Acta Neurologica Belgica

, Volume 112, Issue 1, pp 67–69

Diffusion restriction in the splenium of the corpus callosum in a patient with the syndrome of transient headache with neurological deficits and CSF lymphocytosis (HaNDL): a challenge to the diagnostic criteria?


    • Department of NeurologyZiekenhuis Oost-Limburg
Case Report

DOI: 10.1007/s13760-012-0018-0

Cite this article as:
Raets, I. Acta Neurol Belg (2012) 112: 67. doi:10.1007/s13760-012-0018-0


The diagnosis of transient headache with neurological deficits and CSF lymphocytosis (HaNDL) is essentially based on the clinical and CSF findings, and the absence of MRI abnormalities. We present a young man with the clinical characteristics of HaNDL, but also an area of diffusion restriction in the splenium of the corpus callosum (SCC). When considering the limited experience with this MRI technique in this disorder, we wonder if a normal MRI can be maintained as an indispensable criterion for diagnosis. Similar MRI abnormalities limited to the SCC have been found in mild forms of meningoencephalitis, but their origin remains obscure. In at least some cases not only a clinical, but also a radiological overlap could exist between both disorders.


HeadacheMigraineCorpus callosumCerebrospinal fluidLymphocytic meningitisDiffusion restriction


Transient headache with neurological deficits and cerebrospinal fluid (CSF) lymphocytosis (HaNDL) bears some clinical resemblance with migraine with aura, but abnormal lumbar puncture (LP) and normal brain MRI are considered essential for diagnosis [1, 2]. We present a patient with HaNDL and transient diffusion restriction (DR) in the splenium of the corpus callosum (SCC) on MRI. This finding questions the diagnostic criteria for HaNDL and the relationship between HaNDL and other disorders involving DR in the SCC.

Case report

A 32-year old man, with a blank medical history, had a manipulation of the neck for chronic neck pain. After 12 h, he woke up with unilateral ascending numbness, hemiparesis and dysarthria, lasting some 15 min. A few hours later he noticed a throbbing contralateral headache, nausea, vomiting and photophobia. On admission, neurologic examination and body temperature were normal. MRI demonstrated an area of DR in the SCC, known as the boomerang sign [3] (Fig. 1). Four vessel angiography, blood tests including screening for thrombophilia and autoimmune disease, and EEG were normal. CSF analysis demonstrated 123 leukocytes/ml (94% lymphocytes) and elevated protein (80,00 mg/L). Oligoclonal banding, antibody titers and PCR for CMV, and cultures of the CSF were negative. Intravenous paracetamol induced a quick relief of the headache. On days 2 and 3, similar attacks recurred, again with complete relief after symptomatic therapy. A second LP after 1 week still demonstrated 100 leukocytes/ml (95% lymphocytes). MRI control 5 weeks later was normal, without infarction in the SCC. Clinical evaluation was repeated after 2 months and after 2 years, but the patient remained symptom free.
Fig. 1

DWI with hyperintensity in the splenium of the corpus callosum


Headache with neurological deficits and CSF lymphocytosis is an uncommon benign disorder of unknown etiology, of which the diagnostic criteria have been published by the International Headache Society (IHS) in ICHD-II ( Gómez-Aranda et al. [1] described 50 patients with HaNDL in 1997, the most extensive series until today. Patients present with sensory and motor deficits or dysphasia, lasting from a few minutes to several days. Their headache is mostly throbbing, unilateral or bilateral, with nausea, vomiting, photophobia and sonophobia, and can last for 1 h up to 1 week. Fever may be present, but no meningeal signs. Headache may recur, even without focal deficit, for up to 4 months, but HaNDL is essentially self-limiting. CSF lymphocytic pleocytosis is essential for diagnosis. Most patients have increased protein, but no hypoglycorrhachia or oligoclonal bands. Unlike migraine male gender, prolonged aura, and other than visual aura are more prevalent in HaNDL; more than 15 leukocytes/ml is never found in CSF of migraineurs [1].

Although EEG and brain SPECT can show transient abnormalities, normal MRI is considered essential for diagnosis ([1, 2] Nevertheless, only 18 of 50 patients in the series of Gómez-Aranda et al. [1] had brain MRI. Two of them showed irrelevant T2 abnormalities, but the authors do not mention DWI, which was not yet widely used at that time. More recently, two patients with normal MRI were published, but without information about the lapse of time before imaging [2]. DWI during the aura in one patient with HaNDL did not show any abnormality [4]. Yilmaz et al. [5] found local gray matter swelling and sulcal enhancement on FLAIR and hypoperfusion on PWI in one patient, but no DWI abnormalities; therefore, the data about DWI in HaNDL are scarce.

Different neurologic conditions have been linked to DWI hyperintensities in the SCC: hypoperfusion, hypoglycemia [3], hyponatremia, hypernatremia, renal failure, malnutrition, epilepsy, trauma, mountain sickness, hypertension and preeclampsia, cancer, radiotherapy, chemotherapy, alcohol, infections [6], MS, posterior reversible encephalopathy syndrome, lymphoma, Marchiafava-Bignami disease and extrapontine myelinolysis [7]. Yet abnormalities were most often found also outside the SCC [6]. T2 hyperintensities on MRI, limited to the SCC, were present in 15 Japanese patients with mild encephalitis or encephalopathy, with a proof of a viral origin in 5 of them. DR was present in all seven patients examined with DWI [7]. Infection has been incriminated as a possible cause of HaNDL [1]. We cannot discard a mild meningoencephalitis in our patient, since a viral infection can be hard to proof, but the normalizing between the individual attacks is more in favor of HaNDL.

Those localized DWI abnormalities in the SCC are hard to explain. Their reversibility and the lack of DR elsewhere in the vertebrobasilar territory argue against cytotoxic edema and ischemia. Tada et al. discuss the possibility of a specific affinity of viral antigens or antibodies for receptors on axons or myelin sheaths in the SCC, causing intramyelinic edema or local inflammation, but they admit this is pure speculation [7].

Today the position of HaNDL in the spectrum of headache disorders and infections of the CNS remains unclear. When considering the limited number of MRI and in particular DWI studies that have been performed in this disorder, and the clinical overlap with mild forms of meningoencephalitis, this case suggests that the importance of MRI and DWI in the diagnosis of HaNDL should be reconsidered.

Conflict of interest


Copyright information

© Belgian Neurological Society 2012