Ординатура / Офтальмология / Английские материалы / Retinal Vascular Disease_Joussen, Gardner, Kirchhof_2007

but not MTHFR C677T mutation in patients with retinal artery occlusion. Am J Ophthalmol 134:57 – 61

67.Williams GA, Sarrafizadeh R (2000) Correspondence to Antiphospholipid antibodies and retinal thrombosis in patients without risk factors: a prospective case-control study. Am J Ophthalmol 130:538 – 539

68.Williamson TH et al (1996) Blood viscosity, coagulation and activated protein C resistance in central retinal vein occlusion: a population controlled study. Br J Ophthalmol 80:203 – 208

Core Messages

Retinal vein occlusion is the most frequent primary vascular disorder of the retina

Central retinal vein occlusion (CRVO) is a more subacute chronic rather than an acute entity that occurs at all ages

Cardiovascular risk factors, some forms of thrombophilia, and hyperviscosity may trigger and facilitate the development of CRVO Visual outcome is variable, ranging from benign course to painful blinding

Ischemic (non-perfused) CRVO should be differentiated from non-ischemic (perfused) CRVO, representing entities at the opposite ends of the range

Medical treatment (hemodilution, fibrinolysis) has been tried with some success

Newer modalities for CRVO treatment (surgical, newer steroids) are under investigation but have not yet yielded evidence-based results Neovascular disease cannot be prevented in all cases by panretinal photocoagulation

21.2.1 Background

Retinal vein occlusion is the most frequent primary vascular disorder of the retina and of all the retinal vascular diseases second only to the more common diabetic retinopathy. The first description of a case with central retinal vein occlusion dates from 1855 [109], but it was not until 1878 that von Michel [123] interpreted the finding as central vein thrombosis. The term central retinal vein occlusion (CRVO) is preferred nowadays as we still do not completely understand the pathogenesis of the vein blockage in the optic nerve head. In addition to CRVO, we know the less frequent occlusion of the superior or inferior retinal venous vasculature as hemiretinal vein occlusion (HRVO). The occlusion of a retinal branch vein (BRVO) is slightly more often seen than CRVO and is addressed in Chapter 21.3.

Furthermore, much confusion existed in the past regarding the varieties of CRVO, and terms such as “impending,” “incipient,” “partial,” “incomplete CRVO,” “venous stasis retinopathy,” and “hemorrhagic retinopathy” [64, 66, 97] were used to describe courses ranging from very benign to disastrous. Today the most adequate and widely used designation is “non-ischemic” or “perfused” for the milder and “ischemic” or “non-perfused” for the severe form [68, 165], both representing entities at the opposite ends of the range. The adjunct “indetermi-

nate” may be employed for cases that do not fit easily into the categories mentioned.

21.2.2 Epidemiology

Essentials

Incidence of retinal vein occlusions varies from 2 to 8 per 1,000 persons

CRVO can occur at all ages, with a mean age between 60 and 70 years

The incidence of retinal vein occlusions in popula- tion-based studies varies from 2 to 8 per 1,000 persons and increases with age [23, 96]. CRVO can occur at all ages from 9 months to more than 90 years, but the mean age of patients is between 60 and 70 years. About 10 % are younger than 50 years [10]. It affects males and females equally; yet among patients below 50 years, men prevail [38, 136, 149]. Both eyes are involved equally. Five to 11 % of patients will suffer from CRVO in their other eye within 5 years [10, 75].

21.2.3 Etiology and Pathogenesis

Essentials

Pathogenesis can be represented in three circles: (1) triggering of the venous outflow reduction by narrowing of the vein and par-

21 III tial thrombosis, (2) development of the retinopathy, and (3) transition into severe neovascular disease

The extent of the blood-retinal barrier’s breakdown and ischemia define the course of the retinopathy

Triggering mostly happens at nighttime in the recumbent position probably by low blood pressure, and/or high central venous pressure

Cardiovascular risk factors such as arterial hypertension, arteriosclerotic cardiovascular disease, diabetes mellitus, obesity, hyperlipidemia, hyperhomocysteinemia, and smoking are associated with CRVO

Hyperviscosity and thrombophilia seem to facilitate the development of CRVO

Local risk factors are glaucoma, trauma, and possibly inflammation

Table 21.2.1. Risk factors and associated diseases in CRVO

CRVO’s pathogenesis certainly is multifactorial. Klien [97] speculated that three factors could lead to CRVO: (1) compression by a sclerotic central retinal artery and cribriform plate, (2) hemodynamic disturbances leading to stagnation and primary thrombus formation, and (3) degenerative and/or inflammatory disease within the vein. All these causes must still be considered, supplemented by systemic factors such as thrombophilia and rheological abnormalities [27].

When considering important factors for the triggering, development, and course of CRVO, one must diligently separate the actual cause from the facilitating risk factors and associated diseases (Table 21.2.1) that may be important in the development of the disease and, hence, its prognosis. For instance, a genuine causal relationship for cardiovascular, thrombophilic or rheological risk factors may be difficult to prove particularly as many of these factors have a high incidence in patients without CRVO as well.

21.2.3.1 Systemic Risk Factors

Cardiovascular risk factors (arterial hypertension, cardiovascular disease, diabetes mellitus, obesity, hyperlipidemia, hyperhomocysteinemia, smoking) are associated with the development and course of CRVO [10, 31, 73, 122, 156, 164]. At least one systemic risk factor is seen in two-thirds of patients over 50 years of age, often more frequently in ischemic CRVO than in its non-ischemic variant [49, 61]. The odds ratio for these factors was calculated by Wong et al. [188] and ranges between 3 and 5. Arterial hypertension (32 – 70 %), atherosclerotic heart disease (22 – 50 %), hyperlipidemia (30 – 60 %) and diabetes mellitus (14 – 34 %) are most commonly found. It is worth noting that Doppler imaging of the carotid artery did not reveal a higher percentage of occlusive disease than in age matched controls, although ischemic CRVO is more often associated with carotid artery disease [11, 61, 143, 188] than is the non-ische- mic type.

Cardiovascular risk factors do not seem to be important in younger patients under 50 years of age. Only one-third suffer from arterial hypertension, and 3 – 9 % have diabetes mellitus [10]. Nevertheless, Priluck [133], in a long-term follow-up of young adults with CRVO, observed a mortality of 12 % due to vascular diseases. Migraine and mitral valve prolapse have been encountered in these patients also, although an association remains obscure [38].

Hyperviscosity of the blood caused by elevated cells and plasma proteins (e.g., leukemia, polycythemia, macroglobulinemia, myeloma) can clearly lead to bilateral venostasis with all the features of CRVO (hyperviscosity syndrome), but do minor rheologi-

cal problems play a role in the development of unilateral CRVO Minor viscosity changes have been shown by several groups, indicating increased hematocrit, plasma viscosity [2, 138, 180] and red cell aggregation [47], as well as a reduced red cell deformability [132, 185]. Elevated plasma fibrogen increases the odds by a factor of 3 [188]. All these changes are more or less borderline values and far from being capable of seriously hampering blood flow as occurs in hyperviscosity syndromes; hence, they can only be regarded as risk factors in triggering and evolving CRVO.

Thrombophilia seems a logical candidate in causing or at least contributing to the development of retinal vein “thrombosis.” This has prompted innumerable reports on its role during the last 20 years. A predisposing factor for thrombophilia can be diagnosed in 70 – 80 % of patients with recurrent venous thromboembolic disease outside the eye [34]. On the other hand, these abnormalities cause venous rather than arterial thrombosis, and retinal vein thrombosis is typically associated with risk factors of arterial thromboembolism.

The existence of a thrombus in the central vein was long debated [see 179] until Green et al. [54] examined 29 eyes in which thrombi were found in nearly all of the rubeotic eyes, but we are still unsure about the situation in freshly occluded eyes. In recent years an abundant number of coagulation cascade factors have been examined in retinal vein occlusion. Thrombophilia can be caused by a lack of plasma proteins that slow down the coagulation cascade (e.g., factor XII, antithrombin III, antiphospholipid antibodies, factor V Leiden/APC resistance) and hypofibrinolysis.

Two papers have shown a decreased factor XII [1, 102] and antithrombin III [1, 185], respectively, while anti-phospholipid antibodies are claimed to be a factor by some authors [1, 3, 179] and questioned by others ([7, 45, 46]; for a complete review, see [34]).

Even more controversial is the role of an increased activated protein C (APC) resistance or protein C and S deficiency. Vossen et al. [173] reported on a large protein-C-deficient family with a history of non-ocular venous thrombosis but no signs of retinal vein occlusion. It now seems largely unequivocal that in young patients there is a frequent decrease in the APC resistance [101, 102, 144] although this is not found in all papers [21, 53]. Thus APC resistance may play a definite role in triggering a CRVO in young patients. Most authors agree that in older patients with CRVO, APC resistance is not important [7, 25, 33, 79, 107] although there are a few exceptions [56, 162, 181].

Hyperhomocysteinemia, a risk factor for systemic vascular disease, may, via its deleterious effect on

vascular endothelium, induce increased platelet aggregation, lipid accumulation, and arterial thrombosis [177]. It remains controversial whether hyperhomocysteinemia, mostly dependent on a single mutation of the MTHFR gene (see [34]), is also associated with a higher risk for CRVO [8, 11, 16, 20, 26, 178], although, in a meta-analysis by Jannsen et al.

[89], an overall odds ratio of 8.9 for homocysteine III 21 above the 95th percentile was calculated.

Hypofibrinolysis caused by low levels of plasminogen activator inhibitor was demonstrated by Williamson et al. [185] and confirmed by Gleuk et al. [52].

Anti-phospholipid factors (anticardiolipin antibodies, lupus anticoagulant) activate the coagulation cascade, leading to both arterial and venous thrombosis [see 34] and 6 – 8 % of patients with anti-phos- pholipid syndrome present ocular manifestations. In their meta-analysis, Jannsen et al. [89] calculated an odds ratio of 3.9 for anticardiolipin based on six case-control studies.

In summary, it appears that hyperhomocysteinemia and anti-phospholipid syndrome are risk factors for CRVO and there is evidence that disorders causing hypofibrinolysis may also be important. The common hereditary thrombophilic conditions do not, however, seem to represent strong risk factors [34, 89], but one should be aware of them in young women, especially when additional risks, e.g., oral contraceptives, act synergistically [170]. This highlights the fact that atherosclerosis is an important factor in CRVO, while factors known as risk factors for venous thrombosis are not [82].

21.2.3.2 Local Risk Factors

Trauma may be associated with CRVO. A history of a preceding trauma is found in about 14 % of cases [122]. The mechanism after a serious head trauma may be compression or shearing of the central vein against the lamina cribrosa and constriction by a hematoma within the optic nerve sheath. These mechanisms might induce turbulence and endothelial damage thus provoking thrombus formation (see below). However, one would then expect many cases to occur after retrobulbar anesthesia, but this has only been reported once [158]. My own clinical experience does not lend support to an important role on the part of direct or indirect trauma in triggering CRVO.

Glaucoma is the ocular disease most commonly found in association with CRVO. Glaucoma increases the risk of suffering a CRVO fiveto sevenfold [23]. The incidence of glaucoma in CRVO is reported to be 23 – 69 % when ocular hypertension is included, and vice versa 2 – 8 % of glaucoma patients may develop CRVO [10, 38, 75, 78, 79, 111]. Hayreh pointed out

that CRVO leads to a subsequent fall in the intraocular pressure (IOP) in the involved eye. It thus is important to exclude glaucoma or ocular hypertension in the fellow eye of any patient with CRVO/ HRVO [75]. Furthermore, Cole et al. reported a high prevalence of arterial hypertension and hyperlipidemia in patients with both diseases [22].

21 III Other ocular diseases and factors leading to CRVO such as arteriovenous malformation and optic nerve diseases narrowing the central vessels (drusen, papilledema) may cause a congestion in the retinal veins and precipitate CRVO [19, 43, 163]. Green suggested not only compression, but endothelial damage by the drusen as well [54]. CRVO can be observed in patients with arteriovenous fistulas in the cavernous sinus region [148]. Retinal artery occlusions slow down the flow to stasis, and in case of reperfusion after several hours, a central retinal vein thrombosis might hamper drainage. This may be the cause for combined retinal occlusions.

21.2.3.3 Pathogenesis of CRVO

One main obstacle in clarifying the pathogenesis of vein occlusion is that we have only a few cases of histopathology from fresh cases. Today we are quite certain about the thrombotic character of most cases [54], but we cannot say whether this is the starting point or the end of the story. Thus the real cause of CRVO still remains controversial and is certain to involve different factors. One should not confuse precipitating causes and risk factors, and it will be helpful if the different factors are identified that may

play a role in initiating and maintaining the CRVO and that lead to either a benign course or propel the CRVO to a disastrous outcome. Pathogenesis thus can be represented in three circles: (A) triggering of the venous outflow reduction, (B) development of the retinopathy, and (C) transition into severe neovascular disease (Fig. 21.2.1).

21.2.3.3.1 Precipitating Causes of CRVO (Circle A)

Risk factors are well known, while their clear role in initiating the vicious circle is not well established. Three mechanisms seem to be involved in causing and maintaining the circle of flow turbulence, outflow reduction, and partial thrombosis: (1) a narrowing of the central vein, (2) hemodynamic disturbances, and

(3) thrombophilic conditions such as thrombophilia itself or hemorheological abnormalities.

Narrowing of the central vein can be caused by an enlarged arteriosclerotic central artery following long-standing arterial hypertension and arteriosclerosis, illustrating the high incidence of cardiovascular risk factors [10, 73, 122, 156] in patients with CRVO. In this context, the passage of the vessels through a tiny, non-expandable hole in the cribrose plate seems relevant, but it has never been shown that small papillas are more prone to CRVO, as it has been demonstrated for anterior ischemic optic neuropathy. In addition to a reduced venous outflow, the ensuing turbulences in venous flow may prompt endothelial damage with partial thrombosis. It then depends on additional factors (see below) as to whether a vicious circle develops or not.

B

Ischemia

VEGF

C

Neovascular Glaucoma

Fig. 21.2.1. Pathogenesis of CRVO: A triggering of the venous outflow reduction, B development of the retinopathy, C transition into severe neovascular disease

A narrowing may also occur via optic nerve swelling (papilledema, anterior ischemic optic neuropathy, drusen), but only a few cases of CRVO have been reported in such cases [19, 43, 163]. A long-standing compression of the central vein against the lamina cribrosa by increased IOP has clearly been identified as a risk factor [10, 39, 75, 77, 78, 111] while CRVO has not been reported after acute glaucoma with high IOP.

Damage to the vessel wall by trauma or inflammation with narrowing of the lumen could partly explain CRVO in young patients without cardiovascular risk factors [38, 39]. An inflammatory etiology was suggested to be a cause in young patients with CRVO [67]. More recent studies have not shown a predilection for inflammation in young patients [38, 39].

Increased thrombogenesis may certainly support the vicious circle of flow disturbance and partial thrombosis. The main risk factors for thrombophilia have been presented above (see Sect. 21.2.3.1). One cannot assume that these factors are capable of triggering CRVO as the only cause, as bilateral CRVO has never been reported.

Increased blood viscosity as caused by several minor rheological abnormalities found in patients with CRVO [2, 47, 132, 138, 180, 185] may support the development of the vein occlusion (see Sect. 21.2.3.1), as does thrombophilia. In contrast a severe hyperviscosity by highly elevated cells and/or plasma proteins usually causes a bilateral CRVO (hyperviscosity syndrome), demonstrating in those cases a causal relationship.

Hemodynamic factors seem capable of triggering CRVO by either an increase in central venous pressure (CVP) or a decrease in arterial blood pressure. The poor vision is nearly always detected by the patient the morning after a higher CVP in the recumbent position during the night and low arterial pressure when rising. In addition, the slight improvement of the visual acuity during the daytime supports the concept of better drainage when the CVP falls with the patient in an upright position. In that context, Francis et al. [40] reported that dehydration may trigger CRVO in young patients.

21.2.3.3.2Development of the Retinopathy (Circle B)

The diminished venous outflow can easily be assumed when observing an increase in all transit times in both arterial and venous flow in fluorescein angiography. This has been demonstrated by Sinclair and Gragoudas [153] among others. They measured the time of maximum venous filling as being a good indicator for CRVO’s severity. In addition, laser Doppler velocimetry revealed reduced blood flow in non-ischemic CRVO [17]. Reduced blood flow

increases the viscosity of blood logarithmically, hence worsening microcirculation [146].

Increased pressure within the retinal venous vasculature usually takes some time to damage the vessels. Long-standing engorgement will cause extravasation of blood cells, affect the integrity of the endothelial blood-retinal barrier, and overflow the retina

by exuding plasma. Not until a macular edema has III 21 developed will the drop in visual acuity become evi-

dent. We do not know whether increased vitreomacular adhesion might contribute to chronic edema in CRVO [116, 125].

CRVO not only leads to a decrease in capillary blood flow velocities, but also to an enlargement of perifoveal intercapillary areas [137]. Bleeding into a cyst forming in the fovea will reduce the vision more seriously. Retinal swelling, blood stasis, and occlusion of venules leads to cotton-wool spots and may contribute to ischemic development. Further progression to ischemic disease or a containment of the non-ischemic type of CRVO probably depends on finding a new balance between inflow and outflow, eventually supported by the formation of collateral disk vessels with a new drainage route [159].

21.2.3.3.3Progression to Ischemic CRVO (Circle C)

We do not know why some cases progress to ischemic retinopathy with vast capillary dropout and others not. The risk factors mentioned may play a decisive role in such cases as is supported by the unfavorable prognosis in older people and those with diabetes mellitus [58]. The bad perfusion leads to ischemia of the inner retina with release of growth factors. Hypoxia-induced vascular endothelial growth factor VEGF is, most likely, the link between retinal ischemia and neovascular disease of the iris and retina in CRVO [131]. VEGF-induced endothelial cell hypertrophy, moreover, may cause further capillary closure in this disease [80] and may exacerbate the ischemic retinopathy. Rubeosis iridis and, if not treated, retinal neovascularization and neovascular glaucoma ensue.

21.2.4 Clinical Features

Essentials

Symptoms: blurred vision in the involved eye after getting up, often improvement during daytime

Fundus with hyperemic swollen papilla, engorged veins, streaky retinal bleedings, cotton-wool spots

448

21 III

a

c

Fig. 21.2.2. CRVO of a mild type; b with heavy bleedings; c with few bleedings and engorgement of the veins; d with vitreous bleeding, all with macular edema

III 21

b

d

adapt to the reduced venous outflow, thus permitting the reforming of an intact blood-retinal barrier. A serous retinal detachment over wide areas with hard exudates (Fig. 21.2.7) may then develop.

The first sign proving an ischemic CRVO is neovascular disease of the iris. Untreated rubeosis iridis may lead to occlusion of the chamber angle with secondary glaucoma and hyphema. Interestingly, retinal neovascularization is not encountered that often in late CRVO as in other ischemic retinopathies.

21.2.4.2 Classification of CRVO

As already mentioned, the wide range of clinical features necessitates diligent classification, as that defines the course (see Sect. 21.2.4.4) and treatment. The most important challenge is to detect patients with the ischemic form of CRVO (synonyms: nonperfused CRVO, hemorrhagic retinopathy). This cannot be done on a funduscopic basis and often not at the patient’s first presentation. The ischemic and non-ischemic (venous stasis retinopathy, partial, perfused, hyperpermeable, prethrombosis) types form the two opposite sides of a continuous spectrum rather than distinct entities. Within the first

half year one should try to estimate the degree of ischemia so as not to miss the transition of nonischemic or indeterminate forms into treatable neovascular disease (see Sect. 21.2.4.4). This applies to both CRVO and HRVO [156].

Ischemic features cannot be identified by biomicroscopy of the fundus. One weak indicator may be widespread dense bleedings, although ischemic CRVO may also occur with fewer retinal bleedings (Fig. 21.2.8a–c). All other signs that might signal serious disease, such as severe swelling of the papilla, numerous cotton-wool spots, marked dilation and tortuosity of veins and strong macular edema, cannot be correlated to the degree of ischemia, as illustrated by the example given in Fig. 21.2.8b, c. A reliable sign of ischemia is neovascularization of the iris. Fluorescein angiography (capillary dropout, prolonged transit times) is most helpful in detecting the degree of ischemia but is not always conclusive. Margagal classified CRVO based on the ischemic index, which correlated with the proportion of retinal ischemia determined by fluorescein angiography [115]. An ischemic index of 50 % (corresponding to about 10 disk diameters of retinal capillary non-per- fusion) was considered the threshold for a significant risk of neovascular complications. This value was later picked up by the Central Vein Occlusion Study Group as an ischemia-defining parameter [112] although it was seriously questioned by Hayreh [67, 74]. Apart from these clues, combined information of the four functional tests visual acuity, visual fields, pupillary light response, and electroretinography will enable the ophthalmologist to make a definitive diagnosis and then correctly classify the ischemic risk (Table 21.2.2).

c

Table 21.2.2. Significance of signs pointing to ischemic central retinal vein occlusion

PD papillary diameter; the more + signs, the stronger the association to ischemic CRVO