Background
The long head of the biceps tendon (LHBT) is known to be a relatively frequent cause of anterior shoulder pain [
1‐
3]. Both conservative and surgical treatments can be performed to alleviate symptoms. Treatment method can be selected by taking factors such as comorbid shoulder joint disorders and symptom duration into account [
4]. Biceps tenotomy or tenodesis are currently the most popular surgical methods. Both methods are generally recognized for their advantages such as short surgery time, relatively simple techniques, and low costs [
5‐
7]. Speed’s test [
8] and Yergason’s test [
9] have been developed as physical examinations to detect LHBT pathologies. However, their sensitivities are 32 and 43%, respectively. Their diagnostic performances are considered unsatisfactory [
10]. In general, diagnosing LHBT pathology is difficult as it comes concomitantly with other shoulder pathologies such as labral lesions or rotator cuff tears far more frequently than it comes as a solitary lesion [
11‐
13].
Ultrasonography has been traditionally used as an imaging method to diagnose LHBT pathology. It is low in cost, noninvasive, and has an advantage of dynamic study. Its major disadvantages include low diagnostic ability for partial tears and inability to access intraarticular lesion [
14,
15]. Magnetic resonance imaging (MRI) is more objective than ultrasonography. It can evaluate intraarticular lesions and partial tears more precisely. Therefore, MRI is preferred as a diagnostic tool for various musculoskeletal pathologies including shoulder joint disorders [
11,
16‐
18]. However, only a small number of known studies have examined MRI’s efficacy for diagnosing LHBT pathology. In addition, they were conducted on small cohorts [
17,
19,
20].
In this study, we aimed to determine the diagnostic value of standard non-enhancing magnetic resonance imaging (MRI) for detecting LHBT pathology and identify the most useful diagnostic signs on MRI.
Discussion
LHBT lesions are rarely present as solitary lesions. They are often concomitant with rotator cuff tears or labral lesions. Skendzelet al. [
14] have reported that 96.2% of LHBT tear cases are comorbid with supraspinatus tendon tears based on MRA findings. Murthiet al. [
2] have performed histologic analysis of specimens obtained via tenosynovectomy of LHBT on 200 patients who underwent arthroscopic subacromial decompression for impingement syndrome and reported that normal biceps tendon cases account for only 18% of total cases while 76% of all cases have chronic inflammation or fibrosis. The incidence of biceps pathology is increased in proportion with the extent of rotator cuff tear. Accordingly, abnormalities in biceps tendons can be major causes of shoulder pain with other shoulder lesions. In the present study, MRI findings showed LHBT pathology in 124 (22.4%) cases, different from the incidence reported by in the previous study [
2]. Such difference in incidence of biceps pathology might be due to difference in histological approach or difference in the definition of normal biceps tendon.
It can be a great challenge to identify the pathology of shoulder pain. Because LHBT pathology is highly likely to be comorbid with other shoulder pathologies, physical examinations often show nonspecific results or LHBT is overlooked in deference to another major pathology. Therefore, it is important to detect the presence of LHBT lesion and be prepared before setting up a treatment plan [
2,
3]. Various noninvasive imaging methods have been employed to determine the underlying pathology of shoulder pain and help choose treatment method. Ultrasonography and MRI are broadly and actively applied in orthopedic surgery fields. Diagnostic arthroscopy is generally the last choice because it is an invasive procedure.
As a preoperative diagnostic tool for LHBT pathology, ultrasonography is rapid and cost-effective. It can be conducted in the outpatient department. Skendzel et al. [
14] have analyzed the diagnostic accuracy of preoperative sonography performed on 66 patients and revealed its high diagnostic accuracies for complete tear of the biceps tendon, with sensitivity of 88% and specificity of 98%. However, sonography is inadequate for diagnosing partial tears (sensitivity, 27%; specificity, 100%).
Studies on MRI-assisted diagnosis of LHBT pathology have emerged recently with inconsistent results [
11,
17,
19,
20,
24]. Zanetti et al. [
20] have reported that MRA has adequately high sensitivity for distinction. In their study, two independent observers compared images obtained from 42 patients with their arthroscopic findings to determine the degree of agreement between MRI and arthroscopic findings (observer 1: sensitivity of 92% and specificity of 56%; observer 2: sensitivity of 89% and specificity of 81%). However, the interobserver agreement was low (kappa value: 0.39).
In the present study, both observers who compared interpretations of standard non-enhancing MRI and arthroscopic findings showed high sensitivity and specificity. Interobserver agreements were substantial to almost perfect, with kappa values ranging from 0.68 to 0.81. It was found that diameter change on the axial plane and alteration of signal intensity on the parasagittal plane accurately differentiated LHBT pathologies at a high rate (accuracy ≥80%). Setting the diagnostic criterion as ‘at least two positive signs’ to detect LHBT lesion yielded the most balanced results for all parameters. This suggests that a comprehensive interpretation of various signs in different planes rather than basing diagnosis on one definitive sign can enhance diagnostic accuracy. This makes sense as the biceps tendon runs in the medial-to-lateral direction in the intraarticular region and superior-to-inferior direction when it passes into the bicipital groove.
Diagnostic accuracy was increased when lesion severity was increased. Type II lesion showed the lowest sensitivity. Thus, MRI was considered inadequate for diagnosing this type of lesion. On the other hand, more severe lesions such as type IV and type V lesion that extend into the bicipital groove were more accurately diagnosed. Because tenodesis or tenotomy should be often considered in this type of lesion, it is required to ensure accurate diagnosis prior to surgery. Our analysis showed a high probability for the presence of type III or type IV lesion with diagnostic criterion set as ‘three abnormal signs’. Interpreting MRIs based on these diagnostic criteria can be useful for surgeons to set up a preoperative plan and educate the patient.
The 3 T MR machine was first introduced in 1999. Its advantages over conventional machines with low magnetic field strengths include a higher signal-to-noise ratio, improved spatial resolution, and shorter acquisition time [
25]. Many previous studies that reported low diagnostic efficacy of MRI for detection of LHBT lesions [
11,
20,
24] used 1.0 T or 1.5 T machines which might make image interpretation difficult. The high diagnostic efficacy obtained in this study may be attributed to the use of 3 T machines. Strengths of this study were: 1) we confirmed that standard non-enhancing MRI had sufficient diagnostic efficacy to detect biceps pathology if adequate criteria were used; 2) we presented a novel classification method according to tear progression and demonstrated that MRI sign and plane were the most suitable for diagnosing each classification type; 3) this study was carried out on a much larger group of patients than previous studies. All these advantages are sufficient to prove the diagnostic value of MRI.
However, this study also had some limitations. First, there was only one type V lesion. Thus, statistically significant value could not be obtained. Because type V (complete biceps tendon tear) has already been identified with high diagnostic accuracy in previous studies by both MRI and ultrasonography, we excluded type V from our analysis so that it had little influence on overall results of our study. Second, we described MRI interpretations only as normal or abnormal without differentiating lesion types. As a result, the diagnostic power for lesion type differentiation could not be determined. Third, the lack of evaluation of “hidden lesion” of extra-articular LHBT in arthroscopy is a limitation of this investigation [
26]. Fourth, this study focused on the lesion of biceps tendon itself and thus, we did not include dislocations or subluxations from the bicipital groove as a lesion. Fifth, since this study was conducted in a population undergoing surgery for rotator cuff disease, the prevalence of LHBT pathology probably would be higher than in other populations which affects the positive predictive value and negative predictive value. Lastly, as this study was on the diagnostic accuracy of MRI in diagnosing intra-articular pathology of the long head of the biceps tendon, we did not assess the correlation with clinical signs of biceps pathology. Further studies are needed to address these limitations.
Conclusion
In conclusion, standard non-enhancing MRI using a 3 T MRI machine has highly reliable diagnostic value for preoperative detection of LHBT pathology. Diagnostic accuracy is increased as lesion severity is increased. Among changes in diameter, contour irregularity, and alteration of signal intensity, reading with two or more diagnostic signs showed the highest accuracy. The single diagnostic sign that was the most sensitive for accurate diagnosis was the alteration of signal intensity in the parasagittal view.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.