Relative Motion of the Connective Tissue in Carpal Tunnel Syndrome: The Relation with Disease Severity and Clinical Outcome

Abstract

Excursion of the median nerve and the surrounding subsynovial connective tissue (SSCT) is diminished in patients with carpal tunnel syndrome (CTS). This study sought to determine if SSCT excursion could be utilized to predict surgical outcome. Idiopathic CTS patients were reviewed with ultrasound and electrodiagnostic tests at baseline. A speckle tracking algorithm was used to determine SSCT relative to tendon motion (shear index). Analyses of variance tests were used to compare SSCT motion with disease severity at baseline. Adjusted linear regressions were used to test the association with patient-reported outcome. A total of 90 CTS patients were analyzed and showed an average shear index of 79% (95%CI: 76.3–81.6%). SSCT motion was lower in CTS patients with increasing electrophysiological severity (p=0.0475). There was no significant association of preoperative SSCT motion with symptomatic improvement (p=0.268). Overall, SSCT motion is decreased in CTS patients, but shows limited correlation with clinical severity.

Introduction

Carpal tunnel syndrome (CTS) is a compression neuropathy of the median nerve at the level of the carpal tunnel. The diagnosis is most commonly made based on clinical presentation supported by functional evaluation of the nerve with electrodiagnostic studies. Recently, ultrasound (US) has gained a more profound role in the CTS work-up due to its painless nature, cost-effectiveness (Fowler, et al. 2013), and comparable sensitivity and specificity (Fowler, et al. 2015, Kwon, et al. 2008, Nakamichi and Tachibana 2002, Naranjo, et al. 2007) when using increased nerve cross-sectional area as the parameter of interest (Klauser, et al. 2009). However, these changes in nerve morphology are most likely the physical manifestation of a longstanding process of nerve compression, the initiating factors of which are not well understood. In clinical practice, disease severity is an important factor in the decision to initiate conservative treatment or to proceed to surgical intervention. A non-invasive method of both determining severity and predicting the likelihood of success of operative or non-operative treatment would benefit the individual patient. Although surgery is highly effective, there is a wide variation in reported success rates (Bland 2007), suggesting again that better case selection might improve treatment outcomes. Electrodiagnostic studies are the gold standard, but these tests are invasive and their usage has been a topic of debate (Fowler 2017, Sonoo, et al. 2018).

Besides the well-described increase in median nerve size, a common feature in CTS patients is non-inflammatory fibrotic thickening of the subsynovial connective tissue (SSCT) (Ettema, et al. 2004, Kerr, et al. 1992, Lluch 1992). The SSCT lies below the visceral layers of the bursae in the carpal tunnel and consists of multiple layers of interconnected collagenous fibers, including blood and lymphatic vessels, which directly surround the flexor tendons and the nerve. This structure is a unique feature seen only in the carpal tunnel (Gelberman, et al. 1992, Guimberteau, et al. 2010). As the fingers flex and extend, the SSCT layers are gradually recruited, transmitting and absorbing mechanical stress (Filius, et al. 2014, Vanhees, et al. 2012), and providing a dynamic framework to protect the supply of nutrients and the nerve from excessive strain (Guimberteau, et al. 2010). Normally, this results in a situation where the SSCT moves in coordination with, but at a lower velocity than, the underlying flexor tendon (Ettema, et al. 2006, Oh, et al. 2007, Tat, et al. 2013). Typically, the SSCT is less than 1 mm thick, but in patients with CTS, it thickens and the connections between the layers are disrupted (Ettema, et al. 2004) causing changes in the layered gliding system (Tat, et al. 2015). Peripheral nerves have a physiological reserve when it comes to longitudinal stretching, but with increased perineural fibrosis (Tat, et al. 2015) this may be diminished, and sections of the median nerve could be prone to increases in local strain. With as little as 6% strain already shown to affect nerve function (Kwan, et al. 1992), alterations in mechanical SSCT response are a relevant topic of investigation and may help explain the genesis of nerve compression and ischemia in patients with idiopathic CTS (Festen-Schrier and Amadio 2018, Lluch 1992).

SSCT motion has been measured by Doppler ultrasound as well as by commercial and custom speckle tracking algorithms (Ettema, et al. 2006, Filius, et al. 2015, Korstanje, et al. 2012, Oh, et al. 2007, Tat, et al. 2013, Tat, et al. 2015, Van Doesburg, et al. 2012, Yoshii, et al. 2009), showing the feasibility of these methods as well as increased SSCT/tendon shear strain in patients with CTS. However, these were either done using cadaveric specimens (Ettema, et al. 2006, Oh, et al. 2007) or focused solely on the diagnostic potential of relative SSCT motion (Filius, et al. 2015, Korstanje, et al. 2012, Tat, et al. 2013, Tat, et al. 2015). Our group has recently validated and tested the reliability of an extension of a previously reported speckle tracking algorithm (Korstanje, et al. 2010) optimized to track small, fast-moving tissue such as the SSCT (Bandaru, et al. 2018). The present study describes the application of this algorithm to a large prospectively-gathered dataset of US images in order to evaluate flexor tendon and SSCT motion in CTS patients undergoing carpal tunnel release. The two main aims of this study were to 1) test whether there is a difference in relative SSCT motion between healthy participants and CTS patients of different severities and to 2) determine whether these preoperative measures were associated with clinical outcome after carpal tunnel release.

Materials and Methods

General

This is a prospective clinical (level I) study utilizing the surgical arm of a larger research trial (CT.gov identifier NCT02219555) which was reviewed and approved by our Institutional Review Board (S-1). Written informed consent was obtained from all participants in accordance with the Declaration of Helsinki (Association 2013).

Subjects and disease characteristics

The control group was a non-matched sample of subjects derived from a similar study that ran parallel to the current study using the same US protocol. Participants were eligible for the control group if they were over the age of 18 and had never experienced any classic symptoms of CTS such as numbness or tingling in the distribution area of the median nerve. Exclusion criteria included the presence of rheumatoid arthritis, osteoarthritis or traumatic injuries of the ipsilateral hand or wrist. All imaging was done according to a preset imaging protocol described below.

CTS patients were recruited at our institution at a specialized hand clinic. CTS diagnoses were confirmed by orthopedic or plastic surgeons based on a combination of symptoms at presentation, clinical tests, and nerve conduction studies with/without needle electromyography, referred to as electrodiagnostic studies (EDS). Inclusion criteria were age between 21–80 years, symptoms of numbness or tingling for at least four weeks in at least two digits in one hand including the thumb, index, long or radial border of ring finger, ability to understand spoken and written English, and a decision for surgical release. Exclusion criteria were a previous carpal tunnel release in that hand, tumor, space occupying lesion or deformity in the hand/wrist, pregnancy-induced CTS, and any of these diagnoses: cervical radiculopathy, peripheral nerve disease, thyroid disease, rheumatoid arthritis or other inflammatory arthritis, osteoarthritis in the wrist, diabetes, renal failure, sarcoidosis, amyloidosis or major trauma to the ipsilateral hand or wrist. Clinical evaluations included sensibility (two-point discrimination) in the median nerve distribution of the hand, Phalen’s test, Tinel’s sign, pinch and grip strength, and assessment of thenar muscle atrophy. In cases of bilateral CTS, only the more severe hand was included. As part of the routine diagnostic workup, most patients received electrodiagnostic tests, in each case according to the guidelines of the American Association of Neuromuscular and Electrodiagnostic Medicine (Jablecki, et al. 2002). Disease classification was done with a five scale ordinal categorization (normal, mild, moderate, severe, very severe) (Witt, et al. 2004).

Patient characteristics and CTS data

Patient demographics were collected at baseline and included age, sex, affected side, bilateral presence of symptoms, dominance, disease duration, BMI, visual analog scale for pain, and mood scores assessed using the Center for Epidemiologic Studies Depression scale (Radloff 1977). For this study, disease severity was classified in two ways: EDS-based classification and patient-reported outcome using the Boston Carpal Tunnel Questionnaire (BCTQ) (Levine, et al. 1993). The BCTQ was completed at baseline, and at 1 and 3 months post-surgery, with the 1 month follow-up done via mail and the 3 month follow-up completed in clinic. The BCTQ consists of two subscales, resulting in a symptom severity score (SSS) and a functional severity score (FSS). Scores range between 1 (no symptoms) to 5 (most severe) and for this study were categorized based on a system proposed by Storey et al. (Storey, et al. 2009): 1.0 =‘asymptomatic’, 1.1–2.0 = ‘mild’, 2.1–3.0 = ‘moderate’, 3.1–4.0 = ‘severe’, and 4.1–5 = ‘very severe’. For the outcome measures of the post-surgical association, the difference in average BCTQ scores was used (baseline score - follow-up score) and averaged separately for the two subscales.

Ultrasound acquisition protocol

Patients and controls were tested in the supine position with the upper extremity on a plastic support board at a ~ 45° shoulder abduction angle. Subjects were then asked to flex and extend the third digit to a frequency of fifty beats per minute, paced by a metronome. After a practice session, three clips were recorded in the longitudinal view at the carpal tunnel inlet using the lunate and the distal radius as anatomical landmarks. All recordings were made using a Philips iE33 ultrasound scanner (Royal Philips Electronics, Amsterdam, the Netherlands) equipped with a 15L7 linear array transducer, set at a focal depth of 1.5cm. All measurements were done by ultrasonographers with a minimum of 1 year of experience in this field and two physicians (VS, SE) who were blinded to the clinical outcome and who were trained by the same ultrasonographer using a predetermined protocol.

Image analysis

A full and detailed description of the image analysis protocol has been published previously and showed high reliability (Schrier, et al. 2019). Images were analyzed using a Matlab (R2016a, The MathWorks Inc., Natick, MA, 2000) based custom-made speckle tracking algorithm (Korstanje, et al. 2010) enhanced with a singular value decomposition filter (Bandaru, et al. 2018). A region of interest (ROI) was manually placed on both the third flexor tendon (FDS) and the overlying SSCT (Fig. 1). Then, the ROI was automatically segmented in 40 and 24 kernels for, respectively, the tendon and SSCT. A normalized cross-correlation analysis was applied to calculate the frame-to-frame displacement. Finally, as described by Bandaru et al. (2008), unreliable kernel results were removed based on their correlation values and discordant vectors. The average of the displacement vectors of the reliable kernels gave the final frame-to-frame displacement vector. Analyses were done by a single rater blinded to patient data and clinical outcome. The average length of the x-axis vector was calculated over the course of a single movement based upon three cycles (three flexion and extension). The shear index is a measure introduced in previously published literature (Yoshii, et al. 2009) and describes the relative motion between the FDS and SSCT. It is defined as ([tendon excursion] – SSCT excursion])/[tendon excursion]*100. A value close to 0% indicates a 1:1 movement between the SSCT and the tendon, whereas 100% would indicate a complete dissociation. Additionally, the ratio between maximum velocities (MVR) of the SSCT and FDS was calculated based on the average peak values from three cycles.

Fig. 1:

Longitudinal ultrasound image showing the tendon and overlying hyperechoic SSCT. The overlays represent regions of interest of the third superficial flexor tendon (continuous line) and SSCT (interrupted line). Images were corrected for rotational angle before speckle tracking analysis.

Statistical analysis

Patient demographics were reported based on their distribution either as means including standard deviation or medians including their first and third quartiles. Categorical data is shown as counts and percents. Descriptive statistics of the excursions and US outcomes were expressed in means and standard deviation. To test aim 1: The relation between the shear index or maximum velocity ratio and disease severity was assessed using a one-way analysis of variance or Kruskal Wallis test based on data distribution followed by multiple means comparisons correction (Tukey HSD or Steel Dwass, respectively).

For aim 2, two approaches were used to test the association between baseline US parameters and clinical outcome. First, a linear regression model was applied using the absolute change in separate BCTQ scores as main outcomes. Tested covariates as potential confounders included the baseline BCTQ score, sex, age, disease duration, BMI, and baseline depression score. To obtain consistent estimates of the variance of the model caused by heteroscedastic residuals White robust variance estimators (White 1980) were used. Secondly, we categorized patients based on their shear index as either “low”, “normal”, or “high” using the 95% confidence interval of the healthy control group as thresholds. This was used as the independent variable of a logistic regression model adjusted for the same confounders. Here, we dichotomized the CTS-SSS scale to “success” and “not success” based on the relative minimal clinically important difference (De Kleermaeker, et al. 2019). Type I error rate was set at 0.05. All analyses were done using SAS™ (version 9.4; SAS Institute, Cary NC) and JMP (Version 14, SAS Institute Inc., Cary, NC, 1989–2007.).

Results

Descriptive Overall Results

A total of 124 CTS participants completed the baseline assessment, of whom 90 (71%) had US images which could be analyzed for SSCT movement. The main reason for image exclusion was out-of-plane motion of the SSCT or the tendon. In this sample, follow-up rate was 72% and 90% for 1 and 3 months follow-up, respectively. The demographics and the distribution over the different disease severities are described in Table 1. A total of seventeen participants were included in the control group with an average age of 37.9 (SD: 12.0) of whom eight females (47%). At baseline, there was a mean shear index of 79% (SD: 13%), resulting from 1.2 mm (SD: 0.4 mm) excursion of the tendon and 0.24 mm (SD: 0.15 mm) total excursion of the SSCT (Table 2). On visual inspection, no gross differences were found echographically between SSCT imaging characteristics in healthy or CTS participants. In individual cases, the SSCT would appear as a more hyperechoic, thicker band superficial to the flexor tendon, but without association with the participants’ sex, profession or disease severity.

Table 1.

Baseline characteristics of CTS patients

Age, years (mean, SD)	55.8±14.6
Sex (female)	55 (61%)
Duration of symptoms, months (median, IQ1–3)	24 (12–44)
Surgery on dominant hand (yes)	55 (63%)
Bilateral CTS (yes)	52 (60%)
BMI, kg/m2 (mean, SD)	31.8±6.4
Pain score, visual analogue scale (median, IQ1–3)	3 (2–5)
EDS severity (%)
Normal	2 (2%)
Mild	25 (28%)
Moderate	41 (46%)
Severe	19 (21%)
Very severe	1 (1%)
No NCS available	2 (2%)
Symptom Score severity (%)
Asymptomatic	0 (0%)
Mild	7 (8%)
Moderate	35 (39%)
Severe	40 (44%)
Very severe	7 (8%)
Missing/ incomplete	1 (1%)

BMI: body mass index, IQ: Interquartile, EDS: electrodiagnostic study, SD: Standard deviation. N=90.

Table 2.

Ultrasound parameters of CTS patients at baseline

Tendon excursion, cm (mean, SD)	1.2±0.4
SSCT excursion, cm (mean, SD)	0.24±0.15
Shear index, % (mean, SD)	78.9±12.8
Maximum velocity ratio, cm/s (mean, SD)	0.30±0.15

SD: Standard deviation.

Shear Index and Disease Severity at Baseline

Figure 2 shows the shear indices relative to CTS severity based on either EDS results or patient-reported assessment. With only small sample sizes in the symptom score-based ‘mild’ and ‘very severe’ categories, these groups were merged with ‘moderate’ and ‘severe’ respectively, since these were found to be clinically most alike. A single patient who had a ‘very severe’ EDS was merged with the ‘severe’ group, but the two cases with ‘normal’ EDS were excluded from subsequent analyses. Analyses of variance showed a significant statistical difference in average shear index between categories based upon EDS severity (p=0.0475). Post-hoc mean comparisons showed that the mean shear index in the moderate EDS severity group was 8.6% higher than that of the milder CTS cases (p=0.0389). However, no association of shear index with the initial symptom severity score from the BCTQ (p=0.609) was found. The shear index averages, the 95% confidence intervals per group, and the results of the pairwise comparisons can be found in the Supplements (S-2). For the MVR, overall analysis of variance showed a similar pattern as the shear index, with the EDS based disease severity having a greater association with MVR (p=0.072) compared to the symptoms score comparison (p=0.207). None of the pairwise comparisons with the maximum velocity ratio showed statistical significance.

Fig. 2:

Boxplots showing individual samples of shear index values at baseline. They are categorized by CTS severity defined by electrodiagnostic studies (left) or patient-reported symptom severity score (right). Dashed lines represent mean value of the control group. Overall ANOVA results showed difference in means for EDS based severity (p=0.0475), but not for symptom based severity (p=0.609). *p=0.0389 based on pairwise means comparison corrected for multiple comparisons.

Shear Index and Association with Surgical Outcome

BCTQ scores improved for all patients after surgery for both the SSS (p<0.001) and the FSS (p<0.001); The SSS improved from 3.1 (SD: 0.7) to 1.6 (SD: 0.5) at 1 month and 1.5 (SD: 0.5) at 3 months post-surgery. The FSS improved from 2.5 (SD: 0.8) to 1.7 (SD: 0.5) and 1.5 (SD: 0.6) for the same time points respectively (S-3). Adjusted linear regression models showed no association between either of the US parameters with symptomatic or functional improvement at one or three months follow-up (Table 3). To also test extreme values of the shear index for prognostic potential, the shear index was split into three groups (high, middle, low) and used as an independent variable. The surgical outcome was dichotomized into “success” and “not success” groups using the minimal clinically important difference. Out of those who completed the follow-up time points, 39 patients (61%) were categorized as successful at one month and 52 patients (66%) at three months. Again, no statistically significant associations were found between shear index and surgical outcome at 1 month (p=0.42) or 3 months (p=0.60). More detailed chart reviews were done of the five patients with the highest and lowest shear indices but they did not stand out from the others in terms of symptom severity score changes, intraoperative findings, or nerve size and dynamics.

Table 3.

Linear regression model for baseline SI versus clinical outcome

	Δ Symptom Severity Score				Δ Functional Status Score
	1 month		3 month		1 month		3 month
	Estimate (β)	p-value*	Estimate (β)	p-valueǂ	Estimate (β)	p-valueǂ	Estimate (β)	p-value#
Shear index, %	0.003	0.49	−0.004	0.27	0.007	0.18	−0.004	0.25
Maximum Velocity Ratio	−0.02	0.58	0.51	0.10	−0.15	0.74	0.67	0.06

Modeled with linear regression, adjusted for

^*baseline BCTQ and age

^ǂbaseline BCTQ

^#baseline BCTQ and CES-D score

Discussion

This study looked at the added value of longitudinal dynamic ultrasound in the assessment of disease severity and the association with post-surgical outcome. We found a significant decrease in SSCT motion (i.e. increase in shear index) between mild and moderate CTS patients, suggesting that the SSCT became less adherent to the tendon as severity increased. However, there was no association between pre-operative shear index and self-reported clinical symptom (improvement) after carpal tunnel release surgery.

Current literature suggests that diminished nerve excursion is present in patients with CTS(Ellis, et al. 2017) during finger and wrist motion in both the transverse (Filius, et al. 2015, Nanno, et al. 2017, Nanno, et al. 2015, van Doesburg, et al. 2012, Wang, et al. 2014, Wang, et al. 2014) and in the longitudinal plane (Hough, et al. 2007, Korstanje, et al. 2012, Liong, et al. 2014, Wang, et al. 2014). In addition, mechanical studies in cadaver tissue have shown that the SSCT is prone to damage even with normal finger tendon excursion (for example, if adjacent fingers are moving in opposite directions) (Filius, et al. 2014, Kociolek, et al. 2015, Vanhees, et al. 2012) and that progressive fibrosis of the SSCT is present in synovial biopsies from CTS patients (Ettema, et al. 2006). Combining these observations, it has been suggested that the SSCT could play a central role in the initiation and progression of CTS (Festen-Schrier and Amadio 2018, Werthel, et al. 2014). A previously published pilot study of intra-operatively measured relative SSCT movement in CTS patients reported a bimodal distribution where the SSCT excursion was either very closely related to the tendon motion or close to completely dissociated (Ettema, et al. 2008). This led to a hypothetic model that pathological SSCT changes could be divided into two categories: (1) With increasing fibrosis, the SSCT loses the mechanical response characteristic of a viscoelastic structure and adheres to the tendon during motion and (2) with progressive damage and continued repetitive, high force tendon motion, eventually the fibrotic connections between the closest layers of SSCT and the tendon are disrupted, causing a physical dissociation from the tendon and subsequent absence of SSCT motion (Fig. 3). Both descriptions have been observed in CTS patient samples (Ettema, et al. 2004, Ettema, et al. 2008) but have never been correlated to US findings. The first category of patients would be associated with a lower shear index while the second description would produce a higher shear index value. In our dataset, we did find some patients who showed low shear indices, but most of these values were within the range of the normal controls. The dominant trend was a higher shear index with increasing severity, which fits the concept of SSCT dissociation from the tendon. Currently, the US resolution is not adequate enough to characterize the SSCT in more detail, but in the future, higher-resolution imaging options will become available to enhance image processing.

Fig. 3:

Hypothetic model with two stages during CTS. A) During anatomical rest position, the absence of tendon loading leaves the SSCT in a relaxed state. B) During finger flexion and extension, longitudinal motion of the tendon causes the recruitment of SSCT layers. In CTS, fibrosis of the SSCT could alter this motion to a more closely related movement. C) With time, additional damage to the SSCT disrupts the architecture and causes dissociation between the movement of the tendon and the surrounding structures.

TCL: Transverse carpal ligament, SSCT: Subsynovial connective tissue, FDS: Flexor digitorum superficialis.

The first aim of this study was to test the SSCT as a potential proxy for disease severity, as has been done before for the cross-sectional area of the nerve. Compared to the median nerve, however, non-invasively visualizing and quantifying the SSCT behavior is complicated due to its thinness, viscoelastic properties, and the potential effects of variability in the joint range of motion. So far, multiple studies have assessed SSCT motion longitudinally using either Doppler (Oh, et al. 2007, Tat, et al. 2013, Tat, et al. 2015) or a post-processing speckle-tracking based algorithm either using commercially available software (Van Doesburg, et al. 2012, Yoshii, et al. 2009) or custom-designed (Filius, et al. 2015, Korstanje, et al. 2012). Doppler is easy to use but loses its validity when the moving structures start to approach an angle perpendicular to the probe’s surface. Speckle tracking algorithms are not limited by this since it is a post-acquisition processing technique that follows acoustic signals from frame-to-frame with a pattern matching principle.

Four of the previously-mentioned studies included CTS patients (Filius, et al. 2015, Korstanje, et al. 2012, Tat, et al. 2015, Van Doesburg, et al. 2012). From these, Tat et al. looked at eleven patients with self-reported CTS symptoms and compared them to an equally sized healthy group. Sample sizes were too small to directly associate the BCTQ scores to the shear index, but they reported a statistically significant higher shear index (30.2% versus 21.7%, p=0.016) in the CTS patients (Tat, et al. 2015). This was in agreement with data from van Doesburg et al., who assessed 18 CTS patients and compared the shear index to 22 healthy volunteers. Using commercial speckle tracking software, they found a shear index of 48% in the CTS patients, which also was higher than that of the controls (36%, p<0.05) (Van Doesburg, et al. 2012). Again this study was not aimed at testing a direct association with disease severity and the authors commented that their method required further optimization to increase sensitivity for SSCT tracking. The other two studies were larger in their sample size and aimed at testing disease severity with SSCT measures, utilizing a similar speckle tracking algorithm as was used in this study. Korstanje et al. included 55 CTS patients and compared the most affected with the least affected hand based on clinical and NCS classification. An increase in absolute SSCT excursion was reported in the most affected hand (p=0.025), but this was also the case for the FDS tendon (p=0.008), thus it remains unclear whether the relative shear index was different. Finally, a larger clinical study where US images were analyzed with speckle tracking showed that the majority of patients had higher shear indices than the controls (Filius, et al. 2015). Interestingly, their shear index values ranged between 41–61% whereas the majority of our values ranged between 72–88%. This difference was mostly caused by the higher SSCT displacements Filius et al. registered. Possible explanations for this discrepancy could be the noise filter that was added to our speckle tracking algorithm (Bandaru, et al. 2018), which prevents overestimation of motion detection by eliminating background noise. Another difference lies with the imaging protocol since speckle tracking is dependent on velocity and total excursion amount. Overall, these studies confirm our finding that CTS is associated with increased shear index and thus decreased SSCT motion, albeit with outliers who do not necessarily correlate with different initial severity or subsequent outcome.

The major advantage of this study is the larger clinical population used to specifically address the clinical usefulness of relative SSCT displacement. We aimed to combine pathophysiological findings with innovative non-invasive evaluation tools in order to find a patient-centered non-invasive tool that might correlate a biomechanical property with clinical parameters, including the ability to predict outcomes. Both symptomatic and functional evaluation measures were used as reference points for disease severity at baseline. Limitations of our approach include those intrinsic to the speckle tracking method and the small size of the SSCT, which resulted in a relevant loss of ultrasound data: out of plane motion cannot adequately be corrected and the ROI is a set framework through which the structures move, rather than identifying the speckles and dynamically tracing them. We found our method to have higher reliability in terms of intra- and inter-rater repeatability (Schrier, et al. 2019), a feature which is highly desirable when looking at potential diagnostic and prognostic applications, but the trade-off was that speckles could move outside the ROI, resulting in a less accurate measurement. In addition, speckle tracking is based on the assumption that the SSCT can be viewed as a solid structure, which allows averaging the different read-out layers during speckle tracking. Nonetheless, given the current resolution of the images, this simplification likely had a limited effect. Speckle tracking at this level is currently not incorporated in any commercial ultrasound machines so there remains an element of post-processing making it less suited for direct clinical implementation.

Finally, the dataset was limited to CTS patients undergoing surgical intervention. This could have biased the baseline to be more severe than is found in the average CTS population, for whom conservative treatment plays a prominent role. Carpal tunnel release surgery has been known to be highly effective, although reports on wide variation in success have also been published. During study design, we anticipated a symptomatic improvement almost twice as small as what we actually found. In addition, we observed less variation than expected, so only a relatively small proportion of our patients showed less than optimal results. This complicated our ability to correlate the shear index with outcome. We do plan to assess the shear index in a sample of non-operatively treated CTS patients and in those with other forms of CTS (for example, associated with diabetes, where the fibrosis may be more profound and clinical outcomes less predictable) to further test the possible value of the shear index in identifying the best treatment for the right patient.

Conclusion

The SSCT moves less in patients with CTS, resulting in a higher shear index, which significantly increases with increasing disease severity. This fits the underlying pathophysiological model of SSCT disruption and disconnection from the tendon. The relative movement of SSCT can be measured, but we were unable to show a clear association with patient-reported outcome after surgical intervention, possibly because our patients did exceptionally well after surgery, and also due to the fact that SSCT thickness is close to the limit of resolution of current US technology. Given these findings, dynamic ultrasound of the SSCT seems to have limited prognostic value in patients undergoing surgery for idiopathic carpal tunnel syndrome. We plan to investigate this in an injection cohort to see if this measure performs differently in patients who are treated non-operatively, where treatment failures are substantially more common.

Summary

The connective tissue in the carpal tunnel is less responsive to tendon motion in CTS patients with increased electrophysiological severity. This can be assessed with dynamic ultrasound, but inter-patient values present in a wide variation and seem not to be associated with short-term patient-reported outcome in a surgical cohort.