多節段頸椎椎版成形術之最佳化：動態選擇策略、椎版切除與椎板成形比較及生物力學評估

Abstract

頸椎椎板開窗成形術是治療多節段頸椎脊髓病（cervical spondylotic myelopathy, CSM）的一種成熟、同時可以保留頸椎活動度的手術，其傳統臨床適應症包括：側位 X 光頸椎結構呈現中立或輕度前凸（小於10度）的病人，抑或是K-line陽性之患者。近端節段則常習慣性採用 C3–6 椎板成形，認為維持後位連續骨性結構本質上較佳，對穩定度之維持較理想。這些作法在很大程度上忽略了矢狀面動態後曲程度、近端伸肌與後張力帶在維持彎矩平衡中的關鍵角色，以及不同 C3 處理方式對負荷分佈的實際生物力學影響。 本論文提出一套以力學邏輯為核心的整合框架，將病人篩選（static + dynamic）、C3 手術策略，以及負荷傳遞/生物力學驗證串聯考量。研究包含三個部分：第一部分以系統性文獻回顧與統合分析，比較多節段椎板成形構築中 C3 椎板成形（C3-LP）與 C3 椎板切除（C3-LM）在功能性結果（特別是頸椎活動度與術後頸部疼痛）之差異；第二部分以單一中心前瞻性世代研究，在標準化手術與周術期照護流程下，比較傳統 C3–6 椎板成形與「C3 椎板切除合併 C4–6 椎板成形」之臨床與影像結果，並預先定義「次佳對位族群」（K-line 陰性與/或術前 C2–7 後凸 >10°）以檢驗傳統靜態指標之外的臨床效益；第三部分以有限元素分析（finite element modeling, FEM）探討不同 C3 策略對整體活動度、矢狀面彎矩平衡相關參數，以及鄰近節段椎間盤內壓（intradiscal pressure, IDP）等負荷指標之影響，並與論文中提出的 M=F∙d、θ=(M∙L)⁄EI、σ=F⁄A 等力學關係相互印證。 研究結果支持在排除明確固定畸形的前提下，若以「posterior drift 的力學可行性」作為篩選核心、同時納入動態伸展能力評估，椎板成形適應症具有可控的向外延伸空間。次佳對位患者接受椎板成形後仍可獲得顯著神經功能與失能改善，並呈現頸椎曲度朝生理前凸方向回復的趨勢，符合「可恢復前凸（recoverable lordosis）」概念。在此更精準的適應症定義下，C3 策略對整體手術表現具有關鍵影響：前瞻性結果顯示，相較於傳統 C3–6 椎板成形，採行 C3 laminectomy + C4–6 laminoplasty 可觀察到較短手術時間、較佳前凸恢復與較少早期併發症，而短期 PROMs（例如 mJOA、NDI）改善幅度與傳統術式相當。此外，術前伸展活動度（ext ROM）呈現明確的臨床意義：ext ROM 與術前神經功能狀態呈正向關聯，且 ext ROM 較佳者術後較能維持矢狀排列、progressive kyphosis 風險較低；但 ext ROM 並不顯著影響術後改善幅度（ΔmJOA、ΔNDI）。FEM 結果提供機轉支持：以矢狀面彎矩平衡觀點解釋術後前凸維持時，ext ROM 可被視為節段剛性與「伸展儲備（extension reserve）」的綜合；同時 FEM 也指出，採行 C3 椎板切除可能導致鄰近節段負荷與 IDP 輕度上升，提示需以長期追蹤監測鄰近節段退化風險，而非否定其短期臨床優勢。 綜合而言，本論文將多節段椎板成形由「僅限理想靜態評估」的保守策略，轉化為一套兼顧幾何條件與動態能力的機轉導向決策框架：術前除評估中立位影像與 K-line 外，亦應納入伸展能力（ext ROM）以輔助分層（例如 ext ROM ≥9° 可作為需外部驗證之操作性界值），並在以椎板成形為主要術式時，將近端 C3 椎板切除作為保護伸肌與後方張力帶的優先策略，以期在功能、影像與臨床結果間取得更一致的最佳化。

Multilevel cervical laminoplasty is an established motion-preserving procedure for cervical spondylotic myelopathy (CSM). While laminoplasty provides reliable posterior decompression, its clinical application is often constrained by static radiographic assumptions and an unresolved technical controversy at C3. Traditional patient selection relies heavily on neutral lateral radiographs (e.g., K-line positivity and limited kyphosis), and the proximal construct commonly defaults to continuous C3–6 laminoplasty. These conventions may underemphasize dynamic sagittal capacity, the functional role of the proximal extensor mechanism, and the load-sharing consequences of alternative C3 strategies. This thesis proposes a mechanism-based framework that integrates dynamic patient selection, an evidence-informed solution to the C3 dilemma (laminoplasty vs. laminectomy at C3), and biomechanical validation. Three complementary components were performed. First, a systematic review and meta-analysis compared C3 laminoplasty (C3-LP) versus C3 laminectomy (C3-LM) when incorporated into multilevel open-door laminoplasty constructs, focusing on functional outcomes such as cervical range of motion (ROM) and postoperative axial pain. Second, a single-center prospective cohort study compared conventional C3–6 laminoplasty with a proximal muscle-preserving strategy of C3 laminectomy combined with C4–6 laminoplasty, under standardized surgical and perioperative protocols. A prespecified “suboptimal alignment” subgroup (K-line negative and/or preoperative C2–7 kyphosis >10°) was analyzed to evaluate outcomes beyond traditional static indications. Third, finite element modeling (FEM) examined how alternative C3 strategies influence segmental motion and adjacent-level stresses, including intradiscal pressure (IDP). Collectively, the findings support a cautious expansion of laminoplasty indications when feasibility is assessed by both geometry and dynamic capacity rather than static alignment alone. Patients with suboptimal neutral alignment demonstrated meaningful neurological and disability improvement after laminoplasty-based decompression, consistent with the concept of recoverable lordosis in the absence of fixed deformity. Within this refined indication set, the choice of C3 strategy emerged as a key determinant of construct performance. In the prospective cohort, C3-LM + C4–6 laminoplasty was associated with shorter operative time, greater restoration of cervical lordosis, and fewer early complications, while postoperative improvements in patient-reported outcomes (e.g., mJOA and NDI) were comparable between strategies at short-term follow-up. Preoperative extension range of motion (ext ROM) functioned as a clinically interpretable baseline marker and alignment predictor: greater ext ROM correlated with better preoperative neurological status and a lower tendency toward postoperative progressive kyphosis, yet did not materially influence the magnitude of postoperative improvement in PROMs. The FEM analysis provides mechanistic validation for these observations. Framing sagittal balance as the interaction of bending moments (M=F∙d), angular displacement (θ=(M∙L)⁄EI), and stress distribution (σ=F⁄A), the models show that C3-LM configurations enhance physiological motion and alignment through extensor preservation but modestly increase adjacent-segment stresses in a predictable pattern. This trade-off is consistent with the clinical data and emphasizes the need for long-term surveillance rather than undermining the strategy. Taken together, this thesis reframes multilevel laminoplasty as a mechanism-informed strategy guided by dynamic extension capacity and proximal extensor preservation. Incorporating ext ROM into preoperative assessment (with an operational threshold such as ≥9° to be externally validated) and adopting proximally muscle-preserving C3 laminectomy when laminoplasty is selected offers a focused, evidence-based approach to the long-standing C3 dilemma while balancing short-term benefits against predictable biomechanical trade-offs.