This is the first study in Colombia attempting a molecular classification using both histology, PCR, and immunohistochemistry.
We analyzed 49 cases of medulloblastoma treated in the reference pediatric healthcare facility in the city of Bogotá, Colombia. Most of the clinical, demographic, histological, and molecular characteristics of patients with medulloblastoma described in the present study were similar to those reported in the literature. However, the ratio between males and females was 1:1.34, which is contrary of what was described by the WHO and other groups. We hypothesized that this finding is in tune with population distribution in Colombia. (23;24)
This cohort of patients had the particularity of having a girl with a history of Turner Syndrome and two patients with neurofibromatosis type 1, one of whom also had a diagnosis of Noonan-Neurofibromatosis Syndrome. With regard to these diagnoses, the risk that RASopathies entails for the development of neoplastic pathologies has been widely described, specially solid tumors such as gliomas; however, the development of medulloblastomas is an infrequent finding. Likewise, Turner syndrome does not have a clear causality in the pathophysiology of neoplasms. (9)
The most frequently described histological variant in this cohort was the classic one in 55% of the samples, followed by desmoplastic and extensive nodularity (20% each), which is comparable with the findings of Kools et al., however, these results are discordant with those reported in the literature and in the Gottardo et al., Eid et al., and Ellison cohorts, where the most frequent variant is anaplastic followed by classic and extensive/desmoplastic nodularity. (8;10;11,18)
The most common molecular subgroup was the SHH group, 63% of the total cases, followed by the 3/4 group, which we were unable to separate (30%). Only one tumor was classified as WNT (2%). The remaining case was negative for all immunohistochemistry markers.
We classified cases of the SHH group into p53 Wild Type and with possible variants in p53. Regarding histological classification, the classic variant was the most common in our series. These results are different from those reported by Kool et al. Eid et al. Taylor et al. and similar to the brazilian and russian series. (8;15;18;20;21;36)
These 3/4 groups are associated with metastasis. In our group, 3/4 medulloblastoma has a 2-year survival of 50%. In the literature, a survival of 3/4 group is nearly 90% (33). In general, the survival of children with cancer in Colombia is lower than the survival in developed countries, which can be explained by several socioeconomic factors. (16)
One of our cases was considered indeterminate. We believe it could belong to the embryonal tumor family. BCOR, FOXR2, and other genetic alterations of different members of this tumor group should be explored.
The chosen markers were mostly accurate, except for TACI. The expression of the YAP-1, PIGU, and p75-NGFR markers was implemented for the identification of the SHH group, with reactivity for at least 2 of the markers described, a panel of markers similar to those used in the group of Shuangshoti et al. (34) Betacatenin, YAP-1, p75-NGFR, OTX2 and YAP-1, p75-NGFR, OTX2; the subclassification SHH p53 with variant, presented a nuclear reactivity greater than 50%, with a p < 0.001, a result comparable with the cohort studied by Tabori et al., where sensitivity was determined for immunohistochemistry in medulloblastomas with p53 labeling of 100% with a specificity of 83% and evidence of mutated p53 was associated with a worse survival (0% at 5 years) and greater recurrence.(35)
In the determination of subgroup 3/4, a reactivity greater than 10% of OTX2 was evidenced for a p = 0.04, which was presented together with the absence of YAP1 reactivity and reactivity for Betacatenin and the absence of reactivity for YAP1. MYC amplification was performed in 34 samples that had sufficient material to carry out the PCR. Of these patients evaluated, C-MYC amplification was found in two patients aged 5 and 6 years with molecular subclassification by immunohistochemistry for group 3/4 and N-MYC amplification in a single 2-year-old patient classified as group 3/4 with p53 variant, variants previously described in the cohort of Hill et al. in rapidly progressive medulloblastomas.(14;38)
It is worth highlighting the implementation of this panel of six immunohistochemical markers and the amplification of MYC by PCR technique, which allowed the efficient classification in the different molecularly defined groups; however, secondary to the reactivity of immunohistochemical markers, it is considered necessary to validate the results obtained by running a molecular test that allows evaluating the status of N-MYC, MYC, and CTNNB1. (29;38)
This study proposes an affordable panel of techniques, allowing a precise classification of medulloblastoma. The immunohistochemical panel has shown strength for the determination of the molecular group (19;26;37). However, our results show differences compared with the main series of the world. Although it could be related to specific characteristics in our population or its ancestry, we believe is necessary to validate all our results by sequencing.