Ultrasonography of peripheral nerve tumours: a case series

Peripheral nerve tumours (PNTs) are rare, but important cause of peripheral nerve dysfunction. Usually they present with neurological symptoms (muscle atrophy and weakness, paresthesia or sensory loss), palpable mass or pain.^1,2 On examination neurological deficits can be found distally to PNTs in the affected nerve innervation area. Mass movable perpendicular, but not along the peripheral nerve axis, and sensations along the affected nerve elicited on mass percussion (i.e., Tinel’s sign) are also pointing to possible PNT.^1,2 Even before PNT diagnosis is known, electrodiagnostic (EDx) testing is often performed to evaluate severity of peripheral nerve damage. However, EDx is not useful for PNT diagnosis¹, but imaging studies are much more relevant. Magnetic resonance (MR) is regarded as the most useful method^1,3, although ultrasonography (US) is also gaining support among clinicians.^3,4 Particularly when diagnosis of PNT is not known, US is most useful, because it is cheap and widely accessible. Imaging delineates the lesion, identifies its relation to peripheral nerve, and helps to differentiate various types of PNTs (Table 1).⁵ There are several nice reviews describing US characteristics of PNT.^3,4 However, only few publications written mainly by radiologists describe actual clinical experiences with US diagnosis of PNTs.^5,6,7

In the present study we report a series of consecutive patients referred to our US unit mainly from EDx laboratories in whom we have diagnosed PNTs.

We retrospectively reviewed the electronic medical records of all patients referred from February 2013 to May 2020 to the US laboratory at the Institute of Clinical Neurophysiology, University Medical Centre Ljubljana, Slovenia. Our unit is the only one dedicated to peripheral nerve US in Slovenia, a country with a population of two million. The National Ethics Committee of Slovenia approved the study (approval code: 63/07/17), and at the time of analysis all patients signed written informed consent. During the whole review process, all patients’ personal information was carefully protected.

We were not blinded to the findings of the clinical neurologic examination and EDx testing. Before US examinations we performed a focused neurological examination of patients by ourselves. We used standard US equipment (ProSound Alpha 7, Hitachi Aloka Medical, Ltd., Tokyo, Japan), with a 4–13 MHz linear array transducer. We measured tumor cross sectional areas (CSAs) by a trace method that excluded the hyperechoic rim.⁸ During review of US images we observed PNT features typical for neurofibromas and schwannomas (Table 1).⁵ In patients without histological diagnosis presumptive PNT diagnoses were based on clinical, EDx and particularly US features. In all included patients, we collected data on gender, age, symptoms (including duration), neurological, EDx features, US findings, and PNT location.

In the analyzed period, in our US unit we examined 2845 patients, and we found PNTs in 15 (0.5%) of them. Demographic features of individual patients, PNT anatomic, clinical, EDx and US features are shown in Table 2. Our patients’ median age was 33 years (range: 16–69 years). Slightly less than half of them were male (47%). Four had multiple PNTs; in 2 neurofibromatosis type 1 (NF1), and in 1 neurofibromatosis type 2 (NF2) were previously diagnosed. In another (patient #13, Tables 2–3) gene sequencing results are pending. Of 11 patients with single PNTs, 7 (64%) had lesions on the right side. Ulnar nerve was involved in 4 (36%), median, sciatic and tibial in 2 (18%) patients each, and fibular in 1 (9%) patient with single PNT. Elbow and forearm were each involved in 3 (27%), thigh and ankle in 2 (18%) patients each, and knee in 1 (9%) patient with single PNT. The most common clinical findings in our patients were weakness and sensory abnormalities, each found in 7 (47%) patients, followed by muscle atrophy in 6 (40%), pain in 3 (20%), sensitivity on mass percussion in 2 (13%) and only palpable mass in 1 patient (7%). At the time of US study, EDx report was available for 12 (80%) patients. Affected nerve compound muscle action potential (CMAP) amplitude was markedly reduced or absent in 5 (42%) patients, and sensory nerve action potential (SNAP) amplitude in 9 (75%) patients. In our series CSAs of PNT varied from 24 mm² to 1250 mm² (median, 61 mm²). Ratio of PNT/unaffected segment of the same nerve CSA varied from 2.9 to 156 (median, 6.0). Morphological features of PNTs are presented in Table 3. Ratio of maximum/minimum PNT diameter varied from 1.5 to > 10 (median, 4.5). Majority (64%) of PNTs were of fusiform shape. PNT contour was smooth in 8 (53%), and lobulated in remaining 7 (47%). PNT encapsulation was complete in 7 (47%), partial in 6 (40%), and absent in 2 (13%). PNTs echotexture was heterogenous in 13 (87%). Nerve entrance into PNT was central in 10 (67%), eccentric in 3 (20%), and not possible to asses in 2 (13%). In our series infiltrative nerve-tumor transition was observed in 7 PNTs (58%), poorly defined in 5 (42%), and could not be observed in 3 (20%). Histological diagnosis was available in 5 patients (Tables 2 and 3) with schwannoma. In another patient with NF1 histology of multiple PNTs could be also established – neurofibroma. Based on histological definite, and based on clinical and US features probable diagnoses of schwannoma were made in 6 (40%), neurofibroma in 5 (27%), and perineurioma in 4 (27%) patients.⁵

PNTs are known to be rare, which was also confirmed by our US experience, demonstrating PNTs in only every 200^th examined patient. As in our US laboratory we see referrals from a number of Slovenian EDx laboratories, and examine primarily patients with unclear etiology of peripheral nerve lesions, in reality PNTs are probably even rarer. It was reported that PNTs constitute only 5% of adult, and 2% of pediatric upper extremity tumors.⁹

With median age of 33 years our patients were much younger compared to typical oncological patients, which is similar to previously reported series of patients with PNTs reporting mean age of 36 years¹⁰, and 37 years.⁷ As reported by others^7,10, we also found equal distribution of PNTs between both genders.

We could divide our patient cohort into several groups. The first group consisted of 4 patients with numerous PNTs in several peripheral nerves. Two of patients from this group with numerous neurofibromas had known NF1 and a single patient with schwannomas had NF2. Diagnosis in a remaining patient (#13), with numerous neurofibromas in all US examined peripheral nerves (Figure 1), has not been established yet.

Figure

The second group consisted of three girls (patients #6, 7, 10) presenting in the first decade of life with unilateral foot drop. Each of them had several lumbo-sacral spine MRs that all proved negative, and their PNTs were not diagnosed until they presented to our US laboratory 8–15 years after symptom onset. At that time their clinical features were unchanged for several years, so none of them decided to have a nerve biopsy to establish a histological PNT diagnosis. Nevertheless, we believe these 3 patients most probably have sciatic nerve perineurioma.¹¹ One of them (patient #10) had two PNTs, separated by 20–30 cm segment of US relatively normal sciatic nerve. Using US we demonstrated the more distal PNT of the tibial portion of sciatic nerve (Figure 2). Additional more proximal PNT affecting fibular portion of sciatic nerve, and causing foot drop was found on MR.

Figure

The third group consisted of four patients with PNTs on the ulnar nerve in the elbow segment (Figure 3). The main challenge in this group of patients is differentiation of PNTs from ulnar neuropathy at the elbow (UNE) due to entrapment under the humeroulnar aponeurosis (i.e., cubital tunnel syndrome). The main distinctive feature of PNTs is their large CSA (43 mm², 348 mm², 45 mm², and 368 mm²). We found CSA larger than 40 mm² in only 1 of 202 (0.5%) UNE patients,¹² and is therefore extremely rare. Another characteristic feature of PNTs is rather well preserved nerve function in spite of large nerve thickening. By contrast, large CSA in UNE is as a rule accompanied by severe nerve dysfunction.¹² What is the reason for such high number of PNTs on ulnar nerve in the elbow area is not clear. One possibility would be exposure of the nerve in this segment to mechanical stress. Alternative explanation would be a sampling artefact, as we see a plenty of patients with suspected UNE. If the latter case that would mean that a large number of PNTs on other nerves and locations are still missed.

Figure

Of remaining 4 patients, 1 had definite and another probable tibial nerve schwannoma at the ankle, which is again a region of considerable mechanical stress. Another young woman had a fusiform thickening of the ulnar nerve in the forearm (patient #5, Figure 4). She had surgical release of the ulnar nerve exit from the flexor carpi ulnaris muscle, with no apparent benefit. She might also have perineurioma, or less likely neurofibroma. In the fourth man probable median nerve schwannoma in the forearm was a coincidental finding during US evaluation due to Lewis-Sumner syndrome, and caused no additional symptoms.

Figure

As described previously¹ according to clinical presentation, our patients could be divided into two groups. Eight (53%) patients presented with features of peripheral nerve lesion (e.g., muscle atrophy, weakness, sensory loss). Remaining 7 patient presented by other clinical features (i.e., pain, local sensitivity, palpable mass), or as coincidental PNT finding without symptoms. In the first group CMAP and SNAP amplitudes of affected nerves were severely reduced or absent (42% and 75%, respectively), and in the second group they were mainly preserved (58% and 25%, respectively). The first group consisted of all 4 young women with probable perineurioma, and additional 3 patients with probable neurofibroma. By contrast, majority of the second group consisted of 4 patients with probable or definite schwannoma (Table 2).

Before introduction of US into our institution, we diagnosed PNTs only very rarely. This changed after we started to perform US studies. Some PNTs that we finally diagnosed using US, were causing patients’ unexplained severe nerve dysfunction for more than a decade. Without US we would probably not be able to diagnose PNTs in majority of patients from this series, and some of them would probably remain without diagnosis to this day.

The main limitation of the present study was that in majority of patients histological diagnosis of PNTs was not available. Therefore, in these patients we based our PNT diagnoses on clinical and particularly imaging features of the lesions. For differentiation between schwannomas and neurofibromas we applied US criteria of Ryu et al. that demonstrated high diagnostic accuracy.⁵ Unfortunately, no similar criteria are available for US diagnosis of perineurioma. Another limitation of the present study was its retrospective design; at the time of image analysis all projections needed for optimal differentiation between schwannomas and neurofibromas were therefore not available (Table 3).

In conclusion, the present study confirmed that PNTs are rare, but important cause of peripheral nerve dysfunction. We found US critically important for demonstration of PNTs, and published US criteria as useful to differentiate schwannomas from neurofibromas. Unfortunately, no such criteria are available for perineuriomas. PNTs are most likely when their continuity with peripheral nerves is demonstrated, and discrepancy between lesions’ large size and well preserved nerve function is found.