Cephalalgic Alopecia Areata

Case report

A 34-year-old woman presented with a 30-month history of recurrent episodes of occipital and parietal-temporal jabbing, burning head pain which was preceded by a febrile illness associated with localized pruritus on the neck, a low-grade diffuse headache and extreme fatigue. After a week, the fever and headache resolved but malaise persisted. Two weeks later, the patient began to have a cramping, squeezing pain in the muscles of her neck and upper back and a burning sensation that spread over the occipital region of her head. Superimposed upon the burning pain were intermittent volleys of severe jabbing occipital and temporal pain that lasted several minutes. Within several hours of the episode of jabbing scalp pain, the patient began to experience hair loss at the site of the pain. Even after resolution of the pain, the hair loss continued for days. At first, the episodes of pain and hair loss were separated by days to weeks and were infrequent enough to allow hair regrowth. However, subsequent episodes of jabbing pain resulted in more persistent loss of hair. Small patches of alopecia were observed. Increasing frequency of attacks resulted in large areas of the scalp that were almost completely devoid of hair (Fig. 1). On presentation, the burning scalp dysaesthesia occurred daily; however, the jabbing pain was intermittent and not diurnal. Allodynia was present in area of affected scalp but no vasomotor changes were apparent. The patient adamantly denied pulling or in any way manipulating her hair. There was no family history or prior personal history of alopecia areata. The patient had not observed loss of eyebrow, lash or any other body hair.

Figure 1

Intensive patchy hair thinning and loss characteristic of alopecia areata.

Investigations, including magnetic resonance imaging/angiography of brain, showed that the lateral ventricles were near the upper limit of normal for a patient of her age but were otherwise unremarkable. Biopsy of the affected scalp showed lymphocytic infiltration around the hair bulb characteristic of alopecia areata (Fig. 2). Thyroid-stimulating hormone, free T4, α-androstanediol and glutamate levels were normal. Blood testing for inflammatory markers including erythrocyte sedimentation rate (ESR), rheumatoid factor (RF), antinuclear antibody (ANA) and antibodies to extractable nuclear antigens (ENA) was negative.

Figure 2

Peribulbar and external root sheath infiltration by autoreactive T lymphocytes with resultant hair shaft diminution and loss (H&E, ×100).

Four months after presentation, the frequency of attacks began to increase. From her local dermatologist, the patient received two series of triamcinolone acetonide injections (4 and 6 mg) into the scalp lesions combined with 40 mg intramuscularly, without improvement. Over the course of 5 years, the patient had tried numerous treatments for her scalp pain and hair loss, including the maximally tolerated daily doses of buspirone (55 mg), propranolol (180 mg), and tizanidine (16 mg), indomethacin (225 mg), nortriptyline (50 mg), amitriptyline (50 mg) and gabapentin (2700 mg), all without sustained benefit. Oral opiates including propoxyphene, codeine and hydrocodone provided only transient, incomplete relief. In an attempt to reverse hair loss, the patient also received midodrine 5 mg three times per day in an attempt to decrease access to the hair follicles by circulating autoantibodies. There was no response to this manoeuvre. She was refractory to all attempted therapies and was increasingly disabled by her jabbing pain which became daily. In addition, the patient began to have worsening myalgias and fatigue.

Treatment with botulinum A toxin (BTXN A) was then attempted. BTXN A was injected into procerus, corrugator, frontalis, temporalis, splenius capitus, occipitalis and trapezius muscles (100 units total) with special care taken to inject sites most frequently involved by the neuralgiform pain. BTXN A treatment resulted in a marked reduction in the frequency and intensity of the jabbing head and neck pain, which began approximately 10 days after injection and lasted 6 weeks. The patient reported partial hair regrowth during this remission. However, no improvement was seen in the myalgias and fatigue and she began to experience muscle spasm in the intrascapular and thoracic areas of her back. Three months later, she received a second 100-unit BTXN treatment to the head and neck, as well as 192.5 units of BTXN injected into the paraspinal muscles. This second treatment resulted in complete remission of the back pain for 45 days and the head pain for 60 days. During this period she had significant hair regrowth (see Fig. 3). By the eighth week she began to experience recurrence of the neck and head pain and loss of hair. A third BTXN treatment has resulted in a similar remission.

Figure 3

Marked hair regrowth following botulinum A toxin injection (4 weeks after treatment).

Discussion

With the exception of a single case briefly mentioned in the Journal of Nervous and Mental Disease in 1894 (1), to our knowledge this is the only reported association between recurrent episodes of the head pain and colocalized hair loss from scalp. Unlike our case, facial flushing occurred with the head pain in the 1894 case. In addition, myalgias and muscle spasm in the neck and back eventually developed in the current case, while similar symptoms are not mentioned in the 1894 patient. In our patient, neuralgiform pain did not show a sustained response to gabapentin, amitriptyline or nortriptyline, medications that are often effective in the treatment of occipital and other scalp-based neuralgias. The close association of alopecia areata with cephalalgia in our case suggests that the pain involves the primary afferent neuron, as it is difficult to explain direct effects on the hair follicle by activation restricted to the CNS. Once initiated, however, prolongation of the pain and its evolution from intermittent to chronic probably involves sensitization at second and perhaps higher order neurons. It may be that during the early weeks of evolution, the constant input from the primary afferent nociceptors exceeded the modest central effects of tricyclic antidepressants and gabapentin. Lack of response to steroidal and non-steroidal anti-inflammatory drugs argues against focal peripheral inflammation as the sole basis for the pain. Systemic markers of inflammation (elevated ESR, RF, ANA and ENA) were also conspicuously absent.

A biopsy of the affected scalp revealed lymphocytic peribulbar inflammation consistent with alopecia areata, raising the possibility that the case was simply an atypical presentation of alopecia areata. The biopsy showed no features of psychogenic hair pulling (trichotillomania). Alopecia areata is not accompanied by pain, although paraesthesias and pruritus may occasionally be experienced. The diffuse myalgias and muscle spasms in the back that developed in the patient are also not suggestive of alopecia areata. Even if one assumes that this patient has a variant of alopecia areata, the relationship between the pain and recurrent hair loss still warrants explanation and may provide important information about neural regulation of hair follicles.

Alopecia areata is considered an autoimmune disease mediated by a reversible, tissue-restricted immune reaction involving T lymphocytes that infiltrate and surround the hair bulb and root sheath (2). Hair loss within the affected area can be fairly rapid and profound because all hair bulb follicles are affected (3).

In this particular case, it seems likely that hair loss is initiated and sustained by the recurrent activations of the trigeminal and upper cervical branches innervating hair follicles (Fig. 4). Several factors support this possibility: (i) the patient did not develop alopecia until after the appearance of the neuralgiform pain; (ii) early in the evolution of the syndrome, partial hair regrowth occurred during spontaneous remissions, during which the patient was free of the neuralgiform pain; (iii) treatment with BTXN that caused a remission of the neuralgiform pain resulted in hair regrowth; (iv) at the end of remission, recurrence of the neuralgiform pain resulted in loss of regrown hair.

Figure 4

Neuron projections and innervation of hair follicle. Botulinum A toxin (BTXN) transport and localization in neurons. Neural and hair follicle pathology remedied by BTXN injections.

The anatomy and physiological interactions between nociceptive neurons and the hair follicle are complex. Hair follicles are innervated by neuropeptide-containing, unmyelinated neural plexuses (4) with varicosities located in close proximity to the likely site of hair follicle stem cells (5). Nociceptive unmyelinated C-fibres known to have terminals in skin contain neuropeptides including substance P (SP) and calcitonin gene-related peptide (CGRP), which when released by capsaicin injection are associated with a neurogenic inflammatory response (6). SP release from C and A δ fibres is associated with mast cell degranulation both in skin (7) and dural vessels (8). In the skin, SP receptors have been identified in the walls of perifollicular vessels (9) and SP is also involved in the induction and regulation of hair growth (7). In addition to its vasodilatory actions, CGRP inhibits antigen presentation by Langerhans cells and other dendritic cells (10) and regulates T cell proliferation (11) when released from perifollicular nerves within the epidermis. Disruption of the physiological, tonic release of neuropeptides from perivascular neurons may therefore result not only in derangement of trophic vasomotor regulation but also in altered peribulbar antigen presentation and inhibition of further hair growth.

Response to botulinum toxin

Perhaps most intriguing is the fact that injection of BTXN caused both remission of neuralgiform pain and subsequent hair regrowth. There is no known direct effect of BTXN on the hair follicle; thus, it is likely that the regrowth was related to abolition of the recurrent activation of primary afferent nociceptive neurons. There is evidence to suggest that BTXN A is taken up into the primary afferent neurons innervating the muscles into which it is injected and undergoes retrograde transport within the neuron (12) to exert its effects at the interface of the central and the peripheral nervous systems. In animal studies, injection of BTXN (3 ng/kg) into abducens muscle resulted in a marked decrease in synaptic transmission within abducens nuclei from vestibular afferent inputs to abducens motoneurons within 2 days (13). Between 7 and 15 days post injection, both excitatory and inhibitory postsynaptic potentials were abolished. This was followed by profound ultrastructural changes which included synaptic stripping with a decrease in the number of synaptic boutons as well as widening of the subsynaptic interspace (13). It remains to be determined if transsynaptic effects on the postjunctional fibre are exerted by BTXN A. However, there are preliminary data to suggest that this might occur (14, 15).

Possible mechanism of BTXN

One possible scenario for the evolution of events observed in this case is summarized as follows. The asymptomatic patient harbours an immunogenetic predisposition for alopecia areata and sustains an insult to trigeminocervical C and A δ fibres related to the febrile, possibly viral, illness which activates recurrent episodes of aberrant firing of nociceptive neurons innervating the scalp and neck. When occurring infrequently, activation of these fibres causes the release of neuropeptides (including SP and CGRP) from perifollicular nerve terminals with local activation of mast cells. Because the firing is relatively infrequent, the peripheral terminals are able to undergo repletion of the neuropeptides and there is no effect on hair growth. As the frequency and duration of the episodes of C and A δ fibre activation increase, SP and CGRP are depleted in the perifollicular terminals. The tonic suppression of antigen presentation and immune reactivity by CGRP and the induction of hair growth by SP are lost as the perifollicular terminals are unable to recover neuropeptide concentrations between depolarizations. Antigen presentation is de-repressed and activated local immune cells recognize and target the hair bulb and root sheath. Induction of new hair growth also markedly decreases or ceases completely in the affected areas. Hair loss becomes clinically apparent.

The patient receives BTXN injection into muscles innervated by branches of the affected neurons. BTXN is taken into the nerve terminal and transported proximally, where it disables the synapse between the primary afferent and the second-order neuron within the trigeminal nucleus caudalis and the dorsal horn of upper cervical spine. Whether the blockade of firing at this synapse is based on trans-synaptic migration of BTXN A to affect directly the second-order neuron or is solely a prejunctional effect on the primary afferent neuron is unknown. Aberrant neural activation is halted. The perifollicular peripheral terminals replete their neuropeptides and the tonic release of SP and CGRP resumes. Hair regrowth resumes and continues as long as C and A δ fibre activations are suppressed. After a period of weeks, disabled synapses recover and the injured neurons resume frequent aberrant firing. The perifollicular terminals are again depleted and pain-associated hair loss recurs.

Though this is a rare presentation of alopecia areata closely associated with neuralgiform pain, the marked hair regrowth and pain relief observed following BTXN injections strongly implicate local neural factors in mediating immunological activity of alopecia areata. BTXN may also improve alopecia areata not overtly linked to cephalalgia by similar mechanisms. These observations may also provide a mechanism by which BTXN targets first- and second-order neurons in cephalalgia and other BTXN-responsive headaches.