Study population
We enrolled 85 infants with definitive molecular diagnosis of HA. The median age of onset was 54 days (IQR 10–114 days). The study included 54 male (64%, 54/85) and 31 female (36%, 31/85) patients who were divided into two subgroups according to age of onset: neonate group (32 cases, 38%) and infant group (53 cases, 62%). The dominant phenotypes (a patient may have had more than one clinical phenotype) were neurological abnormality (31%), glucose metabolism disturbance (26%), metabolic acidosis (24%), respiratory failure (15%), and electrolyte disturbance (14%) (Table 1).
Genetic spectrum of infantile HA
Collectively, 136 pathogenic or likely pathogenic (P/LP) variants were identified in 33 genes, with 110 reported variants and 26 novel variants (Supplementary Table 1). Fourteen genes were reported with HA (42%, 14/33), with SLC25A13 and MUT being the top two detected genes (22%, 18/85). Nineteen genes, which have not been previously reported with HA, were detected in this study (58%, 19/33), of which JAG1 and ABCC8 were the most frequently mutated genes (16%, 14/85).
Recurrent genetic disorders included cholestasis (55%, 47/85), OA (14%, 12/85), UCD (7%, 6/85), and FAOD (6%, 5/85), with SLC25A13, MUT, CPS1, and SLC25A20 being the most frequently detected genes in each classification, respectively.
Genetic features of infantile HA depending on onset age
Subgroups with different age of onset showed different genetic disorders. The proportions of OA (p = 0.001), FAOD (p = 0.006), and cholestasis (p < 0.001) showed significant differences between the two subgroups (Table 1). MUT (19 %, 6/32) was the most frequently detected gene in the neonatal group, whereas in the post-neonatal group, the most frequently mutated gene was SLC25A13 (21%, 11/53).
In cases of refractory HA (21%, 18/85), all mutated genes in the neonatal group were reported with HA, with MUT being the most frequently identified gene (Figure 1). In the post-neonatal group, mutated ABCB11, CYP7B1, and MPV17, which have not been previously reported with the HA phenotype, were associated with refractory HA; interestingly, all these cases presented with liver failure.
In controllable or self-limiting HA cases (79%, 67/85), mutations in SLC25A13, JAG1, and ABCC8 were detected in more than five cases, accounting for 37% (25/67) of all cases. Mutated SLC25A13 and most mutated JAG1 (7/8) were detected in the infantile group, and all cases presented a cholestatic phenotype. In the neonatal group, ABCC8 was the most frequently mutated gene, and all cases presented with hypoglycaemia. Notably, mutated JAG1 and ABCC8 with the hypoglycaemic phenotype have not been previously reported with HA.
Clinical characteristics of inherited HA according to the age of onset
To investigate the characteristics of inherited HA according to the age of onset, we compared the clinical phenotypes of the neonatal and infantile groups. We found significant differences in the clinical features, clinical course, and outcomes between the two subgroups.
In the neonatal subgroup, 25% cases (8/32) presented with a peak plasma ammonia level ≥ 500 μmol/L, compared to 2% cases in the infantile group (1/53, p = 0.003). Neonatal patients with HA showed higher rates of neurological abnormalities (p < 0.001), respiratory failure (p < 0.001), metabolic acidosis (p = 0.001), glucose metabolism disturbance (p = 0.016), and electrolyte disturbances (p = 0.010) than infantile patients (Table 1). Infants with HA were more likely to present with hepatic failure than neonates (p = 0.042).
Among the neonatal subgroup, 41% patients (13/32) presented with a refractory clinical course, compared to just 9% patients in the infant group (5/53, p = 0.001). Clinical outcomes in the infant group were generally better than those in the neonatal group, with 83% patients showing improvement (44/53) and only 13% patients (7/53) withdrawing the treatment in consideration of the poor prognosis. In contrast, in the neonatal group, only 38% (12/32) patients had good prognoses (p < 0.001), and 47% (15/32) patients withdrew from the treatment (p = 0.001) given the poor prognosis.