Statistics from Altmetric.com
Head drop is familiar to neurologists, but not widely appreciated by neurosurgeons. There are multiple causes of this condition1 in which the patient is unable to hold their head up because of weakness of the neck extensor musculature. It predominantly results from primary muscle pathologies in the neck extensor muscles, with occasional evidence supporting a neurogenic aetiology.1,2 I describe three patients in whom acute head drop closely followed cervical hyperflexion injury, and suggest that the cause is bilateral traction neurapraxia of one or more cervical dorsal rami.
Patient A was an 84 year old man who enjoyed excellent health prior to falling backwards, striking his occiput on a wall and sustaining forced flexion of the cervical spine. He complained of posterior cervical pain but, when seen in casualty for closure of an occipital laceration, was found to be neurologically intact. Cervical x rays showed only degenerative disease in the mid-lower cervical spine and loss of lordosis. Over 2 weeks the pain in his neck resolved, but he became aware of a difficulty holding his head up as the day progressed and, later, of aching in his neck extensor muscles. He was referred to neurosurgery as a possible case of delayed instability. Cervical x rays demonstrated 5° of forward angulation at C4/C5, which did not change with neck flexion, but were otherwise unchanged. He remained neurologically intact but, fearing progression of the angulation and development of neural injury, posterior segmental fixation at C4/5 with a Hartshill rectangle and sublaminar wiring was advised. Surgery was remarkable only for the absence of significant ligamentous injury or abnormal mobility. Unfortunately, his head ptosis recurred after 2 months. x Rays showed that the sublaminar wires at C5 had “cheese-wired” through the bone and allowed recurrence of angulation. He remained neurologically intact. After some discussion, he submitted to extended fixation from C3–C7, producing good alignment, albeit with restricted neck movements. However, he had ongoing problems with neck pain because of prominence of the metalwork due to profound atrophy of the paraspinal muscles. Three months later, he again developed head drop because of “cheese-wiring”, and the Hartshill rectangle was eroding through the skin, necessitating a third procedure to remove it. At this stage, a muscle biopsy was performed showing end-stage atrophy and fibrosis although no comment could be made as to aetiology. The patient declined further investigation or surgery and was managed in a Philadelphia collar in the long term. Despite all the above, the malalignment at C4/5 never progressed, nor did any neurological deficits develop.
Patient B was a fit 72 year old man who sustained a flexion/extension whiplash injury during a road traffic accident (RTA). In casualty, he had minor neck pain but was neurologically intact and had cervical x rays showing only minor degenerative changes and loss of lordosis. He was managed with analgesics and a rigid collar. Ten days later he returned to casualty complaining of aching in his neck and progressive difficulty in holding up his head throughout the day. Neurological examination remained normal. Cervical x rays showed angulation into 7° of flexion at C5/6, but were otherwise unchanged. He was referred to neurosurgery and at review was strikingly reminiscent of patient A. He had to hold his chin up with a hand to look ahead, had pain in the back of his neck, which developed over the day unless he used his collar, and was neurologically normal, including in the cervical dermatomes. Magnetic resonance imaging (MRI) of his neck revealed normal soft tissue anatomy. A neurological opinion confirmed the normal examination, other than head ptosis. There was no evidence of inflammatory, autoimmune, or endocrine disease clinically or biochemically, the Tensilon test was negative, and serum creatine kinase was normal. There were no features of Parkinson’s disease or amyotrophic lateral sclerosis (PLS). Electroneuromyography (EMG) studies of the neck muscles performed 3 weeks after injury were normal in the ventral muscles, but there were typical features of acute partial denervation in the neck extensors bilaterally, particularly in a band in the mid-to-low cervical spine with more normal EMGs above and below this. However, electrophysiological examination of the limbs was abnormal also and consistent with an asymptomatic peripheral neuropathy. The patient declined muscle or nerve biopsy.
In view of patient A’s course and the evidence in patient B of acute denervation that might recover, patient B was managed expectantly. Physiotherapy was used to maintain range of neck movement and encourage use of the neck extensor muscles. He was given a Philadelphia collar, which was worn by day once he became aware of head ptosis. With this regime he recovered to normal over 4 months, including recovery of the spinal alignment at C5/6, and the Philadelphia collar was withdrawn. There has been no recurrence of head ptosis.
Patient C, a 54 year old man, was similar to patient B. He suffered a whiplash injury in an RTA and developed head ptosis and angulation at C5/6 on cervical x rays 2 weeks later. Investigation and management mirrored patient B. He also had focal abnormality of his neck extensor muscle EMG, which suggests partial denervation, but otherwise was normal clinically, biochemically, and electrophysiologically. We did not suggest muscle or nerve biopsy as it was clear he would be managed conservatively. With physiotherapy and external bracing, patient C made a complete recovery in 2 months, including recovery of spinal alignment at C5/6. There was no recurrence of head ptosis at last contact.
Although there are reports of head drop in conditions predominantly affecting neural rather than muscular elements,3 Umapathi et al1 cite Braun et al,4 who treat refractory torticollis by section of multiple cervical dorsal rami without generating significant functional deficits, as evidence that focal denervation of neck extensor muscles is unlikely to cause head ptosis. This surgical denervation, however, is unilateral and the denervated muscles are also likely to be grossly abnormal because of secondary changes resulting from the underlying condition. The cat neck extensor muscle biventer cervicis (analogous to human semispinalis capitis) has tendinous inscriptions defining serially arranged compartments, each receiving segmental innervation from a cervical dorsal ramus. The muscle only generates useful tension if all compartments are co-stimulated; unstimulated compartments act as weak springs in series and dissipate tension within the muscle.5 There is some evidence for similar architecture in human neck extensors: they receive innervation from several cervical dorsal rami6 and have tendinous inscriptions producing several at least partially serial compartments.7 Denervation of one compartment bilaterally would produce significant weakness and fatigability in such compartmentalised muscles. Additionally, the deeper muscles only traverse one motion segment and are innervated by one posterior primary ramus. Segmental denervation of either type of muscle would lead to angulation at a motion segment, limited in degree by intact joints, ligaments, and disc space.
Whiplash injury can cause neurapraxia of cranial nerve XI,8 XII,9 and branches of the cervical plexus,10 and there are other reports of traction neuropathies in the neck.11,12 In the present cases, the close temporal relationship of the head drop to a forced flexion injury and the EMG findings suggesting acute denervation of neck extensor muscles are consistent with a neurogenic mechanism. Although dystonia of neck flexor muscles can produce head drop, these patients could easily lift their chins and there was no evidence of ventral muscle hypertonia on clinical examination. In addition, in patients B and C, there were normal EMG findings in the ventral neck muscles but abnormal findings in the neck extensors.
Neurapraxia of dorsal primary rami would be expected to recover in time, as happened in patients B and C. Equally, muscle tearing would recover in time, but it is inconceivable that sufficient fibres would have been torn to produce head drop without also producing soft tissue abnormalities on MRI scanning. This is not the case. Only two of the cases were investigated to exclude primary neuromuscular disorders. These were excluded in patient C. Although patient B had evidence of a pre-existing peripheral neuropathy, this may simply have made him more prone to traction neurapraxia after whiplash and his eventual recovery is consistent with the proposed mechanism.
It is unclear why this syndrome has not been described before. Perhaps most whiplash injuries produce insufficient neurapraxia to provoke head drop unless patient factors adversely affect the transmission of forces to the nerves or their susceptibility to injury. In non-predisposed individuals, sufficiently severe injuries might instead produce fractures/dislocations, whose management masks signs of a concomitant neurapraxia. Lesser injuries might produce mild head drop, which is either not recognised or recovers quickly and never requires secondary referral. Furthermore, although motor deficits may be rare after whiplash, sensory symptoms may be common, presenting as the patient’s symptoms in a case of “typical” whiplash syndrome. There is support for this notion from reports of symptomatic relief after cervical nerve blocks in cases of “third occipital nerve headache” and other post-whiplash syndromes.13,14
Finally, Umapathi et al1 suggest that cervical spinal fusion might be useful in optimising head position for patients with head drop. I would caution against this approach, given our experience with patient A, although alternative fixation methods, such as lateral mass plating, with careful attention paid to achieving bony fusion, might be appropriate on occasion if conservative measures have failed.
I would like to acknowledge the contribution of the neurophysiologists involved in these cases, particularly Dr Roger Cull, Consultant Neurophysiologist, Department of Clinical Neurosciences, Western General Hospital, Edinburgh, UK.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.