Clinical manifestations

Presentation and course

Diagnostic criteria of HaNDL according to the new International Classification of Headache disorders (ICHD-3) are listed in Table 1.

Table 1. Diagnostic Criteria for HaNDL

Prodromal symptoms. Premonitory symptoms can occur up to 3 weeks prior to the onset of headache. Cough, rhinitis, diarrhea, and generalized malaise are prodromal symptoms present in 25% to 40% of patients (Table 3).

Headache. The majority of patients with HaNDL do not have a personal history of migraine although the headache associated with HaNDL is “migraine-like”. Head pain is typically described as severe, throbbing, or oppressive (09; 32; 48). If patients do have a history of a headache disorder, it is described as a new type of headache. Some patients have only a mild frontal pain (09; 32; 48). On rare occasions, head pain may also be absent; patients diagnosed with HaNDL have presented only with episodes of transient neurologic deficits (32). Head pain may be unilateral or bilateral, typically occurring over the site of the cerebral region involved in neurologic deficit (09; 32). The duration of pain is variable (ranging from 1 hour to 1 week), and it may be accompanied by nausea, vomiting, photophobia, or phonophobia (32).

Neurologic symptoms and signs. Neurologic deficits can precede, occur during, or occur after the headache phase (12; 09; 32). Sensory or motor deficits typically occur as a slow onset with a “march of symptoms” occurring over 5 to 15 minutes (12). Common neurologic deficits and their manifestations are listed in Table 2. The typical duration of deficits ranges from 5 minutes to 3 days (32). Only 1 patient had aphasia that lasted longer than 1 week (69).

The majority of patients have transient neurologic deficits restricted to 1 cerebral hemisphere, typically the left dominant hemisphere (32). This may be due to higher clinical eloquence of a left dominant cerebral hemisphere, whereas phenomena generated from a right nondominant cerebral hemisphere may pass unnoticed. The majority of patients with multiple episodes resulting from 1 hemisphere have stereotyped neurologic deficits; however, deficits can differ across episodes (32). A minority of patients have reported symptoms in the basilar artery territory or in multiple brain regions (always left and right hemispheres, but never basilar and carotid territories in combination).

The most common combinations of neurologic symptoms ranked in order of frequency are: (1) expressive aphasia and both right-sided sensory and motor symptoms; (2) expressive aphasia and right-sided sensory symptoms; and (3) isolated 1-sided sensory symptoms (32). States of confusion have also been described (49).

Cerebrospinal fluid. Elevation of the white blood cell count with lymphocytic pleocytosis in the cerebrospinal fluid is a cardinal diagnostic feature of HaNDL.

Table 2. Common Neurologic Deficits and Manifestation Patterns in HaNDL

Prognosis and complications

Patients with HaNDL typically demonstrate complete recovery within 12 weeks, with recovery duration ranging from 1 to 84 days (Table 3) (09; 32). The syndrome is self-limiting and recurrence has not been reported. All patients with HaNDL reported to this date recovered completely (09; 32). The only reported complications were related to the diagnostic work-up, especially cerebral angiography, which may induce HaNDL episodes (08; 09).

CSF abnormalities lag behind clinical resolution of patients’ symptoms. During an 8 patient case series with repeated lumbar punctures over the clinical course from initial presentation, patients had elevated opening pressures, CSF protein, and CSF lymphocytes even beyond 197 days after initial presentation (63). A case report of a patient who received 3 lumbar punctures in a month while symptomatic described a lag in the peak of white blood count and lymphocyte count from symptom onset (45).

Table 3. The Characteristic Features and Diagnostic Criteria of Headache Associated with Neurologic Deficits and Cerebrospinal Fluid Lymphocytosis

Clinical vignette

A 24-year-old right-handed man presented with a 2-hour episode of “inability to speak” accompanied by a severe pulsating left temporal-parietal headache, photosensitivity, and nausea and vomiting. In addition, he had right hemi-body ascending paresthesias that had gradually accrued over the course of 10 to 15 minutes, first involving the right lower extremity, then the right upper extremity, and finally the right side of his face. His exam was notable for initial expressive aphasia, with gradual improvement over the course of 1 hour to short sentences with paraphasic errors. He also had initial receptive aphasia with difficulty following simple verbal commands, which also gradually improved over the course of 1 hour. The total duration of any symptom was 4 hours. On resolution of symptoms, there was no residual language or comprehension deficit.

The patient’s history was notable for a similar self-resolving episode that had occurred 2 weeks prior to presentation. At that time, he had been given a diagnosis of “complex migraine.” His family history was negative for migraine, transient neurologic symptoms, strokes, or seizures. Other than this episode, the patient did not have a prior history of headache.

The patient underwent diagnostic testing with noncontrast head CT, MRI brain with and without contrast, MRA head and neck, routine EEG, and serum and CSF analysis.

MRI brain imaging showed subtle gyriform diffuse-weighted hyperintensity in the left parietal area on diffusion-weighted imaging without correlate on apparent diffusion coefficient imaging, which was consistent with MRI imaging that had been done when the patient had presented with similar symptoms 2 weeks prior.

Otherwise, neuroimaging did not show any evidence of acute cerebral hemorrhage, vasogenic or cytotoxic edema, or vasoconstriction.

EEG showed slowing over the left temporal lobe, but no epileptiform discharges.

Serum HIV DNA antibody and antigen and Lyme antibody was nonreactive; serum anti-NMDA receptor antibody-testing was negative. CSF viral and antibody screen for VZV, HSV 1 and 2 DNA, VDRL, West Nile IgG and IgM, and enterovirus did not detect any abnormalities. CSF pathology was notable for predominant lymphocytes showing polymorphous morphology mostly composed of small mature lymphocytes with a few plasmacytoid and plasmablastic/immunoblastic forms, thought to be reactive. Further CSF analysis is noted in Table 4.

Table 4. Clinical Vignette CSF Analysis with Symptoms and 2 Months Following Complete Resolution of All Episodes

Similar self-limiting episodes occurred 4 times over 40 days. The patient then had complete recovery with normalization of lymphocytic pleocytosis. Symptoms have not recurred since that time.

Biological basis

Etiology and pathogenesis

Immunological dysfunction theory. HaNDL has been theorized as an immunological response to a viral agent or an autoimmunity (32; 26; 14; 50; 20; 02; 17; 23). Gomez-Aranda and colleagues observed angiographic changes of “local inflammation of the arterial wall” of patients with HaNDL, which suggest that a viral infection could result in antibodies binding to antigens in the walls of the intracranial vessels (32). It is postulated that inflammation of intracranial vessels could then trigger cortical spreading depression with focal and transient neurologic symptoms (32; 26; 50).

A study of neurofilament light chain levels in 2 patients also showed elevations during the attack which then resolved in 3- and 7-month follow-up lumbar punctures (60). Neurofilament light chains are typically found in myelinated axons; damage to these axons causes neurofilament light chains into CSF and then the blood (42). Therefore, these findings suggest acute, transient axonal injury occurring in patients with HaNDL.

The theory that HaNDL results from an immunological response to a viral agent is supported by the monophasic and self-limited course of HaNDL, viral prodromes in 25% to 40% of patients, and CSF lymphocytic elevation. However, viral isolation in HaNDL has rarely been described despite exhaustive serological screening (02; 60). Only singular cases with echovirus, Epstein-Barr virus (human gammaherpesvirus-4), herpesvirus-6, herpesvirus-7, and rotavirus infections have been described (32; 20; 59; 23).

Autoimmunity. Similarities between the clinical course of HaNDL, with recurrent and self-limiting episodes of transient neurologic deficits and CSF lymphocytosis, and autoimmune encephalitis associated with ion channel antibodies have suggested the theory of HaNDL being mediated by ion channel autoimmunity (40). A study of the sera of 4 HaNDL patients and controls showed antibodies to a subunit of the T-type voltage-gated calcium channel, CACNA1H, in 2 of the 4 HaNDL cases (40). A case report described high levels of antibodies to P/Q-type voltage gated calcium channels containing α1-subunit encoded by CACNA1A in a typical patient with HaNDL who had presented with confusion and intracranial hypertension resulting in bilateral sixth nerve palsies (01). The authors postulated that the antibodies resulted in a gain of function of the channels, activating hyperpolarizing potentials and increasing susceptibility to cortical spreading depression, similar to familial hemiplegic migraine type 1 (01).

Migrainous theory. The theory of HaNDL as a migraine-variant stems from shared features of focal vasomotor changes and cortical spreading depression. However, HaNDL is a monophasic, self-limiting condition whereas migraine attacks are recurrent, making this theory less likely (50; 60).

Imaging studies in patients with HaNDL have shown evidence of focal hypoperfusion, fluctuations in arterial tone, and impairment of neuronal metabolism, which resemble findings occurring during attacks of migraine with and without aura (62; 37; 57). Brain single-photon emission computed tomography (SPECT) studies have demonstrated focal hypoperfusion in brain regions generating neurologic deficits with subsequent recovery in blood flow (13; 26). These findings resembling the cortical spreading depression of migraine aura (13; 26). The slow progression of neurologic deficits in HaNDL is also reminiscent of the phenomenon of cortical spreading depression of Leão, which is believed to initiate migrainous attacks (12; 41; 11).

Induction of clinical episodes of HaNDL by cerebral angiography supports the role of cerebral vessels and vasomotor dysfunction in the pathophysiology of the syndrome, which is also similar to focal vasomotor changes in seen migrainous attacks (08; 09; 16). Transcranial doppler findings of asymmetrical changes in blood flow velocity and pulsatility index of the middle cerebral artery in 2 patients with HaNDL, and of symmetrical middle cerebral artery velocity changes in a patient with acute confessional state and diffuse EEG changes diagnosed with HaNDL, also supports vasomotor dysfunction (37; 35).

A patient with HaNDL was also found to have electrophysiological pattern similar to those of migraine with aura subtypes (27).

Despite these similarities, the entities are distinct in clinical course. Although HaNDL self-resolves, migraine is a lifelong disease.

Epidemiology

The incidence of HaNDL is unknown because no epidemiological studies have been reported. To this date, approximately 100 cases have been reported in the literature suggesting that the syndrome is rare. Some argue, however, that HaNDL is underdiagnosed because of clinician unfamiliarity with this disorder (55). Even though this syndrome has been mainly described in the south of Europe and in the United States, there is documentation of HaNDL cases in many other countries. The first Japanese case was reported in 2003 (65).

Berg and Williams found that HaNDL is more prevalent in women than in men (1.25:1) (09). In contrast, Gomez-Aranda and colleagues reported that HaNDL is more frequent in men (3:1) (32). Reported ages have ranged from 5 to 52 years. Most patients are in the third or fourth decade of life; however, some authors have suggested that HaNDL is underdiagnosed in children (32; 50; 24). A 2019 study of pediatric HaNDL cases evaluated 23 definite cases and 7 “probable” cases and found a 4:1 female to male predominance is patients under 18 years of age (04).

Prevention

Because of the unknown etiology and the transient character of the syndrome, there are no specific prevention strategies for HaNDL.

Differential diagnosis

Confusing conditions

The diagnosis of HaNDL is made after excluding more common conditions that present with headache and transient neurologic signs and symptoms. HaNDL can be confused with migraine, particularly with hemiplegic migraine or migraine with brainstem aura, but migraine is not associated with cerebrospinal fluid lymphocytosis (39). Patients with hemiplegic migraine usually have an autosomal family history (familial hemiplegic migraine), and a search for known familial hemiplegic migraine type I mutations and polymorphisms in the CACNA1A gene in HaNDL cases did not identify mutations (15).

Patients with migraine with brainstem aura may have a similar presentation (07), but recurrent symptoms consistent with brainstem aura, a previous history of headache, and good response to antimigraine therapy will usually help differentiate migraine with brainstem aura from HaNDL. Autosomal dominant meningitic migraine with focal cerebral edema usually recurs over many years and is associated with cerebral edema, which can be easily seen on MRI of the brain and thus be distinguished from HaNDL (31). The first episode of HaNDL is one of the most frequent mimics of acute ischemic stroke, leading to the consideration of systemic thrombolytic therapy (33; 05). Normality of diffusion-weighted and apparent diffusion coefficient images in the acute phase is an important clue for diagnosis in these cases (03), although focal hypoperfusion can be seen in some cases (33) and rare DWI abnormalities have been noted in the literature (53). A case report suggests that a normal cutaneous plantar reflex (Babinski sign) in the hemiparetic side can also be a clinical diagnostic clue supporting the diagnosis of HaNDL rather than stroke (22).

Other conditions that present with headache, cerebrospinal fluid lymphocytosis, and transient neurologic symptoms and signs include viral meningitis, Mollaret meningitis, neuroborreliosis, neurosyphilis, neurobrucellosis, mycoplasma infection, neoplastic meningitis, granulomatous meningitis, autoimmune disease, and HIV infection. HaNDL-like cases related to mumps, herpes virus-6, and cytomegalovirus infection have been reported (20; 29; 68). Appropriate laboratory studies and brain imaging help exclude these conditions. Other individualized conditions, such as anti-NMDA receptor encephalitis or SMART syndrome, a similar syndrome manifesting as delayed stroke-like migraine attacks after radiation therapy for brain tumors, should be included in the differential diagnosis (10; 38; 25).

Multiple sclerosis can be confused with HaNDL, particularly when the first episode of multiple sclerosis manifests as focal neurologic deficits together with headache (47). The presence of oligoclonal bands and abnormal immunoglobulin synthesis in the cerebrospinal fluid can help with the diagnosis. Oligocloncal bands are notably absent from the cerebrospinal fluid in patients with HaNDL (32).

Occasionally, seizures and status epilepticus can present with focal signs, such as Todd paralysis, along with cerebrospinal fluid lymphocytosis (18; 06). When the seizures are witnessed, distinguishing them from HaNDL is not difficult, but in the event of unwitnessed spells, an EEG may be helpful (32).

Diagnostic workup

The initial step in evaluation should include a neurologic exam. Of note, papilledema and sixth nerve palsy can be found in HaNDL; however, signs of meningeal irritation are absent (32; 46; 50; 21).

A good quality imaging study (preferably MRI brain) to rule out space-occupying lesions or ischemia is often included in the initial diagnostic work-up of HaNDL. An abnormality observed relatively infrequently on MRI brain in patients with this syndrome is nonspecific T2 hyperintensities (09; 32), which can also be seen in patients with migraine headaches (36). Local leptomeningeal contrast uptake has also been described (28). MR diffusion-weighted imaging during temporary focal symptoms was initially thought to be normal (30), but reports have shown a complete hemispheric perfusion/diffusion mismatch with normal diffusion (56; 67). Infrequently, diffusion-weighted imaging abnormalities have been reported, as described in the real case mentioned above, and in a case report of a patient with DWI abnormalities in corpus callosum (53). A reduced venous signal in the symptomatic hemisphere was described; the authors postulated that this could indicate a decrease in metabolic demands in the affected tissue (54). A transient lesion on the splenium of the corpus callosum, identified on MRI, which resolved along with CSF lymphocytosis 4 weeks after the initial attack has also been described (58).

Other diagnostic studies may include the following:

Lumbar puncture/CSF analysis. Opening pressure can be elevated in patients with HaNDL; in these cases, patients should be evaluated for papilledema and treated accordingly (46). On CSF analysis, total nucleated cell count elevation is predominantly lymphocytic and ranges from 10 to 760 cells/mm³. Protein elevation up to 250 mg/dl is reported in 91% to 96% of cases (09; 32). In 1 study, 4 of 20 patients (20%) had an elevated IgG level, and none had oligoclonal bands on cerebrospinal fluid electrophoresis (32). When measured, 56% to 73% of patients had elevated opening pressure. Other cerebrospinal fluid studies including bacterial, viral, fungal, and immunologic studies, if performed, are normal.

EEG. When performed, EEG can be abnormal in around 70% of HaNDL patients, usually showing focal slowing corresponding to the area of brain dysfunction (32). Generalized slowing may also occasionally be seen (64). In all patients, the EEG became normal after the symptoms resolved (32).

SPECT. SPECT imaging, when performed during or right after the neurologic episode, can show focal areas of decreased radionuclide tracer uptake, coincident with the neurologic symptoms and focal EEG slowing, or bilateral reduction in brain blood flow (13; 26). During and up to 1 day after the episodes, 99mTc-HMPAO SPECT detects focal areas of decreased radionuclide uptake, corresponding with the neurologic symptoms. SPECT becomes normal when repeated more than 2 days after the transient neurologic symptoms (13; 32; 50).

CT perfusion and multimodal MRI. If stroke is suspected, the existence of a complete hemispheric perfusion/diffusion mismatch with normal diffusion imaging, together with the clinical picture, may lead to a presumed diagnosis of HaNDL (51; 56; 67; 70; 52).

Transcranial Doppler. This study may show changes in blood flow accompanied by changes in vessel pulsatility, with transient narrowing that resolves along with symptom resolution (37; 57; 56).

Cerebral angiography. When performed on patients with HaNDL, cerebral angiography was normal in all but 1 patient, whose images showed “local inflammation” (32). The value of cerebral angiography in HaNDL is not established and is, indeed, controversial. Cerebral angiography may have precipitated an event of focal neurologic deficit or migrainous episode in 2 of 3 patients who were worked-up with this modality (08; 16). Cerebral angiography should be carefully considered, and avoided if possible as the procedure has baseline risks and may also precipitate episodes. Typically, vascular imaging with CTA or MRA is sufficient in patients suspected of HaNDL.

Management

The self-limiting character of this condition, a complete resolution of all symptoms and laboratory abnormalities, and a lack of recurrence in all patients suggests that only supportive treatment for symptom management may be needed. In patients who have signs of increased intracranial pressure affecting the optic nerves, such as papilledema or double vision, acetazolamide can be given (46).

One review of HaNDL suggested empiric management of HaNDL with magnesium and nimodipine for 3 weeks for neuroprotection and vasospasm prevention; however, no clinical data has been collected to support the benefits of this treatment (66). One case report noted that a 7-day course of oral methylprednisolone reduced the opening pressure and cell count, and improved clinical symptoms in a patient with HaNDL (71).

Outcomes

Treatments have only been described in case series and case reports, and it is still unclear if treatments have a direct impact on the syndrome’s outcome. The clinical syndrome does spontaneously resolve, typically within 3 weeks, but can last up to 3 months (09; 32; 34). Recurrence of the syndrome after resolution has not been reported (09; 32).

Special considerations

Pregnancy

There are no reports of HaNDL during pregnancy. One report describes a 23-year-old woman in the early postpartum period (16 days after delivery). She experienced episodes of HaNDL during her first and second pregnancies; both episodes were associated with nursing and resolved after its discontinuation (19).

Anesthesia

No information is available on the effect of general anesthesia on patients with the transient syndrome of headache with neurologic deficits and cerebrospinal fluid lymphocytosis.