Cervical facet pain and regenerative approaches to the neck

Regenerative injection for cervical facet pain is an open question with real data on both sides of it — not a settled treatment, and not a dead end. As of 2026 there is one randomized controlled trial in this specific joint, in which platelet-rich plasma and a corticosteroid produced similar improvements in pain over six months [3], alongside several prospective and registry series reporting meaningful improvement in pain and neck disability at three, twelve, and twenty-four months [4][6][7]. A 2025 systematic review grades the overall evidence for platelet-rich plasma in facet joints as moderate at best and limited overall, with low certainty [8].

That is the honest summary. What the summary hides is who those studies enrolled. Every one of them selected patients before treating them, and the criteria they used screened out a great deal of what actually walks into a chronic pain practice — nerve root involvement, prior neck surgery, recent ablation, unstable psychiatric conditions, and, in most of this literature, anyone whose neck pain did not begin with a whiplash injury.

What can be done well right now, before any of that becomes relevant, is the diagnostic work that decides whether your facet joints are the source of your pain at all. An injection aimed at the wrong structure fails regardless of what is in the syringe.

What cervical facet pain is, and how often the facet joints are the source

Cervical facet pain is neck pain arising from the zygapophysial joints — the small paired joints at the back of each spinal segment that guide and limit motion. It is typically movement-related and associated with localized tenderness, and it overlaps clinically with disc-related pain, muscular pain, and pain referred from adjacent structures.

The facet joints are a common source, but not the only one. In 68 consecutive patients referred for chronic neck pain of more than three months’ duration after a whiplash injury, placebo-controlled diagnostic blocks put the prevalence of cervical zygapophysial joint pain at 60% (95% confidence interval 46% to 73%) [1]. That figure describes a specific population — post-whiplash chronic neck pain referred to a specialist clinic — and should not be read as the rate across all neck pain.

The same study is a caution in the other direction. It used a placebo injection of normal saline alongside two different local anesthetics under double-blinded conditions, precisely because its authors were concerned that simpler comparative blocks might exaggerate the importance of the condition [1]. Careful diagnosis narrows the field both ways.

Why the imaging report is not the answer

Degenerative change at the cervical facet joints is common in people who have no neck pain. In a computed tomography study of 500 cervical facet joints in 50 asymptomatic adults aged 20 to 69 — scanned for non-spinal reasons such as thyroid disease or cancer screening, with any record or history of neck or back pain excluded — 33% had facet arthritis of grade 1 to 3 (grade 1, 19%; grade 2, 11%; grade 3, 3%) [2]. Among subjects older than 40, arthritic change at C6–C7 was present in 78% [2].

So a report describing facet arthropathy describes anatomy. It does not identify your pain generator, and treating the picture rather than the patient is how injections end up aimed at the wrong level. The authors of that study note its own limitation: the design was retrospective, and subjects were not directly questioned about pain that might have been mild or undocumented [2].

How the source of pain is narrowed down

Narrowing it down uses history, examination of movement-provoked pain, review of your existing imaging against your actual symptom pattern, and, where appropriate, targeted diagnostic blocks of the medial branch nerves that supply the joint. Both the randomized trial and the largest prospective series in this area required a positive diagnostic block before anyone was injected [3][4], which tells you how central that step is considered to be.

That sequence is the same whether the eventual plan is regenerative, interventional, or conservative. It is also the step most often skipped.

What the published evidence on regenerative injection for the neck actually shows

There are five relevant publications, and they do not all point the same way. Taken together they describe a treatment that appears safe in small samples, produces improvement that is real in the people measured, and has not been shown to beat what is already being done.

The one randomized trial

A single-site randomized, double-blind trial in London, Ontario compared intra-articular leukocyte-poor, low-concentrate platelet-rich plasma against a corticosteroid injection into cervical facet joints under fluoroscopy, with outcomes collected at one, three, and six months [3]. Fifty-five people were enrolled and 40 were analyzed, of whom 21 received platelet-rich plasma [3][5]. Both treatments produced similar effects on pain intensity across six months, with comparable pain reduction at one month and comparable disability improvement at three and six months; the platelet-rich plasma group reported significantly less procedural pain and greater improvement in pain self-efficacy at one month [3].

Similar is not the same as ineffective. The comparator was an active drug already used for this problem, not a placebo, so the trial answers “is this better than a steroid injection” and does not answer “does this do anything at all.” A trial of this size also cannot exclude a difference smaller than it was built to detect.

Its entry criteria matter at least as much as its result. As registered, participants had to be at least 18 years old, have facetogenic neck pain of at least 4 out of 10 following a whiplash injury, and obtain at least 50% relief of familiar neck pain from dual cervical medial branch blocks [3]. Excluded were suspected serious spinal pathology, fracture or dislocation at the time of injury, nerve root compromise, spinal surgery or radiofrequency neurotomy within the previous 12 months, a corticosteroid cervical facet injection within the previous six months, any history of uncontrolled mental health conditions, and other contraindications to spinal injection [3].

Read that list as a description of who the result applies to: a whiplash-onset, block-positive, surgically untouched, single-site population, followed for six months, receiving one preparation of platelet-rich plasma at one concentration. Degenerative facet pain arising without a traumatic onset was not studied here.

The prospective series, and the patients in them

Three series report on regenerative injection in the cervical spine, and each enrolled a different kind of patient.

In a Calgary clinic series, 44 people with chronic whiplash-associated disorder (82% female, mean age 45.2 years, median symptom duration 24 months) received a single leukocyte-rich platelet-rich plasma injection into cervical facet joints under fluoroscopic or ultrasound guidance [4]. At three months, average pain fell from 5.8 to 3.7 on a 0–10 scale and Neck Disability Index scores fell from 45.2% to 30.7%, both p<0.001; 70% exceeded the minimal clinically important difference for pain and 80% for disability, while 41% reported more than 50% pain relief [4]. No adverse events were reported [4]. Entry required whiplash-associated disorder grade II of more than three months’ duration, failure of at least six weeks of conservative therapy such as physiotherapy, chiropractic or massage, and a positive diagnostic medial branch block [4].

An earlier Canadian case review treated 18 cervical zygapophysial joint sides in patients whose chronic whiplash-related neck pain had already failed other conservative and interventional procedures, placing 0.5 to 1 mL of 20% dextrose intra-articularly under fluoroscopic guidance with contrast confirmation [6]. Fifteen patients completed treatment. Mean Neck Disability Index fell from 24.71 before treatment to 14.21 at two months, 13.45 at six months, and 10.94 at twelve months, an average change of 13.77 from baseline to twelve months (p<0.0001) [6]. Of the 14 patients whose symptoms followed a motor vehicle accident, 13 were in litigation — a group most trials screen out — and the authors reported that litigation was not a barrier to recovery [6].

The third is the paper that gave this area its functional spinal unit framing: a registry case series treating the disc, the facet joints, and the supporting ligaments of a cervical segment as one mechanical unit rather than as isolated targets [7]. Of the patients treated between August 2014 and March 2017 with platelet-rich plasma, platelet lysate, or dextrose prolotherapy, only 14 had complete two-year outcome data [7]. In those 14 (mean age 59.8 years, 64% female), numeric pain scores fell by a mean of 2.8 points (p=0.002) and Functional Rating Index scores by 27.3 points (p=0.004) at 24 months, with 79% meeting the minimal clinically important difference on each [7]. Two of the 14 reported adverse events — a skin rash with increased post-procedure pain, and injection-site itchiness — both self-resolving [7]. The paper reports no formal exclusion criteria, and its authors state that larger randomized controlled trials are needed to validate the findings [7].

Why the preparation is not a detail

A 2025 pooled analysis compared the two preparations directly: 36 people who received leukocyte-rich platelet-rich plasma in the Calgary series against 21 who received leukocyte-poor platelet-rich plasma in the randomized trial [5]. At six months the leukocyte-rich group did better on pain (β = −1.68, p<0.01) and on the Neck Disability Index (β = −3.66, p=0.02); 63.9% versus 28.6% achieved at least a two-point pain reduction (p=0.02), and 90.3% versus 61.9% achieved at least a 10% disability improvement (p=0.03) [5]. There were no significant differences at three months [5].

Take that with the caution its own authors name. Pooling a case series with a randomized trial mixes study designs, patient selection, blinding, preparation methods and injection technique, and platelet counts were not routinely quantified for the leukocyte-poor product [5]. It is a signal that preparation matters, not proof of which preparation to use.

This is a large part of why the field is unsettled. Different platelet concentrations, different leukocyte content, different injectate volumes, different numbers of joints treated, different intervals, small samples — studies using the same three-letter abbreviation are frequently not testing the same thing. Our orthobiologics overview sets out which preparations are used here and for which regions.

How the whole body of evidence is graded

A 2025 systematic review of regenerative medicine for axial spine pain of facet joint origin identified four randomized controlled trials and six observational studies, with lumbar facet joint studies predominating [8]. It graded the evidence for platelet-rich plasma injections in facet joints as Level II, or moderate, and Level IV, or limited, overall, with low certainty, giving platelet-rich plasma a moderate recommendation and mesenchymal stem cells a weak one [8].

Moderate-to-limited with low certainty is a fair description of where this stands. It is not nothing, and it is not enough to promise you an outcome.

Who these studies left out, and why that matters if your case is complicated

The gap between a trial result and your situation is not a matter of opinion. It is written into the methods sections, and it is worth reading them.

The exclusions that are written down

Collected across the cervical studies above, the stated criteria excluded: nerve root compromise; fracture or dislocation at the time of injury; whiplash-associated disorder grades III and IV; known or suspected serious spinal pathology such as metastatic disease of the spine; previous cervical spine surgery; spinal surgery or radiofrequency neurotomy within 12 months; a cervical facet corticosteroid injection within the previous three to six months; post-concussion syndrome; a history of bipolar disorder, psychosis or schizophrenia, or any uncontrolled mental health condition; inability to stop anti-inflammatory medication for three days before and ten days after injection; and inability to understand or complete the questionnaires in English [3][4].

Each of those is defensible as trial design. Together they describe a patient who is comparatively uncomplicated. If you have had a fusion, if you have radicular symptoms alongside axial pain, if your last steroid injection was two months ago, if you take an anti-inflammatory you cannot stop for two weeks, or if you carry a psychiatric diagnosis that is managed rather than resolved, you would not have been enrolled in these studies — and their numbers were never a forecast for you.

That cuts in both directions, and we will say so plainly: being screened out is not evidence that a treatment would work better for you. It is evidence that nobody measured.

The exclusion nobody wrote down

None of the cervical studies cited on this page reports the metabolic status of its participants. Body mass index, hemoglobin A1c, diabetes, insulin resistance, and inflammatory or autoimmune disease do not appear among the reported baseline characteristics of the randomized trial, the prospective series, the case review, or the registry series [3][4][6][7]. The observational radiofrequency study discussed below is the only one of the set that names undocumented comorbidities as a potential source of bias in its own data [9].

That is a verifiable gap in the published record rather than an interpretation of it. Regen.MD evaluates metabolic markers before treating, because the metabolic and inflammatory terrain underlying degenerative joint disease is central to how this practice frames a degenerative problem. But the cervical facet literature cannot tell you how these injections perform in someone with poorly controlled diabetes or significant insulin resistance, in either direction, because that variable was not recorded.

The selection nobody controls

Two of the series name a selection effect that has nothing to do with medicine. In both the Calgary series and the two-year registry series, these treatments were not covered by insurance and many participants funded them personally; both papers caution that because wealth is associated with better health, their results may not generalize to the wider population [4][7]. The registry paper adds that an out-of-pocket investment could, in theory, increase a patient’s motivation to perceive improvement [7].

Coverage is a payer decision, not a clinical finding. Insurers determine what they will pay for on their own terms, and patients here sign an Advance Beneficiary Notice for services that are not covered. A non-covered service is not thereby an ineffective one, and this page does not use a payer’s position as an argument in either direction. It does, however, distort who ends up in the published studies, and that is worth knowing when you read them.

What single-modality trials cannot test

All of these studies tested an injection more or less by itself. That is sound experimental design, and it is not how care is delivered.

The studies themselves keep running into this. In the Calgary series, participants were encouraged to attend physiotherapy after the injection, attendance was not monitored, and the authors note that any improvement would likely reflect the combined effect of platelet-rich plasma and physiotherapy rather than either alone [4]. In the earlier dextrose case review, patients who attended physiotherapy over the course of treatment had better outcomes than those who did not, and the authors reported that concurrent physiotherapy helped reduce post-procedure neck stiffness [6]. The registry series treated disc, facet joints and ligaments together as one unit, and so cannot attribute its result to the facet injection alone [7].

None of that establishes that combination care works better. It does establish that the published evidence is not a clean test of the question most patients are actually asking, which is whether a sequenced program that includes an injection outperforms the same program without it. That study has not been done in the cervical spine.

How long anyone was followed

Follow-up in this literature runs to three months in the Calgary series, six months in the randomized trial and the pooled analysis, twelve months in the dextrose case review, and twenty-four months in the registry series [3][4][5][6][7]. Nothing here speaks to what happens after two years. Durability is a claim none of these studies can support, and we do not make it.

Radiofrequency neurolysis is a different question with different evidence

Radiofrequency neurolysis — thermal ablation of the medial branch nerves that carry pain from the facet joint — is not a regenerative treatment. It does not repair a joint; it interrupts the nerve supply, and the nerve grows back over time. Regen.MD offers it, and the evidence base for it in the cervical spine is larger than the regenerative evidence, so it belongs on this page even though it answers a different question.

The largest recent dataset is the cervical cohort of the RAPID study, an international multicenter prospective observational study sponsored by Boston Scientific Corporation and conducted using that manufacturer’s commercially available radiofrequency system [9]. The funding is disclosed here because it should weigh in how you read the numbers: several authors report consulting or advisory relationships with the sponsor, and three are its employees [9].

Fifty-six patients with chronic cervical facetogenic pain were enrolled — 60.7% female, mean age 59.5 years, mean pain duration 9.5 years, with 75% reporting at least two years of pain — at a mean baseline pain score of 6.0 ± 1.97 and a mean Oswestry Disability Index of 32.7 ± 15.6 [9]. Mean pain scores fell to 3.3 ± 2.8 at six months (n=49) and 3.3 ± 2.2 at twelve months (n=40), both p<0.0001, with Oswestry scores improving by 7.9 and 9.7 points respectively [9]. Response rates of at least 50% pain relief were 81% at one month and 80% at twelve months [9]. No device- or procedure-related serious adverse events were reported in this cohort [9].

The design cuts both ways, and this is the part worth understanding. RAPID’s only meaningful exclusions were cognitive impairment that would prevent a patient from assessing their own pain, and contraindications listed in the device instructions for use [9]. Nobody was screened out for comorbidity, prior surgery, symptom duration or general complexity, and a mean of nearly a decade of pain reflects that. External validity is exactly what a study of this design buys.

What it gives up is control. There was no comparison group, so improvement cannot be separated from natural history or from the effect of being enrolled and followed closely. The authors name unidentified confounders and undocumented comorbidities as potential sources of bias, note that the number of diagnostic blocks was not standardized across participating sites, and report that retention fell to 17 of the original 56 patients at 24 months, with more than half of those having undergone at least one repeat ablation [9]. The two-year figures should be read with that in mind.

What a sequenced pathway looks like for the neck

Conservative care comes first, and for many people it is where the answer is. That means loading and movement work, addressing sleep, workstation and posture contributors, and giving those measures a genuine trial against a defined review point rather than an open-ended one. Both cervical series that reported on it found conservative care and injection entangled rather than opposed [4][6].

If a defined trial of conservative care has not met your goals, the next step is diagnostic rather than therapeutic: establishing whether the facet joints are in fact generating your pain, which in the published studies meant medial branch blocks [3][4]. Only then is there a meaningful conversation about which injection, or about whether an injection is the right move at all.

If conservative measures have been exhausted and your goals are still not met, that conversation includes interventional options and, where indicated, surgery. Dr. Gurpreet Singh Padda, MD, MBA, MHP is a surgeon, and surgery is available here when it is warranted — as a later step in the sequence, never as the opening one.

“My ethos is to treat all of my patients as I would my own family, with the goal of giving them back their quality of life.” — Dr. Gurpreet Singh Padda, MD, MBA, MHP

How risk and benefit are discussed here

Plainly, and with the numbers that exist. Reported adverse events in the cervical regenerative literature have been minor and self-limiting: none among the 44 people in the Calgary series [4], two of 14 in the registry series [7], and comparable rates between platelet-rich plasma and corticosteroid in the randomized trial, in which the platelet-rich plasma group reported less procedural pain [3]. The pooled analysis recorded 11 adverse events — pain, spasms and headache — in the leukocyte-poor group and none in the leukocyte-rich group [5].

Those are small samples. Small samples detect common problems and miss rare ones, and that limitation is stated rather than glossed over. What can be told to you is what has been reported, in whom, and over what period — the same standard applied to the benefit figures above.

What we will not tell you

We will not tell you that an injection rebuilds a cervical facet joint. We will not extrapolate a number from the knee, where the joints are large, accessible and far more heavily studied, to your neck, because results do not carry across joints by analogy. We will not present a six-month comparison as a durability claim, and we will not describe a population nobody studied as a population we know how to treat.

Where the evidence is absent, this page says it is absent. Where it is genuinely unsettled, it says why.

Where to find us

Regen.MD is at 4477 Woodson Rd, Suite 103, St. Louis, MO 63134, next to St. Louis Lambert International Airport. Directions and parking are on the locations page, and you can reach the office at (314) 295-3000 or text (314) 886-5902.

Frequently asked questions

Sources

1. Lord SM, Barnsley L, Wallis BJ, Bogduk N. Chronic cervical zygapophysial joint pain after whiplash. A placebo-controlled prevalence study. Spine (Phila Pa 1976). 1996;21(15):1737-1744. https://doi.org/10.1097/00007632-199608010-00005 (referenced for: 60% prevalence, 95% CI 46–73%, of cervical zygapophysial joint pain among 68 consecutive patients with chronic neck pain after whiplash, established with placebo-controlled diagnostic blocks).
2. Kim JH, Sharan A, Cho W, Emam M, Hagen M, Kim SY. The Prevalence of Asymptomatic Cervical and Lumbar Facet Arthropathy: A Computed Tomography Study. Asian Spine Journal. 2019;13(3):417-422. https://doi.org/10.31616/asj.2018.0235 (referenced for: 33% prevalence of grade 1–3 cervical facet arthritis across 500 facet joints in 50 asymptomatic adults, the grade breakdown, the 78% figure at C6–C7 in subjects older than 40, and the study’s retrospective limitation).
3. Allison DJ, Ebrahimzadeh S, Muise S, Joseph S, Roa Agudelo A, Lawson A, Billias N, Tran J, Smith A, Loh E. Intra-articular corticosteroid injections versus platelet-rich plasma as a treatment for cervical facetogenic pain: a randomized clinical trial. Regional Anesthesia and Pain Medicine. 2026;51(1):33-40. https://doi.org/10.1136/rapm-2024-105889 Trial registration ClinicalTrials.gov NCT04392999, https://clinicaltrials.gov/study/NCT04392999 (referenced for: the randomized double-blind design, 55 enrolled and 40 analyzed, comparable six-month outcomes between leukocyte-poor platelet-rich plasma and corticosteroid, lower procedural pain and greater one-month pain self-efficacy in the platelet-rich plasma group, and the registered inclusion and exclusion criteria).
4. Smith A, Andruski B, Deng G, Burnham R. Cervical facet joint platelet-rich plasma in people with chronic whiplash-associated disorders: A prospective case series of short-term outcomes. Interventional Pain Medicine. 2022;1(2):100078. https://doi.org/10.1016/j.inpm.2022.100078 (referenced for: n=44, demographics and median 24-month symptom duration, inclusion and exclusion criteria, three-month pain and Neck Disability Index results with p-values and responder percentages, absence of reported adverse events, the physiotherapy confound, and the out-of-pocket generalizability caution).
5. Allison DJ, Loh E, Burnham R, Burnham T, Smith A. Leukocyte poor platelet rich plasma vs leukocyte rich platelet rich plasma as a treatment for cervical facetogenic pain: A pooled analysis. Interventional Pain Medicine. 2025;4(1):100566. https://doi.org/10.1016/j.inpm.2025.100566 (referenced for: the 36 versus 21 pooled comparison, six-month pain and disability coefficients and responder percentages, the 11 adverse events in the leukocyte-poor group, and the authors’ limitations regarding pooling designs and unquantified platelet counts).
6. Hooper RA, Frizzell JB, Faris P. Case series on chronic whiplash related neck pain treated with intraarticular zygapophysial joint regeneration injection therapy. Pain Physician. 2007;10(2):313-318. https://doi.org/10.36076/ppj.2007/10/313 (referenced for: 18 treated sides in patients who had failed other conservative and interventional procedures, the 20% dextrose intra-articular technique, Neck Disability Index values at 2, 6 and 12 months with p<0.0001, the litigation finding, and the physiotherapy association).
7. Williams C, Jerome M, Fausel C, Dodson E, Stemper I, Centeno C. Regenerative Injection Treatments Utilizing Platelet Products and Prolotherapy for Cervical Spine Pain: A Functional Spinal Unit Approach. Cureus. 2021;13(10):e18608. https://doi.org/10.7759/cureus.18608 (referenced for: the functional spinal unit framework, n=14 with complete 24-month data, the treatment window and injectates, 24-month numeric pain score and Functional Rating Index changes with p-values and minimal clinically important difference percentages, the two adverse events, the absence of stated exclusion criteria, the out-of-pocket generalizability caution, and the authors’ call for larger randomized controlled trials).
8. Manchikanti L, Abd-Elsayed A, Kaye AD, Sanapati MR, Pampati V, Shekoohi S, Hirsch JA. A Systematic Review of Regenerative Medicine Therapies for Axial Spine Pain of Facet Joint Origin. Current Pain and Headache Reports. 2025;29(1):61. https://doi.org/10.1007/s11916-025-01376-1 (referenced for: four randomized controlled trials and six observational studies identified, the predominance of lumbar facet studies, the Level II moderate and Level IV limited grading with low certainty, and the moderate recommendation for platelet-rich plasma versus weak for mesenchymal stem cells).
9. Provenzano DA, Holt B, Danko M, Atallah J, Iqbal M, Shah B, Singh A, Sachdeva H, Ver Donck E, Liebrand B, Shaw E, Haas S, Sekhar R, Pan A, Halperin DS, Goldberg E. Assessment of real-world, prospective outcomes in patients treated with cervical radiofrequency ablation for chronic pain (RAPID). Interventional Pain Medicine. 2025;4(3):100623. https://doi.org/10.1016/j.inpm.2025.100623 Trial registration ClinicalTrials.gov NCT04673032, https://clinicaltrials.gov/study/NCT04673032 Funding disclosure: this study was sponsored by Boston Scientific Corporation, the manufacturer of the radiofrequency system studied; several authors report consulting or advisory relationships with the sponsor and three are employed by it. (Referenced for: the cervical cohort of 56 patients, demographics and 9.5-year mean pain duration, baseline and follow-up pain and Oswestry Disability Index values with p-values, responder rates, absence of serious device- or procedure-related adverse events, the minimal exclusion criteria, and the limitations regarding confounders, non-standardized diagnostic blocks and 24-month retention. This is radiofrequency ablation, not a regenerative treatment.)