Discussion
APC somatic mutations are not generally found in sporadic MBL, but this tumor may be associated with patients harboring the APC germline pathogenic mutations that cause FAP syndrome17. APC gene mutations generally start from a KRAS proto-oncogene point mutation with DNA hypomethylation activating the proto-oncogene. This is followed by a loss of the APC alleles that serve as a tumor suppressor gene. MBL associated with FAP syndrome comes from neural stem cells containing the homozygous APC mutation18.
MBL accounts for 80% of the brain tumors found in patients with FAP, a condition associated with a 7-fold relative lifetime risk of any brain tumor, and with a 90 times higher risk of MBL19.
The molecular subtype in the context of FAP is Wnt/Wg. APC mutations are involved in around 7% of Wnt/Wg molecular subtype MBL20-25. It is noticeable that in our series, some cases despite APC germline mutations do not show nuclear β‐catenin accumulation, but it must be underlined that there are some limits of immunohistochemistry which make the technique not always reliable26-27.. These patients’ radiological features were superimposable on those reported by Patay in the context of Wnt molecular subgroup MBL14 and later confirmed by other authors28. There are no guidelines for brain tumor surveillance in the context of FAP, probably because there are still no registries available to consult that reliably show how many people with inherited or sporadic colorectal cancer develop MBL, or vice versa29. In members of a family with FAP who do not yet have polyposis, neurological symptoms should prompt radiological examination of the CNS because brain tumors manifest before the polyposis is diagnosed in more than one in two FAP patients30. This was also the case in two of our six patients. A major effort by the SIOP to move in this direction is in progress as part of the PNET5 trial12, with the creation of a registry for all patients with MBL harboring constitutional genetic mutations, including APC.
What we are still lacking, however, is therapeutic guidelines that take into account both the good outcomes reported in a handful of more recent papers15 and the genetic fragility of these patients. They spontaneously develop many different types of tumor and are obviously more susceptible to secondary tumors after adjuvant treatment for MBL. It is sadly remarkable that the only death of a patient, in the series recently reported by a cooperative French group15, was due to a triton tumor 14 years after MBL had been diagnosed.
As already underlined, patients with Wnt-MBL are currently treated with de-escalating protocols aiming at maintaining good outcome with reduced late effects12-13. This FAP patient subgroup is even more fragile, deserving special attention in their disease story. It is moreover to note that a very recent paper has reported a meta-analysis on Wnt -MBL relapses showing the usefulness of adjuvant cyclofosfamide in reducing relapses31.
In our smaller but single-institution series, apart from desmoid and thyroid tumors, three patients developed other three tumors possibly unrelated to FAP, despite having received lower doses of CSI than the French series, and standard chemotherapy.
Few data characterize the cognitive abilities of individuals with FAP but the role of the APC protein in the development of the central nervous system is well known32 and a very recent paper has underlined the significantly lower performance on IQ and in a variety of neurocognitive functions in a population of 24 subjects compared to non-FAP pairs33. The results of our small series in term of cognitive performances, FIQ and social life are therefore not only to be attributed to MBL and its treatment but also to the persasive activity of FAp 33
Collecting these now better-defined cases of MBL in patients with FAP will broaden our understanding of their MBL outcome and hopefully enable their better-tailored treatment and rehabilitation.
Disclosure statement . No relevant interest to be disclosed