Scholarlink

Objective <p>At present, cage subsidence is the most prevalent hardware-related complication of anterior cervical discectomy and fusion (ACDF), which may affect postoperative spinal biomechanics and alignments. Presently, there is no established nomogram model for predicting subsidence based on preoperative parameters. Hence, our aim was to develop nomograms for predicting cage subsidence after ACDF.</p> Methods <p>The clinical data of 476 patients who underwent ACDF at our hospital from January 1, 2014, to June 30, 2023, were retrospectively analyzed. A nomogram prediction model was constructed using 238 patients in the training cohort, which was divided into the subsidence group and the non-subsidence group. Various parameters were recorded, such as age, gender, body mass index, smoking and drinking history, history of hypertension and diabetes, operation time, intraoperative blood loss, bone mineral density, visual analogue scale (VAS) score for arm and neck, Japanese Orthopaedic Association (JOA) score, Neck Disability Index (NDI), Hounsfield Units (HU) of upper and lower vertebral bodies, Modic changes, cervical lordosis, intervertebral segmental height (ISH), functional spinal unit angle (FSUA), range of motion (ROM) for C2-C7, as well as the sagittal vertical axis (SVA) and T1 slope (T1S). Multivariate logistic regression was used to identify independent predictors for developing nomograms. Finally, the predictive values of this model were validated in an external verification cohort consisting of another 238 patients.</p> Results <p>In the training cohort, 23.95% (57 of 238 patients) had cage subsidence during an average follow-up of 23.31 ± 6.14&#xa0;months. Multivariate logistic regression analysis results showed that bone mineral density (OR = 0.499, 95% CI 0.356–0.677), upper HU value (OR = 0.979, 95% CI 0.971–0.986), lower HU value (OR = 0.989, 95% CI 0.982–0.996), ISH (OR = 0.504, 95% CI 0.284–0.856), and T1 slope (OR = 0.877, 95% CI 0.821–0.933) were determined as independent risk factors for postoperative cage subsidence. The predictive model exhibited excellent performance in both the training cohort (AUC = 0.885, 95% CI 0.835–0.935) and the validation cohort (AUC = 0.914, 95% CI 0.873–0.955).</p> Conclusion <p>Our predictive model is set to become a valuable and convenient tool for clinical decision-making, enabling early prediction and timely intervention for cage subsidence after ACDF. Future prospective studies could incorporate our model in different clinical scenarios, thus reducing the burden on patients and healthcare resources.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development and validation of a nomogram prediction model for assessing cage subsidence risk following anterior cervical discectomy and fusion

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月2日

查看全文
  • Dongmei Zhao,
  • Xiaojie Sun

摘要

Objective

At present, cage subsidence is the most prevalent hardware-related complication of anterior cervical discectomy and fusion (ACDF), which may affect postoperative spinal biomechanics and alignments. Presently, there is no established nomogram model for predicting subsidence based on preoperative parameters. Hence, our aim was to develop nomograms for predicting cage subsidence after ACDF.

Methods

The clinical data of 476 patients who underwent ACDF at our hospital from January 1, 2014, to June 30, 2023, were retrospectively analyzed. A nomogram prediction model was constructed using 238 patients in the training cohort, which was divided into the subsidence group and the non-subsidence group. Various parameters were recorded, such as age, gender, body mass index, smoking and drinking history, history of hypertension and diabetes, operation time, intraoperative blood loss, bone mineral density, visual analogue scale (VAS) score for arm and neck, Japanese Orthopaedic Association (JOA) score, Neck Disability Index (NDI), Hounsfield Units (HU) of upper and lower vertebral bodies, Modic changes, cervical lordosis, intervertebral segmental height (ISH), functional spinal unit angle (FSUA), range of motion (ROM) for C2-C7, as well as the sagittal vertical axis (SVA) and T1 slope (T1S). Multivariate logistic regression was used to identify independent predictors for developing nomograms. Finally, the predictive values of this model were validated in an external verification cohort consisting of another 238 patients.

Results

In the training cohort, 23.95% (57 of 238 patients) had cage subsidence during an average follow-up of 23.31 ± 6.14 months. Multivariate logistic regression analysis results showed that bone mineral density (OR = 0.499, 95% CI 0.356–0.677), upper HU value (OR = 0.979, 95% CI 0.971–0.986), lower HU value (OR = 0.989, 95% CI 0.982–0.996), ISH (OR = 0.504, 95% CI 0.284–0.856), and T1 slope (OR = 0.877, 95% CI 0.821–0.933) were determined as independent risk factors for postoperative cage subsidence. The predictive model exhibited excellent performance in both the training cohort (AUC = 0.885, 95% CI 0.835–0.935) and the validation cohort (AUC = 0.914, 95% CI 0.873–0.955).

Conclusion

Our predictive model is set to become a valuable and convenient tool for clinical decision-making, enabling early prediction and timely intervention for cage subsidence after ACDF. Future prospective studies could incorporate our model in different clinical scenarios, thus reducing the burden on patients and healthcare resources.