The three commonly reported techniques for quadratus lumborum blocks are the injection of local anaesthetic on the lateral aspect of the quadratus lumborum muscle (lateral quadratus lumborum block approach; (QLB1),1 superficial and posterolateral to the quadratus lumborum muscle (posterior quadratus lumborum block approach; (QLB2)2 or anterior to the muscle (transmuscular quadratus lumborum block).3,4 However, there is no consensus on the dynamics of injectate spread using the different approaches. It has been suggested that the mechanism of action of the anterior quadratus lumborum approach is primarily due to the lumbar plexus,5,6 although some suggest a thoracic paravertebral mechanism.7–11 Attempts have been made to take advantage of this pathway to definitively demonstrate the mode of action.12 The mechanism of injectate spread following the posterior quadratus lumborum approach might be explained by recent developments in anatomical understanding demonstrating the presence of a lumbar inter-fascial triangle (LIFT) at the posterior aspect of the quadratus lumborum muscle, a potential route for medial, lateral and cranio-caudal injectate spread.13 In addition, the range of approaches to this technique, both in terms of relative location to the quadratus lumborum muscle and the cranio-caudal block level, has not been assessed to determine the association between injection location and route of local anaesthetic spread. Cadaveric evidence is a crucial facet to this understanding.
We therefore undertook a cadaveric study to clarify whether injection of dye posterior to the quadratus lumborum muscle was associated with a distinctive pattern of spread that differed from an anterior injection at the subcostal level. We hypothesised that a posterior approach would lead to medio-lateral and cranial dye spread through the LIFT, whereas an anterior, subcostal approach would lead to greater spread of dye deep to the transversalis fascia under the lateral arcuate ligament to the sub-endothoracic part of the thoracic paravertebral space.
Institutional approval for this cadaveric study was waived, and the study was approved by the director of the cadaveric laboratory. Six fresh human cadavers from the Cleveland Clinic cadaveric laboratory were chosen to represent both genders and a range of body mass indices, while avoiding those with known lumbar or thoracic spinal deformities, thoracic cavity disease or those who had undergone spinal surgery. All cadavers were left at room temperature for at least 12 h before performance of quadratus lumborum blocks.
Injections were performed in the anatomy laboratory by two investigators (H.E. and L.M.) and the dissections were performed by an anatomist (R.D.). Standardised methodology was used for each quadratus lumborum block approach, with patients positioned prone with blankets under the abdomen to flex the thoraco-lumbar spine. Anatomical localisation of the desired location was achieved with a curvilinear ultrasound transducer (2 to 5 MHz, 4C-SC; Mindray M7, DS, USA, Mahwah, New Jersey, USA). In both groups, a 125-mm, 18-ga echogenic Tuohy needle (Pajunk, Germany) was used. After injection of 3 to 5 ml of 0.9% sodium chloride to confirm correct location of the needle tip, a total of 30 ml 0.5% methylcellulose with India ink was injected on each side of the six cadavers (n = 12). Each quadratus lumborum block approach was performed on three right-sided hemispecimens and three left-sided hemispecimens to exclude any effect of the liver, spleen and the wider lumbo-costal triangle on the left, on the cranial spread of the dye.
Posterior quadratus lumborum block technique
Posterior quadratus lumborum block (QLB2) was performed at the L3 to L4 level on the posterolateral aspect of the quadratus lumborum muscle, where the anatomical location of the LIFT13 has been described. The transducer was placed in the mid-axillary line between the iliac crest and subcostal margin and then moved posteriorly until tapering off of the three abdominal wall muscle layers (external oblique, internal oblique and transversus abdominis) towards the quadratus lumborum muscle and the transversalis fascia was visualised. The needle was inserted in plane from lateral to medial towards the postero-lateral aspect of the quadratus lumborum muscle, between the quadratus lumborum and erector spinae, latissmus dorsi and serratus posterior inferior muscles within the middle layer of the thoracolumbar fascia (TLF). Following confirmation of location with 2 to 5 ml of 0.9% sodium chloride, 30 ml of dye was injecteded (Fig. 1; Supplementary Video 1, http://links.lww.com/EJA/A127).
Anterior subcostal quadratus lumborum block technique
The anterior, subcostal quadratus lumborum block (transmuscular) technique was performed with the curvilinear ultrasound transducer in a parasagittal oblique plane, after which the transducer was tilted medially, then the cranial side of the transducer rotated medially with the caudal side rotated laterally, producing an oblique sagittal view (Fig. 2).
The quadratus lumborum muscle was then identified as it tapers and thins cranially before its insertion into the lower margin of the last rib. The transducer was 6 to 8 cm lateral to the lumbar spinous process at the L1 to 2 level, immediately cranial and lateral to the ‘cross over point’ of the erector spinae and the quadratus lumborum. At this point, the erector spinae muscle no longer overlies the quadratus lumborum muscle, allowing the optimum ultrasound visualisation of the muscle as it is covered only by the latissimus dorsi muscle and the serratus posterior inferior muscle. Finally, the transducer was then moved cranially to identify the 12th rib as a rounded hyperechoic structure with an acoustic shadow behind it. At this level, the latissmus dorsi muscle, quadratus lumborum muscle, the diaphragmatic zone of apposition, perinephric fat and fascia and the kidney were identified.
The Tuohy needle was then advanced in plane using a caudal to cranial direction. The dye was injected once the needle tip was positioned anterior to the quadratus lumborum muscle, and between the quadratus lumborum muscle and the anterior layer of the TLF (ATLF). A cranial direction of dye spread was observed towards the 12th rib, with a lunar-shaped distribution and anterior displacement of the ATLF (Fig. 3; Supplementary Video 2, http://links.lww.com/EJA/A128).
Cadaveric anatomic dissection
One-hour following injection, cadavers were dissected bilaterally, in layers, from superficial (posterior) to deep (anterior), by an anatomist (RD), with subsequent mapping of the extent of dye spread in the transversus abdominus plane (TAP), the ATLF and middle TLF (MTLF), within psoas major and in the thoracic paravertebral spaces by dissecting at each level.
A vertical midline incision from the fifth thoracic spinous process to below the iliac crest was made. Transverse incisions were then performed at T5 to 6 levels and below the iliac crest, extending from the midline to the midaxillary line. A flap of skin and subcutaneous fascia was raised. The latissmus dorsi muscle flap was then dissected and removed laterally to the midaxillary line, facilitating TAP dissection. The erector spinae muscle was then exposed and removed as a flap to display the underlying MTLF, transverse process, and quadratus lumborum muscle.
The quadratus lumborum muscle, anterior to the erector spinae muscle, was identified, extending from the 12th rib cranially to the crest of the ilium caudally, and attaching to the transverse processes of the lumbar vertebrae medially, with a free lateral border.
The extent of spread of the dye was evaluated in all directions (cranio-caudal, antero-posterior and medio-lateral) anterior to the quadratus lumborum muscle, within the TAP and around tips of the lumbar transverse processes. To visualise the staining of the thoracic paravertebral space, the ribs were removed around their articulation with the transverse processes from T5 to T12, the thoracic transverse processes were removed and the thoracic paravertebral spaces were opened. The diaphragm, medial and lateral arcuate ligaments and aortic hiatus were all carefully observed for dye staining and were subsequently detached to view any thoracic cavity dye staining. The parietal pleura was then visualised for any break in its integrity indicating possible pneumothorax. Next, the individual thoracic segmental nerve roots and intercostal nerves up to T5 were observed for staining. The quadratus lumborum muscle was then dissected to expose the psoas major muscle, which was examined for staining of its anterior fascial covering. The psoas major was then cut and removed to expose the lumbar paravertebral space and visualise the lumbar plexus within the psoas major muscle and any possible extension of dye staining.
Finally, two anaesthetists (W.E. and S.K.) blinded to the technique used for the injection (not involved in the injection or dissection) then confirmed the dye distribution bilaterally as noted by the anatomist. The key anatomical structures (the quadratus lumborum muscle, TAP, diaphragm and thoracic paravertebral space) were documented and photographed. The degree of dye staining of structures was agreed upon by consensus and defined as either deeply stained, faintly stained or unstained.
A total of six adult cadavers were included in the study (three men, three women). The body mass indices of the cadavers ranged from 24 to 38 kg m−2. Ultrasound-guided quadratus lumborum block injections were successfully performed on each side of the six cadavers (n = 12), each cadaver receiving one posterior quadratus lumborum block (n = 6) and a contralateral anterior subcostal quadratus lumborum block (n = 6).
The paravertebral space, parietal pleura and transverse processes were identified bilaterally in all cadavers. The L1 segmental nerve was the first nerve that could be identified cranial to the iliac crest, but the LIFT was not accurately identified in any specimens.
The posterior quadratus lumborum block approach consistently demonstrated deep staining of the iliohypogastric, ilioinguinal, L1 nerve root, subcostal nerve and the T11 intercostal nerve as well as deep staining of the MTLF and faint staining of the ATLF in all specimens. There was variable staining of the T10 intercostal nerves with deep staining in two specimens (one right-sided specimen, one left), faint staining in three (two right, one left) specimens and no staining in two specimens (Fig. 4). The TAP was equally inconsistent, with one left-sided specimen deeply stained, three with faint staining (two left, one right) and two unstained specimens. There was only faint staining up to T10 inside the thoracic paravertebral space (Table 1).
The anterior subcostal quadratus lumborum approach specimens all demonstrated predictable deep staining of the iliohypogastric and ilioinguinal nerves, L1 and subcostal nerves, as well as dye observed under the lateral arcuate ligaments to involve T9 to 12 thoracic paravertebral spaces and staining of the nerve roots (Fig. 4). Five left-sided specimens revealed deep staining at T8 and one right-sided specimen was faintly stained at this level. There was deep staining at T7 in four specimens, and two left-sided cadaveric specimens were faintly stained. Deep staining of T6 was found in one specimen, faint staining in three, whereas two remained unstained. Only one specimen demonstrated faint staining of T5 (Fig. 5). None of the anterior subcostal injections or the posterior injections led to staining of the lumbar plexus within the psoas muscle, but two anterior subcostal injections led to faint dye involvement anterior to the psoas muscle. In one specimen, the space between the psoas muscle and iliacus muscle was deeply stained caudal to the upper lip of the iliac crest, two were faintly stained and three remained unstained. The MTLF was not involved in any specimens (Table 1).
One specimen in the anterior subcostal quadratus lumborum block group appeared to have sustained a possible pneumothorax, suggested by staining deep to the visceral pleura and surface of the lung.
The current study has demonstrated that performing quadratus lumborum block at different injection points (anterior or posterior to the quadratus lumborum muscle) and at different cranio-caudal levels (lumbar or subcostal) will affect the patterns of injectate spread. We noted that an anterior quadratus lumborum block with a subcostal approach was associated with more significant intra-thoracic distribution, consistently rising to the T7 to 8 level, and less reliably spreading up to T5. This contrasts with the limited cranial spread seen with a posterior quadratus lumborum block, spreading only to T11, and infrequently involving T9 or T10. There were no significant differences in terms of laterality, and the MTLF was consistently stained when a posterior quadratus lumborum block approach was performed.
A fundamental grasp of the anatomical basis of the quadratus lumborum muscle and its surrounding structures is necessary to understand the distribution of injectate following quadratus lumborum blocks and how this might translate clinically. The quadratus lumborum muscle is encapsulated by the TLF, a continuous fascial connective tissue running from the occiput to the sacrum.14–16 The anterior layer of the TLF, lying anterior to the quadratus lumborum muscle, is formed by the medial continuation of the transversalis fascia and the investing fascia of the psoas. It attaches to the anterior aspect of the transverse processes of the lumbar vertebrae, the 12th rib and the iliac crest. Cranially, the anterior layer of the TLF thickens to become the lateral arcuate ligament, and caudally, it joins the iliolumbar ligaments. Posterior to the quadratus lumborum muscle lies the MTLF, a multi-layered inter-muscular septum that attaches to the vertebral transverse processes medially. It is composed of the investing fascia of the quadratus lumborum muscle, the paraspinal retinacular sheath that encapsulates the paraspinal muscles, and the anterior lamina of the fused aponeuroses of transversus abdominis and internal oblique muscles. The posterior TLF is more complex, with contributions from numerous structures including the posterior lamina of the aponeurosis of the abdominal muscles.16 It is composed of deep and superficial laminae and meets the MTLF lateral to the quadratus lumborum muscle to form the lateral raphe (Supplementary Fig. S1, http://links.lww.com/EJA/A129).
At the core of the lateral raphe, sits a recently described anatomical compartment called the LIFT,13 a space representing possibly an important anatomic landmark for quadratus lumborum block (Supplementary Fig. S2, http://links.lww.com/EJA/A129). This is a previously poorly defined, fat-filled inter-fascial triangle situated along the lateral border of the paraspinal muscles from the 12th rib to the iliac crest. The LIFT is formed from the union of several independent fascial sheaths along the lateral border of the TLF. Due to its location, the LIFT has been suggested to be an important structure for distribution tension and viscoelastic forces13 and may be a conduit and potential route for spread of injectate from the lumbar to the thoracic paravertebral space following posterior quadratus lumborum block.
Although we hypothesised that dye injection in this well confined compartment would spread medio-laterally and cranially to the mid-thoracic paravertebral level, our results demonstrated the contrary. Thoracic paravertebral spread up to a level of T9 to 10 was only apparent in a single cadaver. Although the target injection point was within the LIFT, it was difficult to visualise both sonographically and during dissection. Injection near the LIFT may have led to spread of injectate within the substance of the MTLF, superficial to the epimysial fascia of the quadratus lumborum muscle. Our results provide limited evidence that the LIFT might be the route towards the lower thoracic paravertebral space. Clinical evidence of higher dermatomal coverage might be due to blockade of the network of sympathetic fibres within the TLF.17,18 Alternatively, the extent of spread may be limited by the cranio-caudad level at which the quadratus lumborum block is performed.
The fibres of the MTLF thicken as they approach the tips of the lumbar transverse processes, but the fascia is thinner and less organised between each transverse process.19 All posterior quadratus lumborum block specimens demonstrated deep staining of the MTLF with medial spread towards the transverse processes, and it is thus possible that in this area, between the lumbar transverse processes, injectate has free access to the ventral primary rami.16
The anterior subcostal quadratus lumborum block approach was associated with consistent dye spread to the mid-thoracic level in most specimens. The endothoracic fascia lines the diaphragm superiorly, and transversalis fascia lines its inferior surface.20 The costal insertions of the diaphragm create a barrier between these fasciae; however, three defects in the diaphragmatic attachments provide communications between the abdominal and thoracic compartments.21 The fascial layers in thoracic and lumbar regions are sometimes described as being separate; however, the thoracic and lumbar paravertebral spaces are in continuity, signifying that this pathway between the thoracic and lumbar paravertebral space is documented. Dorsally, the endothoracic and transversalis fascia are continuous behind the diaphragm at the lumbocostal arches and the aortic and oesophageal hiatus, providing a communication point between the abdominal and thoracic cavities.22,23 Posteriorly, the crura of the diaphragm form two arcuate ligaments that transmit structures between the abdominal and thoracic cavity. The medial arcuate ligament attaches to the body of L1and L2, as well as the transverse process of LI, and transmits the psoas major muscle. The lateral arcuate ligament forms an arch that transmits the quadratus lumborum muscle, attaching medially to the transverse process of the first lumbar vertebrae and laterally to the 12th rib.
Caudal spread of injectate to the lower extent of the thoracic paravertebral space through these potential pathways has been demonstrated in both cadaveric8 and clinical studies.3,7,9 Our cadaveric study has established that injection of dye anterior to the quadratus lumborum muscle at the subcostal level allows cranial spread of injectate under the lateral arcuate ligament of the diaphragm from the lumbar level to the thoracic paravertebral space, through a tunnel formed by the potential space deep to the transversalis fascia and endothoracic fascia (Fig. 6). This modification of traditional anterior quadratus lumborum block takes advantage of the path between the lumbar and thoracic paravertebral space by injecting as close as possible to the arcuate ligament, encouraging more cranial spread of injectate.8,9,12
The high thoracic spread of injectate in anterior subcostal quadratus lumborum block specimens might have several potential explanations. First, below the level of L2, the structures forming the lateral margin of the paraspinal muscles (TLF, latissmus dorsi muscle, lateral raphe and the LIFT) are reinforced internally by the presence of the quadratus lumborum and its associated fascia; above L2, the lateral margin of the paraspinal muscles is reinforced only by the transversalis fascia. Injection anterior to the quadratus lumborum muscle higher than L2 level will lead to injectate spreading cranial on the transversalis fascia to the endothoracic fascia in the lower thoracic paravertebral space. The lateral border of the quadratus lumborum muscle begins cranially from a point medial to the lateral border of the paraspinal muscle and angles from craniomedial to caudolateral as it descends towards the ilium. At L2, the free border of the quadratus lumborum widens beyond the lateral margin of the paraspinal muscles at a point that can be sonographically visualised as the ‘cross-over point’. A subcostal approach above this landmark may lead to more significant proximal distribution of injectate and easier sonographic identification of the quadratus lumborum muscle.
Our results also demonstrated that there was no difference in distribution of dye between right and left sides, despite potential anatomical discrepancy. Right-sided cranial spread of injectate was not hampered by the presence of the liver on the right or the wider lumbocostal triangle on the left.
Overall, the subendothoracic fascia demonstrably provided a reliable link between the point of injection anterior to the quadratus lumborum muscle subcostally and the thoracic paravertebral space, a phenomenon that we did not see consistently with the posterior quadratus lumborum block approach. This evidence suggests that the anterior subcostal technique is, potentially, an alternative reliable approach to the thoracic paravertebral space from the upper lumbar region.
However, clinical correlation is mandatory in any cadaveric study. Although we used fresh cadavers, tissue architecture might differ from live tissue24 potentially leading to differing injection dynamics. The technical skills required to perform these blocks might not be easily learned, and the true clinical efficacy must be demonstrated in clinical trials. To date, only three randomised controlled clinical trials assessing either a lateral25,26 or posterior quadratus lumborum block approach have been published27; yet, the anterior approach, which we have demonstrated to have the most significant thoracic paravertebral spread, is reported in nothing more than case reports or case series.3,12,28,29,30
We therefore suggest that future trials should be well designed and adequately powered to demonstrate analgesic superiority of an anterior, subcostal quadratus lumborum block approach for abdominal and urological procedures versus TAP blocks, epidurals or thoracic paravertebral blocks.
Our study demonstrates that injection of dye on the posterior aspect of the quadratus lumborum led to spread through the LIFT and TAP, but cranial spread was limited to T10. An anterior quadratus lumborum block approach at the subcostal level was associated with cranial spread up to T6 to 7 with limited involvement of the TAP. This suggests that the point of injection of a quadratus lumborum block may have an impact on the dermatomal coverage achieved; thus, clinical translation of these findings to determine the practical significance is warranted.
Acknowledgements relating to this article
Assistance with the study: all images have been used with permission from Cleveland Clinic art photography department.
Financial support and sponsorship: none.
Conflict of interests: none.
1. Blanco R. TAP block under ultrasound guidance: the description of a ‘no pops’ technique: 271. Reg Anesth Pain Med
3. Børglum J, Jensen K, Moriggl B, et al. Ultrasound-guided transmuscular quadratus lumborum blockade. Br J Anaesth
2013; 111 (eLetters Supplement): doi: 10.1093/bja/el_9919.
4. El-Boghdadly K, Elsharkawy H, Short A, et al. Quadratus lumborum block nomenclature and anatomical considerations. Reg Anesth Pain Med
5. Adhikary SD, El-Boghdadly K, Nasralah Z, et al. A radiologic and anatomic assessment of injectate spread following transmuscular quadratus lumborum block in cadavers. Anaesthesia 2016. Anesthesia
6. Carline L, Mcleod GA, Lamb C. A cadaver study comparing spread of dye and nerve involvement after three different quadratus lumborum blocks. Br J Anaesth
7. Carney J, Finnerty O, Rauf J, et al. Studies on the spread of local anaesthetic solution in transversus abdominis plane blocks. Anaesthesia
8. Saito T, Den S, Tanuma K, et al. Anatomical bases for paravertebral anesthetic block: fluid communication between the thoracic and lumbar paravertebral regions. Surg Radiol Anat
9. Karmakar MK, Gin T, Ho AMH. Ipsilateral thoraco-lumbar anaesthesia and paravertebral spread after low thoracic paravertebral injection. Br J Anaesth
10. Saito T, Tanuma K, Den S, et al. Pathways of anesthetic from the thoracic paravertebral region to the celiac ganglion. Clin Anat
11. Krediet AC, Moayeri N, van Geffen GJ, et al. Different approaches to ultrasound-guided thoracic paravertebral block. Anesthesiology
12. Elsharkawy H. Quadratus lumborum block with paramedian sagittal oblique (subcostal) approach. Anaesthesia
13. Schuenke MD, Vleeming A, Van Hoof T, et al. A description of the lumbar interfascial triangle and its relation with the lateral raphe: Anatomical constituents of load transfer through the lateral margin of the thoracolumbar fascia. J Anat
14. Mirilas P, Skandalakis JE. Surgical anatomy of the retroperitoneal spaces – Part I: Embryogenesis and anatomy. Am Surg
15. Mirilas P, Skandalakis JE. Surgical anatomy of the retroperitoneal spaces Part II: The architecture of the retroperitoneal space. Am Surg
16. Willard FH, Vleeming A, Schuenke MD, et al. The thoracolumbar fascia: anatomy, function and clinical considerations. J Anat
17. Benetazzo L, Bizzego A, De Caro R, et al. 3D reconstruction of the crural and thoracolumbar fasciae. Surg Radiol Anat
18. Blanco R, McDonnell JG. The mechanism of the quadratus lumborum block: A peripheral sympathetic field block? Br J Anaesth
2016; 117 (eLetters Supplement): doi: 10.1093/bja/el_13593.
19. Tesh KM, Dunn JS, Evans JH. The abdominal muscles and vertebral stability. Spine
20. Skandalakis PN, Zoras O, Skandalakis JE, et al. Transversalis, endoabdominal, endothoracic fascia: who's who? Am Surg
21. Frias Vilaça A, Reis AM, Vidal IM. The anatomical compartments and their connections as demonstrated by ectopic air. Insights Imaging
22. Oliphant M, Berne AS, Meyers MA. The subserous thoracoabdominal continuum: Embryologic basis and diagnostic imaging of disease spread. Abdom Imaging
23. Kleinman PK, Brill PW, Whalen JP. Anterior pathway for transdiaphragmatic extension of pneumomediastinum. AJR Am J Roentgenol
24. Martin DC, Medri MK, Chow RS, et al. Comparing human skeletal muscle architectural parameters of cadavers with in vivo ultrasonographic measurements. J Anat
25. Parras T, Blanco R. Randomised trial comparing the transversus abdominis plane block posterior approach or quadratus lumborum block type I with femoral block for postoperative analgesia in femoral neck fracture, both ultrasound-guided. Rev Esp Anestesiol Reanim
26. Blanco R, Ansari T, Riad W, et al. Quadratus lumborum block versus transversus abdominis plane block for postoperative pain after cesarean delivery. Reg Anesth Pain Med
27. Blanco R, Ansari T, Girgis E. Quadratus lumborum block for postoperative pain after caesarean section: a randomised controlled trial. Eur J Anaesthesiol
28. Baidya DK, Maitra S, Arora MK, et al. Quadratus lumborum block: an effective method of perioperative analgesia in children undergoing pyeloplasty. J Clin Anesth
29. Chakraborty A, Goswami J, Patro V. Ultrasound-guided continuous quadratus lumborum block for postoperative analgesia in a pediatric patient. A A Case Rep
30. Ueshima H, Hiroshi Otake H, Lin J. Ultrasound-guided quadratus lumborum block: an updated review of anatomy and techniques. Biomed Res Int
Supplemental Digital Content
© 2017 European Society of Anaesthesiology