OUP user menu

Percutaneous Vertebroplasty: Fractured Opinions

Jerome Schofferman MD
DOI: http://dx.doi.org/10.1111/j.1526-4637.2010.00984.x 1585-1586 First published online: 1 November 2010

Vertebral augmentation (VAG) for vertebral compression fracture (VCF) has become an almost bitterly controversial subject. On one hand, there are very experienced clinicians who remain thoroughly convinced VAG is very effective, and there are clinical research scientists who have shown efficacy in multiple prospective and retrospective outcome series [1,2]. On the other hand, two controlled trials showed percutaneous vertebroplasty (PV) to be no more effective than a so-called sham intervention [3,4]. This dichotomy has prompted editorials, narrative reviews, and rebuttals from both sides of the issue [5–9].

This difference of opinion is not just academic. The clinician treating a patient with a very painful VCF that is not improving is faced with difficult choices. There are no data to show efficacy of medications in this setting. There is good evidence that PV is superior to “conservative care” and pain management [10–12]. So if PV really does not work, there are no evidence-based alternatives for the clinician. When evidence is conflicting or lacking, we are bound to rely on the best published evidence combined with our individual expertise and the patient's unique values and circumstances. Rather than revisit arguments about patient selection, timing of PV, other methodology issue, and other concerns that have been discussed, the papers in this issue of Pain Medicine present an opportunity to look at the data from yet another perspective [13,14] (Table 1).

View this table:
Table 1

Conclusions regarding percutaneous vertebroplasty (PV)

1. Most patients with acute vertebral compression fractures improve without specific intervention over 6 to 12 weeks [10].
2. Percutaneous vertebroplasty (PV) works better than pain management and conservative care [10–12].
3. The mechanism of improvement after PV has not been established.
4. There is no difference in outcome between patients treated with PV versus those treated with so-called sham PV [3,4].
5. The sham used in the randomized controlled trials [3,4] may not have been a proper sham.
6. A small number of patients with vertebral compression fractures have been helped with radiofrequency neurotomy or intra-articular facet joint injections [13,14].
7. The posterior elements appear to play a role in pain production in some patients with vertebral compression fractures [13,14].

Most importantly, the two recent RCTs that sparked the controversy actually did not show PV to be ineffective [3,4]. More correctly, they showed that PV did not work any better than their so-called sham treatment. Both treatment groups improved. Therefore, it appears from these studies that the injection of cement into the vertebral body is not the critical component of vertebroplasty. That is to say, the sham treatment might not have been a proper sham, but instead an active and effective treatment. This point has been alluded to by others [5,6].

In this issue of Pain Medicine, two papers add to our understanding of PV and suggest one possible explanation for the success of both PV and sham. Mirta et al. reported pain relief in two patients with thoracic VCFs after anesthetizing the corresponding facet joints with intra-articular local anesthetic [13]. Although they did not use controlled medial branch blocks (MBB), their cases raise the possibility that at least in some patients, the pain of VCF might arise from the facet joints rather than the vertebral body.

In a second study, Bogduk, MacVicar and Borowczyk used more rigorous testing of the same hypothesis [14]. They report six patients with VCF whose pain was temporarily relieved by controlled diagnostic MBBs at the fracture level. Moreover, in five of their patients, pain was relieved long term by radiofrequency neurotomy (RFN). Based on their results, the authors propose a biomechanical model that might explain why patients with VCF could have pain emanating from the posterior elements.

These articles remind us that to date, the exact structural source of pain in patients with VCF has not been demonstrated. It has been assumed, but never demonstrated, that the pain arises from the vertebral body itself. The fact that VCF pain is purportedly relieved by VP does not prove the pain originates from within the vertebral body, or that the injection of cement is necessary. The biomechanics model offered by Bogduk et al. provides an explanation for one source of pain, the posterior elements [14].

The articles also offer a plausible explanation as to why the sham used in the RCTs might not have been a proper sham. The trochars placed in the pedicles during the PV procedure would be close to where the nerves from the facet joints run. The procedure has the potential for damaging or even transecting of these nerves—a plausible explanation for the relief of pain.

These cases reported in this issue of Pain Medicine neither prove nor even suggest that pain arises from the posterior elements in all patients with VCF. However, they do raise a provocative explanation for pain relief in some patients, an idea that is stimulating and invites more study. If posterior element pain were shown to be common in patients with VCF, there would be an objective test available—MBB. In addition, medial branch RFN might then become an option for those patients, a procedure that is less invasive, safer, and less costly than PV.

This controversy should make us take pause and think when we read the literature, especially when reasonable scientific results appear so divergent from the experience of so many clinicians. Does an RCT always provide the best information to guide clinical care? Are there reasonable alternative explanations for the results? Might procedures work, but not for reasons one might think? Is there not a place for well-conducted, but uncontrolled outcome studies? The papers in this issue of PM provide some grist for these mills.


I must disclose that I do not perform VAG, but have cared for many patients who have done quite well afterward, as well as a few who have not. I have no financial interests with any of the companies involved.


View Abstract