Thoracic paravertebral block (TPVB) is a clinically valuable regional analgesia technique, widely applied for managing pain following thoracic and abdominal surgeries, as well as certain chronic pain conditions (1–6). TPVB involves the targeted delivery of drug solution to the paravertebral space (PVS), blocking the branching thoracic spinal nerves and sympathetic nerve fibers to produce unilateral analgesia (7, 8). This approach offers potential benefits over systemic opioid analgesia by minimizing systemic side effects and avoiding the risks associated with neuraxial techniques (9–12).

There are multiple possible techniques when performing a TPVB, including landmark-based techniques using pressure measurement (9), electrostimulation guidance (13, 14), as well as live ultrasound guidance (15). The increasing use of ultrasound imaging has facilitated the development of various ultrasound-guided TPVB approaches (15). Ultrasound-guided TPVB can be performed using either transverse or sagittal approaches, depending on transducer orientation. The transverse approach requires a longer needle path, making it suitable for in-plane advancement with good visualization, but the needle points toward the epidural space, raises concerns about a greater risk of epidural spread. The sagittal approach involves a shorter, steeper trajectory, which makes in-plane needle tip visualization more difficult but facilitates out-of-plane injections. From a learning perspective, placing the probe about 2 cm lateral to the spinous process in the sagittal plane and using an out-of-plane technique is generally easier for beginners compared with the transverse in-plane technique. These techniques differ mainly in needle trajectory, visualization, and safety considerations (15). However, the difference of injectate distribution patterns between transverse and sagittal approaches is still unclear.

Currently, the choice of TPVB approach with ultrasound guidance mainly depends on the tradition of institution, and preference of performer, rather than the specific spread pattern. Actually, understanding the spread pattern of local anesthetic (LA) following a particular ultrasound-guided TPVB approach is essential. Knowledge of the spread pattern allows for informed selection of the injection site and prediction of the resulting sensory distribution, leading to improved analgesic efficacy. While several studies have investigated the injectate spread within the PVS and nearby anatomical structures, as well as the distribution of sensory block after a specific ultrasound-guided TPVB approach (15), direct comparisons between different approaches with respect to injectate distribution patterns remain limited. Therefore, this study utilized a fresh cadaver model to compare the contrast dye distribution following ultrasound-guided TPVB at T6-7 performed via transverse in-plane (TI) vs. sagittal in-plane (SI) approaches. These two approaches are commonly used. This observational anatomical study aimed to quantify differences in the number of PVS stained, distance of intercostal spread, and sympathetic chain stained. By providing a detailed anatomical comparison of these two common TPVB techniques, this study sought to inform clinical practice, guide technique selection, and ultimately contribute to improved patient outcomes.

The procedures used in this study adhere to the tenets of the Declaration of Helsinki. Anatomical bodies were obtained from Beijing Voluntary Body Donation Registration and Acceptance Station affiliated with Peking Union Medical College. The donors had signed an informed consent for body donation during their lifetime, which explicitly stated that the body would be used for medical education and research. The post-mortem interval before preservation was less than 2 days. We adhered to the ChAracteristics of Cadaver Training and sUrgical Studies (CACTUS) guideline to conduct and report this study (https://www.equator-network.org/reporting-guidelines/reporting-characteristics-of-cadaver-training-and-surgical-studies-the-cactus-guidelines/). CACTUS checklist can be found at online resource (Supplementary material). 5 fresh un-embalmed human cadavers were studied. Prior to the study, the cadavers were preserved in an environment at −4 °C. The duration of preservation was within 24 h. None of cadavers exhibited thoracic or dorsal deformities, traumatic lesions, or surgical scars. Bilateral ultrasound-guided paravertebral injections were conducted at T6-7 in each cadaver in prone position, while in our center this position is more frequently used during TPVB procedure. TI approach was utilized on left side, and SI approach on the right side. All injections were performed by a single experienced regional anesthesia specialist. Subsequently, a detailed layer-by-layer dissection of the paravertebral region from T1 to T12 was conducted by a dedicated anatomist assisted by two trained technicians.

A total of 5 fresh un-embalmed human cadavers (two females and three males) were included in the study. The mean age at death was 85.5 yr (range 82–101). Each cadaver received bilateral paravertebral injections at T6-7, with the TI approach performed on the left side and the SI approach on the right. The PVS were successfully identified and reached in all cadavers. Contrast dye was present in PVS following all injections. An illustrative example of contrast dye spread is shown in Figures 1C, D.

This cadaveric study aimed to compare the spread pattern of ultrasound-guided TPVB at T6-7 using 20 mL of 0.02% methylene blue via TI and SI approaches. While the TI approach resulted in a mean cranio-caudal spread encompassed 3 (3,3) vertebral levels, 55.1 (30.1, 76.0) cm² of intercostal spaces, and 80% of sympathetic chain staining, compared to 2 (1,2) vertebral levels, 38.3 (7.8, 82.6) cm², and 40% with the SI approach, no statistical differences were found. Importantly, regardless of the approach, dye distribution was predominantly cephalad, and a reliably staining of PVS and intercostal space at the injection segment and the immediately adjacent cephalad segment was indicated. Furthermore, we observed a positive correlation between the extent of paravertebral spread and the likelihood of sympathetic chain staining.

The 100% success rate in achieving PVB with two different approaches underscore the feasibility and reliability of ultrasound-guided TPVB. Previous studies have reported variable vertebral level staining after ultrasound-guided PVB with sagittal in-plane approach (2-6 levels) and the transverse in-plane approach (2-10 levels) (20, 21). Ruscio et al. also compared contrast dye spread between the transverse and sagittal approaches, reporting mean vertebral level staining of 3 (range 1-5) and 2 (range 1-4), respectively (22). In conjunction with our findings, these results suggested that the TI approach may result in a slightly more segments of vertebral levels stained compared to the SI approach. Although volunteer studies indicated that the spread of LA in the PVS on MRI scan did not perfectly match the area of sensory block (23), the distribution of injectate within the PVS provided an important anatomical basis for understanding the mechanism of TPVB. More research is warranted to investigate the relationship between LA spread in the PVS and the distribution of sensory blockade in patients.

We found that PVS of injection segment and the immediately adjacent cephalad segment were consistently and reliably stained by contrast dye regardless of injection approach. In addition, the dye distributed significantly more in a cephalad segment compared to a caudal segment. Those results indicated a predominantly cephalad spread of the injectate, irrespective of the approach. This predominant cephalad spread may be attributed to the anatomical constraints imposed by the superior costotransverse ligaments (SCTL), which form the posterior border of the PVS. The SCTL originates from the superior border of the rib neck and extends supero-laterally to attach to the inferior border of the transverse process of the vertebra immediately above (8, 24). This anatomical arrangement results in a larger cephalad space within the PVS compared to the caudal space at the same segmental level, potentially facilitating preferential cephalad injectate diffusion. However, a study by Marhofer et al. (23) in volunteers using ultrasound-guided transverse out-of-plane TPVB technique, with MRI scanning after injection, found a predominantly caudal distribution of LA. We acknowledged that the out-of-plane technique and the use of living volunteers in that study was different from our study. Employing the out-of-plane technique, the needle tip visualization is poor, which could potentially result in a different puncture location. Furthermore, the cadaveric model intrinsically differs from the living patient. These differences include variations in compartmental and tissue pressures, as well as the absence of dynamic uptake of injected solutions. These factors, in aggregate, may contribute to discrepancies between two studies. Nevertheless, it also suggests that variations in technique can influence the pattern of injectate spread. Further studies are needed to validate these observations. Given the observed cephalad spread in our study, selecting a TPVB injection site caudal to the intercostal nerve innervating the surgical incision may be a prudent clinical strategy.

In the present study, we observed the diffusion of dye into the space between the parietal pleura and the posterolateral chest wall following ultrasound-guided TPVB. While with ultrasound guidance the correct needle placement of PVB was confirmed by anterior displacement of the parietal pleura. Consequently, the diffusion of dye beneath the parietal pleura indicates a successful PVB. This observation is consistent with findings reported in previous publications (21, 25–27). The intercostal nerves, after exiting the intervertebral foramen, travel inferior to the pleura before they travel within the plane between the internal intercostal muscle and the innermost intercostal muscle. Therefore, after ultrasound-guided TPVB, LA diffusion beneath the pleura effectively blocks the proximal segment of the intercostal nerve. In this study, the distance and area of intercostal spread was used as an indirect measure of the volume of LA pooling around the intercostal nerve. It is generally accepted that a larger volume of pooled LA correlates with a longer duration of action.

In our study, there were no instances where intercostal spread occurred in the absence of PVS spread. Specifically, when the PVS was stained, the intercostal space was also stained; conversely, when the PVS was not stained, neither was the intercostal space. However, we did observe that intercostal spread sometimes occurred via diffusion from an adjacent intercostal space, rather than directly from the PVB injection in the same intercostal level. This suggested that LA diffusion beneath the parietal pleura may be one of the mechanisms contributing to intercostal nerve blockade, although the direct diffusion from corresponding PVS plays a crucial role.

We also observed a high incidence of sympathetic chain staining, averaging 60% (80% with the TI approach and 40% with the SI approach). This finding was consistent with prior research (21, 28). Furthermore, our study suggested that a more extensive dye spread in PVS correlated with a higher odd of sympathetic chain staining, regardless of the injection approach. While sympathetic blockade may contribute to visceral pain control compared to peripheral intercostal nerve block, it can also lead to side effects such as hypotension and bradycardia (29). Therefore, clinicians should be cognizant of the potential for sympathetic blockade following TPVB and be prepared to manage associated side effects.

There were several limitations in the present study. First, the small sample size in this cadaveric study may have limited our statistical power to detect subtle, yet potentially clinically relevant differences. The scarcity of donated cadaveric specimens presented a significant challenge for larger sample size. Second, the cadaver model inherently differs from the living patient due to variations in compartmental and tissue pressures, and the absence of dynamic uptake of injections. These factors may limit the extent and pattern of injectate distribution compared to living human subjects. The injectate spread observed in cadavers may not fully replicate the situation in living patients. Thus, further clinical studies that enroll volunteers or patients are still required to evaluate the similarities and differences between the two injection approaches with respect to clinical efficacy. We noticed that positioning may influence the spread of injectate after TPVB in living patients (30). In the prone position, the injectate could even reach the anterior thorax. Owing to the different factors between cadaver model and living patient, our study presented the spread pattern without further redistribution. It is also worth noting that previous publications investigating injectate distribution after TPVB with cadaveric model have similarly utilized a prone position (21, 22). Next, we were unable to assess the potential spread of the dye to the epidural space due to experimental limitation. In clinical practice TPVB is exclusively performed for unilateral analgesia purpose. We believed that the distribution pattern on the blocked side is of primary concern. Finally, methylene blue was used as contrast dye because it is inexpensive, universally obtainable, almost non-toxic, and provides high-contrast visualization, which was frequently used in anatomic studies (26, 31, 32). We acknowledged that anatomic dissection may cause mechanical displacement of dye and could overestimate the true spread—a limitation common to all cadaveric staining investigations using anatomic dissection technique. Nevertheless, direct anatomical exposure remains an important method for defining detailed perineural and fascial distribution. Future work using CT or MRI with iodinated or gadolinium-based contrast could corroborate our findings and quantify spread in vivo, although such approaches may not resolve perineural and fascial anatomical details. Despite these limitations, our study provided novel evidence contributing to a better understanding of the differences in spread patterns between two distinct ultrasound-guided TPVB techniques.

This cadaveric study demonstrated the ability to reliably target the thoracic PVS with ultrasound guidance and a predominantly cephalad spread of the injectate, irrespective of the injection approach (TI or SI). Meanwhile, a more segments of PVS stained, a higher possibility of sympathetic chain stained. While the TI approach tended to result in a slightly greater number of stained PVS segments, a longer intercostal spread distance, and a higher incidence of sympathetic chain staining compared to the SI approach, these differences were not statistically significant. These anatomical findings may contribute to optimizing clinical practice and potentially improving patient outcomes.