StudyRareStudyRare
Log in to add personal notes on this page.

Failure of the embryonic forebrain (prosencephalon) to cleave into two hemispheres. The brain phenotype runs a continuous severity spectrum, and the facial midline phenotype tracks the brain phenotype (DeMyer's dictum: "the face predicts the brain"). At the severe end, a single fused ventricle and absent interhemispheric fissure; at the mild end, only the most ventral structures fail to separate, and the patient can be cognitively near-normal.

Two axes structure the work-up:

  1. Severity along the HPE spectrum. Alobar (no hemispheric separation; single midline ventricle; usually incompatible with prolonged survival) → semilobar (posterior separation only) → lobar (mostly separated, except in the most rostral / ventral structures) → middle interhemispheric variant (MIHV; failure of posterior frontal and parietal separation, paradoxically preserved anterior and occipital separation).
  2. Etiologic bucket. Chromosomal vs monogenic vs teratogenic vs syndromic. Roughly a quarter of HPE is chromosomal, a quarter monogenic, and the rest unexplained or multifactorial.

Facial midline severity parallels brain severity. Cyclopia and proboscis cluster at the alobar end. Single central maxillary incisor (SCMI), hypotelorism, midline cleft, and a flat nasal bridge cluster at the lobar / MIHV end. A child or adult with isolated SCMI sits on the HPE spectrum and warrants a microform-HPE work-up in first-degree relatives.

Chromosomal (about 25-50% of HPE)

  • Trisomy 13 (Patau syndrome): the single most common chromosomal cause of HPE. Microcephaly, cleft lip / palate, polydactyly, cardiac defects, scalp aplasia cutis. Get a CMA or rapid aneuploidy panel in any newborn with HPE.
  • 18p deletion syndrome: classical HPE-associated deletion; variable phenotype.
  • 13q deletions, triploidy, other rare aneuploidies.

Monogenic (sonic hedgehog pathway, AD with variable expressivity and incomplete penetrance)

  • SHH: the prototype HPE gene. AD inheritance; same family variant can produce severe HPE in one member and isolated SCMI in another. Counsel for variable expressivity, not "skipping."
  • ZIC2: more often associated with neural tube and midline phenotypes; classical full HPE spectrum.
  • SIX3: severe forms more common; some genotype-phenotype correlation.
  • TGIF1: another SHH-pathway gene; AD with reduced penetrance.
  • GLI2, FGF8, DISP1, CDON, PTCH1: additional HPE-spectrum genes.

Syndromic / multisystem

  • Smith-Lemli-Opitz syndrome (DHCR7, AR): defective cholesterol synthesis. HPE clusters in the severe end of the SLOS spectrum. Y-shaped 2-3 toe syndactyly, micrognathia, microcephaly, hypospadias, low cholesterol, elevated 7-dehydrocholesterol. Always check a sterol profile in an HPE infant with multisystem dysmorphism, because SLOS is recessive (recurrence risk 25%, vs much lower for de novo SHH-pathway variants).
  • Pallister-Hall syndrome (GLI3): hypothalamic hamartoma, polydactyly, occasional HPE.
  • Hartsfield syndrome (FGFR1): HPE + ectrodactyly + hypogonadotropic hypogonadism.

Teratogenic

  • Maternal pregestational diabetes (poorly controlled): the most common teratogenic association. Increases HPE risk roughly 200-fold over baseline. Caudal regression syndrome is the other classic diabetic embryopathy.
  • Retinoic acid (isotretinoin) and other vitamin A excess.
  • Alcohol (fetal alcohol spectrum disorders can include midline anomalies, though full HPE is less typical than the holoprosencephaly mimics).
  • Cholesterol-pathway disrupting drugs (statins; rare).
  • Single central maxillary incisor in an otherwise well child or adult → microform HPE. Examine the rest of the family; a parent may carry the SHH-pathway variant and have only the incisor finding.
  • Cyclopia or proboscis → almost always alobar HPE with a chromosomal cause (most often trisomy 13).
  • HPE + Y-shaped 2-3 toe syndactyly + low cholesterol → Smith-Lemli-Opitz.
  • HPE + hypothalamic hamartoma + central polydactyly → Pallister-Hall (GLI3).
  • HPE in an infant of a diabetic mother with poor first-trimester glycemic control → teratogenic; recurrence risk is low if diabetes is controlled next pregnancy.
  • Apparently isolated HPE in a child with normal CMA → SHH-pathway sequencing, including parental testing for asymptomatic carriers (variable expressivity is the rule, not the exception).
  1. Brain MRI with attention to ventricular separation, basal ganglia fusion, olfactory tract, pituitary, corpus callosum. Distinguishes alobar / semilobar / lobar / MIHV.
  2. CMA (or rapid aneuploidy panel if trisomy 13 suspected on exam). Highest yield first-tier genetic test.
  3. HPE gene panel: SHH, ZIC2, SIX3, TGIF1, GLI2, FGF8, DISP1, CDON, PTCH1. Exome if panel negative and clinical suspicion remains.
  4. Sterol profile (7-dehydrocholesterol) if SLOS features present or any HPE patient with multisystem dysmorphism.
  5. Maternal history: pregestational diabetes (HbA1c first trimester), teratogen exposure, prior pregnancy losses.
  6. Pituitary / endocrine evaluation: central diabetes insipidus, hypopituitarism, and electrolyte instability are common and underdiagnosed in HPE survivors.
  7. Cleft repair / feeding / seizure / hydrocephalus care as indicated; multidisciplinary follow-up.
  • The face predicts the brain. Facial midline severity and brain severity track each other along the HPE spectrum, with rare exceptions.
  • Single central maxillary incisor is HPE-spectrum until proven otherwise. It warrants pedigree review and SHH-pathway testing in the rest of the family.
  • Most monogenic HPE is autosomal dominant with variable expressivity and incomplete penetrance. A "skipping" pedigree is almost always undiagnosed mild carriers, not non-paternity or new mutation.
  • Pregestational diabetes is the teratogen to ask about. Excellent first-trimester glycemic control essentially eliminates the elevated HPE risk in subsequent pregnancies.
  • HPE + multisystem dysmorphism + low cholesterol = Smith-Lemli-Opitz. Recurrence risk is 25%, not negligible; sterol screening changes counseling.