You are here

Article Text

Article menu

Download PDFPDF

Original research
Evaluation of nerve capture using classical landmarks for genicular nerve radiofrequency ablation: 3D cadaveric study
  1. John Tran1,
  2. Philip Peng2 and
  3. Anne Agur1
  1. 1 Surgery (Division of Anatomy), University of Toronto, Toronto, Ontario, Canada
  2. 2 Anesthesia, Toronto Western Hospital, University Health Network, Toronto, Ontario, Canada
  1. Correspondence to Dr John Tran, Surgery (Division of Anatomy), University of Toronto, Toronto, ON M5S, Canada; johnjt.tran{at}mail.utoronto.ca

Abstract

Background and objectives Radiofrequency (RF) denervation of the superolateral genicular nerve (SLGN), superomedial genicular nerve (SMGN) and inferomedial genicular nerve (IMGN) is commonly used to manage chronic knee joint pain. However, knowledge of articular branches captured, using classical landmarking techniques, remains unclear. In order to enhance and propose new RF procedures that conceivably capture a greater number of articular branches, more detailed cadaveric investigation is required. The objectives were to (1) determine which articular branches are captured or spared using classical landmarking techniques, and (2) evaluate the anatomical feasibility of classical landmarking techniques using three-dimensional (3D) modeling technology.

Methods Ultrasound-guided classical superolateral/superomedial/inferomedial landmarking techniques were used to position RF cannulae in five specimens. The articular branches, bony and soft tissue landmarks, and cannula tip position, were meticulously dissected, digitized and modeled in 3D. Simulated lesions were positioned at the cannula tip, on the 3D models, to determine which articular branches were captured or spared. Capture rates of articular branches were compared.

Results In all specimens, classical superolateral/superomedial techniques captured the transverse deep branches of SLGN and SMGN, and articular branches of lateral and medial nerve to vastus intermedius, while sparing distal branches of SLGN/SMGN. The inferomedial technique captured anterior branches of IMGN while sparing the posterior and inferior branches.

Conclusions This study provides anatomical evidence supporting the effectiveness of classical landmarking for genicular nerve ablation; however, each technique resulted in sparing of articular branches. The extensive innervation of the knee joint suggests the use of supplementary landmarks to improve capture rates and potentially patient outcomes.

  • chronic pain
  • pain management
  • lower extremity
  • ultrasonography

https://doi.org/10.1136/rapm-2020-101894

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Radiofrequency (RF) ablation of genicular nerves is commonly used to manage chronic knee joint pain related to osteoarthritis, with several clinical studies reporting positive outcomes.1–4 Classical landmarks to target the superolateral genicular nerve (SLGN) and superomedial genicular nerve (SMGN) are the junction of the midpoint of the femoral shaft and the lateral/medial femoral condyle, whereas the inferomedial genicular nerve (IMGN) is targeted at the junction of the midpoint of the tibial shaft and medial tibial condyle.1 2 More recently, the placement of RF cannulae, to target the SLGN and SMGN, have been repositioned to a more posterior location at the level of the junction of the femoral shaft and lateral/medial femoral condyle.3 4 However, knowledge of nerve targets captured using the classical landmarks is first required prior to modifying and/or developing novel RF procedures that conceivably capture more nerves. Furthermore, the accuracy of classical landmarks has recently been discussed and warrants further anatomical investigation.5–7

    Previous cadaveric studies evaluating genicular nerve capture have used low-volume dye injectate as a marker for the location of the RF cannula tip.5 8 However, the variability of tissue dye spread, with the injection methodology, does not reflect the three-dimensional (3D) volume of an RF lesion, and the precise location of the cannula tip within the area of spread is unclear. A more recent cadaveric study has used 3D modeling technology to simulate accurate lesion volumes based on manufacturer’s dimensions.9 The use of 3D modeling technology represents a more geometrically appropriate method to visualize and evaluate nerve capture in three dimensions. Therefore, the objectives of this pilot study were (1) to determine which articular branches are captured or spared using classical landmarking techniques;and (2) to evaluate the anatomical feasibility of classical landmarking techniques using 3D modeling technology.

    Methods

    Five lightly embalmed cadaveric lower limb specimens (3R/2 L) with no visible signs of injury, pathology or previous surgery were used in this study. No further demographic data were available.

    Ultrasound (US) scanning protocol

    A high-frequency (6–13 MHz) linear probe was used to obtain scans for all approaches. An 18 Ga 100 mm cannula (Diros Technology, Markham, Canada) was positioned at the target site of the superolateral, superomedial and inferomedial approaches. Target sites were

    • Superolateral approach: at the junction of the midpoint of the femoral shaft and lateral femoral condyle.

    • Superomedial approach: at the junction of the midpoint of the femoral shaft and medial femoral condyle.

    • Inferomedial approach: at the junction of the midpoint of the tibial shaft and medial tibial condyle.

    As the position of the cannula cannot be maintained during dissection, a marker deployment system developed in our laboratory was used to demarcate cannula placement location. The marker deployment system consists of a cannula, advancer stylet and stainless steel marker (cylinder rod, 0.8 mm diameter). The dimensions of the marker were designed to allow free passage through the lumen of the cannula. The advancer stylet had an unbeveled end and completely filled the diameter and length of the cannula. As the advancer stylet was carefully advanced into the lumen of the cannula, the marker was deployed into the tissue at the location of the cannula tip. The advancer stylet and cannula were then removed. US scanning was used to confirm that the deployed marker was located at the target site of each approach (figure 1).

    Figure 1
    Figure 1

    Ultrasound scan with deployed marker. The inset shows the cannula with stylet deploying marker. Red arrows indicate deployed marker. Reprinted with permission from Philip Peng Educational Series. F, femur.

    Superolateral approach

    The lower limb specimens were placed on a dissection table with the lateral aspect of the thigh and leg accessible for scanning. The US probe was placed on the lateral aspect of the thigh along the long axis of the femur. The target was the junction of the diaphysis and epiphysis at the location where the fascial expansion between the vastus lateralis/intermedius and femoral shaft could be visualized. The depth of this target was noted on the US machine. Next, the probe was rotated 90° to obtain a short-axis view of the target using the depth as the reference point. The cannula was then inserted in-plane, from anterior to posterior, to the target site and the marker deployed. The orientation of the probe was strictly kept in lateral-to-medial position.

    Superomedial and inferomedial approaches

    The specimens were placed on the dissection table with the medial aspect of the lower limb accessible. For the superomedial approach, the US probe was placed along the long axis of the femur to visualize the fascial expansion between the vastus medialis and femur located at the target site at the junction of the diaphysis and epiphysis. Next, the probe was rotated 90° to view the target site in short axis using the depth as the reference point. In the short-axis view, the cannula was advanced in-plane to the target site and the marker deployed. The orientation of the probe was strictly kept in the lateral-to-medial position.

    For the inferomedial approach, the probe was placed on the inferomedial aspect of the knee joint along the long axis of the tibia and translated to visualize the medial collateral ligament of the knee and the junction of the tibial diaphysis and epiphysis. The plane deep to the medial collateral ligament was the target site for this approach. The depth of the target site in long-axis view was noted on the US machine, and then the probe was rotated 900 to view the target site in short axis at this depth. The cannula was inserted in-plane and positioned at the target site to deploy the marker.

    Dissection, digitization and 3D modeling

    Following deployment of markers, the skin, superficial and deep fascia were removed from the lower limb to expose the underlying muscles. Next, the main trunk of the femoral nerve (FN), tibial nerve (TN) and common fibular nerve (CFN) nerves as well as their branches were identified and exposed proximally. The articular nerves of FN, TN and CFN that coursed along the periosteum to supply the anterior knee joint capsule were dissected and traced distally, as previously described.10 11 The dissected articular nerves included the medial branch of nerve to vastus intermedius (MBNVI) and lateral branch of nerve to vastus intermedius (LBNVI), SMGN, IMGN and SLGN. Each articular nerve was meticulously traced to their termination in the knee joint capsule while maintaining the position of the deployed markers. All specimens were photographed throughout the dissection process.

    The deployed markers, articular nerves and bony surfaces were digitized using a Microscribe G2X Digitizer (Immersion Corporation, San Jose, CA, USA; accuracy,±0.23 mm), along with custom software developed in our laboratory. Following digitization, each specimen was skeletonized, leaving only the femur, patella, tibia, fibula and knee joint capsule. The skeletonized specimens were scanned using a Faro Laser ScanArm (FARO Technologies, Lake Mary, Florida, USA; accuracy ±35 µm) to create a high-resolution surface scan. Since both the Microscribe digitizer and Laser ScanArm collects Cartesian coordinate data, the digitized nerves, deployed markers and bony surfaces were registered with the scanned surface data using Autodesk Maya (Autodesk, San Rafael, California, USA). The completed 3D models included the deployed markers, SLGN, LBNVI, MBNVI, SMGN, and IMGN and high-resolution surface reconstruction. Following reconstruction of the 3D volumetric models, an oval lesion with a radius of 5 mm (18 Ga cannula) was generated (Diros Technologies) and placed at the location of each deployed marker.

    The 3D models were used to visualize and document the course of articular nerves relative to the generated lesion. Nerve capture rates were determined based on the observed overlap of lesion volume with the course of the articular nerves.

    Results

    In all specimens, the articular branches of SLGN, LBNVI, SMGN, MBNVI, IMGN and the deployed markers were found. The two markers targeting the SLGN and SMGN, placed at the junction of the midpoint of the femoral shaft and the lateral/medial femoral condyle, were both located at the periosteal level in close proximity to articular branches from the superior lateral and superior medial genicular arteries respectively (figures 2–5). The marker targeting the IMGN, placed at the junction of the midpoint of the tibial shaft and medial tibial condyle, was also found at the periosteal level in close relationship with articular branches of the inferior medial genicular artery, deep to the medial collateral ligament of the knee (figure 6). The course and branching patterns of the articular nerves will be presented first followed by nerve capture rates.

    Figure 2
    Figure 2

    Dissection of articular nerves supplying the lateral aspect of the knee joint, lateral views. (A) Independent origin of lSBr of the SLGN. (B) Trifurcation of SLGN into DBr, tSBr and lSBr. (C) SBr and DBr of SLGN originating from a common trunk. Blue arrow: origin of DBr of SLGN; red arrow: deployed marker. Reprinted with permission from Philip Peng Educational Series. ABr, anterior branch of the lateral nerve to the vastus intermedius; BF, biceps femoris; DBr, deep branch; E, lateral epicondyle; F, shaft of femur; lSBr, longitudinal superficial branch; P, patella; PBr, posterior branch of the lateral nerve to vastus intermedius; SBr, superficial branch; SLGN, superior lateral genicular nerve; tSBr, transverse superficial branch; VI, vastus intermedius; VL, vastus lateralis.

    Figure 3
    Figure 3

    Dissection of the articular branches of the SLGN, lateral view. (A) Dissection. (B) Enlargement of boxed area indicated in (A), showing DBr of SLGN dividing into tDBr and lDBr. Black arrows indicate neurovascular bundle coursing along the lateral femoral condyle; red arrows: deployed marker. Reprinted with permission from Philip Peng Educational Series. BF, biceps femoris (reflected); DBr, deep branch; F, shaft of femur; lDBr, longitudinal deep branch; lSBr, longitudinal superficial branch; JC, joint capsule (partially reflected); P, patella; SLGN, superolateral genicular nerve; tDBr, transverse deep branch; tSBr, transverse superficial branch; VI, vastus intermedius (reflected anteriorly).

    Figure 4
    Figure 4

    Dissection of articular nerves supplying the medial aspect of the knee joint, medial views. (A) Bifurcation of SBr of SMGN into tSBr and lSBr SBrs just superior to adductor tubercle (*). (B) Independent origin of lDBr of SMGN. (C) High bifurcation of SMGN into SBr and DBr. Blue arrow: origin of DBr of SMGN; green arrows: DBr of SMGN joining the superior genicular transverse artery; red arrows: deployed marker. Reprinted with permission from Philip Peng Educational Series. ABr, anterior branch of medial nerve to the vastus intermedius; DBr, deep branch; E, medial epicondyle; F, shaft of femur; lDBr, longitudinal deep branch; lSBr, longitudinal superficial branch; P, patella; PBr, posterior branch of medial nerve to vastus intermedius; S, sartorius; SBr, superficial branch; SMGN, superior medial genicular nerve; tSBr, transverse superficial branch; VM, vastus medialis.

    Figure 5
    Figure 5

    Dissection of the bifurcation of the SMGN. (A) Retraction of SMGN posteriorly to reveal course of DBr. (B) Exposure of distal bifurcation of DBr of SMGN. (C) Partial reflection of VM to expose independent origins of the tDBr and IDBr deep branch of SMGN. Blue arrows indicate origin of DBrs of SMGN; black arrows: superior genicular transverse artery; green arrows: DBr of SMGN joining the superior genicular transverse artery; Asterisks indicate the adductor tubercle. Reprinted with permission from Philip Peng Educational Series. DBr, deep branch; F, shaft of femur; lDBr, longitudinal deep branch; SBr, superficial branch; SMGN, superior medial genicular nerve; tDBr, transverse deep branch; VM, vastus medialis

    Figure 6
    Figure 6

    Dissection of branches of the IMGN, medial views. (A) Excision of the MCL of the knee to expose branching of IMGN. (B) Reflection of medial head of gastrocnemius (the MG) posteriorly to expose course of articular branches of IMGN superior to margin of PT. (C) Enlargement of boxed area in B, showing the course of ABr, PBr and IBr. Red arrows indicate deployed marker. Reprinted with permission from Philip Peng Educational Series. ABr, anterior branch of IMGN; IBr, inferior branch of IMGN; IMGN, inferior medial genicular nerve; MCL, medial collateral ligament; P, patella; PBr, posterior branch of IMGN; PT, popliteus

    The SLGN coursed inferiorly in fatty tissue deep to the biceps femoris. The SLGN terminated by dividing into deep and superficial articular branches at or superior to the lateral femoral condyle (figure 2). In all specimens, the proximal, deep articular branch further divided into transverse and longitudinal deep branches that coursed in a plane between the periosteum of the femur and deep surface of the capsule to supply the knee joint (figure 3). The transverse and longitudinal deep branches were accompanied by articular branches of the superior lateral genicular artery (SLGA) in all specimens. The superficial branches of SLGN, transverse and longitundinal, coursed anteroinferiorly on the capsule of the knee joint (figures 2 and 3). The transverse superficial branch penetrated the capsule posterolateral to the patella, whereas the longitudinal superficial branch continued inferiorly as far as the joint line prior to penetrating the joint capsule. The superficial branches of SLGN were accompanied by articular branches from the superior lateral genicular vessels in two specimens (figure 2C) and in the remaining three specimens no accompanying vessels were found (figure 2A).

    The LBNVI descended on the periosteal surface of the anterolateral femur where it divided into anterior and posterior branches superior to the base of the patella (figure 2). The anterior branch was shorter and terminated in the region of the suprapatellar bursa, whereas the posterior branch was longer and continued distally along the femur to reach the knee joint (figure 2B).

    The SMGN gave off deep and superficial articular branches as the nerve descended along the adductor magnus tendon (figures 4 and 5). The deep branch of SMGN originated superior to the adductor tubercle and coursed anteroinferiorly to reach the periosteum of the femur (figure 5). The deep branch divided into transverse and longitudinal deep branches that coursed in a plane between the capsule and periosteum of the femur to supply the knee joint. In three specimens, the transverse and longitudinal deep branches shared a common trunk (figure 5A,B), and in two specimens, the longitudinal deep branch originated independently (figure 5C). In all specimens, the transverse and longitudinal deep branches were accompanied by articular branches of the superior medial/descending genicular artery. The superficial branch of SMGN continued inferiorly and divided into transverse and longitudinal superficial branches at or just distal to the adductor tubercle (figure 4). The transverse superficial branch coursed anteroinferiorly to penetrate the capsule posteromedial to the patella, whereas the longitudinal superficial branch continued inferiorly to supply the capsule around the medial femoral epicondyle. The transverse superficial branch was accompanied by articular branches of the superior medial/descending genicular artery in two specimens (figure 4A). The transverse superficial branch, in three specimens, and the longitudinal superficial branch, in all specimens, were not accompanied by articular arteries (figure 4B,C).

    The MBNVI coursed along the anteromedial shaft of the femur and branched into anterior and posterior branches superior to the base of the patella (figure 4C). The anterior branch coursed to the medial aspect of the suprapatellar bursa, whereas the posterior branch continued more distally to the anteromedial region of the knee joint (figure 4).

    The IMGN coursed inferomedially along the superior border of popliteus, deep to the medial head of gastrocnemius, where it gave off anterior, posterior and inferior branches (figure 6). The anterior branch terminated in the anteromedial knee joint capsule, the posterior branch in the posteromedial capsule and the inferior branch in the periosteum of the medial aspect of the shaft of the tibia (figure 6).

    Nerve capture rates

    Consistent innervation of the knee joint capsule by SLGN, LBNVI, SMGN, MBNVI and IMGN was found in all specimens. However, classical landmarking for genicular nerve ablation did not capture all individual articular branches supplied by the nerves listed previously (table 1).

    View this table:
    Table 1

    Capture rates of articular branches using classical landmarks for genicular nerve ablation

    The classical landmarking of the superolateral approach (table 1), at the junction of the midpoint of the femoral shaft and the lateral femoral condyle, captured the transverse deep branch of SLGN and posterior branch of LBNVI in all specimens (figures 2, 3 and 7). The longitudinal deep and superficial branches of SLGN, as well as the anterior branch of LBNVI, were spared (figure 7).

    Figure 7
    Figure 7

    Dissection and 3D reconstruction of articular nerves supplying lateral aspect of knee joint. (A) Dissection showing position of deployed marker (red arrow). (B) Skeletonized specimen with intact joint capsule showing simulated lesion (red) relative to shaft of F, E and P. (C) 3D model showing articular branches and simulated lesion relative to reconstructed bony landmarks. Lesion captures the tDBr of the SLGN and PBr of the lateral nerve to the vastus intermedius. Branches not captured include lDBr, tSBr and lSBr of SLGN. Blue arrow: origin of DBr of SLGN. Reprinted with permission from Philip Peng Educational Series. 3D, three-dimensional; ABr, anterior branch of lateral nerve to vastus intermedius; DBr, deep branch; E, lateral epicondyle; F, femur; lDBr, longitudinal deep branch; lSBr, longitudinal superficial branch; P, patella; PBr, posterior branch; SBr, superficial branch; SLGN, superolateral genicular nerve; tDBr, transverse deep branch; tSBr, transverse superficial branch.

    The superomedial approach (table 1), targeting the junction of the midpoint of the femoral shaft and medial femoral condyle, captured the transverse deep branch of SMGN and spared the longitudinal deep and superficial branches of SMGN in all specimens (figures 4 and 8). In addition, the posterior branch of MBNVI was captured in all specimens with the exception of one where the branch coursed more anteriorly along the femoral shaft (figure 4B). The anterior branch of MBNVI was captured in one specimen where the nerve had a more distal bifurcation (figure 4C).

    Figure 8
    Figure 8

    Dissection and 3D reconstruction of articular nerves supplying medial aspect of knee joint. (A) Dissection showing position of deployed marker (red arrow). (B) Skeletonized specimen with intact joint capsule showing simulated lesion (red) relative to shaft of F, E, adductor tubercle (*), and patella (P). C. 3D model showing articular branches and simulated lesion relative to reconstructed bony landmarks. Lesion captures tDBr of SMGN, ABr and PBr of the medial nerve to the vastus intermedius. Branches not captured include lDBr, tSBr and lSBr of SMGN. Blue arrows: origin of DBr of SMGN; green arrows: DBr of SMGN joining the superior genicular transverse artery; blue dashed line: outline of the adductor tubercle. Reprinted with permission from Philip Peng Educational Series. 3D, three-dimensional; ABr, anterior branch; DBr, deep branch; E, medial epicondyle; F, femur; lDBr, longitudinal deep branch; lSBr, longitudinal superficial branch; PBr, posterior branch; SBr, superficial branch; SMGN, superior medial genicular nerve; tDBr, transverse deep branch; tSBr, transverse superficial branch.

    The inferomedial approach (table 1), targeting the junction of the midpoint of the tibial shaft and medial tibial condyle, captured the anterior branch of IMGN in all specimens while sparing the posterior and inferior branches (figures 6 and 9). The posterior and inferior branches originated posterior to the midline of the tibial shaft and therefore were not captured (figure 9C).

    Figure 9
    Figure 9

    Dissection and 3D reconstruction of articular nerves supplying inferomedial aspect of knee joint. (A) Dissection showing position of deployed marker (red arrow). (B) Skeletonized specimen with intact joint capsule showing simulated lesion (orange) relative to T, JC and MFC. (C) 3D model showing articular branches of IMGN and simulated lesion relative to P and PT. Lesion captures the ABr of IMGN while sparing the PBr and IBr branches. Reprinted with permission from Philip Peng Educational Series. 3D, three-dimensional; ABr, anterior branch; IBr, inferior branch; IMGN, inferior medial genicular nerve; JC, joint capsule; MFC, medial femoral condyle; P, patella; PBr, posterior branch; PT, popliteus; T, tibia

    The locations of the articular nerves and RF lesions for the superolateral (figure 10), superomedial and inferomedial (figure 11) approaches have been correlated with standard fluoroscopic images. This enables correlation of the 3D models with anatomical bony landmarks visible with fluoroscopy.

    Figure 10
    Figure 10

    Fluoroscopic image and three-dimensional reconstruction of simulated lesion (red) correlating articular nerves supplying lateral aspect of knee joint relative to fluoroscopic landmarks. (A) Anterolateral view. (B) Lateral view. Blue arrows: origin of DBr of SLGN; yellow dashed line (yellow): outline of E and P; x (red), classical superolateral landmark. Reprinted with permission from Philip Peng Educational Series. ABr, anterior branch; DBr, deep branch; E, lateral epicondyle; P, patella; PBr, posterior branch; SBr, superficial branch; SLGN, superior lateral genicular nerve.

    Figure 11
    Figure 11

    Fluoroscopic image and three-dimensional reconstruction of simulated lesion (red) correlating articular nerves supplying medial aspect of knee joint relative to fluoroscopic landmarks. (A) Anterior view. (B) Anteromedial view. (C) Medial view. Blue dashed line: outline of adductor tubercle; x (red); classical superomedial landmark; x (teal); classical inferomedial landmark. Reprinted with permission from Philip Peng Educational Series. ABr, anterior branch. E, medial epicondyle; lDBr, longitudinal deep branch; lSBr, longitudinal superficial branch; P, patella; PBr, posterior branch; SBr, superficial branch; SMGN, superior medial genicular nerve; tDBr, transverse deep branch; tSBr, transverse superficial branch.

    Discussion

    In the current cadaveric study, 3D modeling technology was used to document the course and location of articular nerves innervating the knee joint and to simulate RF lesions. This methodology enabled reconstruction of geometric RF lesion volumes in 3D space to visualize nerve capture using classical landmarking approaches for genicular nerve ablation (superolateral, superomedial and inferomedial). Clinically, the correlation of fluoroscopic images with the 3D models, of the location of articular nerves and RF lesions, provides insight into structural relationships not seen with fluoroscopic imaging alone.

    In the anatomical and clinical literature, due to differences in naming of the same articular nerves, it is often difficult to compare the findings between studies, resulting in inconsistencies. Therefore, we will discuss the naming convention of the superolateral aspect of the knee joint to clarify the terminology used in this study. The literature is clear that the superolateral aspect of the knee joint receives innervation from branches of the common fibular and sciatic nerves.10 12–14 The branches originating from the common fibular nerve have been named the lateral articular nerve12 or articular branches of the common fibular (peroneal) nerve.10 13 14 Horner et al 13 also reported branches originating from the sciatic nerve referring to these as the SLGN. Comparison of the innervation of the superolateral aspect of the knee with previous literature can be difficult as more recent studies have collectively referred to all of these branches as the SLGN.5 15 16 To facilitate comparison of the current study, with the recent literature, the term SLGN will be used to refer to all of the branches supplying the superolateral knee joint, regardless of whether they are accompanied by branches of SLGA or originating from the common fibular or sciatic nerves.

    The SLGN, in the current study, was found to bifurcate, superior to the lateral femoral condyle, consistent with Sutaria et al’s findings.15 The deep branch of SLGN, branching more proximally, was found to further divide into transverse and longitudinal branches. In all specimens, the transverse and longitudinal deep branches of SLGN,were accompanied by articular branches of the SLGA, previously described as the deep (condylar) branch.17 Morsy et al 17 reported that the deep condylar branch of SLGA consistently coursed anteriorly in close relation to the surface of the femur where it ramified to supply the lateral femoral condyle. This artery can be used as a surrogate marker of the SLGN18; however, visualization may be difficult in individuals with small or narrowed arteries secondary to atherosclerosis.

    The SMGN, in the current study, divided into deep and superficial branches as it coursed along the tendon of the adductor magnus. The presence of the more proximal deep branch of SMGN, which has been discussed in recent publications,6 7 was found in all specimens. In the current study, the deep branch of SMGN coursed to the periosteal level where it divided into transverse and longitudinal branches. The transverse and longitudinal deep branches were accompanied by the previously described superior genicular transverse and deep branch of the superior medial genicular arteries, respectively.19 The superior genicular transverse artery has previously been used as a surrogate marker for the SMGN,18 although the application is limited due to difficulty with vessel visibility in some patients.

    The course of articular branches of LBNVI and MBNVI, as reported in the current study, is consistant with previous literature. Both Gardner and Sakamoto et al 20 described the articular branches of NVI as coursing inferiorly on the anterior surface of the femur.14 However, Gardner found the branches terminated at the margin of the articular cartilage of the femur, and Sakamoto et al 20 found these terminated at the suprapatellar bursa. In Gardner's illustration of the nerve supply to the knee joint, an articular branch of NVI was depicted coursing along the anterolateral surface of the femur consistent with the course of LBNVI as reported in the current study.

    Recent literature suggested the need to optimize the current landmarks for genicular nerve ablation proposed by Choi et al.5–7 10 21 There have also been several publications describing new landmarks for genicular nerve ablation.5 22–24 To evaluate the potential clinical relevance of these publications, it is imperative to determine (1) the articular branches captured by the classical techniques, (2) the anatomical basis supporting the new landmarks and (3) the required number of articular branches captured to produce optimal clinical results. In the current study, the classical superolateral and superomedial landmarking techniques were found to capture the transverse deep branches of the SLGN and SMGN, respectively. Interestingly, the classical landmarks also captured the posterior branches of lateral and medial nerves to the vastus intermedius. The articular branches captured by the classical RF techniques, as reported in the current study, provide the anatomical basis supporting the analgesic effectiveness of current landmarks1–4; however, increasing nerve capture rates will potentially lead to better clinical outcomes. In recent publications, the relocation of classical landmarks to more distal and posterior positions has been proposed.5 22 24 Based on our current study, these relocated landmarks will likely capture the distal branches of SMGN and SLGN, consistent with previous publications.5 22 24 25 However, due to the extensive innervation of the knee joint,5 10 12–14 21 relocating a limited size RF lesion will inevitably result in the capture of different articular branches while sparing those captured using classical landmarks.6 This is supported by anatomical findings of the current study. Based on the position of the LBNVI and MBNVI, coursing along the anterolateral and anteromedial aspects of the femoral shaft, these articular nerves will likely not be captured with relocation of the lesion more distal and posteriorly. The sparing of the articular branches of LBNVI and MBNVI may be of significance as Gardner previously described these branches coursing to the margin of the articular cartilage of the femur.14 Furthermore, the transverse deep branches of SLGN and SMGN, as reported in the current study, may also be spared. The clinical outcomes and benefits of targeting specific articular branches are unknown and remains to be investigated; however, it is conceivable that greater nerve capture will likely improve the analgesic effect. Therefore, supplementing classical landmarks, which have reported positive clinical outcomes,1–4 with additional anatomically validated landmarks could potentially lead to better outcomes than relocation alone. Alternatively, to accomodate possible anatomical variation, a palisade or stripe lesion may be used to improve capture rates.10 26 The effectiveness of supplementary landmarks and palisade techniques require further anatomical and clinical investigations. Finally, the number of articular branches needed to be captured, for optimal clinical outcomes, remains unclear and further clinical investigation is necessary.

    Limitations of this study include a small sample size, and thus possible anatomical variations may not have been documented. Further, the simulated lesion volumes are based on available literature determined using animal tissue27 28 and therefore may not replicate lesion volumes in vivo. Finally, this is an anatomical study and further clinical investigation is required to determine analgesic effectiveness.

    In conclusion, this study provides anatomical evidence supporting the effectiveness of classical landmarking for genicular nerve ablation. The capture of the transverse deep branches of SLGN/SMGN, articular branches of MBNVI/LBNVI and anterior branch of IMGN supports the effectiveness of genicular nerve ablation using classical landmarks as reported in the clinical literature. The extensive innervation of the knee joint suggests the use of supplementary landmarks to improve nerve capture rates and potentially patient outcomes—pending further clinical investigation.

    Acknowledgments

    The authors thank Ian Bell and Benjamin Kozlowski for their valuable technical assistance. We also thank the individuals who donated their bodies and tissue for the advancement of education and research.

    References

    1. ↵
      2. Choi W-J ,
      3. Hwang S-J ,
      4. Song J-G , et al
      . Radiofrequency treatment relieves chronic knee osteoarthritis pain: a double-blind randomized controlled trial. Pain 2011;152:481–7.doi:10.1016/j.pain.2010.09.029 pmid:http://www.ncbi.nlm.nih.gov/pubmed/21055873
      OpenUrlCrossRefPubMed
    2. ↵
      2. McCormick ZL ,
      3. Korn M ,
      4. Reddy R , et al
      . Cooled radiofrequency ablation of the Genicular nerves for chronic pain due to knee osteoarthritis: six-month outcomes. Pain Med 2017;18:1631–41.doi:10.1093/pm/pnx069 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28431129
      OpenUrlPubMed
    3. ↵
      2. Davis T ,
      3. Loudermilk E ,
      4. DePalma M , et al
      . Prospective, multicenter, randomized, crossover clinical trial comparing the safety and effectiveness of cooled radiofrequency ablation with corticosteroid injection in the management of knee pain from osteoarthritis. Reg Anesth Pain Med 2018;43:84–91.doi:10.1097/AAP.0000000000000690 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29095245
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. El-Hakeim EH ,
      3. Elawamy A ,
      4. Kamel EZ , et al
      . Fluoroscopic guided radiofrequency of genicular nerves for pain alleviation in chronic knee osteoarthritis: a single-blind randomized controlled trial. Pain Physician 2018;21:169–77.pmid:http://www.ncbi.nlm.nih.gov/pubmed/29565947
      OpenUrlPubMed
    5. ↵
      2. Fonkoue L ,
      3. Behets CW ,
      4. Steyaert A , et al
      . Accuracy of fluoroscopic-guided genicular nerve blockade: a need for revisiting anatomical landmarks. Reg Anesth Pain Med 2019;44:950–8.doi:10.1136/rapm-2019-100451 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31451628
      OpenUrlAbstract/FREE Full Text
    6. ↵
      2. Tran J ,
      3. Agur A ,
      4. Peng P
      . Revisiting the anatomical evidence supporting the classical landmark of genicular nerve ablation. Reg Anesth Pain Med 2020;45:393–4.doi:10.1136/rapm-2019-101024 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31806682
      OpenUrlFREE Full Text
    7. ↵
      2. Fonkoue L ,
      3. Behets CW ,
      4. Steyaert A , et al
      . Anatomical evidence supporting the revision of classical landmarks for genicular nerve ablation. Reg Anesth Pain Med 2020;45:672–3.doi:10.1136/rapm-2019-101103 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31806680
      OpenUrlFREE Full Text
    8. ↵
      2. Vanneste B ,
      3. Tomlinson J ,
      4. Desmet M , et al
      . Feasibility of an ultrasound-guided approach to radiofrequency ablation of the superolateral, superomedial and inferomedial genicular nerves: a cadaveric study. Reg Anesth Pain Med 2019;44:966–70.doi:10.1136/rapm-2019-100381 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31409664
      OpenUrlAbstract/FREE Full Text
    9. ↵
      2. Tran J ,
      3. Peng P ,
      4. Agur A
      . Evaluation of suprascapular nerve radiofrequency ablation protocols: 3D cadaveric needle placement study. Reg Anesth Pain Med 2019;44:1021–5.doi:10.1136/rapm-2019-100739 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31527162
      OpenUrlAbstract/FREE Full Text
    10. ↵
      2. Tran J ,
      3. Peng PWH ,
      4. Lam K , et al
      . Anatomical study of the innervation of anterior knee joint capsule: implication for image-guided intervention. Reg Anesth Pain Med 2018;43:407–14.doi:10.1097/AAP.0000000000000778 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29557887
      OpenUrlAbstract/FREE Full Text
    11. ↵
      2. Tran J ,
      3. Peng PWH ,
      4. Gofeld M , et al
      . Anatomical study of the innervation of posterior knee joint capsule: implication for image-guided intervention. Reg Anesth Pain Med 2019;44:234–8.doi:10.1136/rapm-2018-000015
      OpenUrlAbstract/FREE Full Text
    12. ↵
      2. Kennedy JC ,
      3. Alexander IJ ,
      4. Hayes KC
      . Nerve supply of the human knee and its functional importance. Am J Sports Med 1982;10:329–35.doi:10.1177/036354658201000601 pmid:http://www.ncbi.nlm.nih.gov/pubmed/6897495
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      2. Horner G ,
      3. Dellon AL
      . Innervation of the human knee joint and implications for surgery. Clin Orthop Relat Res 1994:221–6.doi:10.1097/00003086-199404000-00034 pmid:http://www.ncbi.nlm.nih.gov/pubmed/8156678
    14. ↵
      2. Gardner E
      . The innervation of the knee joint. Anat Rec 1948;101:109–30.doi:10.1002/ar.1091010111 pmid:http://www.ncbi.nlm.nih.gov/pubmed/18915634
      OpenUrlCrossRefPubMed
    15. ↵
      2. Sutaria RG ,
      3. Lee SW ,
      4. Kim SY , et al
      . Localization of the lateral retinacular nerve for diagnostic and therapeutic nerve block for lateral knee pain: a cadaveric study. PM R 2017;9:149–53.doi:10.1016/j.pmrj.2016.06.017 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27346091
      OpenUrlPubMed
    16. ↵
      2. Fonkoué L ,
      3. Behets C ,
      4. Kouassi Jean-Éric K , et al
      . Distribution of sensory nerves supplying the knee joint capsule and implications for genicular blockade and radiofrequency ablation: an anatomical study. Surg Radiol Anat 2019;41:1461–71.doi:10.1007/s00276-019-02291-y pmid:http://www.ncbi.nlm.nih.gov/pubmed/31338537
      OpenUrlPubMed
    17. ↵
      2. Morsy M ,
      3. Sur YJ ,
      4. Akdag O , et al
      . Anatomic and high-resolution computed tomographic angiography study of the lateral femoral condyle flap: implications for surgical dissection. J Plast Reconstr Aesthet Surg 2018;71:33–43.doi:10.1016/j.bjps.2017.08.012 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28882491
      OpenUrlPubMed
    18. ↵
      2. Kim D-H ,
      3. Choi S-S ,
      4. Yoon S-H , et al
      . Ultrasound-Guided Genicular nerve block for knee osteoarthritis: a double-blind, randomized controlled trial of local anesthetic alone or in combination with corticosteroid. Pain Physician 2018;21:41–52.pmid:http://www.ncbi.nlm.nih.gov/pubmed/29357330
      OpenUrlPubMed
    19. ↵
      2. Sung-Yoon W ,
      3. Lee Y-I ,
      4. Hun-Mu Y
      . Anatomy of superior medial genicular artery: a short cadaveric report. Clin Anat 2016;29:798–9.doi:10.1002/ca.22707 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26940044
      OpenUrlPubMed
    20. ↵
      2. Sakamoto J ,
      3. Manabe Y ,
      4. Oyamada J , et al
      . Anatomical study of the articular branches innervated the hip and knee joint with reference to mechanism of referral pain in hip joint disease patients. Clin Anat 2018;31:705–9.doi:10.1002/ca.23077 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29577432
      OpenUrlPubMed
    21. ↵
      2. Franco CD ,
      3. Buvanendran A ,
      4. Petersohn JD , et al
      . Innervation of the anterior capsule of the human knee: implications for radiofrequency ablation. Reg Anesth Pain Med 2015;40:363–8.doi:10.1097/AAP.0000000000000269 pmid:http://www.ncbi.nlm.nih.gov/pubmed/26066383
      OpenUrlAbstract/FREE Full Text
    22. ↵
      2. Conger A ,
      3. Cushman DM ,
      4. Walker K , et al
      . A novel technical protocol for improved capture of the Genicular nerves by radiofrequency ablation. Pain Med 2019;20:2208–12.doi:10.1093/pm/pnz124 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31131850
      OpenUrlPubMed
    23. ↵
      2. Lash D ,
      3. Frantz E ,
      4. Hurdle MF
      . Ultrasound-Guided cooled radiofrequency ablation of the genicular nerves: a technique paper. Pain Manag 2020;10:147–57.doi:10.2217/pmt-2019-0067 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32129148
      OpenUrlPubMed
    24. ↵
      2. Roberts SL ,
      3. Stout A ,
      4. Dreyfuss P
      . Review of knee joint innervation: implications for diagnostic blocks and radiofrequency ablation. Pain Med 2020;21:922–38.doi:10.1093/pm/pnz189 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31407791
      OpenUrlPubMed
    25. ↵
      2. Yasar E ,
      3. Kesikburun S ,
      4. Kılıç C , et al
      . Accuracy of ultrasound-guided genicular nerve block: a cadaveric study. Pain Physician 2015;18:E899–904.pmid:http://www.ncbi.nlm.nih.gov/pubmed/26431143
      OpenUrlPubMed
    26. ↵
      2. Tran J ,
      3. Peng P
      . Hip and Knee Joint Denervation. In: Peng P , Finlayson R , Lee S , et al. , eds. Ultrasound for interventional pain management. Cham: Springer, 2020.
    27. ↵
      2. Organ L
      . Cervical facet joint medial branch rf neurotomy. 8th ed. Toronto: DIROS OWL, 2017: 32–3 ..
    28. ↵
      2. Cosman ER ,
      3. Dolensky JR ,
      4. Hoffman RA
      . Factors that affect radiofrequency heat lesion size. Pain Med 2014;15:2020–36.doi:10.1111/pme.12566 pmid:http://www.ncbi.nlm.nih.gov/pubmed/25312825
      OpenUrlPubMed

    Footnotes

    • Presented at Interim data from this work was presented at the 2020 American Association of Clinical Anatomists Annual Meeting, 15–19 June 2020.

    • Contributors JT, PP, AA contributed to the experimental design, data acquisition, analysis of data, drafting and revising the manuscript critically for important intellectual content.

    • 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 PP received equipment support from Sonosite Fujifilm Canada. AA is an Anatomy Faculty with Allergan Academy of Excellence.

    • Patient consent for publication Not required.

    • Ethics approval This cadaveric study was approved by the University of Toronto Health Sciences Research Ethics Board (#27210).

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data availability statement All data relevant to the study are included in the article or uploaded as supplementary information. All data relevant to the study are included in the article.