Low-term results from non-conventional partial arthroplasty for treating rotator cuff arthroplasthy.

OBJECTIVE: To evaluate the evolution of the functional results from CTA(®) hemiarthroplasty for surgically treating degenerative arthroplathy of the rotator cuff, with a mean follow-up of 5.4 years.
METHODS: Eighteen patients who underwent CTA(®) partial arthroplasty to treat degenerative arthroplathy of the rotator cuff between April 2007 and June 2009 were reevaluated, with minimum and mean follow-ups of 4.6 years and 5.4 years, respectively. Pre and postoperative parameters for functionality and patient satisfaction were used (functional scale of the University of California in Los Angeles, UCLA). All the patients underwent prior conservative treatment for 6 months and underwent surgical treatment because of the absence of satisfactory results. Patients were excluded if they presented any of the following: previous shoulder surgery; pseudoparalysis; insufficiency of the coracoacromial arch (type 2 B in Seebauer's classification); neurological lesions; or insufficiency of the deltoid muscle and the subscapularis muscle.
RESULTS: With a mean follow-up of 5.4 years, 14 patients considered that they were satisfied with the surgery (78%); the mean range of joint motion for active elevation improved from 55.8° before the operation to 82.0° after the operation; the mean external rotation improved from 18.9° before the operation to 27.3° after the operation; and the mean medial rotation remained at the level of the third lumbar vertebra. The mean UCLA score after the mean follow-up of 5.4 years was 23.94 and this was an improvement in comparison with the preoperative mean and the mean 1 year after the operation.
CONCLUSION: The functional results from CTA(®) hemiarthroplasty for treating rotator cuff arthroplasty in selected patients remained satisfactory after a mean follow-up of 5.4 years.

Introduction

The first author to describe the clinical findings from arthropathy of the rotator cuff was Robert Adams, in 1857. In 1981, Halverson et al. described the “Milwaukee shoulder”, in which crystals of calcium phosphate such as hydroxyapatite were involved in a cellular reaction with release of collagenases and joint destruction. However, Neer was the first to use the term “arthropathy of the rotator cuff”, in 1977, in a study published in 1983. Neer believed that extensive injury to the rotator cuff was the cause of the arthropathy and presented the hypothesis that this pathological condition might be the result of mechanical factors such as anterosuperior instability, and nutritional factors such as loss of the closed joint space, with impairment of nutrient diffusion to the joint surface. Interruption of the bone circulation that is provided by the rotator cuff also contributes toward the metabolic loss at the humeral head. The final result from these mechanical and metabolic alterations, in association with osteopenia through disuse of the glenohumeral joint due to pain, consists of collapse of the glenohumeral joint.1, 2, 3, 4

More recently, in 1997, Collins and Harryman produced a synthesis from the two theories and formulated the hypothesis that cranial migration of the humeral head, resulting from loss of the stability that the rotator cuff provides, leads to abnormal glenohumeral contact and formation of debris in the joint. Thus, an inflammatory cascade caused by the calcium phosphate crystals that are released is developed.

The incidence of rotator cuff injuries increases with age. They are relatively rare before the age of 40 years, become more frequent in the fifth and sixth decades of life and continue to increase in the seventh decade and beyond. Many cases do not present symptoms and approximately 50% of all individuals over the age of 80 years may have asymptomatic rotator cuff injuries.6, 7

Arthropathy of the rotator cuff mainly affects elderly women on their dominant side and it triggers chronic symptoms such as progressive pain, which worsens at night and with activities that require use of the shoulder. Other symptoms include weakness and difficulty in raining the arm, and these give rise to functional limitation. Physical examination reveals signs of extensive injury to the rotator cuff, such as atrophy of the supraspinatus and infraspinatus muscles.2, 8, 9, 10

Radiographs show glenohumeral arthrosis, with cranial displacement of the humeral head, which may give rise to abnormal contact between this and the coracoacromial arch and thus lead to “rounding” of the greater tubercle (“femoralization”) and to concave erosion of the coracoacromial arch (“acetabulization”). Using radiographs in anteroposterior (AP) view, Hamada et al. described the natural evolution of extensive rotator cuff injuries, with the development of degenerative arthropathy, and proposed a classification system consisting of five evolutionary stages. However, these do not guide the therapy.

Seebauer developed a biomechanical, functional and morphological classification system that presents therapeutic relevance and assesses the integrity of the anterior stabilizers of the shoulder and coracoacromial arch, the presence of dynamic stability and the upward migration of the humeral head. Additional examinations, such as computed tomography and magnetic resonance imaging, are not necessary for diagnosing arthropathy of the rotator cuff, but they help in making preoperative assessments to analyze the bone stock and the conditions of the rotator cuff, such as fatty degeneration.12, 13, 14

Treatments for arthropathy of the rotator cuff should be started using non-surgical methods, such as modification of activities, use of analgesic and/or anti-inflammatory medications and use of subacromial corticosteroid infiltration.6, 15

Surgical treatment is indicated for patients who do not respond to conservative treatment. Procedures such as arthroplasty to resect the humeral head and glenohumeral arthrodesis are considered to be salvage methods, to be used in patients presenting multiple surgical failures, deficiency of the deltoid muscle and infection. Arthroscopy for debridement, tenotomy of the biceps and tuberculoplasty can be performed, particularly in elderly patients and those with low functional demands. Conventional total arthroplasty of the shoulder is now contraindicated in patients presenting arthropathy of the rotator cuff because of the high rate of loosening of the glenoid component. The current alternative arthroplasty options for arthroplasty of the rotator cuff are non-conventional (CTA®) partial arthroplasty and use of a reverse prosthesis.3, 16, 17, 18

CTA® partial arthroplasty presents greater lateral extent with coverage of the tubercle and produces better contact and connection with the coracoacromial arch (Fig. 1, Fig. 2). Reverse prostheses are based on the concepts of Gramont et al., involving moving the center of rotation medially and distally, with gains in deltoid muscle function. This principle improved the stability of the implant and the range of motion. Nonetheless, despite the good results from reverse prostheses, this is a technically more complex procedure with higher complication rates (5% to 33%). CTA® hemiarthroplasty presents good results in selected patients, with lower incidence of complications than that of reverse prostheses.3, 20, 21, 22, 23, 24

Fig. 1

CTA prosthesis with its lateral extent, shown on intraoperative photo taken by the author.

Fig. 2

CTA prosthesis shown on postoperative radiograph produced by the author.

Patients who are candidates for CTA® hemiarthroplasty need to be free from pseudoparalysis, present a coracoacromial arch that maintains the relative kinematics of the shoulder joint, without anterosuperior escape (Seebauer types IA, IB and IIA), absence of previous surgery involving resection of the coracoacromial arch, functioning motor (intact deltoid) and sufficient subscapular muscle.3, 25, 26, 27, 28

The objective of this study was to evaluate the evolution of the functional results from CTA® partial arthroplasty for surgically treating degenerative arthropathy of the rotator cuff, after a mean follow-up of 5.4 years.

Methods

Between December 2006 and June 2009, 23 shoulders of 23 patients underwent CTA® partial arthroplasty to treat arthropathy of the rotator cuff. During a mean follow-up of 1.6 years, there were improvements in the clinical parameters and UCLA score, as described in the paper by Brasil Filho et al. These patients were evaluated prospectively in the present study after a mean follow-up of 5.4 years.

Among the 23 patients who were included in the first study, three were excluded from the present study because they had died in the meantime and two because they were lost from the follow-up. Thus, 18 patients remained in the study (Table 1). Among these, there was one patient who evolved with late postoperative infection and required surgery to remove the prosthesis.

Table 1

Patient data.

Patient	Sex	Time since op (years)	Age	Side	UCLA before op	UCLA one year after op	UCLA final evaluation	Elev/ext rot/med rot before op	Elev/ext rot/med rot after op
1	M	6.7	86	D	1 + 2 + 1 + 3 + 0 = 7	10 + 4 + 1 + 3 + 0 = 18	10 + 4 + 2 + 2 + 0 = 18	30/10/L2	40/10/L2
2	M	6.6	69	D	2 + 2 + 2 + 3 + 0 = 9	8 + 4 + 3 + 3 + 5 = 23	10 + 10 + 5 + 4 + 5 = 34	80/40/L3	120/50/L2
3	M	6.5	79	ND	2 + 4 + 3 + 3 + 0 = 12	8 + 6 + 3 + 3 + 5 = 25	8 + 8 + 4 + 3 + 5 = 28	54/0/T11	130/10/T8
4	F	6.0	89	D	2 + 4 + 3 + 2 + 0 = 11	8 + 6 + 3 + 3 + 5 = 22	8 + 6 + 3 + 3 + 5 = 25	40/10/L1	56/10/L1
5	F	5.8	67	ND	2 + 2 + 1 + 3 + 0 = 8	10 + 6 + 2 + 3 + 5 = 26	10 + 8 + 2 + 2 + 5 = 27	50/20/T12	70/40/T11
6	F	5.7	88	D	2 + 2 + 0 + 3 + 0 = 7	8 + 6 + 2 + 4 + 0 = 20	8 + 6 + 3 + 3 + 0 = 20	68/24/T12	80/30/T12
7	F	5.5	77	D	2 + 4 + 1 + 2 + 0 = 9	8 + 4 + 1 + 3 + 5 = 21	10 + 4 + 0 + 2 + 5 = 21	12/10/L5	20/20/L3
8	F	5.4	81	D	2 + 2 + 2 + 3 + 0 = 9	8 + 8 + 3 + 4 + 5 = 28	8 + 6 + 3 + 3 + 5 = 25	60/40/L1	110/44/L1
9	F	5.2	73	D	2 + 2 + 1 + 3 + 0 = 8	4 + 4 + 3 + 3 + 0 = 14	2 + 4 + 3 + 3 + 0 = 12	62/10/L1	90/20/L2
10	F	5.1	84	D	2 + 2 + 2 + 2 + 0 = 8	4 + 6 + 3 + 2 + 0 = 25	6 + 4 + 3 + 4 + 0 = 17	60/20/T11	70/20/T10
11	F	5.0	78	D	2 + 4 + 3 + 2 + 0 = 11	6 + 8 + 5 + 4 + 5 = 28	6 + 8 + 5 + 4 + 5 = 28	70/20/L2	120/36/L2
12	F	4.9	72	ND	2 + 2 + 0 + 2 + 0 = 6	8 + 4 + 2 + 3 + 5 = 22	8 + 8 + 3 + 4 + 5 = 28	56/40/L3	70/44/L2
13	F	4.7	70	ND	2 + 4 + 3 + 2 + 0 = 11	8 + 6 + 2 + 3 + 5 = 24	8 + 6 + 2 + 3 + 5 = 24	50/16/L1	60/30/T12
14	F	4.6	74	D	2 + 2 + 2 + 3 + 0 = 9	4 + 4 + 3 + 4 + 5 = 20	8 + 8 + 5 + 4 + 5 = 30	60/10/T10	110/20/T8
15	F	4.6	83	D	2 + 2 + 1 + 3 + 0 = 8	6 + 4 + 2 + 3 + 5 = 18	6 + 4 + 3 + 3 +5 =21	52/10/Trochanter	70/26/Sacrum
16	F	4.6	81	D	2 + 4 + 3 + 2 + 0 = 11	8 + 8 + 4 + 4 + 5 = 29	8 + 6 + 4 + 4 + 5 = 27	70/10/T10	120/22/T8
17	F	5.2	87	D	1 + 2 + 3 + 3 + 0 = 9	6 + 4 + 3 + 3 + 5 = 21	6 + 4 + 3 + 3 + 5 = 21	40/20/L3	60/30/L3
18	F	6.0	66	ND	2 + 2 + 2 + 3 + 0 = 9	8 + 4 + 3 + 4 + 5 = 24	8 + 6 + 2 + 4 + 5 = 25	90/30/L5	80/30/L5

M, male; F, female; D, dominant side; ND, non-dominant side; Elev, elevation; ext rot, external rotation; med rot, medial rotation; op, operation; UCLA, University of California, Los Angeles.

All the patients were operated by the same surgical team (from the Shoulder and Elbow Group of the State of São Paulo Public Servants’ Hospital). A deltopectoral access route was used.

The length of postoperative follow-up ranged from 4.6 to 6.7 years, with a mean of 5.4. The mean age was 78 years. The dominant limb was affected in 13 patients (72.2%).

The Seebauer classification was used. In stage IA, the head is centered in the glenoid; in IB, the head migrates medially and the glenohumeral space becomes pinched; in IIA, the humeral head migrates superiorly, but is stabilized by the coracoacromial arch, which remains intact; and in IIB, the humeral head migrates anterosuperiorly, due to insufficiency of the coracoacromial arch.

Among the 18 patients included in this study, three were classified before the operation as Seebauer lA, seven as lB and eight as llA.

The inclusion criteria were that the patients needed to be symptomatic and classified as Seebauer lA, lB and llA, who did not improve with conservative treatment over a minimum of six months. The exclusion criteria were situations in which the patients improved through clinical treatment or presented previous surgery or neurological lesions in the limb affected, arthropathy classified as Seebauer llB or insufficiency of the deltoid muscle and subscapularis muscle.

In evaluating the results, the functional scale of the University of California in Los Angeles (UCLA) was used, as modified by Ellman and Kay. To evaluate satisfaction, the Neer criteria were used. To measure the range of motion, the method of the American Academy of Orthopedic Surgeons was used. To compare the UCLA score and range-of-motion results, the nonparametric Friedman test was used.27, 28, 29, 30

The statistical significance of the differences in means between the quantitative variables was ascertained by means of the paired Student's t test and the differences in variance were ascertained by means of analysis of variance (ANOVA). The normality of the variables was tests using the Shapiro–Wilk test. All of the analyses were performed using a significance level of 5%. Results with p-values <0.05 were considered to be statistically significant. Two-tailed optional hypotheses were always envisaged.

The information gathered formed a database that was developed using the Excel® software for Windows and the statistical analysis was performed using the Stata® 11 SE and SPSS® 16.0 software.

Results

After a mean follow-up of 5.4 years, 14 patients considered that they were satisfied with the surgery (78%). Among the four who were dissatisfied, three complained about their lack of gain in range of motion, although they reported having achieved an improvement in pain in relation to before the operation. For one patient, the dissatisfaction was due mainly to pain (Fig. 3).

Fig. 3

Patient distribution according to satisfaction level after the operation.

In relation to the range of motion after a mean follow-up of 5.4 years, there was an improvement in the mean active elevation, which went from 55.8° before the operation to 82° after the operation. The mean external rotation improved from 18.9° before the operation to 27.3° after the operation (Fig. 4). The mean medial rotation remained at the level of the third lumbar vertebra.

Fig. 4

Comparison of the mean angles of elevation and external rotation from before to after the operation.

The mean UCLA score after the mean follow-up of 5.4 years was 23.94 and this was a significant improvement in comparison with the preoperative mean of nine (p < 0.001). A small improvement was observed in relation to the mean after the first postoperative year (22.39), but without statistical significance. The mean pain level was 7.67, with a range from 2 to 10; function was 6.11, ranging from 4 to 10; active flexion was 3.06, ranging from 0 to 5; anterior flexion force was 3.22, ranging from 2 to 4; and satisfaction was 3.89, ranging from zero to 5. There were statistically significant improvements in all the criteria for assessing the UCLA score (Table 2 and Fig. 5).

Table 2

P values for the variables of the UCLA score, compared between before the operation, after one year of follow-up and at final evaluation.

Variables	Before operation		One year afterwards		Final evaluation		p
	Mean (SP)	Min–Max	Mean (SP)	Min–Max	Mean (SP)	Min–Max
Pain	1.89 (0.32)	1–2	7.22 (1.83)	4–10	7.67 (1.97)	2–10	<0.001
Function	2.67 (0.97)	2–4	5.33 (1.53)	4–8	6.11 (1.88)	4–10	<0.001
Active flexion	1.83 (1.04)	0–3	2.67 (0.97)	1–5	3.06 (1.26)	0–5	<0.001
Flexion force	2.61 (0.50)	2–3	3.28 (0.57)	2–4	3.22 (0.73)	2–4	0003
Satisfaction	0.00 (0.00)	0–0	3.89 (2.14)	0–5	3.89 (2.14)	0–5	<0.001
UCLA score	9.00 (1.64)	6–12	22.39 (4.23)	14–29	23.94 (5.30)	12–34	<0.001

Fig. 5

Comparison of the UCLA scores before the operation, one year afterwards and at the final evaluation.

There were significant improvements between the pre and postoperative evaluations, both at one year after the operation and at the end of the follow-up. However, there was no statistically significant change between the two postoperative evaluations, performed at means of one and 5.4 years after the operation (Table 3).

Table 3

P values for the variables of the UCLA score over separate times.

	Pain	Function	Active flexion	Flexion force	Satisfaction	UCLA score
Before operation vs. one year after operation	p < 0.001	p < 0.001	p < 0.002	p < 0.002	p < 0.001	p < 0.001
Before operation vs. final evaluation	p < 0.001	p < 0.001	p < 0.003	p < 0.005	p < 0.001	p < 0.001
One years after operation vs. final evaluation	p = 0.157	p = 0.448	p = 0.207	p = 1.00	p = 1.00	p = 0.303

Discussion

CTA® partial arthroplasty for treating arthropathy of the rotator cuff is a relatively recent procedure, with few studies available in the literature, especially with long-term follow-ups.3, 31, 32

Vitotsky et al. conducted a study with a mean follow-up of 32 months and minimum of two years, on 60 patients who underwent CTA® partial arthroplasty, including Seebauer IA, IB and IIA patients. They obtained satisfactory results in 89% of the cases, with mean improvements of 22° in external rotation and 60° in flexion. In our sample, after a minimum follow-up of 4.6 years and mean of 5.4 years, among 18 CTA® partial arthroplasty procedures in 18 patients, the mean satisfaction rate obtained was 78%, with a mean improvement in elevation from 55.8° to 82° and in external rotation from 18.9° to 27.3°. Just as in our study, Vitotsky et al. did not include Seebauer IIB patients.

Over a mean follow-up of 3.7 years, Goldberg et al. obtained a satisfaction rate of 78%, with mean improvements of 33° in elevation and 23° in external rotation through using conventional hemiarthroplasty. The patients with a minimum elevation of 90° achieved the best results. In our study, patients with elevations of less than 90° were excluded.

In a study with a mean follow-up of 28.2 months on 15 cases of hemiarthroplasty, Zuckerman et al. obtained mean improvements of 17° in elevation and 14° in lateral rotation. The satisfaction rate among the patients was 87% and the UCLA score improved from 11 to 22 points.

Checchia et al. followed up 11 patients who underwent hemiarthroplasty to treat arthropathy of the rotator cuff, for a mean of 69 months. They obtained a pain improvement rate of 81.8%, satisfactory results in 54% and a mean UCLA score of 22.7 points. These authors observed that certain factors were associated with unsatisfactory evolution, such as previous surgery on the shoulder with impairment of the coracoacromial arch and previous injury of the deltoid muscle. In our sample, patients with previous shoulder surgery and those classified as Seebauer IIB were excluded.

In our study, patients whose main preoperative symptom was limitation of movements presented unsatisfactory results after the surgery, such that three of the four dissatisfied patients reported this complaint. This finding is in conformity with the study by Nam et al.

The UCLA functional score, which assesses pain, function, active flexion, anterior flexion force and satisfaction, improved from poor (mean of nine points) before the operation, to reasonable after follow-ups of one year and 5.4 years (means of 22.39 and 23.94 points, respectively), which confirmed that hemiarthroplasty was a good option for surgically treating arthropathy of the rotator cuff in selected patients. There was a statistically significant improvement in UCLA, in relation to before the operation, while the difference between the mean postoperative times of one year and 5.4 years was small and non-significant. This can be understood as maintenance of the positive results from the prosthesis over this postoperative period.

Since this is a surgical procedure indicated for elderly patients, one of the factors that caused difficulty in carrying out the present study was in relation to making long-term reevaluations on all the patients, because of deaths and loss of follow-up.

Conclusion

The functional results from non-conventional CTA® partial arthroplasty for treating arthropathy of the rotator cuff in selected patients remained satisfactory after a mean follow-up of 5.4 years.

Conflicts of interest

The authors declare no conflicts of interest.