Journal of Refractive Surgery

Long-Term Endothelial Cell Loss Following Phacoemulsification: Model for Evaluating Endothelial Damage After Intraocular Surgery

Theodore P Werblin, MD, PhD

Abstract

ABSTRACT

BACKGROUND: Newer concepts of phakic intraocular lens (IOL) surgery present concerns as to their long-term deleterious effect on the endothelium. We examine the behavior of the endothelium for up to 6 years following uneventful phacoemulsification surgery, to establish a baseline for what might be an acceptable level of endothelial cell loss due to intraocular surgical trauma.

METHODS: Ninety-three eyes undergoing phacoemulsification surgery who had multiple endothelial specular microscopy examinations for up to 6 years were examined. Central endothelial specular photomicrographs were analyzed in all cases and the percent of endothelial cell loss as a function of time was calculated. Cell density calculations were corrected for the normal endothelial cell loss as a function of age.

RESULTS: Routine uncomplicated phacoemulsification surgery demonstrated a 9% endothelial cell loss 1 year postoperatively. This is in marked contrast to the 16% average endothelial cell loss cited in the literature after phakic AC-IOL surgery. In general, anterior chamber IOL surgery following cataract removal compares unfavorably to posterior chamber IOL insertion long term (20% cell loss versus 12% cell loss).

CONCLUSIONS: A model is proposed to look at the rate of endothelial cell loss that would be significant enough to create the threat for corneal edema long term following refractive intraocular surgery in young patients. Suggestions for critical comparisons between the observed endothelial aging seen with phakic IOLs and values for more routine IOL surgery are proposed to predict the long-term threat for endothelial cell loss and corneal decompensation, which might result many years after phakic refractive IOL surgery. [Refract Corneal Surg 1993;9:29-35.)

RÉSUMÉ

INTRODUCTION: Les effets secondaires endothéliaux suivant l'implantation de lentille infraoculaire chez les yeux phaques méritent consideration. Afin d'établir un niveau passable de perte de !'endothelium suivant la chirurgie intra-oculaire, nous avons suivi à long échéance l'endothelium apès la phacoemulsification noncompliquée.

METHODES: Nous avons examiné 93 patients qui ont subi la phacoemulsification et qui ont subi de nombreuses examens à la microscopic speculate jusqu'à 6 ans post-opératoire. Les photos de la microscopie speculaire de !'endothelium centrale furent analysés dans tous les cas et le pourcentage de perte des cellules endotheliales avec le temps fut calculé. Les calculations furent corrigées avec la perte normale des cellules en fonction du viellissement.

RÉSULTATS: Nous avons observé la perte de 9% des cellules endotheliales 1 année après la phacoemulsification noncompliquée, par contraste au chiffre publiée de 16% après l'implantation de LIO chez les yeux phaques. On a observé une perte de 20% après l'implantation de LIO dans la chambre antérieure suivant l'extraction de cataracte, par contraste à 12% avec une lentille postérieure.

Abstract

ABSTRACT

BACKGROUND: Newer concepts of phakic intraocular lens (IOL) surgery present concerns as to their long-term deleterious effect on the endothelium. We examine the behavior of the endothelium for up to 6 years following uneventful phacoemulsification surgery, to establish a baseline for what might be an acceptable level of endothelial cell loss due to intraocular surgical trauma.

METHODS: Ninety-three eyes undergoing phacoemulsification surgery who had multiple endothelial specular microscopy examinations for up to 6 years were examined. Central endothelial specular photomicrographs were analyzed in all cases and the percent of endothelial cell loss as a function of time was calculated. Cell density calculations were corrected for the normal endothelial cell loss as a function of age.

RESULTS: Routine uncomplicated phacoemulsification surgery demonstrated a 9% endothelial cell loss 1 year postoperatively. This is in marked contrast to the 16% average endothelial cell loss cited in the literature after phakic AC-IOL surgery. In general, anterior chamber IOL surgery following cataract removal compares unfavorably to posterior chamber IOL insertion long term (20% cell loss versus 12% cell loss).

CONCLUSIONS: A model is proposed to look at the rate of endothelial cell loss that would be significant enough to create the threat for corneal edema long term following refractive intraocular surgery in young patients. Suggestions for critical comparisons between the observed endothelial aging seen with phakic IOLs and values for more routine IOL surgery are proposed to predict the long-term threat for endothelial cell loss and corneal decompensation, which might result many years after phakic refractive IOL surgery. [Refract Corneal Surg 1993;9:29-35.)

RÉSUMÉ

INTRODUCTION: Les effets secondaires endothéliaux suivant l'implantation de lentille infraoculaire chez les yeux phaques méritent consideration. Afin d'établir un niveau passable de perte de !'endothelium suivant la chirurgie intra-oculaire, nous avons suivi à long échéance l'endothelium apès la phacoemulsification noncompliquée.

METHODES: Nous avons examiné 93 patients qui ont subi la phacoemulsification et qui ont subi de nombreuses examens à la microscopic speculate jusqu'à 6 ans post-opératoire. Les photos de la microscopie speculaire de !'endothelium centrale furent analysés dans tous les cas et le pourcentage de perte des cellules endotheliales avec le temps fut calculé. Les calculations furent corrigées avec la perte normale des cellules en fonction du viellissement.

RÉSULTATS: Nous avons observé la perte de 9% des cellules endotheliales 1 année après la phacoemulsification noncompliquée, par contraste au chiffre publiée de 16% après l'implantation de LIO chez les yeux phaques. On a observé une perte de 20% après l'implantation de LIO dans la chambre antérieure suivant l'extraction de cataracte, par contraste à 12% avec une lentille postérieure.

Several authors have reported the effects of cataract surgery on the corneal endothelium over extended time periods.1'5 These reports differ in the type of intraocular lens used, the type of cataract surgery performed, and the frequency or duration for which postoperative evaluations were performed. This study will examine the behavior of the endothelium following uncomplicated phacoemulsification surgery in patients who were examined over a 6-year period. All surgeries were performed by one surgeon using the same operative technique.

Much of the damage to the corneal endothelium during cataract surgery occurs in the area of the wound. However, the resultant endothelial loss may not be reflected by the assessment of central endothelial counts for several years following surgery. The mechanism of repair is cell enlargement and movement to cover denuded or killed areas of peripheral endothelium.2 Therefore, reports that look at the effects of surgical procedures on the endothelium, where only the central endothelium is examined, may not reflect actual cell loss until the endothelium has equilibrated several years after surgery; and studies that purport to reflect true endothelial loss must examine patients for an extended period of time until a plateau or an equilibration level is evident.

Newer techniques of intraocular surgery for refractive purposes, specifically the concept of phakic anterior chamber intraocular lenses (IOLs), have raised concerns about the long-term effects of these devices on the corneal endothelium.6-11 Because long-term studies will be required to estimate whether endothelial damage will eventually result in corneal edema secondary to this intraocular procedure, we will attempt to devise a model, which looks at "acceptable" base line levels of endothelial cell damage following intraocular surgery, ie, seen after routine phacoemulsification surgery and use this model as a means for comparison to other intraocular surgical procedures. Establishment of rates of endothelial dropout soon after innovative intraocular procedures might allow the early detection of unacceptable trends in endothelial cell loss.

Additionally, we will present data obtained from endothelial studies in nonhuman primate eyes following phacoemulsification surgery. This model represents yet another possible mechanism of determining endothelial cell loss related to new intraocular procedures or lens designs without having to risk corneal edema in human subjects.

MATERIALS AND METHODS

Ninety-three consecutive eyes having undergone , an uneventful phacoemulsification procedure were followed with sequential specular microscopy at 4 yearly intervals for up to 6 years.

Fifty-three fellow eyes which did not have cataract surgery were also photographed and counted sequentially, constituting a control population for endothelial cell loss. Although all patients were encouraged to return for yearly studies, this was not always possible. These data include all the eyes who had preoperative and two or more postoperative photographs. The eyes were 69.3 ± 8.0 years of age. Patients who did not have readable preoperative endothelial cell counts and any eyes that, during the surgical procedure, had a complication such as vitreous loss or torn posterior capsule requiring an anterior chamber lens were excluded from the study (about 1.5% of cases).

Figure: The percent of loss of central endothelial cell density as a function of years following phacoemulsification surgery. Uncorrected figures are raw data where total percent loss has been calculated. Corrected figures indicate those numbers which were corrected for the normal one half percent per year drop out of endothelial cells seen with age (cells/yr = cells/mmp 2/yr).Table 1Endothelial Cell Loss (N) Phacoemulsification Surgery Human Study

Figure: The percent of loss of central endothelial cell density as a function of years following phacoemulsification surgery. Uncorrected figures are raw data where total percent loss has been calculated. Corrected figures indicate those numbers which were corrected for the normal one half percent per year drop out of endothelial cells seen with age (cells/yr = cells/mmp 2/yr).

Table 1

Endothelial Cell Loss (N) Phacoemulsification Surgery Human Study

Table

Table 2Endothelial Cell Loss (N = 8) Patients With Consecutive 1-,2-, and 3-Year Counts Following Phacoemulsification Surgery†

Table 2

Endothelial Cell Loss (N = 8) Patients With Consecutive 1-,2-, and 3-Year Counts Following Phacoemulsification Surgery†

The surgical procedure consisted of a routine 6.5-millimeter phacoemulsification scleral wound 1.5 mm posterior to the limbus, with posterior chamber intraocular lens insertion utilizing Healon during the lens insertion procedure and Balanced Salt Solution-Plus (BSS-Plus) irrigating solution during phacoemulsification. Kratz style, lenses with 10° posterior angulated haptics were used throughout. As much of the remaining Healon was irrigated out of the eye prior to and after 10-0 nylon suturing. Topical 5% Isopto-Homatropine and 0.1% dexamethazone were used for 4 to 8 weeks postoperatively. Pre- and postoperative central specular microscopic examinations were done using the Keeler Konan specular microscope. Even though regional endothelial studies would be of interest, it was felt that that technique would have been difficult and unreliable in this elderly patient population.

In our clinic population, only three eyes had uneventful secondary anterior chamber lenses and long-term postoperative endothelial follow-up data. The lens insertion was performed through a 6.5-millimeter temporal incision under Healon and was closed with interrupted 10-0 nylon sutures. The Healon was irrigated out of the eye with BSS (not BSS-Plus).

The counting procedure was standardized so that one individual counted all photographs within a very short time period at the conclusion of the study. A minimum of 100 cells were counted. The specular photography was performed on the central endothelium in each case. Morphometric analysis was not performed even though it may have some hypothetical advantages. Many of the literature studies cited in this article did not perform such analysis and, therefore, this analysis would not have helped our comparative conclusions.

The surgical data was analyzed calculating the observed percentage of endothelial cell loss as a function of time. For some, calculating the "normal" endothelial cell loss seen as a function of age,12 0.5%/year, was added to the postoperative data so as to differentiate only that loss of endothelial cells related to the surgical procedure. Although the control population measured in this article demonstrated a slightly higher rate of spontaneous cell loss (0.8%/year) compared to that seen in the literature (0.5%/year),12 the lower figures were used for analysis in the article since it was felt that the 15% loss was derived from a much larger control population.

Several nonhuman primate studies (cynomolgus and rhesus monkeys) have been conducted for the evaluation of new forms of IOL or IOL materials. Control data from these experiments will be presented. In these studies, polymethylmethacrylate (PMMA) lenses similar to those used in humans were inserted into the capsular bag, under Healon after phacoemulsification surgery. These lenses differed from the human IOL in that the haptics were 12.5 mm instead of the 14 mm used in human eyes. Preoperative and sequential postoperative central endothelial counts were obtained in the nonhuman primate model in a fashion similar to the analysis performed in the human. Only uncomplicated procedures were evaluated in this study, as was done with the human clinical study.

RESULTS

The Figure and Table 1 reflect the data analysis performed on 93 human eyes. One half percent/year cell losses were added in a cumulative fashion to each of the six postoperative time periods which is demonstrated in the "corrected" data analysis. This allows the evaluation to reflect only the cell loss due to surgery and isolates the surgical cell loss from the normal decay of the human endothelium as a function of age. As can be seen from the Table 1 after 3 years postoperatively the loss of cells appeared to stabilize at roughly 11.5%. There were statistically significant differences between all preoperative, 1and 2-year postoperative values, compared to all later postoperative time periods. There were no significant differences between the 3-year data and the later time periods. In a smaller group of eight patients (Table 2) where preoperative, 1, 2, and 3 years preoperative data were present in all eyes, the sequential increase in cell loss after surgery is again demonstrated. The increase in cell loss just after surgery and the 3-year value are similar to that seen in the 93 patient population, confirming the plateauing of cell loss during the first 3 postoperative years.

The nonhuman primate model reflected a similar pattern of cell loss, however the studies were only conducted for a 1-year time period (Table 3). It was also apparent that a significantly larger amount of cell loss was seen in the nonhuman primate as compared to the human model, reflecting a greater sensitivity of the endothelium to surgical manipulation, even though the surgical procedure itself is much easier in the primate since the lenses were not cataractous. The nonhuman primate eye is much smaller and more difficult to work with possibly accounting for the greater "sensitivity."

Table

Table 3Endothelial Cell Loss No Correction for "Normal" Cell Loss Species/Lens/Technique ComparisonTable 4Endothelial Cell Loss (N) Nonhuman Primate Model - Phacoemulsification Surgery

Table 3

Endothelial Cell Loss No Correction for "Normal" Cell Loss Species/Lens/Technique Comparison

Table 4

Endothelial Cell Loss (N) Nonhuman Primate Model - Phacoemulsification Surgery

In the human study, the earliest postoperative time examined for endothelial cell loss was 1 year. Therefore, the time course of cell loss during the first year after surgery cannot be ascertained from these data. In the nonhuman primate model, however, sequential examinations were made during the first year and a reasonable estimate of cell loss as a function of time can be ascertained from these data (Table 4). Roughly 50% of the total cell loss was seen in the first 3 months. Thereafter, a slow progressive increase in cell loss occurs which takes a few years before a plateau is reached.

DISCUSSION

These data have indicated that after uneventful phacoemulsification surgery, roughly an 11.5% endothelial cell loss is seen. It takes approximately 3 years for this cell loss to be reflected in the central endothelial cell count. Undoubtedly, the majority of cell loss occurs near the entrance wound and because of the nature of endothelial repair, the central cell density is not fully equilibrated until the entire mosaic of endothelium has expanded to cover the injured or killed cells roughly 3 years after surgery. It is apparent from the data, however, that after the 3-year time period, there is no additional cell loss due to the surgical trauma. Fellow control eyes reflect a 0.8% annual cell loss during the same time period. If the normal cell loss due to age was not corrected for in the analysis, then it might appear erroneously that there is an ongoing increase in cell loss due to surgery.

Table

Table 5Model For Endothelial Cell Loss Postulated Phakic IOL Surgery at Age 25 Years

Table 5

Model For Endothelial Cell Loss Postulated Phakic IOL Surgery at Age 25 Years

Several recent studies, both in the nonhuman primate and human primate models with phakic IOLs indicate that in certain cases significantly greater and perhaps progressive endothelial cell damage is observed (Table 3).610 This was true for both anterior chamber and iris claw IOLs. A major concern with this refractive surgical technique is the potential for long-term damage to the endothelium caused by the anterior chamber lens, either through direct surgical trauma or through an ongoing inflammatory response creating chronic progressive cell loss. Because it takes several years for the central endothelium to reflect the acute surgical damage, then the long-term evaluation of human endothelial cell loss following phakic IOL surgery could be extremely difficult. Another approach to this problem would be to examine the relative effect of a phakic IOL on the endothelium compared to the "acceptable" amount of cell loss seen with routine phacoemulsification surgery. Specifically, one could statistically compare the decay in endothelial density after phakic IOL or other intraocular surgery to that of routine phacoemulsification surgery. Statistical analysis over a relatively short period of time should enable an earlier determination of a trend toward significant or unacceptable damage. For example, at 1 year, experimental studies of the phakic IOL in nonhuman primates showed a 31% endothelial cell loss6 and human studies showed a 20% or greater loss8"10 in many patients (Table 3). This does not compare favorably to the 9% cell loss seen in the phacoemulsification patient population reflecting the potential of ongoing and exaggerated cell damage due to the phakic IOL.

It has been postulated that newer lens designs which create a larger distance between the optic and the endothelium for phakic IOLs may improve performance vis a vis the endothelium. However, even where a much larger separation of anterior chamber lens and endothelium exist, ie, secondary IOL, poor endothelial tolerances have been noted (Table 3). This argues that the anterior chamber lens creates some chronic additional damage to the endothelium in an ongoing fashion which jeopardizes long-term stability.5

We also looked at a nonhuman primate model for detecting damage to the endothelium (Table 4). Although it appears that the nonhuman primate is more sensitive to intraocular surgery, nevertheless it is possible to evaluate the effects of new anterior chamber lenses in the nonhuman primate prior to implantation in the human. These comparative studies could enable modifications of lens design or lens materials and/or coatings to prevent excessive endothelial cell damage and thus allow perfection a more appropriate phakic IOL for use in human clinical studies.

Previous authors have criticized the analysis of specular microscopy using only central endothelial cell counts.13 The possibility of major sampling errors occurring if the endothelium were not uniform is a valid concern. In this study, we basically analyzed eyes with no preexisting major endothelial pathology and, in general, feel that most of the eyes studied had a fairly uniform mosaic of endothelium prior to and after the surgical procedure.

The following model might be reasonable to set guidelines for the evaluation of endothelial performances with newer forms of IOL surgery. This concept has been previously discussed by Mishima.14 If we assume that at maturity (age 20 years), we have an endothelial cell density of roughly 3400 cells per square millimeter, and the normal 0.5% per year loss of endothelial cells occurs,12 then the remaining cell population at age 25 years would show a density of 3330 cells/mm2 (Table 5). If a patient at that time were to have a form of refractive surgery, then the following consequences, depending on the amount of endothelial cell damage, would be observed. Radial keratotomy surgery, although controversial as to any endothelial damage, may demonstrate a 5% cell loss, which would drop the cell count to 2184 cell/ mm2 by age 95 years assuming the 5% loss due to radial keratotomy plus the 0.5% per year loss with aging. If a phacoemulsification surgery were performed at age 25 years, then the age 95 endothelial cell density would be 1950 cells/mm2. ?? instead of the normal 0.5% per year drop out of cells observed, an increase to 1% per year occurred, ie, in a lens design which caused chronic inflammation/ endothelial drop out, then by age 95 years the patient would be left with 970 cells/mm2 which would probably sustain corneal clarity. On the other hand, if the rate of endothelial aging (cell drop out) were to increase to 1.5% per year, then by age 95 the patient would demonstrate end up with roughly 90 cells/mm2, which would be less than that necessary to maintain corneal clarity.

In conclusion, recent studies with phakic AC-IOLs in both humans and nonhuman primate models have shown significantly greater endothelial cell loss than one would expect in routine human intraocular surgery. However, as can be seen in the cell loss reflected in uneventful phacoemulsification surgery in the nonhuman primate, there appears to be a greater sensitivity of the endothelium to intraocular surgery and thus comparison of the nonhuman primate data directly to the human data is inappropriate. Nevertheless, comparing the nonhuman primate phakic IOL data to nonhuman primate phacoemulsification data or human phakic AC-IOL data to human phacoemulsification data does reflect a major increase in cell loss in the phakic AC-IOL populations. Thus, there is significant concern for the long-term effect on the endothelium of phakic AC-IOL surgery using the lenses and procedures described to date. Before large clinical studies of phakic IOLs are undertaken, careful comparisons of phakic IOL data to existing endothelial cell loss data would be advisable.

REFERENCES

1. Rao GN, Stevens RE, Harris JK, Aquavella JV. Long term changes in corneal endothelium following intraocular lens implantation. Ophthalmology. 1981;88:386-397.

2. Matsuda M, Tsuneji S, Manabe R, Serial alterations in endothelial cell shape and pattern after intraocular surgery. Am J Ophthalmol. 1984;98:313-319.

3. Matsuda M, Miyake K, Inaba M. Long-term corneal endothelial changes after intraocular lens implantation. Am J Ophthalmol. 1988;105:248-252.

4. Oxford Cataract Treatment and Evaluation Team (OCTET). Long-term corneal endothelial cell loss after cataract surgery. Results of a randomized controlled trial. Arch Ophthalmol, 1986;104:1170-1175.

5. Ambrose VM, Walters RP, Batterbury M, Spalton DJ, McGiIl JI. Long-term endothelial cell loss and breakdown of the blood-aqueous barrier in cataract surgery. J Cataract Refract Surg. 1991;17:622-627.

6. Peiffer RL, Porter DP, Eifrig DE, Boyd J. Experimental evaluation of a phakic anterior chamber implant in a primate model. Part 1 . Clinical observations. J Cataract Refract Surg. 1991;17:335-341.

7. Porter DP, Peiffer RL, Eifrig DE, Boyd J. Experimental evaluation of a phakic anterior chamber implant in a primate model. Part II. Pathology. J Cataract Refract Surg. 1991;17:342-352.

8. Fechner PU, Strobel J, Wichmann W. Correction of myopia by implantation of a concave worst-iris claw lens into phakic eyes. Refract Corneal Surg. 1991;7:286-298.

9. Saragoussi JJ, Cotinat J, Renard F, et al. Damage to the corneal endothelium by minus power anterior chamber intraocular lenses. Refract Corneal Surg. 1991;7:282-285.

10. Mimouni F, Colin J, Koffi V, Bonnet P. Damage to the corneal endothelium from anterior chamber intraocular lenses in phakic myopic eyes. Refract Corneal Surg. 1991;7:277-281.

11. Waring GO. Phakic intraocular lenses for the correction of myopia- - where do we go from here? Refract Corneal Surg. 1991;7:275-276.

12. Yee RW, Matsuda M, Schultz RO, Edelhauser HF. Changes in the normal corneal endothelial cellular pattern as a function of age. CurrEyeRes. 1985;4:671-678.

13. Hirst LW, Yamauyoshi K, Enger C, Vogelpohl W, Whittington V. Quantitative analysis of wide-field specular microscopy. II. Precision of sampling from the central corneal endothelium. Invest Ophthalmol Vis Sci 1989;9:1972-1979.

14. Mishima S. Clinical investigations on the corneal endothelium. XXXVIII Edward Jackson Memorial Lecture. Am J Ophthalmol. 1982;93:1-29.

Table 2

Endothelial Cell Loss (N = 8) Patients With Consecutive 1-,2-, and 3-Year Counts Following Phacoemulsification Surgery†

Table 3

Endothelial Cell Loss No Correction for "Normal" Cell Loss Species/Lens/Technique Comparison

Table 4

Endothelial Cell Loss (N) Nonhuman Primate Model - Phacoemulsification Surgery

Table 5

Model For Endothelial Cell Loss Postulated Phakic IOL Surgery at Age 25 Years

10.3928/1081-597X-19930101-08

Sign up to receive

Journal E-contents