Chapter 9: Fracture Instability and Height Restoration
John D. Barr, M.D.
Clinically successful percutaneous vertebroplasty produces durable analgesia, yet the mechanism of pain relief remains undetermined. Pain is not, however, the only significant complication of vertebral compression fracture. Decreased pulmonary function, early satiety, and altered center of gravity with increased propensity for falling are direct consequences of the kyphosis that results from vertebral compression fractures. Height loss and the resulting kyphosis associated with many vertebral fractures are not usually addressed by vertebroplasty. Persistent pain after technically successful vertebroplasty may be secondary to kyphosis and related musculoskeletal derangement. Secondary facet arthropathy is common at levels affected by significant kyphosis, and this may be a significant source of pain. Certainly pain relief afforded by vertebroplasty represents a significant therapeutic advance. In addition, restoration of vertebral height and reduction or elimination of kyphosis are also desirable goals that may be achieved in some cases.
Fluoroscopically Visible Fracture Instability
An important clue to both the mechanism of pain relief and the unrealized potential for vertebral height restoration during vertebroplasty was the discovery of visibly unstable fractures during treatment of two patients (1). One patient was a 71 year-old woman with a three week old T9 fracture; the other was a 53 year old man with an eight month old T11 fracture. After needle placement, intraosseous venography ("a vertebrogram") was performed to detect direct communications with the central or epidural veins. In each case, there was prolonged retention of contrast material within a portion of the compressed vertebrae. Normal saline was injected to wash out the contrast material so that retained contrast would not impair visualization of the cement during its injection. Fluoroscopy was used to ensure that the retained contrast had been expelled. As saline boluses were injected, cyclic changes in the height of the compressed vertebrae were visualized, and this phenomenon recorded on videotape. The contrast material appeared to have been trapped within unfused linear fracture planes.
Cement injection in each patient filled the fracture planes with little or no cement filling the remainder of the vertebrae. After cement injection, slight distraction of the end plates resulting in minor height restoration was evident in one patient. Both patients had virtually complete pain relief after treatment.
Proposed mechanisms of pain relief from vertebroplasty have included neurolysis due to the exothermic reaction of the polymethylmethacrylate cement, and mechanical strengthening of the weakened vertebrae. Fluoroscopically visualized fracture instability suggested that this had been the underlying source of pain in these two patients. Cement injected into the fracture planes was felt to have secured the fracture fragments in both cases. Pain relief following fracture stabilization also suggested that fracture instability had been the cause of the patients' pain. Fluoroscopically apparent fracture instability in patients undergoing vertebroplasty had not previously been reported. Lesser degrees of fracture instability may be common and would probably not be visible with fluoroscopy. Observations in these two patients suggest that chronic pain occurring with some vertebral compression fractures may be due to non-healing, unstable fractures, and that pain relief provided by vertebroplasty is due to internal fixation of such fractures.
In addition to providing a clue to the mechanism of analgesia, these observations also demonstrated that height loss and resulting kyphosis may not be irreversible effects of compression fractures. Acute fractures would be expected to be easily reducible (distractible), as they should not have had time to undergo bony healing. The discovery of a reducible 8-month old fracture in the second patient was surprising! However, if the mechanism of pain generation is indeed related to fracture instability, reducibility of a chronically painful fracture that is several months old might not be unexpected.
Back to Top
Hyperextension During Vertebroplasty
The second patient described above, a 53 year-old male, experienced abrupt onset of severe mid-back pain 10 days after his initial treatment. Plain films revealed a new fracture of the superior aspect of the T12 vertebra with marked increase in his already severe kyphosis. Because of his severe pain and desire for re-intervention, the patient was treated for the new fracture 4 days after its occurrence. Because the fracture was known to be acute and potentially not irreversibly compressed, and because the concept of fracture instability was fresh in my mind, I elected to position the patient in hyperextension during treatment. Treatment was performed under general anesthesia due to the patient's severe pain. Significant fracture reduction with height restoration and reduction of kyphosis was achieved by hyperextension. The patient again reported marked reduction in pain following the second procedure. At the same session, prophylactic vertebroplasties were performed upon the T9, T10, L1 and L2 vertebrae because I felt these levels were at very high risk for fracture for several reasons: the patient's severe osteoporosis, presumably elevated mechanical stresses at the apex of the severe kyphosis, and his rapid presentation with a new fracture adjacent to the first treated vertebra. Prophylactic vertebroplasty is admittedly a controversial issue; it remains of unproven value, and this case is not intended as a blanket endorsement of this treatment indication.
This case validated the concept of height restoration for an acute fracture using the relatively simple method of hyperextension during treatment. Admittedly, this patient was treated at a very early stage before any significant bony healing could have occurred. Examination of multiple subsequent patients under fluoroscopy has been performed to assess for vertebral fracture reducibility, and instability and reducibility have been demonstrated in some additional individuals. One limitation of this technique is that attempted examination of patients during hyperextension is often intolerably painful. The application of hyperextension during treatment may likely require deep conscious sedation or general anesthesia. Investigation of the routine use of hyperextension during percutaneous vertebroplasty might prove beneficial, with the expectation that significant vertebral height restoration and reduction of kyphosis could be achieved in some patients.
Back to Top
Figure 1a: Steroid induced osteoporosis and T11 compression fracture with kyphosis. Dramatic pain reduction following vertebroplasty.
Figure 1b: Acute onset of severe lower thoracic back pain 10 days later. Lateral radiograph reveals new T12 fracture and worsened kyphosis.
Figure 1c: Vertebroplasty was performed at T12 with hyperextension; significant height restoration and reduction in kyphosis. Prophylactic treatment of adjacent T9, T10, L1, and L2.