This is a hospital based study of children diagnosed with OA between 2003 and 2009 before Newborn Screening (NBS) was widely available. The survivors were followed up over the next five years or more. The study was cleared by the Institutional Ethics Committee. Clinical suspicion of an IEM was based on suspect family history, age, presentation, presence of a combination of failure to thrive, developmental delay or regression, recurrent vomiting, abnormal body fluid odours, encephalopathy and organomegaly. The metabolic work up included urine ketones, blood glucose, anion gap, arterial blood gas analysis, blood ammonia and lactate. Diagnosis was confirmed by urine GCMS for organic acids supplemented by blood tandem Mass Spectrometry (MS/MS). Management, immediate outcome, compliance to treatment and recurrence of crises were documented.

The survivors were prospectively followed up over next five years or more for growth and development, recurrence of symptoms, neurologic sequelae if any and neuroimaging. Failure To Thrive (FTT) was defined as weight, height and weight for age less than 5th centile in WHO growth standards. Developmental delay was defined as development quotient less than 70 and when possible intelligence quotient was assessed by Binet Kamat scale by a clinical psychologist. All children underwent vision and hearing assessment and where indicated brain imaging by Magnetic Resonance (MR).

Statistical analyses included median or Mean±SE for numerical data and frequencies (n) with percentages (%) for categorical data.

A diagnosis of IEM was considered in 72 children on the basis of clinical presentation and metabolic profile. Spectrometry analyses of urine and blood confirmed the same in 38 (52.8%) of which 15 (39.5%) had an OA. Except one child with MMA who was diagnosed at 14 months, all were diagnosed by infancy. The Mean±SE and median ages at diagnosis were 5.6±1.16 and 4 months respectively. Male preponderance was found with a ratio of 2:1. The most common OAs were MMA in 6 (40%) and MCD in 5 (33.3%). Others included two (13.3%) children of PA and one each of GA1 and βKT deficiency. [Table/Fig-1a-c,2a,b] depict the characteristic urine organic acid excretion pattern seen in these children. Consanguinity was seen in one and sibling affected/death in five (33.3%).

[Table/Fig-1]:

Urine gas chromatograph mass spectrometry showing characteristic organic acid excretion pattern: a): Methyl malonic acid in methylmalonic academia; b): 3-methyl crotonyl glycine in multiple carboxylase deficiency; c): 3-hydroxy propionic acid, propionyl glycine and methyl citrate in propionic acidemia. Absence of 3-hydroxy propionic acid and methyl citrate in a) is indicative of presentation without ketoacidosis.

[Table/Fig-2]:

Urine gas chromatograph mass spectrometry showing characteristic organic acid excretion pattern: a) 3-OH butyrate, 2-methyl 3-OH butyrate and tiglylglycine in beta ketothiolase deficiency; b) glutaric acid and 3-OH glutaric acid in glutaric aciduria type 1.

Developmental delay was the most common presentation seen in 73.3%. Developmental delay, seizures, alopecia and skin rash were diagnostic of MCD. Three children, one each of developmental delay with microcephaly, failure to thrive with seizures with no ketoacidosis and developmental delay with macrocephaly were confirmed to be MMA, PA and GA1 respectively. More than half (8,53.3%) presented with a metabolic crisis and symptoms at presentation included encephalopathy, seizures and rapid breathing due to metabolic acidosis. One each of MCD and MMA were confirmed by presymptomatic screening in view of sibling history. Two other children presenting as persistent vomiting were evaluated for gastro-oesophageal reflux disease and malrotation in surgical department before a diagnosis of MMA was made. The clinical presentation and metabolic profile at presentation are shown in [Table/Fig-3].

Biochemically, wide anion gap metabolic acidosis, ketosis, lactic acidosis and hyperammonemia prompted the suspicion for diagnosis. The highest ammonia of 5706 μg/dl without ketoacidosis was seen in a child with PA [10]. The Mean ±SE ammonia level observed was 639.0±424.1 μg/dl (median 196 μg/dl) and lactate was 33.6±4.9 mg/dl. Mean pH was 7.27±0.07, arterial bicarbonate 14.1±2.3, base deficit 14.9±3.4 and anion gap 17.9±2.3. Neuroimaging (NI) was diagnostic in GA1 with bilateral widening of sylvian fissure and cerebrospinal fluid space anterior to temporal lobe (bat-wing appearance).

Empiric management in crisis included dehydration correction, megavitamins, carnitine, sodium benzoate when indicated, intravenous sodium bicarbonate based on base deficit in blood gases, broad spectrum antibiotics. Specific management after GCMS and MS/MS reports were available included protein restriction, cofactor supplementation like biotin for MCD and PA; vitamin B12 for MMA and riboflavin for GA1. The course was complicated by multiple crisis in 7 (46.7%) precipitated by infection. Poor compliance was noted only in one. The overall mortality observed was both of PA. Adverse outcome was seen in one of six MMA, three of five MCD and βKT deficiency. [Table/Fig-4] gives the adverse neurodevelopmental outcome in survivors of various categories of OA. Spasticity, extrapyramidal movement disorder, intellectual subnormality, autism spectrum, Attention Deficit Hyperactivity Disorder (ADHD) and sensory neural deafness were seen amongst survivors, in spite of compliance to therapy.

In our study, OA contributed to 20.8% of all children investigated and 39.5% of confirmed IEM. Among high risk children screened for IEM, Nagaraja D et al., and Narayanan MP et al., from Southern India have reported OA in 41.6 and 10.7% respectively [6,11]. Prevalence data of IEM in general and OA in particular from India is limited due to difficulties in establishing the diagnosis in an acutely sick child. Najafi R et al., have documented a high consanguinity rate of 84.7% in Mid-West [12]. In our study, though we did not have significant positive consanguinity, familial pattern in two suggested recessive inheritance pattern. High occurrence of MMA over other OA has been reported by other authors. Narayanan MP et al., and Najafi R et al., in their study report that MMA contributes to 34% and 33% of OA respectively [11,12]. Some of the unusual presentations in our children and reported as well, included PA masquerading as urea cycle disorder and βKT deficiency as diabetic ketoacidosis [10,13-17]. NI is highly useful in the diagnosis of GA1, as these children can have a subacute course and may present only with developmental delay and macrocephaly [18,19].

Neuropsychiatric manifestations occur in IEM either at presentation or as sequelae. It is thought to be due to disruption of normal late neurodevelopment or monoaminergic neurotransmitter systems [20]. A disturbance in mitochondrial function which is the final common pathway of all intermediary metabolisms may also be contributory. Autism spectrum disorder has been described in OA specifically PA, methylcrotonyl-CoA and biotinidase deficiencies [21-24]. The relation between ADHD and phenylketonuria is well documented [25]. ADHD is also associated with urea cycle disorders, maple syrup urine disease and homocystinuria. The autism and ADHD seen in our children with MCD may be part of moderate intellectual disability. NI was normal in this group.

Sensorineural hearing loss and optic atrophy are reported in untreated children with biotinidase deficiency [26-28]. It is proposed that accumulation of biocytin or small biotiny peptides may mediate neuronal injury in these children. Weber P et al., reported that vision and hearing abnormalities were none in infants diagnosed by NBS compared to children where treatment was initiated after onset of symptoms [26]. Wolf B et al., reports that more than three-fourths of symptomatic children have sensorineural hearing loss which does not seem to resolve or improve with biotin treatment [29]. Of the seven children with symptomatic biotinidase deficiency, Venkataraman V et al., report neurological deficit in one and deafness in none [30]. The variable prognosis in our series could be best explained by age and developmental period at diagnosis, severity of enzyme defect as well as phenotypic variations, recurrent metabolic decompensations, response and compliance to cofactor therapy.

Until NBS is a reality, OA should be part of the differential diagnosis in any sick neonate or infant with a negative sepsis screen, persistent metabolic acidosis and ketosis as well as in any infant with unexplained neurological deterioration even in the absence of ketoacidosis. Prognosis is guarded even with prompt institution of emergency treatment and long term cofactor supplementation.