The TP53 gene remains the most frequently altered gene in human cancer, of which variants are associated with cancer risk, therapy resistance, and poor prognosis in several tumor types. To determine the true prognostic value of TP53 variants in non–small cell lung cancer, this study conducted further research, particularly focusing on subtype and tumor stage. Therefore, we determined the TP53 status of 97 non–small cell lung cancer adenocarcinoma patients using next generation deep sequencing technology and defined the prognostic value of frequently occurring single nucleotide polymorphisms and mutations in the TP53 gene. Inactivating TP53 mutations acted as a predictor for both worse overall and progression-free survival in stage II–IV patients and patients treated with DNA-damaging (neo)adjuvant therapy. In stage I tumors, the Pro-allele of the TP53 R72P polymorphism acted as a predictor for worse overall survival. In addition, we detected the rare R213R (rs1800372, minor allele frequency: 0.0054) polymorphism in 7.2% of the patients and are the first to show the significant association with TP53 mutations in non–small cell lung cancer adenocarcinoma patients (p = 0.003). In conclusion, Our findings show an important role for TP53 variants as negative predictors for the outcome of non–small cell lung cancer adenocarcinoma patients, especially for TP53 inactivating mutations in advanced stage tumors treated with DNA-damaging agents, and provide the first evidence of the R213R G-allele as possible risk factor for non–small cell lung cancer.
In recent years, remarkable progress has been made in improving treatment modalities of lung cancer, with a shift from more general to targeted treatments depending on the molecular background of individual tumors. However, despite these promising developments, this devastating disease remains the leading cause of cancer-related death worldwide.1 Approximately 75%–85% of all lung malignancies are currently classified as non–small cell lung cancer (NSCLC), for which the predicted 5-year survival rate for newly diagnosed cases is still below 20%.1 Currently, prognosis of NSCLC patients is mainly based on specific clinicopathologic factors and tumor stage. Nevertheless, within one subgroup, there are considerable differences in patient outcome (i.e. progression-free survival (PFS) and overall survival (OS)), even when patients receive identical treatments.2 A more accurate prognostic assessment, incorporating additional molecular biomarkers, is therefore needed.
The TP53 gene remains the most frequently altered gene in human cancer, and both mutations and single nucleotide polymorphisms (SNPs) have been shown to alter p53’s normal functions. P53 ensures regulation of the cellular response to a variety of stress signals by inducing senescence, cell cycle arrest, and/or apoptosis, thereby playing an important role as tumor suppressor protein. Our group recently reviewed the currently available data on the prognostic and predictive value of TP53 genetic alterations in NSCLC, particularly focusing on the TP53 mutational status and the TP53 R72P polymorphism (rs1042522), of which the Pro variant is associated with reduced apoptotic inducing capacities.3,4 Although these data showed a clear influence of TP53 variants on the patients’ prognosis, further research is warranted, particularly focusing on tumor stage. In addition, only limited data of the effect of TP53 variants on PFS is available.
In this study, next generation deep sequencing technology was used to assess genetic alterations in the TP53 gene, resulting in a higher sensitivity than that achieved by conventional Sanger sequencing mostly used in previous studies.3 In addition, deep sequencing offers the possibility to detect mutations in the presence of normal cell contamination, which often occurs after DNA isolation from formalin-fixed paraffin-embedded (FFPE) tissue by macrodissection techniques. In order to better assess the mutational spectrum of the TP53 gene, the whole coding region consisting of exons 2–11 was sequenced, including splice site regions.
Our first objective was to map all TP53 variants, including SNPs, in a NSCLC adenocarcinoma (AC) patient group and determine possible associations between the SNPs identified in this study, R72P (rs1042522) and R213R (rs1800372), and TP53 mutations. Second, the presence of these variants was linked to clinicopathological characteristics of the NSCLC patients. Third, the prognostic value of frequently occurring SNPs and the TP53 mutational status was determined in both univariate and multivariate OS and PFS analysis.
The prognostic value of TP53 variants in NSCLC has been frequently studied, but has often led to conflicting results, especially for the role of the R72P SNP. Therefore, more in-depth studies are warranted, particularly focusing on NSCLC subtypes and possible differences between tumor stages.3
In this study, we focused on a Caucasian (Belgian) NSCLC AC patient group consisting of stage I–IV tumors. In accordance with previous studies, inactivating TP53 mutations were detected in 44% of all patients.3 In addition, three SNPs were detected, including the frequently occurring R72P polymorphism of which the Pro variant is associated with reduced expression of apoptotic-related target genes.4 We did not observe any association of the R72P SNP with TP53 mutations, although this was previously reported by others.7 The rare synonymous P36P SNP was detected in one patient, and has previously been reported to affect mouse double minute 2 homolog (MDM2)-mediated control of TP53 messenger RNA (mRNA) translation.8 Interestingly, we detected the rare R213R (rs1800372) polymorphism in 7.2% of the patients and are the first to show the significant association with TP53 mutations in NSCLC AC patients. Ganci et al.9 previously reported R213R positive head and neck squamous cell carcinoma (SCC) patients, who were all heterozygous for this locus. In addition, they showed higher expression of the synonymous R213R polymorphic allele compared to the wild-type allele, which was related to TP53 LOH (loss of heterozygosity) positive status, which is frequently observed in the presence of TP53 mutations.10 However, TP53 mutations were only detected in two out of five R213R positive patients, taking into account that their study was limited to exons 5–8. Since we observed R213R-associated mutations in exon 4, and Ganci et al. detected TP53 LOH in these exon 5–8 TP53 wild-type patients, there is a strong possibility that these patients carry on other mutations on exons different than the ones studied by the authors. Despite the association of R213R with inactivating TP53 mutations, which acted as an independent predictor for worse OS and PFS, R213R did not act as a prognostic marker. In addition, the outcome of the multivariate model was not influenced by the R213R polymorphism. Consequently, R213R shows limited potential as a predictor for patients’ outcome. Nonetheless, R213R could acts as a valuable marker for the assessment of the risk for NSCLC development, possibly due to the increased susceptibility for TP53 mutations and high frequency in our patient group compared to the general population (MAF: 0.0054).
TP53 inactivating mutations were not associated with any of the patients’ clinicopathological data, including invasiveness, lymph node infiltration, or metastasis. This shows that TP53 mutations did not seem to affect tumor aggressiveness which could have accounted for the worse OS and PFS observed in our patient group. The fact that the prognostic role for TP53 mutations was unique to stage II–IV tumors might depend on the difference in therapeutic strategy for stage I and stage II–IV patients. Surgical resection remains the main therapeutic option for stage I patients, and most of the stage I patients in our study cohort did not receive (neo)adjuvant therapy, while advanced stages were mostly treated with DNA-damaging agents. Previous studies have shown that in patients who received platinum-based adjuvant chemotherapy, the presence of mutant type TP53 showed a tendency for shorter survival compared to wild-type TP53, depending on the specific type of mutation. Hence, the stronger prognostic value we observed for TP53 mutations in advanced stage NSCLC could be due to therapy resistance/failure, supported by the negative prognostic effect of TP53 mutations in the (neo)adjuvant-treated patients of our study population. The currently available data on the prognostic value of TP53 mutations, recently reviewed by Deben et al.,3 indicated the importance of subtype and tumor stage. Previous studies have shown that the prognostic value of TP53 mutations was limited to AC patients, and not observed in SCC patients, consistent with our results in AC. Both Ahrendt et al.17 and Chien et al. reported a significantly worse OS in stage I patients harboring a TP53 mutation, but not in stage II–III patients. However, these studies did not take subtype into account, which is clearly an influencing factor as discussed above. This could account for the contradictory results observed in this study, which was limited to AC tumors. Only limited studies were available on the predictive role of TP53 mutations on PFS, reporting no effect of TP53 mutations on PFS. However, these studies failed to categorize by tumor stage or subtype.
The Pro-allele of the R72P polymorphism acted as a strong predictor for poorer OS in stage I tumors in our study cohort. One previous study took tumor stage into account and found a similar worse OS in the presence of at least one Pro-allele in stage I tumors, although this effect was not significant and subtype was not taken into account. The R72P Pro-allele did not affect PFS in our patient cohort, even when grouped by tumor stage. Han et al.19 and Liu et al.20 showed a significant association of the Pro-allele with poorer PFS, limited to stage III–IV patients. On the contrary, Shiraishi et al. reported a significant better PFS for patients harboring the Pro/Pro genotype. Again, these studies included both AC and SSC subtypes, making it difficult to compare these findings to our results.
In conclusion, we observed a high frequency of the R213R variant in our patient population, indicating that the R213R G-allele might act as an important risk allele for NSCLC, possibly due to its strong association with TP53 mutations. In addition, our findings clearly show an important role for TP53 variants as predictors for a worse outcome of NSCLC AC patients, especially for TP53 inactivating mutations in advanced stage tumors, possibly due to therapy resistance/failure. Restoring wild-type p53 function using compounds like APR-246 (PRIMA-1MET) or exploiting the presence of mutant p53, as previously discussed, could therefore prove to be promising therapeutic strategies which could greatly improve patients’ outcome.
Authors: Christophe Deben*, Jolien Van den Bossche*, Nele Van Der Steen, Filip Lardon, An Wouters, Ken Op de Beeck, Christophe Hermans, Julie Jacobs, Marc Peeters, Guy Van Camp, Christian Rolfo, Vanessa Deschoolmeester, Patrick Pauwels