Does paravertebral block require access to the paravertebral space?

We read the article by Yoshida et al. on the effects of ropivacaine concentration on the spread of sensory block produced by continuous thoracic paravertebral block 1. Both traditional landmark and ultrasound techniques for thoracic paravertebral block (PVB) assume that the needle tip needs to lie anterior to the superior costotransverse ligament (SCTL) for block success. However, we have achieved successful PVB with deliberate injection posterior to the SCTL, and frequently observe pleural displacement with the needle tip and local anesthetic injection clearly posterior to the SCTL. Surgical experience provides a rationale for this occurrence. The paravertebral area is exposed for many spinal surgical procedures. During procedures requiring the removal of ligaments involving the rib and/or transverse process, the ligaments in the immediate paravertebral area are often not visualised as distinct entities; rather, the proximal ligament complex involving the rib head and transverse process may appear as a single structure. Additionally, there can be variability in the density and the continuity of these ligaments such that they may not form a definitive tissue layer. Consequently, the fascial planes in the paravertebral region may not be truly compartmentalised, resulting in locally-injected anesthetic travelling through septations and fenestrations in the ligamentous tissues. Unpredictable spread of PVB in previous studies could be explained by injections posterior to the SCTL resulting in a single level, cloud-like spread, and injections anterior to the SCTL resulting in a multilevel spread. We use dynamic ultrasound with a parasagittal scan, in-plane needle insertion cephalad to the probe, and injection posterior to the SCTL. In an unembalmed cadaver, we noted that injections of 5 ml of methylene blue 1% posterior to the SCTL from T2 to T7 bilaterally (except injections anterior to SCTL on left T4+5) stained the sympathetic chain (T1–T7 on the left, and T4–T8 on the right), and the intercostal nerves T1–T7, and their nerve roots in the paravertebral spaces, bilaterally. Our ultrasound observations and cadaver investigation suggest that the paravertebral space is not a true anatomical compartment and that the SCTL is not a barrier to diffusion of injectate as previously thought. We believe that successful PVB can be achieved with the needle tip further distant from the pleura than currently described, with greater potential for patient safety. Further studies will be required to confirm our preliminary work.