**Background:** Pituitary stalk interruption syndrome (PSIS) is a rare disorder characterized by MRI findings of absent/ectopic posterior pituitary, absent/interrupted pituitary stalk, and anterior pituitary hypoplasia, leading to anterior pituitary deficiencies (e.g., growth hormone deficiency). Pathogenic variants have been reported in genes involved in pituitary development, holoprosencephaly, and other syndromes. Previous exome sequencing of 52 PSIS patients identified pathogenic variants in 39, mostly with complex phenotypes. This study aimed to identify genetic causes in 16 new sporadic PSIS patients.
**Methods:** This retrospective single-center study included 16 patients (13 boys, 3 girls) aged 0.4 to 13.7 years diagnosed with PSIS by MRI (ectopic posterior pituitary in all but three; interrupted/not seen/thin/malformed stalk; anterior pituitary height low in 9, normal in 5, not measured in 2). Growth hormone deficiency was isolated in 10 cases (62.5%) or associated with TSH deficiency (6 cases), ACTH deficiency (1 case), HH (1 case), or multiple deficiencies (2 cases). Additional phenotypes occurred in 6 cases (37.5%), including ophthalmic disorders (4 cases) and Fanconi anemia (2 cases). Exome sequencing was performed using Agilent SureSelect Human All Exon V4 on Illumina HiSeq2000 (average coverage x50). Variants were filtered for novel/rare (MAF<0.01), analyzed with SIFT and Polyphen2, and classified per ACMG guidelines. Karyotyping was normal in all.
**Key Results:** In 13 of 16 patients, variants potentially contributing to the phenotype were identified. Only one patient (case 4) had a pathogenic variant: a heterozygous CSNK2A1 c.A593G:p.K198R missense variant (recurrent in Okur-Chung neurodevelopmental syndrome), presenting with delayed walking (2.4 years), hypotonia, and autism. One patient (case 3) had a likely pathogenic novel splice site variant in GLI2 (c.1368+1G>A), associated with Culler-Jones syndrome/holoprosencephaly 9, presenting with a second anterior pituitary-like structure on MRI. Patient 16 had two PTCH1 variants (p.Y1165C, p.R1291W) and a GLI2 variant (p.V681M), all VUS, with megacisterna magna. Patient 7 carried novel loss-of-function variants in PITX1 and PITX2 (nonsense in PITX2), but only had PSIS and early puberty. Patient 1 had Fanconi anemia with a heterozygous FANCD2 pathogenic variant (p.K261M) plus DNMT1 and RFWD3 VUS. Other VUS included HH genes (FGF17, HS6ST1, KISS1R, CHD7, IL17RD) and holoprosencephaly genes (GLI2, PTCH1). No candidate variants were found in cases 6, 8, and 15. Overall, diagnostic yield was 12.5% (2/16) for pathogenic/likely pathogenic variants.
**Clinical Implications:** This study confirms that genetic diagnosis in isolated PSIS is challenging, with a low yield (12.5%) compared to syndromic forms. Most variants were VUS, suggesting possible multigenic inheritance or non-coding variants. The identification of a CSNK2A1 pathogenic variant highlights that PSIS can be part of a broader neurodevelopmental syndrome. The GLI2 splice site variant supports the role of holoprosencephaly pathway genes in PSIS. Clinically, these findings emphasize the need for comprehensive genetic evaluation in PSIS patients, especially those with additional anomalies, and the importance of considering syndromic causes. The low yield in isolated cases suggests that future research should explore non-coding regions, oligogenic inheritance, and larger cohorts.