A Commentary on “How to Treat Adjacent-Segment Pathology Following Lumbar Fusion Surgery: Decompression Alone or Decompression With Extended Fusion Surgery?” Authors: Jang W. Yoon Published: 2026 Reoperation after lumbar decompression, with or without fusion, is a well-known problem. In spine surgery, we often tell patients, “Never say this is the last spine surgery.” Adjacent-segment pathology (ASP) is one reason for that statement [1]. When symptomatic stenosis develops above or below a prior fusion, the surgeon must decide whether to decompress the adjacent level alone or extend the fusion. We face this problem frequently, yet there is still no single answer that applies to every patient. The authors [2] should be congratulated for attempting to shed some light on this elusive question using the Swespine registry, which captures approximately 86% of spine surgeries performed in Sweden. This is a well-executed study with a large sample and long-term follow-up. A total of 2,521 patients were included: 1,176 underwent decompression alone and 1,345 underwent decompression with extension of fusion. The mean follow-up after surgery for clinical adjacent-segment pathology (CASP) was 6.1 years. Subsequent surgery was more common after extended fusion at every time point: 16% versus 11% at 2 years, 25% versus 18% at 5 years, and 31% versus 23% at 10 years. After adjustment, extended fusion remained associated with a higher risk of another operation (hazard ratio, 1.26). This is an important observation. It supports the idea that fusion should not be added automatically simply because a prior fusion is present. In a carefully selected patient with central or lateral recess stenosis, preserved alignment, no obvious instability, and no major foraminal collapse, decompression alone may be sufficient. It avoids additional instrumentation and generally permits a smaller operation, which is especially attractive in an older patient with medical comorbidities. Prior studies of adjacent-level stenosis and randomized trials of primary lumbar stenosis have similarly shown that decompression alone can provide comparable outcomes in selected patients [3-6]. However, I would be cautious about interpreting these findings as evidence that decompression alone is generally superior to extended fusion. This was a registry study, not a randomized trial, and the 2 treatment groups were meaningfully different. Patients treated with decompression alone were older, more likely to undergo a single-level operation, and most commonly had central stenosis without substantial spondylolisthesis. The extended-fusion group included many more patients with segment-related pain, degenerative spondylolisthesis, multilevel disease, postoperative instability, foraminal stenosis, and deformity. These are precisely the findings that lead a surgeon to choose fusion. Although statistical adjustment is helpful, it cannot fully eliminate confounding by indication. Treatment selection also reflected individual surgeon judgment and local practice patterns over a study period spanning more than 2 decades. These results therefore need to be taken with a grain of salt when applied to an individual patient. In my opinion, the biggest limitation is the absence of radiographic data. We do not know the degree of translation on flexion-extension radiographs, whether any spondylolisthesis was dynamic, the extent of disc-height loss or foraminal collapse, the presence of facet effusion, the orientation and integrity of the facet joints, the spinopelvic alignment, or the status of the prior fusion. These details are often more important than the diagnostic label entered into a registry. Without them, it is difficult to know whether the decompression and fusion groups were truly comparable. In addition, 10,052 patients were excluded because the index procedure could not be verified, and there was substantial missingness for body mass index and smoking status. The definition of “subsequent surgery” also deserves attention. Once instrumentation is added, there are additional pathways back to the operating room. Several of these events, particularly implant removal, loosening, pseudarthrosis, and implant revision, cannot occur after decompression alone. Combining every reoperation into a single endpoint therefore places the fusion group at some disadvantage. The pattern of failure may be more clinically informative than the overall number. After decompression alone, 7% of patients were eventually converted to fusion and 4% underwent another decompression. Extended fusion, on the other hand, was associated with more recurrent CASP at a new adjacent level (12.6% vs. 7.8%). At the treated segment, fusion appears to have done what it was intended to do, but it may have transferred the problem to the next mobile level. The study may therefore be less about which operation “wins” and more about where the next failure occurs. This tradeoff is consistent with prior clinical and biomechanical studies [7,8]. The practical message is not that decompression is always better or that fusion should be avoided. Rather, the operation should match the pathology. A limited, facet-preserving minimally invasive or endoscopic decompression is reasonable when neural compression is the dominant problem and the segment is stable. Extended fusion remains appropriate for dynamic instability, progressive spondylolisthesis, severe foraminal collapse, deformity, mechanical back pain attributable to the adjacent level, pseudarthrosis, or failure of the existing construct [9]. Future studies should incorporate the radiographic findings that surgeons actually use when making this decision, along with patient-reported outcomes, frailty, perioperative morbidity, cost, and the specific reason and location of each reoperation. A randomized trial may be difficult to perform, but a prospective, imaging-rich study limited to patients in whom true equipoise exists would be valuable. Until then, this study provides strong support for avoiding reflexive extension of fusion. It also reminds us that careful patient selection remains the most important part of CASP management.