Morton's neuroma: Diagnostic accuracy of clinical

19 downloads 0 Views 719KB Size Report
Morton's neuroma: Clinical testing and imaging in 76 feet, compared to a control group ... We compared clinical and MRI findings in 76 feet treated operatively for ...
Morton’s neuroma: Clinical testing and imaging in 76 feet, compared to a control group. Frimley Park Hospital, Frimley, Portsmouth Road, Surrey, GU16 7UJ, United Kingdom. Abstract Background Morton's neuroma is a mechanically induced degenerative neuropathy. Methods We compared clinical and MRI findings in 76 feet treated operatively for Morton’s neuroma and 40 feet with different pathologies (controls). Results In the ‘treatment group’ web space tenderness (WST) was positive in 95%, foot squeeze (SQU) in 88%, plantar percussion (PLP) in 61%, and toe tip sensation deficit (TTSD) in 67%. MRI identified neuromata in 97%. Histological examination confirmed neuroma excision in 99%. Frequency of positive tests was significantly higher in the treatment group compared to controls. MRI revealed (asymptomatic) neuromata in 10/40 (25%) feet in the control group. TTSD was similarly positive in asymptomatic and symptomatic neuromata. TTSD in association with any other test being positive, was significantly more frequent in symptomatic neuromata. Conclusions Diagnosis of Morton’s neuroma, based on clinical and imaging findings, was accurate. Positive clinical testing was more frequent compared to the ‘control’ group. Keywords: Morton; neuroma; foot; diagnosis; MRI. Introduction Morton's neuroma is most likely a mechanically induced degenerative neuropathy which has a strong predilection for the second and third common digital nerves in middle-aged women [Weinfeld, Hassouna]. Patients usually present with metatarsalgia (pain in the sole of the foot in the region of the metatarsal heads) [Weinfeld, Hassouna]. Diagnosis is based on clinical examination, and imaging techniques (MRI and ultrasound) are useful in confirming the diagnosis or exclude other causes of metatarsalgia (eg. joint inflammation and bursitis, arthritis, osteonecrosis) [Weinfeld]. Prior to any operative therapy it is mandatory to perform a detailed analysis of the underlying pathology to avoid persistent pain or recurrence. It is not known how accurate clinical assessment is, and which clinical tests could reliably diagnose this condition. We retrospectively reviewed a cohort of patients who were treated operatively for Morton’s neuroma by a single surgeon (AS), and compared the results of imaging (MRI) and clinical tesing with a cohort of 40 feet of patients who did not have a symptomatic neuroma. Methods From 01/2006 to 04/2009, 77 consecutive patients (83 feet) were treated operatively for Morton’s neuroma by the senior author (AS). No rheumatoid patients / feet were included. Two of the authors (RO,HG), other than the surgeon, retrospectively reviewed the patients’ clinical notes and imaging studies associated with their foot pathology, collected and analyzed the data. All surgeries were

1

performed by the senior author or under his direct supervision. Data regarding clinical presentation (symptoms), clinical examination (signs), imaging, and histology were collected and analyzed. Regarding clinical presentation, we documented the presenting problem (pain, site of pain, sensation alterations, effect of shoe-wear, other symptoms). Documentation of clinical examination included assessment of four clinical signs (positive vs negative): web space tenderness (Figure 1), foot squeeze test (Figure 2), plantar percussion test (Figure 3) and toe tip sensation deficit (Figure 4). In the senior author’s practice all patients with suspicion of a Morton’s neuroma underwent MR imaging of their forefoot, using a specific protocol. The scan reports were reviewed and data were collected regarding the presence of a neuroma and its size and the affected web space. The postoperative histology reports were reviewed and data regarding histological confirmation of neuroma excision and its size were collected. Seven patients (7 feet) had to be excluded from final evaluation, because of missing data. This left 70 patients (60 women, 10 men) with 76 feet treated operatively for Morton’s neuroma. We refer to this group of patients as ‘Group A’. Their mean age at the time of surgery was 51±12 years (range 27 to 73). Five of those patients developed later symptoms in an adjacent interspace, and surgical excision of a new Morton’s neuroma was required. Data regarding simultaneous neuroma excision were not used, as previous diagnosis and surgery could have biased the clinical assessment and the diagnostic process. One of the authors (NG) prospectively evaluated the ‘4 clinical tests’ in a cohort of 34 patients / 40 feet (‘Group B’), who presented with foot pain associated with pathology different than Morton’s neuroma, from 05/2009 to 12/2009. These patients served as the ‘control’ group. Mean age was 52±10 (range 34-73) years. They were examined clinically and were investigated with an MRI scan (using the same imaging protocol). Data were collected regarding the clinical diagnosis and presence of (asymptomatic) Morton’s neuroma on MRI scan. For clinically meaningful comparisons to be possible, feet in group B were subdivided into three categories. B1: Feet with an asymptomatic Morton’s neuroma identified by the MRI scan. B2: Feet with pain associated with lesser rays. B3: Feet with scans negative for Morton’s neuroma. Chi-square (x2) test was used to detect statistical differences between the two groups. Level of statistical significance was set at 0.05.

Results In Group A, from 76 symptomatic feet the right foot was affected in 39 of the cases and the left in 37. The 3rd web space was symptomatic in 51 feet (67%) and the 2 nd in 25 (33%). The presenting problem was pain in 74 out of 76 feet (97.4%) (Table 1). ‘Electric shocks’ were the primary symptom in 2 patients (2.6%). Altered sensation (eg numbness, paresthesia, ‘pins and needles’) was reported in 34 of 76 feet (45%). Wearing shoes made symptoms worse in 65 of 76 feet (86%), two feet were better wearing shoes and in 5 it made no difference. In group B, 12 patients / 16 feet were treated for pathology associated with the lesser rays (synovitis / bursitis of metatrsophalangeal joints 14 feet, metatarsal head degeneration in one foot, plantar fad pad alterations in one foot) , 9 patients / 9 feet for pain in the 1st ray (hallux metatarsophalangeal joint synovitis or degeneration in 6 feet, sesamoiditis in 3 feet), 10 patients / 10 feet for pain in the midfoot

2

(all degenerative changes) and 3 patients / 5 feet for pain associated with the plantar fascia (plantar fasciits 3 feet, plantar fibromatosis 2 feet). Four clinical tests, consistently performed in all feet, were variably positive in feet subsequently treated operatively for Mortion’s neuroma (Table 2). In feet treated for other pathologies the ‘4 tests’ were positive less frequently (Figure 5). Chi-square test statistical analysis was performed in order to assess the differences observed between groups, regarding the four clinical tests (Table 3). Positive ‘foot squeeze’ test values were similar in feet with an asymptomatic neuroma identified by MRI (group B1), and those treated for lesser ray pathology (group B2), compared to those with surgically removed neuromata (group A). The combination of positive ‘foot squeeze’ and ‘web space tenderness’ tests was significantly more common in group A (surgically treated neuromata), compared to group B and any subgroup. ‘Toe tip sensation deficit’ was positive in asymptomatic neuromata (group B1) in a similar rate to symptomatic ones that underwent surgery (Table x). However, positive ‘toe tip sensation deficit’ in association with any other test being positive, was significantly more frequent in group A, in comparison with group B1 (asymptomatic neuromata, found on MRI scan), as well as any other subgroup (Table 3). Preoperative imaging (MRI) identified a neuroma in 74 of 76 feet (97%) in Group A. The same musculoskeletal radiologist (with special interest in Foot/Ankle imaging) had reviewed and reported the scans in 71 of 76 feet (93%). In the two feet with a negative MRI report (false-negative), 3 out of 4 clinical tests were positive. One of those scans was reported by a different radiologist and in retrospective evaluation a neuroma was identifiable. In Group B, the MRI scan revealed a Morton’s neuroma in 10 out of 40 feet (25%). These neuromata were asymptomatic. These patients complained for midfoot (4 feet), 1st ray (4 feet), or plantar fascia associated (2 feet) pain. The size of the neuroma in Group B patients was not significantly different compared to Group A (Table 4). Postoperative histological examination of the excised specimen confirmed excision of a neuroma in 75 of 76 feet (99%), in group A. In one specimen the diagnosis was ‘chronic bursitis’. The preoperative MRI scan had shown a neuroma (size 10mm; false positive) and 3 out of 4 clinical signs were positive (no toe tip sensation deficit). The average maximum thickness of the excised neuromata, measured and reported by the pathologist was 7.8±2.4mm (median 8mm, range 3 to 14mm). Discussion The aim of the current study was to present the clinical manifestations and the incidence of certain clinical signs, as well as results of forefoot imaging in a cohort of patients with Morton’s neuroma. As no clinical signs are pathognomonic for Morton’s neuroma, one could argue that the clinical tests could be associated with other foot pathologies. Therefore, we compared clinical testing in this cohort of operatively treated neuromata, with a prospectively examined ‘control’ group of patients of similar age, who presented with foot pain possibly associated with pathologies other than a neuroma, and had an MRI scan of their forefoot. It is not possible to demonstrate if the diagnosis of Morton’s neuroma was missed in other patients (based on the same criteria), as this would require surgical exploration of any foot in patients with forefoot pain. Such a study would be, obviously, unethical. Thus, it is not possible to detect specificity and sensitivity of the clinical testing we used. One the other hand the strengths of the current study

3

are: a) the large sample size in the treatment group, with loss of only 10% of subjects due to missing data, b) the analysis of data by independent assessors, c) the consistency in clinical testing and its documentation following the instructions of an experienced foot and ankle orthopaedic surgeon (AS), and d) the consistency of imaging reported by the same specialized musculoskeletal radiologist in the vast majority of cases (93%). Furthermore, the histological confirmation of the excision of a neuroma in 99% of feet, is indicative of appropriate patients’ selection and surgical technique. Pathognomonic diagnostic clinical tests for Morton’s neuroma do not exist. Any of the ‘4 tests’ can be found positive in feet with different pathologies (control group). Our study demonstrated, however, that the ‘4 tests’ were more frequently positive in feet with a ‘symptomatic’ neuromata, compared to feet with ‘asymptomatic neuromata’, or other pathologies associated with the lesser rays. The ‘treatment’ and the ‘control’ groups are not directly comparable, therefore no definite conclusions regarding the accuracy of clinical testing can be claimed. However, a combination of positive ‘foot squeeze’ and ‘web space tenderness’ tests implies a very high possibility that the diagnosis is (symptomatic) Morton’s neuroma, potentially requiring surgical excision. Similarly, ‘toe tip sensation deficit’ combined with any other positive tests, is more likely to occur if a symptomatic Morton’s neuroma is present. Mulder’s sign (Mulder) was not used in our patients’ clinical evaluation, as it has been shown to be positive in only 40% of cases (Coughlin). This concurs with the senior author’s experience, although other authors found Mulder’s sign positive in 84% of feet in their series (Sharp). We applied the ‘foot squeeze test’ instead, which was positive in 95% of feet in our study cohort. A key characteristic of the ‘squeze test’ is pain associated with palpation of the web space whilst squeezing the foot (Figure 2), as opposed to a palpable ‘click’ reproduction (Mulder’s sign). Other authors reported findings similar to ours, regarding symptoms and signs in patients with Morton’s neuroma. Coughlin (2001) reported shoe-wear restrictions in 70%, preoperative subjective numbness in 51% and objective toe tip sensation deficit in 41% of his patients. This compares to the 86%, 45% and 49% incidence of symptoms and signs, respectively, in our study. Sharp et al. (2003) reported presence of pain in all their patients and that shoe-wear worsened the symptoms in 76%, whereas paraesthesia was present in 68% in a total of 31 feet they treated. In the same study (Sharp) the mean neuroma size detected on MRI was 7.4±4.6mm and after histological examination 8±3.4mm, which is comparable to our findings. They found, however, that MRI failed to detect a neuroma in 4 out of 24 cases (16%). The higher diagnostic accuracy in our study (97%) may reflect a better imaging protocol or expertise by the examiner. The size of asymptomatic neuromata was not significantly smaller compared to symptomatic ones (Table 4). The sample size in the ‘asymptomatic neuromata’ group was, however, small, therefore no definite conclusions can be drawn, according to our data. Whether bigger lesions are clinically more pronounced, needs to be confirmed by further research. In the experience of Sharp et al. (2003), the size of the neuroma did not correlate with symptoms, whereas other studies suggested that lesions bigger than 5mm are more likely to cause symptoms (Redd, Zanetti). Based on our results we cannot give a definite answer on whether we need imaging in the diagnosis of Morton’s neuroma. Resch et al. (1994) had suggested in the past that MRI is of little value in the diagnosis of Morton’s neuroma. In our practice imaging is mainly obtained to differentiate web space neuroma from other forefoot pathologies, and to increase the certainty regarding the diagnosis,

4

especially if surgical treatment is planned. It has been shown that Morton’s neuromata have been identified in 30% of MRI scans of asymptomatic feet (Studler). This compares to the 25% rate in our ‘control group’, which, however, did not include asymptomatic feet. This definitely implies that careful clinical assessment is important. It would be useful if a future prospective research project would trace a patients’ population who present with forefoot pain and assess clinical testing in relation to imaging that identified other pathologies. The 99% accuracy in the diagnosis (confirmed by histology) in our study, supports our diagnostic protocol. Whether to use ultrasound instead of MRI for imaging is debatable. In our practice, MRI is performed with a particular ‘neuroma protocol’ and provides good quality images. Furthermore, it is less operator depending, although more expensive than ultrasound, which has been shown inaccurate in detecting small size (