male pati.ents, a trend was found for the influence of cholesterol.2 The same finding has been corroborated in more recent studies. 5·8 •2u 6 Triglycerides but not cholesterol have been found to be significant in two studies. 1·4 In a series of 192 patients, type II and IV hyperlipoproteinemia were equally frequent in pa-

64 FRICK ET AL

tients with coronary atherosclerosis. 3 In a recent interesting analysis, 7 type II hyperlipoproteinemia was associated with distal coronary atherosclerosis and frequent involvement of the left main stem, whereas patients with type IV hyperlipoproteinemia more often exhibited proximal disease. Most studies have used a dichotomic approach to coronary atherosclerosis, dividing the patients into categories of normal and diseased arteries.2·4.2s-27 In some series, the changes have been graded by either the most severe stenosis or the distal runoff.1·5 A critical lone stenosis in the left main artery with other arteries intact makes the disease "severe" regarding symptoms and prognosis. Yet the degree of atherosclerosis is inferior to a case with multiple small obstructions in several vessels. This was taken into account in the present approach by analyzing three segments of each major coronary artery. The use of the total score discriminated cholesterol and Lp(a)lpre-beta1 lipoprotein as being associated with the presence and degree of coronary atherosclerosis, whereas no relationship was found between triglycerides and the coronary artery lesions. Our data on the customary lipoprotein phenotyping (Table 2) revealed that the prevalence of type II rose from 11 percent in the group with normal arteries to 14 percent and 34 percent, respectively, in the groups of moderate and severe changes. The prevalence of type IV was 18 percent, 22 percent and 26 percent, respectively. The data suggest that both lipoprotein types played a role. Some earlier studies have addressed themselves to the association between pre-betat lipoprotein and coronary atherosclerosis. Vloedman et al 28 found in 42 patients that the presence of this lipoprotein was an excellent indicator of moderate to severe atherosclerosis. An extended series of 178 patients by the same group strengthened this thesis, 29 which is also supported by the findings of Papadopoulos and Bedynek. 11 The present data expand our earlier experience 12·13 and suggest that Lp (a) I pre-betat lipoprotein can be added to the list of atherogenic factors. An analysis of high-density lipoproteins in a subgroup of 95 patients of the present series showed that low levels of this lipoprotein were also associated significantly with coronary artery lesions.30 The analysis on the interrelationship of the different significant variables strengthened our earlier observation that smoking was related to the presence of pre-beta1 lipoprotein. 12 No causality can be attached to this relationship, but it is interesting that the smokers did not have higher serum cholesterol. Smoking was related significantly to both the presence and degree of coronary obstructions in a series of 224 patients, 1 but was not found to infiuCHEST, 73: 1, JANUARY, 1978

ence the degree of obstructions in a series of 300 patients, all of whom had coronary obstruction.6 In the latter study, smokers were somewhat younger than nonsmokers. In the present study smokers had a lower coronary score than nonsmokers (Table 4). This apparent paradox is without a valid explanation, but patients with severe obstructions and who smoke may be underrepresented at angiography because of a higher early mortality. Alternatively, smokers may have earlier and more severe angina

and come to angiography with less severe obstruc-

tions. This study, and the evidence cited, substantiate the fact that serum cholesterol is associated with coronary atherosclerosis. Of comparable significance are the presence of Lp( a) /pre-beta 1 lipoprotein in the serum and low levels of high-density lipoproteins.

1 Cramer K, Paulin S, Werko L : Coronary angiographic findings in correlation with age, body weight, blood pressure, serum lipids, and smoking habits. Circulation 33:888-900, 1966 2 Proudfit WL, Shirey EK, Sones FM Jr: Selective cine coronary arteriography. Correlation with clinical findings in 1000 patients. Circulation 33:901-910, 1966 3 Heinle RA, Levy RI, Frederickson DS, et al: Lipid and carbohydrate abnormalities in patients with angiographically documented coronary artery disease. Am J ~ol24:178-186, 1969 4 Crowley LV: Serum lipid concentrations in patients with coronary arteriosclerosis demonstrated by coronary arteriography. Clin Chem 17:206-209, 1971 5 Cherrier F, Cherrier P, Ray B, et al: Anomalies lipidiques chez 209 coronariens ayant subi une arteriographie coronaire selective. Coeur Med Intern 13:223228, 1974 6 Fuster V, Frye RL, Connolly DC, et al: Arteriographic patterns early in the onset of coronary syndromes. Br Heart J 37:1250-1255, 1975 7 Bloch A, Dinsmore RE, Lees RS : Coronary arteriographic findings in type-11 and type-IV hwerlipoproteinaemia. Lancet 1:928-930, 1976 8 Barboriak JJ, Rirnm AA, Anderson AJ, et al: Coronary artery occlusion and blood lipids. Am Heart J 87:716721, 1974 9 Berg K: A new serum type system in man-the Lp system. Acta Path Microbiol Scand 59:369-382, 1963 10 Dahlen G, Ericson C, Furberg C, et al: Studies on an extra pre-beta lipoprotein fraction. Acta Med Scand Suppl531, 1972 11 Papadopoulos NM, Bedynek JL : Serum lipoprotein pat-

CHEST, 73: 1, JANUARY, 1978

tern in patients with coronary atherosclerosis. Clin Chim Acta 44 :153-157, 1973 12 Frick MH, Dahlen G, Furberg C, et al: Serum pre-p-1 lipoprotein fraction in coronary atherosclerosis. Acta Med Scand 195:337-340, 1974 13 Dahlen G, Berg K, Frick MH: Lp(a)lipoprotein/pre-p1-lipoprotem, serum lipids and atherosclerotic disease. Clin Genet 9:558-566, 1976 14 Valle M: Postoperative coronary angiography. Acta Radio! Suppl 333, 1973 15 Frick MH, Valle M, Harjola P-T, et al: Changes in native coronary arteries after coronary bypass surgery. Role of graft patency, serum lipids and hypertension. Am J Cardiol 36:744-750, 1975 16 Levine J, Zak B: Automated determination of serum total cholesterol. Clin Chim Acta 10:381-384, 1964 17 Cramp DG, Robertson G: The fluorimetric assay of triglyceride by a semiautomated method. Anal Biochem 25:246-251, 1968 18 Dahlen G, Ericson C : Studies on pre-beta-1 lipoprotein. Opusc Med 18:30-34, 1973 19 Dahlen G, Frick MH, Berg K, et al: Further studies of Lp( a) lipoprotein/pre-{3 1 -lipoprotein in patients with coronary heart disease. Clin Genet 8: 183-189, 1975 20 Berg K, Dahlen G, Frick MH : Lp(a) lipoprotein and pre-p1 -lipoprotein in patients with coronary heart disease. Clin Genet 6:230-235, 1974 ,21 Dahlen G, Berg K, Gillnlis T, et al: Lp(a) lipoprotein/ pre-{3 1 -lipoprotein in Swedish middle-aged males and in patients with coronary heart disease. Clin Genet 7:334341, 1975 22 Dahlen G, Berg K, Ericson C, et al : Lp(a) lipoprotein and pre-p 1-lipoprotein in relation to lipid levels in males. Acta Med Scand 198:263-267, 1975 23 Rider AK, Levy DI, Fredrickson DS: "Sinking" prebeta lipoprotein and the Lp antigen. Circulation 42: Suppl 3, 10, 1970 (abstract) 24 Fredrickson DS, Levy RI, Lees RS: Fat transport in lipoproteins. An integrated approach to mechanism and disorders. N Engl J Med 276:32-44, 94-103, 148-156, 1971 25 Valek J, Grafnetter D, Fabian J, et al: Analysis of lipid disturbances in patients with angiographically confirmed coronary artery disease. Nutr Metab 16:193-202, 1974 26 Cohn PF, Gabbay Sl, Weglicki WB : Serum lipid levels in angiographically defined coronary artery disease. Ann Intern Med 84:241-245, 1976 27 Eisenberg S, Tzivoni D, Stem S: Hyperlipoproteinemia in angiographically documented coronary artery disease. Eur J Cardiol4:469-473, 1976 28 Vloedman DA Jr, Najmi M, Insull W Jr: Relation of the pre-beta lipoprotein subfraction to the severity of coronary artery disease. Clin Chern 18:692, 1972 (abstract) 29 Insull W Jr, Najmi M, Vloedman DA Jr : Plasma prebeta lipoprotein sub-fractions in diagnosis of coronary artery disease. Circulation 45 : Suppl 2, 170, 1972 ( abstract) 30 Berg K, Borresen A-L, Frick MH, et al: Serum-H.D.L. in atherosclerotic heart-disease. Lancet 1:1014, 1976

SERUM LIPIDS IN CORONARY ATHEROSCLEROSIS 65