Brief technical report
Reg Anesth Pain Med: first published as 10.1136/rapm-2019-100896 on 2 November 2019. Downloaded from http://rapm.bmj.com/ on March 3, 2022 at Post Graduate Institute of Medical
►► Additional material is
published online only. To view
please visit the journal online
(http://​dx.​doi.o​ rg/​10.​1136/​
rapm-​2019-​100896).
1 of dye spread. In particular, we investigated whether
Anesthesiology, HELIOS
Medical Center, Schwerin,
Germany
2
Institute of Anatomy, University
of Rostock, Rostock, Germany
Correspondence to
Dr Ronald Seidel,
Anesthesiology, HELIOS Medical
Center, Schwerin 19049,
Germany;
​ronald-​seidel@​t-​online.d​ e
Received 3 August 2019
Revised 22 September 2019
Accepted 6 October 2019
Published Online First
2 November 2019
© American Society of Regional
Anesthesia & Pain Medicine
2020. No commercial re-­use.
See rights and permissions.
Published by BMJ.
To cite: Seidel R,
Wree A, Schulze M.
Reg Anesth Pain Med
2020;45:102–106.
102    Seidel R, et al. Reg Anesth Pain Med 2020;45:102–106. doi:10.1136/rapm-2019-100896
Thoracic-­paravertebral blocks: comparative
anatomical study with different injection techniques
and volumes
Ronald Seidel  ‍ ‍,1 Andreas Wree,2 Marko Schulze2
ABSTRACT
Background and objectives  We hypothesized that
different injection techniques and volumes in thoracic-­
paravertebral blocks (TPVB) lead to different patterns
an alternating injection technique leads to complete
staining of all adjacent intercostal nerves.
Methods  This comparative anatomical investigation
was performed using 10 or 20 mL of dye (Alcian Blue)
in 10 unfixed donor cadavers (54 injections) that were
designated for education or research purposes.
Results  In landmark-­guided TPVB, the thoracic-­
paravertebral space (TPVS) was either not stained at
all (spread of dye in the paraspinal muscles, n=3) or
the dye was predominantly found in the epidural space
(n=3). In ultrasound-­guided TPVB, the TPVS was correctly
identified in all cases (n=48). The sympathetic trunk was
stained in 84.6% of injections (multi-­injection technique:
100%), independent of injection technique and volume.
The epidural space was stained more frequently
(p≤0.001) if both the puncture site (sagittal transducer
position) and guidance of the needle were more medial
(77.8%). Finally, a higher injection volume (20 vs 10 mL)
resulted in a higher number of stained intercostal nerves
(p=0.04).
Conclusion  For ultrasound-­guided techniques, a higher
injection volume resulted in a larger number of stained
intercostal nerves. Staining of the sympathetic trunk
was independent of the injection technique. Epidural
spread was observed significantly less frequently if the
injection was lateral (transducer transversal) or with a
strictly cranial injection direction (transducer sagittal).
Landmark-­guided injections reliably achieved the TPVS
(and the epidural space) only after a needle advance of
2.5 cm after initial contact with the transverse process.
Introduction
A thoracic-­ paravertebral block (TPVB) can be
performed to provide analgesia and/or anesthesia for
different surgeries of the upper extremities (inter-
costobrachial nerve T2-3), thorax and abdomen.1 2
There are landmark-­based and ultrasound-­guided
techniques (transversal transducer position:
in-­plane approach from lateral to medial, sagittal
transducer position: in-­plane approach from caudal
to cranial, out-­of-­plane approach) (online supple-
mentary videos 1-3). Depression of the pleural line
during the injection and the visualized spread of
the local anesthetic in the thoracic-­ paravertebral
space (TPVS) indicates correct positioning of the
needle.1 2 A detailed overview of ultrasound-­guided
techniques has been provided by Krediet et al.2–10
The TPVS communicates with the intercostal
(laterally, apex of the TPVS) and the epidural space
(ES) (medially, base of the TPVS). The adipose
tissue at the anteromedial corner of the TPVS is
continuous over all thoracic levels, which allows
a craniocaudal spread of injected local anesthetic
(multisegmental spread). The TPVS contains fatty
tissue, the intercostal nerve (ventral ramus), the
dorsal ramus, intercostal vessels, rami communi-
cantes and the sympathetic chain.1 2 Furthermore,
the intercostal nerves anastomosize during their
Education & Research. Protected by copyright.
course to the anterior chest wall so that each “skin
segment” receives branches of at least two (usually
three) spinal nerves.11
The multiple injection technique, in which small
volumes of local anesthetic (5 mL or less) are
injected at several contiguous thoracic levels, is
preferable in terms of the reliable blockade of all
adjacent intercostal nerves.1 2 This procedure is,
however, associated with more patient discomfort
and additional risks, including unintended pleural
punctures. We, therefore, used an alternating injec-
tion technique in which the dye was injected at
every second thoracic level.
We hypothesized that different techniques and
volumes lead to different patterns of dye spread.
Thus far, no comparative studies have been
published regarding this. In particular, we investi-
gated whether an alternating injection technique
leads to complete staining of all adjacent intercostal
nerves.
Methods
The study was approved by the Ethics Committee (A
2016 0083) of the University of Rostock, Germany
and performed in 10 unfixed donor cadavers (race:
Caucasian; gender: 9 male, 1 female; age: mean
79 years, range 66–91  years; body mass index:
range 19–34) who had bequeathed their bodies for
education or research purposes at the Institute of
Anatomy, University of Rostock, Germany.
The cadavers were placed for puncture in a
prone position and with laterally positioned arms.
The injections (10 or 20  mL) were performed
using a Sonoplex 21 G 100 mm cannula (Pajunk,
Geisingen, Germany). The dye used was Alcian Blue
8GX (Sigma-­ Aldrich, St. Louis, Missouri, USA).
The ultrasound device was a Sonosite S-­Nerve with
a 38 mm linear transducer (6–13 MHz) (Sonosite,
Bothell, Washington, USA).

Figure 1  Thoracic-­paravertebral space (sonographic representation).
(A) (top panel): sagittal paravertebral view. (B) (middle panel):
sagittal paravertebral view (power Doppler image). (C) (bottom
panel): transverse paravertebral view. Black arrow, intercostal nerve;
eim, external intercostal muscle; esm, erector spinae muscle; pvs,
paravertebral space; red arrow, mirror artifact; red x, pleural line; tm,
trapezius muscle; tp, transverse process; white arrow, intercostal artery;
white x, superior costotransverse ligament; yellow x, internal intercostal
membrane.
In order to select the correct puncture site, the angulus inferior
of the scapula was palpated at the assumed level of the seventh
thoracic vertebral body. From this point, the puncture site
(thoracic level) of the injections was determined by palpation.
The transverse process (TP) and the superior costotrans-
verse ligament were important sonoanatomical landmarks for
TPVB using sagittal transducer guidance (n=30). The TP and
the internal intercostal membrane were important sonoanatom-
ical landmarks for TPVB using transverse transducer guidance
(n=18) (figure 1 and online supplementary files 1–3).1 2
The puncture depth for landmark-­ guided blocks was first
determined sonographically. The depth of the TP was assumed
to be 5 mm. After contact with the TP, the needle was redirected
cranially. The aim was to advance the needle 1 (n=3) or 2 (n=3)
cm beyond the ventral boundary of the TP into the TPVS.
A total of 48 ultrasound-­ guided and six landmark-­ guided
TPVB were performed. The spread of the dye was investigated by
anatomical preparation and photodocumentation immediately
after administration of the injection. For the purpose of this study,
a complete anatomical dissection included the following steps:
preparation of the dorsal muscle layers (trapezoid, rhomboid,
Seidel R, et al. Reg Anesth Pain Med 2020;45:102–106. doi:10.1136/rapm-2019-100896
Brief technical report
Reg Anesth Pain Med: first published as 10.1136/rapm-2019-100896 on 2 November 2019. Downloaded from http://rapm.bmj.com/ on March 3, 2022 at Post Graduate Institute of Medical
Figure 2  Transducer position sagittal. Injection from caudal to cranial
(in-­plane approach). 10 mL Alcian Blue. Staining of intercostal nerves:
white circle—puncture level, red circle—unstained intercostal nerve,
green circle—stained intercostal nerve. White numbers (5, 7, 9)—level
of intercostal space, black numbers (12)—rib. Right side: dye from
thoracic level T5 to level T11, no unstained segments. Left side: dye from
thoracic level T5 to level T9, one unstained segment (T8). The lack of
Education & Research. Protected by copyright.
staining was confirmed during further preparation after removal of the
ribs.
serratus posterior superior and erector spinae muscles), verifi-
cation of the correct puncture level (palpatory counting of the
exposed ribs) and examination of the intercostal, paravertebral,
epidural and spinal spaces in the area of the thoracic spine.
Statistical analyses were performed using GraphPad PRISM
8.0 (GraphPad Software, San Diego, California, USA). The
Fisher exact test was used to compare categorical variables, and
the Mann-­Whitney U test to compare continuous variables. The
significance level was defined as p≤0.05.
Results
The results and statistical analyses are presented as online supple-
mentary tables 1-3.
Due to the divergent positioning between patients and the
cadavers in the present study, the median difference between the
target and the actual puncture level was one segment (range 0–4
segments) with the actual puncture level always lower than that
determined by external anatomical landmarks (angulus inferior
at level T7).
Its wedge-­shaped expansion causes the puncture needle to
reach the TPVS at different points depending on the injection
technique used. Ultrasound-­guided blocks using the transversal
transducer position (in-­plane technique from lateral) reached the
TPVS near its apex (injection site lateral to the midline: mean
7.7 cm, range 6.5–8.5 cm). Landmark-­guided blocks reached the
TPVS near its base (injection site lateral to the midline: mean
1.9 cm, range 1.5–2.5 cm). For ultrasound-­guided TPVS using
the sagittal transducer position, however, the puncture site was
on average 3.6 cm (range 1.5–5.5 cm) lateral to the midline. This
resulted in significantly different staining patterns (figures 2–4).
Landmark-guided thoracic-paravertebral blocks (TPVB)
Six landmark-­guided TPVB, each using 10 mL of dye, were
performed (puncture levels T5, T7, T9). The depth of the
TP was evaluated sonographically. The aim was to advance
the needle 1 or 2 cm beyond the ventral edge of the TP. The
103
 Brief technical report
Figure 3  Transducer position sagittal-­oblique. Injection from caudal/
lateral to cranial/medial (in-­plane approach). 10 mL Alcian Blue. Staining
of intercostal nerves and epidural space: white rectangle—puncture
level, red circle—unstained intercostal nerve, green circle—stained
intercostal nerve. The lack of staining was confirmed during further
preparation after removal of the ribs.
thickness (dorsoventral extension) of the TP was assumed to be
0.5 cm. After bone contact (TP), the needle was redirected crani-
ally and advanced 1.5 cm (n=3) or 2.5 cm (n=3).
Figure 4  A (Top panel): Landmark-­guided injection (needle tip
2 cm beyond the ventral boundary of the transverse process). Right
side. 10 mL Alcian Blue at levels T5, T7, T9. Predominantly epidural
dye spread. B (Bottom panel): Ultrasound-­guided injection. Left side.
Transducer position transversal, injection from lateral to medial. 20 mL
Alcian Blue per injection. Preparation of the st and the gsn within
the TPVS. gsn, greater splanchnic nerve; st, sympathetic trunk; TPVS,
thoracic-­paravertebral space.
104
Reg Anesth Pain Med: first published as 10.1136/rapm-2019-100896 on 2 November 2019. Downloaded from http://rapm.bmj.com/ on March 3, 2022 at Post Graduate Institute of Medical
The deep injections led to a staining of the TPVS and an exten-
sive bilateral epidural distribution of the dye over the entire
thoracic spine including the upper lumbar segments (figure 4A).
The TPVS, on the other hand, was not impregnated with more
superficial injections (difference in puncture depth: 1 cm). There
was limited staining below the medial part of the erector spinae
muscle (iliocostal and longissimus muscles).
Ultrasound-Guided thoracic-paravertebral blocks (TPVB)
The TPVS could be reliably identified for all ultrasound-­guided
injections (n=48). In particular, no intrapulmonary injections
were observed.
The sympathetic trunk (ST) was prepared for 39 of 48 injec-
tions (figure 4B). Staining within the TPVS was found following
33 injections (84.6%). The result was independent of both injec-
tion volume (10 vs 20 mL, p=0.38) and transducer position
(sagittal vs transversal, p=0.18).
The ES was prepared for 42 of 48 injections. Staining (bilat-
eral) was independent of the transducer position (sagittal vs
transversal; p=1.0), but significantly dependent on the injec-
tion direction. With the sagittal transducer position and medial
needle guidance (n=9), the ES was stained in seven cases
(77.8%), but was only stained in one case (4.7%, p<0.001) with
Education & Research. Protected by copyright.
strictly cranial needle guidance (n=21) (figure 3).
No dye was detected in the spinal space or ventral to the verte-
bral body (possible bilateral spread).
A higher injection volume resulted in significantly more
staining of the intercostal nerves. A series of three injections led
to an average of 5 (range 3–7) stained intercostal nerves in those
injected with 10 mL compared with 7 (range 7–8 mL) in those
injected with 20 mL (p=0.04). The injection volume also affected
the staining of intersegments (adjacent segments between two
injection sites). In the 10 mL group 45% of intersegments were
stained compared with 85.7% in the 20 mL group (p=0.03)
(figure 2). Nevertheless, the segmental spread per puncture level
was systematically underestimated, as segments not directly
targeted may have been stained from both sides (cranial and
caudal) but were only scored once.
Discussion
Landmark-guided TPVB
For landmark-­guided injections, a medial puncture site is recom-
mended as it is here the TPVS reaches its maximum dorsoven-
tral extension and also the risk of pneumothorax is minimal.1
This space should be reliably reached 1 cm beyond the ventral
edge of the TP. This injection site is also close to the interverte-
bral foramina, so there is the risk of spreading into the ES. This
applies to both the injectate and hematoma after unintentional
vascular puncture.
We showed that a 1 cm difference in needle advancement led
to considerable differences in dye spread. At a needle advance-
ment of 1.5 cm beyond the dorsal boundary of the TP, the injec-
tate spread between the paraspinal muscles. The clinical effect
corresponds to that of an erector spinae plane block, where in
addition to the dorsal rami of spinal nerves the TPVS can also be
reached through gaps between the costotransverse ligaments.12–14
This propagation pattern was surprising (the TP forms the dorsal
boundary of the TPVS) and probably due to the redirection of
the puncture needle (prolonged injection route) and the elas-
ticity of the costotransverse ligaments. In contrast, with a needle
advancement of 2.5 cm, the dye was almost exclusively found in
the ES (figure 4A).
Seidel R, et al. Reg Anesth Pain Med 2020;45:102–106. doi:10.1136/rapm-2019-100896

Ultrasound-Guided TPVB
In contrast to landmark-­ guided TPVB, real-­ time ultrasound
guidance enables visualization of the needle tip and depression
of the pleural line during the injection process (restricted in case
of pleural effusion or atelectasis).1 2
In all cases (n=48) the puncture needle was correctly placed
in the TPVS. Studies on both transversal (near the apex of the
TPVS) and sagittal transducer guidance have been published and
report comparable efficacy and side-­effect profiles.2–10 Multi-­
injection techniques do not appear to offer any advantage over
single injections when identical injection volumes are used.15 16
Higher injection volumes (20 vs 10 mL) however, resulted in the
staining of a significantly larger number of intercostal nerves and
less frequently unstained intersegments (between two injection
sites).
In accordance with other studies, we observed cranial dye
spread within the TPVS and also spread into the intercostal or
ES.1–10 Epidural dye spread (always bilateral) was particularly
frequent when both transducer position (sagittal) and needle
guidance were more medial, which is consistent with the results
of the cadaver study by Luyet et al.5 Whether this is a side effect
or an essential mechanism of action of TPVB remains debatable
(figures 3 and 4). An injection technique that leads to bilateral
epidural spread should appear clinically as a successful block, but
with bilateral effect and a relevant rate of hypotensions. This has
not been reported so far.
The ST (greater and lesser splanchnic nerve) was reliably
stained within the TPVS, which clinically corresponds to
sympathicolysis in the abdominal region (figure 4B).
In anatomical and clinical studies published to date, a high
rate of incorrect positioning (intrapulmonary, prevertebral or
epidural) is shown with catheters despite a correct initial needle
position. An epidural misplacement is more likely with trans-
versal transducer position and an in-­plane approach from lateral
to medial. In contrast, the use of coiled catheters reduces the risk
of an epidural catheter position.5–8 In a recent study, application
of intermittent boli via a catheter was superior to a continuous
infusion in terms of the analgesic effect.17
Study limitations
The study was performed on unembalmed cadavers using dye
injections. A possible limitation is the secondary impregnation of
anatomical structures that were not primarily stained. To mini-
mize this effect, preparation was started immediately after dye
injection and completed within 2 hours after a unilateral injec-
tion series.
To produce the dye, Alcian Blue (in powder form) was
dissolved in water and then filtered to remove undissolved dye
particles. The result is an aqueous solution with approximately
the same viscosity as the local anesthetics used in everyday clin-
ical practice.
Alternatives are in vivo MRI and CT scans. However, as
MRI is associated with gadolinium deposition in the brain/
other tissues and nephrogenic systemic fibrosis and CT has
possible side effects like cancer or radiocontrast reactions,
their application to healthy volunteers is subject to ethical
constraints.18
Another possibility is the examination of latex-­based injections
in Thiel or alcohol-­fixed cadavers. In this context, it is important
that the expansion of the injectate results from hydrodissection.
This is especially true for fascial plane blocks. We believe that
any fixation process has a significant influence on hydrodissec-
tion. This also applies to factors such as body temperature or
Seidel R, et al. Reg Anesth Pain Med 2020;45:102–106. doi:10.1136/rapm-2019-100896
Brief technical report
Reg Anesth Pain Med: first published as 10.1136/rapm-2019-100896 on 2 November 2019. Downloaded from http://rapm.bmj.com/ on March 3, 2022 at Post Graduate Institute of Medical
muscle activity in vivo. These variables should be systematically
investigated in comparative studies.
Conclusion
The spread of the injectate in TPVB is volume-­dependent. It
occurs craniocaudally, in the intercostal (lateral spread) and in
the epidural (medial spread) space. This confirms earlier work
on this subject. For an exact determination of the targeted inter-
costal space, the authors recommend an orientation based on
inner (sonoanatomical) landmarks (12th rib, 1st intercostal
space).19
In individual cases, the craniocaudal spread is not predictable.
Even with an injectate volume of 10 mL, less than 50% of the
adjacent intersegments were stained (85% with 20 mL). This is
sufficient for an analgesic effect due to the multisegmental supply
of each intercostal nerve. For a complete surgical analgesia of the
thoracic wall, however, the authors recommend an injection of
small volumes into each individual TPVS.
For the first time, we were able to show that the injection tech-
nique has a significant influence on the spread of the injectate. A
medial needle position (landmark-­guided) or a medial injection
direction (ultrasound-­guided) leads significantly more frequently
to epidural injectate spread.
Education & Research. Protected by copyright.
Contributors  All authors listed in the manuscript have contributed substantially
to the design of the study, interpretation of results, revised the manuscript and gave
final approval for publication. RS and MS conducted the study and analyzed the
data. RS drafted the manuscript.
Funding  The authors have not declared a specific grant for this research from any
funding agency in the public, commercial or not-­for-­profit sectors.
Competing interests  None declared.
Patient consent for publication  Not required.
Provenance and peer review  Not commissioned; externally peer reviewed.
ORCID iD
Ronald Seidel http://​orcid.​org/​0000-​0003-​1265-​7993
References
1 Karmakar MK. Thoracic paravertebral block. In: Gray AT, ed. Atlas of ultrasound-­guided
regional anesthesia. Philadelphia: Elsevier, 2018: 372–80.
2 Krediet AC, Moayeri N, van Geffen G-­J, et al. Different approaches to ultrasound-­
guided thoracic paravertebral block. Anesthesiology 2015;123:459–74.
3 Ben-­Ari A, Moreno M, Chelly JE, et al. Ultrasound-­guided paravertebral block using an
intercostal approach. Anesth Analg 2009;109:1691–4.
4 Cowie B, McGlade D, Ivanusic J, et al. Ultrasound-­guided thoracic paravertebral
blockade: a cadaveric study. Anesth Analg 2010;110:1735–9.
5 Luyet C, Eichenberger U, Greif R, et al. Ultrasound-­guided paravertebral puncture
and placement of catheters in human cadavers: an imaging study. Br J Anaesth
2009;102:534–9.
6 Luyet C, Hermann G, Ross S, et al. Ultrasound-­guided paravertebral puncture and
placement of catheters in human cadavers: where do the catheters go? Br J Anaesth
2011;106:246–54.
7 Luyet C, Meyer C, Herrmann G, et al. Placement of coiled catheters into the
paravertebral space*. Anaesthesia 2012;67:250–5.
8 Luyet C, Siegenthaler A, Szücs-­Farkas Z, et al. The location of paravertebral catheters
placed using the landmark technique. Anaesthesia 2012;67:1321–6.
9 Paraskeuopoulos T, Saranteas T, Kouladouros K, et al. Thoracic paravertebral spread
using two different ultrasound-­guided intercostal injection techniques in human
cadavers. Clin Anat 2010;110:1735–9.
10 Renes SH, Bruhn J, Gielen MJ, et al. In-­Plane ultrasound-­guided thoracic paravertebral
block: a preliminary report of 36 cases with radiologic confirmation of catheter
position. Reg Anesth Pain Med 2010;35:212–6.
11 Davies F, Gladstone RJ, Stibbe EP. The anatomy of the intercostal nerves. J Anat
1932;66:323–33.
12 Costache I, Pawa A, Abdallah FW. Paravertebral by proxy – time to redefine the
paravertebral block. Anaesthesia 2018;73:1185–8.
13 Forero M, Adhikary SD, Lopez H, et al. The erector spinae plane block: a novel
analgesic technique in thoracic neuropathic pain. Reg Anesth Pain Med
2016;41:621–7.
105
 Brief technical report
14 Ivanusic J, Konishi Y, Barrington MJ. A cadaveric study investigating the mechanism of
action of erector spinae blockade. Reg Anesth Pain Med 2018;43:567–71.
15 Kasimahanti R, Arora S, Bhatia N, et al. Ultrasound-­Guided single- vs double-­level
thoracic paravertebral block for postoperative analgesia in total mastectomy with
axillary clearance. J Clin Anesth 2016;33:414–21.
16 Uppal V, Sondekoppam RV, Sodhi P, et al. Single-­Injection versus multiple-­injection
technique of ultrasound-­guided paravertebral blocks: a randomized controlled study
comparing dermatomal spread. Reg Anesth Pain Med 2017;42:575–81.
106
Reg Anesth Pain Med: first published as 10.1136/rapm-2019-100896 on 2 November 2019. Downloaded from http://rapm.bmj.com/ on March 3, 2022 at Post Graduate Institute of Medical
17 Chen L, Wu Y, Cai Y, et al. Comparison of programmed intermittent bolus infusion
and continuous infusion for postoperative patient-­controlled analgesia with thoracic
paravertebral block catheter: a randomized, double-­blind, controlled trial. Reg Anesth
Pain Med 2019;44:240–5.
18 Ramalho M, Ramalho J, Burke LM, et al. Gadolinium retention and Toxicity—An
update. Adv Chronic Kidney Dis 2017;24:138–46.
19 Bouzinac A, Delbos A, Rontes O. [Ultrasound location of the first rib confirm the level
of realization of thoracic paravertebral block]. Ann Fr Anesth Reanim 2012;31:571–2.
Education & Research. Protected by copyright.
Seidel R, et al. Reg Anesth Pain Med 2020;45:102–106. doi:10.1136/rapm-2019-100896