All procedures were performed under general anesthesia with the patient in a prone position and the neck slightly flexed to widen the interspinous space. Neuromonitoring was used in all cases. A midline skin incision of approximately 2 cm was made at the target level, guided by preoperative imaging and intraoperative portable radiography. The interspinous space was carefully exposed using blunt dissection. Unlike conventional laminoplasty or Shiraishi’s method, in which the posterior musculature is detached or the spinous process is split, this approach preserved the interspinalis cervicis and semispinalis cervicis muscles. Retraction was achieved with a modified blunt-tipped mini-Gelpi retractor, minimizing muscle trauma. After exposure, the lower half of the superior lamina and the upper half of the inferior lamina were sequentially thinned with a high-speed drill. Decompression was then completed with a Kerrison punch, gradually undercutting the lamina to create adequate space while protecting the dura. In cases of multilevel disease, the spinous process at the intervening level was split to fully expose the lamina, and a total laminectomy was performed. Undercutting of the lower margin of the upper lamina and the upper margin of the lower lamina was subsequently carried out to ensure continuity of decompression. Hemostasis was achieved using bipolar cautery only, avoiding monopolar energy to reduce collateral thermal injury. No instrumentation or additional bone work beyond the planned decompression was required. The wound was closed layer by layer, and a single Jackson-Pratt drain was placed in most cases. Split spinous processes were sutured with 1-0 Vicryl. Postoperatively, patients were mobilized early without cervical bracing, and analgesics were required only for a short period.

Both patients underwent posterior cervical muscle-preserving interspinous process decompression without intraoperative complications or the need for conversion to conventional laminoplasty or laminectomy.

Case 1 (multilevel stenosis at C5–7): Postoperative MRI demonstrated adequate decompression at C5–6 and C6–7, with improvement of spinal cord contour and reappearance of the cerebrospinal fluid space (Figure 1D–F). At 1 month, arm pain had decreased from visual analogue scale (VAS) 7 to 1, neck pain had resolved (VAS 0), and the patient reported marked functional improvement, including a more stable gait. At 12 months, she maintained neurological stability and continued to report high satisfaction (5/5), with only minimal residual arm discomfort (VAS 2) and preserved cervical alignment without dynamic instability (Figure 1G–I).

Case 2 (C4–5 with OYL): Postoperative MRI confirmed complete decompression at the C4–5 level with restoration of the cerebrospinal fluid buffer around the spinal cord (Figure 2E and F). At one month, the patient reported resolution of preoperative neck pain (VAS 0) and significant improvement in left arm weakness (from grade 3 to 4+/5 in shoulder abduction and elbow flexion). At 12 months of follow-up, she maintained neurological stability without recurrence of pain (VAS 1), showed preserved cervical alignment (Figure 2G), and expressed high satisfaction with the outcome (4/5). No adverse events occurred during the rehabilitation period.

Clinical outcomes were summarized in Table 1. Both patients demonstrated meaningful improvement in the modified Japanese Orthopaedic Association (mJOA) and 36-item Short Form health survey (SF-36) scores, along with a marked decrease in VAS pain scores. Cervical alignment and segmental stability were maintained during follow-up.

Overall, both patients achieved meaningful neurological recovery, preservation of cervical lordosis, and minimal axial pain following this minimally invasive decompression. The accompanying surgical videos illustrate the operative steps in detail and highlight the technical nuances that contributed to these favorable outcomes.

This approach preserves both the posterior cervical musculature and the spinous process as an anchor point, thereby maintaining biomechanical stability [1-4]. Compared with laminoplasty, the operative corridor is narrower; however, reduced muscle detachment and preservation of bony landmarks result in less postoperative pain and faster recovery [5-8].

Posterior cervical decompression is most commonly achieved with laminectomy or laminoplasty [7,9-14]. While these procedures can effectively enlarge the spinal canal, they are often associated with detachment of the posterior musculature, disruption of spinous process anchor points, and subsequent axial pain or loss of cervical alignment [5,6,15,16]. To address these drawbacks, several muscle-preserving modifications have been introduced. Among them, Shiraishi’s selective laminectomy has been widely described, but it requires splitting of the spinous process and broad removal of lamina, which may compromise the posterior tension band [2-4,8,17].

The muscle-preserving interspinous process approach used in our 2 cases represents a further refinement of these techniques. By entering through the interspinous space and applying blunt retraction, the semispinalis and interspinalis cervicis are preserved and the spinous process remains intact [1,18,19]. These anatomical considerations are critical for maintaining segmental stability and reducing postoperative axial pain [7, 20]. In both of our patients, decompression was achieved effectively without evidence of postoperative kyphosis or functional decline, and both reported significant neurological improvement and high satisfaction.

Prior to surgery, both patients underwent an adequate period of conservative management including medication, physiotherapy, and nerve-block injections. Despite these treatments, their symptoms progressed—particularly hand clumsiness and gait imbalance in case 1 and progressive motor weakness in case 2—which justified surgical decompression. Although preoperative MRI did not reveal severe canal narrowing, both patients exhibited clear clinical signs of cervical myelopathy, including gait disturbance, hyperreflexia, and positive Lhermitte sign, accompanied by intramedullary cord signal change on T2-weighted images. Therefore, surgical intervention was deemed appropriate to prevent further neurological deterioration.

Although this report includes only 2 illustrative cases, the clinical feasibility and safety of the posterior cervical muscle-preserving interspinous process approach have been demonstrated in a larger cohort previously published by our group [19]. In that study of 20 consecutive patients who underwent the same technique, the mean mJOA score improved from 12.6 to 16.2, and the mean Neck Disability Index decreased from 15.4 to 2.5 at 3 months postoperatively. Neck VAS scores also decreased to 0.8, and patient reported satisfaction averaged 4.7 of 5 points. Radiologic assessment showed preservation of cervical lordosis (C2–7 Cobb angle: 14.9° → 15.0°) and range of motion (39.0° → 37.6°) without instability. These data substantiate the reproducibility and effectiveness of this minimally invasive technique beyond the current 2 cases and support its role as a safe alternative to conventional laminoplasty for patients requiring central decompression.

Furthermore, both of our present patients demonstrated quantifiable improvement in validated patient reported outcome measures, including the mJOA and SF-36 scores, confirming the functional benefit and safety of the muscle-preserving interspinous process approach.

An important technical consideration is that this method requires partial incision of the interspinous ligament to access the interspinous space; however, the spinous processes themselves and the posterior tension band are preserved. The interspinalis and semispinalis cervicis remain intact, and the split spinous processes are re-approximated with absorbable sutures after decompression. Although the spinous processes are not clearly visualized on postoperative MRI, this finding reflects the limited midline exposure corridor rather than bony resection. Postoperative flexion–extension radiographs in the first case demonstrated no dynamic instability (<2-mm translation and <5° angulation), and follow-up MRI confirmed maintained cervical alignment without progressive kyphosis.

Recently, various endoscopic approaches have been increasingly applied to patients with cervical spondylotic myelopathy, including endoscopic laminoplasty and laminectomy [21-23]. While these techniques offer the potential for less tissue disruption, concerns remain regarding their safety in cases of severe spinal cord compression. In particular, the potential risk of cord injury related to water pressure during endoscopic procedures has not yet been fully validated [24,25]. In this context, the muscle-preserving interspinous process approach may serve as a valuable alternative, providing direct decompression under microscopic visualization while minimizing disruption of the posterior musculature.

When compared with other minimally invasive approaches—such as the tubular access technique described by Ross and Ross [26]—the interspinous process approach offers distinct trade-offs. The tubular approach provides a wider working corridor and allows foraminotomy if needed, without disturbing the interspinous ligament. In contrast, the interspinous approach minimizes paraspinal muscle detachment, avoids off-midline dissection, and maintains both midline anatomy and spinous process anchor points. Therefore, the 2 techniques may be considered complementary options depending on the pathology and surgeon’s experience.

Nevertheless, certain limitations must be acknowledged. The operative corridor is narrow and requires experience with microscopic techniques, which may limit its applicability in severe central stenosis or cases requiring extensive foraminal decompression [11,13]. Furthermore, this report is restricted to 2 cases, and broader validation through larger series is necessary.

In summary, this technique enables safe and effective central decompression under direct microscopic visualization, while minimizing injury to the posterior musculature and maintaining the integrity of the posterior tension band. These findings, together with the accompanying surgical videos, may help refine current minimally invasive posterior cervical strategies and serve as an educational reference for spine surgeons.

The posterior cervical muscle-preserving interspinous process approach provides safe and effective central decompression while preserving the posterior musculature and spinous process anchor points. In our 2 cases, this technique resulted in satisfactory neurological recovery, minimal axial pain, and maintained cervical alignment without complications. This minimally invasive strategy may serve as a valuable alternative to conventional laminectomy or laminoplasty in selected patients with cervical spondylotic myelopathy or multilevel stenosis without significant foraminal involvement. The accompanying surgical videos illustrate the technical details and may serve as an educational reference for surgeons considering adoption of this approach in clinical practice.