Sunday, November 6, 2016

Nerve injury in shoulder arthroplasty

The risk of nerve injury during anatomical and reverse total shoulder arthroplasty: an intraoperative neuromonitoring study

These authors reviewed 36 consecutive patients who underwent reverse (RSA) (n = 12) or anatomic (TSA) (n = 24) shoulder arthroplasties with intraoperative neuromonitoring.

They found nearly 5 times as many postreduction nerve alerts per patient in the RSA cohort compared with the TSA cohort (2.17 vs. 0.46). 

Most of the nerve alerts for both groups in this study occurred during humeral (43% for TSA vs. 23% for RSA) and glenoid preparation (29% for TSA vs. 34% for RSA) while the arm is in external rotation, accounting for 65% of all nerve alerts in both groups.

There were 17 unresolved nerve alerts postoperatively, with only 2 clinically detectable nerve injuries, which fully resolved by 6 months postoperatively. 

A preoperative decrease in active forward flexion and the diagnosis of rotator cuff arthropathy were independent predictors of intraoperative nerve alerts.

The authors attributed the higher incidence of intraoperative nerve alerts in the post reduction stage in RSA  to the resultant arm lengthening which they state is "inherent in the Grammont-design RSA".

Comment: In their discussion the authors provide evidence that traction nerve injury is related to strain (6 to 12%) and the duration of application of the strain. They suggest that external rotation is the position placing greatest strain on the brachial plexus.  We agree that the 'danger position' is external rotation and extension. For this reason we limit the time the arm is in this position to under one minute at the time. 

Their findings also support the direct relationship between arm lengthening after reduction and neurologic injury. There is evidence that lengthening of 2 cm increases the risk of plexus injury. They raise the question of whether a non-Grammont design that lateralizes the glenosphere (or the humerus) has less brachial stretch and fewer nerve injuries.  We agree and for that reason we use a RSA technique that avoids excessive lengthening.

They observe that nerve injury may present without obvious sensory or motor dysfunction but as pain. We have seen instances of palm pain after shoulder arthroplasty that cannot be otherwise explained.

With respect to factors that may predispose to nerve injury, preoperative stiffness and methotrexate therapy can be added to the list.

Interested readers are directed to this related post:

Shoulder arthroplasty can be/is hard on the nerves around the shoulder