Denis's 1983 retrospective study of 412 thoracolumbar fractures introduces the three-column spine model, adding a middle column (posterior vertebral wall, PLL, posterior annulus) to Holdsworth's prior two-column framework. The paper asks whether middle column failure — rather than posterior ligamentous disruption alone — better explains fracture patterns, mechanical stability, and neurological risk. A new classification of major fractures into four types with defined subtypes is correlated with stability and treatment indications.
Holdsworth's two-column model held that posterior ligamentous disruption alone caused spinal instability — but scoliosis surgeons routinely stripped those same ligaments without catastrophe, exposing the flaw. Denis resolved this by defining the middle column as the true stabilizing structure, leveraging CT scanning (then newly available) to visualize posterior wall integrity in ways plain radiographs could not.
When you see a burst fracture on CT, the finding that changes management is middle column disruption. Loss of posterior vertebral wall height, retropulsed fragment, widened interpediculate distance. That pattern places the injury in Denis's second-degree instability category, where axial loading during ambulation (even in a brace) carries a 20.3% risk of new neurological deficit.
Do not rely on canal obstruction percentage alone to gauge neurological risk. A T12 burst with 25% obstruction can produce worse deficit than an L3 burst with 60% obstruction. The conus is exquisitely sensitive to impact, while the cauda equina tolerates acute displacement better but is vulnerable to sustained compression.
This framework directly shaped the Magerl AO classification (1994) and the Vaccaro TLICS system (2005), both of which retained Denis's column morphology as a foundation. The three-column model remains the everyday language spine surgeons use to communicate fracture severity across institutions.
Denis's 1983 retrospective study of 412 thoracolumbar fractures introduces the three-column spine model, adding a middle column (posterior vertebral wall, PLL, posterior annulus) to Holdsworth's prior two-column framework. The paper asks whether middle column failure — rather than posterior ligamentous disruption alone — better explains fracture patterns, mechanical stability, and neurological risk. A new classification of major fractures into four types with defined subtypes is correlated with stability and treatment indications.
Holdsworth's two-column model held that posterior ligamentous disruption alone caused spinal instability — but scoliosis surgeons routinely stripped those same ligaments without catastrophe, exposing the flaw. Denis resolved this by defining the middle column as the true stabilizing structure, leveraging CT scanning (then newly available) to visualize posterior wall integrity in ways plain radiographs could not.
When you see a burst fracture on CT, the finding that changes management is middle column disruption. Loss of posterior vertebral wall height, retropulsed fragment, widened interpediculate distance. That pattern places the injury in Denis's second-degree instability category, where axial loading during ambulation (even in a brace) carries a 20.3% risk of new neurological deficit.
Do not rely on canal obstruction percentage alone to gauge neurological risk. A T12 burst with 25% obstruction can produce worse deficit than an L3 burst with 60% obstruction. The conus is exquisitely sensitive to impact, while the cauda equina tolerates acute displacement better but is vulnerable to sustained compression.
This framework directly shaped the Magerl AO classification (1994) and the Vaccaro TLICS system (2005), both of which retained Denis's column morphology as a foundation. The three-column model remains the everyday language spine surgeons use to communicate fracture severity across institutions.