Incidence of Sturge Weber Syndrome and Associated Ocular Involvement in Olmsted County, Minnesota, United States

Abstract

Background:

Sturge Weber syndrome (SWS)is a rare sporadic syndrome characterized by nevus flammeus (port wine stain, PWS) in the trigeminal nerve distribution, diffuse choroidal hemangioma, and brain leptomeningeal hemangioma. We are unaware of previous reports of SWS incidence in the United States. This study investigated SWS incidence and associated ocular involvement in Olmsted County, Minnesota.

Materials and methods:

The Rochester Epidemiology Project database was used to identify SWS cases from January 1, 2000–December 31, 2017. Incidence of SWS was calculated using the Olmsted County census population. A literature review of studies investigating SWS-associated ocular involvement was also performed.

Results:

There were 13 patients with SWS in Olmsted County classified as type 1 (31%) or type 2 (69%). Age and sex adjusted incidence of SWS was 0.19/100,000/year. Race was predominantly Caucasian (85%), with sex female (69%) or male (31%). All patients had PWS, mostly with unilateral distribution in the V1 and/or V2 region (85%). Two cases (15%) had associated Klippel–Trenaunay syndrome. The most common ocular features included: dilated episcleral vessels (46%), glaucoma (46%), retinal detachment (23%), DCH (7.7%), strabismus (31%), and refractive error (38%). PWS in the V1 distribution was associated with all cases of glaucoma, DCH, and neurological involvement. Severe visual impairment (>0.6 LogMAR, Snellen equivalent ≤20/100) was found in (23%) at final follow-up, and one patient (8%) required enucleation for uncontrolled glaucoma.

Conclusions:

SWS affects approximately 0.19/100,000/year in Olmsted County. Early diagnosis, intervention, lifelong follow-up, and multidisciplinary approach should be used to optimize systemic and ocular outcomes.

INTRODUCTION

Sturge Weber syndrome (SWS), also known as encephalofacial hemangiomatosis, is a rare, sporadic syndrome with reported incidence of 1:20,000–1:50,000 infants, with no known sex or racial predilection (1,2). SWS is considered one of the phakomatoses, or oculoneurocutanous syndromes, which are a group of syndromes characterized by systemic hamartomas of the eye, brain, skin, and sometimes viscera and bone (3). Characteristic clinical features of SWS include nevus flammeus (port wine stain, PWS) in the trigeminal nerve distribution, diffuse choroidal hemangioma (DCH), and brain leptomeningeal hemangioma (4–9).

Definitions of SWS have varied. In 1992 Roach described three phenotypes: Type 1 (classic SWS) includes both facial cutaneous and leptomeningeal angiomas with likely glaucoma; Type 2 (SWS) includes facial cutaneous angioma with glaucoma possibly present; and Type 3 (SWS forme fruste) includes leptomeningeal angioma with absent cutaneous facial angioma and glaucoma(10). Of all patients with PWS, up to 8% will show complete features of SWS (classic type) (4,11). In 2013 Shirley et al. published a ground-breaking study describing a somatic mosaic activation mutation in the GNAQ gene, which leads to stimulation of cellular proliferation and inhibition of apoptosis (12).The final phenotype of the syndrome depends on the timing and site of this mutation during embryonic development (12,13). Approximately 50–70% of SWS patients show pathologic ocular findings, usually ipsilateral to PWS, which has a predilection for the ophthalmic nerve distribution (V1) (1,14). Of all SWS related ocular involvement, the most commonly reported is glaucoma, which affects 30 to 75% of patients (10,15–18). Another common ocular finding is DCH, which affects 19 to 71% of patients (1,14,18,19).

We are unaware of any prior published reports describing the incidence of SWS in the United States (USA). Herein, we utilize the Rochester Epidemiology Project (REP) database to determine the incidence of SWS in Olmsted County, Minnesota, USA. This information, along with investigation of disease course and related ocular involvement, will further expand the knowledge of this rare type of neurocutaneous angiomatosis and could identify areas for investigation of new treatments or diagnostic modalities to improve visual prognosis for these patients.

MATERIAL AND METHODS

This study utilized the REP medical records linkage system. The database consists only of residents of Olmsted County, MN, and queries yield only subjects who have provided informed consent for research. Medical care of this population is provided primarily by Mayo Clinic and Olmsted Medical Center, but additional small independent clinics are included in order to capture nearly all Olmsted County residents. There are approximately 148,201 unique patient records available for review at the time of this report.

A retrospective chart review was performed to identify patients with SWS from 1/1/2000–12/31/2017. We searched the REP database using ICD-9 (759.6, 757.32, 228.09), ICD-10 (Q85.8, Q82.5, D18.01, D18.09), and HICDA codes (7598310) for SWS, phakomatosis, PWS, skin or subcutaneous hemangioma, and hemangioma elsewhere in order to capture all patients with a possible diagnosis of SWS. We identified patients based on a documented diagnosis of SWS and further examined the medical records to confirm the diagnosis. We excluded patients with another possible diagnosis. Patients of all ages were included. Residence of Olmsted County at the time of at least one confirmatory eye exam was a requirement for inclusion in the ocular involvement documentation. Residence of Olmsted County at the time of diagnosis was required for inclusion in incidence calculations. Anyone who refused MN research authorization was not included in the study. The initial diagnosis of SWS may have been made prior to 1/1/2000, and records in the previous time period 1976–2000 were reviewed if our primary data were not sufficient. Cases initially diagnosed prior to 2000 were not included in incidence calculations.

We found 1,089 unique patients carrying one of 8 search codes between 1/1/2000 and 12/31/2017. These patients were reviewed by a single investigator (HTR) to eliminate unrelated diagnoses, such as hemangioma not associated with SWS, isolated choroidal hemangioma, other phakomatoses syndromes, and visceral hemangioma. If the diagnosis was uncertain, a senior ocular oncologist and retina specialist (JSP) reviewed the case. After excluding patients with alternative diagnoses, there were 18 remaining patients with a SWS diagnosis. Charts for these patients, including any available images, CT/MRI brain, fundus photographs, B-scan ultrasonography, optical coherence tomography (OCT), and fluorescein angiography (FA) were reviewed by the investigators. Five cases were excluded due to insufficient data and/or imaging to confirm associated ocular and/or neurological involvement.We used the following definition for SWS types in our study: 1) type 1: PWS plus leptomeningeal angioma or seizure (presumed leptomeningeal angioma), 2) type 2: PWS (with ocular involvement), 3) type 3: leptomeningeal angioma or seizure with no PWS (see figure 1).

Figure 1.

Patients in the Rochester Epidemiology Project (REP) database with Sturge Weber Syndrome (SWS). Eight unique ICD-9, ICD-10, and HICDA codes were searched to capture all patients with SWS diagnosis. Patients with unrelated diagnoses (** infantile hemangioma, PWS elsewhere, skin nevi or other lesions, hemangiomas elsewhere), or other phakomatosis syndromes (ataxia telangiectasia, neurofibromatosis, tuberous sclerosis, or Von Hippel-Lindau) were eliminated, leaving 18 patients with possible SWS. Five cases were excluded due to insufficient data and/or imaging to confirm ocular and/or neurological involvement with PWS leaving us with 13 cases for analysis. Only six patients were included in incidence calculations.

The incidence of SWS in Olmsted County was calculated based on the number of confirmed cases identified in each given year from 2000–2017 and in the cumulative 18-year period on review of records from the REP database. These values were divided by the population of Olmsted County, and the incidence was estimated. Populations were included for Olmsted County in each census year, and the population between census years was linearly interpolated. Incidence of all SWS cases with [95% confidence interval (CI)] was calculated. SAS version 9.4 was used for all descriptive statistics and analysis (Cary, NC).

Age at SWS diagnosis was approximated by date first seen in Olmsted County when appropriate. Ocular involvement in the SWS population was calculated as a percentage of the total number of cases [95% CI]. Categorical variables were described using frequency and percentage, whereas continuous variables were described using mean, median, and/or range.

Literature review was performed using a PubMed search for publications on the topic of SWS and ocular involvement as a comparison dataset for our retrospective study data. A PubMed search for keywords included: Sturge Weber Syndrome, phakomatosis, neurocutaneous syndromes, port wine stain, nevus flammeus, eye manifestations, eye diseases, ocular, glaucoma, hemangioma, and epidemiologic studies. Results were limited to peer-reviewed English-language journals published between (1980–2018). All articles were reviewed for study inclusion. Case reports, review articles, articles focusing on specific treatment modalities, and articles focusing on neurologic disease only were excluded.

This study is in compliance with the Health Insurance Portability and Accountability Act (HIPAA), adheres to the Tenets of the Declaration of Helsinki, and was approved by the Mayo Clinic and Olmsted County Medical Center Institutional Review Boards.

RESULTS

There were 13 patients with SWS in Olmsted County classified as type 1 (n=4, 31%), type 2 (n=9, 69%), or type 3 (n=0, 0%), of which 6 were included in incidence calculations. We excluded cases with date of diagnosis prior to 2000 (n=6) or patients who were not residents of Olmsted County at the time of diagnosis (n=1). The age and sex adjusted incidence of SWS in Olmsted County was 0.19 [95% CI 0.04–0.34 -] per 100,000/year.

A summary of demographic and clinical characteristics is found in Table 1. Mean age at date first seen in Olmsted County was 15.3 years (median 10.3, range 0–42 years). Patient race was predominantly Caucasian (n=11, 85%), with approximate female (n=9, 69%) to male (n=4, 31%) ratio of 2.2:1. For cases included in incidence calculations, the sex ratio was almost the same 2:1 (67% females, 33% males). With the most recent case diagnosed in 2017, no patients in this series underwent genetic testing for GNAQ mutation. One patient (n=1, 7.7%) did undergo genetic testing (in 2011) and had chromosomal duplication at 5q12.

Table 1:

Summary of demographic and clinical features for SWS cases in Olmsted County

Demographic/Clinical features	N (%- [95% CI])
Age at presentation with SWS (years)a mean (median, range) (n=13 patients)	15.3(10.3, 0.0–42.2)
Race
Caucasian	11(85– [55,98])
Hispanic	1(7.7– [0.2,36.0])
Unknown	1(7.7– [0.2,36.0])
Gender
Male	4(31– [9,61])
Female	9(69– [36,91])
PWS
Craniofacial PWS	13(100– [75,100])
Right V1	4(31– [8,61])
Right V2	1(7.7– [0.2,36.0])
Right V1+V2	1(7.7– [0.2,36.0])
Left V1	3(23– [5,54])
Left V2	1(7.7– [0.2,36.0])
Left V1+V2	1(7.7– [0.2,36.0])
Left (unknown distribution)	1(7.7– [0.2,36.0])
Bilateral V1+V2	1(7.7– [0.2,36.0])
Extra-cranial PWS	2(15– [2,45])
Leptomeningeal angioma
Right	1(7.7– [0.2,36.0])
Left	3(23– [5,54])
None	9(69– [36,91])
Neurological disease
Epilepsy	4(31– [8,61])
Hemiparesis	3(23– [5,54])
Developmental delay	1(7.7– [0.2,36.0])
Mild cognitive impairment	1(7.7– [0.2,36.0])
Homonymous hemianopia	2(15– [2.45])
None	9(69– [36,91])
Age at presentation to ophthalmology (years) mean (median, range) (n=11)	20(18.5,0.0–58.2 )
Years of ophthalmology follow up (years) mean (median, range) (n= 10)	22.3(136, 0.0–77.7)
Ocular involvement (n=13) b
Dilated episcleral vessels	6(46– [19,75])
Band keratopathy	2(15– [2,45])
Buphthalmos	2(15– [2,45])
Glaucoma	6(46– [19,75])
Cataract	5(39– [14,68]) c
Iridocorneal touch	1(7.7, [0.2,36.0])
Tortuous retinal vessels	1(7.7, [0.2,36.0])
Retinal detachment	3(23– [5,54])
DCH	1(7.7, [0.2,36.0])
Hypotonus maculopathy	1(7.7, [0.2,36.0])
ERM	1(7.7, [0.2,36.0])
ExotropiaStrabismus (Exotropia)	4(31– [8,61])
Refractive error (Myopia, Hyperopia)	2,3(15– [2.45], ],23 – [5.54])
Amblyopia	2(15– [2,45])
IOP at DFS (mmHg) mean (median, range) (n=6 )	14(13,6–22)
IOP at DLS (mmHg) mean (median, range) (n=9 )	17(16,8–34)
Visual acuity in LogMAR at DLS (affected eye) (n= 13) mean (median, range) d	1.2 (0.25,0.0–4.0)
≤0.00–0.20 (20/20–20/30)	4(31– [8,61])
0.3–0.6 (20/40–20/80)	1(7.7– [0.2,36.0])
0.7–1.00 (20/100–20/200)	1(7.7– [0.2,36.0])
NLP	2(15– [2,45])
CSM	1(7.7– [0.2,36.0])
Enucleated	1(7.7– [0.2–36.0])
Unknown	3(23– [5,54])

SWS= Sturge Weber Syndrome, PWS= Port wine stain (nevus flammeus), V=trigeminal nerve branch area, ADHD= attention deficit hyperactivity disorder

DCH= diffuse choroidal hemangioma, IOP= intraocular pressure, DFS= date first seen in ophthalmology, DLS= date last seen in ophthalmology

NLP= No light perception, CSM= central steady and maintained

^a)Age presenting to Olmsted County

^b)In 11 patients who had eye exam, 9 patients showed ocular involvement.ablecaption

^c)Out of these patients, 2 cases developed age related nuclear sclerosis, 1 case in the setting of buphthalmic eye, 1 case with underlying funnel RD, last case developed posterior cortical opacity after trabeculectomy.

^d)In LogMar scale: CF =2.0, HM = 3.0, LP or NLP = 4.0

All study patients had PWS (n=13, 100%), with distribution affecting the sensory branch of the trigeminal nerve unilaterally in the V1 and/or V2 distribution (n=11, 85%), bilaterally in the V1 and V2 distribution (n=1, 7.7%), or in an unspecified distribution (n=1, 7.7%), with no cases involving the V3 distribution (n=0, 0%). Patients underwent evaluation by an ophthalmologist in Olmsted County (n=11, 85%), with (n=1, 7.7%) lost follow up after first evaluation.

Descriptions of systemic involvement for SWS are found in Table 2. SWS type 1 (n=4, 31%) was confirmed with brain imaging through CT and/or MRI. The most common neurological manifestation was seizures (n=4, 31% [95% CI 8–61]). All patients had their first seizure during infancy (age 1–9 months). Other neurological features included hemiparesis (n=3, 23 %), homonymous hemianopia (n=2, 15%), delayed development (n=1, 7.7%), and cognitive impairment (n=1, 7.7%). The intracranial abnormality was ipsilateral to the PWS (n=3, 75%) except in 1 case (n=1, 25%) where the PWS was bilateral (V1 and V2) with unilateral brain involvement. PWS with V1 involvement was noted (n=4, 100%) in all type 1 SWS cases. Klippel–Trenaunay syndrome (KTS) was noted in Subjects 4 and 10 (n=2, 15%).

Table 2:

Dtata of patients with SWS at Olmsted County- Demographic and systemic features

No.	Agea (years)	Sex	Race	PWS	V area	Type of SWS	Neuro signs	Age at seizure onset (months)	Rx of seizures	Neurological imaging	Systemic outcome	Follow up ( years)
1	2	M	W	L	1	1	Seizures, right homonymous hemianopia, Right UL weakness, unsteady tandem gait, action tremor left side	9	Phenobarbital (not well controlled- heavy drinker)	Gyriform calcification+ choroid plexus enlargement	Tonic clonic seizures	58
2	42	F	W	R	1	2	None	None	None	N/A	Stable	30
3	<1	F	W	Bilateral (Laser Rx)	1, 2	1	Seizures	6	Carbamazepine	Left parietal and occipital lobe leptomeningeal enhancement, calcification	Mild cognitive impairment	40
4	2	F	W	R (Laser Rx)	1, 2	2	None	None	N/A	Normal	Extracranial PWS (R side chest, upper and lower limb+ hemihypertrophy)	86
5	11	F	W	L	1, 2	2	None	None	N/A	Normal	Hx of multiple melanoma excision +BCC+bilateral breast mastectomy	72
6	5	F	Unknown	R	1	2	None	None	N/A	Normal	Congestive HF at birth, microcytic anemia	<1
7	25	F	W	L	1	2	None	None	N/A	Normal	None	16
8	36	M	W	R	1	2	None	None	N/A	N/A	None	<1
9	10	F	H	R (Laser Rx)	1	1	Intractable daily Seizures, left Homonymous hemianopia	6	Controlled with Valproic acid + Levetiracetam (was uncontrolled with Carbamazepine)	Leptomeningeal enhancement with adjacent white matter changes, gyriform calcification, cerebral atrophy, enlarged choroid plexus and deep veins	Mild left hemiparesis	1
10	0	F	W	R	2	2	None	None	None	Normal	Extracranial PWS + R lower extremity hemihypertrophy + varicose veins + pectus excavatum	8
11	35	M	W	L	Unknown	2	None	None	None	Normal	Stable	0
12	<1	F	W	L	1	1	Seizure	1	Phenobarbital, oxcarbazepine (not well controlled)	Left cerebral hemisphere leptomeningeal enhancement (sparing left anterior Temporal lobe)	Developmental delay, right spastic hemiparesis	1
13	27	M	W	L	2	2	None	None	None	N/A	None	34

PWS= Port wine stain, SWS= Sturge Weber Syndrome, V= trigeminal nerve dermatome distribution, M= male, F= female, L= left, R= right, W= white, H= Hispanic, UL= upper limb, Hx= history, Rx= treatment, HF= heart failure, BCC= basal cell carcinoma, N/A= not available

^(a)Age first presented to Olmsted County

Description of ocular findings in our cohort is found in Table 3. Ocular features (n=13) found in SWS in this cohort included: dilated episcleral vessels (n=6, 46% [95% CI 19–75]), glaucoma (n=6, 46% [95% CI 19–75]), cataract (n=5, 39% [95% CI 14–68]), refractive error (n=5, 38%), strabismus (n=4, 31% [95% CI 8–61]), retinal detachment (n=3, 23% [95% CI 5–54]), band keratopathy (n=2, 15% [95% CI 2–45]), buphthalmos (n=2, 15% [95% CI 2–45]), amblyopia (n=2, 15% [95% CI 2–45]), DCH (n=1, 7.7% [95% CI 0.2–36]), epiretinal membrane (n=1, 7.7% [95% CI 0.2–36]), and hypotony maculopathy (n=1, 7.7% [95% CI 0.2–36]). Ocular involvement (n=9) was ipsilateral to PWS in (n=7, 78%), bilateral in bilateral PWS in (n=1, 11%), and contralateral to PWS in (n=1, 11%).

Table 3:

Data of patients with SWS at Olmsted County- Ocular features

No.	Agea (years)	Affected eye	Ocular symptoms	BCVA (DFS)	IOP mmHg (DFS)	Ocular signs (DFS)	Final BCVA	IOP mmHg (DLS)	Ocular outcome	Glaucoma onset (years) (if present)	Treatment	Follow up (years)
				OD OS	OD OS		OD OS	OD OS
1	N/A	N/A	Visual field defect	N/A	N/A	End gaze nystagmus	N/A	N/A	N/A	none	N/A	-
2	58	R	Redness and dryness	20/20 20/20	N/A	R- Dilated episcleral vessels and POA Myopia OU	20/20 20/20	14 12	R- POA Bilateral cataract surgery at age 72	57	Medical Rx (Cosopt bid, latanoprost 0.005% qhs)	14
3	19	L	Deviating eye	20/20 20/25	14 18	R- dilated episcleral vessels L- Dilated episcleral vessels , POA, early band KP, Large angle XT	20/20 OU	15 21	L- CDR 0.9	16	Medical uncontrolled on maximum therapy might need surgery (latanoprost, brimonidine tid, cosopt bid and pilocarpine tid.)	22
4	8	R	Decreased vision	20/40 20/20	N/A	R-Buphthalmos, dilated episcleral vessels, unspecified glaucoma, Amblyopia, Myopia	NLP 20/20	34 18	R- Band KP, NS+3, 0.8 CDR, macula-off RD, XT b	<8	Unknown past treatment	78
5	20	L	Decreased vision	20/20 20/100	N/A	R- narrow angle L-Buphthalmos, unspecified glaucoma, total cupping	20/25 Enucleated	18 Enucleated	R- Dry AMD, glaucoma suspect , NS+3 L- Enucleated	< 2	R- latanoprost drops, LPI L- unknown	64
6	5	R	Decreased vision, intermittent eye pain	N/A	9 10	R- Dilated episcleral vessels, iridocorneal touch	N/A	None	EUA showed RD with underlying diffuse orange choroidal hemangioma	None	None	No follow up
7	28	L	None	20/20 20/60	16 6	R- normal L- Dilated episcleral vessels, XT, iredectomy Hyperopia	20/20 20/50	11 8	L- POA, ERM with hypotonous maculopathy, posterior cortical opacity	11	Trabeculectomy at age 12	13
8	36	R	Redness	NLP 20/25	10 10	R- NS+4, dilated episcleral vessels, funnel shape RD	NLP 20/25	10 10	R- No view (B-scan showed funnel shaped RD with no underlying mass)	Unknown	Multiple unknown glaucoma procedures done for right eye, lost at age 15 y/o	<1
9	10	None	None	20/25 20/25	20 20	Normal	20/25 20/25	21 19	Left homonymous hemianopia	None	None	<1
10	<1	L	None	CSM CSM	NA	Persistent pupillary membrane OU Hyperopic astigmatism	20/20 20/200	21 19	L-Anisometropic amblyopia, XT	None	Patching	8
11	-	L	Unknown	N/A					Didn’t show for ophthalmology appointment	Unknown	Timolol drops	-
12	<1	None	None	CSM CSM	7 10	Normal exam Normal refraction for age	CSM CSM	7 14	MRI (showed faint posterior left globe enhancement- choroidal hemangioma), no evidence on exam	None	None	1
13	36	None	None	20/20 20/20	N/A	Normal exam Hyperopia	20/25 20/20	15 16	Mild vascular tortuosity OU, no DR	None	None	23

BCVA= best corrected visual acuity, IOP= intraocular pressure, DFS= date first seen in ophthalmology at Olmsted county, DLS= date last seen in ophthalmology at Olmsted county, OD= right eye, OS= left eye, OU= both eyes, N/A= not available, R= right, L= left, POA= primary open angle, KP= keratopathy, CDR= cup to disc ratio, RD= retinal detachment, NS= nuclear sclerosis cataract, AMD= age related macular degeneration, LPI= laser peripheral iridotomy, EUA= examination under anesthesia, ERM= epiretinal membrane, DR= diabetic retinopathy, XT= exotropia, y/o= year old, bid= twice daily, qhs= at bed time, Rx= treatment

^a)Age when presented to ophthalmology department at Olmsted County

^b)Subject 4 (Onset of ocular complications not clearly found in chart, no imaging for fundus found, last follow up showed RD macula- off with tears inferior and nasally)

Age of onset of glaucoma in those affected (n=6) was 0–8 years old (n=2, 33%), 11–57-years-old (n=3, 50%), or unknown (n=1, 17%). PWS was ipsilateral to the affected eye in (n=5, 83%) or bilateral in the V1 and V2 distribution with glaucoma only in the left eye (n=1, 17%). In 100% of glaucoma cases, PWS involved V1 (V1 only=50%, V1 and V2=50%). Concurrent glaucoma and neurological involvement were present (n=1, 17%). One patient had history of using timolol drops but no definite glaucoma; this case was not included in glaucoma analysis. At last follow-up, significant visual field loss occurred despite medical and surgical interventions (n=2, 33%) (Subjects 4 and 5), with enucleation required (n=1, 17%) (Subject 5). Disease was controlled with medical therapy alone (n=1, 17%), trabeculectomy or other surgical procedures (unspecified) (n=4, 67%), or initial medical therapy with possible future surgical intervention in (n=1, 17%).

Cataract was age-related nuclear sclerosis in both eyes (n=1, 7.7%) (Subject 2), secondary to glaucoma with buphthalmos (n=1, 7.7%) (Subject 4), age related nuclear sclerosis contralateral to PWS (n=1, 7.7%) (Subject 5), posterior subcapsular cataract after trabeculectomy (n=1, 7.7%) (Subject 7), or in the setting of funnel retinal detachment after multiple filtering surgeries for glaucoma (n=1, 7.7%) (Subject 8).

DCH (n=1, 7.7% [95% CI 0.2–36]) (Subject 6) was ipsilateral to PWS in the V1 distribution, with maximal choroidal thickening of 4.9 mm and overlying total exudative retinal detachment. Strabismus ( exotropia) was observed (n=4, 31%) (Subjects 3, 4, 7 and 10). Amblyopia was present (n=2, 15%), with (n=1, 50%) anisometropic amblyopia in the eye contralateral to PWS.

Best corrected visual acuity (BCVA) in the affected eye (or the eye ipsilateral to PWS) at date last seen in ophthalmology was analyzed for the cases with follow-up (n=13). BCVA was categorized as near normal (≤0.2 LogMAR, Snellen equivalent 20/20–20/30) (n=4, 31%), moderately impaired (>0.2-≤0.6 LogMAR, Snellen equivalent 20/40–20/80) (n=1, 7.7%), severely impaired (>0.6 LogMAR, Snellen equivalent ≤20/100) (n=3, 23%), or central steady maintained (CSM) (n=1, 7.7%). Enucleation was required in the setting of uncontrolled glaucoma (n=1, 7.7%).

Our literature review resulted in 408 articles. There were 345 articles excluded after reviewing the abstracts due to irrelevance to the topic of interest. There were 45 excluded after reviewing the abstracts (or manuscripts if needed) as they represented single case reports, review articles, or did not include details about ocular involvement in SWS cases. We ended up with 18 articles (11 retrospective reviews, 3 retrospective and cross-sectional studies, 1 cross sectional study, and 4 prospective studies) in compliance for our inclusion criteria. Detailed information can be found in Table 4 and the discussion section.

Table 4:

Literature review summary t

Author/ Year	Purpose	Methods and study population	Age & inclusion criteria	PWS	Neurological involvement	Ocular involvement	Conclusion	Comments
Enjolras O. et al/ Pediatrics 1985	-To find characteristics for PWS that can predict neuro-ocular involvement	-Retrospective -106 cases with facial PWS and associated neuro-ocular signs and symptoms -Department of Neuroradiology, Hospital Lariboisiere, and Department of Dermatology, Hospital Tarnier, Paris -1978–1984	-Age=N/A - Presence of facial PWS -Presence of facial photographs for topography analysis for PWS	-85% had PWS without neurological involvement or glaucoma (unknown if they had other ocular features) - 43% had only 1 branch of CN V involved -8.5% had associated extracephalic PWS	-11% had SWS type 1(33% of which had associated glaucoma)	-7.5% had glaucoma	-Only patients with V1 CN involvement are at risk for associated neuro-ocular symptoms - Proposed that SWS results from dysmorphogenesis of cephalic neuroectoderm	-Ocular involvement without neurological was not considered as SWS
Jerry Oakes W./Pediatr Neurosurg 1992	-To describe the natural history of patients with SWS	- Retrospective - 36 cases with SWS - Duke University Medical Center - (1930–1980)	-Age= N/A -Patients with pathognomic cortical calcification (Type 1 or 3)	-97% had PWS	-80% presented with seizures (range: 3 days-3.3 years, mean= 6 months) -83% had poorly controlled-uncontrolled seizures	-33% had glaucoma -6 other cases had ocular involvement only and were excluded from analysis	- Presence of seizures during first year of live is a predictor for poor outcome	- Exclusion of cases that were never admitted therefore selecting more severe cases. -Focused on natural history of seizures and developmental problems.
Sullivan J. et al./ J. Pediatr. Ophthalmol Strabismus 1992	-To review the ocular manifestations in patients with SWS	- Retrospective - 51 cases with SWS (53% F, 47% M, 92% Caucasian) - The Hospital for Sick Children, Totonto-Canada - (1980–1991)	-Age=N/A -Patients with complete trisymptomatic SWS or incomplete bisymptomatic SWS	-96% had PWS -51% showed extensive bilateral disease -22% had electrocautary or laser Rx for PWS -31% had associated KTS	-80% had epilepsy -33% hemiplegia -53% mental retardation -34% of patients with seizures required surgery (hemispherectomy or lobectomy) for seizure control	-71% had glaucoma (In 72%, onset was <= 2 y/o). Mean age in early onset= 5.83 months. -Surgery was need for glaucoma control in 78% of cases. -69% had conjunctival or episcleral hemangiomas -55% had choroidal hemangiomas (43% of them were bilateral) -14% had strabismus (75%ET, 25% XT) -10% had Iris heterochromia -10% had RD -67% had final VA >= 20/40 or CSM in each eye	-Goniotomy is the procedure of choice in early onset glaucoma - Medical Rx may be the initial treatment option in late onset glaucoma. However, surgery is often needed. -Presence of dilated episcleral vessels and/or DCH are predictors for high incidence of complications seen with filtering surgeries.	-Focused on ocular manifestations and management for ocular disease mainly glaucoma
Sujansky E.& Conradi S./ Am J Med Genet. 1995	-To analyze outcomes in SWS cases	- Retrospective and cross sectional - 52 adults with SWS (62% F, 38% M, 92% Caucasian) - Sturge Weber Foundation	- Age of participants: Range: 18–63 years, mean= 30.6 years. -Pre-existing confirmed diagnosis of SWS in medical records -Those who responded to questionnaire	- Craniofacial PWS in 98%, unilateral in 46%, bilateral in 54% (V1 was involved in 100%, V2 in 76%, and V3 in 60%). - Extracephalic PWS in 52%	-83% had seizures - Seizures onset given in 42 cases (range: 0–23 years, median= 6 months). - Full seizure control was achieved in 27%, partial control in 49%, and 3 cases needed surgical hemispherictomy for control - Intellectual functioning was normal in 45%	- 60% had glaucoma (unilateral 74%, bilateral in 26%); Age of onset in 30 patients (range: 0–41 y/o, median 5 years) - 44% had both seizure and glaucoma, in 15%= glaucoma only	- Delayed onset of glaucoma in 20% of glaucoma cases (>20 y/o), indicates that all SWS patients without glaucoma need periodic ophthalmology evaluation indefinitely. - Prevalence of glaucoma was highest in those with bilateral V1 & V2 involvement (87.5%) and lowest with unilateral V1 only (20%) -Seizures of earliest onset are the most difficult to control	- No analysis for ocular involvement other than glaucoma - Main focus was neurological, psychosocial and functional aspects - Compared to their previous report of 171 cases as shown below (including these 52 cases), adults showed higher prevalence of glaucoma (60% vs. 48%). - Further analysis for the offspring of these participants resulted in 17 healthy child. Tuberous sclerosis, café-au-lait spot, and a birthmark were found in 1 child each (no genetic testing was performed).
Sujansky E. & Conradi S./ Journal of Child Neurology 1995	-To obtain information about SWS through studying large number of cases (distribution of PWS, frequency and age of onset of seizures and glaucoma, correlation between seizures and developmental delay/retardation)	- Retrospective and cross sectional - 171 patients with SWS (53% F, 47% M, 92% Caucasian) - Sturge Weber Foundation	-Age of participants: range:2 months-59 years, median=8 years -Pre-existing confirmed diagnosis of SWS in medical records -Those who responded to questionnaire	-100% had facial PWS -45% had extracranial PWS -17% had other vascular or pigmentary lesions	-80% had seizures (onset: birth-23 years) -83% had developmental delay when seizure onset < 1 y/o -1 case had SWS type 3 -50% had full control with medical Rx, 39% had partial control -6.6% with seizure reuired surgery (hemispherectomy)	-48% had glaucoma (67% unilateral. 33% bilateral), median age= 12.1 years -61% onset by 1st year, 15% between 5–9 yrs. -4 cases had glaucoma onset >=27 y/o -92% of glaucoma cases had both V1 & V2 involved -1 case had glaucoma contralateral to side of PWS	-All patients with seizures and/or glaucoma had PWS involving V1 +/- V2 -Bilateral PWS had significant higher prevalence of seizures vs. unilateral PWS -Patients with later onset seizures had lower prevalence of developmental delay/mental retardation	-The risk of glaucoma was 2.5 times higher in patients with PWS distribution involving both V1 &V2 vs. isolated V1 -Possible later onset of glaucoma (3rd/4th decades) requires periodic indefinite ophthalmic evaluation - 12% had associated dysmorphic features or additional symptoms not related to SWS
Griffiths P. et al./ AJNR Am J Neuroradiol 1996	-To estimate the prevalence of abnormal ocular enhancement in MRI/CT scans in children with SWS and correlate this with clinical and fundoscopic findings	- Prospective - 15 children with SWS - Neuroradiaology department- Newcastle General Hospital (UK)/The Hospital for Sick Children and University of Toronto (Canada) - (1991–1994)	-Range: 2–48 months, mean= 9 months -The presence of typical PWS with seizures and neurological deficits - All cases were evaluated with enhanced CT or MRI	-100% had facial PWS	-100% neurological involvement - Imaging findings were assessed by 3 independent radioneurologists	-47% had abnormal ocular enhancement	- Both enhanced MRI and CT scan show DCH in cases with SWS. However, MRI is more sensitive and is recommended to aid the detection of abnormalities with preventable late complications.	-Ocular enhancement was shown to be positively correlated with bilateral disease, with the extent of facial involvement, and with glaucoma but not with hemispheric involvement
Awad AH. et al/ Journal of AAPOS 1999	-To review cases of SWS and to elucidate patterns of clinical practice in managing cases with glaucoma	-Retrospective -18 patients with SWS (33% F, 67% M) -Pediatric division, King Khalid Eye Specialist Hospital. Saudi Arabia (1983–1995)	- Range (1 day—20 years), median= 18 months - Minimum of 12 months of F/U was required for inclusion	-100% had PWS	- 13% of glaucoma cases had neurological involvement	-83% had glaucoma (of which 47% had bilateral glaucoma) -93% of glaucoma cases was early onset (with megalocornea) - In 67% surgery was required for glaucoma control -44% had choroidal involvement -22% had dilated episcelra vessels -17% had choroidal detachemnt -11% had retinal detachment	- Glaucoma in SWS is common - Medical Rx of glaucoma is considered the initial Rx of choice -Late post-operative complications can lead to loss of vision	-Mean F/U= 5.2 years (range= 1–12.3 years) - Choroidal detachment and retinal detachment was post-operative complication (3–5 months post op) - All included patients were admitted to hospital at some point -All cases with glaucoma had PWS involving upper eyelid
Celebi S.et al./ Eur. J. Ophthalmol.2000	-To review rare ocular findings in SWS and results of Ahmed valve surgery in cases with glaucoma (followed over 2 years)	- Prospective observational -7 cases with SWS (43% F, 57% M) - Firat Medical Center and Military Hospital in Elazigˇ- Turkey - (1996–1998)	-Range: 18–52 years, mean= 27.7 years	-100% had unilateral PWS -14% had Nevus of Ota	- details= N/A -All cases had CT scans and neurological evaluation	-100% prominent episcleral venous vessels - 43% had DCH -43% had juvenile glaucoma -14% had buphthalmos	- Rare ocular findings such as choroidal hemangioma and nevus of Ota are sometimes seen in SWS. Drainage valve implantation, with an anterior chamber maintainer, is a good choice for treatment when surgery is done in cases with glaucoma	-All glaucoma cases underwent valve implantation with AC maintainer -All patients had comprehensive eye exam (slit lamp, tonometry, gonioscopy, automated VF, anterior segment and fundus photographs)
Comi A. et al/ Arch Neurol 2005	-To develop hypotheses regarding the relationship between SWS and other unexpected abnormalities in a subset of patients	-Retrospective - 28 patients with SWS (18% F, 82% M), out of which 8 (25% F, 75% M) had associated abnormalities -Departments of neurology and pediatrics, The John Hopkins University, Baltimore/Department of Neurology, The University of Texas Southwestern Medical Center, Dallas	-Range:1–35 years, mean= 10.4 years -Review of medical records - Presence of associated unexpected abnormalities	-87.5% had facial PWS (out of 8 cases)	-100% had neurological involvement on neuroimaging	-37.5% had unilateral glaucoma	-Abnormalities (masses or malignancies, other malformations or dysmorphic features) found in (29%) patients and after literature review in association with SWS suggests testable insights regarding pathogenesis - Chromosome 17p1-p13 may be candidate region for genes involved in SWS. Further research is needed in this area	-Data from tertiary care centers -Detailed characteristics for neuro-ocular involvement was described for the 8 cases only -Unexpected associated abnormalities were: inguinal hernias, hydrocele, hypospadias, maxillary fibroma, hereditary neuropathy, cleft palate
Pascual- Castroviejo I. et al./Can. J. Neurol. Sci. 2008	-To review clinical and neuroimaging features of patients with SWS seen over a 40-year period	- Prospective - 55 cases with SWS (45.5% F, 54.5% M) - Pediatric neurology service of the University Hospital La Paz-Spain - (1965–2004)	-Unilateral or bilateral facial PWS - Neuroimaging evidence of ipsilateral hemispheric cerebral disease and occipital leptomeningeal angioma (demonstrated by MRI or SPECT)	- Unilateral PWS (63.5%), Left V1+ V2 involved in 43% of unilateral cases - Bilateral PWS (31%), - Laser therapy was applied in 22% cases (between 1989–2000)- then discontinued due to poor result	- 85.5% had seizures -75% had seizures during 1st year (mean= 8 months, range: birth- 6 yearrs) - Complete seizure control was observed in 42.5% (mainly medical Rx, in 2 cases they required lobectomy) - IQ testing was normal in 25% of patients, not tested in 22% of cases -42% had ADHD	-74.5% had glaucoma (bilateral in 34%) - MRI with GD-DTPA demonstrated choroidal angiomas in SWS with PWS affecting upper eyelid.	-Early onset of seizures, poor response to medical Rx, bilateral cerebral involvement or unilateral sever lesion were indicative of poor prognosis	- Main focus was on neurological/behavioral/intellectual outcomes of SWS -Frequent ophthalmic exam, medical or surgical Rx didn’t prevent vision loss as noted by authors
Hennedige A./Plast. Reconstruct. Surg. Journal 2008	-To verify the incidence of SWS and glaucoma within the population and to determine the response for Pulsed Tunable Dye laser in isolated PWS vs. syndromatic PWS	- Retrospective - 874 cases with PWS (61 % F, 39% M) - Laser department at St. John’s Hospital-Scotland - (1990–2004)	-Range (6 months- 77 years), mean=14.2 years -Pretreatment and post treatment photographs was done -Cases with neurological and/or ocular involvement + facial PWS	-100% had PWS	-Used CT to detect calcification for detecting SWS	-Occasional review by an ophthalmologist was performed in certain cases -Glaucoma was most common in cases with both V1 & V2 involved -No focus on other ocular features in SWS (and in cases with glaucoma, IOP was only used for screening purposes)	-SWS was found in 3.4% of cases -Laser treatment produced unsatisfactory outcomes in patients with facial dermatomal PWS. V3 PWS responded best and V2 worst to laser.	- This incidence could be underestimated as not all SWS cases pursue laser therapy -Identifying SWS patients on the basis of calcifications on CT scans might miss some cases, as some calcifications developing in leptomeningeal venous malformations occur later in life and are best detected by MRI.
Khaier A. et al/ Ophthalmology 2011	-To identify the risk factors sensitivity and specificity for developing glaucoma in patients with PWS	-Retrospective case-control study - 216 patients (57.4% F, 42.6% M) (252 eyes) with unilateral/bilateral periocular PWS - Eye department, Great Ormond St. Hospital, UK - (1981–2005)	-Range: 3 weeks-18.8 years, mean= 2.9 years -Patients were identified from dermatology department database where patient had laser Rx for PWS	-83.3% unilateral PWS -16.7% bilateral PWS -43% had both V1 & V2 involved - Extrafacial PWS was identified in up to 77% of bilateral facial PWS -100% had laser ablation Rx	-11% had definite SWS after neurological evaluation - Cases with neurological involvement only were considered as SWS cases (SWS type1)	-36% had glaucoma at presentation or during the F/U period (67% of SWS had glaucoma) -48% of glaucoma eyes required surgery -23.8% had choroidal hemangioma	- Iris heterochromia is associated with the development of early glaucoma in PWS cases - Cases that lack combined lid involvement or episcleral hemangioma have lower risk of developing glaucoma	-This hospital has a tertiary referral center for epilepsy -Statistically significant predictors for developing glaucoma (Bilateral PWS, both V1 & V2 involvement, episcleral hemangioma, iris heterochromia, DCH)
Piram M. et al/ Pediatr Dermatol 2012	-To assess PWS predictors for developing SWS	-Cross sectional - 259 cases with facial PWS (51% F, 49% M) (including 15 patients with SWS) -Pediatric dermatology or angiodysplasia department, th University Hospital Center of Tours, France - (1993–2005)	- Range: 6 days- 61 years), median= 8 years - Range in SWS: 1 month-39 years (mean 11.2) -Filling a standardized form for each patient with facial PWS -Reviewing the results of neuroimaging performed previously	-100% had facial PWS -40% of SWS had bilateral PWS	-73% of SWS had seizures -60% of SWS had mental retardation -27% of SWS had hemiparesis	-40% of SWS had congenital glaucoma -13% had CH -7% of SWS had strabismus	-5.8% of patients with facial PWS had SWS - All cases with SWS had involvement of V1 involvement -SWS was significantly associated with bilateral PWS involvement, involvement of upper eyelid, or extensive PWS involving at least 2 CN V branches	-Patients were seen in specialized dermatology/angiodysplasia clinics -Cerebral imaging was not performed for all patients and no follow up imaging were available. -All cases with glaucoma had PWS involving V1. However, 12.2% of cases with PWS involving V1 had glaucoma
Jagtap S. et al/ Journal of Child neurology 2012	-Evaluate the clinical manifestations, outcome, and natural history in patients with SWS	- Prospective longitudinal - 30 cases with SWS (50% M, 50% F) - Department of Neurology, Sree Chitra Tirunal Institute of Medical Sciences and Technology- India - (January 1985-Mayo 2010)	-Range (1 month-43 years), mean = 11 years -Presence of facial PWS -Calcification typical of SWS on (CT scan) or leptomeningeal angiomas on (MRI)	-86% had PWS (bilateral in 8%) with majority (47%) involving both V1 & V2 distribution	-100% had seizures -In 82%, seizure onset < 2 y/o (range: 1month-21 years) -Seizures well controlled in 73%. Surgery was needed in 3 patients - Male sex was associated with uncontrolled seizures - 13% had hemiparesis - 43% had normal IQ - 10% had ADHD	-30% had glaucoma (1/3 of them required surgical Rx, other 2/3 was medically controlled) - 56% of glaucoma patients developed vision loss impairing daily activities over a period of 2–6 years	- Further global studies are needed to understand this syndrome better - Severe degree of mental impairment was seen in cases with early onset of seizures and in seizure resistant to medical Rx	- Evaluated the clinical manifestations with special emphasis on seizures, their refractoriness to various drugs, clinical features if any which predicted outcome - CT was available for 23 patients, MRI was available in 13 patients -No details available regarding ocular involvement other than glaucoma
Sinawat S. et al/ J Med Assoc Thai 2014	-To review the clinical manifestations and neuroimaging features of patients with SWS	- Retrospective - 16 cases with SWS (44% F, 56% M) - Srinagarind Hospital- Thailand - (2001–2013)	- Range (0–35 years) at presentation - Patients had facial PWS with neurological and/or ocular complications. -Medical records, photographs, and neuroimaging studies were reviewed	-100% had PWS involving -25% had bilateral PWS - 1 case had KTS, another case was diagnosed with Phacomatosis pigmentovascularis - 2 cases had pulsed dye laser Rx	- 75% had neurological involvement (seizures, hemiparesis, headache, developmental delay) with seizures being most common features. Onset in 83% was during infancy - Epilepsy was controlled with medical Rx except in 1 case who required surgery - 14/16 had neuroimaging	- 62.5% had ocular involvement. -56% had glaucoma (in in 55% onset <= 4 y/o) - 44% had dilated episcleral vessels - 25% had corneal findings (haze, buphthalmos, megalocornea) - 18.75% had DCH (one of them presented with bullous RD)	- PWS, glaucoma, and seizures were the most common clinical features -Complete ophthalmic and neurological evaluation should be done at time of diagnosis - Multidisciplinary approach and lifelong F/U is needed	- Four eyes were finally blind (all had infantile glaucoma). - Genetic testing for patients with SWS and associated unusual findings is recommended
Dutkiewicz A. et al./ J. Am. Acad. Dermatol 2015	-To refine the cutaneous distribution of upper facial PWS as risk for SWS	-Prospective multicenter - 66 cases with SWS cases (47% F, 53% M) - Multicenter from 8 French pediatric dermatology departments - (2006–2012) -Mean of F/U= 30–40 months	-< 1 year -PWS should involve upper face with at least 1 cm2 surface area (V1 distribution) - MRI evidence or leptomeningeal involvement or seizures -Medical records, MRI, and photographs were collected and studied	-Identified 6 patterns of upper facial PWS -Patterns 5 (hemifacial) and 6 (median) PWS showed significant risk for SWS	-100% of SWS had seizures - 6% had suspicion on SWS on their first MRI	N/A	-Specific PWS distribution patterns are associated with increased risk of SWS	- 16.6% had SWS ( defines as upper facial PWS with neurological manifestations and cerebral MRI signs (16.6% had SWS) -Limitations: patients were recruited from referral dermatology centers, there was no specified age for doing MRI so some SWS might have been missed due to negative early MRI, and small number of SWS cases
Koenraads Y. et al./Acta Ophthalmologica 2016	-To describe visual outcomes in SWS patients	- Retrospective and review article - 33 cases with SWS (52%F, 48%M) - Dutch multicenter cohort in Collaboration with Dutch pediatric ophthalmology society - 2015	-Range of age at last F/U (3–26 years), median= 9 years -Neurological involvement -Presence of Ophthalmology visit to a society member	-81% had facial PWS (bilateral in 19%)	- 94% had epilepsy (37% required hemispherectomy or focal resection) -72% had developmental delay -53% had hemiparesis	-55% had glaucoma (bilateral in 25% of cases) -20% had DCH (bilateral in 33%, associated with RD in 38%) -VA was near normal (<=0.3 LogMAR) in 72% of cases -Homonymous hemianopia in 60%	- Description of VA and VF in SWS cases is scarce in literature. Impaired visual function in SWS is due to combination of glaucoma, DCH and cerebral injury. Both glaucoma and DCH are associated with worse VA. Due to progressive character of ocular and neurological involvement, adequate visual follow up (including proper VF) is recommended in these patients.	-Majority of cases were from national tertiary referral center for children with epilepsy, particularly those requiring epilepsy surgery -Young age of patients, cognitive impairment, or decreased VA might explain why most studies lack details about visual function (both through VF testing or VA) in SWS cases
Marana Perez A.I. et al./ Neurologia 2017	-To describe clinical and epidemiological characteristics and different treatments in cohort of SWS cases.	- Comparative retrospective and cross sectional - 13 cases with SWS (46%F, 54% M) - Pediatric neurology department at a tertiary hospital- Spain - (1998–2013)	-Range (1–27 years), mean= 15 months -Confirmed intracranial lesions	-61% had facial PWS present (R 23%, L 38%, and bilateral 7%) -No PWS in 30%	- 100% had seizures. Age of onset ranged (2 months-11 years). Majority presented at age <1 year - 23% had good seizure control with monotherapy. 1 case needed hemispherectomy surgery -46% had mental retardation	-77% had ocular involvement -46% had glaucoma --61.5% of glaucoma patients required surgery needed surgery -23% had strabismus -23% had choroidal hemangioma	- There are multiple clinical manifestations of SWS. Being familiar with all of them is vitally important for diagnosing and for Monitoring and proper treatment which will improve the quality of life for these patients.	- Continuous ophthalmic follow up is necessary even in cases where IOP was initially normal - Prognosis and functional impairment in SWS cases depend on the extension of leptominengeal angiomas and severity of eye impairment. Neurological function might deteriorate with age as well.

^*N/A: not available, F: female, M: male, CN V: trigeminal cranial nerve distribution, SWS: Sturge Weber Syndrome, PWS: port wine stain. ET: esotropis, XT: exotropia, RD, retinal detachment, CSM: constant-steady- maintained, KTS: Klippel–Trenaunay syndrome, VA: visual acuity, F/U: follow up, VF: visual field, SPECT: Single-photon emission computed tomography, ADHD: attention deficit hyperactivity disorder, GD-DTPA: gadolinium-diethylenetriamine pentaacetic acid, Rx: therapy or treatment, CH: choroidal hemangioma, IOP: intraocular pressure.

DISCUSSION

Pathogenesis and molecular genetics

The SWS is characterized by hamartomas of the eye, skin, and brain that manifest at different times during life. Some patients develop all features, while others show partial expression of the syndrome (3,10,19,20). The presumed pathogenesis of SWS is early embryological malformation of vascular development, affecting skin, brain, and ocular structures (16,21). The proximity of the ectoderm destined to form the upper facial skin, the part of the neural tube that forms the parieto-occipital brain and leptomeninges, and the optic vesicle during embryogenesis could explain the triad of clinical manifestations of SWS (22), with variable degrees of involvement of each of these components (23,24). However, the exact genetic molecular basis for SWS was not described until 2013 when Shirely et al. identified a missense somatic mosaic activation mutation in GNAQ gene (c.584 G→A) (codon R183) in 88% of 26 participants with SWS (using port-wine stained skin or brain tissue samples) (25). This mutation was also described in 4 cases with SWS type 3 (forme fruste SWS) by Hildebrand et al. few years later (26). In a recent publication in JAMA ophthalmology by Bichsel et al, the same mutation was found in a choroidal blood vessel sample (the source of which was DCH associated with SWS from an enucleated eye). Francis et al confirmed this observation in their recent report when a case with DCH (associated with SWS) was reported to have same mutation at the R183Q codon in the GNAQ gene. However, GNAQ mutation in another codon was identified in cases with solitary choroidal hemangioma or choroidal nevus (27). In our cohort, GNAQ mutation testing was not performed in any case as this was a retrospective study and most cases presented prior to the era of describing GNAQ testing in SWS cases. However, documentation for genetic testing was found in a single patient (n=1, 7.7%) which revealed chromosomal duplication at 5q12 of uncertain significance. Previous reports described detection of GNAQ mutations in SWS patients through biopsy of the PWS, enucleated eyes, or brain tissue. We agree that the least invasive tissue sample should be taken, and genetic testing might help guide management in cases with incomplete expression of the syndrome by detecting patients at risk for developing ocular disease later in life. Closer follow up might be recommended for patients with known GNAQ mutations.

Incidence

The exact incidence of SWS is unknown (10,22), with estimates of 1 per 20,000–50,000 live births in the literature (1,2,4,9). Herein, we investigated the incidence of SWS in a population based cohort from Olmsted County, Minnesota, USA in order to gain a better understanding of the epidemiological features of SWS. The age and sex adjusted incidence of SWS in Olmsted County was 0.19/100,000/year [0.04–0.34].

Most previous studies report the risk of developing SWS in a subset of cases with facial PWS. In 1985, Enjolras et al. reported that up to 11% of patients with facial PWS developed SWS type 1. Khaier et al. reported similar incidence of SWS type 1 as well. However, Hennedige et al. and Pirma et al. reported a lower incidence of SWS (3.4% −5.8%) in cases with facial PWS who were seen at dermatology departments. This could be an underestimation as not all patients with PWS seek laser treatment, especially those with neurologic or ophthalmic pathologies (28). In most previous studies, an underestimation in detecting SWS can be expected as they included only cases with neuroimaging evidence of brain involvement (SWS type 1). Therefore, cases with ocular involvement only (SWS type 2) might have been missed. Also, delayed onset of neurological involvement might be a contributor to this underreporting as well.

Demographics

In a series of 52 adults with SWS by Sujanski et al, 92% were Caucasian, 62% female, and 38% male (approximately 3:2 F:M ratio). In another series of 171 patients with SWS by the same authors, 92% were Caucasian, 53% female, and 47% male (approximately 1:1 F: M ratio). In a series of 55 patients with SWS by Pascual-Castroviejo et al, 45% were female and 55% male.

In our series, (n=11, 85%) were Caucasian vs. (n=1, 7.7%) Hispanic which is consistent with the Olmsted County population where the majority of residents are Caucasian. There was no clear sex predilection (F: M ratio was 2:1), which was consistent with what was described in the literature.

Systemic features (cutaneous and neurological)

Vascular malformations involving the skin and/or brain lead to development of PWS (birthmark) and characteristic leptomeningeal angioma, consecutively. Brain involvement results in epilepsy and other neurological features such as hemiparesis, hemianopia, or cognitive impairment. Sujanski et al. reported craniofacial PWS in all 52 cases with SWS except 1 (98%), and neurological involvement was reported in 65%. In 58% of cases, all V branches (V1, V2, and V3) were involved, with V1 involved in all cases (28). In another series of 171 patients with SWS by the same authors, seizure was present in 80% with age of onset during the first year of life in 75% and extracranial PWS in 45%. All patients with seizure and/or glaucoma had PWS in the V1 distribution alone or in combination with V2 distribution, and patients with bilateral PWS had higher prevalence of seizure (29). Pascual-Castroviejo I. et al. reported that craniofacial PWS was present with V1 involvement in 95% of cases (with unilateral V1 and V2 combined distribution being the most common). Seizure affected 85.5% of cases (16). In another series of 158 patients with facial PWS by Ch’ng S. et al, 9% had associated neurological and/or ocular involvement. Of cases with entire V1 distribution involvement, 78% had neurological and/or ocular involvement (22), consistent with the known increased risk for ocular and brain disease when PWS involves the eyelids (V1 distribution) (4,16,30,31).

In our series, PWS was identified in all (100%) cases, with isolated V1 as the most common distribution (54%). V1 distribution of PWS was present in all cases with glaucoma, DCH, and neurological involvement. Neurological involvement was noticed in (31%) with seizures being the most common presenting symptom, onset of which was during the first year of life in all cases. Other neurological features included hemiparesis, cognitive impairment, homonymous hemianopia, or tremor. Neuroimaging was not performed in (n=3, 23%) as of the last follow-up visit. Another explanation for fewer cases with neurological involvement compared to previous series might be related to the methodology and inclusion criteria followed (refer to Table 4). In both reviews for Sujanski et al, they used the Sturge Weber Foundation database and further data was obtained through written questionnaires, telephone and reviewing medical records. All included cases had neurological involvement clinically or as evident through neuroimaging excluding cases with ocular involvement only (SWS type 2). This method might have preferentially attracted severely affected individuals or may be biased toward families with higher educational or socioeconomic status. In the Pascual- Castroviejo prospective review of 55 SWS cases in a pediatric neurology service- in Spain, only cases with confirmed neuroradiographic evidence of cerebral involvement were studied and followed over a 40 year period. Again tertiary referral center bias might have shifted the results toward higher prevalence of neurological involvement with less emphasis on ocular involvement. However, it is considered a valuable source for neurological and intellectual outcomes in SWS given the prospective nature and standardized follow up protocol.

Finally, Klippel–Trenaunay syndrome (KTS), a sporadic mesodermal phakomatosis characterized by a triad of capillary malformations, atypical varicosities or venous malformations, and bony or soft tissue hypertrophy affecting limbs, can overlap with SWS as noted in previous reports (6–31% of SWS cases) and was identified in (n=2, 15%) of our cases (32–34).

Ocular features

Vascular abnormalities of the episclera, retina, and choroid constitute the underlying culprit for ocular involvement which can manifest as dilated episcleral vessels, iris heterochromia, glaucoma, and DCH. Other ocular manifestations such as refractive error, strabismus, retinal detachment, megalocornea, or cataract can be secondary outcomes to the underlying ocular pathology (such as glaucoma or DCH) or as a post-operative complication.

In a series of 13 patients diagnosed with SWS by Marana Perez et al, ocular disease was found in 77% of cases. The most common conditions were: glaucoma (46%), strabismus (23%), and choroidal hemangioma (23%) (14). In a series of 16 patients with SWS by Sinawat et al, ocular disease was found in (62.5%). The most common conditions were: glaucoma (56%), refractive error (50%), dilated episcleral vessels (44%), megalocornea and buphthalmos (25%), DCH (19%), and retinal detachment (6%). Sullivan et al. reviewed ocular findings in 51 cases of SWS and found glaucoma (71%), dilated episcleral vessels (69%), DCH (55%), iris heterochromia (10%), retinal detachment (10%), and strabismus (14%) (6). Pathologic ocular changes were usually ipsilateral to PWS (1,2,5,9,14,19).

We found similar ocular features of our study. The most common ocular features were dilated episcleral vessels (46%), glaucoma (46%), retinal detachment (23%), and DCH (7.7%). Ocular involvement was ipsilateral to PWS in most (78%) cases. Tertiary referral center bias (16,17,22,28,35) might explain the high frequency of DCH (70%) and glaucoma (75%) in the literature as a retrospective review of medical records for patients with SWS was done in all previous studies in the setting of tertiary referral hospitals (6,14,18,19,28). The frequencies found in our series might more accurately represent those in population-based study. In the Sullivan et al. study, cases were recalled for re-evaluation. Then, a comprehensive ophthalmologic evaluation (including visual acuity, alignment & motility, corneal diameter, gonioscopy, IOP, direct and indirect ophthalmoscopy, visual field testing) was performed when possible. This approach might explain the higher incidence compared with other studies for bilateral disease, glaucoma, conjunctival & episcleral hemangiomas, and choroidal hemangiomas as all patients had standardized ophthalmic assessment. They also emphasized the importance of including cases with incomplete expression of the SWS triad because they share same disease process and proposed embryologic origin. Hence, SWS cases with incomplete expression of the typical triad have been included in our cohort.

Glaucoma is the most common ocular complication reported in the literature (30–75%), with age of onset from birth to 41 years (28). Approximately 60% of cases present in infancy and 40% develop glaucoma in childhood or early adulthood (4). Glaucoma is not always ipsilateral to PWS and hence both eyes should be evaluated (28). The main mechanism of glaucoma is either angle anomaly or increased episcleral venous pressure due to episcleral arteriovenous shunts or both. This might explain the bimodal age distribution in glaucoma cases (6,18,36). Our cohort showed similar bimodal age distribution.

Another classic ocular manifestation of SWS is DCH, which can be found in 20–71% of SWS cases and might be a risk factor for glaucoma development (1,4,6,14,17–19,37,38). DCH results from angiomatous changes affecting choroidal vessels, and, as it was shown recently by Bichsel et al., DCH shares the same molecular genetic basis as capillary malformation (27,39). This observation led to the proposal that choroidal hemangioma in SWS patients is a misnomer and it should be replaced by the term “choroidal capillary malformation” as it lacks the distinguishing features for infantile hemangiomas such as glucose-transporter 1 marker (39). Clinically, DCH gives the entire posterior choroid a deep red or red-orange appearance (1), which is not always detected on routine ophthalmoscopy without careful comparison to the fellow eye (40). This could have contributed to the low prevalence of DCH in our series. The generalized choroidal thickening caused by DCH is better appreciated by B-scan ultrasonography (3,37). Enhanced depth SD-OCT can show increased choroidal thickness in cases without clinically visible choroidal hemangioma (41). However, in young uncooperative children examination under anesthesia is mandatory for proper evaluation. Contrast-enhanced MRI can also aid detection of DCH with or without overlying RD. DCH is usually ipsilateral to PWS (1,19,42), but has been reported in the contralateral eye (43). In a majority of cases, DCH remains asymptomatic early in childhood with later increase in thickness toward adolescence and adulthood (1,41,44). Vision loss can result due to refractive error (mainly hypermetropic shift), overlying retinal changes, or exudative retinal detachment (1,6). In our cohort, we identified only (n=1, 7.7%) with definite DCH who presented at age of 5 years with decreased vision and intermittent eye pain. DCH in this case was complicated with total exudative retinal detachment. However, glaucoma was not ruled out in this case due to lost follow up. In another case (n=1, 7.7%), enhanced orbital MRI showed evidence for posterior globe enhancement suggestive for possible DCH; however, clinical examination did not confirm the diagnosis and further follow up was recommended.

Visual acuity outcome in SWS with ocular involvement can be poor. Koenraads et al., in a review article of visual outcome in SWS, reported up to 70% of all SWS cases had near normal BCVA (≤0.3 LogMAR, Snellen equivalent 20/20–20/40) (45). This percentage decreased to 47% in patients with glaucoma and 15% in patients with DCH. In our cohort (n=13), (n=6, 46%) had near normal to moderately impaired BCVA or CSM vision at last follow up, with (n=3, 23%) who had no final BCVA recorded due to no follow up or unavailable ophthalmologic exam. Of those patients with glaucoma, (n=2, 50%) had near normal to moderately impaired BCVA which is close to the results from Koenraads et al. Obtaining VA in SWS can be challenging due to young age at presentation or cognitive impairment in certain cases(16,45–47).

Management

Patients with SWS require an interdisciplinary approach to management, with neurology for seizure control, physical and occupational therapy for motility impairment, dermatology for PWS treatment, pediatrics for general medical health follow up, psychiatry and cognitive behavioral therapy for possible associated psychiatric disorders and depression, and ophthalmology for management of ocular complications.

Management options for glaucoma include topical medications, filtering surgery (goniotomy, trabeculotomy, trabeculectomy, tube shunt), and cyclodestructive procedures (48 37,49). Glaucoma secondary to SWS remains a challenging disease with low success rate due to severe surgical complications, including choroidal effusion, expulsive hemorrhage, and exudative retinal detachment, with intraocular pressure remaining refractory to multiple treatments in some cases (50,51). In our series, (n=2, 15%) developed retinal detachment in the setting of previous multiple glaucoma surgeries. In another case (n=1, 7.7%), the patient developed hypotonous maculopathy with ERM post trabeculectomy, but vision was maintained at 20/50 at last follow up visit. Moreover, cases with early onset glaucoma and buphthalmos had worse visual acuity outcome (noting that subjects 4 and 5 with poor final visual outcomes were treated decades ago when modern surgical techniques in treating glaucoma might have not been applied). Therefore, long term follow up is essential to facilitate early intervention and management of complications in these patients, such a cataract, retinal detachment, and glaucoma, to prevent vision-threatening disease progression.

Treatments for exudative retinal detachment secondary to the DCH include oral propranolol, external beam radiation, proton beam therapy, brachytherapy, photodynamic therapy, and possibly anti-VEGF (52–54). Similar to glaucoma, DCH can be refractory to multiple treatments and cause permanent vision loss (52,55,56). The advantage of using proton beam therapy in children is in sparing the contralateral orbit and brain from radiation and with less radiation related complications such as optic neuropathy or retinopathy. Proton beam has been modified lately to be applicable for young uncooperative children and showed success in resolving total exudative retinal detachment and regression of DCH in a few cases (55,56). However, more long term follow up is needed to evaluate outcomes. DCH case in our series did not receive treatment given poor visual prognosis and poor adherence to follow up.

Glaucoma, DCH, and brain pathology have a progressive nature, or possibly delayed onset; hence, adequate follow up with periodic ophthalmic exam and neurological evaluation is the general recommended approach.

Limitations and strengths

Limitations to this study include its retrospective nature and use of the Olmsted County population, which is predominantly Caucasian and may not be representative of the worldwide population. To identify cases of SWS, related diagnosis codes were searched in the REP database, but some cases could have been missed without manual review of all records. Not all patients were evaluated by a multidisciplinary team (including neurologist, ophthalmologist, and pediatrician), which might explain why certain ocular and or neurological findings were not identified or diagnosed at as high of a frequency as other studies. Moreover, SWS children can be neurologically normal for several months or less commonly years after birth, which could mean some cases in young patients were missed (5,9). Thus, the incidence of SWS in Olmsted County is at least as high as reported in this study, but it could be higher. Finally, the small size of our cohort in this series precluded definite conclusions regarding ocular involvement or outcomes of SWS on a population-based level.

Unlike many prior studies, we did not rely on tertiary center or specialty clinic data, which likely have an overrepresentation of advanced disease and might be biased toward complications of interest to the subspecialty. Instead, we utilized the REP database, which can be considered a close approximation for Olmsted County population. Moreover, caution was paid during literature review due to variable definitions and inclusion criteria for SWS throughout the years. We selected inclusion criteria consistent with the most agreed upon standards for SWS. To our knowledge, this is the first community based epidemiological report. Further studies are warranted to determine if these findings hold in other regions.

CONCLUSION

In summary, we found that SWS affects approximately 0.19 per 100,000 individuals per year in Olmsted County. The most common clinical features found in our cohort were: PWS (100%); neurological involvement with seizures (31%); and ocular involvement (69%) of dilated episcleral vessels (46%), cataract (39%), glaucoma (46%), retinal detachment (23%), DCH (7.7%), and strabismus (31%). Early diagnosis, proper intervention, lifelong follow up, and a multidisciplinary approach to care will optimize the systemic and ocular outcomes in these patients. Future studies should focus on refining the definition of SWS, perhaps with the aid of genetic testing, confirming the true incidence in a more diverse population, and developing multidisciplinary screening guidelines.