- What are interspinous process devices?
- Since when are they used, and why?
- What is the biomechanical principle?
- When are they indicated? Is the indication still valid?
- What advantages do they have over decompression alone?
Had spine surgery and the pain returned? Send your studies for a second opinion.
Send my studiesAn interspinous process device (IPD) is a mechanical implant inserted between two adjacent spinous processes — the bony projections at the back of each vertebra — to create distraction between them. The goal is to produce indirect decompression of the spinal canal and neural foramina without removing bone. Theoretically, by pushing the spinous processes apart, the device increases the height of the foramen and the diameter of the canal in extension.
Implanted without bone resection as the sole treatment for neurogenic claudication. The X-Stop (Medtronic) was the prototypical device. FDA approved 2005. Largely abandoned due to inferior results vs. conventional decompression.
Implanted after open decompression as a dynamic stabilizer, marketed as an alternative to fusion. Coflex (Paradigm Spine) is the best-studied example. FDA approved 2012. Also associated with device-specific complications over time.
Other devices include the DIAM, Wallis, and Superion systems. All share the same underlying mechanical principle; they differ in material (titanium, PEEK, polyester), design geometry, and the extent of surgical exposure required for placement. (Gazzeri, 2014)
Interspinous process devices emerged in the 1990s, initially in Europe, as a response to a genuine clinical problem: lumbar spinal stenosis in elderly patients who were poor surgical candidates for conventional laminectomy under general anesthesia. The appeal was straightforward — a device that could be placed under local or regional anesthesia, through a small incision, without bone removal, in a patient population with significant comorbidities.
A second wave of interest emerged in the 2000s with the Coflex-type devices, this time marketed not as a standalone treatment but as an alternative to fusion following decompression. The rationale shifted: if decompression alone causes instability, and fusion eliminates motion and causes adjacent level disease, perhaps a dynamic stabilizer could offer a middle path.
Both waves had something in common: the devices were adopted widely before long-term evidence was available. What followed was a pattern that recurs throughout the history of spinal implants — early enthusiasm, commercial expansion, and then a reckoning with the data. (Epstein, 2012)
In lumbar spinal stenosis, symptoms are classically position-dependent: they worsen in extension (standing upright, walking) and improve in flexion (sitting, leaning forward). This is because extension narrows the spinal canal and foramina — the ligamentum flavum buckles inward, the facet joints approximate, and disc bulge increases posteriorly.
An interspinous spacer placed between the spinous processes prevents the segment from going into full extension, forcing it into a relatively neutral or flexed position. This theoretically maintains foraminal height and canal diameter in positions that would otherwise be symptomatic. (Alfieri, 2012)
This mechanism ONLY applies to extension-dependent, position-sensitive stenosis in a mechanically stable segment. It produces no decompression of fixed structural stenosis (bone osteophyte, hypertrophied ligamentum flavum), provides no benefit in spondylolisthesis (where it fails dramatically), and does nothing for canal compromise that is independent of position. The elegance of the biomechanical concept concealed the narrowness of its actual indication.
The original indication was narrow and specific: mild to moderate neurogenic claudication, clearly position-dependent symptoms, single or two-level involvement, no spondylolisthesis, and a patient who is a poor candidate for conventional surgery. Even within that narrow definition, randomized trials comparing IPDs to conventional decompression have consistently shown IPDs to be inferior. (Strömqvist, 2013; Moojen, 2015; Lonne, 2015)
| Study | Design | Finding |
|---|---|---|
| Strömqvist 2013 | RCT: X-Stop vs decompression, 2-year follow-up | Conventional decompression significantly superior; higher reoperation rate in X-Stop group |
| Moojen 2015 | Double-blind RCT: IPD alone vs conventional decompression | IPD without decompression inferior at 2 years; no benefit over conventional surgery |
| Verhoof 2008 | Prospective: X-Stop in stenosis with spondylolisthesis | High failure rate: 60% reoperation rate — device contraindicated in spondylolisthesis |
| Lonne 2015 | RCT: MIS decompression vs X-Stop | MIS decompression significantly superior at all time points |
Is the indication still valid? In most spine centers that have followed the long-term data, the answer is no. The standalone X-Stop has been withdrawn from many markets. The 2022 JAMA review on lumbar stenosis management does not list IPDs as a recommended treatment option for most patients. The Coflex device retains a narrow indication in selected patients after decompression, but even there the evidence is contested. (Katz, 2022)
This is where the framing matters: the advantages of IPDs have consistently been compared to fusion, not to decompression alone. When compared to conventional laminectomy or laminotomy, the picture is very different.
- +Shorter operative time and less blood loss than instrumented fusion
- +Preserves segmental motion — theoretically avoiding adjacent level disease associated with fusion
- +Potentially implantable under local or regional anesthesia in frail patients
- +Reversible in theory — device can be removed if it fails
- −Network meta-analysis: decompression alone superior to standalone IPD for functional outcomes and symptom relief (Zhang, 2021)
- −Systematic reviews (Wu, 2014; Zhao, 2017; Phan, 2016): IPD alone does not achieve better outcomes than conventional decompression and carries higher reoperation rates
- −The 'advantages' are specific to the comparison vs. fusion — not vs. decompression, which is simpler, less expensive, and does not add a device with its own failure modes
The longer the follow-up, the less favorable the picture for interspinous devices. This is the pattern across the literature — early results sometimes reasonable, medium-term results deteriorating, and long-term data revealing complication rates that are difficult to justify. (Li, 2023; Zheng, 2021)
The heterotopic ossification problem
One of the most telling findings in long-term IPD follow-up is heterotopic ossification — new bone formation around and between the spinous processes at the site of the device. Over time, this bone bridges across the interspinous space and produces a spontaneous fusion of the segment. The device that was placed to preserve motion gradually converts the segment into a fused one — without the structural integrity of a proper surgical fusion, and with all the additional surgical exposure of the device implantation as sunk cost. (Tian, 2013; Lee, 2016)
The Coflex data tells a similar story at 8 years: a minority of patients maintain the theoretical biomechanical benefit of the device; the majority have either had complications, required reoperation, or experienced results no better than decompression alone would have produced — without the device and its associated risks. (Zheng, 2021; Du, 2020)
The complication profile of interspinous devices is distinct from conventional spine surgery: it includes both standard surgical complications AND device-specific failure modes. Analysis of real-world adverse events from the FDA MAUDE database revealed a broad spectrum of complications not captured in industry-sponsored trials. (Aggarwal, 2021; Epstein, 2012)
- Spinous process fracture — the most common device-specific complication; particularly frequent in osteoporotic patients, where the spinous processes cannot withstand the distracting forces; often requires device removal and reoperation
- Device migration or displacement — the implant shifts from its original position, potentially causing new nerve compression or rendering the decompression ineffective
- Device fracture or breakage — implant structural failure; more common with titanium U-shaped designs (Tamburrelli, 2011)
- Heterotopic ossification — new bone progressively bridges the interspinous space, defeating the purpose of the device and creating a non-structural spontaneous fusion (Tian, 2013; Lee, 2016)
- Wound complications and infection — rates higher than expected for a supposedly minimally invasive procedure; the device creates a foreign body nidus
- Exacerbation of Baastrup's disease (kissing spines) — IPDs placed in segments with pre-existing interspinous contact can worsen the kissing spine syndrome
- Reoperation — the most clinically significant outcome: rates consistently higher than conventional decompression across multiple registries and multicenter studies (Gazzeri, 2015; Barbagallo, 2009; Bowers, 2010)
A European multicenter study of interspinous process device failure rates found clinically significant failure rates across all device types, with reoperation as the most common endpoint. The study identified that many of the complications were inherent to the device concept rather than to any specific technical error. (Gazzeri, 2015)
— Our position: why we do not use interspinous process devices
In our practice, we do not use interspinous process devices. This is not a reflexive or uninformed position — it is the conclusion we have reached after following the evidence over more than a decade, which has been consistent and, at this point, difficult to argue with.
The evidence does not support their use. Multiple randomized trials, systematic reviews, meta-analyses, and multicenter registries converge on the same finding: interspinous devices alone are inferior to conventional decompression; interspinous devices as an adjunct to decompression are not superior to decompression alone in the medium and long term; and the complication and reoperation rates are consistently higher than those of well-performed conventional surgery.
The heterotopic ossification problem is particularly telling. A device implanted to preserve motion that progressively generates new bone and produces spontaneous arthrodesis is not fulfilling its purpose. The patient ends up with a segment that is functionally fused — without the structural advantage of a surgical fusion, and with the added costs of the device, the reoperation for its complications, and the heterotopic bone itself.
The reoperation burden is not acceptable. When reoperation rates for IPDs exceed those of conventional decompression — a procedure that already carries a real but manageable revision rate — the device has created a net negative. A patient who undergoes IPD placement and then requires reoperation has experienced two surgical procedures when one well-performed decompression would have been sufficient.
We recognize that the original clinical problem — an elderly frail patient with neurogenic claudication who cannot tolerate general anesthesia and major surgery — is real. Our answer to that problem is not an interspinous device: it is minimally invasive decompression under spinal or local anesthesia, which addresses the structural problem without adding a device with an independent failure mode. The SPORT trial data (Weinstein, 2008) and the Cochrane review (Overdevest, 2015) are clear that when patients are appropriately selected, well-executed decompression produces durable results. The interspinous device literature has never demonstrated the same.