AANEM Connect

Having recently seen uncommon nerve injuries a few more than I would have, on 30-years past form, expected, I am prompted to share some thoughts. 

In headline, here are three: 

Man, seventies (intentionally not specific), felt numbness only at the tips of the little and ring fingers straight on coming round after open-heart surgery. He had no impairment of dexterity, ulnar SNAPs only a little reduced, and the referring spinal surgeon “struggling to think that the cervical spine is likely to be the cause of polyradiculopathy like this”. Cannulation through veins on the back of the ulnar side of the wrist is likely the culprit.
 
Man, forties, felt numbness and dull pain in the little and ring fingers and on the ulnar side of the hand and forearm immediately after – for anaesthesia for surgery on the knee – cannulation, inserted into a vein on the back of the ulnar side of the wrist. Symptoms lasted over two years.
 
Woman, fifties, felt numbness and pain in the middle finger straight after trapeziectomy. Few days later, the surgeon decompressed the carpal tunnel, with no benefit. SNAPs from index, middle and little fingers and APB’s amplitude severely reduced, with no slowing. Injury to the median nerve is likely to have resulted from regional nerve block anaesthesia.
 
These nerve injuries are unique for several reasons:
 

• Pain (“neuropathic”) alongside numbness is a distinctive feature.
• In most, the referring clinician is not only not considering what the culprit might be (as transpires – on the balance of probabilities – only at the end of the examination) but thinking along what might prove red herrings: mal-positioning of the elbow or neck, aggravation of pre-existing cervical radiculopathy, injury to the brachial plexus, and other. These red herrings may all too easily be “supported” by confirmation bias. The commonest is “compression” at the elbow, likely to be followed by another surgery, for “release”, not infrequently making matters worse.
• As the patient may not know or recall (having been anaesthetised), history may need digging out from records, or talking to the anaesthesiologists – as I have on numerous occasions and always found invaluable. 
• Sometimes already, or become litigious, they are uniquely educational, demanding relearning (and illustrating) minutia of anatomy beyond what us called for in academic tests. They can be fiercely contested – by lawyers and doctors, necessitating exclusion and inclusion in rigor beyond what is needed in routine clinics.
• Demonstrate the importance not only of detailed history but of comprehensive examination that includes both sides. To make a conclusion on a 45m/s across the elbow, wrist, or knee without knowing the asymptomatic side’s, is to court (as an “expert witness”) discredit – or worse.

 
Two references worth considering:
 
Nerves and Nerve Injuries. Sir Sydney Sunderland. Churchill Livingstone. Second Edition 1978.
I. Injection injuries. Section 5. Nerve injuries caused by the destructive or toxic action of certain therapeutic agents. Chapter 10. Causative agents. Pp. 173-6. 
Perioperative Peripheral Nerve Injuries. Abdul Ghaaliq Lalkhen, Kailash Bhatia. Cont Edu Anaesth Crit Care & Pain. 2012;12(1):38-42. 2012 Oxford University. 

There are others. From patients I have studied, mechanisms of anesthesia-related nerve injury include:

• Cutting action by the tip (crowned, in some designs, by multi-faceted bevel) of the injection needle as it passes through the nerve. Or injury by a cannula that, on being advanced through winding or networking veins at sharp angles, may breach the vein wall and allow extravasation of anaesthetic agents (sometimes containing adrenaline) or antibiotics coming into direct contact with the nerve or with its nutrient blood vessels, resulting in toxic or ischaemic injury.
• Volume-expansion effect of infiltration (sometimes inadvertent) of the anesthetic inside nerves.
• Tetanic electrical nerve stimulation to monitor neuromuscular block during or after surgery. The stimulator (one lent to me by anaesthesiologists) looks innocuous, hand-held (similar to stimulators on some EMG equipment) with two prongs to be placed on a nerve (commonly the ulnar at the wrist or elbow, on the side opposite to where the blood pressure monitor cuff is). Although AA-battery powered, it can deliver up to 80mA, 0.2ms wide stimuli at up to 200Hz, thus ensuring hand muscles undergo effective tetanic contraction that visibly fluctuates with the depth of the block. At even mid-range setting, such stimulation would be intolerable. But as the patient is anesthetised, and although stimulation trains can be delivered in various modes, stimulation can also be applied continuously (or with little interruption) for long periods. When thus applied, it can (as learned from Professor Erik Stålberg, 25 years ago) generate heat, which might be at least one mechanism of injury. Most cases resolve within few weeks; it is the severe, sometimes irreversible injuries that come to our clinics.  
• Ischemia to nerves, muscles – and other tissue – from prolonged cumulative constriction of the thigh, leg or foot, from application of inflatable garments (or boots) to prevent deep vein thrombosis. Although inflation of these garments (applied independently on the three sections of the lower limbs) is meant to be programmable – with cycles of inflation and deflation throughout the surgical procedure, operations (on the thorax, for example) may last over six hours, and inflation / deflation may fail to avoid ischemic (segmental infarction) injury to nerves – such as the sciatic or its divisions, right into the sole. Aside from uniform compression within the boot, additional constriction, on nerves or on their blood vessels, may result just under the rims of the boot if close to where nerves are superficial or over bone.
• Lithotomy stirrups: in gynecological and obstetric procedures, weight of the legs in the lithotomy position taken largely by acutely flexed knees placed on stirrups. Although stirrups (or knee crutches or loop leg supports) are padded, the fibular and tibial nerves – and their blood vessels – can still be subjected to prolonged external compression especially if the upper rim of the boot or support abuts on the head of the fibula or across the popliteal fossa. In these cases, however, I would exclude epidural first.
• Unrelated to anesthesia but sharing mechanism of injury with regional block is venepuncture, which I have described in some detail in relation to the antebrachial nerves under ‘True Neurogenic Thoracic Outlet Syndrome’ – but appending a selection here in a (large) footnote.[1]
• Although uncommon (in the severity that persists weeks after operation), these potential nerve injury mechanisms need to be considered notwithstanding the tendency to forcibly contest forcibly – for they may become litigious.

Amongst the arguments for dismissal: These procedures are used in hundreds of operations every day, so how come we do not see nerve injuries more often? The answer: While the argument is (partly) true, such injuries do occur more frequently but remain largely submerged, in the early postoperative weeks, below what the patient – and medical staff – are concerend about most of the various elements of recovery: pain, immobilisation, wound healing, etc. And if patients do point out numbness on, for example, the outer side of the foot, or inner of the hand, or inability to move the foot, they might be reassured ‘it is one of the things not unexpected after such a lengthy operation and will resolve shortly’. By and large, this forecast comes true.

Nerve injuries, however, that come to our attention two weeks or more after surgery are those (tip of the iceberg) that have exceeded injury-threshold (beyond neuropraxia, to infarction) to remain noticeable after the other postoperative concerns have receded.

Nerve injuries I am describing are what have come to my attention, and there might be more I am unaware of, but other colleagues are, and from whom I hope to learn.

For update on devices, procedures, and injuries, I defer to anesthesiologists. Nevertheless, these nerve injuries should be kept in mind amongst what (physician Arthur Conan Doyle’s) to eliminate so that what remains, however improbable, must be (or most likely is) the culprit.

• [1] On normal anatomy, the medial antebrachial enters the forearm in two branches: anterior and posterior. Around where the nerve can optimally be recorded, antidromically 7cm distal to the medial epicondyle or elbow crease (stimulated just proximal to the crease), the two branches can be 4cm apart. In most patients I can reliably record only one branch, settling for maximal amplitude (usually 2-3cm medial to the medial border of the forearm) as either representing an undivided nerve or the larger branch. But there is a caveat. The nerve usually divides in mid upper arm, such that unless stimulation is also (separately) applied to the anterior branch (more medial to the posterior and stimulation thus encroaching on the median nerve) the anterior branch cannot be recorded. This may be acceptable in routine practice, although ideally recording the two branches – and adding up their individual amplitudes – allows a more sensitive index of the total axon content of the nerve before it divides. Which means we also need to compare the two arms. While this is perhaps pedantic and beyond the routine, it is worth being mindful of in cases of selective injury to the anterior branch (with division of the parent nerve in mid upper arm). Such injury can result from difficult venepuncture for blood donation: 16G, 75mm long needle, possibly retracted and redirected a few times. While the patient may have numbness and difficult-to-describe ‘strange’ sensation over the medial 2/3 of the front of the forearm including numbness, felt at the same instant of injury, in the little finger. Although the ulnar nerve might be suspected, this is not necessarily: inclusion of the little finger likely mediated by terminal connection of the antebrachial with the palmar cutaneous branch of the ulnar – both zones within the territory of the anterior branch. (Such terminal plexus formation applies to the lateral antebrachial with the superficial terminal branch of the radial nerve, which can be confused with injury to the radial.)

  Back to the selective injury to the anterior branch of the medial antebrachial, the posterior branch may have been spared and recording it at ‘normal’ reference amplitude may be confusing – and misleading: if, on symptoms as described above, the medial antebrachial is pronounced normal based on recording its posterior branch only. This is not just pedantic; in litigious cases, it can have grave implications. 

  A case I have seen included separate selective fascicular injury to the musculocutaneous nerve. This nerve supplies articular branches to the elbow joint, and a branch to the humerus, and injury to the musculocutaneous nerve is thus likely to underlie pain at the elbow – which can be confused with pain directly related to the puncture site and can also be incorrectly considered “tennis elbow”). Nevertheless, pain of musculocutaneous nerve injury is unique and is characteristically relieved by maintaining the elbow in acute flexion.     

  I use bar electrodes (with shortened – 15mm – felt inserts) as these are easier to slide in small sideways steps, imperative for bracketing variability in the course of nerves and minimising overlap between adjacent nerves. Furthermore, although these nerves (lateral and medial) come from different trunks and have different (but adjacent) root representations, the latter may overlap. Thus, assessing these nerves is enhanced by looking at them together on both sides.