Metatarsalgia is a common problem in the general population that significantly reduces independence and quality of life, and it is also a known risk factor for locomotor disability, impaired balance, and increased risk of falls [1,2]. Various etiologies can present similar symptoms and similar physical examination findings, but they must be distinguished from each other because the treatment can be different [3-5]. As the clinical diagnosis based on the symptoms and physical examination can be misleading, ultrasonography (US) can be a problem-solving tool to make the correct diagnosis [6]. However, without knowing the differential diagnoses and the structures to evaluate, the radiologist might need to invest a considerable amount of time and effort to find the true pathology. Therefore, the purpose of this article is to review the possible etiologies of central metatarsalgia and their US findings to aid radiologists in performing effective and efficient US examinations.

Common causes of metatarsalgia are Morton’s neuroma, intermetatarsal bursitis, plantar plate injury, submetatarsal bursitis, synovitis of metatarsophalangeal (MTP) joint, and metatarsal fractures [7,8]. As there can be multiple pain sources, the forefoot should be evaluated with US according to each possible location of the disease. A suggested checklist is provided in the Table 1. Furthermore, careful history-taking and a basic radiographic evaluation should be done before US imaging to guide the examiners to focalize the US examination [8].

The US examination to evaluate the forefoot can be done with the patient in either the supine or sitting position according to the examiner’s preference. When the patient’s knee is flexed and feet flat on the bed, the dorsal side of the foot can be easily examined. When the patient’s knee is extended and leg rests on the bed, both the plantar and dorsal surfaces of the foot can be evaluated without changing the position. This position also allows the most dynamic assessments and facilitates an easy comparison with the contralateral foot.

Morton's neuroma is a common problem of the foot that presents metatarsalgia [9,10]. Histologically, it is a neuroma-bursal complex, which consists of an enlarged nerve with degeneration, perineural scarring, tangled vessels, and scarred intermetatarsal bursa around the nerve [11-13]. The main symptoms are burning pain in the plantar aspect of the forefoot and a feeling of fullness, or "walking on a pebble." Patients sometimes complain of tingling, numbness, or radiating pain in the affected toes [3,9,10] (Fig. 1). Entrapment, ischemia, overuse, and repetitive trauma have been suggested as possible causes of Morton’s neuroma, and variations in the anatomy of the forefoot may contribute to the development of symptoms [9,10]. Morton’s neuroma most commonly affects the third web space, followed by the second web space [14-17], because the spaces are narrow in those regions [18].

US has been considered as an accurate modality to identify Morton's neuroma [19,20]. The patient can be examined with either a dorsal or plantar approach depending on the examiner’s preference. First, in the dorsal approach, the intermetatarsal web space can be scanned with plantar flexion of the toes to visualize neuromas in the intermetatarsal space more superficially [21,22]. Second, in the plantar approach with dorsiflexion of the toes, the plantar surface of the foot can be scanned with pressure on the dorsal intermetatarsal spaces [17,23]. Compressing the opposite side of the intermetatarsal space helps to splay the metatarsal heads and increases the diagnostic accuracy for finding masses in this region [17,22,23]. The authors prefer the plantar approach because of the wider field of view from the intermetatarsal space to the MTP joint/plantar plate, which helps to differentiate Morton’s neuroma from plantar plate injury.

On US, Morton’s neuroma appears as a hypoechoic mass located in the intermetatarsal space at the level of metatarsal heads [21,23] (Fig. 2). The long-axis view often allows the identification of continuity between the common plantar digital nerve and the neuroma. Detection of the plantar digital nerve entering the neuroma can raise the diagnostic accuracy for Morton's neuroma relative to other intermetatarsal masses such as complicated ganglion cysts or tendon sheath fibromas (Fig. 3).

Dynamic US originating from Mulder’s clinical test can be done by the examiner squeezing or clasping the forefoot with his or her unused hand [24] (Fig. 4A). While applying some pressure at the plantar surface of intermetatarsal space on the short-axis view, the examiner can observe a hypoechoic mass, compressed between the metatarsal heads (Fig. 4B). When the examiner squeezes the forefoot further, with mild release of the pressure given by the transducer at the plantar surface, the mass is displaced superficially from the intermetatarsal space to the plantar aspect [24], resembling the shape of a gingko leaf [25] (Fig. 4C, Video clip 1). In patients with symptomatic neuroma, this maneuver may reproduce the characteristic shocking pain with a clicking sensation. This is called the "sonographic Mulder sign" [24].

The transverse dimension of the neuroma is more important to report than the longitudinal dimension because a lesion that is larger than 5 mm in the transverse dimension is usually considered symptomatic [21,26]. However, since several studies have found symptomatic neuromas smaller than 5 mm [27-29], it is important to search carefully and to apply a proper dynamic study with correlation of typical symptoms in small lesions. Meanwhile, asymptomatic neuromas are also known to be commonly found in adults on imaging [16,26]. Therefore, the diagnosis of Morton’s neuroma should be made with caution in asymptomatic patients.

Intermetatarsal bursitis is a potential cause of metatarsalgia that is often underdiagnosed or misdiagnosed [30]. It develops mostly due to blunt trauma, but sometimes it can be the first manifestation of inflammatory arthritis such as rheumatoid arthritis (RA) or gout [30,31]. Moreover, it is known that intermetatarsal bursitis contributes to the development of Morton’s neuroma [32].

The intermetatarsal bursa is located in each intermetatarsal space at the dorsal side of the deep transverse intermetatarsal ligament (DTML) and common plantar digital nerve. Because the intermetatarsal bursa projects distally beyond the margin of the DTML at the second and third web space closely abutting the neurovascular bundle [32], intermetatarsal bursitis is often found simultaneously with and very close to Morton's neuroma (Fig. 5), making the neuroma seem oversized on imaging [13]. Dynamic examination with local pressure on the other side of the intermetatarsal space helps to differentiate bursitis from neuroma since the bursa changes in shape or collapses (Video clip 2), whereas the neuroma maintains its nodular shape and size [8].

On US, the intermetatarsal bursa is shown as a cystic lesion between the metatarsal heads and dorsal side of the DTML. Intermetatarsal bursitis appears as hypoechoic thickening of the bursal wall with or without associated effusion. The bursal distension up to 3 mm in transverse dimension is not uncommon in asymptomatic populations and has been considered as a physiologic change [26]. However, in a recent large cohort study of RA patients with magnetic resonance imaging (MRI), most patients had intermetatarsal bursitis measuring 3 mm or smaller in the transverse dimension, and a size larger than 3 mm worked as a very specific cut-off value for RA (specificity, 98%; sensitivity, 13%) [33]. The authors additionally investigated the dorsoplantar dimension of intermetatarsal bursitis, but also there was a large overlap below 15 mm between RA and normal controls, such that a dorsoplantar dimension of 15 mm or larger served as a very specific cutoff for RA (specificity, 100%; sensitivity, 16%). Therefore, in patients with prominent intermetatarsal bursitis (Fig. 6), a thorough evaluation is needed to find other features of inflammatory arthritis.

The plantar plate is a rectangular trapezoid shape fibrocartilaginous structure that is firmly attached to the base of the proximal phalanx and resists hyperextension of the MTP joint [34-36]. The plantar plate and joint capsule may become attenuated and injured because of hyperextension forces to the MTP joint of smaller toes. A plantar plate injury can cause forefoot pain and some degree of deformity, resulting in dorsomedial subluxation or crossover toes in the advanced stage [35].

Because the most common site for injury is the lateral aspect of distal plantar plate insertion at the second MTP joint, the pain related to plantar plate injury is usually localized around the second toe proximal phalangeal base [36] (Fig. 1). When the lateral aspect of the second distal plantar plate insertion is injured, the affected toe deviates medially, and splaying of the second and third toes occurs. Further injury of the plantar plate results in hyperextension of the MTP joint with subsequent subluxation and dislocation. The MTP drawer test or vertical stress test (Hamilton-Thompson test) permits the provocative clinical assessment of MTP joint instability and grading [34] (Fig. 7), but a low-grade injury might demonstrate little or no instability upon stress [37]. A plantar plate injury can also produce symptoms similar to Morton’s neuroma when there is reactive pericapsular soft tissue thickening (pericapsular fibrosis) pressing on the adjacent common plantar digital nerve or a concomitant Morton’s neuroma [38]. Plantar plate injuries should be distinguished from Morton’s neuroma, because a misdiagnosis would worsen the toe deformity and dysfunction.

Although MRI is accepted as the imaging modality of choice, US also shows a good correlation with the intraoperative finding of plantar plate injury [39-41]. On US long-axis view, the plantar plate normally appears as a homogeneously hyperechoic structure attached at the proximal phalangeal base [42]. It is important to tilt the transducer along with the distal insertion of the plantar plate to avoid the anisotropy artifact (Fig. 8) [43]. A hyperechoic triangular area can normally be seen in the central distal portion of the plantar plate [43]. A dynamic assessment with dorsiflexion of the toe can be performed to increase the sensitivity and accuracy of US for diagnosing plantar plate injuries [40,44].

In plantar plate injuries, the plantar plate can show a focal full or partial thickness defect, diffuse thinning, or complete absence (Fig. 9) [42]. On a dynamic study, the true defect can be identified as an anechoic or hypoechoic tissue gap with dorsiflexion of the toe, sometimes interposed with fluid or herniated fat tissue (Video clips 3, 4). Diffuse thinning of the plantar plate is considered significant when the distal plantar plate is thinner than its proximal or middle portion [42] (Video clip 5). When the plantar plate is completely torn and retracted, the flexor tendon can be found directly on the metatarsal head.

Known indirect signs of a plantar plate injury are the cartilage interface sign, flexor tendon/proximal phalangeal abnormalities, and pericapsular fibrosis. The flexor tendon can show subluxation into the torn gap of a plantar plate injury or medial subluxation from the central plantar plate groove [42]. Tenosynovitis can also occur if the synovial fluid passes through a full-thickness tear of the plantar plate [8]. Enthesophytes, cortical irregularities, and focal avulsion fragments at the proximal phalangeal base are also abnormalities known to be associated with plantar plate injuries [42,45]. Pericapsular fibrosis (pseudoneuroma) is a frequently noted reactive soft tissue lesion associated with plantar plate injuries [46]. As pseudoneuroma appears as a low-echoic soft tissue mass in the intermetatarsal space, it can mimic Morton’s neuroma on US. Because it abuts the lateral side of the MTP joint near the torn plantar plate [37], the key distinguishing features of pericapsular fibrosis are a more elongated lesion with a broad base at the MTP joint, eccentrically located in the second web space (Fig. 10). In comparison, Morton’s neuromas show more ovoid masses in continuity with the plantar digital nerve, concentrically located in the third web space [42,46]. Other previously described imaging findings, such as cortical irregularities at the base of the proximal phalangeal base, flexor tendon subluxation, tenosynovitis, and splaying of the second and third toes, can also be clues to help to differentiate plantar plate injuries from Morton’s neuroma [37].

Metatarsalgia may develop in patients with RA, gout, or other inflammatory processes in the MTP joints. On US, normal joints have smooth cortical surfaces, thin synovial recesses at the dorsal and plantar aspects of the small joints, and no vascularity on Doppler US [47]. When there is an increased amount of synovial-fluid complexes or synovial hypertrophy with intra-articular Doppler flow, active synovitis can be diagnosed on US [31].

High-resolution US is a more sensitive method to detect early RA in regard to synovitis and erosions in the bone when compared to clinical examinations or radiographs [48,49]. Every single small joint of the foot is well accessible by US, especially the first and fifth MTP joints and interphalangeal joints. Pressing a small-footprint hockeystick transducer into the interdigital space may help to evaluate regions with limited exposure for US examination, such as the radial and ulnar sides of the second, third, and fourth MTP joints [47]. Common findings in RA are synovitis, joint effusion, bone erosions, and tenosynovitis [47].

Synovitis can be diagnosed if there is abnormally hypertrophied synovial tissue that may exhibit Doppler signals, regardless of effusion [50] (Fig. 11). Bone erosion is an intraarticular discontinuity of bone surface, visible in two perpendicular planes [51]. A normal smooth, shallow depression at the dorsal edge of the metatarsal head should not be mistaken for erosion [47]. Because synovial inflammation and proliferation is the primary pathologic process, resulting in marginal erosions at "bare areas," no bone damage occurs in the absence of synovitis [52]. Therefore, a focal depression of bone that does not have synovitis adjacent to it is unlikely to be true erosion. Tenosynovitis can be diagnosed when there is abnormal anechoic/hypoechoic tendon sheath widening with abnormal tenosynovial fluid or synovial hypertrophy [51]. Plantar tenosynovitis is more frequently found in patients with early inflammatory arthritis, even without MTP joint synovitis, than in the healthy population [53].

Synovitis, small joint effusion, or tenosynovitis can also be seen in non-inflammatory arthropathies such as overuse or degenerative arthritis, and it is difficult to differentiate these conditions from those caused by early inflammatory arthritis. However, inflammatory arthritis has a typical location (e.g., the fifth MTP joint in RA) where it develops in its early stages, and it often involves the bilateral extremities. Furthermore, typical bone erosion or joint deformities can be clues in the chronic stage [54]. Another strength of US is the flexible field of view or site of examination. Other typical locations of inflammatory arthritis elsewhere in the body (e.g., the wrist joints in RA, knee joints in gout) are also readily accessible with US to increase the diagnostic confidence.

High-resolution US has also been recognized as a good imaging modality for crystal-induced arthropathies, including gout and calcium pyrophosphate dihydrate (CPPD) crystal deposition disease. Particularly in gouty arthritis, US plays a valuable role in the early diagnosis and monitoring treatment response [55]. In gouty arthritis, deposited monosodium urate (MSU) crystal on the surface of the hyaline cartilage shows another hyperechoic line added to the hyperechoic line formed by the subchondral bone, which is named the "double-contour sign" [56] (Fig. 12A). This sign should not be confused with the normal thin hyperechoic line at the surface of the cartilage in the area where the US beam is perpendicular to the cartilage surface [57]. MSU deposition at the cartilage surface is usually stippled, irregular in thickness, as bright as the subchondral bone, and visible on any angle independent of reflectivity of US [58]. The reported sensitivity of the double-contour sign for diagnosing gout ranges from 25% to 95% [59,60]. Effusion in gouty arthritis can be variably anechoic to heterogeneously hyperechoic (Fig. 12B). In some cases, hyperechoic aggregates float in the synovial recess with a gentle compression resulting in a "snow-storm appearance" [57]. Gout tophi can show various morphologies, from heterogeneous hypoechoic to hyperechoic nodular or amorphous lesions accumulated in the intra-/extra-articular bone or soft tissues (Fig. 12C), sometimes outlined by a hypoechoic halo that might correspond to the surrounding loose fibrovascular zone of tophi [61]. The bone erosions associated with gout may be closer to the diaphysis, deeper, and more irregular in shape with overhanging edges when compared to those in RA [47,62].

CPPD arthropathy is caused by the deposition of CPPD crystals in the articular and periarticular tissues. CPPD arthropathy commonly involves the knee, wrist, symphysis pubis, and hip, leaving the foot as a rare site [63]. Similar to gouty arthritis, CPPD arthropathy can demonstrate synovitis with echogenic aggregates or floating echogenic foci in the synovial fluid. Linear calcifications in the tendon or ligaments also can be found in CPPD arthropathy. However, unlike gout, CPPD crystals tend to accumulate in the middle layer of the cartilage, parallel to the bony cortex. They appear as an echogenic line or dotted foci inside the hypoechoic hyaline cartilage with a preserved hyaline cartilage surface. In addition, CPPD crystals are frequently found in fibrocartilaginous tissues, such as the menisci of the knee and the triangular fibrocartilage of the wrist. These characteristic features and screening of frequently involved sites with radiographs or US allow a differential diagnosis between gout and CPPD arthropathy in the majority of cases [57] (Fig. 13).

Fractures of the lesser metatarsals usually occur at the mid to distal shaft or subchondral region of the metatarsal heads [64]. Metatarsal shaft fractures frequently involve the second and third metatarsals, whereas subchondral fractures mostly involve the second metatarsal head, followed by the third and fourth metatarsal heads [65]. Subchondral fractures related to osteochondrosis of the metatarsal heads (Freiberg’s infraction) in adolescents are commonly discussed in this category, because their imaging findings are known to be similar to those of subchondral fractures in adults.

Radiographs are often unremarkable at the onset of symptoms because metatarsal fractures may be subtle or occult (Fig. 14A), and cortical stress reaction or callus formation can obscure the lucent fracture lines. In such cases, MRI or bone scintigraphy has been used to find clinically suspected, but radiographically occult fracture lines. Although MRI permits the best visualization of stress-related bone marrow edema and parosteal soft tissue changes even in the early stage, high-resolution US has become an emerging initial diagnostic method since it is commonly available in sports clinic offices and is easily accompanied by a physical examination on-site [66].

Metatarsal shaft fractures are frequently found on the dorsal side of the foot because of the superficial location and because the cortical bone is best evaluated when the US beam is perpendicular to the cortex. The dorsal side of the metatarsal head, where subchondral fractures usually occur, can be evaluated with or without plantar flexion of the toes. However, US has some limitations in evaluating the central and plantar sides of the metatarsal head. Therefore, radiographs and US should be used complementarily to each other.

Normal cortical bone appears as a smooth echogenic line with a reverberation artifact and posterior acoustic shadowing on US [66]. The earliest US findings of a stress reaction or fracture of metatarsal shaft are hypoechoic periosteal elevation and hypervascularity. When the stress reaction progresses into an overt fracture, the following US findings can be seen: interruption of the smooth cortical line (step off, gap, or displacement of the cortex), cortical irregularity or thickening by callus formation, and hyperechogenicity of the surrounding soft tissue because of edema [66] (Fig. 14B, C). Joint effusion is usually present with cortical abnormalities in recent metatarsal head and neck fractures, whereas degenerative changes of the joint can be seen in a remote fracture [64] (Fig. 15).

Nonetheless, the examiner should also be familiar with other cortical irregularities that may mimic fractures, such as nutrient vessels, physeal plates, cortical erosions, sesamoid bone, or ossicles. The differential points are the typical location, absence of proper trauma history, no other signs of injury, and no significant pain when adequate pressure is loaded over the site with the transducer [66].

Submetatarsal fat-pad abnormalities of the forefoot, including adventitious bursitis (submetatarsal bursitis), are usually caused by repeated friction and pressure on the foot during walking or prolonged weight-bearing [67]. Submetatarsal bursitis is one of the causes of metatarsalgia [67,68] and must be differentiated from the others. On US, submetatarsal bursitis is visible as an ill-/well-defined low-echoic lesion in the plantar fat pad, which is often compressible and sometimes contains fluid. Gregg et al. described that ill-defined bursitis is more compressible (Video clip 6), whereas well-defined bursitis is less compressible and even mobile. Hyperemia on power Doppler can be found in submetatarsal bursitis [69].

Abnormalities of the submetatarsal fat pad are also commonly found in asymptomatic patients, yet the distribution and imaging findings are different. An MRI study of submetatarsal fat-pad lesions revealed that asymptomatic lesions were most frequently found beneath the first and fifth MTP joints and showed T1 isointensity and T2 hypointensity, which suggests that these lesions were predominantly fibrosis. However, symptomatic lesions were often seen under the second or third MTP joint, and they were larger in size and showed mixed to hyperintense T2 signal intensity, suggesting a fluid component [67]. Similarly, a recent study showed that early RA patients tended to have submetatarsal bursitis more frequently than healthy controls. Furthermore, centrally located (beneath second to fourth MTP joints) bursitis is not uncommon in RA patients, whereas normal controls rarely have an abnormality in these regions [33].

There are limited studies regarding the pathogenesis or imaging findings of submetatarsal fat-pad lesions in patients with metatarsalgia. However, submetatarsal fat-pad lesions are occasionally seen on US imaging and are worth reporting. Therefore, when there is submetatarsal bursitis that is correlated with the location of pain, showing pain provocation by palpation/compression with the US probe might allow radiologists to suggest a diagnosis.

US is increasingly used to evaluate patients who present with metatarsalgia. Unlike cross-sectional imaging modalities such as MRI, US examinations are usually concluded during scanning, which requires a high degree of intention to scan the proper place where the true pathology is located. Therefore, radiologists should be familiar with the various causes of metatarsalgia, possible differential diagnoses, and their US imaging features.