Work

Conversations on mutism: risk stratification for cerebellar mutism based on medulloblastoma subtype

Year: 2019

Type: article

Abstract: See the article by Jabarkheel and Amayiri et al. in this issue, pp. 290–297. An unfortunate and somewhat mysterious postoperative complication of posterior fossa tumor resection is cerebellar mutism syndrome (CMS), classically defined as reduced speech or mutism with emotional lability that begins within days of posterior fossa surgery and typically improves over weeks to months.1 CMS is often conflated with terms for comorbid syndromes with overlapping symptoms, such as posterior fossa syndrome, cerebellar cognitive affective syndrome, cerebellar motor syndrome, and pseudobulbar palsy. Recent attempts at consensus rectify these disparate terms into one overarching definition that emphasizes the delayed onset and transient nature of the speech and emotional symptoms of CMS, which may be accompanied by cognitive, cranial nerve, or motor deficits.2 This array of associated symptoms aligns with our ever-evolving understanding of the non-motor roles of the cerebellum in cognition and emotion. CMS is relatively common, estimated at about 24% of patients after resection of medulloblastoma, by one prospective study,3 as well as a more recent multicenter retrospective study.4 Thus, 1 in 4 children who already suffers from a posterior fossa tumor and the stresses of neurosurgical intervention will develop additional, disabling comorbidity from CMS. Yet despite its high frequency, our understanding of the pathophysiology and predisposing risk factors for CMS remains limited. Current thinking suggests that CMS results from disruption of the dentato-thalamo-cortical pathway, particularly the efferent tracts that exit the cerebellar dentate nuclei via the superior cerebellar peduncles.5 The risk factors that lead these pathways to be vulnerable in some subsets of patients remain elusive, even though multiple studies have tried to illuminate how patient characteristics, radiographic features, surgical techniques, or tumor type might predict CMS. One surgical study, for example, suggested that a high level of retraction or a vermian incision might increase the risk of postoperative CMS,6 but this has failed to be consistently upheld in the literature.4,7 Another study analyzed radiologic aspects of posterior fossa tumors to develop a preoperative scoring system to predict patients at high risk of developing CMS.8 The authors concluded that bilateral middle cerebellar peduncle invasion, dentate nucleus invasion, and age greater than 12.4 years predicted an increased risk of postoperative CMS, while cerebellar hemispheric location was protective.8 However, a limitation of this model is that some of the subtler radiologic features did not demonstrate strong interobserver agreement.8 Thus, while modest headway has been made toward understanding surgical practices that minimize CMS or radiographic biomarkers that predict CMS, we have yet to explain how tumor biology itself contributes to CMS risk. Medulloblastoma has typically been reported as the posterior fossa tumor with the highest risk of postoperative CMS4,5,7 and is now known to comprise 4 molecularly, radiographically, and clinically distinct subgroups. Two subgroups are named after neurodevelopmental signaling molecules—wingless (WNT) and sonic hedgehog (SHH)—and 2 with less certainty regarding developmental driver genes are nebulously named Group 3 and Group 4.9 In this issue of Neuro-Oncology, Jabarkheel et al systematically investigated how molecular properties of medulloblastoma correspond to risk of developing CMS in an international multicenter study of 370 patients across 5 sites.9 First they confirmed an incidence of CMS in 23.8% of cumulative medulloblastoma patients,9 consistent with previous reports.3,4 They also found that younger age, larger tumor volume, and midline location were associated with an increased risk of CMS,9 again in agreement with previously identified risk factors.3,5,7,8 After controlling for age, tumor volume, and tumor location, multivariate analysis revealed that molecular subcategory remained predictive of CMS development. Highest rates of CMS were observed in children with Group 4 tumors (35%), followed by Group 3 (31%), and then WNT (21%).9 SHH tumors had the lowest risk of CMS development (7%).9 Moreover, within Group 4, tumor volume continued to have a significant positive association with the development of CMS.9 These findings demonstrate that tumor subtype is indeed predictive of CMS risk, which begs the question of how to explain this correlation. Does molecular subtype predict features of medulloblastoma, such as tumor size and location, which in turn directly increase the risk of CMS? Previous work by members of the same research team implies that this is likely to be the case. An earlier study examined radiographic features that were predictive of the medulloblastoma subtype, which was then confirmed on pathology.10 Tumor location and contrast enhancement correctly identified medulloblastoma molecular subtype in two-thirds of cases.10 Groups 3 and 4 medulloblastomas were typically located in the midline fourth ventricle, WNT medulloblastomas in the cerebellar peduncle or cerebellopontine angle, and SHH medulloblastomas in the cerebellar hemispheres.10 Midline Group 4 tumors were distinguished additionally by lack of contrast enhancement.10 Another relevant clue toward the link between medulloblastoma subtype, radiographic appearance, and ultimately CMS risk is the distinct neurodevelopmental domains from which each subtype is derived. Group 3 tumors likely arise from Nestin-expressing cells, Group 4 tumors from unipolar brush cells, WNT tumors from the lower rhombic lip, and SHH tumors from the external granule cell layer.9 One can start to see how these pieces fit together: the genetics of a particular medulloblastoma subtype and its neurodevelopmental domain likely confer properties on the budding tumor that dictate its eventual temporal course, location, and size. Its location and size in turn correlate with its radiographic appearance, the surgical strategy needed for resection, the odds of disrupting the dentato-thalamo-cortical pathway, and ultimately the postoperative likelihood of developing CMS. The results of this study provide insight into how modifying clinical practice could minimize the undue burden of CMS on pediatric patients with posterior fossa tumors. Previously, molecular subtype could only be confirmed by postoperative pathology—too late for patients who have already developed CMS. A crucial application of this research group’s cumulative work is that radiographic features characteristic of medulloblastoma subtype can be combined with other predictive models8 to identify preoperatively those patients at highest risk for CMS. Though surgical technique does not appear convincingly to alter the chance of developing CMS, risk stratification will still be helpful in guiding future medical and surgical interventions by individualizing their treatment options to minimize development of CMS. Herein, Jabarkheel et al lay the groundwork for such individualized risk assessment in the treatment of medulloblastoma, another step toward reducing the incidence of the enigmatic and debilitating symptoms of cerebellar mutism syndrome. The Weiss lab is supported by R01CA159859, R01NS106155, R01NS091620; R01CA221969; P01CA217959; P30CA82103; P50CA097257; U01CA217864, U54CA243125, Brain Tumour Charity (GN356); Cancer Research UK (A28592) the Evelyn and Mattie Anderson Chair; The Children’s Tumor-, V- (T2017-020), and Samuel G. Waxman Foundations.

Source: Neuro-Oncology

DOI: 10.1093/neuonc/noz228

Authors: Jeffrey B Russ, William A Weiss

Institutions: University of California, San Francisco, UCSF Helen Diller Family Comprehensive Cancer Center

Language: English

Topic: Glioma Diagnosis and Treatment

Subfield: Genetics (clinical)

Field: Medicine

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: green

Funders: Cancer Research UK, Brain Tumour Charity, National Cancer Institute, National Institute of Neurological Disorders and Stroke