Jiang, Liuyan

7047 Background: In addition to secondary AML (sAML), important lifestyle and environmental exposures associated with increased risk of AML development have been identified in recent case-control studies including obesity (BMI>30kg/m2), smoking, regular acetaminophen use, and rural/farm habitat. The association of these exposures with AML cytogenetic risk groups is unknown. We therefore evaluated relevant exposures in a cohort of 301 AML patients with confirmed central cytogenetics analysis diagnosed and treated at Mayo Clinic FL and AZ since 1995. Methods: Documented patient exposures were extracted from central electronic medical records, including: prior chemotherapy/radiation or hematologic malignancy, occupation, select medications, “toxins” (esp. benzene) and solid organ transplantation (SOT). Standard cytogenetic risk categories (including intermediate abnormal) were applied. The association of epidemiologic exposures with cytogenetic risk was evaluated on multivariable analysis using logistic regression models. Results: sAML was significantly associated with cytogenetic risk groups, as was prior SOT, healthcare occupation, farm habitat and toxin exposures [see Odds Ratios (OR), Table). In contrast, prior radiation for epithelial malignancies, smoking and BMI were not independently associated with specific cytogenetic risk categories. Conclusions: Specific epidemiologic exposures are significantly associated with AML cytogenetic risk categories suggesting a unique clinical phenotype. This supports a link to leukemogenesis and requires validation in a controlled prospective therapeutic study. [Table: see text]

AbstractPrimary appendiceal lymphomas (PAL) are a type of primary gastrointestinal non‐Hodgkin lymphoma (PGINHL) with an incidence of <1%. There is considerable discordance with regard to the optimal management of PGINHL. We describe two cases of PAL, perform a literature review, and discuss the available evidence for optimal treatment.

Introduction Novel high-risk groups have been identified in adult ALL, including secondary (sALL) and Philadelphia-like ALL (Ph-like, based on CRLF2, IgH, ABL2, JAK2 and other tyrosine kinase translocations), and those with minimal residual disease >0.1% (MRD+) after induction therapy. Novel targeted therapies are now routinely incorporated into 1st line regimens, including tyrosine kinase inhibitors (BCR-ABL1-pos), rituximab (CD20+) and blinatumomab (Blina) for MRD+. The impact of these novel high risk groups and therapies after alloHCT is unknown; therefore, we evaluated their impact on overall survival (OS), relapse rate (REL), non-relapse mortality (NRM) and acute and chronic GVHD. Methods We evaluated pts receiving 1st Allo-HCT for ALL at Mayo Clinic (Rochester, Phoenix, and Jacksonville) from 2008-2018 for outcomes of interest, specifically the impact of novel therapies and risk groups. Associations of patient factors with outcomes were examined using univariable (UVA) and multivariable (MVA) Cox proportional hazards regression models, where the cause-specific hazard of the given outcome was modeled to account for the competing risk of death. Results We identified 261 consecutive AlloHCT recipients during the study period. Median age at transplant was 48 years (18-72) and 147 (56.3%) were male. The median comorbidity (HCT-CI) score was 2 (0-8). 213 pts (81.6%) had B-lineage ALL, of which 85 (32.6%) were BCR-ABL-pos, 17 (6.5%) Ph-like (identified by FISH), 16 (6.1%) hypoploidy/near triploidy (Hy/Tri), and 67 (25.7%) pre-B ALL NOS. The remaining 48 (18.4%) had T-ALL. 30 pts (11.5%) had sALL (i.e. prior chemo/radiotherapy for another malignancy). HyperCVAD was the most common 1st line regimen (68.2%). 243 (93.1%) pts achieved Complete Remission (CR1) after induction therapy, and 203 (77.8%) were in CR1 at the time of alloHCT. Blina was administered for MRD+ in 14 pts (5.4%), and for relapsed/refractory ALL (R/R) in 13 (27% of R/R pts), 7 of whom received Blina as initial therapy for R/R. Donors were matched unrelated in 149 (57.1%), matched related in 98 (37.5%), and haploidentical in 14 (5.4%). Peripheral blood (PB) grafts were used in 233 (89.3%). 103 (54.5%) were donor:recipient (D:R) sex-matched, and 86 D:R mismatched [47 (24.9%) M:F; and 39 (20.6%) F:M]. Myeloablative conditioning was used for the majority (78.5%) mostly with Cy/TBI (60.5%). Standard GVHD prophylaxis regimens were used. Outcomes Median follow-up after transplant was 22.4 months (0.5-135), and 51 (19.5%) had REL. The 1, 2 and 5-year survival rates were 71.9%, 64.9%, and 54.1%, respectively (Figure 1). Acute GVHD developed in 144 (55.2%) and chronic GVHD in 100 (38.3%). Ph-like ALL, Blina for MRD+, Blina for R/R, sALL and CD20-pos had no independent impact on OS. In contrast, age>60, Hy/Tri, and >CR1 at alloHCT were associated with worse OS in UVA, however, in MVA only pre-B ALL NOS was associated with better OS. Female:male D:R status was associated with inferior OS. Blina for R/R disease was associated with increased risk of REL in UVA [HR 5.26 95% CI (1.33, 20.00), p=0.017], whereas other novel high risk groups had no impact on REL. In contrast, T-ALL, Hy/Tri and >CR1 at AlloHCT were associated with increased REL in UVA, but only T-ALL and Hy/Tri continued to predict for increased REL in MVA. Secondary ALL was associated with increased NRM in UVA [HR 1.96 95% CI (1.07, 3.57), p=0.028], whereas other novel high risk groups had no impact on NRM. In contrast, age>60, >CR1 at AlloHCT and D:R sex mismatch were associated with higher NRM in UVA, but only sex mismatch and >CR1 at AlloHCT were associated with higher NRM in MVA. TBI use was associated with higher risk of acute GvHD (p=0.008) and ATG use with lower risk chronic GVHD (p<0.001). Similarly non-PB grafts were associated with a lower risk of chronic GVHD (p=0.005). Results for OS, REL, NRM, acute and chronic GVHD analysis are shown in Table 1. Conclusion Novel high risk groups (CD20+, Ph-like and sALL) do not appear to adversely impact OS after alloHCT, although sALL was associated with increased risk of NRM. Interestingly, pre-B-ALL NOS appear to be associated with favorable OS. Novel targeted therapies also do not independently predict outcome, with the exception of Blina for R/R ALL which may be associated with REL after subsequent alloHCT (a subgroup for whom novel maintenance strategies should be explored). Our analysis highlights the importance of allo-HCT for novel high risk ALL subgroups. Disclosures Patnaik: Stem Line Pharmaceuticals.: Membership on an entity's Board of Directors or advisory committees. Kharfan-Dabaja:Daiichi Sankyo: Consultancy; Pharmacyclics: Consultancy. Foran:Agios: Honoraria, Research Funding.

Introduction There have been few recent therapeutic advances in T-acute lymphoblastic leukemia/lymphoma (T-ALL/LBL), a rare and aggressive malignancy. Intensive chemotherapy +/- allogeneic transplantation (AlloHCT) can effect cure, realized by only a minority of adults. In distinction to B-ALL, there have been no safe/effective targeted therapies developed in T-ALL/LBL. We therefore performed a focused tissue array to identify novel tumor proteins, including CCR4; evaluated together with established immunophenotypic (IP) T-ALL/LBL subtypes [early-T-precursor (ETP), early non-ETP (Pre/Pro-T), common & late thymocyte]. We evaluated the clinical phenotype and outcome together with clinical and epidemiologic data in patients diagnosed at the 3-site Mayo Clinic Cancer Center (MCCC: Rochester, MN; Phoenix, AZ; and Jacksonville, FL). Methods Following IRB approval, we identified 50 consecutive T-ALL/LBL patients from 1997 -2019 diagnosed at the 3-site MCCC. Available remnant leukemia/lymphoma paraffin blocks were identified (n=28). Using immunohistochemistry (IHC) we evaluated expression of leukemic antigens of interest (CCR4, CD47, BCL2, BCL6, PD-L1, CD38 and CD123); these were selected based on clinical availability of novel targeted agents, approved or in development for other indications. Tissue array studies were performed and interpreted by a single hematopathologist (L.J) blinded to clinical data, using positive and negative controls, according to standard techniques. We also evaluated the clinical and epidemiological parameters, including different IP subtypes [ETP: CD1a-, sCD3- & CD8-, Pre/Pro-T: CD1a-, sCD3- & CD8+ or CD8- (11 marker T-ALL IP score was used to distinguish CD8- ETP vs. Pre/Pro-T), thymic: CD1a+ and mature: CD1a- & sCD3+] (Khogeer et al., Br J Haematol., 2019) & the presence of the unique MLLT10-PICALM fusion (n=6), as well as therapy and outcome. Comparisons of characteristics and outcomes of interest were made using a Wilcoxon rank sum test (continuous characteristics) or Fisher's exact test (categorical characteristics), unadjusted Cox proportional hazards regression models (overall survival), or unadjusted logistic regression models (complete remission after first line treatment). Results We identified the expression of novel targetable tissue proteins in all adult T-ALL/LBL cases [Table 1]. BCL2, CD38 & CD47 were expressed in majority of the cases, and CCR4 [Fig. 1] was expressed in 11/28 (39.3%) cases. CCR4 expression was significantly more common in the Pre/Pro-T (75%) vs. other IP subtypes (19%) (p=0.011); there was no other association of CCR4 expression with clinical characteristics/outcomes. We also observed differences in T-ALL vs. T-LBL, specifically T-ALL patients were more likely to be male (71% vs. LBL 38.9%, p=0.038). Surprisingly, T-LBL patients were more likely to have a family history of leukemia/lymphoma (41.2% vs. T-ALL 0%, p<0.001). In addition, a significantly different distribution of IP subtype in T-LBL was noted [Table 2]. Not surprisingly, T-LBL patients were more likely to receive 1st line Hyper-CVAD therapy (88.2% v. T-ALL 43.3%, p=0.005), & had a longer delay from diagnosis to initiation of therapy (median 14 vs. 4 days, p=0.012), however this did not appear to impact clinical outcome. In contrast, we observed an important association of IP subtype with outcome after therapy, including complete remission (p=0.035), & a trend suggesting differences in survival (p=0.069) [Table 3]. However, MLLT10-PICALM was not associated with any unique clinical features or with outcome (death, p=0.85; CR, p=0.98), although 4/6 patients underwent AlloHCT, suggesting an important role of AlloHCT in this subgroup. No other significant association of epidemiologic risk factors with phenotype or outcome was noted. Conclusion Using focused tissue array we have identified expression of tumor associated proteins in T-ALL/LBL including CCR4 (Pre/Pro-T IP) & others, which may be targets of therapeutic agents like mogamulizumab. Furthermore, our analysis demonstrates an important clinical impact of IP subtype in T-ALL/LBL, as well as important clinical & therapeutic differences in T-ALL vs. LBL. This study will help guide development of targeted clinical trials in T-ALL/LBL (CCR4 trial already in development). Further analyses are planned to determine the clinical & epidemiologic association of genetic lesions in this cohort. Disclosures Foran: Agios: Honoraria, Research Funding.

Abstract Background: We previously demonstrated that putative clinical epidemiologic exposures associated with leukemia risk are prevalent among AML patients, and some are associated with unique cytogenetic risk group and with clinical phenotype (Finn, Cancer Epidemiol, 2015). Herein, we studied genome-wide DNA methylation in a cohort of AML patients to evaluate the association of individual hyper- and hypo-methylated CpG sites with epidemiologic exposures and overall survival. Methods: The Mayo Clinic AML Epidemiology Cohort is a highly annotated retrospective case series of 295 consecutive patients (pts) with AML diagnosed and treated at Mayo Clinic Florida and Arizona, with central cytogenetics performed in all cases. The prevalence of clinical epidemiologic exposures, past medical and family history as well as medication use and lifestyle was systematically obtained. After IRB approval, we interrogated the cytogenetic database and successfully obtained leukemia DNA from available remnant diagnostic cytogenetic cell pellets in a cohort of 148 AML patients in the Mayo epidemiology case series and performed an assessment of genome-wide DNA methylation using the Infinium HumanMethylation450K BeadChip. Samples were processed using the R Bioconductor package 'minfi' using Subset Within Array Quantile Normalization (PMID: 22703947). Individual CpGs that did not reach a detection p-value of <0.05 were filtered out. An internal control was included on each array to control for significant batch effects. To determine differential methylation status of 473,864 individual CpG sites (after exclusion of 11,648 that did not pass detection p-value QC), CpG sites with low interpatient variability, a standard deviation <0.05 of methylation values, and an interquartile range <0.05 of methylation values and X/Y chromosome probes were excluded from analysis. Epidemiologic and important clinical variables occurring in at least 10 pts or previously-identified as of unique interest but in fewer than 10 pts were evaluated, including sex, BMI, performance status and comorbidities, tobacco and alcohol use, family history of hematologic malignancy, medications of interest, history of toxin exposure, and history of immunosuppression and/or solid organ transplant or of secondary and therapy-related AML (t-AML). Associations of differential hypo or hypermethylation at 281,259 CpG sites with epidemiologic exposures/clinical variables were evaluated using Spearman's test of correlation (continuous or ordinal exposures), a Wilcoxon rank sum test (dichotomous exposures), or a Kruskal-Wallis rank sum test (multi-category exposures). A Bonferroni correction was applied for multiple testing, after which p <1.8 x 10-7 was considered as significant. To reduce the likelihood of false-negative findings, we additionally considered p <5 x 10-6 as indicating suggestive evidence of an association. CpG sites and associated gene and function was determined using Illumina manifest file & www.ncbi.nlm.nih.gov/gene. Results: Statistically significant associations (p <1.8 x 10-7) with individual epidemiologic and clinical exposures were identified for 109 unique CpG sites, corresponding to differential methylation in the genes listed in Table. Specifically, obesity (predominantly hypomethylation), specific cytogenetic lesions and risk group, and gender were highly significantly associated with unique CpG methylation status, but not smoking, toxin exposure, family history, comorbidity or performance status, secondary or t-AML, immunosuppression, medication use, of family history in the analysis. A further 353 unique CpG sites had defined suggestive associations (ongoing analysis). We also identified 8 additional CpG sites where differential methylation was associated with overall survival (p <5 x 10-6, Table). Conclusion: Obesity, cytogenetic lesions, and sex are associated with differential methylation of unique CpG sites at AML diagnosis, using stringent univariate statistical criteria. Significant CpG sites were identified in genes previously linked to AML biology and prognosis (DOCK6, HOXB3, MIR10A, FOXN3/CHES1, GPX1, MYST2/KAT7, PTPRD), but some represent novel findings in AML. These preliminary results suggest an association of some AML risk factors and clinical variables with unique gene methylation and will undergo validation in a prospective AML epidemiology dataset. Disclosures Cerhan: Jannsen: Other: Scientific Advisory Board; Nanostring: Research Funding; Celgene: Research Funding. Foran:Agios: Research Funding; Xencor, Inc.: Research Funding.

Abstract Background: Large case-control studies have identified common epidemiologic exposures associated with risk of AML development, including obesity [BMI ≥30 kg/m2] associated with 2-fold increase risk. However it is not known whether individual risk factors are associated with unique AML genetic lesions. We have recently identified a significant association of some common exposures including obesity with unique disease cytogenetic subgroups in a cohort of n=295 consecutive AML patients diagnosed and treated at Mayo Clinic [Cancer Epidemiol 39:1084, 2015]. Aberrant DNA methylation is a hallmark of cancer and has been reported in AML, and we therefore performed an analysis of global DNA methylation in patients in this cohort to determine whether specific epidemiologic exposures are associated with a unique pattern of DNA methylation. Methods: Patients in the initial cohort were previously reported. An available baseline AML bone marrow cytogenetic pellet from diagnosis was identified for n=148 patients (50.2%). Following IRB approval and using previously reported techniques [Leukemia 24:1283, 2010], leukemia DNA was isolated from the remnant diagnostic cytogenetic samples and processed on the HumanMethylation450 BeadChip array (Illumina). Individual CpG dinucleotides were assigned a Beta value (i.e. methylation value) ranging from 0 (unmethylated) to 1 (methylated). We excluded those with minor methylation changes (<0.2 between any two samples), resulting in 383,524 CpG's. For the purposes of this first analysis we have focused only on CpG's with an inter-quartile range (IQR, i.e. difference between 75th and 25th percentiles) of greater than 0.25, comprising a total of 39,657 CpG's. An unsupervised cluster analysis of differential methylation was generated ("Heatmap", Figure). Univariate and multivariable analyses were performed adjusting for age as a continuous variable to determine the association of individual exposures with unique methylation clusters (primary endpoint), and we also evaluated the association of clinical characteristics and clinical outcomes after therapy (secondary endpoint). The median age of patients was 69 yrs (range 19-89), male 65%, and Caucasian 89%. The median BMI was 27.2 (range 16.9-59.9), and n=47 patients (32%) had BMI ≥30. Standard cytogenetic risk groups were applied [favorable (n=8), normal (n=61), intermediate abnormal (n=32) and poor (n=47)]. Results: We identified 4 distinct methylation clusters (Figure: Cluster #1, Black horizontal bar; Cluster #2, Green; Cluster #3, Blue; and Cluster #4, Red). Multivariate analysis of select variables is shown in Table. Patients in cluster #1 were significantly more likely to have a BMI ≥30 [Odds Ratio (OR) 2.49, p=0.017] and more likely to have favorable risk cytogenetics although this was not statistically significant [OR 6.71, p=0.088], while those in cluster #2 were significantly more likely to be have a history of regular NSAID use [OR 3.26, p=0.03] and less likely to have normal [OR 0.40, p=0.044] and possibly intermediate abnormal cytogenetics [OR 0.34, p=0.051]. Those in cluster #3 were significantly younger [median age 48 yrs, p<0.001] and were also more likely to have prior history MDS although this was not statistically significant [OR 3.7, p=0.082]. Cluster #4 was not correlated with any distinct exposure apart from possible decreased alcohol use [OR 0.3, p=0.079]. History of past or current tobacco use, regular acetaminophen use, documented myelotoxin exposure, family history, and secondary AML did not correlate with a unique methylation cluster. There was no independent association of methylation cluster with overall survival (p=0.46), although among patients receiving intensive therapy cluster #2 was associated with lower complete remission rate [OR 0.28 (95% CI 0.09-0.86), p=0.026]. Conclusion: This first analysis of AML molecular epidemiology and clinical outcome suggests that specific disease risk factors (esp. obesity and age) are associated with unique leukemia DNA methylation clusters. In addition, aberrant methylation is associated with some clinical features and with outcome after therapy, suggesting that epidemiologic risk factors may have a prognostic impact. This provides novel insight into leukemogenesis, and will guide prospective studies of AML epidemiology and outcome. Figure. Figure. Disclosures Foran: medscape: Honoraria; pfizer: Honoraria; Millennium Pharmaceuticals, Inc.: Research Funding; novartis: Honoraria; karyopharm: Honoraria; boehringer ingelheim: Research Funding; agios: Research Funding; Cellerant: Research Funding.

AbstractAnaplastic large cell lymphomas (ALCLs) represent a relatively common group of T-cell non-Hodgkin lymphomas (T-NHLs) that are unified by similar pathologic features but demonstrate marked genetic heterogeneity. ALCLs are broadly classified as being anaplastic lymphoma kinase (ALK)+ or ALK−, based on the presence or absence of ALK rearrangements. Exome sequencing of 62 T-NHLs identified a previously unreported recurrent mutation in the musculin gene, MSCE116K, exclusively in ALK− ALCLs. Additional sequencing for a total of 238 T-NHLs confirmed the specificity of MSCE116K for ALK− ALCL and further demonstrated that 14 of 15 mutated cases (93%) had coexisting DUSP22 rearrangements. Musculin is a basic helix-loop-helix (bHLH) transcription factor that heterodimerizes with other bHLH proteins to regulate lymphocyte development. The E116K mutation localized to the DNA binding domain of musculin and permitted formation of musculin–bHLH heterodimers but prevented their binding to authentic target sequence. Functional analysis showed MSCE116K acted in a dominant-negative fashion, reversing wild-type musculin-induced repression of MYC and cell cycle inhibition. Chromatin immunoprecipitation–sequencing and transcriptome analysis identified the cell cycle regulatory gene E2F2 as a direct transcriptional target of musculin. MSCE116K reversed E2F2-induced cell cycle arrest and promoted expression of the CD30–IRF4–MYC axis, whereas its expression was reciprocally induced by binding of IRF4 to the MSC promoter. Finally, ALCL cells expressing MSCE116K were preferentially targeted by the BET inhibitor JQ1. These findings identify a novel recurrent MSC mutation as a key driver of the CD30–IRF4–MYC axis and cell cycle progression in a unique subset of ALCLs.

Cutaneous involvement by precursor T-cell lymphoblastic leukemia/lymphoma (T-ALL/LBL) is rare, and almost all cases are seen in association with bone marrow, blood, and/or lymph node involvement. Presentation with isolated skin involvement is very rare. Literature review revealed only one case report of primary cutaneous T-cell LBL. We discuss here another patient diagnosed with primary cutaneous T-cell LBL at our institute. This patient was initially misdiagnosed as having peripheral T-cell lymphoma NOS. Cytogenetic analysis showed the CDKN2A deletion (−9p21×2) in addition to three intact copies of ABL1 (+9q34). Although she failed multiple lines of intensive chemotherapy, her disease remained confined to the skin. We believe that this presentation of T-LBL is underreported, and many patients are likely misdiagnosed as having high-grade cutaneous T-cell lymphomas. With this case and literature review, we would like to highlight the importance of keeping lymphoblastic lymphoma on the differential diagnosis of cutaneous T-cell lymphoma-like lesions to avoid delay in diagnosis and inappropriate treatment of this aggressive disease.