TP53 Arg72 as a favorable prognostic factor for Chinese diffuse large B-cell lymphoma patients treated with CHOP

Background TP53 Arg72Pro (SNP rs1042522) is associated with risk of non-Hodgkin lymphoma (NHL). Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of NHL. However, the relationship between this SNP and prognosis of DLBCL in Asians is unknown. Methods Genotyping of TP53 Arg72Pro was done in 425 Chinese DLBCL patients. Two hundred and eighty-nine patients were treated with R-CHOP, and 136 patients received CHOP or CHOP-like as frontline regimen. Three hundred and ninety-six patients were assessable for the efficacy. Results Patients with Arg/Arg and Arg/Pro at codon 72 of TP53 had a higher complete response rate (61% vs. 44%, P = 0.007) than those with Pro/Pro. In the subgroup treated with CHOP or CHOP-like therapy, patients with Arg/Arg and Arg/Pro showed a higher 5-year overall survival (OS) rate than those with Pro/Pro (68.8% vs. 23.2%, P = 0.001). Multivariate Cox regression analysis revealed TP53 Arg72 as a favorable prognostic factor in this group. However, the combination of rituximab with CHOP significantly increased the 5-year OS rate of patients with Pro/Pro to 63%. Conclusion This study revealed TP53 Arg72 as a favorable prognostic factor for Chinese DLBCL patients treated with CHOP or CHOP-like as frontline therapy.

Several reports demonstrated that TP53 Arg72Pro was associated with tumorigenesis and clinical outcomes [5][6][7][8]. Several meta-analyses of this SNP in cancer risk revealed that the TP53 Arg72Pro polymorphism is associated with an increased risk of cancer. In the subgroup analysis, significantly increased cancer risk was observed among Asians in homozygous and recessive models, while in Americans increased cancer risk was observed only in dominant and recessive models [9,10]. Moreover, a significantly increased non-Hodgkin lymphomas (NHL) risk was found in carriers of the TP53 72Pro allele, including in Chinese patients [11][12][13][14]. However, the association of TP53 Arg72Pro with clinical outcomes and prognosis in lymphoma is still uncertain [15,16].
DLBCL is the most common subtype of NHL [17]. Combined treatment of rituximab and chemotherapy has resulted in improved clinical outcomes [18][19][20][21]. However, one-third of responding patients become refractory to treatment and no-responders to second line therapy or immune-chemotherapy-based third line therapy [22,23]. To evaluate the influence of TP53 Arg72Pro on the prognosis of NHL in the Chinese population, this retrospective study was done in 425 DLBCL patients treated with CHOP or CHOP plus rituximab (R-CHOP) therapy.

Patients population and response evaluation
The clinical research protocol was approved by the Institutional Review Board and the Ethical Committee of Peking University School of Oncology, Beijing, China. All patients participating in this study signed the informed consent.
Four hundred and twenty-five patients with DLBCL confirmed by our Department of Pathology according to the World Health Organization classification were included in this study. Of the patients, 289 received rituximab in combination with a chemotherapy regimen between January 2000 and January 2015 at the Beijing Cancer Hospital, Peking University School of Oncology. Another 136 patients received CHOP or CHOP-like therapy (e.g. COP, CCOP, CHO or CHOPE) as the frontline chemotherapy. Based on the expression levels of Bcl-6, CD10, and MUM-1 measured by immunohistochemistry, cases were subdivided into germinal center B-cell (GCB) and non-GCB types according to the Hans algorithm [24,25]. The response to chemotherapy was evaluated after completion of 2 to 3 courses of therapy and 1 to 2 months after completion of all treatment plans, then every 3 months for the first year and every 6 months thereafter until progression.
Overall survival (OS) was calculated from the date of disease confirmation to the date of last follow-up or death. Progression free survival (PFS) was identified as the period between the disease confirmation and progression (relapse and refractory) or disease-related death. Disease status was evaluated via clinical findings and computed tomography and classified as complete response (CR), unconfirmed complete response (CRu), partial response (PR), stable disease (SD), progressive disease or relapse according to the revised response criteria for malignant lymphoma [26,27]. Patients who had heterozygous (GC) or homozygous G (GG) genotype of TP53 SNP rs1042522 were designated as G carriers.

DNA extraction and genotyping
Genomic DNA was extracted from whole blood using the Whole Blood Genome DNA isolation Kit according to the manufacturer's instructions (Qiagen, Nussloch, Germany). DNA was diluted in AE buffer to a final stock concentration of 20 ng/μl, and 2 μl was used in each PCR reaction.
Following an initial denaturation step at 94°C for 3 min, amplification was carried out by 40 cycles of denaturation at 94°C for 30s, annealing at 62°C for 40s, and extension at 72°C for 40s. This was followed by a final extension at 72°C for 5 min. Amplified products were analyzed by gel electrophoresis on 1.5% agarose gels and were sequenced using an ABI 3730XL Avant Genetic Analyzer (Applied Biosystems Inc., USA). Determination of the TP53 rs1042522 genotype was achieved blindly on coded specimens by Sanger chain termination sequencing with the Seqman software (DNASTAR, USA).

Statistical analysis
All statistical analysis was carried out using SPSS software for Windows (version 19.0). An effect was considered statistically significant at P < 0.05. Genotype frequencies and clinical parameters were compared using a χ2 test. The Kaplan-Meier method was used to construct survival curves, and results were compared using a log-rank test. Multivariate Cox regression analyses were used to assess associations between survival time and potential risk factors. The Hardy-Weinberg equilibrium was used to test for deviation of allele and genotype frequency.

Patients' characteristics
The general characteristics of the 425 DLBCL patients (175 male and 250 female) in this study are summarized in Table 1. The median age at diagnosis was 54 years (range, 15-90 years). Two hundred and forty-three (57.2%) patients were in stage 3 or 4, and 136 (32.0%) patients had intermediate-to-high or high international prognostic index (IPI) scores. One hundred and twelve (26.4%) patients were classified into GCB subgroup, 251 (59.1%) patients were classified into non-GCB subgroup, and 62 patients had incomplete records. Two hundred and thirty-four (55.1%) patients exhibited B symptoms at diagnosis and 109 (28.2%) patients showed an elevated β 2 -MG level. Two hundred and eighty-nine (68%) patients were treated with R-CHOP therapy and 136 patients were treated with CHOP or CHOP-like therapy only.

TP53 SNP rs1042522 in 425 DLBCL patients
We detected the genotype of TP53 SNP rs1042522 in 425 patients. As shown in Table 2, 28% patients carried the homozygous GG genotype (Arg/Arg), 53.2% patients had the heterozygous GC genotype (Arg/Pro), and18.8% patients carried the homozygous CC genotype (Pro/Pro). The frequency of the G allele in 425 patients was 55%, and the frequency of the C allele was 45%. The genotype distribution of SNP rs1042522 in the DLBCL population analyzed in this study was in Hardy-Weinberg equilibrium (P = 0.135), and the allele distribution was close to the frequency distribution seen in the Asian population based on the dbSNP database.

Correlations between SNP rs1042522 and clinical features of DLBCL patients
As shown in Table 1, patients with genotype GG and GC of SNP rs1042522 had a lower positive rate for β 2 -MG than those with genotype CC (25.1% vs. 42.3%, P = 0.004). Although the G allele carriers showed a higher positive rate for B symptoms (58.1% vs. 42.5%, P = 0.012), the univariate analysis revealed that B symptoms is not an independent prognostic factor for overall survival (P = 0.983). The genotype distribution in CHOP or CHOP-like and R-CHOP treated subgroups is unbiased.
Clinical response according to the genotype of TP53 SNP rs1042522 Of the 396 patients evaluable for response to CHOP or CHOP-like therapy with or without rituximab, the OR rate was 84.1% (333 of 396 patients), including a CR rate of 57.8% (229 of 396 patients) and a PR rate of 26.3% (104 of 396 patients). As shown in Table 3, of the 396 patients, patients with genotypes GG and GC exhibited higher CR and OR rates than those with the genotype CC (61% vs. 44%, P = 0.007; 86% vs. 76%, P = 0.033). The combination of rituximab in treatment significantly increased the CR rate (65% vs. 38%, P < 0.001). In the subgroup treated without rituximab, a relatively higher CR rate was achieved in patients with genotype GG and GC than in those with genotype CC (45.78% vs. 12.5%; P = 0.004). However, this significant difference vanished in the subgroup treated with combination of rituximab. In the subgroup treated with R-CHOP or R-CHOP-like, patients with genotypes GG and GC exhibited similar CR (66.4% vs. 58.8%, P = 0.304) rates and OR (87% vs. 76.5%, P = 0.056) rates to those with the CC genotype.
Survival analyses according to the genotype of TP53 SNP rs1042522   the genotype CC showed a median OS of 39.9 months and a median PFS of 18.1 months. In the subgroup treated with CHOP or CHOP-like therapy (Fig. 1), patients with genotype GG and GC had higher 5-year OS and PFS rates than those with genotype CC (68.8% vs. 23.2%, P = 0.001; 56.1% vs. 25.4%, P = 0.002, respectively). However, the integration of rituximab in treatment significantly increased the 5-year OS and PFS rates (57.1% vs. 72.8%, P = 0.001; 49.4% vs. 61.3%, P = 0.017) in the overall population. Therefore, in the subgroup treated with R-CHOP therapy (Fig. 2), the 5-year OS and PFS rates of CC patients are only about 10% lower than those of G allele carriers (63.0% vs. 74.9%, P = 0.218; 51% vs. 63.5%, P = 0.05) and did not reach the statistical significance.

Discussion
In this study, a retrospective analysis was done to evaluate the influence of TP53 Arg72Pro on the prognosis of 425 Chinese DLBCL patients treated with CHOP or R-CHOP therapy. Patients with genotype GG (Arg/Arg) and GC (Arg/Pro) of SNP rs1042522 had a lower positive rate for β 2 -MG and higher CR and OR rates for treatment than those with genotype CC (Pro/Pro). In the subgroup treated without rituximab, a significant higher CR rate and higher 5-year OS and PFS rates were achieved in patients with Arg/Arg and Arg/Pro than in those with Pro/Pro. Multivariate analysis revealed TP53 Arg72 as a favorable prognostic factor in this group. As the integration of rituximab in treatment significantly increased the CR, 5-year OS and PFS rates in the subgroup treated with R-CHOP therapy these significant differences vanished between two genotype groups.  The previous study in European Caucasians demonstrated no influence of TP53 Arg72Pro on survival of DLBCL patients [15]. However, we observed better survival in patients with Arg/Arg and Arg/Pro than those with Pro/Pro, when treated with CHOP or CHOP-like therapy. Meta-analysis revealed that ethnicity may modulate the penetrance of TP53 Arg72Pro in cancer susceptibility [9,10]. According to the phase 3 data of 1000 Genome project, the C allele frequency is 28.53% in Europeans and 41.37% in East Asians. The C allele frequency is 25% in 205 Germany DLBCL patients and 45% in 425 Chinese DLBCL patients. Therefore, the genetic background may account for the discrepancy of clinical outcomes in two studies. Furthermore, patients in the European study were from the NHL-B1 and B2 studies, which treated aggressive lymphoma in elderly patients and good-prognosis young patients by CHOP with or without etoposide [28,29]. However, how the 205 DLBCL patients were stratified according to good or poor prognosis and treated with or without etoposide was not clarified. In our subgroup treated with R-CHOP, the difference in survival between two genotype groups was not significant.
CHOP regimen has always the backbone of treatment strategies in DLBCL. Clinical trials had confirmed that the overall survival of patients was estimated at 50% in young and elderly patients [30][31][32]. However, in our study, the 5-year survival rate of patients with genotype CC (Pro/Pro) was found to be 23.2%. It is unclear exactly how TP53 polymorphism affects the survival to CHOP chemotherapy. Previous studies reported that when cells were exposed to doxorubicin, apoptosis was always higher in cells expressing the TP53 72Arg variant than those expressing the TP53 72Pro [33]. However, the mechanisms underlying the influence of SNP on the response to chemotherapy is still needed to further investigate in different cancers and in different populations. In general, the recombination of rituximab with CHOP therapy might be highly beneficial for Chinese patients with Pro/Pro at TP53 codon 72.

Conclusion
In summary, our study revealed TP53 Arg72 as a favorable prognostic factor in Chinese DLBCL patients treated with CHOP/CHOP-like as frontline therapy. Combination of rituximab with CHOP could optimize the survival for the Chinese patients with Pro/Pro, therefore reducing the predictive value of this biomarker with the current standard of care. This is the first report to evaluate the influence of TP53 Arg72Pro on clinical outcomes of DLBCL patients from Asia. The prognostic implication of this SNP in other lymphoma subtypes, as well as in other cancers needs to be further studied.