Kitada, Shinichi

Abstract We previously demonstrated that apoptosis-inhibiting proteins (e.g., Bcl-2, Mcl-1, and XIAP) are important in predicting response to chemotherapy and antibody therapy for CLL. Additionally, genetic abnormalities and epigenetic control of gene expression have become increasingly important in understanding the prognosis and potential therapeutic opportunities in this disease. Greater than 80% of patients with CLL have genetic abnormalities detectable by FISH. Important subsets include those with a less favorable prognosis (deletions of 17p, 11q), an intermediate prognosis (trisomy 12), or a favorable outcome (deletion of 13q). PTPRO, encoding a receptor-type tyrosine phosphatase, is a specific tumor suppressor gene located in chromosomal region 12p12.3. PTPRO has been highly characterized as inhibiting anchorage-independent tumor growth in human lung carcinomas. This gene has been found to be silenced by promoter methylation, and can be re-expressed by hypomethylating agents in human lung tumor models. Thus, we examined the methylation profile of the promoter region of PTPRO in primary leukemic cells from patients with CLL, and in B cells obtained from normal volunteers. Analysis of the PTPRO CpG island by methylation-specific PCR and combined bisulfite restriction analysis (COBRA) demonstrated that this region was methylated in 91% of samples from 46 patients, whereas it was unmethylated in control B cells from normal volunteers. Specific quantitative determinations of important proteins related to apoptosis in the leukemic cells from 20 of these patients showed that, in the group with methylation of the PTPRO promoter, there was a significant elevation of Bcl-2 (p=0.079), Mcl-1 (p=0.004), XIAP (p=0.048), Bcl-2/BAX ratio (p=0.0024), and Mcl-1/BAX ratio (p=0.0017). There was no apparent relationship with FISH category or immunoglobulin VH mutational status. Thus, PTPRO promoter methylation analysis may identify a subset of CLL patients less likely to respond to therapy, and provide a potential reversible target for new therapeutic approaches to this disease. Finally, analysis of methylation of CLL samples by restriction landmark genome scanning also revealed methylation of genes for three additional protein tyrosine phosphatases: PTPRN2, PTPRZ2, and PTPN11. Increases in methylation of two of these (PTPRN2 and PTPRZ2) in CLL samples relative to control B cells was confirmed by COBRA. These results provide additional potential molecular targets for CLL therapy.

6521 Background: Genomic features including lack of IgVH mutations, del(11q), del(17p), and p53 mutations have been reported to predict clinical course and overall survival in CLL patients (pts). Bcl-2 family proteins and ZAP-70 have also been explored as predictors in CLL. Methods: We prospectively evaluated the prognostic significance of clinical features and laboratory variables on response and progression-free survival (PFS) following treatment with fludarabine (F, n=132) or fludarabine plus cyclophosphamide (FC, n=137) as part of the US Intergroup Trial E2997 for previously untreated CLL. Results: FC therapy had higher complete response (CR) (23.4% versus 4.6%), overall response (OR) (74.3% versus 59%), and median PFS (31.6 mos versus 19.2 mos) compared to F. CR and OR were not significantly different based on interphase cytogenetics, IgVH status, or p53 mutation. IgVH status or levels of ZAP-70, Bcl-2, Bax, Mcl-1, XIAP, Caspase-3, and Traf-1 proteins were not associated with clinical response or PFS. IgVH status and ZAP-70 levels were associated with time from diagnosis to treatment. Using a model including treatment arm, pts with del(17p) or (11q) had significantly shorter PFS (hazard ratios 3.54 and 2.05 respectively). In pts with a p53 mutation without del(17p) there was no enhancement of the model predicting poorer outcome, nor did IgVH status enter the model for predicting PFS. Conclusions: Combination chemotherapy is associated with a higher CR, OR, and PFS. Also, initiation of therapy for pts with Rai stages 0/1 resulted in a higher response rate. Del(17p) and del(11q) are highly predictive of shortened PFS with fludarabine-based chemotherapy. IgVH and p53 mutational status, as well as expression of ZAP-70, Bcl-2 family proteins, and CD38, did not predict response or PFS. ZAP-70 expression is associated with cytogenetic subsets predicted to have a worse overall prognosis, but did not identify pts who will do poorly with therapy. A combination of clinical staging and cytogenetics provide support for future risk-stratified treatment of CLL. Pts with a projected short response duration can be identified for future investigational strategies. No significant financial relationships to disclose.

The critical importance of specific proteins including Bcl-2, Mcl-1, XIAP, Bax, Caspase-3, Traf-1, and ZAP70 in predicting disease progression, prognosis, and response to therapy in CLL have been extensively studied through in vitro experiments. We now report molecular data obtained from a large prospective clinical trial involving 278 previously untreated patients with CLL. This ECOG-led intergroup trial involved participation from CALGB and SWOG, and currently provides quantitative information regarding protein levels on 217 cases assessed immediately prior to treatment. NCI criteria for therapy in CLL were met before patients were randomized to receive either Fludarabine or Fludarabine plus Cytoxan. On-study analyses included: interphase cytogenetics by Fluorescence In Situ Hybridization (FISH), p53 mutational analysis, immunoglobulin variable heavy chain (IgVH) gene rearrangement, and quantitative analysis of the aforementioned specific proteins. Peripheral blood mononuclear cells were obtained from each patient at registration before starting therapy. Lysates were prepared from the patient cells and separated by SDS-PAGE. Proteins were detected by immunoblot analysis using the multiple antigen detection method previously developed by our group. Detection was performed by an enhanced chemiluminescence method and quantified by densitometry. Expression levels were calculated relative to expression in the Jurkat T cell line (for ZAP70) or the RS11846 lymphoblastoid cell line (for apoptosis-related proteins). Using a two-sided Wilcoxon rank sum test, molecular data were compared with FISH and IgVH mutational status. Patients with either p53 mutations or del(17p) had significantly lower levels of XIAP (p=0.022). Patients with del(11q) had significantly higher Bcl-2/Bax ratios (p=0.052). Patients with trisomy 12 had lower levels of Bcl-2 (p=0.066); higher Mcl-1 (p=0.026); higher Bax (p=0.09); and higher Caspase-3 (p=0.019) levels. Interestingly, these specimens also had lower ratios of Bcl-2/Bax (p=0.0016). Patients with del(13q) had significantly higher levels of Bcl-2 (p=0.0059) lower levels of Caspase-3 (p=0.047), and higher Bcl-2/Bax ratios (p=0.0569). IgVH mutated patients had significantly lower levels of Caspase-3 using either a 98% cut-off (p=0.0008) or the 97% cut-off (p<0.0001). Finally, patients with higher ZAP70 protein expression had significantly higher levels of Caspase-3 (p=0.0099) and Traf-1 (p=0.016). In 163 patients evaluable for both response and the molecular studies, either elevation of Bcl-2 protein (p= 0.01) or an abnormality of p53 (either del 17p or a p53 mutation) (p=0.08) correlated significantly with attainment of a complete remission. Potential opportunities to predict response to therapy in CLL will hopefully identify those patients likely to benefit from therapy. Based upon our previous work, these molecular parameters can be modulated both in vitro and in vivo. Therefore, understanding the relationship of these important proteins to therapeutic response can facilitate strategic plans to circumvent drug resistance to chemotherapy and immunotherapy in CLL.

Abstract Abstract 924 Background: Therapy-related myeloid neoplasia (t-MN), including myelodysplastic syndrome and acute myeloid leukemia, has been reported at a higher frequency with chlorambucil + fludarabine as compared to fludarabine alone, but has not been rigorously studied in the context of cyclophosphamide as an alkylator agent. The intergroup prospective randomized Phase 3 trial, E2997, compared FC with F alone as initial therapy for patients (pt) with chronic lymphocytic leukemia (CLL). FC therapy led to higher complete and overall response rates and longer progression-free survival at the initial analysis (Flinn et al JCO 2007). One rationale for combining F with C is that F can inhibit repair of DNA damage induced by C. FC did cause more myelosuppression that could lead to more long-term effects on myeloid hematopoietic function, including t-MN. Methods: E2997 enrolled 278 pts, 141 on FC and 137 on F. These cases were assessed for t-MN by required reporting of these events. Baseline genetic and molecular features of CLL were available through a companion lab correlative trial for 235 pts, 122 on FC and 113 on F. Results: With median follow-up of 6.4 yrs, there have been 13 (4.7%) reported cases of t-MN, 9 after FC and 4 after F. By cumulative incidence methodology, adjusting for the competing risk of death, the rates of t-MN at 7 yr are 8.2% after FC and 4.6% after F (p=0.18 by the Gray test). Median age at study entry of the t-MN pts was 60 (range 45–80) yrs vs 61 (range 33–86) yrs among those not reported to have t-MN. Median time from initial therapy to diagnosis of t-MN was 5 (range 0.7–8) yrs and did not differ between FC and F. Ten of 13 affected pts received 6 chemotherapy cycles. Additional therapy prior to occurrence of t-MN was given to only 2 of 9 FC pts, in contrast to 3 of 4 F pts. Thus, t-MN occurred in only 1 pt treated with F and no further therapy vs 7 who received FC and no further therapy. Ten of 12 pts with available cytogenetics at diagnosis of t-MN had an abnormality of chromosome 5 and/or 7, common to alkylating agent-induced t-MN, usually (n=8) with complex karyotype. Of the 9 pts with t-MN after FC, all 7 with available CLL data had lower risk IgVH mutated disease, in contrast to 1 of 4 with t-MN after F and 44% in the entire cohort. Conclusion: Our analysis suggests that a higher incidence of t-MN has occurred after FC than after F. t-MN after FC occurred most often without additional therapy and in IgVH mutated CLL which is associated with more favorable outcome. The increased incidence of t-MN after FC in this study, usually in the absence of additional treatment, suggests that FC is more leukemogenic than F. This finding emphasizes a need for longer follow-up of toxicity and survival before concluding that FC is preferable to F as the chemotherapy backbone for initial therapy of both low and high risk CLL. Disclosures: No relevant conflicts of interest to declare.

Purpose Genomic features including unmutated immunoglobulin variable region heavy chain (IgV H ) genes, del(11q22.3), del(17p13.1), and p53 mutations have been reported to predict the clinical course and overall survival of patients with chronic lymphocytic leukemia (CLL). In addition, ZAP-70 and Bcl-2 family proteins have been explored as predictors of outcome. Patients and Methods We prospectively evaluated the prognostic significance of a comprehensive panel of laboratory factors on both response and progression-free survival (PFS) using samples and data from 235 patients enrolled onto a therapeutic trial. Patients received either fludarabine (FL; n = 113) or fludarabine plus cyclophosphamide (FC; n = 122) as part of a US Intergroup randomized trial for previously untreated CLL patients. Results Complete response (CR) rates were 24.6% for patients receiving FC and 5.3% for patients receiving FL (P = .00004). PFS was statistically significantly longer in patients receiving FC (median, 33.5 months for patients receiving FC and 19.9 months for patients receiving FL; P < .0001). The occurrence of del(17p13.1) (hazard ratio, 3.428; P = .0002) or del(11q22.3) (hazard ratio, 1.904; P = .006) was associated with reduced PFS. CR and overall response rates were not significantly different based on cytogenetics, IgV H mutational status, CD38 expression, or p53 mutational status. Expression of ZAP-70, Bcl-2, Bax, Mcl-1, XIAP, Caspase-3, and Traf-1 was not associated with either clinical response or PFS. Conclusion These results support the use of interphase cytogenetic analysis, but not IgV H , CD38 expression, or ZAP-70 status, to predict outcome of FL-based chemotherapy. Patients with high-risk cytogenetic features should be considered for alternative therapies.