Molecular, genetic factors influence treatment response in children, adolescents with medulloblastoma

Hyperfractionated craniospinal radiotherapy, combined with induction and maintenance chemotherapy, conferred positive OS outcomes in children and adolescents with metastatic medulloblastoma, according to prospective study results published in Journal of Clinical Oncology.

Further, researchers identified several biologic and treatment-related parameters — including biologic subgroup status, response to induction chemotherapy and MYCC/MYCN status — that may improve treatment stratification.

Medulloblastoma represents 20% of pediatric brain tumors, and roughly one-third of children have metastatic disease at diagnosis.

The condition is associated with generally poor outcomes, with historical 3- to 5-year EFS rates below 50%. However, 5-year EFS rates have risen above 60% in recent years, due in part to new treatment strategies.

“In the precursor trial HIT ’91, patients aged 3 years to 18 years were randomly assigned to receive postoperative chemotherapy followed by craniospinal irradiation (sandwich strategy) vs. postoperative craniospinal irradiation followed by chemotherapy (maintenance strategy),” Stefan Rutkowski, MD, director of the department of pediatric hematology and oncology at University Medical Center Hamburg-Eppendorf, and colleagues wrote. “We observed a moderate trend toward better survival for children with M2/M3 disease treated with the sandwich strategy compared with the maintenance strategy; therefore, a sandwich option concept was chosen for the subsequent HIT 2000 trial.”

The current trial protocol included the introduction of an intensified neoadjuvant systemic and intraventricular chemotherapy regimen; intensified hyperfractionated craniospinal radiation therapy; and four cycles of maintenance chemotherapy.

The study included data from 123 patients (median age, 8.2 years; range, 4-21; 76.4% male) diagnosed with metastatic medulloblastoma between 2001 and 2007.

Median follow-up was 5.38 years (range, 1.15-9.48).

Sixty-two percent (95% CI, 52-72) of the study population achieved 5-year EFS and 74% (95% CI, 66-82) achieved 5-year EFS.

Children and adolescents with desmoplastic or nodular medulloblastoma (n = 11) had the best EFS and OS outcomes (89%; 95% CI, 67-100 for both). Children and adolescents with classic medulloblastoma — which represented the majority of patients (n = 107) — had an EFS of 61% (95% CI, 51-71) and OS of 75% (95% CI, 65-85).

Poorer outcomes occurred among children and adolescents with large-cell or anaplastic medulloblastoma (n = 5), who had an EFS of 20% (95% CI, 0-55) and an OS of 40% (95% CI, 0-83).

When researchers compared outcomes of patients with metastatic medulloblastoma treated in the two consecutive trials (HIR 2000, n = 119; HIT ’91, n = 62), they found 5-year OS rates in the current study surpassed outcomes presented in the HIT ’91 trial (73% vs. 55%; P = .037), whereas EFS outcomes were similar in both trials (60% vs. 47%).

Independent EFS prognostic factors identified by a multivariate analysis included histology (desmoplastic or nodular medulloblastoma, HR = 0.19; large-cell or anaplastic medulloblastoma, HR = 45.57) and lack of response to first induction chemotherapy (HR = 1.97).

Sixty-six percent of patients (n = 81) had tumor material available for immunohistochemical and gene amplification analysis. Based on these analyses, researchers identified patients as low risk (WNT pathway activation, n = 4), high risk (MYCC/MYCN amplification, n = 5) and intermediate risk (neither, n = 72).

EFS and OS outcomes varied by risk stratification. Low-risk patients had 5-year EFS and OS outcomes of 100%, whereas high-risk patients had 5-year EFS and OS rates of 20%. Intermediate-risk patients had EFS outcomes of 63% and OS outcomes of 73%.

Sixty-nine patients (56%) had paraffin material available for molecular subgroup analysis (WNT, n = 4; sonic hedgehog [SHH], n = 4; group 3, n = 20; group 4, n = 41). EFS and OS outcomes did not significantly differ between these groups. However, survival rates differed in a model based on five molecular groups: WNT, SHH, group 4, and group 3 with or without MYCC/MYCN amplification.

The researchers did not observe a significant association between treatment nonresponse and WNT, MYCC or MYCN amplification, molecular subgroup or histologic subgroup.

Forty-seven patients experienced disease progression or recurrence, the majority of whom (n = 34) had distant metastases.

The researchers acknowledged the small patient population, particularly among patients with molecular subgroups or genetic alterations, as a potential limitation.

“Our prospective data from homogeneously treated patients with clinically and biologically well-defined, high-risk medulloblastoma provide evidence for improved risk-adapted treatment stratification by clinical, biologic and histologic parameters into a favorable risk group (metastatic WNT patients), a high-risk group (group 3 with MYCC or MYCN amplification [need to be considered very high risk] and nonresponders to induction chemotherapy) and a large group of intermediate-risk patients (responders to induction chemotherapy without other low- or high-risk criteria),” Rutkowski and colleagues wrote. “Long-term treatment-related morbidity and neurocognitive outcomes need to be further addressed.” – by Cameron Kelsall

Disclosure: Rutkowski reports research funding from Riemser Pharma and a consultant role with Novartis. Please see the full study for a list of all other researchers’ relevant financial disclosures.