Transfalcial and transtentorial herniation and the engagement of the cerebellar tonsils into the foramen magnum are well recognised consequences of increased intracranial pressure. They were discovered through both experimental findings carried out in the 19th century,1 and pathological and clinical experiences reported in the first decades of this century.2 Harvey Cushing (1869–1939) is usually acknowledged as the first clinician who, in 1902, correlated the clinical picture of neurological disturbances and cardiorespiratory failure with the herniation of the cerebellar tonsils into the foramen magnum.3 However, when in his book on pontocerebellar angle tumours Cushing made use of the notion of “cerebellar pressure cone”,4 he was employing a term first introduced into the English language by Collier,5 as Cushing himself acknowledged in an earlier article.6

Here we report on the observations of Pierre Marie (1853–1940) that bear on the same issue and antedate those of both Cushing and Collier. On l July 1899, during a session of the Societé de Biologie, Pierre Marie7 reported on two cases of cerebral haemorrhage that resulted in secondary compression of the cerebellum. In the first case, a haemorrhage of the external segment of the lenticular nucleus flattened the superior face of the left cerebellar lobe and shifted the vermis to the right. In the second case, a thalamic haemorrhage resulted in compression of the superior face of the left cerebellar hemisphere and caused a protrusion of the cerebellar tonsils that “appeared to have engaged the occipital foramen”.

Pierre Marie drew attention to the pathophysiological research of Leonard Hill. “In his Hunterian Lectures in 1896, he showed that the brain does not transmit the pressure on one point in all directions. Indeed, a clear pressure difference is created between the main cranial cavity and the posterior fossa. According to Hill, the pressure gap is due to: (1) the viscosity of the cerebral substance; (2) the interposition of the falx cerebri and the tentorium cerebelli. This explains why a haemorrhage lesion exerts a localised compression on a particular part of the cerebrum, instead of creating a uniform pressure growth on the whole cerebral mass, as some think. In the present circumstance, the compression is exerted on a cerebellar hemisphere.

Furthermore, according to Leonard Hill

“the compression exerted by the haemorrhaged cerebral hemisphere upon the superior surface of the cerebellum can have very serious consequences. In fact, if the compression is sufficiently strong, the downward push on the cerebellum causes the engagement of its inferior surface into the foramen magnum. In this process the cerebellum takes the form of two cones constituted by the cerebellar tonsils. As a result the medulla oblongata is trapped within the foramen magnum, and if the pressure is strong enough its vessels will be compressed to the point of failing to supply the organ. Hence a severe bulbar ischemia will ensue.”⁷

Pierre Marie concluded his communication by considering the possibly severe consequences of cerebellar tonsillar penetration into the occipital foramen and the resulting bulbar hypoperfusion, exactly devising the mechanism of cerebellar herniation. In the next year, Pierre Marie reported the pathological findings of another two cases of raised intracranial pressure.8 In the first case with a thalamic haemorrhage the rise in intracranial pressure caused the caudal shift of a cerebellar tonsil, that consequently “engaged into the occipital foramen”. The second case was of increased intracranial pressure of unknown origin, disclosed by pronounced hydrocephalus. “Both cerebellar tonsils had shifted into the occipital foramen such as to form a sort of cone enclosing the bulbus for more than half of its circumference.” In his conclusions, Pierre Marie suggested that the compression of the medulla could have contributed to the fatal outcome.

A few years later, in 1905, Louis Pierre Marie Alquier (1872–1956) described two further cases of cerebellar tonsils herniation into the occipital foramen.9 In Alquier’s article the discussion of Jean Athanase Sicard (1872–1929) and Pierre Marie on the topic is reported. Mentioning his personal experience in similar cases, Pierre Marie also made the somewhat surprising suggestion that Alquier’s pathological findings might be attributed to a postmortem process. We could find no further contribution by Pierre Marie on this topic in his subsequent scientific production.

Recently, Fisher10 has pointed out that cerebellar herniation is not necessarily a terminal event, except, in cases in which cerebellar herniation occurs in posterior fossa infarcts, acute subdural haematomas, or during a lumbar puncture. In such cases, acute herniation is justifiably considered among the determinant factors of clinical outcome. The first clinical description of this important phenomenon was made by Pierre Marie.