Charged particles can act on and interfere with the mechanisms involved in visual information processing. Evidence in this respect comes from early reports by space crew members and later by patients undergoing heavy ions treatment of tumors of the eye or skull who reported phosphenes, i.e., unstructured positive visual phenomena (light flashes) occurring in isolation and in the absence of alterations of consciousness. Structured or complex positive phenomena (photopsias, kinetopsia, visual distortions or hallucinations) have not been described. Patients undergoing proton radiotherapy occasionally reported non-visual sensory percepts, often associated with phosphenes and with the brain regions receiving the highest doses corresponding to the anatomical structures associated with each type of percept (1–4). Non-visual percepts, either isolated or concomitant to phosphenes in multisensory perception, appear less frequent or exceptional in space (2, 3, 5–7). The origin of phosphenes in radiation activating structures of the retina or the optic nerve and some peculiar sensitivity of the visual system to charged particles have been hypothesized. Comparable percepts were reproduced in early human experiments in accelerator (8–14). The mechanisms of generation are conceivably complex and remain to be investigated in full detail. Occurrence, potential interference with CNS functions, and clinical relevance with respect to the pathophysiological conditions in which phosphenes are observable remain underestimated.

Phosphenes are uncommon but can occur spontaneously also in a variety of physiological conditions such as transient deformation of the eye, accommodation, saccades, vivid lights, rhythmic light stimulation, etc. (15). The mechanical deformation of the eyeball, for instance, causes phosphenes by activating both center-on and center-off retinal ganglion cells (16). Excitation or the damage or functional impairment at any portion of the visual system (retina, afferent pathways or cortex) can result in positive visual phenomena, including phosphenes. These are reported in severe visual loss and by subjects using neuroactive drugs (mostly alcohol and hallucinogens), and occasionally in other disorders of the CNS (15, 17). The incidence of simple phosphenes in epilepsy is low (<1%) and limited to the focal involvement of visual cortex. They are by contrast quite common in the aura preceding pain in migraine attacks. Most studies agree in estimating the occurrence of phosphenes in a portion up to ~90% of adult migraneurs, irrespective of their experiencing visual auras with or without migraine. Functional testing of the visual cortex by transient magnetic stimulation (TMS) has documented a neuronal hyperexcitability in migraine with aura compared to both migraineurs without it and controls (18–23). In contrast, the threshold of TMS-induced phosphenes did not differ in a study comparing migraine without aura, migraine with aura, and control groups (68 ± 9.5% vs. 75 ± 12%, vs. 80 ± 11%, respectively) (24). The visual aura in migraines is associated with reduced cerebral blood flow and increased BOLD signals (25–27). Migraine, migraine aura without pain and headache are reportedly common during space traveling (with unclear incidence) and have been mimicked in studies on Earth (28, 29).

Colors and the perception of movement can result from the stimulation of basal occipital areas in the fusiform and lingual gyri and of the parieto-occipital and basal temporo-occipital junctions, respectively. Activation/damage of specific portions of the visual system has been correlated to structured or complex positive visual phenomena (hallucination, etc.), but no topographic selectivity (or significantly higher incidence) has been found thus far for simple percepts like phosphenes (15, 17, 30, 31). Bilateral cortical activation covering the entire visual cortices was observed by fMRI in patients with Leber’s hereditary neuropathy of optic nerve reporting induced phosphenes (32). Phosphenes have been reported during hypothalamic deep brain stimulation (33).

Retinal photoreceptors and neurons in the retina and cortex are sensitive to ionizing agents and can respond to their action with transient functional changes. Phosphenes have been reported by subjects exposed to X-rays, with the electroretinogram amplitude correlating with the X-rays intensity; an action on the rhodopsine proteic component opsin starting a cascade of reactions leading to a visual sensation has been suggested (49, 50). Focal electrical stimulation of striate visual cortices V₁ and V₂ during neurosurgery evoked simple phosphenes, while stimulation of the extrastriate areas or temporal regions evoked complex visual phenomena (15, 51, 52). In early studies, electrical stimulation of discrete points of the human visual cortex produced corresponding punctuated sensation of light in both sighted and blind subjects. Systematic investigation on the size, luminosity and position in visual space of phosphenes induced by discrete electrical stimulation of the retina and optic nerve has been instrumental in the tentative development of prostheses for the blind; electric stimulation is thought to evoke phosphenes by opening voltage-sensitive ion channels and by-passing the chemically gated channels in stimulated neurons (53–58). Magnetic stimulation of the retina or cortex induces phosphenes via complex mechanisms of action and depending on the stimulate position and functional status (59–62).

A suggested alternative mechanism of generation is a local disinhibition of neuronal structures which can result in increased neuronal excitability due to deafferentation (the release phenomenon). In this regard, any local cause of deafferentation within the visual system able to interfere with information processing via functional impairment (e.g., miscoded neuronal function) or due to anatomic damage (macular degeneration, glaucoma, cataract, macular holes and other retinopathies, lesions of visual pathways or cortices) can result in a release mechanism and in the perception of positive visual phenomena (14, 63). Retinal or cortical damage has not been observed in in vivo rodent studies with direct retinal stimulation below the Cerenkov threshold (47). However, detrimental effects on synaptic density and myelination in response to exposure to high-energy charged particles have been reported (64). Significant retinal damage (apoptosis in vascular endothelial cells, significant changes in regulated protein expression, cellular structure, immune response and metabolic function) have been observed in mice after 35 days in space and were reduced by artificial gravity (65).

The findings collectively suggest some action of heavy ions on photoreceptors and a parallel activation of postsynaptic currents in inner retina, with modalities not necessarily comparable to those of light, but nevertheless able to trigger the retinal cascade to cortical activation (48). Studies on patients with retinitis pigmentosa and mutant mice have documented activation of the visual cortex in response to light also after photoreceptors extensive damage (49, 66–68). Subsets of intrinsically photosensitive retinal ganglion cells have been identified in animals. In primates, these cells project to the lateral geniculate nucleus and reportedly merge with the retinal pathways to cortex that transfer visual information to be processed into visual images (69–71); some role in the generation of phosphenes has been suggested (38). Transient depression of neuronal activity (often referred to as spreading depression in rodent models) can be associated to increased released of K⁺, propagation of Ca²⁺ between glia and neuronal cells or glutamate action (15, 72) and has been observed in cortex, hippocampus, and in superior colliculus and retina.

Evidence suggests that photon emission can mediate in networks function as an additional signal enabling cell-to-cell communication and coupling depending on the neuronal physiological state (73). Bovine rhodopsin irradiated with ¹²C ions in suspension was activated by ions as it was by light. The process was mediated by lipid peroxidation and successive chemiluminescence and the bleaching process proved reversible under these experimental conditions. These findings indicate a ¹²C ions-induced recombination of radicals possibly responsible for the release of photons with subsequent bleaching of rhodopsin and the triggering of mechanisms generating phosphenes (74, 75).

Different mechanisms are suggested to mediate in the production of phosphenes and are individually or collectively eligible as generating processes (38). However, the interest on the issue has encompassed different scientific contexts, with varying technological background, rationale, and experimental approaches. Research on the topic has been fragmentary rather than systematic, and the available evidence may suggest limited depth of investigation. The different radiation and conditions of irradiation causing phosphenes propose the transfer of energy as a conceivable common mechanism activating photoreceptors and/or neuronal cells and their interaction in the retinal cascade. Interactions between the charged particles action and the retina and/or the CNS functional status of the stimulated cells or network are also conceivable. The Cerenkov radiation measured in particle accelerator experiments was compatible with the estimated threshold sensitivity to photons of the retina photoreceptors, with mechanisms of generation mimicking light (10, 12). Indirect Cerenkov light due to nuclear interactions/radioactive decay or direct activation/radical excess within the retina are also possible. Models and experiments nevertheless estimated a threshold number of ionizations per sensitive volume below the Cerenkov threshold, indicating effects on the rod outer segment or photochemical molecules at the energies compatible with those used for ocular proton therapy (11) or able to activate network functions (1, 2, 42, 63). A direct action of ionizing particles on photoreceptors or neural tissues in the retina appears compelling but is not conclusive. Direct effect on brain neurons remains undocumented and in vivo and in vitro experiments implicate mechanisms of action on the visual system that could be more complex than hypothesized and depend to a relevant extent on the activated cells or structures. The bilateral cortical activation covering the entire visual cortices observed by fMRI when patients reported phosphenes for instance indicates a relevant impact on visual function mediated by magnification factors (16, 32). Several subjective characteristics of the phosphenes reported in space, in in vivo/in vitro experiments, and during hadron therapy are similar, but a scale problem result of the wide range of the ions number needed to generate a subjective phosphene or obtain a chemical or electrophysiological response. Small numbers of particles may elicit a subjective phosphene in space, although not every single ion can, with and estimated efficiency in the order of 10⁻²–10⁻³. The threshold in vivo animal models is ~10³ ions; higher flux appears necessary both in in vitro models and in patients undergoing hadron therapies (2). The functional status of the stimulated areas and network are also conceivable. For instance, about 25% of subjects do not report perceiving phosphenes upon magnetic occipital stimulation (76). The estimated radiation release in hadron therapy accounts for both normal tissues and target tumors, so that single pulse delivery or full irradiation are to be considered in in vivo and in vitro studies, respectively.

Evidence from laboratory tests suggests that the phosphenes reported by astronauts may reflect transient functional impairment. Their relevance as potential early markers of later biological tissue damage remain largely unknown and would add health hazard to the effects of microgravity, isolation, disruption of circadian rhythm, impaired sleep dynamics, and hypercapnia associated with space travel [e.g., (77)]. Long-duration space travels nevertheless impose conditions that are known to interfere with and may have detrimental effects on the visual system and CNS functions in humans also irrespective of heavy ions exposure (7, 78–82). The limited number of particles triggering phosphenes in space may indicate a limited risk or reinforce the hypothesis of higher sensitivity of the visual system to radiation direct/indirect action due to the prolonged permanence in microgravity (73), and combined effects of particles and microgravity in space appear therefore peculiarly probable. Strategies to counterbalance the effects of microgravity have been suggested and methods to shield astronauts from cosmic radiation have been devised and are being tested (7, 83–87). On the contrary, the effects of therapeutic irradiation on the photoreceptors and nervous system and the mechanisms originating phosphenes in patients have been largely overlooked thus far, most likely because of the low priority of these phenomena in the estimated cost–benefit ratio of a therapeutic approach with limited or no alternatives. Acute side-effects on encephalic nervous tissues have been described in patients during and immediately after ¹²C treatments of skull tumors (41), but long-term monitoring is not yet available and the possible relevance of phosphenes as early markers of visual impairment has not been and should be investigated in full detail. The action of particles on the visual system may be deemed of limited relevance in the extreme conditions of space travel, but further research may devise application in therapeutic procedures that are becoming routinary and better tolerated by patients. The correlation of phosphenes and their conditions of occurrence in hadron therapies with quantitative neurophysiological measures may provide additional information on the pathophysiology of charged particles effects on CNS tissues. More significant could be this information in further improving the procedures of hadron treatments and the prediction of outcome, and in reducing the patients’ discomfort (88).