I remember my early years in interventional physiatry with a clarity that only comes from repetition. A young patient, often in their twenties or thirties, would walk in with a single, unremarkable complaint: axial low back pain, non-radiating. The exam told a consistent story: pain reproduced with extension and rotation, no radicular findings, and no neurological deficits. Imaging was equally quiet: maybe a mildly bulging disc, some early facet hypertrophy, nothing alarming.

There was real satisfaction in those visits. I could walk a patient through the findings and offer a clear plan: fluoroscopically guided lumbar facet injections. Most of the time, it worked. Relief became the foundation for a rehabilitation process that succeeded, because the pain was quiet enough to let patients move, strengthen, and recondition muscles that had been guarding and compensating for months.

Then the algorithm changed. Insurers began requiring medial branch blocks as a prerequisite for facet joint injections, ostensibly to confirm the pain generator before committing to treatment. On paper, that sounds like reasonable diagnostic rigor. In practice, it restructured the pathway. Once a patient had a positive medial branch block, the natural next step, the one reimbursement structure nudged everyone toward, was no longer a simple intra-articular injection. It was radiofrequency ablation (RFA). RFA started being used more indiscriminately, applied to patients who, a few years earlier, would have been managed with a single injection and physical therapy.

I don’t think anyone intended this shift to be reckless. But it changed who received RFA and at what age, without a proportional change in how carefully we thought about what RFA actually does.

RFA targets the medial branch of the dorsal ramus, the same nerve that innervates the facet joint we’re trying to quiet and the multifidus muscle, the deepest and most important segmental stabilizer of the lumbar spine. When we lesion that nerve for pain relief, we are not selectively silencing pain fibers. We are denervating a muscle.

This isn’t a new observation. Dreyfuss and colleagues described it in 2009, in a small but important prospective study of five patients who underwent unilateral RFA. Blinded radiologists reviewing MRIs 17 to 26 months later found diffuse multifidus atrophy, though in that study, the atrophy didn’t clearly correlate with recurrent pain during that window. More recent work has caught up with that early signal. A 2024 systematic review consolidated the evidence connecting lumbar RFA to measurable multifidus atrophy and dysfunction, and a 2025 narrative review went further, naming this a clinically significant, understudied area deserving more rigorous prospective research. Newer data, including a 2024 study by Guven and colleagues, points to increased fatty infiltration and muscle wasting on the treated side following RFA, with plausible links to reduced stabilizing strength.

None of this makes RFA a bad procedure. For a patient with degenerative, facet-mediated pain later in life, where multifidus reserve is already limited, trading some muscle function for durable pain relief may be entirely reasonable. That calculus looks very different in a 30-year-old with a healthy multifidus and forty more years of spinal loading ahead.

Here is a real gap, not just in the literature but in daily practice. Most multifidus research is built on populations with established degenerative disc disease and years of chronic pain. In those patients, multifidus atrophy on MRI reads as expected, just another line in the general pattern of an aging spine. Nobody pauses to ask whether it predated treatment or was caused by it.

In a younger patient, that assumption doesn’t hold. A healthy 25- or 30-year-old typically has a robust, well-preserved multifidus. If atrophy appears on a follow-up MRI after RFA, it is far more likely to be a direct consequence of the procedure than a feature of natural spinal aging. Yet this is precisely the population least likely to have that atrophy flagged, because it isn’t part of the standard MRI checklist, and because nobody is routinely looking for it.

Part of why this atrophy goes unflagged is practical: manually segmenting and quantifying a muscle like the multifidus, distinguishing lean tissue from fatty infiltration slice by slice, is time-consuming and has never been part of a routine lumbar MRI read. That may be changing. Deep-learning segmentation models are now being trained specifically for this task. Recent multicenter work published in Radiology: Artificial Intelligence demonstrated automated multifidus and paraspinal muscle segmentation approaching the accuracy of manual expert measurement, and a similar deep-learning approach achieved comparably high accuracy in segmenting the multifidus and erector spinae on T2-weighted imaging. Broader tools such as TotalSegmentator MRI are extending automated, sequence-independent segmentation across dozens of anatomic structures, muscles included. If muscle volume and fat-infiltration ratio can be generated automatically every time a lumbar MRI is obtained, multifidus assessment could become a routine, quantitative part of the report rather than an omitted afterthought, an objective baseline for a younger RFA candidate, and an objective way to track change afterward, without depending on a radiologist happening to comment on it.

A few things, practically. First, a more deliberate consent conversation: when we offer RFA to a younger patient, we should be explicit that we are denervating a stabilizing muscle, not simply turning off a pain signal, and that this trade-off carries different weight at 30 than at 65. Second, multifidus assessment, whether manual or AI-assisted, should become a standard element of imaging for younger RFA candidates, at baseline and follow-up. Third, rehabilitation shouldn’t stop once pain improves; targeted multifidus-specific stabilization exercise deserves a sustained role after RFA, not just after injections.

I still think about those early years, when the plan was simple: quiet the joint, let the patient move, and let rehabilitation do the rest. I’m not arguing for a return to a simpler time; the diagnostic rigor of medial branch blocks has real value. But we owe our younger patients, the ones with the most years of spinal function ahead of them, a treatment algorithm that weighs durability of relief against long-term structural cost, not one shaped primarily by prior-authorization requirements.

The multifidus doesn’t appear prominently on a standard radiology report. With the right tools, that may finally change.

Francisco M. Torres is an interventional physiatrist specializing in diagnosing and treating patients with spine-related pain syndromes. He is certified by the American Board of Physical Medicine and Rehabilitation and the American Board of Pain Medicine and can be reached at Florida Spine Institute and Wellness. 

Dr. Torres was born in Spain and grew up in Puerto Rico. He graduated from the University of Puerto Rico School of Medicine. Dr. Torres performed his physical medicine and rehabilitation residency at the Veterans Administration Hospital in San Juan before completing a musculoskeletal fellowship at Louisiana State University Medical Center in New Orleans. He served three years as a clinical instructor of medicine and assistant professor at LSU before joining Florida Spine Institute in Clearwater, Florida, where he is the medical director of the Wellness Program.

Dr. Torres is an interventional physiatrist specializing in diagnosing and treating patients with spine-related pain syndromes. He is certified by the American Board of Physical Medicine and Rehabilitation and the American Board of Pain Medicine. He is a prolific writer and primarily interested in preventative medicine. He works with all of his patients to promote overall wellness.

Emory Hubbard is an undergraduate student.