A wide range of prevalences of TP53
mutations in RMS were reported in the literature, with larger studies tending to report lower mutation frequencies. To address these inconsistencies, we have analyzed TP53
mutations in exons 4–9 in 75 pediatric RMS tumor samples, as the majority of TP53
mutations occur in exons 5–8. We detected only a single mutation in exon 7 (c.775G/T, p.D259Y), associated with loss of p53 activity [30
]. This occurred in a 1-year-old patient with ERMS. ERMS are characterized by an earlier age of onset compared to ARMS, with approximately half of all cases diagnosed before the age of 5 years. Interestingly, the tumor of this patient was located in the lower extremity, a site that is more frequently associated with ARMS tumors.
All tumor samples in this study came from the CHTN, which includes a central pathology review. However, the majority of samples did not have accompanying data on the percentage of tumor present in each sample or the prevalence of necrosis. We therefore performed hematoxilin staining of the available histopathology slides and reviewed all histopathology slides to complete these data. This helped us address the possibility that the observed low prevalence of mutations in these tumors may be due to a large content of normal tissue compared to tumor tissue in our samples, which might undermine the detection of mutations that may be present in tumor. However the pathology review determined that, except for two samples, that had approximately 60% of the sample composed of tumor tissue, at least 80% of all other samples was tumor tissue. All reviewed samples had over 90% viable cells present. Therefore, it is unlikely that the composition and integrity of our samples may have affected the observed low prevalence of TP53 mutations.
Differences in the methods used to identify mutations in earlier compared to more recent studies may partially account for the observed differences in mutation frequencies. All earlier studies used single strand conformation polymorphism (SSCP), a method shown to more frequently detect TP53
mutation detection, compared to sequencing, which is current gold standard [31
Some clinicopathologic characteristics of our samples were in line with well-established differences between the two major RMS subtypes, including differences in primary sites, and equal male to female ratio among ARMS cases [2
]. In contrast, earlier age of onset of ERMS compared to ARMS was not reflected in our samples (). Despite slightly older age of ERMS cases than anticipated, the proportion of very young children (up to 5 years of age) was substantial (34/75), indicating that the age may not explain the observed low mutation prevalence. Namely, among germline TP53
carriers, RMS occured very early in life and many developed RMS before the age of 5 years [33
]. By analogy, if sporadic RMS tumors harboring TP53
mutations were more likely to occur in younger children, a study not including this age category would likely report lower prevalence of TP53
mutations. There was a disproportionately higher number of males than females with ERMS in our study compared to the reported 1.3 male to female ratio. As we observed no gender differences, gender distribution was unlikely to affect our results.
To explore whether increased levels of MDM2 may provide an alternative mechanism to TP53
gene mutation, we have analyzed MDM2
gene amplifications. Overexpression or amplification has been observed in a number of human cancers [34
] suggesting that it may act as an alternative mechanism to attenuate wild-type p53 function. Although MDM2
amplifications were frequently observed in soft tissue sarcoma [34
], the only study examining this in pediatric RMS has shown very low prevalence of MDM2
amplifications (2 out of 20 samples) [22
]. In addition, a study that included both pediatric and adult RMS patients showed a similar prevalence (3 out of 18 patients) [21
]. Our study confirms the finding that the prevalence of MDM2
amplifications in pediatric RMS may be low, as we detected no MDM2
amplifications in the 22 samples we analyzed. In contrast to the previous suggestion that MDM2
amplifications may be more likely found in ARMS tumors compared to ERMS [21
], we observed no amplifications in either tumor subtype.
Among the variants that affect p53 degradation, MDM2
SNP309 was associated with earlier age of cancer onset among germline TP53
mutation carriers [10
]. In addition, this SNP was reported to be associated with younger age of onset of sporadic soft tissue sarcoma [11
], but this study included insufficient number of RMS samples to determine whether this association would be also observed for RMS. In contrast, our study included a much larger number of pediatric RMS samples, and we found no association between MDM2
SNP309 and younger age of tumor onset. Likewise, TP53
SNP72 was reported to be associated with age of tumor onset in carriers, where it was also shown to interact with MDM2
SNP 309 to further reduce the age of tumor onset. Such interactions have also been described in several sporadic human malignancies; therefore, we examined whether TP53
SNP72 may act as a modifier in RMS. We found no effect of TP53
SNP72 on the age of tumor diagnosis, nor any evidence of interaction between TP53
SNP72 and MDM2
SNP309 on the age of tumor onset.
Although our study was relatively large, it was focused on pediatric RMS. Therefore, the range of ages at tumor onset was narrower compared to other studies that reported modifying effects of the variants affecting p53 degradation in either germline TP53 carriers or other types of sporadic cancer. In addition, molecular characteristics of RMS tumors that occur in children may differ from those that occur in adults. Therefore, it remains possible that any modifying effects of the variants analyzed here would have been observed if adult cases had been included in our study. Further studies involving both pediatric and adult RMS patients would be needed to explore this.
In conclusion, we observed low prevalence of TP53 mutations and no MDM2 amplifications in pediatric RMS. This is different from what has been found before, and due to the sample size, ours is likely a more accurate finding. Variants TP53 SNP72 and MDM2 SNP309 did not accelerate tumor development. TP53 may therefore not play an important role in pediatric RMS development. The possible inactivation of p53 function by other mechanisms thus remains to be elucidated.