INTRAVITREAL BEVACIZUMAB IN THE TREATMENT OF CHOROIDAL NEOVASCULAR MEMBRANE DUE TO ANGIOID STREAKS CENGIZ ALAGÖZ, MD,* NESxE ALAGÖZ, MD,* ABDULLAH ÖZKAYA, MD,* UGUR ÇELIK, MD,† MIRAY F. TURAN, MD,* AHMET T. YAZICI, MD,* OSMAN ÇEKIÇ, MD, PHD,‡ AHMET DEMIROK, MD* Purpose: To investigate the results of intravitreal bevacizumab for choroidal neovascularization (CNV) secondary to angioid streaks and to assess the factors influencing disease progression. Methods: Patients treated with intravitreal bevacizumab (1.25 mg/0.05 mL) for CNV secondary to angioid streaks were reviewed retrospectively. In addition to demographic findings, ophthalmologic findings at baseline and during follow-ups were recorded. Results: Twenty-three eyes of 20 patients (mean age, 45.7 years; 7 women) were included in the study. Mean follow-up was 23 months. Mean number of injections was 5.1. Initial and final logMAR visual acuity was not different (0.53 ± 0.33 and 0.60 ± 0.40 logMAR, P = 0.79). At the last examination, patients with final active CNV (N = 14) were younger (mean age, 42 years) than patients with final inactive CNV (N = 9) (mean age, 52 years). The former group required higher number of injections (6.6 vs. 2.8, P = 0.013). Eyes with pseudoxanthoma elasticum (N = 10) needed injections every 4.4 months while the others (N = 13) every 7.2 months (P = 0.072). Pseudoxanthoma elasticum positivity ended up with active membranes in 70% of the cases, composing half of the overall final active CNVs in this study. Conclusion: Intravitreal bevacizumab stabilized vision in eyes with CNV and angioid streaks. At younger ages, CNV behaved more aggressively and seemed to be more resistant to treatment. RETINA 35:2001–2010, 2015
A
ngioid streaks were first defined by Doyne1 in 1889 as small breaks in Bruch membrane. In fundus examination, these are observed as irregular gray to reddish brown lines that extend outward from the optic disk. Histologically, it has been shown that patients with angioid streaks have an abnormally thickened and calcified Bruch membrane, which is thought to be prone to developing breaks in response to intrinsic or extrinsic mechanical stress such as that resulting
from the ocular muscles and trauma.2 Angioid streaks may develop either idiopathically or in association with a number of systemic diseases, such as pseudoxanthoma elasticum (PXE), Paget disease, Ehlers–Danlos syndrome, and hemoglobinopathies.3 Among these, PXE is the most common systemic disease encountered in association with angioid streaks, reported at a rate of 59% to 86% in large series.4 The most important complication threatening vision in eyes with angioid streaks is the development of choroidal neovascularization (CNV). Between 72% and 86% of patients develop CNV, whereas 71% of these patients also develop CNV in the contralateral eye.3,5 Because the disease generally occurs at a younger age compared with agerelated CNV, severe loss in visual acuity results in an economic burden because of the visual disability of this relatively younger population.3
From the *Beyoglu Eye Research and Training Hospital, Istanbul, Turkey; †Gaziosmanpasxa Taksim Training and Research Hospital, Istanbul, Turkey; and ‡Department of Ophthalmology, Marmara University Medical School, Istanbul, Turkey. Paper partly presented at the 14th ESASO Retina Academy meeting, Istanbul, Turkey, November 13–15, 2014. None of the authors have any financial/conflicting interests to disclose. Reprint requests: Cengiz Alagöz, MD, Beyo glu Eye Research and Training Hospital, Bereketzade Camii sok, Kuledibi, Beyo glu, Istanbul 34000, Turkey; e-mail: [email protected]
2001
2002 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
The break in the Bruch membrane itself has no effect on visual function.6,7 However, with disease progression, degenerative changes in the retina pigment epithelium and overlying neurosensory retina may occur, decreasing the visual acuity. Even without CNV progression, there may be low visual acuity in eyes with angioid streaks involving the fovea, which makes the visual acuity gain as the treatment goal impractical in some of the cases. Also for the same reason, in the decision to retreat, anatomical changes such as leakage on fundus fluorescein angiography and intra/subretinal fluid observed by optical coherence tomography (OCT) might be referred other than the visual function. Various methods such as laser photocoagulation and photodynamic therapy with verteporfin have been used in the treatment of CNV due to angioid streaks. However, these treatment methods did not yield the desired success rate because of high recurrence rates, atrophic changes, and ongoing growth of CNV.5,8–10 As in age-related CNV and CNV secondary to high myopia, various intravitreal, antivascular endothelial growth factor (anti-VEGF) agents have been shown to be successful in the treatment of CNV due to angioid streaks.11–22 Yet, there is no optimal treatment protocol for this group of patients, and the risk factors for the disease progression could not be clearly identified. In this study, we investigated the long-term results of intravitreal bevacizumab injection in eyes with CNV secondary to angioid streaks and assessed the factors influencing disease progression.
Materials and Methods All patients who were treated with intravitreal bevacizumab (1.25 mg/0.05 mL; Altuzan, Roche, Mannheim, Germany) for CNV secondary to angioid streaks in Beyo glu Eye Research and Training Hospital between July 2009 and March 2014 were investigated retrospectively using patient records. The principles of the study were in accordance with the Declaration of Helsinki, and a written informed consent was obtained from every patient before every injection. Subjects with corrected baseline visual acuity of ,20/ 200 were excluded, as were subjects who were followed up for less than 6 months, and those with any other disease that can compromise visual acuity. Bevacizumab injection was not applied to patients with uncontrolled systemic hypertension or with a history of thromboembolism or cerebrovascular accident.
2015 VOLUME 35 NUMBER 10
In addition to demographic findings, ophthalmologic examination findings of all patients before the injection and during follow-ups were recorded from their files. For statistical calculations, visual acuity values were converted from the Early Treatment Diabetic Retinopathy Study chart to logarithm of minimal angle of resolution (logMAR). Biomicroscopic examination findings, Goldmann applanation tonometry measurements, and findings of dilated fundus examination performed with 90-diopter lens were recorded. Optical coherence tomography scans (either Stratus OCT; Carl Zeiss Meditec Inc, Dublin, CA, or Spectralis HRA-OCT; Heidelberg Engineering, Heidelberg, Germany) of all subjects before the injection and at the last follow-up were analyzed. Initial and final fundus fluorescein angiography (either Zeiss FF450; Zeiss, Oberkochen, Germany, or Spectralis HRA-OCT) images were examined. Localization of CNV was determined by angiographic images before the injection. Initially, intravitreal bevacizumab injection was applied on subjects either with newly diagnosed CNV or on subjects with an old CNV scar demonstrating activity; further treatment was administered on a pro re nata regimen. The patients received a followup 4 to 5 weeks after the injections and then at regular intervals at the discretion of the treating retina specialist. In this study, intravitreal bevacizumab was administered as monotherapy, where the combination of intravitreal bevacizumab with any other treatment modality on the same or different session was not performed. However, a previous treatment history of photodynamic therapy (N = 2) was not an exclusion criterion because the photodynamic therapy was applied at least 3 years before the first injection. The decision to retreat was made if at least one of the following signs was present: occurrence of a new intraretinal or subretinal macular hemorrhage from CNV as detected on fundoscopic examination, leakage on angiography, or presence of intraretinal or subretinal fluid on OCT. When any sign of membrane activity, such as increasing metamorphopsia, new or persistent leakage on fluorescein angiography, retinal hemorrhage, or edema documented by OCT was detected, intravitreal injection was repeated regardless of the visual acuity level. Figure 1 shows a series of OCT images of a patient with CNV due to angioid streaks that showed multiple recurrences and received 20 injections over a period of 40 months. Primary end points of the study were stabilization or improvement in visual acuity between baseline and the last follow-up, and final activity of CNV. Stabilization of visual acuity was considered as visual acuity change of 2 ETDRS lines and improvement as gain in visual
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2003
Fig. 1. A series of OCT images of a 41-year-old male patient with PXE (Patient 8). At baseline, the patient had a juxtafoveal CNV that demonstrated multiple reactivation episodes and increased in size despite the frequent injections of bevacizumab during the follow-up. Both subretinal and intraretinal fluids are observed on baseline time domain OCT (A). At 14 months of follow-up, after receiving 7 injections, the lesion did not demonstrate any activity; no fluid was observed on the time domain OCT image (B). After 30 months of follow-up, the patient received a total of 13 injections and the lesion extended subfoveally; arrows show subretinal fluid indicating activity; the arrowhead shows intraretinal cysts located over the fibrous scar; and the asterisks show outer retinal tubulation (ORT) (C). The patient received another 3 injections and at 34 months of follow-up, only the intraretinal cysts (arrowhead) and ORT (asterisks) were observed on the OCT image; no additional injection was administered because those findings did not change in response to treatment and therefore were not regarded as an activity sign (D). At the next follow-up (1 month later), new subretinal fluid collection appeared at the edges of the lesion (arrow) (E-1 and E-2) and the patient received further injections. The OCT images of the same patients at the last follow-up (40 months) receiving a total of 20 injections; the lesion activity is present as observed by the subretinal fluid (arrows, F-1 and F-2). Note that the intraretinal cysts (arrowhead) and ORTs (asterisks) appeared almost similar at different time points of follow-up (C, D, E-1, and F-1). Images (E-2 and F-2) are from different sections (superior to the fovea) of the same OCT imaging as the images (E-1 and E-2), respectively.
acuity of 2 ETDRS lines. The cases with new intraretinal or subretinal macular hemorrhage from CNV and/ or leakage on fluorescein angiography and/or the presence of intraretinal or subretinal fluid on OCT at the last follow-up were considered to have final active CNV. The secondary end points of the study were the factors affecting the disease progression and complications of the treatment. Further analysis was performed for the following conditions: the eyes with final active membrane (N = 14) were compared with the eyes with final inactive membrane (N = 9), and the subjects with PXE (N = 10) were compared with the others (N = 13) for age, follow-up period, number of injections, initial and final visual acuity, and visual acuity change. For statistical analysis, SPSS 20.0 for Windows was used. Continuous variables were presented as mean ± standard deviation. Frequency and percentage (%) were used for categorical variables. For quantitative variables, the Wilcoxon test was used for the comparison of two
related groups, whereas the Mann-Whitney U test was used for the comparison of two independent groups. The Fisher’s exact test was used for categorical variables. P , 0.05 was considered to be statistically significant. Results No ocular or systemic complication was encountered in the follow-up period related to the intravitreal bevacizumab. Twenty-three eyes of 20 patients (7 female, 13 male) were included in the study. The mean (±standard deviation) age of patients was 45.7 ± 9.1 (range, 29–65) years, and the follow-up after the treatment was 23.2 ± 16.2 (range, 6–58) months. The average number of injections performed during follow-up was 5.1 ± 4.3 (range, 1–20). Two eyes had a history of ocular photodynamic therapy (one eye 3 times and the other eye only once) dating back at least 3 years before the start of this study. Localization of CNV due to angioid streaks was subfoveal in
2004 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
15 eyes (65.2%) and juxtafoveal in 8 eyes (34.8%) on initial examination. All patients with a diagnosis of angioid streaks were referred to the dermatology department; and 10 patients (43.5%) received a biopsy-confirmed diagnosis of PXE, whereas 1 patient (4.3%) had Ehlers–Danlos syndrome. The mean logMAR visual acuity of patients was 0.53 ± 0.33 (20/63, Snellen equivalent) at baseline and 0.60 ± 0.40 (20/80) on the last follow-up (P = 0.79). On the last examination, visual acuity was stabilized in 7 eyes (30.4%), improved in 8 eyes (34.8%), and worsened in 8 eyes (34.8%), compared with the baseline. The mean change in VA was −0.65 ± 5.01 (range: −10 to +9) ETDRS lines. An active membrane was present in 14 of 23 eyes (61%) on the last follow-up. When the eyes with a final active CNV (N = 14) were compared with those with a final inactive CNV (N = 9), the mean age was 41.8 ± 6.2 years in the first group and 51.8 ± 9.7 years in the latter group (P = 0.007). The average number of injections was 6.6 ± 4.7 and 2.8 ± 2.3, respectively (P = 0.013). There was no significant difference in the duration of follow-up between the 2 groups (P = 0.829). The mean interval between the injections (follow-up time/injection number) was 4.1 ± 2.4 and 9.0 ± 4.8 in the groups, respectively (P = 0.001). Both the baseline and final visual acuity values were found to be similar between the groups (P = 0.179, P = 0.926, respectively) (Table 1). Figure 2 represents an example of patients with final active CNV, and (Figures 3 and 4) of patients with final inactive CNV. Nine of 14 eyes (64.3%) with final active CNV had a subfoveal CNV on initial examination, and 5 (35.7%) had a juxtafoveal CNV initially. Three eyes with initial juxtafoveal CNV demonstrated subfoveal extension at the final follow-up (Figures 1 and 2). In subjects with final inactive CNV, 6 of 9 eyes (66.6%) had a subfoveal CNV and 3 eyes (33.3%) had a juxtafoveal CNV initially. None of the eyes with juxtafoveal membrane in this group of patients showed involvement of the fovea during the follow-up (Figure 4).
2015 VOLUME 35 NUMBER 10
Characteristics of all cases were summarized in Table 2. Eleven of 12 eyes (91.7%) of patients who were younger than 46 years (the mean age of subjects in the study population) had signs of membrane activity on the last examination, whereas activity rate of eyes of the patients older than 46 years was 27.3% (3 of 11 eyes) (P = 0.003). Taking into account the patients with PXE (10 eyes of 10 patients), the mean age at presentation was 43.5 ± 10.1 years, the follow-up was 28.3 ± 18.2 months, and the average number of injections was 7.5 ± 5.5. The mean (follow-up time/injection number) ratio was 4.4 ± 2.5. The baseline logMAR visual acuity was 0.53 ± 0.31 (20/63), and the final visual acuity was 0.72 ± 0.37 (20/100) (P = 0.352). Final membrane activity was present in 7 of 10 eyes (70%) in this group. The patients without PXE (13 eyes of 10 patients) had a mean age of 47.4 ± 8.1 years. The mean follow-up was 19.4 ± 14.1 months, and the mean number of injections was 3.2 ± 1.9. The mean (follow-up time/injection number) ratio was 7.3 ± 4.9. The baseline visual acuity was 0.53 ± 0.37 (20/63), and the final visual acuity was 0.50 ± 0.42 (20/63) (P = 0.552). Final membrane activity was present in 7 of 13 eyes (53.8%). Although the difference in the follow-up between the groups with and without PXE was nonsignificant, the injection number was higher in the PXE group (P = 0.018). There seemed to be a difference between the follow-up/injection number ratio (4.4 vs. 7.2, for subjects with and without PXE, respectively); however, the statistics could not reveal any significance (P = 0.072) (Table 3).
Discussion In this study, the mean follow-up time of subjects with CNV due to angioid streaks was 23 months, and the mean number of repeated intravitreal bevacizumab
Table 1. Comparison of Eyes With Final Active CNV (N = 14) to Those With Final Inactive CNV (N = 9)
Age, years Membrane Localization Number of injections Follow-up, months Follow-up/injection, months Initial VA, logMAR Final VA, logMAR
Subfoveal Juxtafoveal
Data were expressed as mean ± standard deviation. *Mann–Whitney U test. VA, visual acuity.
Final Active CNV (N = 14)
Final Inactive CNV (N = 9)
P*
41.8 ± 6.2 9 (64.3%) 5 (35.7%) 6.6 ± 4.7 24.6 ± 16.7 4.05 ± 2.4 0.46 ± 0.34 0.61 ± 0.42
51.8 ± 9.7 6 (66.6%) 3 (33.3%) 2.8 ± 2.3 21.1 ± 16.2 9.01 ± 4.8 0.65 ± 0.31 0.58 ± 0.40
0.007 1.00 0.013 0.829 0.001 0.179 0.926
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2005
Fig. 2. Images obtained from a 29-year-old male patient (Patient 1) with PXE. Color fundus picture at baseline shows cracks radially extending from the optic disk (A). Note the juxtafoveally localized choroidal neovascular membrane and multiple subretinal hemorrhages surrounding the membrane. Intraretinal and minimal subretinal fluid is observed on OCT image (B). One month after the first injection of intravitreal bevacizumab, the lesion and hemorrhage regressed as seen on the color fundus picture (C). The corresponding FA did not show leakage from the lesion (D), and fluid is not observed on OCT examination (E). After a long quiescent period (20 months), the lesion reactivated (F–H). Color fundus image (F), leakage on the late frames of FA (G), and intraretinal and subretinal fluid on OCT (H) indicating the activity of the membrane. After 12 injections (47 months), the membrane still demonstrated activity (I–K); color fundus photograph (I), FA (J), and OCT (K) images of the patient at the last follow-up. FA, fluorescein angiography.
injections needed during follow-up was 5. The final visual acuity showed no significant change when compared with the baseline. The mean change in VA was −0.65 ± 5.01 (range: −10 to +9) ETDRS lines. Sixtyfive percent of eyes with CNV due to angioid streaks demonstrated an improvement or stabilization in visual acuity, whereas visual acuity worsened in 35% of eyes. With respect to visual acuity improvement, various studies using anti-VEGF agents to treat CNV secondary to angioid streaks yielded similar results.12–22 In a series of 35 subjects, ranibizumab injections were performed and the subjects were followed up for a mean of 24 months. The same study reported either an improvement or a stabilization (±3 lines of change) in visual acuity in 85.7% of subjects and a mean change in visual acuity of −0.17 (−10 to +7) lines.18 In the study conducted by Battaglia Parodi et al,19 intravitreal bevacizumab was
performed on 15 eyes and the subjects were followed up for 1 year. The study reported a 33% increase and 46.6% stabilization in visual acuity, whereas the mean number of injections was 2.5 and the mean initial visual acuity was preserved at the same level throughout the study period. In a study by Çekiç et al,13 intravitreal pegaptanib sodium was used in 5 eyes in repeated doses. After 15 months to 24 months of follow-up, 2 of the subjects had an improvement in visual acuity, whereas the remaining 3 were reported to have a stable visual acuity throughout the study. Our study reflects the treatment results achieved by an initial dose of intravitreal bevacizumab followed by repetitive doses as membrane activity was detected. However, a series where monthly ranibizumab was injected for 12 months regardless of membrane activity reported quite successful short-term results.20
2006 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 10
Fig. 3. Images obtained from a 56-year-old female patient (Patient 17) with PXE. In addition to an area showing retina pigment epithelium changes and multiple tiny hemorrhages in the macula, angioid streaks are clearly visible on the initial fundus photograph image (A). Initial angiographic image showing fluorescein blockage because of both retina pigment epithelium hypertrophy and subretinal hemorrhage, leakage from the active choroidal neovascular membrane, and a cystoid macular edema secondary to the membrane (B). The intraretinal fluid is visible on initial OCT (C). The patient received 3 intravitreal bevacizumab injections during a 26-month follow-up period. The final fundus photograph image shows the areas of retina pigment epithelium changes (atrophy and hypertrophy) and a scar involving the macula (D). Hyperflourescence because of staining of the scar tissue without leakage is noted on late frames of the final fluorescein angiography (E). Atrophy of the neurosensory retina overlying the subretinal scar is seen on final OCT of the patient (F). Neither subretinal nor intraretinal fluid was present at final visit.
In this prospective study of 7 eyes, visual acuity increased from 20/63 to 20/32 at the 12th month and was preserved at the same level for 3 months after the last injection. However, long-term results of frequent anti-VEGF treatment remain unclear. There is a doubt that the frequent use of anti-VEGF agents in patients with angioid streaks may accelerate the atrophic changes in the retinal pigment epithelium and neurosensory retinal layers because the drug can readily reach the choriocapillaris through the breaks in the Bruch membrane.22,23 Finger et al22 reported an increase in retina pigment epithelial atrophy and
subretinal fibrosis in fundus autofluorescence imaging of patients with angioid streaks, despite the frequent anti-VEGF treatment. In this study, we assessed the membrane activity of the subjects at the last follow-up and observed that activity tends to persist in patients of younger age despite repetitive injections. Ninety-two percent of patients who were 46 years of age or younger had an active CNV on the last examination, whereas the activity rate among the patients older than 46 years was 27%. Likewise, the mean age of subjects who had signs of membrane activity at the last follow-up was
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2007
Fig. 4. Juxtafoveal CNV secondary to angioid streaks in a 49-year-old female patient without PXE (Patient 14) (A–C). Fundus photograph (A); fluorescein angiography at late frames demonstrating leakage from CNV (B); subretinal fluid seen at the border of the lesion (arrow, C). The lesion was inactive after one injection of bevacizumab. A series of OCT images from the same patient during the follow-up of 19 months (D at 1 month, E at 6 months, and F at 19 months).
significantly lower than the remaining subjects. In addition, the mean number of injections was higher in the group of patients with final active CNV. Although the duration of follow-up periods was similar between the 2 groups, the higher number of injections in patients with final active CNV and the tendency of the juxtafoveal lesion to involve the fovea suggest that the disease progression is more aggressive in this group of patients (Figures 1 and 2). The most frequent systemic association in angioid streaks is PXE.4 Ten of 20 patients (50%) (10 of 23 eyes) had PXE in our series. In PXE, because of the absence of a systemic antimineralization factor, calcification and fragmentation of the elastic fibers in a number of organs, including the skin, eye, and the blood vessels, occurs. Involvement of the blood vessels, besides being a risk factor for cardiovascular diseases such as acute myocardial infarction and cerebrovascular occlusion,24,25 also may constitute a risk factor for abnormal wound healing. Viljoen et al26 reported tissue friability, poor wound healing, and keloid formation after cosmetic surgery in patients with PXE. In a previous report by Mimoun et al,18 no difference was found between the treatment results of subjects with and without PXE. This study drew attention to the PXE group that seemed to require a higher number of injections. Although nonsignificant, there was a tendency for frequent injections in the treatment of subjects with PXE, which is demonstrated by the shorter treatment intervals in this group. Our study demonstrated that the PXE group required injections
in every 4 months in contrast to every 7 months for the non-PXE group. This may be an important finding for the clinical practice. Because of an increased cardiovascular risk profile in PXE patients and the anti-VEGF treatment itself possessing risk for systemic complications such as cerebrovascular occlusion, the treatment should be considered meticulously and followed up carefully.27,28 Acute cerebrovascular accident after bevacizumab injection was reported in a 54-year-old PXE patient with no other systemic disease.29 In our series, no ocular or systemic complication because of bevacizumab injection was encountered. Because of high recurrence rates and rapid progression, CNV due to angioid streaks has a poorer prognosis when compared with other etiologies.3 The disease may show multiple recurrences after inactive periods or a new CNV in a different area may develop.30 Intravitreal bevacizumab has a short activity period, and repetitive injections are needed. Therefore, patients should be monitored frequently and informed both of the possibility of recurrence and of an increased cardiovascular risk after the treatment. The limitations of the study include the retrospective study design. Apart from this, the series is small because of the rarity of the disease and includes patients with different etiologies for angioid streaks, such as PXE, Ehler-Danlos syndrome, and idiopathic ones. Among the subjects, only 1 patient (Patient 1) did not attend his scheduled follow-ups for a period of 20 months in the quiescent period after the first
Patient No.
M F M M M M M M M M M M M F M F F F F M M M F
L L L L R R L L R R L L L L R R L R L L R L R
PXE PXE PXE — — — — PXE PXE — — PXE — EDS PXE — — PXE PXE PXE — — —
Juxtafoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal
Final Foveal Involvement of CNV*
Previous Treatment
+ − − + PDT + − − −
·3 PDT
Number of Injections
Baseline VA, Final VA, logMAR logMAR
12 1 2 4 5 1 3 9 20 6 2 8 3 6 7 1 5 5 3 7 2 1 1
*Foveal involvement for the membranes initially localized juxtafoveally. EDS, Ehlers–Danlos syndrome; F, female; L, left eye; M, male; PDT, photodynamic therapy; R, right eye; VA, visual acuity.
0.30 0.22 0.15 1.00 1.00 0.40 0.15 0.52 0.15 0.30 0.05 0.80 0.70 0.52 0.70 1.00 0.10 0.80 0.70 1.00 0.40 0.30 1.00
1.30 0.22 0.15 0.70 0.40 0.15 0.40 0.80 1.00 1.30 0.05 0.40 0.40 1.30 0.70 0.15 0.22 1.00 1.00 0.70 0.15 0.52 0.80
Follow-up, months
Final Membrane Activity
47 6 6 9 29 10 10 33 40 30 8 31 14 33 12 19 56 19 26 58 10 10 14
+ − + + + + + + + + + + − − + − + + − − − − −
2015 VOLUME 35 NUMBER 10
29 31 35 39 39 40 40 40 41 43 43 44 47 47 48 49 50 54 56 57 57 57 65
Baseline CNV Localization
1 2 3 4 5 6 6 7 8 9 9 10 11 12 13 14 15 16 17 18 19 19 20
Age, Systemic years Gender Eye Disease
2008 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
Table 2. Clinical Features of All Cases Affected by CNV Due to Angioid Streaks Treated by Intravitreal Bevacizumab
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2009
Table 3. Comparison of the Eyes of Subjects With and Without PXE
Age, years Number of injections Follow-up, months Follow-up/injection, months Initial VA, logMAR Final VA, logMAR Final active CNV
Eyes With PXE (N = 10)
Eyes Without PXE (N = 13)
P*
43.5 ± 10.1 7.5 ± 5.5 28.3 ± 18.2 4.39 ± 2.5 0.53 ± 0.31 0.72 ± 0.37 7 (70.0%)
47.4 ± 8.1 3.2 ± 1.9 19.4 ± 14.1 7.28 ± 4.9 0.53 ± 0.37 0.50 ± 0.42 7 (53.8%)
0.446 0.018 0.284 0.072 1.000 0.166 0.669
Data were expressed as mean ± standard deviation. *Mann–Whitney U test. VA, visual acuity.
injection, and he presented as soon as he noticed the first symptoms of disease reactivation. He was regularly followed up then. In conclusion, intravitreal bevacizumab stabilized visual acuity in eyes with CNV due to angioid streaks with a mean follow-up of 23 months. In younger ages, CNV behaved more aggressively and tended to be more resistant to the treatment. Eyes with PXE contributed to half of the overall final active CNVs in the study. Although we could not find a difference in the treatment intervals between eyes with and without PXE (P = 0.07), an increased number of the study population might reach the statistical significance. Prospective, randomized long-term studies are warranted to determine the optimal frequency of anti-VEGF treatment in CNV due to angioid streaks and to justify the use of intensive anti-VEGF treatment, especially in patients who are expected to show aggressive progression. Key words: angioid streaks, choroidal neovascular membrane, intravitreal bevacizumab, pseudoxanthoma elasticum. References 1. Doyne RW. Choroidal and retinal changes. The results of blows on the eyes. Trans Ophthalmol Soc U K 1889;9:128. 2. Adelung JC. Genesis of angioid streaks (Knapp). Klin Monbl Augenheilkd Augenarztl Fortbild 1951;119:241–250. 3. Clarkson JG, Altman RD. Angioid streaks. Surv Ophthalmol 1982;26:235–246. 4. Connor PJ Jr., Juergens JL. Pseudoxanthoma elasticum and angioid streaks. A review of 106 cases. Am J Med 1961;30: 537–543. 5. Pece A, Avanza P, Galli L, Brancato R. Laser photocoagulation of choroidal neovascularization in angioid streaks. Retina 1997;17:12–16. 6. Charbel Issa P, Finger RP, Holz FG, Scholl HP. Multimodal imaging including spectral domain OCT and confocal near infrared reflectance for characterization of outer retinal pathology in pseudoxanthoma elasticum. Invest Ophthalmol Vis Sci 2009;50:5913–5918. 7. Dreyer R, Green WR. The pathology of angioid streaks: a study of twenty-one cases. Trans Pa Acad Ophthalmol Otolaryngol 1978;31:158–167.
8. Singerman LJ, Hatem G. Laser treatment of choroidal neovascular membranes in angioid streaks. Retina 1981;1:75–83. 9. Chung AK, Gauba V, Ghanchi FD. Photodynamic therapy (PDT) using verteporfin for juxtafoveal choroidal neovalkoscularization (CNV) in angioid streaks (AS) associated with pseudoxanthoma elasticum: 40 months results. Eye (Lond) 2006; 20:629–631. 10. Browning AC, Chung AKK, Ghanchi F, et al; United Kingdom PDT Users Group. Verteporfin photodynamic therapy of choroidal neovascularization in angioid streaks: one year results of a prospective case series. Ophthalmology 2005;112: 1227–1231. 11. Yilmaz I, Ozkaya A, Alkin Z, Yazici AT. Intravitreal ranibizumab for bilateral choroidal neovascularisation in a patient with angioid streaks. BMJ Case Rep 2014;29:2014. 12. Finger RP, Charbel Issa P, Ladewig M, et al. Intravitreal bevacizumab for choroidal neovascularization associated with pseudoxanthoma elasticum. Br J Ophthalmol 2008;92:483–487. 13. Çekiç O, Göçmez E, Kocabora MS. Management of CNV in angioid streaks by intravitreal use of specific anti-VEGF165 Aptamer (pegaptanip sodium): long-term results. Curr Eye Res 2011;36:492–495. 14. Kang S, Roh YJ. Intravitreal ranibizumab for choroidal neovascularisation secondary to angioid streaks. Eye (Lond) 2009; 23:1750–1751. 15. Molina Guilabert I, Calvo-Gonzalez C, Reche-Frutos J, et al. Intravitreal pegabtanib sodium in choroidal neovascularization secondary to angioid streaks. Acta Ophthalmol 2009;87: 581–582. 16. Schiano Lomoriello D, Parravano MC, Chiaravalloti A, Varano M. Choroidal neovascularization in angioid streaks and pseudoxanthoma elasticum: 1 year follow-up. Eur J Ophthalmol 2009;19:151–153. 17. Wiegand TW, Rogers AH, McCabe F, et al. Intravitreal bevacizumab (Avastin) treatment of choroidal neovascularisation in patients with angioid streaks. Br J Ophthalmol 2009;93:47–51. 18. Mimoun G, Tilleul J, Leys A, et al. Intravitreal ranibizumab for choroidal neovascularization in angioid streaks. Am J Ophthalmol 2010;150:692–700. 19. Battaglia Parodi M, Lacono P, LaSpina C, et al. Intravitreal bevacizumab for nonsubfoveal choroidal neovascularization associated with angioid streaks. Am J Ophthalmol 2014;157: 374–377. 20. Finger RP, Charbel Issa P, Hendig D, et al. Monthly ranibizumab for choroidal neovascularizations secondary to angioid streaks in pseudoxanthoma elasticum: a one-year prospective study. Am J Ophthalmol 2011;152:695–703.
2010 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 21. Gliem M, Finger RP, Fimmers R, et al. Treatment of choroidal neovascularization due to angioid streaks: a comprehensive review. Retina 2013;33:1300–1314. 22. Finger RP, Charbel Issa P, Schmitz-Valckenberg S, et al. Long-term effectiveness of intravitreal bevacizumab for choroidal neovascularization secondary to angioid streaks in pseudoxanthoma elasticum. Retina 2011;31:1268–1278. 23. Gibran SK, Sachdev A, Stappler T, et al. Histological findings of a choroidal neovascular membrane removed at the time of macular translocation in a patient previously treated with intravitreal bevacizumab treatment (Avastin). Br J Ophthalmol 2007;91:602–604. 24. Combrinck M, Gilbert JD, Byard RW. Pseudoxanthoma elasticum and sudden death. J Forensic Sci 2011;56:418–422. 25. Bock A, Schwegler G. Intracerebral haemorrhage as first manifestation of pseudoxanthoma elasticum. Clin Neurol Neurosurg 2008;110:262–264.
2015 VOLUME 35 NUMBER 10
26. Viljoen DL, Bloch C, Beighton P. Plastic surgery in pseudoxanthoma elasticum: experience in nine patients. Plast Reconstr Surg 1990;85:233–238. 27. Wong TY, Liew G, Mitchell P. Clinical update: new treatments for age-related macular degeneration. Lancet 2007;370:204–206. 28. Van den Berg JS, Hennekam RC, Cruysberg JR, et al. Prevalence of symptomatic intracranial aneurysm and ischaemic stroke in pseudoxanthoma elasticum. Cerebrovasc Dis 2000;10:315–319. 29. Besozzi G, Ferrara A, Epifani E, et al. Acute stroke after intravitreal bevacizumab to treat choroidal neovascularization due to angioid streaks in pseudoxanthoma elasticum: a severe systemic adverse event after an off-label procedure. Int Ophthalmol 2013;33:181–183. 30. Sawa M, Gomi F, Tsujikawa M, et al. Long-term results of intravitreal bevacizumab injection for choroidal neovascularization secondary to angioid streaks. Am J Ophthalmol 2009; 148:584–590.
A
ngioid streaks were first defined by Doyne1 in 1889 as small breaks in Bruch membrane. In fundus examination, these are observed as irregular gray to reddish brown lines that extend outward from the optic disk. Histologically, it has been shown that patients with angioid streaks have an abnormally thickened and calcified Bruch membrane, which is thought to be prone to developing breaks in response to intrinsic or extrinsic mechanical stress such as that resulting
from the ocular muscles and trauma.2 Angioid streaks may develop either idiopathically or in association with a number of systemic diseases, such as pseudoxanthoma elasticum (PXE), Paget disease, Ehlers–Danlos syndrome, and hemoglobinopathies.3 Among these, PXE is the most common systemic disease encountered in association with angioid streaks, reported at a rate of 59% to 86% in large series.4 The most important complication threatening vision in eyes with angioid streaks is the development of choroidal neovascularization (CNV). Between 72% and 86% of patients develop CNV, whereas 71% of these patients also develop CNV in the contralateral eye.3,5 Because the disease generally occurs at a younger age compared with agerelated CNV, severe loss in visual acuity results in an economic burden because of the visual disability of this relatively younger population.3
From the *Beyoglu Eye Research and Training Hospital, Istanbul, Turkey; †Gaziosmanpasxa Taksim Training and Research Hospital, Istanbul, Turkey; and ‡Department of Ophthalmology, Marmara University Medical School, Istanbul, Turkey. Paper partly presented at the 14th ESASO Retina Academy meeting, Istanbul, Turkey, November 13–15, 2014. None of the authors have any financial/conflicting interests to disclose. Reprint requests: Cengiz Alagöz, MD, Beyo glu Eye Research and Training Hospital, Bereketzade Camii sok, Kuledibi, Beyo glu, Istanbul 34000, Turkey; e-mail: [email protected]
2001
2002 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
The break in the Bruch membrane itself has no effect on visual function.6,7 However, with disease progression, degenerative changes in the retina pigment epithelium and overlying neurosensory retina may occur, decreasing the visual acuity. Even without CNV progression, there may be low visual acuity in eyes with angioid streaks involving the fovea, which makes the visual acuity gain as the treatment goal impractical in some of the cases. Also for the same reason, in the decision to retreat, anatomical changes such as leakage on fundus fluorescein angiography and intra/subretinal fluid observed by optical coherence tomography (OCT) might be referred other than the visual function. Various methods such as laser photocoagulation and photodynamic therapy with verteporfin have been used in the treatment of CNV due to angioid streaks. However, these treatment methods did not yield the desired success rate because of high recurrence rates, atrophic changes, and ongoing growth of CNV.5,8–10 As in age-related CNV and CNV secondary to high myopia, various intravitreal, antivascular endothelial growth factor (anti-VEGF) agents have been shown to be successful in the treatment of CNV due to angioid streaks.11–22 Yet, there is no optimal treatment protocol for this group of patients, and the risk factors for the disease progression could not be clearly identified. In this study, we investigated the long-term results of intravitreal bevacizumab injection in eyes with CNV secondary to angioid streaks and assessed the factors influencing disease progression.
Materials and Methods All patients who were treated with intravitreal bevacizumab (1.25 mg/0.05 mL; Altuzan, Roche, Mannheim, Germany) for CNV secondary to angioid streaks in Beyo glu Eye Research and Training Hospital between July 2009 and March 2014 were investigated retrospectively using patient records. The principles of the study were in accordance with the Declaration of Helsinki, and a written informed consent was obtained from every patient before every injection. Subjects with corrected baseline visual acuity of ,20/ 200 were excluded, as were subjects who were followed up for less than 6 months, and those with any other disease that can compromise visual acuity. Bevacizumab injection was not applied to patients with uncontrolled systemic hypertension or with a history of thromboembolism or cerebrovascular accident.
2015 VOLUME 35 NUMBER 10
In addition to demographic findings, ophthalmologic examination findings of all patients before the injection and during follow-ups were recorded from their files. For statistical calculations, visual acuity values were converted from the Early Treatment Diabetic Retinopathy Study chart to logarithm of minimal angle of resolution (logMAR). Biomicroscopic examination findings, Goldmann applanation tonometry measurements, and findings of dilated fundus examination performed with 90-diopter lens were recorded. Optical coherence tomography scans (either Stratus OCT; Carl Zeiss Meditec Inc, Dublin, CA, or Spectralis HRA-OCT; Heidelberg Engineering, Heidelberg, Germany) of all subjects before the injection and at the last follow-up were analyzed. Initial and final fundus fluorescein angiography (either Zeiss FF450; Zeiss, Oberkochen, Germany, or Spectralis HRA-OCT) images were examined. Localization of CNV was determined by angiographic images before the injection. Initially, intravitreal bevacizumab injection was applied on subjects either with newly diagnosed CNV or on subjects with an old CNV scar demonstrating activity; further treatment was administered on a pro re nata regimen. The patients received a followup 4 to 5 weeks after the injections and then at regular intervals at the discretion of the treating retina specialist. In this study, intravitreal bevacizumab was administered as monotherapy, where the combination of intravitreal bevacizumab with any other treatment modality on the same or different session was not performed. However, a previous treatment history of photodynamic therapy (N = 2) was not an exclusion criterion because the photodynamic therapy was applied at least 3 years before the first injection. The decision to retreat was made if at least one of the following signs was present: occurrence of a new intraretinal or subretinal macular hemorrhage from CNV as detected on fundoscopic examination, leakage on angiography, or presence of intraretinal or subretinal fluid on OCT. When any sign of membrane activity, such as increasing metamorphopsia, new or persistent leakage on fluorescein angiography, retinal hemorrhage, or edema documented by OCT was detected, intravitreal injection was repeated regardless of the visual acuity level. Figure 1 shows a series of OCT images of a patient with CNV due to angioid streaks that showed multiple recurrences and received 20 injections over a period of 40 months. Primary end points of the study were stabilization or improvement in visual acuity between baseline and the last follow-up, and final activity of CNV. Stabilization of visual acuity was considered as visual acuity change of 2 ETDRS lines and improvement as gain in visual
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2003
Fig. 1. A series of OCT images of a 41-year-old male patient with PXE (Patient 8). At baseline, the patient had a juxtafoveal CNV that demonstrated multiple reactivation episodes and increased in size despite the frequent injections of bevacizumab during the follow-up. Both subretinal and intraretinal fluids are observed on baseline time domain OCT (A). At 14 months of follow-up, after receiving 7 injections, the lesion did not demonstrate any activity; no fluid was observed on the time domain OCT image (B). After 30 months of follow-up, the patient received a total of 13 injections and the lesion extended subfoveally; arrows show subretinal fluid indicating activity; the arrowhead shows intraretinal cysts located over the fibrous scar; and the asterisks show outer retinal tubulation (ORT) (C). The patient received another 3 injections and at 34 months of follow-up, only the intraretinal cysts (arrowhead) and ORT (asterisks) were observed on the OCT image; no additional injection was administered because those findings did not change in response to treatment and therefore were not regarded as an activity sign (D). At the next follow-up (1 month later), new subretinal fluid collection appeared at the edges of the lesion (arrow) (E-1 and E-2) and the patient received further injections. The OCT images of the same patients at the last follow-up (40 months) receiving a total of 20 injections; the lesion activity is present as observed by the subretinal fluid (arrows, F-1 and F-2). Note that the intraretinal cysts (arrowhead) and ORTs (asterisks) appeared almost similar at different time points of follow-up (C, D, E-1, and F-1). Images (E-2 and F-2) are from different sections (superior to the fovea) of the same OCT imaging as the images (E-1 and E-2), respectively.
acuity of 2 ETDRS lines. The cases with new intraretinal or subretinal macular hemorrhage from CNV and/ or leakage on fluorescein angiography and/or the presence of intraretinal or subretinal fluid on OCT at the last follow-up were considered to have final active CNV. The secondary end points of the study were the factors affecting the disease progression and complications of the treatment. Further analysis was performed for the following conditions: the eyes with final active membrane (N = 14) were compared with the eyes with final inactive membrane (N = 9), and the subjects with PXE (N = 10) were compared with the others (N = 13) for age, follow-up period, number of injections, initial and final visual acuity, and visual acuity change. For statistical analysis, SPSS 20.0 for Windows was used. Continuous variables were presented as mean ± standard deviation. Frequency and percentage (%) were used for categorical variables. For quantitative variables, the Wilcoxon test was used for the comparison of two
related groups, whereas the Mann-Whitney U test was used for the comparison of two independent groups. The Fisher’s exact test was used for categorical variables. P , 0.05 was considered to be statistically significant. Results No ocular or systemic complication was encountered in the follow-up period related to the intravitreal bevacizumab. Twenty-three eyes of 20 patients (7 female, 13 male) were included in the study. The mean (±standard deviation) age of patients was 45.7 ± 9.1 (range, 29–65) years, and the follow-up after the treatment was 23.2 ± 16.2 (range, 6–58) months. The average number of injections performed during follow-up was 5.1 ± 4.3 (range, 1–20). Two eyes had a history of ocular photodynamic therapy (one eye 3 times and the other eye only once) dating back at least 3 years before the start of this study. Localization of CNV due to angioid streaks was subfoveal in
2004 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
15 eyes (65.2%) and juxtafoveal in 8 eyes (34.8%) on initial examination. All patients with a diagnosis of angioid streaks were referred to the dermatology department; and 10 patients (43.5%) received a biopsy-confirmed diagnosis of PXE, whereas 1 patient (4.3%) had Ehlers–Danlos syndrome. The mean logMAR visual acuity of patients was 0.53 ± 0.33 (20/63, Snellen equivalent) at baseline and 0.60 ± 0.40 (20/80) on the last follow-up (P = 0.79). On the last examination, visual acuity was stabilized in 7 eyes (30.4%), improved in 8 eyes (34.8%), and worsened in 8 eyes (34.8%), compared with the baseline. The mean change in VA was −0.65 ± 5.01 (range: −10 to +9) ETDRS lines. An active membrane was present in 14 of 23 eyes (61%) on the last follow-up. When the eyes with a final active CNV (N = 14) were compared with those with a final inactive CNV (N = 9), the mean age was 41.8 ± 6.2 years in the first group and 51.8 ± 9.7 years in the latter group (P = 0.007). The average number of injections was 6.6 ± 4.7 and 2.8 ± 2.3, respectively (P = 0.013). There was no significant difference in the duration of follow-up between the 2 groups (P = 0.829). The mean interval between the injections (follow-up time/injection number) was 4.1 ± 2.4 and 9.0 ± 4.8 in the groups, respectively (P = 0.001). Both the baseline and final visual acuity values were found to be similar between the groups (P = 0.179, P = 0.926, respectively) (Table 1). Figure 2 represents an example of patients with final active CNV, and (Figures 3 and 4) of patients with final inactive CNV. Nine of 14 eyes (64.3%) with final active CNV had a subfoveal CNV on initial examination, and 5 (35.7%) had a juxtafoveal CNV initially. Three eyes with initial juxtafoveal CNV demonstrated subfoveal extension at the final follow-up (Figures 1 and 2). In subjects with final inactive CNV, 6 of 9 eyes (66.6%) had a subfoveal CNV and 3 eyes (33.3%) had a juxtafoveal CNV initially. None of the eyes with juxtafoveal membrane in this group of patients showed involvement of the fovea during the follow-up (Figure 4).
2015 VOLUME 35 NUMBER 10
Characteristics of all cases were summarized in Table 2. Eleven of 12 eyes (91.7%) of patients who were younger than 46 years (the mean age of subjects in the study population) had signs of membrane activity on the last examination, whereas activity rate of eyes of the patients older than 46 years was 27.3% (3 of 11 eyes) (P = 0.003). Taking into account the patients with PXE (10 eyes of 10 patients), the mean age at presentation was 43.5 ± 10.1 years, the follow-up was 28.3 ± 18.2 months, and the average number of injections was 7.5 ± 5.5. The mean (follow-up time/injection number) ratio was 4.4 ± 2.5. The baseline logMAR visual acuity was 0.53 ± 0.31 (20/63), and the final visual acuity was 0.72 ± 0.37 (20/100) (P = 0.352). Final membrane activity was present in 7 of 10 eyes (70%) in this group. The patients without PXE (13 eyes of 10 patients) had a mean age of 47.4 ± 8.1 years. The mean follow-up was 19.4 ± 14.1 months, and the mean number of injections was 3.2 ± 1.9. The mean (follow-up time/injection number) ratio was 7.3 ± 4.9. The baseline visual acuity was 0.53 ± 0.37 (20/63), and the final visual acuity was 0.50 ± 0.42 (20/63) (P = 0.552). Final membrane activity was present in 7 of 13 eyes (53.8%). Although the difference in the follow-up between the groups with and without PXE was nonsignificant, the injection number was higher in the PXE group (P = 0.018). There seemed to be a difference between the follow-up/injection number ratio (4.4 vs. 7.2, for subjects with and without PXE, respectively); however, the statistics could not reveal any significance (P = 0.072) (Table 3).
Discussion In this study, the mean follow-up time of subjects with CNV due to angioid streaks was 23 months, and the mean number of repeated intravitreal bevacizumab
Table 1. Comparison of Eyes With Final Active CNV (N = 14) to Those With Final Inactive CNV (N = 9)
Age, years Membrane Localization Number of injections Follow-up, months Follow-up/injection, months Initial VA, logMAR Final VA, logMAR
Subfoveal Juxtafoveal
Data were expressed as mean ± standard deviation. *Mann–Whitney U test. VA, visual acuity.
Final Active CNV (N = 14)
Final Inactive CNV (N = 9)
P*
41.8 ± 6.2 9 (64.3%) 5 (35.7%) 6.6 ± 4.7 24.6 ± 16.7 4.05 ± 2.4 0.46 ± 0.34 0.61 ± 0.42
51.8 ± 9.7 6 (66.6%) 3 (33.3%) 2.8 ± 2.3 21.1 ± 16.2 9.01 ± 4.8 0.65 ± 0.31 0.58 ± 0.40
0.007 1.00 0.013 0.829 0.001 0.179 0.926
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2005
Fig. 2. Images obtained from a 29-year-old male patient (Patient 1) with PXE. Color fundus picture at baseline shows cracks radially extending from the optic disk (A). Note the juxtafoveally localized choroidal neovascular membrane and multiple subretinal hemorrhages surrounding the membrane. Intraretinal and minimal subretinal fluid is observed on OCT image (B). One month after the first injection of intravitreal bevacizumab, the lesion and hemorrhage regressed as seen on the color fundus picture (C). The corresponding FA did not show leakage from the lesion (D), and fluid is not observed on OCT examination (E). After a long quiescent period (20 months), the lesion reactivated (F–H). Color fundus image (F), leakage on the late frames of FA (G), and intraretinal and subretinal fluid on OCT (H) indicating the activity of the membrane. After 12 injections (47 months), the membrane still demonstrated activity (I–K); color fundus photograph (I), FA (J), and OCT (K) images of the patient at the last follow-up. FA, fluorescein angiography.
injections needed during follow-up was 5. The final visual acuity showed no significant change when compared with the baseline. The mean change in VA was −0.65 ± 5.01 (range: −10 to +9) ETDRS lines. Sixtyfive percent of eyes with CNV due to angioid streaks demonstrated an improvement or stabilization in visual acuity, whereas visual acuity worsened in 35% of eyes. With respect to visual acuity improvement, various studies using anti-VEGF agents to treat CNV secondary to angioid streaks yielded similar results.12–22 In a series of 35 subjects, ranibizumab injections were performed and the subjects were followed up for a mean of 24 months. The same study reported either an improvement or a stabilization (±3 lines of change) in visual acuity in 85.7% of subjects and a mean change in visual acuity of −0.17 (−10 to +7) lines.18 In the study conducted by Battaglia Parodi et al,19 intravitreal bevacizumab was
performed on 15 eyes and the subjects were followed up for 1 year. The study reported a 33% increase and 46.6% stabilization in visual acuity, whereas the mean number of injections was 2.5 and the mean initial visual acuity was preserved at the same level throughout the study period. In a study by Çekiç et al,13 intravitreal pegaptanib sodium was used in 5 eyes in repeated doses. After 15 months to 24 months of follow-up, 2 of the subjects had an improvement in visual acuity, whereas the remaining 3 were reported to have a stable visual acuity throughout the study. Our study reflects the treatment results achieved by an initial dose of intravitreal bevacizumab followed by repetitive doses as membrane activity was detected. However, a series where monthly ranibizumab was injected for 12 months regardless of membrane activity reported quite successful short-term results.20
2006 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2015 VOLUME 35 NUMBER 10
Fig. 3. Images obtained from a 56-year-old female patient (Patient 17) with PXE. In addition to an area showing retina pigment epithelium changes and multiple tiny hemorrhages in the macula, angioid streaks are clearly visible on the initial fundus photograph image (A). Initial angiographic image showing fluorescein blockage because of both retina pigment epithelium hypertrophy and subretinal hemorrhage, leakage from the active choroidal neovascular membrane, and a cystoid macular edema secondary to the membrane (B). The intraretinal fluid is visible on initial OCT (C). The patient received 3 intravitreal bevacizumab injections during a 26-month follow-up period. The final fundus photograph image shows the areas of retina pigment epithelium changes (atrophy and hypertrophy) and a scar involving the macula (D). Hyperflourescence because of staining of the scar tissue without leakage is noted on late frames of the final fluorescein angiography (E). Atrophy of the neurosensory retina overlying the subretinal scar is seen on final OCT of the patient (F). Neither subretinal nor intraretinal fluid was present at final visit.
In this prospective study of 7 eyes, visual acuity increased from 20/63 to 20/32 at the 12th month and was preserved at the same level for 3 months after the last injection. However, long-term results of frequent anti-VEGF treatment remain unclear. There is a doubt that the frequent use of anti-VEGF agents in patients with angioid streaks may accelerate the atrophic changes in the retinal pigment epithelium and neurosensory retinal layers because the drug can readily reach the choriocapillaris through the breaks in the Bruch membrane.22,23 Finger et al22 reported an increase in retina pigment epithelial atrophy and
subretinal fibrosis in fundus autofluorescence imaging of patients with angioid streaks, despite the frequent anti-VEGF treatment. In this study, we assessed the membrane activity of the subjects at the last follow-up and observed that activity tends to persist in patients of younger age despite repetitive injections. Ninety-two percent of patients who were 46 years of age or younger had an active CNV on the last examination, whereas the activity rate among the patients older than 46 years was 27%. Likewise, the mean age of subjects who had signs of membrane activity at the last follow-up was
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2007
Fig. 4. Juxtafoveal CNV secondary to angioid streaks in a 49-year-old female patient without PXE (Patient 14) (A–C). Fundus photograph (A); fluorescein angiography at late frames demonstrating leakage from CNV (B); subretinal fluid seen at the border of the lesion (arrow, C). The lesion was inactive after one injection of bevacizumab. A series of OCT images from the same patient during the follow-up of 19 months (D at 1 month, E at 6 months, and F at 19 months).
significantly lower than the remaining subjects. In addition, the mean number of injections was higher in the group of patients with final active CNV. Although the duration of follow-up periods was similar between the 2 groups, the higher number of injections in patients with final active CNV and the tendency of the juxtafoveal lesion to involve the fovea suggest that the disease progression is more aggressive in this group of patients (Figures 1 and 2). The most frequent systemic association in angioid streaks is PXE.4 Ten of 20 patients (50%) (10 of 23 eyes) had PXE in our series. In PXE, because of the absence of a systemic antimineralization factor, calcification and fragmentation of the elastic fibers in a number of organs, including the skin, eye, and the blood vessels, occurs. Involvement of the blood vessels, besides being a risk factor for cardiovascular diseases such as acute myocardial infarction and cerebrovascular occlusion,24,25 also may constitute a risk factor for abnormal wound healing. Viljoen et al26 reported tissue friability, poor wound healing, and keloid formation after cosmetic surgery in patients with PXE. In a previous report by Mimoun et al,18 no difference was found between the treatment results of subjects with and without PXE. This study drew attention to the PXE group that seemed to require a higher number of injections. Although nonsignificant, there was a tendency for frequent injections in the treatment of subjects with PXE, which is demonstrated by the shorter treatment intervals in this group. Our study demonstrated that the PXE group required injections
in every 4 months in contrast to every 7 months for the non-PXE group. This may be an important finding for the clinical practice. Because of an increased cardiovascular risk profile in PXE patients and the anti-VEGF treatment itself possessing risk for systemic complications such as cerebrovascular occlusion, the treatment should be considered meticulously and followed up carefully.27,28 Acute cerebrovascular accident after bevacizumab injection was reported in a 54-year-old PXE patient with no other systemic disease.29 In our series, no ocular or systemic complication because of bevacizumab injection was encountered. Because of high recurrence rates and rapid progression, CNV due to angioid streaks has a poorer prognosis when compared with other etiologies.3 The disease may show multiple recurrences after inactive periods or a new CNV in a different area may develop.30 Intravitreal bevacizumab has a short activity period, and repetitive injections are needed. Therefore, patients should be monitored frequently and informed both of the possibility of recurrence and of an increased cardiovascular risk after the treatment. The limitations of the study include the retrospective study design. Apart from this, the series is small because of the rarity of the disease and includes patients with different etiologies for angioid streaks, such as PXE, Ehler-Danlos syndrome, and idiopathic ones. Among the subjects, only 1 patient (Patient 1) did not attend his scheduled follow-ups for a period of 20 months in the quiescent period after the first
Patient No.
M F M M M M M M M M M M M F M F F F F M M M F
L L L L R R L L R R L L L L R R L R L L R L R
PXE PXE PXE — — — — PXE PXE — — PXE — EDS PXE — — PXE PXE PXE — — —
Juxtafoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Juxtafoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal Subfoveal
Final Foveal Involvement of CNV*
Previous Treatment
+ − − + PDT + − − −
·3 PDT
Number of Injections
Baseline VA, Final VA, logMAR logMAR
12 1 2 4 5 1 3 9 20 6 2 8 3 6 7 1 5 5 3 7 2 1 1
*Foveal involvement for the membranes initially localized juxtafoveally. EDS, Ehlers–Danlos syndrome; F, female; L, left eye; M, male; PDT, photodynamic therapy; R, right eye; VA, visual acuity.
0.30 0.22 0.15 1.00 1.00 0.40 0.15 0.52 0.15 0.30 0.05 0.80 0.70 0.52 0.70 1.00 0.10 0.80 0.70 1.00 0.40 0.30 1.00
1.30 0.22 0.15 0.70 0.40 0.15 0.40 0.80 1.00 1.30 0.05 0.40 0.40 1.30 0.70 0.15 0.22 1.00 1.00 0.70 0.15 0.52 0.80
Follow-up, months
Final Membrane Activity
47 6 6 9 29 10 10 33 40 30 8 31 14 33 12 19 56 19 26 58 10 10 14
+ − + + + + + + + + + + − − + − + + − − − − −
2015 VOLUME 35 NUMBER 10
29 31 35 39 39 40 40 40 41 43 43 44 47 47 48 49 50 54 56 57 57 57 65
Baseline CNV Localization
1 2 3 4 5 6 6 7 8 9 9 10 11 12 13 14 15 16 17 18 19 19 20
Age, Systemic years Gender Eye Disease
2008 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
Table 2. Clinical Features of All Cases Affected by CNV Due to Angioid Streaks Treated by Intravitreal Bevacizumab
INTRAVITREAL BEVACIZUMAB IN ANGIOID STREAKS ALAGÖZ ET AL
2009
Table 3. Comparison of the Eyes of Subjects With and Without PXE
Age, years Number of injections Follow-up, months Follow-up/injection, months Initial VA, logMAR Final VA, logMAR Final active CNV
Eyes With PXE (N = 10)
Eyes Without PXE (N = 13)
P*
43.5 ± 10.1 7.5 ± 5.5 28.3 ± 18.2 4.39 ± 2.5 0.53 ± 0.31 0.72 ± 0.37 7 (70.0%)
47.4 ± 8.1 3.2 ± 1.9 19.4 ± 14.1 7.28 ± 4.9 0.53 ± 0.37 0.50 ± 0.42 7 (53.8%)
0.446 0.018 0.284 0.072 1.000 0.166 0.669
Data were expressed as mean ± standard deviation. *Mann–Whitney U test. VA, visual acuity.
injection, and he presented as soon as he noticed the first symptoms of disease reactivation. He was regularly followed up then. In conclusion, intravitreal bevacizumab stabilized visual acuity in eyes with CNV due to angioid streaks with a mean follow-up of 23 months. In younger ages, CNV behaved more aggressively and tended to be more resistant to the treatment. Eyes with PXE contributed to half of the overall final active CNVs in the study. Although we could not find a difference in the treatment intervals between eyes with and without PXE (P = 0.07), an increased number of the study population might reach the statistical significance. Prospective, randomized long-term studies are warranted to determine the optimal frequency of anti-VEGF treatment in CNV due to angioid streaks and to justify the use of intensive anti-VEGF treatment, especially in patients who are expected to show aggressive progression. Key words: angioid streaks, choroidal neovascular membrane, intravitreal bevacizumab, pseudoxanthoma elasticum. References 1. Doyne RW. Choroidal and retinal changes. The results of blows on the eyes. Trans Ophthalmol Soc U K 1889;9:128. 2. Adelung JC. Genesis of angioid streaks (Knapp). Klin Monbl Augenheilkd Augenarztl Fortbild 1951;119:241–250. 3. Clarkson JG, Altman RD. Angioid streaks. Surv Ophthalmol 1982;26:235–246. 4. Connor PJ Jr., Juergens JL. Pseudoxanthoma elasticum and angioid streaks. A review of 106 cases. Am J Med 1961;30: 537–543. 5. Pece A, Avanza P, Galli L, Brancato R. Laser photocoagulation of choroidal neovascularization in angioid streaks. Retina 1997;17:12–16. 6. Charbel Issa P, Finger RP, Holz FG, Scholl HP. Multimodal imaging including spectral domain OCT and confocal near infrared reflectance for characterization of outer retinal pathology in pseudoxanthoma elasticum. Invest Ophthalmol Vis Sci 2009;50:5913–5918. 7. Dreyer R, Green WR. The pathology of angioid streaks: a study of twenty-one cases. Trans Pa Acad Ophthalmol Otolaryngol 1978;31:158–167.
8. Singerman LJ, Hatem G. Laser treatment of choroidal neovascular membranes in angioid streaks. Retina 1981;1:75–83. 9. Chung AK, Gauba V, Ghanchi FD. Photodynamic therapy (PDT) using verteporfin for juxtafoveal choroidal neovalkoscularization (CNV) in angioid streaks (AS) associated with pseudoxanthoma elasticum: 40 months results. Eye (Lond) 2006; 20:629–631. 10. Browning AC, Chung AKK, Ghanchi F, et al; United Kingdom PDT Users Group. Verteporfin photodynamic therapy of choroidal neovascularization in angioid streaks: one year results of a prospective case series. Ophthalmology 2005;112: 1227–1231. 11. Yilmaz I, Ozkaya A, Alkin Z, Yazici AT. Intravitreal ranibizumab for bilateral choroidal neovascularisation in a patient with angioid streaks. BMJ Case Rep 2014;29:2014. 12. Finger RP, Charbel Issa P, Ladewig M, et al. Intravitreal bevacizumab for choroidal neovascularization associated with pseudoxanthoma elasticum. Br J Ophthalmol 2008;92:483–487. 13. Çekiç O, Göçmez E, Kocabora MS. Management of CNV in angioid streaks by intravitreal use of specific anti-VEGF165 Aptamer (pegaptanip sodium): long-term results. Curr Eye Res 2011;36:492–495. 14. Kang S, Roh YJ. Intravitreal ranibizumab for choroidal neovascularisation secondary to angioid streaks. Eye (Lond) 2009; 23:1750–1751. 15. Molina Guilabert I, Calvo-Gonzalez C, Reche-Frutos J, et al. Intravitreal pegabtanib sodium in choroidal neovascularization secondary to angioid streaks. Acta Ophthalmol 2009;87: 581–582. 16. Schiano Lomoriello D, Parravano MC, Chiaravalloti A, Varano M. Choroidal neovascularization in angioid streaks and pseudoxanthoma elasticum: 1 year follow-up. Eur J Ophthalmol 2009;19:151–153. 17. Wiegand TW, Rogers AH, McCabe F, et al. Intravitreal bevacizumab (Avastin) treatment of choroidal neovascularisation in patients with angioid streaks. Br J Ophthalmol 2009;93:47–51. 18. Mimoun G, Tilleul J, Leys A, et al. Intravitreal ranibizumab for choroidal neovascularization in angioid streaks. Am J Ophthalmol 2010;150:692–700. 19. Battaglia Parodi M, Lacono P, LaSpina C, et al. Intravitreal bevacizumab for nonsubfoveal choroidal neovascularization associated with angioid streaks. Am J Ophthalmol 2014;157: 374–377. 20. Finger RP, Charbel Issa P, Hendig D, et al. Monthly ranibizumab for choroidal neovascularizations secondary to angioid streaks in pseudoxanthoma elasticum: a one-year prospective study. Am J Ophthalmol 2011;152:695–703.
2010 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES 21. Gliem M, Finger RP, Fimmers R, et al. Treatment of choroidal neovascularization due to angioid streaks: a comprehensive review. Retina 2013;33:1300–1314. 22. Finger RP, Charbel Issa P, Schmitz-Valckenberg S, et al. Long-term effectiveness of intravitreal bevacizumab for choroidal neovascularization secondary to angioid streaks in pseudoxanthoma elasticum. Retina 2011;31:1268–1278. 23. Gibran SK, Sachdev A, Stappler T, et al. Histological findings of a choroidal neovascular membrane removed at the time of macular translocation in a patient previously treated with intravitreal bevacizumab treatment (Avastin). Br J Ophthalmol 2007;91:602–604. 24. Combrinck M, Gilbert JD, Byard RW. Pseudoxanthoma elasticum and sudden death. J Forensic Sci 2011;56:418–422. 25. Bock A, Schwegler G. Intracerebral haemorrhage as first manifestation of pseudoxanthoma elasticum. Clin Neurol Neurosurg 2008;110:262–264.
2015 VOLUME 35 NUMBER 10
26. Viljoen DL, Bloch C, Beighton P. Plastic surgery in pseudoxanthoma elasticum: experience in nine patients. Plast Reconstr Surg 1990;85:233–238. 27. Wong TY, Liew G, Mitchell P. Clinical update: new treatments for age-related macular degeneration. Lancet 2007;370:204–206. 28. Van den Berg JS, Hennekam RC, Cruysberg JR, et al. Prevalence of symptomatic intracranial aneurysm and ischaemic stroke in pseudoxanthoma elasticum. Cerebrovasc Dis 2000;10:315–319. 29. Besozzi G, Ferrara A, Epifani E, et al. Acute stroke after intravitreal bevacizumab to treat choroidal neovascularization due to angioid streaks in pseudoxanthoma elasticum: a severe systemic adverse event after an off-label procedure. Int Ophthalmol 2013;33:181–183. 30. Sawa M, Gomi F, Tsujikawa M, et al. Long-term results of intravitreal bevacizumab injection for choroidal neovascularization secondary to angioid streaks. Am J Ophthalmol 2009; 148:584–590.
