Optic pathway gliomas are low-grade tumors that arise anywhere along the visual system, from the optic nerve behind the eye to the chiasm and the radiation of fibers carrying signals toward the visual cortex. They account for roughly three to five percent of all pediatric brain tumors and are most commonly pilocytic astrocytomas, slow-growing World Health Organization Grade I lesions with excellent overall survival. The catch is morbidity rather than mortality. A large share of affected children carry a diagnosis of neurofibromatosis type 1, and among children with this genetic condition, approximately fifteen to twenty percent will develop these tumors at some point. Progressive visual loss is the most common and most debilitating consequence, whether the tumor arises sporadically or in the setting of NF1, and it can profoundly disrupt a child’s neurodevelopment and quality of life.

Current therapy is aimed squarely at tumor control rather than at rescuing vision. Under established pediatric protocols, younger children typically receive chemotherapy while radiotherapy is generally reserved for those older than eight years. What has been missing is any disease-specific treatment for the visual dysfunction itself. The biological target that the Italian research team behind the new study focused on is the retinal ganglion cell, the neuron whose long axon forms the optic nerve. Children with optic pathway gliomas often show thinning of the retinal nerve fiber layer, the bundled axons of these cells, driven by apoptosis, a form of programmed cell death. In NF1-associated tumors, this cell death may stem not only from mechanical compression by the tumor mass but also from a functional deficit caused by loss of the neurofibromin protein, which triggers hyperactivation of the Ras signaling pathway, phosphorylation of atypical protein kinase C, and a drop in intracellular cyclic AMP. Crucially, work in rodent models has shown that ganglion cell apoptosis is preceded by a phase of axonal damage, opening a therapeutic window during which intervention might still prevent irreversible cell loss.

Nerve growth factor, the prototypical neurotrophin, has emerged as a leading candidate for exploiting that window. The molecule signals through two classes of receptors on cell surfaces: the high-affinity tyrosine kinase receptor TrkA and the low-affinity p75 neurotrophin receptor. Both are expressed throughout the visual system, from retina to cortex, and on retinal ganglion cells themselves, where endogenous NGF normally serves as a survival factor. Preclinical studies have repeatedly shown that supplying extra NGF can attenuate ganglion cell death after optic nerve injury, and more recent experiments with optic nerve crush models demonstrated that both recombinant human NGF injections and eye drop formulations can promote axonal regrowth, apparently by interfering with apoptotic machinery, including inhibition of the NogoA/p75NTR complex and suppression of Rho/ROCK2 signaling.

Those laboratory findings set the stage for the original clinical trial that the new report follows up. In that double-blind, randomized, placebo-controlled phase II study, eighteen children and young adults with optic pathway gliomas, aged between two and twenty-three years and with stable disease but severe visual impairment, were assigned to receive murine NGF as eye drops at a dose of 0.5 milligrams three times daily for ten consecutive days, or a matched placebo. The treatment course was strikingly short, yet it produced statistically significant improvements in electrophysiological measures of retinal and optic pathway function compared with placebo, along with a notable enlargement of the visual field in treated patients where reliable measurement was possible. No severe ocular or systemic adverse events were recorded, and MRI showed no tumor progression in either group.

The new publication extends that story to four years. Fifteen of the original eighteen participants, nine from the NGF arm and six from the placebo arm, agreed to continued clinical and neuroradiological monitoring, and importantly, no additional NGF, chemotherapy, or radiotherapy was given during the extended follow-up. Every six months the patients underwent general clinical and neuro-ophthalmological examinations, including visual acuity testing, Goldmann perimetry, flash visual evoked potentials, and the photopic negative response of the electroretinogram, a signal that specifically reflects the function of retinal ganglion cells. Brain MRI was performed annually and assessed by board-certified pediatric neuroradiologists using the Response Assessment in Neuro-Oncology criteria for low-grade glioma. Eleven of the fifteen patients had NF1, the median age at enrollment was 12.6 years, and most had previously received chemotherapy.

The headline result is stability. Comparing baseline and end-of-follow-up values, the researchers found no statistically significant change in median visual acuity, in the amplitude of visual evoked potentials, in the photopic negative response amplitude or timing, or in the radius of the visual field. At the level of individual eyes, only five of thirty showed any change in visual acuity over four years, and just three of those crossed the pre-specified threshold for clinical meaningfulness, a shift of at least 0.3 logMAR units, roughly fifteen letters on an eye chart. On imaging, stable disease was confirmed in all fifteen patients throughout the entire observation period. Two children showed minor tumor volume increases of about ten percent, but these fell short of the criteria for progressive disease and were not accompanied by clinical deterioration.

One intriguing wrinkle deserves attention. When the investigators re-analyzed individual visual field data, they found that the change from baseline differed significantly between the original treatment groups, both at three months and at the four-year endpoint: every patient originally assigned to NGF showed an increase or no change in visual field radius, while every placebo-assigned patient showed a decrease. Yet the authors are careful, and rightly so. Baseline visual field radius was itself substantially larger in the placebo group, a statistical imbalance that raises the specter of regression to the mean, and Goldmann perimetry is known to have within-visit test-retest variability that can approach twenty percent, with reliability that depends heavily on a child’s age and cooperation. The team explicitly labels any interpretation of a persistent treatment effect on the visual field as hypothesis-generating rather than confirmatory, a level of caution that stands out in a field often tempted by optimistic readings of small datasets.

Perhaps the most consequential question surrounding growth factor therapy in a tumor context is whether flooding the visual system with a pro-survival molecule might feed the neoplastic glial cells themselves. Here the evidence points the other way. Pediatric pilocytic astrocytomas characteristically downregulate TrkA, the receptor whose activation is associated with tumor cell proliferation, while upregulating the pro-apoptotic p75NTR receptor. Laboratory studies have shown that NGF inhibits the proliferation of rat glioma cells and drives them toward a more differentiated state, and more recent work demonstrated that ectopic NGF prevents proliferation in pediatric low-grade glioma cell lines by inducing cellular senescence. The four years of stable MRI scans in this cohort, spanning both treated and placebo patients, provide clinical reassurance consistent with that biology, although the authors note that the indolent natural history of these tumors is itself a plausible explanation for the observed stability.

The study’s limitations are acknowledged candidly. Fifteen patients is a small cohort, a consequence of the rarity of the condition and the pragmatic constraints of following up a completed trial, and the observational extension lacked the randomization that gave the original phase II study its power. Both eyes of most patients were included in analyses despite not being fully independent observations, and optical coherence tomography, a sensitive structural marker of ganglion cell integrity, could not be systematically repeated in young children for whom reliable scans often require sedation. What the study does establish is the sustained safety and tolerability of a short course of topical NGF in children, with no late ocular adverse events, no corneal sensitivity changes, and no evidence of tumor stimulation over four years. Whether nerve growth factor eye drops can genuinely preserve vision in children facing optic pathway glioma-related blindness will now depend on larger, adequately powered, prospective randomized trials with standardized outcome measures, for which this four-year safety record removes a significant barrier.

Subject of Research: Long-term safety and visual outcomes of topical nerve growth factor treatment in children with optic pathway gliomas

Article Title: Long-term outcomes after a randomized phase II trial of nerve growth factor eye drops for optic pathway gliomas: four-year follow-up findings in children

Article References: Ruggiero, A., Attinà, G., Placidi, G., D’Agostino, E., Amato, A., Verdolotti, T., Mastrangelo, S., & Falsini, B. (2026). Long-term outcomes after a randomized phase II trial of nerve growth factor eye drops for optic pathway gliomas: four-year follow-up findings in children. Journal of Neuro-Oncology, 179(2), Article 67. https://doi.org/10.1007/s11060-026-05777-z

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05777-z

Keywords: nerve growth factor, optic pathway glioma, neurofibromatosis type 1, pediatric neuro-oncology, retinal ganglion cells, neuroprotection, visual evoked potentials, pilocytic astrocytoma, eye drops, phase II trial, visual field, low-grade glioma

Pilocytic astrocytoma is a WHO grade 1 glioma, a tumor class that behaves far more indolently than the aggressive glioblastomas that dominate public awareness of brain cancer. These tumors are driven largely by alterations in the MAPK signaling pathway and, in children, are often cured by surgery alone. In adults, however, the disease is less well characterized. Adult tumors tend to arise in different locations, may be less amenable to complete resection, and can recur years or even decades after an initial operation. When they sit in the brainstem, the optic pathway, the hypothalamus, or deep midline ventricular structures, the risk of repeated surgery becomes unacceptable, leaving radiotherapy as the principal local option.

The Heidelberg team, led by Fabian J. K. Allmendinger and colleagues, assembled a retrospective cohort of fourteen adults treated with intensity-modulated proton radiotherapy between 2011 and 2025. The patients were young, with a median age of 23.5 years at treatment, and their tumors occupied some of the most anatomically treacherous territory in the brain: the cerebellum, the frontal and temporal lobes, the third and fourth ventricles, the suprasellar compartment, the mesencephalon, and the pineal region. Ten of the fourteen had undergone prior surgery, and in nine of those cases the operation had achieved only subtotal removal. For ten patients, proton therapy was salvage treatment for recurrent disease; for four, it was the primary local therapy when surgery was judged too risky.

The technical premise of proton therapy is elegantly simple. Unlike conventional photon beams, which deposit radiation along their entire path through tissue and exit on the far side of the target, protons carry a finite range. They release most of their energy in a sharply defined burst, the Bragg peak, at a depth determined by their initial energy, and deliver essentially nothing beyond it. By scanning pencil-thin proton beams across the tumor volume, clinicians can conform the dose to irregular targets while dramatically reducing the radiation bath received by uninvolved brain, the temporal lobes, the hippocampi, the optic apparatus, the hypothalamic-pituitary axis, and cerebral blood vessels. For patients expected to live for decades, that reduction in integral dose is not a cosmetic advantage; it is the difference between preserving and progressively eroding neurocognitive, endocrine, and visual function over a lifetime.

The prescribed dose in the cohort was a median of 54.0 Gy(RBE), delivered in roughly 29 fractions, with planning that co-registered treatment CT scans with contrast-enhanced T1-weighted and T2/FLAIR MRI and applied QUANTEC-based constraints for organs at risk. What followed was, by the standards of brain tumor therapy, remarkable. At last follow-up, bidirectional T2-weighted tumor measurements showed complete regression in three patients, partial response of at least fifty percent in seven, and a minor response in two. Only two patients experienced progressive disease. Perhaps most intriguingly, shrinkage frequently continued beyond the twelve-month landmark: ten of the twelve non-progressive patients kept improving after the standardized one-year assessment, a pattern consistent with the slow growth kinetics of grade 1 tumors and a warning against prematurely declaring treatment failure.

Survival outcomes matched the radiographic enthusiasm. Estimated overall survival was one hundred percent at both three and five years, with no deaths recorded within the first five years after therapy. Progression-free survival stood at 83.3 percent at both timepoints, and most progression events occurred early, meaning that patients who remained disease-free in the initial years tended to stay that way. Median clinical follow-up extended to nearly ten years, with vital status confirmed through national registry data, lending unusual depth to the survival analysis for a cohort of this size.

Toxicity was predominantly low grade. During treatment, the most common acute effects were focal alopecia in 64 percent of patients, fatigue and headache in 43 percent each, and radiodermatitis in 29 percent, and nearly all of these resolved in follow-up. Alopecia and skin reactions vanished entirely beyond the first months. Fatigue was the most persistent complaint, affecting three patients across all follow-up intervals, while mild cognitive impairment appeared in two. One patient with pre-existing postoperative visual deficits developed late grade 1 visual impairment, though optic nerve dose constraints had been met and imaging showed no optic neuropathy. Critically, no grade 4 or grade 5 events occurred at any point.

The study also shines a light on a subtle diagnostic trap known as radiation-induced contrast enhancement, or RICE. Within months to years after proton therapy, new areas of contrast uptake can appear on MRI within or near the irradiated volume, mimicking tumor progression on a scan. In this cohort, three patients developed RICE, yielding a three- and five-year RICE-free probability of 76.2 percent, with all events occurring in the early post-treatment period. One patient was initially misclassified as having progressed and was started on temozolomide chemotherapy, which had to be abandoned after a single cycle due to blood toxicity; retrospective review showed the enhancing lesion was radiation injury, not tumor, and bevacizumab sufficed. The authors emphasize that distinguishing RICE from true progression, using imaging morphology, temporal evolution, spatial relation to the dose distribution, and clinical course, is essential to avoid unnecessary chemotherapy in patients who otherwise face decades of life.

The findings arrive amid an evolving treatment landscape. Current EANO-EURACAN-SNO guidelines recommend maximal safe resection when feasible, with molecular testing for BRAF and MAPK alterations, and acknowledge reoperation, radiotherapy, and targeted agents such as BRAF and MEK inhibitors as options for recurrent or unresectable tumors. Recent multicenter data show chemotherapy with temozolomide achieves a median progression-free survival of only about twenty months in adult pilocytic astrocytoma, a figure that looks modest against the durable control seen here. The Heidelberg authors are careful, however, not to overreach: the study was retrospective, involved only fourteen patients, lacked a photon comparator, and carried incomplete molecular annotation, so it cannot prove that protons outperform photons, only that they deliver excellent control with acceptable morbidity in a selected, anatomically complex population.

Even with those caveats, the message for patients and clinicians is consequential. For adults facing a progressive pilocytic astrocytoma that surgery cannot safely reach, proton radiotherapy now has dedicated, adult-specific evidence behind it: near-universal radiographic response, complete five-year survival, and side effects that were mostly transient and mild. Because prospective trials are unlikely in a disease this rare, single-center cohorts like this one carry disproportionate weight. The study also reinforces a broader principle of modern radiation oncology, that in long-surviving patients with benign-behaving tumors, the true measure of a treatment is not only whether it controls the disease, but what quality of brain, and of life, it leaves behind ten, twenty, and thirty years later.

Subject of Research: Proton radiotherapy for progressive or surgically unfavorable adult pilocytic astrocytoma

Article Title: Efficacy and safety of proton radiotherapy for progressive or surgically unfavorable adult pilocytic astrocytoma: a single-center experience

Article References: Allmendinger, F. J. K., Cherniienko, Y., Lischalk, J., Walter, J., Mose, L., Wickert, R., Deng, M., Regnery, S., Wessel, L., Kozyra, K., Tessonnier, T., Krieg, S., Debus, J., König, L., & Adena-Eichkorn, T. (2026). Efficacy and safety of proton radiotherapy for progressive or surgically unfavorable adult pilocytic astrocytoma: a single-center experience. Journal of Neuro-Oncology, 179(3), Article 89. https://doi.org/10.1007/s11060-026-05798-8

Image Credits: AI Generated

DOI: 10.1007/s11060-026-05798-8

Keywords: proton radiotherapy, pilocytic astrocytoma, adult glioma, radiation oncology, RICE, tumor control, low-grade glioma, Bragg peak, neuro-oncology, radiation toxicity, MAPK pathway, brain tumors