Optic Nerve Sheath Fenestration in Management of Anterior Ischemic Optic Neuropathy
The Role of
To the Editor.\p=m-\I was very interested to read the expedited publication in the December 1989 issue of the Archives by Sergott et al1 claiming that surgical optic nerve sheath decompression improved visual function in patients with "progressive" nonarteritic anterior ischemic optic neuropathy (NA-AION). They hypothesized that in NA-AION, "progressive visual loss after the initial ischemic event could be due to interference with rapid axoplasmic transport produced by [cerebrospinal fluid] CSF pressure within the anatomically restricted confines of the perineural optic nerve space. Drainage of the CSF might allow recovery of fast axoplasmic transport with subsequent improvement in vision." This theory raises some very important issues that need comment before they are accepted as scientifically
valid concepts.
My comments are based on the following: (1) my prospective studies on NA-AION in more than 600 patients; (2) my studies of the blood supply of the optic nerve head for the past 35 years; (3) experimental and clinical studies on the pathogenesis of AION (20 years), and the pathogenesis of optic disc edema in raised intracranial pressure and other conditions (30 years); (4) studies of the role of axoplasmic flow in optic disc edema due to various causes as well as in acute ocular hypertension; and (5) experimental studies on decompression of the optic nerve sheath in optic disc edema due to raised intracranial pressure.2 Based on this wide ex¬ perience, I raise the following questions about the hypoth¬ esis of Sergott et al.1 In NA-AION, is the visual loss due to blockage of the ax¬ oplasmic flow? Optic disc edema in raised intracranial pressure is due to axoplasmic flow stasis.3 Optic disc edema due to intracranial hypertension, even of a very severe de¬ gree, does not by itself cause visual dysfunction. When vi¬ sual dysfunction is seen in these cases, it is secondary to is¬ chemie changes4; an eye with optic disc edema blacks out briefly when the blood pressure drops. This is further evi¬ dence that the mechanism is ischemia. Axoplasmic flow plays no role in the conduction of nerve impulses. Thus, there is not a shred of evidence that axoplasmic flow stasis in optic disc edema per se causes any clinically detectable visual loss. The visual loss in AION is due to ischemia of the nerve fibers and not to axoplasmic flow stasis. Can normal CSF pressure with normal intraocular pres¬ sure produce axoplasmic flow stasis and optic disc edema? We have no evidence for this. The authors' statement that "drainage of the CSF might allow recovery of fast axoplas¬ mic transport, with subsequent improvement of vision" has no scientific basis. Axoplasmic flow stasis in NA-AION is due to ischemia and is in no way related to the CSF pressure.
Thorough neurologic investigation, including CSF pressure measurement, of patients with NA-AION reveals no such abnormality. While it is true that swollen axons (due to ax¬ oplasmic flow stasis from any cause) in the restricted space in the optic disc may develop secondary capillary vascular changes and ischemia, resulting in further ischemie dam¬ age, the important factor to combat is the primary cause of axoplasmic flow stasis, which in NA-AION is ischemia and not CSF pressure. The authors further added, "If the initial event in progressive NAION [NA-AION] is indeed second¬ ary to an ischemie, vaso-occlusive process, optic nerve sheath decompression may be effective because it decreases the perineural pressure exerted by the CSF. This pressure under normal circumstances may not interfere with optic nerve blood flow; however, in progressive NAION this subdural CSF might produce additional compromise of vascu¬ lar perfusion and impair axoplasmic transport." This state¬ ment is simply not supported by the basic hemodynamics of the optic nerve head. Decompression of the sheath of the optic nerve cannot influence ischemia of the optic disc in NA-AION because the optic disc is for the most part sup¬ plied by the peripapillary choroid. Our fluorescein fundus angiographie studies, in more than 500 patients seen during the very early stages of acute NA-AION, have overwhelm¬ ingly shown peripapillary choroidal filling defects. Thus, the authors' statement that "optic nerve sheath decom¬ pression represents a rational neurophysiologic approach" in NA-AION has no scientific validity. What does a "distended" optic nerve sheath represent? The authors stated, "At surgery, the optic nerve sheaths were distended, similar in appearance to the nerve sheath we have observed in pseudotumor cerebri.... Incision of the retrobulbar meningeal sheath resulted in the release of a
considerable amount of clear CSF such
as
observed with
pseudotumor cerebri. Incisions into multiple areas of the nerve sheath and lysis of presumed arachnoidal adhesions
within the subdural space resulted in additional CSF drainage, similar to the operative findings associated with chronic papilledema caused by pseudotumor cerebri." I studied the normal anatomy of the optic nerve sheath in 80 human subjects5 and also in vivo in more than 200 normal rhesus monkeys. I found that just behind the globe the sheath is normally bulbous and much larger in diameter than the retrobulbar optic nerve. Since the durai sheath is made up of collagen tissue and not of elastic tissue it cannot "distend abnormally" at a level of CSF pressure compatible with life. The arachnoidal adhesions are a common normal finding. The amount of CSF released on incision of the ret¬ robulbar part of the sheath depends on the capacity of the bulbous part of the sheath and on the extent of communi¬ cation between the sheath and the intracranial subarachnoid space through the optic canal region, which is highly
Downloaded From: http://archopht.jamanetwork.com/ by a Michigan State University User on 06/13/2015
variable.5 Thus, the authors have described as abnormal perfectly normal findings on the retrobulbar part of the sheath of the optic nerve. What is the natural history of visual loss in NA-AION? There is no doubt that there is a progressive form of NA-AION, but it is quite unusual; the incidence in our pro¬ spective series is far lower than that cited by Sergott et al. We have seen spontaneous improvement of vision of the type claimed by the authors in some cases, without any treatment. Over the years, we have given the option of sys¬ temic corticosteroid therapy to these patients; some of the treated patients improved, a few from finger counting visual acuity and marked visual field defects to 20/40 or better, with almost normal visual fields. The statement by Sergott et al that patients showing beneficial influence of cortico¬ steroid therapy in my previous study6 "could have had di¬ abetic papillopathy" is totally erroneous, and their claim that AION in diabetic patients is an "often self-limited condition" with visual recovery is contradicted by our large prospective study on the subject. Sergott et al, after surgery, reported visual improvement in two patients in the contralateral eye with long-standing decreased vision due to NA-AION and attributed it once again to "a decreased perineural CSF pressure around the previously damaged optic nerve which had not been operated on." As discussed above, the assumption that perineural CSF pressure has anything to do with the visual loss in NA-AION is entirely mistaken. I have seen similar visual recovery without any treatment whatsoever. Thus, the natural history of NAAION is more variable than is generally thought; a few pa¬ tients progress and a few spontaneously recover some vision. The authors may contend that they had a group of control cases in their series, but their sample size is far too small, both in the treated and control groups, to give accu¬ rate information, and, as the authors pointed out, the study was neither randomized nor double-masked. There is another confounding factor that one must keep in mind when recording visual acuity and visual fields—patients show a certain learning curve so that the first visual acu¬ ity and visual fields may not be representative of the ac¬ tual visual loss in all patients. Variability of visual func¬ tions on repeated testing in NA-AION is a not uncommon
finding. Thus, I have serious reservations about the claims of Sergott et al, and I find no scientific basis for the hypoth¬ esis they put forward to explain their claim. My biggest concern, however, is that the authors may have unwittingly and with the best of intentions opened a Pandora's box with this publication for these reasons: (1) NA-AION is a very common disease in middle-aged and elderly people. (2) Pa¬ tients who have suffered a dramatic visual loss are desper¬ ate people who will try anything, and are therefore most vulnerable. New surgical procedures always have a special glamour, and many patients with AION will happily submit themselves to such procedures, whether their visual loss is progressive or not. (3) Surgical decompression of the sheath of the optic nerve is not altogether a benign procedure; in addition to complications associated with general anesthe¬ sia, serious blinding complications are possible, especially in patients with AION who are particularly vulnerable to further ocular ischemie disorders. (4) We know that surgi¬ cal procedures are a very lucrative business. The combina¬ tion of all these factors, added to the understandable enthusiasm of even honest ophthalmologists to undertake a new procedure, could create a serious ethical problem. Although Sergott et al have stressed that surgical decom¬ pression of the sheath should be "considered only for that small group of patients with progressive visual loss," I fear
that proviso will be quickly forgotten, as ethics and caution are pushed aside by expediency and fashion. Sohan Singh Hayreh, MD, PhD, DSc, FRCS Iowa City, Iowa 1. Sergott RC, Cohen MS, Bosley TM, Savino PJ. Optic nerve decompression may improve the progressive form of nonarteritic ischemic optic neuropathy. Arch Ophthalmol. 1989;107:1743-1754. 2. Hayreh SS. Pathogenesis of oedema of the optic disc (papilloedema). Br J Ophthalmol. 1964;48:522-543. 3. Hayreh SS. Optic disc edema in raised intracranial pressure, V: pathogenesis. Arch Ophthalmol. 1977;95:1553-1565. 4. Hayreh SS. Optic disc edema in raised intracranial pressure, VI: assocated visual disturbances and their pathogenesis. Arch Ophthalmol. 1977; 95:1566-1579. 5. Hayreh SS. The sheath of the optic nerve. Ophthalmologica. 1984;189:54\x=req-\ 63. 6. Hayreh SS. Anterior ischaemic optic neuropathy, III: treatment, prophylaxis, and differential diagnosis. Br J Ophthalmol. 1974;58:981-989.
In
Reply.\p=m-\Wewould like to address each of Dr Hayreh's In NAION, is the visual loss due to blockage of the axoplasmic flow? Dr Hayreh states, and we agree, that in papilledema visual dysfunction is secondary to ischemia. Because optic nerve sheath decompression (ONSD) has been demonstrated to reverse severe visual loss in pseudotumor cerebri,1 we considered ONSD for other optic neuropathies in which ischemia may be involved. Dr Hayreh continues that "axoplasmic flow plays no role in the conduction of nerve impulses." While this statement is true in the strict neurophysiologic sense, alteration of axoplasmic flow has produced blockage of impulse conduction in several carefully studied experimental and human conditions. First, giant axonal swellings have been produced in rats by concerns.
the administration of \g=b\-\g=b\1-iminodipropionitrile,a toxin that disrupts the cytoskeletal elements of axons.2 In this model, iminodipropionitrile blocks slow axonal transport more than the fast component, resulting in massive axonal edema in the proximal region of almost all -motor neurons without segmentai demyelination or fiber loss. Both monosynaptic reflex pathway and intracellular records have re¬ vealed substantial reduction in conduction velocity, with conduction block developing 2 to 4 days after the adminis¬ tration of iminodipropionitrile. Parallel neuropathologic studies demonstrated that the terminal loops of myelin at the nodes of Ranvier maintained axonal contact but were displaced from the paranodal region to the internode. There¬ fore, in this model, a pure blocker of axoplasmic transport produced nerve impulse conduction abnormalities. In the specific circumstance of chronic experimental neu¬ ral ischemia, Sladky et al3 have produced neurophysiologic and histopathologic changes identical to the iminodipropi¬ onitrile model. Therefore, in these two circumstances, axo¬ plasmic transport and impulse conduction have been dem¬ onstrated to be inseparably related neurophysiologic pro¬ cesses. We believe that such a dynamic relationship may exist for NAION. Can normal CSF pressure with normal intraocular pres¬ sure produce axoplasmic flow stasis and optic disc edema? Dr Hayreh states that "axoplasmic flow stasis in NA-AION is due to ischemia and is in no way related to the CSF pres¬ sure." We assume that the CSF pressure measurements re¬ ferred to are those determined by lumbar puncture. How¬ ever, there is not always a direct correlation between intracranial pressure and optic nerve function as illus¬ trated by the following: (1) patients with pseudotumor cerebri may continue to lose vision despite absolutely nor¬ mal intracranial pressure and perfectly functioning lumboperitoneal shunts, and (2) ONSD reverses visual loss in pseudotumor cerebri but does not lower the intracranial
Downloaded From: http://archopht.jamanetwork.com/ by a Michigan State University User on 06/13/2015
The Role of
To the Editor.\p=m-\I was very interested to read the expedited publication in the December 1989 issue of the Archives by Sergott et al1 claiming that surgical optic nerve sheath decompression improved visual function in patients with "progressive" nonarteritic anterior ischemic optic neuropathy (NA-AION). They hypothesized that in NA-AION, "progressive visual loss after the initial ischemic event could be due to interference with rapid axoplasmic transport produced by [cerebrospinal fluid] CSF pressure within the anatomically restricted confines of the perineural optic nerve space. Drainage of the CSF might allow recovery of fast axoplasmic transport with subsequent improvement in vision." This theory raises some very important issues that need comment before they are accepted as scientifically
valid concepts.
My comments are based on the following: (1) my prospective studies on NA-AION in more than 600 patients; (2) my studies of the blood supply of the optic nerve head for the past 35 years; (3) experimental and clinical studies on the pathogenesis of AION (20 years), and the pathogenesis of optic disc edema in raised intracranial pressure and other conditions (30 years); (4) studies of the role of axoplasmic flow in optic disc edema due to various causes as well as in acute ocular hypertension; and (5) experimental studies on decompression of the optic nerve sheath in optic disc edema due to raised intracranial pressure.2 Based on this wide ex¬ perience, I raise the following questions about the hypoth¬ esis of Sergott et al.1 In NA-AION, is the visual loss due to blockage of the ax¬ oplasmic flow? Optic disc edema in raised intracranial pressure is due to axoplasmic flow stasis.3 Optic disc edema due to intracranial hypertension, even of a very severe de¬ gree, does not by itself cause visual dysfunction. When vi¬ sual dysfunction is seen in these cases, it is secondary to is¬ chemie changes4; an eye with optic disc edema blacks out briefly when the blood pressure drops. This is further evi¬ dence that the mechanism is ischemia. Axoplasmic flow plays no role in the conduction of nerve impulses. Thus, there is not a shred of evidence that axoplasmic flow stasis in optic disc edema per se causes any clinically detectable visual loss. The visual loss in AION is due to ischemia of the nerve fibers and not to axoplasmic flow stasis. Can normal CSF pressure with normal intraocular pres¬ sure produce axoplasmic flow stasis and optic disc edema? We have no evidence for this. The authors' statement that "drainage of the CSF might allow recovery of fast axoplas¬ mic transport, with subsequent improvement of vision" has no scientific basis. Axoplasmic flow stasis in NA-AION is due to ischemia and is in no way related to the CSF pressure.
Thorough neurologic investigation, including CSF pressure measurement, of patients with NA-AION reveals no such abnormality. While it is true that swollen axons (due to ax¬ oplasmic flow stasis from any cause) in the restricted space in the optic disc may develop secondary capillary vascular changes and ischemia, resulting in further ischemie dam¬ age, the important factor to combat is the primary cause of axoplasmic flow stasis, which in NA-AION is ischemia and not CSF pressure. The authors further added, "If the initial event in progressive NAION [NA-AION] is indeed second¬ ary to an ischemie, vaso-occlusive process, optic nerve sheath decompression may be effective because it decreases the perineural pressure exerted by the CSF. This pressure under normal circumstances may not interfere with optic nerve blood flow; however, in progressive NAION this subdural CSF might produce additional compromise of vascu¬ lar perfusion and impair axoplasmic transport." This state¬ ment is simply not supported by the basic hemodynamics of the optic nerve head. Decompression of the sheath of the optic nerve cannot influence ischemia of the optic disc in NA-AION because the optic disc is for the most part sup¬ plied by the peripapillary choroid. Our fluorescein fundus angiographie studies, in more than 500 patients seen during the very early stages of acute NA-AION, have overwhelm¬ ingly shown peripapillary choroidal filling defects. Thus, the authors' statement that "optic nerve sheath decom¬ pression represents a rational neurophysiologic approach" in NA-AION has no scientific validity. What does a "distended" optic nerve sheath represent? The authors stated, "At surgery, the optic nerve sheaths were distended, similar in appearance to the nerve sheath we have observed in pseudotumor cerebri.... Incision of the retrobulbar meningeal sheath resulted in the release of a
considerable amount of clear CSF such
as
observed with
pseudotumor cerebri. Incisions into multiple areas of the nerve sheath and lysis of presumed arachnoidal adhesions
within the subdural space resulted in additional CSF drainage, similar to the operative findings associated with chronic papilledema caused by pseudotumor cerebri." I studied the normal anatomy of the optic nerve sheath in 80 human subjects5 and also in vivo in more than 200 normal rhesus monkeys. I found that just behind the globe the sheath is normally bulbous and much larger in diameter than the retrobulbar optic nerve. Since the durai sheath is made up of collagen tissue and not of elastic tissue it cannot "distend abnormally" at a level of CSF pressure compatible with life. The arachnoidal adhesions are a common normal finding. The amount of CSF released on incision of the ret¬ robulbar part of the sheath depends on the capacity of the bulbous part of the sheath and on the extent of communi¬ cation between the sheath and the intracranial subarachnoid space through the optic canal region, which is highly
Downloaded From: http://archopht.jamanetwork.com/ by a Michigan State University User on 06/13/2015
variable.5 Thus, the authors have described as abnormal perfectly normal findings on the retrobulbar part of the sheath of the optic nerve. What is the natural history of visual loss in NA-AION? There is no doubt that there is a progressive form of NA-AION, but it is quite unusual; the incidence in our pro¬ spective series is far lower than that cited by Sergott et al. We have seen spontaneous improvement of vision of the type claimed by the authors in some cases, without any treatment. Over the years, we have given the option of sys¬ temic corticosteroid therapy to these patients; some of the treated patients improved, a few from finger counting visual acuity and marked visual field defects to 20/40 or better, with almost normal visual fields. The statement by Sergott et al that patients showing beneficial influence of cortico¬ steroid therapy in my previous study6 "could have had di¬ abetic papillopathy" is totally erroneous, and their claim that AION in diabetic patients is an "often self-limited condition" with visual recovery is contradicted by our large prospective study on the subject. Sergott et al, after surgery, reported visual improvement in two patients in the contralateral eye with long-standing decreased vision due to NA-AION and attributed it once again to "a decreased perineural CSF pressure around the previously damaged optic nerve which had not been operated on." As discussed above, the assumption that perineural CSF pressure has anything to do with the visual loss in NA-AION is entirely mistaken. I have seen similar visual recovery without any treatment whatsoever. Thus, the natural history of NAAION is more variable than is generally thought; a few pa¬ tients progress and a few spontaneously recover some vision. The authors may contend that they had a group of control cases in their series, but their sample size is far too small, both in the treated and control groups, to give accu¬ rate information, and, as the authors pointed out, the study was neither randomized nor double-masked. There is another confounding factor that one must keep in mind when recording visual acuity and visual fields—patients show a certain learning curve so that the first visual acu¬ ity and visual fields may not be representative of the ac¬ tual visual loss in all patients. Variability of visual func¬ tions on repeated testing in NA-AION is a not uncommon
finding. Thus, I have serious reservations about the claims of Sergott et al, and I find no scientific basis for the hypoth¬ esis they put forward to explain their claim. My biggest concern, however, is that the authors may have unwittingly and with the best of intentions opened a Pandora's box with this publication for these reasons: (1) NA-AION is a very common disease in middle-aged and elderly people. (2) Pa¬ tients who have suffered a dramatic visual loss are desper¬ ate people who will try anything, and are therefore most vulnerable. New surgical procedures always have a special glamour, and many patients with AION will happily submit themselves to such procedures, whether their visual loss is progressive or not. (3) Surgical decompression of the sheath of the optic nerve is not altogether a benign procedure; in addition to complications associated with general anesthe¬ sia, serious blinding complications are possible, especially in patients with AION who are particularly vulnerable to further ocular ischemie disorders. (4) We know that surgi¬ cal procedures are a very lucrative business. The combina¬ tion of all these factors, added to the understandable enthusiasm of even honest ophthalmologists to undertake a new procedure, could create a serious ethical problem. Although Sergott et al have stressed that surgical decom¬ pression of the sheath should be "considered only for that small group of patients with progressive visual loss," I fear
that proviso will be quickly forgotten, as ethics and caution are pushed aside by expediency and fashion. Sohan Singh Hayreh, MD, PhD, DSc, FRCS Iowa City, Iowa 1. Sergott RC, Cohen MS, Bosley TM, Savino PJ. Optic nerve decompression may improve the progressive form of nonarteritic ischemic optic neuropathy. Arch Ophthalmol. 1989;107:1743-1754. 2. Hayreh SS. Pathogenesis of oedema of the optic disc (papilloedema). Br J Ophthalmol. 1964;48:522-543. 3. Hayreh SS. Optic disc edema in raised intracranial pressure, V: pathogenesis. Arch Ophthalmol. 1977;95:1553-1565. 4. Hayreh SS. Optic disc edema in raised intracranial pressure, VI: assocated visual disturbances and their pathogenesis. Arch Ophthalmol. 1977; 95:1566-1579. 5. Hayreh SS. The sheath of the optic nerve. Ophthalmologica. 1984;189:54\x=req-\ 63. 6. Hayreh SS. Anterior ischaemic optic neuropathy, III: treatment, prophylaxis, and differential diagnosis. Br J Ophthalmol. 1974;58:981-989.
In
Reply.\p=m-\Wewould like to address each of Dr Hayreh's In NAION, is the visual loss due to blockage of the axoplasmic flow? Dr Hayreh states, and we agree, that in papilledema visual dysfunction is secondary to ischemia. Because optic nerve sheath decompression (ONSD) has been demonstrated to reverse severe visual loss in pseudotumor cerebri,1 we considered ONSD for other optic neuropathies in which ischemia may be involved. Dr Hayreh continues that "axoplasmic flow plays no role in the conduction of nerve impulses." While this statement is true in the strict neurophysiologic sense, alteration of axoplasmic flow has produced blockage of impulse conduction in several carefully studied experimental and human conditions. First, giant axonal swellings have been produced in rats by concerns.
the administration of \g=b\-\g=b\1-iminodipropionitrile,a toxin that disrupts the cytoskeletal elements of axons.2 In this model, iminodipropionitrile blocks slow axonal transport more than the fast component, resulting in massive axonal edema in the proximal region of almost all -motor neurons without segmentai demyelination or fiber loss. Both monosynaptic reflex pathway and intracellular records have re¬ vealed substantial reduction in conduction velocity, with conduction block developing 2 to 4 days after the adminis¬ tration of iminodipropionitrile. Parallel neuropathologic studies demonstrated that the terminal loops of myelin at the nodes of Ranvier maintained axonal contact but were displaced from the paranodal region to the internode. There¬ fore, in this model, a pure blocker of axoplasmic transport produced nerve impulse conduction abnormalities. In the specific circumstance of chronic experimental neu¬ ral ischemia, Sladky et al3 have produced neurophysiologic and histopathologic changes identical to the iminodipropi¬ onitrile model. Therefore, in these two circumstances, axo¬ plasmic transport and impulse conduction have been dem¬ onstrated to be inseparably related neurophysiologic pro¬ cesses. We believe that such a dynamic relationship may exist for NAION. Can normal CSF pressure with normal intraocular pres¬ sure produce axoplasmic flow stasis and optic disc edema? Dr Hayreh states that "axoplasmic flow stasis in NA-AION is due to ischemia and is in no way related to the CSF pres¬ sure." We assume that the CSF pressure measurements re¬ ferred to are those determined by lumbar puncture. How¬ ever, there is not always a direct correlation between intracranial pressure and optic nerve function as illus¬ trated by the following: (1) patients with pseudotumor cerebri may continue to lose vision despite absolutely nor¬ mal intracranial pressure and perfectly functioning lumboperitoneal shunts, and (2) ONSD reverses visual loss in pseudotumor cerebri but does not lower the intracranial
Downloaded From: http://archopht.jamanetwork.com/ by a Michigan State University User on 06/13/2015
