1260

Efficacy and Safet of Perhexiline Maleate in

Refractory AnginaA Double-Blind Placebo-Controlled Clinical Trial of a Novel

Antianginal Agent

Patricia L. Cole, MD, Andrew D. Beamer, MD, Noreen McGowan, RN,

Catherine 0. Cantillon, RN, Kathleen Benfell, RPh, Ralph A. Kelly, MD,

L. Howard Hartley, MD, Thomas W. Smith, MD, and Elliott M. Antman. MD

Despite large gains in the medical and surgical treatment of angina pectoris in the past twodecades, many patients are refractory to conventional medical therapy and are unsuitable fora first or, more commonly, repeat coronary revascularization procedure. We evaluated theefficacy of perhexiline maleate, a drug with an antianginal mechanism of action in humans thatis as yet unknown, by using a randomized double-blind placebo-controlled crossover design in17 patients with refractory angina who continued to receive maximal antianginal therapy,typically including nitrates, a ,B-blocker, and a calcium channel antagonist. In view ofperhexiline's potential for hepatic and neurological toxicity, plasma drug levels were monitoredand maintained in the 150-600 ng/ml range. Sixty-three percent of patients were judgedperhexiline responders by objective exercise testing criteria, as compared with 18% of patientson placebo (p<0.05). By blinded review of subjective measures of anginal frequency andseverity, 65% of patients noted an improvement while on perhexiline, whereas no patientidentified the placebo phase with improvement. Side effects observed in 29% of patients wereminor and related to transient elevations of blood levels of more than 600 ng/ml; no patientsuffered hemodynamic or cardiac conduction abnormalities attributable to perhexiline. Withattention to the pharmacokinetics of perhexiline's elimination in individual patients, this novelantianginal agent seems to be safe and effective and deserves further evaluation in patientsalready receiving maximal antianginal therapy who are not candidates for revascularizationprocedures. (Circulation 1990;81:1260-1270)

D uring the last 20 years, we have witnessed adramatic increase in the range of therapeu-tic options, both medical and surgical, for

the treatment of angina pectoris. The size of thatsubgroup of patients who remain refractory to con-ventional therapeutics progressively declined as eachnew class of antianginal drugs or surgical approachwas introduced. Nevertheless, even now in the 1990s,most clinicians can identify in their practice patientswho are not surgery or angioplasty candidates andwhose frequency of anginal episodes is poorly con-

From the Cardiovascular Division, Department of Medicine,Brigham and Women's Hospital and Harvard Medical School,Boston, Massachusetts.

Supported in part by NRSA Training grant 1T32-HL07604 ofthe NHLBI and an American Heart Association Clinician-Scientist award to R.A.K.Address for correspondence: Elliott M. Antman, MD, Cardio-

vascular Division, Brigham and Women's Hospital, 75 FrancisStreet, Boston, MA 02115.

Received July 7, 1989; revision accepted December 6, 1989.

trolled on maximal medical therapy. The reemer-gence of anginal symptoms in previously revascular-ized coronary artery bypass patients who are notcandidates for a second or third operation suggeststhat this subgroup of patients will not decline but,rather, might increase in the future."Maximal medical management" for angina pecto-

ris in contemporary practice consists of the judicioususe of nitrates, ,3-adrenergic receptor blockers, andcalcium channel blockers, in addition to an antiplate-let agent (usually aspirin) and, with a view to the longterm, a lipid-lowering regimen. For patients withrefractory symptoms or patients intolerant of one ofthese major classes of drugs, the therapeutic options,particularly in the United States, are limited. Ironi-cally, a drug introduced in Europe nearly 20 yearsago might provide a safe and effective adjunct to themodern treatment of angina.1

Perhexiline maleate gained a reputation forefficacy in the medical management of angina in the

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Cole et al Perhexiline Maleate in Refractory Angina 1261

1970s and seemed to have no significant effect onmyocardial contractility or on hemodynamics.2-11 Itwas also found, however, to cause infrequent butunpredictable serious hepatic and neurologicaltoxicity.10-14 Despite its effectiveness, use of the drughad become limited in the early 1980s, in partbecause of its toxicity but also because of marketingdecisions by its manufacturers. Originally labeled as acalcium antagonist although the experimental evi-dence was far from clear that this was its clinicallyrelevant mechanism of action in humans, the incen-tive for marketing perhexiline declined as less toxiccalcium channel blockers were developed.Within the past decade, however, it has become

evident that toxicity is directly related to perhexilineblood levels and that the drug has a saturable rate ofhepatic metabolism that is genetically determined. Asubset of patients demonstrates saturable clearanceof the drug at standard therapeutic doses, in somecases leading to a dramatically lengthened elimina-tion half-life.213'15 With the advent of straightforwardand reliable methods for quantifying blood levels ofperhexiline,16-18 individualization of dosing has led toa pronounced decrease in the incidence of seriousside effects.15"17

Therefore, to investigate rigorously the safety andefficacy of perhexiline, a prospective randomizeddouble-blind placebo-controlled crossover trial wasinitiated. Its purposes were twofold, as follows: 1) tostudy the efficacy of perhexiline as an antianginaltherapy when added to the existing medical regimenin patients whose symptoms were refractory to othertherapeutic modalities, and 2) to determine whetherdrug-related side effects could be minimized andantianginal efficacy maintained when the dosage ofperhexiline was adjusted to maintain plasma levels inthe proposed therapeutic range of 150-600 ng/ml.15

MethodsStudy Design

This was a double-blind randomized crossover trialin which patients served as their own controls. Base-line measurements were made as described herein.Patients were then randomly assigned to receiveeither placebo or perhexiline for an initial 3-monthperiod. At the conclusion of 3 months, end-pointmeasurements were made, and the patient wascrossed over to receive the alternative medication(perhexiline or placebo). One physician (E.M.A.)remained unblinded and used the results of theperhexiline assay to adjust drug dosage to maintainperhexiline blood levels in the therapeutic range(150-600 ng/ml) suggested by Horowitz et al.15When it was necessary to adjust the dose of activedrug, this physician "informed" the patient of thechange through one of the blinded physicians orstudy nurses although placebo dosages were carefullyadjusted in a similar manner so as to maintain thedouble-blind nature of drug administration.

Entry CriteriaThe study population consisted of patients who

were referred by cardiologists from the medical ser-vice, coronary care unit, or outpatient cardiologypractice at Brigham and Women's Hospital and whowere suffering from refractory angina pectoris. Allpatients were required to have diagnostic evidence ofsignificant coronary artery disease at coronary arte-riography, a documented history of acute myocardialinfarction (by electrocardiographic or enzyme crite-ria, or both), or both the diagnostic evidence and thedocumented history. Additionally, all patients hadrecurrent episodes of chest discomfort typical ofmyocardial ischemia (retrosternal or upper extremitydiscomfort, or both, relieved by sublingual nitrates)despite medical therapy with a combination of /3-adrenergic receptor blockers, nitrates, and calciumchannel blockers that, in the opinion of the treatingcardiologist, had been advanced to the maximum-tolerated doses. Maximum-tolerated therapy wasdefined as the highest doses of /3-blockers, nitrates,and calcium antagonists used in combination that didnot produce symptomatic bradycardia, hypotension,or congestive heart failure. Patients also had coronaryanatomy that rendered them unsuitable candidates forinitial or repeat coronary artery bypass graft surgery orpercutaneous transluminal coronary angioplasty.

Exclusion CriteriaPatients were excluded if one or more of the

following conditions were present: suboptimal medi-cal therapy; Q wave myocardial infarction during thepreceding 3 months; severe left ventricular failure(New York Heart Association class III or IV); hemo-dynamically significant valvular disease, hypotension(systolic blood pressure, <85 mm Hg), or hyperten-sion (systolic or diastolic pressures, >210 and >120mm Hg, respectively); evidence of severe renal (cre-atinine, >2.5 mg/dl) or hepatic (SGOT, >50 units/land bilirubin, >2.0 mg/dl) dysfunction, brittleinsulin-dependent diabetes mellitus; or knownperipheral neuropathy.

All patients were enrolled with the approval oftheir primary physician, and all patients signed aninformed consent statement using the guidelines setforth by the hospital's Committee for the Protectionof Human Subjects.

Study ProtocolBefore randomization, all patients underwent a

detailed history and physical examination by one ofthe investigators (P.L.C. or A.D.B.), including ascreening neurological exam. Baseline laboratorymeasurements included serum electrolytes, bloodurea nitrogen, creatinine, glucose, liver functiontests, uric acid, and a complete blood count. Addi-tionally, a chest radiograph and electrocardiogramwere performed.

All patients had previously undergone exercisetests and were familiar with the standard procedures

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

1262 Circulation Vol 81, No 4, April 1990

TABLE 1. Characteristics of Study Group

Age MIs CAD EF CABG Grafts PTCA 13-blk TNG CCBA 1 CCBA 2 DigPatient (yr) (n) (vessels) (%) (n ops) (n) (n) (mg/day) (mg/day) (mg/day) (mg/day) (mg/day)

1 65 2 3 47 1 2 0 ... ISDN 320 DZ 240 ... 0.125/0.252 62 1 3 50 2 2 0 Aten 50 ISDN 160 DZ 240 NIF 40 0.1253 71 1 3 49 2 2 0 Prop 160 ISDN 160 DZ 240 ...

4 67 2 3 33 1 3 0 Prop 160 ISDN 80 DZ 2405 59 0 1 56 1 1 0 ... DZ 90 NIF 160 ...

6 74 1 3 70 1 1 0 Aten 75 ISDN 320 DZ 2407 61 2 3 69 1 2 0 Aten 75 ISDN 240 DZ 908 57 1 1 63 0 0 0 Metop 50 ISDN 160 ... VER 2409 76 2 3 ... 0 0 1 Prop 320 ISDN 80 ... NIF 80 ...

10 68 1 3 68 1 3 0 Metop 200 ISDN 160 ... NIF 160 ...

11 81 1 3 53 1 3 0 Prop 160 ISDN 80 DZ 240 ... ...

12 72 1 3 36 1 3 0 Aten 50 ISDN 160 ... NIF 80 ...

13 43 1 2 34 1 2 0 ... ISDN 80 ... NIF 40 0.12514 65 0 1 62 0 0 0 Metop 200 ISDN 160 DZ 360 ... ...

15 54 1 3 57 1 3 0 Aten 50 ... ... NIF 80 ...

16 76 1 ... ... 0 0 0 Aten 50 ISDN 240 DZ 60 VER 240 0.125

17 60 2 3 35 2 3 0 Metop 50 ISDN 90 DZ 240Mean 65.4 1.18 2.56 52.1 0.9 1.8 0.06 .......... ... ...

SD 9.5 0.6 0.8 13.1 0.7 1.2 0.2 ... ...

MI, myocardial infarction; CAD, coronary artery disease; EF, left ventricular ejection fraction; CABG, coronary artery bypass graft; PTCA,percutaneous transluminal coronary angioplasty; /3-blk, ,-blocker; TNG; CCBA, calcium channel blocking agent; Dig, digoxin; A/C,anticoagulants; ops, operations; ISDN, isosorbide dinitrate; DZ, diltiazem; Warf, warfarin; Thy, thyroid; Aten, atenolol; NIF, nifedipine; Fur,furosemide; Dipy, dipyridamole; Prop, propranolol; Tr/HCTZ, triamterene/hydrochlorothiazide; ASA, aspirin; Mdopa, methyldopa; Chlor,chlorpropamide; Nico, nicotinic acid; Metop, metoprolol; VER, verapamil; Tolb, tolbutamide; Capt, captopril; Cholest, cholestyramine.

in the exercise laboratory used during the trial. Allpatients underwent maximum symptom-limitedgraded exercise testing using a bicycle ergometer,except in one case where an arm ergometer was used.The initial exercise load for all tests was 5 W, and theload was increased by 10-20 W every 5 minutes. Foreach patient, the identical protocol was used duringthe baseline test, at the end of the 3-monthperhexiline-phase exercise test, and at the end of the3-month placebo-phase exercise test.

In addition to the above information, patients wereasked to complete a specific-activity scale question-naire that helped to quantify exercise capacity andactivity level.1819 Patients were also asked to keep adetailed diary of anginal episodes, including fre-quency and duration of angina, precipitating events,and consumption of sublingual nitroglycerin.

Drug AdministrationAfter baseline data were collected, patients were

randomly assigned by a research pharmacist (K.B.) toreceive either perhexiline or placebo. The initial doseof perhexiline was 100 mg p.o. daily for 3 days and100 mg p.o. twice daily thereafter. Placebo tabletsidentical to perhexiline were supplied by MerrellDow Pharmaceuticals Inc. (Cincinnati, Ohio), andthe placebo was dosed in a manner similar to activedrug. Beginning at 1 week and then at approximately2-4-week intervals, "study drug" blood levels weremeasured and dosage adjusted accordingly by theunblinded physician.

At the end of the initial 3-month period, a completephysical examination, including neurological evalua-tion, electrocardiogram, and the laboratory measure-ments previously noted, was repeated. Additionally,each patient underwent a repeat interview and assess-ment of subjective response, performed a repeat exer-cise tolerance test, and had a study drug blood leveldrawn. The groups were then crossed over to the othermedication, dosing was initiated in a manner similar tothe initial 3-month period, and serial blood leveldeterminations were obtained. At the end of thissecond 3-month block, the physical examination, elec-trocardiogram, and laboratory measurements wererepeated, and patients again underwent an assessmentof subjective response, an exercise tolerance test, anda final drug level determination.

AssayPerhexiline maleate levels in plasma or serum were

measured using the technique described by Horowitzet al16 with minor modifications. Plasma samplesspiked with hexadiline HCI as an internal standardwere extracted in hexane, dried, reconstituted in asodium bicarbonate buffer, and derivatized with dan-syl chloride exactly as previously described.16 Chro-matography was performed in 3 -,m reverse-phaseRadial-Pak cartridges (Waters ChromatographyDiv., Millipore Corp., Milford, Massachusetts) usingisocratic (100% methanol) conditions, with detectionby a fluorescence detector using 360- and 510-nmexcitation and emission filters. Perhexiline maleate

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Cole et al Perhexiline Maleate in Refractory Angina 1263

TABLE 1. Continued

Diuretics Vasodil(dose/day) (dose 3/day) A/C Other

Fur 80 ... Warf Thy USPFur 120 ... Dipy InsulinTr/HCTZ 2 Dipy, ASA ...

Fur 40 ... Dipy, ASA Mdopa 500

Fur 160 Dipy, ASA Chlor 500... ... ASA Insulin, Nico... ... ... Nico

Fur 80 Dipy, ASA AllopurinolFur 60 Dipy, ASA Tolb

ASA ...

Tr/HCTZ 1 Captl2.5 ... ...

... ... ... Indo

... ... ASA CholestDipy, ASA ...

... Capt37.5 ASA Insulin

. . . .

.

standards and hexadiline HCl were donated by Mer-rell Dow Pharmaceuticals Inc. The limit of detectionwas 1 ng perhexiline/ml plasma. The coefficient ofvariation of perhexiline 250 ng/ml added to blankplasma samples and analyzed over a 1-month periodwas 9.4% (n = 12), and that of replicate samplesperformed on the same day was 1.7% (n=5). The useof 3 -,m radial compression cartridges allowed excel-lent resolution with rapid separation (12 minutes) atlow column pressures.

Criteria for Antianginal EfficacyDefinitions of drug efficacy were established

before initiation of the trial and included assessmentof both objective and subjective responses. Objectiveevidence of drug efficacy was defined as the fulfill-ment of one or more of the following criteria ascompared with baseline exercise testing: 1) theabsence of angina at an equivalent workload; 2) theabsence of angina at an equivalent exercise duration;3) an increase of more than 20% in peak exerciseworkload; or 4) increase of more than 20% in peakexercise duration. If a patient met any of thesecriteria, he or she was classified as a responder.Categories 3 and 4 were deemed necessary criteria incase a patient did not have angina during baselineexercise testing.

Subjective criteria for efficacy were established bya review of angina diaries, nitroglycerin consumptionassessments, and a specific-activity scale by both thepatient and a study physician (P.L.C. or A.D.B.),each of whom remained blinded to the assignedsequence of therapies. A determination was made

regarding which of the two 3-month periods, ifeither, resulted in a significant reduction in thefrequency or severity, or frequency and severity, ofangina and requirement for nitroglycerin. Patientswere classified as responders based on subjectivecriteria if the blinded review identified a 3-monthperiod that was clearly superior to the interval on thealternative medication.18'19

Statistical AnalysesStatistical analyses were performed to compare the

proportion of patients classified as responders basedon the pretrial definitions of objective or subjectivecriteria for efficacy as noted previously during thevarious treatment phases. Because the drugresponses in individual patients represented paireddata measured on a nominal scale, the proportion ofresponders during the perhexiline phase and duringthe placebo phase were entered into a 2x 2 contin-gency table and analyzed using the McNemar testwith Yates' correction for continuity.20 The 95%confidence intervals (95% CI) for the proportion ofpatients responding during each treatment phasewere determined from standard statistical tables ofconfidence limits of the binomial distribution.21Group results for continuous variables are presentedas mean+1 SD. Comparisons between group meanswere performed with a two-tailed paired t test.Because of the crossover nature of the trial design, ananalysis of variance (ANOVA) with repeated mea-sures was performed to examine the effect of drugadministration on exercise performance. Results ofall analyses were considered statistically significantwhen p values were less than 0.05.

ResultsClinical Characteristics of Study GroupA total of 19 patients (13 men and six women) were

enrolled during an 18-month period (Table 1). Twopatients initially enrolled did not complete the study,one because of an intercurrent myocardial infarction,and one because of a fractured leg (while on placebo)that precluded exercise testing. Seventeen patientscompleted the study and form the basis of this report.They had a mean age of 65±9.4 years (range, 43-81years), and they had suffered a mean of 1.19+0.7myocardial infarctions before enrollment. Sixteen ofthese 17 patients had undergone coronary arteriogra-phy, and 13 patients had undergone previous coronaryartery bypass surgery (three of them twice). Allpatients were on combination antianginal therapy; 14patients were on p-blockers, 15 were on long-actingnitrates, and 14 were on calcium channel blockers(three patients were on two calcium channel block-ers). Other medications included digoxin (fourpatients); diuretics (eight patients); vasodilators (twopatients); aspirin, dipyridamole, or both (11 patients);and insulin (three patients). Three patients werereceiving lipid-lowering agents.

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

1264 Circulation Vol 81, No 4, April 1990

Objective Criteria for Drug Efficacy (ExerciseTolerance Testing)

All 17 patients completed each of the three exer-

cise tests. In the case of patient 12, the baseline test

was terminated after 540 seconds because of the

development of progressive hypotension and fatigue.This was judged to be a result of having taken his

maintenance antianginal medications within 30 min-

utes of exercise. During the perhexiline and placebo

phases, peak exercise durations were 1,080 and 1,020

seconds, respectively, when the same antianginalmedications were taken more than 2 hours before

exercise testing. Because of the noncomparable clin-

ical circumstances present during the three tests, this

patient's exercise data were excluded from analysis,

although the results are reported in Table 2.

During baseline exercise testing, 10 of 16 patients

experienced angina at an exercise load of 29±t28 W(5-90 W), which developed at 520±312 seconds

(range, 180-1,020 seconds) of exercise. The mean

peak exercise load achieved by the group was 38±35

W at a peak exercise duration of 662±t386 seconds.

The six patients who did not experience angina in the

baseline exercise test all had their test stoppedbecause of fatigue.

During the exercise test performed at the conclu-

sion of the perhexiline phase of treatment, seven of

the 10 patients who experienced angina during the

baseline exercise test were able to exercise to at least

the same exercise load, duration, or both as duringthe baseline test without experiencing angina. As

compared with the baseline exercise test, seven of 16

patients achieved at least a 20% increase in peakexercise duration. Based on the pretrial definition of

a positive exercise response, 10 of 16 patients (63%;95% Cl, 35-82%) were classified as responders to

perhexiline. The peak exercise load achieved by the

group was 42±35 W and peak exercise duration was

675 ±322 seconds (p=NS, as compared with base-

line). One of the six nonresponders (patient 2 in

Table 2) during the perhexiline phase had undetect-

able drug levels while on active treatment.

During the exercise test performed at the conclu-

sion of the placebo phase of treatment, two of 16

patients were able to exercise to at least the same

exercise load, duration, or both as during the base-

line test without experiencing angina. Compared with

the baseline exercise test, three of 16 patientsachieved at least a 20% increase in peak exercise

duration. Based on the pretrial definition of a posi-tive exercise response, three of 16 patients (18%;95% CI, 5-43%) were classified as responders to

placebo. The mean peak exercise load achieved bythe group was 36±t36 W, and peak exercise duration

TABLE 2. Summary of Subjective and Objective Responses to Perhexiline and Placebo

Double- Baseline ETT Perhex ETT Plac ETTblind diary Onset of Peak Onset of Peak Onset of Peak

Drg review angina exercise angina exercise Perhex angina exercise Pa

Patient order Perhex Plac (W) (sec) (W) (sec) (W) (sec) (W) (sec) response (W) (sec) (W) (sec) response1 2 + ... 12.5 330 12.5 450 25 711 abcd 25 690 25 720 abcd2 1 ... ..15 180 15 210 5 1 5 1* 5 130 5 1653 1 + ..30 870 30 1,200 10 420 30 960 10 540 30 1,0904 2 ..... 60 1,020 60 1,170 60 1,050 ab 20 390 20 6005 1 ..... 10 330 10 390 5 180 10 531 d 5 88 5 2406 2? + ..5 210 5 176 5 285 d7 1 5 240 10 360 10 675 abd 5 120 5 2088 2 + 5 270 5 300 5 300 10 390 bd 5 1 5 789 1 + ..10 405 20 660 20 660 abcd 15 600 15 720 abcd

10 2 + ..20 563 40 735 cd 20 5601,1 1 + ... 40 769 60 1,012 60 960 60 1,110 ab 40 720 60 1,02012 1... [20] [540] [60] [1,080] [60] f 1,020]13 2 ..90 720 60 480 90 720 20 60 90 68814 1 + ..80 1,369 90 677 90 70015 2 + ... 90 822 120 1,020 120 1,050 ab 90 900 120 96016 1 + ..40 376 40 371 40 42517 2 + ... 20 367 40 840 40 450 60 690t abc 20 570 40 780n 11 0 10 10 16 16 8 8 16 16 10 12 12 16 16 3Mean 29 520 38 662 26 431 42 657 22 401 36 577SD 28 312 35 386 24 293 35 311 24 309 36 316

Drug order, number indicates which 3-month block was randomized to perhexiline.Criteria for drug response as compared with baseline exercise testing, as follows: a, absence of angina at equivalent workload; b, absence

of angina at equivalent exercise duration; c, increase of >20% in peak exercise workload; d, increase of >20% in peak exercise duration.ETT, exercise tolerance test; Perhex, perhexiline; Plac, placebo.*Undetectable drug levels while on active treatment. tlnitial workload during test was 30 W.

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Cole et al Perhexiline Maleate in Refractory Angina 1265

Perhexiline Placebo

* Respondero NonResponder

U)

0CL

E

* Respondera NonResponder

Perhexiline Placebo

Blinded Diary Review ETT vs Baseline

FIGURE 1. Bar graphs ofsubjective and objective measures of responsiveness to perhexiline. Results ofsubjective evaluations (leftpanel) and objective evaluations (right panel) of antianginal efficacy are shown. Patients judged to be responders are representedby solid bars, and patients judged to be nonresponders are indicated by open bars. On double-blind review of angina diaries andspecific-activity scale, 11 patients indicated that one 3-month treatment block resulted in definite improvement as compared withopposite block, and six felt there was no difference. In all 11 cases of responders, by subjective criteria, the treatment block selectedcorresponded to perhexiline ( p<0.005). Based on objective evidence of improvement durting exercise testing (ETT) as comparedwith baseline, 10 patients were classified as responders during perhexiline treatment as compared with only three patients duringplacebo (p<0.05).

was 577±316 seconds (p=NS, as compared withbaseline).To control for the effects of multiple exercise tests,

the effects attributed to the natural history of thedisease and the placebo effect, the exercise testresponses during the perhexiline and placebo phaseswere each compared with the baseline exercise test(Figure 1). Applying the criteria of a positive exerciseresponse as compared with baseline if no anginaoccurred at an equivalent exercise load or durationand/or a 20% increase in peak exercise load or

duration was achieved, the proportion of individualsresponding to perhexiline was significantly greaterthan that observed during placebo (p<0.05) (Figure1). Restricting the definition of a positive exerciseresponse as compared with baseline to the absence ofangina at an equivalent exercise load or duration, theproportion of individuals responding to perhexilinewas also significantly greater than observed duringplacebo (p <0.05). Restricting the definition of a

positive exercise response as compared with baselineto at least a 20% increase in peak exercise load or

duration, there was a trend toward a greater propor-tion of responders on perhexiline as compared withplacebo, although this did not achieve statisticalsignificance (p<0.25). The analysis of variance withrepeated measures did not reveal any significantrelation between the order of drug administrationand any of the indices of exercise performancepreviously discussed. There was no statistically signif-icant difference in the rate-pressure product at rest(before exercise) or at peak exercise duration amongthe baseline, perhexiline, and placebo exercise tests(p=NS, by ANOVA). Finally, there was no evidenceof a period effect for either the objective or subjectiveresponses, suggesting a lack of any significant alter-ation in the patients' conditions attributable to thepassage of time.22

Subjective criteria of efficacy. Based on a double-blind review of angina diaries, nitroglycerin con-

sumption estimates, and the specific-activity scale, 11of 17 (65%; 95% CI 41-83%) were able to identify a

treatment phase associated with subjective improve-ment, whereas six patients were unable to distinguishbetween the two treatments. In all 11 cases wheredefinite subjective improvement occurred during one

of the treatment phases, the patients were receivingperhexiline. Thus, no patients identified the placebophase as the period during which they had subjectiveimprovement. The proportion of responders basedon subjective criteria was, therefore, significantlygreater during the perhexiline phase as comparedwith the placebo phase (p<0.005) (Figure 1).

Side effects. Five of 17 (29%) patients developedside effects, as follows: Four patients developedtransient ataxia; additionally, there were two epi-sodes of temporary nausea, one episode of dizziness,and one patient developed Beau's lines (horizontalstreaking of the nailbed). All of these side effectswere mild, occurred while under treatment withactive drug, were associated with blood levels of drugtransiently in excess of 600 ng/ml, and resolved in 2weeks or less with reduction of the maintenance doseof perhexiline and the ensuing decrease in perhexi-line blood levels. No patient developed other subjec-tive side effects previously attributed to perhexiline10such as anorexia, weakness, changes in visual acuity,rash, or weight loss. During the course of the study,no significant deviations from baseline measurementsoccurred in serum creatinine, blood urea nitrogen,alkaline phosphatase, bilirubin, glucose, or AST.During the course of the study, no patients developedsymptomatic bradycardia, hypotension, decompensa-tion of left ventricular function, or worsening ofventricular arrhythmias.

Pharnacokinetics of perhexiline in study group. Allpatients were started on a dose of perhexiline 100 mgp.o. daily for 3 days, followed thereafter by 100 mgp.o. twice daily. Sixteen of 17 patients (94%)required an alteration in the maintenance dose of

2010)0

0

Q

D

Ez

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

1266 Circulation Vol 81, No 4, April 1990

FIGURE 2. Bar graph of mean

daily dose of perhexiline during 3-month period of treatment withactive drug.

9 1 7 1 0 1 3 1 6 7 6 1 1 1 5 5

Patient Number

3 14 4 2 8 12

perhexiline during the 3-month period based on theresults of blood level determinations. Thirteenpatients required a lower maintenance dose to main-tain perhexiline blood levels between 150-600 ng/ml;on average, the lowering of dosage occurred on day27. Three patients required an increased dose tomaintain blood levels in the desired range. The mean

daily dose during the 3-month period of treatmentwith perhexiline was calculated for each patient, andranged widely at 23-278 mg/day (mean, 135±79mg/day) (Figure 2).While monitoring for side effects or making adjust-

ments to the maintenance dose of perhexiline, or

both, sufficient specimens were obtained in 11 of 17patients to permit estimation of the half-life of theterminal phase of drug elimination. The range ofhalf-lives was 35-864 hours and exhibited a bimodaldistribution. The terminal phase in all individuals,except patients 1 and 13, had a half-life of less than144 hours; the terminal phase in the remaining twopatients had half-lives of 402 and 864 hours, respec-tively (Figure 3).

Figure 4 illustrates the necessity for careful moni-toring of perhexiline blood levels. The left paneldepicts the results of serial perhexiline blood leveldeterminations in patient 1 (a presumed poor metab-olizer) who had a prompt rise in levels to nearly 2,500ng/ml in less than 1 week of treatment. No furtherperhexiline was administered until study day 90 whenthe blood level had declined back to the desiredrange. The estimated terminal half-life of excretionof perhexiline calculated from the data between day50 and day 100 was 864 hours. The right panel depictsthe results of drug-level monitoring in patient 12 (apresumed extensive metabolizer) who received pro-

gressive increases in the daily dose of perhexiline tomaintain blood levels of 150-600 ng/ml. Dosing withperhexiline was terminated on day 90 according to

the protocol, allowing an estimation of the terminalhalf-life of excretion at about 36 hours. Neitherpatient experienced drug-related side effects duringperhexiline treatment. Both cases demonstrate thevalue of therapeutic drug monitoring because thepresence of initially unacceptably high or low levelsand the slowly progressive rise in levels over timewould not have been suspected clinically before theemergence of an adverse reaction.

DiscussionThis report provides strong evidence that perhexiline

maleate administration results in an incrementalimprovement in patients with refractory angina pectoris

1-__Oa ._

._,

900

800

700

600

500

400

300

200

100

]U 1 1'

Patient Number

FIGURE 3. Bar graph showing terminal half-life of excretionofperhexiline. In 11 patients, terminal half-life of excretion ofperhexiline could be estimated from serial blood levels. Therewas a bimodal distribution of results, with terninal half-livesof excretion in patients 1 and 13 being distinctly longer,probably reflecting genetically determined poor hepatic metab-olism ofperhexiline.

61)(n0

61)

= (8

._x X-s- ECL

CO

1o0 9 1O 1D b

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Cole et al Perhexiline Maleate in Refractory Angina

600

500

400

Px Level(ngeml)

t 1/2= 864 hrs

300

200

100400 Px Dose200 (mg/dag)

a l w l | l w l 0

20 40 60 80 100

Study Day Number

Px Dose(mg/day)

Studg Dog Number

FIGURE 4. Graph showing plasma perhexiline levels in patients demonstrating poor and extensive metabolizer phenotype. Leftpanel: Results of serial perhexiline blood level determinations in patient 1 (presumed poor metabolizer). Right panel: Results ofdrug-level monitoring in patient 12 (presumed extensive metabolizer).

already receiving maximal medical therapy. Althoughthere was little doubt of the drug's efficacy in untreatedor minimally treated angina patients based on studiesperformed in the 1970s,2-11,23-26 this is the only studyusing a double-blind placebo-controlled crossoverdesign to document efficacy and safety when perhexi-line was added to maximally tolerated doses of a

conventionally prescribed contemporary drug regimen.The study population was treated more aggressivelywith standard antianginal therapeutic modalities thanin any other previously reported clinical trial of perhex-iline and included a number of patients who hadalready undergone one or more coronary artery bypassor angioplasty procedures.

Perhexiline was used with a negligible impact on thehemodynamic profile of individual patients. No exac-erbations of congestive heart failure or conductiondisturbances developed despite the fact that the drugwas administered in conjunction with 3-adrenergicreceptor antagonists and calcium channel blockingdrugs. The absence of any deleterious interaction withthe latter class of drugs is particularly relevant, con-

sidering 70% of patients were receiving verapamil,diltiazem, or both, when prescribed perhexiline.

EfficacyAlthough subjective evaluations of a drug's

antianginal effectiveness as compared with placebocan be reliable if both patient and examining clini-cian are blinded as in this study, we felt that an

objective measure of perhexiline's efficacy was nec-

essary as well, particularly because the patients werealso receiving a number of other antianginal medica-tions concurrently. Based on the pretrial criteria forestablishing drug efficacy, 63% of the patientsimproved their bicycle ergometer performance whilereceiving active drug, as judged by a blinded investi-gator (P.L.C. or A.D.B.), as compared with just 18%in the placebo phase (p<0.05). The fact that therewere no significant changes in the group means forexercise test performance during the perhexiline or

placebo phase versus baseline testing seems to reflectthe considerable interindividual variability in theresponse to exercise, as indicated by the magnitudeof the standard deviations.

Criteria 1-4, noted in "Methods," were selected as

measurement end points for objective evidence ofdrug efficacy because the elimination of angina withexercise or the increase in workload tolerated duringexercise are clinically meaningful drug responses.Other criteria, such as the rate-pressure products atonset of angina or 1 mm or more ST segmentdepression, were not selected as objective end pointsbecause previous drug trials with fl-blockers andcalcium antagonists failed to show a consistent drug-related increase in double product at the onset ofischemia. The majority of the drug benefit reportedin these cases seemed to be related to an increase inworking capacity before the development of isch-emia. Because we anticipated that many patients whowould meet the entry criteria for this trial would havean abnormal baseline electrocardiogram or be receiv-ing medications that would obscure the interpreta-tion of ST segment deviation with exercise, or both,the time to onset of 1 mm or more ST segmentdepression was also not selected as a measurementend point.

It is possible that a residual perhexiline effect afterthe crossover to placebo phase might have obscureda more pronounced perhexiline effect in patients whorandomly received active drug first. We believe,however, this is unlikely for several reasons. Based on

the estimated half-life of excretion of the drug, evenindividuals with the lowest clearance rates wouldhave had undetectable perhexiline plasma levelswithin 3 months after crossover at the time of thefinal exercise test. All patients had undetectableperhexiline levels during the placebo phase. Further-more, by ANOVA, there was no effect of the order ofdrug administration on exercise performance.Using subjective measures of efficacy including

angina diaries and estimates of sublingual nitroglyc-

3000-

2000Px Level(ongml)

1000 -

00

1267

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

1268 Circulation Vol 81, No 4, April 1990

erin consumption reviewed by a blinded investigator,and a specific-activity scale, again, the data areconvincing. Sixty-five percent (11 of 17) of patientshad a subjective improvement in anginal symptomsduring the perhexiline phase, whereas none noted animprovement in symptoms during the placebo phase(p<0.005). No patient reported a worsening of base-line angina during the treatment phase that corre-sponded to therapy with perhexiline, and nonerequired any adjustment of other antianginal medi-cations during the active drug phase. This remark-able difference in the patients' blinded subjectiveassessment of drug efficacy is noteworthy, consider-ing the complexity and severity of most patients'underlying coronary disease in this study and theirintensive regimens of other concomitantly adminis-tered antianginal agents.

SafetyThe results of the present investigation support the

conclusions of Horowitz et al,15 based on evidencefrom an uncontrolled long-term study of a patientpopulation similar to ours, that therapeutic efficacycan be maintained with a minimum of toxicity whendosage is adjusted to maintain plasma levels within arange of 150-600 ng/ml. In our study, side effectsattributable to the drug were relatively minor, wererelated to transient elevations in plasma perhexilinelevels to more than 1,000 ng/ml, and disappearedwhen the dosage was adjusted to bring plasma levelswithin the therapeutic range.

In the report by Horowitz et al1 concerning 18patients with refractory angina given 50-400 ng ofperhexiline daily with dose determined by relief ofsymptoms rather than by any arbitrary therapeuticrange of plasma levels, 74% of patients eitherbecame asymptomatic or had a clear reduction inangina frequency but an unacceptable 47% devel-oped hepatic or neurological toxicity. The averagemaximal perhexiline plasma level in these patientswas 1,170 ng/ml overall, and 1,281 ng/ml in patientsdeveloping side effects. Importantly, no patient inthis group developed side effects during as long as 27months of drug therapy if plasma levels were below700 ng/ml. When these investigators studied a secondgroup of 19 patients whose plasma drug levels weremaintained below 600 ng/ml, no clinical or biochem-ical evidence of toxicity attributable to perhexilinedeveloped during durations of treatment lasting up to42 months. Our study supports the recommendationof Horowitz et al15 that the therapeutic range belimited to 150-600 ng/ml.

Pharmacokinetics of Perhexiline EliminationPerhexiline is rapidly absorbed from the gastroin-

testinal tract after oral administration and is excretedin urine and stool as the parent compound plus itsmore water-soluble metabolites, mono- and dihydroxy-perhexiline. Bioavailability is variable, in the range of52-95%.10 The elimination half-life of perhexilinehas been reported to be in the range of 2-5 days for

most patients, although the clearance rate is widelyvariable and nonlinear with respect to dosage.12Importantly, hepatic metabolism has been demon-strated to be saturable at "therapeutic" doses of thedrug. For a small subset of patients receiving perhex-iline, nominally classified as "poor hydroxylators"who have a genetic defect in the metabolism ofperhexiline and of other drugs, saturation of hepaticmetabolism and a switch from first-order to zero-order elimination kinetics occurs at relatively lowdoses of the drug.12'7 These patients can have dra-matically lengthened elimination half-lives once thehepatic metabolic threshold is exceeded, as illus-trated by our patient 1 (Figure 4, left panel), in whomthe perhexiline elimination half-life was longer than5 weeks. These patients seem to be at greatest riskfor toxicity.

Perhexiline undergoes classic "phase F' oxidativemetabolism in the liver by a specific isoform of thecytochrome P-450 family of related enzymes. Knowl-edge of the molecular biology of the P-450 enzymeshas rapidly expanded in the last several years and isthe topic of several recent extensive reviews.27-29 Thespecific P-450 enzyme responsible for perhexilinemetabolism is now known to be a member of the typeII, class D family, denoted P-450dbl.30-32 The sub-script "dbl" refers to "debrisoquine type 1," afterthe now rarely used sympatholytic antihypertensivedrug that, in very low doses, serves as a useful agentto determine hydroxylator phenotype in humans.33-35This enzyme is also responsible for the initial metab-olism of many other drugs, including many of thenewer type Ic antiarrhythmics (e.g., encainide andflecainide), several /3-blockers (e.g., timolol, meto-prololol, and propranolol), and some tricyclicantidepressants.2829 The poor hydroxylator pheno-type is transmitted as an autosomal recessive traitwith a prevalence of 5-10% among Caucasians, andis presumably because of inheritance of two mutantalleles for the P-450dbl gene, recently localized tochromosome 22.31,32

It would be useful if the poor hydroxylator patientscould be identified before initiating therapy withperhexiline. This can be done by measuring, in theurine, the ratio of debrisoquine and its major P-450dbl metabolite after a subtherapeutic oral dose, arelatively simple task for most laboratories withaccess to standards. Alternatively, a patient's geno-type can be established by identifying characteristicrestriction fragment length polymorphisms on South-ern analyses of genomic DNA. Although this sort ofanalysis would aid in establishing a safe therapeuticdose of perhexiline, it is not necessary for routineclinical practice and cannot replace plasma drugmonitoring. It should be emphasized that 94% ofpatients in this study, not just the presumed two poorhydroxylators, required adjustment of the mainte-nance dose of perhexiline, reflecting the slow devel-opment of saturation of hepatic metabolism and thetransition from first-order to zero-order eliminationkinetics in many patients. Nevertheless, we believe

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Cole et al Perhexiline Maleate in Refractory Angina 1269

that, with reasonable attention to therapeutic drugmonitoring, only slightly more rigorously than isappropriate for digoxin or procainamide and which isless demanding than monitoring prothrombin timeswith warfarin therapy, perhexiline is a safe drug. Thedose should be monitored at weekly intervals until astable plasma drug level is achieved and, then,monthly or less frequently, as needed, thereafter.

Perhexiline's Mechanism ofActionOn an extensive review of the literature up to 1980,

William E. Vaughans concluded that experimentalpharmacologists in academia and industry had failedto elucidate the means by which perhexiline acted asan antianginal agent in humans. He concluded thatthe experimental evidence did not support thedescription of perhexiline as a coronary vasodilatorbut, rather, that its antianginal efficacy might bebecause of a change in myocardial metabolism, per-haps by altering the primary source of carbon andreducing equivalents for oxidative phosphorylationfrom fatty acids to glucose.5 There is evidence thatthe enhancement of carbohydrate uptake and metab-olism could have protective effects during isch-emia.36-38 This might be because there is an approx-imate 6% increase in ATP produced per 02

consumed when glucose is used rather than fattyacids as substrate, thus providing a critical margin ofincreased efficiency when oxygen supply to the myo-cardium is limited by obstructive coronary arterydisease. Within the past decade, other agents such asoxfenicine, an inhibitor of carnitine acyltransfer-ase,39-41 and the adenosine derivative, N(6)-phenyl-N(6)-allyladenosine,42,43 both of which reduce fattyacid oxidation in favor of glucose oxidation, havebeen demonstrated to have protective effects inexperimental models of ischemic myocardium with-out altering hemodynamics.

Clinical ImplicationsWe have shown, in a double-blind prospective

randomized study using a crossover design betweentreatment and placebo phases, that perhexiline male-ate is effective in reducing the frequency and severityof anginal attacks and in increasing exercise toler-ance in patients with refractory angina, many ofwhom had already undergone at least one coronaryrevascularization procedure. Perhexiline's efficacywas superior to placebo despite the fact that patientswere receiving maximally tolerated doses of theconventional antianginal regimen currently availablein the United States. We also confirmed that the riskof potentially serious toxic effects of perhexiline can

be greatly reduced, if not eliminated, by appropriateattention to plasma drug levels.15Although it is difficult to estimate the size of the

relevant population, we believe that the number ofrefractory angina patients unsuitable for furtherrevascularization or angioplasty will increase as thepool of patients who have already undergone one ormore procedures age and their atherosclerosis

progresses. Additionall, perhexiline might proveuseful for those patients whose coronary disease isextensive and diffuse or inaccessible to surgery or tothe angioplasty catheter. Finally, perhexiline hasbeen used, in addition to conventional antianginalagents, in Europe, New Zealand, and Australia toreduce anginal symptoms in patients waiting forscarce semielective positions on cardiovascular sur-gery operative schedules.Although perhexiline has been shown to be effec-

tive when given alone as single-drug antianginaltherapy, we would not recommend its being used as afirst-line drug for the treatment of angina because ofits potential for serious toxicity. Nevertheless, theabsence of clinically important hemodynamic or neg-ative inotropic effects of the drug, or of knowndeleterious drug interactions with the many othermedications these patients were receiving confirmsthe experience of Horowitz et als that, if dosedproperly, perhexiline can be safely used in addition toother therapeutic modalities.

Perhexiline's AvailabilityPerhexiline is unavailable in the United States, in

large part, because of the fact that in recent yearsperhexiline has not had a strong advocate. Theinstitutions that would normally perform that func-tion, clinical investigators in academic medicine, andthe pharmaceutical industry, have been reluctant tosupport the drug for several reasons. These reasonsinclude the drug's reputation for severe hepatic andneurological toxicity that developed before the genet-ics of its metabolism were understood and therapeuticdrug monitoring became widely available. Anotherreason is the original marketing of the drug as acalcium antagonist, only to be superceded by effectiveand less toxic drugs of this class. Finally, the fact thatpatent rights have expired for this agent in many partsof the world (although not in the United States, wherethe drug has never been marketed). Although we weregenerously supplied with perhexiline maleate tabletsand look-alike placebos, as well as standards for thehigh-performance liquid chromatography assay, thisstudy received no funding, directly or indirectly, fromthe pharmaceutical industry.

Perhexiline remains commercially available in sev-eral European and South American countries, Aus-tralia, and New Zealand. Based on the average cost ofperhexiline in these countries, an estimated monthlysupply of the drug in U.S. funds would cost $10-20.The results of this report, when viewed in the contextof other recent uncontrolled studies, support a moreextensive examination of the safety and efficacy ofperhexiline in patients with refractory angina.

References1. Chamberlain DA: The medical treatment of angina pectoris.

Br Heart J 1987;58:547-5512. Wallace DC: Perhexiline maleate in the treatment of angina

pectoris. Med JAust 1978;2:466-495

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

1270 Circulation Vol 81, No 4, April 1990

3. Mir MA, Kafetzakis EM: Assessment of perhexiline maleatein angiographically proven intractable angina: A double-blindtrial. Am Heart J 1978;96:350- 354

4. Horgan JH, O'Callaghan WG, Teo KK: Therapy of anginapectoris with low-dose perhexiline. J Cardiovasc Pharmacol1981;3:566-572

5. Vaughan WE: Antiarrhythmic Action and the Puzzle of Perhex-iline. London, Academic Press, Inc, 1980

6. Bounhoure JP: Treatment of severe coronary angina withperhexiline maleate: Long-term results, in: Perhexiline Maleate.Proceedings of a Symposium. Amsterdam, Excerpta Medica,1978, pp 131-139

7. Morledge J, Adams D, Hudak WJ, Powell RL, Kuzma RJ:Effects of perhexiline maleate versus placebo in angina pec-

toris. Symptomatic and treadmill exercise double-blind evalu-ation, in: Perhexiline Maleate. Proceedings of a Symposium.Amsterdam, Excerpta Medica, 1978, pp 178-190

8. Pilcher J, Chandrasekhar KP, Rees JR, Boyce MJ, Pierce H,Ikram H: Long-term assessment of perhexiline maleate inangina pectoris. Postgrad Med J 1973;49:115-118

9. Pepine C, Schang SJ, Bemiller CR: Effects of perhexiline on

coronary hemodynamic and myocardial metabolic responses

to tachycardia. Circulation 1974;49:887- 89310. Pexid Product Summary for Physicians. Cincinnati, Ohio, Mer-

rell Dow Pharmaceuticals Inc, 198211. Lockhart JDF, Masheter HC: Report of a multicentre moni-

tored release study of perhexiline maleate in the prevention ofangina pectoris. Br J ClGn Pract 1976;30(a):172

12. Horowitz JD, Morris PM, Drummer 0, Goble AJ, Louis WJ:Saturable metabolism of perhexiline maleate: Correlationswith long term toxicity. Am J Cardiol 1981;47:399

13. Singlas E, Goujet MA, Simon P: Pharmacokinetics of perhex-iline maleate in angina patients with and without peripheralneuropathy. Eur J Clin Pharmacol 1978;14:195-201

14. Laplane D, Bousser MG, Bouche P, Touboul PJ: Peripheralneuropathies caused by perhexiline maleate, in: PerhexilineMaleate. Proceedings of a Symposium. Amsterdam, ExcerptaMedica, 1978, pp 89- 96

15. Horowitz JD, Sia STB, MacDonald PS, Goble AJ, Louis WJ:Perhexiline maleate treatment for severe angina pectoris-correlations with pharmacokinetics. Int J Cardiol 1986;13:219-229

16. Horowitz JD, Morris PM, Drummer OH, Goble AJ, Louis WJ:High performance liquid chromatographic assay of perhexilinemaleate in plasma. J Pharn Sci 1981;70:320-322

17. Pilcher J, Cooper JDH, Turnell DC, Matenga J, Paul R,Lockhart JDF: Investigations of long-term treatment withperhexiline maleate using therapeutic monitoring and electro-myography. Ther Drug Monit 1985;7:54-60

18. Feinstein AR, Wells CK: A new clinical taxonomy for ratingchange in functional activities of patients with angina pectoris.Am Heart J 1977;93:172-182

19. Goldman L, Hashimoto B, Cook EF, Loscalzo A: Comparativereproducibility and validity of systems for assessing cardiovas-cular functional class: Advantages of a new specific activityscale. Circulation 1981;6:1227-1234

20. Glantz SA: Primer of Biostatistics, ed 2. New York, McGraw-Hill Book Co, 1987, pp 276-280

21. Ingelfinger JA, Mosteller F, Thibodeau LA, Ware JH: Biosta-tistics in Clinical Medicine, ed 2. New York, MacMillan Publi-cations Inc, 1987, pp 309-315

22. Woods JR, Williams JG, Tavel M: The two-period crossover

design in medical research. Ann Intern Med 1989;110:560- 56623. Armstrong ML: A comparative study of perhexiline, beta-

adrenergic blocking agents and placebos in the management ofangina pectoris. Postgrad Med J 1973;49:108-111

24. Horowitz JD, Mashford ML: Perhexiline maleate in thetreatment of severe angina pectoris. Med J Aust 1979;1:485-488

25. White HD, Lowe JB: Antianginal efficacy of perhexilinemaleate in patients refractory to beta-adrenoreceptor block-ade. Int J Cardiol 1983;3:145-155

26. Teo KK, Kelly JG, Carby JF, Ennis JT, Horgan JH: Perhexi-line during exercise training in coronary heart disease. ClinPharmacol Ther 1983;34:744-748

27. Nebert DW: P450 genes: Structure, evolution, and regulation.Annu Rev Biochem 1987;56:945-993

28. Clark DWJ: Genetically determined variability in acetylationand oxidation. Therapeutic implications. Drugs 1985;29:342-375

29. Nebert DW, Gonzalez FJ: P450 genes and evolutionary genet-ics. Hosp Pract (Off) 1987;22:63-74

30. Gonzalez FJ, Skoda RC, Kimura S, Umeno M, Zanger UM,Nebert DW, Gelboin HV, Hardwick JP, Meyer UA: Charac-terization of the common genetic defect in humans deficient indebrisoquine metabolism. Nature 1988;331:442-446

31. Gonzalez FJ, Vilbois F, Hardwick JP, McBride OW, NebertDW, Gelboin HV, Meyer UA: Human debrisoquine 4-hydroxylase (P45011lD1): cDNA and deduced amino acidsequence and assignment of the CYP2D locus to chromosome22. Genomics 1988;2:174-179

32. Skoda RC, Gonzalez FJ, Demierre A, Meyer UA: Two mutantalleles of the human cytochrome P-450 dbl gene (P450C2D1)associated with genetically deficient metabolism of debriso-quine and other drugs. Proc Natl Acad Sci USA 1988;85:5240-5243

33. Kupfer A, Schmid B, Pfaff G: Pharmacogenetics of dex-tromethorphan O-demethylation in man. Xenobiotica 1986;16:421-433

34. Shah RR, Oates NS, Idle JR, Smith RL, Lockhart JDF:Impaired oxidation of debrisoquine in patients with perhexi-line neuropathy. Br Med J 1982;284:295-299

35. Distlerath LM, Guengerich FP: Characterization of a humanliver cytochrome P-450 involved in the oxidation of debriso-quine and other drugs by using antibodies raised to analogousrat enzyme. Proc NatlAcad Sci USA 1984;81:7348-7352

36. Vik-Mo H, Mjos OD: Influence of free fatty acids on myocar-dial oxygen consumption and myocardial ischemic injury. AmJ Cardiol 1981;48:361-365

37. Liedtke AJ, Nellis S, Neely JR: Effects of excess free fattyacids on mechanical and metabolic function in normal andischemic myocardium in swine. Circ Res 1978;43:652-661

38. Opie LH: The glucose hypothesis: Relation to acute myocar-dial ischemia. J Mol Cell Cardiol 1970;1:107-115

39. Higgins AJ, Morville M, Burges RA, Gardiner DG, Page MG,Blackburn KJ: Oxfenicine diverts rat muscle metabolism fromfatty acid to carbohydrate oxidation and protects the ischemicheart rat. Life Sci 1980;27:963 970

40. Korb H, Hoeft A, Hunneman DH, Schraeder R, Wolpers HG,Wober W, Hellige G: Changes in myocardial substrate utili-zation and protection of ischemic stressed myocardium byoxfenicine [(S)-4-hydroxyphenylglycine]. Naunyn Schmiede-bergs Arch Pharmacol 1984;327:70-74

41. Bergman G, Atkinson L, Metcalfe J, Jackson N, Jewitt DE:Beneficial effects of enhanced myocardial carbohydrate utili-zation after oxfenicine (L-hydroxyphenylglycine) in anginapectoris. Eur Heart J 1980;1:247-253

42. Kahles HW, Reigger GAJ: Prolonged ischemic tolerance dur-ing coronary angioplasty by H(6)-phenyl-N(6)-allyladenosine.Circulation 1988;78:II-454

43. Kahles H, Junggebwith J, Lick T, Schafer W, Kochsiek K:

Effects of a N(6)-disubstituted adenosine derivative on myo-cardial metabolism and ischemic stress following coronaryocclusion. Arzneimittelforschung 1987;37:1137-1140

KEY WORDS * angina * myocardial ischemia * perhexilinemaleate * clinical pharmacology * slow metabolizer * cytochromeP-450

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from

Smith and E M AntmanP L Cole, A D Beamer, N McGowan, C O Cantillon, K Benfell, R A Kelly, L H Hartley, T W

placebo-controlled clinical trial of a novel antianginal agent.Efficacy and safety of perhexiline maleate in refractory angina. A double-blind

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1990 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.81.4.1260

1990;81:1260-1270Circulation. 

http://circ.ahajournals.org/content/81/4/1260the World Wide Web at:

The online version of this article, along with updated information and services, is located on

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on October 5, 2017

http://circ.ahajournals.org/D

ownloaded from