Retrospective analysis of fractionated intensity-modulated radiotherapy (IMRT) in the interdisciplinary management of primary optic nerve sheath meningiomas

As ONSM are benign tumors and are sometimes stable in size and symptoms over years, one crucial step in the treatment of these patients is the selection of patients to be treated and the timing of therapy. Patients with clinical suspicion of ONSM based on slow decline of vision or incidental imaging finding underwent detailed neuro-ophthalmological examination including best corrected visual acuity, visual field, pupil function, ocular motility, slitlamp, ophthalmoscopy (photography) and OCT. These findings were correlated with gadolinium-enhanced MR imaging. Typical imaging findings for ONSM include sheath-like contrast enhancement around the optic nerve or an intraorbital mass with close spatial relationship to the optic nerve and gadolinium enhancement. Typical ophthalmological findings were reduced visual acuity, relative afferent pupillary defect, visual field loss, normal, swollen or pale optic disc. In case of normal visual acuity and visual field, patients underwent active surveillance with three-monthly ophthalmological assessment and MRI follow-up. Patients with mild or moderate symptoms were only scheduled for treatment after documented progressive loss of vision. Radiotherapy was planned in cases with typical findings in somatostatin-receptor-analoga PET-CT. Typically, ONSM show high uptake of somatostatin receptor analogon compared to normal tissue (except for the hypophysis). In the case of atypical findings, biopsy was considered. For patients with severe symptoms, the integrity of retinal neural structures was assessed by OCT. OCT is helpful in treatment decision making, although interpretation may be difficult. Oedema of the nerve fibre layer can be misinterpreted as normal thickness of nerve fibre and ganglion cell layer although a considerable ganglion cell and axonal loss may be present. If nerve fibre and ganglion cell layer thickness is reduced irreversible optic nerve damage is confirmed. A careful interpretation of morphological findings and correlation with function and history is necessary.

Absence of ganglion cell layer and nerve fibre loss usually led to the indication for radiotherapy, even in cases with very poor vision. Therapy was initiated in a timely manner because of the chance for recovering of visual function. Patients with neural degeneration and no useful visual function underwent active surveillance for the visual function of the contralateral eye and were scheduled for radiotherapy or surgery in the case of threatened or declining visual function contralaterally (Fig. 1).

In our patient cohort 22 patients (85%) had been diagnosed during the workup for substantial vision loss and were referred for radiotherapy after diagnosis. Three patients underwent active surveillance for 1.3 years, 2.5 years and 15 years respectively before undergoing radiotherapy following first signs of declining visual function. One patient had undergone three surgical procedures over 15 years prior to radiotherapy, which was started at progression of visual symptoms.

Equivalent dose in 2 Gy dose per fraction (EQD2) was calculated with the linear quadratic model for prescribed doses for optic nerve and PTV51 / PTV54. For tumor control (meningioma) α/β was considered to be 3.76 Gy [28]. Thus, the prescribed doses are equivalent to EQD2_3.76 of 48.34 Gy and 52.13 Gy for PTV51 and PTV54, respectively. A maximum dose to the optic nerve of 54 Gy in 1.8 Gy fractions corresponds to 51.30 Gy EQD2₂ with α/β = 2 Gy as described for the optic chiasm [29].

Mean target volumes for radiotherapy planning were 4.09 ± 0.70 cm³, 7.80 ± 1.06 cm³ and 15.78 ± 1.66 cm³ for GTV, PTV54 and PTV51, respectively. A typical example of PET findings and radiation plan is depicted in Fig. 2. As our prescribed dose of 54 Gy was rather high and to optimize functional outcome, the main planning objective was not to exceed the tolerance dose to the optic nerve or other organs at risk (Tbl. 2). Maximal physical dose to the optic nerve was 53.92 ± 0.05 Gy for all patients, corresponding to EQD2₂ of 51.18 ± 0.07 Gy. D2 of the optic nerve did not exceed 54 Gy (physical dose) in any patient.

As EUD of PTV51 and PTV54 was not used as planning objective, but might be a measure of plan quality and tumor control, we evaluated EUDs for both volumes. EUD was 52.27 ± 0.08 Gy, corresponding to 96.8% of the prescribed dose for PTV54 and 49.72 ± 0.19 Gy, corresponding to 97.5% of the prescribed dose for PTV51. As shown in Fig. 3, plotting EUD of PTV51 and PTV54 against start of treatment, resulted in increasing EUDs for the first patients, stabilizing after a learning curve. Comparison of EUDs for the 10 first treated patients (Cohort A) compared to the other 16 (Cohort B) showed highly significantly higher EUDs for patients treated after the learning curve (Fig. 3). At the same time, Dmax for the optic nerve was significantly lower in cohort B with 53.82 ± 0.04 Gy versus 54.07 ± 0.09 in cohort A (p = 0.01, data not shown).

At time of treatment start patients had significant functional ophthalmological deficits with a median visual acuity of 0.45 (0.01–1.5) and a median visual field loss of 65% (4–100%). These functional parameters improved significantly at first control (up to three months) after therapy to 0.70 (0.01–1.6; p = 0.04) and 42.5% (0–100%; p = 0.001) respectively (Fig. 4). Notably, respecting the above-mentioned selection criteria for radiotherapy, even patients with severely impaired function at start of treatment had a reasonable chance for improvement. Seven patients with a visual acuity < 0.1 before treatment (mean 0.06 ± 0.02) showed a significant improvement to 0.43 ± 0.14, p = 0.04. Nine patients had visual field loss > 70% before radiotherapy (mean 84.7% ± 3.6%), which showed a non-significant reduction to 66.1% ± 10.4%, p = 0.11.

Changes in visual field loss and visual acuity showed a moderate correlation for the evaluated patients. Data was available for 24 patients, one additional patient was excluded due to complete blindness after radiotherapy (R² = 0.41, Fig. 4). Ophthalmological outcome did not correlate with age, volume of PTV54 or EUD of PTV51 and PTV54 (data not shown). Patients with sheathlike growth of ONSM showed significantly greater reduction of visual field deficits compared to those with fusiform tumors (23.2% ± 5.6% vs. 2.1% ± 3.1%, p = 0.02).

No local progression was observed in follow-up MR imaging in any patient. Long term results of visual acuity and visual field loss were assessable for all patients in follow-up using the same method as for the first examination after radiotherapy. Results are shown in Fig. 5. Improved visual acuity was recorded for 9 patients, 12 patients showed stable function, in 5 patients visual acuity declined. Visual field was improved in 14 patients, stable in 8 patients and worsened in 2 patients (data missing for two patients). The two patients with increasing loss of visual field also had severely impaired visual acuity, thus the results in these patients might have been compromised by overall visual function. In total, 16 patients had improved overall visual function, 6 were stable and 4 patients declined (1 patient rated as stable with a decline in visual acuity and improved visual field).

Possible prognostic factors tested for the influence on long term results of visual function included age at start of radiotherapy (classified as above or below the median age of 48.3 years), size of PTV above or below the median of 4.1 cm³, sheathlike versus fusiform growth and involvement of the optic canal. For statistical analysis visual acuity and visual field was classified in decreased / stable versus improved and decreased versus stable / improved, respectively. The only factor significantly influencing decreased visual function was higher patient age at start of radiotherapy (Fig. 5). Improvement of visual field was significantly more frequent in patients with sheathlike tumor growth compared to fusiform tumors (Fig. 5). Of the five patients with a decline in visual acuity after treatment one had additional facial nerve palsy with corneal erosion, one had pre-existing localized systemic sclerosis. One patient showing typical signs of radiation induced optic neuropathy (RION) had pre-existing auto-antibodies to aquaporin 4. Steroid treatment did not improve visual function. Patients with decline of visual acuity are summarized and discussed in Table 3. In total, after interdisciplinary discussion, findings in two patients were classified as radiation toxicities (2/26, 8%). Interdisciplinary discussion involved additional diagnosis, ophthalmologic findings, time course between diagnosis, radiotherapy and decline in visual function as well as radiation dose distribution.