Plantar digital neuroma 

Digital neuroma is an old problem and yet new analysis is warranted. This article teases out some of the theories that contribute to this foot metatarsalgia. The source of papers come from orthopaedic, radiology, neurology and podiatry sources. Overall there are some 1200 papers available written in English but how many in other languages is an unknown. You can read the No.1 first article here.

Key words – PDN plantar digital neuroma. MTPJ metatarsophalangeal joint

No.2 – Concepts of cause and pathology

A picture emerges that the plantar digital neuroma, assigned with the name Morton’s Neuroma or Civinini-Morton, is a nerve entrapment associated with chronic trauma. My earlier article discussed the fact that Morton in fact while recording a number of clinical histories did not really know what he was dealing with and his surgery was crude. Civinini at least understood a soft lesion was present and Durlacher (1845) recognised the neuralgic side of the condition. In contrast Morton thought it joint related. I have written a little more about these intrepid clinicians from the 19th Century in Back storyPathology can only be determined by histology, ultrasound and magnetic resonance imaging methods. Clinical methods of identifying the condition are reliable for thumb-finger pressure and Mulder’s click, the latter test cited back in 1951. The click is pathognomic for the nerve over a bursa because the bursa is more flexible. As the nerve size increases, more for the third intermetatarsal space, the transverse intermetatarsal ligament acts as a trap for the nerve to slip over. Characteristically the measurable size is more likely to be greater than 5mm to show dramatic signs but paraesthesia can arise earlier without the two clinical signs being evident. Neuritis changes nerve conduction sensitivity and may well be the cause of the bizzare shooting sensations (Satkevicuite, 2018). As the lesion increases in size symptoms may escalate and can even break through sleep, albeit not so commonly reported. Bencardino (2000) found that on retrospective MRI review of 85 patient images that 19 had positive MRI findings without symptoms after correlating with patients by questionnaire. Sizes varied from 2-9mm for the asymptomatic group.

Differential diagnosis

Common differential diagnosis includes synovitis of the MTPJ, avascular necrosis (Freiberg’s infraction), bursitis, gout, bone bruising to the metatarsal head,  stress fracture (March #) as well as true neoplasia. The presence of sensory changes can arise from other causes including nutritional, chemical and drugs, although more general. Radiating pain toward the ankle and even knee are not so uncommon. The cause of a neuroma are formed in part from theories, observations, tissue examination at surgery and under the microscope. Thomas G Morton was the first to undertake surgery (Morton, 1876) but this was crude and resulted in bone removal, while Hoadley (1893) accessed the foot through the sole to gain sight of the neuroma.

Renaut’s bodies

The intermetatarsal transverse ligament has often been considered a common reason for both the clinical sign of Mulder (1951) and compression of the nerve. Can the nerve compress on its own? From the earliest reports footwear and injury are the main causes of nerve damage and this damage unabated will lead to changes in the bundles around nerves lined by the perineurium, epineurium or the endoneurium. The resultant effect is scar tissue and so fibrosis of any of these linings is reported on histology as well as an increase in size. DeHeer (2020) draws attention to the spidery fibroblasts called Renaut bodies which are found near sites of nerve compression. Autopsy and tissue studies show presence of Renaut bodies in peripheral nerves associated with thickened sub perineurial capillaries. They are not present in all nerves but the association with mechanical factors are important in the pathogenesis of nerve alterations which further suggests they do not protect nerve fibres from mechanical stress.

In an autopsy study of the pathology of chronic subclinical nerve entrapment Renaut bodies showed a strong predilection for sites of nerve entrapment. They were present at these sites in 43 of 74 peripheral nerves but in none of the control sections of the same nerves. Renaut bodies were most frequently encountered in the median nerve at the wrist and in the lateral femoral cutaneous nerve at the inguinal ligament but were rarely seen in sections of the common peroneal nerve at the neck of the fibula. In several nerves Renaut bodies showed a similar orientation within adjacent fascicles, suggesting that mechanical factors were related to their pathogenesis; despite this finding there was no relationship between their numbers at entrapment sites and the presence of pathological changes in myelinated nerve fibres at the same level. These findings suggest that while mechanical factors are important in the pathogenesis of Renaut bodies there is no evidence to support the theory that these structures protect nerve fibres from mechanical stress Jefferson (1981).

Ischaemic influence

Blood vessels run through nerve fascicles and they show endarterial thickening as well as vascular proliferation as can be seen in gross samples taken from live subjects during surgery. There is also the presence of oedema in the endoneurium. One is drawn to ask why we need to know the detail. The relevance of pathology will emerge as an important part of understanding how to plan treatment and assess the best approach. Unfortunately all the main systematic (meta-analysis) reviews do not draw together treatment and pathology, and treatments often emerge with little scientific critique except reporting upon the basis of percentage success. New treatments based around substance injection are in fact reviewed on the basis of reversing some of the fibroblastic activity. We need to turn our attention to the principle causes that create neural fibroblastic degeneration that leads to axonal degeneration. To date no study has shown any outcome based on vascular changes to the nerve but injections aimed at neurolysis or sclerosing with ethanol in rat sciatic nerves undertaken by Mazoch (2014) and cited by DeHeer (2021) might suggest some changes arise. Data is far too small to draw any meaningful conclusion.

What causes fibrosis?

Most clinicians understand the pathology of tissue healing and damage together with axonal degeneration but in the case of the PDN there has to be inciting factors. Infection has never been suggest so it remain mechanical in nature. The Mulder test creates both pain and a pop sound as the neuroma rides over the adjacent ligament lying above the nerve. It is for this reason that many surgeons prefer a plantar approach so as not to disturb this structure. Equally there are minimalists who cut the ligament alone to relieve the pressure on the nerve and hopefully reverse the changes that are slow and progressive. As a sufferer myself and having had surgery after conservative care, the length of time from inception to resolution can take >1 year to 7 years or more. Morton reported a case that had tolerated pain for 30 years. There is no standard baseline for symptoms much as reported by Bencardino (2000). The most commonly assumed reason for compression comes from shoes and the group who suffer most are females by a factor of anything from 3-10 times. In my own case I used a tight cycle shoe that clipped into the pedal and the damage escalated, despite using orthoses successfully for 4 years. I deduced the shoe caused the final damage whereas before such damage was controlled as guided by the symptoms. The neuroma was injected with steroid and local anaesthetic which yielded no benefit and ultrasound showed a lesion in the third intermetatarsal space. Too little, too late!

Injury leading to haematoma can theoretically cause scarring with the same fibroblastic pathogenesis but normally haematomas resolves. Spasm of the interossei are not reported in the literature, but again my affected foot suffers from paroxysmal spasms that are inordinately painful. These small muscles, be they lumbricals or interossei seem innocuous but could contribute to compression. The presence of a bursa is not an uncommon finding as most foot surgeons will report and without histology we make blind assumptions as tissue cannot always be classified at the time of surgery without the aid of the dedicated histologist or cytologist.

Junction box theory


Type II Govsa (2005) 20% of cases n=50

There would be some injustice however to say that all PDNs are caused by footwear as injury, either directed at the sole, or through a fall or twist affecting the metatarsals can incite the same changes. The idea that metatarsals 2-4 are mobile and are the cause if excessive movement arises has more validity when we realise that the anatomical relationship around the connecting branch between medial and lateral plantar nerve promote a potential obstructive impingement if damaged. Govsa’s (2005) classification of anatomical communicating nerve branches is persuasive where there were only 11% of his 50 male specimens that had a communicating branch between the lateral and medial nerve. This is where the branch arises laterally. There are branches in 8% of cases from the medial nerve toward the lateral nerve ending in a junction box between the 4th-5th metatarsals. We know that the incidence of 4th interspace neuromata arise between 2-8% of the time (Raouf 2019, Bencardino, 2000, DiCaprio 2018) and the first space hardly ever recorded. So why is the 3rd space more prone to symptoms and change than the 4th space. The independence of the 1st and 5th rays are almost part of basic podiatric-orthopaedic foot knowledge and so can move out of the way compared to the central 3 metatarsals. Naraghi (2017) were unable to demonstrate any relationship in this study between radiographic metatarsal length and angular measurements for symptomatic PDN when compared to a control group.


Arguments over cost and imaging

Mallina – UK (2017) and Raouf – USA (2019) are critical of the unwarranted costs of undertaking investigations on the basis that clinical signs and symptoms are conclusive and sufficient for undertaking surgery. This is useful because health-economics must be balanced between what is appropriate and what is not in terms of value added. Neither studies suggest that tests do not have a place, especially where the diagnosis remains unclear. There is a belief that unless imaging or histology can change the course of management then routine imaging or histology should not be undertaken. One of the driving factors that has affected medicine and surgery is the rampant cases that seek judicial settlement, often demarcated to consent and decision making. The ideals of the two papers by Mallina and Raouf are not misplaced but proof of activity can often only arise from validating the aetiology. However the need for vigilance arises where treatment has led to complications as a fall back part of medical documentation. Mallina and Raouf are correct that savings can be made but the cost of litigation settlement can also add to the cost of management disproportionately.

Counter arguments

Contrary to those who believe clinical signs are sufficient, pathognomic (clinical) testing can lead to false-positives Santiago (2018). Radiological imaging has brought value to our understanding of neuroma, and histology to the stages of pathogenesis cannot be contradicted. We must interrogate some of the evidence because we can learn about quantifying the pathology and the sensitivity applied to each test. Of some 50 papers reviewed for these articles, over half had an imaging basis and nearly 20% of studies were undertaken by radiology rather than orthopaedic departments. When we learn that many techniques today use a needle to inject an agent into the neuroma or around the location, whether by corticosteroid, ethanol, phenol, hyaluronic acid, ultrasound (US) or fluoroscopy is often the preferred method to accompany these injections. These highly desirable methods fall outside the wasteful cost argument because they are part of the technique and add safety.

False diagnoses

There are perhaps two fundamental questions clinicians would desire. Is a test accurate and which test should I use. As a clinician I can advise that I have undertaken an MRI scan for a neuroma and the result as reported by a musculoskeletal radiologist was negative. I decided to operate and found a neuroma. This is never an easy decision but highlights the reality of false negatives. An orthopaedic colleague swore by MRI while I placed most of my own faith on ultrasound (US). Which one of us were right?

Key imaging techniques

Santiago (2018) points out that a nerve 1mm at the level of the intermetatarsal heads can be determined on high US and MRI.

Ultrasound is more difficult to read in static form compared to dynamic states and so is dependent on experience from the operator. The position of the transducer head and quality of the equipment can make a difference. As technology has improved the growing inclusion of ultrasound has grown exponentially in podiatry. Many favour a plantar approach. A well-defined round or ovoid area is seen on the short axis, while on the long axis a fusiform shape is evident for a neuroma. The term hypoechoic is applied relative to the adjacent tissues. If injected the area distends where a bursa is present. Lesions longer than 20mm should be excluded as unlikely PDN lesions (Santiago).

Magnetic resonance imaging is more expensive and time consuming but less reliant on the operator, although carried out by trained professionals nonetheless. A well demarcated ovoid or dumbbell shaped intermetatarsal mass is evident. T1 and T2 weighted scans show a low signal for a neuroma. Gadolinium is injected to improve the quality of the image and only enhances a bursa rather than the PDN. T2 is hyperintense while T1 is isointense for the bursa. The MRI provides better soft tissue differentiation and visualisation.

Sensitivity & specificity

Sensitivity relates to how often a test shows positive i.e true positivity (TP) as opposed to a false negative (FN); this was the case cited above in my own experience with the MRI result. TP/TP+FN=sensitivity.

Specificity is where a person does not have the disease and the calculation is related to how often the test turns out negative, or represented by a true negative (TN). If you have a test as in the case of the MRI and it shows positive then you have the condition but it could be a false positive (FP). TN/TN+FP=specificity. 

Sensitivity MRI US
Xu 93% 90%
Bignotti 90% 91%
Claassen 84%
Benecardino 100% 98%
Pastides 88% 98%
Raouf 80% 93%


Specificity MRI US
Xu 68% 88%
Bignotti 100% 85%
Claassen 33%
Benecardino 68%

See Claassen’s sensitivity graph (below) which helps visualise data from the perspective of true positive and true negative.


Claassen (2014) Role of MRI in neuroma

Bignotti (2015) looked at past papers in order to run a systematic review (meta-analysis) comparing MRI (3 studies) to US (5 studies) for PDN. Sensitivity and specificity were criteria important. Of 277 studies they narrowed the group down to 27 papers from which 14 articles were selected. Six studies were prospective and eight were retrospective. Bignotti assumes multiple studies in the table above rather than single articles shown by others. In their meta-analysis they concluded that there was only a small margin of difference in terms of sensitivity, both being highly sensitive. This paper suggests that at the time of the review MRI and US is equally accurate, US has more advantages toward cost. Claassen (2014) highlighted the use of MRI alone and found negative MRI when Morton’s neuroma existed (Figure 2). This meant specificity was poorer compared to the sensitivity. In this case the authors based in Germany felt clinical signs of PDN could be more reliable. The value  associated with imaging excludes differential diagnoses. Pastides (2012) found sensitivity high for US compared to MRI and concluded that in the presence of clear clinical signs imaging was not required.  As with other authors there were indications for ruling out other problems or determining foot problems with more than a single PDN. Xu (2015) concluded that US could provide better accuracy than MRI. Some 7 years earlier a study by Lee (2007) contradicted this finding suggesting both imaging method were equal. Bignotti certain draws the same conclusion with their 14 article comparitive meta-analysis.

In order to answer the question Which one of us were right? – there is no absolute correct approach presently even though these reports come from an on-line search (30/10/21). Cost and flexibility must be considered as well as the confidence surrounding the presenting symptoms. 

Continuing reading the series on plantar digital neuroma

Claassen L, Bock K, Ettinger M, Waizy H, Stukenborg-Colsman C, Plaass C. Role of MRI in Detection of Morton’s Neuroma. Foot & Ankle International. 2014;35(10):1002-1005. doi:10.1177/1071100714540888
Lee MJ, Kim S, Huh YM, Song HT, Lee SA, Lee JW, Suh JS.   Morton Neuroma: Evaluated with Ultrasonography and MR Imaging.   Korean J Radiol. 2007 Mar-Apr;8(2):148-155.
Jefferson D, Neary D, Eames RA. Renaut body distribution at sites of human peripheral nerve entrapment. J Neurol Sci. 1981 Jan;49(1):19-29. doi: 10.1016/0022-510x(81)90184-2. PMID: 7205316.
Naraghi R, Bremner A, Slack-Smith L, Bryant A. Radiographic Analysis of Feet With and Without Morton’s Neuroma. Foot & Ankle International. 2017;38(3):310-317. doi:10.1177/1071100716674998

A full set of references is available click here

Thanks for reading ‘Digital Neuroma Old Problem New Analysis’ by David R Tollafield

Published by Busypencilcase Communications (Est. 2015)

Published 6th December 2021 for Consultingfootpain.