Radiation-induced optic neuropathy (RON), a form of delayed radionecrosis of the central nervous system, is an uncommon but devastating complication of cranial irradiation. Based mostly on single-case or small series reports, it consists of painless, monocular, relatively acute, and irreversible visual loss after fractionated radiation usually exceeding a total dose of 50 Gy in greater than 2 Gy fractions (1–17). The incidence of RON is rare and depends on the nature of the irradiated tissue: 0.53% among partially resected nonfunctioning adenomas (18), 2.04% among partially resected anterior visual pathway meningiomas (7), and 8.7%–9.0% among tumors of the nasopharynx, nasal cavity, and paranasal sinuses (19,20). The latency from treatment to vision loss has ranged from 3 months to 9 years, clustering at 18 months (12,21,22). The optic disc at the time of initial vision loss has usually appeared normal or pale (4,22). Visual field abnormalities consist of nerve fiber bundle or chiasmal defects (22). Magnetic resonance imaging (MRI), critical in excluding compressive, inflammatory, and infiltrative etiologies, has revealed a characteristic postgadolinium enhancement along a discrete segment of the prechiasmatic (intracranial) optic nerve or optic chiasm (1–10,12–15). Although published cases have described these MRI abnormalities, they have not been displayed frequently in detail (1–17). The largest published case series that contains documentation of accompanying MRI abnormalities had only 6 patients (9 eyes) (4). This report contributes 12 new cases (15 eyes) with emphasis on imaging abnormalities in correlation with clinical findings.
All cases were accrued either from the University of Michigan Neuro-Ophthalmology Clinic files or from cases coded as “radiation optic neuropathy” at the University of Michigan Medical Center between 1994 and 2017. The study interval was chosen for availability of patient data and adequate radiologic studies. All patients had undergone 3D-conformal linear accelerator (photon) external beam radiation, although radiation protocol and dosage were not available on all patients. Clinical follow-up ranged from 3 weeks to 52 months after initial presentation with vision loss (average: 13.5 months). We excluded patients in whom there were insufficient ophthalmic or radiologic data, or evidence of multiple sclerosis, optic disc edema, lumbar puncture–elevated opening pressure or abnormal cerebrospinal fluid constituents, stereotactic radiosurgery, or confounding abnormalities on MRI that would suggest an alternative explanation for the optic neuropathy. This study was approved by our institutional review board.
From the available records, we documented age at time of radiation, patient sex, patient vasculopathic risk factors, the pathology for which radiation was applied, date and protocol of radiation, date of onset and laterality of vision loss, evolution of vision loss, pre-treatment and post-treatment visual acuity, presence of an afferent pupillary defect, Humphrey visual field mean deviations, and ophthalmoscopic optic disc appearance.
A single neuroradiologist (E.A.L.) evaluated all MRI scans, noting the date of the imaging study in relation to the first date of documented RON, as well as enhancement, expansion, and thinning of the optic nerves or chiasm, corresponding T2 signal abnormalities, and the absence of demyelination or confounding compressive lesions. All studies were performed with and without gadolinium contrast.
Most studies were completed at our institution within several weeks of onset of symptoms or diagnosis. Two studies were completed at an outside institution, so that MRI machine specifics were not available (Cases 1, E1). Two studies were completed on the 1.5-T General Electric Signa Excite, 8 channel brain coil (Cases E6, 4); 1 study was completed on the 1.5-T General Electric Signa, head coil (Case E4); 2 studies were completed on the 1.5-T Philips Achieva, 8 channel brain coil (Cases E2, E5); 1 study was completed on the 1.5-T Philips Ingenia (Case E3); 1 study was completed on a 1.5-T Signa HCxt, head coil (Case E4); and 4 studies were completed on the 1.5-T Genesis Signa, 8 channel brain coil (Cases 2, 3, E10, E11). In one case (Case E2), we relied on the outside report of a study performed at the time of initial vision loss, but a repeat study at our institution 17 months later showed persistence of the abnormal findings reported in the initial study.
Twelve patients met inclusion criteria. Four are described here, and the details and imaging studies of the other 8 patients can be found in Supplementary Digital Content, Cases E1-E8, https://links.lww.com/WNO/A349.
There were 12 patients (8 women) and 15 affected eyes. Patient age at the time of radiation ranged from 50 to 81 years (average: 61 years). Six of the 12 patients had no conventional vasculopathic risk factors. Two patients (Cases E3, E4) had more than one vasculopathic risk factor, specifically hypertension and hyperlipidemia. In total, 5 patients suffered from hypertension, 1 from myeloproliferative disorder, 2 from hyperlipidemia, and 1 was a current smoker (Table 1).
Vision loss was monocular in 9 patients and binocular in 3 patients. When vision loss was binocular, it occurred consecutively in the 2 eyes with an interval of 3–6 months. In 4 eyes, vision loss was sudden and nonprogressive (Cases 1, E5, 3, and E6). In 3 eyes, sudden vision loss was followed by progression over 4–7 months (Cases E5, E7, and E8). In 2 eyes, decline of vision was subacute over 1–4 weeks (Cases E1 and E2). In 6 eyes, visual decline was slowly progressive over 2–15 months (Cases 2, E3, E4, 3, and E7). In 1 eye, sudden vision loss was followed by mild improvement (Case E1). At the last available visit, visual acuity was typically devastated: 5 eyes had no light perception (NLP), 1 had hand motion (HM), and 4 had counting fingers (CFs). The other 5 affected eyes had less impaired visual acuity (20/150, 20/60, 20/60, 20/40, and 20/20) but with markedly impaired visual fields. The latency of vision loss after completion of radiation therapy was wide, spanning 7–48 months (average 18 months).
Relative Afferent Pupillary Defect
A relative afferent pupillary defect (RAPD) was noted at the initial visit in all patients, involving the worse-seeing eye.
Where automated (Humphrey 24-2 protocol), visual fields were performed on the affected eye at the initial visit, and all cases disclosed either nerve fiber bundle defects or severe constriction. There were no hemianopic defects.
The ophthalmoscopic appearance of the optic disc was documented in 14 of 15 eyes at the first visit. In 1 eye, dense corneal opacification precluded an adequate fundus view (Case E8). Among 11 affected eyes examined within 4 weeks of onset of vision loss, 6 eyes had optic disc pallor (Cases 1, 2, E2, E4, E7, and E8), and 5 eyes had normal-appearing optic discs (Cases E1, E4, E5, and 3). In all but 1 of the 5 eyes with initially normal-appearing optic discs, optic disc pallor developed over the coming weeks to months. In the single eye that had not developed optic disc pallor, the interval from awareness of vision loss to our examination was only 3 weeks (Case 3). Among the 3 affected eyes examined at least 2 months after onset of vision loss, all had optic disc pallor (Cases E3, 3, and E6).
Duration of Follow-up
Final neuro-ophthalmologic examinations occurred 3 weeks to 52 months (average 13.5 months) after initial presentation with vision loss.
Radiation had been administered for metastatic lung carcinoma (2 patients), partially resected pituitary adenoma (2 patients), recurrent pituitary adenoma (1 patient), partially resected anterior clinoid meningioma (1 patient), left temporal lobe glioblastoma multiforme (1 patient), basal cell carcinoma with lacrimal system extension (1 patient), facial squamous cell carcinoma with perineural invasion and cavernous sinus extension (1 patient), nasopharyngeal carcinoma (1 patient), sinonasal squamous cell carcinoma (1 patient), and sinonasal adenocarcinoma (1 patient).
Ten of 12 patients had undergone surgery of the brain, nasopharynx, paranasal sinuses, orbit, or eyelid region before receiving radiation therapy (all cases except E4 and E5).
The 3D conformal radiation protocol was available in only 6 patients. In those patients, total whole brain dose was 37.5 Gy (Case 1) and 30 Gy (Case E4), total left temporal lobe dose was 60 Gy (Case E3), total nasopharynx and neck dose was 133 Gy (Case E5), and total pituitary region dose was 46 Gy (Case 3). One patient developed RON after a total dose delivered in 3 epochs: 60 and 70 Gy to the maxillary region, followed by an unknown dose to the cavernous sinus (Case 2). Another patient developed RON after 2 epochs: 1 to the orbit and 1 to the paranasal sinus region, but dose details were not available (Case E2). Daily dose fractions, available in only 2 patients, were 3 and 2.5 Gy (Case 2).
Fourteen of 15 affected eyes displayed ipsilateral enhancement of the optic nerve (Figs. 1–4). Among the 14 enhancing optic nerves, 13 displayed the enhancement in the prechiasmatic portion and 1 in the intraorbital portion (Case E8). The length of optic nerve enhancement, as measured in the axial plane, ranged from 2 to 14 mm (average 8.3 mm). One optic nerve displayed volume loss without enhancement (Case 4).
In 3 affected eyes, prechiasmatic enhancement was apparent on imaging completed 3, 4, and 6 weeks before the onset of vision loss (Cases 1, 2, and E3, respectively). In 8 affected eyes, prechiasmatic and intraorbital enhancement was appreciable on imaging completed at the time of vision loss (Cases E1, E2, E4, E5, 3, E7, and E8). In 2 affected eyes, prechiasmatic enhancement was present on imaging completed no sooner than 3 and 4 months after the onset of vision loss (Case 3 and E6, respectively). In one patient with bilateral optic nerve involvement (Case E4), the first affected optic nerve did not display enhancement on imaging obtained at the time of vision loss; it was not until 7 months later that segmental prechiasmatic enhancement appeared.
Of the 14 enhancing optic nerves, 9 (64%) also displayed expansion of the enhancing segment. High T2 signal was present in the enhancing segments of 8 (57%) patients. Seven (50%) enhancing optic nerves displayed both expansion and high T2 signal within the affected segment. Four enhancing optic nerves displayed neither expansion nor high T2 signal.
There were 9 eyes with follow-up MRIs (Cases 2, E1–E6); the average duration of enhancement among these patients was at least 6 months, ranging from 2 to 17 months. Information regarding the persistence of accompanying T2 hyperintensity was more limited because many follow-up studies did not include T2 coronal images for adequate evaluation. In 2 cases, T2 hyperintensity persisted for at least 2 and 4 months (Cases E4 and E6, respectively). Volume loss of the affected optic nerve segment developed within 4–11 months among 4 affected eyes (Case 2, E1, and E4).
A 50-year-old woman with a heavy smoking history underwent craniotomy and resection of a right parieto-occipital lesion, followed by whole brain radiation (total dose 37.5 Gy) for metastatic lung carcinoma. Approximately 21 months later, she developed sudden vision loss in her left eye. On presentation, visual acuity was 20/20 in the right eye (unaffected eye), with a nasal hemianopic defect attributed to a right optic radiation metastasis that had been previously resected. Visual acuity of the affected eye (left eye) was HM (unknown baseline). A left RAPD was present. The left optic disc was pale; the right appeared normal. MRI at the time of presentation revealed enhancement of the left prechiasmatic optic nerve, along with increased T2 signal and subtle expansion of the affected segment. Imaging failed to reveal any compressive lesions along the anterior visual pathways. An MRI performed 3 weeks before the onset of vision loss also displayed enhancement of the left prechiasmatic optic nerve (Fig. 1). The patient was last evaluated 2 months after the onset of vision loss without change in visual function.
A 54-year-old woman with no known vasculopathic risk factors received 60 Gy of local radiation for a partially resected squamous cell carcinoma of the right face, nose, and cheek. She underwent repeat radiation to a total dose of 70 Gy (1.25 Gy fractions, twice daily) after an anterior medial maxillectomy with skin flap. A third dose of radiation was given 2 years later for perineural trigeminal invasion. After the third radiation treatment, the patient suffered from a gradual loss of vision in her right eye over 1 year. Visual acuity declined from a baseline of 20/20 to 20/40 in the right eye, with a new superior altitudinal defect in the right visual field. A right RAPD was detected in the right eye, as well as mild optic disc pallor. Visual acuity worsened over the next year, falling to 20/60 with a stable visual field deficit in the right eye. The left visual field remained full, and optic disc appearance in the left eye remained normal. MRI performed 4 weeks before the onset of right vision loss revealed enhancement of the right prechiasmatic optic nerve without expansion of the affected segment (Fig. 2). T2 abnormalities could not be assessed. Imaging failed to reveal any compressive lesions along the anterior visual pathways. Based on follow-up MRI studies, enhancement was present for at least 8 months, resolving by 11 months. The patient was last evaluated 15 months after the onset of vision loss without change in visual function.
A 63-year-old man with no known vasculopathic risk factors underwent pituitary radiation for residual tumor after surgery (total dose 46 Gy). Approximately 7.5 months after radiation, he experienced progressive decline in the vision of his right eye over 3 months. Initially visual acuity was 20/20 in the right eye, but a superior nerve fiber bundle visual field defect was present; visual acuity in his left eye was 20/20. Over that 3-month course, right visual acuity deteriorated to CF. At that time, he noted a sudden decline in vision in his left eye, now measured at 20/40. The right optic disc was diffusely pale; the left was normal. MRI revealed bilateral prechiasmatic optic nerve enhancement (Fig. 3). The left optic nerve segment was expanded, with an accompanying increased T2 signal. There was no evidence of tumor recurrence. The patient was last evaluated 3 weeks after initial presentation without change in visual function.
An 81-year-old woman with systemic hypertension and myelodysplastic syndrome underwent numerous surgical excisions, followed by local radiation (total dose unknown) for recurrent basal cell carcinoma of the left medial canthus extending into the left lacrimal system and orbit. She experienced sudden painless loss of vision in her left eye 24 months after completing radiation. Visual acuity dropped to CF, eventually declining to NLP over the subsequent 7 months. There was an ipsilateral APD but no view of the optic disc because of dense corneal opacification. MRI performed at the time of initial vision loss failed to reveal any abnormal optic nerve enhancement but displayed volume loss of the left prechiasmatic optic nerve (Fig. 4). There was no evidence of compressive lesions along the anterior visual pathways. The patient was last evaluated 19 months after the onset of vision loss without change in her visual function.
The clinical and imaging profiles of our 12 patients generally conform to previously reported cases of RON (1–17,21,23). Our report extends this description by providing a larger series of patients with new clinical information and more extensive imaging documentation.
In common with previous reports, the onset of vision loss in our patients was often acute, but 9 (60%) of the affected eyes suffered progressive visual loss stretching out over many months, leading to finger counting or worse visual acuity in most affected eyes. Both optic nerves were eventually affected in 3 (25%) of the 12 patients. Visual loss was irreversible over a lengthy follow-up (average: 13.5 months). In the single patient who displayed mild improvement in visual acuity over a 4-week period, recovery may have been the result of a learned ability to search within a small remaining central island of fixation (Case E1).
Our patients share the previously documented wide latency span from radiation to first symptoms of RON, with an average of 18 months (12,22) Most reported cases have been exposed to a total dose of radiation of at least 50 Gy, (2–8,10–16,24) but 2 of our patients had received only 30 Gy (Case E4) and 37.5 Gy (Case 1) whole brain radiation. Although these 2 patients received radiation distributed across the entire brain, both displayed prechiasmatic optic nerve enhancement, suggesting that this anatomical area may be relatively vulnerable to delayed radionecrosis (1,2,4–10,12–15). There are no previous reports of whole brain radiation leading to RON. Conventional arteriosclerotic or diabetic risk factors have often been implicated in the development of RON, (12–16,20,22) but such factors could be documented in only half of our cohort.
Nerve fiber bundle visual field defects were characteristic in affected eyes, correlating with the prechiasmal location of the imaging abnormality (Cases 2, E1, E3, E5, and 3). Although these imaging abnormalities were close to the optic chiasm, no cases displayed temporal hemianopic defects as published in earlier reports (2,4,6,10,21–23).
Affected optic discs have been described as appearing normal at the onset of vision loss (4). Indeed, 5 eyes in our cohort had normal-appearing optic discs at examinations conducted within 4 weeks of symptom onset (Cases E1, E4, E5, and 3). In all but 1 of those affected eyes, optic disc pallor developed over the ensuing weeks to months (follow-up was merely 3 weeks in one eye of Case 3). As indicated in previously published cases, among patients evaluated shortly after the onset of vision loss, the affected optic disc may also appear pale (Cases 1–4, 8, 13–14, 16, and 23). Among 11 eyes evaluated within 4 weeks of vision loss, 6 eyes already displayed optic disc pallor (Cases 1, 2, E2, E4, E7, and E8), suggesting that preclinical axonal damage had been present (12).
We found a prechiasmatic imaging abnormality (Cases 1, 2, 4–10, and 12–15) in the affected optic nerves of 13 (87%) of 15 affected eyes in our cohort. This abnormality was most readily apparent on axial projections and was often associated with expansion and abnormal T2 signal of the affected segment. These imaging abnormalities are critical in confirming the clinical diagnosis of RON because a discrete area of enhancement accompanied by expansion or T2-high signal does not typically occur in neoplastic infiltrative optic neuropathies. Compressive optic neuropathy is excluded by the lack of tumor displacing the optic nerve. Inflammatory optic neuropathy could produce this imaging abnormality, but the setting, the lack of visual recovery, and the progressively worsening vision would be clues to RON.
One patient treated for a recurrent pituitary lesion unexpectedly displayed enhancement of the intraorbital segment of the affected optic nerve (Case E8). Intraorbital enhancement has not been frequently reported (11,17), although pathologic evidence of radionecrosis involving the intraorbital optic nerve of an enucleated eye has been described (25). In our only patient without optic nerve enhancement (Case 4), there was prechiasmatic optic nerve volume loss, suggesting that the damage to the blood–brain barrier had resolved after a prolonged period of vision loss. A previous report listed “atrophy” in 3 affected optic nerves without further details (17).
In a small portion of affected eyes, prechiasmatic enhancement was apparent on imaging performed 3–6 weeks before the onset of vision loss, as previously documented (12). Delayed development of enhancement occurred in one affected eye in our series, being evident only 7 months after the onset of vision loss (Case E4).
The duration of enhancement among 9 eyes with follow-up MRI studies (Cases 2, E1–E6) was at least 2 months, averaging 6 months, but extended in one case to at least 17 months (Case E2), consonant with published reports indicating persistent enhancement ranging from 3 to 13 months after vision loss (1,4,9,10,12,13).
In summary, our report further highights the following characteristics of RON:
- Not all affected patients have vasculopathic risk factors.
- Whole brain radiation can be responsible, even at relatively low doses.
- The characteristic MRI finding of RON is a discrete region of enhancement of the prechiasmatic optic nerve, often accompanied by expansion and T2 hyperintensity in the enhancing segment. Intraorbital optic nerve enhancement and volume loss without enhancement may rarely occur as the only imaging abnormalities.
- Enhancement may precede vision loss and may rarely be absent at the time of vision loss, developing only months afterward. Enhancement can persist for at least 17 months after the onset of vision loss.
We acknowledge some limitations of this report. First, a time delay from symptom onset to initial visit in our tertiary care institution may have obscured the time of onset of visual symptoms in some cases. Second, we were unable to obtain details of the radiation protocol in most cases. Despite these limitations, we believe our data contribute to a fuller understanding of RON.
STATEMENT OF AUTHORSHIP
Category 1: a. Conception and design: E. L. Archer, E. A. Liao, and J. D. Trobe; b. Acquisition of data: E. L. Archer, E. A. Liao, and J. D. Trobe; c. Analysis and interpretation of data: E. L. Archer, E. A. Liao, and J. D. Trobe. Category 2: a. Drafting the manuscript: E. L. Archer, E. A. Liao, and J. D. Trobe; b. Revising it for intellectual content: E. L. Archer, E. A. Liao, and J. D. Trobe. Category 3: a. Final approval of the completed manuscript: E. L. Archer, E. A. Liao, and J. D. Trobe.
1. Guy J, Mancuso A, Quisling RG, Beck R, Moster M. Gadolinum-DPTA-enhanced magnetic resonance imaging in optic neuropathies. Ophthalmology. 1990;97:592–600.
2. Zimmerman CF, Schatz NJ, Glaser JS. Magnetic resonance imaging of radiation optic neuropathy. Am J Ophthalmol. 1990;110:389–394.
3. Tachibana O, Yamaguchi N, Yamashima T, Yamashita J. Radiation necrosis of the optic chiasm, optic tract, hypothalamus, and upper pons after radiotherapy for pituitary adenoma, detected by gadolinium-enhanced, T1-weighted magnetic resonance imaging: case report. Neurosurgery. 1990;27:640–643.
4. Guy J, Mancuso A, Beck R, Moster ML, Sedwick LA, Quisling RG, Rhoton AL Jr, Protzko EE, Schiffman J. Radiation-induced optic neuropathy: a magnetic resonance imaging study. J Neurosurg. 1991;74:426–432.
5. Hudgins PA, Newman NJ, Dillon WP, Hoffman JC Jr. Radiation-induced optic neuropathy: characteristic appearances on gadolinium-enhanced MR. AJNR Am J Neuroradiol. 1992;13:235–238.
6. Young WC, Thornton AF, Gebarski SS, Cornblath WT. Radiation-induced optic neuropathy: correlation of MR imaging and radiation dosimetry. Radiology. 1992;185:904–907.
7. Goldsmith BJ, Rosenthal SA, Wara WM, Larson DA. Optic neuropathy after irradiation of meningioma. Radiology. 1992;185:71–76.
8. Borruat FX, Schatz NJ, Glaser JS, Feun LG, Matos L. Visual recovery from radiation-induced optic neuropathy. J Clin Neuroophthalmol. 1993;13:98–101.
9. McClellan RL, Gammal TE, Kline LB. Early bilateral radiation-induced optic neuropathy with follow-up MRI. Neuroradiology. 1995;37:131–133.
10. Piquemal R, Cottier JP, Arsene S, Lioret E, Rospars C, Herbreteau E, Jan M, Renard P. Radiation-induced optic neuropathy 4 years after radiation: report of a case followed up with MRI. Neuroradiology. 1998;40:439–441.
11. Wijers OB, Levendag PC, Luyten GP, Bakker BA, Freling NJ, Klesman-Bradley J, Woudstra E. Radiation-induced bilateral optic neuropathy in cancer of the nasopharynx: case failure analysis and a review of the literature. Strahlenther Onkol. 1999;175:21–27.
12. Lessell S. Friendly fire: neurogenic visual loss from radiation therapy. J Neuroophthalmol. 2004;24:243–250.
13. Garrott H, O'Day J. Optic neuropathy secondary to radiotherapy for nasal melanoma. Letters to the Editor. Clin Exp Ophthalmol. 2004;32:330–333.
14. Danesh-Meyer H, Savino PJ, Sergott RC. Vision loss despite anticoagulation in radiation-induced optic neuropathy. Letters to the Editor. Clin Exp Ophthalmol. 2004;32:333–335.
15. Demizu Y, Murakami M, Miyawaki D, Niwa Y, Akagi T, Sasaki R, Terashima K, Suga D, Kamae I, Hishikawa Y. Analysis of vision loss caused by radiation-induced optic neuropathy after particle therapy for head-and-neck and skull-base tumors adjacent to optic nerves. Int J Radiat Oncol Biol Phys. 2009;75:1487–1492.
16. Ballian N, Androulakis II, Chatzistefanou K, Samara C, Tsiveriotis K, Kaltsas GA. Optic neuropathy following radiotherapy for Cushing's disease: case report and literature review. Hormones. 2010;9:269–273.
17. Zhao Z, Lan Y, Bai S, Shen J, Xiao S, Lv R, Zhang B, Tao E, Liu J. Late-onset radiation-induced optic neuropathy after radiotherapy for nasopharyngeal carcinoma. J Clin Neurosci. 2013;20:702–706.
18. van den Bergh AC, Schoorl MA, Dullaart RPF, van der Vliet AM, Szabo BG, ter Weeme CA, Pott JW. Lack of radiation optic neuropathy in 72 patients treated for pituitary adenoma. J Neuroophthalmol. 2004;24:200–205.
19. Jiang GL, Tucker SL, Guttenberger R, Peters LJ, Morrison WH, Garden AS, Ha CS, Ang KK. Radiation-induced injury to the visual pathway. Radiother Oncol. 1994;30:17–25.
20. Bhandare N, Monroe AT, Morris CG, Bhatti MT, Mendenhall WM. Does altered fractionation influence the risk of radiation-induced optic neuropathy? Int J Radiat Oncol Biol Phys. 2005;62:1070–1077.
21. Kline LB, Kim JY, Ceballos R. Radiation optic neuropathy. Ophthalmology. 1985;92:1118–1126.
22. Danesh-Meyer HV. Radiation-induced optic neuropathy. J Clin Neurosci. 2008;15:95–100.
23. Mayo C, Martel MK, Marks LB, Flickinger J, Nam J, Kirkpatrick J. Radiation dose-volume effects of optic nerves and chiasm. Int J Radiat Oncol Biol Phys. 2010;76:S28–S35.
24. Roden D, Bosley TM, Fowble B, Clark J, Savino PJ, Sergott RC, Schatz NJ. Delayed radiation injury to the retrobulbar optic nerves and chiasm. Ophthalmology. 1990;97:346–351.
25. Ross HS, Rosenberg S, Friedman AH. Delayed radionecrosis of the optic nerve. Am J Ophthalmol. 1973;76:683–686.