Pathophysiology

Three types of brainstem hemorrhage are typically recognized: (1) nontraumatic spontaneous hemorrhages confined within the brainstem parenchyma, (2) spontaneous hemorrhages that spread superiorly into the midbrain or thalamus or that rupture into the ventricular system, and (3) petechial brainstem hemorrhages that arise from trauma due to supratentorial mass effect and subsequent transtentorial cerebral herniation. The pons is by far the most common location within the brainstem to be affected by intraparenchymal hemorrhage (IPH), regardless of where the hemorrhage originates. ¹

Brainstem hemorrhage is a relatively rare condition characterized nearly exclusively by primary pontine hemorrhage (PPH) and thalamic hemorrhage (TH). Clinicopathologic investigations of cases of spontaneous IPH have largely focused on the much more common supratentorial forms of the disease. Although much of the accepted understanding of the mechanisms leading to supratentorial IPH can reasonably be applied to PPH and TH, each type of hemorrhage has a distinct system of blood supply with different cell populations that may react differently to IPH. It is therefore essential to continue efforts to develop specific animal models of PPH and TH. ^{2 , 3 , 4}

Long-term hypertension results in pathologic change within the arterial walls of vulnerable cerebral vasculature and is thus a major risk factor for IPH. This general relationship has been posited for more than a century, but the exact mechanism by which rupture occurs remains a point of contention. In 1868, Charcot and Bouchard related IPH to the rupture of small pseudoaneurysms that form preferentially within the walls of the small arteries most often implicated in large spontaneous intracranial hemorrhages. ² Recent ultrastructural investigation has demonstrated that Charcot-Bouchard’s proposed mechanism is responsible for only a few IPH cases. Instead, most IPHs arise from arterial bifurcation points weakened by arteriosclerosis. ⁵

Takebayashi and Kaneko ⁵ were the first to report on an examination of these rupture points with electron microscopy. Lenticulostriate arteries were removed from patients who underwent surgery within 4 hours of an IPH, and for comparison, bilateral lenticulostriates and middle meningeal arteries were removed at autopsy from patients who had died. The small temporal cortical branches assessed as controls had normal smooth muscle cells within the media. However, the middle and distal portions of involved lenticulostriate arteries were severely atrophied or had complete obliteration of medial smooth muscle cells at rupture points, which occurred at or near vessel bifurcations. All the rupture sites demonstrated luminal dilatation, intimal thickening, and atrophied medial smooth muscle cells with a moth-eaten appearance due to increased amounts of cellular debris within the intercellular matrix.

These small intracerebral vessels undergo degenerative changes largely due to chronic hypertension, which accelerates lipid infiltration, and the processes that reduce compliance in the vessel wall, such as arteriosclerosis with lipohyalinosis and fibrinoid necrosis. These degenerative changes leave vessels vulnerable to rupture from spontaneous fluctuations in blood pressure and seem to most severely affect the lenticulostriate arteries of the middle and anterior cerebral arteries. However, the Takebayashi and Kaneko ⁵ study did demonstrate similar findings in four cases of PPH and TH.

Takebayashi and Kaneko ⁵ concluded that degeneration of medial smooth muscle cells at the branch points of small penetrating arteries is the primary mechanism resulting in IPH. Of the 61 cases examined, 2 cases involved ruptures of small segmental dilatations consistent with Charcot-Bouchard aneurysms. However, it was ultimately concluded that, although these aneurysms are a real phenomenon, they do not represent the main mechanism driving arterial rupture in patients with IPH.

The Takebayahsi and Kaneko ⁵ study focused mainly on ruptured lenticulostriate arteries causing supratentorial IPH treated by surgical decompression and resection. Although some of the mechanisms underpinning this phenomenon might be expected to also apply to PPH and TH, it would be inappropriate to assume that these pathways are identical. The development of a brain-stem hemorrhage model with which to study the same processes in these distinct anatomical locations has proven challenging. ^{4 , 5}

One autopsy study of 18 patients with brainstem hemorrhage created a classification based on anatomical location ( Fig. 25.1 ⁶ ). Most PPH cases (15 of 18) were found to originate in the region of the medial border between the basilar and tegmental parts of the middle level of the pons. This bleeding pattern was termed tegmentobasilar, as evident on magnetic resonance imaging ( Fig. 25.2 ), and it was further subdivided into “upper tegmentobasilar” and “lower tegmentobasilar,” with nearly equal incidence. Few patients (3 of 18) demonstrated a bleeding origin in the tegmentum at the middle level of the pons, termed tegmental ( Fig. 25.3 ).

Tegmentobasilar pontine hemorrhages were found to spread in predictable patterns, namely via contralateral extension, ventrodorsal extension, and upward extension. ⁶ In 15 of 18 patients, the bleeding extended contralaterally to form a PPH with bilateral involvement. Ventrodorsal extension also occurred in 15 patients, and 12 of these hemorrhages broke into the fourth ventricle. In 13 patients, upward extension was noted, with 11 of these hemorrhages penetrating into the midbrain and 2 reaching as far as the thalamus. Among all 18 patients, 5 had hemorrhages that were confined to the pons. Direct extension into the medulla was not noted in any bleeding pattern. The cases of tegmentobasilar PPH were concluded to be arterial in nature because of the characteristic rapid onset and large volume (< 40 mL) of hemorrhage. The tegmental cases were believed to represent capillary or venous bleeding because of the small volume of hemorrhage (< 10 mL) and their typical unilateral confinement within the pons.

Regardless of subtype, the hemorrhage initially forms when a single vessel ruptures, subsequently increasing tension and inducing rupture in adjacent arteries, rapidly giving rise to a larger hemorrhage involving multiple rupture points in an “avalanche” fashion. ⁵ Histologic sections of brain affected by IPH are characterized by the presence of edema, neuronal damage, macrophages, and neutrophil infiltration of surrounding tissue. ² Hemorrhage is found to spread between planes of intact white matter embedded in the hematoma.

The disruption of the blood-brain barrier and the process of neuronal death result in vasogenic and cytotoxic edema that usually persists for 5 to 14 days. ⁷ Although mechanical compression of surrounding tissue secondary to hematoma was once widely thought to be the primary factor in both initial and secondary ischemic injuries after IPH, blood and plasma by-products of neuronal death are now believed to drive secondary injury to neural tissue. ²

Incidence and Epidemiology

Spontaneous cases of IPH occur at a rate of about 10 to 20 cases per 100,000 persons worldwide. ² This rate represents spontaneous bleeding in all sites throughout the brain. A 30-year review of all IPH-related hospital admissions in the United States identified 1,545,000 cases of IPH, representing 0.15% of all hospitalizations. ⁸ A sharp increase in IPH-related admissions between 1979 and 1988 was attributed to the advent of computed tomography for neurologic emergencies and the 1979 introduction of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Since this initial increase, the rate of admission for IPH has essentially remained stable as of 2008, the last year included in the study.

Within this larger group, PPH is a relative rarity, representing a yearly incidence of 1 case per 50,000 people, accounting for nearly 140,000 cases per year worldwide. Although PPH accounts for only about 7 to 10% of the 2 million annual cases of IPH worldwide, it is characterized by the highest rates of mortality and morbidity of any stroke subtype. ^{2 , 3 , 4 , 9 , 10 , 11 , 12 , 13 , 14} TH is a more common phenomenon, accounting for up to 30% of all cases of IPH, and it is also characterized by high morbidity and mortality. ^{15 , 16 , 17}

Patients with chronic hypertension have a fourfold risk of IPH. ¹⁸ As might be expected, in light of the association of IPH with chronic hypertension, its incidence also increases as access to regular medical care decreases, and it is thus more common throughout impoverished and poorly educated populations. ² However, the rate of IPH is probably much higher in at-risk groups such as men, persons older than 55 years, and certain ethnic groups. ^{2 , 3 , 4 , 9 , 10 , 11}

Several groups, such as Japanese and African American populations, have a disproportionately higher rate of PPH and TH. A study on the incidence of intracranial hemorrhage by anatomical location in a midwestern American population of African Americans and American whites found that African Americans had higher rates of IPH (risk ratio = 1.9). ¹⁹ Despite this difference, the rate of lobar IPH was only nominally increased in African Americans (risk ratio = 1.4). The most important contributor to increased risk in the African American population arose from the increase in PPH and TH (risk ratio = 3.3). Furthermore, when the rate among younger African Americans (aged 35–54 years) was examined, the relative risk was substantially higher (risk ratio = 9.8).