Prolotherapy for persistent low back pain after surgery

Marc Darrow, MD, JD

Persistent low back pain after surgery does not automatically mean that another operation is needed. It also does not establish that prolotherapy, PRP, or a cell-based injection is the right next step. The useful starting question is what explains the pain now.

Arthur raised this problem on Living Pain-Free. He described an unsuccessful back operation, repeated scans reporting mild disc herniations at L4–L5 and L5–S1, extensive conservative treatment, and temporary relief from spinal injections. His symptoms remained centered in his lower back rather than extending down a leg.

He wanted to know whether regenerative injections could help after so many other treatments had not. His question is a reasonable one. Answering it requires matching the treatment to the current diagnosis, rather than assuming that the disc named on the MRI must be the cause—or that it cannot be the cause.

Does a disc herniation on MRI explain the pain?

Disc degeneration, bulges, and protrusions are common in people without pain. A systematic review of 33 studies included 3,110 asymptomatic people and found that many degenerative imaging features became more prevalent with age. The authors emphasized interpreting imaging in the context of the person’s clinical condition. (1)

This helps explain why a scan is not a pain measurement. A disc abnormality can be important, incidental, or one contributor among several. Finding the same abnormality on repeated scans does not resolve which of those possibilities applies.

Imaging remains valuable. After surgery it may help assess recurrent nerve compression, narrowing, instability, hardware problems, fracture, infection, or another condition. A physical examination and an MRI answer different questions; neither should be dismissed simply because the other is informative.

Arthur’s scan could not be interpreted adequately during a radio call. His surgical history, symptom pattern, neurologic findings, and prior treatment responses would all affect the assessment.

Pain in the back and pain down the leg are different patterns

Back-centered pain can involve discs, facet joints, the sacroiliac region, muscles, or connective-tissue attachments. Leg pain may involve an irritated nerve root, but it can also be referred from another structure. More than one problem can coexist.

Weakness, changes in sensation, reflex abnormalities, and the pattern of aggravating movements can help distinguish possibilities. The absence of leg pain does not rule out disc-related pain, and pain near a ligament attachment does not by itself prove that the ligament is responsible for every symptom.

Marc Darrow emphasized examining where muscles and ligaments attach to bone. These attachment sites are called entheses, and a disorder there is an enthesopathy. This is a useful part of the differential diagnosis. It is not evidence that nearly all low back pain has the same source.

The question is whether the proposed explanation accounts for the person’s symptoms and limitations. A previous operation can address one problem while another source of pain remains or develops later.

What does “failed back surgery” mean?

The term describes persistent or recurrent symptoms after an operation; it does not identify one mechanism. A person may have ongoing nerve-related pain, recurrent structural disease, altered movement, deconditioning, or a separate pain source around the pelvis or spine.

It is also useful to distinguish pain that never improved from pain that returned after a period of relief. A new injury, progressive weakness, or a change in the distribution of symptoms can create a different clinical question from stable longstanding pain.

The site’s discussion of persistent pain after spinal surgery provides additional context. Further surgery may be appropriate when there is a clear problem it can address. It should not be recommended solely because pain persists. Similarly, an injection needs a defined purpose rather than being offered simply because surgery was disappointing.

What is prolotherapy?

Prolotherapy is an injection-based treatment, commonly using dextrose, sometimes with a local anesthetic. The term comes from “proliferation therapy,” and related approaches have also been called regenerative injection therapy.

Higher-concentration dextrose injections are intended to influence local tissue responses around selected painful structures. Other dextrose approaches use lower concentrations and different targets. Proposed effects include changes in connective-tissue repair and pain signaling, but those mechanisms are not interchangeable or established by symptom improvement alone. (2)

The concentration, injection location, guidance method, and diagnosis should be specified when discussing evidence. A study of dextrose inside the sacroiliac joint is not a trial of injections into the lumbar disc, an epidural procedure, or treatment of every postoperative back problem.

Prolotherapy, PRP, and bone marrow are different preparations

PRP is prepared from the patient’s blood and contains concentrated platelets and signaling proteins. Bone marrow aspirate and bone marrow concentrate contain mixtures of marrow cells and other components. They are different from PRP and from purified or culture-expanded cell preparations.

These treatments should not be ranked automatically as progressively stronger versions of the same injection. Different ingredients and biological rationales do not establish greater effectiveness for a particular pain source.

Nor should all three be described simply as irritants that create inflammation and then grow cartilage. A proposed mechanism is not proof that a disc, ligament, or joint has regenerated. Clinical improvement and structural restoration are separate outcomes.

The site’s PRP injections for back pain and sacroiliac joint pain discussion address related but distinct treatment questions.

Sacroiliac treatment: ligament and joint studies should be separated

The sacroiliac joint connects the sacrum and pelvis. Its supporting ligaments and the joint itself can be relevant to pain in the lower back or buttock. Symptoms in this region do not establish sacroiliac dysfunction without further assessment.

A prospective descriptive study enrolled 25 patients with deficient sacroiliac load transfer. They received three CT-guided dextrose injections into the dorsal interosseous ligament, six weeks apart. Functional measures improved among patients assessed at follow-up, but attendance declined over time and there was no comparison group. These findings cannot supply a reliable general success rate. (3)

A separate randomized study enrolled 48 patients whose sacroiliac pain had responded to a diagnostic anesthetic block. It compared fluoroscopy-guided injections of dextrose into the joint with steroid injections. Both groups improved initially. Sustained pain relief favored prolotherapy during longer follow-up. (4)

This is useful evidence for a selected sacroiliac population. It does not prove that ligament injections and joint injections are equivalent, or that prolotherapy is superior to steroids for every cause of low back pain. Neither study establishes that Arthur’s symptoms were sacroiliac in origin.

What has been studied after back surgery?

A 2019 consecutive case series evaluated 79 patients with persistent symptoms after back surgery. Participants had at least six months of symptoms and had not responded to a further course of conservative care. Pain and disability scores improved following 5% dextrose injections using posterior and lateral approaches. (5)

The study supports further investigation, but it had no randomized comparison group. It cannot determine how much improvement was due to the injections rather than time, other care, or differences in patient selection. It did not demonstrate that a disc regenerated or establish that revision surgery could always be avoided.

Because it used 5% dextrose, its findings should not be presented as proof for every higher-concentration ligament protocol. The actual injection method matters as much as the label “prolotherapy.”

New 2026 evidence: prolotherapy versus epidural steroid injection

A pragmatic randomized trial analyzed 104 people with chronic low back pain and associated leg pain. It compared dextrose prolotherapy, with up to five monthly treatments, against epidural steroid injection, with up to three monthly injections. Pain and function outcomes favored prolotherapy at one year after treatment completion. (6)

This is a useful comparative finding. However, treatment was unblinded, the allowable treatment schedules differed, and the study concerned back pain with leg pain. It cannot establish the same benefit for isolated back-centered pain or every postoperative diagnosis.

Arthur reported no leg pain during the call. That distinction should remain visible when discussing whether the trial is relevant to his situation. A favorable study can strengthen a treatment discussion without answering every patient’s question.

What does the broader prolotherapy literature show?

A 2026 systematic review included 12 studies covering several chronic spinal pain conditions. Most reported improvement in pain, and some described improvements in function or quality of life. However, only three were randomized trials; the remainder included case series and other nonrandomized designs. Diagnoses, concentrations, and injection targets varied. (2)

The review supports prolotherapy as an approach deserving further research and consideration in selected circumstances. It does not establish one success rate, one treatment schedule, or proof of tissue regeneration across all spinal disorders.

The review searched through June 2025, so the newer 2026 randomized trial should be considered separately. Its publication date does not mean every study published during 2026 was included.

Intradiscal PRP: a different target and mixed trial results

Injecting PRP inside a disc is different from injecting around a ligament, into a facet joint, or into the sacroiliac joint. Evidence from one location should not be used to promise results at another.

A small randomized trial analyzed 47 patients with selected chronic discogenic pain. Intradiscal PRP improved certain pain, function, and satisfaction measures over eight weeks compared with control. Control patients could later cross over to PRP, limiting the longer-term comparative conclusions. (7)

Another randomized trial assigned 98 patients with discogenic pain without Modic changes on MRI to intradiscal PRP or a saline-based control. It analyzed 89 participants with complete data and found no significant advantage in pain or function at one year. One case of spondylodiscitis, an infection involving the disc and adjacent vertebrae, occurred after PRP. (8)

These findings show why intradiscal PRP cannot be described as consistently effective or risk-free. They also do not establish that the disc is the right target for someone with back pain after surgery.

What does temporary relief from a spinal injection mean?

Arthur reported that steroid or nerve-block injections made him feel somewhat better for a time. The meaning of that response depends on where the injection was placed, which medications were used, and whether the purpose was diagnostic or therapeutic.

Temporary relief does not by itself prove that the underlying tissue has healed. It also does not establish that the injection caused tissue damage or that another type of injection will provide permanent relief.

Likewise, unsuccessful acupuncture, chiropractic treatment, massage, or traction does not guarantee that prolotherapy or PRP will succeed. Prior treatment may have targeted a different mechanism, provided incomplete benefit, or been insufficient for the person’s particular condition. The current diagnosis still needs to make sense.

How many treatments are needed, and what if there is no improvement?

There is no single evidence-based number for all back pain injections. The studies above used different preparations, targets, and schedules. Their protocols are not a universal prescription.

Repeat treatment should have a reason based on the diagnosis, meaningful response, and the balance of benefit and risk. Persistent symptoms may indicate that the original explanation needs reconsideration. Lack of improvement should not be blamed automatically on a patient stopping too soon or being too active.

Progress can be assessed through tolerable walking, sitting, sleep, daily activities, strength, and activity recovery—not only a pain score immediately after an injection. Meaningful symptom improvement remains valuable even when no structural regeneration has been demonstrated.

Soreness is possible, but spinal procedures have other risks

Transient soreness or stiffness can follow an injection. Bleeding, infection, medication reactions, nerve injury, and other complications are also possible, with risks depending on the site and technique. Appropriate guidance and sterile technique matter.

Small studies reporting no serious complications cannot exclude uncommon harms. Safety findings from an injection around a peripheral ligament should not be generalized to an epidural or intradiscal procedure. A description such as “natural” or “regenerative” does not establish that a procedure is harmless.

Rehabilitation and urgent warning signs remain important

Rehabilitation can address strength, movement, and activity tolerance, but it should fit the diagnosis. Severe unexplained pain is a reason to reassess, rather than simply push harder through exercise. A symptom flare during loading is also different from evidence of new structural damage.

New or progressive weakness, loss of bladder or bowel control, numbness around the saddle region, fever, or severe pain after a new injury warrants urgent assessment. These concerns should not be managed as a routine question about another injection.

Does persistent pain mean another operation is inevitable?

No. Nonsurgical care may still be useful, and some patients have a treatable pain source outside the area addressed by the original operation. Others have recurrent compression, instability, or another condition that deserves a surgical discussion.

Before selecting another procedure, the explanation should identify what is being treated, which findings support that diagnosis, what the treatment is expected to change, and what evidence applies. The useful question is more specific than “Can stem cells fix my back?”

Arthur’s call illustrates the value of revisiting a diagnosis after unsuccessful surgery. It does not establish that his pain was an easy fix. The best opportunity to choose a useful treatment comes from understanding the current problem and setting realistic functional goals.

References