t;ye(1990) 4, 25-41

Anterior Ischaemic Optic Neuropathy Differe ntiation of arteritic from non-arteritic type and its management

SORAN SINGH HAYREH Iowa City, Iowa, USA

Summary Anterior ischaemic optic neuropathy (AION), a common, visually crippling dis­ order, is discussed, with particular emphasis on differentiating AION due to giant cell arteritis (arteritic AION)from that not due to it (non-arteritic AION). Giant cell arteritis is an ophthalmic emergency because of imminent danger of bilateral total blindness, which is almost always preventable if the disease is quickly identified and treated urgently and aggressively. My studies have revealed that the best means of differentiating arteritic from non-arteritic AION is a combination of information from the following: systemic and visual symptoms of giant cell arteritis, high eryth­ rocyte sedimentation rate and C-reactive protein, early massive visual loss, chalky­ white optic disc swelling, associated cilio-retinal artery occlusion, massive non-filling of the choroid on fluorescein fundus angiography and temporal artery biopsy. Management of giant cell arteritis and of arteritic AION is discussed. Current mis­ conceptions about AION are pointed out.

Anterior ischaemic optic neuropathy (AION) Iowa and have collected data on about 600 is one of the most prevalent visually crippling cases of AION. diseases in the middle aged and elderly-a fact not fully recognised by ophthalmologists Pathogenesis of AION and neurologists. AION is due to acute ischaemia of the The account of AION given in this paper is anterior part of the optic nerve, which consists essentially based on my personal experience of the optic nerve head and adjacent retro­ of the disease and not on a review of the laminar optic nerve. 1-3 This part of the optic extensive literature on the subject. I have nerve is supplied mainly by the peripapillary studied the subject of AION clinically and choroid and by some direct branches from the experimentally over the past 20 years. For the short posterior ciliary arteries, with minor and +-x past 15 years we have been prospectively differing contributions from other sources investigating various aspects of AION in our (Fig. 1). Thus, posterior ciliary artery (peA) Ocular Vascular Clinic at the University of circulation is the dominant source of blood

These studies have been supported by a research grant from the NatiOnal Institutes of Health (Grant No. BY-1151), and unrestricted grants from Research to Prevent Blindness, Inc, New York, and from the Alcon Research InstItute

Correspondence to: Soh an Singh Hayreh, MD, PhD, DSc, FRCS, Department of Ophthalmology, University Hospitals and Chnics, Iowa City, Iowa 52242, USA. 26 SOHAN SINGH HAYREH -'---- �6-� --:-::-' - '" - -< :-"..: - - NFL J R

�RA c � ) PLR s

-LC

peA P Normal

Fig. 1. Schematic representation of blood supply of the optic nerve head and retrolaminar optic nerve.

C = choroid, CRA = central retinal artery, LC = lamina cribrosa, NFL = surface nerve fibre layer of the optic disc, ON = optic nerve, P = pia, PCA = posterior ciliary artery, PLR = prelaminar region, R = retina,

RA = retinal arteriole, S = sclera. (Reproduced from Hayreh5) . supply to this region. Usually there are two or tension, collagen vascular diseases, massive/ three peAs arising from the ophthalmic recurrent systemic haemorrhages,14 valvular artery, arranged medial and lateral to the heart disease, carotid artery disease, optic nerve and designated as medial peA migraine, marked chronic optic disc oedema and lateral peA respectively.9,10 The peAs in of any causelS and raised intraocular pres­ vivo physiologically behave as end-arterieslO•11 sure,16 with giant cell arteritis and diabetes and the watershed zone between the peAs mellitus as the most important causes. may be situated anywhere between the fovea From the aetiology and pathogenesis point and the nasal border of the optic disc;6-8,10the of view, AION can usually be subdivided into location of the watershed zone plays an two types: (i) due to thrombotic or embolic important role in determining the site and lesions of the peAs or their subdivisions, and extent of optic nerve head ischaemia.6-8 The (ii) due to temporary hypoperfusion or non­ blood supply of the anterior part of the optic perfusion of the nutrient vessels in the nerve shows a marked inter-individual varia­ anterior part of the optic nerve, usually as a tion,6-8 which explains variation in the pattern result of fall in the perfusion pressure in the of visual loss associated with AION in differ­ peripapillary choroid; the second type of ent eyes, AION is by far the commonest variety. We In many patients no definite cause can be have discussed the pathogenesis of AION , in discovered for the development of AION; however, the known causes of AION include generae,3,17,18 and in specific condi­ giant cell arteritis (temporal or cranial arter­ tions,12-16,19,20 in detail elsewhere. itis)3, diabetes mellitus,12 arteriosclerosis, atherosclerosis, arterial hypertension, malig­ Clinical classification of AION nant arterial hypertension,13 arterial hypo- Based on the aetiology of AION, and from ANTERIOR ISCHAEMIC OPTIC NEUROPATHY 27 the clinical management point of view, AION Management of AION can also be classified as arteritic (caused by Once a patient in the late fifties or older is giant cell arteritis), and non-arteritic (from diagnosed to have AION, the first, most other causes )-the latter includes those cases important step is to rule out giant cell arteritis where there is no apparent cause. because of imminent danger of bilateral total blindness which is almost entirely preventable Clinical features of AION if the disease is quickly identified and treated AION usually presents with classical symp­ toms and signs, so that it is one of the easiest urgently and aggressively. In fact, Jennings,24 diagnoses to make in ophthalmology. 3,17,21 more than half a century ago, stressed that There is sudden, painless visual deterioration giant cell arteritis 'ranks as the prime medical in one eye, usually discovered on waking in emergency in ophthalmology, There being no the morning. On examination, the visual other disease in which the prevention of blind­ acuity may vary from better than 6/6 to no ness depends so much on prompt recognition light perception (Table),22 Perimetry usually and early treatment. ' reveals relative or absolute inferior altitudinal With this abbreviated introduction on defect, inferior nasal sectoral defect or central AION, I wish to deal briefly with the fol­ scotoma, and less commonly other optic disc lowing three aspects of AION: related field defects (Fig, 2),22 Fundus exam­ (AJ Differentiation of arteritic from non-arter­ ination initially shows optic disc oedema, itie AION usually with splinter retinal haemorrhage(s) 7),3.17,23 - 12e at the disc margin (Figs, 3, 4a, 5a, 6a, -14e The disc oedema usually resolves spontane­ --- V 4e ously in about two months and is replaced by sectoral or more often generalised optic atro­ phy, The fellow, normal optic disc shows either no cup or only a very small CUp,20 Fluorescein angiography during the initial phase usually shows fillingdefects in the optic disc, peripapillary choroid and choroidal watershed zones (Figs, 4b, 5b), and in giant cell arteritis much more extensive choroidal filling defects (Figs. 6b, C).3,6,17,21,23 AION usually starts as a unilateral condition but 140 __1_5 5_,,��L_��_ 300 afterdays, months or years may become bilat­ eral-the estimated 25th-percentile time to development of bilateral AION in our series was 0.4 months for untreated arteritic AION and 32.4 months for non-arteritic AION,19

Table Visual acuity in arteritic and non-arteritic AWN (from Hayreh & PodhajskJ2)

Arteritic Non-arteritic AION AION Visual acuity (per cent) (per cent)

Normal to fairly good visual acuity: 6/6 or better 14 26 6/12 or better 26 41 6/30 or better 40 63 Poor visual acuity to blind: Count fingers 14 17 Fig. 2. Visual field defects in AION with Goldmann Hand motion 11 5 perimeter: (a) showing inferior nasal quadrantic defect Light perception 6 1.5 with /4, (heavy line) and V4, (broken line) , and inferior No light perception 29 2 altitudinal defect with /2, (thin line) ; (b) showing infer­ ior altitudinal defect with all the 3 isopters, 28 SOHAN SINGH HAYREH

Fig. 3. Fundus photographs a/right eye with non-arteritic AlON showing optu; disc oedema and flame-shaped haemorrhages during the acute phase.

(B) Management of arteritic AION and giant prevent any further loss of vision. The criteria cell arteritis for such a differentiation discussed here are (C) Currently prevalent misconceptions about mainly drawn from the studies conducted in AION our clinic. In addition to the 600 patients of AION (about 75 arteritic and 525 non-arter­ itic) studied in our clinic prospectively and in (A) Differentiation of Arteritic From Non­ detail, we have also followed prospectively Arteritic AION about 30 cases of giant cell arteritis without Since giant cell arteritis is an ocular emer­ any AI ON or any other visual complaint. This gency, it is imperative to differentiate arteritic clinical material, combined with my experi­ from non-arteritic AION at the earliest poss­ ence over the past two decades, has revealed ible moment, to enable us to start cortico­ that the following criteria collectively provide steroid therapy immediately in order to information which can help to differentiate Fig. 4. Right eye with non-arteritic AION (a) fundus photograph showing optic disc oedema and a splinter haemorrhage, and (b) fluorescein fundus angiogram showing non-filling of temporal parts of the peripapillary choroid and optic disc and of upper half of the choroidal watershed zone. (Fig. 4b reproduced from Hayreh6).

Fig. 4b. Fig. 5. Right eye with non-arteritic AION (a) fundus photograph showing optic disc oedema and fine splinter haemorrhages, and (b) fluorescein fundus angiogram showing non-fillingof the choroidal watershed zone between the lateral and medial peAs and of the temporal part of optic disc. (Fig. 5b reproduced from Hayreh6).

Fig.5b. Fig.6. Left eye with arteritic AION and associated cilio-retinal artery occlusion (a) fundus photograph showing chalky-white optic disc oedema and a patch of retinal infarction in the distribution of cilio-retinal artery; (b-d) fluoresceinfundus angiograms showing normal fillingof the central retinal artery and of the choroid supplied by the lateral peA, but no fillingof the choroid and optic disc supplied by the medial peA and of the cilio-retinal artery (in b and c) , and late staining of the disc (d). (Fig. 6c reproduced from Hayreh'7) .

Fig.6b. Fig. 6c.

Fig. 6d. ANTERIOR ISCHAEMIC OPTIC NEUROPATHY 33

Fig. 7. Fundus photograph of right eye with arteritic AlON, showing chalky-white swelling of the optic disc. arteritic from non-arteritic AION almost arteritis. It is not at all uncommon in our invariably. experience to find patients who have repeat­ edly given a classical history of this disease to 1. Systemic symptoms of giant cell arteritis their physicians and yet the diagnosis was Typically the patient is sixty or older, more totally missed and the patient went blind in often female than male, and complains of one or both eyes. However, not all patients vague aches and pains, malaise, anorexia, with giant cell arteritis have systemic symp­ 'fiu'-like symptoms, weight loss, fever of toms; in occult giant cell arteritis, the patient unknown origin, headaches, scalp tender­ is perfectly fitand healthy and yet has classical ness, neck pain, jaw claudication, anaemia or evidence of giant cell arteritis. Tender, hard, other vague systemic symptoms; the patient nodular, prominent and non-pulsatile tempo­ feels tired, sleepy and generally unwell. Paul­ ral arteries are described as a classical finding ley and Hughes25 admirably summarised the in these patients; however, perfectly normal symptoms when they stated that 'When and pulsatile temporal arteries are not elderly people begin to fail mentally and uncommon in typical giant cell arteritis. physically, this disorder should be one of the These are some of the pitfalls, and it is impor­ first to be considered'. We have seen many tant to be aware of them. patients who have been subjected to exten­ sive, highly expensive and totally unnecessary 2. Visual symptoms of giant cell arteritis systemic investigations to rule out malignan­ Amaurosis fugax is an important symptom of cies and other systemic causes for their symp­ giant cell arteritis, because it is rarely noticed toms and the missed diagnosis ultimately in non-arteritic AION. The presence of resulted in blindness. I findwidespread lack of amaurosis fugax is an ominous sign of awareness among physicians of this impor­ impending AION in patients with giant cell tant, typical symptom-complex of giant cell arteritis, and these eyes are at risk of develop- 34 SOHAN SINGH HAYREH ing blindness very quickly. Typically, there is a cell arteritis and in their follow-up on systemic sudden, painless, spontaneous and perman­ corticosteroids. A normal CRP level (:::;1mg) ent visual loss. A few patients may complain with equivocal ESR strongly indicates of diplopia. Some patients with giant cell absence of giant cell arteritis or good control arteritis are somewhat euphoric, so that in with treatment, even when ESR is elevated; spite of bilateral total blindness, the patient however, a markedly elevated CRP with so­ may seem not to be very concerned about the called 'normal ESR' would suggest giant cell visual loss and may even deny blindness. arteritis. In some previous studies28 on CRP in giant cell arteritis, the results were given 3. High erythrocyte sedimentation rate (ESR) simply as 'positive' or 'negative' CRP and and C-reactive protein (CRP) conclusions based on those could be mislead­ These are two simple, quick, inexpensive and ing unless exact levels of CRP are given. readily available tests. Although both tests are non-specific,levels are classically elevated 4. Early massive visual loss in giant cell arteritis. In arteritic AION there is usually early mas­ If the ESR (as measured by Westergren sive visual loss, so that the eye has either no method) is markedly elevated, e.g. about light perception or can barely see light or hand 80 mm or more, this is highly suggestive of motion. In contrast to that in non-arteritic giant cell arteritis; however, our experience AION the visual loss is usually not so severe has shown that the so-called 'normal ESR' in to begin with. Some time ago we analysed our no way rules out this disease. This is because in data on the initial visual acuity in the two our experience, 'normal ESR' varies from one types of AION (Table). 22 This revealed that in to over 40 mm in the age-popUlation at risk of arteritic AION there was no light perception developing giant cell arteritis, and we could in 29%, hand motion or worse vision in 46%; find no haematologic abnormality to explain on the other hand, in non-arteritic AION such a wide variation. I have seen AION with hand motion or worse vision was in only 9%. active giant cell arteritis (confirmed on tem­ Thus, early massive visual loss is highly sug­ poral artery biopsy) where the ESR was S mm gestive of arteritic AION but normal visual with the Westergren method. It is very impor­ acuity does not rule this out (Table). tant to bear in mind that there is a tremendous variation in the normal levels of ESR in the S. Chalky-white optic disc swelling elderly and this has not been adequately If the optic disc swelling is chalky-white stressed in the literature, resulting in missing initially (Fig. 7), it is extremely suggestive of giant cell arteritis; ESR in the teens, twenties arteritic AION because this is exceedingly or even thirties has often been dismissed as rare in non-arteritic AION.3,23 However, 'normal'. This serious pitfall regarding ESR about 50% of eyes with arteritic AION do not should be kept in mind. show this and ophthalmoscopic disc appear­ For the past few years, in addition to ESR, ance is similar to that of non-arteritic AION we have used CRP to evaluate and manage (Fig. 3, 4a, Sa). these patients. A rise in CRP level in serum and other body fluids is produced by inflam­ 6. AlON associated with cilio-retinal artery matory processes in the body.26 Usually occlusion changes in CRP correlate with the ESR; how­ In our experience, when optic disc oedema of ever, in some aspects it behaves differently AION is initially associated with cilio-retinal from ESR. CRP does not rise during some artery occlusion (Fig. 6), it is almost always viral diseases, some severe toxic states, and in diagnostic of arteritic AION; I have yet to see some forms of chronic inflammatory arthri­ this in a non-arteritic AION . 3,23 In giant cell tis,27 when ESR does rise. Also, we have arteritis the primary lesion responsible for found much less fluctuation in CRP than in AION is thrombotic occlusion of the PCA. If ESR. In our experience, combined infor­ a PCA is occluded, it infarcts not only the area mation from CRP and ESR is very useful in of the anterior optic nerve supplied by it diagnosis and evaluation of patients with giant (Fig. 1) but also cuts out blood flow in the ANTERIOR ISCHAEMIC OPTIC NE UROPATHY 35 corresponding cilio-retinal artery. In contrast pulse amplitude and calculated ocular blood to that, our fluorescein fundus angiographic flow were reduced as compared to the studies have consistently shown that non­ normals and non-arteritic AION. This is not arteritic AION is almost always due to tran­ surprising; this is just another way to evaluate sient non-perfusion or hypoperfusion of the the peA occlusion caused by giant cell arter­ nutrient vessels in the anterior part of the itis. However, unfortunately the authors did optic nerve because of a fall in perfusion pres­ not perform fluorescein fundus angiography sure in the peripapillary choroid, and not due and provided no information on the state of to peA occlusion6,17,21 (Fig. 4b, 5b). In spite of the choroidal circulation on angiography in my repeatedly stressing the latter fact over the any of their patients. Four of the 11 eyes years, I am constantly being misquoted by reported with visual loss in this series had cen­ various authors as stating that non-arteritic tral retinal artery occlusion and the remaining AION is due to peA occlusion-this is simply seven had AION. In my anatomical studies I not so in the vast majority of cases. That is have shown that the central retinal artery and why in non-arteritic AION there is no asso­ one or more peAs may arise by a common ciated occlusion of the cilio-retinal artery. trunk from the ophthalmic artery in 60% of human cases.9,30 In giant cell arteritis patients, I have seen eyes which present with what, 7. Massive choroidal non-filling of the ophthalmoscopically, looks like central ret­ choroid on fluorescein fundus angiography inal artery occlusion, but in fact, on fluores­ If angiography is performed during the first cein fundus angiography, turns out to have days after the onset of arteritic AI ON, it peA occlusion too, because of the thrombosis almost always shows massive choroidal non­ of their common trunk by giant cell arter­ filling, corresponding to the occluded peA itis. 3,23 It is well known that giant cell arteritis (Fig. 6b, C).3,6,17,21,23 This is almost always has a special predilection to involve peAs. diagnostic of arteritic AION. In our experi­ Bosley et al. 29 reported that ocular pulse ence the medial peA is involved much more amplitude improved with the passage of time frequently than the lateral peA by giant cell and (erroneously) attributed this to cortico­ arteritis (Fig. 6b, c). The area supplied by the steroid therapy; our fluorescein angiographic medial peA shows a marked inter-individual studies, in patients with arteritic AION and in variation, so that it may supply the entire rhesus monkey eyes after experimental peA nasal choroid up to the foveal region, includ­ occlusion, have consistently shown that the ing the entire optic nerve head (Fig. 6b, c), or choroidal vascular bed in the involved area it may stop short nasal to the medial peri­ starts to fill progressively with the passage of papillary choroid so that the optic nerve head time because of establishment of collateral is supplied entirely by the lateral peA; or any circulationll which has nothing to do with variation between the two extremes.6-8 With corticosteroid therapy. Since ocular pneumo­ the passage of time after the onset of arteritic plethysmography essentially evaluates the AION, we have found that collateral circula­ choroidal circulation, other conditions caus­ tion in the choroid progressively results in fill­ ing choroidal insufficiency would give false ing of the involved choroid so that fluorescein positive results; the authors reported eight angiography performed a week or more after false positive results in non-arteritic AION. I the onset may not show any such massive fill­ find that fluorescein fundus angiography, a ing defect. In non-arteritic AION, however, safe, routine and readily available test, such massive choroidal non-filling is usually gives much better and more detailed extremely rare.6 information on the peA, choroidal and optic Recently, Bosley et al. 29 described the role disc circulations in giant cell arteritis than of ocular pneumoplethysmography in the ocular pneumoplethysmography is likely to diagnosis of giant cell arteritis. They studied do. Moreover, there are some contraindi­ nine patients with giant cell arteritis (with cations to doing ocular pneumoplethysmo­ abnormal temporal artery biopsy) and visual graphy in persons of this age group. loss (uniocular in seven and binocular in two). 8. Temporal artery biopsy They stated that in giant cell arteritis ocular This must be performed in every patient having 36 SOHAN SINGH HAYREH a high index suspicion of giant cell arteritis or strong clinical suspicion for giant cell arteritis of arteritic AIO N, to confirm and establish the we biopsy the second side if one side shows no diagnosis even though all other parameters granulomatous lesion. Biode6 found no case may be classical of the disease. This mor­ of false negative biopsy in his series. Hedges phological documentation of giant cell arter­ et at.37 found it in 5% of their cases, and Hall itis is essential to justify committing a patient et at.38 in 6%. In my experience in our clinic, I to protracted systemic steroid therapy, in view saw four such patients during the early stages of the possible serious complications of such a of our study who met all the remaining criteria long-term treatment and to safeguard against for arteritic AION mentioned above, and possible litigation emerging from such com­ whose therapeutic response and follow-up plications. Temporal artery biopsy should be course was exactly like giant cell arteritis, but performed as soon as possible, but there is who had a negative biopsy on both sides; at absolutely no justification to withhold that time we were not very particular about systemic corticosteroid therapy until biopsy is removing a large piece of the temporal artery done, because giving such treatment before and doing serial sectioning. The high inci­ the biopsy does not alter the outcome of the dence of false negative temporal artery biopsy biopsy. reported in some of the papers in the liter­ Since Birkhead et al.31 in 1957 pointed out ature may be due either to removing only a that involvement of the temporal artery by small segment of the artery, not examining the giant cell arteritis may be focal and segmental, entire piece by serial sectioning, or problems the topic of 'skip lesions' in temporal artery in in defining giant cell arteritis cases in the first this disease has been raised from time to time. place. Thus, I feel temporal artery biopsy, if Cohen and Smith32 did not find any example done properly and examined correctly, has of 'skip lesions' in their study of temporal the highest sensitivity and specificity for artery biopsies from 42 patients. A number of detection of giant cell arteritis. studies, however, have shown the presence of , 'skip areas 33,34 strongly indicating that such Conclusions areas do exist. This is obviously of great clini­ Our experience shows that none of the para­ cal importance because examination of a 'skip meters mentioned above has 100% reliability area' alone, showing no granulomatous in differentiating arteritic AION from non­ lesions, would give a false negative result and arteritic AION. Sometimes there are false cause the physician to miss the diagnosis of negative or even false positive findings. Com­ giant cell arteritis. To minimise the risk of a bined information from all the parameters, false negative temporal artery biopsy and to however, helps make such a differentiation in diminish the chance of erroneous information almost all cases. from a 'skip area', our routine is to remove a long piece of the temporal artery and our (B) Management of Arteritic AION and pathologist serially sections it. He has dis­ Giant Cell Arteritis covered skip areas located both circumfer­ There is universal agreement that arteritic entially and longitudinally but serial AION should be treated with systemic sectioning of a long segment of the artery corticosteroids as an emergency, to prevent removed at biopsy reduces markedly the any further visual loss. However, there is a danger of a false negative biopsy. Studies by good deal of controversy and variation in the Wells et at. 35 from our Department have literature in the therapeutic regimen advo­ shown that direct immunofluorescencemicro­ cated in these cases. I shall describe here only scopy of temporal artery biopsy sections is not what is our practice, which has evolved from more sensitive or specific than light micro­ my experience of dealing with the disease for scopy, thus they questioned its routine use as a the past 20 years. diagnostic test. In a rare case, however, they If there is a reasonable index of suspicion found immunofluorescencemicroscopy iden­ that the patient has giant cell arteritis, with or tifiedtemporal arteritis despite negative find­ without visual loss, we start the patient on ings by light microscopy. In patients with a systemic corticosteroids immediately as a ANTERIOR ISCH AEMIC OPTIC NEUROPATHY 37 prophylaxis, even before doing temporal but to stall at that dosage for a much longer artery biopsy or getting its results. This is time than usual. important; it may be unsafe to wait till the Determination of the maintenance dose is a biopsy diagnosis is available, since the patient slow, laborious and painstaking job over could further lose vision during this interval. many months. The guiding principle in deter­ If the temporal artery biopsy result and other mining the dosage is to titrate the dosage of evidence indicates that there is no giant cell prednisone with the levels of ESR and CRP, arteritis, the drug can be stopped. The dosage the object being to have the lowest levels of and mode of its administration depends upon ESR and CRP with the lowest level of pred­ the severity of visual loss and systemic mani­ nisone. There is no formula or any other way festations. If a patient presents with no visual to predict the maintenance dosage required symptoms or visual loss, we start with at least by a particular patient; it is only by trial and 80 mg of oral prednisone daily or even up to error that one can determine this. Once a 120 mg, depending upon the severity of maintenance dosage is established, the systemic symptoms, ESR and CRP. We do, patient is followed much less frequently while however, give megadose corticosteroid ther­ on that dose. Our experience of following apy intravenously to start with on an emer­ these patients for up to 15 years has shown gency basis when a patient with symptoms of that they require maintenance dosage for giant cell arteritis has any one of the following years-the concept that giant cell arteritis ocular signs or symptoms: (i) visual symptoms burns itself out after about 9-12 months is not (particularly amaurosis fugax), (ii) a complete at all true and is based more on clinical loss of vision in one eye, or (iii) early evidence impression than on any valid scientific data. of second eye involvement (amaurosis fugax, Alternate-day corticosteroid therapy is fre­ visual field defect, optic disc oedema, cilio­ quently advocated and even practised for the retinal artery occlusion, or sluggish circula­ management of giant cell arteritis. I find that tion in the central retinal artery). Our mega­ there is a good deal of misunderstanding of dose therapy consists of giving usually the this therapy. Fauci39 has reviewed the subject equivalent of one gram of prednisone slowly admirably. He rightly stated that 'The primary by intravenous drip, every 6-8 hours and 3-4 goal in the clinical use of corticosteroids' 'is to such doses. This is followed by 120 mg pred­ balance maximal therapeutic efficacy with a nisone orally daily. We estimate ESR and minimum of deleterious side effects.'Accord­ CRP daily initially to evaluate the response of ing to him, alternate-day corticosteroid ther­ the patient to the therapy. ESR and CRP apy 'will not interfere with the normal initially show a rapid drop-the response ACTH-cortisol cycle, nor will it expose the varies a good deal in different patients. We tissues to sustained levels of hormone keep the patient on the high doses of pred­ throughout the day'. However, the question nisone till the marked and rapid initial fall in arises whether such a regimen can effectively ESR and CRP changes to a mild downfall control active inflammatoryprocess such as in (Fig. 8); following this we slowly start to taper giant cell arteritis? Fauci3Y states that 'it is down the dosage of prednisone in steps of quite difficult, if not impossible, to satisfac­ about 10 mg every 3-4 days. The response is torily induce a remission with alternate-day judged by repeated ESR and CRP, at least corticosteroid therapy in a fulminant or once a week till the dosage is about 60 mg grossly active disease'. All the available evi­ daily. No generalisation is possible, because dence seems to indicate that such a regimen of the wide variation in response by different has no place in management of active giant patients to the treatment from this point on. cell arteritis. Obviously the next question is Some patients respond dramatically, and one whether such a therapy can maintain suppres­ can go down with the prednisone dosage sion of inactive giant cell arteritis? Available faster than in others who respond very slowly. evidence once again seems to indicate that I have seen some patients in whom any reduc­ this is most probably not possible. Moreover, tion of dosage to less than 80 mg results in a the foremost tactical issue is converting from rise of ESR and CRP, and one has no option daily to alternate-day therapy. One cannot do 38 SOHAN SINGH HAYREH

ESR IN MM. 100

­ - - - . .---_ .

o 1 2 3 4 5 6 7 8 910111217242731394550677590 DAYS SINCE START OF STEROID THERAPY 1 -- 2 --- 3 -- 4 -- 5 - 6 ESR IN ARTERITIC AION

Fig. 8. Graph of (Westergren) erythrocyte sedimentation rates of six patients with arteritic AION, showing their initial responses to high doses of systemic corticosteroid therapy.

it immediately. Fauci39 advocates that in such spite of the aggressive treatment and even a conversion, one should immediately double mega dose intravenous corticosteroid therapy, the dose on the 'on' days, preferably as a did suffer further loss of vision, even complete single morning dose, and the dose on the 'off' blindness, but none after that. We, therefore, days is then gradually tapered off. He advises feel that one can virtually guarantee that if a close supervision of the patient during this person does not suffer any further visual loss conversion phase because of the risk of for one week after starting treatment with the flare-up of inflammatory process. According high doses suggested above, the chance of to him 'the process of conversion may be diffi­ visual loss after that is practically none. This is cult, at times discouraging and, in some cases, proof of the effectiveness of adequate impossible'. I hope this places the alternate­ systemic corticosteroid therapy in the preven­ day steroid therapy in its true perspective and tion of visual loss in giant cell arteritis. eliminates all the misunderstandings about it. (2) It has been claimed by some authors, from We have not used alternate-day therapy. time to time, that aggressive treatment with The objectives of corticosteroid therapy: In systemic corticosteroids can restore visual loss the management of these patients with giant in arteritic AlaN. Our experience does not cell arteritis, we have found it helpful to support this view at all, because, on the explain at length at the very beginning the fol­ whole, we have not seen any worthwhile lowing objects of treatment: visual recovery in spite of early, aggressive, (1) The principal object of the treatment is to megadose steroid therapy. Only a very rare prevent blindness. Our experience indicates patient has shown some visual improvement. that during the first week of the therapy one I have frequently found that patients have the should give a guarded prognosis about any impression that aggressive steroid therapy is further visual loss because we have found that really meant to reverse their visual loss. To about 5% of patients during this period, in avoid serious disappointment, it is essential to ANTERIOR ISCHAEMIC OPTIC NEUROPATHY 39 stress at the outset that the object of the treat­ was instead diagnosed as optic neuritis. We ment is not to recover the lost vision but to have found that AION is a disease of all ages prevent any further visual loss-if the patient and no age is immune from it. is not aware of this fact, he/she is likely to co­ operate poorly with continued treatment and 2. Eyes with AION must be blind or almost thereby to risk losing further vision. blind (3) The most important point to stress is that In our studies in non-arteritic AION, 41 % of the treatment is essentially suppressive and the eyes had 6112 or better visual acuity (26% not curative, and will be required for years to had 6/6) and 63% had 6/30 or better (Table). prevent any further visual loss. This infor­ Even in arteritic AION, the visual acuity was mation is vital for regular compliance of the 6112 or better in 26% (14% had 6/6) and 40% treatment, without any break and as directed. had 6/24 or better (Table)22. Thus, these eyes At the same time it is essential to discuss in do not have to be stone blind. detail the various complications of prolonged steroid therapy and how to avoid some of 3. AION always has pale optic disc oedema them or manage them, so that the patient is Our experience has shown that during the not only aware of those but also emotionally very early stages, optic disc oedema in non­ and physically prepared for them. It is also arteritic AION on ophthalmoscopy may look vital to involve their local physician in moni­ no different from any other type of optic disc toring and management of systemic side­ oedema, and the disc may be even hyper­ effects of corticosteroid therapy. aemic. It is only after a couple of weeks or so I have found that most ophthalmologists that the disc starts to show pallor with and physicians have either distorted or inade­ oedema. Presumably this misconception quate information on the management of crept in because the published reports have patients with giant cell arteritis, with or with­ been from referral centres where the patients out visual loss. The management of this were usually not seen early enough. In our disease is very taxing both for the patient and series, about half of the eyes with arteritic for the physician. AION even showed a pattern similar to that Unfortunately, we have no effective and seen in non-arteritic AION, and only about well-established therapy for non-arteritic half of the eyes with arteritic AION really AION-either for visual recovery or for showed pale oedema of the disc from the very prophylaxis against the involvement of the start. second eye, the incidence of which is quite high, particularly in diabetics. 19 4. Asymptomatic optic disc oedema is not seen in AION (C) Misconceptions about AI ON My studies18 have shown conclusively that A number of serious misconceptions have asymptomatic optic disc oedema may precede developed over the years about AION, and the visual loss in AION. I have collected over these are responsible for misdiagnosis of 25 such cases over the years. AION and consequent unnecessary and expensive investigation in those patients. It is 5. Neural ischaemia is an 'allor none' important to comment briefly on these mis­ phenomenon conceptions, which include the following: It is well-known that neural ischaemia can vary from sub-clinical to an extreme type. 1. AION is exclusively a disease of the Consequently, the ischaemic damage to the elderly optic nerve head can vary from sub-clinical18 This is because it is assumed that ischaemic to total infarction and the clinical manifesta­ disorders do not occur in young persons. tions would differ accordingly. Based on this misconception, in the previous studies on AION reported in the literature, 6. AION is an early complication of patients younger than 50 or so were excluded hypertensive arterial disease from the studies so that AION in those cases This misconception is based on a report by 40 SOHAN SINGH HAYREH

11 Ellenberger40 and his theory was based on a Hayreh SS and Baines JAB: Occlusion of the pos­ retrospective review of records of only 18 terior ciliary artery I. Effects on choroidal cir­ culation. Br J OphthalmoI1972, 56: 719-35. cases of non-arteritic AION! Our prospec­ 1 2 Hayreh SS and Zahoruk RM: Anterior ischemic tive, detailed studies on about 525 patients of optic neuropathy VI. In juvenile diabetics. Oph­ non-arteritic AION lends no support at all to thalmologica 1981, 182: 13-28. 1 this view. 3 Hayreh SS, Servais GE, Virdi PS: Fundus lesions in malignant hypertension V. Hypertensive optic 7. Non-arteritic AION patients with systemic neuropathy. Ophthalmology 1986, 93: 74--87. 1 hypertension have increased risk of 4 Hayreh SS: Anterior ischemic optic neuropathy cerebrovascular accidents and myocardial VIII. Clinical features and pathogenesis of post­ hemorrhagic amaurosis. Ophthalmology 1987, infarction 94, 1488-502. 1 This opinion was based on a retrospective 5 Hayreh SS: Optic disc edema in raised intracranial review of records of 212 patients with non­ pressure VI. Associated visual disturbances and arteritic AION.41 This has created a certain their pathogenesis. Arch Ophthalmol 1977, 95: 1566-79. amount of unnecessary fear in AION 1 patients. From a prospective follow-up of 6 Hayreh SS: Anterior ischemic optic neuropathy IV. Occurrence after cataract extraction. Arch Oph­ about 525 patients in our Clinic up to 15 years, thalmol1980, 98: 1410-16. 1 I have so far detected no support for this view. 7 Hayreh SS: Anterior ischemic optic neuropathy. Arch Neuro11981, 38: 675-8. 1 I am grateful to my wife Shelagh for her help with the 8 Hayreh SS: Anterior ischemic optic neuropathy V. manuscript, to Mrs. Ellen Ballas and Mrs. Georgiane Optic disc edema an early sign. Arch Ophthalmol Parkes-Perret for their secretarial help, and to the 1981,99: 1030-40. photography department for the illustrations. 1 9 Beri M, Klugman MR, Kohler JA, Hayreh SS: Anterior ischemic optic neuropathy VII. Inci­ References 1 dence of bilaterality and various influencing fac­ Hayreh SS and Baines JAB: Occlusion of the pos­ tors. Ophthalmology 1987, 94: 1020-8. terior ciliary artery III. Effects on the optic nerve 0 2 Beck RW, Servais GE, Hayreh SS: Anterior head. Br J OphthalmoI1972, 56: 754--64. ischemic optic neuropathy IX. Cup-to-disc ratio 2 Hayreh SS: Anterior ischaemic optic neuropathy I. and its role in pathogenesis. Ophthalmology1987, Terminology and pathogenesis. Br J Ophthalmol 94: 1503-8. 1974, 58: 955-63. 1 2 Hayreh SS: Ischemic optic neuropathy. Int Ophthal­ 3 Hayreh SS: Anterior ischemic optic neuropathy. mol1978, 1: 9-18. New York. Springer-Verlag, 1975. 22 Hayreh SS and Podhajsky P: Visual field defects in 4 Hayreh SS: Pathogenesis of visual field defects­ anterior ischemic optic neuropathy. Doc Ophthal­ Role of the ciliary circulation. Br J Ophthalmol mol Proc Ser 1979, 19: 53-71. 1970, 54: 289-311. 23 Hayreh SS: Anterior ischaemic optic neuropathy II. 5 Hayreh SS: Structure and blood supply of the optic Fundus on ophthalmoscopy and fluorescein nerve. In Heilmann K, Richardson KT eds. Glau­ angiography. Br J Ophthalmol1974, 58: 964--80. coma: Conceptions of a disease. Pathogenesis, 24 Jennings GH: Arteritis of the temporal vessels. diagnosis, therapy, Stuttgart: Thieme 1978: Lancet 1938, 1: 424--8. 78-96. 25 Paulley JW and Hughes JP: Giant-cell arteritis, or 6 Hayreh SS: Inter-individual variation in blood arteritis of the aged. Br Med J 1960, 2: 1562-7. supply of the optic nerve head. Its importance in 26 Hedlund P: Clinical and experimental studies on various ischemic disorders of the optic nerve C-reactive protein (Acute phase protein). Acta head, and glaucoma, low-tension glaucoma and Med Scand SuppI1961, 361: 1-71. allied disorders. Doc Ophthalmol 1985, 59: 27 Lawlor GJ and Fischer TJ: Manual of allergy and 217-46. immunology-diagnosis and therapy. Boston. 7 Hayreh SS: Blood supply of the anterior optic nerve. Little Brown & Co, 1984: 277. In Ritch R, Shields MB, Krupin T eds. The Glau­ 28 Eshagian J and Goeken JA: C-reactive protein in comas, Vol. I. St Louis: CV Mosby 1989: 133-61. giant cell (cranial, temporal) arteritis. Ophthal­ 8 Hayreh SS: Blood supply of the optic nerve head in mology 1980, 87: 1160-6. health and disease. In Lambrou GN, Greve EL 29 Bosley TM, Savino PJ, Sergott RC, Eagle RC, eds. Ocular blood flowin glaucoma. Amsterdam: Sandy R, Gee W: Ocular pneumoplethysmo­ Kugler & Ghedini 1989: 3-48. graphy can help in the diagnosis of giant-cell 9 Hayreh SS: The ophthalmic artery III. Branches. Br arteritis. Arch Ophthalmol1989, 107: 379-81. 0 J Ophthalmol1962, 46: 212-47. 3 Singh S and Dass R: The central artery of the retina 10 Hayreh SS: Physiological anatomy of the choroidal I. Origin and course. Br J Ophthalmol1960, 44: vascular bed. Int Ophthalmol1983, 6: 85-93. 193-212. ANTERIOR ISCHAEMIC OPTIC NEUROPATHY 41

31 Birkhead NC, Wagener HP, Shick RM: Treatment 36 Blodi FC: The temporal artery biopsy as a diagnostic of temporal arteritis with adrenal corticosteroids: procedure in ophthalmology. Trans Aust Coli results in fifty-five cases in which lesion was Ophthalmol1969, 1: 26-33. proved at biopsy. JAm Med Assoc 1957, 163: 37 Hedges TR III , Gieger GL, Albert OM: The clinical 821-7. value of negative temporal artery biopsy speci­ 32 Cohen DN and Smith TR: Skip areas in temporal mens. Arch Ophthalmol1983, 101: 1251-4. arteritis: myth versus fact. Trans Am Acad Oph­ 38 Hall S, Lie JT, Kurland LT, Persellin S, O'Brien PC, thalmol Otolaryngol1974, 78: OP772-83. Hunder GG: The therapeutic impact of temporal 33 Albert DM, Ruchman MC, Keltner JL: Skip areas in artery biopsy. Lancet1983, 2: 1217-20 . temporal arteritis. Arch Ophthalmol 1976, 94: 39 Fauci AS: Alternate-day corticosteroid therapy. Am 2072-7. J Med 1978, 64: 729-31. 34 Klein RG, Campbell RJ, Hunder GG, Carney JA: 4{) Ellenberger C: Ischemic optic neuropathy as a poss­ Skip lesions in temporal arteritis. Mayo Clin Proc ible early complication of vascular hypertension. 1976,51: 504--10 . Am J Ophthalmol1979, 88: 1045-51. 1 35 Wells KK, Folberg R, Goeken JA , Kemp JD: Tem­ 4 Guyer DR, Miller NR, Auer CL , Fine SL: The risk poral artery biopsies correlation of light micro­ of cerebrovascular and cardiovascular disease in scopy and immunofluorescence microscopy. patients with anterior ischemic optic neuropathy. Ophthalmology 1989, 96: 1058-64. Arch Ophthalmol1985, 103: 1136-42.