Disc degeneration, discogenic pain, and what randomised trials of cell therapy have actually found.
## ⚠️ Symptoms that need urgent assessment, today
Most back pain is not dangerous. A small number of features are emergencies, and they are the reason this
notice comes first rather than last.
Seek immediate medical assessment if you have:
- Loss of bladder or bowel control, difficulty passing urine, or numbness around the genitals, buttocks or inner thighs — these can indicate cauda equina syndrome, in which delay of even hours can cause permanent loss of bladder, bowel and sexual function
- Progressive weakness in a leg or foot, or a foot that is dropping
- Back pain with fever, unexplained weight loss, or a history of cancer
- Back pain following significant trauma, or in someone taking long-term steroids
No injection, and nothing on this page, is an alternative to urgent assessment for any of the above.
What Degenerative Disc Disease Is, and Is Not
Low back pain is among the leading contributors to years lived with disability worldwide — one of the largest single causes of disability in the Global Burden of Disease analysis [1]. That scale is worth holding in mind, because it means a very large number of people are looking for something that works.
"Degenerative disc disease" is a misnomer. It is not a disease in the ordinary sense but a description of age-related change in the intervertebral discs. Those changes are close to universal with age, and most people who have them have no symptoms at all.
Sources: Roughley 2004 [2]; Kepler et al. 2013 [3]
The disc is the largest avascular structure in the body: nutrients reach it only by diffusion through the cartilaginous endplates. That limited supply is why discs repair poorly and why degeneration, once established, tends to persist [3].
These are different diagnoses, with different evidence
Back pain is not one condition, and treatments that help one presentation may be irrelevant to another.
⚠️ *The cell-therapy evidence below is for discogenic pain specifically* — patients with axial back pain attributed to a degenerated disc, who have failed conservative treatment. It is not evidence for non-specific back pain, for sciatica, for facet pain or for stenosis.
Why Placebo Matters More Here Than Almost Anywhere
This section governs how everything below should be read.
Back pain — and discogenic pain in particular — responds strongly to placebo. A sham injection, an attentive clinician, the expectation of improvement, and the natural tendency of pain to fluctuate all produce measurable benefit. Sham-controlled trials in this field routinely show substantial improvement in the sham arm.
Two consequences follow:
- An uncontrolled study showing improvement tells you very little. If a series of patients receives an injection and reports less pain a year later, that result contains the treatment effect, the placebo effect, regression to the mean, and natural history — with no way to separate them.
- Sham-controlled trials are the only ones that answer the question, and this field has some. They are reported first below.
What the Randomised, Sham-Controlled Evidence Shows
The meta-analysis of sham-controlled trials
Rahyussalim et al. 2026 pooled seven randomised controlled trials comparing intradiscal MSC injection with sham or placebo in degenerative disc disease [8]:
- Visual analogue scale pain: mean difference 6.67, 95% CI −9.31 to −4.02 — a significant reduction against sham
- Oswestry Disability Index also significantly improved
- Conclusion: intradiscal MSC injection "appears to be a safe and effective therapy, offering modest improvements in pain and function for patients with chronic low back pain due to DDD who do not respond to conservative treatments"
(Note: the published abstract states the point estimate as a positive 6.67 alongside a negative confidence interval; the interval indicates a reduction of approximately that magnitude.)
This is the strongest evidence on this page, and it is genuinely positive. The authors' own word for the size of the benefit is modest.
The single randomised trial with a sham control
Noriega et al. 2017 randomised 24 patients with chronic discogenic back pain unresponsive to conservative treatment [4]:
- Treatment: allogeneic bone marrow MSCs, 25 × 10⁶ cells per segment, injected intradiscally under local anaesthetic
- Control: a sham infiltration of the paravertebral muscles with the anaesthetic
- Follow-up: 12 months
Findings:
- MSC-treated patients showed "a quick and significant improvement in algofunctional indices versus the controls"
- The improvement was restricted to a responder group comprising about 40% of the cohort — a post-hoc subgroup, not a prespecified analysis
- Degeneration measured by Pfirrmann grading improved in treated patients and worsened in controls — a structural finding, and one of the few in this literature
- Feasibility and safety confirmed
Twenty-four patients, roughly twelve per arm. A real sham-controlled result, in a very small trial, with the benefit concentrated in a minority.
The efficacy meta-analysis — and its safety finding
Xie et al. 2021 meta-analysed randomised controlled trials of MSC therapy in degenerative disc disease [7]:
⚠️ And the finding from the same analysis that must travel with those results:
"Our further statistical analysis showed that MSC therapy may induce AE of TEAE related to study treatment (OR = 3.05, 95% CI = 1.11 ~ 8.40, P = 0.03)." [7]
Rates of back pain, arthralgia and muscle spasm did not differ significantly between groups — but treatment-related adverse events were approximately three times more likely, and the confidence interval excludes 1. This is the clearest safety signal in the disc literature and it comes from the same paper as the efficacy figures.
The Uncontrolled Studies, and How to Read Them
Two frequently cited studies have no control group. Given the section above, their results cannot be attributed to the treatment.
Pettine et al. 2015 — an open-label pilot study, 26 patients, autologous bone marrow concentrate injected intradiscally [6]:
- ODI fell from 56.5 at baseline to 25.0 at 12 months; VAS from 79.3 to 33.2 (p ≤ .0001 within group)
- Eight of twenty patients improved by one modified Pfirrmann grade at one year
- A cell-concentration relationship: patients receiving more than 2,000 colony-forming units per mL improved by 69.5% (ODI) and 70.6% (VAS), versus 33.7% and 29.1% in patients over 40 receiving fewer (p = .03 and p = .01)
Orozco et al. 2011 — a pilot study of 10 patients, autologous expanded bone marrow MSCs injected into the nucleus pulposus [5]:
- Rapid improvement in pain and disability, reaching "85% of maximum in 3 months"
- Disc height was not recovered; water content on MRI was significantly increased at 12 months
Both are described by their authors as feasibility and safety studies with early indications of efficacy — which is the correct reading. Neither can establish that the cells caused the improvement.
Intradiscal Injection Is a Procedure, With Procedural Risks
Putting a needle into a disc is not equivalent to an intravenous infusion or a joint injection, and it carries specific hazards that any patient should hear before consenting [9][10]:
- Discitis — infection of the disc space. Rare, but serious and difficult to treat, because the disc has almost no blood supply to deliver antibiotics
- Accelerated degeneration after annular puncture — a documented concern with needle entry into the disc, and the reason discography itself has become less used
- Post-procedural pain flare — common in the first days
- Disc herniation — reported in the Pettine series in one patient [6]
- Nerve or vascular injury — rare with image guidance, not zero
The route also determines what the evidence means. The randomised evidence above is for intradiscal injection. Intravenous administration has not been tested for discogenic pain in a comparable trial, and a result obtained by one route is not evidence for another.
What Else Works, Reported at Full Strength
Exercise therapy
This has the strongest and broadest evidence of anything on this page. A 2026 network meta-analysis of 24 randomised controlled trials across 28 intervention arms in non-specific chronic low back pain found that all exercise modalities tested — stabilisation, resistance, traditional, combined and motor-control — significantly improved pain compared with control conditions [11].
The best-performing combination was traditional exercise, 15–20 minutes per session, 6–7 times weekly, sustained for at least 16 weeks. Duration and consistency mattered more than modality.
Epidural steroid injection
Effective in certain patient populations, with rare but real neurological complications [12]. The FDA issued a safety warning in 2014 about corticosteroid injection into the epidural space. The complication risk is concentrated in particulate corticosteroids given by the transforaminal route; non-particulate formulations given by other routes carry lower risk, with most comparative studies finding similar efficacy [12].
Surgery
Adjacent segment pathology after lumbar fusion is real and has been systematically reviewed [13]. Figures for its incidence vary widely between studies and definitions, and this page does not quote a single percentage as though it were settled.
The Comparison Nobody Has Made
No trial has compared cell therapy with spinal fusion, with microdiscectomy, or with any surgical procedure. No trial has compared it with a structured exercise programme.
That means:
- It cannot be said that cell therapy is an alternative to surgery. It can be said that it is a different treatment, tested against sham, with modest benefit in a selected population
- Anyone who has been told they need surgery should have that conversation with the surgeon who assessed them, on the basis of their imaging and their symptoms
- The two most common surgical indications — nerve compression causing progressive deficit, and instability — are situations in which cell therapy has not been studied and is not appropriate
Realistic Expectations
Will cells regrow a disc? No. Current evidence does not support restoration of a degenerated disc to its former state. Orozco found disc height was not recovered [5]; the structural signals that do exist are modest changes in MRI water content and Pfirrmann grade [4][5][6]. The realistic goal is reduction in pain and improvement in function.
Who does the evidence apply to? Patients with discogenic axial back pain, confirmed by appropriate assessment, who have failed at least several months of conservative treatment, without instability, without significant nerve compression, and with preserved disc height.
Who should not be considering it? Anyone with the red-flag features at the top of this page. Anyone with cauda equina syndrome, progressive neurological deficit, significant instability, spinal stenosis requiring decompression, or active infection. Anyone expecting structural regeneration.
Questions Worth Asking
- "What exactly is my diagnosis?" — non-specific back pain, discogenic pain, radiculopathy, facet pain and stenosis are different, and the evidence differs for each
- "Was the evidence for this sham-controlled?" — in a condition this placebo-responsive, uncontrolled results mean little
- "By what route, and what is the evidence for that route?" — the randomised evidence is intradiscal
- "What are the risks of putting a needle in my disc?" — discitis and post-puncture degeneration should be named
- "Have I done a proper course of exercise therapy?" — 16 weeks, several times weekly, is what the evidence describes
- "Is this treatment licensed here, or is it unapproved?"
The Bottom Line
There is real randomised evidence here, and it should not be dismissed: seven sham-controlled trials, pooled, show significantly better pain and function with intradiscal MSC injection than with sham [8]. That is more than most regenerative indications can claim.
The same literature says the benefit is modest [8], that it was concentrated in a 40% responder subgroup in the one trial with individual results [4], and that treatment-related adverse events were about three times more common than in controls [7].
Discs are not regrown. Nothing has been compared with surgery. And exercise therapy — cheap, available, and supported by 24 randomised trials — improves pain in every form it has been tested in [11].
A patient with discogenic pain who has genuinely exhausted conservative treatment, understands the procedure carries risk, and expects modest rather than transformative benefit, is making a defensible decision. Those conditions are the point.