Prospective randomized comparative study of macular thickness following phacoemulsification and manual small incision cataract surgery

Introduction

Globally, cataract is the leading cause of blindness and leading cause of low vision in all sub-regions (Resnikoff et al. 2004). Worldwide cataract surgery is the major workload of most eye care units and is a major health care expense (Ellwein & Urato 2002). Phacoemulsification has become the treatment of choice for cataract surgery over the past two decades. Manual small incision cataract surgery (MSICS) with non-foldable poly methyl methacrylate (PMMA) lens implantation is a low-cost surgical technique that has become popular for high volume surgeries in developing countries. Both phacoemulsification and MSICS are sutureless surgeries with low complication rates and satisfactory visual outcomes (Ruit et al. 2007). However, phacoemulsification is involved with higher capital investment, higher cost of consumables and higher average surgical time than MSICS (Gogate et al. 2005). Phacoemulsification machine is not required in MSICS in which nucleus is removed through a 6.5-mm sclerocorneal tunnel. Several centres have reported success with MSICS (Hennig et al. 2003; Venkatesh et al. 2005).

Cystoid macular oedema (CME) is a cause of vision loss after cataract surgery. Surgical trauma to the iris, ciliary body and lens epithelial cells causes release of phospholipids and thereby release of prostaglandins or other inflammatory mediators. Post cataract surgery anterior chamber inflammation can be minimized by topical 0.1% bromfenac and 0.1% betamethasone (Miyanaga et al. 2009). The inflammatory mediators diffuse through the vitreous and disrupt the blood retinal barrier to cause pooling of serum in retina resulting in macular oedema (Miyake & Ibaraki 2002; Yavas et al. 2007). Peak incidence of CME varies between 4 and 12 weeks after surgery (Ginsburg et al.1995). Incidence of clinical CME has been reported to be 0–6%. An incidence of macular oedema in 8.1% cases has been reported after phacovitrectomy (Li et al. 2009). Diseases that affect the blood retinal barrier like uveitis and diabetes increase the risk for CME. The risk factors also include surgical complications like rupture of posterior capsule, vitreous loss and iris incarceration (Rossetti & Autelitano 2000). However, several authors have reported sub-clinical perifoveal oedema following uneventful phacoemulsification. It can be detected by measuring macular thickness values with optical coherence tomography (OCT). The thickness value of the perifoveal 3.0 and 6.0 mm sectors alone or averaged together with the foveal value provides useful description of the postoperative macular oedema (Biro et al. 2008). Macular thickness values following MSICS have not been reported in literature (Medline search). Compared to phacoemulsification, MSICS is associated with more tissue trauma because of a larger incision size and more iris manipulation. These may increase the chance of postoperative macular oedema following MSICS compared to phacoemulsification. Although cystoid changes in CME are visible by slit-lamp biomicroscopy and fluorescein angiography, only OCT can quantitatively assess the retinal thickness with demonstration of any associated structural anomalies of retinal pigment epithelium (Voo et al. 2004; Sander 2009). Linear OCT is a non-contact and non-invasive technique for high resolution cross-sectional imaging that can identify and confirm oedematous changes in retina at selected points on the macula with high sensitivity, specificity and better reproducibility than angiography (Hee et al.1995). Spectral domain OCT has been demonstrated in one study to be superior to time domain OCT in the investigation of retinal response to cataract surgery (Stock et al. 2009). The localized intraretinal fluid accumulation in CME appears on OCT as intraretinal cystoid spaces of low reflectivity with highly reflective septa separating these spaces.

The aim of our study was to compare postoperative macular thickness following uneventful phacoemulsification with acrylic lens and MSICS with PMMA lens implantation. The main outcome variable was macular thickness value by OCT.

Methods

This study was carried out in a tertiary eye care hospital in Kolkata, India between April 2007 and April 2008 with approval from the Institutional Ethics Committee. Patients between 50 and 75 years with age-related cataract attending the outpatient department were included in the study with their informed consent. All applicable regulations concerning the ethical use of human volunteers were followed.

Patients with fasting plasma glucose ≥ 126 mg/dl or any treatment history for diabetes, history of previous eye surgery, present or past history of uveitis, ocular disease other than cataract, history of significant eye trauma and axial length more than 26.5 mm were not invited to participate. Patients with 3+ or more flare (Standardization of Uveitis Nomenclature Working Group) on the 1st postoperative day were also excluded from study.

Sample size was calculated assuming 1:1 randomization, 80% power (two-tailed beta of 0.20), 1% precision error and 18.43 μm standard deviation (Tewari et al. 2004) to detect a difference in two groups of 10 μm or more in macular thickness. The sample size was calculated to be 102 in each group, i.e. 204 in total. To account for loss in follow-up (up to 10%), a total of 224 patients were recruited in the study.

Each patient was allocated to either MSICS or phacoemulsification group by drawing ballots from a sealed envelope. Each group had 112 patients. In each patient, the eye with more advanced cataract was included as study eye. Allocation to each surgeon was also performed during drawing of ballots, the two procedures being equally distributed among two surgeons (PNB, SG). Each surgeon performed 112 surgeries, 56 each of MSICS and phacoemulsification. The participating surgeons were not involved with the allocation procedure and were masked concerning the method of surgery until the patients were prepared on the table. The patients were masked to the allocation code until surgery was performed. Five patients in MSICS group (corneal oedema 1, iridodialysis 1, posterior capsular rupture 1, sulcus fixation IOL 1, 3+ flare 1) and nine patients in phacoemulsification group (corneal oedema 4, posterior capsular rupture 4, 3+ flare 1) were further excluded from the study because of various peroperative and postoperative complications necessitating alteration in management protocol. The two planned methods were phacoemulsification and MSICS with implantation of foldable hydrophobic acrylic lens (Aurolab, Madurai, India) and non-foldable polymethyl methacrylate (PMMA) lens (Aurolab), respectively.

Besides all relevant preoperative examinations, A and B scan ultrasonography (USG) was performed to detect any posterior vitreous detachment (PVD) and patients were classified as having no PVD, complete PVD and incomplete PVD. In our study, USG, OCT and visual acuity measurement were performed by a single masked investigator (LM). All patients received preoperative treatment with topical ofloxacin (0.3%) 3 days prior to surgery. We used combination of phenylephrine (5%) and tropicamide (0.8%) eye drop to dilate pupil on the day of surgery. We instilled povidone iodine solution (5%) in the conjunctival fornix prior to surgery. Anaesthesia was given by peribulbar injection of 3 ml each of 2% lignocaine and 0.5% bupivacaine along with hyaluronidase. Phacoemulsification was performed by stop-and-chop technique after 2.6 mm clear corneal section with in-the-bag implantation of hydrophobic acrylic foldable lens. In MSICS, 6.5-mm sclerocorneal tunnel was prepared with frown-shaped scleral incision. Nucleus was delivered out of anterior chamber by irrigating vectis. Capsulorhexis was performed in all cases with in-the-bag lens implantation. Post operation, all patients received a 21-day course of prednisolone acetate (1%) four times a day for 1 week followed by twice daily dosage for 14 days along with ofloxacin (0.3%) and tropicamide (1%) eye drop. Patients were examined on 1st, 7th, 42nd and 180th postoperative day. All patients completed the 1st-day and the 7th-day follow-up. In phacoemulsification group, 97 patients came for the 42nd-day follow-up and 94 patients completed the 180th-day follow-up. In MSICS group, 100 patients completed the 42nd-day follow-up and 99 came for the 180th-day follow-up. Four cases of phacoemulsification and one case of MSICS, although excluded from the study owing to corneal oedema, had macular thickness measurement performed on the 42nd and the 180th day after their cornea became clear. The values of these five patients were not included in the average thickness readings for 42nd and the 180th day. The cases excluded from the study because of 3+ flare (two cases) and posterior capsular rupture (five cases) were also followed up with serial measurements by OCT. Dilated fundus examination was performed at each visit. Best Corrected visual acuity was recorded at each visit with Snellen’s visual acuity chart and converted to logMAR value.

OCT scan was performed with Stratus OCT-3 (Carl Zeiss Meditec, Inc, Dublin, CA, USA). The scan protocol used was the Fast Macular Thickness Map algorithm. The best quality macular thickness maps (signal strength five or more) at each visit were analysed to produce the Macular Thickness Map, consisting of thickness values in nine subfields distributed in three concentric circles with diameters of 1, 3 and 6 mm.

Data were entered into a Microsoft Excel spreadsheet and analysed using spss (version 12.0, SPSS, Inc., Chicago, IL, USA). Macular thickness value in central fovea, four quadrants of inner retina and four quadrants of outer retina were compared by using student’s t-test. Same test was used for age, gender, eye laterality and surgeon variations. Correlation test was performed between macular thickness and logMAR visual acuity. Relation of PVD and macular thickness was tested by chi-square test. Spearman’s rank correlation test was used to compare visual acuity after two methods of surgery.

Results

The mean age of patients was 62 ± 6 and 61 ± 6 years in MSICS and phacoemulsification group, respectively. Eighty-eight (78.6%) patients in MSICS group and 91 (81.2%) patients in phacoemulsification group had preoperative visual acuity of 6/36 or less. Number of male patients was 125 (55.8%; 95% CI 53.8–57.8%). The rest were female. Fifty-eight per cent in MSICS and 53.5% in phacoemulsification group were male. There was no significant difference in age (p = 0.535) and sex (p = 0.501) between two groups. Right eye was operated in 49.1% of MSICS group and 50.9% of phacoemulsification group with no significant difference (p = 0.798). Clinical macular oedema was diagnosed in no patient at any visit. On the 1st postoperative day, central subfield mean thickness (CSMT) in MSICS group was 192.8 ± 17.9 and that in phacoemulsification group was 192.1 ± 27.4 μm, with no significant difference (p = 0.12). After comparing the values in all subfields, significant difference was found in none on the 1st day (Table 1). On the 7th day, CSMT in MSICS group (198.9 ± 21.4 μm) was significantly (p = 0.04) more than in phacoemulsification group (193.1 ± 19.3 μm) (Table 2). Significant difference was also found in superior inner (p = 0.04), inferior inner (p = 0.04), temporal inner (p = 0.04) and temporal outer (p = 0.04) subfields. On the 42nd day, CSMT in MSICS group was 207.8 ± 26.3 μm and that in phacoemulsification group was 198.3 ± 23 μm, the difference being significant (p = 0.007). Besides central 1 mm, significant difference was observed in nasal outer (p = 0.02) and inferior inner (p = 0.05) subfields. On 180th day, CSMT in MSICS group (198.7 ± 17.3 μm) was significantly more (p = 0.04) than phacoemulsification group (193.1 ± 21.2 μm). Besides central 1mm, significant difference was also observed in superior inner (p = 0.01), nasal inner (p = 0.02) and inferior inner (p = 0.03) subfields. The macular thickness of the five cases that were excluded for corneal oedema was evaluated on the 42nd and the 180th day after clearance of oedema. When these values along with the values of two patients with 3+ flare and five patients with capsular rupture were included in the analysis, the difference in CSMT between the two groups was still statistically significant.

On the 1st postoperative day, CSMT in male and female patients were 195.8 ± 17.9 μm and 192.1 ± 27.4 μm, respectively, with no significant difference (p = 0.12). In phacoemulsification group, superior inner subfield demonstrated significantly (p = 0.04) higher thickness in male (263.6 ± 16.8 μm) compared to female (255.1 + 22.3 μm) on the 7th day. Other sectors did not show any significant difference between the male and female patients throughout the study. In MSICS group, superior inner subfield demonstrated significantly higher thickness in male compared to female on the 7th (267.1 ± 17.9 μm versus 257.7 ± 21.9 μm; p = 0.02), the 42nd (272.6 ± 17.7 μm versus 263.1 ± 20.3 μm; p = 0.002) and the 180th day (265.3 ± 17.4 μm versus 258.7 ± 18.8 μm; p = 0.02). In MSICS group, other subfields did not show significant difference between the male and female patients throughout the study except inferior inner on the 7th day (p = 0.01) and temporal inner on the 42nd day (p = 0.03). The two cases (one in each group) with 3+ flare were followed up with OCT and neither of them developed CME. When their values were included in the analysis, the difference in CSMT between two groups was still statistically significant. Ninety-six cases had no PVD, 68 cases had complete PVD and 46 cases had incomplete PVD as diagnosed on USG. There was no significant difference in CSMT in complete PVD, incomplete PVD and no PVD patients following either type of surgery.

Best corrected visual acuity of 6/12 or more was achieved on the 42nd day in 100% patients of both groups. Comparing visual acuity after two methods, rank correlation coefficients (r) on the 7th day and the 42nd day were 0.69 and 0.66, respectively, indicating no significant correlation. On the 180th day, there was perfect correlation (r = 1) signifying identical visual outcome. Trend chi-square test for visual acuity on the 42nd day also showed no significant difference (p = 0.4). Correlation coefficient estimated for visual acuity and macular thickness in different sectors ranged between 0.01 and 0.36 on the 42nd and the 180th day.

Discussion

We did not observe clinical macular oedema in any of our study population. Similar observation was carried out in other studies after uncomplicated phacoemulsification (Mentes et al. 2003; Biro et al. 2008). Final visual outcome was identical at 6 months follow-up with two methods in our study. Other studies have reported similar result (Gogate et al. 2005; Ruit et al. 2007).

The fast macular thickness map protocol was used as this widely used protocol provides regional macular thickness measurements with good repeatability and reproducibility (Polito et al. 2005; Danis et al. 2008). We excluded patients with large axial length as positive correlation of foveal thickness with axial length has been reported (Lam et al. 2007). On the 1st day, there was no significant difference in macular thickness in any sector between two groups. Maximum thickness was observed in both groups on the 42nd day. This is consistent with the findings of Biro et al. (2008), who observed maximum mean foveal and perifoveal thickness at 1 and 2 months follow-up. We did not observe any relation between PVD and macular thickness as was also found by Gulkilik et al. (2006).

In our study, significantly increased CSMT was observed in MSICS group compared to phacoemulsification group on all follow-up days except on the 1st day. This was probably attributed to more tissue trauma and iris manipulation associated with MSICS. Larger incision size was associated with more severe blood aqueous barrier break down in a study comparing phacoemulsification with extracapsular cataract extraction (Pande et al.1996). No such comparative data between phacoemulsification and MSICS are available in literature.

In our series, the material of intraocular lens was different (PMMA and hydrophobic acrylic) in two groups. We did not use same (acrylic) lens in both groups, because in practice, it is a PMMA lens that is most commonly used in MSICS and our aim was to compare the outcome of two methods as they are actually performed.

The study was internally valid as results from patients who were excluded in the postoperative period because of 3+ or more flare, capsular rupture and significant corneal oedema interfering with first postoperative day OCT, when analysed on intent to treat basis, did not make any qualitative change in the results. The participating surgeons had ten or more years of postresidency experience, with a minimum of 750 phacoemulsification and 1500 MSICS. Hence, the study can also be considered externally valid.

One weakness of our study was the absence of preoperative thickness value because of the presence of significant media opacity in many patients interfering with good quality OCT scan. Macular thickness on the first postoperative day was taken as baseline value in our study as no significant change in the foveal and perifoveal thickness values occurs on the first postoperative day (Degenring et al. 2007; Biro et al. 2008). Another weakness was the absence of control value from un-operated fellow eye. In most cases, fellow eye either had cataract or was pseudophakic.

It is already reported that post phacoemulsification, significant increase in macular thickness can be detected on the postoperative 7, 30 and 60 days in the perifoveal 3.0 and 6.0 mm sectors (Biro et al. 2008). Literature search did not reveal any such study with MSICS patients. In our study, the increase in macular thickness was sub-clinical and did not affect final visual outcome in any patient. Our study shows that in spite of the greater chance of iris and lens manipulation and theoretical risk of increased postoperative inflammation, there is no evidence of CME on OCT following MSICS. However, we excluded all patients who were at higher risk of developing macular oedema. Our study results emphasize the need to investigate further the influence of these two surgical procedures on macula in those patients who are at higher risk of developing macular oedema.

Acknowledgement

This was presented in part at 67th annual conference of All India Ophthalmological Society at Jaipur, India, 5–8 February 2009.