British Journal of Ophthalmology, 1984, 68, 553-560

Severe ocular side effects of perhexilene maleate:case report

J. M. GIBSON,' A. R. FIELDER,' A. GARNER,2 AND P. MILLAC3

From the 'Department of Ophthalmology, University of Leicester School of Medicine, the2Department of Pathology, Institute of Ophthalmology, Cayton Street, London, and the3Department of Neurology, Leicester Royal Infirmary

SUMMARY We report a case of perhexilene maleate (PEXID) toxicity in which the presentingfeature was loss of vision secondary to chronic papilloedema. Vortex keratopathy similar to thatseen in amiodarone keratopathy was present, and corneal and conjunctival biopsy findings are

presented. To our knowledge this is the first case report of a keratopathy occurring in perhexilenetoxicity. After withdrawal of the drug the papilloedema and keratopathy subsided, but some visualdeficit remains. The properties of perhexilene maleate and other amphiphilic drugs are described,and the possible aetiology of vortex keratopathy is discussed.

Perhexilene maleate (PEXID) is an antianginal drugwhich has been in general use in Britain since 1975.'Several side effects have been associated with its use,including hypoglycaemia, hepatic dysfunction,peripheral neuropathy,4 and papilloedemra.5" Wereport the case of a patient who developed raisedintracranial pressure and papilloedema while takingthis drug and who presented with severe visual loss.Ophthalmic examination revealed a bilateral vortexwhorl keratopathy indistinguishable from the verti-cillate keratopathy seen with chloroquine and amio-darone. The histopathological findings of cornealepithelial and conjunctival biopsy are presented, andother drugs with similar amphiphilic properties toperhexilene are discussed.

Case report

A 62-year-old Caucasian man attended the Oph-thalmic Department of the Leicester Royal Infirmarywith a one-week history of blurred vision affecting hisright eye. Corrected visual acuity was 6/6 in each eye,but the Goldmann perimeter showed enlargement ofthe blind spots in each eye and inferior field loss in theleft. Fundus examination showed bilateral optic discoedema and attenuation ofthe retinal arterioles. Plainskull and orbital x-rays and head CT scan showed noabnormality.Correspondecnc to Mr J. M. Gibson. FRCSEFd Dcpairtimicnt ofOphthailmology. Clinicall Scicnccs Buildling. L-ciccstcr Royail In-uir.iary. PO Box 65. Lciestcr LE2 7LX.

Three weeks later his corrected visual acuity haddeteriorated to 6/18 in the right eye and 6/24 in theleft eye. Slit-lamp examination revealed bilateralwhorl-like epithelial keratopathy (Fig. 1). Bilateraloptic disc oedema with choroidal folds and attenuatedarterioles were found on fundus examination. Apresumptive diagnosis of chronic papilloedema wasmade, and he was admitted to hospital for furtherinvestigation.The patient was taking perhexilene maleate initially

in a dose of 300 mg daily for a week before reducing ithimself to 200 mg daily, and he had been on this dosefor eight months at the time of presentation. He was

Fig. I High-power reiroillitu iiitiio .slit-l(Illp)phologra,lpoJ leji coriea showing vortex. whorl-like keratopathl

553

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

J. M. Gibson, A. R. Fieldler, A. Garner, and P. Millac

Fig. 2 Right fluorescein (anigiogramil in earlyv venous phasesshowinig dilated( capillaries ont (diCsurf(c(ac antt (choroidalfOldlsliperiorlv.

also taking a thiazide diuretic and penicillamine forgout, but the patient had never taken amiodarone,chloroquine, or indomethacin.The patient had suffered frontal headaches over

the preceding few weeks, increased by bending downand coughing. There were no other neurologicalsymptoms, but he had lost 13 kg in weight over theprevious two months. Clinical examination failed toelicit any abnormal neurological or systemic signs.Intracranial pressure measured by lumbar puncturewas raised at 32() mm cerebrospinal fluid. Fundusfluorescein angiography showed leatkage from dilatedoptic disc capillaries in the late stage and confirmedthe presence of choroidal folds (Figs. 2, 3). Thesefindings supported the diagnosis of long-standingpapilloedema.

Investigations gave the following results: haemo-globin 17-0 g/dl, erythrocytes 5*66x 10('/I, urea 7Tmmol/l, sodium 139 mmol/l, potassium 4-3 mmol/l,bicarbonate 23 mmol/l, creatinine 142 mmol/l,alkaline phosphatase 132 IU/I, alanine transaminase302 IU/I, gamma glutamyl transferase 206 IU/I, totalbilirubin 19 mmol/l.

Plasma viscosity was 1-89 cp, VDRL and Tre-ponema pallidiutni haemagglutination (TPHA) testswere negative, and an autoantibody screen wasnegative. Plasma protein electrophoresis gave normalresults. Urine microscopy showed no excess ofbirefringent material.

Fig. 3 Left fSlioresceini aigiogrami in late stage showingleakage of d'tvefromn dila(tet disc capillaries and smalllhae1torrhliige at (isc margr,ini. A clloroidlalfol(l can he seen (itto/)ofp ratie.

Cerebrospinal fluid (CSF) analysis: protein lessthan 0.1 g/dl, glucose 4 3 g/dl, Pandy test negative,VDRL and TPHA tests negative, erythrocytes 77x1()"/, leucocytes nil. The CSF sample was furtheranalysed by the Department of Chemical Pathologyat the National Hospital, Queen Square, London. Nooligoclonal immunoglobulins were detected, and theproteins were consistent with moderate transudationof high-moleculir-weight proteins across a damagedblood/CSF endothelium barrier.

Plasma perhexilene levels were assayed by gaschromatography, and the initial level was 3-3 mg/l(normal therapeutic range () 1 1 mg/l). Patternreversal visual evoked responses showed reducedamplitudes and delayed P. latencies (right 168 ms,left 145 ms).

In the absence of any other causes of raised intra-cranial pressure a diagnosis of perhexilene toxicitywas made and the drug withdrawn. A short course ofsystemic corticosteroids and chlorthalidone wasstarted in an attempt to reduce CSF production. Ninedays after withdrawal of the drug intracranial pressurehad fallen to 140mm of CSF, liver function tests werestill abnormal, and the serum level of perhexilenewas estimated as 3 1 mg/I. Over the ensuing twoweeks, however, the patient's visual acuity improvedand tests of liver function partially improved.

Six weeks after drug withdrawal the vision hadimproved to 6/9 right and 6/12 left; papilloedema had

554

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

Severe ocutlar side effects ol perhexvilene inatleate: casereport55

subsided, but both optic discs showed temporal pallorwith arteriolar attenuation. Liver function was stillmarginally abnormal. Six months after stoppingperhexilene the liver function had returned to normalano the visual acuity was 6/9 right and 6/12 left. Thevisual fields have improved but there remains someloss to small targets inferiorly in the left eye. Con-junctival and corneal epithelial biopsies were takenshortly before the drug was withdrawn and light andelectron microscopy performed.

LIGHT MICROSCOPYThe conjunctival epithelium wais generally thickened,

showing squamous metaplasia and a loss of mucus-secreting goblet cells. There were sporadic foci ofcystic degeneration, aind the flattened cytoplasm ofresidual cells surrounding rounded intraepithelialcavities were filled in part by variably fragmented andcondensed debris of defunct cells. The corneal epi-thelium presented similatr degenerative changes.

ELECTRON MICROSCOPY

Many conjunctival epithelial cells at all levels, butmore especially in the basal cell layer, containedmultiple electron-dense cytoplasmic inclusions whichwere either randomly distributed or in clusters (Fig.

-t

ll..tt:t

..........

4-1

............

..X

Fig. 4 .S~everal epithelial (c/Iells i/ihe COnjIunctiva Inc/lude electromnicrographI, X 7800).

.VW.

.....

A

.

oil -dense botlies in their cYtoplasin. (Transmission electron

555

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

J. M. Gibson, A. R. Fielder, A. Gartner, and P. Millac

4). Each inclusion was round or oval, with maximumdiameters ranging from 0( I to 1 () !tm; in someinstances they were seen to be membrane-bound.Often the inclusions were amorphous or so electron-dense as to make any internal structure unrecognis-able (Fig. 5), but in some there was a well-definedconcentric laminar arrangement (Fig. 6), each laminameasuring approximately 9 nm thick. Occasionallythere was a less-ordered membranous structure (Fig.5), and exceptionally there was at clear halo betweenthe membranous whorl and the limiting membrane ofthe inclusion (Fig. 7); some were completely empty.

Mitochondria were numerous, and there was anormal complement of tonofilaments.

Structurescorrespondingto the intraepithelial cystsobserved by light microscopy consisted of amorphousdebris surrounded by the attenuated cytoplasm ofadjaicent intact cells (Fig. 8). (Comment: the intra-epithelial inclusions are closely similar if not identicalto those described in amiodarone keratopathy byD'Amico et al.' It is reasonable to associate the cysticdegeneration prevalent in the more superficial epi-thelial layers with the presence of the lipid inclusionsin the basal cells.)

Fig. 5 Embedded in ihe cytoplasm of (cirt of cI coin uncli'cil e/pitlleil'l cell there are mnembran e-holinid inclusions whic h arelvariablv eleciron dense w it/h1 10 (cl(irIl esolvacible inf rastru c (.solid (irro ws) or ha ve an irregulalr nehrnranous .iruc ture(opeiu crrow). ( Trans,ini.s.ioni c/cc ho i(i TogLrap)hl, x 48 000).

556

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

Severe ocular side effects ofperhexilene maleate: case report

Discussion

The patient presented here suffered a long-standing(and probably permanent) visual defect as a result ofchronic papilloedema following perhexilene adminis-tration. Perhexilene is an effective antianginal drug,recommended for use in patients with severe anginawhen other drugs fail to achieve adequate sympto-matic control.8 Its mode of action is not fully under-stood, but exercise tolerance is improved." Perhexi-lene is rapidly and completely absorbed by mouthand is metabolised by the liver. Its half-time of elim-ination is 2-5 days. "'

Serious adverse side effects have been reported inpatients taking perhexilene, including neuropathy,liver damage, and weight loss. Raised intracranialpressure and papilloedema may appear concurrentlywith neuropathy.'" So far there have been 10associated deaths with hepatic toxicity.'2 and recentlythe Committee on Safety of Medicines drew attentionto these adverse reactions.

The peripheral neuropathy appears to be due to ademyelinating process.'3 It is interesting to note,however, that clioquinol (iodochlorhydroxyquin),also an amphiphilic drug, has been implicated incausing subacute myelo-opticoneuropathy (SMON)in Japan. '4 In this condition there is a demyelinationof the optic nerves and spinal cord columns, thoughraised intracranial pressure is not usually a feature.Why perhexilene should occasionally cause raisedintracranial pressure and papilloedema remainsunexplained.

Ocular side effects of perhexilene are few, andthere has been no previous report of corneal involve-ment. However, a case of keratoconjunctivitis siccawas reported by Turut et al. which improved after theperhexilene was withdrawn.'The cause of the serious toxic reactions to thera-

peutic doses of perhexilene in some patients is notclear, but there is evidence that they may result fromimpaired drug metabolism. Singlas et al. '" demon-strated that patients with peripheral neuropathy had

Fig. 6 Several ofthe inclusions in this cell have an organised concentric lamellar ahrangement, with width ofeach lamellaraveraging 9nm. (Transmission electron micrograph, x 78 000).

557

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

J. M. Gibson, A. R. Fielder, A. Garner, and P. Millac

higher plasma perhexilene concentrations and longerclearance times than those patients taking the drugwho did not suffer neuropathy. It is likely that ourpatient was unable to metabolise perhexilenenormally, a condition reflected by his high serumconcentration and slow rate of drug clearance.The ability to metabolise perhexilene efficientlyappears to be an individual genetically determinedcharacteristic. "

Perhexilene is a cationic amphiphilic compound-that is, the hydrophilic and hydrophobic moieties arein very close proximity within the same molecule.Owing to their physicochemical properties amphi-philic drugs display high affinities to polar lipids andform tight but not irreversible complexes with thelipids. This results in the alteration of certain proper-ties of the polar lipids such that they are less readilycatabolised and accumulate within lysosomes. 7 Theresulting inclusion bodies resemble the pathological

changes seen in some inherited metabolic disorders.Ultrastructurally the drug-induced phospholipidosesare characterised by membrane-bound residualbodies with a lamellated structure called myelinfigures or multilamellar bodies. These have beenobserved in liver biopsy specimens from patients withperhexilene toxicity. '8 They appear to be identical tothe membrane-bound bodies which we have demon-strated in the conjunctival and corneal epitheliumfrom this patient, and are identical to those describedin amiodarone keratophy.7 It is not surprising,therefore, to find that the other drugs which havebeen implicated in causing verticillate keratopathyare all amphiphilic despite being pharmacologicallydifferent-for example, amiodarone, chloroquine,mepacrine, hydroxychloroquine, amodiaquin, andindomethacin. "' Furthermore similar intralysosomalinclusion bodies occur in Fabry's disease, a rare in-herited metabolic disease in which there is a

Fig. 7 Ofseveral vacuoles in this cell one contains a central whorledlaminarstructure (arrow). There is an unusually densetonofilament content allied with a relative dearth ofmitochondria. (Transmission electron micrograph, x8300).

558

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

Severe ocularside effects ofperhexilene maleate: case report

deposition of glycolipid in the skin, kidneys, heart,gastrointestinal tract, and central nervous system aswell as in the cornea.20

Besides corneal involvement it is well known thatother ocular tissues can be affected by amphiphilicdrugs. Thus studies have shown intralysosomal in-clusion bodies in the retinal pigment epithelium ofrhesus monkeys,' while in rats similar changes havebeen induced by amiodarone.2 To our knowledge noretinopathy associated with amiodarone or perhexi-lene has so,far been described in man.

Franqois23 first used the term cornea verticillata todescribe the typical keratopathy involving the cornealepithelium in the characteristic vortex whorl. Theyhave been comprehensively reviewed by Bron,'4 who

described the characteristic vortex patterns seen intoxic keratopathies, Fabry's disease, corneal ironlines, and in certain local corneal disturbances. Hehas suggested that the formation of the vortex patternis due to the growth and repair patterns of the cornealepithelium. There is a continuous centripetal andradial slide of epithelial cells from the limbal epi-thelium. The vortex pattern is seen because affectedcells have lost their transparency. What is not clear iswhy the vortex pattern is almost always in the lowerhalf of the cornea. It is possible that this could be amechanical or temperature-dependent phenomenonand develop more readily in the relatively coolerzone of the cornea.25The case we describe is important in several

Fig. 8 A cystic cavity containing amorphous debris andsurrounded by the attenuated cytoplasm ofadjacent epithelial cells.(Transmission electron micrograph, x3900).

559

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from

J. M. Gibson, A. R. Fielder, A. Garner, and P. Millac

respects. A patient taking perhexilene should bereviewed regularly, as indeed the manufacturerrecommends, to screen for serious side effects. Raisedintracranial pressure and papilloedema can lead tolong-standing and perhaps permanent visual loss. Thefinding of a vortex keratopathy should therefore alertthe ophthalmologist to the possibility of otherpotentially more serious side effects such as raisedintracranial pressure and papilloedema.

Wc thaink Dr J. Coopcr, of the Coventry and Warwickshirc Hospital,for pcrforming the drug assays, Miss D. Callaghan for hcr photo-graiphic expcrtisc, aind Mrs J. Wood for her sccrctarial help. Thecicctron micrograiphs werc taken ait the Institutc of Ophthalmologyand i saimpic ofCSFwas analyscd in thc Chemical Pathology Depairt-mcnt ait the Naitionail Hiospital., Quccns Squairc, London.

References

I Shath RR, OaitcsNS, IdlcJR, Smith RL, LockhartJDF. Impatircdoxidaition of dchrisoquinc in patients with pcrhcxilencncuropathy. Br MedJ 1982; 284: 295-9.

2 Houdcnt CE. Wolf LM. Corriait A. Livcr during perhexilenchypoglycaamia. Lancet 1977; ii: 10(28.

3 Lcwis D, Watinwright HC, Kcw MC, Zwi S. Isaacson C. Livcrdaimaigc aissocialted with perhcxilcnc mailcitc. Gut 1979; 20:186-9.

4 Laplanc D, Bousscr MG. Bouchc P. Touboul PJ. Pcripheralncuropaithics catused hy perhexilcne mailcatc. In: Ititertnationial.ss'nposiuin oni perhexilenie inaleate. Amstcrdam: ExccrptaMcdicia, 1978: 89-96.

5 Hutchinson WM, Williaims J, Gawlcr J. Papilloedemat in paticntstaiking perhcxilcnc malcitc. Br Med J 1978; i: 30)5.

6 Atkinson AB, McArcivey D, Tropc G. Paipilloedcmat aindhepaitic dysfunction apparently induced hy perhexilcnc malcatc(Pcxid) Br Heart J 198); 43: 49)- 1.

7 D'Amico DJ, Kcnyon KR. RuskinJN. Amiodaironc kcraitopaithy.A rch Ophithlalnol 198 1 ; 99: 257-6 1.

8 Anonymous. Perhexilcnc in anginai pcctoris. Drug Tlier Bull1976;14: 97-8.

9 Pcpinc CJ. Schang SJ, Bcmillair CR. Altcrattion of lcft ventricular

responses to ischaemia with oral perhcxilene. Postgrad Med J1973; 49 (suppl 3): 43-6.

It) Datcy KK, Bagri AK, Kelkar PN, Varma SR, Bhootra RK,Amin BM. Perhexilenc malcate: a new antianginal drug. Post-grad MedJ 1973; 49 (suppl 3): 75-8.

11 Stephens W, Eddy JD, Parsons LM, Singh SP. Raised intra-crainial pressure due to perhexilene malcatc. Br MedJ 1978; i: 2 1.

12 Committee on Safety of Medicincs. Perhexilene malcate (Pexid):aidversc reactions. Current Problems. August 1983; 11.

13 Bouchc P, Bousscr M-G, Peytour MA, Cathala HP. Perhexilenemaleatc and peripheral neuropathy. Neurology 1979; 29: 739-43.

14 Shigematsu 1. Suhacutc myclo-optico-neuropathy (SMON) andclioquinol. Jptn J Med Sci Biol 1975; 28 (suppl): 10)7-64.

15 Turut P, Haiche JC, Francois P, Arnott G. Les complicationsophthalmologiques du Pexid. Bull Soc Ophtalmol Fr 1977; 77:1(()3-7.

16 Singlas E, Goujet MA, Simon P. Pharmaicokinctics of perhexi-lcnc mailcatc in anginal patticnts with aind without pcripheralncuropathy. EurJ Cliii Phlarmacol 1978; 14: 195-21)1.

17 Drcnckhahn D, Lullmain-Rauch R. Drug induced retinallipidosis: differcntiall susceptibilities of pigment epithclium andncurorctina toward sevcral ampiphilic cationic drugs. Exp MolP(lthol 1978; 27: 111)3.

18 Riddell RH. Pathology of drug induced and toxic diseases.London: Churchill-Livingstonc, 1982: 472-3.

19 D'Amico DJ, Kcnyon KR. Drug-induced lipidoses of the corncaand conjunctiva. htit Ophthalmol 1981; 4: 67-76.

2(0 Kcnyon KR. Ocular ultrastructural of inherited mctaholicdiscisc. In: Goldberg F, cd. Genetic and metabolic eye disease.Boston: Little, Brown, 1974: 139-85.

21 Roscnthal AR, Kolh H, Bergsma D. Chloroquine retinopathy inrhesus monkey. Invest Ophthalmol Visual Sci 1978; 17: 1158-75.

22 Bockhardt H, Drcnckhahn D, Lullmann-Rauch R. Amiodaronc-induced lipidosis-likc altcrations in ocular tissues of rats. Albrechtvon1 Graefes Arch Klini Ophthalmol 1978; 207: 91-6.

23 Francois J. Cornca verticillata. Doc Ophthalmol 1969; 27:235-51.

24 Bron AJ. Vortex pattcrns of the corncal epithclium. Trans Oph-thalmol Soc UK 1973; 113: 455-72.

25 Fielder AR, Winder AF, Cookc ED, Bowcock SA. Arcus senilisand corneal temperature in man. In: Trevor-Roper PD, ed. Thecornea in health amid disease. London: Academic Press and RoyalSocicty of Medicine, 1981.

560

on 2 May 2018 by guest. P

rotected by copyright.http://bjo.bm

j.com/

Br J O

phthalmol: first published as 10.1136/bjo.68.8.553 on 1 A

ugust 1984. Dow

nloaded from