\[Background\] It has been reported that approximately 85% of sciatica cases result from nerve root compression secondary to lumbar disc herniation or lumbar spinal stenosis, a clinical entity defined as radicular pain that manifests as radiating pain involving the buttocks and lower extremities. For patients with an inadequate response to conservative management, surgical decompression of the compressed nerve root serves as a conventional and highly effective therapeutic option. Furthermore, if the pain generator can be confirmed to arise from a single nerve root, targeted single-level decompression of that individual nerve root is generally sufficient for pain relief. However, limitations of conventional imaging examinations coupled with overlapping innervation territories of multiple lumbosacral nerve roots often make accurate identification of the symptomatic compressed nerve root challenging in certain patient populations. Misidentification of the pathological nerve root may consequently lead to surgical failure. To address this clinical dilemma, diagnostic selective nerve root block (SNRB) is regarded as the most valuable adjunct modality. In routine clinical practice, complete pain remission following SNRB indicates that the nerve root infiltrated by local anesthetics is the primary pain source.
Regrettably, prior relevant studies have demonstrated that conventional SNRB exhibits relatively low overall diagnostic accuracy, failing to achieve satisfactory sensitivity and specificity simultaneously. Uncontrolled diffusion of injectable agents is recognized as the predominant contributor to such poor diagnostic performance. Specifically, unintended anesthetic contamination of adjacent unaffected nerve roots may trigger false-positive outcomes, while insufficient delivery of local anesthetics to the targeted nerve root can result in false-negative findings. At present, no techniques or strategies for precisely regulating drug distribution during SNRB procedures have been reported in the existing literature.
Interestingly, during the nerve root fluorography in some patients in our team, when linear striation opacities are visualized within the nerve root, which indicates occurrence of intraperineural, the contrast medium typically diffuses only inside and around the targeted nerve root. This phenomenon can also be observed in the illustrative images from a previous study on therapeutic SNRB, yet it seems to have attracted little attention from researchers. Additionally, the findings of this study demonstrated that intraperineural injection occurred in approximately 30% of patients undergoing therapeutic SNRB with accidental intraperineural injection, and no cases of neurological injury were documented during follow-up. Therefore, it is reasonable to hypothesize that intraperineural injection is a safe, feasible approach that enables precise distribution of agents to the targeted nerve root.
Using the postoperative efficacy of single-segment single-nerve-root decompression as the gold standard for identifying the compressed nerve root, the present study aimed to investigate the diagnostic accuracy, safety, and technical feasibility of selective intraperineural nerve root block (SINRB) in patients with radicular pain. A double-blind approach was implemented, with participants and assessors masked to the nerve root status (responsible vs. non-responsible) during evaluations. Meanwhile, to improve the success rate of intraperineural injection, to the best of our knowledge, the present study is the first to adopt three-dimensional computed tomography multiplanar volume reconstruction (3D-CT MPVR) imaging to visualize the anatomical course of the nerve root within the intervertebral foramen.
\[Sample Size Estimation\] To ensure adequate statistical power for accurately evaluating the diagnostic accuracy of selective intraperineural nerve root block (SINRB) in identifying the responsible compressed nerve root (RCNR) in patients with radicular pain, sample size calculation was performed based on diagnostic test design principles, incorporating the following key parameters: (from preliminary pilot data) an expected sensitivity of 95%, specificity of 96%, 95% confidence level, ±5% margin of error, and a 10% allowance for potential missing data or patient dropout. Using the single-proportion estimation method, the required numbers of positive and negative events were calculated separately for sensitivity and specificity. The sample size required for sensitivity was 80 cases, and for specificity was 66 cases. Accordingly, we plan to enroll 66 patients with single-level lesions (anticipated to yield one positive and one negative result each) and 14 patients with multi-level lesions (anticipated to yield one positive result each). This sample size sufficiently meets the statistical power requirements for the primary study objective-evaluating the sensitivity and specificity of SINRB for RCNR identification-ensuring the scientific rigor, stability, and generalizability of the study findings.
