- Email: [email protected]
A comparison of the effects of phacoemulsification and nucleus expression on endothelial cell density
articles A comparison of the effects of phacoemulsification and nucleus expression on endothelial cell density John M. Graether, M.D. Gilbert W. Harris, M.D. James A. Davison, M.D. Russell H. Watt, M.D. Russell R. Widner, M.D. Marshalltown, Iowa Vincent Sposito, Ph. D. Ames, Iowa ABSTRACT A series of patients undergoing extracapsular cataract extraction by phacoemulsification (1,126) or nucleus expression (632) were studied to evaluate the relative effects of emulsification time, irrigating volume, and surgical technique on endothelial cell density. The collective data revealed a higher cell loss with phacoemulsification (15%) than with nucleus expression (12%). Patients over 69.5 years appeared to be more vulnerable to the effects of phacoemulsification than those below 69.5 years. Although the data were subjected to combinations and permutations of age, sex, time, and volume, no direct correlation between emulsification time and/or irrigating volume and enothelial cell loss was shown. When data sets for individual surgeons were analyzed, the most striking difference was between the endothelial effect of anterior chamber phacoemulsification (26% average reduction) and that of posterior chamber phacoemulsification (13% average reduction). Key Words: anterior chamber phacoemulsification, emulsification, endothelial cell loss, irrigating volume, nucleus expression, posterior chamber phacoemulsification
The effects of various types of cataract extraction and intraocular lens implantation on endothelial cell density have been studied by numerous investigators. I -7 In general, phacoemulsification results in greater cell loss than does extracapsular cataract extraction (ECCE) or intracapsular cataract extraction (ICCE).1,2,4 However, no attempt to relate the degree of cellular loss to the emulsification time or to the volume of fluid used
during the procedure has been made. This study will investigate those relationships in a series of patients. MATERIALS AND METHODS Since 1978, patients undergoing cataract extraction or lens implantation at the Wolfe Clinic have routinely had preoperative and postoperative endothelial cell counts. The Heyer-Schulte endothelial camera with
This study was supported by a research grant from the Wolfe Foundation, Marshalltown, Iowa. Statistical analysis of the data was performed under the direction of Dr. Vincent Sposito of Iowa State University, Ames, Iowa. Reprint requests to John M. Graether, M.D., Wolfe Clinic, P.C., 309 East Church Street, Marshalltown, Iowa 50158.
420
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
35-mm black and white film was used for endothelial cell counts. At least five pictures were taken at each observation and the best three were selected for counting. The usual interval for postoperative endothelial cell counts was three months for all surgeons, but counts were accepted when done between two and eight months. The patients studied had ECCE by phacoemulsification or by nucleus expression and implantation of a posterior chamber lens. The posterior chamber lens style was uniplanar and the risk oflens-corneal contact during insertion was minimal. Therefore, insertional trauma would not have contributed significantly to endothelial cell loss. We waited until a large series of patients had been examined before analyzing the data statistically to obtain the best correlation. We studied consecutive patients who had the necessary data recorded, i. e., preoperative and postoperative cell counts, emulsification time, and irrigating volume of balanced salt solution. Patients were omitted only if data was not complete or they were lost to follow-up. No attempt to eliminate patients with corneal disease, such as cornea guttata, was made as long as preoperative cell counts could be obtained. In the few cases where difficulties with postoperative corneal clarity prevented postoperative cell counts, the count was recorded as zero. No sodium hyaluronate (Healon ®) was used during this study. Patients from all six clinic surgeons were included in the study and the raw data were submitted to an independent statistician who supervised the data processing and statistical analyses. 8,9 Univariate analyses of both data sets, identifying extreme values, averages, and standard deviations were performed.
RESULTS Our data included 1,126 phacoemulsification patients and 632 nucleus expression patients. These represented 53% of patients operated on in that period. We first analyzed the data as a whole, and then related the data to specific surgeons to isolate the effects of surgical technique. The average percentage reduction of endothelial cells after surgery was significantly greater than zero at the 0.0005 significance level for both groups. The reduction and percentage loss of endothelial cells after phacoemulsification was significantly greater than the reduction and percentage loss after nucleus expression at the 0.01 significance level; 15% for phacoemulsification and 12% for nucleus expression. The average volume of irrigating solution was larger in the phacoemulsification group than in the nucleus expression group at the 0.0005 significance level. There was no significant correlation between the percentage of endothelial cells lost and the emulsification time, or the irrigating volume, or a combination of the two (Table 1). The mean age in the phacoemulsification group was 68 years; in the nucleus expression group it was 73.2 years (Table 2). When the data were analyzed for sex and age differences, the following observations were made (Tables 3, 4, 5, and 6): The average patient in the nucleus expression group was older than the average patient in the phacoemulsification group at the 0.0005 Significance level. Preoperative counts of patients over 69.5 years were, in general, lower than those of patients under 69.5 years at the 0.0005 significance level.
Table 1. Analysis of 1,126 phacoemulsification and 632 nucleus expression patients.
Table 3. Preoperative cell counts in phacoemulsification and nucleus expression patients.
Phacoemulsification Nucleus Expression Time (min)
2.12 ± 0.84
Volume (ml)
185.9 ± 63.3
140.6 ± 63.0
Mean preoperative count (cell/mm 2)
2701 ± 485.4
2709.4 ± 473.1
Mean postoperative count (cell/mm2)
2287 ± 603.8
2377 ± 559.4
423 ± 488
332.2 ± 454.9
15.1 ± 19.6
11.9 ± 17.9
Cell reduction Percent reduction
Age (Years)
< 69.5 > 69.5
Type of Surgery
Age (Years)
Phacoemulsification
68.0 ± 10.2
Nucleus expression
73.2 ± 8.7
Mean
S.D.*
Nucleus Expression Number
Mean
S.D.*
594
2763.93 445.59
213
2793.85 406.61
532
2651.74 520.33
419
2666.51
498.57
*S.D. = standard deviation
Table 4. Percentage of cell reduction in phacoemulsification and nucleus expression patients. Age (Years)
Table 2. Age of patients studied.
Phacoemulsification Number
< 69.5 > 69.5
Phacoemulsification Number
Mean
594
13.81
532
16.61
S.D.*
Nucleus Expression Number
Mean
S.D.*
19.78
213
10.97
13.65
19.36
419
12.31
19.66
*S. D. = standard deviation
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
421
Table 5. Preoperative cell count by sex in phacoemulsification and nucleus expression patients. Phacoemulsification Number
Mean
Nucleus Expression
Variance* Number
Mean
Variance
Table 8. Univariate analysis of nucleus expression""~ons.
Surgeon Number
3
46
4
379
141.9 ± 55.9 326.3 ± 412.7
11.1 ± 18.9
Table 6. Percent cell reduction by sex in phacoemulsification and nucleus expression patients.
5
24
167.5 ± 103.4 273.3 ± 476.8
10.2 ± 17.3
6
19
161.1 ± 54.3 413.9 ± 308.4
16.5 ± 12.0
497
2766.80 183048.90
367 264
Phacoemulsification Number
2678.03 257520.35 2750.73 174128.50
78
10.6± 14.1 16.0 ± 16.6 12.1 ± 18.9
Nucleus Expression
Mean Variance * Number
Mean
Variance
Female 628
15.37
491.91
367
12.14
375.07
Male
14.77
249.42
264
11.48
243.33
497
2
Percent Reduction
*Variance = standard deviation squared
Male
2668.12 272342.80
86
Cell Count Reduction (cells/mm 2 )
149.2 ± 82.5 297.2 ± 409.1 127.3 ± 49.3 423.3 ± 399.6 113.3 ± 63.1 288.7 ± 476.7
Female 628
1
Irrigating Volume (cell/mm 2)
Table 9. Comparison of anterior chamber phacoemulsification (surgeon 5) with posterior chamber phacoemulsification (other surgeons). Surgeon 5
*Variance = standard deviation squared
Patient age (yrs)
In the nucleus expression group, there was no significant cell loss difference between the patients over 69.5 years and those under 69.5 years. However, in the phacoemulsification group, patients over 69.5 years exhibited a larger cell loss at the 0.05 significance level. Preoperative counts of female patients were, in general, lower than those of male patients at the 0.05 significance level. No significant difference between the percentage reduction in cells in female patients and that in male patients could be determined. . Tables 7, 8, and 9 present data of the individual surgeons and yield the following observations: The average percentage postoperative reduction of endothelial cells for each surgeon was significantly greater than zero for both groups at the 0.0005 significance level. The average percentage decrease between phacoemulsification and nucleus expression was similar for all surgeons except surgeon 5. The average percentage decrease for surgeon 5 was larger than the other surgeons at the 0.05 significance level. The average percentage reduction of endothelial
Irrigating volume (ml)
68.70 ±
8.10
67.90 ± 10.50
203.90 ± 65.60
183.20 ± 62.50
Emulsification time (min) Percent cell reduction
Other Surgeons
2.46 ±
0.88
26.40 ± 23.40
2.07 ±
0.82
13.40 ± 18.40
cells was not similar among surgeons in the phacoemulsification group; moreover, for surgeon 5, the reduction was significantly larger (at O. 001) than it was for other surgeons. The average percentage reduction of endothelial cells for surgeons in the nucleus expression group was approximately the same. The average phacoemulsification time among surgeons in the phacoemulsification group was variable, with surgeon 5 showing a significantly greater time than other members of the group at the 0.01 significance level. There was no significant difference among the surgeons in the average age of the patients. Table 9 demonstrates the comparison of the data from surgeon 5 who employed anterior chamber phacoemulsification and the pooled data of the other surgeons who used posterior chamber phacoemulsification.
Table 7. Univariate analysis of phacoemulsification by surgeon. Surgeon
422
Number
Emulsification Time (min)
Irrigating Volume (ml)
Cell Count Reduction (cell/mm 2)
Percent Reduction
1
143
2.04 ± 0.74
194.8 ± 64.1
286.8 ± 367.1
2
156
2.08 ± 0.93
210.2 ± 53.9
349.7 ± 426.5
13.0 ± 17.0
3
462
1.95 ± 0.74
153.9 ± 50.8
367.5 ± 482.5
12.8 ± 20.5
4
77
1.73 ± 0.61
212.5 ± 63.6
483.9 ± 521. 7
17.6 ± 19.6
5
151
2.46 ± 0.88
203.9 ± 65.6
730.0 ± 553.9
26.4 ± 23.4
6
136
1.85 ± 0.65
222.5 ± 58.4
461.1 ± 447.7
16.7 ± 15.4
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
9.9 ± 13.0
There was a higher percentage of cell loss in the former group of patients.
DISCUSSION The most surprising observation of this study is the apparent lack of correlation between the irrigating volume and/or the emulsification time and the reduction in endothelial cells. Although these data were studied and subjected to age, time, and volume combinations and permutations for data sets of individual surgeons (Tables 7 and 8), as well as the data as a whole, no relationship could be established. Even when quadratic terms were introduced, no acceptable correlations were found. As noted, there was a correlation between the loss of endothelial cells during emulsification and age, the older endothelium being more vulnerable. This confirms the observation of Waltman and Cozean. 1 There was also a definite increase in average cell loss in the phacoemulsification group when compared with the nucleus expression group, as observed by several authors.l- 7 Although the average cell loss with each procedure was modest, the phacoemulsification procedure produced a much wider range of cell loss, with alarming decrease in cell density in some patients. This observation suggests careful patient selection, especially in the more vulnerable, older patient. The procedure should be avoided where pupil size, shallow chamber, posterior pressure, or other factors will result in the procedure being performed close to the endothelium. One of the six surgeons in this study (surgeon 5) had continued to emulsifY the nucleus as originally taught by Charles Kelman, i. e., by displacing the nucleus into the anterior chamber prior to emulsification. The other surgeons employed techniques whereby most of the nucleus was emulsified prior to dislocation so most of the emulsification took place behind the iris plane. Surgeon 5 averaged a cell loss of26%; combined average cell loss for the other surgeons was 13%. Surgeon 5 also had higher average emulsification times and irrigating volumes. However, attempts to obtain correlations between cell loss and volume and/or time were unsuccessful in his isolated data set as in the data as a whole. Therefore, it appears that location of the nucleus vis a vis the endothelium during emulsification is the cause of the increased endothelial damage, an observation made previously by KrafT 4 and Hoffer. 7
If the first eye experienced significant cell loss with phacoemulsification, a different approach should be selected for the fellow eye. For the occasional surgeon, nucleus expression is likely to be a safer choice. SUMMARY A group of patients undergoing cataract extraction with posterior chamber lens implantation was studied for the effects of various techniques on preoperative and postoperative cell density. The nucleus was removed by phacoemulsification in 1,126 patients and by simple expression of the nucleus in 632 patients. The endothelial cell loss in the phacoemulsification group (15%) was moderately higher than in the nucleus expression group (12%), but both were in the expected range for this operation. 5 No correlation between the irrigating volume or the phacoemulsification time and the endothelial cell loss could be found in either group. There was a correlation between the patient's age and the endothelial cell loss during phacoemulsification, with older corneas (over 69.5 years of age) showing a greater percentage loss than corneas under 69.5 years. There was also a correlation between the location of the nucleus and the endothelium during emulsification; significantly greater loss occurred when the nucleus was emulsified in the anterior chamber. REFERENCES 1. Waltman SR, Cozean CH Jr: The effect of phacoemulsification on the corneal endothelium. Ophthalmic Surg 10(1):31-33, 1979 2. Abbott RL, Forster RK: Clinical specular microscopy and intraocular surgery. Arch Ophthalmol 97:1476-1479, 1979 3. Stanley JA, Shearing SP, Anderson RR, Avallone AN: Endothelial cell density after posterior chamber lens implantation. Ophthalmology 87:381-384, 1980 4. KraffMC, Sanders DR, Lieberman HL: Specular microscopy in cataract and intraocular lens patients. A report of 564 cases. Arch Ophthalmol 98:1782-1784, 1980 5. Bourne WM, Waller RR, Liesegang TJ, Brubaker RF: Corneal trauma in intracapsular and extracapsular cataract extraction with lens implantation. Arch Ophthalmol 99:1375-1376, 1981 6. KraffMC, Sanders DR, Lieberman HL: Monitoring for continuing endothelial cell loss with cataract extraction and intraocular lens implantation. Ophthalmology 89:30-34, 1982 7. Hoffer KJ: Effects of extracapsular implant techniques on endothelial density. Arch Ophthalmol 100:791-792, 1982 8. Statistical Analysis System (SAS), SAS Institute Inc., Cary, NC, 1979 9. Snedecor G, Cochran WG: Statistical Methods, 6th ed., Ames, Iowa, Iowa State University Press, 1967
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
423
The effects of various types of cataract extraction and intraocular lens implantation on endothelial cell density have been studied by numerous investigators. I -7 In general, phacoemulsification results in greater cell loss than does extracapsular cataract extraction (ECCE) or intracapsular cataract extraction (ICCE).1,2,4 However, no attempt to relate the degree of cellular loss to the emulsification time or to the volume of fluid used
during the procedure has been made. This study will investigate those relationships in a series of patients. MATERIALS AND METHODS Since 1978, patients undergoing cataract extraction or lens implantation at the Wolfe Clinic have routinely had preoperative and postoperative endothelial cell counts. The Heyer-Schulte endothelial camera with
This study was supported by a research grant from the Wolfe Foundation, Marshalltown, Iowa. Statistical analysis of the data was performed under the direction of Dr. Vincent Sposito of Iowa State University, Ames, Iowa. Reprint requests to John M. Graether, M.D., Wolfe Clinic, P.C., 309 East Church Street, Marshalltown, Iowa 50158.
420
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
35-mm black and white film was used for endothelial cell counts. At least five pictures were taken at each observation and the best three were selected for counting. The usual interval for postoperative endothelial cell counts was three months for all surgeons, but counts were accepted when done between two and eight months. The patients studied had ECCE by phacoemulsification or by nucleus expression and implantation of a posterior chamber lens. The posterior chamber lens style was uniplanar and the risk oflens-corneal contact during insertion was minimal. Therefore, insertional trauma would not have contributed significantly to endothelial cell loss. We waited until a large series of patients had been examined before analyzing the data statistically to obtain the best correlation. We studied consecutive patients who had the necessary data recorded, i. e., preoperative and postoperative cell counts, emulsification time, and irrigating volume of balanced salt solution. Patients were omitted only if data was not complete or they were lost to follow-up. No attempt to eliminate patients with corneal disease, such as cornea guttata, was made as long as preoperative cell counts could be obtained. In the few cases where difficulties with postoperative corneal clarity prevented postoperative cell counts, the count was recorded as zero. No sodium hyaluronate (Healon ®) was used during this study. Patients from all six clinic surgeons were included in the study and the raw data were submitted to an independent statistician who supervised the data processing and statistical analyses. 8,9 Univariate analyses of both data sets, identifying extreme values, averages, and standard deviations were performed.
RESULTS Our data included 1,126 phacoemulsification patients and 632 nucleus expression patients. These represented 53% of patients operated on in that period. We first analyzed the data as a whole, and then related the data to specific surgeons to isolate the effects of surgical technique. The average percentage reduction of endothelial cells after surgery was significantly greater than zero at the 0.0005 significance level for both groups. The reduction and percentage loss of endothelial cells after phacoemulsification was significantly greater than the reduction and percentage loss after nucleus expression at the 0.01 significance level; 15% for phacoemulsification and 12% for nucleus expression. The average volume of irrigating solution was larger in the phacoemulsification group than in the nucleus expression group at the 0.0005 significance level. There was no significant correlation between the percentage of endothelial cells lost and the emulsification time, or the irrigating volume, or a combination of the two (Table 1). The mean age in the phacoemulsification group was 68 years; in the nucleus expression group it was 73.2 years (Table 2). When the data were analyzed for sex and age differences, the following observations were made (Tables 3, 4, 5, and 6): The average patient in the nucleus expression group was older than the average patient in the phacoemulsification group at the 0.0005 Significance level. Preoperative counts of patients over 69.5 years were, in general, lower than those of patients under 69.5 years at the 0.0005 significance level.
Table 1. Analysis of 1,126 phacoemulsification and 632 nucleus expression patients.
Table 3. Preoperative cell counts in phacoemulsification and nucleus expression patients.
Phacoemulsification Nucleus Expression Time (min)
2.12 ± 0.84
Volume (ml)
185.9 ± 63.3
140.6 ± 63.0
Mean preoperative count (cell/mm 2)
2701 ± 485.4
2709.4 ± 473.1
Mean postoperative count (cell/mm2)
2287 ± 603.8
2377 ± 559.4
423 ± 488
332.2 ± 454.9
15.1 ± 19.6
11.9 ± 17.9
Cell reduction Percent reduction
Age (Years)
< 69.5 > 69.5
Type of Surgery
Age (Years)
Phacoemulsification
68.0 ± 10.2
Nucleus expression
73.2 ± 8.7
Mean
S.D.*
Nucleus Expression Number
Mean
S.D.*
594
2763.93 445.59
213
2793.85 406.61
532
2651.74 520.33
419
2666.51
498.57
*S.D. = standard deviation
Table 4. Percentage of cell reduction in phacoemulsification and nucleus expression patients. Age (Years)
Table 2. Age of patients studied.
Phacoemulsification Number
< 69.5 > 69.5
Phacoemulsification Number
Mean
594
13.81
532
16.61
S.D.*
Nucleus Expression Number
Mean
S.D.*
19.78
213
10.97
13.65
19.36
419
12.31
19.66
*S. D. = standard deviation
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
421
Table 5. Preoperative cell count by sex in phacoemulsification and nucleus expression patients. Phacoemulsification Number
Mean
Nucleus Expression
Variance* Number
Mean
Variance
Table 8. Univariate analysis of nucleus expression""~ons.
Surgeon Number
3
46
4
379
141.9 ± 55.9 326.3 ± 412.7
11.1 ± 18.9
Table 6. Percent cell reduction by sex in phacoemulsification and nucleus expression patients.
5
24
167.5 ± 103.4 273.3 ± 476.8
10.2 ± 17.3
6
19
161.1 ± 54.3 413.9 ± 308.4
16.5 ± 12.0
497
2766.80 183048.90
367 264
Phacoemulsification Number
2678.03 257520.35 2750.73 174128.50
78
10.6± 14.1 16.0 ± 16.6 12.1 ± 18.9
Nucleus Expression
Mean Variance * Number
Mean
Variance
Female 628
15.37
491.91
367
12.14
375.07
Male
14.77
249.42
264
11.48
243.33
497
2
Percent Reduction
*Variance = standard deviation squared
Male
2668.12 272342.80
86
Cell Count Reduction (cells/mm 2 )
149.2 ± 82.5 297.2 ± 409.1 127.3 ± 49.3 423.3 ± 399.6 113.3 ± 63.1 288.7 ± 476.7
Female 628
1
Irrigating Volume (cell/mm 2)
Table 9. Comparison of anterior chamber phacoemulsification (surgeon 5) with posterior chamber phacoemulsification (other surgeons). Surgeon 5
*Variance = standard deviation squared
Patient age (yrs)
In the nucleus expression group, there was no significant cell loss difference between the patients over 69.5 years and those under 69.5 years. However, in the phacoemulsification group, patients over 69.5 years exhibited a larger cell loss at the 0.05 significance level. Preoperative counts of female patients were, in general, lower than those of male patients at the 0.05 significance level. No significant difference between the percentage reduction in cells in female patients and that in male patients could be determined. . Tables 7, 8, and 9 present data of the individual surgeons and yield the following observations: The average percentage postoperative reduction of endothelial cells for each surgeon was significantly greater than zero for both groups at the 0.0005 significance level. The average percentage decrease between phacoemulsification and nucleus expression was similar for all surgeons except surgeon 5. The average percentage decrease for surgeon 5 was larger than the other surgeons at the 0.05 significance level. The average percentage reduction of endothelial
Irrigating volume (ml)
68.70 ±
8.10
67.90 ± 10.50
203.90 ± 65.60
183.20 ± 62.50
Emulsification time (min) Percent cell reduction
Other Surgeons
2.46 ±
0.88
26.40 ± 23.40
2.07 ±
0.82
13.40 ± 18.40
cells was not similar among surgeons in the phacoemulsification group; moreover, for surgeon 5, the reduction was significantly larger (at O. 001) than it was for other surgeons. The average percentage reduction of endothelial cells for surgeons in the nucleus expression group was approximately the same. The average phacoemulsification time among surgeons in the phacoemulsification group was variable, with surgeon 5 showing a significantly greater time than other members of the group at the 0.01 significance level. There was no significant difference among the surgeons in the average age of the patients. Table 9 demonstrates the comparison of the data from surgeon 5 who employed anterior chamber phacoemulsification and the pooled data of the other surgeons who used posterior chamber phacoemulsification.
Table 7. Univariate analysis of phacoemulsification by surgeon. Surgeon
422
Number
Emulsification Time (min)
Irrigating Volume (ml)
Cell Count Reduction (cell/mm 2)
Percent Reduction
1
143
2.04 ± 0.74
194.8 ± 64.1
286.8 ± 367.1
2
156
2.08 ± 0.93
210.2 ± 53.9
349.7 ± 426.5
13.0 ± 17.0
3
462
1.95 ± 0.74
153.9 ± 50.8
367.5 ± 482.5
12.8 ± 20.5
4
77
1.73 ± 0.61
212.5 ± 63.6
483.9 ± 521. 7
17.6 ± 19.6
5
151
2.46 ± 0.88
203.9 ± 65.6
730.0 ± 553.9
26.4 ± 23.4
6
136
1.85 ± 0.65
222.5 ± 58.4
461.1 ± 447.7
16.7 ± 15.4
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
9.9 ± 13.0
There was a higher percentage of cell loss in the former group of patients.
DISCUSSION The most surprising observation of this study is the apparent lack of correlation between the irrigating volume and/or the emulsification time and the reduction in endothelial cells. Although these data were studied and subjected to age, time, and volume combinations and permutations for data sets of individual surgeons (Tables 7 and 8), as well as the data as a whole, no relationship could be established. Even when quadratic terms were introduced, no acceptable correlations were found. As noted, there was a correlation between the loss of endothelial cells during emulsification and age, the older endothelium being more vulnerable. This confirms the observation of Waltman and Cozean. 1 There was also a definite increase in average cell loss in the phacoemulsification group when compared with the nucleus expression group, as observed by several authors.l- 7 Although the average cell loss with each procedure was modest, the phacoemulsification procedure produced a much wider range of cell loss, with alarming decrease in cell density in some patients. This observation suggests careful patient selection, especially in the more vulnerable, older patient. The procedure should be avoided where pupil size, shallow chamber, posterior pressure, or other factors will result in the procedure being performed close to the endothelium. One of the six surgeons in this study (surgeon 5) had continued to emulsifY the nucleus as originally taught by Charles Kelman, i. e., by displacing the nucleus into the anterior chamber prior to emulsification. The other surgeons employed techniques whereby most of the nucleus was emulsified prior to dislocation so most of the emulsification took place behind the iris plane. Surgeon 5 averaged a cell loss of26%; combined average cell loss for the other surgeons was 13%. Surgeon 5 also had higher average emulsification times and irrigating volumes. However, attempts to obtain correlations between cell loss and volume and/or time were unsuccessful in his isolated data set as in the data as a whole. Therefore, it appears that location of the nucleus vis a vis the endothelium during emulsification is the cause of the increased endothelial damage, an observation made previously by KrafT 4 and Hoffer. 7
If the first eye experienced significant cell loss with phacoemulsification, a different approach should be selected for the fellow eye. For the occasional surgeon, nucleus expression is likely to be a safer choice. SUMMARY A group of patients undergoing cataract extraction with posterior chamber lens implantation was studied for the effects of various techniques on preoperative and postoperative cell density. The nucleus was removed by phacoemulsification in 1,126 patients and by simple expression of the nucleus in 632 patients. The endothelial cell loss in the phacoemulsification group (15%) was moderately higher than in the nucleus expression group (12%), but both were in the expected range for this operation. 5 No correlation between the irrigating volume or the phacoemulsification time and the endothelial cell loss could be found in either group. There was a correlation between the patient's age and the endothelial cell loss during phacoemulsification, with older corneas (over 69.5 years of age) showing a greater percentage loss than corneas under 69.5 years. There was also a correlation between the location of the nucleus and the endothelium during emulsification; significantly greater loss occurred when the nucleus was emulsified in the anterior chamber. REFERENCES 1. Waltman SR, Cozean CH Jr: The effect of phacoemulsification on the corneal endothelium. Ophthalmic Surg 10(1):31-33, 1979 2. Abbott RL, Forster RK: Clinical specular microscopy and intraocular surgery. Arch Ophthalmol 97:1476-1479, 1979 3. Stanley JA, Shearing SP, Anderson RR, Avallone AN: Endothelial cell density after posterior chamber lens implantation. Ophthalmology 87:381-384, 1980 4. KraffMC, Sanders DR, Lieberman HL: Specular microscopy in cataract and intraocular lens patients. A report of 564 cases. Arch Ophthalmol 98:1782-1784, 1980 5. Bourne WM, Waller RR, Liesegang TJ, Brubaker RF: Corneal trauma in intracapsular and extracapsular cataract extraction with lens implantation. Arch Ophthalmol 99:1375-1376, 1981 6. KraffMC, Sanders DR, Lieberman HL: Monitoring for continuing endothelial cell loss with cataract extraction and intraocular lens implantation. Ophthalmology 89:30-34, 1982 7. Hoffer KJ: Effects of extracapsular implant techniques on endothelial density. Arch Ophthalmol 100:791-792, 1982 8. Statistical Analysis System (SAS), SAS Institute Inc., Cary, NC, 1979 9. Snedecor G, Cochran WG: Statistical Methods, 6th ed., Ames, Iowa, Iowa State University Press, 1967
AM INTRA-OCULAR IMPLANT SOC J-VOL. 9, FALL 1983
423