LIVER PHYSIOLOGY AND DISEASE

Transcription

1 GASTROENTEROLOGY 68: , 1975 Copyright 1975 by The Williams & Wilkins Co. Vol. 68, No.2 Printed in U.S.A. LIVER PHYSIOLOGY AND DISEASE EFFECTS OF LOW DOSE CHENODEOXYCHOLIC ACID FEEDING ON BILIARY LIPID METABOLISM RONALD D. ADLER, M.D., LYNN J. BENNION, M.D., WILLIAM C. DUANE, M.D., AND SCOTI M. GRUNDY, M.D., PH.D. Phoenix Clinical Research Section, National Institute of Arthritis, Metabolism, and Digestive Diseases, Phoenix Indian Medical Center, Phoenix, Arizona To better define the mechanisms by which chenodeoxycholic acid (CDCA) alters the lipid composition of bile and promotes dissolution of cholesterol gallstones, we have studied the effects of relatively low doses of CDCA on the metabolism of biliary lipids in 10 subjects without gallstones. In 5 subjects, the effects of CDCA feeding were compared to those of cholic acid, a bile acid that apparently does not dissolve cholesterol stones. The following measurements were carried out during control and treatment periods: (1) composition of biliary lipids, (2) hepatic secretion rates of biliary lipids, (3) pool sizes of bile acids, and (4) specific composition of bile acids. Low doses of CDCA consistently reduced the lithogenicity of gallbladder bile by decreasing the proportion of cholesterol relative to the solubilizing lipids-bile acids and lecithin. This decrease in lithogenicity was associated with a selective reduction in hepatic secretion rates of cholesterol. At the doses of CDCA given, secretion rates of bile acids and lecithin and pool sizes of bile acids were not significantly changed; also conversion of cholesterol into bile acids was not completely inhibited. Cholic acid feeding appeared to increase the total size of the bile acid pool, but it did not affect the lithogenicity of bile or secretion rates of cholesterol. The data show that at low doses CDCA lowers lithogenicity of bile by reducing hepatic secretion of cholesterol, while cholic acid does not have a similar effect. Oral administration of chenodeoxycholic acid (CDCA) has been shown to cause dissolution of cholesterol gallstones in man. I - 3 Stone dissolution appears to be associated with a reduction in lithogenicity Received January 25,1974. Accepted July 31, Address requests for reprints to: Secretary, Phoenix Clinical Research Section, National Institute of Arthritis, Metabolism, and Digestive Diseases, Phoenix Indian Medical Center, Phoenix, Arizona The authors gratefully acknowledge helpful discussions with Dr. Robert S. Gordon and Dr. Peter H. Bennett. They also thank Dr. Eunice Flock for biochemical consultation. Excellent technical assistance of bile, i.e., a reduction in concentration of cholesterol relative to the solubilizing lipids-bile acids and lecithin. I, 4, 5 Theoretically, a decrease in lithogenicity could be achieved through an increase in hepatic secretion of bile acids and/or lecithin, or by was provided by Mr. Robert Collins, Mr. Elliott Groxzek and Mr. James Hobza. Mrs. Margaret Hendrikx and Mrs. Marjorie Kennel rendered excellent dietetic and secretarial service. Finally, the authors are grateful to the Indian Health Service and the Phoenix Indian Medical Center for cooperation and provision of excellent nursing services. 326

2 February 1975 CHENODEOXYCHOLIC ACID 327 a decrease in secretion of cholesterol. It has been generally assumed that CDCA feeding decreases lithogenicity by increasing the secretion of bile acids since pool sizes of bile acids are abnormally low in many patients with cholesterol gallstones,6, 7 but the possibility that CDCA could also interfere with cholesterol secretion has not been excluded. To elucidate the mechanisms by which CDCA lowers the relative cholesterol content of bile, we have examined the effects of this bile acid on hepatic secretion rates of biliary lipids and pool sizes of bile acids; these studies were carried out before and after the feeding of relatively low doses of CDCA in subjects without gallstones. In addition, the results were compared to those obtained by the feeding of cholic acid (CA), a bile acid that has failed to improve bile composition or to dissolve gallstones. 3, 4 Our results show that bile lithogenicity is reduced even with low doses of CDCA; this fall in lithogenicity is primarily due to a decrease in cholesterol secretion, and may occur without an increase in bile acid secretion or pool size. In contrast, CA feeding failed to significantly improve bile lithogenicity or to decrease hepatic secretion rates of cholesterol. Methods Patients Studies were carried out on the metabolic ward of the Phoenix Indian Medical Center, Phoenix, Arizona. The age, sex, and race of the subjects are given in table 1. Six of the subjects were white and 4 were American Indians. All were in good health, and without evidence of metabolic or hepatobiliary disorders. Since all subjects were asymptomatic, oral cholecystograms were not routinely carried out. Each patient gave informed consent before initiation of studies. Diets Six subjects were admitted to the metabolic ward and were fed a mixed solid food and formula diet containing approximately 40% of calories as lard. The composition of this diet has been presented previously." In 4 Indian subjects, studies were carried out on an outpatient basis; these patients were instructed to maintain their previous dietary habits. Bile Acid Feeding Five of the subjects admitted to the metabolic ward were studied over a period of 3 months after a crossover design: during the 1st month, which was the control period, no bile acids were given; during the 2nd month, either CDC A (Weddel Pharmaceuticals, London, England) or CA (Matheson, Coleman, and Bell, Norwood, Oh.) was given as a single dose daily in gelatin capsules; during the 3rd month, each patient received the other bile acid. When CDCA and CA were subjected to gas-liquid chromatography, more than 98% of bile acids were in the form of the primary bile acid; lithocholic acid was not detectable in either preparation. The dosage given throughout the study is indicated in table 1. One subject stayed for only 2 months, and he received CDCA after a control period of 1 month. Outpatients received CDCA for 1 month, and they did not take CA. One of the outpatients (no. 10) was able to take only 250 mg of CDCA per day because of symptoms of diarrhea and cramping at the 500-mg dose. Another outpatient (no. 9), who weighed 105 kg, tolerated 1500 mg with no symptoms whatsoever. Liver Function Tests Plasma cholesterol, SGOT, bilirubin, and alkaline phosphatase were measured at the beginning of the study and on a weekly basis on all subjects during bile acid feeding. Biliary Lipids In the patients on the metabolic ward, the following parameters were measured at the end of each monthly period: (1) lipid composition of gallbladder bile (aspirated from the duodenum Subject TABLE 1. Clinical data Age/race/sex CDCA Daily dose" mg CA 1 20/Cau/F /Cau/M /Cau/M /Cau/F /Cau/M /Cau/M /Ind/F /Ind/F /1nd/F /1nd/F 250 a CDCA, chenodeoxycholic acid; CA, cholic acid.

3 328 ADLER ETAL. Vol. 68, No.2 after stimulation of gallbladder contraction}; (2) hourly outputs of biliary cholesterol, bile acids, and phospholipids; (3) pool sizes of bile acids; and (4) specific composition of bile acids. Bile acids were not fed on the day these measurements were carried out. The 4 Indian subjects, who were studied as outpatients, had these same measurements before and after 1 month of CDCA feeding; they underwent the same studies as the inpatients except that they did not take CA. The following studies were carried out at the end of each period. Lipid composition of gallbladder bile. After an overnight fast, a three-lumen tube was positioned in the duodenum with the most proximal outlets adjacent to the ampulla of Vater, and the other outlet 10 to 12 cm distally. The tube was placed in the correct position with X-ray guidance. Gallbladder contraction was stimulated by the intra duodenal infusion (through the distal lumen) of a dilute solution of amino acids (casein hydrolysate, Cutter Laboratories, Berkeley, Calif.). Gallbladder bile mixed with intraduodenal contents was then collected by siphonage through the two proximal lumens over the next 30 min. The collected bile was thoroughly mixed, and a 10-ml sample was retained for analysis; the remainder was returned to the patient via the tube. Cholesterol and lecithin content of the bile was measured, as previously described. 9 Bile acids were determined by an automated modification of a standard enzymatic procedure. lo, 11 In previous studies, the term "lithogenic" bile has been used to denote bile that is supersaturated with cholesterol For this reason, we introduced the term "lithogenic index" as a numerical expression for the relative lithogenicity of bile, 12 According to our definition, the lithogenic index indicates the degree to which bile is saturated with cholesterol; the maximum solubility of cholesterol in bile was based on the criteria of Admirand and SmaiL II, 13 More recently, Carey and Small l4 have redefined this original limit of cholesterol saturation as the metastable-labile limit of cholesterol solubility; above this limit, cholesterol exists in a labile state and it will precipitate out of solution in less than 8 hr. Immediately below the limit is the metastable zone in which precipitation will eventually occur after a more prolonged period. The Imver limit of the metastable zone is the equilibrium solubility line, as described by Holzbach et al 15 and by Hegart and Dam 16; this limit defines the maximum solubility of cholesterol at equilibrium. Thomas and Hofmann l7 have recently presented mathematical equations to describe the metastable-labile limit and the equilibrium solubility line. They propose that lithogenic indices could be based on either definition of the limit of cholesterol solubility, and calculations of these indices can easily be carried out by computer using their equations. In this study, lithogenicity of bile was calculated in three ways: (a) as the molar percentage of cholesterol, according to Admirand and Small 1i ; (b) as the lithogenic index based on the metasta ble-iabile limit of Admirand and SmalJ1\ and (c) as the lithogenic index based on the equilibrium solubility line defined by Holzbach et al.,15 by Hegart and Dam,16 and, more recently, by Carey and SmaiL 14 The molar percentage of cholesterol was calculated as follows: moles of cholesterol/[moles of cholesterol + moles of bile acids + moles of lecithin] x 100. Lithogenic indices were calculated according to the equations of Thomas and Hoffmann. 17 The lithogenic indices and the term "lithogenic" are meant to be used in a comparative sense, and no attempt has been made to demarcate precisely between lithogenic and nonlithogenic bile. Hepatic secretion rates of biliary lipids (cholesterol, bile acids, and lecithin) were determined by the marker dilution method of Grundy and Metzger.9 In brief, after gallbladder bile was obtained as described above, liquid formula containing 40% fat and a dilution marker ([14C ]cholesterol) were infused continuously through the most proximal lumen. After allowing 3 to 4 hr for stabilization of bile secretion, hourly samples were obtained for 8 hr from the middle and distal lumens by slow and continuous aspiration. Evidence has previously been presented to show that duodenal bile obtained during this latter period is derived predominately from hepatic secretion without admixture of gall bladder bile. 8, 9 Estimations of bile acids removed after gallbladder contraction and throughout this perfusion study showed that less than 10% of the total pool of bile acids was withdrawn. Since the rate of marker infusion was known, measurement of the ratio of cholesterol to marker at the distal outlet gave the rate of output of biliary cholesterol These data, combined with measurements of the concentration of bile acids and lecithin relative to cholesterol at the proximal outlet, permitted calculation of the hourly secretion rates of bile acids and lecithin, There was no tendency for a reduction in bile acid outputs throughout the ste'ady state period, as might have occurred if a significant portion of the bile acid pool had been removed. Bile ac id pool size. On the evening before the measurement of biliary lipid outputs, and 4 to 6

4 February 1975 CHENODEOXYCHOLIC ACID 329 hr after his last meal, each subject swallowed the three-lumen tube. After allowing time for the distal outlet to pass into the duodenum, 3 to 5 /Joc of ["C ]cholic acid (New England Nuclear Corp., Boston, Mass.) in a small volume of ethanol were injected through the distal lumen; the tube was then flushed with 44 meq of sodium bicarbonate in 50 ml of water. On the following morning after the overnight fast, a portion of the sample of gallbladder bile was taken for estimation of bile acid pool size. Measurement of total bile acid concentration (C IlA ) and concentration of radioactivity (Cdpm ) in this sample allowed calculation of the bile acid pool according to the formula: Total bile acid pool (mg) = (CSA)/CdPm) [radioactivity (dpm) administered] In our laboratory, we have carried out extensive studies to validate this estimation of pool size from a single bile sample (unpublished data); our results can be summarized as follows: 15 subjects had two separate estimates of pool size by this technique to test its reproducibility. The average difference between these two estimates was only 5.2% (range, 0.4 to 14.9%). In 14 subjects, 16 measurements of pool size by this method were compared to estimates by the method of Lindstedt. 18 Because of the extrapolation of bile acid specific activity across a 12-hr fasting interval to zero time by the Lindstedt method, 18 all pool size estimates by our method were higher than the corresponding estimate by the Lindstedt procedure. The average difference between the two methods was 13.7% with a range of 1.3 to 27.9%. This difference depends largely on the definition of zero time. According to Lindstedt, 18 the zero time is the time of isotope administration, while our method defines the zero time as approximately 10 hr after the isotope is given. Since the true zero time must depend on a variety of factors involving relations between mixing time, new synthesis, and loss of bile acids from the pool, we can see no a priori reasons for choosing one time in preference to the other. Specific bile acid composition. The relative proportions of the individual species of bile acids present in the gallbladder bile were determined by gas-liquid chromatography of the trimethylsilyl ethers of the methyl esters of bile acids on 1% Hi-Eff 8BP (Supelco, Inc., Bellefonte, Pa.), as previously described. 19 On Hi-Eff 8BP columns, all hydroxy bile acids were found to give the same detector response on a weight basis l9 ; therefore, no corrections were required for differences in responses between the bile acids.. Cholesterol Absorption and Fecal Neutral Steroids Estimations of cholesterol absorption were made once during each period by oral administration of a single dose of [4-C 14 ]cholesterol and [22,23-3 H]/3-sitosterol. These labeled compounds were obtained from New England Nuclear Corporation, and they were purified by thin layer chromatography before use!o Our procedure for estimating cholesterol absorption by this method has been as described by Quintao et at!1 Ten days before the end of each period, the subjects who were studied on the metabolic ward were given 1.6 /JoC of both [4-"C lcholesterol and [22,23-3 H 1/3-sitosterol. The compounds were administered at 2 PM, just before the intake of the regular liquid formula meal. Thereafter, all stools were collected for 10 days, and ratios of C":H3 were determined for each day's collection. Cholesterol absorption was calculated according to the following equation: Percentage of cholesterol absorbed = 1-(C":H 3 in feces) x [C":H 3 administered)] x 100 In accordance with the method of Quintao et a1.,21 the quantities of radioactivity in the above calculations represented the total amounts excreted over an 8-day period. On the stool samples obtained during the last 7 days of each period, fecal neutral steroids were also estimated. The method used was essentially that described by Miettinen et a1. 20; however, in this procedure, isolation of neutral steroid fractions ( ~, 3t10H; ' 5aH, 3t10H; and 5aH, 3-keto compounds) by thin layer chromatography was eliminated; omission of this step was possible because of the small amounts of the plant sterol, campesterol, in the fecal samples. Simultaneous analyses with and without thin layer chromatography were shown to give identical results. Excretions of neutral steroids were corrected for losses and for variations in fecal flow with /3-sitosterol; 22 each patient received 200 mg of /3-sitosterol twice daily in capsule form throughout the entire study. 8 Results Lipid composition of gallbladder bile. CDCA feeding reduced lithogenicity of gallbladder bile in 9 of the 10 subjects studied (table 2). Paired ranked analysis showed this reduction to be significant at the P < 0.05 level for molar percentage of cholesterol and lithogenic indices. The molar percentage of cholesterol averaged

5 330 ADLER ETAL. Vol. 68, No.2 24% lower after CDCA feeding than after the control period, and lithogenic indices showed a comparable decrease. CA feeding, on the other hand, did not significantly reduce the relative proportion of cholesterol in gallbladder bile. Hepatic secretion rates. The administration of CDCA was consistently associated with a decreased rate of hepatic secretion of cholesterol. Secretion rates for cholesterol, shown in table 3, averaged 23% lower after CDCA feeding than after the control period, and rates were reduced in every subject studied (P < 0.01 by paired ranked analysis). The significance of this observation is further supported by the results from the 5 subjects who received both CDCA and CA in a crossover design under controlled conditions on the metabolic ward. As shown in figure 1, every subject had a lower cholesterol secretion after CDCA than after control or CA, regardless of the order of administration. In contrast, there was no significant difference between rates of cholesterol secretion in control and CA periods. Rates of hepatic secretion of TABLE 2. Relative cholesterol content and lithogenic indices of gallbladder bile during three periods a Patient Cholesterol Lithogenic index' Control COCA CA Control COCA molar % Means a CDCA, chenodeoxycholic acid; CA, cholic acid., Admirand and Small. 11, Holzbach et al Lithogenic index' CA Control COCA CA TABLE 3. Biliary lipid secretion rates in normal volunteers during control period and after i-month periods of chenodeoxycholic acid (CDCA) or cholic acid (CA) feeding Subject Cholesterol secretion Bile acid secretion Lecithin secretion Control COCA CA Control COCA CA Control COCA CA mg/hr mg/hr mg/hr 1 35 a b b b a Each value represents the mean of eight hourly determinations. b Secretion studies after CDCA not completed because of equipment failure.

6 February 1975 CHENODEOXYCHOLIC ACID o Pollent Control _ Cheno CJ Cholic FIG. 1. Cross-over study of the effects of COCA and CA feeding on biliary cholesterol secretion. Each bar represents the mean ± SE of 8 hourly determinations. Left to right sequence of bars for each subject indicates temporal sequence of study periods. Subject 5 did not complete secretion study following COCA because of equipment failure. In every other subject, cholesterol output is lower after COCA than after control or CA. bile acids and lecithin, unlike those of cholesterol, showed no consistent or significant change with feeding of either CDC A or CA. Bile acid pool size. As shown in table 4, no significant increase in total bile acid pools occurred after administration of CDCA. In 7 of 10 subjects, the total pool size of bile acids was numerically smaller after CDCA feeding. A different result was found with CA feeding; in the 5 subjects studied on the metabolic ward, CA produced a mean increase in pool size of about 1 g over that of the control value (3.63 versus 2.74 g; P < 0.05). Individual bile acids. Both CDCA and CA produced significant changes in relative proportions of individual bile acids in bile, as summarized in table 5. Even at relatively low doses of exogenous bile acids, all the subjects showed an increased proportion of the bile acid being administered. Neither of the exogenous bile acids, however, completely replaced the other primary bile acid. It is therefore unlikely that bile acids given in these low doses completely inhibit endogenous bile acid synthesis. Lithocholic acid, which is produced by bacterial dehydroxylation of CDCA in the colon and is potentially hepatotoxic,23 was not increased after CDCA feeding and was never greater than 2% of total bile acids. As shown in table 5, ursodeoxycholic acid, the 7 -{1 epimer of CDCA, was present in small amounts during the control period and significantly increased (P < 0.05, Student's t-test) after CDCA feeding, but not after CA feeding. (Identification of ursodeoxycholic acid was accomplished by combined gas-liquid chromatography and mass spectrometry, kindly performed by Drs. Gerald Salen and G. S. Tint.) Cholesterol absorption and fecal neutral steroids. Values for fecal neutral steroid TABLE 4. Total bile acid pool size in normal volunteers before bile acid feeding (control), and after 1-month periods of chenodeoxycholic acid (CDCA) or cholic acid (CA) Subject Pool size Control CDCA CA ll Mean ± SE 2.77 ± ± ± 0.33 g TABLE 5. Bile acid composition before and after feeding of chenodeoxycholic acid and cholic acid in normal subjects Period Per cent of total bile acids ± SE Chenodeoxycholic Cholic Beoxycholic Lithocholic Ursodeoxycholic Control (N = 10) 34.8 ± ± ± ± ±.5 CDCA(N = 10) 73.1 ± ± ± ± ± 1.3 CA (N = 5) 15.1 ± ± ± ± ± 1.1

7 332 ADLER ETAL. Vol. 68, No. 2 excretion and for cholesterol absorption are presented in table 6. Neutral steroid excretion consists of endogenous neutral steroids plus unabsorbed dietary cholesterol. The dietary intake for patients of this study averaged about 200 mg per day. Since absorption of cholesterol averaged 35%, fecal excretion of unabsorbed dietary cholesterol should have been on the order of 130 mg per day. Thus, the bulk of fecal neutral steroids was derived from endogenous cholesterol. No significant differences were found in excretion of neutral steroids between any of the three balance periods, and values for cholesterol absorption were not constantly altered by bile acid feeding. However, in this small series, 3 of the subjects receiving CDCA had relatively low values for cholesterol absorption, as compared to the control and CA feeding period. Toxicity and side effects. All subjects were asymptomatic at the dosage shown in table 1. Attempts to administer twice this dose early in the 1st month of bile acid feeding caused abdominal cramps and diarrhea in all the subjects except 1 (no. 9). The dose was therefore reduced to 500 mg per day in all subjects, with the exception of no. 9 who received 1500, and no. 10 who received 250. Although serum alkaline phosphatase and bilirubin remained unchanged in every case, weekly determinations of SGOT revealed transient increases in 6 subjects during CDCA feeding. In only TABLE 6. Cholesterol balance and absorption data Neutral steroid excretion and cholesterol absorption Subject Control CDCN CN mg/day excretion (% absorption) 419 (29) 503 (18) (39) 494 (26) (32) 593 (47) (23) 632 (16) (59) 456 (60) Mean steroid excre- 485 ± ± 33 tion (mg/day ± SE Mean absorption (36 ± 6) (33 ± 8) (% ± SE) a CDC A, chenodeoxycholic acid period. b CA, cholic acid period. 457 (57) 538 (43) 547 (52) 487 (45) 415 (34) 488 ± 25 (46 ± 4) 2 of these subjects did these elevations exceed the upper limit of normal; both these subjects remained asymptomatic, and repeat determinations done within 1 week were again normal. Plasma cholesterol and triglyceride concentrations did not change from control val ues during the 1 month of CDCA or CA feeding. Discussion The primary purpose of this study was to elucidate the mechanisms by which CDCA decreases bile lithogenicity. This effect could result from either enhanced hepatic secretion of bile acids and/or lecithin or from decreased secretion of cholesterol. The most obvious possibility is that CDCA feeding expands the bile acid pool and consequently increases bile acid secretion. Although there is evidence from recent studies of Danzinger et al. 24 that expansion of pool sizes can occur in some patients given CDCA, our findings indicate that CDCA feeding also reduces lithogenicity by decreasing the secretion of cholesterol, independent of any effect on bile acid secretion or pool size. CDCA feeding reduced the hepatic secretion rates of cholesterol in all 10 of our subjects. Moreover, in the 4 subjects whose secretion rates were measured both after CDCA and after CA feeding, cholesterol secretion was always lower after CDCA than after CA, regardless of the order of administration. Bile acid secretion and pool size, meanwhile, showed no regular pattern of change, despite consistent improvement in bile composition. These findings are in accord with those reported by Northfield et al. 25; these workers measured 24-hr biliary outputs of cholesterol, bile salts, and lecithin in 4 patients taking CDCA, and they also found a decrease in cholesterol secretion with little or no change in bile acid or lecithin secretion. Although this study suggests that CDCA improves bile composition by reducing the hepatic secretion of cholesterol, the doses of CDCA used in this study, although effective in improving 'bile composition, were considerably smaller than those used in clinical trials so far reported. I - 5 More-

8 February 1975 CHENODEOXYCHOLIC ACID 333 over, most of our subjects had bile acid pools of normal size, whereas subjects in previously reported clinical trials were comprised of gallstone patients, whose bile acid pools were presumably small. 6, 7 Danzinger et al 24 have recently shown that feeding of 0.75 to 4.5 g per day of CDCA expands bile acid pools in patients with gallstones and small bile acid pools. Thus, higher doses of bile acids in patients with reduced pools probably decreases lithogenicity by enhancing bile acid secretion, in addition to reducing cholesterol secretion. There are several mechanisms by which CDCA might decrease hepatic secretion of cholesterol. First, it might inhibit synthesis of cholesterol, as suggested by several studies in animals Interestingly, Schoenfield et al. 28 have reported that CDCA causes a greater inhibition of cholesterol synthesis in hamster liver than does CA. Studies in man have also shown that bile acids regulate cholesterol synthesis Thus, if CDCA is relatively more potent in inhibiting cholesterol synthesis than CA, hepatic secretion of cholesterol could be reduced when CDCA is the predominant bile acid in the enterohepatic circulation. Second, CDCA might alter biliary secretory mechanisms so as to reduce secretion of cholesterol relative to other biliary lipids. If the primary effect of CDCA is to alter the mechanisms for hepatic cholesterol secretion, two important questions must be raised: namely, (1) whether reduced cholesterol outputs are due to a toxic effect of CDCA on the liver, and (2) whether inhibition of cholesterol secretion might lead to an expansion of body pools of cholesterol. The possibility of hepatic toxicity must be considered in view of the transient mild elevations in SGOT seen in our subjects, as well as those of Thistle and Hofmann 3 ; however, bile acid and lecithin secretion remained unchanged, and there were no elevations in serum bilirubin, alkaline phosphatase, or cholesterol to suggest cholestasis. Moreover, cholesterol secretion was decreased during CDCA feeding in all 10 subjects, whereas SGOT values exceeded normal in only 2. Therefore, it seems unlikely that decreased outputs of cholesterol can be explained by hepatic toxicity. The question of the effects of CDCA on the body pool of cholesterol has not been resolved, but in a preliminary report by Thistle et ai., 32 no change in total pool size of cholesterol was found during long term CDCA feeding. A third mechanism for reduced outputs of biliary cholesterol could be a decreased reabsorption of cholesterol; by this mechanism, less cholesterol would be returned to the liver from the gut. Actually, results of this study are compatible with this possibility. Despite a decreased biliary secretion of cholesterol by the liver, fecal excretion of cholesterol (neutral steroids) was not decreased; this finding is consistent with a reduction in the quantity of endogenous cholesterol reabsorbed. Direct measurements of cholesterol absorption during CDCA feeding were not conclusive. Three of our 5 patients showed less absorption during CDCA feeding as compared to the control period and treatment 'Yith CA. Clearly, additional studies are needed to elucidate the effects of CDCA feeding on cholesterol absorption. In this study, we have shown that reduced lithogenicity of bile can be obtained with relatively low doses of CDCA. The higher doses employed in previous human studies have been found to suppress endogenous bile acid synthesis,24 which is one of the major routes of cholesterol removal. Our finding that CA was still present in the bile after 1 month of low dose CDCA, suggests that the conversion of cholesterol to bile acids was not abolished, thus reducing the potential for accumulation of cholesterol in body pools. This finding also strengthens the possibility that therapeutic effects of CDCA can be achieved at doses that will avoid possible adverse effects. In our view, serious consideration should be given to the use of relatively low doses of CDCA in clinical trials designed to test the usefulness of this bile acid of gallstone dissolution. REFERENCES l. Danzinger RG. Hofmann AF, Schoenfield LJ, et al: Dissolution of cholesterol gallstones by chenodeoxycholic acid. N Engl J Med 268:1-8, 1972

9 334 ADLER ETAL. Vol. 68, No.2 2. Bell GD, Whitney B, Dowling RH: Gallstone dissolution in man using chenodeoxycholic acid. Lancet 2: , Thistle JL, Hofmann AF: Efficacy and specificity of chenodeoxycholic acid therapy for dissolving gallstones. N Engl J Med 289: , Thistle JL, Schoenfield LJ: Induced alterations in composition of bile of persons having cholelithiasis. Gastroenterology 61: , Thistle JL, Schoenfield LJ: Lithogenic bile among young Indian women. N Engl J Med 284: , Vlahcevic ZR, Bell CC, Gregory DH, et al: Relationship of bile acid pool size to the formation of lithogenic bile in female Indians of the southwest. Gastroenterology 62:73-83, VIahcevic ZR, Bell CC, Buhac I, et al: Diminished bile acid pool size in patients with gallstones. Gastroenterology 59: , Grundy SM, Metzger AL, Adler RD: Mechanisms of lithogenic bile formation in American Indian women with cholesterol gallstones. J Clin Invest 51: , Grundy SM, Metzger AL: A physiologic method for estimation of hepatic secretion of biliary lipids in man. Gastroenterology 62: , Talalay P: Enzymic analysis of steroid hormones. Methods Biochem Anal 8: , Admirand WH, Small DM: The physiochemical basis of cholesterol gallstone formation in man. J Clin Invest 47: , Metzger AL, Heymsfield S, Grundy SM: The lithogenic index-a numerical expression for the relative lithogenicity of bile. Gastroenterology 62: , Small DM: Gallstones. N Engl J Med 279: , Carey MC, Small DM: Solubility of cholesterol in aqueous bile salt-lecithin solutions. Importance of metastability, lipid concentration, and temperature (abstr). Gastroenterology 64:706, Holzbach RT, Marsh M, Olszewski M, et al: Cholesterol solubility in bile: evidence that supersaturated bile is frequent in healthy man. J Clin Invest 52: , Hegart FG, Dam H: The solubility of cholesterol in aqueous solutions of bile salts and lecithin. Z Ernaehrungswiss 10: , ThomasPJ, HofmannAF: Asimple calculation of the lithogenic index of bile: expressing biliary lipid composition on rectangular coordinates. Gastroenterology 65: , Lindstedt S: The turnover of cholic acid in man. Acta Physiol Scand 40:1-9, Grundy SM, Ahrens EH Jr, Miettinen TA: Quantitative isolation and gas-liquid chromatographic analysis of total fecal bile acids. J. Lipid Res 6: , Miettinen TA, Ahrens EH Jr, Grundy SM: Quantitative isolation and gas-liquid chromatographic analysis of total dietary and fecal neutral steroids. J Lipid Res 6: , Quintao E, Grundy SM, Ahrens EH Jr: An evaluation of four methods for measuring cholesterol absorption by the intestine in man. J Lipid Res 12: , Grundy SM, Ahrens EH Jr, Salen G: Dietary i3-sitosterol as an internal standard to correct for cholesterol losses in sterol balance studies. J Lipid Res 9: , Palmer RH: Bile acids, liver injury and liver disease. Arch Intern Med 130: , Danzinger RG, Hofmann AF, Thistle JL, et al: Effect of oral chenodeoxycholic acid on bile acid kinetics and biliary lipid composition in women with cholelithiasis. J Clin Invest 52: , Northfield TC, LaRusso NF, Thistle JL, et al: Effect of chenodeoxycholic acid therapy on biliary lipid secretion in gallstone patients (abstr). Gastroenterology 64:780, Hamprecht B, Roscher R, Waltinger G, et al: Influence of bile acids on the activity of rat liver 3-hydroxy-3methylglutaryl coenzyme A reductase. 2. Effect of cholic acid in lymph fistula rats. Eur J Biochem 18:15-19, Shefer S, Hauser S, Lapar V, et al: Regulatory effects of sterols and bile acids on hepatic 3-hydroxy-3-methylglutaryl CoA reductase and cholesterol 7a-hydroxylase in the rat. J Lipid Res 14: , Schoenfield LJ, Bonorris GG, Ganz P: Induced Alterations in the rate-limiting enzymes of hepatic cholesterol and bile acid synthesis in the hamster. J Lab Clin Med 82: , Grundy SM, Hofman AF, Davignon J, et al: Human cholesterol synthesis is regulated by bile acids (abstr). J Clin Invest 45:1018, Grundy SM, Ahrens EH Jr, Salen G: Interruption of the enterohepatic circulation of bile acids in man: comparative effects of cholestyramine and ileal exclusion. J Lab Clin Med 78:94-121, Salen G, Nicolau G, Shefer S: Chenodeoxycholic acid (CDCA) inhibits elevated HMG-CoA reductase activity in subjects with gallstones (abstr). Clin Res 21 (3}:522, Thistle JL, Hoffman NE, Hofmann AF: With the technical assistance ofd.t.a. Belobaba. Effect of bile acid administration on cholesterol kinetics in patients with radiolucent gallstones (abstr). Gastroenterology 65:572, 1973