Abstract
Despite the refinement of modern cataract surgery, postoperative inflammation still constitutes a substantial amount of visual morbidity worldwide. A surrogate for intraocular inflammation and blood–aqueous barrier breakdown can be objectively quantified by Laser flare photometry (LFP). This review outlines the utility of LFP in assessing the assessment of post-cataract surgery inflammation. It highlights the impact of preoperative pathological states such as uveitis and diabetes, intraoperative techniques, including efficient phacoemulsification and direct comparisons between postoperative anti-inflammatory regimes. There is a large interobserver variation in the subjective flare measurement after cataract surgery and the continued use of LFP amongst other objective, noninvasive measurements of intraocular inflammation, particularly in the further development of cataract surgery, is recommended.
Introduction
Senile cataracts are the leading cause of visual impairment worldwide. 1 Cataract surgery is one of the most commonly performed surgical procedures, with over 4.7 million operations performed across European Union member states in 20172 and over 400,000 carried out by the National Health Service in the United Kingdom alone. 3 Surgical techniques have advanced dramatically over the last 60 years since the advent of modern small-incision phacoemulsification cataract extraction, first described by Kellman in 1967. 4 Its high safety profile and excellent visual outcomes allow cataract surgery to treat glaucoma5,6 and refractive errors using premium intraocular lenses (IOLs). 7
Intraocular surgery causes a breakdown in the blood–aqueous barrier and leakage of inflammatory material to the anterior chamber (AC). 8 Consequently, cataract surgery produces distinctive alterations in aqueous concentrations of proteins. 9 Increased aqueous protein concentrations result in an aqueous humour haze called AC flare. 8 This is assessed during slit lamp biomicroscopy and is the subject of various clinical grading systems,10,11 the most popular of which is the Standardization of Uveitis Nomenclature (SUN) grading system. 12 While this classification is widely accepted in quantifying AC inflammation to guide treatment, limitations include inherent subjectivity and marked interobserver variability, especially at low levels, primarily seen in clinical practice. It is, therefore, imperative to develop objective and noninvasive measurements of AC inflammation.
Laser flare photometry (LFP), described as early as 1988, 13 has been commercialized since its emergence in 1989 and is currently marketed by KOWA (Kowa Company, Tokyo, Japan). It measures the level of scattered light from a laser beam directed through a measurement window at the AC 14 (Figure 1). Successive models of laser flare meters (Kowa Co. Ltd, Tokyo, Japan) are differentiated by the laser medium (e.g. Helium-Neon 632.8 nm laser for FC-1000 and diode 670 nm for FM-500), the area of AC studied as well as the ability to quantify cell as well as flare. Published comparisons are rare but show reproducibility and comparability. 15
Schematic of laser flare photometry.
Measurements cannot be recorded in eyes with opaque media (e.g. corneal scarring) or in shallow ACs. Reliable measurements may also not be obtained in eyes with mature cataracts or extensive posterior synechiae due to increased background light scatter. LFP measurements are expressed in photons per millisecond (ph/ms), and the grading ranges from 3 to 1000 ph/ms. 16 Because of its ability to be a more precise tool for aqueous flare quantification, LFP has been utilized in measuring postsurgical inflammation in ophthalmic surgical procedures such as trabeculectomy 17 and pars plana vitrectomy, 18 as well as laser procedures such as argon laser panretinal photocoagulation 19 and laser capsulotomy. 20 Background flare (2.9–3.9 ph/ms between 20 and 40 years of age, increasing to 5.0–6.5 ph/ms between 70 and 80) is present in physiological eyes,21,22 but is undetectable at the slit lamp. Studies have shown a high correlation between LFP values and protein concentrations in plasma and aqueous humour samples obtained in patients undergoing intraocular surgery. 23 LFP values also appear not to be significantly affected by mydriatic agents such as tropicamide and phenylephrine. 24 This review aims to identify and summarize the studies of LFP quantifying AC inflammation (via aqueous flare measurements) in the use of materials and surgical procedures used in cataract extraction as well as regimens to control postoperative inflammation.
Methods
We conducted a search using the following databases: PubMed (all years), the Web of Science (all years), Ovid MEDLINE (R) (1980–31 December 2021), Ovid MEDLINE (R) Daily Update 31 December 2021, MEDLINE and MEDLINE non-indexed items, Embase (1980–2021, week 52), Ovid MEDLINE (R) and Epub Ahead of Print, in-Process & Other Non-Indexed Citations and Daily (1980–31 December 2021), CENTRAL (including Cochrane Eyes and Vision Trials Register; Cochrane Library: Issue 12 of 14 December 2021), metaRegister of Controlled Trials (mRCT) (https://www.controlled-trials.com), ClinicalTrials.gov (www.clinicaltrial.gov) and the WHO International Clinical Trials Registry Platform (www.who.int/ictrp/search/en). Search terms included ‘aqueous flare’, ‘anterior chamber inflammation’, ‘tyndallometry’, ‘laser flare photometry’ combined with ‘cataract surgery’, ‘extra capsular’, ‘phacoemulsification’, ‘laser’ and ‘femtosecond’. Published articles in English were preferentially selected for consideration in the review. Articles selected for eligibility were screened and deemed eligible if the aqueous flare was measured with LFP both before and after cataract surgery. All the articles were screened by two authors (CW and AS) for eligibility in this narrative review. The data were collected in a spreadsheet (MicroSoft Excel, MicroSoft Corp, USA). The flare data in relation to the preoperative patient’s factors/comorbidities, intraoperative steps (incision, method of cataract removal, ophthalmic viscosurgical device, irrigating solutions and intraocular lenses) and postoperative regime for control of inflammation were collected.
Results
Articles from 1990 to June 2022 were accessed, identifying 155 articles for further analysis. Of these, 24 articles were excluded for the following reasons: 15 were duplicate studies, 1 did not involve cataract surgery and 8 measured flare using either clinical grading systems such as the SUN or an enzyme-linked immunosorbent assay instead of LFP (Figure 2).
Flow diagram outlining eligible studies.
All studies showed an increase in flare after cataract surgery, which is at its highest and most variable in the immediate postoperative hours. 25 No study recorded flare with more than one model of flare meter, with the most popular meters used being the FC-1000, FM-500 and FM-600. The influence of pre-, intra- (including incision, nucleus dissassembly, ophthalmic viscosurgical devices, irrigating solutions and intraocular lenses) and postoperative variables are discussed below.
Preoperative
Patient factors and comorbidities are associated with the increased postoperative flare. Non-pathological states include older age 26 and pathological states include pseudoexfoliation syndrome,27–29 chronic uveitis17,30 and diabetes both with and without retinopathy.31–34 The density of the nucleus, according to the LOCS III grading system, did not correlate with higher flare preoperatively. 26 However, it was associated with a higher postoperative flare, 26 possibly due to a higher phacoemulsification energy requirement. 34 Though iris colour does not affect preoperative values,26,35 darker irises have a greater postoperative flare response at 2 weeks than lighter colours. 36
Intraoperative
Incision
The introduction of phacoemulsification techniques and foldable intraocular lenses in the early 1990s allowed for smaller primary incisions into the AC. This was shown in 1991 by Gills and Sanders to reduce flare as well as surgically induced astigmatism at postoperative day 1. 37 Though this is confounded by the different lenses that each group received, it is later discussed that foldable lenses do not adversely affect flare outcomes. The clear corneal incision is the preferred choice among cataract surgeons compared with scleral tunnels. 38 There is contrasting evidence regarding the impact of incision on postoperative flare; Dick et al. 39 found significantly lower flare in those who underwent clear corneal incisions in the first three postoperative days, but Kruger et al. 40 found similar flare levels at almost all time points to 90 days postoperatively. Nguyen et al. 41 also found no postoperative difference between the incisions, however this was a secondary outcome measure No studies have found clear corneal incisions to worsen flare levels compared with scleral tunnels.
Nucleus disassembly
Phacoemulsification is the dominant method of nucleus disassembly in modern cataract surgery. Several studies comparing phacoemulsification with extracapsular surgery in the mid-1990s demonstrated that phacoemulsification resulted in a statistically significant reduction in peak flare values and a shorter postoperative recovery to normal flare values.42–44 Kruger et al. 40 showed higher flare levels in patients subjected to longer phacoemulsification times (over 80 s) at postoperative week one. These differences had disappeared by 1 month. A similar influence of phacoemulsification time on postoperative flare was seen by Kaur et al., 45 who found phacoemulsification time is affected by different femtosecond laser pretreatment patterns. Though the aim of phacoemulsification and irrigation/aspiration is to fully remove both nucleus and cortex, retained cataract fragments are often seen in the postoperative setting. Though retained fibres were not found to increase postoperative flare in a small study, 46 Nishi et al. found a prolonged postoperative flare in eyes with retained lens epithelial cells at days 6–14.
An observational study on femtosecond laser cataract surgery showed elevated flare levels for 1 month postoperative, 47 which returned to preoperative levels by 3 months. Comparative studies of flare measurements between femtosecond laser fragmentations of the lens compared to manual phacoemulsification shows mixed results. Femtosecond use was associated with a statistically significant decrease in flare at all time points up to 1 month by Abell et al. 48 and Chen et al., 49 however it was not associated with a statistically significant difference in aqueous flare in a study of 110 patients by Pahlitzch et al. 50 or Favuzza et al. 51 in their case series of 40 eyes undergoing femtosecond laser pretreatment. A matrix lens fragmentation pattern was associated with reduced postoperative flare values compared with a phacoemulsification ‘chop’ method. Proponents of femtosecond suggest that it can facilitate more efficient lens removal and less phacoemulsification time, which is known to decrease postoperative flare. 45 Femtosecond laser-assisted posterior capsulotomy at the end of surgery (to avoid posterior capsule opacification) did not increase flare compared to controls. 52
Ophthalmic viscosurgical devices
Ophthalmic viscosurgical devices (OVDs) are essential in modern cataract surgery in protecting structures, compartmentalizing key areas and maintaining pressure gradients. 53 Complete removal of OVD at the end of the case is important to avoid postoperative intraocular pressure (IOP) spikes. 54 In the early 1990s, studies investigated flare in postoperative patients with retained OVD, assumed by the absence of the physiological ‘warm current’ detected at the slit lamp. It identified that those with loss of the warm current had higher flare levels for up to 1 week postoperatively compared to those with a normal warm current.55,56 Use of the viscoadaptive OVD Healon 5 (Johnson & Johnson, New Jersey, USA) in cataract surgery was associated with lower flare values than the dispersive OVD Viscoat (Alcon laboratories, Fort Worth, Texas, USA) up to 8 h postoperatively. 57
Irrigating solutions and antibiotics
Irrigation solutions are used during phacoemulsification to facilitate cataract removal and maintain AC stability. Roberts found that adding the miotic carbachol at the end of surgery increased postoperative flare, probably due to a delayed re-establishment of the blood–aqueous barrier, while acetylcholine reduced flare compared to administration of a balanced salt solution control. 58 However, this was likely due to increased iris manipulation of the unconstructed pupil at the end of surgery. Kohnen et al. 59 also found the addition of heparin to the irrigation solution reduced postoperative flare on days 1 and 3, a difference that disappeared after 1 year. Unfortunately there are no studies that investigate the influence of intracameral antibiotic prophylaxis on aqueous flare as measured by LFP.
Intraocular lenses
LFP found similar flare rates across the previous generation of rigid poly (methyl methacrylate) (PMMA) IOLs inserted during extracapsular cataract extraction. 60 Modern IOL technology encounters materials that include silicone and acrylic with hydrophobic or hydrophilic properties. Foldable silicone and acrylic lenses which could be inserted through a smaller corneal incision than the rigid PMMA lens was not associated with a statistically significant difference in flare in patients with no medical history,61–63 diabetes 64 and diabetic retinopathy. 65 Initially, it was thought that more postoperative flare was seen in hydrophobic than hydrophilic IOLs, 66 however the results of this short follow-up study (3 months) have not been replicated. No differences in postoperative flare have been seen among three different IOL models by Monnet et al. 67 and four types of foldable lenses from Schauersberger et al. 68 Similar flare values were seen in the 1CU (HumanOptics Ag, Erlangen, Germany) accommodative IOL. 69 Abela-Formanek et al. 70 studied six different IOLs in 2002 and found the hydrophobic acrylic AR40 lens (Allergan) to have statistically less flare than hydrophilic acrylic or hydrophobic silicone counterparts at 1 year postoperative. However, the authors could not identify an explanation for this. In addition, there were no correlations with other inflammatory parameters and these results were not replicated. There were no differences in postoperative flare in patients with uveitis who were given hydrophobic acrylic, hydrophilic acrylic or silicone IOLs.70,71 Early evidence showed that IOLs with Heparin Surface Modification (HSM) were thought to reduce the inflammatory response, particularly in eyes at higher risk of postoperative inflammation such as uveitis. 72 Indeed, HSM of hydrophobic IOLs was associated with less flare on postoperative day 1 in two studies,73,74 but both of these results were not statistically significant in subsequent visits. By contrast, there were no differences in flare in patients with diabetes receiving a HSM PMMA IOL versus an uncoated hydrophobic acrylic IOL 64 or in patients with either diabetes and/or pseudoexfoliation syndrome receiving either a coated or uncoated IOL. 75 In 1997 Alió et al. investigated the flare associated with haptic placement post extracapsular cataract surgery. They found that on day 1, the highest flare was in the eyes with IOL implantation in the sulcus, and the lowest flare was when IOL implantation was in-the-bag implantation. 62 In patients without capsular support, Cellini et al. found higher postoperative flare in scleral-fixated IOLs compared with iris fixation or control in-the-bag placement. 76 Amino and Yamakawa found that sulcus-to-sulcus fixation resulted in higher flare values than the sulcus-to-ciliary body or in-the-bag fixation. 77
Postoperative regimen
The postoperative regime after cataract surgery aims to minimize infection, inflammation and the development of pseudophakic cystoid macular edema (CME), which often impedes visual recovery.
Aqueous flare is often seen alongside CME as both result from a compromised blood–aqueous barrier.78–81 Therefore, flare assessment is important to those investigating methods of reducing the incidence of postoperative CME. The major therapeutic classes are corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs), which have been shown either alone or combined to decrease postoperative flare compared with vehicle controls.82,83 They are particularly important for patients at higher risk of postoperative inflammation, such as uveitis, pseudoexfoliation syndrome, or diabetes. 84 A summary of relevant studies is found in Table 1.
Selected studies comparing the effect of two different postoperative medications on post-cataract surgery aqueous flare. Comparisons are between the medication in question in the vertical rows and comparator medication in the columns. Upward arrows indicate an increase in flare compared to the comparator medication and downward arrows indicate a decrease. Sideways arrows indicate no statistically significant difference between the medications.
Corticosteroids are well-known anti-inflammatory molecules and topical use has been a historical mainstay of ocular anti-inflammatories. Several steroid injections administered at the end of surgery do not appear to reduce flare compared with topical steroid controls, including subconjunctival betamethasone111,112 or methylprednisolone 113 and a single perioperative subtenon dose of triamcinolone (either 20 or 30 mg) 114 in uncomplicated cataract surgery, subconjunctival dexamethasone in normal eyes and those with diabetes115,116 or the intracameral dexamethasone delivery system IBI-10090 (Dexycu®, Massachusetts, USA). 117 There were no statistically significant differences in postoperative flare between postoperative prednisolone 1% drops and a dexamethasone 0.1% gel (Dexagel®, Bausch & Lomb, Ireland). 118 The AC dexamethasone delivery system Surodex® (Oculex, California, USA) did reduce postoperative flare compared with a 4-week course of dexamethasone eye drops in patients undergoing extracapsular extraction.119,120
The side effects of steroids are well recognized, 121 and LFP has allowed for investigating alternative NSAIDs in the postoperative period. As a sole treatment, Indomethacin appears to reduce flare over flurbiprofen85,86 and when combined with prednisolone, reduces flare more than combination diclofenac and prednisolone for the first three postoperative days. 87 Indomethacin also appears non-inferior to ketorolac. 88 Diclofenac 0.1% is associated with less flare than flurbiprofen86,90 or bromfenac 92 and is similarly effective as ketorolac 0.5%. 91 Ketorolac has an equivalent effect on flare as the now discontinued rimexolone. 95 Nepafenac 0.1% reduces aqueous flare more than ketorolac 0.4% and bromfenac 0.09% at the 4-week interval 99 and reduces flare more than diclofenac 0.1% when both are combined with steroids at days 15 and 30. 98 It appears comparable to preservative-free diclofenac 0.1% in diabetic and nondiabetic patients. 97 Twice daily bromfenac 0.09% does not improve postoperative flare compared to once daily dosing. 122 Flare comparisons between topical corticosteroids and NSAIDs have yielded no statistically significant differences between the following: dexamethasone versus flurbiprofen, 101 dexamethasone versus diclofenac,102–104 dexamethasone versus bromfenac 105 ; bromfenac versus betamethasone 108 or bromfenac versus fluorometholone 0.1% 109 ; diclofenac versus 1% prednisolone, 123 fluorometholone/levofloxacin versus tobramycin/dexamethasone. 107 Erichsen et al. found no difference in postoperative flare between those taking combination prednisolone 1%/ketorolac 0.5% versus ketorolac 0.5% alone, 96 as did Ylinen et al. with diclofenac 0.1%/dexamethasone 0.1% versus diclofenac 0.1%. 97 Intraoperative administration of subtenon dexamethasone depot does not reduce flare as effectively as a topical regime of ketorolac 0.5% with or without prednisolone 1%. 96
Several NSAIDs have been shown to reduce flare more effectively than topical steroids. Bromfenac was shown to reduce flare more than a course of betamethasone followed by a fluorometholone wean in both patients with diabetes and their healthy controls; a difference was seen in the control arm at 2 weeks and the diabetes arm at 4 and 6 weeks. 110 Nepafenac 0.1% appears to reduce flare more strongly than fluorometholone resulting in reduced fluorescein angiographic evidence of CME and hastened visual recovery. 100 Bromfenac 0.09% is associated with lower flare levels at 3 and 6 months postoperatively compared with dexamethasone 0.1%. 106 Diclofenac reduced postoperative flare compared to fluorometholone up to 8 weeks postoperatively. 93 It also reduces flare significantly compared to dexamethasone in both normal eyes and those with pseudoexfoliation.29,94 By contrast, Zhang et al. 89 identified that dexamethasone reduced postoperative aqueous flare more than indomethacin. The intravitreal dexamethasone drug delivery system Ozurdex® (Allergan, Dublin, Ireland) reduced postoperative flare for weeks 1–4. However, the standard of care comparison group received no anti-inflammatory therapies. 124
Postoperative antibiotic prophylaxis are frequently used to reduce the risk of endophthalmitis. Though this is a well-studied area, the authors cannot find studies that investigate the influence of postoperative antibiotic medication on aqueous flare, nor are there comparisons between different postoperative antibiotic regimes that utilize LFP to assess anterior segment inflammation after cataract surgery. Some studies as described in this review investigate combination antibiotic/anti-inflammatory medication but no meaningful conclusions about the contribution of the antibiotic component can be drawn.
Ophthalmic solution preservatives contribute to conjunctival inflammation and disrupt the corneal epithelium, but these effects are primarily seen in their long-term use. 125 Hessemer et al. 126 found no statistically significant difference in aqueous flare between groups given preservative-free diclofenac 0.1% postoperatively compared to preserved diclofenac 0.1%. The addition of preservative-free diclofenac, when given both pre- and postoperatively, appeared to reduce flare compared to when given postoperatively alone. However, the study arms were limited to 30 patients only and follow-up extended to 7 days postoperatively. In patients with diabetic retinopathy, Yasuda et al. 127 found patients treated with preservative-free diclofenac postoperatively showed a faster return to preoperative flare values than preserved formulations.
Traditional Chinese and Sino-Japanese herbal remedies have shown statistically significant reductions in the postoperative flare on days 1, 3 and 5 with Orengedoku-to in uncomplicated eyes and Kakkon-to in both uncomplicated and uveitic eyes.128,129 However, these results have since not been replicated, follow-up was limited to 7 days postoperatively and the control group in one of these studies received no therapies. 128 Topical miotics given postoperatively do not appear to affect flare. 130
Discussion
LFP is an established technology in assessing intraocular inflammation after cataract surgery. It identifies those at risk of postoperative inflammation, including older age and the presence of diabetes or uveitis. Ursell et al. showed that postoperative flare is also related to preoperative flare (p = 0.001, r2 = 9.0%) as well as the preoperative colour of the cataract (p = 0.038, r2 = 3.3%). 26 However, the properties of the cataract may falsely change the flare due to increased background light scatter. 131 Cataract surgery technique influences postoperative flare and can be minimized with small incisions, efficient phacoemulsification and in-the-bag IOL placement. The majority of IOLs used today do not affect flare differently from one another. Finally, laser flare measurements are critical in the ongoing pursuit of the optimum postoperative anti-inflammatory regime. A recent phase III, multicentre, randomized trial has demonstrated the non-inferiority of a new topical antibiotic/anti-inflammatory agent levofloxacin 5 mg/mL + dexamethasone 1 mg/mL for 1 week in comparison to tobramycin 3 mg/mL + dexamethasone 1 mg/mL (Tobradex®, Novartis Pharmaceuticals, UK) for 2 weeks after cataract surgery but still relied on subjective aqueous flare measurements. 132 Many other studies investigating topical medications postoperatively rely on the clinical grading of flare. These studies find little to no flare postoperatively but this is probably more due to the low sensitivity of the clinical grading system as opposed to representing a true absence of flare.133,134
Levels of aqueous flare negatively correlate with visual function in other conditions such as uveitic macular edema 135 and retinitis pigmentosa. 136 Aqueous flare levels mostly seen after cataract surgery are unlikely to be of visual significance to most patients. However, as a marker of intraocular inflammation, flare is strongly associated with visually significant outcome measures such as cystoid macular edema. 81 In patients with pseudophakic cystoid macular edema, LFP did not correlate with retinal thickness but did correlate significantly with reduced visual acuity. 79 LFP was analysed in 42% of 187 studies investigating both AC inflammation and macular changes after cataract surgery in a review by De Maria et al. 81 It was found to correlate with the frequency of CME postoperatively better than other methods, including clinical grading and aqueous humour sampling. This is probably because of its superior ability to quantify the low levels of inflammation often seen after uncomplicated cataract surgery. Its routine use post-cataract surgery may help predict those at risk of inflammatory decompensation and allow treatment before significant morbidity occurs. Its validation as a sensitive, objective marker of postoperative inflammation has allowed for new associations with intraocular inflammation. For example, it is now shown that cases with raised LFP values at day 1 postoperative tended to also demonstrate increased flicker electroretinogram (ERG) responses at week 1. 137 Increased flare is also seen in eyes with anterior capsular contraction 138 and is interestingly detected in the fellow non-operated eye on day 1 postoperatively. 139
This review consists mainly of well-designed prospective, randomized studies with clear outcome measures and statistically significant results. However many studies did not include a flare measurement protocol. Pupillary dilatation prior to flare measurement may influence values14,23,140 and it is unknown in many whether this was performed. Many followed a similar protocol to that of Ursell et al. 26 whereby seven dilated readings with <15% between background readings were taken, the two extreme values were discarded and the mean and standard deviation were taken from the remaining five. However, many studies included different approaches to background readings and extreme values before calculating mean and standard deviation and a number did not include a protocol at all. Methods for flare measurement must be included in each study and a standardized protocol in this area is welcome. Longer patient follow-up is also required, with very few studies having follow-up of greater than 12 months; despite evidence that aqueous flare in the eyes post-cataract surgery can persist for up to 36 months. 141 Though the majority of research has been in the assessment of topical postoperative anti-inflammatory agents, little has been done to investigate the influence of antibiotic prophylaxis on anterior segment inflammation, either through intracameral antibiotics at the end of surgery or through postoperative drop regimes. Future studies should utilize LFP to measure the influence of cataract surgery antibiotics on blood–aqueous breakdown.
Conclusion
Despite its success, the worldwide frequency of cataract surgery creates substantial morbidity associated with blood–aqueous breakdown and inflammatory sequelae. The optimization of an already refined operation with modest postoperative changes cannot rely on clinical grading, which is qualitative and fraught with interobserver variability, particularly at the lower levels seen after an uncomplicated cataract surgery. In this review we highlight the influence of preoperative pathological states such as uveitis and diabetes where the blood–aqueous barrier is compromised prior to surgery. Various intraoperative strategies can minimize postoperative flare, including clear corneal incisions, efficient phacoemulsification and in-the-bag IOL placement. The optimum postoperative anti-inflammatory regimen is subject to ongoing study and care must be taken to include the influence of antibiotic prophylaxis. Aqueous flare is far from the only consideration in the assessment of cataract surgery strategies but it is only with objective and sensitive outcome measures of intraocular inflammation that the refinement of modern cataract surgery can continue.
