Abstract

Transformation of follicular lymphoma (FL) to a more aggressive clinical and histological phenotype, typically diffuse large B-cell lymphoma (DLBCL), occurs frequently. It is associated with a number of recurrent genomic insults, including the acquisition of TP53 mutations in a subset of patients (pts). The use of novel agents targeting p53 and mdm2 appears attractive given the resistance of transformed DLBCL to conventional therapies. Tailoring these therapies will require precise characterisation of mutation status and functional consequence in transformation. The frequency and temporal relationship of TP53 mutation gain to transformation was analysed in DNA from sequentially collated lymph node biopsies taken pre and post transformation (n=91) obtained from 29 pts. A median of 3 samples (range 2–5) was available from each pt (13 taken at FL presentation). Transformation was documented a median of 3.8 years (range 0.2 to 15.2) from diagnosis, and median follow up from diagnosis for all pts at the time of analysis was 6.7 years (range 2 to 19.1). The entire coding sequence of TP53 was screened by PCR, fluorescent-SSCP and sequencing. Loss of heterozygosity (LOH) was examined at 5 common polymorphic sites with in TP53. Immunocytochemistry for p53, mdm2 and p21 was performed on slides obtained from 77 available paraffin blocks. Ten mutations were detected in 8 pts (28%), of which 5 were missense. The remaining was accounted for by two nonsense mutations, a splice mutation, a branch site mutation and a single base insertion. All mutations were within the genomic region covered by primer sets exon 5–7 inclusive. Mutated TP53 was first documented only at the time of histologic transformation in 4 pts, in the remainder latency between documentation in FL sample and transformation was variable (0.5–6 years). For pts with mutations, time from documentation to death ranged from 1 month-12 years (median 37 months), with 2 pts alive 8.5 and 13.5 years following initial documentation. LOH occurred in 2 pts, both at the time of transformation and was associated with short survival (1 and 17 months). Overall survival from diagnosis or histological transformation was not significantly different between pts with mutated TP53 and wtTP53. Five TP53 mutated pts. recurred post transformation (either with FL or DLBCL); in 4 pts the identical mutation was detected at this time. p53 staining was positive in 82% (9/11) of biopsies with missense mutations, and negative in 71% (45/63) with wtTP53. Mdm2 expression was predominantly centroblastic in FL and was correspondingly higher in DLBCL samples (mean 72%; 95% CI 68–76%) compared to FL (mean 58%; 95% confidence interval: 54–62%) (p<0.001). Mdm2 expression did not correlate with TP53 mutation status. Expression of p21 antigen was positive in 19/71 (27%) cases and did not correlate with histology. Absence of p21 occurred in both wtTP53 (66%) and mutated TP53 (94%) samples. TP53 mutations were associated with transformation in only a subset of pts; the potential of individual mutations to induce phenotypic change was variable and thus may influence the potential success of novel TP53 directed therapies.

Author notes

Corresponding author