Proc. Natl. Acad. Sci. USA Vol. 81, pp. 2538-2542, April 1984 Medical Sciences

Inhibition of 3-hydroxy-3-methylglutaryl-CoA reductase by mevinolin in familial hypercholesterolemia heterozygotes: Effects on cholesterol balance (cholesterol balance/low-density lipoprotein receptors/fecal neutral and acidic steroids)

SCOTT M. GRUNDY AND DAVID W. BILHEIMER Department of Internal Medicine and Center for Human Nutrition, University of Texas Health Science Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75235 Communicated by Michael S. Brown, January 9, 1984

ABSTRACT Patients with heterozygous familial hyper- through endocytosis mediated by a specific LDL receptor cholesterolemia (FH) have a deficiency of receptors for plasma (8). low-density lipoprotein (LDL) that impairs removal of LDL When human fibroblasts or cultured liver cells are incu- from plasma. In these patients, mevinolin, an inhibitor of 3- bated with compactin or mevinolin, HMG-CoA reductase is hydroxy-3-methylglutaryl-CoA reductase [mevalonate:NAD' competitively inhibited, and the de novo synthesis of choles- oxidoreductase (CoA-acylating), EC 1.1.1.88], increases re- terol is blocked (9, 10). This block triggers a dual regulatory ceptors for LDL and decreases LDL concentrations. To deter- response that seems designed to provide more cholesterol mine whether mevinolin also causes severe decreases in total for the cell: there is a simultaneous increase in HMG-CoA body synthesis of cholesterol, fecal excretions of neutral ste- reductase and in the number of LDL receptors (9, 11, 12). roids and acidic steroids were determined in five FH heterozy- In dogs, mevinolin has been shown to decrease plasma gotes before and during treatment with mevinolin. The drug LDL by causing an increase in hepatic receptors for LDL produced an average decrease in plasma total cholesterol of (2); the same response probably occurs in humans (7). Me- 23% and in LDL cholesterol of 24%. Mevinolin caused a sig- vinolin thus seems to have the same effect in vivo as in cul- nificant decrease in the output of neutral and acidic steroids in tured cells; i.e., it inhibits HMG-CoA reductase and triggers three patients, but it caused no alterations in two others. a regulatory response that increases LDL receptors and Changes in fecal output of steroids did not correlate with the probably also increases the amount of HMG-CoA reductase. degree of lowering of the patients' LDL-cholesterol level. In The question that arises is whether HMG-CoA reductase can none of the patients did the output of fecal steroids fall below increase sufficiently to return cholesterol synthesis to nor- the values seen in normal subjects studied under similar condi- mal or whether cholesterol synthesis must remain depressed tions. One patient had a previous ileal exclusion operation and for mevinolin to maintain an increase in LDL receptors. In had a massive output of acidic steroids in the control period; the current study, we have attempted to distinguish between mevinolin therapy caused a slight decrease in excretion of acid- these two possibilities in humans by use of the cholesterol- ic steroids, but the output was still markedly above normal. balance technique. We conclude that the LDL lowering action of mevinolin does Normal humans excrete more steroids in the feces than not appear to require a severe decrease in cholesterol synthesis can be accounted for by the cholesterol that they ingest. that might lead to depletion of vital body stores of cholesterol. These steroids emerge as neutral steroids and bile acids. Since the steroid nucleus is not degraded in the body, and An exciting class of drugs for treatment of hypercholester- since the feces are the predominant route of steroid excre- olemia consists of fungal metabolites that are competitive tion, the net fecal steroid excretion (cholesterol intake minus inhibitors of 3-hydroxy-3-methylglutaryl-CoA reductase steroid output) in the steady state approximates the total [HMG-CoA reductase; mevalonate:NAD' oxidoreductase amount of cholesterol synthesized in the body (13). (CoA-acylating), EC 1.1.1.88], a rate-controlling enzyme in In the current investigation we treated five heterozygotes cholesterol synthesis. The prototype compound, compactin, for familial hypercholesterolemia with mevinolin and mea- and its analogue, mevinolin, decrease the levels of low-den- sured the net excretion of fecal steroids. This cholesterol sity lipoprotein (LDL) in normal animals (1, 2), normal hu- balance study was conducted as part of another study in mans (3), and in heterozygotes with familial hypercholester- which turnover of LDL was measured (7). The results indi- olemia (FH) (4-7). Recently, we demonstrated that mevino- cate that mevinolin can cause a moderate decrease in total- lin achieves this effect by increasing the fractional rate of body synthesis of cholesterol, but a marked overall decrease removal of LDL from the circulation of FH heterozygotes in synthesis seemingly is not necessary for mevinolin to en- (7). The higher clearance rate appeared to be due to en- hance the activity of LDL receptors. hanced receptor-mediated catabolism of LDL. A theoretical basis for the ability of mevinolin to stimulate METHODS receptor-mediated catabolism of LDL comes from studies of Patients. Five patients with heterozygous FH were studied human and animal cells in tissue culture. These cells have a on the General Clinical Research Center at Parkland Memo- dual source of the cholesterol required for synthesis of new rial Hospital or in the Metabolic Unit at the Veterans Admin- membranes. Cells can synthesize cholesterol de novo in a istration Medical Center (Dallas, TX). Two patients (M.B. pathway that requires the action of HMG-CoA reductase; and J.P.) were men of ages 36 and 37 yr, respectively; J.P. or, the cells can obtain cholesterol from plasma LDL Abbreviations: FH, familial hypercholesterolemia; HMG-CoA re- The publication costs of this article were defrayed in part by page charge ductase, 3-hydroxy-3-methylglutaryl-CoA reductase; LDL, low- payment. This article must therefore be hereby marked "advertisement" density lipoprotein; VLDL, very-low-density lipoproteins; HDL, in accordance with 18 U.S.C. §1734 solely to indicate this fact. high-density lipoproteins. 2538 Downloaded by guest on September 23, 2021 Medical Sciences: Grundy and Bilheimer Proc. Natl. Acad. Sci. USA 81 (1984) 2539

Table 1. Plasma lipids and lipoproteins Plasma cholesterol,* mg/dl SD Plasma Patient Period Total LDL HDL triglyceride* M.B. Control 340 ± 32 272 ± 24 24 ± 1 208 ± 53 Mevinolin 285 ± 20 239 ± 16 22 ± 6 77 ± 15 J.P. Control 267 ± 17 207 ± 18 24 ± 3 221 ± 21 Mevinolin 226 ± 20 156 ± 4 26 ± 3 205 ± 39 J.C. Control 320 ± 6 281 ± 9 19 ± 2 103 ± 12 Mevinolin 233 ± 9 201 ± 11 20 ± 6 105 ± 10 M.Ba. Control 306 ± 11 254 ± 6 26 ± 3 118 ± 14 Mevinolin 247 ± 12 190 ± 11 29 ± 2 144 ± 8 C.C. Control 319 ± 21 228 ± 18 18 ± 4 311 ± 42 Mevinolin 211 ± 9 153 ± 13 24 ± 5 178 ± 47 Mean + SEM Control 310 ± 12 248 ± 13 22 ± 2 192 ± 38 Mevinolin 240 ± 13 188 ± 16 24 ± 2 142 ± 23 P valuet < 0.005 < 0.002 NS NS NS, not significant. *Mean ± SD of six or seven observations during the LDL turnover study in control period and with mevinolin treatment. tP values were calculated by paired t test.

had severe coronary heart disease, but M.B. had no clinical Medical Center. Informed consent was obtained from each evidence of atherosclerosis. The remaining three patients patient. (C.C., J.C., and M.Ba.) were women of ages 52, 32, and 55 Lipid and Lipoprotein Analysis. In each patient, the kinet- yr, respectively. C.C. had coronary heart disease, but J.C. ics of LDL turnover were measured in each study period, and M.Ba. were clinically free of ischemic heart disease. and the results have been reported separately (7). These J.C. is the daughter of M.Ba.; M.B. and C.C. are first cous- turnover studies were initiated about 3 weeks into each peri- ins. od. During the kinetic study, which lasted 20 days, blood Diets. The patients were fed a diet of solid food and liquid samples were obtained daily, and total cholesterol was mea- formula containing 40% of calories as fat, 40% as carbohy- sured on each sample. In this period, samples were taken 6 drate, and 20% as protein (7). The major fat of the diet was or 7 times for estimation of triglycerides, LDL cholesterol, lard, and the polyunsaturated/saturated fat ratio was 0.33. and high-density lipoprotein (HDL) cholesterol. Each sam- Cholesterol intake varied between 170 and 280 mg per day. ple was subjected to ultracentrifugation at density <1.006 Weights were checked daily and caloric intake was adjusted g/ml for isolation of very-low-density lipoprotein (VLDL). to maintain a constant weight throughout the study. Cholesterol was measured on the infranatant both before and Drug Therapy. Capsules containing mevinolin and placebo after precipitation of LDL with heparin-manganese accord- were supplied by Merck Sharp & Dohme (courtesy of J. To- ing to Lipid Research Clinic procedures (14). Cholesterol bert). Mevinolin was given in doses of 20 mg twice daily. and triglyceride concentrations were determined using a Two patients (J.C. and M.B.) received mevinolin first and Boehringer Mannheim enzyme kit. then placebo; the other three (J.P., C.C., and M.Ba.) were Sterol Balance. Cholesterol balance was performed ac- given placebo before mevinolin. The protocol was approved cording to described methods (15, 18). Measurements were by the U.S. Food and Drug Administration and by the Hu- not made in one patient (T.P.), who had been studied previ- man Research Review Boards of the University of Texas ously (7). Serial stool collections were obtained and com- Health Science Center and the Veterans Administrations bined into 4-day pools during each study period (mevinolin

Table 2. Cholesterol balance results Days: Neutral Acidic Total number Chol intake, steroids, steroids, steroids, Chol balance, Patient Period determ* mg/day mg/day + SD mg/day ± SD mg/day ± SD mg/day ± SD M.B. Control 21:5 178 623 ± 85 575 ± 170 1198 ± 158 1020 ± 159 Mevinolin 20:5 178 513 ± 28 243 ± 76 750 ± 96 572 ± 96 P valuet <0.025 <0.005 <0.001 <0.001 J.P. Control 36:9 249 1207 ± 245 442 ± 70 1630 ± 235 1381 ± 235 Mevinolin 28:7 249 920 ± 193 413 ± 50 1333 ± 220 1084 ± 249 P valuet <0.02 NS <0.025 <0.025 J.C. Control 32:8 280 655 ± 82 210 ± 56 867 ± 118 587 ± 118 Mevinolin 36:9 280 612 ± 114 193 ± 56 805 ± 147 525 ± 147 P valuet NS NS NS NS M.Ba. Control 40:10 171 1203 ± 123 214 ± 46 1417 ± 118 1246 ± 118 Mevinolin 31:8 171 1198 ± 93 237 ± 46 1480 ± 122 1309 ± 122 P valuet NS NS NS NS C.C. Control 36:9 170 580 ± 30 2701 ± 376 3281 ± 363 3111 ± 363 Mevinolin 26:7 170 452 ± 24 2133 ± 504 2637 ± 410 2467 ± 380 P valuet <0.001 <0.025 <0.01 <0.01 NS, not significant; Chol, cholesterol. *Days in period/number of measurements. The ratio gives the number of days in each pool, which usually was four. tP values were calculated by Students' t test. Downloaded by guest on September 23, 2021 2540 Medical Sciences: Grundy and Bilheimer Proc. Natl. Acad. Sci. USA 81 (1984)

J.P. 0 M.B. .5 4001- qU) So 0D 0 X- o X0 30C L) CP E a E E E 0 In a 20C 0L le

xa.In~ 160C 0 140C tLi o >, . a os 120C _ C.I-..ln '. _ a lOOC 76E 6E 80C 60CD"n 40C 20C 12 16 20 24 28 32 12 16 20 24 28 32 4 8 12 16 20 24 28 32 36 40 4 8 12 16 20 24 28 Days on Mevinolin Days on Placebo Days on Placebo Days on Mevinolin

.5 400r 0 M.Ba 0- 300 400 J.C. o E 0~~ 350 0 a Wot la -S.--o X 300 0 cp E 250 0 l 200 1600 _ 10 I UV ~0> 0 1200 0c, 1200 mvX~m a.aS 1000 CW 800 800 - U.- 0 a 600 400 4001_ 200 4 8 12162024283438424 8 121620242832 4 8 12 1416 202428323640 4 8 12 162024283032 Days on Mevinolin Days on Placebo Days on Placebo Days on Mevinolin

0... 41-5 - 0 X

a_ ca.a 1 Acidic Steroids

0 0 E Neutral Steroids . a en -- oa E

4 8 12162024283236404 8 12 16202428 Days on Placebo Days on Mevinolin

FIG. 1. Cholesterol balance studies in five patients treated with mevinolin. Patients J.P., M.Ba., and C.C. received placebo first, and M.B. and J.C. received mevinolin first. Plasma cholesterol levels are those obtained during the LDL turnover (7). These turnover studies generally for were started after the patient had been on mevinolin for about 2 weeks. Generally, stool collections were started after allowing 4 days equilibration on internal markers. However, in M.B., balance studies were not begun until the time of the LDL turnover. Acidic steroids are equivalent- to bile acids; neutral steroids include cholesterol and its bacterial conversion products, coprostanol and coprostanone. Downloaded by guest on September 23, 2021 Medical Sciences: Grundy and Bilheimer Proc. Natl. Acad. Sci. USA 81 (1984) 2541

Table 3. Comparison of cholesterol balance between current patients and control subjects Cholesterol Fecal neutral Fecal acidic Fecal total Cholesterol Patient Period intake, mg/day steroids, mg/kg/day steroids, mg/kg/day steroids, mg/kg/day balance, mg/kg/day M.B. Control 2.5 8.9 8.2 17.1 14.6 Mevinolin 2.5 7.3 3.4 10.7 8.17 J.P. Control 2.3 11.4 4.2 15.4 13.0 Mevinolin 2.3 8.6 3.9 12.6 10.2 J.C. Control 4.0 9.4 3.0 12.4 8.4 Mevinolin 4.0 8.7 2.8 11.5 7.5 M.Ba. Control 1.3 9.2 1.6 10.8 9.5 Mevinolin 1.3 9.1 1.8 11.2 10.0 C.C. Control 2.6 9.2 42.9 52.1 49.4 Mevinolin 2.6 7.1 33.9 41.9 39.2 Normal adult mean* (n = 14) 1.8 ± 0.5 6.5 ± 0.5 4.9 ± 0.5 11.4 ± 0.7 9.6 ± 0.6 *These data are expressed as mean ± SEM and have been presented previously (19).

and placebo). On each pool, measurements were made of tal conditions (19). In none of the FH patients did mevinolin neutral steroids (15) and acidic steroids (16). These measure- cause cholesterol balance to fall below the range seen in nor- ments were carried out entirely by chemical means using mal subjects. gas-liquid chromatography for quantification of steroids. Table 4 compares the percent change in LDL-cholesterol Excretions of neutral and acidic steroids were calculated us- with the percent change in cholesterol balance induced by ing f3sitosterol (17) and chromic oxide (18) as internal stan- mevinolin in each patient. Of interest, patient M.B., who dards. For controls, 14 normal men were studied on a similar had the greatest decrease in cholesterol balance (44%), had diet, and stool samples were analyzed by the same methods. the least decrease in LDL cholesterol (12%); overall there Ages of these men ranged from 29 to 63 yr (mean, 52 yr); was no correlation between the change in cholesterol bal- their weights were 95-105% of ideal. Sterol balance data ance and the fall in LDL cholesterol during mevinolin treat- from the normal men have been presented (19). ment.

RESULTS DISCUSSION In a recent study we demonstrated that mevinolin lowers Mevinolin therapy produced a decrease in plasma total cho- plasma LDL by enhancing receptor-mediated clearance of lesterol in all patients (Table 1); the mean decrease for the lipoprotein (7). This action presumably is due to an increased five patients was 22.6%. Plasma triglyceride levels were not synthesis of LDL receptors secondary to inhibition of HMG- changed for the group, although two individuals (M.B. and CoA reductase activity. The latter action should be associat- C.C.) had sizable decreases. Plasma LDL cholesterol fell ed with a decrease in cholesterol synthesis; the degree of during mevinolin therapy by 24.2%, while levels of HDL inhibition of cholesterol synthesis required to produce the cholesterol were unchanged. The turnover studies demon- observed effect on plasma LDL is unknown. If a severe de- strated that this drop in LDL cholesterol is due to enhanced crease in cholesterol synthesis is required, a decreased avail- receptor-mediated removal of LDL from plasma (7). ability of cholesterol for vital functions-cellular membrane The effects of mevinolin on cholesterol balance are shown structure and synthesis of steroid hormones and bile acids in Table 2 and Fig. 1. One patient (C.C.) had undergone a might result. The purpose of the present study was to deter- previous ileal bypass operation and therefore showed a large mine the extent to which mevinolin inhibits whole-body syn- output of bile acids during the control period. Mevinolin pro- thesis of cholesterol and particularly to learn whether syn- duced significant decreases in outputs of neutral steroids in thesis is seriously decreased. three patients (M.B., J.P., and C.C.), but no significant The method used to estimate cholesterol synthesis was the changes were noted in two others (M.Ba. and J.C.). Changes balance method. This method has the advantage of giving a in output of acidic steroids paralleled those of neutral ste- daily measurement of steroid excretion over a period of sev- roids in all patients. Two patients (M.B. and C.C.), who eral weeks, from which one can estimate cholesterol synthe- were first cousins, had the greatest decreases in total steroid sis. It has the disadvantage of being unable to distinguish excretion during mevinolin treatment. Patient C.C., who had between fecal steroids derived from newly synthesized cho- the ileal exclusion, had a slight decrease in output of acidic lesterol and from preexisting tissue cholesterol. Any reduc- steroids when receiving mevinolin, but this output was still tion in total steroid excretion during mevinolin treatment al- markedly above normal despite mevinolin therapy. most certainly would reflect a decrease in total cholesterol Patient J.C. had no overall reduction in steroid excretion when the mevinolin treatment period as a whole was consid- Table 4. Comparison of changes in plasma LDL cholesterol and ered, but there was a downward trend in steroid output dur- cholesterol balance (mevinolin period vs. control period) the + ing last 12 days of the period (642 31 mg/day); the % change in plasma % change in latter was significantly lower than that of the whole control LDL cholesterol* cholesterol balance* period (867 ± 118 mg/day, P < 0.005). However, a down- Patient (P value)t (P value)t ward similar trend was not noted in J.C.'s mother (M.Ba.), and one of the other patients who had reduced steroid out- M.B. -12 (<0.02) -44 (<0.001) puts during mevinolin treatment showed downward trends in J.P. -25 (<0.001) -22 (<0.001) steroid excretion rates during the treatment period (i.e., the J.C. -28 (<0.001) -11 (NS) mevinolin effect appeared to be maximal within the first M.Ba. -25 (<0.001) +5 (NS) week). C.C. -32 (<0.001) -24 (<0.01) In Table 3, the balance data are expressed per kg of total NS, not significant. body weight. The results are compared to those for 14 nor- *Mevinolin period minus control period. mal men studied in this laboratory under similar experimen- tp values calculated by Student's t test. Downloaded by guest on September 23, 2021 2542 Medical Sciences: Grundy and Bilheimer Proc. Natl. Acad Sci. USA 81 (1984)

synthesis. A severe decrease in steroid output would signify ance and support of Drs. Michael S. Brown and Joseph L. Goldstein a corresponding reduction in synthesis of cholesterol and in the design and implementation of this project. We thank Dr. Jona- would be a cause for concern. On the other hand, a severe than Tobert of Merck Sharp & Dohme for making mevinolin avail- reduction in cholesterol synthesis might be masked by mobi- able. This research was supported by a grant from the National In- lization of cholesterol from tissue pools, so that cholesterol stitutes of Health (HL-29252), by the Veterans Administration, and balance would continue to be normal. This response could by the Moss Heart Foundation. not be continued indefinitely, or tissue pools would become totally depleted; eventually fecal steroid excretion would 1. Endo, A., Tsujita, Y., Kuroda, M. & Tanzawa, K. (1977) Eur. show a decline. J. Biochem. 77, 31-36. 2. Kovanen, P. T., Bilheimer, D. W., Goldstein, J. L., Jaramillo, The data of this study indicated that mevinolin does cause J. J. & Brown, M. S. (1981) Proc. Natl. Acad. Sci. USA 78, a detectable decrease in cholesterol synthesis in some pa- 1194-1198. tients. The percentage reductions in cholesterol excretion in 3. Tobert, J. A., Bell, G. D., Birtwell, J., James, I., Kukovetz, three patients (22%, 24%, and 44%) represent a minimum W. R., Pryor, J. S., Buntinx, A., Holmes, I. B., Chao, Y.-S. decrease in whole-body synthesis of cholesterol. The find- & Bolognese, J. A. (1982) J. Clin. Invest. 69, 913-919. ings in these patients thus support the concept that the LDL 4. Yamamota, A., Sudo, H. & Endo, A. (1980) Atherosclerosis lowering action of mevinolin is mediated by the action of the 35, 259-266. drug on cholesterol synthesis. 5. Mabuchi, H., Haba, T., Tatami, R., Miyamoto, S., Sakai, Y., Although these reductions in cholesterol balance were not Wakasugi, T., Watanabe, A., Koizumi, A. J. & Takeda, R. (1981) N. Engl. J. Med. 305, 478-482. marked, we can ask whether cholesterol synthesis might 6. Mabuchi, H., Sakai, T., Yoshimura, A., Watanabe, A., Waka- have been reduced severely, and yet this change was ob- sugi, T., Koizumi, J. & Takeda, R. (1983) N. Engl. J. Med. scured by mobilization of cholesterol from preexisting pools. 308, 609-613. The results in patient C.C. make this very unlikely. This pa- 7. Bilheimer, D. W., Grundy, S. M., Brown, M. S. & Goldstein, tient had a previous ileal exclusion operation (20) and thus J. L. (1983) Proc. Natl. Acad. Sci. USA 80, 4124-4128. had a massive increase in synthesis and excretion of bile ac- 8. Goldstein, J. L. & Brown, M. S. (1977) Annu. Rev. Biochem. ids (21). During the period of mevinolin therapy, massive ex- 46, 897-930. cretion of bile acid continued with no sign of progressive 9. Brown, M. S., Faust, J. R., Goldstein, J. L., Kaneko, I. & diminution in output. The net loss of steroid during 40 days Endo, A. (1978) J. Biol. Chem. 253, 1121-1128. 10. Alberts, A. W., Chen, J., Kuron, G., Hunt, V., Huff, J., Hoff- of mevinolin therapy was 99 g. Since total-body stores of man, C., Rothrock, J., Lopez, M., Joshua, H., Harris, E., Pat- cholesterol are only =100 g (22) and do not greatly exceed chett, A., Monaghan, R., Cprrie, S., Stapley, E., Albers- this value even in heterozygotes for FH (23), this patient Schonberg, G., Henseus, O., Hirshfield, J., Koogsteem, K., would have become severely depleted of cholesterol during Leisch, J. & Springer, J. (1980) Proc. Natl. Acad. Sci. USA 77, the mevinolin treatment period if cholesterol synthesis had 3957-3961. been profoundly inhibited. 11. Goldstein, J. L., Helgeson, J. A. & Brown, M. S. (1979) J. The data from the other patients also are strongly sugges- Biol. Chem. 254, 5403-5409. tive that their synthesis of cholesterol was not markedly in- 12. Pangburn, S. H., Newton, R. S., Chong, C. M., Weinstein, Two and M.Ba.) had no overall D. B. & Steinberg, D. (1981) J. Biol. Chem. 256, 3340-3347. hibited. patients (J.C. 13. Grundy, S. M. & Ahrens, E. H., Jr. (1969) J. Lipid Res. 10, change in cholesterol balance on mevinolin and yet showed 91-107. decreases in LDL-cholesterol of 25% and 28%. It seems 14. U.S. Department of Health Education and Welfare (1974) highly unlikely that a severe decrease in cholesterol produc- Manual ofLaboratory Operations, Lipid Research Clinic Pro- tion could have been replaced in fecal steroid excretion en- gram: Lipid and Lipoprotein Analysis (NIH/75-628) (Govern- tirely by cholesterol mobilized from tissue pools. In J.P. and ment Printing Office, Washington, DC). M.Ba., a total of 30 and 41 g of cholesterol, respectively, 15. Grundy, S. M., Ahrens, E. H., Jr., & Miettinen, T. A. (1965) were lost in feces with no sign of progressive diminution in J. Lipid Res. 6, 397-410. in none of the did cholesterol 16. Miettinen, T. A., Ahrens, E. H., Jr., & Grundy, S. M. (1965) output. Furthermore, patients J. Lipid Res. 6, 411-424. balance fall below the values seen in normal adult men. 17. Grundy, S. M., Ahrens, E. H., Jr., & Salen, G. (1968) J. Lipid Therefore, while we cannot rule out the possibility that a Res. 9, 374-387. portion of the steroids excreted during mevinolin therapy 18. Davignon, J., Simmonds, W. J. & Ahrens, E. H., Jr. (1968) J. may have been derived from preexisting pools, it is doubtful Clin. Invest. 47, 127-138. that cholesterol synthesis was curtailed to a dangerously low 19. von Bergmann, K., Mok, H. Y. I., Hardison, W. G. M. & level. Grundy, S. M. (1979) Gastroenterology 77, 1183-1192. While studies of much longer duration will be required to 20. Buchwald, H. & Varco, R. (1966) J. Am. Med. Assoc. 196, confirm the current conclusion, we feel that the data strong- 627-630. ly suggest that treatment with mevinolin can lower LDL lev- 21. Grundy, S. M., Ahrens, E. H., Jr., & Salen, G. (1971) J. Lab. stores of cho- Clin. Med. 78, 94-121. els without leading to severe depletion of vital 22. Goodman, DeW. S., Smith, F. R., Seplowitz, A. H., Ramak- lesterol in familial hypercholesterolemia heterozygotes. rishnan, R. & Dell, R. B. (1980) J. Lipid Res. 21, 699-713. Michael Ahern, Michael Funk, and Marorie Whelan provided ex- 23. Samuel, P., Perl., W., Holtzman, C. M., Rochman, N. D. & cellent technical assistance. We particularly appreciate the assist- Lieberman, S. (1972) J. Clin. Invest. 51, 226-278. Downloaded by guest on September 23, 2021