537
Atherosclerosis,
@ Elsevier
24 (1976)
Scientific
537-549 Publishing Company,
Amsterdam
HALOFENATE IN THE TREATMENT HYPERLIPOPROTEINEMIA DOUBLE BLIND COMPARISON
in The Netherlands
OF TYPE II
WITH CLOFIBRATE
E. DOUGLAS REES, RONALD D. HAMILTON SHIRLEY WASSON and THOMAS HEARN Clinical Research Laboratory and Lipid University of Kentucky Medical Center,
- Printed
*, IRVING
F. KANNER
Clinic, Departments of Medicine Lexington, KY., (U.S.A.)
**, and Pharmacology,
(Received 23rd December, 1975) (Revised, received 30th March, 1976) (Accepted 30th March, 1976)
Summary A double-blind study comparing halofenate, a new lipid-lowering investigational drug, with an established drug, clofibrate, was conducted on 33 clinic patients with Type II hyperlipoproteinemia for a period of 48-96 weeks. All but 10 patients had some type of symptomatic major vascular disease. With respect to serum cholesterol levels, a comparable proportion (56-59%) of patients in each group responded to the respective treatment but the magnitude of lowering was substantially less for the halofenate responders (12% mean decrease versus 25%). Type IIa patients in both groups were more likely than Type IIb patients to have a favorable cholesterol-lowering response. Weight gain of 5% or greater was prejudicial to cholesterol lowering. In the case of serum triglycerides, the proportion of patients responding to clofibrate treatment was somewhat greater (87% versus 57% for halofenate) but the mean magnitude of lowering (27-34%) was comparable for responders in the two groups. Weight gain did not influence appreciably the triglyceride-lowering effect. Elevated concentrations of triglyceride (Type IIb) in the control period favored a triglyceride lowering response by clofibrate but was only a moderate influence on the response to halofenate. This study was supported by grants from Merck, Sharp and Dohme and from the National Institutes of Health (RR-158-06). * Present Address: Wright State University School of Medicine, Dayton, Ohio, U.S.A. * * Deceased. Reprint requests to Dr. E.D. Rees, Department of Medicine, University of Kentucky, Lexington, Kentucky 40506. U.S.A.
538
The lowering (35-42%) of serum uric acid by halofenate was prompt and sustained in most patients. An adverse reaction (generalized rash and hives) necessitated cessation of halofenate in only 1 patient. In all, halofenate is a convenient drug which is comparable in effectiveness to clofibrate in the lowering of serum triglyceride concentrations and, to lesser degree, in the lowering of serum cholesterol concentration for patients with Type II hyperlipoproteinemia. Key
words:
Cholesterol - Clofibrate - Halofenate - Hyperlipoproteinemia sponders -Responders - Triglycerides - Type II - Uric acid
-
Nonre-
Introduction Halofenate, 2-acetamidoethyl @-chlorophenyl) (m-trifluoromethylphenoxy) acetate, is a new lipid-lowering investigational drug developed by Merck, Sharp and Dohme. The uricosuric effect of this drug has been stressed [l-6] thus far in clinical investigations. Additional metabolic effects include the displacement of thyroid hormone from specific binding proteins with consequent lowering of total but not free thyroid hormone concentration in the plasma [ 2,7], an antidiuretic affect which may prove useful in the management of pitressin-sensitive diabetes insipidus [ 81, potentiation of the hypoglycemic effect of sulfonylureas [8a], and a lowering effect on the concentration of serum lipids, especially triglycerides. The serum lipid-lowering effects on patients with Types III, IV and V hyperlipoproteinemias have been studied in considerable detail [9--141, but no study has been confined to patients with Type II abnormalities. The present study was conducted on patients with Type II hyperlipoproteinemia and a double-blind comparison with an established drug, clofibrate, was made. Subjects
and Methods
Thirty-three patients with a Type II lipoprotein serum electrophoretic pattern and serum cholesterol values of 270 mg/dl or greater participated in this double blind study. Each patient was randomly assigned to the halofenate or clofibrate group; a placebo identical in appearance to the alternate drug was also given. No patient weighed under 45.5 kg, so most halofenate patients received 1 g (2 tablets) once a day; a few patients weighed over 109 kg and received 1.5 g once a day. Clofibrate was administered in a dose of 1 g (2 capsules) twice each day. Tablet and capsule counts were performed. Seventeen patients (6 women, 11 men) received halofenate and 16 patients (8 women, 8 men) received clofibrate. Mean age was 45.4 (range of 35-65 years) and 47.1 (range of 39-60) years at the onset of treatment for those on halofenate and clofibrate, respectively. Of the 33 patients, 16 (7 halofenate, 9 clofibrate) had documented coronary artery heart disease, 7 (3 halofenate, 3 clofibrate) had peripheral vascular disease, 4 (3 halofenate, 1 clofibrate) had cerebral vascular disease and 10 (6 halofenate, 4 clofibrate) were symptomatically free of vascular disease. Three (1 halofenate and 2 clofibrate) of these 33 patients had vascular disease involving more than one region. All patients were informed about the nature of the study and signed a consent form before admitted to the study. They continued their usual diets during the study.
539
After 5 Lipid Clinic visits during an &week control period, drug therapy was immediately initiated and visits were then made at end of the second and fourth weeks and subsequently at 4-week intervals through the 48th week of treatment. Most patients had never received a lipid-lowering drug and none received treatment for at least 2 months prior to onset of the control period. On each visit patients were examined by a physician and serum lipids determined. Total serum cholesterol was determined by the procedure of Franey and Mador [ 151 and in all instances standards and patients’ and control serums were analyzed in duplicate. Reconstituted Lipitrol@ (Dade) and frozen serum obtained from the Center for Disease Control (CDC), Atlanta, Georgia, served as the control serums. The coefficients of variation for Lipitrol and CDC control serums were +3.7% and *3.9%, respectively, and the mean values were 296 mg/dl versus a stated value of 300 mg/dl for Lipitrol and 304 versus 296 for the CDC control. The manual method of Levy [ 161 was used for triglycerides. Standards (triolein), control serums, and serum from patients were analyzed together and in duplicate. The same control serums were used as for cholesterol analyses and the coefficients of variations were ?6.7% and +8.8%, respectively, for Lipitrol and CDC preparations. The mean value obtained for Lipitrol was 208 mg/dl versus the stated value of 218 mg/dl, and for the CDC control serum a mean of 226 mg/dl was obtained for a stated value of 247 mg/dl. Lipoprotein electrophoresis was carried out on paper in a human albumin containing buffer according to the method of Lees and Hatch [ 171; the classification of abnormalities was made according to the system of Fredrickson and Lees [18]. On a periodic basis, red and white blood cell counts and hematocrits were performed by a Coulter counter and sizer. A blood urea nitrogen and serum sugar were obtained by Technicon-Autoanalyzer techniques, serum uric acid spectrophotometrically at 292 nm using the uricase method (Sigma Chemical Corporation), creatine phosphokinase according to Rosalki [ 191, total serum protein by the biuret method, and albumin by electrophoresis and quantitative densitometry. Alkaline phosphatase activity was determined by the nitrophenol method [ 201, bilirubin according to Stoner and Weisberg [ 211, glutamic-pyruvate and glutamic-oxalacetate transaminases by spectrophotometry at 340 nm (Sigma Chemical Corporation). Electrocardiograms were obtained during the control period and after the 24th and 48th weeks of drug therapy. Eye examinations were generally performed on the same day as the ECG by members of the Department of Ophthalmology at the University of Kentucky Medical Center. Results Serum
lipids: halofenate
versus clofibrate
The 5 lipid determinations obtained per patient during the control period were averaged and these individual mean values covered a substantial range, 270-518 mg/dl for cholesterol and 77-435 mg/dl for triglycerides. In view of the wide variation in control period mean values between patients, treatment values were analyzed in terms of percent of the mean control values for indivi-
540
dual patients. The standard deviation associated with the mean control values was *6.4% and the standard error ?0.5%. The serum concentrations of cholesterol and triglyceride were obtained every 4 weeks, but for summary purposes treatment periods were separated into l-8, 9-24, 25-36, 37-48 weeks after start of drug treatment. With respect to the individual drugs and the individual serum lipid determinations, there were 2 overt groups of treated patients -those who responded to treatment and those who did not. Responders to treatment showed in at least 2 of the 4 treatment periods a mean concentration decrease at least 10% below the mean value for the control period. Except for one patient in the first period, no responder had an increase in lipid concentration above the control mean for any period during course of treatment. The demarcation provided by this criterion separated without overlapping the group of mean treatment values for patients responding to therapy from the mean values for patients not responding to therapy. Among responders, the mean cholesterol-lowering effect of halofenate was about one-half that of clofibrate (Table 1). For both drugs, the lowering effect in responders was generally sustained throughout the treatment period. The effect on serum triglyceride level for responders was comparable (approximately a 30% decrease) for both halofenate and clofibrate (Table 1). Again, the lowering effect when present was usually sustained over the treatment period. In most cases, if a patient did not respond within the first 8 weeks of treatment, a significant response did not occur later. Mean lipid concentrations during the control period and during the 2 periods of greatest lowering response are presented (Table 2) for individual patients responding to drug treatment. Porportions of responders Patients with Type II hyperlipoproteinemia TABLE MEAN Percent
can be further
divided into 1 of
1 CHANGES
IN SERUM
change from control
Drug
Response
LIPIDS
AFTER
START
Period (weeks of treatment)
a
9-24
l-8 Cholesterol Halofenate
Clofibrate
Triglycerides Halofenate
Clofibrate
OF TREATMENT
period mean f SE.
25-36
31-48
responders(S) nonresponders(7)
-11.0 -3.0
f 4.1 f 2.1
-13.6 -t2.5 -3.9 r 3.0
-11.9 + 2.8 +2.6 f 3.5
-12.2 f 3.5 +5.4 f 4.9
responders (9) nonresponders
(7)
-22.1 -3.3
f 1.9 ? 1.2
-24.8 t 3.6 -1.9 fi2.6
-24.2 + 2.8 -4.1 + 2.1
-28.0 f 3.0 -2.4 k 2.8
responders (8) nonresponders
(7)
-27.0 + 7.2 +fi.of 13.8
-33.7 + 2.3 +26.3 t 18.1
-30.1 + 4.5 +35.8 f 20.1
-15.0 +15.7
responders (14) nonresponders (2)
-28.5 i 3.2 -7 f 2
-33.7 k 4.3 -4 + 2
-28.6 It 5.8 -5 +12
-24.4 f 5.5 +1 + 29
a Comparisons between responders and nonresponders the nonparametric Mann-Whitney U-test.
are aI significant
+ 11.1 + 13.8
at the P < 0.01 or less level by
541
TABLE
2
CHANGES IN THE WITH TREATMENT
MEAN
Control
Halofenate period
period
Cholesterol 333 275 315 353 297 284 284 302 281 425
LIPID
treatment
CONCENTRATIONS
control
period
INDIVIDUAL
Clofibrate period
293 217 246 311 235 244 244 263 235 a 369 a
271 316 342 356 359 356 518 436 283
211 224 239 228 302 260 357 244 232
94 72 97 85 92 76 141 135 164 142 a
435 128 169 88 298 209 210 233 124 187 117 241 168 157
165 78 78 55 200 146 118 165 93 120 68 113 136 108
(mg/dl)
a Treatment
was stopped
during the first period.
as the mean for two periods
RESPONDERS
treatment
The treatment
period
value of other patients is presented
of greatest lowering.
3
RESPONDERS PROTEINEMIA
AND
Type of hyperlipoproteinemia
NONRESPONDERS
Males
FOLLOWING
TREATMENT
Females
Lowering a of serum cholesterol Yes
Halo feenote IIa IIb Total Clofibrate IIa IIb To
OF
(mg/dl)
Triglycerides 144 128 170 129 192 106 205 264 216 169
TABLE
SERUM
ta1
a Responders mean control
FOR
No
TYPE
II HYPERLIPO-
Lowering of serum triglycerides Yes
5 6
4 2
7 3
2 5
5 5
4 3
11
6
10
7
10
7
2 7
4 3
6 3
0 7
4 10
2 0
9
7
9
7
14
2
had to show a decrease in the respective serum lipid concentration value in two or more of the 4 treatment periods.
-
NO
at least 10% below
-
the
542
2 subgroups: IIa (elevated cholesterol only) and IIb (elevated triglyceride as well as cholesterol). Patients with mean triglyceride values above 150 mg/dl in the control period were classified as IIb. Patients representing each of the 2 subtypes were found in both treatment groups. On grouping according to subtype, responders and nonresponders to halofenate were found in both the IIa and IIb group, though a tendency for cholesterol lowering to occur more often in IIa patients and for triglyceride lowering to occur more often in IIb patients seems present (Table 3). This latter tendency was more pronounced in patients treated with clofibrate. The proportion of responders were comparable among halofenate and clofibrate patients. (In Tables 2 and 3, two patients in the halofenate group who had substantial decrease in serum cholesterol are included among the responders, though one dropped from the study after 3 treatment weeks (allergic rash) and the other after 12 weeks due to an accidental shooting). The analysis was carried a further step (Table 4) by examining various combinations of cholesterol-lowering and triglyceride-lowering effects according to treatment and subtype of lipoprotein disorder. With either drug, over 80% of patients showed a lowering at least 10% below control values in one or both of the serum lipids studied. The 5 patients with xanthelasma had considerable or nearly complete clearing of these lesions during the 48-week treatment period; this effect was not dependent on the presence of a serum lipid-lowering response to treatment. The large tendinous xanthomas of one patient regressed to approximately one-fourth pretreatment size during halofenate treatment, even though serum cholesterol remained in the 400-500 mg/dl range. (Also, this patient’s xanthelasma almost completely disappeared).
Body weight After 48 weeks of treatment
the mean increase
in body weight for patients
TABLE 4 RELATIONSHIP BETWEEN CHOLESTEROL-LOWERING AND TRIGLYCERIDE-LOWERING EFFECTS ACCORDING TO TREATMENT AND SUBTYPE OF LIPOPROTEIN DISORDER Drug
Halofenate Ha IIb
Combination of lowering effects Chol-TG
Chol-TG
Chol-TG
Chol-TG
Yes
Yes No (number of patients) 4 0 -
No
No
(23.5%)
(23.5%)
Yes 3 3 -
(35.3%) Clofibmte Ha IIb
4 3 (43.8%)
2 0 (12.5%)
Yes 2 2
0 7 (43.8%)
No 0 3 -
(17.6%)
0 0 (0%)
Each number in parenthese indicates, for a given treatment, the percent of patients have a particular combination of responses.
543 TABLE MEAN
5 INCREASE
% Over weight
IN BODY
in control
Period
Halofenate Clofibrate
WEIGHT
DURING
STUDY
period. of treatment
(weeks)
l-8
9-24
25-36
3748
0.9% 0.8%
2.2% 2.6%
4.1% 2.9%
4.7% 2.7%
on clofibrate was 2.7% above the mean weight of the control period; for halofenate the mean increase was 4.7% (Table 5). Two patients on clofibrate and one patient on halofenate lost 5% or more body weight. Five patients on clofibrate and 6 patients on halofenate showed a gain of 5% or more in body weight. In order to evaluate the possible effect of weight gain on the response to treatment, the proportion of responders among those with the weight gain of 5% or more were determined. Weight gain reduced the proportion of patients having a cholesterol-lowering response to either halofenate or clofibrate (Table 6); but the small number of patients in each group should be noted. In contrast, weight gain influenced the triglyceride-lowering effect of treatment slightly if at all. Serum uric acid
The lowering of serum uric acid by halofenate was prompt and sustained (Table 7); this effect was evident on the first visit 2 weeks after start of treatment. All but 3 of 15 halofenate-treated patients showed a decrease greater than 20%, and the mean decrease for these responders was 35-42%. Of the
TABLE
6
INFLUENCE Weight gain
Yes No
OF WEIGHT
GAIN
Treatment
and response
RESPONSE
b
Halofenate
Clofibrate
Cholesterol-lowering l/7 (14%) 6/8 (75%)
215 7/11
p < 0.10 c
N.S.
Triglyceride Yes No
a ON TREATMENT
(40%) (64%)
lowering
517 (71%) 5/8 (63%)
5/5 9/11
N.S.
N.S.
(100%) ( 82%)
a Patients who had a weight gain of 5% or greater during the 48-week treatment period were considered to have gained significant weight. b The numerator designates the number of patients who had a lipid-lowering response and the denominator presents the number of patients who gained weight (“yes”) or did not gain weight (“no”). ’ Using Yate’s Chi-Square correction for small numbers to assess these 2 X 2 tables; N.S. indicates nonsignificance at the 0.10 level.
544 TABLE
7
MEAN
LOWERING
Percent
change
OF SERUM
from control
URIC ACID
period,
Drug
mean f. S.E.
Period (weeks after start of treatment) 9-24
l-8
36-48
25-36
Halofenate (6.22 f 0.25 mg/dl) all patients (15) -21.0 responders only (12) -39.0
* 4.5 + 2.3
-28.2 -41.7
? 5.3 f 1.8
-21.5 -35.2
?r 9.5 2 3.2
Clofibrate (5.91 + 0.27 mg/dl) alI patients
k 2.3
-1.5
+ 2.3
-9.1
f 3.3
-8.0
-27.6 -36.6
+0.5
+ 5.8 + 3.1
k 6.1
Values in parenthesis following the drug are the mean f SE values of serum uric acid levels for the control period: other parentheses denote number of patients.
3 nonresponders, none had an elevated serum uric acid concentration and, in fact, one had a mean value of 3.0 mg/dl for the control period. Clofibrate did not produce an effect on uric acid comparable to that of halofenate. Although 3 patients in the halofenate group and 3 patients in the clofibrate group had slightly or moderately elevated levels of uric acid in their serums, no patient in either group had a history of symptomatic gout, nor did gouty symptoms develop in any patient during course of treatment. Monitoring of selected blood and serum components Analyses for a number of blood or serum components (Table 8) were used to monitor possible effects of treatment on various organ systems or on other metabolic processes. Besides effects on the serum lipids and serum uric acid, both described separately above, a few other appreciable effects were noted. Serum albumin concentration increased modestly (0.3-0.4 mg/dl) in addition to a slightly greater increase in serum total protein (0.5-0.6 mg/dl) which must reflect slightly increased (0.2 mg/dl) concentration of serum globulins as well. These effects on serum protein concentration fell in both halofenate- and clofibrate-treated patients. There was a modest increase in the mean serum creatine phosphokinase activity for halofenate-treated patients. Patients treated with clofibrate had a slight but significant decrease in mean serum alkaline phosphatase activity. Adverse reactions Only one adverse reaction necessitating cessation of treatment occurred during the study. This patient, a 25 year-old nurse who had a history of allergic reactions to propoxyphene, merthiolate and keflin, developed a generalized rash and urticaria after 2 weeks of halofenate treatment. The drug was immediately stopped; both rash and urticaria responded to symptomatic treatment. Acute situations developed during the treatment of 4 other patients. Bleeding from an unsuspected duodenal uicer developed after 4 weeks of halofenate treatment in a 49-year-old farmer with a history of peptic ulcer. The drug was discontinued for one month and then resumed without furt’ler difficulty for 92
(12.5-15.5.F: 14-17.M)
(38-46.F: 41--5O.M)
(4-12)
Hemoglobin (g/dl)
Hct(%)
WBC
ha1 cl0 ha1 cl0 ha1 cl0 ha1 cl0 ha1 cl0
(8-19)
(6-8.2)
(3.5-4.5)
(0.2-1.0)
(640.F; 7-54.M)
(0.7-2.7)
(O-35)
(O-28)
Semmureanitrogen (mg/dl)
Serum protein (g/d0
Serum albumin (g/dl)
Totalbilirubin (mg/dl)
CPK (mIU/ml)
Alk.phos. (BLB)
SGPT (Karmen)
SGOT (Karmen)
0.9 1.0
0.3 0.4
* *
0.1 0.1
6 5
33 35
0.9 1.1
0.3 0.4
f *
* 1 -t 1
14 18 + +
14 16
1 1
+1 5 1
9 12 i- 1 i 1
12 12
0.1 0.1 b
1.7 c 1.3 c
i' 5 *lo
0.9 0.04
0.43 + 0.39 + 40 49
0.1 0.1
0.1 0.1
4.0 * 4.0 f
7.4 + 7.6 *
1.1 0.9
2 3
f14 f36
14.2 + 14.9 +
103 101
265 245
0.4 0.5
0.7 1.1
0.3 0.4
+ +
0.08 0.06
0.44 f 0.39 +
1 1
t * * *
11 13 14 18
1 1
0.1 0.1 b
2.0 + 1.2 f
?; 5a + 6
0.1 0.1
4.2 f 4.2 *
49 44
0.2 0.1 a
7.5 f 7.8 +
0.7 0.8
2 3
*13 f29
14.0 + 15.3 *
102 105
252 338
6.7 f 8.5 ?
45.2 f 42.6 +
15.2 * 14.2 f
9-24
treatment (weeks)(means
6.9 ? 0.5 8.3 f 0.6
45.1 * 43.6 +
15.2 + 14.6 +
l-8
Period of
2.1 * 0.2 1.9 t 0.1
+ +
0.06 0.01
0.58 + 0.47 c
4.0 + 0.1 4.0 -r 0.1
7.4 + 7.5 *
0.8 0.7
4 2
+ 22 f 27
13.4 * 15.4 *
107 105
272 320
6.9 -t 0.5 8.5 + 0.5
45.2 + 44.3 *
15.2 + 14.9 +
Control
PERIODS
f +
t + + i
11 11 15 19
1.8 + 1.2 +
47 48
0.36 + 0.26 f
4.1 t 4.4 ?
2 1
2 2
0.2 0.1 b
8 9
0.08b 0.05b
0.1 0.1 b
7.5 ?; 0.1 7.9 f 0.1 b
0.9 1.1
?ll ? 4 15.3 f 15.2 +
114 109
+ +
+ *
8a 8
O.Olb O.lb
O.la 0.1 b
0.1 b 0.1 b
0.6 1.0
7 3
14 18
10 12
* +
f +
1 1
1 1
2.3 -t 0.3 1.2 f 0.1 b
55 42
0.22 + 0.26 +
4.3 f 4.4 f
7.9 + 8.1 f
14.7 f 14.9 *
110 100
0.5 0.6
0.9 1.1
0.3 0.4
*19 * 23
237 307
*17 +19
246 305
44.9 f 43.5 f 6.9 f 8.4 +
1.0 1.1
43.6 + 42.9 +
15.1 f 14.6 i
37-48
6.3 i- 0.4 8.2 ? 0.5
0.4 0.4
14.9 f 14.5 f
25-36
* SE)
aP < 0.05:b P < 0.01 for comparisons between treatment and control values.F = range for normalfemales,and M = range for normal.males.
ha1 cl0
ha1 cl0
ha1 Cl0
ha1 cl0
(70-110)
Serum glucose (fasting.mg/dl)
ha1 Cl0
(150450)
ha1 cl0
ha1 Cl0
ha1 Cl0
Drug
Platelets (thousands/mm3)
(thousands/mm3)
Normal (range)
component (units)
LEVELSOFVARIOUSBLOODANDSERUMCOMPONENTSDURINGCONTROLANDTR~ATMENT
TABEL8
a
546
weeks (i.e. 44 weeks remaining in the initial study plus 48 weeks in an extension period). After 40 weeks of halofenate treatment a 53-year-old housewife, who had taken aspirin preparations chronically for an arthritic condition, developed symptoms of gastric ulcer which was verified by X-ray, gastroscopy and biopsy. The ulcer resolved on removal of aspirin and on addition of antacid therapy. Halofenate was stopped for one week, then resumed for the final 8 weeks and there was nearly complete healing of the ulcer during that time as observed by endoscopy and X-ray. Two patients receiving clofibrate had intercurrent illnesses during course of their treatment. A 64-year-old farmer with a previous history of stroke suffered a myocardal infarction after 40 weeks of therapy; clofibrate was discontinued for 5 days immediately after the occurrence and, was then resumed uneventfully for the remaining 8 weeks of treatment. After 36 weeks of clofibrate treatment a 44-year-old mother of 3 was hospitalized for depression. Symptoms of cholecystitis developed and a gall bladder containing multiple stones of mixed composition was removed. Therapy was resumed after a ‘I-week hiatus and the subsequent 12 weeks went without difficulty. No deaths occurred during the course of this study. ‘Extended period of halofenate treatment The treatment period for 10 patients receiving halofenate was extended from 48 weeks to 96 weeks. No adverse reactions nor untoward changes in laboratory values occurred during the extended period. The response of the serum lipids and uric acid during the extension remained in accord with that in the initial treatment period. During the extension period, one patient (a husky, 34-year-old male suffering from incapacitating coronary artery disease with severe angina) demonstrated clearly the dependence of the serum lipid response on halofenate dose. (In addition to halofenate, he was taking nitroglycerine, long-acting nitrates and propranolol.) The initial dose based on a body weight of 96 kg was 1 g daily and in the first 24 weeks serum cholesterol decreased 15% (from 310 to 263 mg/dl) and triglyceride serum decreased 32% (from 205 to 139 mg/dl). Subsequently, his weight increased to 106.5 kg and the lipids returned to control period values. During the first 28 weeks of the extension period he attained 110 kg and halofenate dose was increased to 1.5 g daily. This higher dose again decreased the serum cholesterol by 16% and triglyceride by 22% below the control period values. Discussion The lowering of serum uric acid levels by halofenate is well-known [l-6,14, 221. Halofenate increases the urinary excretion of uric acid but does not increase urinary oxypurines (as does allopurinol). These findings suggest halofenate interference with tubular reabsorption of uric acid [ 1,2,5]. Progressive decrease in serum total bilirubin with halofenate and with clofibrate treatment is in accord with previous observations [3,9,23] and this is accredited in part to the displacement of bilirubin from plasma binding proteins, albumin especially. Alkaline phosphatase activity in clofibrate-treated patients decreased moderately and this has been observed previously [23]. Though halofenate-treated
patients in our study did not show a mean lowering of serum alkaline phosphatase activity, such an effect was noted in another study [ 31. Hutchison and Wilkinson [3] observed a modest decrease in serum albumin concentration (but not total protein concentration) in their series of halofenate-treated patients, but our patients had a mean increase in concentration of both albumin and total protein. Different analytical methods may account for this discrepancy. Clofibrate-treated patients also showed a similar small increase in serum albumin and total protein concentration. Appreciable changes in transaminases, serum urea and glucose, red cell count and size, hemoglobin concentration and white blood cell count did not occur in either clofibrate- or halofenate-treated patients in the course of our study. The moderate increase in mean creatine phosphokinase activity for halofenate patients was due to sporadic increases in a few patients and we could not relate these episodes to any change in clinical signs or symptoms. Thus laboratory monitoring revealed few, if any, indication of adverse treatment effects by either halofenate or clofibrate in our study patients. The only pronounced and unequivocal adverse effect for halofenate was a generalized rash in a patient who previously had an allergic reaction to other drugs and who several years before had a short term of clofibrate therapy. Although the chemical structure of halofenate is substantially different from that of clofibrate, both clofibrate and halofenate are tightly bound to serum proteins and may be effective haptens in certain patients. Thus, it seems wise to use halofenate circumspectly and with forewarning of possible allergic response in any patient who has previously been treated with clofibrate and who has an allergic history. The development of a bleeding duodenal ulcer in a man with a history of duodenal ulcer and of a gastric ulcer in a woman who had been taking aspirin for an arthritic condition may or may not represent a complication of halofenate treatment, only further controlled observations will provide an answer. In this regard, however, gastrointestinal complaints were uncommon in our halofenate treated patients - even those treated for 96 weeks; the drug was tolerated well. In the Type III, IV and V patients investigated in other halofenate studies [l-3,13,14] increased serum triglyceride concentration was the most prominent aspect of the lipid abnormality. There is agreement in those studies that halofenate lowered mean serum triglyceride levels by 16-20% or more. Our group of patients differed from those in the previous studies by being limited to patients with Type II hyperlipoproteinemia and in this condition the primary serum lipid abnormality is considered to be the hypercholesterolemia. Despite the apparent primacy of cholesterol elevation, substantial triglyceride lowering was noted in both halofenate- and clofibrate-treated patients. Even patients with normal triglyceride levels (Type IIa) often experienced triglyceride lowering. The magnitude of lowering was comparable among responders for the two drugs, but the chances of a patient responding favorably seemed somewhat greater for clofibrate than for halofenate. Among nonresponders, halofenate treatment in contrast to clofibrate resulted in an elevation of triglyceride levels of some patients such that a substantial mean elevation for nonresponders occurred. Presently we have no explanation for this apparent difference in effect of the two drugs.
548
In the case of cholesterol, the data obtained from patients with Type II hyperlipoproteinemia afforded some similarities and contrasts with data obtained by others from patients having other types of hyperlipoproteinemias. In accord with an earlier study [ 91, halofenate in general is not as proficient as clofibrate in lowering serum cholesterol levels. However, both halofenate and clofibrate appreciably lowered (by 10% or more) cholesterol levels in over one-half of treated patients; a substantially higher proportion than expected from the previous studies on patients with other types of hyperlipoproteinemias. Both halofenate and clofibrate elicited a cholesterol-lowering effect in nearly all Type IIa patients but in only about one third of Type IIb patients. It is now demonstrably clear that the types of hyperlipoproteinemias categorized by the criteria of the Fredrickson and Lees’ system do not constitute genetically homogeneous groups of patients [ 241. Both Type IIa and IIb phenotypes can occur in the genetic disorders of monogenic familial hypercholesterolemia, monogenic combined hyperlipidemia, and polygenic hypercholesterolemia which were recognized in the study of Goldstein et al. [ 251. Since the gene product responsible for the aberrancy in these disorders has not been identified (except for deficiency or defectiveness in cellular receptors for low-density lipoproteins [26]), the extent of metabolic heterogeneity within even the monogenic disorders can not be determined. Given even a limited amount of heterogeneity within a metabolic disorder, differing therapeutic responses to a given drug are not surprising. Until more is known about the biochemical basis of the hyperlipidemic disorders, little can be definitely stated as to why a drug such as halofenate or clofibrate brings about or fails to bring about a lowering of serum cholesterol or triglyceride in an individual patient. Acknowledgements The assistance of Norma Slone, Evangeline Lear, Judy Campbell, Helen Lipscomb, Catherine Fine and Bernetta Mason is acknowledged with pleasure. We are grateful to Dr. Grant Somes for statistical assistance. References 1 2 3 4 5 6 7 8 8a
Jam. A., Ryan, J.R., Hague, D. and McMahon. F.G., The effect of MK-185 on some aspects of uric acid metabolism, CIin. Pharmacol. Therap. 11 (1970) 551. Morgan, J.P., Bianchine, J.R., Hsu, T.H. and Margolis. S.. Hypolipidemic, uricosuric, and thyroxinedisplacing effects of MK-185 (halofenate), CIin. Pharmacol. Therap., 12 (1971) 517. Hutchison. J.C. and Wilkinson, W.H., The uricosuric action of halofenate (MK-185) in patients with hyperuricemia or uncomplicated primary gout and hyperlipidemia, Atherosclerosis, 18 (1973) 353. Aronow. W.S., Harding, P.R., Khursheed, M., Vangrow, J.S. and Papageorge’s, N.P., Effect of halofenate on serum uric acid, CIin. Pharmacol. Therap., 14 (1973) 371. Ravenscroft, P.J., Sands, J.M. and Emmerson, B.T., Studies of the uricosuric action of the hypoIipidemic drug halofenate. CIin. Pharmacol. Therap., 14 (1973) 547. Wolfram, G., Keller. C., KiIami, S. and Zollmar, First cIinicaI experiences in therapy of hyperuricemia and hyperhpidemia, Verhandl. Dtsch. Ges. Inn. Med., 79 (1973) 2191. Ryan. J.R., Jain. A., Maha, G.E. and McMahon, F.G., The effects of MK-185 on thyroid function, Clin. Pharmacol. Therap.. 12 (1971) 464. Gattereau, A., Davingnon, J., Verdy, M. and Lewis, W., Halofenate versus clofibrate in the management of true diabetes insipidus, Canad. Med. Ass. J. 110 (1974) 1275. Jain, A., Ryan, J.R. and McMahon, F.G., Potentiation of hypoglycemic effect of sulfonylureas by halofenate, New En%. J. Med., 293 (1975) 1283.
549
9
Jepson,
E.M.,
Small,
new drug MK-185 10 11
12 13 14
15 16 17 18 19 20 21 22
23 24 25 26
Aronow,
W.S.,
E.,
Grayson,
with clofibrate
Harding,
P.R.,
M.F.,
Gillian, B. and Billimoria.
in the treatment
Khursheed,
of hyperlipidemias,
M., Vangrow.
J.S.
J.D.,
A comparative
Atherosclerosis,
Papageorge’s,
study 16 (1972)
of a 9.
N.P. and Mays, J., Effect
of halofenate on serum lipids, Clin. Pharmacol. Therap.. 14 (1973) 358. Aronow, W.S.. Vangrow. J.S.. Nelson, W.H., Pagano, J., Papageorge’s, N.P.. Khursheed, M., Harding, P.R. and Khemka, M.. Halofenate, serum lipids, and exercise performance in coronary heart disease, Clin. Pharmacol. Therap.. 15 (1974) 67. Hutchison, J.C. and Wilkinson, W.H., Halofenate -Effectiveness of two dosage forms and two dose schedules. Atherosclerosis, 19 (1974) 417. Sirtori, C., Hurwitz, A., Khalid. S. and Azarnoff, D.L., Clinical evaluation of MK-185 -A new hypolipidemic drug, Lipids, 7 (1972) 96. Lisch, H.J., Patsch, J., Sailer, S. and Braunsteiner, H., Comparison of the effects of halofenate pa(MK-185) and clofibrate on plasma lipid and uric acid concentration in hyperlipoproteinemic tients, Atherosclerosis. 21 (1975) 391. Franey, R.J. and Amador, E., Serum cholesterol measurement based on ethanol extraction and ferric chloride-sulfuric acid, Clin. Chim. Acta, 21 (1968) 255. Levy. A.L., Measurement of triglycerides using nonane extraction and calorimetry, Ann. Clin. Lab. Sci., 2 (1972) 474. Lees, R.S. and Hatch, F.T., Sharper separation of lipoprotein species by paper electropboresis in albumin-containing buffer, J. Lab. Clin. Med., 16 (1963) 518. Fredrickson. D.S. and Lees, R.S., A system for phenotyping hyperlipoproteinemia, Circulation. 31 (1965) 321. Rosaliki, S.B.. An improved procedure for serum creatine phosphokinase determination, J. Lab. Clin. Med., 69 (1967) 696. Bessey, O.A.. Lowry, O.H. and Brock. M.J.. A method for the rapid determination of alkaline phosphatase with five cubic millimeters of serum, J. Biol. Chem., 164 (1947) 321. Stoner, R.A. and Weisberg. H.F.. Ultramicro method for serum bilirubin by diazo blue reaction, Clin. Chem., 3 (1957) 22. Aronow. W.S., Vangrow, J.S., Nelson, W.H., Pagano, J.. Papageorge’s, N.P., Khursheed. M., Harding, P.R. and Mahaveer, K., Halofenate - An effective hypolipemiaand hipouricemia-inducing drug, Curr. Therap. Res., 15 (1973) 902. Coronary Drug Project Research Group, Clofibrate and niacin in coronary heart disease, J. Amer. Med. Ass., 231 (1975) 360. Lees, R.S., A progress report on lipoprotein phenotyping, J. Lab. Clin. Med., 82 (1973) 529. Goldstein, J.L.. Schrott, H.G., Hazard. W.R., Bierman. E.L. and Motulsky, A.G.. Hyperlipidemia in coronary heart disease, J. Clin. Invest., 52 (1973) 1544. Goldstein, J.L., Dana, S.E., Brunschede, G.Y. and Brown, M.S., Genetic heterogeneity in familial hypercholesterolemia - Evidence for two different mutations affecting functions of low-density lipoprotein receptor, Proc. Nat. Acad. Sci. (Wash.), 72 (1975) 1092.
Atherosclerosis,
@ Elsevier
24 (1976)
Scientific
537-549 Publishing Company,
Amsterdam
HALOFENATE IN THE TREATMENT HYPERLIPOPROTEINEMIA DOUBLE BLIND COMPARISON
in The Netherlands
OF TYPE II
WITH CLOFIBRATE
E. DOUGLAS REES, RONALD D. HAMILTON SHIRLEY WASSON and THOMAS HEARN Clinical Research Laboratory and Lipid University of Kentucky Medical Center,
- Printed
*, IRVING
F. KANNER
Clinic, Departments of Medicine Lexington, KY., (U.S.A.)
**, and Pharmacology,
(Received 23rd December, 1975) (Revised, received 30th March, 1976) (Accepted 30th March, 1976)
Summary A double-blind study comparing halofenate, a new lipid-lowering investigational drug, with an established drug, clofibrate, was conducted on 33 clinic patients with Type II hyperlipoproteinemia for a period of 48-96 weeks. All but 10 patients had some type of symptomatic major vascular disease. With respect to serum cholesterol levels, a comparable proportion (56-59%) of patients in each group responded to the respective treatment but the magnitude of lowering was substantially less for the halofenate responders (12% mean decrease versus 25%). Type IIa patients in both groups were more likely than Type IIb patients to have a favorable cholesterol-lowering response. Weight gain of 5% or greater was prejudicial to cholesterol lowering. In the case of serum triglycerides, the proportion of patients responding to clofibrate treatment was somewhat greater (87% versus 57% for halofenate) but the mean magnitude of lowering (27-34%) was comparable for responders in the two groups. Weight gain did not influence appreciably the triglyceride-lowering effect. Elevated concentrations of triglyceride (Type IIb) in the control period favored a triglyceride lowering response by clofibrate but was only a moderate influence on the response to halofenate. This study was supported by grants from Merck, Sharp and Dohme and from the National Institutes of Health (RR-158-06). * Present Address: Wright State University School of Medicine, Dayton, Ohio, U.S.A. * * Deceased. Reprint requests to Dr. E.D. Rees, Department of Medicine, University of Kentucky, Lexington, Kentucky 40506. U.S.A.
538
The lowering (35-42%) of serum uric acid by halofenate was prompt and sustained in most patients. An adverse reaction (generalized rash and hives) necessitated cessation of halofenate in only 1 patient. In all, halofenate is a convenient drug which is comparable in effectiveness to clofibrate in the lowering of serum triglyceride concentrations and, to lesser degree, in the lowering of serum cholesterol concentration for patients with Type II hyperlipoproteinemia. Key
words:
Cholesterol - Clofibrate - Halofenate - Hyperlipoproteinemia sponders -Responders - Triglycerides - Type II - Uric acid
-
Nonre-
Introduction Halofenate, 2-acetamidoethyl @-chlorophenyl) (m-trifluoromethylphenoxy) acetate, is a new lipid-lowering investigational drug developed by Merck, Sharp and Dohme. The uricosuric effect of this drug has been stressed [l-6] thus far in clinical investigations. Additional metabolic effects include the displacement of thyroid hormone from specific binding proteins with consequent lowering of total but not free thyroid hormone concentration in the plasma [ 2,7], an antidiuretic affect which may prove useful in the management of pitressin-sensitive diabetes insipidus [ 81, potentiation of the hypoglycemic effect of sulfonylureas [8a], and a lowering effect on the concentration of serum lipids, especially triglycerides. The serum lipid-lowering effects on patients with Types III, IV and V hyperlipoproteinemias have been studied in considerable detail [9--141, but no study has been confined to patients with Type II abnormalities. The present study was conducted on patients with Type II hyperlipoproteinemia and a double-blind comparison with an established drug, clofibrate, was made. Subjects
and Methods
Thirty-three patients with a Type II lipoprotein serum electrophoretic pattern and serum cholesterol values of 270 mg/dl or greater participated in this double blind study. Each patient was randomly assigned to the halofenate or clofibrate group; a placebo identical in appearance to the alternate drug was also given. No patient weighed under 45.5 kg, so most halofenate patients received 1 g (2 tablets) once a day; a few patients weighed over 109 kg and received 1.5 g once a day. Clofibrate was administered in a dose of 1 g (2 capsules) twice each day. Tablet and capsule counts were performed. Seventeen patients (6 women, 11 men) received halofenate and 16 patients (8 women, 8 men) received clofibrate. Mean age was 45.4 (range of 35-65 years) and 47.1 (range of 39-60) years at the onset of treatment for those on halofenate and clofibrate, respectively. Of the 33 patients, 16 (7 halofenate, 9 clofibrate) had documented coronary artery heart disease, 7 (3 halofenate, 3 clofibrate) had peripheral vascular disease, 4 (3 halofenate, 1 clofibrate) had cerebral vascular disease and 10 (6 halofenate, 4 clofibrate) were symptomatically free of vascular disease. Three (1 halofenate and 2 clofibrate) of these 33 patients had vascular disease involving more than one region. All patients were informed about the nature of the study and signed a consent form before admitted to the study. They continued their usual diets during the study.
539
After 5 Lipid Clinic visits during an &week control period, drug therapy was immediately initiated and visits were then made at end of the second and fourth weeks and subsequently at 4-week intervals through the 48th week of treatment. Most patients had never received a lipid-lowering drug and none received treatment for at least 2 months prior to onset of the control period. On each visit patients were examined by a physician and serum lipids determined. Total serum cholesterol was determined by the procedure of Franey and Mador [ 151 and in all instances standards and patients’ and control serums were analyzed in duplicate. Reconstituted Lipitrol@ (Dade) and frozen serum obtained from the Center for Disease Control (CDC), Atlanta, Georgia, served as the control serums. The coefficients of variation for Lipitrol and CDC control serums were +3.7% and *3.9%, respectively, and the mean values were 296 mg/dl versus a stated value of 300 mg/dl for Lipitrol and 304 versus 296 for the CDC control. The manual method of Levy [ 161 was used for triglycerides. Standards (triolein), control serums, and serum from patients were analyzed together and in duplicate. The same control serums were used as for cholesterol analyses and the coefficients of variations were ?6.7% and +8.8%, respectively, for Lipitrol and CDC preparations. The mean value obtained for Lipitrol was 208 mg/dl versus the stated value of 218 mg/dl, and for the CDC control serum a mean of 226 mg/dl was obtained for a stated value of 247 mg/dl. Lipoprotein electrophoresis was carried out on paper in a human albumin containing buffer according to the method of Lees and Hatch [ 171; the classification of abnormalities was made according to the system of Fredrickson and Lees [18]. On a periodic basis, red and white blood cell counts and hematocrits were performed by a Coulter counter and sizer. A blood urea nitrogen and serum sugar were obtained by Technicon-Autoanalyzer techniques, serum uric acid spectrophotometrically at 292 nm using the uricase method (Sigma Chemical Corporation), creatine phosphokinase according to Rosalki [ 191, total serum protein by the biuret method, and albumin by electrophoresis and quantitative densitometry. Alkaline phosphatase activity was determined by the nitrophenol method [ 201, bilirubin according to Stoner and Weisberg [ 211, glutamic-pyruvate and glutamic-oxalacetate transaminases by spectrophotometry at 340 nm (Sigma Chemical Corporation). Electrocardiograms were obtained during the control period and after the 24th and 48th weeks of drug therapy. Eye examinations were generally performed on the same day as the ECG by members of the Department of Ophthalmology at the University of Kentucky Medical Center. Results Serum
lipids: halofenate
versus clofibrate
The 5 lipid determinations obtained per patient during the control period were averaged and these individual mean values covered a substantial range, 270-518 mg/dl for cholesterol and 77-435 mg/dl for triglycerides. In view of the wide variation in control period mean values between patients, treatment values were analyzed in terms of percent of the mean control values for indivi-
540
dual patients. The standard deviation associated with the mean control values was *6.4% and the standard error ?0.5%. The serum concentrations of cholesterol and triglyceride were obtained every 4 weeks, but for summary purposes treatment periods were separated into l-8, 9-24, 25-36, 37-48 weeks after start of drug treatment. With respect to the individual drugs and the individual serum lipid determinations, there were 2 overt groups of treated patients -those who responded to treatment and those who did not. Responders to treatment showed in at least 2 of the 4 treatment periods a mean concentration decrease at least 10% below the mean value for the control period. Except for one patient in the first period, no responder had an increase in lipid concentration above the control mean for any period during course of treatment. The demarcation provided by this criterion separated without overlapping the group of mean treatment values for patients responding to therapy from the mean values for patients not responding to therapy. Among responders, the mean cholesterol-lowering effect of halofenate was about one-half that of clofibrate (Table 1). For both drugs, the lowering effect in responders was generally sustained throughout the treatment period. The effect on serum triglyceride level for responders was comparable (approximately a 30% decrease) for both halofenate and clofibrate (Table 1). Again, the lowering effect when present was usually sustained over the treatment period. In most cases, if a patient did not respond within the first 8 weeks of treatment, a significant response did not occur later. Mean lipid concentrations during the control period and during the 2 periods of greatest lowering response are presented (Table 2) for individual patients responding to drug treatment. Porportions of responders Patients with Type II hyperlipoproteinemia TABLE MEAN Percent
can be further
divided into 1 of
1 CHANGES
IN SERUM
change from control
Drug
Response
LIPIDS
AFTER
START
Period (weeks of treatment)
a
9-24
l-8 Cholesterol Halofenate
Clofibrate
Triglycerides Halofenate
Clofibrate
OF TREATMENT
period mean f SE.
25-36
31-48
responders(S) nonresponders(7)
-11.0 -3.0
f 4.1 f 2.1
-13.6 -t2.5 -3.9 r 3.0
-11.9 + 2.8 +2.6 f 3.5
-12.2 f 3.5 +5.4 f 4.9
responders (9) nonresponders
(7)
-22.1 -3.3
f 1.9 ? 1.2
-24.8 t 3.6 -1.9 fi2.6
-24.2 + 2.8 -4.1 + 2.1
-28.0 f 3.0 -2.4 k 2.8
responders (8) nonresponders
(7)
-27.0 + 7.2 +fi.of 13.8
-33.7 + 2.3 +26.3 t 18.1
-30.1 + 4.5 +35.8 f 20.1
-15.0 +15.7
responders (14) nonresponders (2)
-28.5 i 3.2 -7 f 2
-33.7 k 4.3 -4 + 2
-28.6 It 5.8 -5 +12
-24.4 f 5.5 +1 + 29
a Comparisons between responders and nonresponders the nonparametric Mann-Whitney U-test.
are aI significant
+ 11.1 + 13.8
at the P < 0.01 or less level by
541
TABLE
2
CHANGES IN THE WITH TREATMENT
MEAN
Control
Halofenate period
period
Cholesterol 333 275 315 353 297 284 284 302 281 425
LIPID
treatment
CONCENTRATIONS
control
period
INDIVIDUAL
Clofibrate period
293 217 246 311 235 244 244 263 235 a 369 a
271 316 342 356 359 356 518 436 283
211 224 239 228 302 260 357 244 232
94 72 97 85 92 76 141 135 164 142 a
435 128 169 88 298 209 210 233 124 187 117 241 168 157
165 78 78 55 200 146 118 165 93 120 68 113 136 108
(mg/dl)
a Treatment
was stopped
during the first period.
as the mean for two periods
RESPONDERS
treatment
The treatment
period
value of other patients is presented
of greatest lowering.
3
RESPONDERS PROTEINEMIA
AND
Type of hyperlipoproteinemia
NONRESPONDERS
Males
FOLLOWING
TREATMENT
Females
Lowering a of serum cholesterol Yes
Halo feenote IIa IIb Total Clofibrate IIa IIb To
OF
(mg/dl)
Triglycerides 144 128 170 129 192 106 205 264 216 169
TABLE
SERUM
ta1
a Responders mean control
FOR
No
TYPE
II HYPERLIPO-
Lowering of serum triglycerides Yes
5 6
4 2
7 3
2 5
5 5
4 3
11
6
10
7
10
7
2 7
4 3
6 3
0 7
4 10
2 0
9
7
9
7
14
2
had to show a decrease in the respective serum lipid concentration value in two or more of the 4 treatment periods.
-
NO
at least 10% below
-
the
542
2 subgroups: IIa (elevated cholesterol only) and IIb (elevated triglyceride as well as cholesterol). Patients with mean triglyceride values above 150 mg/dl in the control period were classified as IIb. Patients representing each of the 2 subtypes were found in both treatment groups. On grouping according to subtype, responders and nonresponders to halofenate were found in both the IIa and IIb group, though a tendency for cholesterol lowering to occur more often in IIa patients and for triglyceride lowering to occur more often in IIb patients seems present (Table 3). This latter tendency was more pronounced in patients treated with clofibrate. The proportion of responders were comparable among halofenate and clofibrate patients. (In Tables 2 and 3, two patients in the halofenate group who had substantial decrease in serum cholesterol are included among the responders, though one dropped from the study after 3 treatment weeks (allergic rash) and the other after 12 weeks due to an accidental shooting). The analysis was carried a further step (Table 4) by examining various combinations of cholesterol-lowering and triglyceride-lowering effects according to treatment and subtype of lipoprotein disorder. With either drug, over 80% of patients showed a lowering at least 10% below control values in one or both of the serum lipids studied. The 5 patients with xanthelasma had considerable or nearly complete clearing of these lesions during the 48-week treatment period; this effect was not dependent on the presence of a serum lipid-lowering response to treatment. The large tendinous xanthomas of one patient regressed to approximately one-fourth pretreatment size during halofenate treatment, even though serum cholesterol remained in the 400-500 mg/dl range. (Also, this patient’s xanthelasma almost completely disappeared).
Body weight After 48 weeks of treatment
the mean increase
in body weight for patients
TABLE 4 RELATIONSHIP BETWEEN CHOLESTEROL-LOWERING AND TRIGLYCERIDE-LOWERING EFFECTS ACCORDING TO TREATMENT AND SUBTYPE OF LIPOPROTEIN DISORDER Drug
Halofenate Ha IIb
Combination of lowering effects Chol-TG
Chol-TG
Chol-TG
Chol-TG
Yes
Yes No (number of patients) 4 0 -
No
No
(23.5%)
(23.5%)
Yes 3 3 -
(35.3%) Clofibmte Ha IIb
4 3 (43.8%)
2 0 (12.5%)
Yes 2 2
0 7 (43.8%)
No 0 3 -
(17.6%)
0 0 (0%)
Each number in parenthese indicates, for a given treatment, the percent of patients have a particular combination of responses.
543 TABLE MEAN
5 INCREASE
% Over weight
IN BODY
in control
Period
Halofenate Clofibrate
WEIGHT
DURING
STUDY
period. of treatment
(weeks)
l-8
9-24
25-36
3748
0.9% 0.8%
2.2% 2.6%
4.1% 2.9%
4.7% 2.7%
on clofibrate was 2.7% above the mean weight of the control period; for halofenate the mean increase was 4.7% (Table 5). Two patients on clofibrate and one patient on halofenate lost 5% or more body weight. Five patients on clofibrate and 6 patients on halofenate showed a gain of 5% or more in body weight. In order to evaluate the possible effect of weight gain on the response to treatment, the proportion of responders among those with the weight gain of 5% or more were determined. Weight gain reduced the proportion of patients having a cholesterol-lowering response to either halofenate or clofibrate (Table 6); but the small number of patients in each group should be noted. In contrast, weight gain influenced the triglyceride-lowering effect of treatment slightly if at all. Serum uric acid
The lowering of serum uric acid by halofenate was prompt and sustained (Table 7); this effect was evident on the first visit 2 weeks after start of treatment. All but 3 of 15 halofenate-treated patients showed a decrease greater than 20%, and the mean decrease for these responders was 35-42%. Of the
TABLE
6
INFLUENCE Weight gain
Yes No
OF WEIGHT
GAIN
Treatment
and response
RESPONSE
b
Halofenate
Clofibrate
Cholesterol-lowering l/7 (14%) 6/8 (75%)
215 7/11
p < 0.10 c
N.S.
Triglyceride Yes No
a ON TREATMENT
(40%) (64%)
lowering
517 (71%) 5/8 (63%)
5/5 9/11
N.S.
N.S.
(100%) ( 82%)
a Patients who had a weight gain of 5% or greater during the 48-week treatment period were considered to have gained significant weight. b The numerator designates the number of patients who had a lipid-lowering response and the denominator presents the number of patients who gained weight (“yes”) or did not gain weight (“no”). ’ Using Yate’s Chi-Square correction for small numbers to assess these 2 X 2 tables; N.S. indicates nonsignificance at the 0.10 level.
544 TABLE
7
MEAN
LOWERING
Percent
change
OF SERUM
from control
URIC ACID
period,
Drug
mean f. S.E.
Period (weeks after start of treatment) 9-24
l-8
36-48
25-36
Halofenate (6.22 f 0.25 mg/dl) all patients (15) -21.0 responders only (12) -39.0
* 4.5 + 2.3
-28.2 -41.7
? 5.3 f 1.8
-21.5 -35.2
?r 9.5 2 3.2
Clofibrate (5.91 + 0.27 mg/dl) alI patients
k 2.3
-1.5
+ 2.3
-9.1
f 3.3
-8.0
-27.6 -36.6
+0.5
+ 5.8 + 3.1
k 6.1
Values in parenthesis following the drug are the mean f SE values of serum uric acid levels for the control period: other parentheses denote number of patients.
3 nonresponders, none had an elevated serum uric acid concentration and, in fact, one had a mean value of 3.0 mg/dl for the control period. Clofibrate did not produce an effect on uric acid comparable to that of halofenate. Although 3 patients in the halofenate group and 3 patients in the clofibrate group had slightly or moderately elevated levels of uric acid in their serums, no patient in either group had a history of symptomatic gout, nor did gouty symptoms develop in any patient during course of treatment. Monitoring of selected blood and serum components Analyses for a number of blood or serum components (Table 8) were used to monitor possible effects of treatment on various organ systems or on other metabolic processes. Besides effects on the serum lipids and serum uric acid, both described separately above, a few other appreciable effects were noted. Serum albumin concentration increased modestly (0.3-0.4 mg/dl) in addition to a slightly greater increase in serum total protein (0.5-0.6 mg/dl) which must reflect slightly increased (0.2 mg/dl) concentration of serum globulins as well. These effects on serum protein concentration fell in both halofenate- and clofibrate-treated patients. There was a modest increase in the mean serum creatine phosphokinase activity for halofenate-treated patients. Patients treated with clofibrate had a slight but significant decrease in mean serum alkaline phosphatase activity. Adverse reactions Only one adverse reaction necessitating cessation of treatment occurred during the study. This patient, a 25 year-old nurse who had a history of allergic reactions to propoxyphene, merthiolate and keflin, developed a generalized rash and urticaria after 2 weeks of halofenate treatment. The drug was immediately stopped; both rash and urticaria responded to symptomatic treatment. Acute situations developed during the treatment of 4 other patients. Bleeding from an unsuspected duodenal uicer developed after 4 weeks of halofenate treatment in a 49-year-old farmer with a history of peptic ulcer. The drug was discontinued for one month and then resumed without furt’ler difficulty for 92
(12.5-15.5.F: 14-17.M)
(38-46.F: 41--5O.M)
(4-12)
Hemoglobin (g/dl)
Hct(%)
WBC
ha1 cl0 ha1 cl0 ha1 cl0 ha1 cl0 ha1 cl0
(8-19)
(6-8.2)
(3.5-4.5)
(0.2-1.0)
(640.F; 7-54.M)
(0.7-2.7)
(O-35)
(O-28)
Semmureanitrogen (mg/dl)
Serum protein (g/d0
Serum albumin (g/dl)
Totalbilirubin (mg/dl)
CPK (mIU/ml)
Alk.phos. (BLB)
SGPT (Karmen)
SGOT (Karmen)
0.9 1.0
0.3 0.4
* *
0.1 0.1
6 5
33 35
0.9 1.1
0.3 0.4
f *
* 1 -t 1
14 18 + +
14 16
1 1
+1 5 1
9 12 i- 1 i 1
12 12
0.1 0.1 b
1.7 c 1.3 c
i' 5 *lo
0.9 0.04
0.43 + 0.39 + 40 49
0.1 0.1
0.1 0.1
4.0 * 4.0 f
7.4 + 7.6 *
1.1 0.9
2 3
f14 f36
14.2 + 14.9 +
103 101
265 245
0.4 0.5
0.7 1.1
0.3 0.4
+ +
0.08 0.06
0.44 f 0.39 +
1 1
t * * *
11 13 14 18
1 1
0.1 0.1 b
2.0 + 1.2 f
?; 5a + 6
0.1 0.1
4.2 f 4.2 *
49 44
0.2 0.1 a
7.5 f 7.8 +
0.7 0.8
2 3
*13 f29
14.0 + 15.3 *
102 105
252 338
6.7 f 8.5 ?
45.2 f 42.6 +
15.2 * 14.2 f
9-24
treatment (weeks)(means
6.9 ? 0.5 8.3 f 0.6
45.1 * 43.6 +
15.2 + 14.6 +
l-8
Period of
2.1 * 0.2 1.9 t 0.1
+ +
0.06 0.01
0.58 + 0.47 c
4.0 + 0.1 4.0 -r 0.1
7.4 + 7.5 *
0.8 0.7
4 2
+ 22 f 27
13.4 * 15.4 *
107 105
272 320
6.9 -t 0.5 8.5 + 0.5
45.2 + 44.3 *
15.2 + 14.9 +
Control
PERIODS
f +
t + + i
11 11 15 19
1.8 + 1.2 +
47 48
0.36 + 0.26 f
4.1 t 4.4 ?
2 1
2 2
0.2 0.1 b
8 9
0.08b 0.05b
0.1 0.1 b
7.5 ?; 0.1 7.9 f 0.1 b
0.9 1.1
?ll ? 4 15.3 f 15.2 +
114 109
+ +
+ *
8a 8
O.Olb O.lb
O.la 0.1 b
0.1 b 0.1 b
0.6 1.0
7 3
14 18
10 12
* +
f +
1 1
1 1
2.3 -t 0.3 1.2 f 0.1 b
55 42
0.22 + 0.26 +
4.3 f 4.4 f
7.9 + 8.1 f
14.7 f 14.9 *
110 100
0.5 0.6
0.9 1.1
0.3 0.4
*19 * 23
237 307
*17 +19
246 305
44.9 f 43.5 f 6.9 f 8.4 +
1.0 1.1
43.6 + 42.9 +
15.1 f 14.6 i
37-48
6.3 i- 0.4 8.2 ? 0.5
0.4 0.4
14.9 f 14.5 f
25-36
* SE)
aP < 0.05:b P < 0.01 for comparisons between treatment and control values.F = range for normalfemales,and M = range for normal.males.
ha1 cl0
ha1 cl0
ha1 Cl0
ha1 cl0
(70-110)
Serum glucose (fasting.mg/dl)
ha1 Cl0
(150450)
ha1 cl0
ha1 Cl0
ha1 Cl0
Drug
Platelets (thousands/mm3)
(thousands/mm3)
Normal (range)
component (units)
LEVELSOFVARIOUSBLOODANDSERUMCOMPONENTSDURINGCONTROLANDTR~ATMENT
TABEL8
a
546
weeks (i.e. 44 weeks remaining in the initial study plus 48 weeks in an extension period). After 40 weeks of halofenate treatment a 53-year-old housewife, who had taken aspirin preparations chronically for an arthritic condition, developed symptoms of gastric ulcer which was verified by X-ray, gastroscopy and biopsy. The ulcer resolved on removal of aspirin and on addition of antacid therapy. Halofenate was stopped for one week, then resumed for the final 8 weeks and there was nearly complete healing of the ulcer during that time as observed by endoscopy and X-ray. Two patients receiving clofibrate had intercurrent illnesses during course of their treatment. A 64-year-old farmer with a previous history of stroke suffered a myocardal infarction after 40 weeks of therapy; clofibrate was discontinued for 5 days immediately after the occurrence and, was then resumed uneventfully for the remaining 8 weeks of treatment. After 36 weeks of clofibrate treatment a 44-year-old mother of 3 was hospitalized for depression. Symptoms of cholecystitis developed and a gall bladder containing multiple stones of mixed composition was removed. Therapy was resumed after a ‘I-week hiatus and the subsequent 12 weeks went without difficulty. No deaths occurred during the course of this study. ‘Extended period of halofenate treatment The treatment period for 10 patients receiving halofenate was extended from 48 weeks to 96 weeks. No adverse reactions nor untoward changes in laboratory values occurred during the extended period. The response of the serum lipids and uric acid during the extension remained in accord with that in the initial treatment period. During the extension period, one patient (a husky, 34-year-old male suffering from incapacitating coronary artery disease with severe angina) demonstrated clearly the dependence of the serum lipid response on halofenate dose. (In addition to halofenate, he was taking nitroglycerine, long-acting nitrates and propranolol.) The initial dose based on a body weight of 96 kg was 1 g daily and in the first 24 weeks serum cholesterol decreased 15% (from 310 to 263 mg/dl) and triglyceride serum decreased 32% (from 205 to 139 mg/dl). Subsequently, his weight increased to 106.5 kg and the lipids returned to control period values. During the first 28 weeks of the extension period he attained 110 kg and halofenate dose was increased to 1.5 g daily. This higher dose again decreased the serum cholesterol by 16% and triglyceride by 22% below the control period values. Discussion The lowering of serum uric acid levels by halofenate is well-known [l-6,14, 221. Halofenate increases the urinary excretion of uric acid but does not increase urinary oxypurines (as does allopurinol). These findings suggest halofenate interference with tubular reabsorption of uric acid [ 1,2,5]. Progressive decrease in serum total bilirubin with halofenate and with clofibrate treatment is in accord with previous observations [3,9,23] and this is accredited in part to the displacement of bilirubin from plasma binding proteins, albumin especially. Alkaline phosphatase activity in clofibrate-treated patients decreased moderately and this has been observed previously [23]. Though halofenate-treated
patients in our study did not show a mean lowering of serum alkaline phosphatase activity, such an effect was noted in another study [ 31. Hutchison and Wilkinson [3] observed a modest decrease in serum albumin concentration (but not total protein concentration) in their series of halofenate-treated patients, but our patients had a mean increase in concentration of both albumin and total protein. Different analytical methods may account for this discrepancy. Clofibrate-treated patients also showed a similar small increase in serum albumin and total protein concentration. Appreciable changes in transaminases, serum urea and glucose, red cell count and size, hemoglobin concentration and white blood cell count did not occur in either clofibrate- or halofenate-treated patients in the course of our study. The moderate increase in mean creatine phosphokinase activity for halofenate patients was due to sporadic increases in a few patients and we could not relate these episodes to any change in clinical signs or symptoms. Thus laboratory monitoring revealed few, if any, indication of adverse treatment effects by either halofenate or clofibrate in our study patients. The only pronounced and unequivocal adverse effect for halofenate was a generalized rash in a patient who previously had an allergic reaction to other drugs and who several years before had a short term of clofibrate therapy. Although the chemical structure of halofenate is substantially different from that of clofibrate, both clofibrate and halofenate are tightly bound to serum proteins and may be effective haptens in certain patients. Thus, it seems wise to use halofenate circumspectly and with forewarning of possible allergic response in any patient who has previously been treated with clofibrate and who has an allergic history. The development of a bleeding duodenal ulcer in a man with a history of duodenal ulcer and of a gastric ulcer in a woman who had been taking aspirin for an arthritic condition may or may not represent a complication of halofenate treatment, only further controlled observations will provide an answer. In this regard, however, gastrointestinal complaints were uncommon in our halofenate treated patients - even those treated for 96 weeks; the drug was tolerated well. In the Type III, IV and V patients investigated in other halofenate studies [l-3,13,14] increased serum triglyceride concentration was the most prominent aspect of the lipid abnormality. There is agreement in those studies that halofenate lowered mean serum triglyceride levels by 16-20% or more. Our group of patients differed from those in the previous studies by being limited to patients with Type II hyperlipoproteinemia and in this condition the primary serum lipid abnormality is considered to be the hypercholesterolemia. Despite the apparent primacy of cholesterol elevation, substantial triglyceride lowering was noted in both halofenate- and clofibrate-treated patients. Even patients with normal triglyceride levels (Type IIa) often experienced triglyceride lowering. The magnitude of lowering was comparable among responders for the two drugs, but the chances of a patient responding favorably seemed somewhat greater for clofibrate than for halofenate. Among nonresponders, halofenate treatment in contrast to clofibrate resulted in an elevation of triglyceride levels of some patients such that a substantial mean elevation for nonresponders occurred. Presently we have no explanation for this apparent difference in effect of the two drugs.
548
In the case of cholesterol, the data obtained from patients with Type II hyperlipoproteinemia afforded some similarities and contrasts with data obtained by others from patients having other types of hyperlipoproteinemias. In accord with an earlier study [ 91, halofenate in general is not as proficient as clofibrate in lowering serum cholesterol levels. However, both halofenate and clofibrate appreciably lowered (by 10% or more) cholesterol levels in over one-half of treated patients; a substantially higher proportion than expected from the previous studies on patients with other types of hyperlipoproteinemias. Both halofenate and clofibrate elicited a cholesterol-lowering effect in nearly all Type IIa patients but in only about one third of Type IIb patients. It is now demonstrably clear that the types of hyperlipoproteinemias categorized by the criteria of the Fredrickson and Lees’ system do not constitute genetically homogeneous groups of patients [ 241. Both Type IIa and IIb phenotypes can occur in the genetic disorders of monogenic familial hypercholesterolemia, monogenic combined hyperlipidemia, and polygenic hypercholesterolemia which were recognized in the study of Goldstein et al. [ 251. Since the gene product responsible for the aberrancy in these disorders has not been identified (except for deficiency or defectiveness in cellular receptors for low-density lipoproteins [26]), the extent of metabolic heterogeneity within even the monogenic disorders can not be determined. Given even a limited amount of heterogeneity within a metabolic disorder, differing therapeutic responses to a given drug are not surprising. Until more is known about the biochemical basis of the hyperlipidemic disorders, little can be definitely stated as to why a drug such as halofenate or clofibrate brings about or fails to bring about a lowering of serum cholesterol or triglyceride in an individual patient. Acknowledgements The assistance of Norma Slone, Evangeline Lear, Judy Campbell, Helen Lipscomb, Catherine Fine and Bernetta Mason is acknowledged with pleasure. We are grateful to Dr. Grant Somes for statistical assistance. References 1 2 3 4 5 6 7 8 8a
Jam. A., Ryan, J.R., Hague, D. and McMahon. F.G., The effect of MK-185 on some aspects of uric acid metabolism, CIin. Pharmacol. Therap. 11 (1970) 551. Morgan, J.P., Bianchine, J.R., Hsu, T.H. and Margolis. S.. Hypolipidemic, uricosuric, and thyroxinedisplacing effects of MK-185 (halofenate), CIin. Pharmacol. Therap., 12 (1971) 517. Hutchison. J.C. and Wilkinson, W.H., The uricosuric action of halofenate (MK-185) in patients with hyperuricemia or uncomplicated primary gout and hyperlipidemia, Atherosclerosis, 18 (1973) 353. Aronow. W.S., Harding, P.R., Khursheed, M., Vangrow, J.S. and Papageorge’s, N.P., Effect of halofenate on serum uric acid, CIin. Pharmacol. Therap., 14 (1973) 371. Ravenscroft, P.J., Sands, J.M. and Emmerson, B.T., Studies of the uricosuric action of the hypoIipidemic drug halofenate. CIin. Pharmacol. Therap., 14 (1973) 547. Wolfram, G., Keller. C., KiIami, S. and Zollmar, First cIinicaI experiences in therapy of hyperuricemia and hyperhpidemia, Verhandl. Dtsch. Ges. Inn. Med., 79 (1973) 2191. Ryan. J.R., Jain. A., Maha, G.E. and McMahon, F.G., The effects of MK-185 on thyroid function, Clin. Pharmacol. Therap.. 12 (1971) 464. Gattereau, A., Davingnon, J., Verdy, M. and Lewis, W., Halofenate versus clofibrate in the management of true diabetes insipidus, Canad. Med. Ass. J. 110 (1974) 1275. Jain, A., Ryan, J.R. and McMahon, F.G., Potentiation of hypoglycemic effect of sulfonylureas by halofenate, New En%. J. Med., 293 (1975) 1283.
549
9
Jepson,
E.M.,
Small,
new drug MK-185 10 11
12 13 14
15 16 17 18 19 20 21 22
23 24 25 26
Aronow,
W.S.,
E.,
Grayson,
with clofibrate
Harding,
P.R.,
M.F.,
Gillian, B. and Billimoria.
in the treatment
Khursheed,
of hyperlipidemias,
M., Vangrow.
J.S.
J.D.,
A comparative
Atherosclerosis,
Papageorge’s,
study 16 (1972)
of a 9.
N.P. and Mays, J., Effect
of halofenate on serum lipids, Clin. Pharmacol. Therap.. 14 (1973) 358. Aronow, W.S.. Vangrow. J.S.. Nelson, W.H., Pagano, J., Papageorge’s, N.P.. Khursheed, M., Harding, P.R. and Khemka, M.. Halofenate, serum lipids, and exercise performance in coronary heart disease, Clin. Pharmacol. Therap.. 15 (1974) 67. Hutchison, J.C. and Wilkinson, W.H., Halofenate -Effectiveness of two dosage forms and two dose schedules. Atherosclerosis, 19 (1974) 417. Sirtori, C., Hurwitz, A., Khalid. S. and Azarnoff, D.L., Clinical evaluation of MK-185 -A new hypolipidemic drug, Lipids, 7 (1972) 96. Lisch, H.J., Patsch, J., Sailer, S. and Braunsteiner, H., Comparison of the effects of halofenate pa(MK-185) and clofibrate on plasma lipid and uric acid concentration in hyperlipoproteinemic tients, Atherosclerosis. 21 (1975) 391. Franey, R.J. and Amador, E., Serum cholesterol measurement based on ethanol extraction and ferric chloride-sulfuric acid, Clin. Chim. Acta, 21 (1968) 255. Levy. A.L., Measurement of triglycerides using nonane extraction and calorimetry, Ann. Clin. Lab. Sci., 2 (1972) 474. Lees, R.S. and Hatch, F.T., Sharper separation of lipoprotein species by paper electropboresis in albumin-containing buffer, J. Lab. Clin. Med., 16 (1963) 518. Fredrickson. D.S. and Lees, R.S., A system for phenotyping hyperlipoproteinemia, Circulation. 31 (1965) 321. Rosaliki, S.B.. An improved procedure for serum creatine phosphokinase determination, J. Lab. Clin. Med., 69 (1967) 696. Bessey, O.A.. Lowry, O.H. and Brock. M.J.. A method for the rapid determination of alkaline phosphatase with five cubic millimeters of serum, J. Biol. Chem., 164 (1947) 321. Stoner, R.A. and Weisberg. H.F.. Ultramicro method for serum bilirubin by diazo blue reaction, Clin. Chem., 3 (1957) 22. Aronow. W.S., Vangrow, J.S., Nelson, W.H., Pagano, J.. Papageorge’s, N.P., Khursheed. M., Harding, P.R. and Mahaveer, K., Halofenate - An effective hypolipemiaand hipouricemia-inducing drug, Curr. Therap. Res., 15 (1973) 902. Coronary Drug Project Research Group, Clofibrate and niacin in coronary heart disease, J. Amer. Med. Ass., 231 (1975) 360. Lees, R.S., A progress report on lipoprotein phenotyping, J. Lab. Clin. Med., 82 (1973) 529. Goldstein, J.L.. Schrott, H.G., Hazard. W.R., Bierman. E.L. and Motulsky, A.G.. Hyperlipidemia in coronary heart disease, J. Clin. Invest., 52 (1973) 1544. Goldstein, J.L., Dana, S.E., Brunschede, G.Y. and Brown, M.S., Genetic heterogeneity in familial hypercholesterolemia - Evidence for two different mutations affecting functions of low-density lipoprotein receptor, Proc. Nat. Acad. Sci. (Wash.), 72 (1975) 1092.
