Ultrasound Cross-Sectional Area in Median Nerve Axonal Loss and Demyelination in Carpal Tunnel Syndrome

Abstract

Introduction

Ultrasound is an effective diagnostic tool for carpal tunnel syndrome (CTS). However, it is unclear how ultrasound correlates with axonal loss and/or demyelination on electrodiagnostic studies (EDS). The objective of this study is to determine whether ultrasound cross-sectional area (CSA) of the median nerve varies between patients with axonal loss or demyelination.

Methods

A retrospective review was completed of patients who presented to an orthopaedic hand clinic with numbness/paresthesias over a 6-year period. Demographics, CTS symptoms scale 6 (CTS-6) scores, Boston Carpal Tunnel Syndrome Questionnaire (BCTQ) scores, EDS results, and ultrasound results were collected. Median neuropathies were graded as normal, demyelination, or axonal loss using EDS reports. The data were analyzed with chi-square and t-tests.

Results

In all, 383 hands were included (92 axonal loss, 182 demyelination only, and 108 neither). The average patient age was 52.2 and the average body mass index (BMI) was 31.7. The group consisted of 70.7% females, and 23.2% had diabetes. Patients with either axonal loss or demyelination had larger CSA and higher CTS-6 and BCTQ scores than patients with negative EDS. Patients with axonal loss also had larger CSA and higher CTS-6 and BCTQ scores than patients with demyelination only. The rates of positive ultrasound results between axonal loss and demyelination groups did not differ until the ultrasound cutoff was increased from 10 to 12 mm².

Conclusion

Rates of positive ultrasound results (CSA ≥ 10 mm²) do not differ between wrists with axonal loss or demyelination alone. Therefore, the character of carpal tunnel neuropathy does not affect ultrasound’s diagnostic ability. Additionally, CSA increases as wrists develop axonal loss, and an increased ultrasound cutoff of 12 mm² is correlated with this pathology.

Introduction

Carpal tunnel syndrome (CTS) is the most common peripheral compressive neuropathy globally.¹ Although prevalence estimates vary widely, it is thought that up to 1 in 10 people experience it at some point in their lives.² Like any compressive neuropathy, CTS is characterized by gradual demyelination of the nerve at the site of compression (in this case, the median nerve as it passes through the carpal tunnel).³ As the disease progresses, axonal loss begins to occur.4, 5, 6 Patients may present anywhere along the spectrum of early demyelination to chronic axonal loss. Axonal loss is generally considered irreversible, and its presence has implications for outcomes after carpal tunnel release.⁷

Electrodiagnostic studies (EDS) have been well established as a diagnostic tool for CTS.^8,9 Although more costly and uncomfortable for the patient than clinical diagnosis alone, EDS offers the benefit of grading severity of nerve disease and determining whether the nerve has experienced demyelination or axonal loss.^10,11 Nerve conduction studies (NCS) primarily evaluate sensory nerve and compound motor action potentials (sensory nerve action potential [SNAP] and compound muscle action potential [CMAP]), evaluating their amplitude and latency for aberrancy.¹⁰ Electromyography (EMG) is often used to supplement NCS, and certain abnormalities in the abductor pollicis brevis muscle such as positive sharp waves and fibrillations can indicate axonal loss in the median nerve.¹¹

In recent years, ultrasound examination has been increasingly utilized to aid in the diagnosis of CTS. Studies have shown that ultrasound may offer similar diagnostic accuracy when compared to NCS without added cost or patient discomfort.^12,13 In general, median nerves affected by CTS have a larger cross-sectional area (CSA) at the carpal tunnel inlet.^14,15 Some more recent studies have suggested that, in more severe disease, ultrasound may be less reliable.16, 17, 18 This finding has been postulated to be due to nerve shrinkage in the setting of axonal loss with perineural fibrosis rather than swelling. The purpose of this study is to investigate median nerve CSA at the carpal tunnel in nerves with EDS evidence of demyelination versus axonal loss to better understand ultrasound’s diagnostic ability in more advanced disease.

Materials and Methods

Institutional review board approval was obtained prior to the initiation of this study. A retrospective review was completed on a database of all patients who presented to an outpatient orthopaedic hand clinic with a chief complaint of hand numbness or paresthesias from October 2014 to September 2020. Demographic data, CTS symptoms scale 6 (CTS-6) score, Boston CTS Questionnaire (BCTQ) scores, EDS results, and ultrasound CSA measurements were collected. The ultrasound examinations were performed utilizing a 15-6 MHz linear-array transducer (SonoSite M Turbo; SonoSite, Quest Imaging Solutions, Las Vegas, NV, United States). Patients with isolated ulnar neuropathies by EDS were excluded. Any patients who did not undergo either EDS or ultrasound examination were also excluded. Additionally, patients with bifid median nerves were excluded, as no standardized CSA measurement could be used. Many patients had bilateral diseased extremities included, and many had their well hand included in the database, including separate symptom scores. If all necessary data were present, contralateral hands were included as separate entities.

Nerves were categorized as normal, demyelination only, or axonal loss based on preoperative EDS results. For the purpose of this study, any nerve with axonal loss was placed in the axonal loss category, regardless of the presence or absence of concomitant demyelination. Where provider impressions were available, these were used for categorization. Seventy-four percent of cases had a complete report that explicitly stated the presence or absence of demyelination and axonal loss. In patients for whom expert impressions were not available or did not explicitly comment on the presence or absence of axonal loss (even though most did comment on demyelination), further classifications were made by one of the authors based on the EDS data. A second author weighed in on any unclear cases. Subjects with the presence of fibrillations, positive sharp waves, or polyphasic action potentials on EMG were classified as having axonal loss.¹⁹ Demyelination was identified on NCS using reference values for conduction velocities less than the 3rd percentile and distal latencies greater than the 97th percentile based on the patient’s age group as established by Chen et al.²⁰

Ultrasound examination was performed by the senior author on all wrists included in the study, and the CSA measurements were included in the database. CSA measurements were taken at the carpal tunnel inlet at the level of the pisiform, with the wrist in neutral position and the fingers in a resting cascade (Fig. 1). A positive ultrasound result was considered a CSA of 10 mm² or more. Chi-square and Student’s t-tests were performed to compare data between EDS negative and EDS positive (either axonal loss or demyelination) hands, as well as between axonal loss and demyelination groups, with p-value less than 0.05 considered statistically significant.

Fig. 1.

Ultrasound images of median nerves at the carpal tunnel inlet. Cross-sectional areas observed: 24 mm² (top) and 10 mm² (bottom).

Results

Out of 337 patients in the database, 259 patients met the inclusion criteria, for a total of 383 hands included in the study. Bilateral hands were included in 47.5% of patients. Female patients accounted for 70.7% of the patients and the average age was 52.2 years (SD: 14.6). Most patients were Caucasians (74.1%) and nondiabetic (76.8%). The average body mass index (BMI) was 31.7. In all, 182 hands (47.5%) were identified as having demyelination only, 93 (24.3%) as having axonal loss, and 108 (28.2%) as having neither.

Overall, 275 of 383 hands (71.8%) were EDS positive. Patients with either axonal loss or demyelination had higher CTS-6 scores compared to those with negative EDS results (14.2 ± 6.0 vs. 11.5 ± 7.3, p < 0.001). Similarly, the BCTQ scores were significantly higher in the EDS positive group, although only for the symptom severity and not the functional status scales (Symptom Severity Scale [SSS] 2.9 ± 0.8 vs. 2.7 ± 0.9, p = 0.033). Seventy-six percent of EDS positive hands tested positive on ultrasound, significantly more than the EDS negative group (45.4%). The mean CSA in the EDS positive group was significantly higher than in the EDS negative group (11.7 ± 3.7 vs .9.5 ± 3.5 mm², p < 0.001; Table 1).

Table 1.

Axonal loss or demyelination versus neither

	EDS positive (N = 275)	EDS negative (N = 108)	p-Value (CI = 95%)
	Average (standard deviation)	Average (standard deviation)
CTS-6	14.2 (6.0)	11.5 (7.3)	< 0.001
BCTQ SSS	2.9 (0.84)	2.7 (0.89)	0.033
BCTQ FSS	2.3 (0.88)	2.2 (0.85)	0.164
CSA	11.7 (3.7)	9.5 (3.5)	< 0.001
Ultrasound	n (%)	n (%)
≥ 10 mm2	209 (76.0)	49 (45.4)	< 0.001
< 10 mm2	66 (24.0)	59 (54.6)

Abbreviations: BCTQ, Boston Carpal Tunnel Questionnaire; BMI, body mass index; CSA, cross-sectional area; FSS, Functional Severity Scale; SSS, Symptom Severity Scale.

Patients diagnosed with axonal loss in particular had higher CTS-6 scores (15.3 ± 6.0 vs. 13.6 ± 6.0, p = 0.022) and higher BCTQ SSS scores (3.1 ± 0.81 vs. 2.9 ± 0.85, p = 0.045) compared to patients with demyelination alone. The mean ultrasound CSA measurements of the median nerve were significantly greater in axonal loss compared to demyelination (13.0 ± 4.6 vs. 11.1 ± 3.0 mm², p < 0.001). However, there was no difference in the proportion of patients with ultrasound results deemed positive between the groups (Table 2). When the positive ultrasound cutoff was increased to CSA of 12 mm² or greater, however, nerves with axonal loss were significantly more likely to be ultrasound positive than nerves with demyelination alone (p = 0.002).

Table 2.

Axonal loss versus demyelination

	Axon loss (N = 93)	Demyelination (N = 182)	p-Value (CI = 95%)
	Average (standard deviation)	Average (standard deviation)
CTS-6	15.3 (6.0)	13.6 (6.0)	0.022
BCTQ SSS	3.1 (0.81)	2.9 (0.85)	0.045
BCTQ FSS	2.4 (0.93)	2.3 (0.85)	0.196
CSA	13.0 (4.6)	11.1 (3.0)	< 0.001
Ultrasound	n (%)	n (%)
≥ 10 mm2	73 (78.5)	136 (74.7)	0.296
< 10 mm2	20 (21.5)	46 (25.3)

Abbreviations: BCTQ, Boston Carpal Tunnel Questionnaire; BMI, body mass index; CSA, cross-sectional area; FSS, Functional Severity Scale; SSS, Symptom Severity Scale.

Discussion

In the early stages of CTS, the median nerve undergoes demyelination resulting in slowed signal conduction. As the disease progresses, axonal loss may occur. While ultrasound of the median nerve is an effective first-line diagnostic tool when there is clinical concern for CTS, it may not necessarily distinguish between axonal loss and demyelination. Furthermore, nerve size may change over the course of the disease with the development of axonal loss, which may affect the utility of ultrasound examinations. Given the safety, effectiveness, and widespread accessibility of ultrasound as an outpatient diagnostic tool, there is considerable appeal in being able to distinguish axonal loss from demyelination in CTS using ultrasound criteria. This study explores ultrasound findings in nerves affected by axonal loss versus demyelination alone as defined by EDS in order to compare the diagnostic ability of ultrasound in either setting.

Unsurprisingly, the presence of CTS on EDS (either axonal loss or demyelination alone) was found to be associated with higher CTS-6 and BCTQ scores and greater nerve CSAs on ultrasound. When comparing nerves with axonal loss versus demyelination alone, we found that axonal loss is similarly associated with higher CTS-6 and BCTQ scores, and, notably, greater CSA. Using the commonly cited cutoff of 10 mm², there was no significant difference in the percentage of nerves deemed ultrasound positive between the axonal loss and demyelination groups. When the cutoff was raised to 12 mm², nerves were deemed ultrasound positive significantly more often in the axonal loss group.

Our study’s findings are somewhat in contrast to those of Moghtaderi et al,¹⁷ who found that although ultrasound-measured median nerve CSA was greater in patients affected with CTS than those without, CSA was significantly greater in nerves categorized as having only moderate CTS than those with severe CTS by EDS. It is difficult to directly compare findings, as Moghtaderi et al did not divide subjects into a binary (axonal loss or demyelination), but rather a graded severity of electrodiagnostic results. However, given the pathogenesis of CTS, it can be assumed that patients with EDS evidence of axonal loss would fall into the moderate or severe categories. Moghtaderi et al’s study did not report whether there was any difference in ultrasound sensitivity or specificity in different disease severity, and they were limited by a small sample size with only 36 nerves affected by CTS. Because of Mogthtaderi et al’s findings, we initially hypothesized that in extremely advanced CTS with axonal loss, the nerve would undergo shrinkage at the carpal tunnel. However, our findings suggest that nerve CSA continues to increase as the disease progresses, including into a period of axonal loss, at least within the symptom duration by which the majority of patients have presented for evaluation.

A recent study by Martikkala et al,¹⁸ on the other hand, supports our findings. In their study, Martikkala et al compared the median nerve CSA to NCS-graded CTS severity including negative, mild, moderate, and severe categories with a sample size of 259 wrists. They found a positive correlation between CTS severity and CSA at the wrist. In pairwise comparisons, CSA in negative versus mild, moderate, or severe groups differed significantly, but the CSA differences between the NCS positive groups did not reach statistical significance. Martikkala et al additionally included analysis of wrist-to-forearm ratio (WFR), a ratio of the CSA of the median nerve at the wrist and at the forearm. WFR was positively correlated with CTS severity, and there was a significant difference found in pairwise comparisons of negative to mild and mild to moderate disease. Similar to Martikkala et al, El Miedany et al²¹ found that median nerve CSA at the wrist increases with worsening EDS severity of disease, and this correlation was strong enough in their study to support distinct CSA cutoff ranges for each category: mild, moderate, and severe (10–13, 13–15, and > 15 mm², respectively).

Another study by Martikkala et al,¹⁶ similar to ours, evaluated the relationship between axonal loss and median nerve CSA on ultrasound. In addition to reporting wrist CSA and WFR as in the aforementioned study, they also reported forearm CSA. Axonal loss was graded based on EMG studies. They found forearm CSA correlated negatively with the severity of EMG changes—the worse the axonal loss, the smaller the nerve at the forearm. The WFR correlated positively with the severity of the EMG findings in moderate CTS, but lost its correlation in severe and extreme CTS. Altogether, WFR was the highest when slight axonal loss was present. This indicates that, as axons are lost, the median nerve likely shrinks. However, at the site of compression (the wrist), perineural edema causes the CSA to remain enlarged even into moderate axonal damage. This corroborates our findings that the median nerve at the wrist can be expected to be even larger in many cases of axonal loss than in demyelination alone.

Our findings of increased wrist CSA in the setting of axonal loss demonstrate that if nerve shrinkage occurs in advanced CTS, this does not happen immediately as axonal loss begins. The average hand surgeon who performs ultrasound as a diagnostic adjunct for CTS can expect to see patients with axonal loss test positive with enlarged CSA with the same frequency as nerves that have undergone demyelination only. Although ultrasound cannot definitively distinguish between axonal loss and demyelination, a wrist CSA measurement of 12 mm² or more may be a clue that axon loss is occurring. These findings reaffirm ultrasound’s ability to detect and, to a certain extent, grade CTS, even in the majority of cases that have progressed to axonal loss.

This study has a few limitations. First, there is some variability in the methods used to place subjects into axonal loss or demyelination categories. Interpretations in the reports were relied upon when present, and these came from multiple providers. When expert interpretation was not present or unclear, the authors assigned classifications based on limited EMG and NCS criteria. Therefore, classifications were not completely standardized, leaving room for error. Another limitation of this study is that it treats demyelination and axonal loss as clearly distinct categories, when the two exist on a spectrum with gradual disease progression. Our results may have differed with subcategories for mild versus severe axonopathy; however, this would have decreased sample sizes and presented an additional classification challenge. Additionally, ultrasound measurements were taken by a single operator (the senior author) who was not necessarily blinded to the EDS results when taking the measurements. However, in our experience, most patients present to the clinic for initial evaluation without having undergone electrodiagnostic testing, so the majority of ultrasound measurements were taken without the knowledge of these results.

In future studies, it would be beneficial to determine if, and at what point, wrist CSA begins to decrease in CTS. Current studies show mixed reliability of ultrasound in late stages of CTS, and the reason for this trend remains unclear. While this study supports the diagnostic capacity of ultrasound for CTS even in the setting of axonal loss, some provider interpretation is still required to identify and optimize treatment for this condition in a clinical context. In that vein, it would be interesting to compare outcomes after carpal tunnel release for different stages of axonal loss along with their corresponding preoperative CSA. Future studies could investigate these questions further.

Conclusion

Ultrasound of the median nerve at the carpal tunnel demonstrates enlarged CSA in the setting of both demyelination and axonal loss. The rate of nerves testing positive on ultrasound with a cutoff of 10 mm² does not differ between these two pathologies. When axon loss is present, CSA is higher, and a cutoff of 12 mm² captures significantly more wrists with axon loss than demyelination. Axonal loss is also associated with older age and worse CTS-6 scores. Providers using ultrasound as a diagnostic adjunct for CTS can rely on the modality to pick up patients with both axonal loss and demyelination alike.