Author: John F. Kerrigan, MD, Co-Director, Hypothalamic Hamartoma Program Barrow Neurological Institute
Most conditions that cause epilepsy affect the cortex of the brain (the gray matter lining the surface of the brain), and therefore that is where most seizures originate.
Hypothalamic hamartomas (HH) are unique as a cause of epilepsy, as seizures begin in the hypothalamus (located at the base of the brain, below the cortex). As a result, the gelastic (laughing) seizures associated with HH are regarded as “sub-cortical”. HH are the exception to the rule.
There are practical consequences for evaluating HH patients (or possible HH patients) with brain imaging studies. Because HH occurs in a unique location for an epilepsy-causing abnormality, and because there are very few patients with HH, the hypothalamus may be over-looked on brain imaging as an area of concern for a patient who is newly diagnosed with seizures.
In our series of patients with HH, we would estimate that the HH was missed on the first imaging study (even though apparent on the films) in 5-10% of patients because the radiologist did not include the hypothalamus as an area where they focused their attention. (In contrast, the temporal lobes of the brain are a very common region for epilepsy, and would always be very carefully scrutinized.)
While our experience has taught us to always look at the hypothalamus in patients diagnosed with epilepsy, it is critically important for patients who are experiencing gelastic seizures. Not every patient with gelastic (laughing) seizures has an HH, but most do. From an imaging point of view, one should consider that a patient with gelastic seizures has an HH until proven otherwise.
The imaging study itself must be adequate to the task of properly evaluating the hypothalamus for the presence (or absence) of an HH. The good news is that if modern imaging studies are performed, an HH can be excluded if it is not there. Stated another way, a good imaging study will diagnose the HH if it is present.
Head CT (computed tomography) scans may detect larger HH, but are not adequate for visualizing smaller HH lesions. Probably 10-20% of our patients experienced a delay in diagnosis because CT was initially the only brain imaging study performed.
In our opinion, MRI (magnetic resonance imaging) is required for evaluating patients with HH (or possible HH). MRI has much higher resolution (that is, can see smaller abnormalities) and also has the capability to look at the brain in different projection planes (slices of the brain), making it easier for the radiologist to visualize the brain in a three-dimensional manner.
Even with MRI, the technical features are important. Ideally, studies would include coronal sequencing with attention to the region of the hypothalamus, with thin cuts and very little gap (or skip) between individual slices.
Barrow Neurological Institute recommends the following protocol for MRI imaging:
1. 3D T1 SPGR, axial 1mm isotropic voxels
2. Sag T1 – min TE; 3mm slice, 0.5mm gap; FOV 20cm
3. Sag T2(FSE) – 2mm slice no gap; FOV 20cm
4. Cor T2(FSE) – 2mm slice no gap; FOV 16cm
5. Cor T1 – 3D SPGR; 2mm slice; FOV 24cm – recon for axial
6. Axial T2(FSE) – routine brain
7. Contrast may be necessary for first evaluation to exclude other lesions. The Sag and Cor T2 are the critical sequences.
8. The MRI must include the whole brain cortical evaluation to exclude other causes for seizure.
Some other questions about HH and imaging include:
Should HH be considered tumors?: Yes. Hypothalamic hamartomas are congenital tumors (masses) and during fetal development have an abnormal growth capacity. Some are small (the size of a pea) and some are large (the size of a plum), but most are somewhere in-between (the size of a grape). However, once a patient is born, the lesions no longer grow, except in proportion to the normal growth of the brain itself. That is, the relative size of the HH to the rest of the brain will be the same for the patient’s entire life.
If HH are tumors, are they cancer?: No. We have never encountered an HH patient, or an HH patient reported in the medical literature, in which the HH turned into a more aggressive tumor. That is, HH do not expand after birth and do not spread to other locations in the body (as would be expected for a tumor that is cancerous).
Is there a need for intravenous MRI (gadolinium) contrast?: It is good medical practice to get an MRI scan with contrast when the diagnosis is established. However, HH lesions do not enhance with intravenous contrast agents. If a hypothalamic lesion enhances (“lights up”) with contrast, it is not an HH, and other tumors that can occur in this region should be considered.
Should routine follow-up MRI studies be performed once an HH is diagnosed?: We think it is a good idea to obtain at least one follow-up MRI study after the initial diagnosis is made, approximately 6 months after the original study. However, once the diagnosis is secure (and it usually should be after the first and second imaging studies) then repeated MRI studies are not required unless unexpected symptoms occur, or the clinical circumstances of the patient are otherwise not typical.
Is the location (or size) of the HH important?: Yes. There is now a large amount of research that relates the location of the HH to the symptoms that it causes. In brief, HH that are located anteriorly in the hypothalamus (near the stalk of the pituitary gland) result in precocious puberty (abnormally early puberty) and those that occur more posteriorly in the hypothalamus (near the mammillary bodies) result in epilepsy. Larger HH lesions (those that attach both anteriorly and posteriorly in the hypothalamus) can cause both.