Sufficient mydriasis is required for safe and efficient cataract surgery. Traditionally, topical mydriatic agents have been the preferred way to achieve mydriasis. However, there are some disadvantages to using topical mydriatics. These include slow penetration of mydriatic substances through the cornea rendering slow pupil enlargement,1 which delays preoperative pupillary dilatation; diminishing mydriatic effect intraoperatively especially in prolonged surgery; and significant systemic absorption with additional risk of systemic side effects.2 The concept of using intracameral mydriatics of various formulations has gradually gained acceptance,3-5 as this can avoid some of the disadvantages of topical mydriatics.
In cataract surgery, the risk of corneal endothelial damage is determined by several preoperative and intraoperative parameters. The preoperative parameters include older age,6 small pupil diameter, firmness of the nucleus,6 and shorter axial length (AL).6,7 Intraoperative parameters include phacoemulsification time,6-9 phacoemulsification technique,9,10 type of ophthalmic viscosurgical device (OVD) used,11 and toxic intraoperative medications, for example, the use of intracameral solution.12 Intracameral phenylephrine is one of the mydriatic agents used in small pupil cataract surgery for mydriasis of the pupil, maintaining dilation throughout the surgery,13 and preventing complications due to intraoperative floppy iris syndrome (IFIS), induced by tamsulosin14,15 or other α-1 adrenergic antagonists.16 However, the safety of intracameral phenylephrine remains a concern. The toxicity to corneal endothelial cells may be attributed to the pH of the solution or its chemical composition.
Most studies have used endothelial cell count as the marker for corneal health, yet others17,18 have shown that morphometric analysis of individual cell size and shape provides a more sensitive indicator of endothelial cell damage than cell density measurements alone. Therefore, analysis of quantitative and morphologic parameters can provide insight into the potential toxic effects of intracameral phenylephrine.
Although there is increasing use of intracameral mydriatics for cataract surgery, not many studies have been done to compare the risk of postoperative corneal endothelial cell loss and morphological changes between eyes receiving intracameral phenylephrine during phacoemulsification and eyes which do not. Studies that were conducted had very small sample sizes. Therefore, we performed a study in a larger sample size to ascertain whether intracameral phenylephrine 1.5% (from Minims phenylephrine 10%) has toxic side effects on the corneal endothelium in Malaysian eyes.
MATERIALS AND METHODS
This was a randomized controlled trial comprising 295 patients who underwent phacoemulsification surgery at Penang General Hospital from June 23, 2014, to May 31, 2016. This included 68 patients who had enrolled in the clinical trial “Intracameral Mydriatics versus Topical Mydriatics for Pupillary Dilatation in Phacoemulsification Cataract Surgery,” which was conducted at Penang General Hospital from June 23, 2014, until December 31, 2014. Patients aged from 18 to 70 years old with immature cataract of nuclear opacity of grade V or less according to Lens Opacity Classification System III (LOCS) were recruited. Exclusion criteria included precious or single-eyed patients, patients with history of uveitis and pupillary abnormalities such as posterior synechiae, shallow anterior chamber less than 2.5 mm, previous intraocular surgery, past ocular trauma, pseudoexfoliation, polar and dense cataracts, combined surgery (glaucoma, vitreoretina), chronic miotic or α blocker usage, and systemic diagnoses that were thought to affect pupillary dilatation such as proliferative diabetic retinopathy patients with history of laser panretinal photocoagulation or Horner syndrome. Conditions affecting the corneal endothelium including pre-existing corneal scar or dystrophy, low preoperative endothelial cell density less than 1500 cells/mm2, complicated phacoemulsification surgery such as posterior capsule rupture or zonular dialysis with vitreous loss, and additional methods needed to dilate pupil were also excluded. Informed consent was taken from all patients. Ethical approval for this study (NMRR-14-1536-23600 and FF-2015-097) was provided in accordance with the Declaration of Helsinki by the Medical Research and Ethics Committee of Malaysia on April 24, 2015, and the Ethics Committee of The National University of Malaysia on March 25, 2015.
Baseline preoperative central corneal endothelial cell density (ECD), coefficient of variation (CV), and percentage of hexagonal cells were measured with specular microscope CEM-530 (Nidek Co, Ltd). An average of 3 readings were taken per eye. Coefficient of variation was defined as the degree of variation in the size of the endothelial cells, which is an index of “polymegathism.” It was calculated as the standard deviation of the cell sizes for a specimen divided by the mean cell size for the same specimen. A CV less than 40 was normal.19 Percentage of hexagonal cells was defined as the variability in hexagonal cell shape over time. It was an index of “pleomorphism.” Hexagonality above 50% was taken to be normal.19 The lens nuclear density and color were graded preoperatively using the LOCS III protocol under slit lamp examination to rule out preoperative differences between the 2 groups.
The study subjects were randomized into 2 groups, which were the topical group and the intracameral group (ICM). Randomization was done using the envelope technique. The topical group was subjected to topical dilating eye drops, 2 drops each of tropicamide 1% and phenylephrine 2.5% at 15-minute intervals for 60 minutes before the operation. The ICM group received placebo topical eye drops (normal saline 0.9%) while waiting for surgery and a sterile intracameral solution, which contained 0.2 mL of a preservative-free mixture of lignocaine 1% and phenylephrine 1.5% (Minims phenylephrine hydrochloride 10% preservative-free; Chauvin Pharmaceuticals Ltd, Surrey, UK), was injected into the anterior chamber at the beginning of the surgery after paracentesis. The OVD of Healon GV (Abbott Medical Optic Inc, CA) or Discovisc (Alcon Laboratories Inc, Fort Worth, TX) was injected after 60 seconds of pupillary dilatation. All patients were subjected to clear corneal incision using 2.75-mm microkeratome and Trypan Blue 0.06% (Aurolab, Veerapanjan Madurai, India). Both groups were subjected to balanced salt irrigating solution with 1:1000 adrenaline.
The cocktail of intracameral lignocaine 1% and phenylephrine 1.5% was prepared under sterile technique, using preservative-free 0.3 mL phenylephrine 10% with 1.0 mL lignocaine 2% (preservative-free) plus 0.7 mL balanced salt solution into 2.0 mL solution. The pH of the intracameral phenylephrine 1.5% and lignocaine 1.0% solution was 6.0. All operations were performed using standard technique by 1 of 3 surgeons, with a minimum of 3 years' experience of performing cataract surgery (V.N.M., A.E.L., F.S.W.) using the Infinity Phacoemulsification Machine with Ozil torsional technology (Alcon Laboratories Inc, Fort Worth, TX). The total duration of surgery, phacoemulsification time, and power were recorded at the end of the surgery.
Central corneal ECD, CV, and percentage of hexagonal cells were measured using specular microscopy at postoperative weeks 1 and 6 and at 3 months.
Patient characteristics were summarized as means and standard deviations for continuous normally distributed variables, as medians and interquartile ranges for non-normally distributed variables, and as frequencies and percentages for categorical variables. The characteristics of the patients in both groups were compared using independent t test, Pearson χ2 test, or Fisher exact test. Student t test was used to analyze the univariate significance, whereas mixed analysis of variance (ANOVA) was used to analyze the between- and within-group effect on ECD, percentage of hexagonal cells, and CV. The correlation between endothelial cell loss and associated risk factors was done using Pearson correlation analysis. The threshold for statistical significance was set at P < 0.05.
A total of 295 patients were recruited for this study, of which 148 patients were given topical mydriatics before phacoemulsification surgery and 147 patients were given intracameral mydriatics during surgery. The median age was 65 years old (interquartile range, 7 years). The distribution of age was found to be negatively skewed using the Shapiro-Wilk test for normality (W = 0.921, P < 0.001).
There was no significant difference in the median age (65 years vs 65 years, P = 0.838) and sex between the 2 groups (Table 1). Both groups had a higher proportion of females (51% vs 56.8%) (χ2 = 0.977, P = 0.323). However, there was a significant difference in patient ethnicity between the 2 groups (χ2 = 13.737, P = 0.002). There was a major difference of Indian (4.1% vs 13.5%) and other (0% vs 3.4%) races between the 2 groups.
There was also no significant difference in phacoemulsification time (82.53 vs 89.15 seconds, P = 0.157) and phacoemulsification power (25.50 vs 25.48, P = 0.981) between the 2 groups. However, there were significant differences between the 2 groups for lens color (3.35 vs 3.23, P = 0.027), lens opalescence (3.35 vs 3.24, P = 0.035), anterior chamber depth (3.25 vs 3.15, P = 0.017), and operation time (17.63 vs 15.44 minutes, P = 0.002).
Endothelial Cell Density
Univariate analysis using t test showed that there was no significant difference between the ECD and the percentage of endothelial cell (EC) loss in patients in the topical mydriatics group or intracameral mydriatics group at the 4 individual time points (Tables 2, 3). At 3 months, the mean ECD was 2129.76 ± 423.53 cells/mm2 for the ICM group versus 2100.54 ± 393.00 cells/mm2 for the topical group (P = 0.539) and there was a reduction in EC of 18.60 ± 12.79% for the ICM group versus 19.44 ± 11.24% for the topical group (P = 0.550).
A mixed between-within subject analysis of variance (ANOVA) was then conducted to assess the impact of 2 different interventions on corneal ECD across 4 time periods (pre-intervention and 1-week, 6-week, and 3-month follow-up). Mauchly test indicated that the assumption of sphericity had been violated [χ2(5) = 2018.22, P < 0.001]; therefore, degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity (ε = 0.34). Overall, there was no significant interaction and difference between the 2 intervention types (ICM and topical groups) and the 4 time periods (preoperatively, postoperatively at 1 week, 6 weeks, and 3 months) [F (1.03, 301.07) = 0.912, P = 0.343, ηp 2 = 0.003]. Analysis was also done for the 2 main effects, which were time (across the 4 time periods) and intervention type (intracameral and topical). As expected, there was a significant difference across the 4 time periods [F (1.03, 301.07) = 38.19, P < 0.001, ηp 2 = 0.115] with both groups showing a reduction in ECD across the 4 time periods. However, between the 2 types of interventions, there was no significant difference [F (1, 293) = 0.796, P = 0.373, ηp 2 = 0.003].
Because it was noted that 4 factors were significantly different between the 2 groups, analysis was conducted to adjust for lens color, lens opalescence, operation time, and anterior chamber depth. There was no interaction among each of the 4 covariates. Furthermore, there was no significant interaction between intervention type and time period in terms of patient EC density [F (1.642, 474.63) = 0.633, P = 0.501, ηp 2 = 0.002] with mixed ANOVA analysis. The main effect comparing the 2 types of intervention was not significant [F (1, 289) = 0.946, P = 0.332, ηp 2 = 0.003], suggesting no difference in the ECD of the 2 interventions.
Percentage of Hexagonal Cells (Pleomorphism) and Coefficient of Variation (Polymegathism)
Univariate analysis showed no significant difference in the percentage of hexagonal cells and CV of patients between the 2 interventions at the 4 individual time points (Tables 4, 5).
A mixed ANOVA was then conducted to assess the impact of the 2 different interventions on the percentage of hexagonal cells and CV across the same 4 time periods.
For assessment of pleomorphism, correction done using Greenhouse-Geisser estimates showed no significant interaction between intervention type and the 4 time periods in terms of pleomorphism [F (1.03, 302.95) = 1.032, P = 0.313, ηp 2 = 0.004]. There was a significant difference across the 4 time periods [F (1.03, 301.07) = 9.106, P < 0.001, ηp 2 = 0.03] with both groups showing a general reduction in percentage of hexagonal cells across the 4 time periods. However, between the 2 types of interventions, there was no significant difference [F (1, 293) = 0.893, P = 0.345, ηp 2 = 0.003].
After adjusting for confounders there was still no significant interaction between intervention type and time period in terms of endothelial cell percentage of hexagonality [F (2.85, 823.56) = 0.33, P = 0.794, ηp 2 = 0.001]. The main effect comparing the 2 types of intervention was not significant [F (1, 289) = 0.222, P = 0.638, ηp 2 = 0.001], suggesting that there was no difference in the percentage of hexagonality of the 2 interventions.
For assessment of polymegathism, after Huynh-Feldt correction, there was no significant interaction between intervention type and the 4 time periods in terms of polymegathism [F (2.75, 804.88) = 0.406, P = 0.731, ηp 2 = 0.001]. Although there was a significant main effect for time [F (2.75, 804.88) = 111.93, P < 0.001, ηp 2 = 0.276] in both groups, the main effect comparing the 2 types of intervention did not suggest any significant difference in CV between the 2 interventions [F (1, 293) = 0.53, P = 0.818, ηp 2 < 0.001].
Similar results were obtained after adjustment of all 4 confounders (lens color, lens opalescence, operation time, and anterior chamber depth) between intervention type and time period in terms of patient coefficient of variation [F (2.79, 806.68) = 0.647, P = 0.574, ηp 2 = 0.002]. The main effect was also not significantly different between the 2 groups after adjustment of confounders [F (1, 289) = 0.581, P = 0.447, ηp 2 = 0.002], suggesting no difference in the CV of the 2 interventions.
Association of Endothelial Cell Loss With Phacoemulsification Time, Phacoemulsification Power, and Duration of Surgery
Spearman correlation showed that there was no significant association between endothelial cell loss and phacoemulsification time (r2 = 0.01, P = 0.078) and duration of surgery (r2 = 0.001, P = 0.772).
However, there was significant association between phacoemulsification power and endothelial cell loss (r2 = 0.027, P = 0.004).
The use of intracameral mydriatics for phacoemulsification has gained widespread acceptance as many studies were able to demonstrate the effectiveness of intracameral mydriatics in producing pupillary dilatation during cataract surgery.3-5,13 However, the safety profile remains a concern. Hull et al20 reported that intracameral adrenaline with its preservative sodium bisulfite causes damage to corneal endothelial function and ultrastructure in animal eyes; therefore, corneal effects have to be assessed in intracameral mydriatic injection.
Intracameral phenlyephrine complications include EC destruction syndrome, toxic anterior segment syndrome, and endophthalmitis.21 Mean corneal ECD has been reported to be 2400 (range, 1500-3500) cells/mm2 in normal adults.22 The corresponding preoperative value in the present study was 2608 ± 256 cells/mm2 in the ICM group and 2600 ± 280 cells/mm2 in the topical group. This is comparable to a study done in Malaysia,23 which showed a mean ECD of 2648 ± 310 cell/mm2. At present, corneal endothelial injury associated with phacoemulsification is usually assessed by specular microscopy and viewed in terms of the changes in cell density and morphology. The results of our study clearly showed no significant difference between the ECD and the percentage of endothelial cell loss between the ICM group and the topical group. This result was consistent with previous studies. A study by Lundberg and Behndig13 comparing intracameral mydriatics (cyclopentolate 0.1%, phenylephrine 1.5%, and lidocaine 1%) with topical mydriatics (cyclopentolate 1% and phenylephrine 10%) found no significant difference in rates of EC loss and changes in CV and hexagonality among patients. Another study done in Japan24 investigated the efficacy and safety of intracameral injection of 0.5% tropicamide and 0.5% phenylephrine hydrochloride. It showed no significant reduction in corneal ECD at 3 months and 1 year after cataract surgery. The results from 40 of the cases from Lundberg and Behndig's study13 on intracameral mydriatics showed corneal thickness and EC loss did not differ significantly between the 2 treatment groups after 6 years. Furthermore, the EC morphology showed no statistical differences between the cases where intracameral mydriatics and traditional topical dilatation were used during surgery.
It is inevitable that endothelial cell damage will occur during the phacoemulsification procedure. This study demonstrated that both the ICM and topical groups showed a reduction in ECD over time. However, the ICM group had less endothelial cell loss compared with the topical group. Although this was not statistically significant between the 2 groups, it clearly showed that there was no aggravation of EC loss even with the administration of intracameral phenyelphrine. The mean endothelial cell loss in the ICM group at 3 months was slightly less than that at 1 month. This finding may be due to migration of peripheral endothelial cells into the central cornea to compensate for the loss.7
The percentage of hexagonal cells and CV between the 2 groups in our study were not significantly different. This shows that intracameral phenylephrine does not increase the risk of endothelial cell damage. Matsuda et al25 found that the frequency of hexagonal cells decreases from 68% preoperatively to 52% at 4 weeks after cataract surgery and to 64% at 24 weeks after surgery. This was relatively similar to our study in which the percentage of hexagonal cells reduced from 70% to 66% at 6 weeks and then to 68% at 3 months. This implies that the rate of reduction may reflect the normal EC decrement expected after uneventful cataract surgery and is not due to the interventions.
Various factors are associated with EC loss during cataract surgery.6-12 However, we did not find a significant correlation of endothelial cell loss with phacoemulsification time and duration of surgery, both of which were expected to have positive correlation. In experienced hands, EC loss is not accelerated in cataract surgery of standard duration. Our study showed significantly fewer Indian subjects in the ICM group as compared with the topical group. The degree of melanin pigment in the iris can affect the onset and duration of an intraocular drug. The melanin pigment binds the drug, thus in a darkly pigmented iris the onset of dilatation is slower and maximal pupil dilatation is limited. Slow release from the melanin-bound drug reservoir allows for more prolonged effects of the drug.26 Indians with highly pigmented irides require a longer time for pupillary dilatation when intracameral phenylephrine is injected. A study comparing pupillary dilatation using intracameral mydriatics and topical mydriatics showed that the ICM group had a smaller pupil size than the topical group in early cataract surgery.13 Therefore, 4 Indian patients were excluded in the ICM group, where 2 of the patients needed additional ICM adrenaline, 1 patient developed posterior infusion syndrome, and 1 had Descemet membrane tear during surgery.
There were also significant differences between the 2 groups in lens color and opalescence, anterior chamber depth, and operation time, although the mean value between the 2 groups was only approximately a 2-minute difference in operation time and a 0.1-mm difference in anterior chamber depth. The difference in anterior chamber depth may be caused by different methods to obtain biometry [A-scan and IOL Master 500 (Carl Zeiss Meditec, Inc)]. Both lens color and lens opalescesce showed statistically significant differences despite mean differences of only 0.12 and 0.11, respectively. This is most likely due to an inadequate sample size. However, after adjusting for these confounders (lens color and opalescence, anterior chamber depth, and operation time), our results indicate no significant interaction between intervention type and time period with regards to endothelial cell loss, hexagonality, and coefficient of variation.
In conclusion, intracameral phenylephrine was not associated with increased risk of postoperative endothelial cell loss or morphological changes. It can be safely injected into the anterior chamber for pupil dilatation before phacoemulsification cataract surgery.
The authors would like to thank Dr. Amelia Lim Lay Suan for her guidance and invaluable support in completion of this dissertation.
1. Lovasik J. Pharmacokinetics of topically applied cyclopentolate HCl and tropicamide. Am J Optom Physiol Opt.
2. Haaga M, Kaila T, Salminen L, et al. Systemic and ocular absorption and antagonist activity of topically applied cyclopentolate in man. Pharmacol Toxicol.
3. Cionni RJ, Barros MG, Kaufman AH, et al. Cataract surgery without preoperative eyedrops. J Cataract Refract Surg.
4. Soong T, Soultanidis M, Claoué C, et al. Safety of intracameral mydriasis in phacoemulsification cataract surgery. J Cataract Refract Surg.
5. Nikeghbali A, Falavarjani KG, Kheirkhah A, et al. Pupil dilation with intracameral lidocaine during phacoemulsification. J Cataract Refract Surg.
6. Hayashi K, Hayashi H, Nakao F, et al. Risk factors for corneal endothelial injury during phacoemulsification. J Cataract Refract Surg.
7. Walkow T, Anders N, Klebe S. Endothelial cell loss after phacoemulsification: relation to preoperative and intraoperative parameters. J Cataract Refract Surg.
8. Baradaran-Rafii A, Rahmati-Kamel M, Eslani M, et al. Effect of hydrodynamic parameters on corneal endothelial cell
loss after phacoemulsification. J Cataract Refract Surg.
9. Pirazzoli G, D'Eliseo D, Ziosi M, et al. Effects of phacoemulsification time on the corneal endothelium using phacofracture and phaco chop techniques. J Cataract Refract Surg.
10. Kosrirukvongs P, Slade SG, Berkeley RG. Corneal endothelial changes after divide and conquer versus chip and flip phacoemulsification. J Cataract Refract Surg.
11. Storr-Paulsen A, Nørregaard JC, Farik G, et al. The influence of viscoelastic substances on the corneal endothelial cell
population during cataract surgery: a prospective study of cohesive and dispersive viscoelastics. Acta Ophthalmol Scand.
12. Baradaran-Rafii A, Rahmati-Kamel M, Eslani M, et al. Effect of hydrodynamic parameters on corneal endothelial cell
loss after phacoemulsification. J Cataract Refract Surg.
13. Lundberg B, Behndig A. Intracameral mydriatics
in phacoemulsification cataract surgery. J Cataract Refract Surg.
14. Cantrell MA, Bream-Rouwenhorst HR, Steffensmeier A, et al. Intraoperative floppy iris syndrome associated with alpha1-adrenergic receptor antagonists. Ann Pharmacother.
15. Gurbaxani A, Packard R. Intracameral phenylephrine
to prevent floppy iris syndrome during cataract surgery in patients on tamsulosin. Eye (Lond).
16. Williams GP, Tsaloumas MD. The use of intracameral phenylephrine
in the management of intraoperative floppy-iris syndrome with doxazosin. Eye (Lond).
17. Schultz RO, Matsuda M, Yee RW, et al. Corneal endothelial changes in type I and type II diabetes mellitus. Am J Ophthalmol.
18. Rao GN, Aquavella JV, Goldberg SH, et al. Pseudophakic bullous keratopathy: relationship to pre-operative corneal endothelial status. Ophthalmology.
19. Bonnell AJ, Cymbor M. Under the specular microscope: take a closer look at the clinical and financial benefits that non-contact specular microscopy could offer your practice. Review of Optometry.
20. Hull DS, Chemotti MT, Edelhauser HF, et al. Effect of epinephrine on the corneal edothelium. Am J Ophthalmol.
21. Tinley C, Bates AK. Intracameral 0.5% phenylephrine—a safe solution? Eye (Lond).
22. American Academy of Ophthalmology. Examination techniques for the external eye and cornea. Basic and Clinical Science Course.
23. Mohammad-Salih PA. Corneal endothelial cell
density and morphology in normal Malay eyes. Med J Malaysia.
24. Mori Y, Miyai T, Kagaya F, et al. Intraoperative mydriasis by intracameral injection of mydriatic eye drops: in vivo efficacy and in vitro safety studies. Clin Exp Ophthalmol.
25. Matsuda M, Suda T, Manabe R. Serial alterations in endothelial cell shape and pattern after intraocular surgery. Am J Ophthalmol.
26. De Santis L, Patil, PN. Pharmacokinetics. In: Mauger T, Craig, EL, ed. Havener's Ocular Pharmacology
. St. Louis: Mosby; 1994:30-32.