ALBERT

Description: Introduction: Standard chemotherapy represented by the R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone) regimen is successful in about 60% of patients (pts) with diffuse large B-cell lymphoma (DLBCL). Pts who do not benefit from this treatment, due to the development of tumor drug resistance, have a very poor prognosis. Currently, knowledge on reasons of treatment related failures in DLBCL are scanty and predictive biomarker of response are largely unknown. We hypothesized that polymorphisms of gene involved in the pharmacokinetics and pharmacodynamics of drugs included in R-CHOP regimen may play a role in predicting the outcome in DLBCL pts.Thus, we designed a multicentre prospective pharmacogenetic trial aimed at identifying gene polymorphisms potentially predictive of drug efficacy/resistance in DLBCL pts treated with R-CHOP. An interim analysis on the first 80 enrolled ptswas planned and has been performed. Methods: The study included chemonaive DLBCL pts at various stages of disease candidate to an R-CHOP standard treatment. The Ethical Committee of each participating centre approved the pharmacogenetic protocol, and all pts signed a written informed consent. According to the aims of this interim analysis, the impact of single nucleotide polymorphisms (SNPs) on R-CHOP efficacy was evaluated by objective response (OR) rate at the end of treatment. The efficacy of R-CHOP was evaluated according to the Cheson criteria by performing standard hematochemical and instrumental (TC and FDFG-PET) tests and defining complete remission (CR), partial remission (PR), non response or progressive disease (PD). Genomic DNA wasextracted from peripheral blood of 80 pts. Twentysingle nucleotide polymorphisms (SNPs) from18candidate genes (ABCB1, ABCC1, ABCC2, ABCG2, CYBA, CYP2C9, FCGR2A, GSTP1, IL2, MLH1, NCF4, NQO1, NQO2, RAC2, TNF, TOP2A, TP53, TUBB)involved in pharmacokinetics and pharmacodynamics of R-CHOP (www.pharmgkb.org) have been analysed by a genotyping array based on Affimetrix methodology. Univariate analysis was performed to evaluate associations between polymorphisms and clinical/pathological characteristics or OR (Fisher exact test). Multivariate logistic regression analysis was performed to estimate adjusted odds ratios along with the corresponding 95% confidence intervals for the polymorphisms and OR. Results: Median age was 63 years. There were 37 men and 43 women. 47.5 % of pts were in stage I-II,52.5 % of pts in stage III-IV. 27.5% of ptshad bulky disease, 43.8 % of pts had involvement of extranodal site. 47.5% of pts had pathological LDH value. According to the revised IPI, 15 % pts were in the low risk group, 58.7 % in the intermediate risk group, and 26.3 % in the high risk group.Overall, 468 courses of R-CHOP had been administered (mean: 5.85 courses, range: 4-6). 81% of pts had CR to R-CHOP whereas the remaining showed PR (14%) or PD(5%). No statistically significant correlation was found between OR and clinical characteristics of pts.However, stage III-IV pts showed a worst OR than stage I-II pts (77% vs 87% of CR, respectively); pts with bulky disease had worst OR than non-bulky disease pts(73% vs 84.5% of CR, respectively); ptswith R-IPI 3-5 a worst OR than pts with R-IPI 0-2 (71.5% vs 85% of CR, respectively). Univariate and multivariateanalysis identified TOPOII rs13695as a predictor of OR (p=0.042). Pts with CT or TT genotypesshowed worst OR than CC wild-type homozygous pts (odds ratio 3.070, CI95% 1.113-13.457). Also, a statistical trend toward significance was observed for MLH1 rs1800734 polymorphism (p=0.062): ptswith homozygous genotype for the mutant allele showed a better OR than wild-type and heterozygous pt genotypes. Conclusions: No significant relationship between clinical/pathological characteristics and OR was observed. Our preliminary data show that SNPs affecting a gene involved in doxorubicin pharmacodynamics, i.e. the drug target TOPOII, as well asone of the major components of DNA mismatch repair, i.e. MLH1 gene,may predict response in DLBCL pts treated with R-CHOP. These preliminary results from the interim analysis are promising and warrant completion of pt accrual to reach the planned number of cases at the end of our study. Acknowledgments This work was supported by a grant from the Associazione Giacomo Onlus, Castiglioncello (LI), Italy to E.M. and Cassa di Risparmio di Firenze, Firenze, Italy to S.N. Disclosures No relevant conflicts of interest to declare.

Description: Abstract 2478 Diffuse large B cell lymphoma (DLBCL) is one of the most common types of non-Hodgkin's lymphoma. Approximately half of patients will be cured of their disease by primary therapy, including the R-CHOP regimen (rituximab, doxorubicin, cyclophosphamide, vincristine, desamethasone). The remaining die of the disease, mainly because of the occurrence of tumor drug resistance. Many efforts have been made to explain the biochemical and molecular mechanisms involved in resistance to the drugs used in the treatment of cancer patients, including those with DLBCL. A dose-intense therapy regimen (e.g. R-CHOP14) may help to improve the treatment outcome of DLBCL patients. We have carried out a retrospective study aimed at correlating the mRNA expression levels of genes involved in metabolism, mechanisms of action and resistance to doxorubicin (i.e. MDR1, GSTP1, TOPO-2a, Bcl-2, PKC-b2) that represents the backbone of the R-CHOP regimen with treatment outcome data of 54 patients at various stages of disease. The expression of the 5 above mentioned genes was determined in formalin-fixed paraffin-embedded samples from DLBCL using real time RT-PCR. A threshold analysis to identify a cut-off distinguishing recurrent or non-recurrent disease was used. The correlations between gene expression data and clinical/pathological characteristics as well as survival parameters have been evaluated by standard statistical tests. The case series included 32 males and 22 females; 6 patients had follicular lymphoma grade IIIb and 48 diffuse large B cell lymphoma; 19 presented symptoms at diagnosis. Thirty patients showed abnormal LDH values, the IPI was intermediate-high risk or high risk in 14 patients. Forty-six patients (85.2%) obtained a complete remission and 8 (14.8%) a partial response. The median overall survival (OS) as well as the median progression free survival (PFS) have not yet been reached after a median follow-up of 43.6 months. The mRNA expression levels of TOPO-2a and GSTP1 were detectable in all samples, that of PKC-b2 in 52 samples, that of MDR1 and bcl-2 in 34 and 29 samples, respectively. A high degree of interpatient variation in relative tumor expression of the study gene was observed: from 0.008 for TOPO-2a to 〉100.000 for PKCbII. Threshold analysis indicated significant inverse relationships between PKC-b2 and PFS (p=0.046): higher gene expression was associated with shorter PFS. Conversely, higher expression of ABCB1 was associated with prolonged PFS (p=0.039). This kind of analysis also showed associations between OS and TOPO-2a, GSTP1and PKC-b2: higher gene expression was associated with shorter OS. Overall, our results confirm that the high expression of some genes such as TopoIIa, GSTP1 and PKCβII may represent a prognostic factor in case of an intensified anthracycline-based chemotherapy with immunotherapy. Moreover, our results suggest that intensified immunochemotherapy could affect the role of bcl2, ABCB1, GSTP1 and TopoIIa in predicting tumor response. These results and others from related studies may help to identify gene profiles useful for selecting patients eligible for more intensified or personalized chemotherapy. Prospective larger studies are warranted. Supported by a grant from Associazione Giacomo Onlus, Castiglioncello (LI). Disclosures: No relevant conflicts of interest to declare.