A scattergram shows the best-corrected preoperative Snellen visual acuity vs the final visual acuity in groups 1 (patients with high myopia) (left) and 2 (patients with high myopia and presumed ocular histoplasmosis syndrome) (right). HM indicates hand movements; LP, light perception; and NLP, no light perception. Points below the line represent eyes with an improved final visual acuity.
Patient 1. Preoperative fundus photograph (left) and fluorescein angiogram (right) of a 39-year-old man with degenerative myopia. His visual acuity is 20/400.
Patient 1. Postoperative fundus photograph 1 (left) and 42 (right) months after surgery. His visual acuity is 20/80, and he has marked chorioretinal atrophy.
Patient 24. Preoperative fundus photograph (left) and fluorescein angiogram (right) of a 49-year-old man with high myopia and presumed ocular histoplasmosis. His visual acuity is 20/300.
Patient 24. Postoperative fundus photograph 4 years after surgery. His visual acuity is 20/25, and he has chorioretinal atrophy.
Uemura A, Thomas MA. Subretinal Surgery for Choroidal Neovascularization in Patients With High Myopia. Arch Ophthalmol. 2000;118(3):344-350. doi:10.1001/archopht.118.3.344
To analyze the visual outcome in patients undergoing surgical removal of subfoveal choroidal neovascularization (CNV) in eyes with high myopia.
We retrospectively reviewed the medical records of 48 consecutive patients with high myopia (≥6 diopters [D]) who underwent vitrectomy with surgical removal of subfoveal CNV. The patient population consisted of 2 groups. Group 1 included 23 patients with findings only of myopic degeneration, and group 2 included 25 patients with presumed ocular histoplasmosis syndrome and myopia of 6 D or more.
In group 1, the visual acuity improved by 2 or more Snellen lines in 9 eyes (39%), decreased in 8 eyes (35%), and remained unchanged in 6 (26%), with a mean follow-up of 24 months (range, 8-60 months). The preoperative visual acuity was 20/40 or better in only 1 eye (4%), but 8 (35%) achieved a final visual acuity of 20/40 or better. In group 2, the visual acuity improved in 16 eyes (64%), was stable in 4 (16%), and deteriorated in 5 (20%), with a mean follow-up of 18 months (range, 6-44 months). Only 3 eyes (12%) had a preoperative visual acuity of 20/40 or better, but 11 (44%) achieved a final visual acuity of 20/40 or better. Recurrence occurred in 13 (57%) of the 23 eyes in group 1 and in 9 (36%) of the 25 eyes in group 2. Univariate analysis demonstrated a significant relation between younger patient age (group 1) and absence of postoperative CNV recurrence (group 2) and an improvement of visual acuity (P<.01).
Surgical removal of CNV may provide visual benefit in selected cases of subfoveal CNV associated with high myopia. The determination of whether surgical intervention is appropriate in these cases requires a prospective, randomized, clinical trial.
DEGENERATIVE myopia is an ocular disease characterized by excessive axial length, abnormal visual function, and several changes in the ocular tissues.1- 4 The major complication leading to decreased central vision in these eyes is choroidal neovascularization (CNV).5- 8 Treatment of myopic CNV has been controversial. It is still uncertain whether laser photocoagulation is beneficial in the treatment of CNV associated with high myopia.9- 12 Laser treatment of subfoveal CNV usually results in an absolute scotoma and a permanent, sudden decrease in visual acuity.13 Thomas et al14 described the surgical results in 10 patients with high myopia in 1994. Since then, a few other investigators15,16 have reported their clinical experience with subfoveal surgery related to degenerative myopia. The visual results did not seem to be satisfactory, although the small number of patients and the short-term follow-up precluded a definitive assessment of the effectiveness or ineffectiveness of the surgery.
We have reviewed retrospectively 48 consecutive patients who underwent surgical removal of subfoveal CNV associated with high myopia, and analyzed postoperative visual outcomes as a primary end point after dividing the patients into groups according to the causes.
We retrospectively reviewed the medical records of patients with high myopia who underwent vitrectomy with surgical removal of CNV by a single surgeon (M.A.T.) between April 1, 1991, and June 30, 1997. Eyes were classified as highly myopic if the preoperative refractive error showed 6 or more diopters (D) of myopia (in spherical equivalents). We referred to the refraction data from the referring ophthalmologist for those patients who had a history of cataract extraction or refractive surgery for myopia. One patient with a history of radial keratotomy was ineligible because the refraction before radial keratotomy was unknown. We (A.U. and M.A.T.) independently reviewed preoperative fluorescein angiograms and confirmed angiographic evidence of CNV under the geometric center of the foveal avascular zone in all eyes. No review of angiograms was obtained elsewhere.
Sixty-seven eyes of 67 patients were identified that met our selection criteria except for the follow-up criterion. Six-month follow-up visits or longer were obtained in 52 eyes of 52 patients. One additional patient was excluded because the patient had a history of submacular surgery that was performed elsewhere in the study eye. Therefore, we evaluated the surgical results in 51 eyes of 51 patients. The patient population was divided into 3 groups according to the presumed cause of the CNV.
Group 1 included 23 patients with findings only of degenerative myopia. All eyes in this group had the stigmata of degenerative myopia on clinical fundus examination and fluorescein angiography: fundus pallor and tessellation, diffusely thinned retinal pigment epithelium (RPE) and/or focal retinal pigment attenuation, diffuse or focal chorioretinal atrophy, posterior staphyloma, or lacquer cracks.7 Group 2 included 25 patients with high myopia who also had presumed ocular histoplasmosis syndrome (POHS) as an underlying diagnosis. All the patients in this group had atrophic scars, peripapillary atrophic lesions characteristic of POHS, or both in addition to myopic changes such as tigroid fundus. Group 3 included 3 patients with subfoveal CNV in eyes with high myopia who also had other potential causes for the neovascularization. We excluded the patients in group 3 from the analysis because they had various potential causes for the neovascularization: congenital toxoplasmosis, multifocal choroiditis, and posttrauma each in 1 eye.
All patients underwent a complete preoperative ocular examination, including determination of the best-corrected visual acuity by Snellen visual acuity chart, which was recorded as one of the gradations shown in Table 1. Change in visual acuity was defined as better (increase of ≥2 lines of Snellen acuity), same (within 1 line of the baseline Snellen acuity), or worse (loss of ≥2 lines of Snellen acuity). After obtaining appropriate informed consent, pars plana vitrectomy and removal of the subfoveal CNV were performed according to techniques described previously.14 Following pars plana vitrectomy with removal of the posterior hyaloid, a small retinotomy was created using a 36-gauge subretinal pick. The neurosensory retina was elevated by infusing balanced salt solution with a 33-gauge subretinal infusion cannula. A 36-gauge pick was used to break adhesions between the membrane and overlying neurosensory retina and then to elevate the edge of the membrane from underlying tissue. Angled, horizontal, subretinal forceps were used to grasp and extract the neovascular complex, often with the preexisting laser scar (if any). An air-fluid exchange was performed by aspirating over the optic disc and/or through the retinotomy site. Balanced salt solution was reinstilled at the end of the procedure, leaving only 15% to 50% of the vitreous cavity full of air. If the retinotomy was large, a nonexpansile concentration of sulfur hexafluoride gas was used occasionally.
The initial size of the lesion was measured using a transparency developed by the Macular Photocoagulation Study Group with circles enclosing areas ranging from 1.0 to 16.0 disc areas (1 disc area equals 1.77 mm2). Recurrent CNV (including persistence) was defined by angiographic criteria as new or persistent fluorescein leakage within or at the margin of the surgical scar. We (A.U. and M.A.T.) reviewed all fluorescein angiograms postoperatively. A χ2 test or Fisher exact test was used to compare the 2 groups for dichotomous variables. The mean visual acuity status was measured using the logarithm of the minimum angle of resolution (logMAR) scale.
Twenty-three consecutive patients (13 men and 10 women) underwent surgical removal of CNV. The patient age ranged from 19 to 77 years (median, 41 years). The median refraction was −10.00 D (range, −6.00 to −19.00 D). Fourteen patients (61%) had undergone laser photocoagulation for CNV, and surgery was performed for persistent or recurrent CNV. The preoperative visual acuity ranged from 20/40 to 20/400 (median, 20/80; mean, 20/100). Only 1 eye had a preoperative visual acuity of 20/40 or better. The median best postoperative visual acuity was 20/40 (mean, 20/60), with a median interval of 6 months (range, 1-36 months). With a mean follow-up of 24 months (range, 8-60 months), the final postoperative visual acuity ranged from 20/20 to 20/400 (median, 20/100; mean, 20/100). Of the 23 eyes, 8 (35%) had a final visual acuity of 20/40 or better; 5 (22%), between 20/50 and 20/100; and 10 (43%), between 20/200 and 20/400 (Table 2 and Figure 1). Overall, the visual acuity improved in 9 eyes (39%), decreased in 8 (35%), and remained unchanged in 6 (26%). Of 9 eyes followed up for 2 years or more, 4 (44%) had an improvement of visual acuity and 3 (33%) had a final visual acuity of 20/40 or better.
No intraoperative complications occurred. Postoperative complications included the development of nuclear sclerosis in 5 eyes (22%); transient elevated intraocular pressure in 2 (9%); and endophthalmitis, retinal detachment, and vitreous hemorrhage each in 1 (4%). Of 5 patients who developed postoperative nuclear cataract, 4 were older than 50 years. Postoperative recurrent CNV developed in 13 eyes (57%), with a median interval of 5 months (range, 1-18 months). Recurrent CNV was observed in 9 (64%) of 14 eyes with a history of laser treatment before surgery, while 4 (44%) of 9 eyes without a history of laser treatment developed recurrent CNV (P =.31). Of 11 eyes with a 2-line decrease of Snellen visual acuity from the best visual acuity to the final visual acuity, 10 (91%) had recurrent CNV.
Twenty-five consecutive patients (13 men and 12 women) underwent surgical removal of CNV. The patient age ranged from 17 to 62 years (median, 41 years). The median refraction was −7.75 D (range, −6.00 to −12.75 D). Eleven patients (44%) had undergone laser photocoagulation for CNV, and surgery was performed for persistent or recurrent CNV. The preoperative visual acuity ranged from 20/30 to 3/200 (median, 20/80; mean, 20/100). Three eyes (12%) had a preoperative visual acuity of 20/40 or better. The median best postoperative visual acuity was 20/30 (mean, 20/45), with a median interval of 3 months (range, 1-31 months). With a mean follow-up of 18 months (range, 6-44 months), the final postoperative visual acuity ranged from 20/20 to 3/200 (median, 20/50; mean, 20/70). Of the 25 eyes, 11 (44%) had a visual acuity of 20/40 or better; 7 (28%), between 20/50 and 20/100; 6 (24%), between 20/200 and 20/400; and 1 (4%), worse than 20/400 (Table 3). Overall, the visual acuity improved in 16 eyes (64%), was stable in 4 (16%), and deteriorated in 5 (20%). Of 8 eyes followed up for 2 years or more, 5 (63%) had an improvement of visual acuity and 4 (50%) had a final visual acuity of 20/40 or better.
The only intraoperative complication encountered was a full-thickness retinal break in the macula during subretinal manipulation in 1 eye because of the weakness of the retina over a prior photocoagulation scar. Postoperative complications included the development of nuclear sclerosis in 3 patients (12%) older than 49 years, transient intraocular pressure elevation in 2 (8%), and subcapsular cataract development in 1 (4%). We did not encounter retinal detachment during the follow-up period. Recurrent CNV was observed in 9 (36%) of the 25 eyes, with a median interval of 2 months from surgery (range, 0.5-12 months). Of 11 eyes with a history of laser treatment before surgery, 5 (45%) developed recurrent CNV, while 4 (29%) of 14 eyes without a history of laser treatment developed recurrent CNV (P =.32). Of 8 eyes with a 2-line decrease of Snellen acuity from the best visual acuity to the final visual acuity, 7 (88%) had recurrent CNV.
Several patient characteristics were evaluated for a statistically significant effect on an individual's chance for improvement of 2 or more Snellen lines or for a final visual acuity of 20/40 or better. These included age, sex, presence or absence of previous laser treatment, symptom duration, size of CNV, size of total lesion (including CNV, blood, laser scar, and/or blocked fluorescence), preoperative visual acuity, preoperative refraction, and presence or absence of postoperative recurrence of CNV. The statistically significant effects on visual acuity improvement of 2 lines or more were age 40 years or younger (P =.003) in group 1 and absence of recurrent CNV (P =.001) in group 2. Although not to a statistically significant degree, absence of recurrent CNV (P =.07) in group 1 and age 40 years or younger (P =.05) in group 2 were likely to improve visual acuity. Patients aged 40 years or younger were more likely to achieve a final visual acuity of 20/40 or better in group 1 than were patients older than 40 years (P =.06). The absence of recurrent CNV (P =.02) was also statistically significantly associated with a final visual acuity of 20/40 or better in group 2.
A 39-year-old white man was referred with an 8-month history of blurred central vision in his right eye. Preoperatively, the best-corrected visual acuity was 20/400 OD, with refraction of −11.50 D sphere. The results of a fundus examination revealed an obvious pigmented neovascular membrane extending nasally from an atrophic lesion through the center of the fovea (Figure 2). During vitrectomy, the neovascular complex was gently dislodged and was peeled up in 1 piece. The underlying RPE seemed to be attenuated or absent after the complex was extracted. His visual acuity improved to 20/60 OD at 12 months after surgery. Despite the presence of marked chorioretinal atrophy, the visual acuity remained at 20/80 OD after 60 months of follow-up, and there was no recurrent neovascularization (Figure 3).
A 49-year-old white man had an approximately 6-week history of reduced vision in the left eye. Preoperatively, the best-corrected visual acuity in the affected eye was 20/300 OS, with a refraction of −8.00 D sphere. Posteriorly, peripapillary atrophy and punched out peripheral lesions were present, consistent with POHS. A thin layer of blood surrounded the neovascular complex (Figure 4). Angiographically, subretinal hemorrhage blocked the border of the membrane (Figure 4). During vitrectomy, the membrane peeled back, leaving the subfoveal RPE intact, and the membrane was successfully extracted. He had recurrence of the neovascularization just temporal to the fovea. Because the recurrent membrane did not extend within the foveal avascular zone, the patient was treated successfully with laser photocoagulation. His visual acuity improved to 20/25 OS 3 months after surgery. The visual acuity has remained unchanged during 44 months of follow-up, and there was no additional neovascularization (Figure 5).
We divided patients with high myopia into 3 groups in an attempt to ascertain the influence of degenerative myopia or other presumed disorders on the visual results of subfoveal surgery. We occasionally see CNV in patients with high myopia, but it is often difficult to distinguish CNV caused by degenerative myopia itself from CNV caused by another reason. In this series, we identified patients with findings only of degenerative myopia as group 1, and if the patient had POHS as a probable cause in addition to myopia, we identified these patients as group 2. We assigned patients with typical findings corresponding to POHS to group 2 even if they had some minimum myopic change in the posterior pole, although none of the patients in group 2 had severe degenerative myopic changes such as diffuse or focal chorioretinal atrophy, posterior staphyloma, or lacquer cracks.
The clinical management of subfoveal neovascularization in patients with high myopia, especially typical in those with degenerative myopia, produced significant scar expansion after laser therapy, without a significant loss of visual acuity during the first year after treatment.17 The fact that the CNV in patients with degenerative myopia is likely to remain relatively small because of the attenuation of the choroid and the knowledge that the laser scar is likely to expand over time make laser photocoagulation treatment of CNV in patients with degenerative myopia of limited value.
Several previous reports14- 16 have documented visual results following surgical removal of CNV associated with high myopia. Of 5 eyes with myopic degeneration studied by Adelberg et al,15 significant improvement in visual acuity after surgery (to 20/80 or better) was seen in 2 eyes (40%), with a median follow-up of 12 months. Bottoni et al16 performed surgical removal of CNV in 21 eyes with high myopia; 10 (48%) improved, and 13 (62%) were at a visual acuity of 20/200 or better after a median follow-up of 12 months. These results appeared to suggest some advantage over laser photocoagulation or observation for restoration or preservation of visual function, although the number of patients and the follow-up period were limited. In our series, 15 (65%) of 23 eyes with only myopic degeneration experienced relatively stable or improved visual acuity at the final follow-up visit. In addition, the preoperative visual acuity was 20/40 or better in only 1 eye (4%) in group 1, but 8 eyes (35%) achieved a final visual acuity of 20/40 or better. It is difficult to compare these results with those of previous reports, but it seems that these results are better than those obtained with laser photocoagulation or natural history.
Visual outcome was better in patients in group 2 compared with those in group 1, although it is difficult to compare these groups simply because some of the preoperative findings are different. In patients with degenerative myopia, the choroid and the Bruch membrane around the macular lesion are diffusely thinned and abnormal. On the other hand, the patients in group 2 had POHS as the underlying disease. Presumed ocular histoplasmosis syndrome is a disease with focal atrophic chorioretinal scars in an otherwise normal fundus. These differences in pathologic features might explain the difference in visual outcome.18
The visual results in group 2 were comparable with those of our earlier work in patients with POHS-associated CNV.19 In that report, 40% of patients had improvement of 3 or more Snellen lines, and 35% had a postoperative visual acuity of 20/40 or better after surgical removal of POHS-associated CNV. It is gratifying that significant myopia in the eyes with POHS did not lead to visual outcomes that were any worse than those found in the previous series of surgery for POHS-associated CNV. This may imply that POHS was more responsible for the development of CNV in group 2 than myopia per se, and that high myopia does not demonstrably influence the visual results in POHS-associated CNV.
Significant postoperative complications were seen following surgical removal of CNV. Postoperative recurrent CNV developed in 57% (13/23) of eyes in group 1 and in 36% (9/25) of eyes in group 2. These recurrence rates in patients with degenerative myopia are higher than those given in previous reports.15,16 On the other hand, a high incidence (72%) of recurrent CNV after laser photocoagulation for CNV in eyes with pathologic myopia has been reported.11 These recurrence rates seem to be higher in patients with degenerative myopia than in patients with CNV secondary to other causes.19,20 Recurrence was one of the principal causes of limited visual improvement in both groups, especially in group 2. In addition, the visual acuity decreased by 2 lines or more from the best visual acuity level to the final visual acuity level in most of the eyes with recurrent CNV. Preventing recurrences with a pharmacological agent (if that were possible) might make the visual results better than described herein.
We encountered postoperative rhegmatogenous retinal detachment in only 1 of the 51 patients. The patient developed retinal detachment within 1 month after surgery and underwent retinal reattachment after scleral buckling surgery. The prevalence of abnormalities of the vitreoretinal interface, such as lattice degeneration, increases in eyes with myopia.21 Careful scleral depression intraoperatively with treatment of the pathologic characteristics is also mandatory.
Multiple preoperative, intraoperative, and postoperative factors impact the visual result following surgery for CNV in patients with high myopia, although our number of patients is too small to assess statistical significance. The presence of recurrent CNV and a patient age of 40 years or older were negatively associated factors in visual improvement. Regardless of the underlying cause, age may be an important factor in the growth pattern of CNV. Gass18 supposed that the choroidal neovascular membrane in older patients may develop in the subpigment epithelial space rather than in the subsensory retinal space. Berger et al20 reported that younger patient age is one of the factors for better visual outcome in the results of subretinal surgery for POHS. These factors should be considered when counseling patients with CNV associated with high myopia.
There are several limitations of this study, some of which are inherent to a retrospective review of surgical cases, including lack of a control group, variable patient follow-up, and the use of best-corrected Snellen visual acuity (rather than best-corrected Early Treatment Diabetic Retinopathy Study visual acuity, with which low visual acuity levels are measured well). Some follow-up visual acuity results and other data were obtained from referring ophthalmologists. These visual acuities may or may not represent the best possible corrected acuities, as were obtained for patients seen in our office. In addition, the follow-up in this series is relatively short (range, 6-60 months). A longer follow-up may lead to deterioration of visual function in these patients. Highly myopic eyes may develop atrophy of the RPE and choroid with or without laser photocoagulation. It is unknown whether subretinal surgery may precipitate or accelerate this pathologic process. Delayed loss of RPE and choriocapillaris has been reported following subretinal surgery for CNV associated with POHS.22 Furthermore, the high rate of recurrent neovascularization and its association with a worse visual outcome seen in this series also raise serious concern regarding longer-term outcomes.
We surmise that the visual outcomes in most patients in this series are either better or no worse than if the subfoveal CNV had been photocoagulated. We cannot claim superiority of this therapy over observation without conducting a randomized, prospective trial. Given these modest short-term results, a randomized, prospective trial appears reasonable to pursue. However, in the general population, the incidence of subfoveal CNV secondary to high myopia is low enough that recruitment for a randomized, prospective trial would be problematic.
Accepted for publication October 29, 1999.
Reprints: Matthew A. Thomas, MD, Barnes Retina Institute, One Barnes Hospital Plaza, Suite 17413 East Pavilion, St Louis, MO 63110 (e-mail: firstname.lastname@example.org).