TP53 Gene Status Affects Survival in Advanced Mycosis Fungoides

TP53 is frequently mutated in different types of neoplasms including leukemia and lymphomas. Mutations of TP53 have also been reported in mycosis fungoides (MF), the most common type of cutaneous lymphoma. However, little is known about the frequency, spectrum of mutations, and their prognostic significance in MF. In this study, we have optimized the protocol for Sanger sequencing of TP53 using DNA extracted from archival paraffin-embedded biopsies. Of 19 samples from patients with stage IIB MF or higher, 31% harbored mutations in TP53. Overall survival of the patients with mutated TP53 was significantly shorter than median survival in the age- and stage-matched patients treated in our Institution. Distribution of mutations was heterogenous in TP53 exons; however, C > T transitions were common suggesting the causal role of ultraviolet radiation. We propose that TP53 mutation status would be useful for risk stratification of patients with advanced MF.

TP53 is frequently mutated in different types of neoplasms including leukemia and lymphomas. Mutations of TP53 have also been reported in mycosis fungoides (MF), the most common type of cutaneous lymphoma. However, little is known about the frequency, spectrum of mutations, and their prognostic significance in MF. In this study, we have optimized the protocol for Sanger sequencing of TP53 using DNA extracted from archival paraffin-embedded biopsies. Of 19 samples from patients with stage IIB MF or higher, 31% harbored mutations in TP53. Overall survival of the patients with mutated TP53 was significantly shorter than median survival in the age-and stagematched patients treated in our Institution. Distribution of mutations was heterogenous in TP53 exons; however, C > T transitions were common suggesting the causal role of ultraviolet radiation. We propose that TP53 mutation status would be useful for risk stratification of patients with advanced MF.
Keywords: p53 mutation, mycosis fungoides, survival rate, sequencing data analysis, cutaneous lymphoma inTrODUcTiOn Mycosis fungoides (MF) is the most common form of cutaneous T-cell lymphoma (CTCL) and in most cases runs an indolent course. However, approximately 20% of patients would progress to a widespread disease with multiple skin tumors and extracutaneous involvement with dire outcome (1). Male gender, young age, and folliculotropic subtype are risk factors for progression, but their predictive value is low (1,2). There are currently no robust biomarkers that predict the course of the disease.
The tumor suppressor gene TP53 is central in tumorigenesis and regarded as a master regulator of several signaling pathways involved in this process. TP53 is mutated in more than 50-70% of all solid tumors and in approximately 10-20% in hematological malignancies (3). While low-grade lymphoid neoplasms reveal low p53 mutation rates, lymphomas and leukemias (in particular chronic lymphatic leukemia) with an aggressive clinical course demonstrate higher frequencies (4)(5)(6)(7). To this end, a strong correlation was found between p53 functional status and clinical outcomes in lymphoma, such as mortality or resistance to chemotherapy (8,9). TP53 may be mutated in a proportion of patients with another type of CTCL, Sezary syndrome, but mutations are not predictive for the course of the disease (10). Less is known about TP53 gene status in MF. Previous studies suggested that p53 overexpression and mutation is uncommon in early MF (11)(12)(13) but can be found in advanced stages indicating prognostic significance (14,15). It has also been suggested that p53 mutations in MF are caused by ultraviolet radiation (11,14,16). Therefore, in this study we wished to elucidate the   Table 1) are compared against the control group of 30 patients with stage IIB MF or higher. The difference in survival is significant (p = 0.05) after adjustment for stage, age, and gender by Cox logistic regression.
x-axis -years, y-axis -probability of survival.  resUlTs In 6 out of 19 cases (31%), 1 or several mutations could be identified, exhibiting a heterogeneous pattern with mutations in different exons at various codon sites ( Table 1).
To investigate whether TP53 mutations had impact on survival, we compared patients with mutated and normal TP53 by Kaplan-Meier analysis. Because the sample size was small, we decided to compare survival of the 6 patients with mutated p53 to all 30 patients with stage >IIB MF treated in our center in the same time frame (14 women, 16 men; mean age 71.8 years, SD 12.6). As shown in Figure 1, patients with mutated TP53 had a shorter survival than the controls. The statistically significant difference (p = 0.05) was maintained when data were adjusted to age, stage, and gender by Cox logistic regression. Adjusted odds ratio for death in the presence of mutated P53 was 2.99 (95% confidence interval 1.12-8.00, Cox regression) comparing to the control group.

DiscUssiOn
The protocol developed in this study allowed for PCR amplification and Sanger sequencing of TP53 in archival formalin-fixed and paraffin-embedded samples of MF. Only in 4 out of 19 samples, we were unable to amplify one or several exons, which is most probably due to sample age and DNA crosslinking.
The frequency of mutations in stage IIB or higher was 31%, which is similar to what has been reported previously (11,16). In contrast, no mutations have been identified in a recent study of McGirt et al. (14) who reported Pro72Arg polymorphism to be more common in MF. Mutation analysis did not reveal any distinctive pattern of alterations in TP53. None of the earlier reported mutations were identical to our findings, and our cases harbored only one of the known hotspot mutations (R273H in case 18). In a recent study, Choi et al. reported 7 different mutations located at different sites in 7 out of 40 analyzed patients, including the well-known R273H hotspot (22). Our cases revealed mainly missense single nucleotide mutations resulting in stop codons (cases 1 and 17). Frequent C > T transitions seem to support the hypothesis that TP53 mutations demonstrate an UV B signature, which might be related to the effect of phototherapy given repeatedly in the earlier stages of the disease (11,14,16).
TP53 mutations seem to be associated with worse outcome (higher overall mortality). However, this analysis should be interpreted with caution since our sample is small, and the presence of mutations may reflect more intensive use of phototherapy in patients with aggressive, difficult to control disease. Larger sample and careful matching with the control patients with normal TP53 status will be required to elucidate this issue.

aUThOr cOnTriBUTiOns
Design of the study: RG, LG, and LM; data collection: LM, ER, LG, and RG; analysis of data and full access to raw data: RG, LG, and LM; drafting of the manuscript: GW, LM, LG, and RG; and revision of manuscript for important intellectual content: RG, LG, LM, GW, and ER.

FUnDing
This study was funded from the operational grant from Bispebjerg Hospital and research grant from Aage Bangs Foundation, Copenhagen, Denmark.