Long-term outcomes and late toxicity of adult medulloblastoma treated with combined modality therapy: A contemporary single-institution experience

Abstract

Background

Medulloblastoma (MB) is a rare central nervous system malignancy of adults, with limited contemporary studies to define treatment guidelines and expected late toxicity.

Methods

A single-center, retrospective study was conducted of patients age ≥18 years from 1997–2019 with MB and who were treated with postoperative radiotherapy. Late toxicity was defined as a minimum of 18 months from diagnosis. Overall survival (OS) and progression-free survival (PFS) were characterized using Kaplan-Meier and Cox regression analyses.

Results

Fifty-nine patients met criteria, with median age of 25 years (range 18–62 y) and median follow-up of 6.5 years (range 0.7–23.1 y). At diagnosis, 68% were standard-risk, 88% Chang M0, and 22% with anaplastic histology. Gross total resection was achieved in 75%; median craniospinal irradiation dose was 30.6 Gy (relative biological effectiveness [RBE]), median total dose was 54.0 Gy (RBE), 80% received proton radiotherapy; 81% received chemotherapy. 5 year PFS and OS were 86.5% and 95.8%, respectively; 10 year PFS and OS were 83.9% and 90.7%, respectively. Anaplastic histology was associated with worse PFS (P = .04). Among eight recurrences, 25% presented after 5 years. Most common grade ≥2 late toxicities were anxiety/depressive symptoms (30%), motor dysfunction (25%), and ototoxicity (22%). Higher posterior fossa radiation dose was associated with increased risk of late toxicity, including worse cognitive dysfunction (P = .05).

Conclusions

Adults with MB have favorable survival outcomes, but late failures and toxicity are not uncommon. Better understanding of prognostic factors, possibly from molecular subtyping, may help to define more personalized treatments for patients with high risk of recurrence and long-term treatment sequelae.

Key Points.

1. Long-term survival outcomes in adult medulloblastoma are excellent.

2. Recurrences are uncommon but can occur beyond 5 years from diagnosis.

3. Up to 30% of patients experience toxicity in the years following treatment.

Importance of the Study.

Adult medulloblastoma treatment strategies are largely based on pediatric studies and guidelines, though recent molecular studies have shown a significant difference between adult and pediatric medulloblastoma. This single institutional experience of contemporary combined modality therapy demonstrates excellent clinical outcomes. Reported incidence and diversity of late toxicity may help to guide optimal strategies for survivorship care in this patient population.

Medulloblastoma (MB) is an aggressive primary central nervous system (CNS) malignancy with rapid growth and tendency to disseminate within the CNS and systemically. MB is the most common malignant brain tumor in children but is rare in adults, accounting for less than 1% of all adult CNS malignancies.¹ Management of adult medulloblastoma is typically derived from retrospective data and pediatric experiences; however, natural history and molecular profiles of MB differ substantially.^2,3

Management of MB is driven by risk stratification of disease, currently stratified into standard- and high-risk based on residual disease after surgery and/or evidence of disseminated disease. Modern studies have shown that adult MB with standard-risk can experience late recurrences and extraneural relapses, which are less common in the pediatric population.^4,5 Molecular profiling in pediatric and adult MB has led to the classification of four unique subtypes of MB: SHH group, WNT-group, Group 3 and Group 4.^6–10 Adult MB is a genetically distinct disease process, where the most common subtype is the more favorable SHH-subtype with distinct molecular and genomic differences within subtypes, including rare MYC/MYCN amplifications compared to other markers such as CDK6 amplification and chromosomal aberrations.^10–12

Management typically consists of maximal safe resection followed by craniospinal irradiation (CSI) and chemotherapy.^2,3 CSI doses and techniques have evolved over time and may vary substantially across institutions, with modern strategies achieving greater target conformality and exploring lower CSI doses.^13–15 The reduction in collateral normal tissue irradiation using protons as compared to photon radiation therapy has been associated with improved acute and long-term toxicity.¹⁶ Series reporting long-term clinical outcomes beyond 5 years have suggested improved toxicity with proton RT in children and adults with MB, including neurocognition, ototoxicity, neuroendocrine deficit, cardiac, pulmonary, and gastrointestinal toxicity.^4,16–18

Prospective studies evaluating management of MB in adult patients are limited, with management and survivorship driven from extrapolating from heterogeneous retrospective series and pediatric MB experiences.^2,3,19,20 We present here a contemporary single institutional retrospective study including a large cohort of patients treated with proton CSI, assessing long-term survival outcomes and incidence of late toxicity.

Methods

Patient Population

A retrospective review was performed of adult patients (age ≥18 years) with MB treated from 1997–2019 at a single tertiary-care institution. All patients had histopathologic diagnosis confirmed at our center, received postoperative RT at our center, and had a minimum of 6 months of follow-up. A total of 73 adult MB patients were initially identified. Of these, 14 were ultimately excluded from final analysis: seven did not have sufficient follow up, three patients received RT at outside institutions, two had insufficient information in medical records, and two were deemed to have a different diagnosis after secondary pathologic review (both poorly differentiated malignant epithelial tumor). All research was conducted in accordance with our Institutional Review Board (IRB) approved protocol.

Clinical and Pathologic Variables

Surgical extent was defined by neuroradiologist interpretation of findings on postoperative MR scan within 48 hours of surgery; if scan was not available then extent of resection was extrapolated from surgeon’s operative report. All patients were separated into standard-risk, defined as no evidence of metastasis on MRI or CSF and less than 1.5 cm² of residual tumor, versus high-risk, defined as presence of metastatic disease (Chang stage M1 or higher) or more than 1.5 cm² of residual tumor on postoperative imaging.¹⁴ Molecular subtyping, where available, was performed by combination of immunohistochemistry, DNA methylation array, next generation sequencing, or targeted RT-PCR mRNA expression.

Toxicity

Toxicity was defined by the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 5.0 with scale of grade 2 (G2), 3, 4, or 5. Motor dysfunction was defined as the presence of ataxia and/or weakness. Limited data exists for a validated, reproducible definition of late toxicity in adult medulloblastoma. Therefore, a timepoint for late toxicity was selected based on our clinical workflow that would allow for sufficient time beyond a full course of treatment, up to 12 months in total in some patients, and at least one follow-up scan, to minimize any potential overlap with acute, reversible toxicity. Late toxicity was defined as toxicity arising at least 18 months after diagnosis; we additionally assessed toxicity assessed among a subset of patients with at least 5 years of follow up. Toxicity analysis was performed on maximum toxicity within the entire follow-up period.

Statistical Analysis

Survival outcomes (progression-free survival and overall survival), and clinical outcomes (toxicity, including neurologic symptoms, mood, etc.) were evaluated. Progression-free survival (PFS) was defined as time interval between the date of diagnosis and the date of recurrence of disease (local, craniospinal, or extraneural) or death. Overall survival (OS) was defined as time interval between the date of diagnosis and the date of death of any cause. Kaplan-Meier method was used to estimate PFS and OS. We compared survival curves among different patient subgroups using the log-rank test with statistical significance defined as P < .05. Univariate actuarial analysis was performed with Cox proportional hazards model, with P < .05 defined as statistically significant. Multivariate analysis was not performed due to limited number of events. Statistical analyses were performed with Stata (Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC.).

Results

A total of 59 patients were included in this study with a median follow up of 78 months (range 8–277 months) (Table 1). Median age at diagnosis was 25 years (range 18–62) and 56% of patients were male. At diagnosis, 68% of patients were standard-risk and 88% of patients were Chang stage M0 (Supplementary Table 1). Classification by histologic subtype was 36% desmoplastic histology, 42% classical histology, and 22% anaplastic histology. Of the 23 patients with available molecular profiling, 14 were SHH subtype, four were Group 4 subtype, three were WNT subtype, and two were Group 3.

Table 1.

Baseline Patient and Tumor Characteristics, and Patterns of Treatment

Patient or tumor characteristic (n = 59)	n (%)
Age at diagnosis (median (range))	25 (18–54)
Sex
Female	26 (44)
Male	33 (56)
Risk category
Standard risk	40 (68)
High risk	19 (32)
Chang M stage
M0	52 (88)
M1-4	7 (12)
Histology
Desmoplastic	21 (36)
Classical	25 (42)
Anaplastic	13 (22)
Molecular subtype
SHH	14 (24)
WNT	3 (5)
Group 3	2 (3)
Group 4	4 (7)
Unavailable	36 (61)
Extent of resection
Gross total resection	44 (75)
Subtotal resection/biopsy	15 (25)
Radiation dose, median, (range)
CSI, Gy (RBE)	30.6 (23.4–36)
Boost, Gy (RBE)	23.4 (18–36)
Total, Gy (RBE)	54 (54–60)
Radiation modality
Photon	9 (15)
Proton	47 (80)
Mixed	3 (5)
Chemotherapy
Neoadjuvant only	1 (2)
Concurrent only	2 (3)
Adjuvant only	10 (17)
Concurrent + Adjuvant	33 (56)
Neoadjuvant + Concurrent + Adjuvant	2 (3)
None	11 (19)

All patients had multi-modality therapy with some combination of maximal safe resection, CSI, and chemotherapy (Table 1). Gross total resection (GTR) was achieved in 75% of patients. Second surgery was performed in seven patients, and five of those patients achieved GTR after a second surgery. Of the two patients with STR after a second surgery, one had anaplastic histology and had local recurrence months after surgery and the other STR had limited follow-up at 15 months.

Median time from surgery to RT initiation was 39 days (range 19–409 days). Median CSI dose was 30.6 Gy (RBE) (range 23.4–36.0) with a median boost to 23.4 Gy (RBE) (range 18.0–36.0), with a median total dose of 54.0 Gy (RBE) (range 54.0–60.0). Proton therapy was the predominant modality, with 80% receiving proton RT, 15% received photon RT, and 5% receiving a mix of proton and photon RT.

A total of 48 patients received chemotherapy, with 3/48 patients (5%) receiving neoadjuvant chemotherapy prior to surgery, 37/48 patients (64%) receiving chemotherapy concurrently with CSI, and 42/48 patients (72%) receiving adjuvant chemotherapy after CSI (Table 1). The most common chemotherapy treatment approach was to administer concurrent and adjuvant chemotherapy (56%), followed by adjuvant only (17%), concurrent only (3%), neoadjuvant + concurrent + adjuvant (3%), and neoadjuvant only (2%). The most common concurrent chemotherapy regimen was vincristine alone (29/37 patients, 78%), followed by vincristine + carboplatin (8/37 patients, 22%). The most common adjuvant chemotherapy regimen was vincristine, cisplatin, and cyclophosphamide (30/42 patients, 71%). Among patients who received adjuvant chemotherapy, median time from RT to chemotherapy was 1.5 months (range 1.0–2.5 months), and number of cycles was 6 (range 2–9). Among 9/42 patients who did not complete planned adjuvant chemotherapy cycles (median cycles completed = 4), five were stopped due to neuropathy, one due to persistent pancytopenia, one due to stroke, one due to fatigue, and one due to early recurrence.

Outcomes

The 5 year and 10 year OS were 95.8% (95% CI 84.0–99.0) and 90.7% (95% CI 76.9–96.6), respectively (Figure 1a). The 5 year and 10 year PFS were 86.5% (95% CI 73.5–93.4) and 83.9% (95% CI 70.1–91.7), respectively (Figure 1b). When stratified by anaplastic histology (Figure 1C), the 5 year and 10 year PFS for non-anaplastic MB were 95.0% (95% CI 81.2– 98.7) and 89.3% (95% CI 73.7–95.9), compared to 76.9% (95% CI 44.2–91.9) and 65.9% (95% CI 31.5–86.0), respectively, for anaplastic MB.

Fig. 1.

Survival of total population. (A) Kaplan-Meier Curve for overall survival from diagnosis of medulloblastoma to event, with 5 year OS of 95.8% and 10 year OS of 90.7%. (B) Kaplan-Meier Curve for progression-free survival from diagnosis of medulloblastoma to recurrence, with 5 year PFS of 86.5% and 10 year PFS of 83.9%. (C) Kaplan-Meier Curve for progression-free survival from diagnosis of medulloblastoma to event, stratified by anaplastic histology. The 5 year PFS for non-anaplastic MB was 89.3% compared with 76.9% for anaplastic MB.

Recurrence

There were eight recurrences in this cohort, with four recurrences in the posterior fossa, two within the craniospinal axis outside the posterior fossa, and two extraneural (Table 2, Figure 2). Seven of the eight recurrences (88%) were in patients classified as standard-risk, and all were Chang stage M0. 50% of recurrences were anaplastic histology, followed by desmoplastic histology (38%) and classical histology (13%). Most patients with recurrences had received upfront chemotherapy, with 5/8 patients (63%) receiving concurrent and adjuvant therapy followed by one patient (13%) with adjuvant therapy; 2/8 patients (25%) with recurrent disease had not received any prior chemotherapy. Of note, both patients with extraneural disease recurrences received concurrent and adjuvant chemotherapy.

Table 2.

Patterns of Recurrence Among All Patients

Category	N	%
Recurrences	8	100
Site of recurrence
Posterior Fossa	4	50
Craniospinal	4	50
Extraneural	2	25
Risk
Standard risk	7	88
High risk	1	12
Chang M stage
M0	8	100
Histology
Desmoplastic	3	38
Classical	1	13
Anaplastic	4	50
Chemotherapy
Adjuvant only	1	13
Concurrent + Adjuvant	5	63
None	2	25

Fig. 2.

Patterns of Failure in Adult Medulloblastoma. Forty-seven-year-old male with standard-risk, anaplastic, SHH-activated MB treated with GTR and craniospinal proton RT (36.0 Gy [RBE] CSI, 19.8 Gy [RBE] PF boost) without concurrent or adjuvant chemotherapy. He experienced concurrent local failure of posterior fossa, as seen on axial and coronal MRI Brain images (A) as well as nodular recurrence within cord at C1-2, as seen on axial and sagittal MRI Spine images (B), 65 months after diagnosis. Twenty-four-year-old female with standard-risk, desmoplastic, SHH-activated MB treated with GTR, craniospinal proton RT (23.4 Gy [RBE] CSI, 30.6 Gy [RBE] PF boost) with concurrent (weekly vincristine) and adjuvant (vincristine/lomustine/cisplatin) chemotherapy for six cycles. She experienced extraneural recurrence within left parotid bed, as seen on coronal and axial PET images (C) and fused axial PET/CT and axial CT images (D), 56 months after diagnosis.

Several patient and tumor factors were investigated as potential risk factors for disease progression (Table 3). The results indicate a higher risk of disease progression for anaplastic histology (HR = 4.2, 95% CI 1.04–14.7, P = .04). Risk category, M stage, extent of resection, total RT dose, time to start RT, and sequencing of chemotherapy were not associated with disease progression. With only three deaths a regression analysis of factors associated with overall survival was not feasible.

Table 3.

Univariate Cox Regression Analysis Results for Progression-Free Survival Among All Patients

Category	HR (95% CI)	P-value
Risk category (high v low)	0.26 (0.09–2.19)	0.3
Extent of resection (STR/biopsy v GTR)	0.42 (0.05–3.38)	0.4
Time to start RT (continuous)	1.00 (0.98–1.02)	0.8
RT duration (continuous)	0.99 (0.87–1.12)	0.9
Adjuvant chemotherapy (no vs yes)	0.71 (0.14–3.54)	0.7
Concurrent chemotherapy (no vs yes)	0.87 (0.21–3.65)	0.8
Histology (desmoplastic vs others)	0.94 (0.23–3.96)	0.9
Histology (anaplastic vs others)	4.18 (1.04–16.74)	0.04
CSI (≥36 Gy (RBE))	7.38 (0.91–60.04)	0.06
Total dose (continuous)	1.21 (0.78–1.88)	0.4

Late Toxicity

Among 56/59 patients with minimum of 18 months follow up, 42 patients (75.0%) had at least one ≥G2 late toxicity and 13 patients (23.2%) had at least 1 ≥G3 late toxicity (Supplementary Table 2). The most common ≥G2 toxicities were anxiety/depressive symptoms (30%), motor dysfunction (25%), ototoxicity (22%), cognitive dysfunction (21%), hypothalamic-pituitary dysfunction (20%), and cranial nerve dysfunction (14%). Posterior fossa (PF) RT boost dose greater than 54 Gy (RBE) was associated with increased risk of neuropathy (P = .01), hypothalamic-pituitary-adrenal (HPA) axis dysfunction (P = .03), and total cognitive dysfunction (P = .05) (Supplementary Table 3, 4). Adjuvant chemotherapy was associated with better perceived short-term memory (P = .02). The most common ≥G3 toxicities were ototoxicity (9%), motor dysfunction (7%), and cognitive dysfunction (5%). There were three secondary malignancies: two papillary thyroid carcinoma and one osteosarcoma of the bone; all were within RT fields. Among patients with at least 60 months follow-up, 28.6% of patients were prescribed medication for at least 3 months for anxiety or depressive symptoms, and 69.6% of patients were gainfully employed or in school full-time.

Discussion

Our single-institutional analysis demonstrates adult patients with MB have excellent survival rates but late and distant relapses are not uncommon. Our survival outcomes are similarly favorable compared to contemporary pediatric MB survival rates.^21,22 In our cohort, eight (13.6%) patients had recurrences at a median follow up of 26 months; however, most patients with recurrences were classified as standard-risk or M0 disease at diagnosis, suggesting the need for better prognostication. This is the largest contemporary single-institutional series of adult MB to date to our knowledge, as well as the largest single-institutional cohort of adult MB treated primarily with proton RT.

Our series consisted of patients with primarily favorable features, as most patients were M0 and standard-risk at diagnosis with non-anaplastic histology. A second surgery was employed to achieve GTR whenever feasible. Metastatic disease at presentation was seen in 12% of patients, which is similar to what has been described in other series.²³ These survival results are similar to other series of adults with favorable MB, with similar clinical outcomes utilizing modern treatment strategies (Table 4). Majd et al report the largest experience of 200 adult MB patients (80 patients receiving surgery and/or postoperative care at MD Anderson Cancer Center) treated from 1978–2017 with a 5 year OS of 74% for all patients and 62% for high-risk patients.²³ Use of RT, use of chemotherapy, and GTR were associated with improved OS, and late recurrences were more frequent than in pediatric studies, similar to our series. Our study intentionally excluded older data to assess outcomes based upon modern technology and standard-of-care treatment. In our study, rates of RT use were higher (100% vs 92%), chemotherapy use was higher (81% vs 43%) and GTR was higher (75% vs 55%) although incidence of high-risk (32% vs 30%) and presence of metastatic disease at diagnosis (12% vs 11.5%) was similar.

Table 4.

Current Study Examining Clinical Outcomes of Adult Medulloblastoma Patients in the Context of Other Published Adult Medulloblastoma series

Author	Dates of inclusion	No. Patients	GTR	Standard risk	M0	Desmoplastic/ classic	Anaplastic histology	Chemotherapy	5 y DFS	5 y OS	10 y DFS	10 y OS
Brandes, 200734	1989–1995	36	31%	28%	64%	-	-	61%	72%	75%	-	-
Padovani, 200735	1975–2004	253	63%	55%	92%	96%	0%	58%	65%	72%	55%	55%
Silvani, 201236	1991–2001	28	50%	-	86%	100%	0%	100%	58%	80%	-	56%
Balducci,201237	1990–2008	13	46%	31%	77%	100%	0%	31%	84%	76%	84%	76%
Friedrich, 201338	1998–2009	70	63%	-	71%	96%	4%	70%	68% (4 y)	89% (4 y)	-	-
Call, 201439	1969–2008	66	62%	50%	67%	98%	2%	95%	59%	74%	-	-
Buglione, 201540	1975–2006	16	69%	69%	81%	100%	0%	13%	67%	75%	60%	67%
Beier, 2018 (NOA-07)41	2009–2014	30	50%	-	67%	100%	0%	100%	66% (3 y)	70% (3 y)	-	-
Majd, 202123	1978–2017	200	55%	55.6%	63%	67%	4%	48%	55%	74%	42%	48%
Quinones, 202142	2006–2017	26	85%	-	54%	88%	12%	81%	77%	80%	-	-
Liu, 202129	2008–2020	20	60%	55%	75%	90%	10%	75%	90% (4 y)	95% (4 y)	-	-
Current study	1997–2019	59	75%	68%	88%	78%	22%	81%	86%	96%	83%	93%

Median time to recurrence was 26 months, with half of recurrences occurring after 3 years and two recurrences occurring after 5 years (both were local recurrences). Two patients had extraneural recurrence (one in the left parotid 56 months after diagnosis, who also had a history of Li-Fraumeni Syndrome, another in the right cervical lymph nodes 17 months after diagnosis); both were M0 and standard-risk on initial diagnosis and received concurrent and adjuvant systemic therapy. Brandes et al reported long-term results of their prospective study of adult MB, and found recurrence continued to increase after 7 years of follow-up in standard-risk patients, and after 10 years in high-risk patients.⁴ De et al. report their series with 9.0 year follow-up with one of five recurrences occurring at 7.2 years, and was classified as standard-risk at diagnosis.¹⁸ This suggests long-term follow up is crucial in this population, and improved risk stratification is needed to better identify patients at risk for recurrence. Anaplastic histology was associated with worse PFS, which is consistent with other series^24–26; this further adds to the consideration of adding anaplastic histology to high-risk classification and treatment intensification.²⁷ Due to a limited number of events in our cohort, it was not feasible to perform univariate analyses of factors associated with overall survival.

Rates of late toxicity were similar among patients with at least 18 months follow up compared with patients with at least 60 months follow up, which suggests any late toxicity at 18 months may continue to persist and therefore is a clinically meaningful endpoint for symptom management and survivorship considerations.^18,28,29 Rates of toxicity were also similar to other modern series: De et al. reported an MSKCC long-term experience with 48% ototoxicity (21% ≥G2) and 21% hypothyroidism, similar to our rates of 49.2% ototoxicity (22.0% ≥G2) and 14.3% hypothyroidism, as well as high rates of anxiety/depressive symptoms and cognitive disturbance.¹⁸

Studies have demonstrated lower CSI doses are associated with decreased late toxicity in pediatric MB, though survival may be inferior in some patients^22,30. While unable to adequately assess toxicity related to CSI dose in this study, future studies should investigate which patients may be candidates for lower CSI with equivalent survival outcomes and decreased toxicity given excellent outcomes in adult MB. We found that a posterior fossa dose >54 Gy (RBE) was associated with increased toxicity, including cognitive dysfunction, memory deficits, and neuropathy. While the standard dose typically is not greater than 54 Gy (RBE), this adds an important consideration of diverse symptomatology of posterior fossa toxicity and the therapeutic ratio. Finally, it is important to note a high incidence of proton CSI in this study (80% protons and 5% mixed photon/proton), which is unique to other adult MB studies. While studies have demonstrated the decreased rate of late toxicity in children receiving proton CSI, limited robust data exists for adult MB patients; only one other study (Liu 2021) in our literature review (Table 4) mentions proton RT^29,31–33.

Adjuvant chemotherapy was the most common sequencing of chemotherapy in this study, with most patients receiving vincristine, cisplatin, and cyclophosphamide. Among patients who received adjuvant chemotherapy, 19% did not complete planned chemotherapy cycle due to toxicity, which is similar to other proton CSI series (15%) and slightly lower than other historical series of 40–52%^18,29,34. Most common toxicity leading to chemotherapy discontinuation were neuropathy, cytopenia, and fatigue, respectively, while is similar to other reported series^3,18,34. There is a growing understanding of a difference in outcomes with different molecular subgroups of adult MB, however it is not yet clear which groups may benefit from changes in conventional chemotherapy regimens or targeted therapy^35–37. Future studies should investigate modification or deferment of chemotherapy for selected patients with more favorable overall prognosis.

Our results further add to the literature of the high rate of mood disturbances in the cancer survivorship population, suggesting an urgent need for greater emotional/coping support integrated into long-term care. Clinicians should consider continually addressing common late toxicity in this patient population within first few years of follow-up, followed by referral to specific services, such as psychology, endocrinology, physical therapy, and occupational therapy. Clinicians should also consider referral to survivorship clinics or support groups as early as at time of diagnosis, noting significant challenges in each of coping with a new diagnosis, tolerating diagnostic testing and treatments, and adjusting to survivorship.

Limitations of this study include the retrospective study design, with potential for inclusion bias and limited ability to evaluate treatment efficacy through routine surveillance. Even with the intentional limitation of patients treated in the last 25 years and exclusion of patients with noncontemporary management, this series still spans a period of evolving practices in management. Anxiety and depression were defined per CTCAE, however more standard metrics such as Patient Health Questionnaire or other objectives measurements were not routinely available during the period of study and therefore these results may be difficult to interpret; more robust instruments measuring anxiety and depression should be incorporated into future studies of adult MB. Finally, follow up data may be limited by our reliance upon single institution records, as some patients may have returned to their local providers for long-term care.

Our study highlights the limited utility of our current use of metastatic burden and standard- versus high-risk classification in predicting disease recurrence. Thus, future studies are needed to clarify patient and tumor related factors affecting prognosis. This may additionally inform tailored treatment strategies. Molecular subgrouping of MB has demonstrated differences between pediatric and adult populations, both in incidences of subtypes such as greater SHH subtypes in adults, but also in clinical behavior of disease and outcomes¹⁰. SHH-inhibitors such as vismodegib and sonidegib have shown promising results in early clinical trials.^20,38–40 EORTC 1634-BTG/NOA-23 trial, investigating the addition of sonidegib to upfront therapy, and the Alliance AMBUSH study, investigating maintenance sonidegib, are two trials in development to investigate the use of targeted therapy in the management of patients with SHH subgroup MB.⁴¹ Long-term survivorship series have shown late morbidity with neurocognitive deficits, psychological distress, and impaired quality of life.^4,16,18 With improved biomarkers, personalized treatment can lead to improvements in both survival outcomes and treatment-related toxicity.

Conclusion

This is the largest contemporary single institutional experience of adults with MB to date, showing favorable survival outcomes similar to pediatric patients with MB. The favorable overall survival in our cohort suggests a benefit of multi-modality therapy used in treating the majority of these patients. Many patients will achieve long-term survival after treatment, while select patients are at risk for late recurrences. Long-term survivors exhibit significant and diverse sequelae of treatment such that future survivorship considerations should include attention to patients’ long term emotional/coping needs and incorporate social support among other multi-disciplinary care. Further research is needed to identify prognostic factors to better define personalized treatments for adults with MB.

Funding

None declared.

Conflict of interest statement. TIY reports in kind support from MIM software for pediatric radiation registry. WTC reports board of director for IMRIS, Deerfield Ventures. DAF reports minority shareholder of Eli Lilly, LLC. NJT reports Mevion (advisory board). JSL reports co-editor of Neuro-Oncology section of UpToDate, Advanced Oncolotherapy (MAB), Mevion (advisory board).

Authorship statement. Study Design: AS, HAS, AN. Implementation and analyses (including statistical): AS, HAS, AN. Interpretation of data: all. Manuscript writing, feedback, and revisions: all.