Refractory Seizure in a Patient With Griscelli Syndrome: A Unique Case With One Mutation and a Novel Deletion

Griscelli syndrome (GS) is a rare syndrome characterized by hypopigmentation, immunodeficiency, and neurological features. The genes Ras-related protein (RAB27A) and Myosin-Va (MYO5A) are involved in this condition's pathogenesis. We present a GS type 1 (GS1) case with developmental delay, hypotonia, and refractory seizures despite multiple medications, which included clobazam, cannabinol, zonisamide, and a ketogenic diet. Lacosamide and levetiracetam were added to the treatment regimen, which decreased the seizures' frequency from 10 per day to five per day. The patient had an MYO5A mutation and, remarkably, a deletion on 18p11.32p11.31. The deletion was previously reported in a patient with refractory seizures and developmental delay. We reviewed all cases of GS that presented with seizures. We reviewed other cases of GS and seizures described in the literature and explored possible seizure mechanisms in GS. Seizure in GS1 seems to be related directly to the MYO5A mutation. The neurological manifestations in GS2 seem to be caused indirectly by the accelerated phase of Hemophagocytic syndrome (HPS), which is characteristic of GS2. By having the MYO5A gene mutation combined with the 18p11.32p11.31 deletion, the prognosis and severity of the patient's condition are poor. This is the first report of GS1 with a deletion in 18p11.32p11.31.


Introduction
Griscelli syndrome (GS) is an autosomal recessive disorder characterized by hypopigmentation, immunodeficiency, and neurological features [1]. The syndrome was first described in 1978 [2]. Only 150 cases have been reported to date, with a higher prevalence in Mediterranean countries such as Turkey [2][3]. GS usually manifests from four months to four years, but other studies have reported a range from one month to eight years [2].
Two genes on chromosome 15q21, Ras-related protein Rab-27A (RAB27A) and Myosin-Va (MYO5A), are the cause of GS [1]. GS type 1 (GS1) presents with partial albinism and neurological features while GS type 2 (GS2) presents with partial albinism and immunodeficiency [4]. GS type 3 (GS3) is caused by a mutation in chromosome 2q37.1, which encodes melanophilin; the disease is usually benign [1]. RAB27A has an immunologic effect. The gene is expressed in cytotoxic T cells and natural killer cells [4]. The gene regulates the docking of proteins and exocytosis of granules containing granzyme and perforins. The dysfunction in cytotoxic T cells and natural killer cells explains the immunodeficiency seen in GS2. The gene MYO5A regulates the organelle transport in melanocytes and neuronal cells [5].
GS is an extremely rare disorder. Knowing the peculiarities of the clinical presentations is difficult because there is not sufficient information. We present a case of GS1 with an additional deletion (18p11.32p11.31). This deletion was previously reported by Verroti et al. [6]. This patient presented with similar clinical features as our patient, including drug-resistant atypical absence epilepsy and severe developmental delay [6]. The main concern of the patient was refractory seizures. We have presented the differential diagnosis of the disorder, differentiated each type of GS, and reviewed cases of GS with seizures. Finally, we have discussed the possible causes of seizures and the prognosis in each type of GS.
A three-year-old male was brought to the hospital by his mother due to increased seizure frequency despite being on multiple anti-seizure medications. The patient was admitted for further assessment. Past medical history was relevant for epilepsy, infantile spasms, developmental delay, gastroesophageal reflux, laryngomalacia, and dysphagia. The mother denied any prenatal or neonatal complications. A genetic test showed a mutation in the MYO5A gene of chromosome 15q21, suggesting GS.

Seizure history
The child was diagnosed with infantile spasms at 11 months of age, and he was successfully treated with prednisone. He was started on clobazam for seizure prophylaxis, but he developed myoclonic seizures after two months. Clobazam dose was adjusted, which helped reduce the frequency of the seizures.
At 16 months of age, he started having tonic-clonic seizures and was started on a ketogenic diet, which decreased the seizures' frequency from 20 episodes to 10 episodes a day. He was also given cannabinol at age two, which decreased the seizures' frequency to less than five per day. However, seizures recurred again with frequencies above 10 per day.
At two years and three months, he was started on levetiracetam without any significant change. Three months later, zonisamide was given without any significant effects. The child's neurologist felt that the ketogenic diet was no longer helpful, and, therefore, the patient was weaned off the diet. The dose of zonisamide was increased, which reduced the seizures' frequency to four to five episodes a day.

Magnetic Resonance Imaging (MRI)
There was poor differentiation of the gray-white matter interface in the temporal and frontal lobes, with atrophy in the frontal lobe and infratentorial region. Secondary ex-vacuo ventriculomegaly was also seen. Figure 1 shows the MRI findings in this patient.

Electroencephalogram (EEG)
At times, there was high amplitude with irregular spikes and waves. Then, the spikes and waves were immediately followed by brief diffuse attenuation +/-overlying fast activity, often without a clear clinical correlation but at times with an associated myoclonic jerk. At least eight of these seizures were noted within 24 hours. Impression: The EEG was markedly abnormal when awake and asleep. There were abundant multi-focal and diffuse epileptiform discharges and multiple recorded seizures with multifocal onset. Findings were consistent with epileptic encephalopathy and a predominance of myoclonic or brief tonic seizures.

Genetic testing
The first mutation was already reported, and an additional test shows an additional deletion of unknown significance.
Impression and treatment: The ketogenic diet was weaned due to the lack of efficacy. He was discharged with the following medications: lacosamide 40 mg BID, clobazam 10 mg TID, levetiracetam 500 mg BID, cannabinol 500 mg, zonisamide 50 mg, and diazepam PRN.
The prognosis of the patient is mainly unknown due to the natural course of the disease. The patient also had an additional mutation [loss 18p11.32p11.31 (2.1 MB)], which could be worsening the condition.

Discussion
Chédiak-Higashi syndrome, Elejalde syndrome, Hermansky-Pudlak syndrome type 2, and GS share similar features. All the syndromes present with partial albinism. Neurological features can be seen in all the differentials, except GS3. Immunodeficiency is seen in all the disorders except in Elejalde syndrome and GS1.
Making the differential diagnosis of these three syndromes could be challenging. Table 2 showed the main differences among these three disorders [4,[7][8][9].

Clinical Manifestations Genetics Epidemiology
Chédiak-    [4,18]. Interestingly, in one report, two siblings had the same mutation, and both were diagnosed with GS1, but just one sibling had epilepsy, showing that different phenotypes could be present with the same mutation [18]. Five out of eight cases of GS2 died shortly after developing seizures. One patient developed seizures and progressed to a vegetative state [16]. Another patient continued to have refractory epilepsy a year later; the first episode.
There is no description of what medication was tried [11]. Our patient has refractory epilepsy, and he was treated with more medications than other previously described cases.
There have only been 20 cases reported of GS1 based on our literature review, so information is limited. The most common neurological manifestations are hypotonia, seizures, neurodevelopmental delay, and ophthalmological features [4]. In our case, the patient did not have ocular manifestations. Once the diagnosis of GS1 has been made, it is important to differentiate it from GS2, which can also cause neurological symptoms. Congenital cerebellar atrophy is the most common MRI sign in GS1 [10]. The main findings in our patient were cortical atrophy, cerebellar atrophy, and ex-vacuo ventriculomegaly.
The neurological features of GS1 are related to the gene MYO5A, which regulates the organelle transport in melanocytes and neuronal cells [5]. The abnormal movement of melanosomes altered intracellular vesicle transport, so fast axonal transport in nerve cells is reduced. Neuronal development, axonal transport, dendritic spine structure, and synaptic plasticity altered [19]. The neurological features of GS2 are related to the accelerated phase of HPS due to histiocytic infiltration in the central nervous system [10]. The gene regulates the docking of proteins and exocytosis of granules containing granzyme and perforins. The lack of granzymes and perforins explained the immunodeficiency in GS2 [4]. Neurological manifestations in GS2 can be a sign of an accelerated phase of HPS [10].
Microscopic examination of the hair shafts allows differentiating GS from similar disorders like Chediak-Higashi and Elejaldes syndrome [1]. In GS, there are large clumps of melanin distributed irregularly. In Chediak-Higashi, there are small clumps of melanin regularly distributed, and in Elejaldes syndrome, the clumps of melanin are distributed very similarly to GS [1]. Overall the prognosis is poor, and the limited information about GS1 left physicians without many options. Currently, bone marrow transplantation seems to be the only definitive treatment for GS [15]. Cannabinol was the anti-seizure medication more effective in the patient. There are no studies about GS and cannabinol. However, in other epilepsy syndromes of childhood, such as Dravet's syndrome and Lennox Gastau, cannabinol reduced the frequency of seizures by 37.2% in Lennox Gastau and 38.9 in Dravet's syndrome [20].
Interestingly, our patient also had a deletion on 18p11.32p11.31. This finding is remarkable because there is a case report by Verroti et al. describing a case with a deletion in 18p11.32p11.31. This patient presented with similar clinical features as our patient, including drug-resistant atypical absence epilepsy and severe developmental delay [6]. One crucial gene found in this region is the TGIF-1, expressed very early in the nervous system. Mutation in the TGIF-1 has been related to severe mental retardation and holoprosencephaly [6]. While seizures are expected in GS1. The frequency and severity of the seizures in our patient are more severe than other cases of GS1 with seizures. We believe that mutation in the MYOV5A gene combined with the deletion 18p11.32p11.31 in our patient explained the severity of the case.

Conclusions
We found four cases of GS1 with seizures and eight cases with GS2 and seizures. The neurological manifestation is GS1 is related to the MYO5a while the GS2 neurological manifestations are indirectly related to the HPS when there is an accelerated phase. The differential diagnosis between Chédiak-Higashi syndrome, Elejalde syndrome, and GS could be challenging. The microscopic finding is the most accurate method to differentiate among these disorders. The prognosis of these patients is poor, the same as in our patient. The mutation in the MYO5A gene and the 18p11.32p11.31 deletion could explain the severity of the seizures in our case. This is the first report of GS1 with a deletion in 18p11.32p11.31.