The purpose of this study is to investigate the potential insults of rotator cuff muscle and the functional recovery of upper extremity function after antegrade nailing of humeral fractures.

Background:

With improved implant design and surgical technique, operative treatment of humeral shaft fractures increasingly has become accepted. Such treatment offers several advantages. The rigidly fixed, fractured limb can be used early without external support, and the patient can sleep, as normal, in the supine position. In addition, the angular deformity often seen with conservative treatment effectively is prevented. Although plate osteosynthesis can afford a rigid fixation and good functional recovery, its disadvantages have been reported. By comparison, locked nailing, which provides sufficient fixation stability and prevents slipout of the nail, offers the advantages of less soft tissue injury, a lower infection rate, and no need for radial nerve extrication. Locked nailing is preferred especially for open fractures, comminuted fractures, pathologic fractures, and fractures associated with osteoporosis. The use of Seidel locked nails with spreading fins can be complicated by inadequate fixation, intraoperative bone comminution, and unsatisfactory recovery of shoulder function. Biomechanically, locked nails with transfixing locking screws allow much better rotational control of the distal fragment than do Seidel nails. Most rigid humeral nails are inserted antegrade, but a disadvantage to antegrade nailing has been the risk of shoulder function impairment, which partially negates the advantages of closed nailing and of loadsharing mechanical properties. The disadvantages of retrograde nailing are potential risk of elbow joint injury and nonlinear entry portal of nailing.

Gaullier O et al did a study of 23 cases with rotator cuff evaluation in 1999. Only humeral diaphyseal fractures were enrolled. Fracture type classification was according to AO. Clinical shoulder assessment was carried out using the Constant score. Treatment was Seidel intramedullary locked nailing.

Seidel intramedullary locked nailing was too big for patients in Taiwan. Compared to patients in western countries, our patients had small stature. The humeral locked nail, invented by Jinn Lin, provided a proper treatment option for humeral fractures in Asian people.

Although H-L nail has the advantages of minimal tissue trauma, it still has the possibility of rotator cuff injury. The purpose of our study is to evaluate rotator cuff injury in antegrade humeral locked nailing for humeral fractures by means of sonographic evaluation and shoulder functional examination.

Materials and methods:

I. Implant design:

The Humeral Locked nail is a uniform closed cylindrical tube with a 5° bend at 70 mm from the threaded end. It is manufactured from 316L stainless steel (Carpenter Technology Corporation, Indianapolis, IN) in 4 sizes: 7 (diameter) x 190 (length) mm, 7 x 210 mm, 8 x 210 mm, and 8 x 230 mm. The thickness of the nail wall is 1.5 mm for 8 mm nails and 1 mm for 7 mm nails. There are 3 parallel screw holes (the 2 outermost holes usually are used) at the threaded end, 45° oblique to the axis of the nail, and 2 transverse screw holes at the other end. Cortical screws are used as locking screws that are inserted under the guide of an L shaped, nail mounted targeting device.

II. Patient selection:

The inclusion criteria for patients in the present study were listed as follows: (1) subjects must sign and date informed consent; (2) subjects must be over fifteen years old, of either sex; (3) subjects suffered from acute humeral fractures or non-unions; (4) subjects were treated by antegrade H-L nailing (United, Taipei, Taiwan); 5) subjects had solid union of fractures; 6) postoperative time was at least 12 months; 7) subjects were operated between 2003 and 2006. The exclusion criteria for patients in the present study were listed as follows: (1) subjects refusing to take the examinations, (2) subjects lost to follow up, (3) subjected with pre-existing shoulder problems.

III. Study design:

Our study design is a retrospective, cross-sectional study. Patients with acute humeral fractures or nonunions treated by antegrade H-L nailing were recruited and studied. Sonographic evaluation of rotator cuff was done by an experienced orthopedist and shoulder functional scores such as Neer score, Constant score and QuickDASH score was assessed at the same time.

Real-time sonography was performed with a Sonoline Sienna scanner (HDI-5000, Philips, USA) equipped with a 5.5-9.4-MHz linear-array transducer. The patient was seated on a stool with the examiner sitting in front of him. Lateral transverse view and lateral longitudinal view were acquired. For these two views, the arm of the examined patient was placed in adduction and internal rotation, which gave a better exposure of the supraspinatus tendon under the acromion. Criteria for full-thickness rotator cuff tears were nonvisualization of the rotator cuff and hypoechoic or anechoic discontinuity. Besides, criteria for partial- thickness rotator cuff tears included focal heterogenous hypoechogenecity and incomplete hypoechoic cleft. Measurement of rotator cuff tears was done on lateral transverse and longitudinal views. The width of rotator cuff tears was measured on lateral transverse view, while the height of rotator cuff tears was measured on lateral transverse view. Moreover, the length of rotator cuff tears was measured on lateral longitudinal view. The thickness of rotator cuff at the dome of humeral head was measured on lateral transverse and compared between the operative and sound sides. We calculate the ratio of the thickness of supraspinatus on the operative side to that on the sound side.

a. Thickness rotator cuff tear:

Size (full or partial):

i. width (mm) ii. height (mm) iii. length (mm)

b. visualization of nail

c. scar: nature and size

d. supraspinatus thickness(mm) :

IV. Baseline variables and primary endpoints:

We included patient's age and gender, fracture location, fracture cause, presence of nail removal, post-operative period and comorbidity or associated disease as independent variables. Fracture location was either humeral neck or humeral shaft. Humeral neck fractures were defined as fractures above surgical neck of the humerus. Humeral neck fractures were further categorized into three groups: two-part, three-parts or four-parts, according to Neer classification. Meanwhile, humeral shaft fractures were defined as fractures below surgical neck of the humerus and 5 cm above the olecarnon fossa. Humeral shaft fractures were separated into three groups: proximal third shaft fractures, middle third shaft fractures and distal third shaft fractures. Fracture causes included simple falls and traffic accidents. We calculate post-operative period according to the latest operation involving rotator cuff injury, which included antegrade H-L nailing and nail removal. For example, if patient X was treated by antegrade H-L nailing two years ago and received nail removal one year ago, his (or her) post-operative period was twelve months rather than twenty-four months.

The primary endpoints of the present study were functional scores and sonographic findings. As to functional scores, we chose Neer score, Constant score and QuickDASH score as our endpoints. Neer score consists of pain component, function component, range of motion component and anatomy component. The interpretation of Neer score are listed as follows: excellent when Neer score is above 89 units, satisfactory when Neer score is between 80 to 89 units, unsatisfactory when Neer score is between 70 to 79 units, failure when Neer score is below 70 units. Constant score consist of pain, activity level, arm positioning, strength of abduction and range of motion. We examined Constant score of affected and healthy shoulder and calculated the difference of Constant scores between two shoulders. The interpretation is according to the difference of Constant scores: excellent (0~5 points), very good (6~10 points), good (11~20 points), fair (21~30 points) and poor (above 30 points). The QuickDASH outcome measure is a questionnaire which asks about a patient's symptoms as well as his (or her) ability to perform certain activities. Patients should answer this questionnaire based on their conditions in the last week.

V. Statistics:

We include age, gender, fracture location, fracture cause, the presence of nail removal, post-operative period and comorbidity as independent variables. We use Neer score, Constant score, QuickDASH score, the presence of full-thickness rotator cuff tear, partial-thickness rotator cuff tear and scar , supraspinatus thickness ratio as dependent variables. Univariate analyses were conducted with Student's t tests for continuous variables and Chi square tests for binary variables. We also do multivariate regression analyses to adjust the confounding factors. The difference is considered statistically significant when p value is under 0.05.

VI. Sample size calculations (power analysis):

Neer score:

The primary endpoint is used for calculation of sample size. We assume that the standard deviation of Neer score in our patients is 12 points. How many patients are required to have a 80% power of detecting a 10-point difference in the mean of functional score between the two groups at the 5% level of significance? The required sample size is 48 patients (24 patients for each group).

Inclusion Criteria:

• subjects must sign and date informed consent
• subjects must be over fifteen years old, of either sex
• subjects suffered from acute humeral fractures or non-unions
• subjects were treated by antegrade H-L nailing (United, Taipei, Taiwan)
• subjects had solid union of fractures
• postoperative time was at least 12 months subjects were operated between 2003 and 2006.

Exclusion Criteria:

• subjects refusing to take the examinations
• subjects lost to follow up
• subjected with pre-existing shoulder problems.

Patients operated in National Taiwan University hospital