Abstract
This was a retrospective study examining 60 surgically excised Morton's neuromas from 53 patients over a period of three years at Southend University Hospital, Essex. The initial diagnosis of Morton's neuroma was based on history and examination findings. In only one-third of cases was the neuroma palpable. However, many patients had difficulty localizing the pain and had atypical presentation. We attempted to assess the reliability of ultrasound in diagnosing Morton's neuroma in the 60 histologically confirmed cases. Our second objective was to assess whether size estimate of a neuroma seen on ultrasound correlated with subsequent real specimen measurement. In our study, preoperative ultrasound reliably diagnosed Morton's neuroma in 97% of the cases. In contrast, there was poor correlation in real specimen size measurements when compared with ultrasound reports.
In 1876, Thomas Morton 1 described a ‘peculiar and painful affection of the fourth metatarsal-phalangeal articulation’. However, it was Durlacher 2 in 1845 who first described clinical symptoms of the condition now known as Morton's neuroma. Although named a neuroma, there is no proliferation of axons as seen in a true neuroma. The nerve is enlarged due to a degenerative process of perineural fibrosis. 3 The position of the lesion typically lies just proximal to the bifurcation of the plantar digital nerve (Figure 1). It is also distal and deep to the transverse intermetatarsal ligament. 4 In most studies, the third intermetatarsal space is most commonly affected followed by the second intermetatarsal space. 5
Drawing showing the typical location of a Morton's neuroma (plantar view of left foot). m, medial plantar nerve; l, lateral plantar nerve; n, Morton's neuroma
The diagnosis of Morton's neuroma is currently based on the history and physical findings that include Mulder's and Tinel-Hoffman's signs. The former involves exacerbation of pain and sometimes a ‘click’ is heard when compression is applied over the metatarsal heads. The latter is a result from tapping over a nerve trunk on the dorsum of the foot which can cause a sensation of ‘pins and needles’ in its distribution. Preoperative localization of a suspected lesion is often difficult in atypical presentation or when multiple lesions are suspected.4,6
Ultrasound had previously been shown to be a valuable adjunct to the diagnosis of Morton's neuroma. In experienced hands, the sensitivity of diagnosis using ultrasound had been reported as between 96% and 98%. 3 , 6 , 7 We reviewed ultrasound scans of 53 individuals who had undergone surgery and had histological confirmation of their lesion.
Redd 8 described typical ultrasound characteristics of a Morton's neuroma as being an ovoid, hypoechoic mass just proximal to the metatarsal heads within a metatarsal space. Kaminsky et al.3 stated that pathological size measurements did not correlate well with ultrasound size measurements mainly due to additional tissue that was frequently incorporated into surgical specimens. Our study supported this finding.
Materials and Methods
We obtained 60 pathological reports of Morton's neuroma from April 2006 to March 2009 using our laboratory's Systematized Nomenclature of Medicine coding system. This is an organized computer system that allows easy storing and retrieval of clinical records based on medical terminology.
All 53 patients had one preoperative ultrasound scan followed by surgery in our institution. Six patients had two lesions on the same foot, both were excised at the same time. One patient presented with a single lesion on each foot and both were seen on ultrasound. In approximately 80% of our cases, surgery was performed by one consultant orthopaedic surgeon. We specifically excluded recurrent neuromas and patients with any prior surgery to the affected region. Patients ranged in age from 28 to 77 years with a mean of 56 years. All patients were symptomatic at presentation. Prior to ultrasound, all patients underwent routine plain radiography to exclude bony pathology, such as a stress fracture that could produce symptoms of a Morton's neuroma. In our orthopaedic department, lesions less than 5 mm were treated conservatively with a single intralesional corticosteroid injection. Patients who failed to respond to conservative treatment or had neuromas 5 mm or greater in size on ultrasound were offered surgery.
All scans were performed on a Philips iU22 ultrasound machine (Philips, Guildford, Surrey, UK) with a high-resolution broadband linear array transducer (L17-5). The vast majority of scans were performed by two musculoskeletal radiologists with many years of experience. Patients were either sitting or supine with knees extended and the affected foot dorsiflexed. The scans were performed from the plantar aspect with the transducer held transverse to the metatarsal heads. The transducer was moved from proximal to distal along the distal one-third of the metatarsals while exerting pressure on the dorsal aspect of the intermetatarsal space. All intermetatarsal spaces were examined in sequence. The views were supplemented by longitudinal scans when additional confirmation or confidence was needed. This was also performed by scanning along the intermetatarsal space while exerting finger pressure on the dorsal aspect. 9
The normal intermetatarsal space has a bright fatty signal. Morton's neuroma appears as a hypoechoic mass in between the metatarsal heads which replaces this fatty space (Figures 2a and 2b). An anechoic mass most likely represents an inflamed intermetatarsal bursa. The fluid in the bursa can be displaced by exerting pressure on the dorsal aspect of the foot and thus can be differentiated from a neuroma. 9 Neuromas usually appear rounded but may sometimes appear lobulated on a transverse view and ovoid on a longitudinal view (Figure 2b). Once a neuroma was located, exerting pressure along the intermetatarsal space with the index finger on the dorsum of the foot would quite often replicate the patient's symptoms of pain and sometimes tingling (the ultrasound Tinel's sign). A positive sign strongly guides the diagnosis of a Morton's neuroma. The transverse diameter of a neuroma was measured and included in the ultrasound report. After surgery, the size of a real specimen was measured by taking the longest diameter from any single fragment (Figure 3).
(a) Transverse scan of a left foot showing a hypoechoic mass representing a neuroma (borders marked by crosses). (b) Longitudinal view of an intermetatarsal space showing an ovoid mass. The digital nerve is visualized leading into the mass (arrow). P, plantar surface; D, dorsal surface
(a) A collection of haphazardly arranged nerves on low-power magnification. (b) Seen on higher power magnification, the size of these haphazardly arranged nerves are far in excess of what is expected of a normal digital nerve. Their expanded perineuria can be seen – the perineuria is the fibrotic capsule on the outer surface of the nerve. Such changes are typical of a Morton's neuroma
Results
Forty-five (85%) patients were women and eight (15%) were men. The majority (77%) had solitary lesions while seven (13%) patients had double lesions. There was a slight predominance of the left foot (58%, n = 35). Just over half (57%) of the neuromas were located in the second intermetatarsal space while the remaining 26 (43%) were in the third metatarsal space. In this study, no neuroma was found in the first and fourth intermetatarsal spaces. The average time delay between ultrasound and surgery was 23 weeks.
The transverse diameter of neuromas identified on ultrasound ranged from 3 to 23 mm. The median size was 5 mm and the mean size was 7.3 mm. In our study, only three neuromas (5%) from three patients measured less than 5 mm on ultrasound. They were the only patients in our study who were treated with single intralesional corticosteroid injections prior to surgery. Their clinical response to conservative treatment was not evaluated in our study, but all three real specimens measured larger than seen on ultrasound. Only two (3%) specimens were received as a single fragment. The real specimen size ranged from 5 to 30 mm with a mean of 11 mm and a median size of 10 mm. It appeared that there were definite discrepancies between sizes ascertained on ultrasound when compared with resected specimens. The difference in size measured macroscopically compared with ultrasound varied from −15 mm (real specimen smaller than ultrasound) to +22 mm (real specimen bigger than ultrasound). Most (83%) of the real specimens measured larger than when visualized on ultrasound. Only 10 (17%) specimens were smaller than the ultrasound measurement. The median of this difference was +4 mm and the mean was very similar at +3.8 mm (Figure 4).
Size difference between ultrasound and pathological measurements. (–): Real specimen measuring smaller than on ultrasound; (+): real specimen measuring larger than on ultrasound
The preoperative ultrasound reliably detected the presence and location of neuromas in 97% of our cases. Two scans gave false-negative results. One predicted a foreign body granuloma while the other was inconclusive. Both lesions were later excised and confirmed to be Morton's neuromas. Metatarsal heads were clearly identified in all scans. One scan reported an additional fluid-filled bursa and another mentioned capsular thickening. However, both gave the correct diagnosis of Morton's neuroma. In the ultrasound reports, all neuromas were classified as being hypoechoic masses. Fifty-two (87%) were classified as well-defined while eight (13%) were ill-defined.
One patient proceeded to have preoperative magnetic resonance imaging (MRI) of the affected foot. We did not evaluate the clinical indication in this case. MRI confirmed the presence of a neuroma with evidence of bursal inflammation. This was an additional finding not detected by ultrasound. On MRI, transverse diameter of the lesion was 6 mm but on ultrasound it measured 10 mm. Histological assessment determined the maximum diameter as 14 mm.
Discussion
Morton's neuroma is a disorder of the plantar digital nerve. A clear pathogenesis of this disorder has not yet been determined, but many theories have been advanced. The most widely accepted theory involves repetitive trauma to the neurovascular bundle while it is being trapped by the intermetatarsal ligament. The plantar digital nerve is then pinched between the metatarsal heads during the ‘push off’ phase of walking. 10 Women are more commonly affected than men. It is presumed that the pliability of a woman's foot together with the culture of wearing highheeled, narrow-toed shoes contributes to the development of this lesion. 4 , 8 , 11 In this study, middle-aged patients in their 50s were most commonly affected.
Although Thomas Morton initially described the abnormality in the fourth intermetatarsal space, subsequent studies showed that the third space was most commonly affected, followed by the second space. 5 One study revealed a prevalence of 53% in the third intermetatarsal space and 46% in the second intermetatarsal space. 6 Our study showed a higher prevalence in the second space.
The normal plantar digital nerve is no bigger than 2 mm. The majority of neuromas in this study were between 5–6 mm and this diameter had been proposed as a threshold for surgical treatment. However, the size of a neuroma does not always correlate with symptoms. 7 , 8 , 12
While ultrasound is known for its high sensitivity and specificity for detection of Morton's neuromas, 13 we demonstrated that there was poor correlation between ultrasound size and histological size of these lesions. The longest diameter of a real specimen may not correlate with the transverse diameter measured on preoperative ultrasound. The rationale for measuring the transverse diameter on ultrasound was due to two factors. First, margins of a neuroma are usually well-defined between two metatarsal heads. Hence, this measurement was easy to reproduce. Second, a neuroma was more likely to be compressed in this axis between two metatarsal heads during weight bearing. 10 This diameter may correlate better with symptoms and indications for surgery.
Formaldehyde fixation can produce up to a 20% decrease in linear dimension of tissues. 14 In our study, surgery rarely removed a neuroma as a single specimen. Excision of a lesion frequently incorporated normal adjacent tissue. This has also been described by others. 3 None of the pathological reports in our study described the margins of excision. This in combination with surgical practice (multifragments) is unlikely to yield the true measurement of a single-excised neuroma. In the minority of cases where ultrasound measurements were greater than histological measurements (17%), ultrasound may have identified adjacent inflamed bursa or oedematous tissue as being part of the neuroma itself.
The average time between ultrasound and surgery was 23 weeks. This is a significant length of time in which a neuroma may have continued to grow and may help to explain why 83% of real specimens were larger than seen on ultrasound. Reasons for this delay were not investigated and none of our patients had a repeat scan closer to the time of surgery.
It is not always easy to clinically localize a neuroma. The accuracy of Mulder's sign is variable and Nissen reported that the pain can be referred to adjacent intermetatarsal spaces. 15 Ultrasound had repeatedly demonstrated its reliability in detecting lesions down to size of 3 mm. 3 Ultrasound is readily available in most institutions and is cost-efficient. Such imaging has direct clinical merit as it potentially eliminates simultaneous exploration of adjacent interspaces, which could pose a risk of traumatic vasospasm. 8
Conclusion
This predefined pilot study demonstrated further that ultrasound, in the hands of an experienced operator, is a safe and reliable tool for detecting Morton's neuroma. 16 Ultrasound also has a role in reducing unnecessary surgical exploration. There was poor correlation between sonographic size and histological size of neuromas in this study. The reliability of ultrasound in measuring the true size of a neuroma remains questionable especially if measurement from a different axis is taken during histological examination. Furthermore, measuring fragmented excised neuromas will remain problematic.