Comment

www.thelancet.com/oncology Vol 14 November 2013 1147

Genetics of recurrent medulloblastoma

Published OnlineOctober 17, 2013http://dx.doi.org/10.1016/ S1470-2045(13)70482-0

See Articles page 1200

Zeph

yr/S

cienc

e Ph

oto

Libr

ary

Medulloblastomas are tumours of primitive neuro-ectodermal origin that arise in the cerebellum and disseminate through the cerebrospinal fl uid in the leptomeningeal space. Clonal genetic selection has identifi ed four main subgroups1 with distinct clinical characteristics: group 1 tumours show mutations in SHH and its receptors; group 2 tumours are driven by changes in WNT signalling, generally through its main eff ector gene, β-catenin; group 3 tumours are driven by changes in TGFβ–OTX2 signalling; and group 4 tumours are driven by mutations MYC and MYCN.2

In The Lancet Oncology, Vijay Ramaswamy and colleagues3 assess the genetic and clinical characteristics of these genetically defi ned subgroups in recurrent medulloblastoma. Importantly, they show that recurrent tumours belong to the same genetic subgroup as the primary tumour. This fi nding contrasts with other solid brain tumours, such as glioblastoma, in which genetic subgroup affi liation can change at recurrence, with patterns linked to tumour location.4 Another key fi nding is that SHH tumours almost always recur locally (eight of nine cases; 89%), whereas metastatic recurrence is common in tumours from groups 3 and 4 (17 of 20; 85%). As the investigators note, this fi nding suggests that additional therapies for SHH tumours should be aimed at the posterior fossa, whereas for patients with tumours of group 3 or 4, the focus should be on metastatic disease.

Ramaswamy and colleagues also identifi ed an increased incidence of local recurrence of group 4 tumours in non-irradiated patients. Existing clinical practice is not to irradiate infants with group 4 tumours, and further investigations will be necessary to show whether diff erent therapeutic approaches can prevent local recurrence in these patients.

An understanding of genetic affi liation through matching data from primary and metastasic tumours in the same patient will be key to enhancing treatment options for medulloblastoma. Further work to identify the gene signatures of these four independent tumour groups will be crucial, especially in relation to mutations that give metastatic properties to malignant cells. Data presented by Pugh and coworkers,5 who used whole-exome hybrid capture and deep sequencing to identify somatic mutations in the coding regions of medulloblastomas, might show the way forward.

Key questions remain, such as determining how brain tissue behaves in relation to the tumour micro-environment and how do the genetically driven subgroups aff ect clonigenic recurrent tumorigenic cells? The question of how remaining cancer cells can be eradicated is potentially the most important of all.

Additional issues are relevant to groups 3 and 4, which are the most prevalent subgroups, accounting for 40% of medulloblastomas. Investigations are needed to identify genes and processes involved in metastases in primary disease, and whether these are the same genes that are activated in recurrent tumours. Important questions will then be whether it is possible to design drugs against these particular functions, and whether these agents will be suffi cient to suppress the tumour’s metastatic properties. Present clinical trials for previously irradiated relapsed metastatic medulloblastoma are very heterogeneous and are commonly phase 2 trials. Refi nement of therapies for patients with tumours of groups 3 and 4 should focus on metastatic disease, because it is the near universal cause of death; trials in these patients might fail if only the primary tumour is taken into account. Importantly, clinical studies should consider chemoimmunotherapy. In melanoma, trials with ipilimumab and dacarbazine,6 abscopal radio therapy,7 or ipilimumab and fotemustine,8 have shown positive eff ects on brain metastases.

The release of tumour antigens, together with amplifi cation of T-cell regulatory actions within the blood–brain barrier, will contribute to an environment in which malignant progression is possible.9 Epi-genetics will also aff ect the recurrence pattern in medulloblastoma, silencing genes that promote remission of disease. Thus, future genetic analyses in medulloblastoma must also consider the epigenotype.10 However, these analyses are only the start, there is an urgent need for improved subgroup-specifi c therapies for recurrent medulloblastoma.

Massimo Zollo Department of Molecular Medicine and Medical Biotechnologies, University Federico II, 80131 Naples, Italy; and CEINGE, Biotecnologie Avanzate, 80145 Naples, Italymassimo.zollo@unina.it

Comment

1148 www.thelancet.com/oncology Vol 14 November 2013

Published OnlineOctober 11, 2013

http://dx.doi.org/10.1016/S1470-2045(13)70478-9

This online publication has been corrected.

The corrected version fi rst appeared at

thelancet.com/oncology on October 28, 2013

See Articles page 1208

Zeph

yr/S

cienc

e Ph

oto

Libr

ary

Chemotherapy and resection for colorectal metastasesThe past two decades have seen the outlook for patients with hepatic colorectal metastases improve substantially, from near-certain death within 1 year to more than 40% of patients being cured.1 Chemotherapy plays a major part in this improvement in outcome; 15–20% of patients with unresectable disease at diagnosis were reported to have resectable disease after receiving neoadjuvant chemotherapy, with resultant favourable long-term survival.2 Chemotherapy is also used in the adjuvant setting to improve outcomes in patients with advanced, resected metastatic colorectal cancer.3 Nevertheless, routine use of neoadjuvant chemotherapy remains controversial,4 as does the choice of chemotherapy to give after hepatectomy.

The European Organisation for Research and Treatment of Cancer (EORTC) Intergroup 40983 phase 3 study, which randomly assigned 364 patients with resectable liver metastases to surgery alone or neoadjuvant and adjuvant FOLFOX4 (folinic acid, fl uorouracil, and oxaliplatin), is a landmark study that partly addresses these questions. The initial results from this trial, published in 2008,5 showed that perioperative FOLFOX4 signifi cantly improved progression-free survival compared with surgery alone, and guided clinical practice. The long-term overall survival results, reported by Bernard Nordlinger and colleagues6 in The Lancet Oncology, have been greatly anticipated. The major fi nding of this follow-up study was that overall survival was not signifi cantly greater in patients who received perioperative FOLFOX4 than in those who received surgery only. At a median follow-up of

8·5 years, 107 (59%) of 182 patients in the perioperative chemotherapy group had died versus 114 (63%) of 182 patients in the surgery-only group (hazard ratio 0·88, 95% CI 0·68–1·14; p=0·34). This result is surprising and should be interpreted carefully.

Most importantly, adjuvant chemotherapy should not be disregarded in the treatment of metastatic colorectal cancer. Third-party payers and national regulatory agencies that control access to health-care resources should not deny access to adjuvant chemotherapy solely on the basis of the outcome of this study. As discussed well by Nordlinger and colleagues,6 their follow-up study was underpowered to prove the signifi cance of the 5% between-group diff erence in 5-year overall survival. Furthermore, only 115 (63%) patients in the perioperative chemotherapy group received postoperative chemotherapy. Thus, the inability to show a signifi cant diff erence in overall survival might be a result of inadequate duration of chemotherapy, especially since adjuvant chemotherapy has previously been shown to be important in the treatment of resected, advanced stage colorectal cancer.3

Additionally, Nordlinger and colleagues6 report that second-line treatment with chemotherapy for disease progression was given more frequently to patients in the surgery-only group than in the perioperative chemotherapy group. An interesting question, therefore, is whether the fi ndings mean that administration of chemotherapy at minimal recurrence of disease is just as benefi cial as administering it

I am supported by the AIRC Associazione Italiana per la ricerca sul Cancro. I declare that I have no confl icts of interest.

1 Gibson P, Tong Y, Robinson MC, et al. Subtypes of medulloblastoma have distinct developmental origins. Nature 2010; 468: 1095–99.

2 Northcott PA, Shih DJ, Peacock J, et al. Subgroup-specifi c structural variation across 1,000 medulloblastoma genomes. Nature 2012; 488: 49–56.

3 Ramaswamy V, Remke M, Bouff et E, et al. Recurrence patterns across medulloblastoma subgroups: an integrated clinical and molecular analysis. Lancet Oncol 2013; published online Oct 17. http://dx.doi.org/10.1016/S1470-2045(13)70449-2.

4 Sottoriva A, Spiteri I, Piccirillo SG, et al. Intratumor heterogeneity in human glioblastoma refl ects cancer evolutionary dynamics. Proc Natl Acad Sci USA 2013; 110: 4009–14.

5 Pugh TJ, Weeraratne SD, Archer TC, et al. Medulloblastoma exome sequencing uncovers subtype-specifi c somatic mutations. Nature 2012; 488: 106–10.

6 Robert C, Thomas L, Bondarenko I, et al. Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N Engl J Med 2011; 364: 2517–26.

7 Postow MA, Callahan MK, Barker CA, et al. Immunologic correlates of the abscopal eff ect in a patient with melanoma. N Engl J Med 2012; 366: 925–31.

8 Di Giacomo AM, Ascierto PA, Pilla L, et al. Ipilimumab and fotemustine in patients with advanced melanoma (NIBIT-M1): an open-label, single-arm phase 2 trial. Lancet Oncol 2102; 13: 879–86.

9 Spano D, Heck C, De Antonellis P, et al. Molecular networks that regulate cancer metastasis. Semin Cancer Biol 2012; 22: 234–49.

10 Batora NV, Sturm D, Jones DT, Kool M, Pfi ster SM, Northcott PA. Transitioning from genotypes to epigenotypes: why the time has come for medulloblastoma epigenomics. Neuroscience 2013; published online July 20. DOI:10.1016/j.neuroscience.2013.07.030.