with 45 per cent of the calories derived from fat; insulin, one (Ber) with familial hypercholesteremia

Transcription

1 Effect of Antibiotics on the Serum Cholesterol Concentration of Patients with Atherosclerosis By ALFRED STEINER, M.D., ELLIOTT HOWARD, M.D., AND SUAT AKGUN, M.D. A PRELIMINARY STUDY' from this laboratory indicated that the oral administration of neomycin was associated with a significant decrease in serum cholesterol concentration in all of 1 patients. This investigation was prompted by the chance observation that a fall in the level of serum cholesterol occurred in a patient coincident with the oral administration of neomycin, 2 Gm. daily, in the treatment of a gastrointestinal infection due to Salmonella Tennessee. Additional studies are now presented together with experiments designed to clarify the mechanism of action of neomycin in lowering the serum cholesterol level. Evaluation of the effects of other antibiotics on the serum cholesterol levels is also reported. Method Fifteen male and nine female patients, aged 21 to 72 years, represent the subject material. Twentyone were hospitalized and three were observed as outpatients. The clinical diagnosis of each subject is listed in table 1. Total serum cholesterol levels were determined by the method of Abell et al.2 twice weekly in the hospitalized subjects and at weekly or biweekly intervals in the outpatients. The ester fraction was determined by the method of Schoenheimer and Sperry.3 Total serum lipid values were obtained by a gtavimetric procedure.4 lipid phosphorus by the method of Fiske and Subbarow,5 and alpha- and betalipoproteins by paper electrophoresis.6 Seventeen of the 21 hospitalized patients were on the regular hospital diet of 2,2 calories, From the First (Columbia University) Division, Goldwater Memorial Hospital, Department of Hospitals and Department of Medicine, College of Physicians and Surgeons, Columbia University, New York. Aided in part by the Health Research Council of the City of New York under contract no. U-117, the Revlon Fund, Thomas Phillips Fund, and Albert and Mary Lasker Foundation. 729 with 45 per cent of the calories derived from fat; one (Nas) received a diabetic diet of carbohydrate 18 Gm., protein 8 Gm., fat 8 Gm. without insulin, one (Ber) with familial hypercholesteremia was on a low-fat diet in which 25 per cent of the calories were derived from fat; two (Lew, Les) were maintained on a similar low-fat diet after a preliminary period of 6 weeks on the regular hospital diet. The three outpatients ate an unrestricted and unknown diet during the study. Average control serum lipid values were established during periods of 4 to 14 weeks (mean 6.7 weeks) prior to administration of the antibiotics. The number of control determinations averaged 12.7 per patient. Oral neomycin was the initial drug studied and was given at the following daily dose level:.5 Gm. during five periods, 1 Gm. during five periods, 1.5 Gm. during two periods, 2 Gm. during 15 periods, and 4 Gm. during two periods. The duration of the periods of drug administration varied from 4 to 36 weeks, the mean was 8.5 weeks. In four patients, successive increasing amounts of neomycin were given after 4 to 6 weeks at each dose level without intervening control periods. At the termination of the oral neomycin, a second control period of 5 to 12 weeks was carried out. Many of the subjects were then tested with other oral antibiotics including kanamycin, bacitracin, achromycin, aureomyein, chloromycetin, mycostatin, and streptomycin as well as the following antibiotics given by intramuscular injection: neomycin, streptomycin, and penicillin.* In each instance, a control period of observation preceded the administration of each drug, or each special study when it was desirable to evaluate the effects of a combination of antibiotics. An alteration of 1 per cent or more from the mean serum cholesterol value of the control period has been found to be necessary to have statistical significance. In a previous report7 from this laboratory, it was shown that the serum cholesterol level in patients with coronary atherosclerosis is inconstant and varies widely. The standard deviation of the mean serum cholesterol values in *The authors are indebted to the Upjohn, Squibb, Lederle, Pfizer, and Parke Davis Companies for a generous supply of antibiotics.

2 73 STEINER, HOWARD, AKGUN Effect of Oral Neomycin on Serum Cholesterol Table 1 Control Neomycin Average No. of total No. of Average Average determin- cholesterol determin- total fall Patient, sex, age, diagnosis Weeks ations mg./1 ml. Gm. Weeks ations cholesterol % 1. PUP, F, 59, cerebral thrombosis ± 37* ± PUP, F, 59, cerebral thrombosis ± ± PUP, F, 59, cerebral thrombosis ± ± WAL, M, 58, hyperten. heart dis ± ± WAL, M, 58, hyperten. heart dis ± ± WAL, M, 58, hyperten. heart dis ± ± LEW, M, 69, cerebral thrombosist ± LEW, M, 69, cerebral thrombosis ± ± NAS, M, 5, rheumatic heart dis ± ± NAS, M, 5, rheumatic heart dis ± ± BER, F, 52, familial hypercholesteremia, coronary art. dis.t ± ± LES, M, 72, coronary art. dis.t ± ± LES, M, 72, coronary art. dis.t ± ± OST, M, peripheral neuritis ± ± OST, M, peripheral neuritis ± ± BER, F, 21, familial hypercholesteremia ± ± HAM, F, 36, familial hypercholesteremia ± WAL, M, 64, coronary art. dis ± SIL, M, 68, coronary art. dis ± ± GOF, M, 62, coronary art. dis ± ± HRO, F, 68, cerebral thrombosis ± ± HRO, F, 68, cerebral thrombosis ± ± SCA, M, 6, emphysema ± ± KAT, F, 52, coronary art. dis ± CON, F, 65, cerebral thrombosis ± ± GOL, M, 64, coronary art. dis ± ± TRL, M, 54, coronary art. dis ± ± BAL, F, 68, cerebral thrombosis ± SCH, F, 7, cerebral thrombosis ± ± 26 3 *Maximum deviation from the mean. tlow-fat diet during control period. patients with coronary atherosclerosis was almost three times the standard deviation of control subjects. The difference between mean serum cholesterol values of any control and treatment period should thus approximate 1 per cent in order to have statistical significance. In the present study, there was no exclusion of serum lipid values during transition periods when the antibiotics tested had not become fully effective. Results Neomycin The results of the oral administration of neomycin to 2 patients during 29 periods are summarized in table 1. The amount of oral neomycin given daily varied from.5 to 4 Gm. The duration of therapy varied from 4 to 36 weeks, the mean was 8.5 weeks. After the initiation of oral neomycin, the serum cholesterol level decreased significantly and usually reached the lowest level after 2 weeks (figs. 1 and 2). This decrease in serum cholesterol was maintained for the duration of the neomycin period. In some instances, to keep the serum cholesterol at the low point, the amount of neomycin was increased from.5 or 1 Gm. daily to 1.5 or 2 Gm. daily (fig. 1). Increasing the amount of oral neomycin from.5 to 2 Gm. tended to increase the average fall in the level of serum cholesterol. At a dose of.5 or 1 Gm. daily (1 periods) the

3 ANTIBIOTICS AND SERUM CHOLESTEROL NEOMYCIN DO- NEOMYCIN. 1.5 GM. G. MYCOSTAT1W PUP, Q.59 YR 3 K1ANAMYCI1N NAS.d. 52 YR. NEOMYCIN...,_. _J _i s2 cj )(-19% ) (-5) (-25%(28) h WI I- X U 2( ( lo - -2 % ) -998) Figure 1 Effect of oral neomycin on serum cholesterol. average decrease of the mean serum cholesterol concentration of each period was 17 per cent, at 1.5 or 2 Gm. daily (15 periods) the average fall in the mean serum cholesterol concentration was 2 per cent. When the amount of the drug was increased to 4 Gm. (2 periods) the average fall in the mean serum cholesterol concentration was 22 per cent. The neomycin was not increased further because this 2 per cent additional fall did not appear large enough to warrant further study. It was also considered that possible toxic effects of the drug might appear. In the 29 periods of neomycin administration to 2 patients, the fall in average serum cholesterol of each period varied from 1 to 29 per cent, the average being 21 per cent. When oral neomycin was discontinued, the serum cholesterol level returned to control values, usually in 2 weeks (figs. 1 and 2). The prior lowering of the serum cholesterol by a low-fat diet did not prevent a further decrease in serum cholesterol by oral neomycin. In three patients (Lew, Les, Ber), a lowfat diet (25 per cent of calories from fat) for 4 to 6 weeks resulted in an average fall in the serum cholesterol level of 1, 12, and 18 per cent. Neomycin (1.5 Gm. daily) was then added to this program and caused a further fall in the average serum cholesterol concentration of 2, 18, and 17 per cent, respectively. In contrast unpublished studies have demonstrated that the feeding of unsaturated fats to patients on low-fat diets will not fur Figure 2 Effect of oral antibiotics on serum cholesterol. ther decrease the serum cholesterol values.8 Oral neomycin is poorly absorbed from the gastrointestinal tract, and it has been determined that 97 per cent of the neomycin so administered can be recovered in the stools.9 The action of neomycin may be due to the effect locally in the intestinal wall or on the intestinal contents. Two patients were given 6 mg. of neomycin intramuscularly daily for 2 weeks without any appreciable change in the serum cholesterol level. This amount of neomycin was calculated to exceed that proportion of the drug that is absorbed from the gastrointestinal tract. In addition to a fall in serum cholesterol concentration, oral neomycin resulted in a decrease in total serum lipids in each of the patients, the average fall being 23 per cent. The esterified fraction of serum cholesterol decreased in the serum in proportion to the fall in total serum cholesterol. The serum phospholipids also decreased in a similar proportion to the fall in serum cholesterol with the cholesterol/phospholipid ratio remaining unchanged. The alpha- and beta-lipoprotein ratios were not significantly altered by neomycin. Kanamycin Kanamycin is chemically closely related to neomycin and is also poorly absorbed from the gastrointestinal tract. Ten periods of oral kanamycin administration have been studied in seven patients. After a prior control pe-

4 732 STEINER, HOWARD, AKGUN Table 2 Effect of Oral Kanamycin on Serum Cholesterol Control Kanamycin Average No. of total No. of Average determi- cholesterol determi- total Difference Dose Patient Weeks nations mg./1 ml. Weeks nations cholesterol % Gm. 1. Nas ± ± Les ± ± Les Ost Ost ± Lew ± Sea ± Wal ± ± Wal ± ± Con ± ± 38 2 riod, the drug was given daily in amounts of 2 Gm. for 3 to 5 weeks in seven instances. In two of the drug periods, a significant fall in the serum cholesterol level occurred, averaging 13.5 per cent (table 2 and fig. 2). At the 4-Gm. dose level in three trial periods, there was one instance of a significant fall (greater than 1 per cent) in the level of serum cholesterol. Although kanamycin produced a fall in serum cholesterol in 3 of 1 instances, its effect is smaller and less consistent than that of neomycin. Streptomycin and Bacitracin The combination of intramuscular streptomycin plus oral bacitracin was thought likely to result in a similar effect as oral neomycin on the bacterial flora of the gastrointestinal tract. One gram of streptomycin (intramuscular) plus 2,5 units of oral bacitracin were given to three patients for 2 weeks after ar initial control period. No significant alteration of the serum cholesterol level occurred. At the end of 2 weeks, 2 Gm. of oral neomycin was added to the streptomycin and bacitracin and in each instance, a significant fall in serum cholesterol level occurred. After 2 weeks of the three medications, the bacitracin and streptomycin were discontinued and the patients continued on the oral neomycin for one additional week. The fall in serum cholesterol level was maintained in each of the three patients (table 3 and fig. 3). Aureomycin In the previous experiment, when the oral neomycin was stopped, the three patients were transferred to oral aureomycin, 1 Gm. daily for 2 weeks. In one patient, the serum cholesterol level fell further, while in the remaining two patients, the serum cholesterol level increased but did not return to the control level. Additional studies on the effect of aureomycin on the serum cholesterol level are contained in table 4 and figure 4. It can be seen that oral aureomycin (1 Gm. daily for 2 weeks) was prescribed for five periods to four patients. In three of the five periods, a significant fall (greater than 1 per cent) in the serum cholesterol occurred averaging 17 per cent. In two periods, the decrease in serum cholesterol was not significant. Aureomycin had a similar but less marked effect than neomycin on the serum cholesterol level. Achromycin Achromycin was administered orally for six periods, 1 Gm. daily for 2 weeks, to four patients. In two instances, the average fall in serum cholesterol level was at borderline significance, while in four periods, the average decrease was less than 1 per cent. Chloromycetin Chloromycetin was given orally to three patients at the dose level of 1 Gm. daily for 2 weeks without any significant alteration of the serum cholesterol level.

5 ANTIBIOTICS AND SERUM CHOLESTEROL STREPTOMYCIN I GM(I.M) BACITRACIN 25(4 U1 NEOMYCII 2 GM. AUREO GOFF df.s YR n. 3 W 2- en a AUREOM CIN AUREOM CIN GM GM. 1~~~~~ 1, 1G LES d 73YR. W (-221) (-148) 1-.) ' ' h b X 8.. Figure 3 Effect of oral antibiotics on serum cholesterol. Mycostatin (-49) (-199)(-25tkrlg Oral mycostatin has been administered alone and in conjunction with oral neomycin. Two million units of mycostatin were given for seven periods of 2 weeks to five patients. In three instances, decreases of borderline significance in serum cholesterol level occurred. When mycostatin was given in conjunction with oral neomycin, the same decrease in serum cholesterol concentration occurred as when neomycin alone was given (fig. 2). Mycostatin did not appear to potentiate or decrease the effect of neomycin. Streptomycin Oral streptomycin (2 Gm. daily) was administered to two patients for 2 weeks without producing an alteration of the level of the serum cholesterol. Penicillin Intramuscular crystalline penicillin, 1.2 million units daily, was given to three patients for 2 weeks and resulted in no effect on the level of the serum cholesterol. Discussion The results presented revealed that oral neomycin consistently produces a significant, sustained fall in the level of serum cholesterol. Of the other antibiotics tested, in a 12 O Figure 4 Effect of oral antibiotics on serum cholesterol. limited number of patients, aureomycin and kanamycin resulted in less consistent and less marked decreases in the level of serum cholesterol. Oral achromycin, chloromycetin, bacitracin, streptomycin, mycostatin as well as intramuscular streptomycin, penicillin, and neomycin were ineffective. A number of studies were designed to throw some light on the possible mechanism of action of neomycin. Preliminary studies of bacteriologic content of stools reveal that oral neomycin produces a profound effect upon the intestinal flora. Stool cultures during control periods were compared with those obtained during the neomycin period. Escherichia coli were most effected; with a marked reduction to absence of these organisms from the stool. There is also a suppression of the other enteric organisms except for the enterococei and Clostridium welchii which tend to become more abundant. Further studies of the effect of neomycin upon the intestinal flora are necessary before the mechanism of action of neomycin can be attributed to its bacteriostatic action. An alteration in the absorption of fat from the gastrointestinal tract has been considered as one of the possible mechanisms of action of neomycin. For this reason, in cooperation with Dr. A. R. Wertheim, oral fat tolerance tests were carried out in eight patients prior to and during the oral administration of neomycin. The test meal consisted of 3 ml. of medium cream (28 per cent fat). In addition

6 734 STEINER, HOWARD, AKGUN Effect of Antibiotics on Serum Cholesterol Table 3 Streptomycin 1 Gm./day Bacitracin 25, Units/day (4 weeks) Control (2 Weeks) Neomycin 2 Gm./day (3 weeks) Aureomycin (2 Weeks) (1 Week) 1 Gm./day (2 weeks) Average Average Average Average Average total No. total No. total No. total No. total No. cholesterol determi- cholesterol determi- cholesterol determi- cholesterol determi- cholesterol determimg./1 ml. nations mg./1 ml. nations mg./1 ml. nations mg./1 ml. nations mg./1 ml. nations Gof 23 ± ± Nas ± ± Sea 22 ± ± ± Effect of Oral Aureomycin on Serum Cholesterol Table 4 Control Aureomycin Control Before medication (1 Gm. daily) After medication Average Average Average Difference No. of total No. of total No. of total % determi- cholest'"rol determi- cholesterol Difference determi- cholesterol (from Weeks nations mg./1 ml. Weeks nations mg./1 ml. %1 Weeks nations mg./1 ml. first control) Arr ± ± Ost ± ± Les ± ± Les Gof to determination of the serum lipids, the turbidity of the serum was measured. During the neomycin period the response of the serum turbidity was altered so that the wide range of turbidity exhibited by the patients with atherosclerosis was replaced by the response obtained in healthy "control" subjects. The various serum lipid moieties revealed less response to the fat meal when the patients were on neomycin than in the absence of the drug. This difference, however, was not remarkable. Jacobson et al.1 have reported the production of a malabsorption syndrome induced by the ingestion of 12 Gm. of neomycin daily. This large amount of medication produced a steatorrhea as well as diarrhea in most patients. Goldsmith et al.1' has reported increased fecal excretion of bile salts coincident with oral neomycin administration. In the present study, significant diarrhea did not occur. Several loose stools daily for short periods were recorded in three patients. It was not necessary to discontinue the medication in any individual. Steatorrhea was not encountered. The weights of our subjects were not altered significantly during this study. No evidence of renal or hepatic damage due to neomycin was demonstrated by the usual laboratory determinations. Summary Oral neomycin, at the dose level of.5 to 2 Gm. daily, produced a significant fall in the level of serum cholesterol in all of 29 periods in 2 patients studied. The effect reached a maximum in approximately 2 weeks and persisted for the duration of the experiments up to 9 months. After the medication was discontinued, the serum cholesterol levels returned to control values within 2 weeks. No significant toxic reaction to oral neomycin was discernible. Neomycin by injection in excess of the amount calculated to be absorbed from the gastrointestinal tract failed to lower the serum cholesterol level. The mechanism of action of neomycin was not determined, but it is

7 ANTIBIOTICS AND SERUM CHOLESTEROL7 735 thought to be a local effect in the gastrointestinal tract or its contents. Oral aureomyein and oral kanamycin had a similar but less marked and less consistent effect on the level of serum cholesterol in a number of patients. In a limited number of patients, oral achromycin, bacitracin, chloromycetin, streptomycin, and mycostatin did not have a significant effect on the serum cholesterol level; likewise, intramuscular penicillin and streptomycin failed to alter the serum cholesterol level. The prior lowering of the serum cholesterol concentration by a low-fat diet did not prevent a further decrease in the serum cholesterol by oral neomycin. Acknowledgment The authors are indebted to Drs. J. Bakke and M. Finkel for their helpful assistance in the clinical care of the patients in this study. References 1. SAMUELS, P., AND STEINER, A.: Effect of neomycin on serum cholesterol level of man. Proc. Soc. Exper. Biol. & Med. 1: 193, ABELL, L. L., LEVY, B. B., BRODIE, B. B., AND KENDALL, F. E.: A simplified method for the estimation of total cholesterol in serum. J. Biol. Chem. 195: 357, SCHOENHEIMER, R., AND SPERRY, N. M.: Micro method for determination of free and combined cholesterol. J. Biol. Chem. 16: 745, BLOOR, W. R.: Determination of small amounts of lipids in blood plasma. J. Biol. Chem. 77: 53, FISKE, C. H., AND SUBBAROW, Y.: Calorimetric determination of phosphorus. J. Biol. Chem. 66: 375, KUNKEL, K., AND TISELIUs, A.: Electrophoresis of proteins on filter paper. J. Gen. Physiol. 35: 89, STEINER, A., AND DOMANSKI, B.: Serum cholesterol level in coronary arteriosclerosis. Arch. Int. Med. 71: 397, STEINER, A., HOWARD, E., AND AKGUN, S.: Unpublished observations. 9. WAKSMAN, S. A.: Neomycin. Baltimore, Williams & Wilkins Company, JACOBSON, E. D., CHODOS, R. B., AND FALOON, W. W.: An experimental malabsorption syndrome induced by neomycin. Am. J. Med. 38: 524, GOLDSMITH, G. A., HAMILTON, J. G., AND MILLER,. N.: Lowering of serum lipid concentrations. Arch. Int. Med. 15: 512, 196. Experience In beginning teaching twenty and more years ago, I determined that the basis of what I taught should be that which I myself had seen and proved to be true. A second ideal that I have striven hard to attain is simplicity in teaching; this was not so easy during the years in which my work lay largely amongst pulse and galvanometric curves. But I have become more and more convinced of the need for simplicity; and it was apparent that little of the graphic work could profitably be taught to men entering practice, whose foremost interest must always be in observations they themselves can make upon their patients.-sir THOMAS LEWIS. Diseases of the Heart. New York, The MacMillan Company, 1933, p. v.