Key words radiation retinopathy - radiation choroidopathy - uveal melanoma - ocular radiation
Schlüsselwörter strahlenbedingte Retinopathie - strahlenbedingte Choroidopathie - uveales Melanom - Aderhautmelanom - Augenbestrahlung
Introduction
The Collaborative Ocular Melanoma Study (COMS) published its first report in 1990 [1 ]. Since then, the study has demonstrated comparable mortality among patients treated with brachytherapy and enucleation. Thus, globe-preserving therapy with radiation has developed into the standard of care when treating uveal melanomas that are small or medium in size [2 ]. Radiation retinopathy is a well-known, slowly progressive, and predictable consequence of radiation exposure, with vision loss being the most important complication to the patient. Clinical evidence of radiation retinopathy has been shown to occur in up to 42% of patients 5 years after radiation treatment for posterior uveal melanomas [3 ]. The COMS histopathologic review demonstrated vascular abnormalities in 55% of irradiated eyes that were enucleated secondarily to plaque failure [4 ]. Chorioretinal anastomosis,
vascular dilatation, and polypoidal changes have been occasionally clinically noted in eyes with radiation retinopathy, but the histopathologic findings after plaque brachytherapy are not well described in the literature [5 ], [6 ], [7 ], [8 ]. Herein, we review the pathology findings of 26 enucleated eyes after treatment of ocular melanoma with plaque brachytherapy aiming to describe the vascular changes observed in the choroid.
Materials and Methods
Patient selection
After approval by our Institutional Review Board, the study patient population was identified from our patient database at Mayo Clinic from June 7, 2007 to June 7, 2017 (10 years). A total of 101 patients had undergone enucleation for uveal melanoma. The diagnosis was confirmed in all patients by our pathology department at Mayo Clinic. Of these 101 patients, we identified 26 eyes that had been enucleated after receiving radiation therapy, with a mean time to enucleation of 7.2 years (range 0 – 30 years). Radiation included brachytherapy (25 eyes) and proton beam radiation (1 eye). All patient charts were reviewed for demographic features (age, sex, age at diagnosis), treatment type, radiation dose, and time from radiation exposure to enucleation.
A complete review of clinical exam findings, ancillary testing, and specimen slides were performed on all patients. The specimen slides from all patients were available and were reviewed by a single pathologist. Our histopathologic assessment included retrospective evaluation of the enucleation specimens for the presence of radiation retinopathy and other radiation-associated findings. More specifically, each specimen was analyzed for characterization of vascular changes in the retina and choroid, including the presence of ectatic blood vessels, hyalinization of vascular walls, and presence of neovascularization. Location of choroidal vasculopathy was recorded as intratumoral or extratumoral. Other features recorded were presence of fibrosis, inflammatory infiltrates, and atrophy of the retina and/or the choroid. All information was placed on a secured, password-protected, encrypted database.
Treatment
Small and medium melanomas were generally treated with radiation (65 – 85 Gy to a prescription point) and enucleation for large melanomas. Radiation therapy was administered using two different techniques: proton beam (done elsewhere) or iodine-125 brachytherapy.
Histopathologic evaluation
The histology slides from the 26 cases were evaluated by a board-certified pathologist who had special training in ocular pathology (D. R. S.). The ocular pathologist was not aware of the prior history of any patient aside from the fact that the patient had uveal melanoma and had plaque brachytherapy. D. R. S. was asked to evaluate the eye for evidence of vascular changes in the retina and choroid.
Statistical analysis
Total number of cases were determined and percentages of the total number of cases, depending upon the numerator and denominator, were then assessed.
Results
Patient characteristics
A total of 101 included patients with uveal melanoma that underwent enucleation that were seen at the Mayo Clinic from June 2007 to June 2017 were identified ([Table 1 ]).
Table 1 Patient Characteristics.
Characteristic
Number of patients (%)
Total number included patients
18
Age at diagnosis
Range
30 – 85 years
Mean
67.2 years
Male
8 (44%)
Female
10 (56%)
Site of involvement
Right eye
10 (56%)
Left eye
8 (44%)
Radiation dose
Range
65 – 85 Gy
Mean
83.3 Gy
Time from treatment to enucleation
Range
0 – 30 years
Mean
7.2 years
Median
5.5 years
Of the 101 eyes, 26 (26%) had received radiation therapy and were enucleated, so they were available for pathologic evaluation. Of these 26 eyes, 18 (69%) had clinical or pathologic evidence of radiation-induced vasculopathy. There was a slight predominance of female patients: 10 patients were female (56%) and 8 patients were male (44%), with a mean age of 67.2 years.
Radiation Outcomes
Clinical findings
The average dose of radiation where the radiation dose was known was 83.3 Gy. The average time from treatment to enucleation was 7.2 years ([Table 2 ]). Nine patients had vitreous hemorrhage at the time of enucleation. When visualization to the fundus was present, nine patients had evidence of radiation retinopathy and, in five patients, large dilated choroidal vasculature was clinically evident.
Table 2 Patient findings.
Location of vasculopathy
Location of vasculopathy
Histopathologic findings
Rad dose (Gy)
Time to Enucleation
Vasculopathy type
Other
Patient #
Sex
Age
Vitreous hemorrhage
Retina
Choroid
Intratumoral
Extratumoral
Ectatic
Hyalinization
1
M
59
+
+
+
Scarring
Unknown
7
2
F
83
+
+
+
+
+
85 Gy
1
3
F
30
+
+
+
+
+
+
Scarring
85 Gy
2
4
M
65
+
+
+
+
Scarring
85 Gy
8
5
M
62
+
Scarring
85 Gy
3
6
F
62
+
Choroidal Inflammation
85 Gy
1
7
F
77
+
+
+
+
+
Unknown
14
8
F
48
+
+
Subretinal Hemorrhage
65 Gy
0
9
F
67
+
Phthisis
Unknown
18
10
M
57
+
+
+
+
+
85 Gy
1
11
M
68
+
+
+
+
+
+
85 Gy
3
12
F
80
+
+
+
+
+
Choroidal Inflammation
Unknown
7
13
M
76
+
+
+
+
+
+
+
85 Gy
9
14
F
81
+
+
+
+
Choroidal Inflammation
85 Gy
3
15
F
83
+
+
+
+
Unknown
10
16
F
73
+
+
+
+
+
+
Choroidal Inflammation
85 Gy
4
17
M
54
+
+
+
+
+
+
+
Unknown
9
18
M
85
+
+
+
Unknown
30
Histopathological findings ([Table 2 ])
A histopathologic review was conducted with emphasis on the choroidal vascular changes within the tumor and in the choroid at the base of the treated tumor. Eighteen of the 26 eyes that were enucleated demonstrated evidence of radiation vasculopathy. Of the 18 eyes, 10/18 (55%) had radiation retinopathy and 16/18 (89%) had radiation choroidal vasculopathy. One patient had a phthisical eye, and the choroid could not be evaluated because the characteristics of the vasculature could not be determined. This patient was also the only patient that had radiation administered via proton beam elsewhere. Nine cases had vitreous hemorrhage (50%), all cases had radiation retinopathy, and 8/9 (89%) had radiation choroidopathy.
Of the 16 cases with radiation choroidal vasculopathy, 3/16 (19%) had only intratumoral radiation choroidal vasculopathy, 3/16 (19%) only had extratumoral radiation choroidal vasculopathy, and, thus, 10/16 (32%) had both intratumoral and extratumoral radiation choroidal vasculopathy. In the patients with radiation choroidal vasculopathy, 2/16 (13%) had hyalinization of the choroidal vessels ([Fig. 1 a ]). Another 3/16 (19%) cases with radiation choroidal vasculopathy had ectatic vessels ([Fig. 1 b ]). The other 11/16 (68%) had evidence of both hyalinization of the choroidal vessels as well as ectatic vessels in the choroid. Polypoidal choroidal vasculopathy, choroidal neovascularization, and retinal choroidal anastomoses (RAP-type lesions) were seen in 12 of the 16 eyes (75%) ([Figs. 1 c ] and [d ]).
Fig. 1 a Hyalinization. Histological section shows marked hyalinization of vascular walls in the choroidal vessels (arrows), subretinal accumulation of histiocytes, and reactive changes in the overlying retina (hematoxylin-eosin stain; original magnification × 100). b Ectatic vessels. Low-power view of the choroidal melanoma demonstrates dilated intratumoral vessels focally coalescing to form hemorrhagic lakes (arrows). The overlying retina shows marked atrophy (hematoxylin-eosin stain; original magnification × 20). c Polyps. Low-power view of the choroidal mass with markedly dilated intratumoral, resulting in an elevated, polypoid mass (hematoxylin-eosin stain; original magnification × 20). d Ectatic vessels and neovascularization. High-power view shows ectatic choroidal vessels and areas of neovascularization (arrow) (hematoxylin-eosin stain; original magnification × 200).
During our analysis, certain histologic features became evident. Hyalinization or thickening of the vascular wall was a common early feature seen on histology ([Fig. 1 a ]). Next, the presence of ectatic or thin-walled vessels appeared as the severity of presentation progressed ([Fig. 1 b ]). The next stage of severity appeared to be with the formation of choroidal polyps as well as choroidal neovascularization. The size of the polyps varied from subtle polypoidal dilatation to large polypoidal complexes ([Fig. 1 c ]). It was sometimes difficult to distinguish polypoidal vessels from choroidal neovascularization. The most severe cases had vitreous hemorrhage and polyps or neovascularization in the choroid.
Currently, a histological grading system for radiation-associated choroidal vasculopathy has not been proposed. It is important for a grading system to be available as we learn more about the pathogenesis of this process. It will be important to organize this data for future research ([Table 3 ]). Grade 1 radiation choroidal vasculopathy is the presence of hyalinization of the choroidal vessels ([Fig. 1 a ]). Grade 2 radiation choroidal vasculopathy is the development of ectatic or thin-walled vessels ([Fig. 1 b ]). Grade 3 is the presence of polyps or choroidal neovascularization ([Fig. 1 c ] and [d ]). Grade 4 is the presence of vitreous and/or subretinal hemorrhage in the presence of Grade 3 vessels and no retinal neovascularization to account for the vitreous hemorrhage.
Table 3 Choroidal vasculopathy grading system.
Grade
Hyalinization
Ectatic vessels
Polyps/neovascularization
Vitreous hemorrhage
Figure
0
–
–
–
–
1
+
–
–
–
[Fig. 1 a ]
2
+
+
–
–
[Fig. 1 b ]
3
+
+
+
–
[Fig. 1 c ]
4
±
±
±
+
Discussion
The incidence of radiation retinopathy has been extensively reviewed in the literature, with rates generally ranging from 10 to 63% [3 ], [4 ], [9 ], with proliferative changes reported to develop in 7% of eyes within 10 years after plaque therapy [10 ]. Reviewing the literature and their own institutionʼs experiences treating radiation retinopathy, Wen and McCannel noted that anti-VEGF and intraocular steroids had an effect on macular edema, but its use for visual recovery over time was limited [9 ]. In contrast, Bascom Palmerʼs review of over 5000 injections for radiation retinopathy (54.2%) demonstrated stability or improvement of vision [11 ], [12 ].
Avery et al. previously studied histopathologic characteristics of patients from the COMS treated with iodine-125 versus radiation-naïve tumors. Noted findings included inflammation, fibrosis, vascular changes within the tumor, and damage to the retinal vasculature, which was consistent with our findings [4 ]. In contrast to our study, we also noted changes to the choroidal vasculature that localized to intratumoral and extratumoral locations. Spaide et al. conducted a retrospective imaging review of 193 patients who underwent radiation treatment for subfoveal choroidal neovascularization. Spaide first described choroidal vascular dilatations or “blebs” in 19 patients outside the border of previous CNV [6 ]. This description is similar to the ectatic choroidal vascular changes seen histologically in our study. Additionally, Peiretti et al. and Midena et al. noted, with indocyanine green angiography chorioretinal
anastomosis, atypical dilated choroidal vessels and choriocapillaris drop out, respectively [5 ], [13 ]. A comparable histopathologic study by Egbert et al. looked at 17 eyes enucleated after radiation treatment for retinoblastoma. Their study found radiation-induced abnormalities to the posterior ciliary arteries but did not identify radiation-induced choroidal changes [14 ]. The myointimal proliferation significantly constricted the posterior ciliary vessels and, in some cases, caused complete occlusion [14 ]. This suggests that posterior ciliary vessels, just like retinal vessels, are radiosensitive. Radiation-induced ischemia to the choroid from posterior ciliary artery occlusion, as well as free radical damage to the choroid, would cause a multifactorial effect on the choroidal vasculature.
There are two studies in the literature that have evaluated indocyanine green angiography of the choroidal vasculature after radiation, which suggested that there is not only vascular closure but also dilation of existing vessels, leakage of the vessels near the site of radiation, and possible choroidal neovascularization [15 ], [16 ]. Finally, another prior histopathologic study of cases treated with plaque brachytherapy discussed that some cases had hyalinization of vessels, dilation of some vessels, and intratumoral as well as subretinal hemorrhages, but the evaluation was not as specifically directed to evaluating the choroidal vasculature as this study; but the fact that some cases had similar findings suggests that the findings that we have described have been seen by others by ICG as well as by histology [17 ].
Our study is limited by being a retrospective study, but it was especially designed to evaluate the choroidal vasculature. Additionally, we have graded the response of the radiation choroidopathy but cannot state that it actually conforms to stages since we could not prospectively follow the cases. Regardless, the study does conform to the little that is known about radiation choroidopathy in the literature and adds that choroidopathy is an important aspect of radiation effects to the eye [4 ], [5 ], [6 ], [13 ], [14 ], [15 ], [16 ], [17 ].
Even with intravitreal anti-VEGF treatment, proliferative radiation-induced retinopathy can still be difficult to treat and sometimes recalcitrant to anti-VEGF therapy. Our study shows that, like polypoidal choroidal vasculopathy, radiation-induced choroidal vasculopathy has similar histologic findings, which may explain why anti-VEGF treatments can be unsuccessful.
Conclusion
Our histopathologic analysis demonstrated choroidal vasculopathy to be a common histologic feature associated with the late effects of radiation exposure. Although the presence of these choroidal changes has been previously recognized by ICG angiography [5 ], [6 ], [15 ], [16 ], histologic analysis of these findings has not been well described. Currently, a histologic grading system for radiation-associated choroidal vasculopathy does not exist. The role of radiation choroidopathy, in the subsequent visual loss following radiotherapy and the role of anti-VEGF therapy to treat it, needs to be recognized and distinguished from radiation retinopathy. As we try to determine the cause and treatment, a grading system is needed. Therefore, we would like to propose the following three grades: Grade 1 radiation choroidal vasculopathy is the presence of hyalinization of the
choroidal vessels; Grade 2 radiation choroidal vasculopathy is the development of ectatic or thin-walled vessels; Grade 3 is the presence of polyps or choroidal neovascularization; and Grade 4 is the presence of a vitreous and/or subretinal hemorrhage in the setting of Grade 3 vessels and no retinal neovascularization to account for the vitreous hemorrhage.
