Article Text
Statistics from Altmetric.com
Editor,—We were interested to read Kirwanet al’s report of acute angle closure glaucoma (AACG) associated with the antidepressant paroxetine (Seroxat),1 as we have reported a similar case.2 In Kirwan’s report, AACG occurred within 24 hours of the first dose, suggesting an anticholinergic mechanism. Our case became symptomatic some 2 weeks after daily dosage was commenced, leading us to postulate that the effect could have been mediated by serotoninergic pathways. This raises the possibility of other important ocular side effects with drugs of this class.
Paroxetine is an antidepressant of the selective serotonin reuptake inhibitor (SSRI) class. SSRIs act by inducing a gradual rise in postsynaptic levels of serotonin (5-hydroxytryptamine, 5-HT) via desensitisation of the feedback systems which control the rate limiting enzyme in 5-HT synthesis.3 The 5-HT receptors involved have not yet been fully elucidated, and their role in ocular physiology is a subject of ongoing research. In animal studies, serotoninergic stimulation may cause mydriasis, and can have an independent effect in raising the intraocular pressure (IOP).4 Receptors for 5-HT have been demonstrated in the human eye.5
It is therefore possible that SSRIs could have an effect on IOP in humans. Consequently, we should look for raised IOP and open angle glaucoma as a side effect of SSRIs. A recent study has shown a significant short term rise in IOP after a single oral dose of the SSRI fluoxetine (Prozac). Twenty depressed patients were given either fluoxetine 20 mg or placebo in a randomised crossover blinded study, and fluoxetine was associated with a mean IOP elevation of over 4 mm Hg, lasting 6–8 hours.6 We have been unable to find any publication regarding the effect on IOP in the longer term, but we have recently become aware of previously unpublished data which partly address this issue.
In premarketing and subsequent clinical trials of fluoxetine, 585 adult patients of a wide age range have been assessed. Ophthalmological examinations were made at baseline, and again after a treatment period which ranged between less than 1 month and 24 months. “Five out of the 585 patients had a change in intraocular pressure... . One had a decrease in IOP which was determined to be clinically insignificant [29 year old female, dosage 20 mg/day, examined at 1 month]. Four patients experienced an elevation: in one patient the elevation was considered a normal diurnal variation [74 year old male, dosage 80 mg/day, examined at 2 months], in a second patient the elevation was felt to be related to angle closure [59 year old male, dosage 80 mg/day, examined at 1 month], the third patient acknowledged cocaine use prior to the examination and this was considered the probable explanation [37 year old male, dosage 80 mg/day, examined at 12 months], and the fourth patient had a l mm Hg rise with no probable extenuating circumstances [26 year old female, dosage 80mg/day, examined at 2 months]” (Dista Products Limited, personal communication).
Voluntary reporting of suspected adverse events with fluoxetine has identified a total of 63 cases of “glaucoma” in an estimated patient population of 21 million (Dista Products Limited, personal communication). The manufacturers of paroxetine are aware of four cases of AACG, six of “glaucoma (unspecified)”, and one of raised IOP, in a UK patient population of over one million (SmithKline Beecham Pharmaceuticals, personal communication).
These data indicate that our understanding of the effect of SSRIs on IOP is still unclear. The demonstration of a short term IOP rise after a single fluoxetine dose6 implies that chronic dosage could lead to a sustained elevation of IOP. However, the manufacturer’s own data suggest that this is not the case, in that less than 1% of patients showed any IOP change after treatment. The low incidence of reported glaucoma with SSRIs does not exclude a real effect: many clinicians may not suspect a particular drug to be a contributory factor when diagnosing a particular condition, especially if a causal relation has not been suggested in the literature. This is particularly true of open angle glaucoma, which is common and usually idiopathic, and of AACG, which is rarer but occurs in anatomically predisposed eyes.
We feel that this area merits further study and clarification, particularly regarding the effect of long term SSRI administration on IOP. In the meantime, we would encourage colleagues to report cases of glaucoma or raised IOP which may be associated with SSRIs to the Committee on Safety of Medicines.
References
Reply
Editor,—In 1991 Ahmadl first reported the occurrence of glaucoma, probably an acute angle closure glaucoma (AACG), in a 35 year old man as a result of fluoxetine administration, during the fifth week of therapy. More recently, also Kirwan et al and Ekeet al described two cases of AACG associated with the administration of paroxetine, an antidepressant which acts as a selective inhibitor of the serotonin neuronal uptake (SSRI). In 1996 we described the effect of fluoxetine oral administration on intraocular pressure (IOP) in 20 consecutive depressed patients and demonstrated that after a single dose of 20 mg IOP was significantly increased.
Serotonin (5-HT) is present in mammalian iris ciliary body and cornea at higher concentration than in non-mammalian species.While a transmitter role for serotonin in the retina has been established, conflicting data exist in the literature on whether or not activation of serotonin receptors can cause changes in IOP. Experimental evidences demonstrated that topical application of serotonin increased IOP in rabbit eyes, and that 5-carboxamidotryptamine, a 5-HTla receptor agonist, is even more effective than 5-HT itself in elevating IOP. These results confirm the involvement of serotonin receptors in the regulation of IOP, and the fact that ketanserin, a compound with serotonergic blocking properties, reduces IOP in animals and humans emphasises the role exerted by 5-HT on IOP.
To date, no long term studies regarding the effect of SSRI on IOP have been published. We know the prevalence of primary open angle glaucoma (POAG) (conservatively estimated to be approximately 1/200 of the general population over the age of 40 years) and that of AACG (approximately 1/1000 individuals over the age of 40 years), while the incidence of new cases within the population treated with SSRI has not yet been evaluated, with the exception of the manufacturer’s own data, which do not show a significant occurrence of IOP changes induced by SSRI treatment (Dista Products Limited, personal communication; SmithKline Beecham Pharmaceuticals, personal communication). POAG is an asymptomatic disease, until significant visual field loss occurs, and a single IOP reading does not necessarily exclude or confirm the diagnosis. The rise in IOP after a single dose of SSRI was about 4 mm Hg in a non-glaucomatous population, equivalent to the diurnal fluctuations of IOP found in normal subjects, and probably is without long term effect on the ganglion cell integrity. On the contrary, this same variation, together with the small changes in pupil size reported by other investigators, in the so called “predisposed eyes” (that is, patients with positive family history of glaucoma and/or patients with physiological risk factors—relatively anterior location of the iris-lens diaphragm, shallow anterior chamber, and narrow entrance to the chamber angle) is able to modify the hydrodynamic homeostasis of the eye, leading to the acute angle closure.
These considerations suggest the need to include an ophthalmological examination in the protocol of depressed patients before starting SSRI administration.
Authors’ reply
Editor,—Eke et al’s letter raises several interesting points. The rapid onset of acute angle closure glaucoma in our case implies an anticholinergic mechanism. The patient in Eke’s report may simply have had an anticholinergic response to commencement of the higher dose 3 days before. However, the serotoninergic mode of action with this agent does raise the issue of both a mydriatic effect and an effect in eyes with open angles. As with any recently introduced class of drug, it is important for clinicians to monitor closely possible adverse effects.
So far the data reported from the manufacturers reveal a minimal number of cases of raised intraocular pressure, not different from the probable incidence of primary open angle glaucoma in the general population. However, a small study with careful monitoring of intraocular pressure demonstrated a small but significant elevation of IOP over 12 hours. Subjects were, however, young adults and the study was over a short period. We have little idea what happens over the longer term. For example, tricyclic antidepressants have profound anticholinergic side effects on commencement of treatment to which patients physiologically adapt to the point that sudden withdrawal leads to a rebound cholinergic syndrome. The effect in older patients, those with anatomically predisposed narrow angles, or indeed glaucoma patients may be very different. If this class of agents does increase intraocular pressure by 4 mm Hg in a sustained fashion this could potentially have clinical implications in population terms for glaucoma risk. In the year ending August 1994, British general practitioners wrote 13 million prescriptions for antidepressants. A significant proportion of these were SSRIs.
Should the possibility of this class of agent causing an elevation of intraocular pressure modify their current use? In general the answer should be no. Depression is common; the estimated lifetime prevalence of “major depression” in women is between 4.9% and 8.7% and in men, 2.3–4.4% with even higher figures applying to the less severe end of the spectrum. Depression is a severe, disabling disease. It has a high mortality. It is estimated that severe depressive illness increases the risk of suicide by a factor of 30 such that 15% die by suicide. Alternative treatments with other antidepressants are generally less well tolerated, especially in the elderly. Despite the possibility of raised intraocular pressure the risk/benefit ratio will almost always favour treating depression with the optimum agent.
Until the real effect of SSRIs on IOP has been ascertained it is difficult to make suggestions on management and further evidence on this subject is required. However, it would seem prudent to closely monitor glaucoma patients who have recently commenced treatment with an SSRI. Given limited ophthalmic resources, until we know more about the long term effects of these drugs on IOP it does not seem reasonable to recommend ophthalmological screening of all patients commenced on SSRIs. Like Eke and Carr, we encourage colleagues to report cases of glaucoma or raised IOP that may be due to therapy with SSRIs. Additionally, we would stress the importance of communication between disciplines so that no doctor is unaware of prescribed medication.