Sunflower Syndrome: A Survey of Provider Awareness and Management Preferences

Abstract

Objective:

Sunflower syndrome is a rare photosensitive pediatric epilepsy characterized by stereotyped hand-waving in response to bright lights. These stereotyped movements with maintained awareness can be mistaken for a movement disorder. This study assessed neurology providers’ diagnostic reasoning, evaluation, and treatment of Sunflower syndrome.

Methods:

A 32-question anonymized electronic survey, including a clinical vignette and video of hand-waving in sunlight, was distributed to child neurology providers to assess: (1) initial diagnosis and workup based on clinical information; (2) updated diagnosis and management after electroencephalogram (EEG); and (3) prior experience with Sunflower syndrome.

Results:

Among 277 viewed surveys, 211 respondents provided information about initial diagnosis and workup; 200 about updated diagnosis; 191 about management; and 189 about prior clinical experience. Most providers (135, 64%) suspected seizure, while fewer suspected movement disorders (29, 14%) or were unsure of the diagnosis (37, 22%). EEG was recommended by 180 (85%). After EEG, 189 (95%) diagnosed epilepsy, 111 of whom specifically diagnosed Sunflower syndrome. The majority (149, 78%) recommended antiseizure medications (ASM) and sun avoidance (181, 95%). Only 103 (55%) had managed Sunflower syndrome. Epileptologists and those with prior clinical experience were more likely to suspect a seizure, order an EEG, and offer ASMs than those without prior experience.

Conclusions:

Though many providers had not managed Sunflower syndrome, the majority recognized this presentation as concerning for epilepsy. Epilepsy training and prior clinical experience are associated with improved recognition and appropriate treatment. Educational initiatives that increase awareness of Sunflower syndrome may improve patient care.

Introduction

Sunflower syndrome is a rare and often pharmacologically resistant photosensitive epilepsy syndrome beginning in childhood. Affected children demonstrate a strong attraction to bright light, particularly sunlight, and will wave their hands with fingers outstretched in front of their eyes while facing the light source. Hand-waving is associated with onset of generalized electroencephalogram (EEG) spikes. During the spiking, children may have slowed behavioral responses or epileptic myoclonus of the eyelids (eyelid fluttering), but they typically retain consciousness. Historically, this was considered a self-induced epilepsy. The hand-waving was thought to induce a flicker effect and epileptiform activity (Geenen, Patel, et al., 2021); some children also present with more subtle, flicker-inducing movements such as rolling eyes upward or rubbing the forehead (Barnett et al., 2021; Baumer & Porter, 2018). However, recent studies have debated whether the repetitive movements are part of the seizure or are a form of self-induction (Baumer & Porter, 2018; Sourbron et al., 2020).

The semiology of Sunflower syndrome overlaps significantly with common pediatric movement disorders, including tics and stereotypies, leading to initial misdiagnosis for some patients (Baumer & Porter, 2018). While hand-waving seizures can occur numerous times per day, a large minority (38%) of children never develop other seizure types (Barnett et al., 2021). Therefore, familiarity with the syndrome’s presentation is important to ensure timely and appropriate diagnostic testing and treatment. In contrast to Sunflower syndrome, motor tics and stereotypies are among the most frequently seen disorders in the child neurology clinic, with a reported 1%–3% prevalence of transient tic disorders in children (Tourette Association of America; Knight et al., 2012) and 3%–4% or potentially higher prevalence of complex motor stereotypies (Foster, 1998; Singer, 2010; Mackenzie, 2018). These movement disorders are typically diagnosed by clinical history and exam and do not require routine ancillary testing.

Considering this semiological overlap with movement disorders as well as the skewed prevalence, we hypothesized that a child with sunlight-triggered hand-waving episodes might not be recognized as having an epilepsy syndrome by providers at time of initial presentation. To test this hypothesis, we conducted an online survey distributed through several professional societies. As debate remains about the nature of the hand-waving episodes and best treatment practices, our survey then disclosed EEG data and assessed providers’ opinions about (1) the nature of the hand-waving episodes; (2) the need for treatment of hand-waving episodes; and (3) treatment preferences.

Methods

Survey Development:

A set of 32 questions was prepared by the Sunflower syndrome working group, which is part of the Pediatric Epilepsy Research Consortium (PERC). Members of the working group included epileptologists, neuro-geneticists, and movement disorder specialists from five large pediatric hospitals (Lucile Packard Children’s Hospital, Colorado Children’s Hospital, Rady Children’s Hospital, Dell Children’s Medical Center, and Massachusetts General Hospital). The first aim of the survey was to assess whether providers recognize the symptoms of Sunflower syndrome as an epilepsy syndrome. The second aim was to determine how providers would classify and treat this condition if they received updated information about its associated EEG abnormalities.

The survey consisted of four pages of questions. Once a page was completed, it could not be revisited. Most questions had closed-ended choices, though there were free-text options. On the first page, respondents reviewed a clinical vignette with accompanying video that described a previously healthy 9-year-old girl with hand-waving episodes in sunlight presenting for initial neurological evaluation (Figure 1a). Respondents were queried regarding most likely diagnosis (choices: movement disorder, seizure disorder, other, or undecided) and whether and which ancillary testing was required (neuroimaging, EEG, lab work, and/or other testing).

Figure 1:

(A) Vignette and still frame from the video presented at the beginning of the survey. (B) Electroencephalogram (EEG) tracing and report presented on page 2 of the survey, after responses regarding initial diagnostic impression and recommended testing were gathered. OCD, obsessive-compulsive disorder.

On the second page, respondents were shown an EEG tracing and report indicating epileptiform discharges during hand-waving episodes (Figure 1b). They were asked: (1) Does this patient have epilepsy? (2) Is the hand-waving movement itself a seizure? (3) Is the patient self-inducing seizures? and (4) Does this patient have an epilepsy syndrome you recognize? Those who endorsed recognizing the syndrome were asked to name the syndrome. On the third page, respondents were asked whether they would recommend pharmacological therapy or lifestyle interventions. On the fourth page, respondents were queried about their clinical practice and experience with similar patients.

Survey Administration:

The Stanford University Institutional Review Board approved this study. Responses were collected anonymously after participants reviewed study information. A child neuro-oncologist from Stanford University served as the study contact to avoid biasing respondents. This survey was distributed via REDCap to: (1) the Child Neurology Society message boards (approximately 2000 members), (2) the American Epilepsy Society message board (3600 United States members, with 1600 neurologists), (3) the PERC mailing list (approximately 145 members), and (4) the International Parkinson and Movement Disorder Society Moving Along newsletter (391 active newsletter readers). Inclusion criteria included physicians or other providers who independently evaluate and manage children with neurologic disorders. Trainees were excluded. The survey was conducted from December 2020 to July 2021.

Survey Inclusion:

We reviewed surveys for completeness. Since each of the four survey pages was intended to address a different question (initial diagnosis and management, updated diagnosis and assessment with EEG data, treatment preferences, and provider’s prior clinical experience), we included surveys where at least the first page was completed.

Statistical Analysis:

Statistical analyses were performed using Statistical Analysis System (SAS) OnDemand for Academics (Cary, NC). We calculated descriptive statistics (number and percentage) for each response. To assess if formal training or prior clinical experience with such patients influenced diagnosis or management, we compared those with and without epilepsy training and those with and without prior clinical experience with a similar patient. We used chi-square to test if groups differed on initial suspicion for a seizure disorder, intention to order an EEG, and prescription of anti-seizure medications (ASMs). We considered p-values of <0.05 significant for these independent tests.

Results

Survey Completion:

Of the 277 surveys that were accessed, 211 (76%) completed at least page 1 indicating initial diagnosis and recommended testing, and hence were included. Of these 211, 200 (95%) completed only the first two pages, 191 (91%) completed only the first three pages, and 188 (89%) completed all four pages of the survey.

Initial Impression and Recommended Workup:

Leading differential after review of the clinical vignette and video was (a) seizure disorder for 135 (64%); (b) movement disorder for 29 (13.7%) (10 [34.5%] of whom diagnosed a tic disorder, 19 [65.5%] of whom identified the movement as a stereotypy); (c) undecided for 38 (18%); and (d) other for 9 (4.3%). Of the nine respondents who chose other, two were concerned for compulsive behavior, four for self-stimulatory behavior, and three specified Sunflower syndrome. Of the three respondents who named Sunflower syndrome, two later endorsed it to be a seizure disorder and one did not.

Overall, 182 (86%) of respondents felt that ancillary testing was necessary; 180 (85%) requested an EEG, 19 (9%) magnetic resonance imaging (MRI), and nine (4%) laboratory testing. Only nine of the 29 (31%) respondents who were initially concerned for a movement disorder ordered extra testing; two ordered MRIs, seven ordered EEGs, and one ordered laboratory tests. In contrast, 134 of the 135 (99%) initially concerned for a seizure disorder ordered extra testing; 11 requested imaging, 134 an EEG, and six laboratory tests. Of the 38 who were undecided on the diagnosis, 33 (87%) ordered additional tests; six requested an MRI and 33 an EEG. Surprisingly, of the three who diagnosed Sunflower syndrome, only one requested an EEG. Of the two who considered it compulsive behavior, one requested laboratory studies and one requested an EEG. Finally, of the four who identified self-stimulatory behavior, all requested an EEG. Therefore, based on clinical impression (135), or results of the EEG ordered at the initial visit (46), 181 of the 211 (86%) of respondents would be expected to obtain information regarding the EEG abnormalities associated with the hand-waving episodes after a single visit.

Updated Impression and Data Interpretation:

Of the 200 respondents who reviewed the EEG tracing and report, 189 (94.5%) classified the condition as epilepsy and 11 (5.5%) did not. Of the 11 respondents who did not classify this as epilepsy, their original diagnosis had been movement disorder (4), epilepsy (1), undecided (5), and other/Sunflower syndrome (1). Interestingly, four of these 11 respondents had requested the EEG at initial presentation but did not update their diagnosis to epilepsy based on the EEG findings.

Nature of Hand-Waving Episodes:

Of the 189 respondents who classified this as epilepsy, only 43 (23%) thought that the hand-waving movement itself was a seizure. When asked separately, 144 (76%) felt that the patient was self-inducing seizures. Notably, of the 11 respondents who did not classify this as epilepsy, one classified the hand-waving movement as seizure and three felt the patient was self-inducing seizures.

Syndrome Recognition (Figure 2):

Figure 2:

Recognition of the syndrome. Respondents (n=156) who reported recognizing the syndrome were asked to name it. The majority named Sunflower syndrome, though some chose an overlap with reflex epilepsy or Jeavons syndrome.

Of the 189 respondents who classified this presentation as epilepsy, 156 provided a syndrome name via free text. The majority of these 156 respondents labeled this condition as either Sunflower syndrome (111, 71%), Jeavons syndrome (17, 11%), or photosensitive epilepsy (18, 12%). Few respondents described this as reflex (4, 2%), self-induced (2, 1%), primary generalized epilepsy (2, 1%) or benign occipital epilepsy (2, 1%). Respondents noted areas of overlap, as shown in Figure 2. When offered a list of possible syndrome names, the vast majority of the 191 respondents chose Sunflower syndrome (151, 79%) while a minority chose Jeavons syndrome (15, 8%) or absence epilepsy (4, 2%). Eighteen (9%) were unsure that the syndrome was consistent with one of the choices and three (2%) felt it was a different diagnosis (benign occipital, self-induced photoparoxysmal epilepsy, and photoparoxysmal response).

Treatment Choice (Figure 3):

Figure 3:

Treatment choices. Columns represent the first-line antiseizure medications chosen by each of the 148 respondents who recommended a specific antiseizure medication (ASM). Subdivisions of each column represent the second-line ASM (see text for numbers of respondents who chose each medication as the second-line). Levetiracetam and valproic acid were common first- and second-line therapies. Lamotrigine was more frequently chosen as a second-line than first-line therapy. Benzodiazepines were never chosen as a first-line therapy but were a common second-line choice.

First-Line Therapy:

Of 191 respondents who provided their treatment recommendations, 151 (79%) recommended pharmacological therapy. Most (149 of 151, 99%) recommended an ASM, typically a broad-spectrum one, as first-line therapy. The 149 recommended ASMs included levetiracetam (68, 45%), valproic acid (46, 30%), ethosuximide (15, 10%), or lamotrigine (13, 9%); acetazolamide (1, 0.7%), topiramate/zonisamide (3, 2%), or fenfluramine (2, 1%) were occasionally recommended. One respondent was unsure of which ASM to recommend. Few respondents recommended medications typically given for movement disorders. One chose an alpha-agonist (e.g., guanfacine, clonidine) and one a selective serotonin reuptake inhibitor (e.g., fluoxetine, sertraline). Anti-dopaminergic agents (e.g., pimozide) or other medications were not recommended.

Second-Line Therapy:

Most of the 149 respondents who recommended ASMs also chose a broad-spectrum agent as a second-line therapy. Lamotrigine (36, 24%), valproic acid (30, 20%), levetiracetam (25, 17%), topiramate/zonisamide (20, 13%), clobazam/clonazepam (17, 11%), and ethosuximide (8, 5%) were the most common choices. Three respondents (2%) who chose levetiracetam as a first-line agent recommended brivaracetam as second line. Lacosamide (3, 2%), oxcarbazepine (2, 1%), vigabatrin (1, 1%), the ketogenic diet (1, 1%), and no additional treatment or other(3, 2%) were other choices.

Non-pharmacological Therapies:

The majority (181, 95%) of the 191 respondents recommended sun avoidance via sunglasses (140, 73%), hats (90, 47%), or Zeiss blue lenses (113, 59%). A sizable minority (57, 30%) recommended formal behavioral therapy.

Prior Experience with Similar Patients:

Of 188 respondents who completed the survey, 85 (45%) had no previous experience with a similar patient, 87 (46%) had cared for fewer than five such patients, 14 (7%) for 5–10 patients, and two (1%) for > 10 patients. The vast majority with prior experience (98, 95%) always ordered EEGs on these patients, while only two (2%) never did; the remaining three providers ordered EEGs intermittently. EEGs were generally high yield; of the 101 providers who ordered EEGs, 67 (65%) noted they were abnormal in all patients, 28 (27%) in greater than half of patients, and only six (6%) in less than half of patients. No one reported that EEGs were normal in all similar patients.

Clinical Practice Patterns:

Of the 188 who provided data on their clinical practice, 69 (37%) identified as general neurologists and 119 (63%) as subspecialists (defined as having formal training or seeing > 50% of patients with a specific neurologic problem). Sixty-four (34%) subspecialized in epilepsy, eight (4%) in movement disorders, eight (4%) in headache medicine, six (3%) in neuromuscular, six (3%) in behavioral/cognitive neurology, five (3%) in neuro-oncology, five (3%) in neuro-inflammatory disorders, four (2%) in neurointensive care, three (2%) in neonatal/fetal neurology, three (2%) in genetics, two (1%) in neurodevelopmental diseases, two (1%) in vascular neurology, and three (2%) in other (neuro-ophthalmology, neuro-rehabilitation, and internal medicine). Most (178, 95%) identified as child neurologists (> 75% of patients younger than 18 years), two (1%) as adult neurologists (> 75% of patients older than 18 years), and eight (4%) as seeing a mix of ages.

Care by Clinical Expertise and Prior Experience (Table 1):

Table 1:

Influence of Subspecialty Expertise and Prior Experience on Clinical Approach. Number and percentage of providers who were concerned for seizures/ordered electroencephalogram (EEG)/started antiseizure medications (ASMs), divided by subspecialty training and prior experience. Epileptologists and providers with prior experience treating patients with Sunflower syndrome were more likely to order the appropriate diagnostic tests and treat with medications.

	n (%)	n (%)	Group Difference % Difference (95% CI)	X2	p
Epilepsy Training	Yes (n=64)	No (n=124)
Concerned for Seizure	57 (89)	72 (58)	31.0% (19.4–42.6%)	18.8	<0.0001
Ordered EEG	64 (100)	100 (81)	19.4% (12.4–26.3%)	14.1	0.0002
Started ASM	53 (83)	94 (76)	7.0% (−5.0–18.9%)	1.2	0.27
Clinical Experience	Yes (n=103)	No (n=85)
Concerned for Seizure	86 (84)	43 (50)	32.9% (20.0–45.7%)	23.4	<0.0001
Ordered EEG	100 (97)	64 (75)	21.8% (12.1–31.5%)	19.9	<0.0001
Started ASM	90 (87)	57 (67)	20.3% (8.5–32.2%)	11.3	0.0008

Epileptologists were more likely than providers without epilepsy training to be concerned for seizures [57/64, 89% vs. 72/124, 58%; X²(1 n = 188) =18.8, p<0.0001] and to order an EEG [64/64, 100% vs. 100/124, 81%; X²(1, n = 188) =14.1, p=0.0002], but were not more likely to start a medication. Providers with prior clinical experience were more likely than those without experience to be concerned for seizures [86/103, 84% vs. 43/85, 50%; X²(1 n = 188) =23.4, p<0.0001], order an EEG [100/103, 97% vs. 64/85, 75%;X²(1 n = 188) =19.9, p<0.0001], and start an ASM [90/103, 87% vs. 57/85, 67%; X²(1 n = 188) =11.3, p=0.0008].

Discussion

Sunflower syndrome is a rare childhood-onset epilepsy syndrome with significant phenomenological overlap with common movement disorders including tics and stereotypies. Our survey found that most child neurology providers recognize the symptoms of hand-waving episodes provoked by bright light/sunlight as warranting workup for a seizure disorder. Even though only 64% of respondents thought seizure disorder was the most likely diagnosis based on clinical history and exam, 85% would order an EEG at the first visit and thus learn about the epileptiform discharges that accompany hand-waving episodes. Those with formal training in epilepsy and those with clinical experience with similar patients are more likely to be concerned for a seizure disorder and evaluate with an EEG than those without this training or experience. Finally, we note that there is still controversy regarding whether hand-waving episodes constitute seizures and whether the disorder should be treated with an ASM; even with knowledge of EEG abnormalities, less than 80% of respondents recommended daily ASMs.

Given the overlap in semiology between rare Sunflower syndrome and common movement disorders, we were surprised to find that most respondents recognized the presentation as a seizure disorder. A sizable majority were able to label the presentation as Sunflower syndrome without prompting, and 80% chose Sunflower syndrome as the most likely diagnosis from a list, suggesting that many providers have heard of Sunflower syndrome. While most providers (64%) recognized that the child should be evaluated for a seizure disorder, those with subspecialty training in epilepsy or with prior clinical experience with similar patients were more likely to order appropriate diagnostic testing and recommend treatment with an ASM. These findings demonstrate a knowledge gap about this condition within the child neurology community. Providing education and awareness to the general pediatric and child neurology community about the features of Sunflower syndrome could improve timely diagnosis, workup, and management. Furthermore, limited data exist about adult outcomes in Sunflower syndrome, though case reports suggest it can persist into young and even middle adulthood (Belcastro et al., 2021). Therefore, expanding educational efforts to include adult neurologists is also warranted.

The study furthermore shows that, even among those who recognize this as epilepsy, the clinical presentation of hand-waving is not universally recognized as pathognomonic for Sunflower syndrome. Notably, 10% of providers felt that the presentation was consistent with Jeavons syndrome. While children with Sunflower syndrome can have eyelid myoclonia and absence seizures in addition to hand-waving (Baumer & Porter, 2018), children with Jeavons syndrome do not have hand-waving episodes or heliotropism. Many other providers commented that there is overlap with reflex or photosensitive epilepsies more generally. Based on our clinical experience, Sunflower syndrome and Jeavons syndrome are distinct clinical entities; however, the International League Against Epilepsy has recently renamed Jeavons syndrome as “epilepsy with eyelid myoclonia (EEM)” and suggested that Sunflower syndrome be subsumed in this category as “EEM with prominent photic induction” (Specchio et al., 2022).

The survey also highlights that there is disagreement about whether the hand-waving itself is a seizure. This is clinically relevant, because antiseizure medications are often titrated based on the persistence of hand-waving episodes. Notably, similar disagreement about the nature of the hand-waving episodes remains in the literature. For many years, the hand-waving has been described as a voluntary movement used to induce a flicker effect and hence epileptiform activity (for a review of the literature, see (Geenen, Patel, et al., 2021). One case series reported that some children continued to have hand-waving even after the EEG had normalized, suggesting that the movement could be disentangled from the seizure itself (Baumer & Porter, 2018) and was perhaps a compulsion. Another report found that a patient could produce the abnormal EEG discharges when voluntarily performing the hand-waving (Belcastro & Striano, 2014). In contrast, Sourbron et al. reviewed the EEGs of five children with Sunflower syndrome and found onset of epileptiform activity within one second of hand-waving, which would generally be too short of an induction period for a photoparoxysmal response (Sourbron et al., 2020). The authors concluded that the hand-waving episodes may be part of the seizure semiology rather than a self-induction technique. It is possible that the hand-waving is part of the seizure semiology initially but develops into a compulsory behavior over time, and more detailed study of these episodes over time in this disorder is warranted.

Sunflower syndrome is often refractory to medications (Belcastro et al., 2021), and there is no consensus about best treatment. In several retrospective case series, seizure improvement has been reported in 60%–66% with valproic acid (Barnett et al., 2021; Baumer & Porter, 2018) and up to 80% with ketogenic diet therapies (Barnett et al., 2021), whereas levetiracetam monotherapy was ineffective in all patients in one study (Baumer & Porter, 2018) but was effective for some in another study (Barnett et al., 2021). Recently, an open-label study of patients with Sunflower syndrome reported improvement in hand-waving episodes and other seizures with fenfluramine (Geenen, Doshi, et al., 2021; Patel et al., 2022). Most survey respondents felt that a broad-spectrum ASM should be initiated first, with the majority recommending levetiracetam. Further studies in larger cohorts are needed to determine whether some ASMs should be recommended as a first-line therapy in this condition. Until consensus or evidence-based treatment recommendations are formulated, medical education highlighting the mostly poor response of Sunflower syndrome to levetiracetam monotherapy could help providers more efficiently identify effective treatments for their patients. Interestingly, some survey respondents did not recommend treatment with ASMs despite classifying this as seizures and/or epilepsy, possibly related to prioritizing stimulus avoidance in the setting of reflex epilepsy. We feel strongly children with Sunflower syndrome should both be treated with ASMs and counseled on stimulus avoidance.

There are several limitations in this study. First, while we made all efforts to ensure that respondents represented a wide swath of the child neurology community, we may have inadvertently captured clinicians with more expertise in epilepsy as approximately one-third of respondents identified as having subspeciality training in epilepsy. However, the child neurology clinical workforce study in 2016 also reported that a high percentage of child neurologists have this expertise (Kang et al., 2016). We tried to limit bias by distributing the survey through multiple platforms, including the International Parkinson and Movement Disorder Society. A second issue is missing data. Only 211 of the 277 (76%) people who opened the study filled out at least the first page, and only 188 (68%) completed the entire survey. The response rate is similar to that seen in other publications using survey methodology (Wu et al., 2022). These data may be missing at random, for example due to survey fatigue (especially as the survey was administered during the first year of the COVID pandemic), but it is also possible that those who did not recognize the disorder discontinued the survey early. If that is the case, we may be overestimating clinicians’ recognition of Sunflower syndrome. If this were true, it would further strengthen the suggestion for additional medical education on this disorder. Finally, we noted that responses to several questions seemed contradictory or confusing. For example, some respondents classified hand-waving as seizures but did not consider the picture consistent with an epilepsy syndrome. Given that the survey was administered anonymously, we could not ask follow-up questions and speculate that some individuals may not classify reflex seizures as consistent with epilepsy. Fortunately, there were very few such responses, and hence we think it is unlikely misinterpretation is significantly altering our results.

Conclusion

A high percentage of clinicians recognize that children presenting with hand-waving episodes when exposed to bright lights or sunlight should undergo workup for epilepsy. There is disagreement about whether and how these episodes should be treated. The results also demonstrate that providers with epilepsy subspecialty training or prior experience with such patients are more likely to recommend appropriate diagnostic workup and treatment, suggesting that additional medical education about Sunflower syndrome may improve patient outcomes. Further work is needed both to determine best management practices as well as to increase provider knowledge and awareness of Sunflower syndrome.