Introduction

Cavernous malformations (CMs) are the most common angiographically occult vascular malformations. Brainstem and thalamic CMs comprise approximately 15% and 9% of cerebral CMs, respectively. Hemorrhagic ictus in these eloquent areas often causes neurologic deterioration, and prior hemorrhage is widely considered to be a predictor of further hemorrhage. The annual hemorrhage rate of untreated CMs in these regions ranges from 2.3 to 21.5%, and a functional outcome of improved or unchanged ranges from 46.7 to 95.7%. Comprehensive preoperative clinical and radiologic evaluation is mandatory to guide selection of treatment strategy and surgical approach for optimizing neurologic outcomes. Although there are no clear treatment guidelines for the timing of surgical intervention, the agreed-upon surgical indications of repeated hemorrhagic episodes and aggravated neurologic status justify surgery in most cases. This chapter focuses on patient selection, surgical pearls, and outcomes of patients with lesions located in the thalamus and brainstem.

Pathophysiology, Incidence, Epidemiology, and Natural History of Disease

CMs, also known as cavernomas or cryptic vascular malformations, are low-flow vascular malformations. ¹ CMs are vascular hamartomas with a multilobulated, mulberry-like appearance that have a wide range in size. The lesions are composed of endothelium-lined sinusoidal chambers without mural elements of mature vascular structures. The lesions lack tight junction, elastic lamina, and smooth muscle. Unlike arteriovenous malformations (AVMs), CMs rarely intralesionally intermix with neural tissue. ¹ Pathologic examination of CMs reveals thrombosis, calcification, hemosiderin-laden macrophages, and recanalization.

CMs are the most common angiographically occult vascular malformations (62-96%), and their prevalence ranges from 0.4 to 0.6%, based on natural history and autopsy studies. ² Mathiesen et al ³ reported that incidental deep and brainstem CMs are detected in 0.8 persons per million per year. Brainstem and thalamic CMs comprise approximately 15% and 9% of cerebral CMs, respectively, without evident male or female predilection. ^{2 , 4} CMs usually present as sporadic and familiar forms, and the latter account for at least 6% of all cases. ¹ Approximately one-fifth of patients have familial disease with an autosomal-dominant pattern of inheritance. CMs are usually associated with mutations in the genes encoding any of three structurally distinct proteins, including KRIT1 (CCM1), CCM2, or programmed cell death 10 (PDCD10; CCM3). ⁵

Because of the relative rarity of brainstem and thalamic CMs, the natural history of the disease is often described together with that of other cerebral CMs or deep CMs, and specific data for prospective hemorrhage rates and neurologic outcomes are limited. The annual hemorrhage rate of brainstem and thalamic CMs varies widely among studies and ranges from 2.3% per patient per year (22 hemorrhages in 943 patient-years) ^{2 , 6 , 7} to 21.5% per lesion per year (with neither the number of hemorrhages from brainstem CMs nor the number of lesions reported). ⁸ Early in a prospective nonfamilial natural history study by Aiba et al, ⁸ the rehemorrhage rate was cited as 11% and 21.5% per lesion per year (numerators and denominators not reported) in basal ganglia (n = 7) and in brainstem (n = 15) CMs, respectively. In 1995, Kondziolka et al ⁹ reported annual hemorrhage rates of 2.4% per patient and 2.9% per patient for brainstem (n = 43) and basal ganglia or thalamus (n = 20) CMs, respectively, and this rate increased to 5% per patient for brainstem CMs with prior hemorrhage (n = 27) (numerators and denominators not reported). In 1997, Porter et al ¹⁰ cited an annual hemorrhage rate of 4.1% per patient (7 hemorrhages in 170 patient-years) for deep CMs (n = 64), which included CMs involving the brainstem, cerebellar nuclei, thalamus, and basal ganglia, and an annual rate of neurologic events of 10.6% per patient (18 hemorrhages in 170 patient-years). In 1999, Moriarity et al ¹¹ grouped thalamus or basal ganglia (n = 9) and brainstem (n = 16) CMs with deep CMs, for which the annual hemorrhage rate was 3.1% per patient (2 hemorrhages in 64.5 patient-years) in patients with a single deep lesion. In a dedicated natural history study of brainstem CMs (n = 37) by Kupersmith et al, ¹² the annual rehemorrhage rate was 5.1% per patient (8 hemorrhages in 157 patient-years). In recent studies by Li et al, ^{6 , 7} the annual hemorrhage rate was estimated to be 13.6% per patient (185 hemorrhages in 1,364.3 patient-years) in 331 adult patients with brainstem CMs and 11.7% (47 hemorrhages in 401.6 patient-years) in 85 pediatric patients with brainstem CMs. Results of retrospective studies have been summarized by Gross et al, ^{13 , 14} who calculated an annual hemorrhage rate of 2.3% (22 hemorrhages in 943 patient-years) for brainstem CMs (n = 25) and 2.8% (10 hemorrhages in 355 patient-years) for deep hemispheric CMs (n = 13). Barker et al, ¹⁵ described the clustering of CM-related hemorrhagic events within the first 2 years after the initial hemorrhage as indicating self-limited hemorrhage; this phenomenon was demonstrated by Li et al ^{6 , 7} both in symptomatic, hemorrhagic brainstem CMs and in pediatric brainstem CMs.

Various adverse hemorrhagic predictors for cerebral CMs have been reported, including sex, ^{6 , 8 , 11} age, ⁶ lesion location, ¹⁰ presence of edema, ^{6 , 7} and prior hemorrhage. ^{6 , 8 , 9 , 10 , 12} Prior hemorrhage is considered to be a widely accepted hemorrhagic risk, and the annual hemorrhage rate increases in patients with a history of hemorrhagic brainstem or thalamic CMs. Prior hemorrhage often justifies treatment.

The neurologic outcomes of patients with untreated brain-stem or thalamic CMs are documented in only a few studies with relatively small numbers of patients. The improved or unchanged functional outcomes range from 46.7 to 95.7%, and a poor outcome is reported in 20% of patients for whom long-term follow-up is available. ^{6 , 7 , 12} Li et al ⁷ recently reviewed 7 studies that included 123 untreated brainstem CMs and found that 88 patients (71.5%) were improved or unchanged and that 33 patients (26.8%) had worsened; 2 patients who died of myocardial infarction were not included in the analysis. They also found that in 11 studies, mortality from rehemorrhage was 4.1% (9/218). In Li et al’s ⁷ cohort of 85 pediatric patients with brainstem CMs who had a mean follow-up of 4.7 years, the improved and worsened outcomes were 76.5% (n = 65) and 23.5% (n = 20), respectively, and 25.9% (n = 22) had complete recovery of neurologic function. Prospective hemorrhage was verified as an independent risk factor prohibiting full recovery. In a large series by Li et al ⁶ that included 331 patients with untreated brainstem CMs, after a mean follow-up of 6.5 years, 307 patients (92.7%) were improved or had stable neurologic deficits, 268 (81.0%) were living independently, and 95 (28.7%) had completely recovered. Factors related to complete recovery were extended hemorrhage-free follow-up, younger age, and small lesion size. Overall, the outcomes of untreated brainstem CMs are acceptable but are not as poor as previously believed. However, Aiba et al, ⁸ reported that 4 of 7 patients with basal ganglia or thalamic CMs had excellent outcomes, whereas 3 patients were moderately disabled. The neurologic morbidity observed by Pozzati ¹⁶ was 40% (2/5 patients), and the rate of morbidity per hemorrhage was 29% (2/7 hemorrhages).

Case reports suggest an increased risk of hemorrhage with pregnancy. However, two large studies, including a prospective study from Barrow Neurological Institute, suggest that pregnancy is not generally a risk factor for hemorrhage in patients with CMs. ^{17 , 18}

Clinical Presentation

The clinical symptoms of patients with CMs are referable to the anatomic location of the lesion. In most prior surgical series, the clinical course of patients with brainstem and thalamic CMs who had hemorrhages was aggressive and led to severe neurologic deterioration, as only a few patients were diagnosed incidentally or because of other unrelated causes ( Table 30.1, Table 30.2 ). ^{3 , 4 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 , 44 , 45 , 46 , 47 , 48 , 49 , 50 , 51 , 52 , 53} Initially, without magnetic resonance imaging (MRI) evidence, the clinical presentation of patients with angiographically occult lesions might be confused with that of patients with demyelination, neoplasms, infection, or infarction. ⁴ Some symptoms that lead to suspicion of infarction are similar to symptoms of Wallenberg, Millard-Gubler, Parinaud, Benedikt, or Weber syndromes. ²

In the largest surgical series of thalamic CMs (n = 46), ⁴⁸ the most common symptom at presentation was hemorrhage (59%, n = 27), followed by headache (41%, n = 19), contralateral body numbness or paresthesia (22%, n = 10), diplopia (7%, n = 3), and contralateral facial nerve paralysis (4%, n = 2). Of the 12 patients with thalamic CMs reported by Pozzati, ¹⁶ most (67%, n = 8) presented with hemorrhage. Other authors have reported a hemorrhagic presentation in 100% of 27 patients harboring thalamic CMs, ⁵³ with the most common symptoms at presentation being weakness (63%, n = 17) followed by paresthesia (56%, n = 15), headache (48%, n = 13), visual disturbance (22%, n = 6), and language deficits (19%, n = 5). The pooled percentage of patients with hemorrhage in the 33 thalamic CMs reviewed by Gross et al ⁴ was 45% (15/33 patients), and other symptoms included hemiparesis, hemisensory deficit, hemianopsia, hemidystonia, diplopia, and thalamic pain syndromes. Parkinsonism and extra-pyramidal symptoms were rare.

In a large series with 176 patients with 179 CMs involving the brainstem, thalamus, and basal ganglia, most patients presented with multiple signs and symptoms. ³⁸ The most common presenting symptoms in the 176 patients were cranial nerve deficits (51.1%, n = 90), hemiparesis (40.9%, n = 72), face or body numbness (34.7%, n = 61), and cerebellar symptoms (38.6%, n = 68). Other unspecific symptoms included headache (36.4%, n = 64), diplopia (38.1%, n = 67), vertigo (11.4%, n = 20), swallowing difficulty (7.9%, n = 14), and seizures (5.7%, n = 10). In other large series of patients with only brainstem CMs, ^{19 , 20 , 25 , 39 , 44 , 51 , 52} the most common presenting signs and symptoms were cranial neuropathy (63-82.6%), followed by pyramidal sign and other relatively rare symptoms of dysarthria, dysmetria, refractory hiccup, hydrocephalus, and respiratory difficulty. Wang et al ⁴⁴ described the clinical presentation of 137 patients in their series on the basis of lesion locations, which indicated the level of brain-stem affected. In 29 patients with midbrain CMs, diplopia (69%, n = 20) was the most common symptom, followed by hemiparesis (48%, n = 14), ataxia (38%, n = 11), and increased intracranial hypertension (38%, n = 11). In 90 patients with pontine CMs, the most common symptoms were CNs V-VIII deficits (76%, n = 68), followed by hemiparesis (57%, n = 51), hemianesthesia (49%, n = 44), and vertigo (44%, n = 40). In 18 patients (n = 18) with medullary CMs, dysphasia was observed in all 18 patients.