Table 2.

Spinal cord AHT pathological findings: evidence from neuropathology (case series)

Study	Cases	Spinal cord level	Spinal cord injuries	Muscolo-skeletal spinal injuries	Author’s statement
Serinelli et al., 2017 [58]	51 homicide victims (42 AHT) (<3 yo)	All levels	Spinal cord injuries at toracholumbar (33%) > lumbosacral 27.5 > cervical 15.5%.	Not mentioned.	When considering the distribution of SC injuries (EDH, SDH, SAH), the thoracic SC was the most frequently involved area of the SC.
Brennan et al., 2009 [36]	41 AHT and 11 accidental head trauma (<2 yo)	Cervical	71% AHT had primary cervical spinal cord injuries: 72% parenchymal, 83% meningeal haemorrhages, 55% nerve roots (avulsion, haemorrhages).	21% among those AHT with spinal cord injuries had soft-tissue injuries.	Cervical spinal cord injury is a frequent but not universal finding in AHT. Parenchymal/dorsal nerve roots injuries can occur without ligamentous cervical injuries.
Geddes et al., 2001 (II) [59]	37 AHT (<9 mo, ma 2.4 mo) 14 non-traumatic (<11 mo, ma 3 mo)	Cervical	3/28 AHT showed βAPP positivity in cervical cord and/or dorsal nerve roots 8/28 AHT showed βAPP positivity in the lower pons and medulla 0/14 non-traumatic had βAPP positivity.	Not mentioned.	The predominant histological abnormality in AHT is diffuse hypoxic brain damage not axonal injury.
Geddes et al.,2001 (I) [32]	53 AHT (0.5–97 mo, ma 4 mo)	Cervical	3/53 had EDH 3/53 AHT showed βAPP positivity in cervical cord 8 AHT showed βAPP positivity in the lower pons and medulla.	Not mentioned.	AHT damage is age-related: infants (ma 2–3 mo) had thin bilateral intracranial SDH and higher incidence of skull fractures; AI is seen in craniocervical junction. Children >1 yo had larger intracranial SDH collection and higher incidence of extracranial damage: AI is seen in hemispheric with matter.
Saternus et al., 2000 [31]	4 AHT (shaking only) (3 autopsied and 1 survived) (4–30 mo, ma 13.7 mo)	Cervical	1/3 had cervical EDH (dorsal, C2/C3–C5/T2). Survived children had blood-stayed CSF.	No skeletal fractures were found at skeletal survey in 4/4 spine radiographs 2/3 had cervical soft-tissue injuries.	In every case of child autopsy, it is mandatory to look at cervical spinal cord.
Shannon et al., 1998 [60]	13 AHT (shaken only) (<2 yo, ma 5 mo) 7 hypoxia, 6 sudden asphyxia children	Cervical	7/11 AHT showed βAPP positivity in cervical cord and in spinal nerve roots, as opposed to none in the control groups.	Not mentioned.	Cerebral axonal injury is common in shaken babies and may be due in part by hypoxic/ischaemic mechanism. Cervical cord inj. is also common and cannot be attributed to HIE, so a traumatic mechanism may play a crucial role.
Feldman et al., 1997 [40]	5 AHT (1.3–34.1 mo, ma 5.8 mo)	Cervical	1/5 had SDH at the upper cervical cord, in association with cranial SDH 3/5 had SAH at the cervical cord level, in association with similar intracranial findings.	No skeletal fractures were found at skeletal survey in 12/12 spine radiographs.	Routine cervical MRI is not convenient to identify cervical spinal cord injuries as well as to recognize abused babies.
Hadley et al., 1989 [28]	13 AHT (shaken only) (ma 3 mo) (6 autopsied)	Cervical	5/6 had EDH at cervicomedullary junction 4/6 had SDH at cervicomedullary junction 4/6 had ventral spinal contusions at high cervical levels.	Not mentioned.	Haemorrhages and contusions of the high cervical cord may contribute to morbidity and mortality in shaken baby syndrome.

AHT abusive head trauma, βAPP β amyloid precursor protein, EDH epidural haematoma, yo years old, ma mean age, mo months old, MRI magnetic resonance, SAH Subarachnoid haematoma, SDH subdural haematoma