Cataract is the most common cause of blindness (accounting for 51% of blindness) and cataract surgery involving removal of fogged lenses and replacement with synthetic intraocular lenses is one of the most common surgical procedures performed worldwide.1,2 This intervention is hugely important in tackling otherwise increasing levels of vision loss or poor vision quality resulting from cataracts, especially in the elderly and in people with diabetes.3–6 A negative aspect of cataract surgery, however, is pseudophakic macular oedema (PMO). PMO is considered by some ophthalmologists to be a fairly infrequent complication of cataract surgery, but it has a reported incidence that ranges from 0.2–20% in different studies.1,7 This variation arises for several reasons, especially the diagnostic method used, such as confirmation by clinical examination only, optical coherence tomography (OCT) or fluorescein angiography.1,7 In addition, the changing definition of PMO and various treatment and patient variables affect this rate. The use of phacoemulsification methods in particular has reduced this incidence range to 0.2–2.35%7–9 Even at this low incidence, the increasing scale of cataract surgeries (estimated to be >20 million performed worldwide in 2015)10 makes PMO a significant problem that should be managed effectively. PMO is detected more frequently when eyes are examined using angiography, rather than relying on clinical symptoms.
When PMO occurs, it may result in suboptimal or impaired vision after surgery.8,9,11 It is important therefore that preventive treatments are used to diminish this risk as much as possible and ensure a favourable outcome. Among the various treatments used to achieve this aim is the preventive use of the non-steroidal anti-inflammatory drug (NSAID), nepafenac (NPF). This is administered as 0.1% eye drops and has shown notable efficacy in reducing the incidence of post-cataract PMO in several clinical trials12–18 and in clinical use. It is now approved in this indication for patients with or without diabetes.19,20 This article will consider the problem of PMO following cataract surgery and its consequences. It will also provide an update and consider newer evidence supporting a higher dose of NPF (0.3%) that can be dosed less frequently than the 0.1% formulation and will compare the efficacy and safety of both doses.
The pathophysiology and consequences of pseudophakic macular oedema
The pathogenesis of PMO following cataract surgery is not completely understood, but the causes are likely to be multifactorial.7,8 These are believed to include: vitreous traction resulting from manipulation and incisions in the eye,21 functional disturbance of the blood–retinal barrier (BRB)22 and inflammation mediated by prostaglandins.23 Surgery is known to cause some degree of irritation and trauma within the eye, especially to the iris. It is likely that various mediators of inflammation (prostaglandins, cytokines and other vascular permeability factors) are released from the anterior segment and move to the posterior, including the vitreous and retina, where they stimulate breakdown of the BRB and leakage of fluids to the perifoveal retinal tissues leading to inflammation.8,24 Damage to the BRB leads to leakage and fluid accumulation in the central retina, which can result in vision impairment and even vision loss.8,11,24–26
In most patients, PMO resolves without intervention.24 In some patients, however, more extensive leakage from the retina causes chronic inflammation that can lead to severe damage, which is irreversible.27 The reasons why most cases of PMO resolve but some become chronic are, as yet, unclear but they may be related to the presence of risk factors, as discussed below. PMO has traditionally been diagnosed using fluorescein angiography, but OCT accurately reveals swelling and thickening of the retina and cystic spaces (Figure 1) and is now the method of choice. Other methods include vitreous fluorometry and the retinal leakage analyser.
Risk factors and incidence of pseudophakic macular oedema
The risk of PMO is increased by various factors, especially rupture of posterior capsule, vitreous loss, iris incarceration, use of iris fixated lenses, active uveitis and diabetes.11 In one comparative cohort study (n=93) in eyes with uveitis, active inflammation within 3 months prior to surgery significantly increased the risk of PMO compared with eyes without prior inflammation.28 A retrospective database study involving 81,984 eyes undergoing cataract surgery in the UK identified the highest risk factors for PMO. These included: epiretinal membrane (hazard ratio [HR]: 5.596), retinal vein occlusion (HR: 4.471) previous retinal detachment repair (HR: 3.926) and uveitis (HR: 2.878) (Figure 2).1
The reported incidence of PMO varies from approximately 0.1% to 2.35%, but the rate of angiographic/asymptomatic PMO (with visible leakage from perifoveal capillaries) is up to 10 times greater. Use of a large incision instead of a small incision with phacoemulsification also increases the rate of PMO.8 This incidence does not appear to be affected with the use of laser or manual surgery. This was demonstrated in a study of 1,390 patients receiving cataract surgery, which found PMO in 15 eyes (1.1%; 11 patients). There was also no difference in the incidence of PMO in patients receiving laser or conventional cataract surgery.29
The incidence of PMO appears to decrease with increasing age, as was highlighted in a database study of 13,556 patients (19,980 eyes) who received uncomplicated phacoemulsification during a 1-year period at community-based outpatient clinics within a region in France.30 This study found a trend of decreasing PMO in older age groups: 2.38% in those aged 40–54 years, 0.86% in those aged ≥85 years; p=0.0009. The incidence of PMO was also higher in men than women (1.25% versus 0.73%; p=0.002), but there was no difference in incidence between those with a history of diabetes than those without it (1.05% versus 0.90%; p=0.48). This gender difference is supported by the UK database study (81,984 eyes) that showed a higher incidence of PMO among males versus females (1.391% versus 1.023%; p=0.0019).1
Management of pseudophakic macular oedema
The American Academy of Ophthalmology (AAO) and the American Society of Cataract and Refractive Surgery (ASCRS) consider that there is insufficient strong evidence to support the routine prophylactic use of anti-inflammatory drugs to prevent PMO following cataract surgery.31,32 Furthermore, a Cochrane review concluded that the value of adding topical NSAIDs to steroids or using them as an alternative to topical steroids with a view to reducing the risk of poor visual outcome after cataract surgery is uncertain.33 However, many ophthalmologists accept that prophylactic use of anti-inflammatory drugs is important in preventing PMO and report that this approach is effective.34–37 Such treatments include various topical NSAIDs or corticosteroids, and combinations of the two. These treatments have also been used in combination with anti-vascular endothelial growth factor (VEGF) agents such as bevacizumab in PMO.38 Prophylactic NSAIDs have shown varied efficacy in the prevention and treatment of PMO and have been used in these indications for many years. Frequently used examples of NSAIDs include: ketorolac, bromfenac, diclofenac, indomethacin, flubiprofen, nepafenac and suprofen.7,8 Some studies have provided strong evidence supporting such treatments. One example included 1,659 consecutive cataract surgeries performed by residents.39 In treating post-surgical cystoid macular oedema, the combination of an NSAID and a steroid resulted in significantly shorter resolution times than no treatment. On the other hand, some other studies evaluating the use of anti-inflammatory drugs before and after cataract surgery have been less well designed. As a result, the treatment is still considered controversial by some physicians and more studies are needed to verify the efficacy of anti-inflammatory treatments in PMO to justify their continued use.8,33 Once PMO becomes chronic, treatment has a high cost; prophylactic treatments significantly reduce these costs and reduce patient burden, particularly in those with preoperative diabetic macular oedema whose rates of PMO are substantially higher than those without it.40 These costs were emphasised by a retrospective analysis of the US Medicare files of 78,949 patients undergoing cataract surgery over a 2-year period. This study found that patients with diabetic macular oedema (DMO) had mean ophthalmic charges of US$10,410 versus US$5,950 for those without it (p<0.0001).
Nepafenac 0.1% – an established preventive treatment for macular oedema following cataract surgery
Among the multiple NSAIDs that are available, topical 0.1% NPF has shown efficacy in preventing PMO when started shortly before and continued after cataract surgery.41 Various clinical trials in recent years have demonstrated the benefits of NPF in this indication.13,14,18,42 NPF is a unique prodrug that rapidly penetrates the cornea and is converted to the active compound, amfenac. This molecule is target-specific and inhibits prostaglandin formation and consequent inflammation in both the anterior and posterior segments of the eye. The prodrug structure may reduce surface damage to the eye, which can be an issue with non-prodrug medications. NPF also inhibits retinal VEGF secretion and reduces neovascularisation.43–45
Several key phase II and III trials and a pooled analysis have supported use of NPF in the prevention and treatment of PMO (Table 1). Among these was a phase II randomised vehicle-controlled study (n=263) that included patients who were treated with topical 0.1% NPF or vehicle from day 1 to day 90 post-surgery.17 Significantly fewer patients developed PMO in the NPF group than vehicle group (3.2% versus 16.7%; p<0.001) and fewer had best-corrected visual acuity (BCVA) reductions of >5 letters on days 30, 60 and 90 (p=0.001–0.006). Mean central subfield macular thickness (CSMT) mean and change in macular volume from baseline were also significantly lower with NPF (the mean maximum change for NPF was 18.9 ± 19.5 μm and for vehicle was 40.8 ± 49.0 μm; p<0.001). In addition, changes in macular volume were significantly lower with NPF (p<0.001 for days 30, 60 and 90). No safety issues were identified.
A further key study was a phase III randomised vehicle-controlled study that included 175 patients with diabetes requiring cataract surgery and assigned to 0.1% NPF (three times daily, n=87) or vehicle (three times daily, n=88) (all received tobramycin and dexamethasone).15 At 90 days post surgery, significantly fewer NPF-treated patients developed PMO than vehicle-treated patients (5.0% versus 17.5%; p=0.01). The mean change in BCVA from baseline to day 90 was numerically greater for NPF but not significantly different (+17.7 versus +14.3 letters; p=0.14). The 90-day treatment regimen with NPF effectively prevented PMO after cataract surgery and identified no new safety issues or trends.
Comparative trials of nepafenac in the prevention of macular oedema following cataract surgery
Several key clinical trials have compared NPF with another treatment or compared concomitant treatments used in addition to NPF. One such study was a retrospective investigation of patients receiving cataract surgery with phacoemulsification (n=65) who were all given dexamethasone and netilmicin and were divided into treatment groups according to the additional NSAIDs that they received. Only patients receiving 0.1% NPF plus steroids showed a reduction in macular thickness (-1.3%; p=0.048) whereas it remained the same for those receiving steroids plus bromfenac, indomethacin or diclofenac (-1.1% to +1.2%; p=0.19–0.74) but was increased in those treated with dexamethasone and netilmicin alone (+4.3%, p=0.04).46 No side effects were observed in any group. It was concluded that NPF, bromfenac and indomethacin enhanced the efficacy of steroids used to reduce PMO.
A phase II double-blind active controlled study directly compared 0.09% bromfenac with 0.1% NPF in terms of bioavailability and efficacy in patients undergoing phacoemulsification (n=139).47 After 30 days, there was a significant difference in central retinal thickness when the two treatments were compared (252.0 versus 264.0 μm; p=0.022), indicating that NPF provided a greater effect than bromfenac on this parameter. However, there were no clinically relevant changes in the presentation of PMO and no significant alterations in intraocular pressure, lissamine green or fluorescein staining. For other parameters, pain, photophobia, conjunctival hyperaemia, chemosis, cellularity and corneal oedema had all disappeared by day 30 with both treatments. Overall in this study, bromfenac 0.09% and nepafenac 0.01% used after phacoemulsification showed similar clinical efficacy.
In a Japanese study, 59 patients receiving small incision cataract extraction and foldable intraocular lens implantation were randomised to fluoromethalone 0.1% or NPF 0.1%.14 At 5 weeks, with NPF compared with fluoromethalone, the incidence of fluorescein angiographic cystoid MO was lower (14.3% versus 81.5%, respectively; p<0.0001), the fovea was thinner (p=0.0055) and anterior chamber flare was lower (p=0.0304). In addition, the visual acuity recovery was significantly greater with NPF (80.0%) than fluorometholone (55.2%; p=0.0395). This study therefore showed that NPF was more effective than fluorometholone in preventing angiographic PMO and that NPF provided a more rapid recovery in vision.
A study in India compared NPF with ketorolac in PMO.16 This study included 200 eyes in 200 patients undergoing phacoemulsification (including some patients with diabetes) with a mean age of 58–59.3 years and a low PMO risk (without ocular inflammation, diabetic or hypertensive retinopathy, ocular vascular disease, retinitis pigmentosa, previous intraocular surgery, vitreoretinal interface abnormalities or hypersensitivity to NSAIDs). Patients were randomised to post-operative 0.1% NPF 0.15 or ketorolac 0.4%. At 1 month, post-operative macular thickening was similar in both groups. After 1 week, diabetic patients showed significantly less central macular thickness difference with NPF than with ketorolac (p=0.04), but this was not significant at 1 month (p=0.09). Macular thickening after cataract surgery therefore appeared similar with ketorolac and NPF, although the authors noted that a larger study is needed to validate these findings.
A meta-analysis of 11 randomised controlled trials that included a total of 1,165 patients with 1,175 cataract surgeries showed that topical 0.1% NPF is superior to ketorolac for ocular inflammation in terms of patient tolerability (conjunctival hyperaemia and ocular discomfort) following cataract surgery.48 However, the two drugs were found to be equally suitable for the management of anterior chamber inflammation, visual rehabilitation and intraoperative mydriasis.
Evidence supporting a higher dose, less frequent dosing regimen of nepafenac (0.3%) in the prevention of macular oedema following cataract surgery
The 0.1% NPF formulation for PMO prevention/treatment is given three times daily. More recently, a higher concentration 0.3% formulation has been developed and approved to enable a more convenient once daily dosing.42 Some key phase III trials have compared the efficacy and safety of the 0.3% dose with the established 0.1% dose.
The first key trial for 0.3% dosing was a double-masked,vehicle- and active-controlled study at 65 centres in the US and Europe (NCT01109173), in which patients receiving cataract surgery were randomised to 0.1% NPF three times daily (n=819) or 0.3% NPF once daily from day 1 to day 14.49 Other patients (n=200 and n=206 received the respective vehicles only). For patients receiving 0.3% NPF versus vehicle at day 14 significantly more had no inflammation (68.4% versus 34.0%; p<0.0001) and significantly more were pain free (91.0% versus 49.7%; p<0.0001; pain free was defined as an ocular pain assessment score of 0). The proportions that were cured or pain free by visit are given in Figure 3. With 0.3% NPF there were also significantly fewer treatment failures (p≤0.0012) and more treatment successes (p≤0.0264) compared with vehicle. 0.3% NPF was non-inferior to 0.1% NPF for inflammation (95% confidence interval [CI]: -5.73–3.17%) and for proportions who were pain free (95% CI: -3.08–2.70%). NPF 0.3% was well tolerated and the safety profile was similar to that of 0.1% NPF. This indicated that a higher dose of NPF could be given at more convenient, less frequent intervals without decreasing efficacy, tolerability or safety.
More recent key evidence in favour of the 0.3% NPF formulation comes from a pair of phase III randomised, multicentre vehicle-controlled trials (study 1: NCT01853072; study 2: NCT01872611) that were initiated to support the approval of 0.3% NPF for use in PMO in Europe.50 Adult patients (n=615 and n=605) had planned phacoemulsification and intraocular lens implantation and had BCVA of ≤73 ETDRS letters (<26/63 Snellen). In both studies, significantly fewer patients receiving 0.3% NPF developed PMO than those receiving vehicle (both studies pooled: 4.1% versus 15.9%; p<0.001) (PMO was defined as ≥30% increase from the preoperative baseline in central subfield macular thickness within 90 days following cataract surgery) (Figure 4). The proportion showing ≥15-letter improvement from baseline to day 14 and maintained to day 90 was greater with 0.3% NPF versus vehicle (both studies pooled: 55.4% versus 46.7%; p=0.003).
In study 1, a greater proportion receiving 0.3% NPF showed a BCVA improvement of ≥15 letters from baseline to day 90 versus vehicle only (77.2% versus 67.7; p=0.009) but in study 2 there was no difference (65.4% versus 65.9%; p=0.888) (Figure 4). This lack of difference was considered by the authors to be most likely due to type II errors (false negative) that diminished the apparent treatment effect in terms of improvements in BCVA. The study populations were similar, but differences in medical practices such as cataract surgery technique and patient compliance may also have affected the findings.Sub-analyses by location, age, sex, race, retinopathy severity, visual acuity, medical history and concomitant medications showed no differences to the findings of the main study populations. There were no unanticipated adverse events; in study 1, 35.9% of NPF-treated and 41.1% of vehicle-treated patients had at least one treatment-emergent adverse event (TEAE). In study 2, these proportions were 34.8% and 46.1%, respectively. As with the US/Europe phase III study, these two European studies showed the superiority of the 0.3% NPF dose over vehicle in terms of efficacy.
An additional randomised study of 1,000 eyes, conducted in the US recently, showed that prophylactic once-daily 0.3% NPF for 5 weeks after cataract surgery significantly reduced the incidence of cystoid macular oedema compared with placebo (p=0.0001).51 This effect, however, was only seen in patients with preoperative risk factors (mainly pseudophakic cystoid macular oedema in the previously operated eye, diabetic retinopathy and vein occlusion).
Further evidence for the 0.3% NPF formulation comes from a 2013 literature search and analysis of clinical studies of NPF to reduce pain after cataract surgery, which concluded that the 0.1% and 0.3% NPF formulations appeared to be equally effective, but the 0.3% dose was more convenient yet with a slightly higher adverse event rate than the 0.1% dose.42
Discussion and future directions
Suboptimal results after phacoemulsification are not well tolerated in today’s world with advanced intraocular lenses, advanced intraoperative techniques and increasingly active patients with high demands. One such suboptimal outcome, PMO, is a comparatively uncommon condition; however, the increasing frequency of cataract surgery and the growing need for lens replacement in ageing populations has increased its incidence and raised the importance of preventive treatments.1,2,7–9 PMO prevention is particularly important in patients with diabetes and/or uveitis who are at a greater risk than those without these factors.1,24,26 Uveitis is an important sign in patients with a history of retinal vein occlusion. It is also an important sign in patients with glaucoma who should stop receiving prostaglandin therapy prior to cataract surgery.11,24,28 Various studies have indicated the benefits of NSAIDs in the prevention of PMO but there is a lack of level I evidence that supports the long-term benefit of this therapy in preventing vision loss after cataract surgery, even though it may increase the speed of vision recovery in the first few weeks after surgery.32 Topical NSAIDs have variable efficacy in PMO prevention but there is extensive lower-level evidence that they do provide benefits when used alone or in combination with corticosteroids.32,38,46,52
More compelling evidence in favour of NSAIDs in PMO prevention comes from a systematic review that identified 15 randomised trials that met the selection criteria.53 The analysis found high-quality evidence that topical NSAIDs are more effective than corticosteroids in preventing PMO after cataract surgery (PMO occurred in 3.8% of NSAID-treated patients versus 25.3% of corticosteroid-treated patients; risk ratio 5.35;95% CI: 2.94–9.76). This analysis also found low- to moderate-quality evidence that topical NSAIDs are more effective than corticosteroids in preventing post-operative inflammation after cataract surgery.
Among the NSAIDs, in the clinical trials discussed, 0.1% NPF (three times daily) has consistently shown significant efficacy in reducing PMO incidence in terms of evidence of angiographic leakage, macular thickness and measures of visual acuity, and the treatment is well tolerated with little difference in adverse events between the treatment and vehicle only. To reduce the required frequency of dosing and improve convenience, phase III clinical trials have shown similar findings in that the higher 0.3% NPF formulation is effective in PMO prevention when dosed once daily.
The two recent European phase III studies did not compare the 0.3% NPF dose regimen with the 0.1% NPF dose regimen, but comparisons between studies suggest that the former achieves similar efficacy outcomes without changing the safety profile.50 The more convenient 0.3% NPF dosing schedule also has the potential to improve patient compliance and outcomes. However, the earlier phase III trial conducted in the US and Europe49 was only powered to show non-inferiority of 0.3% NPF compared with the 0.1% NPF regimen, so it is not possible to state whether the 0.3% dose is superior.
Further evaluation of NSAIDs in PMO prevention is needed. This would further justify their use and identify which are more effective and tolerable. Data from such studies may encourage greater use of PMO preventive medication in the future and help decrease inflammation leading to reduced vision damage in a sizeable population of patients who are about to undergo or have recently undergone cataract surgery.