Knee Pain Examined under Musculoskeletal Ultrasonography: Experience from a Community-based Rheumatology Practice – A Cross-sectional Study

2.1. MSUS Examination of Knee and Terminology used to Define the Structural Changes

Patients lie supine on the examination table with the knee flexed at 20-30 degrees [22]. The anterior, medial, lateral, and posterior knee was examined on the short and longitudinal axis using a high-frequency linear probe (Samsung Accuvix 2010, Probe 12A, South Korea). Smooth probe gliding without any risk of slippage using the 'afternoon tea' technique was followed [23].

Echogenicity: The echogenicity of the tissue refers to the ability to reflect or transmit US waves in the context of surrounding tissues [23, 24]. Anechoic means no transmission of waves, for example, in cartilage, bone, and fluid-filled structures (Baker's cyst, BC, vessels). Fat is almost anechoic, while the fascia and other connective tissue strands and fascicles appear as hyperechoic lines [24]. Muscles are hypoechoic with striate structures. Proximal nerves are hypo-anechoic, and the distal nerves are hyperechoic, with a stippled (“honeycomb”) structure (with hypo-anechoic fascicles on the hyperechoic background of connective tissue surrounding them). Ligaments and tendons are hyperechoic with a fibrillar pattern but not a “honeycomb” like a nerve [24, 25]. Acoustic shadow typically arises from deep to strong hyperechoic reflectors, for example, cortical bone, larger foreign bodies, and calcifications with posterior acoustic shadowing. Enhanced through transmission deep to a fluid-filled structure occurs due to weak sound wave attenuation within a simple fluid (for example, Baker's cyst), resulting in a localized area of increased echo posteriorly to the interface (posterior acoustic enhancement) [26].

Anisotropy and how to improve it?: The anisotropic artifact refers to a darkening and loss of resolution of the image, which occurs when the approach of the sound wave is less than perpendicular (i.e., angle of incidence greater than 0 degrees) [24]. At the insertion of tendon and ligament attachments or course of the tendon, the structures are usually anisotropic; however, changing the angle of incidence with toggling, heel-toe effects of the probe can improve the anisotropy [24, 26].

Doppler activity (DA): Color Doppler (CD) distinguishes blood moving within vessels. When the angle of incidence is close to 90°C, and the flow is low, there will be no color on the screen, producing a false negative indication of no flow; in a situation like in small flow, power Doppler (PD) is more sensitive [24, 25]. Increased DA within the lesion helps differentiate inflammatory from non-inflammatory and monitor treatment outcomes with anti-inflammatory [26].

Anechoic compressible versus non-compressible pathology: joint effusion is usually anechoic and compressible and is found in the distended bursa, knee joint recess, and can be successfully differentiated from non-compressible anechoic structure, for example, ganglion cyst and articular cartilage [27, 28]. Veins are more compressible than an artery with a Doppler effect [29].

Knee regions scanned: Anteriorly, the quadricep's tendon (QT), supra-patellar recess, pre-femoral fat pads, patella, patellar tendon, pre-patellar bursa, infra-patellar bursa, and patellar retinacula were examined for any loss of tendon integrity, alteration of thickness, effusion in the bursa and recess, loose body in the cavity, synovial thickening, hypertrophy, and synovial plica [30, 31]. In full knee flexion, femoral trochlea and medial femoral condylar cartilage smoothness, echogenicity, and thickness are best reviewed [22].

The distal iliotibial band, lateral collateral ligament (LCL), popliteal tendon, common peroneal nerve (CPN), fibular head, and meniscus were laterally scanned. Medial scanning revealed medial collateral ligament (MCL), pes anserinus tendons bursa, and medial meniscus pathology. Valgus and varus stresses were helpful in examining the integrity and laxity of LCL and MCL, respectively [22].

Posteriorly, the patient was examined in a prone position and the knee extended to access the dynamic fat-filled popliteal fossa. First, sartorius, gracilis, and semitendinosus (SMT) were scanned medially. Then, assess whether the semimembranosus gastrocnemius bursa and Baker's cyst (BC) arise between the medial gastrocnemius muscle and SMT tendons. At SMT insertion, the anechoic, compressible collection was checked around the tendon. Postero-laterally, the biceps femoris was checked at insertion into the fibular head; and CPN compressing around the fibular neck was checked [22].

2.2. Outcome Measures

There were two types of outcome measures documented in the study: IA and EA pathologies [1, 22, 30-43].

Intra-articular (IA): The anterior joint recess usually collects uncomplicated anechoic compressible synovial effusions. The anechoic, compressible simple collection is also seen in BC. Sometimes, complicated and heterogenous effusion is seen in infection and post-traumatic hemorrhagic arthropathy [33, 34]. Mobile or impacted loose bodies can also be found in the anterior and posterior joint recesses and BC [33, 35].

In inflammatory arthritis (InA), hypoechoic and thickened synovial membranes with high DA due to hyperemia can be seen in the anterior joint recess and BC [36]. Frond-like synovial mass is characteristic of lipoma arborescence [35]. Anterior knee pain may be associated with synovial plica [30, 36].

Articular cartilage is usually smooth, non-compressible, and anechoic [37]. In early KOA, CPPD crystals depositing within the cartilage make it hyperechoic [37]; however, advanced KOA includes reduced articular cartilage thickness (<2.2 mm) [37]. Podlipská et al. describe the MSUS-based qualitative grading for cartilage regeneration in knee OA: grade-0, normal (smooth, non-compressible, anechoic cartilage); grade I (increased echogenicity of articular cartilage); grade IIa (grade-I plus reduced cartilage thickness < 50%); grade-IIb (grade-IIa plus reduced cartilage thickness > 50%); Grade-III (grade-IIb plus reduced cartilage thickness 100%) [37]. Depositing MSU crystals over the articular cartilage forms DCS, diagnostic for gout [1]; however, DCS may confuse the cartilage interface signs. Hyperechoic DCS, chondrocalcinosis, and subchondral bone creating a ‘triple contour sign’ may be revealed in the OA knee [38].

Medial meniscal (MM) tear is common in clinical practice, and patients report medial knee pain with a history of preceding knee injury. In the MM tear, the explicit hypoechoic cleft is seen. Sometimes, MM tear may be complicated with an anechoic, compressible cyst, and hyperechoic calcification [39]. When MM protrudes two mm beyond the joint line, it is called MME and is significantly associated with knee pain in OA [40].

Extra-articular (EA): Fluid-filled anechoic cystic lesions with or without increased wall DA are seen before the patella (prepatellar bursitis) and before and after the patellar tendon (superficial and deep infra-patellar bursitis) [41]. SMT bursitis signifies anechoic, compressible collection around the fibrillar SMT tendon [42]. Superimposed infection may complicate bursal collection [41]. Osgood-Schallater disease (OSD) and Sinding Larsen Johansen syndrome (SLJS) are seen in children [32]; however, describing them is beyond the scope of the paper. MCL injury is associated with MM tear and anterior cruciate ligament (ACL) injury, an ‘unhappy sign’ [42, 43].

2.3. Statistics

A prefabricated data sheet was used to document clinical and US data. Numerical data were presented with frequency, mean, and standard deviation (SD), and categorical data were reported in frequency and percentage. As the questionnaire included information on categorical variables and the data were not distributed normally, the Mann-Whitney test was used to assess the correlation between independent categorical variables. Statistical Package for the Social Sciences (SPSS, v.28.0.1.1) was used for data analysis. Tables 1-5 and Fig. (1a-f and 2a-f) are used to present data. SPSS was set to mark missing values with a roman number in the list. There was a number everywhere in the listing with a blank in the data.

4.1. Reduced Cartilage Thickness, Osteophyte Formation, and Effusion in OA

In the present study, we depicted articular cartilage increased echogenicity and thickness, reduced cartilage thickness, MME, osteophyte formation, and synovial effusion in KOA. Podlipska et al. documented US-depicted IA structural changes in KOA, and the most common pathologies were reduced cartilage thickness, osteophyte formation, and MME [37]; global femoral cartilage degeneration grade was strongly associated with pain, and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) index [37]. The presence of lateral femoral osteophytes (irrespective of size) was also associated with increased pain and WOMAC index. All osteophyte grades in the MFC were associated with worsened WOMAC clinical symptoms [37].

The ZAGAZIG, US scale constitutes a validated tool of five domains [46]: severity of KOA (depending on the osteophyte severity grade, 0-4), effusion (0-3), grey-scale synovitis (0-3), pes anserine tendonitis/bursitis (0-2), and BC (0-2) [46]. A recent cross-sectional observational study that included 245 KOA was assessed with the ZAGAZIG-US tool, and positive correlations were depicted between the ZAGAZIG score and WOMAC subscales (pain, stiffness, and function) [46]. The ZAGAZIG scale and VAS pain and Health Assessment Questionnaire-II (HAQ-II) had strong positive correlations [46]. The total US score can distinguish between mild and moderate-severe KOA [46-48]. Another study with KOA depicted a strong correlation between the radiographic medial tibiofemoral (TFJ) narrowing and the US-MFC grade (p< 0.0001) [49]; radiographic and US-femoral osteophyte (medial and lateral) (p< 0.0001) [49]. The US could delineate osteophytes in EOA where X-ray was negative for osteophytes [49, 50]. The correlation of cartilage thickness severity between US and arthroscopy varied from insignificant to significant depending on the site involved: at the sulcus - the correlation was highest (p<0.001); at MFC was notable (p=0.003); at LFC, no significant correlation (p=0.103) [51].

In another study, Basha et al. reported US and MRI findings in anterior knee pain (AKP); joint effusion was the most common (38%), followed by trochlear cartilage defect (20.6%) and superficial infrapatellar subcutaneous edema (20%) [52]. Draghi et al. demonstrated the US's sensitivity and specificity of 81.3% and 100% specificity in knee effusion [41]. The US could approximate as low as 7 to 10 ml joint effusion [53]. US is a suitable alternative to MRI in AKP [52].

Mild-moderate radiographic KOA was examined under the dynamic US using a standardized OMERACT scanning protocol and assessed clinical severity (numerical rating scale for pain and Knee Injury, Osteoarthritis Outcome Score (KOOS), radiographic OA severity (KL score), and MRI Osteoarthritis Knee Score (MOAKS) of US pathologies [54]. All US scores (except for cartilage grade) demonstrated significant associations with worse KOOS symptoms, whereas only PD and meniscal extrusion were associated with worse KOOS pain [54]. US pathologies revealed a moderately good correlation with their MOAKS counterparts. Articular cartilage calcification is seen with advanced age [55]. There is a positive correlation between cartilage calcification and grade 1 and 2 but not above 2 OA grades [55]. In the present study, reduced MFC cartilage thickness in the KOA was significantly associated with MME but not with clinical effusion, osteophyte size, or MM tear.

4.2. Synovial Membrane Increased Thickness, Increased Doppler Activity in Inflammatory Arthritis

High-resolution MSUS has been increasingly used in inflammatory arthritis in RA [56]. MSUS delineates synovitis and bone erosion, which correlate with clinical disease activity in RA [56]. A longitudinal pilot study explored synovitis in reactive arthritis (ReA) with MRI and US [57]. In an interesting study, 25 first-degree relatives (FDRs) of RA without clinical arthritis and negative for RF and ACPA were examined under MSUS, which depicted inflammatory synovial changes in 40% of cases [58]. Another large cohort study investigated 273 FDRs: 8% were positive for anti-CCP, 21% had unclassified arthritis (UA), and 35% had MSUS-depicted active synovitis [59]. US use in FDRs without clinical RA was not impressive, except in those with UA [59]. In a multicenter study, PD signal and morning stiffness (over thirty minutes) at baseline were independently associated with incident InA in clinically suspected arthralgia [60].

US and MRI are comparable regarding synovitis and tenosynovitis diagnosis [61]. US can predict synovial inflammation in preclinical RA with at-risk populations [62]. US-depicted joint synovial thickening, bone erosion, tendon sheath synovial thickening, tendon sheath, bone, and enthesis PD signal, and bone erosion were higher in PsA than in non-PsA patients (p<0.05) [63].

US is comparable to dual-energy CT (DECT) in gout involving the knee, as revealed in a prospective study [64]. DCS was considered the most valuable US sign in gout [64]. OMERACT-US could be a helpful tool to monitor urate-lowering therapy (ULT) in gout and monitor the significant change in DCS and tophus sum score at follow-up (p<0.05) [65]. The US also is valuable in CPPD diagnosis [66]. In the present study, we documented only four InA with synovial effusion and increased synovial thickening in the anterior joint recess. BC with high DA was also noted. In one patient with a previous diagnosis of gout, DCS was seen.

4.3. Medial Meniscal Extrusion (MME), MM Tear, Cysts (Meniscal and Para-meniscal), and Calcification

Clinically diagnosed early KOA was explored in the US to study MM dynamically [67]. MME > 2 mm can occur in 75% of early medial KOA cases [67]. Clinically suspected MM tears may be correctly diagnosed with US and MRI [68]. The sensitivity, specificity, and positive and negative predictive values of MM tears appeared to be the same for both US and MRI [68]. Regarding MM tear, US and MRI findings were comparable to arthroscopy [68]. Patients negative for US-MM tear should not go for MRI [68].

In an observational study with chronic knee pain, the US showed excellent sensitivity (95% and 96% for each reader) and good specificity (82% and 70% for each reader) in MME diagnosis [69]. US is useful in monitoring cartilage thickness and MME pre-and post-IA interventions, as documented in a study with unilateral KOA [70]. In the present study, we did not quantify the MME and MFC thickness; instead, we confirmed whether MME was present. MM extrusions beyond 2 mm registered as MME; we further measured MFC cartilage thickness in the study participants using the Podlipski technique [37]: out of 34, the maximum 73.5% had the grade IIa degeneration, and only one had a severe decline of articular cartilage thickness. We did not document IA cartilage thickness and MME changes after IA injection, and it was not our study aim.

A lateral meniscal cyst could cause knee pain with snapping [71]. The MRI and US-depicted well-circumscribed cyst can be revealed [71, 72]. Para-meniscal cyst of MM with medial knee pain may cause popliteal artery compression that can be relieved with US-guided aspiration [72]. In the present study, we documented MM cysts and MM calcification in 4.3% and 2.2% of cases of KOA, respectively.

Meniscal calcifications have been revealed in human cadaveric menisci with high-spatial-resolution radiography and 3.0-T MR (ultrashort echo time, UTE-MR) imaging and classified as punctate, linear, or spherical and could be pathognomonic in OA [73], could be meniscal a new target for OA [74]. We noted MCL calcification in all cases with MCL strain.

4.4. Synovial Chondromatosis or Loose Body

Synovial chondromatosis was first described by Leannac in 1813 [74]; however, Jaffe's current description was applied in 1958 [74]. An identical process can also involve the synovium extending along with tendons and bursae and is referred to as tenosynovial and bursal chondromatosis, respectively [74]. Secondary synovial chondromatosis is associated with trauma, osteochondritis dissecans, neuropathic osteoarthropathy, OA, and septic or InA. Primary synovial chondromatosis typically affects the adult knee [74]. In the present study, we found the loose body in six patients, and all were in OA knee: four were found anteriorly, and two were in the knee back.

Clinical symptoms typically include pain, swelling, and restricted joint ROM [74]. Radiographic features include multiple IA chondral bodies with “ring-and-arc” chondroid mineralization and extrinsic erosion of bone on both sides of the joint [74]. Radiology is normal in 5%–30% of cases [73]. Limited case reports described the US features of synovial chondromatosis - heterogeneous mass containing foci of hyperechogenicity with posterior acoustic shadowing being seen in the joint recess, bursa, tendon sheath, and synovium, either impacted or loose with the change of position during dynamic US examination [74]. Arthroscopic treatment is successful with low recurrence rates [75]. We managed synovial chondromatosis with IA steroid and lidocaine injection and referred the patient for an orthopedists' opinion and further action.

4.5. Synovial Plica and Lipoma Arborescence

Synovial plica of the knee is common in clinical practice. Patients usually report intermittent pain, swelling, and snapping sensation affecting the knees, related to increased patella-femoral joint loading [30]. MRI reveals plica as bands of low signal intensity within the high-signal-intensity synovial fluid; however, the medial plica is challenging to see with effusion, an important cause of medial knee pain [76].

In suspected medial plica, dynamic sonography had diagnostic accuracy, sensitivity, and specificity of 88%, 90%, and 83%, respectively [77, 78]. The US is useful in detecting supra-patellar and all medial patellar plicae but not infra-patellar plicae [52]. In the present study, we defined synovial plica in over 70% of patients. The correlation between synovial plica and US-depicted anterior recess effusion was significant (p<0.05).

Lipoma arborescence is a rare, benign IA lesion typically of the knee joint and is primary and secondary to OA, RA, trauma, or diabetes mellitus [35]. Frond-like, pliable, hyperechoic lipoma arborescence can be seen in US in the supra-patellar recess with or without compressible anechoic effusion and increased DA [35, 79]. In the present study, lipoma arborescence was noted in five patients with (four KOA and one SpA); we treated them with IA methyl-prednisolone. We did not examine them on histopathology.

4.6. Baker’s Cyst

A BC is an abnormal fluid distension in the gastrocnemius semimembranosus (G-S) knee bursa [80]. G-S bursa is located in the medial aspect of the popliteal fossa of the knee, at the intersection between the semimembranosus tendon (SMT) and the tendon of the medial head of the gastrocnemius (MHG). BC is found in about 4% to 7% of asymptomatic adults, 38% of patients with internal knee derangements (IKDs), and 22% to 47% with KOA. Like MRI, the US also has a diagnostic value for BC [80].

A BC may have a broad base than the body, neck between the MHG and SMT, and septa within the cyst [81]. BC is further complicated by ruptured walls and the dissemination of their contents around [80, 82]. Giant BC could find in InA [81, 83]. In this study, we documented BC in 33 patients; however, BC was not significantly correlated with anterior recess effusion, synovial plica, reduced cartilage thickness, increased cartilage echogenicity, or increased MCL thickness. We did not find any ruptured cysts. BC was approached with a 20G needle for aspiration, followed by IL methyl-prednisolone and lidocaine injection.

4.7. Bursitis

The serous bursae consist of a synovial membrane enveloping a fluid film [84]. Bursae are native and adventitious - native bursae are lined with a synovial membrane and present at predictable anatomical sites; they are easily distinguished from pathological entities [85]. The adventitious bursa forms at friction sites but with similar US features to native bursae (unilocular, simple, compressible, and anechoeic) [85]. Unlike bursal cysts, ganglion cysts are usually multilocular and non-compressible [85].

A variety of bursae are found around the knee: anteriorly (supra-patellar, pre-patellar, superficial infra-patellar, and deep infra-patellar), medially (pes anserine, medial collateral ligament, and posteriorly (semimembranosus–medial collateral ligament bursae, BC), laterally (lateral compartment Iliotibial and lateral collateral ligament–biceps femoris bursae). Overuse, trauma, infection, bleeding, arthropathy, IKDs, villonodular synovitis, and synovial osteochondromatosis contribute to bursal inflammation [84, 85]. Draghi et al. depicted the diagnostic value of the US as compared to MRI [41]. The diagnostic sensitivity and specificity of US for supra-patellar bursitis are 71.4% and 100%, respectively [41]. Anechoic, compressible, septated bursitis with or without increased DA around the SMT may sometimes be confused with BC [86]. The US accurately delineates the anserine bursa, including its anatomical variations [87]. We documented bursitis in eight patients in the present study: six SMT bursitis and two anserine bursitides. We had no pre-patellar and infra-patellar (superficial and deep) bursitis. We successfully managed all of them with IL methylprednisolone and lidocaine injection.

Study limitation: the study was conducted in a single community-based rheumatology facility with a limited number of cases. Only one physician with expertise in MSUS was involved in the patient examination, so we could not check the inter-observer variability of the diagnosis.