- Email: [email protected]
Posterior Vitreous Detachment and Neovascularization in Diabetic Retinopathy
Posterior Vitreous Detachment and N eovascularization in Diabetic Retinopathy JUN AKIBA, MD, CARLOS W. ARZABE, MD, CLEMENT L. TREMPE, MD
Abstract: The authors evaluated the effect of posterior vitreous detachment (PVD) in 172 eyes (114 patients) with nonproliferative diabetic retinopathy on the development of retinal or optic disc neovascularization. Patients were followed for more than 6 months after the initial vitreous examination (mean, 32 months). None of the patients 39 years of age or younger had PVD, but this condition became more prevalent thereafter with age. In 124 eyes of patients 40 years of age or older, neovascularization developed eventually in 20 (22%) of 93 eyes without PVD but in only 1 (3%) of 29 eyes with complete PVD (P < 0.05). A vitreous examination was invaluable in predicting the development of retinal or optic disc neovascularization in diabetic patients 40 years of age or older. Ophthalmology 1990; 97:889-891
The development of retinal or optic disc neovascularization is a critical stage of diabetic retinopathy that could cause severe visual loss. It is well recognized that retinal ischemia, shown as capillary non perfusion by fluorescein angiography, predisposes diabetic retinopathy patients to neovascularization. I - 4 However, other factors that affect the development of neovascularization in these patients are not well defined. The role of the vitreous in advanced proliferative diabetic retinopathy is well established. Partial posterior vitreous detachment (PVD) with traction on the neovascular tissue is a major risk factor for the progression of proliferative diabetic retinopathy to vitreous hemorrhage or tractional retinal detachment. s-9 However, the effect of PVD on the development of neovascularization is still unclear. We studied the vitreous condition in 172 eyes with non proliferative diabetic retinopathy to assess the effect of PVD on diabetic neovascularization. Originally received: October 17, 1989. Revision accepted: April 2, 1990. From the Eye Research Institute, Retina Associates, and Department of Ophthalmology, Harvard Medical School, Boston. Dr. Arzabe is currently affiliated with the Ophthalmic Foundation of the Orient, Santa Cruz, Bolivia. Dr. Akiba is currently affiliated with the Department of Ophthalmology, Asahikawa, Medical College, Hokkaido, Japan. Reprint requests to Library, Eye Research Institute, 20 Staniford St, Boston, MA 02114.
PATIENTS AND METHODS
We reviewed the records of 114 patients ( 172 eyes) with nonproliferative diabetic retinopathy who were followed for more than 6 months (mean, 32 months; maximum, 102 months) after the initial vitreous examination. There were 54 women and 60 men. Patient ages ranged from 17 to 79 years (mean, 53 years). The average age (mean) when diabetes was first diagnosed was 35 years, and the average duration of diabetes (mean) was 18 years. Included in the study were 85 insulin-dependent patients and 29 patients whose diabetes was controlled with oral hypoglycemic agents or diet. Eyes with a history of cataract surgery or retinal surgery and intraocular disease that might influence the development of neovascularization such as retinal vascular occlusion were excluded. Since the objective of this study is to investigate the effect of PVD on the development of diabetic neovascularization, eyes that received prophylactic panretinal photocoagulation were excluded from the study to preclude the prevention effects oflaser photocoagulation on neovascularization. When retinal or disc neovascularization developed during the study, panretinal photocoagulation was carried out, and we included only the period until neovascularization developed as followup data. 889
OPHTHALMOLOGY
100
•
JULY 1990
rr-
80
r-
!!
60
-
II)
~
40
20
~
~
0
17·39
40·49
50·59
[] No
~ ~ ~
I"JD
•
Partial PVC
D
Compie1e PVC
~ v::
-
60·69
70·79
Aoe (years) Fig 1. Proportion of eyes with each vitreous condition in nonproliferative diabetic retinopathy by age group. PVD = posterior vitreous detachment.
Table 1. Relationships of Posterior Vitreous Detachment, Patient Age, and Onset and Duration of Diabetes Diabetic Patient Characteristics Age at vitreous examination (mean ± SO, yrs) Age at onset of diabetes (mean ± SO, yrs) Duration of diabetes (mean ± SO, yrs) PVD
=
No PVD
Complete PVD
49 ± 15*
63 ± 7*
31 ± 8*
46 ± 11*
18 ± 10
17 ± 8
posterior vitreous detachment; SO
* P< 0.0001.
=
standard deviation.
Table 2. Relationships Among Posterior Vitreous Detachment, Neovascularization, and Age in Diabetic Conditions Prevalence of Neovascularization
Age at examination (range, yrs) 17-39 40-79 Age at onset of diabetes (range, yrs) 5-39 40-70
No PVD
Complete PVD
No. of Eyes (%)
No. of Eyes (%)
15/48 (31) 20/93 (22)*
0/0 (0) 1/29 (3)*
23/76 (30) 12/65 (18)*
1/8 (13) 0/21 (0)*
PVD = posterior vitreous detachment. *
P< 0.05.
Measurements of corrected visual acuity, biomicroscopic examination, indirect ophthalmoscopy, and fundus photography were performed at each visit. Three independent observers diagnosed the presence of retinal or optic disc neovascularization based on results from binocular indirect ophthalmoscopy and fundus biomicros890
•
VOLUME 97
•
NUMBER 7
copy. fluorescein angiography, performed on 88 (51%) of 172 eyes, also was helpful in making the diagnosis. Eyes were assigned to one of three groups, according to the initial vitreous examination: no, partial, or complete PVD. The classification was proposed by Takahashi and associates7 and based on the position of the vitreous cortex relative to the retina posterior to the equator. The vitreous was examined and photographed by two of the authors (JA and CLT) using an EI Bayadi-Kajiura lens (+58.6diopter aspheric preset lens) mounted on a photographic slit lamp. Statistical analysis was performed using the MannWhitney U test and Fisher's exact test. A finding was considered significant if P < 0.05.
RESULTS At the initial vitreous examination, no PVD was noted in 141 (82%) of 172 eyes, a partial detachment was noted in 2 eyes (1 %), and a complete detachment was noted in 29 eyes (17%). No patient 39 years of age or younger had PVD, but thereafter the prevalence of complete or partial PVD increased with age (Fig 1). The average age when the vitreous examination was performed was significantly higher in eyes with complete PVD than without (P < 0.0001) (Table 1). In addition, the average age for onset of diabetes was significantly higher in eyes with complete PVD than in eyes without this condition (P < 0.0001). However, the average duration of diabetes was not significantly different in these groups. During a mean follow-up of 32 months, retinal or optic disc neovascularization developed in 36 (21 %) of 172 eyes. In 48 eyes of patients 39 years of age or younger who had no PVD, neovascularization developed in 15 eyes (31 %) (Table 2). In 124 eyes of patients 40 years of age or older, retinal or optic disc neovascularization developed in 20 (22%) of93 eyes with no PVD and in the peripheral retina near the vitreous base in only 1 (3%) of 29 eyes with complete PVD. The prevalence of optic disc or retinal neovascularization in eyes without PVD was significantly higher than that in eyes with complete PVD in patients 40 years of age or older (P < 0.05). The follow-up durations between both groups were not significantly different. Complete PVD developed during follow-up in seven eyes initially without this condition, but neovascularization did not occur. Therefore, 36 eyes (21 %) had complete PVD at the end of follow-up and retinal neovascularization developed eventually in one (3%) of these eyes. Neovascularization did not develop in the two eyes with partial PVD.
DISCUSSION Our study showed that PVD may prevent retinal or optic disc neovascularization in patients with diabetic retinopathy, although the mechanism is unknown. Kado and Trempe lO reported that compared with eyes with no PVD,
AKIBA et al
•
VITREOUS DETACHMENT AND DIABETIC NEOVASCULARIZATION
eyes with complete PVD and branch retinal vein occlusion had significantly less risk of neovascularization developing even if large areas of nonperfusion existed. We believe that vitreoretinal attachment is strongly associated with retinal or optic disc neovascularization in retinal ischemic disease. Retinal neovascularization regresses if the posterior hyaloid membrane is separated from the new vessel spontaneously5 or surgically I I in proliferative diabetic retinopathy. A'histologic study conducted by Faulborn and Bowald l2 found that the fibrous material of the vitreous cortex was interconnected with and incorporated into the newly formed proliferative tissue in proliferative diabetic retinopathy. We speculate that vitreoretinal attachment may provide the necessary support for neovascularization. Two factors should be considered in evaluating the vitreous condition in patients with diabetic retinopathy. First, the vitreous changes with age. In normal eyes, liquefaction of the vitreous may begin in childhood. 13 The volume of liquefied vitreous increases with age, which may lead to the development ofPVD in patients 40 years of age or older. 14,15 Second, diabetic retinopathy may cause vitreous shrinkage. Plasma leakage or subclinical hemorrhage in the vitreous gel, observed occasionally in eyes with diabetic retinopathy, may cause the pathologic vitreous shrinkage, which may lead to PVD even in young patients. Our results demonstrate that PVD may be related more to patient age than duration of diabetes in eyes with nonproliferative diabetic retinopathy (see Table 1). In contrast to the proliferative stage, aging may influence the vitreous condition more than diabetes in patients in the nonproliferative stage. Our results demonstrated that the majority of patients whose onset age of diabetes was less than 40 years had no PVD (see Table 2). Thus, type I diabetic persons with nonproliferative diabetic retinopathy are unlikely to have PVD. However, retinal or optic disc neovascularization is more likely to develop in association with type I than type II diabetes. We assume that the high prevalence of no PVD in patients with type I diabetes may contribute to the higher risk of neovascularization. Although the EI Bayadi-Kajiura lens we used to examine the vitreous condition does not allow observation of the vitreous anterior to the equator, this technique has many advantages over the conventional method with the Goldmann contact lens. 16 We believe that the observation of the vitreous condition in the posterior pole and midperiphery is useful in predicting the development of neovascularization in diabetic retinopathy because neovascularization primarily develops in this area. 3 In this study, neovascularization developed in only one eye with complete PVD from the peripheral retina. In this case, we speculate that the vitreous might have been attached to the retina anterior to the equator where neovascularization developed.
Predicting the possibility of retinal or optic disc neovascularization is important in the management of diabetic retinopathy. Our study indicates that (1) the prevalence of PVD increases with age in patients 40 years of age or older, and (2) the development of neovascularization is significantly rare if the vitreous is detached completely in diabetic retinopathy. We recommend a vitreous examination in non proliferative diabetic retinopathy patients 40 years of age or older, since it may be beneficial in detecting which patients are protected from neovascularization because ofPVD and because PVD is virtually nonexistent in patients younger than 40 years of age.
REFERENCES 1. Wise GN. Factors influencing retinal new vessel formation. Am J Ophthalmol1961; 52:637-50. 2. Bresnick GH, DeVenecia G, Myers FL, et al. Retinal ischemia in diabetic retinopathy. Arch Ophthalmol1975; 93:1300-10. 3. Shimizu K, Kobayashi Y, Muraoka K. Midperipheral fundus involvement in diabetic retinopathy. Ophthalmology 1981; 88:601-12. 4. Patz A. Clinical and experimental studies on retinal neovascularization. XXXIX Edward Jackson Memorial Lecture. Am J Ophthalmol 1982; 94:715-43. 5. Davis MD. Vitreous contraction in proliferative diabetic retinopathy. Arch Ophthalmol1965; 74:741-51. 6. Tolentino FI, Lee P-F, Schepens CL. Biomicroscopic study of vitreous cavity in diabetic retinopathy. Arch Ophthalmol1966; 75:238-46. 7. Takahashi M, Trempe CL, Maguire K, McMeel JW. Vitreoretinal rela· tionship in diabetic retinopathy: a biomicroscopic evaluation. Arch Ophthalmol 1981; 99:241-5. 8. Jalkh A, Takahashi M, Topilow HW, et al. Prognostic value of vitreous findings in diabetic retinopathy. Arch Ophthalmol1982; 100:432-4. 9. Tagawa H, McMeel JW, Trempe CL. Role of the vitreous in diabetic retinopathy. II. Active and inactive vitreous changes. Ophthalmology 1986; 93:1188-92. 10. Kado M, Trempe CL. Role of the vitreous in branch retinal vein occlusion. Am J Ophthalmol1988; 105:20-4. 11. Federman JL, Boyer D, Lanning R, Breit P. An objective analysis of proliferative diabetic retinopathy before and after pars plana vitrectomy. Ophthalmology 1979; 86:276-82. 12. Faulbom J, Bowald S. Microproliferations in proliferative diabetic ret· inopathy and their relationship to the vitreous: corresponding light and electron microscopic studies. Graefes Arch Clin Exp Ophthalmol 1985; 223:130-8. 13. Balazs EA, Denlinger JL. Aging changes in the vitreous. In: Sekular R, Kline D, Dismukes K, eds. Aging and Human Visual Functions. New York: Alan R Liss, 1982; 45-57. (Modern Aging Research; vol. 2.) 14. Foos RY. Posterior vitreous detachment. Trans Am Acad Ophthalmol Otolaryngol1972; 76:480-97. 15. Takahashi M. Posterior vitreous detachment as aging process: analysis of 1,077 normal eyes. Jpn J Clin Ophthalmol 1982; 36: 1137-41. 16. Buzney SM, Weiter JJ, Furukawa H, et al. Examination of the vitreous. A comparison of biomicroscopy using the Goldmann and EI BayadiKajiura lenses. Ophthalmology 1985; 92:1745-8.
891
Abstract: The authors evaluated the effect of posterior vitreous detachment (PVD) in 172 eyes (114 patients) with nonproliferative diabetic retinopathy on the development of retinal or optic disc neovascularization. Patients were followed for more than 6 months after the initial vitreous examination (mean, 32 months). None of the patients 39 years of age or younger had PVD, but this condition became more prevalent thereafter with age. In 124 eyes of patients 40 years of age or older, neovascularization developed eventually in 20 (22%) of 93 eyes without PVD but in only 1 (3%) of 29 eyes with complete PVD (P < 0.05). A vitreous examination was invaluable in predicting the development of retinal or optic disc neovascularization in diabetic patients 40 years of age or older. Ophthalmology 1990; 97:889-891
The development of retinal or optic disc neovascularization is a critical stage of diabetic retinopathy that could cause severe visual loss. It is well recognized that retinal ischemia, shown as capillary non perfusion by fluorescein angiography, predisposes diabetic retinopathy patients to neovascularization. I - 4 However, other factors that affect the development of neovascularization in these patients are not well defined. The role of the vitreous in advanced proliferative diabetic retinopathy is well established. Partial posterior vitreous detachment (PVD) with traction on the neovascular tissue is a major risk factor for the progression of proliferative diabetic retinopathy to vitreous hemorrhage or tractional retinal detachment. s-9 However, the effect of PVD on the development of neovascularization is still unclear. We studied the vitreous condition in 172 eyes with non proliferative diabetic retinopathy to assess the effect of PVD on diabetic neovascularization. Originally received: October 17, 1989. Revision accepted: April 2, 1990. From the Eye Research Institute, Retina Associates, and Department of Ophthalmology, Harvard Medical School, Boston. Dr. Arzabe is currently affiliated with the Ophthalmic Foundation of the Orient, Santa Cruz, Bolivia. Dr. Akiba is currently affiliated with the Department of Ophthalmology, Asahikawa, Medical College, Hokkaido, Japan. Reprint requests to Library, Eye Research Institute, 20 Staniford St, Boston, MA 02114.
PATIENTS AND METHODS
We reviewed the records of 114 patients ( 172 eyes) with nonproliferative diabetic retinopathy who were followed for more than 6 months (mean, 32 months; maximum, 102 months) after the initial vitreous examination. There were 54 women and 60 men. Patient ages ranged from 17 to 79 years (mean, 53 years). The average age (mean) when diabetes was first diagnosed was 35 years, and the average duration of diabetes (mean) was 18 years. Included in the study were 85 insulin-dependent patients and 29 patients whose diabetes was controlled with oral hypoglycemic agents or diet. Eyes with a history of cataract surgery or retinal surgery and intraocular disease that might influence the development of neovascularization such as retinal vascular occlusion were excluded. Since the objective of this study is to investigate the effect of PVD on the development of diabetic neovascularization, eyes that received prophylactic panretinal photocoagulation were excluded from the study to preclude the prevention effects oflaser photocoagulation on neovascularization. When retinal or disc neovascularization developed during the study, panretinal photocoagulation was carried out, and we included only the period until neovascularization developed as followup data. 889
OPHTHALMOLOGY
100
•
JULY 1990
rr-
80
r-
!!
60
-
II)
~
40
20
~
~
0
17·39
40·49
50·59
[] No
~ ~ ~
I"JD
•
Partial PVC
D
Compie1e PVC
~ v::
-
60·69
70·79
Aoe (years) Fig 1. Proportion of eyes with each vitreous condition in nonproliferative diabetic retinopathy by age group. PVD = posterior vitreous detachment.
Table 1. Relationships of Posterior Vitreous Detachment, Patient Age, and Onset and Duration of Diabetes Diabetic Patient Characteristics Age at vitreous examination (mean ± SO, yrs) Age at onset of diabetes (mean ± SO, yrs) Duration of diabetes (mean ± SO, yrs) PVD
=
No PVD
Complete PVD
49 ± 15*
63 ± 7*
31 ± 8*
46 ± 11*
18 ± 10
17 ± 8
posterior vitreous detachment; SO
* P< 0.0001.
=
standard deviation.
Table 2. Relationships Among Posterior Vitreous Detachment, Neovascularization, and Age in Diabetic Conditions Prevalence of Neovascularization
Age at examination (range, yrs) 17-39 40-79 Age at onset of diabetes (range, yrs) 5-39 40-70
No PVD
Complete PVD
No. of Eyes (%)
No. of Eyes (%)
15/48 (31) 20/93 (22)*
0/0 (0) 1/29 (3)*
23/76 (30) 12/65 (18)*
1/8 (13) 0/21 (0)*
PVD = posterior vitreous detachment. *
P< 0.05.
Measurements of corrected visual acuity, biomicroscopic examination, indirect ophthalmoscopy, and fundus photography were performed at each visit. Three independent observers diagnosed the presence of retinal or optic disc neovascularization based on results from binocular indirect ophthalmoscopy and fundus biomicros890
•
VOLUME 97
•
NUMBER 7
copy. fluorescein angiography, performed on 88 (51%) of 172 eyes, also was helpful in making the diagnosis. Eyes were assigned to one of three groups, according to the initial vitreous examination: no, partial, or complete PVD. The classification was proposed by Takahashi and associates7 and based on the position of the vitreous cortex relative to the retina posterior to the equator. The vitreous was examined and photographed by two of the authors (JA and CLT) using an EI Bayadi-Kajiura lens (+58.6diopter aspheric preset lens) mounted on a photographic slit lamp. Statistical analysis was performed using the MannWhitney U test and Fisher's exact test. A finding was considered significant if P < 0.05.
RESULTS At the initial vitreous examination, no PVD was noted in 141 (82%) of 172 eyes, a partial detachment was noted in 2 eyes (1 %), and a complete detachment was noted in 29 eyes (17%). No patient 39 years of age or younger had PVD, but thereafter the prevalence of complete or partial PVD increased with age (Fig 1). The average age when the vitreous examination was performed was significantly higher in eyes with complete PVD than without (P < 0.0001) (Table 1). In addition, the average age for onset of diabetes was significantly higher in eyes with complete PVD than in eyes without this condition (P < 0.0001). However, the average duration of diabetes was not significantly different in these groups. During a mean follow-up of 32 months, retinal or optic disc neovascularization developed in 36 (21 %) of 172 eyes. In 48 eyes of patients 39 years of age or younger who had no PVD, neovascularization developed in 15 eyes (31 %) (Table 2). In 124 eyes of patients 40 years of age or older, retinal or optic disc neovascularization developed in 20 (22%) of93 eyes with no PVD and in the peripheral retina near the vitreous base in only 1 (3%) of 29 eyes with complete PVD. The prevalence of optic disc or retinal neovascularization in eyes without PVD was significantly higher than that in eyes with complete PVD in patients 40 years of age or older (P < 0.05). The follow-up durations between both groups were not significantly different. Complete PVD developed during follow-up in seven eyes initially without this condition, but neovascularization did not occur. Therefore, 36 eyes (21 %) had complete PVD at the end of follow-up and retinal neovascularization developed eventually in one (3%) of these eyes. Neovascularization did not develop in the two eyes with partial PVD.
DISCUSSION Our study showed that PVD may prevent retinal or optic disc neovascularization in patients with diabetic retinopathy, although the mechanism is unknown. Kado and Trempe lO reported that compared with eyes with no PVD,
AKIBA et al
•
VITREOUS DETACHMENT AND DIABETIC NEOVASCULARIZATION
eyes with complete PVD and branch retinal vein occlusion had significantly less risk of neovascularization developing even if large areas of nonperfusion existed. We believe that vitreoretinal attachment is strongly associated with retinal or optic disc neovascularization in retinal ischemic disease. Retinal neovascularization regresses if the posterior hyaloid membrane is separated from the new vessel spontaneously5 or surgically I I in proliferative diabetic retinopathy. A'histologic study conducted by Faulborn and Bowald l2 found that the fibrous material of the vitreous cortex was interconnected with and incorporated into the newly formed proliferative tissue in proliferative diabetic retinopathy. We speculate that vitreoretinal attachment may provide the necessary support for neovascularization. Two factors should be considered in evaluating the vitreous condition in patients with diabetic retinopathy. First, the vitreous changes with age. In normal eyes, liquefaction of the vitreous may begin in childhood. 13 The volume of liquefied vitreous increases with age, which may lead to the development ofPVD in patients 40 years of age or older. 14,15 Second, diabetic retinopathy may cause vitreous shrinkage. Plasma leakage or subclinical hemorrhage in the vitreous gel, observed occasionally in eyes with diabetic retinopathy, may cause the pathologic vitreous shrinkage, which may lead to PVD even in young patients. Our results demonstrate that PVD may be related more to patient age than duration of diabetes in eyes with nonproliferative diabetic retinopathy (see Table 1). In contrast to the proliferative stage, aging may influence the vitreous condition more than diabetes in patients in the nonproliferative stage. Our results demonstrated that the majority of patients whose onset age of diabetes was less than 40 years had no PVD (see Table 2). Thus, type I diabetic persons with nonproliferative diabetic retinopathy are unlikely to have PVD. However, retinal or optic disc neovascularization is more likely to develop in association with type I than type II diabetes. We assume that the high prevalence of no PVD in patients with type I diabetes may contribute to the higher risk of neovascularization. Although the EI Bayadi-Kajiura lens we used to examine the vitreous condition does not allow observation of the vitreous anterior to the equator, this technique has many advantages over the conventional method with the Goldmann contact lens. 16 We believe that the observation of the vitreous condition in the posterior pole and midperiphery is useful in predicting the development of neovascularization in diabetic retinopathy because neovascularization primarily develops in this area. 3 In this study, neovascularization developed in only one eye with complete PVD from the peripheral retina. In this case, we speculate that the vitreous might have been attached to the retina anterior to the equator where neovascularization developed.
Predicting the possibility of retinal or optic disc neovascularization is important in the management of diabetic retinopathy. Our study indicates that (1) the prevalence of PVD increases with age in patients 40 years of age or older, and (2) the development of neovascularization is significantly rare if the vitreous is detached completely in diabetic retinopathy. We recommend a vitreous examination in non proliferative diabetic retinopathy patients 40 years of age or older, since it may be beneficial in detecting which patients are protected from neovascularization because ofPVD and because PVD is virtually nonexistent in patients younger than 40 years of age.
REFERENCES 1. Wise GN. Factors influencing retinal new vessel formation. Am J Ophthalmol1961; 52:637-50. 2. Bresnick GH, DeVenecia G, Myers FL, et al. Retinal ischemia in diabetic retinopathy. Arch Ophthalmol1975; 93:1300-10. 3. Shimizu K, Kobayashi Y, Muraoka K. Midperipheral fundus involvement in diabetic retinopathy. Ophthalmology 1981; 88:601-12. 4. Patz A. Clinical and experimental studies on retinal neovascularization. XXXIX Edward Jackson Memorial Lecture. Am J Ophthalmol 1982; 94:715-43. 5. Davis MD. Vitreous contraction in proliferative diabetic retinopathy. Arch Ophthalmol1965; 74:741-51. 6. Tolentino FI, Lee P-F, Schepens CL. Biomicroscopic study of vitreous cavity in diabetic retinopathy. Arch Ophthalmol1966; 75:238-46. 7. Takahashi M, Trempe CL, Maguire K, McMeel JW. Vitreoretinal rela· tionship in diabetic retinopathy: a biomicroscopic evaluation. Arch Ophthalmol 1981; 99:241-5. 8. Jalkh A, Takahashi M, Topilow HW, et al. Prognostic value of vitreous findings in diabetic retinopathy. Arch Ophthalmol1982; 100:432-4. 9. Tagawa H, McMeel JW, Trempe CL. Role of the vitreous in diabetic retinopathy. II. Active and inactive vitreous changes. Ophthalmology 1986; 93:1188-92. 10. Kado M, Trempe CL. Role of the vitreous in branch retinal vein occlusion. Am J Ophthalmol1988; 105:20-4. 11. Federman JL, Boyer D, Lanning R, Breit P. An objective analysis of proliferative diabetic retinopathy before and after pars plana vitrectomy. Ophthalmology 1979; 86:276-82. 12. Faulbom J, Bowald S. Microproliferations in proliferative diabetic ret· inopathy and their relationship to the vitreous: corresponding light and electron microscopic studies. Graefes Arch Clin Exp Ophthalmol 1985; 223:130-8. 13. Balazs EA, Denlinger JL. Aging changes in the vitreous. In: Sekular R, Kline D, Dismukes K, eds. Aging and Human Visual Functions. New York: Alan R Liss, 1982; 45-57. (Modern Aging Research; vol. 2.) 14. Foos RY. Posterior vitreous detachment. Trans Am Acad Ophthalmol Otolaryngol1972; 76:480-97. 15. Takahashi M. Posterior vitreous detachment as aging process: analysis of 1,077 normal eyes. Jpn J Clin Ophthalmol 1982; 36: 1137-41. 16. Buzney SM, Weiter JJ, Furukawa H, et al. Examination of the vitreous. A comparison of biomicroscopy using the Goldmann and EI BayadiKajiura lenses. Ophthalmology 1985; 92:1745-8.
891