Vertebroplasty: What Sham-Controlled Trials Reveal

Vertebroplasty can produce a dramatic before-and-after story: a person with a painful vertebral compression fracture receives cement and reports less pain soon afterward. That sequence does not prove that cement caused the improvement: fracture pain often improves with time, local anesthetic can help, attention and expectation change symptom reporting, and co-interventions occur around the procedure.

Sham-controlled trials were designed to isolate the added effect of cement. In several trials, pain and function improved in both groups without a clinically important advantage for vertebroplasty. VAPOUR reported an early benefit in a narrower group with severe pain and very recent fractures. To read the full evidence you have to ask which patients, what timing, which comparator, and which endpoint produced each result.

What the procedure changes#

Percutaneous vertebroplasty places a needle through the pedicle or another safe route into the fractured vertebral body under imaging guidance. Polymethylmethacrylate cement is injected and hardens. The proposed effect is mechanical stabilization of painful micro-motion.

Balloon kyphoplasty adds a balloon cavity before cement, aiming to restore some height and control cement placement; the two procedures are related but not identical, and evidence for one should not be transferred automatically to the other. The procedure treats the fractured vertebra. It does not reverse systemic osteoporosis, strengthen unfractured bones, correct every spinal deformity, or prove that the treated level caused all back pain.

Diagnosis comes before the procedure debate#

A vertebral compression deformity on an x-ray can be old and painless. Acute pain may instead come from muscle, facet joints, or disc. It may come from stenosis, infection, cancer, or another fracture. Clinical localization and imaging timing matter.

Magnetic resonance imaging can show marrow edema consistent with a recent or unhealed fracture and can evaluate malignancy, infection, canal compromise, and other pathology. Nuclear imaging can help when MRI is unsuitable. Imaging findings still need to match the pain location and onset.

Neurologic weakness, bowel or bladder change, or saddle sensory loss demands urgent evaluation. So does fever, cancer history, major trauma, or progressive systemic illness. A cement procedure should not delay treatment of cord compression, infection, or an unstable fracture.

Why usual-care comparisons can mislead#

In an open trial, participants know whether they received a procedure. Those assigned to usual care may be disappointed and seek crossover, while treated participants may expect rapid benefit. Clinicians may change analgesics, activity advice, and follow-up based on allocation.

A procedure also bundles several active experiences: local anesthetic, sedation, and positioning. It bundles skin incision, periosteal contact, monitoring, and intensive attention. Comparing that bundle with an office visit does not isolate cement. Open-label trials can answer whether offering a procedural pathway changes outcomes in practice, but they are weaker for the mechanistic question of whether cement adds benefit beyond a credible simulated procedure.

What a vertebroplasty sham includes#

A credible sham brings participants to the procedure room, uses similar positioning and sterile preparation, administers local anesthetic, and recreates procedural sounds, smells, or pressure without injecting cement into the vertebra. Staff who perform the intervention cannot be blinded, but participants and outcome assessors can be.

The sham must be credible enough that participants cannot reliably guess assignment. Trials can ask them afterward which group they believe they entered. If blinding fails, expectation can still bias patient-reported pain.

Sham procedures are not risk free. Needle placement, anesthetic, sedation, and withholding a desired procedure require ethical justification. Their scientific value is greatest when a procedure is invasive, widely used, and its specific benefit remains uncertain.

The two 2009 trials changed the argument#

The Australian trial led by Buchbinder enrolled 78 participants with one or two painful osteoporotic fractures of less than 12 months' duration, confirmed as unhealed on MRI, and participants received vertebroplasty or a sham. Pain and quality-of-life outcomes did not show a significant advantage for cement over the simulated procedure.

The INVEST trial also randomized participants to vertebroplasty or a simulated intervention, and at one month, improvement occurred in both groups without significant differences in the primary pain and disability outcomes. Crossover after the initial period and sample size affected longer interpretation. These trials did not prove that nobody can benefit. They showed that large uncontrolled improvements could occur without cement and that routine efficacy was less certain than observational series suggested.

VERTOS IV tested a more acute population#

VERTOS IV enrolled 180 adults with one to three osteoporotic compression fractures, MRI edema, and mostly pain of six weeks or less; the multicenter study compared vertebroplasty with a carefully simulated procedure and followed pain, quality of life, and function.

Both groups improved substantially. Vertebroplasty did not produce greater pain relief over 12 months. The day after the procedure, similar proportions in each group believed they had received cement, supporting the credibility of blinding; the median timing was still several weeks after onset, and recruitment excluded many screened patients. The finding applies to the selected population, not every hospitalized patient within days of a fracture.

VAPOUR found a different early result#

VAPOUR enrolled 120 participants with acute fractures and severe pain, many hospitalized, with symptom duration generally six weeks or less, and the primary endpoint was the proportion with pain below 4 on a 10-point scale at 14 days.

At 14 days, 44 percent of the vertebroplasty group and 21 percent of the placebo group met that endpoint, a 23-point difference. This result suggests that very early intervention in a severe, narrowly defined population may differ from broader or later use. VAPOUR also differed in sham technique, cement volume, inpatient severity, timing, and endpoint. The contrast with other trials should prompt effect-modifier questions rather than a vote count.

Timing may be an effect modifier#

If cement works by stabilizing painful movement at an unhealed fracture, benefit could be largest before natural consolidation; very acute, severe pain may define a window in which the specific mechanical effect is easier to detect.

Timing is entangled with severity and selection. People still in severe pain after two months may represent a different biological group from those hospitalized within two weeks, and very early studies can also capture a period of rapid natural recovery, making a sham essential. Subgroup claims need prespecification and enough participants. A positive result in one acute trial does not establish an exact day cutoff, and a negative average in mixed-duration trials does not exclude every early effect.

The Cochrane conclusion prioritizes placebo comparisons#

The Cochrane review included five placebo-controlled studies with 541 participants, plus trials against usual care and other procedures, and it gave the sham comparison greatest weight because it is least vulnerable to performance and detection bias for pain outcomes.

At one month, pooled pain was about 0.7 points better on a 10-point scale with vertebroplasty, while the prespecified minimal clinically important difference was 1.5 points. Disability, quality of life, treatment success, and new fractures did not show clinically important routine benefit. The review therefore concluded that high-quality evidence did not support routine vertebroplasty. That population conclusion leaves room for research and selected decisions without making the procedure first-line care.

Statistical difference is not enough#

Pain scales are noisy and responsive to context. A small mean difference can reach statistical significance in a large study while remaining too small for most patients to notice. Conversely, responder outcomes can conceal the full distribution.

What you should be shown is the mean change, the between-group difference, and the responder threshold. You should also be shown the confidence intervals, disability, and mobility. You should be shown medicine use, quality of life, and serious harms too. A threshold chosen after looking at results is less credible than one prespecified and clinically justified. Early pain relief matters, especially when severe pain prevents breathing deeply or mobilizing. Its value still needs to exceed procedural risk and be sustained long enough to change function.

Procedure risks are uncommon but real#

Cement can leak outside the vertebral body. Most leakage is asymptomatic, but cement can compress a nerve root or spinal cord, enter veins, and embolize to the lungs. Bleeding, infection, and fracture are possible. So are allergic reaction, anesthesia complications, and death.

Osteoporotic patients may be older, frail, anticoagulated, or have cardiopulmonary disease, increasing procedural burden. Holding an anticoagulant can also create thrombotic risk. These decisions require coordinated planning. New adjacent fractures occur after vertebral fracture with or without augmentation, and whether cement materially increases that risk remains uncertain because underlying osteoporosis and altered spinal mechanics already create high risk.

Conservative care is active care#

Initial management can include diagnosis-specific pain treatment, short-term bracing in selected people, and mobility support. It can include physical therapy and prevention of complications from bed rest. Prolonged immobility increases deconditioning, thrombosis, pneumonia, constipation, and loss of function.

Analgesic choice accounts for kidney function, gastrointestinal bleeding, and falls. It accounts for sedation, interactions, and prior opioid use. Calcitonin has been studied for short-term acute fracture pain, but its role and duration require clinician guidance. Rehabilitation progresses spinal extensor strength, balance, safe movement, and daily function. High-force flexion or lifting may need temporary modification, but fear-driven permanent avoidance can worsen disability.

Treat the osteoporosis behind the fracture#

A low-trauma vertebral fracture often establishes very high future fracture risk even if bone-density measurement is not in the osteoporosis range. Evaluation can include bone density and calcium and vitamin D status when indicated. It can include kidney and thyroid factors, medicines, falls, and secondary causes.

Antiresorptive and anabolic osteoporosis medicines reduce future fracture risk in appropriate populations. Choice depends on fracture pattern, kidney function, and prior therapy. It depends on cancer history, route, cost, and sequencing. Cement does not replace this treatment. The care plan should also cover nutrition, smoking, and alcohol. It should cover vision, home hazards, and the medicines that make you more likely to fall.

Guidelines can appear to conflict#

The 2019 ASBMR task force concluded that vertebral augmentation should not be a first treatment and emphasized the lack of demonstrable benefit over sham for most patients, and some spine and interventional guidelines allow augmentation in selected acute, severe fractures with persistent pain.

Differences can reflect evidence weighting, specialty perspective, publication date, threshold for a recommendation, and the population emphasized. A guideline citation should include its exact selection criteria rather than only its conclusion, and no one document should be handed to you as universal consensus while important uncertainty remains.

A disciplined selection discussion#

Confirm a recent symptomatic fracture and exclude cancer, infection, neurologic compression, and another pain source. Document severity, duration, MRI findings, functional loss, and response to a structured initial plan. Say plainly why your patient resembles or differs from VAPOUR and the negative sham trials.

Discuss expected absolute early pain benefit, uncertainty after the early window, and cement and anesthesia risks. Discuss alternatives and the osteoporosis plan. Define in advance what outcome would justify the procedure, and how you will measure it.

The sham-controlled evidence does not make pain less real. It shows why improvement after a procedure is not enough to establish that the procedure's specific component worked.

References#

1. Cochrane review of vertebroplasty for osteoporotic fractures
2. Buchbinder and colleagues' 2009 sham-controlled trial
3. INVEST randomized simulated-procedure trial
4. VAPOUR randomized placebo-controlled trial
5. VERTOS IV full text
6. ASBMR task-force report on vertebral augmentation

For your own health, talk with your clinician.*