Magnetic resonance imaging of sellar and juxtasellar abnormalities in the paediatric population: an imaging review

TL;DR

This narrative imaging review summarises the MRI features of congenital/developmental and acquired sellar and juxtasellar abnormalities in children, framing MRI as the "gold standard" modality because of its high contrast resolution. It provides entity-by-entity imaging descriptions (signal characteristics, enhancement, location) alongside brief clinical, embryological and epidemiological context.

Key findings

• Normal paediatric pituitary imaging evolves: both lobes are T1-hyperintense at birth, with the anterior lobe dropping to near isointensity with the posterior pons by 6 weeks of age, while the posterior lobe retains T1 hyperintensity throughout adulthood. Normal pituitary height is typically 2–6 mm during childhood and up to 10 mm or slightly higher at puberty; preterm infants have taller glands than term infants.
• Developmental anomalies covered include adenohypophyseal agenesis and hypoplasia, ectopic posterior pituitary, pituitary duplication, hypothalamic hamartoma, persistent craniopharyngeal canal, Rathke's cleft cyst, cephaloceles, epidermoid/dermoid, arachnoid cyst, lipoma and empty sella. Reported frequencies include Rathke's cleft cysts on MRI in childhood at 1.2%, arachnoid cysts in 0.16% of all newborns with 9–15% in the sellar region, and trans-sphenoidal meningoencephaloceles at 1 per 700,000 live births.
• Neoplastic and inflammatory entities reviewed include pituitary adenoma (≈2.7% of all supratentorial tumours in children; only 3–6% non-functioning versus one-third in adults), craniopharyngioma (3% of all intracranial tumours), optic pathway glioma (4–6% of paediatric brain tumours, median diagnosis 5–9 years, ~half with NF1), pilomyxoid astrocytoma, germinoma and teratoma, tuberculous meningitis, osteosarcoma, Ewing sarcoma, esthesioneuroblastoma, fibrous dysplasia/McCune-Albright, Langerhans' cell histiocytosis (2 per 100,000 children under 15), lymphocytic hypophysitis, Rosai-Dorfman disease, leptomeningeal lymphoma and vascular malformations.
• Several discriminating features are highlighted: non-enhancing T2-hyperintense lesions suggest hypothalamic hamartoma or Rathke's cleft cyst; restricted diffusion distinguishes epidermoid from arachnoid cyst; intra-tumoural haemorrhage favours pilomyxoid over pilocytic astrocytoma (25% vs 8%); a parasellar T2 dark sign helps separate lymphocytic hypophysitis from adenoma; stalk thickness >3 mm with loss of posterior pituitary T1 bright spot suggests LCH.

Methods

• Design: narrative, image-based educational review; no systematic search strategy, inclusion criteria, or evidence grading is described.
• Data: no patient cohort or primary data; the article is illustrated with numbered figures (Figs. 1–31) of example cases and synthesises published literature, with 79 cited references.
• Analysis: qualitative synthesis organised by anatomical/embryological background, normal paediatric imaging characteristics, developmental anomalies, pituitary lesions, and juxtasellar lesions.

Limitations

• The paper states no limitations explicitly.
• Evident from the text: as a narrative review there is no stated search strategy, study selection, or critical appraisal of the cited literature; epidemiological figures are quoted from heterogeneous prior series without quality assessment.
• Also evident: illustration is via selected single-institution example images rather than systematic imaging–pathology correlation, and the review does not address diagnostic accuracy, sensitivity/specificity, or comparative performance of MRI against other modalities.