The pressure angle of the median nerve as a new magnetic resonance imaging parameter for the evaluation of carpal tunnel

1. Introduction 

Carpal tunnel syndrome (CTS) is the most frequently encountered peripheral compressive neuropathy. It affects 0.5% of the general population. One population-based study reported a symptomatic CTS prevalence rate of 3%, which was confirmed with nerve conduction studies [1], [2].

The combination of medical history, physical examination, and nerve conduction study findings is usually sufficient to diagnosis CTS. Nevertheless, there are cases in which nerve conduction study results or symptoms are ambiguous. Tinel’s sign and Phalen’s sign are moderately sensitive (20–70%) and specific (70–83%) [3], [4], [5], [6]. Electrodiagnostic tests such as electromyography (EMG) and nerve conduction study are 85–90% accurate in CTS patients, with a false-negative rate of 10–15% [7], [8]. Moreover, they may be invasive, painful, or uncomfortable for some patients. Nonetheless, while nerve conduction studies often indicate the severity of the lesion, they do not provide spatial information about the nerve or its surroundings, which could help in determining etiology. In recent years, imaging techniques, including magnetic resonance imaging (MRI) and ultrasonography (US), have been shown to be of value in the diagnosis of CTS. MRI offers the advantage of providing direct visualization of the carpal tunnel contents with high resolution and post-operative MRI might also be useful by demonstrating that the median nerve is adequately decompressed [9], [10], [11].

We compared clinical, electrophysiological, and MRI findings in patients with idiopathic CTS. Additionally, we present a new MRI parameter, the pressure angle of the median nerve (PAMN), for the evaluation of CTS.

2. Patients and methods 

This study employed a prospective case–control design and included 36 wrists with clinically diagnosed CTS in 21 patients that were followed-up for a mean duration of 24 months (range: 6–32 months). The control group consisted of 19 wrists in 10 healthy volunteers. All patients in this study had a diagnosis of CTS made by clinical history, physical examination, and electrophysiological investigation. The study protocol was approved by the hospital ethics committee and informed consent was obtained from all participants after they received an explanation of the purpose of the study. The patients were diagnosed with CTS considering at least one of the criteria (using clinical signs or sensory/motor nerve conduction evaluations) and were included in the study.

Upon the clinical diagnosis of CTS the following recruitment criteria were used [7]: (1) history of nocturnal paresthesias and pain in the hand; (2) exacerbation of symptoms with repeated flexion and extension movements of the wrist; (3) positive Phalen’ test result; (4) positive Tinel’ sign; (5) sensory deficit in the hand region innervated by the median nerve; (6) thenar atrophy.

The following criteria were used for the electrophysiological diagnosis of CTS [7], [12]:

•Median sensory nerve action potential (SNAP) peak distal latency>3.2ms.

•Difference between the SNAP peak distal latencies of the median and ulnar nerves ≥0.5ms, with ring electrodes recording the 2nd and 5th digits.

•Difference between the SNAP peak distal latencies of the median and ulnar nerves ≥0.4ms upon examination of the 4th digit.

•Decrease in the speed of median nerve sensory and motor conduction (<50ms) in the wrist.

•Median nerve compound muscle action potential distal latency>3.5ms.

•Spontaneous pathological activity recorded in needle EMG of the abductor pollicis brevis muscle.

All patients were classified into three groups according to the severity of subjective symptoms and objective clinical findings [13] as follows:

Patients meeting any one of the following criteria were excluded from the study: (1) age <18 or >70 years; (2) diabetes mellitus; (3) chronic renal deficiency; (4) thyroid disease; (5) polyneuropathy; (6) plexopathy; (7) radiculopathy; (8) pacemaker or other device that could interfere with or be adversely impacted by magnetic fields; (9) cognitive disorders; (10) rheumatoid arthritis; (11) history of trauma and previous surgery of the wrist; (12) double crush injury.

Nerve conduction studies and needle electromyography were performed by a clinical neurophysiology specialist (G.S.). All patients were evaluated using the same protocol and the electrophysiological assessments were carried out with a Medelec Synergy 5 canal NCS/EMG/EPS system electroneuromyography unit (Oxford Instruments, London). Skin temperature of the hand was kept constant above 32°C. Neurophysiological tests were used to grade CTS into the following categories according to American Association of Electrodiagnostic Medicine (AAEM) criteria (12):

MRI evaluation was performed on 36 wrists diagnosed with CTS based on history, clinical, and electrophysiological criteria. All cases were treated conservatively at the initiation of the study. Eleven patients responded to conservative therapy. Surgery was recommended to 25 patients who did not respond to conservative therapy; however, among these patients 18 gave consent to be operated. The pre- and post-operative clinical findings, electrophysiological findings, clinical staging, and MRI parameters were compared.

One neuroradiologist (F.S.) interpreted all of the images while blind to the clinical and electrophysiological findings. All MRI examinations were performed on a Philips Gyroscan Intera 1.5T MRI apparatus. All the wrists were imaged in the supine position and in the neutral position using a superficial C3 coil. The following parameters were used in the axial position: (a) 3-D-T1 FFE (gradient echo), TR: 25ms; TE: 4.7ms; flip angle: 35° (cross-sectional thickness 0.7mm), FOV: 120mm×120mm; NSA: 2 (cross-sectional thickness 1.5mm, interslice gap 0.8mm); (b) 3-D-T2 FFE (gradient echo), TR: 19ms; TE: 12ms; flip angle: 50° (cross-sectional thickness 0.7mm); FOV: 120mm×120mm; NSA: 2 (cross-sectional thickness 1.5mm, interslice gap 0.8mm); (c) STIR/TSE (spin echo), TR: 1500ms; TE: 15ms; T1 (160); FOV: 120mm×120mm; TSE factor: 4; NSA: 2 (cross-sectional thickness 3mm). T1-weighted sequences and T2-wighted fast spin echo sequences were used. All data were evaluated in the same center. The following parameters were evaluated during the MRI examinations:

View Large Image Download to PowerPoint

Fig. 1. MRI axial section of the wrist (T1-weighted sequence). Carpal arch width (A), carpal arch height (B) and the pressure angle of the median nerve (C).

Median nerve flattening was evaluated using MR1/MR2 and MR3/MR4 ratios.

3. Results 

3.3. MRI findings 

Details of the data concerning median nerve width and diameter at the pisiform and hamate levels, diameter of the carpal tunnel at the pisiform and hamate levels, median nerve flattening, and the pressure angle of the median nerve are shown in Table 1. Median nerve width (MR2) and carpal tunnel diameter (MR8) at the pisiform bone level were greater in the CTS wrists than in the controls. Median nerve diameter at the hamate bone level (MR3), carpal arch height (MR6), pressure angle of the median nerve (MR7), and median nerve flattening at the hamate bone level (MR3/MR4) were less in the CTS wrists than in the controls (P<0.000). Carpal arch height and PAMN significantly increased after surgery (3.5–7.5mm and 20–42° pre- and post-operative, respectively). On the other hand, median nerve diameter at the pisiform bone level (MR1), median nerve width at the hamate bone level (MR4), carpal arch width (MR5), carpal tunnel diameter at the hamate bone level (MR9), and median nerve flattening at the pisiform bone level (MR1/MR2) were not statistically different between the CTS and control wrists (P>0.05).

Table 1. Comparison of MRI findings in CTS group, control group and post-operative group

	Pre-operative	CTS group		Control group		MW	P
		Median	Mean±S.D.	Median	Mean±S.D.
	Comparison of MRI findings in CTS and control group
	MR1	4	4.14±0.49	4	3.92±0.53	251.5	0.084
	MR2	5	4.92±0.51	4.5	4.37±0.57	174	0.002
	MR3	3	3.01±0.33	4.5	4.45±0.55	18.5	0.0001
	MR4	4.5	4.47±0.51	4	4.24±0.63	270	0.186
	MR5	23	22.89±1.37	23	22.45±1.01	271.5	0.199
	MR6	3.5	3.5±0.7	5	4.71±1.12	129.5	0.0001
	MR7	20	20.28±2.04	26	26.95±3.52	9	0.0001
	MR8	13	12.85±1.51	11	11.03±1.21	121	0.0001
	MR9	11	10.82±1.16	11	10.89±1.33	340.5	0.978
	MR1/MR2	0.88	0.85±0.11	0.66	0.92±0.22	2.92	0.378
	MR3/MR4	0.66	0.68±0.06	1.12	1.08±0.25	10.5	0.0001

		Pre-operative		Post-operative		Z	P
		Median	Mean±S.D.	Median	Mean±S.D.
	Comparison of MR data obtained in the median nerve before and 8 weeks after surgical decompression
	MR1	4	4.14±0.49	4	3.86±0.38	−2.2	0.028
	MR2	5	4.92±0.51	4.8	4.5±0.64	−2.475	0.013
	MR3	3	3.01±0.33	4.5	4.64±0.48	−3.766	0.0001
	MR4	4.5	4.47±0.51	4.5	4.44±0.66	−0.618	0.537
	MR5	23	22.89±1.37	24	23.67±1.24	−1.705	0.088
	MR6	3.5	3.5±0.7	8	7.57±1.05	−3.739	0.0001
	MR7	20	20.28±2.04	42	42.39±2.73	−3.733	0.0001
	MR8	13	12.85±1.51	14	13.33±0.84	−0.655	0.513
	MR9	11	10.82±1.16	15	14.72±1.6	−3.735	0.0001
	MR1/MR2	0.88	0.85±0.11	1	1.06±0.16	−0.052	0.959
	MR3/MR4	0.66	0.68±0.06	1.12	0.88±0.18	−3.75	0.0001

		Pre-operative group		Control group		MW	P
		Median	Mean±S.D.	Median	Mean±S.D.
	Comparison of MRI findings in post-operative and control group
	MR1	4	3.86±0.38	4	3.92±0.53	158.5	0.683
	MR2	4.8	4.5±0.64	4.5	4.37±0.57	145.5	0.414
	MR3	4.5	4.64±0.48	4.5	4.45±0.55	144	0.39
	MR4	4.5	4.44±0.66	4	4.24±0.63	142.5	0.369
	MR5	24	23.67±1.24	23	22.45±1.01	73	0.002
	MR6	8	7.57±1.05	5	4.71±1.12	11.5	0.0001
	MR7	42	42.39±2.73	26	26.95±3.52	0	0.0001
	MR8	14	13.33±0.84	11	11.03±1.21	26	0.0001
	MR9	15	14.72±1.6	11	10.89±1.33	13	0.0001
	MR1/MR2	1	1.06±0.16	0.80	1.08±0.26	162	0.783
	vMR3/MR4	1.12	0.88±0.18	0.89	0.92±0.23	352	0.480

Comparison of pre- and post-operative MRI findings showed that the differences in MR1, MR2, MR3, MR6, MR7, MR9, and MR3/MR4 were statistically significant (Table 1). In the post-operative period, median nerve diameter (MR1) and width (MR2) at the pisiform level decreased, whereas median nerve diameter (MR3) at the hamate bone level increased. Post-operative median nerve width (MR4) and carpal arch width (MR5) at the hamate bone level did not change. In the post-operative period carpal arch height (MR6) and PAMN (MR7) significantly increased (P=0.0001). Additionally, carpal tunnel diameter (MR9) and the median nerve-flattening ratio (MR3/MR4) increased in the post-operative period (P=0.0001). Carpal tunnel diameter (MR8) at the pisiform bone level and median nerve flattening at the pisiform bone level (MR1/MR2) did not change significantly in the post-operative period.

4. Discussion 

In recent years imaging techniques using US and MRI have been shown to be of value in the diagnosis of CTS. Configuration of the median nerve is the parameter most widely investigated, not only with MRI, but with US, because delineation of this nerve is easy. With the advent of MRI there now exists a non-invasive technique that can accurately image the contents of the carpal tunnel and the anatomic relationships between these structures, as well as accurately measure the cross-sectional areas. Compared to MRI, US has the potential advantages of lower cost and shorter examination time [11]. The major advantage of MRI is that, due to high soft tissue contrast, it allows detailed imaging of both bone and soft tissue. This characteristic makes MRI far superior to either computerized tomography or real-time US for the valuation of the median nerve within the confines of the carpal tunnel. MRI offers the advantage of providing direct visualization of carpal tunnel contents and post-operative MRI might also be useful by demonstrating that the median nerve is adequately decompressed [9], [10].

This prospective study aimed to define objective imaging criteria for the diagnosis of CTS. The findings demonstrate that MRI is an optimal paraclinical tool for describing median nerve changes and for the diagnosis of idiopathic CTS.

We defined a new parameter, the pressure angle of the median nerve, which could be used along with previously known parameters for the evaluation of the median nerve and carpal tunnel with MRI. This angle was significantly lower in CTS wrists than in control wrists (P=0.000). PAMN showed a marked improvement 8 weeks after surgery compared to pre-operative values, although it was not statistically significant. It is possible that these parameters would have been significantly better if the post-operative period had been longer.

The best region for screening the median nerve diameter is generally the level of the pisiform bone because of the proximal diffusion of median nerve inflammation at the proximal carpal tunnel [8], [14]. Enlargement of the cross-sectional area of the median nerve at the proximal carpal tunnel, or at the tunnel entrance, was reported to be of diagnostic value for idiopathic CTS in the majority of cases studied [14], [15], [16]. Our findings of increased median nerve width at the pisiform bone level support this notion. We observed a significant reduction of the median nerve diameter and nerve flattening at the hamate hook level in CTS wrists. In particular, findings related to median nerve flattening show variation. Some studies reported no significant difference between idiopathic CTS cases and controls [17], whereas others reported more flattening at the hamate hook level in CTS cases [9], [14]. That is, median nerve enlargement seems to be more apparent in the proximal carpal tunnel, whereas median nerve flattening is a more important parameter in the distal carpal tunnel.

Our study revealed that carpal arch height and PAMN were significantly lower in all CTS wrists and that these parameters showed significant differences post-operatively. The other important point related to MRI is its use for post-operative evaluation. Surgical release of the nerve is frequently required when conservative measures fail. Despite adequate surgical decompression some patients remain symptomatic. Since the clinical signs and electrodiagnostic tests have a positivity ratio of 60–80% and a false negativity ratio of 15–20% in CTS patients [5], [6], [14], the identification of an objective and specific MRI finding for determining the necessity of surgery in cases with persistent symptoms or with suspected clinical and electrophysiological findings is expected to be useful. In 1989 Richman et al. [18] suggested measuring the carpal arch angle to determine post-operative morphological changes. They determined the carpal arch width and carpal arch angle measured between the hamate hook and trapezium beak, in addition to pre- and post-operative carpal canal measurements. The carpal arch angle was measured by connecting a line along the palmar ulnar and dorsal radial poles of the trapezium with a line drawn through the hook of the hamate. Carpal canal volume increased post-operatively; however, there was no significant difference between pre- and post-operative carpal arch width and carpal arch angle. They suggested that increased carpal canal volume was due to newly formed TCL in the anterior displacement, but not to the enlargement of the bone canal arch. This study was designed in consideration of the fact that surgery targeted the carpal ligament, but not the bony structure.

We thought that evaluating the distance of the median nerve (compressed by carpal ligament) to bony structures and its angle, in addition to the distance between bony structures would be objective and important in the pre- and post-operative periods. For this purpose we defined three new measurement parameters: carpal arch width, carpal arch height, and PAMN. We observed that carpal arch width was not significantly different between the CTS and control wrists; however, carpal arch height was significantly less (P=0.000) and the pressure angle of the median nerve was markedly less (P=0.000) in the CTS wrists than in the controls. In addition to these, symptoms either disappeared or were reduced, improvement of nerve conduction velocities was confirmed, and patient satisfaction was good post-surgery in 75% of the patients with severe stage CTS. In particular, post-operative improvement in clinical and electrophysiological findings was confirmed by the improvement of the pressure angle of the median nerve, as assessed radiologically. In contrast to some studies [9], [14] we did not observe a significant change in median nerve width at the hamate bone level, or carpal tunnel diameter and median nerve flattening at the pisiform bone level. Except for these two parameters, all other measured parameters changed significantly.

Pasternack et al. [19] investigated studies relating to the use of MRI for evaluating CTS. Although it is difficult to evaluate the sensitivity and specificity of the findings due to the heterogeneity of the studies, among 13 studies the comparison of MRI findings with values obtained from asymptomatic individuals revealed that the sensitivity and specificity, respectively, were as follows: enlargement of the cross-sectional area (35 and 84%), median nerve flattening (54 and 95%), flexor retinaculum bowing (70 and 93%), and increased T2 signal intensity of the median nerve (75 and 66%).

In a prospective blind study conducted in Turkey [20], median nerve flattening and increased median nerve intensity at the distal radioulnar joint level were significantly higher in patients with advanced-stage disease compared to those with early-stage disease. Similarly, median nerve areas at the hamate bone level correlated with clinical and electrophysiological parameters. It has been suggested in several studies that median nerve configuration and signal intensity are the indicators of disease severity [17], [21]. Contrary to these studies, we did not observe a statistically significant relationship between pre- and post-operative MRI findings, and clinical or EMG staging.

The present investigation has several drawbacks. First, the study population was small–36 CTS wrists and 19 control wrists. Secondly, our post-operative evaluation period was short. We believe that the findings might have reached a higher level of significance if there had more cases and a longer post-operative period.

Nowadays, MRI is not an ideal test since it is expensive, time-consuming, and not as readily available as US. MRI may be of use in the investigation of unconfirmed or complex cases, and in cases in which surgical treatment has failed. In conclusion, the results of our study show that MRI is the ideal paraclinical test for evaluating patients with idiopathic CTS. We think that the pressure angle of the median nerve will be used as a non-invasive method, especially for post-operative evaluation of CTS.