The Prevalence of Scapholunate Signal Abnormalities on Magnetic Resonance Imaging

Abstract

Background:

Although trauma is a major cause of symptomatic scapholunate interosseous ligament (SLIL) pathology, many patients do not recall a specific injury or repetitive trauma. We report on: (1) the prevalence of SLIL signal changes in patients who underwent wrist magnetic resonance imaging (MRI) for various indications; and (2) the prevalence of SLIL signal changes on MRI in patients without prior wrist trauma.

Methods:

This is a retrospective study evaluating 1021 patients who underwent wrist MRI or magnetic resonance arthrogram. We collected data on SLIL signal changes on MRI. Patients were divided into 6 groups, based on age, to calculate the proportions of SLIL signal changes across different age groups.

Results:

There were a total of 317 (31%) patients with SLIL signal changes, of whom 264 (26% from 1021 and 83% from 317) had a documented low clinical suspicion of SLIL pathology. The prevalence was 15% among 18 to 30 years old and increased to 50% in those above 70 years old. Of the 317 patients with SLIL signal changes, 161 (51%) had no documented prior wrist trauma. The prevalence of SLIL signal changes in this group was 28% in 18 to 30 years old and increased to 80% in patients older than 70 years.

Conclusions:

Magnetic resonance imaging signal changes demonstrating SLIL pathology in patients of a younger age may be more clinically meaningful as there is a lower prevalence of incidental SLIL pathology in these patients. Furthermore, it should be kept in mind that SLIL changes on MRI in older patients are common and may not represent acute pathology.

Introduction

Scapholunate interosseous ligament (SLIL) dissociation is the most common type of traumatic wrist instability.^1-5 Magnetic resonance imaging (MRI) can detect signal changes in the SLIL associated with SLIL pathologies such as degenerative changes, sprains, partial tears, or complete tears.^6-8

Although trauma is 1 of the major causes of symptomatic SLIL pathology,^2-4,3,9-11 many patients do not recall a specific antecedent event or injury, nor any repetitive trauma that may have directly led to injury. ¹ However, changes in the SLIL may result from degeneration over time or may be associated with a multitude of conditions such as osteoarthritis, infection, malignancy, and inflammatory arthropathies.^12-14 As such, it is suspected that some of the SLIL changes seen on MRI may not result from a specific trauma.

The objective of this study was to report the prevalence of SLIL signal changes in patients who underwent wrist MRI for various indications, including traumatic and nontraumatic causes. Furthermore, we aimed to report the prevalence of SLIL signal changes on MRI in patients who do not report prior wrist trauma. Finally, we investigated the factors associated with SLIL signal changes in these 2 groups.

Material and Methods

Study Design

The study has been approved by the institutional review board (Mass General Brigham IRB) under protocol number 2019P000635 and uses a database spanning 5 academic hospitals. We requested all MRI reports of the wrist conducted between January 2018 and December 2020 through our Research Patient Data Registry.

Data Collection

Data were collected by 2 independent researchers. Any discrepancies in data collection were resolved by secondary review and consensus by the collectors. We included all wrist MRIs, of which 9% (n = 92) were magnetic resonance (MR) arthrograms. The MRI scanners used in this study were 3T Siemens MAGNETOM Vida scanner, 3T General Electric Premier scanner, and 1.5T Siemens MAGNETOM Avanto scanner. Dedicated hand/wrist coils were used in all scanners. All wrist MRIs were evaluated for SLIL signal changes by different musculoskeletal radiologists outside of the study team, including evaluation of volar, membranous, and dorsal components of the SLIL. The radiologists systematically reviewed the MRI/MR arthrogram, providing a comprehensive description of the SLIL ligament in the report. Intact SLIL in the MRI report was considered as no signal change. All other SLIL findings were denoted as signal change. As SLIL signal changes were described with heterogeneity, it was not possible to differentiate between these options. Therefore, we chose to collect “SLIL signal changes” as a binary variable—being present or not present. If the patient previously had multiple wrist MRIs performed, we limited our analysis to the oldest report available. The second oldest report was used when the first MRI was aborted for technical reasons. We excluded: (1) patients below 18 years of age; (2) 22 patients due to missing data on clinical suspicion of wrist or hand pathology in both the clinical notes and MRI orders; (3) patients who had previously undergone SLIL, lunate, or scaphoid surgery; and (4) 3 patients who had a malignant neoplasm located near the SLIL at the time the MRI was performed. The radiologist readers explicitly noted that the SLIL could not be assessed due to the presence of a tumor near the ligament. Pathology confirmed a malignant diagnosis in all 3 cases. We also excluded (5) patients whose MRIs showed deficiencies such as incomplete sequencing, poor quality, or the presence of artifacts as described in the radiologist’s report.

Demographic Characteristics of the Study Cohort

Demographic characteristics, including sex, race, age at the time of the MRI, and prior wrist trauma sustained by the wrist area due to external force or impact, were retrospectively extracted from the medical record. Indications for MRIs were collected from physician orders (orthopedic surgeons, plastic surgeons, hand surgeons, (orthopedic) trauma surgeons, general practitioners, and internists) and relevant clinic notes that were directly related to the MRI request (Table 1). If suspicion for SLIL pathology was described in the clinical note or MRI order, the patient was classified as having “high clinical suspicion” for SLIL pathology. In cases where SLIL pathology was not listed in the physician’s differential diagnosis, the patient was classified as having “low clinical suspicion.” The indications for low clinical suspicion are listed in Table 1. We classified indications as low clinical suspicion when there was a specific non-SLIL diagnosis as the indication (eg, tumor or nerve pathology). Generalized wrist pain was also considered a nonspecific diagnosis because the MRI was not specifically ordered with the suspicion that an SL injury was present. Missing data are reported in Table 1.

Table 1.

Demographics.

Variable	n (%)
Age, y, mean (±SD)	47 (16)
Male	476 (47)
Race a
Caucasian	778 (79)
Black	98 (10)
Asian	50 (5)
Others	62 (6)
Prior wrist injury b	468 (46)
Indications
High clinical suspicion for SLIL pathology	109 (11)
Low clinical suspicion for SLIL pathology	912 (89)
General hand/wrist pain	121 (13)
Soft tissue/neoplasm	153 (17)
Bone pathology/fracture	196 (22)
Non SLIL ligament/tendon/muscle pathology	299 (33)
Infection/inflammation	113 (12)
Nerve pathology/impingement	30 (3)

Note. y = years; SD = standard deviation; SLIL = scapholunate interosseous ligament.

^aMissing = 33.

^bMissing = 3.

Statistical Analysis

The normality of the data was assessed with a histogram and quantile-quantile plot. As demographic descriptions were normally distributed, the data are presented with mean ± standard deviation (SD). Categorical data are presented with frequencies and percentages.

Patients were subdivided into 6 groups based on age: (1) 18 to 30 years (mean, 24 years; SD, 3 years); (2) 31 to 40 years (mean, 35 years; SD, 3 years); (3) 41 to 50 years (mean, 45 years; SD, 3 years); (4) 51 to 60 years (mean, 56 years; SD, 3 years); (5) 61 to 70 years (mean, 65 years; SD, 3 years); and (6) >71 years (mean, 77 years; SD, 4 years). Subsequently, we calculated the proportion of MRIs displaying signal changes in SLIL in each subgroup categorized by high and low clinical suspicion. Furthermore, we calculated the proportion of MRIs showing SLIL signal changes in the nonprior wrist trauma group within each of the 6 age groups. We used the Locally Weighted Scatterplot Smoothing (LOWESS) ¹⁵ method to illustrate the frequency of SLIL signal changes in the overall population, high clinical suspicion, and low clinical suspicion groups, as it varies with age. A multivariate logistic regression model was used to analyze the association between the clinical indicators and SLIL signal changes in wrist MRIs in: (1) all the patients; and (2) in the nontraumatic wrist patients. A P value of <.05 was considered statistically significant. We used StataCorp. 2019 (Stata Statistical Software, Release 16; StataCorp LLC, College Station, Texas).

Results

This study included 1021 patients with a mean age of 47 ± 16 years, of which 47% (n = 476) were men and 79% were Caucasian (n = 778). In the included cohort, 46% (n = 468) had a prior wrist trauma. One hundred nine (11%) MRI scans were conducted for suspected SLIL pathology: high clinical suspicion. The most common reason for MRI requests with low clinical suspicion was the presence of ligamentous, tendinous, or muscular pathology (33%, n = 299), excluding SLIL pathology. The demographic data are presented in Table 1.

Among 1021 patient MRIs, 317 (31%) had SLIL signal changes. Of all the patients, 26% (264 of 1021) had a low clinical suspicion of SLIL pathology and 5% (53 of 1021) had a high clinical suspicion for SLIL pathology. In the group with SLIL signal changes, 83% (264 of 317) of the patients had a low clinical suspicion of SLIL pathology and 17% (53 of 317) of the patients had a high clinical suspicion for SLIL pathology. Exploring the MRIs showing SLIL signal changes (n = 317), we found that in the low clinical suspicion group, the prevalence of SLIL signal changes was 15% (32 of 218) among patients aged 18 to 30 years and increased to 50% (41 of 82) in patients older than 70 years (Table 2 and Figure 1). In the overall group, we found that 47% (53 of 109) of the wrists indicated as a high clinical suspicion had SLIL signal changes on MRI, compared with 29% (264 of 912) of the wrists that were indicated as low clinical suspicion (P < .001 by Fisher exact test).

Table 2.

Prevalence of SLIL Signal Changes in 1021 Patients Who Underwent MRI of the Wrist.

Variable	Age group, y
	18-30	31-40	41-50	51-60	61-70	>70	Total
Wrist MRI, no.	218	207	164	212	138	82	1021
Proportion of MRIs showing SLIL signal changes, overall, % (no.)	20 (43)	23 (47)	30 (50)	39 (82)	39 (54)	50 (41)	31 (317)
High clinical suspicion	5 (11)	6 (12)	9 (14)	6 (12)	3 (4)	0 (0)	5 (53)
Low clinical suspicion	15 (32)	17 (35)	22 (36)	33 (70)	36 (50)	50 (41)	26 (264)

Note. y = years; MRI = magnetic resonance imaging; SLIL = scapholunate interosseous ligament.

Figure 1.

Signal changes and no signal changes in the scapholunate interosseous ligament on all wrist MRI scans (n = 1021). MRI = magnetic resonance imaging.

Of all the MRIs with SLIL signal changes, 51% (161 of 317) had no prior wrist trauma. In this group, the prevalence of SLIL signal changes was 28% (12 of 43) in 18 to 30 years old and increased to 80% (33 of 41) in patients older than 70 years (Figure 2).

Figure 2.

Prior and nonprior wrist trauma in comparison with MRI scans that show SLIL signal changes (n = 317). SLIL = scapholunate interosseous ligament; MRI = magnetic resonance imaging.

Multivariable logistic regression analyses showed that increasing age (odds ratio [OR] = 1.03 per year increase per age; 95% confidence interval [CI] = 1.03-1.05; P < .001), male sex (OR = 1.46; 95% CI = 1.10-1.95; P = .009), and a high clinical suspicion for SLIL signal changes (OR = 3.59; 95% CI = 1.38-9.38; P = .009) were independent predictors for SLIL signal changes on MRI (Table 3). The LOWESS curve demonstrated a steadily increasing trend in SLIL signal changes on MRI correlated with age (Figure 3).

Table 3.

Multivariable Logistic Regression Analysis of Factors Associated With SLIL Signal Changes on MRI (n = 986).

Variable	OR	Lower (95% CI)	Upper (95% CI)	SE	P value
Age	1.03	1.03	1.05	0.00	<.001
Men	1.46	1.10	1.95	0.21	.009
Race (reference: others)
Caucasian	0.67	0.38	1.19	0.20	.17
Black	0.91	0.45	1.83	0.32	.79
Asian	0.84	0.36	1,96	0.36	.69
Prior wrist injury	0.98	0.83	1.15	0.08	.79
Indications
High clinical suspicion for SLIL pathology	3.59	1.38	9.38	1.76	.009
Low clinical suspicion for SLIL pathology (reference: nerve pathology/impingement)
General hand/wrist pain	1.65	0.65	4.18	0.79	.29
Soft tissue/neoplasm	0.91	0.36	2.3	0.43	.84
Bone pathology/fracture	1.39	0.55	3.47	0.65	.48
Non SLIL ligament/tendon/muscle pathology	1.08	0.44	2.64	0.49	.87
Infection/inflammation	1.07	0.42	2.75	0.52	.88

Note. Boldfaced values indicate statistical significance (P < .05). OR = odds ratio; CI = confidence interval; SE = standard error; MRI = magnetic resonance imaging; SLIL = scapholunate interosseous ligament.

Figure 3.

A Locally Weighted Scatterplot Smoothing (LOWESS) graph depicts the gradual changes of overall, high, and low clinical suspicion scapholunate interosseous ligament signal changes on magnetic resonance imaging as it varies with age.

In patients without prior wrist trauma, multivariable logistic regression showed that a higher age (OR = 1.04 per year increase in age; 95% CI = 1.03-1.06; P < .001) and a high clinical suspicion for SLIL signal changes (OR = 6.09; 95% CI = 1.59-23.28; P = .008) were independent predictors for SLIL signal changes on MRI (Table 4).

Table 4.

Multivariable Logistic Regression Analysis of Factors Associated With SLIL Signal Changes on MRI in Nontrauma Patients (n = 536).

Variable	OR	Lower (95% CI)	Upper (95% CI)	SE	P value
Age	1.04	1.03	1.06	0.01	<.001
Men	1.49	1.00	2.22	0.30	.051
Race (reference: others)
Caucasian	0.84	0.35	1.98	0.37	.68
Black	1.09	0.38	3.15	0.59	.88
Asian	0.45	0.12	1.68	0.30	.24
Indications
High clinical suspicion for SLIL pathology	6.09	1.59	23.28	4.17	.008
Low clinical suspicion for SLIL pathology (reference: nerve pathology/impingement)
General hand/wrist pain	2.57	0.81	8.18	1.52	.11
Soft tissue/neoplasm	1.53	0.50	4.67	0.87	.46
Bone pathology/fracture	3.07	0.94	10.04	1.86	.06
Non SLIL ligament/tendon/muscle pathology	1.51	0.50	4.58	0.86	.46
Infection/inflammation	1.55	0.51	4.73	0.88	.44

Note. Boldfaced values indicate statistical significance (P < .05). OR = odds ratio; CI = confidence interval; SE = standard error; MRI = magnetic resonance imaging; SLIL = scapholunate interosseous ligament.

Discussion

A total of 317 patients (31%) of 1021 had SLIL signal changes, of whom 262 (26% from 1021 and 83% from 317) had low clinical suspicion of SLIL pathology prior to the MRI being obtained. In both the overall and the low clinical suspicion groups, the SLIL signal changes increased with age. In the full cohort, we observed that SLIL signal changes were associated with older age, male sex, and a documented high clinical suspicion. Furthermore, 161 of all patients with SLIL signal changes (51%) had no documented prior wrist injury. Also in this group, the SLIL signal changes increased with age. Predictors for SLIL signal changes in this group were higher age and high clinical suspicion.

There are several limitations to this study. First, our data do not differentiate between different stages of SLIL pathology (eg, degenerative change, sprains, partial tears, or complete tears). ⁶ Radiologists described SLIL signal changes in different ways. ⁶ In many cases, it was not possible to discriminate between different verbal descriptions. Therefore, we chose to collect “SLIL signal changes” as a binary variable—being present or not present. In terms of this study, a binary treatment is reasonable to establish a Bayesian prior for interpretation of MRI/MR arthrogram; however, these data are not intended to describe partial versus complete SLIL injury.

Second, the presence of SLIL signal changes was determined by wrist MRI, but we did not examine the MRI images ourselves; we instead relied on radiology reports, intending to decrease bias. ¹⁶ Furthermore, the images were reviewed by several musculoskeletal radiologists with varying levels of experience across 5 hospitals. The examination of SLIL signal changes may vary by radiologist and institution.

Third, there is no control for the type of examination or imaging parameters. There could be variation between MRI scanners, ¹⁷ imaging techniques ¹⁸ and the use of gadolinium for MR arthrograms.^19,20 Potter et al ²¹ and Anderson et al ²² described that MRIs with a higher resolution permitted more accurate depiction and localization of tears of the triangular fibrocartilage complex. Furthermore, previous studies with different MRI scanners, using arthroscopy as a reference, described that the sensitivity MRI of the wrist to identify SLIL tears has been found to range between 46% and 63%, with specificity ranging between 86% and 91%.^23,24 Regarding MR arthrograms, Kader et al described differences in sensitivity and specificity between MRIs and MR arthrograms for the detection of SLIL signal changes. The use of gadolinium in MRI showed a higher degree of sensitivity and specificity. ²³ Both MRIs and MR arthrograms were combined to reflect real-world clinical practice, where both imaging modalities are commonly used to evaluate SLIL pathology. Because the study is intended to establish a Bayesian prior for interpreting advanced imaging studies, and considering the fact that the performance characteristics of MRI and MR arthrogram vary widely across studies, we chose not to treat these as separate groups.

Fourth, physicians with different specialties or subspecialties (orthopedic surgeons, plastic surgeons, hand surgeons, (orthopedic) trauma surgeons, general practitioners, and internist) ordered the MRIs. The reason for patients’ clinic visits sometimes focused on other possible wrist or hand pathologies. Overall, these results should be understood within a Bayesian statistical framework. Incidental SLIL signal changes are common, and when identified on MRI in a low-probability setting, the SLIL finding rarely warrants treatment.

Fifth, we included a heterogeneous group of patients, which may have introduced variability in imaging types, diagnostic accuracy, and clinical coding. This heterogeneity could impact the internal validity of the findings. However, it also allowed for the inclusion of a larger and more representative cohort, thereby enhancing the generalizability of the results to broader clinical settings.

To our knowledge, there is a limited understanding of the prevalence of SLIL signal changes on radiological images. Based on radiographic gap views, Akahane et al ²⁵ described that the incidence of asymptomatic scapholunate dissociation was significantly higher in patients above 30 years than in those below 30 years. Furthermore, studies have shown age-related changes to the carpal ligaments. ²⁶ These results were consistent with those in our study, which used MRI scans to show that, across our cohort, SLIL signal changes steadily increased with age. This trend was present both for patients with and without prior injury to the wrist. This suggests likely age-related and degenerative processes ongoing in the ligament, which can be visualized on MRI.

The SLIL signal changes in patients with low clinical suspicion increased progressively with age. Among patients aged 18 to 30 years, 75% (15 of 20) showed SLIL signal changes, whereas in the >70 years age group, 100% (50 of 50) exhibited SLIL signal changes. Our overall data indicated that most patients with SLIL signal changes had low clinical suspicion (84%).

Overall, these data suggest that there is a relatively high prevalence of patients with SLIL signal changes on MRI or MR arthrogram who are asymptomatic. However, we found that less than half of the patients in the high suspicion group had an SLIL pathology detected on MRI. This could result from 2 possibilities: (1) patients who had an actual SLIL injury were not identified by their imaging study; or (2) our ability to determine SLIL injury from clinical history and examination is quite imprecise. Most studies suggest that there is a meaningful false-negative rate ²⁷ ; however, it is also well established that the mechanism of SLIL injury overlaps with other wrist pathologies such as fracture or other sprains, and physical examination findings are relatively nonspecific. Most likely, both of these factors account for this result.’

Consequently, the pretest suspicion of SLIL pathology is very important in determining whether an observed signal change on MRI is clinically meaningful. In other words, an MRI for the evaluation of SLIL signal changes in the wrist could be useful in patients of younger age with a discrete history of trauma for clinical decision-making. Similarly, in an older patient without a history of antecedent trauma, the presence of SLIL signal change may reflect expected degenerative change or other chronic pathology and may not warrant treatment.