SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Aim  We report on seizures, paroxysmal events, and electroencephalogram (EEG) findings in four female infants with pyridoxine-dependent epilepsy (PDE) and in one female with pyridoxine phosphate oxidase deficiency (PNPO).

Method  Videos and EEGs were analysed and compared with videos of seizures and paroxysmal events archived from 140 neonates. PDE and PNPO were proven by complete control of seizures once pyridoxine or pyridoxal 5′-phosphate was administered and by recurrence when withdrawn. Mutations in the antiquitin gene were found in three patients and in the PNPO gene in one child.

Results  Seizures began within 48 hours after birth in four newborns and at age 3 weeks in one. Frequent multifocal and generalized myoclonic jerks, often intermixed with tonic symptoms, abnormal eye movement, grimacing, or irritability, were observed in all infants with PDE and PNPO, but rarely in the other archived videos of neonates. EEGs were inconstant and frequently no discernable ictal changes were recorded during the seizures and the paroxysmal events. In addition, interictal EEGs were inconclusive, with normal and abnormal recordings. In older children tonic–clonic seizures, abnormal behaviour, inconsolable crying, frightened facial expression, sleep disturbance, loss of consciousness, paraesthesia, or intermittent visual symptoms were described during controlled and uncontrolled withdrawal or insufficient dosage.

Interpretation  PDE or PNPO should be considered in infants with prolonged episodes of mixed multifocal myoclonic tonic symptoms, notably when associated with grimacing and abnormal eye movements.

List of Abbreviations
PDE

Pyridoxine-dependent epilepsy

PNPO

Pyridoxine phosphate oxidase deficiency

Pyridoxine-dependent epilepsy (PDE) and pyridoxine phosphate oxidase deficiency (PNPO) are rare diseases that necessitate rapid diagnosis and appropriate treatment. Recently, genetic and biochemical diagnostic tools have become available to confirm the diagnosis.1–3 However, as these results are not available for several hours, administration of pyridoxine and pyridoxal 5′-phosphate is still the method of choice to treat patients suspected of having the diseases and to make a preliminary diagnosis (folinic acid-responsive seizures have recently been shown to be identical to PDE and to respond to pyridoxine).4 Administration of pyridoxine in neonates with drug refractory seizures is widely accepted. However, pyridoxal 5′-phosphate is not licensed in many countries and often not readily available in the hospitals.

In 2004 we retrospectively reviewed 105 neonatal seizures and paroxysmal events archived in our video database.5 Videos of two patients differed markedly from the others and both patients became free from symptoms immediately after administration of pyridoxine. Their unique pattern of paroxysmal events prompted us to suspect PDE in other neonates with similar symptoms. Since 2004 we have identified two additional children with PDE and one with PNPO. The diagnosis of PNPO was made because the paroxysmal events were typical of PDE but were unresponsive to pyridoxine. Here we report on the seizures, paroxysmal symptoms, and electroencephalograms (EEGs) of five patients with PDE or PNPO.

Method

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

We reviewed the case histories and EEG reports of four patients with PDE and one patient with PNPO retrospectively and analysed their ictal videos and their EEGs. In addition, we anonymized all digital ictal and interictal EEGs, and one of us (GW) reanalysed them (blinded). Results were compared with the original reports and the unblinded analysis and dissenting opinions were discussed. Some details of our patients have already been published in other papers: patients 1 and 3 correspond to patients 9 and 8 in Plecko et al.,3 patient 3 corresponds to ‘family G’ in Mills et al.,1 and patient 5 corresponds to patient 3 in Hoffmann et al.6 The study took place at the University Children’s Hospital in Zurich (Switzerland). Approval to perform the study was obtained from the institutional review board of the hospital. Patients and carers have given informed consent to publication of the results and the videos.

The video-EEGs of patients 2, 3, 4, and 5 were recorded with a 21-channel standard system (for details see Schmitt et al.5). The EEGs of patient 1 were recorded on paper by a 21-channel machine and stored on microfilm. The video was recorded simultaneously but not synchronized with the EEG. The amplitude-integrated EEG (aEEG) in patient 4 was recorded with a two-channel device (BRAINZ; GE Healthcare, Chalfont St Giles, UK).

In addition, we reanalysed seizures and paroxysmal events in 140 neonates videotaped and archived between 1991 and 2009 (Table I) and compared them with the symptoms observed in our study patients (Table II).

Table I.   Symptoms of seizures and paroxysmal events in the reference group
AetiologyNumber of patientsClonicTonic–clonicMyoclonicTonic and spasmsAbnormal eye movementOralComplex movementsTremorMixedc
  1. aHealthy neonates with sleep myoclonia, neonates with hyperekplexia or Marcus Gunn phenomenon. bParenthesis: numbers of patients with very frequent and multifocal myoclonic jerks. cMyoclonic/clonic jerks mixed with tonic movements or spasms or abnormal eye movements.

Total14043749 (10)b1110356183
Hypoxic–ischaemic encephalopathy64272101416412 
Intracranial haemorrhage111  3 (1)122 1 
Intracranial infection9   2522 2 
Metabolic disease32  2 (2)1 1  1
Hypoglycaemia21  1  2   
Developmental defects94  2 (1)314111
Benign neonatal seizures111       
Abstinence syndrome6   5 (4)  1   
Non-epileptic eventsa20  18      
Unknown aetiology1574 6 (2) 17121
Table II.   Symptoms reported (R) and video-recorded (V)
 Patient 1Patient 2Patient 3Patient 4Patient 5
RVRVRVRVRV
  1. aRegistered in electroencephalogram.

Neonatal
 Myoclonic jerks××××× ××× 
 Twitching mouth, eyes×  ×× ×   
 Clonic   ×    × 
 Tonic, dystonic movement ×××      
 Startle/Spasms××  × × ××
 Hiccups, diaphragmatic      ×× ×
 Erratic movements×  ×    × 
 Tremor× × ×     
 Abnormal eye movements  ××  ××××
 Grimacing, frowning ×××    ××
 Frightened facial expression  ××      
 Oral automatisms×         
 Irritability, crying×××××   ××
 Vegetative symptoms        × 
Infants/children
 Tonic or tonic–clonic× × ××  × 
 Startle responses×   ×a     
 Myoclonic jerks, hiccups× ×  ×    
 Erratic movements         ×
 Abnormal eye movements×    ×  × 
 Grimacing, frowning     ×   ×
 Visual symptoms   ?    × 
 Frightened facial expression   ×     ×
 Headache, vomiting×         
 Loss of consciousness×         
 Irritability, crying   × ×   ×
 Sleep disturbance  ×       
 Dysaesthesia        × 

Results

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Clinical and laboratory data verifying the diagnosis of PDE or PPNO, initial leading symptoms, and response to pyridoxine or pyridoxal 5′-phosphate are listed in Table III. Case histories of the four females (patients 1, 2, 3, 4) and the male (patient 5) are presented in ‘Case histories’ section. In all the children PDE or PNPO was proven by the immediate disappearance of symptoms once pyridoxine or pyridoxal 5′-phosphate was administered. Genetic testing revealed mutations in the antiquitin gene in patients 1, 3, and 4 and in the PNPO gene in patient 5. In patient 2 no mutation in the antiquitin gene was detected and both pipecolic acid and alpha-aminoadipic semialdehyde were also normal, suggesting a cause of PDE other than antiquitin deficiency. Recently it was shown that mutations in the antiquitin gene are usually the cause of neonatal-onset PDE but not of all later-onset cases.7 However, PDE was confirmed in patient 2 by severe behavioural abnormalities, putative hallucinations, and sleeplessness after withdrawal of pyridoxine, which promptly ceased after the reintroduction of pyridoxine.

Table III.   Verifying the diagnosis of PDE or PPNO, initial leading symptoms, and response to pyridoxine or pyridoxal 5′-phosphate
Patient (diagnosis)Diagnostic criteriaStart of neurological symptomsInitial leading symptomsResponse to pyridoxine or pyridoxal 5′-phosphate at the first administration
  1. PDE, pyridoxine-dependent epilepsy; PNPO, pyridoxine phosphate oxidase deficiency; CTG, cardiotocography; PA, pipecolic acid; α-AASA, α-aminoadipic semialdehyde.

1 (PDE)Response to treatment, relapse after withdrawal, mutation in the antiquitin gene, elevated PA in urine1st dayPreterm 35th week gestational age, abnormal CTG, severe respiratory distress, necrotizing enterocolitis, apnoea, drug-refractory seizuresAge 5d: 50mg intravenously, prolonged sleep, muscle hypotonia, feeding problems
2 (PDE)Response to treatment, relapse after withdrawal, mutation not found, normal PA and α-AASA1st weekAbnormal CTG, irritability, crying, drug-refractory seizuresAge 28d: 100mg intravenously, periodic breathing for several hours, sleep for 2d, marked muscle hypotonia, tube feeding necessary
3 (PDE)Response to treatment, relapse after withdrawal, mutation in the antiquitin gene, elevated PA and α-AASA3rd weekAbdominal pain, irritability, myoclonic jerksAge 2mo: 60mg orally, no side effects reported
4 (PDE)Response to treatment, mutation in the antiquitin gene, elevated plasma PA2nd daySevere respiratory distress, apnoea, artificial ventilation, drug-refractory seizuresAge 3d: 50mg intravenously, drop in blood pressure (intravenous adrenaline necessary), prolonged sleep, muscle hypotonia
5 (PNPO)Response to treatment, relapse after withdrawal, mutation in the PNPO gene2nd daySlight respiratory distress, drug-refractory seizuresAge 14d: pyridoxine 100mg intravenously, no effect. Age 27d: pyridoxal 5′-phosphate 4×50mg orally, increased sleep, feeding difficulties, very low muscle tone for several days

Clinical seizures and EEG features before diagnosis

During pregnancy the mothers of patients 1 and 5 felt rhythmic fetal movements in the 36th and 38th weeks.

Symptoms began within 48 hours after birth in four patients (1, 2, 4, 5) and at age 3 weeks in patient 3. Multiple or mixed symptoms were reported in all neonates, consisting of irritability and shrill or inconsolable crying (patients 1, 2, 3, 5), excessive startles (patients 1, 3, 5), tremor (patients 1, 2, 3), erratic movements (patients 1, 5), erratic multifocal or generalized myoclonic jerks (all patients), twitching of eyelids and mouth (patients 1, 3, 4), hiccups (patient 4), clonic seizures (patient 5), opisthotonus (patient 2), abnormal eye movements (strabismus, setting sun phenomenon, staring, nystagmus; patients 2, 4, 5), oral automatisms (patient 1), grimacing (patients 2, 5), forehead frowning (patient 5), panicky facies (patient 2), apnoea (patient 4), and flushing and sweating (patient 5). Some episodes were reported to be long-lasting.

Video-EEGs were recorded in patients 2, 3, 4, and 5 before the first administration of pyridoxine or pyridoxal 5′-phosphate.

Patient 2

Forty-two hours of EEG and 50 minutes of video-EEG were analysed. At age 27 days, several episodes with frightened facial expression, grimacing, frowning, periocular twitching, repeated forced eye closure, irregular eye movements, nystagmus, intermittent grotesque convergent strabismus, myoclonic jerks (focal, multifocal, generalized), generalized tonic posturing, dystonic movements, and erratic head movements were recorded. The symptoms were variably combined, fluctuated in intensity, and lasted between a few seconds and several minutes. No ictal EEG change was visible. In addition, one clonic seizure associated with centro-median spike waves (Fig. 1) and postictal amplitude attenuation was recorded in the EEG. Interictal EEG demonstrated trace alternant in sleep and a few sharp waves.

image

Figure 1.  Patient 2, age 27 days. Ictal EEG during clonic seizure. First arrow, awakening; second arrow, start of clonic seizure.

Download figure to PowerPoint

Patient 3

At age 2 months, 4 hours of EEG and 7 minutes of video-EEG were archived. The seizures started with one or two generalized myoclonic jerks and were followed by variable symptoms such as grimacing (‘smiling’), crying, eye deviation, eye blinking, or generalized tonic posturing (Video S2; additional material). Ictal EEG revealed right occipital rhythmic theta-sharp-waves (Fig. 2), or bilateral frontal irregular spike wave discharges. Interictal EEGs were normal or showed rare multifocal sharp waves.

image

Figure 2.  Patient 3, age 2 months. Ictal EEG during intermixed seizures with right occipital rhythmic theta-sharp-waves.

Download figure to PowerPoint

Patient 4

At age 3 days, 36 hours of aEEG, 3 hours of EEG, and 62 minutes of video-EEG were archived. Multifocal myoclonic jerks were recorded intermittently at age 2 days and continuously at age 3 days (Video S3; additional material), not associated with ictal EEG changes. The eyes were mostly open and showed intermittent abnormal movements. aEEG starting from age 20 hours revealed electrographic status epilepticus for about 80 minutes (Fig. 3) which was stopped by administration of phenobarbital. Interictal EEG at age 2 days revealed intermittent periodic epileptic discharges alternating with bursts and periods of suppression (Fig. 4) and repeated 1Hz right fronto-central sharp wave series. At age 3 days the EEG was more continuous with multifocal sharp waves.

image

Figure 3.  Patient 4, age 20 hours, aEEG. Lower two tracings, left and right hemisphere, increase in the lower and upper margins (80min). Black arrow, administration of phenobarbital. Upper two tracings, ‘raw EEG’ at the time of the red bark in the lower tracings.

Download figure to PowerPoint

image

Figure 4.  Patient 4, age 2 days, interictal EEG, periodic epileptic discharges alternating with bursts.

Download figure to PowerPoint

Patient 5

Forty-three video-sequences made by the parents, 53 hours of EEG, and 32 minutes of video-EEG were archived. Videos from the parents showed abnormal eye movements, frowning, (Video S4; additional material) hiccups, irritability, crying, rhythmic elevation of both eyelids, grimacing of the mouth (‘grinning’), and startle reactions. The video-EEG age 22 days showed serial spasms at intervals of 1 to 2 minutes each, followed by crying (Video S5; additional material). They persisted for more than 2 hours. Ictal EEG revealed artefacts only. At age 25 days, left fronto-temporal sharp wave discharges lasting 30 to 60 seconds were recorded in EEG during sleep without clinical symptoms. Several normal and abnormal interictal EEGs were recorded. The abnormal EEGs showed discontinuous activity and intermittent prolonged periods of suppression (Fig. 5) and multifocal sharp waves.

image

Figure 5.  Patient 5, age 4 days, interictal EEG, intermittent periods of suppression.

Download figure to PowerPoint

Clinical and EEG response to pyridoxine or pyridoxal 5′-phosphate

In all patients, seizures and paroxysmal events promptly ceased after the administration of pyridoxine or pyridoxal 5′-phosphate. The EEGs showed no significant modifications in the patients 2, 3, and 5 and a stepwise normalization in patient 4. In patient 1 the first EEG (performed shortly after a single dose of pyridoxine and several hours after hemicolectomy) showed a burst suppression pattern which became more and more continuous at ages 8 and 16 days.

Clinical symptoms and EEG after withdrawal

Patient 1

Pyridoxine was not continued after age 5 days and seizures relapsed at age 26 days. The video (Video S1; additional material) showed frequent, erratic, multifocal, and bilateral myoclonic jerks intermixed with tonic spasms, groaning, crying, and grimacing. In addition, the technician reported very abnormal eye movements, difficulties with falling asleep, and seizures continuing throughout the recording (2.5h). The ictal EEG (microfilm) revealed no ictal discharge and only few sharp waves. Unmonitored withdrawal at age 6 months was followed by tonic–clonic seizures and at age 11 years 7 months by vomiting, headache, hiccups, tonic seizures, startle reactions, intermittent eye deviation, and loss of consciousness.

Patient 2

Ten days after withdrawal at age 14 months, 18 hours of EEG and 22 minutes of video-EEG were recorded before and after the reintroduction of pyridoxine. Beginning 8 days after withdrawal the child’s behaviour became increasingly abnormal. She slept distinctly less with frequent awakenings and, when awake, most of the time she cried inconsolably and wanted to be carried. Often she pointed, frightened, to something imaginary in the distance, which was assumed to be hallucinatory. EEG was normal during these episodes. After reintroduction of pyridoxine she fell asleep for about 12 hours. After awakening her behaviour was normal.

Patient 3

Withdrawal (age 8mo) was followed by generalized tonic seizures after 17 days. Interictal EEG before the reintroduction of pyridoxine was normal.

Patient 5

The pyridoxal 5′-phosphate was withdrawn at age 7 months. Nineteen hours of EEG and 33 minutes of video-EEG were archived. Seven hours after the last dose the child’s behaviour became increasingly abnormal with irritability, inconsolable crying, frightened facial expression, erratic movements, and agitated eye movements. Five minutes after restarting oral pyridoxal 5′-phosphate the patient exhibited his typical seizure symptoms with frowning. A few minutes later the previously normal EEG changed to irregular generalized and later centro-parietal theta–delta activity. During this 80-s ictal discharge the child merely interrupted drinking. After full reintroduction of pyridoxal 5′-phosphate the EEG remained normal.

Clinical and EEG features at the time of long-term follow-up

During febrile infections, generalized tonic–clonic seizures (patients 1, 2, 5) or startle reflexes (patient 3) were reported. During the startles the EEG was normal. Regular follow-up EEGs revealed isolated generalized spike waves in patient 1 and centro-median sharp waves in patient 4.

In patient 5 the dose of pyridoxal 5′-phosphate had to be limited because of abnormal liver function tests at higher dosages. At age 4 years the patient described pre-seizure symptoms consisting of tingling and numbness in his fingers, legs, and tongue, and seeing coloured spots. Once he lost his vision before having a tonic–clonic seizure. He also experienced intermittent eye deviations and head movements to the right which were exacerbated by watching television or by doing tasks requiring concentration, and sometimes he reported that he could not see the features of faces, or that objects seemed far away or disproportionate for a few minutes.

Case histories

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Patient 1

The female infant of unrelated parents was delivered by Caesarean section because of breech presentation and abnormal cardiotocography. Apgar scores were 2, 8, and 8 at 1, 5, and 10 minutes respectively; the amniotic fluid was meconium stained and cord blood pH was 7.2. Spontaneous breathing was insufficient after artificial respiration. Because of seizures and frequent startle reactions within the first few hours of life, phenobarbital and later phenytoin were administered, but without effect. At age 5 days necrotizing enterocolitis was diagnosed and hemicolectomy followed. A few hours after surgery a single dose of pyridoxine 50mg was administered intravenously, but the effect was not documented in the case history. Magnetic resonance imaging (MRI) of the brain at age 18 days revealed subacute haemorrhages from the choroidal plexus of both posterior ventricle horns and cystic lesions adjacent to both frontal horns. At this time, prolonged seizures reoccurred and did not respond to diazepam and clonazepam. After another intravenous dose of pyridoxine 50mg at age 4 weeks, seizures ceased again. Laboratory analysis (before pyridoxine) revealed elevated pipecolic acid in the urine. The association of pipecolic acid with PDE was not yet known in 1991 and Zellweger syndrome was suspected but was excluded by normal very-long-chain fatty acids. Because of the imaging abnormalities and the perinatal and postnatal complications, the diagnosis of PDE was discarded and pyridoxine was stopped again. Eight days later, seizures reoccurred and pyridoxine was reintroduced. From that time, the child received a daily low dose of pyridoxine (20–40mg) and phenobarbital up to the age of 14 years. Three relapses were reported: one after missing drug intake for 2 days, one during viral infection, and one after temporary cessation of pyridoxine and phenobarbital at age 11 years and 7 months. At age 14 years she was receiving pyridoxine 40mg (0.35mg/kg) and phenobarbital 15mg (0.3mg/kg). Re-evaluation at this age revealed normal pipecolic acid in urine and plasma. She was compound heterozygous for the common mutation E399Q (exon 14) and for a single base deletion (c.749delt; Val 250 frameshift) in exon 9. Pyridoxine was increased to 10 to 15mg/kg. At age 18 years she was seizure free with monotherapy pyridoxine 12mg/kg and in a stable healthy and social condition.

Patient 2

This female was delivered by forceps because of abnormal cardiotocography and failure to progress in labour. Apgar scores were 6, 7, and 8 at 1, 5, and 10 minutes respectively. After 24 hours she became irritable and at age 7 days recurrent and prolonged seizures were observed which did not respond to phenobarbital or diazepam. The child was transferred to our hospital at age 19 days. Several ictal and interictal EEGs were recorded with inconclusive results. Analysis of the seizure videos at age 28 days raised suspicion of PDE. Biochemical analysis in cerebrospinal fluid (CSF), plasma, and urine revealed mainly normal results (lactate, CSF/plasma glucose ratio, amino- and organic acids, neurotransmitter, α-aminoadipic semialdehyde); only pipecolic acid in CSF was minimally elevated (0.31μmol/l, controls: 0.01–0.12μmol/l). Pyridoxine was started and no more seizures and irritability were observed. Breakthrough seizures occurred only during febrile illnesses. PDE was confirmed by a controlled withdrawal trial at age 14 months, but no mutations in the antiquitin gene were found. At age 2 years and 6 months, Bayley Scales of Infant Development 2nd edition (Bayley II) gave normal results (Mental Developmental Index 103). At her last visit at age 4 years and 7 months her development was normal but neurological examination revealed marked muscle hypotonia and hyperlordosis.

Patient 3

This female of consanguineous parents was born at term. Apgar scores were 8, 9, and 9 at 1, 5, and 10 minutes respectively. The child was normal until age 3 weeks, when the parents reported frequent episodes of abdominal pain, and two or three times per week several minutes of generalized myoclonic jerks followed by crying and irritability. Between these episodes the child was neurologically normal. Further investigations were delayed until age 2 months when the duration and frequency of these events increased and intermittent loss of eye contact was reported by the parents. Phenobarbital was effective for only a few days and was replaced by valproic acid (20mg/kg) and pyridoxine 10mg/kg. In the next 5 months the child was seizure free. Only recurrent startle reactions were reported during a febrile illness and incomplete drug intake. At age 8 months pyridoxine was withdrawn over 2 weeks and valproic acid continued. Seventeen days after the last pyridoxine dose, seizures recurred. Laboratory analysis revealed elevated pipecolic acid in CSF (0.59μmol/l; controls, 0.01–0.12μmol/l) and plasma (4.91μmol/l; controls, 0.54–2.46 μmol/l) as well as elevated α-aminoadipic semialdehyde in plasma (4.1μmol/l; controls, <0.2μmol/l) and urine (13mmol/mol creatinine; controls, <1mmol/mol creatinine). γ-Aminobutyric acid (GABA) was normal in CSF. She was homozygous for the missense mutation A171V in exon 6 of antiquitin. The parents were both heterozygous. Brain MRI was normal. Pyridoxine 10mg/kg was reintroduced and valproic acid stopped. In the follow-up, the child remained seizure free. At age 5 years her motor development was normal. Her cognitive development (SON-R) was in the lower normal range with dysfunctions in expressive language.

Patient 4

Born at term, the 4Apgar scores for this female were 9, 9, and 9 at 1, 5, and 10 minutes respectively; cord blood pH was 7.34. The amniotic fluid was meconium stained. Because of severe respiratory distress and relapse into fetal circulation, artificial ventilation was started on the second day of life. At the same time, seizures were observed and status epilepticus was recorded in the two-channel aEEG. Seizures improved after the administration of phenobarbital and phenytoin. Only a few myoclonic jerks remained during routine EEG on the same day. MRI revealed features of hypoxic–ischaemic encephalopathy. In the following hours, multifocal myoclonic jerks again became more frequent and later continuous. Because multifocal myoclonic seizures are unusual in neonates with hypoxic–ischaemic encephalopathy, a metabolic disease was suspected and pyridoxine started. No more myoclonic seizures were observed. Plasma pipecolic acid was elevated (62μmol/l; age-matched controls, 0.55–10.8μmol/l) and normal in urine. Sequence analysis of the 18 exons and intron–exon boundaries of the antiquitin gene revealed a homozygous deletion of four bases (c.787+3delAAGT) within intron 9 near the donor splice site. Up to the age of 17 months (last follow-up) the child was seizure free, showed a low muscle tonus, and almost normal development but dysfunctions in expressive language.

Patient 5

The paternal grandfather of this male patient showed fine pigmentary change in the maculae, a characteristic that is consistent with people of Hebridean (group of Scottish islands) ancestry, owing to a different rate of absorption of vitamin B6. During pregnancy the mother felt rhythmic fetal movements in the 36th and 38th weeks. The patient was born at 39 weeks’ gestation. Apgar scores were 8, 9, and 10 at 1, 5, and 10 minutes respectively; the amniotic fluid was meconium stained and cord arterial blood pH was 7.34. The infant suffered acute respiratory distress syndrome, but artificial ventilation was not necessary. Seizures started on the second day and increased in frequency and duration. Phenobarbital was started and the infant was seizure free for 7 days, but hypotonic, sleepy, and difficult to feed. MRI of the brain and metabolic investigations of CSF and plasma (including pipecolic acid) were normal. PDE was suspected after seizure relapse and pyridoxine (100mg) was administered intravenously at age 14 days, without effect. Hence, phenobarbital was increased and again the infant was seizure free for 6 days. When seizures relapsed, clonazepam, phenytoin, vigabatrin, and an increased dose of pyridoxine (54mg/kg/day) were tried, without effect. Because the seizures were very similar to our infants with PDE, pyridoxine phosphate oxidase deficiency (PNPO) was suspected. We obtained pyridoxal 5′-phosphate, which is not licensed in Switzerland and which was not available in our pharmacy. The pyridoxine was replaced by pyridoxal 5′-phosphate 4×50mg/day orally and the seizures improved immediately. In the following days, the patient slept a great deal, had feeding difficulties, and marked muscle hypotonia. During the next 4 weeks the patient rarely showed abnormal eye movements, myoclonic jerks, or frowning. In the following months, phenobarbital and vigabatrin were withdrawn stepwise without relapse, but a trial of complete withdrawal of pyridoxal 5′-phosphate at age 7 months was followed by relapse within 7 hours. At age 9 months, testing by the Bayley Scales of Infant Development showed normal results (psychomotor developmental index 101, reference range 100 [SD 15]). Molecular analysis revealed heterozygosity for a missense mutation, D33V (c.98A>T), and a single base-pair deletion, c.246delT, in the PNPO gene.

CSF studies initially only showed increased levels of 3-methoxytyrosine and serine, and slightly increased taurine and histidine. All of these parameters normalized after pyridoxal 5′-phosphate replacement. Because of rare breakthrough seizures at the end of the 6-hour drug interval, pyridoxal 5′-phosphate was increased from 60 to 100mg/kg per day and divided into six doses, but had to be reduced at age 2 years 6 months because of abnormal liver function tests. The patient, now 4 years 7 months old, is neurologically and developmentally normal. However, a few seizures and episodes of visual symptoms and abnormal eye movements have occurred under the reduced dose (53mg/kg/day in seven doses) of pyridoxal 5′-phosphate.

Discussion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Seizures and paroxysmal events

Frequent multifocal and erratic or generalized myoclonic jerks have been the most prominent seizures in our patients and are the most frequently mentioned seizure type in neonates and infants with PDE and PNPO.1–3,8–20 In our neonatal video archive most myoclonic jerks occurred in isolation and during sleep. Frequent multifocal and erratic myoclonic jerks were exceptional and were recorded in neonates with abstinence syndrome (four neonates), intracranial haemorrhage (one neonate), developmental defects (one neonate), unknown aetiology (two neonates), and metabolic diseases (one with propionic acidemia, one with methylmalonic aciduria).

As with those patients reported by Nabbout et al.,15 our infants exhibited a great variety of intermixed seizures consisting of myoclonic jerks, tonic seizures, spasms, abnormal eye movements, grimacing, crying, and irritability. In our archive only three other neonates (one with developmental defect, one with methylmalonic aciduria, one unknown) showed comparable intermixed seizures. As in our patients, seizures in PDE have been reported to be prolonged or to proceed to status epilepticus.8,9,12,14–16,19–21 Irritability, agitation, excessive crying, and sleeplessness are common symptoms in infants with PDE.10,15 In our experience, these symptoms are often intermixed with other paroxysmal symptoms or they are associated with frightened facial expressions.

Focal or generalized clonic, tonic, or tonic–clonic seizures, which have been mentioned in various series,1–3,8,12,14,16,19,20 were seen in only one of our neonates (patient 2), but were reported when they were older (patients 1, 2, 3, 5). Generalized tonic–clonic seizures are very exceptional in neonates. It might be supposed that the intermixed tonic–myoclonic seizures in patients with PDE were classified as generalized tonic–clonic.

Abnormal eye movements and grimacing have been described in PDE12,13,17 and PNPO2,18 and were prominent symptoms in our patients. In our video archive we observed abnormal eye movements in 10 neonates (four with hypoxic–ischaemic encephalopathy, two with intracranial haemorrhage, two with intracranial infection, one with developmental defects, one with unknown aetiology). Only one preterm infant with intraventricular haemorrhage showed such grotesque eye deviations as were observed in patients 2 and 5, but these eye deviations were associated with ictal EEG changes.

Other symptoms reported are apnoea, cycling movements, atonic seizures, staring or atypical absence, hiccups, myoclonic seizures provoked by intermittent photic stimulation, startle reactions to touch and sound, and spasms.2,3,9–12,14–16,19,20 Startle reactions were reported in patient 1 before the introduction and after the withdrawal of pyridoxine, and in patient 3 during febrile infection. Serial spasms were recorded in patient 5. Although these spasms are similar to infantile spasms, inter-spasm intervals of 1 to 2 minutes, the absence of ictal pattern, and almost normal interictal EEGs are not readily compatible with such a diagnosis.

Visual symptoms as suspected in patient 2 and reported by patient 5 are described in the recent literature.16,19,20 They comprise intermittent loss of vision, flashing lights, and visual hallucinations, and they are sometimes followed by generalized tonic–clonic seizures. Probably they are more frequent in PDE or PNPO than so far reported.

EEG

The EEG during the seizures and/or paroxysmal events was surprisingly inconclusive. We observed continuous or intermittent focal or generalized discharges of sharp waves or rhythmic sharp theta waves during the seizures, but more frequently, we observed no discernable ictal changes. The lack of ictal EEG discharge during distinct paroxysmal events is perhaps more suggestive of PDE or PNPO than the inconstantly obvious ictal discharges. In neonates, myoclonic seizures and other symptoms may occur with and without associated EEG seizure activity.22 Not all paroxysmal events in PDE and PNPO are necessarily of epileptic origin. Vitamin B6 is a required enzyme in the biosynthesis of dopamine and serotonin. Abnormal eye movements, myoclonic jerks, flexor spasms, orofacial dystonia, irritability, startle responses, or behavioural abnormalities might be non-epileptic and could be due to impaired dopaminergic transmission. Some of the clinical features in patients with aromatic l-amino acid decarboxylase deficiency, a disorder of monoamine neurotransmitter metabolism, are similar to those seen in our patients.23,24 Even here, EEG often revealed no apparent ictal change.

In the light of our results, the inconsistent reports on ictal recordings in infants with PDE are not surprising. Nabbout et al.15 presented three ictal EEG records during intermixed spasms, myoclonic, and partial seizures which did not differ greatly. Other ictal EEGs in patients with PDE were described as bilaterally synchronous, lateralized or focal sharp waves, or as focal rhythmic sharp theta waves.11 Spike and poly-spike waves were observed during focal, multifocal, or generalized myoclonic, tonic–clonic, or tonic seizures8,9,12,20 and burst suppression pattern during status epilepticus, infantile spasms, generalized myoclonic jerks, or hiccups.9,12,20 Electrographic seizures without symptoms have also been found.8,13 Hellstrom-Westas et al.17 described a ‘saw-tooth’ pattern in the aEEG of patients with PDE. We performed aEEG in patient 4 and detected a continuous increase in the lower and upper margins over a period of 80 minutes, which is typical of neonatal status epilepticus.

In addition, interictal EEG was inconclusive with normal and abnormal registrations. Often we recorded normal and abnormal EEGs in the same patient before pyridoxine or pyridoxal 5′-phosphate. Normal EEGs have been reported repeatedly,9,11,12,15–17 but burst suppression pattern, bilaterally synchronous sharp and slow waves, intermittent or continuous slow waves, focal dysrhythmia, focal or multifocal spike and poly-spike waves, hypsarrhythmia, and photoparoxysmal response have also been reported.1–3,8,11–15,17–21 Burst suppression characterized as periods of inactive background (lower than 10–15μV) interrupted by bursts of activity which persist in both sleep states and even in wakefulness25 was seen in only one neonate, shortly after a single dose of pyridoxine and several hours after hemicolectomy. It is conceivable that preterm birth, acute respiratory distress syndrome, and antiepileptic or sedative drugs might have contributed to the burst suppression pattern in some of the reported patients with PDE or PNPO. The most frequent abnormal EEG patterns in our patients were multifocal sharp waves and discontinuous activity. We observed no impressive EEG changes after the first administration of pyridoxine or pyridoxal 5′-phosphate.

Conclusion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Long-lasting, frequent multifocal, and generalized myoclonic jerks intermixed with tonic seizures, spasms, abnormal eye movement, grimacing, or irritability are prominent symptoms in neonates with PDE and PNPO and should remind clinicians of these rare diseases. This even applies when the case history points to perinatal distress or other aetiology. Whereas the recommendation to administer pyridoxine in all neonates and infants with drug-refractory seizures is widely accepted, PNPO is often not considered. The typical symptoms should prompt clinicians to consider this rare disease and hence shorten the time between seizure manifestation and the administration of the life-saving pyridoxal 5′-phosphate. In older children, visual symptoms are probably more frequent in PDE and PNPO than have been reported so far and they are, possibly, a clinical marker for insufficient treatment. Ictal and interictal EEGs in infants with PDE and PNPO were variable, not pathognomonic, and probably not helpful in diagnosis.

What this paper adds

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information
  • • 
    Neonates with pyridoxine-dependent epilepsy and pyridoxine phosphate oxidase deficiency showed typical seizures and paroxysmal events otherwise rarely seen in neonates.
  • • 
    They often occur without ictal changes in EEGs.
  • • 
    Ictal and interictal EEGs in pyridoxine-dependent epilepsy and pyridoxine phosphate oxidase deficiency were variable, not pathognomonic, and often normal.
  • • 
    In older children, visual symptoms are probably more frequent than reported.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

We thank the parents who allowed us to publish the videos of their children. Research undertaken by PBM is supported by the Wellcome Trust.

References

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Case histories
  6. Discussion
  7. Conclusion
  8. What this paper adds
  9. Acknowledgements
  10. References
  11. Supporting Information
FilenameFormatSizeDescription
DMCN_3660_sm_VideoS1.mov6722KSupporting info item
DMCN_3660_sm_VideoS2.mov7520KSupporting info item
DMCN_3660_sm_VideoS3.mov4005KSupporting info item
DMCN_3660_sm_VideoS4.mov2610KSupporting info item
DMCN_3660_sm_VideoS5.mov4484KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.