Vertebroplasty for Spinal Fractures — the Cement Fix Two Sham Trials Killed in 2009

Summary

In 1987 the French neuroradiologist Hervé Galibert reported injecting acrylic bone cement into a single cervical vertebra eaten away by a hemangioma, and by the mid-1990s that salvage technique had been repurposed into a booming outpatient business — percutaneous vertebroplasty for the painful spinal fractures of ordinary osteoporosis — on the strength of nothing but uncontrolled case series in which patients reported feeling better; the gap between that universal operator conviction that the procedure plainly worked and what a blinded comparison actually showed is the entire case. When the procedure was finally tested against a credible fake in 2009, the cement turned out to do nothing the placebo did not do.

The clinical claim was seductive and mechanically intuitive: drive a needle through the back under imaging, inject polymethyl methacrylate (PMMA) into the collapsed vertebral body, stabilize the fracture, and abolish pain — often, operators said, on the table. By the mid-2000s the operation and its cousin kyphoplasty were a multibillion-dollar global market; U.S. Medicare alone was paying for vertebral augmentation in roughly a fifth to a quarter of compression-fracture patients, on the order of tens of thousands of procedures a year. The evidence underneath was almost entirely uncontrolled. Pain from an acute vertebral fracture improves substantially on its own over weeks, and a needle in the back is a powerful theatrical placebo — two facts the case series could not separate from any true effect of the cement.

On 6 August 2009 the New England Journal of Medicine published, in a single issue, two independent randomized double-blind sham-controlled trials. David Kallmes's multicenter INVEST trial (131 patients) and Rachelle Buchbinder's Australian trial (78 patients) both gave control patients the full ritual — the same positioning, the same local anesthetic, the same room, the smell of mixed cement — but no PMMA. Both found no meaningful difference. Pain and disability fell sharply in both arms, by roughly the same amount, at every follow-up point.

No agency banned vertebroplasty and no court enjoined it. It was discredited by its own pivotal trials and then slowly throttled by guidelines, payers, and a 2018 Cochrane review, surviving today only as a restricted, narrow-indication option rather than the routine fracture treatment it had been. It stands as the textbook modern lesson that a procedure can feel like it works to every operator and every patient and still be a placebo — and that the only way to know is the sham control almost nobody wanted to run.

Timeline

1984–1987

Galibert injects the first vertebra

Neuroradiologist Hervé Galibert and colleagues in Amiens, France, inject acrylic cement to stabilize a C2 vertebra destroyed by an aggressive hemangioma, reporting dramatic pain relief; published 1987. The salvage technique is born for tumors and angiomas, not osteoporosis.

Early 1990s

Indication creep to osteoporosis

Operators in France and then the United States extend cement injection to the far larger population of painful osteoporotic compression fractures, on the basis of open case series.

1994–1997

U.S. uptake accelerates

Percutaneous vertebroplasty diffuses through interventional radiology and spine practice; balloon kyphoplasty (a cavity-creating variant) is commercialized later in the decade, expanding the market further.

2001

Reimbursement and codes mature

Dedicated procedure coding and Medicare payment normalize vertebroplasty and kyphoplasty as standard outpatient care, fueling volume despite the absence of any blinded trial.

2005–2008

Peak utilization

Vertebral augmentation is performed in roughly a fifth to a quarter of U.S. Medicare compression-fracture patients; the combined global market for the procedures runs to billions of dollars annually.

Aug 6, 2009

The two sham trials land together

NEJM publishes Kallmes (INVEST, 131 patients) and Buchbinder (78 patients) on the same day. Both blinded, both sham-controlled, both null: cement is no better than a fake injection for pain or disability.

2009–2010

Specialty backlash

Interventional and spine societies dispute the trials — enrollment was slow, populations were less acute, sham wasn't truly inert — while referrals to vertebroplasty fall at academic centers almost immediately.

2010 / 2016

The defenders' trials (VERTOS II, VAPOUR)

Open-label or selectively-positive trials report benefit in very acute, severely painful fractures, keeping the controversy alive and a narrow indication open.

Apr 24, 2013

NICE narrows it

UK guidance (TA279) recommends vertebroplasty and kyphoplasty only for severe ongoing pain after a recent unhealed fracture despite optimal pain management — not routine use.

May 2018

VERTOS IV confirms the null

A Dutch double-blind sham-controlled trial (180 patients with acute fractures) again finds no significant difference between cement and sham at any point through six months.

Nov 2018

Cochrane closes the door on routine use

An updated Cochrane review concludes high-to-moderate-quality evidence shows no clinically important benefit of vertebroplasty for osteoporotic fractures and warns of potential harm; routine use collapses while restricted use persists.

2012–2017

The volume curve falls

U.S. Medicare vertebroplasty volume drops sharply (one analysis: about 48% from 2012 to 2017); the procedure contracts from default care to a contested niche.

A Tumor Salvage Repurposed Into a Mass Market

Vertebroplasty began as a narrow, defensible rescue: cement to shore up a vertebra hollowed out by cancer or a vascular tumor, where there was little to lose and visible structural collapse to fix. Its migration to osteoporosis was an act of analogy, not evidence. The reasoning was mechanical and persuasive — a broken bone, filled and hardened, ought to hurt less — and it matched what operators saw with their own eyes: patients who had been immobilized by pain sitting up and walking after the injection. That observation, repeated across hundreds of uncontrolled series, hardened into clinical certainty. What the series could not see was their own confounders. Acute osteoporotic fracture pain is largely self-limited, fading over four to twelve weeks as the bone heals; a dramatic interventional procedure carries one of medicine's strongest placebo signatures; and patients were selected, positioned, and assessed by the enthusiasts performing the operation. Every ingredient of a false positive was present, and the surrogate everyone trusted — "the patient feels better afterward" — measured the natural history of the fracture and the theatre of the needle at least as much as the cement.

The Day Two Sham Trials Arrived Together

The disconfirmation was unusually clean because it was doubled. Editors at NEJM published Kallmes and Buchbinder in the same 6 August 2009 issue, two teams on two continents who had independently done the one thing the field had avoided for fifteen years: blind the patient by giving controls a convincing fake. In both trials the sham arm received the full sensory ritual — local anesthetic to the periosteum, the same imaging and positioning, even cement mixed in the room so its odor was present — but no PMMA entered the bone. Kallmes's INVEST trial randomized 131 patients with one to three painful fractures; Buchbinder's randomized 78. The results converged: large improvements in pain and function in both arms, no statistically significant advantage for cement at one month or beyond. The procedure that "always worked" worked exactly as well when nothing was injected. The trials did not prove cement was inert chemically; they proved the part everyone had been selling — the pain relief — could not be distinguished from a placebo.

Discredited, Disputed, and Throttled Rather Than Banned

No regulator pulled vertebroplasty, because no regulator had to clear it; like most procedures it had never faced a premarket efficacy gate. The 2009 trials therefore triggered not a recall but a slow contest. Proponent societies attacked the trials' generalizability — too few patients, fractures too old, a sham that itself injected anesthetic — and ran their own studies (VERTOS II, VAPOUR) that reported benefit in narrowly selected, very acute, severely painful patients, preserving a foothold. But the disconfirming evidence kept accumulating: the 2018 VERTOS IV trial, 180 patients with acute fractures, again found no difference from sham at any timepoint through six months, and the November 2018 Cochrane review judged the high-to-moderate-quality evidence to show no important benefit for routine use, plus a non-trivial risk of cement leakage and adjacent fractures. Payers tightened, NICE and clinical references restricted the indication, and U.S. procedure volumes fell steeply. The outcome was not abolition but demotion: vertebroplasty survives as a restricted option for a small, severely affected subgroup, no longer the reflexive answer to a painful spine.

Contributing Factors

01

Surrogate validated by observation, never by blinding

The entire case for vertebroplasty rested on operators and patients reporting less pain after the injection — a surrogate that conflated the cement's effect with the fracture's natural healing and the procedure's placebo power. Without a blinded comparator, "the patient improved" measured everything except the variable in question. An effect that is large, real-feeling, and uncontrolled is the precise signature of a confound, not of efficacy.

02

A procedure with no premarket efficacy gate

Drugs must prove benefit against placebo before sale; a percutaneous procedure could diffuse into routine, reimbursed practice on case series alone. The regulatory vacuum meant the first credible test came fifteen years and hundreds of thousands of procedures after adoption. When the approval mechanism is reimbursement coding rather than controlled evidence, scale precedes proof.

03

The strongest available placebo went unblinded for years

A needle in the back under imaging is one of the most potent theatrical placebos in medicine, and the target symptom — pain — is the most placebo-responsive endpoint there is. The field declined for over a decade to run the sham control that alone could separate ritual from cement, partly because operators were certain it worked and partly because a convincing sham is expensive and ethically fraught. The harder the placebo is to rule out, the more indispensable the sham arm — and the more tempting it is to skip.

04

Indication creep from a rare rescue to a common complaint

A technique legitimate for tumor-destroyed vertebrae was extended by analogy to the vastly larger osteoporosis population, where the natural history is benign and self-limited. The structural rationale that fit a hollowed-out cancer vertebra did not transfer to a bone that would largely heal on its own. Borrowing credibility from a narrow, defensible use to justify a broad, untested one is how a niche salvage becomes a mass market.

05

No clean exit once an industry and identity formed

By 2009 vertebroplasty was a billion-dollar market, a professional identity, and a service line; the disconfirming trials met organized rebuttal, rival positive-leaning studies, and a retreat to ever-narrower indications rather than withdrawal. With no regulator to force the issue, a discredited procedure does not stop — it shrinks slowly to the smallest subgroup its defenders can still claim, dragging the abandonment out across a decade.

Aftermath

The material consequence is enormous: hundreds of thousands of needles driven into spines worldwide, billions of dollars spent, and a documented though uncommon toll of cement leakage, pulmonary embolism, and adjacent-level fractures — all for a pain benefit that two pairs of blinded trials could not distinguish from a sham. The durable ripple is methodological. Vertebroplasty became the canonical modern argument for sham-controlled surgical and procedural trials: if a procedure cannot beat a convincing fake, its observed benefit is placebo plus natural history, and the only honest way to find out is to build the fake and blind the patient. It is now cited alongside arthroscopic knee debridement as proof that intuitive, mechanically sensible, universally "effective" procedures can be placebos at scale. What remains is a restricted indication and a permanent caution: the procedure was not banned, it was simply caught — and "vertebroplasty" has become shorthand among evidence methodologists for the operation everyone knew worked until someone finally faked it.

Lessons

Blind the patient before you believe the procedure. If a treatment's only proven endpoint is subjective pain and you have never compared it to a convincing sham, you have not shown it works — you have shown that patients feel better after a dramatic intervention, which they do regardless. Build the fake and run it before you scale.
Distrust the surrogate that improves on its own. When the condition heals naturally and the outcome is the most placebo-responsive symptom in medicine, "the patient got better afterward" is the weakest possible evidence. Demand a control arm that captures natural history and placebo, or assume your effect is both.
Audit indication creep back to the original justification. A technique earned its credibility in a narrow, structurally clear use (tumor-destroyed bone); confirm that rationale still holds before applying it to a common, self-limited complaint. Borrowed legitimacy does not transfer across populations.
Close the gate that drugs face but procedures don't. When a procedure can reach routine reimbursed use without ever beating placebo, the system has outsourced proof to billing codes. Treat the absence of a premarket efficacy requirement as a reason for more scrutiny, not less.
Design the retirement before the market forms. A discredited procedure with an industry behind it will not stop — it will retreat to its narrowest defensible subgroup and litigate the evidence for a decade. Define in advance the disconfirming trial result that ends the practice, or the abandonment will be partial, slow, and contested.