Advances in classification-based treatment of Kümmell disease

Background Kümmell disease is a special type of delayed vertebral collapse after osteoporotic vertebral compression fracture, characterized by vertebral ischemic osteonecrosis, intravertebral cleft, fracture nonunion, and pseudoarthrosis-like instability. Its treatment should shift from simple pain relief to classification-guided stability reconstruction. Methods This narrative review was conducted with a structured literature search focusing on the pathological mechanism, imaging classification, vertebral augmentation, bone cement distribution, anchoring techniques, fixation strategies, and complication management of Kümmell disease. Main findings Classification-based treatment is important because patients differ in vertebral collapse, intravertebral cleft morphology, posterior wall integrity, dynamic instability, spinal canal compromise, and neurological status. Stable or early-stage cases may be treated with conservative management, percutaneous vertebroplasty (PVP), or percutaneous kyphoplasty (PKP). Unstable cases with obvious intravertebral clefts, cement aggregation, or high risk of cement loosening and migration may require bone cement distribution optimization, anchoring techniques, bone-filling mesh containers, or intravertebral support devices. Advanced or complex cases with severe posterior wall defect, severe spinal canal compromise, neurological deficit, rigid kyphotic deformity, or poor reducibility may require posterior fixation, decompression, or deformity correction. The treatment boundary for stage III patients with posterior wall disruption but without neurological symptoms remains controversial and requires individualized decision-making. Conclusions Kümmell disease treatment should shift toward classification-guided decision-making, vertebral stability reconstruction, and whole-process osteoporosis management. Emerging techniques may provide additional options for selected patients, but current evidence remains limited and further multicenter prospective studies are needed