Less treatment for Wing less medulloblastoma: germline data re-emphasize this

See the article by Surun et al in this issue, pp. 128–138.

Medulloblastoma (MB), the most common malignant pediatric brain tumor, comprises 4 distinct molecular subgroups. WNT/Wingless MB accounts for 10–15%, with a peak incidence in late childhood, are infrequently metastatic, and are associated with an excellent survival in children younger than 16 years.¹ This likely relates to these tumors’ increased sensitivity to current therapies, possibly due to a “leaky” vasculature secondary to the presence of WNT ligand.² However, due to their location, they have a higher risk of cerebellar mutism syndrome following surgical resection.³ As their survival may not be impacted by the residual tumor volume, less aggressive approaches may reduce morbidity.¹ In addition, reducing radiation doses can improve neurocognition without compromising survival.⁴

Around 90% of WNT MBs harbor somatic activating mutations in exon 3 of catenin beta-1 (CTNNB1), which encodes for β-catenin.^1,5 The majority (9–10%) of the remainder harbor loss-of-function mutations of the tumor-suppressor gene adenomatous polyposis coli (APC).⁶ The latter forms a complex with glycogen synthase kinase 3 beta and axin, promoting phosphorylation-dependent ubiquitination and degradation of β-catenin. Upregulation of β-catenin by either mechanism leads to its translocation into the nucleus, where it acts as a coactivator of T-cell factor/lymphoid enhancer factor (TCF/LEF) transcription factors, regulating genes involved in cell cycle, apoptosis, and differentiation. APC can also bind and enhance the export of β-catenin from the nucleus, thereby blocking its interaction with TCF/LEF.⁷

Being mutually exclusive, testing for APC mutations in the germline is recommended for WNT MBs lacking CTNNB1 mutations.⁶ Germline mutations in APC are associated with familial adenomatous polyposis (FAP) syndrome, a rare (1:7000–30 000 live births), autosomal dominant cancer predisposition syndrome. It is highly penetrant but may have variable phenotypic expression linked to mutations in specific regions of the relatively large APC gene. In addition to WNT MBs, germline APC mutations predispose to precancerous intestinal polyps, thyroid neoplasms, hepatoblastoma, pancreatic carcinoma, adrenal tumors, osteoma, fibromas, pilomatrixomas, desmoid tumors, as well as epidermal cysts, dental anomalies, and congenital hypertrophy of the retinal pigment epithelium.⁸ Although 20–30% of mutations arise de novo, screening and surveillance of individuals and “at-risk” family members is highly recommended. It is important to emphasize that older phenotypic terminologies like Gardner and Turcot syndromes should be avoided, as the former is a part of the FAP spectrum, while the latter includes cases having constitutional mismatch repair deficiency (type 1), as well as those within the FAP spectrum (type 2).⁸

Although the association between FAP and MB has been known for over 20 years, the prognosis of these patients is currently indeterminant. As such, efforts to de-escalate therapy for WNT MBs have so far not taken this small but important group into account. To address whether sporadic and inherited WNT MBs share the same excellent prognosis, the French cooperative group reports a retrospective multicenter review of 12 patients with germline APC-mutated MB (proven/ highly suspicious due to family history) diagnosed over 30 years.⁹ The authors report an excellent survival for FAP associated MB, despite the heterogeneity of the treatment regimens over the decades, consistent with previous work demonstrating that WNT MBs have an excellent survival across multiple treatment regimens. Although detailed molecular profiling was not available in all cases, a strong nuclear immunopositivity for β-catenin in 75% of cases was highly suggestive of these tumors being WNT MB. Not surprisingly, they report a high risk of second neoplasms, many of which may have been treatment related (radiation and/or surgery). The only mortality was related to radiation-induced malignant triton tumor. Coupled with the almost universal late neurocognitive sequelae associated with craniospinal irradiation, the burden of achieving cure seems significant, mandating consideration for de-escalation of treatment.

Waszak et al had previously reported on germline APC mutations in 1% of all MB (n = 7/1022).⁶ They included 5/68 (7%) children with WNT MB, 1/236 (0.4%) infant with sonic hedgehog MB and retention of the wild-type allele, and 1 adult (molecular subgroup not available). Most importantly, germline APC mutations were mutually exclusive with somatic CTNNB1 mutations. Monosomy 6 was detected in 3/5 (60%) and was not useful in differentiating somatic from germline WNT MBs. Additional malignancies were reported in 43% of those with APC germline mutations, highlighting that the excellent MB-free survival is significantly diminished by second neoplasms. Taken together, the studies by Surin et al and Waszak et al provide a rational approach for identifying germline predisposition within patients diagnosed with WNT MB (Fig. 1).

Fig. 1.

A diagnosis of WNT MB, confirmed by a combination of strong nuclear immunopositivity for β-catenin and/or presence of monosomy 6 by fluorescence in situ hybridization, should be followed by sequencing of the tumor for CTNNB1 mutation. Absence of a somatic CTNNB1 mutation should prompt germline testing for APC mutation. Presence of APC mutation should lead to enrollment in WNT MB specific trials, surveillance for second neoplasms, and genetic counseling. Both germline and somatic WNT MB share the same excellent prognosis following current multimodal therapy, albeit with neurocognitive sequelae. However, the overall survival continues to diminish beyond 10 years from the diagnosis due to higher risk of second malignancies, both related to the APC tumor predisposition syndrome, as well as due to therapy-related neoplasms.

Second malignancies in a high proportion of FAP associated WNT MBs further support our current attempts to de-escalate therapy for the entire group. Ongoing studies are attempting to reduce the dose of radiation to either 15 Gy or 18 Gy in WNT MB. However, a study of chemotherapy only following surgery had to be prematurely discontinued due to an unacceptable number of relapses.¹⁰ The observations from Surin et al, confirming the excellent prognosis in APC-mutated WNT MB, suggest that further efforts at reducing radiation and alkylating agents in therapy should continue, and would strongly benefit this group.

Desmoid tumors, given their local invasiveness and unpredictable aggressive behavior, are the second important cause of morbidity and mortality in FAP. Surgery is a well-established risk factor for developing desmoid tumors, especially in the presence of APC germline mutations.¹¹ Recent advances in radiogenomics and sequencing of cell-free DNA has the potential to allow for robust diagnosis of MB subgroups without surgery. As such, applying nonsurgical diagnosis of WNT MBs may allow for a surgery-free approach, and avoid the morbidity of mutism for all WNT MBs and the risk of desmoid tumors in germline APC-mutated patients.

In conclusion, the report by Surin et al reiterates the importance of diagnosing germline APC mutations in WNT MBs, specifically in those lacking somatic CTNNB1 mutations. Rarity of this entity highlights the need for international cooperation, beyond collaborative group trials, to devise the best strategies to minimize the late treatment–related mortality and morbidity in patients with an otherwise curable malignancy.