Comparison of shoulder Magnetic Resonance Imaging findings between patients with stage 4 chronic kidney disease and hemodialysis patients with healthy controls

1. Introduction

Chronic kidney disease (CKD) is defined as the persistent reduction of glomerular filtration rate (GFR). It can lead to detectable alterations in renal function, and the level of reduction in GFR can be classified in five stages [1]. In patients with end stage renal failure, an abnormal electrolyte status and metabolic condition may contribute to tissue damage, resulting in the destruction of the musculoskeletal system [2]. In addition to reduced renal function, clearance of β 2 microglobulin is decreased. β 2 microglobulin then accumulates in the plasma and is stored in tissues, leading to a gradual accumulation. The normal serum level of β 2 microglobulin is 1.5–3 mg/l, but this may increase more than 60 times due to disturbed renal catabolism in cases of renal failure [3]. In general, when GFR is lower than 15 ml/min, the level of β 2 microglobulin in the blood is approximately 20 mg/1 [4]. Following five to seven years of induction through dialysis concomitant with the accumulation of β 2 microglobulin, amyloidosis can be detected, and its incidence increases as the duration of dialysis increases [5]. Amyloidosis can be also detected in CKD patients who are naïve to dialysis [6].

Dialysis-related arthropathy (DRA) is one of the serious possible complications of long-term hemodialysis and the related accumulation of β 2 microglobulin [7]. However, histological accumulation of amyloid within the tissue occurs at a very early stage, before any clinical or radiological signs of disease appear [8]. The symptoms of DRA range from painless effusion on the affected joint to complete destruction of the joint [9]. Histological identification of β 2 microglobulin amyloid deposits in samples from tissue biopsies is still the gold standard for diagnosis. In clinical practice, non-invasive methods such as MRIs are preferred [10]. MRI is superior to any other imaging method in terms of showing joint contour, bone marrow lesions and the size of soft tissue mass [11].

Shoulder involvement is among the best-known forms of DRA. Shoulder pain may occur following short-term dialysis, and similar morphological changes can be observed in asymptomatic patients. Shoulder stiffness and restricted movement are associated with dialysis lasting longer than ten years [12]. In cases with DRA, a shoulder MRI reveals certain signs of DRA; these are joint effusion, thickening of the rotator cuff tendons, thickening of the biceps tendon sheath, effusion at the subdeltoid bursa, hyper- and hypo-echoic deposits at the rotator cuff and bursa, and a lack of homogeneity in the tissue [13].

Although shoulder involvement in dialysis patients has been examined in many studies [6, 7, 8, 9, 10, 11, 12, 13], no data are available comparing shoulder involvement in CKD patients who are not on dialysis. Our aim is to compare the above signs in the shoulder MRIs of dialysis patients, of dialysis-naïve patients with stage 4 CKD and of healthy controls. This study will also examine the correlation between β 2 microglobulin level and shoulder signs in an MRI.

The study was conducted between January 2015 and February 2016 at the Corum Training and Research Hospital, with the approval of the Local Ethics Committee. Fifty patients undergoing dialysis three times a week were included: 20 patients with shoulder pain (Group 1) and 30 asymptomatic patients (Group 2). The study also included 20 patients with stage 4 CKD who were not under dialysis (Group 3) and 30 healthy control subjects (Group 4). In the dialysis patients, the shoulder without a fistula was accepted as dominant, and patients were included in the study based on their dominant and non-dominant shoulder. The exclusion criterias were as follows: Patients with systemic inflammatory diseases such as diabetes mellitus, rheumatoid arthritis, and ankylosing spondylitis were excluded, as were patients with musculoskeletal system diseases which might cause shoulder pain and patients with a history of trauma. In addition, patients with lymphoproliferative disorders which might increase the β 2 microglobulin level and those with cirrhosis, primary amyloidosis, or infections such as tuberculosis and osteomyelitis were also excluded.

Each subject’s age (year), height (cm), weight (kg), the time since dialysis started (months), and pain in the side and fistula zone were recorded.

3. Statistical analysis

Continuous quantitative data were expressed as n, mean and standard deviation, and qualitative data were expressed as n, median 25 th percentile and 75 th percentile. Continuous data from independent measurements and showing normal distribution were analyzed using a one-way ANOVA test (Tukey’s method was used for multiple comparison of this test) and a t-test; data with abnormal distribution or including score variables were analyzed using a Kruskal-Wallis test (Dunn’s method was used for multiple comparison of this test) according to their number. A Spearman correlation test was used for variables with abnormal distribution, as determined by the normality test, to demonstrate the correlations among variables. A chi-squared test was used for the categorical data set. Probability of 0.05 was accepted as significant. All data analysis was done with IBM SPSS Statistics 21.0 software programs.

4. Results

Table 1 shows the mean age, body mass index (BMI) values and β 2 microglobulin levels of the groups. The mean age did not differ significantly ( $p>$ 0.05) across the groups. The BMI value was significantly different between Groups 1 and 4, and the BMI of the control group was significantly higher than that of dialysis patients with painful shoulders. The duration of dialysis duration for Group 1 was 116.85 ± 48.84 months and 94.40 ± 50.02 months for Group 2; these differences are not statistically significant ( $p>$ 0.05). The duration of the disease for Group 3 was 71.40 ± 48.07 months. β 2 microglobulin levels did not differ significantly between Groups 1 and 2, but they did differ significantly among all other groups. The β 2 microglobulin level was significantly higher in dialysis patients than in non-dialysis patients or the control group; it was also significantly higher in non-dialysis patients than in the control group ( 0.001) (Table 1). Gender distribution and the ratio of dominant to non-dominant shoulder did not differ among the groups; 50% of the subjects were female and 50% of the affected shoulders were dominant in all the groups.

The supraspinatus tendon was significantly thicker in Group 1 than in any of the other three groups ( 0.001). The infraspinatus tendon was also significantly thicker in Group 1 than in Group 2 or Group 4 ( 0.05). Biceps tendon thickness differed significantly only between Groups 1 and 4 ( 0.05). Although all tendons were thicker in Group 3 than in Group 4, this difference was too small to be significant ( $p>$ 0.05) (Table 2).

Table 3 compares the presence of effusion in the groups. In general, effusion was significantly higher in Group 1 than in all the other groups ( 0.05). Group 1 also had significantly more effusion areas than the other groups, and Groups 2 and 3 had more than the control group ( 0.05) (Table 3).

There was no difference among the groups in the assessment of the supraspinatus muscle in terms of partial rupture/tendinosis/non-homogeneity, the presence of osseous lesions, or the acromioclavicular joint ( $p>$ 005).

The correlations among age, β 2 microglobulin level, GFR, total number of effusion areas and tendon thickness were also examined. There were significant correlations between age and the total number of effusion areas ( 0.05), GFR and supraspinatus tendon thickness ( 0.05) and infraspinatus tendon thickness ( 0.01). There were also significant correlations between β 2 microglobulin levels and the total number of effusion areas ( 0.01), infraspinatus tendon thickness ( 0.01) and subscapularis tendon thickness ( 0.05) (Table 4).

In our study we compared signs in the shoulder MRIs of CKD patients with and without dialysis and of healthy control groups, although effusion was more common in the dialysis group with pain, it was also significantly higher in the painless dialysis and CKD groups than in the control group. In the dialysis group with pain, the supraspinatus, infraspinatus and biceps tendons were thickened, while no difference was detected in tendon thickness of the painless dialysis and CKD groups compared to the control group.

In dialysis patients, shoulder pain is correlated with effusion and tendon thickness [14]. Kamel et al. divided 32 hemodialysis patients into two groups: patients with clinical and radiological DRA symptoms ( n = 12) and patients without symptoms ( n = 20); these groups were compared with 15 healthy control subjects. In the groups with and without DRA, the subdeltoid bursa effusion and subacromial bursa thickness were 91.7%–80% and 83.3%–75%, respectively. Subacromial bursa thickness was only 40% in the healthy controls [8]. Similarly, Slavotinek et al. studied 22 hemodialysis patients, 15 of which had shoulder pain, and reported subacromial-subdeltoid effusions in three patients (13.6%) and subacromial-subdeltoid thickness in six patients (27.2%) [7]. In another MRI study, Escobedo et al. examined five symptomatic dialysis patients (12 years), four asymptomatic dialysis patients (5 years) and nine non-dialysis healthy subjects, assessing eight shoulder MRIs from five symptomatic patients. Abnormal joint and bursa liquid accumulation were observed in all shoulders. The bursal collections frequently showed locution, thickness and irregular margins, corresponding to synovitis. Effusion was detected in five subchorocoid bursa, in four biceps tendon sheaths and in eight shoulders from four asymptomatic dialysis patients. Joint effusion was also detected in two of the MRIs, subacromial subdeltoid bursal effusion in two and effusion in biceps tendon sheath [15]. In our study, subdeltoid-subacromial effusion was 70%, 40%, 35% and 13.3% in the painful dialysis group, painless dialysis group, CRF group and control group, respectively. We assessed effusion in five separate areas and ranked four the four groups – painful dialysis, painless dialysis, CRF and control group. However, subchorocoid effusion was higher in the control group than in the CRF group. The total number of effusion areas was significantly greater in the dialysis and CKD groups than in the healthy controls. This sign demonstrated that changes in the shoulder joint had initiated in the non-dialysis and asymptomatic CKD patients as well as those with pain.

Similar results on effusion have also been found in ultrasonography studies. Sommer et al. used ultrasonography to assess 14 shoulders in dialysis patients who had been on dialysis for a minimum of ten years. Biceps tendinitis, subscapularis tendinitis and subdeltoid bursitis were detected in 50%, 27.5% and 35.7% of the patients, respectively [16]. Barısıc et al. conducted a shoulder USG trial with 54 hemodialysis patients (27 symptomatic, 27 asymptomatic) and 50 healthy controls. The biceps tendon effusion rate in their subjects was 42%, 17% and 5% in symptomatic, asymptomatic and healthy subjects, respectively; subdeltoid effusion was present in only 31% of the symptomatic group [12].

Thickening of the supraspinatus and biceps tendons correlates with DRA [8]. In our previous study, we compared 23 painful dialysis patients and 20 painless dialysis patients and detected significant thickening of the supraspinatus tendon in the group with pain [14]. Slavotinek et al. detected thickening in the supraspinatus tendon in nine patients (45%), eight of which reported pain. The same study found thickening in the supraspinatus tendon in ten patients (45%), including those with and without pain. The duration of dialysis was longer for patients with tendon thickening than other subjects (mean 18.5 years) [7]. In their study, Kamel et al. reported 75%–80% and 75%–80% thickening of the supraspinatus and biceps tendons in group with and without DRA, respectively. Thickening of the supraspinatus tendon was not present in healthy controls, while thickening of the biceps tendon was present in 33.3% of the control group [8]. Escobedo detected significantly more supraspinatus thickness in all symptomatic dialysis patients than in the asymptomatic dialysis patients or the healthy controls [15]. Barısıc et al. reported a significantly higher rate of thickening in the supraspinatus and biceps tendons in symptomatic dialysis patients than in asymptomatic patients or healthy controls, and a significant increase was also present in asymptomatic patients compared to healthy controls [12]. In our study, the thickening of the supraspinatus, infraspinatus and biceps tendons was significantly higher in painful dialysis patients, and all tendon thickening was higher in the CKD than in the control group, but not significantly.

Certain studies detected a correlation between β 2 microglobulin and the duration of dialysis [8, 17], while others couldn’t show any correlation [14, 18]. Although no study exists on shoulder joint involvement and amyloidosis in CKD without dialysis, asymptomatic β 2 deposits can be observed before dialysis treatment or within one year of the initiation of dialysis [8]. Zingraff et al. showed β 2 microglobulin on sternoclavicular joint of a female patient aged 69 years with chronic renal insufficiency and reported that β 2 microglobulin amyloidosis can occur in uremic patients before dialysis treatment, regardless of the association with dialysis [6]. Another report detected β 2 microglobulin in clavicular cysts in a patient aged 55 years with chronic glomerulonephritis [4]. In addition, Nakazawa et al. [19] examined surgical specimens from subacromial bursa and biceps tendon sheaths in patients with DRA and found an accumulation of β 2 microglobulin amyloid. Therefore, peri-articular fluid collection can be associated with this disease.

In our study, we wanted to determine whether increased β 2 microglobulin levels in patients with chronic kidney disease not yet on dialysis led to any changes in the shoulder joint. In the literature, several results are associated with β 2 microglobulin. In a study by Kamel et al., dialysis patients with and without DRA were compared to a healthy group, and β 2 levels were similar in the dialysis groups. In the healthy control group, the β 2 microglobulin level was 1.9 ± 0.3 mg/L, while this level was 84.9 ± 9.1 mg/L in the group with DRA and 75.1 ± 14.8 mg/L in the group without DRA [8]. There are also studies indicating that the absence of any correlation between DRA initiation and β 2 levels, which may suggest that DRA can initiate due to β 2 microglobulin accumulation following long-term dialysis [20, 21]. Although the serum β 2 microglobulin level is similar in hemodialysis patients with and without amyloidosis [22], the β 2 microglobulin level was higher in patients with symptoms of DRA than in asymptomatic patients [9, 13, 23]. Again, there are studies indicating that β 2 microglobulin level also correlates with tendon thickness [8, 14]. Our study found that both pain and β 2 microglobulin levels, in addition to GFR, correlate with thickness of the rotator cuff tendon.

Due to the sectional design of our study, this study cannot demonstrate functional insufficiencies and quality of life problems related to the chronological process of the disease, which is a significant limitation of the study. Prospective studies assessing this would be valuable.

In conclusion, tendon thickening and increased effusion in the shoulder joint are significantly higher in symptomatic dialysis patients than in asymptomatic dialysis patients and stage 4 CKD patients. However, tendon thickness and effusion are also common in asymptomatic dialysis patients and stage 4 CKD patients who are not yet on dialysis, but this is not statistically significant. These results suggest that shoulder joint involvement can initiate in CKD patients who are not yet on dialysis.