Marchiafava-Bignami Disease

Updated: Jun 27, 2017
  • Author: Cortney Lyford, MD; Chief Editor: Tarakad S Ramachandran, MBBS, MBA, MPH, FAAN, FACP, FAHA, FRCP, FRCPC, FRS, LRCP, MRCP, MRCS  more...
  • Print


Marchiafava-Bignami disease (MBD) is a rare condition characterized by demyelination of the corpus callosum. It is seen most often in patients with chronic alcoholism. (See Etiology and Pathophysiology.)

In 1903, Italian pathologists Marchiafava and Bignami described 3 alcoholic men who died after having seizures and coma. In each patient, the middle two thirds of the corpus callosum was found to be severely necrotic. Through the years, the medical literature accumulated hundreds of cases of MBD. [1] Most of these cases were found in alcoholic men.

With the advent of computed tomography (CT) scanning and magnetic resonance imaging (MRI), more cases of MBD have been recognized than before. Analyses of such cases have revealed several patterns, including scattered lesions or cysts observed at intervals from the front to the back of the callosum. Nearby areas (eg, anterior commissure, posterior commissure, brachium pontis, other white-matter tracts) and the centrum semiovale are frequently involved. (See Workup.)

Subtypes of MBD

In 2004, Heinrich et al described 2 clinical subtypes of MBD as follows, based on a review of 50 radiologic cases diagnosed in vivo [2] :

  • Type A - Has predominant features of coma and stupor; this subtype is associated with a high prevalence of pyramidal-tract symptoms; radiologic features include involvement of the entire corpus callosum
  • Type B - Characterized by normal or mildly impaired mental status; radiologic features are partial or focal callosal lesions (see the image below).
    T2-weighted axial image in a patient with Marchiaf T2-weighted axial image in a patient with Marchiafava-Bignami disease showing a high-signal lesion in the corpus callosum.

Etiology and Pathophysiology

It has long been accepted that the etiology of Marchiafava-Bignami disease (MBD) is likely either toxic or nutritional, as it is seen predominantly in malnourished alcoholics. After a literature review of 100 studies in 2017, Fernandes et al. [3] suggested a synergism between ethanol-induced neurotoxic effects and hypovitaminosis B, particularly B1. While alcoholism and poor nutrition remain the greatest risk factors for MBD, there have been a few cases reported in individuals who were not malnourished and did not drink alcohol. For example, several case reports of MBD have been reported in patients with dramatic fluctuations in serum glucose in the setting of poorly controlled diabetes. [4, 5, 6] This suggests that abrupt changes in serum osmolality may lead to myelinolysis of the corpus callosum, similar to the mechanism implied in central pontine myelinolysis. In addition to alcoholism, malnutrition, and wide fluctuations in serum glucose, Jorge et al. describe a case of MBD in a young trauma patient in 2015. [7]

Although the callosal lesions are the hallmark of the disease, for years some cases of MBD were known to be associated with cortical damage in addition to damage to the white matter tracts of the corpus callosum. Generally, the cortical damage was in the lateral frontal and the temporal lobes, mainly in the third (although sometimes also in the fourth) cortical layer. In these areas, the neurons degenerated and were replaced by glial cells. In 1939, Morel described this as cortical laminar sclerosis (now known as Morel cortical laminar sclerosis). [8]

Although Morel did not report an association between cortical laminar sclerosis and MBD, many subsequent authors did, including Jequier and Wildi in 1956 [9] and Delay et al in 1959. [10, 11] Indeed, Ropper et al stated in 2005, [12] in Adams and Victor's Principles of Neurology, that Jequier and Adams (in an otherwise unpublished review) reexamined Morel's slides and found evidence of MBD in all of those cases. Thus, the prevailing view has generally been that Morel cortical laminar sclerosis is secondary to MBD.

Nevertheless, in 1978, Naeije et al reported a case of Morel cortical laminar sclerosis in an alcoholic woman who did not have MBD. [13] In addition, Okeda et al reported 3 cases of cortical laminar sclerosis in 1986 in patients who had various combinations of pontine and extrapontine myelinolysis but who did not have MBD. [14] One of these patients had alcoholic cirrhosis and 2 had malignancies.



Although this disease occurs in both sexes, most cases are found in men. Most cases of MBD occur in persons older than 45 years.

Alcohol abuse is such a common problem that underdiagnosis of MBD seems likely (although now, with the availability MRI, fewer cases are going undiagnosed). In addition, many cases of MBD may be diagnosed but not reported, and autopsies are largely not performed. Hence, the disease may be more common than thought, and the overall outcome may be better than previously believed.

Occurrence in the United States

MBD is a very rare condition. In 2001, Helenius et al wrote that they had found approximately 250 cases in published reports, although they also suggested that many cases had gone undiagnosed. [15]

The authors of this article have estimated that approximately 300 cases of MBD turned up in published reports between 1966 and November 2008. Another 40 or 50 cases have been mentioned in textbooks that are too old to have been included in the author's PubMed search.

International occurrence

International cases of MBD are similar to US cases, but 1 additional detail deserves mention. Some of the old literature on MBD suggested that this condition was more common in Italians. This was solely an artifact of the initial cases having been found in Italy and the fact that, at first, Italian physicians were apparently the only investigators interested in finding such cases. MBD has since been found in persons from all over the world.

It is now firmly believed that no national, geographic, ethnic, or racial predilection is known for MBD. However, with such few reports, the numbers of cases reported from each country could not be expected to be exactly in proportion to the population size of each country. In 2006, Staszewski et al described the first case in Poland, which was detected by MRI. [16]



In the era before CT scanning, MBD was found almost exclusively at autopsy. Patients with the condition usually died from the effects of alcoholism and typically had severe neuropsychological deficits before death. Helenius et al reported in 2004 that among approximately 250 known patients with MBD, 200 died, 30 remained severely demented or bedridden, and only 20 had a favorable outcome. If the underlying cause of MBD is alcoholism, the prognosis is poor unless the patient adheres to an alcohol treatment program.

However, modern CT scanning and MRI have allowed the detection of mild cases of the disease, and some patients have recovered with minimal deficits. Moreover, data suggest an improved overall prognosis for MBD.

The prognosis for MBD is correlated with the subtype, as follows:

  • Type A - Has a long-term disability rate of 86% and a mortality rate of 21%
  • Type B - Has a long-term disability rate of 19% and a mortality rate of 0%

Radiologic findings

In a 2004 review of acute and chronic cases of MBD, Heinrich et al separated most cases into 2 groups. Group A included the worst cases, in which patients presented with coma or other severe impairment of consciousness. On MRI scans, their lesions typically involved most or all of the corpus callosum. For example, in the acute phase, the entire corpus callosum was commonly hyperintense on T2-weighted MRI scans. As the lesions evolved, considerable necrosis occurred, and cystic areas of necrosis were present in most or many regions of the corpus callosum. The death rate for patients with such presentations was high (21%), and those who lived frequently had severe deficits.

In group B, patients had little or no impairment of consciousness. Their deficits were subtle and included various cognitive difficulties and signs of impaired interhemispheric information transfer, gait disturbances, dysarthria, limb hypotonia, and rare seizures or upper motor neuron signs. Initial hyperintense lesions on T2-weighted MRI scans were limited to a few areas of the corpus callosum. Some cystic necrotic areas developed over time, but they were fewer and smaller than those in type A. No deaths occurred in this group, and patients frequently had good recoveries.

The authors did not attempt to correlate the severity of the cases with the presumed causes. Patients with the most severe alcoholism might have been in group A, but this is speculation. In both groups, the amount of early callosal edema in the acute phase often markedly exceeded the areas of ultimate cystic necrosis.

In 2006, Menegon et al reported 6 patients with MBD in whom (1) the entire corpus callosum appeared to be affected by a reduced apparent diffusion coefficient, as seen on diffusion-weighted imaging studies, and (2) lateral and frontal cortical lesions were also detected by diffusion-weighted imaging. Menegon et al suggested, on the basis of the outcomes of their patients, that such a combination of findings was a harbinger of poor outcome for cognitive recovery and for survival. [17]

However, as pointed out by Khaw et al in 2006, [18] the older literature, such as that by Brion, from 1977, [19] does not support a correlation between laminar sclerosis and bad outcome. In addition, studies such as that by Hlaihel et al from 2006 [20] do not support a correlation between reduced apparent diffusion coefficient and poor prognosis or even with irreversibility of the lesion.

Finally, they noted that cortical MRI findings have not been definitively correlated with the specific pathology of Morel cortical laminar sclerosis. However, if indeed they represent laminar sclerosis, the fact that this is present in the acute or subacute stages of MBD may force a reevaluation of the thought that laminar sclerosis is a secondary consequence of the MBD.