Abstract
Rhabdomyosarcoma (RMS) is a well-described cancer in Li-Fraumeni Syndrome (LFS), resulting from germline TP53 pathogenic variants (PVs). RMS exhibiting anaplasia (anRMS) have been associated with a high rate of germline TP53 PVs. This study provides an updated estimate of the prevalence of TP53 germline PVs from a large cohort of patients (n=239) enrolled in five Children’s Oncology Group (COG) clinical trials. Although the prevalence of germline TP53PVs in anRMS patients in this series is much lower than previously reported, this prevalence remains significantly elevated. Germline genetic evaluation for TP53 PVs should be strongly considered in patients with anRMS.
INTRODUCTION
Rhabdomyosarcoma (RMS) is the most common soft tissue sarcoma of childhood1. An association between RMS and other early-onset cancers was first described in 19692, became known as Li-Fraumeni Syndrome (LFS), and was subsequently shown to be caused by germline pathogenic variants (PVs) in the TP53tumor suppressor gene3. Evidence suggests that RMS patients who harbor germline TP53 pathogenic variants (PVs) are more likely to exhibit anaplasia (anRMS)4. In 2014, Hettmer et al. reported in a retrospective limited series that the overall frequency of germline TP53 PV in pediatric patients with anRMS was 73% (11 of 15 cases)5. These findings formed the basis of recommendations to test for TP53 germline PV in patients with anRMS in the most recent Chompret Criteria and were adapted for testing criteria by the National Comprehensive Cancer Network (NCCN)6, 7.
We sought to expand on the prior analysis of anaplasia and germlineTP53 PV and to avoid the potential for selection bias inherent in a limited institution study. We investigated a large RMS study population derived from Children’s Oncology Group (COG) clinical trials which included central pathology review for the presence of anaplasia and germline TP53 PV status from exome sequencing8. From this cohort, we estimated the prevalence of germline TP53 PVs with and without anaplasia, as well as assessed TP53 PV associations with other RMS tumor and patient characteristics.
METHODS
The study population was derived from patients enrolled on one of the five COG clinical trials for which central pathology review was performed prospectively to determine the presence of anaplasia and for which germline TP53 data from exome sequencing was recently reported8. Two hundred and thirty-nine patients were identified from the five following COG studies: D9602 (n=18), D9802 (n=14), D9803 (n=29), ARST0331 (n=61), ARST0531 (n=117).
RESULTS
TP53 germline status, tumor histology, anaplasia status, sex of patient, age at diagnosis, primary tumor site, tumor size, nodal status, and tumor FOXO1 fusion status were evaluated. Histology, including anaplasia status, was performed through central review by three expert pediatric pathologists. Anaplasia was defined as the presence of enlarged hyperchromatic nuclei with or without multipolar mitotic figures9, 10. Focal anaplasia was defined as anaplastic cells loosely scattered among non-anaplastic cells, whereas diffuse anaplasia was defined as anaplastic cells that were aggregated in clusters or that formed continuous sheets9. Exome sequencing, variant filtering, and identification of TP53 PVs, followed methods described in Li and colleagues8.
The Fisher’s Exact test was used to examine the association between categorical characteristics. The two-sample t-test was performed to compare age at diagnosis between TP53 PV and wild type (WT) cohorts. Statistical significance was considered at the 0.05 level. The software SAS 9.4 was used for analysis.
The prevalence of germline TP53 PVs among the entire RMS cohort was 3% (n=7/239). The median age of diagnosis was 2.8 years (range 0.9 to 3.7 years) and 5.8 years (range 0.2 to 28.3 years) (P=0.0003) for theTP53 PV and the TP53 WT patients, respectively. Among the entire cohort, histology was classified as alveolar (n=73), embryonal (n=122), botryoid (n=25), spindle cell (n=15), mixed (n=1), not otherwise specified (n=2), and unknown (n=1) (Table 1). Among patients with germline TP53 PVs, histology included: embryonal (n=3), botryoid (n=2), and spindle cell (n=2); none had alveolar histology. There was a statistically significant difference in histology type between the germline TP53 PV patients and those without PVs (P=0.04).
Similar to other clinical reports10, anaplasia was present in 19% (n=46) of the 239 patients (Table 2); 34 of which had diffuse anaplasia while 12 demonstrated focal anaplasia s (Table 3, supplement). Among the 46 patients with anRMS, 11% (n=5) carried a germline TP53 PV compared to 1% (n=2) among patients without anaplasia (P=0.003). The proportion of TP53 PVs in those with diffuse anaplasia and focal anaplasia were 9% (n=3) and 17% (n=2), respectively. Among the seven patients with TP53 PV patients, 71% (5/7) exhibited anaplasia.
DISCUSSION
Using a population of 239 patients, this report provides the most comprehensive estimate to date of the prevalence of germline TP53PVs in pediatric patients with anRMS. We reconfirm the association between RMS and germline TP53 PVs, and specifically, the association between anRMS and germline TP53 PVs. However, our findings suggest that the prevalence of TP53 PVs in anRMS is lower than previously estimated5. Specifically, we found in our combined cohort that approximately 11% of patients with anRMS have germline TP53 PVs compared to the prior description where 75% anRMS patients had TP53 PVs5. It is possible that smaller sample size and selection bias inherent to a limited institution study explains the discordance between our results. Furthermore, we describe the rate of a germline TP53 PV in RMS without anaplasia to be approximately 1%.
Consistent with other reports5, 11, we observed a high prevalence of anaplasia (71%) in patients harboring germlineTP53 PVs. We found a statistically significant difference in median age at diagnosis between patients with germline TP53 PVs and TP53 WT, in line with previous published reports5, 11, 12. A detailed comparison between tumor and clinical features in the germline TP53 PV versus TP53WT cohorts revealed no other significant differences.
Limitations to our current study include the retrospective nature of this investigation and the lack of reporting of heterozygous deletions which may make up 1-5% of TP53 PVs. Additionally, the number of cases was limited to those in which we had both anaplasia andTP53 PV information, and therefore could bias the true prevalence of TP53 germline PVs in anRMS.
Although the prevalence of TP53 PV in anRMS is lower than prior reports and leads to a decrease in the pretest probability in screening for germline TP53 PVs in anRMS patients, the 11% prevalence of germline TP53 PVs in anRMS still exceeds the threshold for recommendation for germline genetic evaluation13. This study provides the ongoing rationale to evaluate germline TP53 in the context of RMS so that oncologists can improve the clinical approach to RMS patients and optimize screening practices for LFS in the RMS population.