Kyphoplasty

Two Level II studies examined the use of kyphoplasty compared to conservative treatment.^{54, 73} One study examined subacute fractures⁵⁴ while the other study examined chronic fractures.⁷³ In the study of patients with subacute fractures, clinically important benefits in pain were found at 1 week and 1 month, with possibly important effects at 3 and 6 months. There was no clinically important benefit in pain at 12 months. The study also found possibly clinically important benefits in physical function (at 1 and 3 months only) and the SF-36 physical component score (at 1, 3, and 6 months only). Clinically important improvement in quality of life was present at 1 month, and it was possibly clinically important at 3, 6, and 12 months.

In the chronic fractures study, all patients had fractures that were greater than one year old, raising the question as to whether the fracture was responsible for all of the pain. There was a statistically significant and possibly clinically important improvement in pain at 3, 6 and 12 months.

There were also three Level II studies which compared kyphoplasty to vertebroplasty.^{60, 61, 74} These studies were inconsistent in design and outcome. In the first study, patients were treated at a median of 8 weeks after a fracture.⁶⁰ No conservative treatment control group was included.  Kyphoplasty was favored over vertebroplasty when pain was measured out to two years. Repeat kyphoplasty in this study was a confounding factor. In the second study 21 patients were treated.⁷⁴ Both groups experienced similar pain relief at 6 months, although there was insufficient power to find a difference.  In the third and most recent study, 100 patients  received either kyphoplasty or vertebroplasty within 43 days of fracture.⁶¹ There was no difference in pain outcomes between the treatment groups at 3 days and 6 months.

When considering the technical similarities between kyphoplasty and vertebroplasty and the unique recommendations for their use within this guideline, several points deserve mention.

• The comparison of vertebroplasty to a sham procedure confirms the lack of benefit from vertebroplasty for critical outcomes.
• Both procedures were compared to similar control groups. In the case of kyphoplasty the comparison to conservative treatment resulted in possible clinically important differences for critical outcomes up to 12 months whereas vertebroplasty compared to conservative treatment showed only possible clinically important differences for critical outcomes at 1 day.
• The direct comparison between vertebroplasty and kyphoplasty is logically consistent with the previous two points in as much as it shows a possibly clinically important advantage in critical outcomes for kyphoplasty at durations up to 2 years.

These points alone merit a moderate strength recommendation for kyphoplasty due to the two Level II studies which compared kyphoplasty to conservative treatment. However, the comparisons between vertebroplasty and kyphoplasty are important. The results of the direct comparisons between kyphoplasty and vertebroplasty are not repeated across all studies which lowers our confidence that future studies will confirm the results of the current evidence. Thus, the recommendation is downgraded from moderate to limited and kyphoplasty is an option, for patients who present with an osteoporotic spinal compression fracture on imaging with correlating clinical signs and symptoms and who are neurologically intact.