Clinical relevance of MDM2 SNP 309 and TP53 Arg72Pro in follicular lymphoma

Tumor protein 53 (TP53) is critical to cell cycle control and is the most common mutational target in germinal center lymphomas. However, these mutations occur infrequently at diagnosis (<10%) in follicular lymphoma (FL), and are more commonly associated with disease progression or transformation to more aggressive histology.1–4 Inactivation of TP53 may also occur by upregulation of the human homolog of the Murine Double Minute 2 protein (MDM2), which targets TP53 for degradation by the proteasome and is frequently amplified in FL and other malignancies; indeed Mdm2 haplo-insufficiency in mice leads to enhanced TP53 function with delayed onset of lymphoma.5 While many studies have examined the genomic events targeting these loci, less is known as to the functional role of polymorphic variants at the MDM2 or TP53 loci. The single nucleotide polymorphisms (SNPs), MDM2 SNP 309 and TP53 Arg72Pro, have at least in some studies an additive effect on cancer susceptibility with MDM2 SNP 309 also predicting advanced disease at diagnosis.6 Consequently, we investigated the impact of MDM2 SNP 309 and TP53 Arg72Pro on the clinical outcome of FL. MDM2 SNP 309 characterizes a T>G substitution at nucleotide 309 of intron one and leads to higher MDM2 mRNA and protein expression with lower apoptotic response. Significantly, homozygosity for the G allele correlates with earlier onset of de novo diffuse large B-cell lymphoma (DLBCL) in females7 Its role in CLL is less clear as two recent studies provide contrasting results.8,9 TP53 Arg72Pro is a non-synonymous SNP involving G>C substitution at nucleotide 466 of exon 4 in TP53 which creates a TP53 protein with reduced potential to induce apoptosis or suppress cell transformation.10 These SNPs are also associated with poor prognosis or increased cancer risk in certain solid malignancies.11,12

Table 1.Assessment of MDM2 SNP 309 and TP53 Arg72Pro against clinical parameters

We investigated the relationship of the allelic combinations for both SNPs with the following clinical variables in cases of FL: age, gender, stage at diagnosis, response to first line therapy, best response to therapy, progression-free survival, relapse-free survival, overall survival and time to transformation (Details of each variable are available in the Online Supplementary Table S1). A real-time polymerase chain reaction Allelic Discrimination multiplexed endpoint assay protocol using the ABI PRISM™ 7700 Sequence Detector (Applied Biosystems, Foster City, CA, USA) was used to determine the SNP alleles present in DNA of 226 patient samples from bone marrow (n=207), peripheral blood (n=14) or lymph nodes (n=5). (Oligonucleotide sequences used are available on request.) Samples were obtained from the tissue archive at St Bartholomew’s Hospital under Institutional Ethical Approval obtained from the Local Ethics Committee. Reactions were performed in duplicate including homozygous and heterozygous controls for each SNP genotype and a no DNA template control. Positive controls were confirmed by direct sequencing.

In this patient cohort, allelic frequencies were similar to previously published results at 44%, 43%, 13% for MDM2 SNP 309 TT, TG, GG and at 46%, 46%, 8% for TP53 Arg72Pro GG, GC, CC respectively (Online Supplementary Table S2) and both polymorphisms satisfied the Hardy-Weinberg equilibrium. Median age at diagnosis was 46 years, which is younger than that previously described for FL, and reflects our status as a tertiary referral center for patients with FL. There was no difference in the median age at diagnosis for each of the 3 allelic combinations of MDM2 SNP 309 (both for the whole study population and for gender in contrast to findings in DLBCL7) and of TP53 Arg53Pro (Online Supplementary Table S2).

Figure 1.Kaplan-Meier plots and log rank test statistics for overall survival (A) and time to transformation (B) by MDM2 SNP 309 and TP53 Arg72Pro alleles. Illustrated are results for four groups of allelic combinations of the two SNPs. The four groups Neither rare, Rare TP53, Rare MDM2 and Both rare are defined in Table 1. In addition, no associations were seen between the 3 allelic combinations for each SNP and the survival parameters, nor were associations seen on informal analyses of all possible allelic combinations between the two SNPs (data not shown).

Relationships between SNP alleles and clinical parameters were studied using χ test for stage and gender, analysis of variance (ANOVA) for age and Kruskal-Wallis test for responses to therapy. The 3 allelic combinations for each SNP were informally assessed as well as formally assessing two groups of patient samples consisting of those homozygous for the common allele or those with at least one rare allele, enabling any dominant effects of the MDM2 SNP 309 G or TP53 Arg72Pro C alleles, as described in other malignancies, to be assessed in FL. As MDM2 SNP 309 can lead to earlier onset or more advanced disease6 and TP53 negatively regulates MDM2 expression, we assessed MDM2 SNP 309 alleles in isolation or in combination with TP53 Arg72Pro alleles against these clinical parameters; there was no evidence of any association (illustrated in Table 1 for gender, stage and responses to therapy; p≥0.13). Kaplan-Meier plots and log rank test statistics demonstrated no evidence for associa-ion oft MDM2 SNP 309 or TP53 Arg72Pro allelic variants alone, or in combination, with overall survival (Figure 1A), progression free survival, relapse free survival or time to transformation (Figure 1B) (p≥0.17).

Consequently, whilst genomic lesions targeting the MDM2-TP53 axis are an important feature of FL, MDM2 SNP 309 and TP53 Arg72Pro do not predict clinical outcome. In contrast to other malignancies these polymorphisms are not significant in FL.

Footnotes

• Funding: this work was supported by Cancer Research UK, the Cancer Committee of Barts and the London NHS Trust and the Research Advisory Board of Barts and the London Medical School