To assess the relationship between the phenotype of the “visual snow” syndrome, comorbid migraine, and typical migraine aura on a clinical basis and using functional brain imaging.
To assess the relationship between the phenotype of the “visual snow” syndrome, comorbid migraine, and typical migraine aura on a clinical basis and using functional brain imaging.
Patients with “visual snow” suffer from continuous TV-static-like tiny flickering dots in the entire visual field. Most patients describe a syndrome with additional visual symptoms of the following categories: palinopsia (“afterimages” and “trailing”), entopic phenomena arising from the optic apparatus itself (floaters, blue field entoptic phenomenon, photopsia, self-light of the eye), photophobia, nyctalopia (impaired night vision), as well as the non-visual symptom tinnitus. The high prevalence of migraine and typical migraine aura in this population has led to the assumption that “visual snow” is caused by persistent migraine aura. Due to the lack of objective measures, alternative diagnoses are malingering or a psychogenic disorder.
(1) The prevalence of additional visual symptoms, tinnitus, and comorbid migraine as well as typical migraine aura was assessed in a prospective semi-structured telephone interview of patients with “visual snow.” Correlations were calculated using standard statistics with P < .05 being considered statistically significant. (2) Areas with increased brain metabolism in a group of “visual snow” patients in comparison to healthy controls were identified using [18F]-2-fluoro-2-deoxy-D-glucose positron emission tomography and statistical parametric mapping (SPM8 with whole brain analysis; statistical significance was defined by P < .001 uncorrected for multiple comparisons).
(1) Of 120 patients with “visual snow,” 70 patients also had migraine and 37 had typical migraine aura. Having comorbid migraine was associated with an increased likelihood of having palinopsia (odds ratio [OR] 2.8; P = .04 for “afterimages” and OR 2.6; P = .01 for “trailing”), spontaneous photopsia (OR 2.9; P = .004), photophobia (OR 3.2; P = .005), nyctalopia (OR 2.7; P = .01), and tinnitus (OR 2.9; P = .006). Typical migraine aura was associated with an increased likelihood of spontaneous photopsia (OR 2.4; P = .04). (2) After adjusting for typical migraine aura, comparison of 17 “visual snow” patients with 17 age and gender matched controls showed brain hypermetabolism in the right lingual gyrus (Montreal Neurological Institute coordinates 16-78-5; kE = 101; ZE = 3.41; P < .001) and the left cerebellar anterior lobe adjacent to the left lingual gyrus (Montreal Neurological Institute coordinates -12-62-9; kE = 152; ZE = 3.28; P = .001).
—Comorbid migraine aggravates the clinical phenotype of the “visual snow” syndrome by worsening some of the additional visual symptoms and tinnitus. This might bias studies on “visual snow” by migraineurs offering study participation more likely than non-migraineurs due to a more severe clinical presentation. The independence of entoptic phenomena from comorbid migraine indicates “visual snow” is the main determinant. The hypermetabolic lingual gyrus confirms a brain dysfunction in patients with “visual snow.” The metabolic pattern differs from interictal migraine with some similarities to migrainous photophobia. The findings support the view that “visual snow,” migraine, and typical migraine aura are distinct syndromes with shared pathophysiological mechanisms that need to be addressed in order to develop rational treatment strategies for this disabling condition.
[18F]-2-fluoro-2-deoxy-D-glucose positron emission tomography
Patients with “visual snow” (VS) describe a visual disturbance that consists of tiny dynamically flickering dots in the entire visual field resembling the “static” or “snow” of a badly tuned analogue television. The symptoms are continuous and can persist over years. Persistent visual disturbance is mentioned sporadically in the literature without larger systematic studies.1-3 Patients are often diagnosed as having persistent migraine aura, malingering, or psychogenic disorder because objective measures for the condition are not available to date. A recent study of a substantial cohort of subjects with VS confirmed that the visual disturbance is often associated with migraine and migraine aura. However, not every patient with VS has a history of migraine. Further, VS starts only rarely with migraine aura, and the phenotypical description as well as the clinical course of VS by no means resembles typical migraine aura, which is in general homonymous, often presents with moving zigzag lines, and typically lasts less than 60 minutes. This suggests that VS is a unique condition different from migraine aura.4-6 Importantly, VS should be seen as a syndrome since it is almost always associated with additional visual complaints including palinopsia, entoptic phenomena that arise from the optic apparatus itself (ie, floaters, blue field entoptic phenomenon, self-light of the eye and photopsia), poor night vision (nyctalopia), and photophobia. A large proportion of VS patients has bilateral continuous tinnitus.
To investigate the role of migraine and typical migraine aura mechanisms underlying VS, we sought to assess whether the presence of migraine or aura is associated with different phenotype of the VS syndrome. We prospectively recorded accompanying visual and auditory symptoms in a large cohort of patients with VS and correlated these symptoms with comorbid migraine and typical migraine aura. To assess potential pathophysiological correlates, we further studied brain metabolism in patients with the hypothesis that VS is associated with regional hypermetabolism distinct from previous findings in migraine.[8, 9]
The study was approved by the Institutional Review Board (# 11-07270 and # 11-07431) and the radiation safety committee (58605-RU-04-URH) of the University of California, San Francisco. Patients were recruited via advertisements in social media with the support of a self-help group on VS (Eye on Vision Foundation; http://www.eyeonvision.org/). After being contacted by the patient, eligibility was assessed during telephone interviews.
After being approached by the patient, verbal consent was obtained and subjects with self-suspected VS underwent a semi-structured telephone interview. The following items were covered during the interview:
SPSS (v20, IBM Corp., Armonk, NY, USA) was used for the statistical analysis of the clinical data. Standard descriptive statistics were applied. If appropriate, data are presented as mean ± standard deviation. Nominal data were compared using chi-square or Fisher's exact test, ratio data using t-test. Statistical significance was defined as P < .05.
All subjects participating in the positron emission tomography (PET) study gave written informed consent. Inclusion criterion was VS with at least 2 additional visual symptoms as defined previously. Control subjects did not have VS, associated visual symptoms, tinnitus, a history of frequent migraine attacks (more than 1 every 2 months), or of migraine aura. Exclusion criteria for both groups were ophthalmological pathology other than refraction anomalies, any lifetime history of intake of hallucinogenic drugs, and pregnancy in women. Each subject underwent a detailed personal interview with a focus on visual symptoms, migraine history including typical aura and general past medical history. On the scanning day, each subject had a fasting period of at least 6 hours prior to the acquisition of a high-resolution T1-weighted anatomical MR image (MPRAGE sequence) on a General Electric Signa HDxT 3.0 Tesla scanner (GE Healthcare, Fairfield, CT, USA). Afterwards, a [18F]-2-fluoro-2-deoxy-D-glucose PET ([18F]-FDG PET) scan was acquired using standard parameters, with injection of 10 mCi via an antecubital vein and 45 minutes distribution period in a dark room with eyes closed, on a GE Discovery VCT PET/CT scanner (GE Healthcare) in three-dimensional (3D) mode with septa retracted. Images were reconstructed by 3D iterative reconstruction into 47 image planes (separation 3.27 mm) and into a 128 by 128 image matrix (pixel size: 2.1 × 2.1 mm2). The structural magnetic resonance imaging (MRI) was coregistered to the PET using SPM8 (Wellcome Department of Imaging Neuroscience, http://www.fil.ion.ucl.ac.uk/spm). The coregistered MRI was automatically segmented into gray matter, white matter, and cerebrospinal fluid and normalized into standard stereotaxic space. The spatial normalization parameters from this step were applied to spatially normalize the PET image. Final voxel size was 2 × 2 × 2 mm3. PET images were then smoothed with a Gaussian Kernel of 12 mm full-width at half maximum. The group of VS patients was compared with controls using a 2-sample t-test with masking of non-brain tissue (whole brain explicit mask generated with WFU PickAtlas from Advanced Neuroscience Imaging Research Laboratory, Department of Radiology of Wake Forest University School of Medicine, http://fmri.wfubmc.edu/), and using proportional scaling. Due to the high prevalence of typical migraine aura in patients with VS, the presence of migraine aura was used as a covariate of no interest. According to the clinical manifestation of VS, we suspected hypermetabolism in VS patients. We therefore assessed brain areas with increased metabolism in VS patients compared with controls in a voxel-wise fashion. In line with previous studies on migraine,[12, 13] we considered voxels reaching a significance threshold of P < .001 uncorrected for multiple comparisons to be significant.
Of the 142 patients who contacted the Headache Center at the University of California, San Francisco, 120 subjects (mean age 31 ± 12 years; 62 female) met criterion A for “visual snow,” ie, presence of dynamic, continuous, tiny dots in the entire visual field lasting longer than 3 months.
Palinopsia with “afterimages” from stationary scenes was present in 84%, and with “trailing” in 58%. Excessive floaters were the most common entoptic phenomenon with a prevalence of 83%. Second most common was the blue field entoptic phenomenon (76%). Spontaneous photopsia and consistent self-light of the eye occurred in half of patients. About two thirds of patients had photophobia and nyctalopia. In addition to these visual symptoms, 64% of patients noted continuous bilateral and mainly high-pitched tinnitus (Table 1).
|All Visual Snow Patients|
|N = 120|
|Blue field entoptic phenomenon||91||76%|
|Self-light of the eye||64||53%|
The presence of migraine was associated with an increased prevalence of the additional symptoms palinopsia (odds ratio [OR] 2.8 for “afterimages” and OR 2.6 for “trailing”), spontaneous photopsia (OR 2.9), photophobia (OR 3.2), nyctalopia (OR 2.7), and tinnitus (OR 2.9). Spontaneous photopsia was more prevalent in patients with typical migraine aura (OR 2.4, Table 2).
|Patients With “Visual Snow”|
|N = 120|
|Migraine||Typical Migraine Aura|
|Yes||No||P||OR (95% CI)||Yes||No||P||OR (95% CI)|
|n = 70||n = 50||n = 37||n = 83|
|Palinopsia (“afterimages”)||63||38||.04||2.8 (1.0; 7.8)||34||67||.12||—|
|Palinopsia (“trailing”)||47||22||.01||2.6 (1.2; 5.5)||25||44||.14||—|
|Blue field entoptic phenomenon||53||38||.97||—||28||63||.98||—|
|Self-light of the eye||38||26||.81||—||20||44||.92||—|
|Spontaneous photopsia||45||19||.004||2.9 (1.4; 6.2)||25||39||.04||2.4 (1.0; 5.3)|
|Photophobia||57||29||.005||3.2 (1.4; 7.2)||30||56||.13||—|
|Nyctalopia||51||25||.01||2.7 (1.2; 5.8)||26||50||.29||—|
|Tinnitus||52||25||.006||2.9 (1.3; 6.2)||27||50||.18||—|
Seventeen patients (10 female, mean age ± standard deviation 31 ± 7 years) with VS and at least 2 additional visual symptoms were recruited for the imaging study. Seven had VS as long as they could remember. Mean age of onset in the remaining was 25 ± 8 years. Fourteen (82%) had a history of migraine. Five of those had migraine with typical aura, and 1 had typical migraine aura without history of migraine. All 3 patients without history of migraine had a positive family history of migraine. Besides headache, past medical history included depression, Graves' disease, hypothyroidism, acne, and attention deficit hyperactivity syndrome, each present only in 1 subject. The current regular medication as well as the past medication trials for VS are shown in Table 3. All subjects stated having normal ophthalmological exams except for some refraction anomalies. The 17 controls had the same age and gender distribution (10 female, 31 ± 7 years). Since history of migraine and typical migraine aura were exclusion criteria for controls, they differed significantly from VS patients in respect of history of migraine (P < .001, Fisher's exact test) and history of typical migraine aura (P = .02, Fisher's exact test).
|Current Medication||Medication for “Visual Snow”|
|2||—||Sertraline, fluoxetine||6 months||None|
|4||Dexlansoprazole, bupropion, zolpidem, topiramate, dicyclomine||—||—||None|
|5||Methimazole||Fluoxetine, verapamil, lamotrigine||—||None|
|6||—||Amitriptyline, propranolol||2 months||None|
|9||Throid (porcine), vitamin D, fexofenadine||—||—||None|
The voxel-wise [18F]-FDG PET group comparison evidenced hypermetabolism of the right lingual gyrus (Montreal Neurological Institute coordinates: 16-78-5; cluster size kE = 101; ZE = 3.41; P < .001) and a trend for the anterior lobe of the left cerebellum (Montreal Neurological Institute: -12-62-9; kE = 152; ZE = 3.28; P = .001) ( Figure) in patients with VS when compared with healthy controls after adjusting for the presence of typical migraine aura.
“Visual snow” (VS) is a disabling disorder with patients complaining about TV-snow-like tiny flickering dots in the entire visual field. The symptoms can be continuous and might persist over years. In a recent study, almost all patients with VS had additional visual symptoms, such as palinopsia, entoptic phenomena (floaters, blue field entoptic phenomenon, and others), nyctalopia (impaired night vision), photophobia, and tinnitus suggesting that VS is likely a clinical syndrome. In our study population, the majority of patients with VS had comorbid migraine (58%), and 31% had typical migraine aura. This high comorbidity, when compared with the general population, has led to the assumption that VS might represent persistent migraine aura as often discussed in the initial case series,1-3 although the clinical presentation is clearly different from typical migraine aura.
Here, we sought to understand whether the VS syndrome manifests differently in patients with migraine or typical aura. For that, a cohort of VS patients was carefully phenotyped in respect to the clinical presentation and comorbidities. We found that VS patients, who also have migraine according to International Classification of Headache Disorders – 2nd edition had a significantly higher likelihood of having palinopsia, photophobia, nyctalopia, and tinnitus. Of the entoptic phenomena, ie, visual perceptions arising from the optic apparatus itself, only spontaneous photopsia was more prevalent in VS patients with migraine history, while floaters, blue field entoptic phenomenon, and self-light of the eye were equally distributed. Three major conclusions might be drawn from this: First, the presence of migraine might aggravate the manifestation of the VS syndrome by worsening some, but not all additional visual symptoms. Second, our study population was recruited via a self-help group, and it is possible that patients with a more severe clinical manifestation are more eager to participate in a research study. Therefore, a more severe manifestation of the VS syndrome in migraineurs indicates that the high prevalence of migraine in our VS study population might be subject to a selection bias suggesting that the relevance of migraine for VS pathophysiology might be overrated as well. In contrast, the presence of typical migraine aura, ie, the putative correlate of cortical spreading depression that presents with a homonymous, centrifugally moving scintillating scotoma shaped in zigzag lines,[16, 17] does not substantially alter the distribution of the additional visual symptoms in the VS syndrome. Typical migraine aura may thus not influence the VS phenotype suggesting that the high prevalence of aura is less subject to selection bias than migraine. Although VS is clearly not persistent migraine aura, typical migraine aura might share some pathophysiological background with the VS syndrome. Third, the impressive entoptic phenomena floaters, blue field entoptic phenomenon, and self-light of the eye are present in VS patients independently of a history of migraine, suggesting that these symptoms are probably not mediated or facilitated by a migrainous mechanism. In contrast, they might depend solely on the presence of VS.
Some of the additional visual symptoms in patients with VS can also be found in migraineurs. This might, at least in part, explain how a migrainous, but not typical migraine aura, comorbidity might potentiate these symptoms in VS patients. For migraineurs without VS, the higher prevalence of palinopsia when compared with healthy controls seems to be of minor relevance since it affects only 14.2% of the group and occurs only episodically. However, this predisposition to palinopsia in migraineurs might perpetuate mechanisms of palinopsia in VS resulting in a higher prevalence and continuous presence. For the key migraine symptom photophobia, recent studies have suggested a pain-mediated increase in light sensitivity. In VS, such mechanism is unlikely due to the low prevalence of chronic headache in patients with continuous VS and photophobia. In contrast, photophobia as a symptom of the VS syndrome might be perpetuated by comorbid migraine in a non-pain-mediated manner. This is less clear for tinnitus, which is not a classical migrainous symptom although migraine attack-associated episodes of tinnitus have been reported. Tinnitus could be interpreted as noise within the acoustic system. The similarity to “TV-snow,” ie, “TV-noise,” has previously led to the interpretation that tinnitus might be the clinical correlate of the affection of the acoustic system by VS-like mechanisms. In our study, tinnitus was also more prevalent in VS patients with comorbid migraine and thus behaved like the additional visual symptoms supporting that the VS syndrome might indeed include the non-visual symptom tinnitus.
In [18F]-FDG PET, the right lingual gyrus and the anterior lobe of the left cerebellum were metabolically more active in patients with VS when compared with healthy controls. This first objective correlate of VS strongly suggests the VS syndrome is a neurological condition. This has important consequences for communication with patients, who have been frequently diagnosed as having a psychogenic disorder or as being malingerers. The relevance of the (trend) hypermetabolism of the left cerebellum is unclear. The cerebellum's key function for vision is extraocular motility. Only little is known about its role in visual perception, but cerebellar disease has been associated with difficulties in depth perception or with a phenomenon called upside-down vision.[24, 25] When analyzed visually, this area seems to extend laterally and rostrally to the left lingual gyrus (Figure) possibly reflecting the relatively low spatial resolution of PET. Such bilateral hypermetabolism in the lingual gyrus might be a signature of hyperactivity of the visual system in VS. Interestingly, the same area showed hyperperfusion in [15O]-water PET during high luminous stimulation in migraineurs and during low light stimulation in spontaneous migraine attacks indicating relevance for the migrainous phenomenon photophobia. To put our finding into a broader neurobiological context, it has to be stated that the lingual gyrus is also involved in visual memory and different higher order functions of vision, such as the perception of color, the identification of facial expressions of emotions, or grapheme-color synesthesia. This broad involvement of the lingual gyrus in visual post-processing including photophobia during migraine attacks indicates that VS might also be a disorder of visual post-processing.
One limitation of the imaging part of the study is the higher prevalence of migraineurs in the VS group in comparison with the control group. This could potentially bias the results by showing an effect from migraine rather than from VS – or by “masking” VS correlates in PET by the presence of migraine in the VS group. To address this issue, future studies with pure VS patients without history of migraine or with migraineurs without VS as controls would be necessary. However, we believe that the hypermetabolism in our patients is VS related and not a migraine effect since not all subjects with VS had a history of migraine and, importantly, several recent studies were not able to show hypermetabolism in interictal migraineurs in comparison with controls despite including only migraineurs.[8, 9, 32] In addition, it is unlikely the metabolism data were biased by the higher number of patients with history of migraine aura in the VS group since only one third of VS patients had comorbid aura. Further, the analysis was adjusted for migraine aura and none of our subjects had experienced an episode of typical migraine aura during the distribution period of the tracer or during the scanning.
In a substantial cohort of patients with the “visual snow” (VS) syndrome, migraine is associated with an increased prevalence of the additional symptoms of palinopsia, photopsia, photophobia, nyctalopia, and tinnitus suggesting a more severe phenotype, although not with entoptic phenomena. VS patients with migraine might thus be more interested in participating in studies on VS than patients without migraine, creating a bias of migraine prevalence in such studies and an overestimation of the relevance of migraine for VS pathophysiology. In contrast to migraine, comorbidity of typical migraine aura did not alter the phenotype of the VS syndrome. The high prevalence of typical migraine aura in VS patients therefore is not associated with a worsening of the additional visual symptoms and thus not with an overestimation of aura prevalence in VS. This might indicate a pathophysiological overlap of VS and typical migraine aura despite the distinct clinical presentation. [18F]-FDG PET revealed an objective correlate for VS symptoms. The unique pattern of hypermetabolism in the lingual gyrus in patients with VS has not been shown for interictal migraineurs alone. VS is thus a syndrome distinct from migraine, although the hyperperfusion of this area during migrainous photophobia indicates a potential pathophysiological overlap of both conditions and possibly reflects the perpetuation of the additional visual symptoms in VS patients by comorbid migraine. Understanding this overlap in more detail will be crucial to develop treatment strategies for this disabling neurological disorder in the future.
We thank all patients who have taken part in the interview and the imaging study. Without the interest, participation, and dedication of the patients, this study would not have been possible. The study was supported by the self-help group for “visual snow” (Eye On Vision Foundation) by communicating the study to “visual snow” sufferers. Jan Hoffmann, MD, Department of Neurology, University of California, San Francisco, San Francisco, CA, USA, helped conduct the interviews.