Pain and Paresthesia Patterns in Degenerative Cervical Myelopathy: Neuropathic Sensory Features and Postoperative Transitions in a Prospective Cohort.

Oshima, Yasushi; Sumitani, Masahiko; Matsubayashi, Yoshitaka; Nakamoto, Hideki; Nakarai, Hiroyuki; Kodama, Hiroyasu; Shirokoshi, Tomohiko; Kumanomido, Yudai et al. · Spine J · 2026

prospective_cohort · Level II

Where this comes from

Abstract

Pain and paresthesia are common in degenerative cervical myelopathy (DCM), yet their relationship and postoperative course remain uncertain. To categorize patients undergoing surgery for DCM using preoperative pain and paresthesia severity, characterize neuropathic sensory signatures, and quantify postoperative transitions among groups. Prospective cohort study at a single academic institution. Of 408 eligible surgical DCM cases, 306 with complete preoperative and 1-year postoperative datasets were analyzed. NRS for hand pain and hand paresthesia; EuroQol 5-Dimension (EQ-5D); the Japanese Orthopaedic Association Cervical Myelopathy Evaluation Questionnaire (JOACMEQ); Neuropathic Pain Symptom Inventory (NPSI); and painDETECT questionnaire (PDQ). Using an NRS threshold of 7 to define "high" symptoms, patients were classified as Group A (low pain/low paresthesia), Group B (low pain/high paresthesia), and Group C (high pain/high paresthesia). High pain/low paresthesia was rare (n=2, 0.6%) and excluded. Group differences were assessed using parametric/nonparametric tests with multiplicity-adjusted post hoc comparisons. Associations between NRS pain/paresthesia and NPSI items were assessed by Spearman correlation. Postoperative transitions were summarized using a transition matrix and tested for asymmetry. Preoperatively, 159 (51.9%) patients were Group A, 85 (27.8%) Group B, and 60 (19.6%) Group C. Groups B and C had worse preoperative EQ-5D and JOACMEQ scores than Group A, whereas Groups B and C did not differ. At 1 year, EQ-5D and JOACMEQ improved and converged across preoperative groups, despite residual pain/paresthesia in a subset. Paresthesia NRS correlated most closely with NPSI tingling, whereas pain NRS showed higher scores across multiple NPSI domains. Groups B and C shared prominent tingling, but Group C showed broader NPSI elevation and higher PDQ scores. Postoperative transition analysis showed net movement toward Group A, with Group B reaching Group A more frequently than Group C. We investigated pain-paresthesia patterns in DCM patients. Among patients with severe paresthesia, concomitant severe pain was associated with broader neuropathic sensory signatures. Although EQ-5D and JOACMEQ converged after surgery, the likelihood of reaching the low pain/low paresthesia state differed between groups.