Failure after ulnar nerve decompression at the elbow can be defined as either no change in the patient’s symptoms or an initial improvement with recurrence, making the patient history essential in the work-up. Failure may be due to diagnostic, technical, or biologic factors. Technical errors and the development of perineural fibrosis necessitate revision surgery, while nerve damage due to chronic severe compression should be observed. We do not believe any one procedure is superior in the revision setting as long as a complete decompression is achieved with a compression free, stable transposition of the surgeon’s choice.
Failure after ulnar nerve decompression at the elbow can be defined as either no change in the patient’s symptoms after surgery or an initial improvement with recurrence of symptoms.
Failure may be attributable to diagnostic (incorrect diagnosis or coexistent diagnoses), technical (incomplete decompression, creation of a new area of compression or subluxation, or injury to the cutaneous or ulnar nerves), or biologic (perineural fibrosis or significant preoperative nerve damage) factors.
Diagnosis must be verified with a basic history and physical examination ± further studies, such as a nerve study if one was not performed initially.
Technical errors and the development of perineural fibrosis necessitate revision surgery, whereas nerve damage owing to chronic severe compression should be observed.
No one procedure is superior in the revision setting as long as a complete decompression from the Arcade of Struthers to the flexor carpi ulnaris (FCU) and flexor/pronator aponeurosis is achieved with a compression-free, stable transposition of the surgeon’s choice.
Compression of the ulnar nerve around the elbow, also known as cubital tunnel syndrome, is the second most common compressive neuropathy in the upper extremity after carpal tunnel syndrome. Unlike carpal tunnel syndrome, however, there are multiple sites of compression as well as multiple surgical techniques with cubital tunnel syndrome. Surgical options include open in situ decompression, endoscopic decompression, anterior transposition (subcutaneous, intramuscular, submuscular), and medial epicondylectomy. The subsequent evaluation and treatment of a failed ulnar nerve decompression at the elbow remain similar to the evaluation and treatment for a new diagnosis of cubital tunnel syndrome, but the outcomes for this problem are more unpredictable. Discussion of surgical treatment failure for cubital tunnel syndrome warrants defining the term “failure,” and reviewing its causes and the optimal workup and treatment for this condition.
Defining a surgical “failure” after a cubital tunnel release, with or without transposition, requires a standardized approach. The 2 main types of failures include (1) patient’s symptoms are unchanged, and (2) patient’s symptoms initially improved and then recurred ( Table 1 ). Failure may be attributable to an error in diagnosis (incorrect diagnosis or coexistent diagnoses), technique (incomplete decompression, creation of a new area of compression or subluxation, or injury to the cutaneous or ulnar nerves), or biologic (perineural fibrosis or significant preoperative nerve damage) factors ( Table 2 ). All these causes of failure except for perineural fibrosis, will lead to unchanged symptomatology. Fibrosis may be the cause of symptom recurrence after a period of improvement after the initial operation.
|Patient Symptoms||Likely Etiologies||Workup||Treatment Options/Recommendations|
|• No change in symptoms||• Incorrect diagnosis||• History and physical examination, nerve study (if one had not been done) or repeat study, cervical and/or wrist magnetic resonance imaging||• Treatment for secondary compression or alternative diagnosis|
|• Incomplete decompression or new site of compression||• Consider repeat nerve study||• Revision decompression|
|• Significant nerve damage before index procedure||• Repeat nerve study to document nerve improvement||• Patient counseling, observation|
|• Resolved, then recurred||• Perineural fibrosis||• Repeat nerve study||• Revision decompression|
• Incorrect diagnosis
• Coexisting sites of compression
• Inadequate decompression
• Creation of new sites of compression during transposition
• Persistent or new ulnar nerve subluxation/instability
• Injury to ulnar nerve or medial antebrachial cutaneous nerves
• Perineural fibrosis/cicatrix
• Chronic, severe distal sensory and motor changes
• Elbow stiffness
Incidence of Failed Decompression
In his review of more than 2000 patients undergoing ulnar nerve release, Dellon found that nearly 90% of patients with moderate compression achieved excellent results, but 20% to 35% of patients with severe compression noted a recurrence or “worsening” symptoms. He did add that in recurrent cases, the most favorable results were achieved when an internal neurolysis was performed in addition to a submuscular transposition. Bartels reported on more than 3100 surgeries in which patients with unchanged symptoms or worsening of neurologic deficits occurred in 10.9% of simple decompressions, 14.6% of subcutaneous transpositions, 10.0% of intramuscular transpositions, and 21.0% of submuscular transpositions. In a meta-analysis reviewing 903 cases, Mowlavi and colleagues found that in moderate-staged patients, no one treatment was statistically superior, with a recurrence rate of 4% and total relief in 80% of patients with submuscular transposition; however, in the severe-staged, all modalities were noted to provide similarly poor results, with 25% recurrence with decompression alone or transposition.
Determining Cause of Initial Failed Release
First, it must be considered that if there is little or no relief from the initial ulnar nerve release at the elbow that the pain and dysfunction are a result of another condition that causes ulnar nerve irritation or symptoms similar to ulnar nerve problems. If the surgery was based on clinical examination findings only, without a confirmatory electrodiagnostic study, the initial assumed diagnosis of an isolated cubital tunnel syndrome may be incorrect. Even if an electrodiagnostic study was performed, the reported diagnostic accuracy ranges from 20% to 100%. Because conduction velocities are variable, there is no absolute sensory and motor conduction value that is diagnostic of ulnar neuropathy at the elbow. Motor studies are clearly less useful in the evaluation of patients with absent or subtle examination findings. A study by Britz and colleagues found that in clinically mild cubital tunnel neuropathies, 73% had electromyography findings consistent with an ulnar neuropathy, but only 68% had findings at the elbow. Even in patients with “definite” clinical symptoms of cubital tunnel syndrome, abnormalities in sensory or mixed nerves identified only 86% as having an ulnar nerve problem. In patients with more subtle clinical presentations, they found these abnormalities in only 68% of patients.
Alternative compressive neuropathic diagnoses
Distally, compression of the ulnar nerve at the Guyon canal by either a ganglion cyst or aneurysm can produce similar symptoms in the hand as compression at the elbow, although this is estimated to have a 20 times lower incidence than cubital tunnel syndrome. Distinguishing between these either clinically or with nerve conduction studies may be difficult, as the motor nerve fibers to muscles in the forearm are often spared with compression at the elbow, whereas the motor and sensory fibers to the hand are more affected (which would also be affected by compression at the wrist). Hypothenar hammer syndrome may also present with ulnar nerve symptoms of the hand, although the primary etiology is an injury to the ulnar artery, leading to cold ulnar-sided digits but the ulnar nerve may also be involved. The dysvascular digits may present as subjective numbness and tingling. A history of repetitive occupational trauma in which the heel of the hand is often used as a hammer, can help make this diagnosis.
Compression occurring at the C8 nerve root from arthritis, or much more rarely a Pancoast tumor, can affect fibers that distally become the ulnar nerve, which would mimic symptoms of cubital tunnel syndrome. History-taking that includes inquiring about neck and radicular symptoms, as well as physical examination maneuvers, such as the “Spurling” test will help make the diagnosis. An electrodiagnostic study will also aid in the diagnosis. Furthermore, patients may have compression at the elbow and the cervical spine, which leads to the so-called, “double crush” phenomenon, in which release at the elbow would not allow for complete resolution of symptoms. Compression of the neurovascular bundle at the superior thoracic outlet, which is known as thoracic outlet syndrome, can also mimic cubital tunnel syndrome. These patients present with pain in the arms and hand, which may also be similar to a burning sensation, as well as possible discoloration of the hand. Lastly, a failure to diagnosis ulnar nerve subluxation over the medial epicondyle as the cause of ulnar nerve symptoms will lead to continued problems if in situ decompression is performed and subluxation is not evaluated intraoperatively.
Alternative noncompressive neuropathic diagnoses
Raynaud disease or phenomenon may also cause tingling and pain in the hand, but the accompanying skin blanching and association with cold, helps to differentiate it from a compressive neuropathy. Usually Raynaud disease affects all the fingers but may be isolated to particular fingers, and if only the ring and small are involved, this can be confused with cubital tunnel syndrome. Other causes of ulnar nerve irritation at and distal to the elbow include medial epicondylitis and FCU tendonitis, respectively, which can continue after ulnar nerve release. Another diagnosis that should be kept in mind is brachial plexus neuritis (Parsonage-Turner syndrome), which usually resolves without intervention. This diagnosis is made when the history and examination reveal abnormalities of the median, radial, and ulnar nerves. Last, polyneuropathy attributable to diabetes or a neurologic condition may be mistaken for cubital tunnel syndrome. Again, an electrodiagnostic study will be helpful in making this diagnosis.
The technical factors that lead to failure include failure to fully decompress the nerve, creation of new sites of compression or ulnar nerve subluxation, and injury to branches of the medial antebrachial cutaneous nerve or the ulnar nerve itself.
Amadio described 5 major anatomic sites of compression in cubital tunnel syndrome: the arcade of Struthers, the medial intermuscular septum, the medial epicondyle, the cubital tunnel, and the deep flexor pronator aponeurosis. In his case series of 41 patients, 3 had constriction at the arcade, 2 at the medial intermuscular septum, 13 at the region of the medial epicondyle, 17 at the cubital tunnel, and 4 at the distal flexor pronator aponeurosis, whereas 12 had no apparent constrictive focus. All these sites need to be considered when a revision is being considered. However, there have never been any large studies reporting the incidence of the location of compression in the primary setting, although the cubital tunnel and the 2 heads of the FCU are reported as the most common sites of compression. We consider the arcade of Struthers a potential site of compression if not resected with transposition, and it is unclear if it is ever a primary site of compression. Of note, Wehrli and Oberlin believe that the arcade of Struthers should be called the internal brachial ligament, as Struthers originally described it in 1854. In this article, they felt the medial intermuscular septum is only a site of compression if it is not resected with transposition. As for the medial epicondyle, it is unclear to us how it is a primary site of compression. Other potential sites include a hypertrophied triceps, a dislocation (snapping) of the medial portion of the triceps, an anomalous epitrochleo-anconeus muscle (also known as anconeus-epitrochlearis, subanconeus, or accessory anconeus), and the ligament of Spinner, which is a distinct aponeurosis between the flexor digitorum superficialis of the ring finger and the humeral head of the FCU.
Failures following incomplete decompression without transposition range from 6% to 20%. Gabel and Amadio found that during 30 revision cubital tunnel cases an average of 2.2 sites of compression were found: 24 had compression at the level of the medial epicondyle; 16 at the medial intermuscular septum; 13 at the cubital tunnel or the FCU arcade; and 7 at the arcade of Struthers and deep flexor pronator aponeurosis. Failure to recognize osteophytes at the medial epicondylar groove as a result of significant osteoarthritis may also result in incomplete decompression. Even if the overlying fascia is resected, the nerve may be compressed from within the floor of the cubital tunnel.
Creation of new areas of compression
Anterior transposition can move the nerve from the environment of the cubital tunnel and medial epicondyle, but can create other levels of mechanical impingement on the nerve. Kinking proximally at the arcade of Struthers or medial intramuscular septum, and distally at the fascial bands between the two heads of the FCU can lead to poor results after transposition. Subcutaneous transposition may uniquely leave the nerve exposed to external trauma, and the nerve has been frequently found directly overlaying the medial epicondyle at revision surgery. A medial epicondylectomy is purported to potentially result in scarring and fibrosis from the exposed cancellous bone, which could lead to recurrent ulnar nerve symptoms.
As reported before, among 30 revisions, Gabel and Amadio attributed the surgical failures to persistent compression at the medial intermuscular septum (16), cubital tunnel or FCU arcade (13), and the arcade of Struthers and deep flexor aponeurosis (7). In 13 cases of previous transpositions, Rogers and colleagues noted that the septum had been left intact in 12 cases. A severe kinking of the nerve over an unresected septum was observed in 6 of 10 cases by Filippi and colleagues, 1 whereas Caputo and Watson reported an incompletely resected septum in 15 of 20 cases. Vogel and colleagues found retained medial intermuscular septum (10), common flexor aponeurosis (9), and arcade of Struthers (5) as the cause of failure in their 18 patients who failed cubital tunnel surgery. Most recently, Mackinnon and Novak found the most common site of compression in 100 revision cubital tunnel surgeries to be distally at the fascial septum between the FCU and the flexor/pronator muscle mass. In their study, it was twice as common at this site when compared to the site of the medial intermuscular septum.
The overall incidence of anterior instability in the setting of simple decompressions is cited to range from 2.4% to 17.0% of cases. In 5 of 10 failures in the patients who had undergone decompression without transposition, it was found that dynamic subluxation was occurring with deep elbow flexion, which was assumed to be causing persistent irritation to the nerve. Nerve instability was found in 8 patients in Vogel and colleagues’ case series of 18 patients. Additionally, in cases where an inadequate subcutaneous sling is constructed, the ulnar nerve has been noted to fall back into the epicondylar groove. We suggest closing the cubital tunnel after transposition to prevent this from occurring.
Direct injury to the ulnar nerve, while rare, can be a catastrophic complication. More commonly, branches of the medial antebrachial cutaneous nerve are injured in the superficial dissection as they cross the incision within 6 cm proximal and distal to the medial epicondyle. In Mackinnon and Novak’s review of 100 failed ulnar nerve cases, they saw 73 medial antebrachial cutaneous neuromas, and felt this to be one of the major causes of failed decompression and transposition. However, some would characterize this as a complication of the procedure and not a failure. Some patients could interpret their symptoms from the neuroma as similar to their preoperative ulnar nerve symptoms. Mackinnon and Novak urged meticulous technique in the primary setting to prevent neuroma formation and electrocautery “capping” of excised neuroma nerve endings in the revision setting to prevent recurrence of painful neuromas. If the medial antebrachial cutaneous branch is injured, we would recommend primary repair using miscrosurgical technique.
Broudy and colleagues found in 10 revision cases that scarring at the transposition site was frequent. Gabel and Amadio’s finding of an average of 2.2 sites of compression and its various locations, which were reviewed previously, were all described as fibrosis. Rogers and colleagues reviewed 14 patients who had failed ulnar nerve release and found dense scarring of the nerve to the medial epicondylectomy site in 7, scarring in the cubital tunnel in the 1 patient who had in situ decompression, and 1 patient who had an intramuscular transposition had thick scarring within the flexor pronator mass. In a report of 22 patients who had revision ulnar nerve decompression surgery, Filippi and colleagues attributed 14 of the failures to perineural fibrosis. They found this in cases of both simple decompressions as well as transpositions. In 20 patients, Caputo and Watson found an average 1.9 sites of compression: FCU aponeurosis and deep flexor pronator mass (14), medial intermuscular septum (15), and the medial epicondyle (6). Vogel, and colleagues reported on 18 patients in whom most (15) had subcutaneous transposition, and found the most common operative findings to be perineural scarring (16). Dagregorio and Saint-Cast found in their 9 patients who failed a primary submuscular transposition, dense scarring at the fascial-muscular flap in all cases.
Preoperative nerve damage
Another reason for a lack of improvement in symptoms is significant preoperative nerve damage from chronic compression. The patient may see the index procedure as a failure if there is no change in symptoms or the improvement is mild or negligible. These outcomes have to do with the status of the nerve and are largely out of the hands of the surgeon and patient. The surgeon must recognize the possibility of irreversible nerve damage, communicate this to the patient, and manage the patient’s expectations. The patient needs to understand that in this setting complete recovery may not occur. Also, intrinsic muscle atrophy once present will most likely not recover. Most likely, the patient will experience some improvement in symptoms in terms of degree and areas involved, and function continues to improve beyond 2 years after surgery. The goal of surgery in chronic severe compression cases should be to prevent worsening of motor and sensory dysfunction and symptoms. Dellon found that in cases of long-standing symptoms and severe deficits, the results of surgical treatment are not satisfactory regardless of the surgical technique, with fewer than 50% of severe cases experiencing significant sensory improvement and just 25% experiencing sensory improvement.
Results in the past following ulnar nerve release were complicated by elbow stiffness resulting from excessive postoperative immobilization. Although technically not a failure of the nerve release, postoperative elbow stiffness can affect the functional use of the arm. Mackinnon and Novak argue that early elbow motion is essential to preventing recurrence by preventing scarring and additional sites of compression in the revision setting. Weirich and colleagues, however, did not find differences in pain, muscle strength, or overall satisfaction in a randomized study of early motion versus delayed motion after primary subcutaneous transposition. Currently for simple decompressions, most surgeons immobilize only until the wound has healed, or not at all. If a submuscular transposition is performed and the flexor pronator mass is reflected and then repaired, however, some would immobilize to allow the repair to heal.
Work-up of Failed Ulnar Nerve Release
Understanding the patient’s symptoms both before and after the initial surgery, including subtle difference in pain and parasthesias, will guide the appropriate workup. In cases of residual or recurrent pain, a thorough examination with the other differential diagnoses in mind must be performed. Starting proximally, provocative testing for cervical nerve root impingement (Spurling test and Lhermitte sign), and thoracic outlet maneuvers (Adson, Wright, and Roos stress test) should be evaluated. Along the course of the ulnar nerve, a Tinel that is moving distally from the site of the positive Tinel on initial presentation is a good prognostic sign of reinnervation. A Tinel at the Guyon canal can indicate compression occurring at this site, although it is much less common than at the elbow. Intact sensation of the dorsal ulnar hand (via the dorsal ulnar sensory branch) may help distinguish compression at the Guyon canal rather than at the elbow. Tenderness at the medial epicondyle or along the FCU may indicate medial epicondylitis or FCU tendonitis, respectively, rather than cubital tunnel syndrome.
The role of repeat electrodiagnostic studies remains controversial. Most agree that the clinical examination and the patient’s symptoms should dictate treatment decisions. In difficult cases, however, a worsening nerve study may help confirm the need for reexploration, whereas an improving electrodiagnostic study can be reassuring. Cervical magnetic resonance imaging (MRI) may confirm a clinical diagnosis of cervical nerve root impingement, whereas a wrist MRI can confirm a cyst or mass in the Guyon canal. In terms of grading systems, to date none of the grading systems can completely capture the patient experience and in the revision setting where many factors come into play; these grading systems have little prognostic value.
Revision nerve surgery is challenging, as tissue planes may be disrupted and the nerve may be encased in scar. The first step is to identify the nerve. To aid in accurate identification of the nerve, the prior incision should be extended, thereby allowing identification of the ulnar nerve in normal tissue planes. The dissection of the nerve should then be moved into the previous surgical field. If it remains difficult to dissect the nerve free from the surrounding scar tissue, we recommend including a cuff of scar tissue around the nerve to avoid damaging it.
When performing a revision release, we recommend a “maximally” invasive approach in which the nerve is fully exposed at all potential sites of compression from the arcade of Struthers to the 2 head of the FCU and the deep flexor pronator aponeurosis. This exposure and nerve dissection will most likely not allow the nerve to stay stable in its native location (assuming an in situ decompression was performed during the index operation), and it will therefore most likely need to be transposed if it has not been already. The best case scenario is if there is one focal area of perineural fibrosis or compression, which would explain the failure of the index operation, such as an unresected medial intermuscular septum after an anterior transposition. This is not always the case, however, as even during primary cubital tunnel surgery it is not always obvious where the compression is located. As Gabel and Amadio (2.2 sites) and Caputo and Watson (1.9 sites) showed in their case series, there is usually more than one site of compression found during the revision surgery.
As in primary cubital tunnel surgery, there is no one technique that has been definitively proven superior to the rest. Dellon stated, “that little more than personal bias is available for guidance in selecting treatment.” The same can be applied to revision cubital tunnel surgery. Different investigators have reported results for various procedures for revision surgery, including simple neurolysis, subcutaneous transposition, intramuscular transposition, and submuscular transposition. Most revision case series have reported on and supported submuscular transposition for unclear reasons. We believe it may be partly because it is the most technically demanding operation, leaving no other procedure that is more complex or intricate to chose. We agree with Caputo and Watson when they reviewed the articles by Broudy and colleagues, Gabel and Amadio, and Rogers and colleagues, 3 and stated that, “the rationale behind choosing a submuscular versus subcutaneous transposition for revision procedure in the 3 previous reports is not well described.”
We believe that success after revision surgery depends on finding and removing any external compression on the ulnar nerve and then placing the nerve in a stable bed with no subluxation. An environment that inhibits scar formation would be ideal but there is no proven method to inhibit this process. Some have suggested that vein wrapping may insulate the peripheral nerve from developing scar tissue. If the index procedure was a minimally invasive approach, such as endoscopic or in situ release of just the cubital tunnel, then the nerve will be in its native location. Nerve subluxation should be assessed, as it may be the cause of the failure, after complete decompression of all the previously reviewed potential sites of decompression. As stated previously, after revision decompression, the nerve will most likely not be stable in its native location and will need to be transposed. Our preferred transposition is subcutaneous with the reasoning that no structure is placed over the nerve (ie, the fascia of the flexor-pronator mass or the whole mass itself as in the intramuscular and submuscular transposition, respectively) and the nerve has a large potential space: all the space anterior to the fascia of the anterior elbow deep to the subcutaneous adipose tissue.
If the index procedure was subcutaneous transposition and scarring at that site was the etiology of the compression, then external neurolysis with revision of the subcutaneous transposition (fascia to subcutaneous flap) is an option so that the nerve sits stably in the transposed state; however, conversion to intramuscular or submuscular can be performed also. Likewise, if a submuscular or intramuscular site is excessively scarred, moving it out of the scarred bed to a subcutaneous site may be reasonable. Although revision ulnar nerve surgery at the elbow will require some external neurolysis, we see no role for internal neurolysis, as it most likely will injure the nerve further. Others agree that internal neurolysis may cause ischemia and infarction of the ulnar nerve.