Diagnostic utility of anteroposterior measurements of the median nerve on sagittal ultrasonographic images and their correlation with clinical findings in carpal tunnel syndrome

To investigate the efficacy of sagittal ultrasonography of the median nerve in diagnosing carpal tunnel syndrome (CTS).


| INTRODUCTION
2][3] CTS causes symptoms such as numbness in the region innervated by the median nerve, hand or wrist pain, and fine motor deficits.Despite being such a common disease, the gold standard diagnostic criteria for CTS have not yet been established.
][6][7][8][9][10][11][12][13] Some scholars believe that CTS should be diagnosed based on clinical symptoms and does not require objective examinations, such as EDX testing. 14However, some conditions may cause symptoms similar to CTS, including cervical radiculopathy, peripheral neuropathy, and tendinitis and tenosynovitis around the hand.Therefore, we believe that objective and quantitative auxiliary tests, such as EDX testing and ultrasonographic examination (both of which have good diagnostic ability), are necessary to distinguish CTS from these differential diagnoses.It is important to correctly diagnose CTS because the treatment of CTS involves invasive surgery.
The evaluation of morphological changes in the median nerve using ultrasonographic techniques reportedly aids in diagnosing CTS.][12][13] Diagnostic criteria for CTS determined from the patient's manifestations and physical examination findings have also been reported.

Proposed diagnostic criteria for CTS include the Boston Carpal Tunnel
Questionnaire and the Carpal Tunnel Syndrome 6 (CTS-6) score.The Boston Carpal Tunnel Questionnaire diagnoses CTS based on the results of a patient questionnaire, while the CTS-6 score diagnoses CTS based on the medical history, clinical symptoms, and physical findings. 15,16Compared with the Boston Carpal Tunnel Questionnaire, the CTS-6 is shorter but has similarly high clinical validity and reliability, and is broadly used in diagnosing CTS. 17,18In patients with a positive CTS-6 result, ultrasonographic examination is reportedly more effective than EDX testing for the definitive diagnosis of CTS. 19rthermore, as ultrasonographic examination and the CTS-6 are sufficient in diagnosing CTS, EDX testing is not necessary in most cases. 20asuring of the cross-sectional area (CSA) of the median nerve by ultrasonography is a widely used method of diagnosing CTS.2][23][24] In contrast, it is unknown whether sagittal ultrasonographic measurements of the median nerve are correlated with the clinical symptoms of CTS and EDX test results, and whether these measurements are useful in diagnosing CTS.
The purpose of the present study was to investigate the efficacy of sagittal ultrasonography of the median nerve in diagnosing CTS.We compared the diagnostic performance of sagittal versus axial ultrasonographic imaging and examined the relationship between sagittal ultrasonographic imaging measurements and the EDX test results and CTS-6 assessment.
We hypothesized that the diagnostic performance of sagittal ultrasonographic imaging would be superior to that of axial ultrasonographic imaging.We also hypothesized that the results of sagittal ultrasonographic examination would correlate with the EDX test findings and CTS-6 score.

| MATERIALS AND METHODS
The present study was approved by the ethics committee of our hospital.The study was a case-control observational study.
The CTS group composed of 76 wrists (18 from men and 58 from women) of 59 patients with CTS, consisting of 14 men and 45 women with a mean age of 72.67 years (range 47-90 years).Of the 76 hands with CTS, 41 were the dominant hand and 35 were the non-dominant hand.The mean duration of CTS symptoms was 2.23 years (range 2 months-6 years), as determined through history taking from the patients.All included patients were continuously assessed from October 2011 to October 2015 at a single center.
Idiopathic CTS was diagnosed based on clinical and electrophysiological findings.One physician performed all the clinical evaluations.The clinical evaluation findings included sensory disturbance in the area innervated by the median nerve, nocturnal symptoms, a positive Phalen's test result, Tinel's sign, sensory test showing loss of two-point discrimination, fine motor impairment, decreased pinch strength, and thenar atrophy.EDX testing was performed in patients with clinical findings suggestive of CTS.All EDX testing was done by one investigator.The EDX test was defined as positive when the median distal motor latency (DML) was 4.2 ms or more, or the median sensory nerve conduction velocity (SCV) was 45 m/s or less. 10The exclusion criteria were: previous treatment for CTS comprising corticosteroid injection, splinting, or surgery; history of wrist trauma, wrist surgery, rheumatoid arthritis, connective tissue disorder, diabetes mellitus, or hemodialysis; and clinical diagnosis of radiculopathy, polyneuropathy, or mononeuropathy.
To determine reference values for the ultrasonographic examina- The CTS-6 is calculated as the total score for six questions regarding symptoms and physical findings (Table 1).CTS was diagnosed based on a CTS-6 score of 12 or higher. 16,20We calculated the CTS-6 score for all patients in the CTS group.

Findings Points
Numbness predominantly or exclusively in the median nerve territory We carried out ultrasonographic examination of the CTS and control groups using an ultrasound machine equipped with a 6-12 MHz linear array transducer (LOGIQ P6, GE Healthcare, Chicago, IL, USA).
For all study participants, ultrasonographic examinations were performed in the sitting position, with the forearm in the external rotated position and the wrist in the middle position.A sagittal image of the median nerve was acquired by placing the transducer parallel to the median nerve at the level of the carpal tunnel (Figure 1A).On a sagittal image, the maximum median nerve diameter (MND) in the proximal portion of the carpal tunnel and the minimum MND at the distal part of the carpal tunnel were measured at the distal radioulnar joint level and the middle of the capitate level, respectively.On a sagittal image, we identified the distal (middle of the capitate) and proximal (distal radioulnar joint) portions of the carpal tunnel using only ultrasonography, without using x-ray imaging.
To obtain an axial image of the median nerve, the transducer was placed perpendicular to the median nerve at the level of the pisiform (Figure 1B).We measured the CSA on an axial image of the median nerve using ImageJ (National Institutes of Health, Bethesda, MD, USA).All study participants underwent three ultrasonographic examinations, and the mean values were used in the analysis.All ultrasonography was performed by one orthopedic specialist with 15 years of experience.

| Statistical analysis
The distinctions between the CTS and control groups in age, height, and weight were evaluated using the t-test.The sex ratio of the CTS and control groups was compared using Fisher's exact test.
Receiver operating characteristic (ROC) curves were created Pearson's product moment correlation coefficient was also applied to assess the associations between the CTS-6 score and the ultrasonographic measurements.The Mann-Whitney U-test was applied to compare the ultrasonographic results between the CTS-6-positive and CTS-6-negative subgroups within the CTS group.
All statistical analyses were performed using EZR software, version 1.27 (Saitama Medical Center, Jichi Medical University, Saitama, Japan).Values of P < .05were considered statistically significant.

| RESULTS
The characteristics of the CTS and control groups are shown in Table 2.The CTS and control groups were similar regarding sex, age, height, and weight.
Among the total 76 hands with CTS, 63 hands were CTS-6-positive and 13 were CTS-6-negative.All 63 CTS-6-positive hands also had positive EDX test results, while 11 of 13 hands CTS-6-negative hands had positive EDX test results.For the two patients who were negative for both the CTS-6 and the EDX testing, the diagnosis of CTS was made based on the relief of symptoms following corticosteroid injection into the carpal tunnel.
Representative ultrasonographic images are presented in Figure 2. Figure 2A shows the sagittal ultrasonographic image of a patient in the CTS group.In the sagittal ultrasonography, the mean maximum MNDs were 0.252 cm and 0.202 cm, the mean minimum MNDs were 0.145 cm and 0.165 cm, and the mean MNSRs were 41.83% and 17.35% in the CTS and control groups, respectively.
Figure 2B shows the axial ultrasonographic image of a patient in the CTS group.In the axial ultrasonography, the mean CSAs were 0.138 and 0.093 cm 2 in the CTS and control groups, respectively.
The ROC curves were applied to calculate the ideal cut-off values for identifying CTS (Table 3).Among the ultrasonographic measurements, the MNSR showed the highest area under the ROC curve of 0.955.The ideal cut-off value for the CSA was 0.113 cm 2 , with a sensitivity of 84.2% and specificity of 90.0%.Compared with the CSA, the ideal cut-off value for the MNSR of 34.0% showed a higher sensitivity and specificity of 85.5% and 92.5%, respectively.
The transducer is placed longitudinally (A) and perpendicular (B) to the median nerve, respectively, to obtain sagittal and axial images of the median nerve in the proximal portion of the carpal tunnel.
We observed that the severity of delay in the EDX test correlated with the ultrasonographic findings.Table 4 shows the correlations between the ultrasonography results and the EDX test results in the CTS group.The CSA, maximum MND, and MNSR showed significant correlations with the DML and SCV.

| DISCUSSION
The present study assessed the usefulness of sagittal ultrasonographic imaging of the median nerve in diagnosing CTS, and evaluated the correlations between the ultrasonographic findings, EDX testing, and CTS-6 score.The CSA, maximum MND, and MNSR evaluated on ultrasonographic examination were significantly correlated with the CTS-6 and the EDX test results.Among the ultrasonographic measurements, the MNSR had the highest area under the ROC curve among the ultrasonographic measurements, and the cut-off value for the MNSR of 34.0% showed high a sensitivity and specificity of 85.5% and 92.5%, respectively.As far as we know, this is the first report to assess the correlations between sagittal ultrasonography of the median nerve, EDX testing, and the CTS-6 in patients with CTS.
We showed that sagittal ultrasonography of the median nerve was helpful in diagnosing CTS.
2][23][24] The present study showed that not only the CSA but also the maximum MND and the MNSR (ratio of the distal MND to the proximal MND in the carpal tunnel) correlated with the EDX test result.Animal studies have shown that mechanical compression of a peripheral nerve causes ischemic changes, resulting in intraneural edema and enlargement of the nerve in its proximal region. 25,26In CTS, the degree of nerve damage caused by constriction of the median nerve is reportedly related to anatomical changes in the median nerve, namely swelling proximal to the nerve stenosis. 21ese observations support the correlation between the degree of nerve damage and the MNSR, a quantitative representation of median nerve distal compression and proximal swelling in the carpal tunnel.
Therefore, the maximum MND and the MNSR were correlated with the EDX test result.
In the present study, the CSA, maximum MND, and MNSR evaluated on ultrasonographic examination were correlated with the CTS-6, which is used to evaluate the severity of CTS manifestations.
These ultrasonographic examination findings also significantly differed between the CTS-6-positive and CTS-6-negative subgroups within the CTS group.Similar to our study findings, Fowler et al. reported that the CSA in the proximal portion of the carpal tunnel was significantly larger in the CTS-6-positive group than the CTS-6-negative group among patients with CTS. 19Recently, it has also been reported that the CTS-6 scores are significantly greater in ultrasonography-positive patients with CTS than in ultrasonographynegative patients with CTS. 27When the CTS group was divided into ultrasound-positive and ultrasound-negative subgroups based on the ultrasonographic results in the current study, only the MNSR significantly differed between the two subgroups; however, the mean CTS-6 score was larger in the ultrasound-positive subgroup than the ultrasound-negative subgroup (data not shown).These previous study findings and our results lead us to speculate that, when assessing the shape of the median nerve in CTS, the CTS-6 score is related not only to the CSA determined on an axial image of the median nerve but also to the swelling of the median nerve in the proximal part of the carpal tunnel evaluated on sagittal ultrasonographic images.
Although one study reported that sagittal ultrasonographic examination was inconsistent in diagnosing CTS, 28 Okura et al. examined the intra-and inter-rater reliabilities of sagittal ultrasonographic measurements and reported that the correlation coefficients were greater than 0.8, indicating sufficient reliability. 1111][12][13] In a study similar to ours, the ideal cut-off values of the MND for the diagnosis of CTS on sagittal ultrasonography were reported to be 2.235 mm for the proximal carpal tunnel and 1.755 mm for the distal carpal tunnel. 7rthermore, another study reported that these morphological changes of the median nerve on sagittal images were observed in 90% of patients with CTS. 29

| CONCLUSION
We evaluated the merit of sagittal ultrasonographic imaging of the median nerve in diagnosing CTS, and evaluated the associations between the ultrasonographic findings, EDX testing, and CTS-6 score.
There was a significant correlation between the results of the sagittal ultrasonographic examination and the EDX test, which are both objective tests.The sagittal ultrasonographic findings also correlated with the CTS-6 results and appeared to reflect the severity of clinical symptoms.Among the ultrasonographic findings, the evaluation of the median nerve on sagittal images, especially the MNSR, had higher sensitivity and specificity compared with the CSA evaluated on an axial image.Thus, the MNSR on sagittal ultrasonography may be considered the first choice for the ultrasonographic diagnosis of CTS.
tion, 80 wrists (19 from men and 61 from women) of 47 asymptomatic volunteers, comprising 11 men and 36 women with a mean age of 71.41 years (range 48-94 years) were included as the control group.The control group consisted of hospital employees and their relatives, and comprised 33 bilateral hands and 14 unilateral hands (8 dominant hands, 6 non-dominant hands).The control group had no clinical manifestations and signs of CTS or neurological disorder.Control subjects were excluded if they had a history of wrist trauma or wrist surgery, history of diagnosis of CTS, rheumatoid arthritis, connective tissue disorder, diabetes mellitus, hemodialysis, or clinical diagnosis of radiculopathy, polyneuropathy, or mononeuropathy.

Note:
The corresponding point values for all positive findings are summed to obtain a total score.A score of 12 or more points was defined as positive for carpal tunnel syndrome.measured the median DML over the abductor pollicis brevis orthodromically while stimulating the elbow and the wrist 7 cm central to the measuring site.We recorded median sensory responses over the index finger antidromically and calculated the median SCV across the wrist.
using the mean values of the CSA, maximum MND, minimum MND, and MNSR obtained in the three ultrasonographic examinations of the CTS and control groups; the ideal cut-off values for the CSA, maximum MND, minimum MND, and MNSR of the median nerve in differentiating between the CTS and control groups were then calculated.The correlations between the ultrasonographic results, and the SCV and DML obtained from the EDX test were examined using Pearson's product moment correlation coefficient.

F I G U R E 2
Sagittal (A) and axial (B) ultrasound images of the median nerve (indicated by the arrow) in CTS.Distance (a) is the maximum median nerve diameter (MND), while (b) is the minimum MND.P = pisiform, R = radius, L = lunate, C = capitate.T A B L E 3 Optimal cut-off value, sensitivity, specificity, positive predictive value, and negative predictive value of the ultrasonographic measurements in distinguishing patients with CTS from controls.
30e present study had some limitations.First, the sample size was small.Second, we could not examine the usefulness of sagittal ultrasonographic measurements of the median nerve in patients who had CTS manifestations but had negative EDX test results, as 97.37% (74/76) of the patients with CTS in this study had positive EDX test results.This may be because the prevalence of positive EDX test results in patients with CTS reportedly increases with age, with abnormal neurophysiological changes detected in most patients 65 years or older.30Therefore,as the average age of patients with CTS in the present study was 72.67 years, most patients had positive EDX test results.The present cohort may have included many patients with severe CTS.Further research is needed to evaluate the usability of sagittal ultrasonographic measurements of the median nerve in patients with early-stage CTS who have mild symptoms but normal EDX test results.
13rt ¸ea et al. recently reported a method of diagnosing CTS using almost the same formula as ours based on measurements obtained from a sagittal ultrasonographic image of the median nerve.13Theyreported that a stenosis rate of