Allmer, Cristine

Abstract Background: Follicular lymphoma (FL) is an indolent lymphoma generally considered incurable with current standard therapies. Recent advances have resulted in prolongation of overall survival (OS) for patients with FL such that death from competing causes may now limit the mortality impact of FL in some patient groups. Identification of patients for whom FL related mortality is expected to be minimal, or conversely those at high risk of disease related mortality, is highly clinically relevant. We recently reported (Maurer et al, JCO 2014;32:1066-73) that event-free survival (EFS) at 24 months from diagnosis is a robust endpoint for disease related outcome in patients with diffuse large B cell lymphoma (DLBCL). Here we use the same approach to assess if landmark timepoints of EFS can stratify subsequent OS in FL. Methods: All newly diagnosed grade 1-3a FL patients enrolled on the University of Iowa/Mayo Clinic Lymphoma SPORE Molecular Epidemiology Resource (MER) from 2002-2011 were eligible. Patients with grade 3b FL, transformation, or presence of DLBCL at diagnosis were excluded. EFS was defined as time from diagnosis to progression, relapse, re-treatment, or death due to any cause. EFS12 was defined based on EFS status 12 months after date of diagnosis; we also assessed EFS at 24 months. OS was defined as time from a specific timepoint (diagnosis, event, or EFS12) until death due to any cause. OS was compared to age and sex matched survival in the general US population using an expected survival approach and standardized mortality ratios (SMR) and 95% confidence intervals (CI). Outcomes were examined in all patients as well as selected subsets defined by initial treatment. Replication was performed using a Lyon, France hospital registry of newly diagnosed grade 1-3a FL patients enrolled from 2002-2010 with survival compared to French population data. Results: 936 patients with grade 1-3a FL were eligible from the MER. The median age at diagnosis was 60 years (range 19-91) and 53% were male. 86% had grade 1-2 disease and 14% grade 3a disease. 23% had high-risk disease (FLIPI score ≥3). 347 (37%) were initially treated with alkylator or anthracyline based immunochemotherapy (IC), 113 (12%) rituximab monotherapy (RM), and 318 (34%) were initially observed; the remaining patients received radiation for limited stage disease (8%), chemotherapy without rituximab (4%), or other therapy (5%). At a median follow-up of 59 months (range 1-131), 403 patients (43%) had an event and 120 patients (13%) had died; 155 patients (17%) failed to achieve EFS12. From diagnosis, patients with FL had inferior OS compared to the general population (SMR=1.24, 95% CI: 1.03-1.48, p=0.018); results were similar for patients with grade 1-2 (SMR=1.23) and grade 3a (SMR=1.29) disease. The survival deficit was eliminated in patients who achieved EFS12 (SMR=0.85, 95% CI: 0.66-1.09, p=0.20, Fig 1a), and so we did not pursue modeling EFS24. In contrast, patients with a non-death event (relapse, retreatment, or progression) within 12 months of diagnosis had inferior subsequent OS (SMR=3.90, 95% CI: 2.89-5.25, p=4.8x10-19, Fig 1b). These results were replicated in 153 FL patients from the Lyon registry: patients who achieved EFS12 had equivalent subsequent OS to the general French population (SMR=0.82, 95% CI: 0.39-1.72, p=0.60, Fig 1c), while patients who failed to achieve EFS12 had poor subsequent OS (SMR=4.16, 95% CI: 1.34-12.91, p=0.013, Fig 1d). In subset analysis of the MER, patients achieving EFS12 initially treated with IC (SMR=0.98, 95% CI: 0.63-1.54, p=0.93), RM (SMR=0.66, 95% CI: 0.31-1.38, p=0.27), or observation (SMR=0.73, 95% CI: 0.47-1.15, p=0.18) had equivalent OS to the population. In contrast, for patients not achieving EFS12, OS after event was inferior for those treated with IC (SMR=15.0, 95% CI: 10.22-22.05, p=2.0x10-43) or RM (SMR=4.96, 95% CI: 1.24-19.84, p=0.023), while trending towards inferior for those initially observed (SMR=1.55, 95% CI: 0.81-2.99, p=0.19). Summary: Events within the first 12 months of diagnosis are associated with poor OS in FL, especially in patients treated with IC, while patients achieving EFS12 have excellent subsequent survival, overall and in the main treatment subsets. Prognosis for FL patients should be recalibrated at 12 months from diagnosis. EFS12 will be useful in counseling and should be considered as an endpoint in studies of newly diagnosed FL. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Abstract Purpose: The purpose of this study was to examine whether quality of life (QOL) scores at diagnosis predict survival among patients with aggressive lymphoma, controlling for clinical variables and histological subtype. Patients and Methods: Newly diagnosed lymphoma patients were prospectively enrolled within 9 months of diagnosis in the University of Iowa/Mayo Clinic SPORE Molecular Epidemiology Resource. Eligibility criteria included age ≥18 years; exclusion criteria included HIV-positive, non-US residency, and inability to participate in an interview conducted in the English language. All pathology was reviewed by a lymphoma hematopathologist. For purposes of this analysis, we considered “aggressive lymphoma” to include the histological subtypes diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL) grade III, mantle cell lymphoma, T-cell lymphoma excluding mycosis fungoides and cutaneous types, and other (Burkitt lymphoma, high grade B-cell lymphoma NOS, grey zone, etc). Baseline QOL was measured at time of registry enrollment with the Functional Assessment of Cancer Treatment-General scale (FACT-G), which measures 4 QOL domains: physical, social/family, emotional, and functional well-being (WB); a single item Linear Analogue Self Assessment (LASA) for measuring overall QOL; and a LASA measuring spiritual WB. Patients who did not complete ≥80% of FACT-G QOL questions were excluded. Clinical data were abstracted and patients were systematically followed for event-free and overall survival (EFS, OS). QOL scores were transformed to a 0-100 scale for modeling. Generalized additive models were used to account for timing of QOL assessment in relation to treatment and diagnosis; residuals from these models were then used to assess association between QOL measures and outcome via univariate and multivariable (adjusted for IPI and subtype) Cox proportional hazards models. Results: From September 2002 to December 2009, 701 patients with aggressive lymphoma who completed baseline QOL questionnaires were enrolled and eligible for this analysis. ; Subtype was 60% DLBCL, 14% mantle cell, 11% FL grade III, 13% T-cell, and 1% other subtypes. The median age was 62 years (range 18-92), and 58% were male. At a median follow-up of 71 months (range 6-128), 316 patients (45%) had an event and 228 patients (33%) died. Patient-reported QOL was prior to therapy in 330 patients (47%) and after initiation of therapy in 371 (53%). The median overall FACT-G total score was 83 (IQR 70-93), and the median scores on the functional WB, physical WB, emotional WB, and social/family WB were 18 (13-23), 22 (17-26), 18 (16-21), and 25 (22-27), respectively. 42 patients (6%) reported a clinically deficient QOL score on the FACT-G (score ≤50 on 0-100 scale). When comparing QOL scores between those who were on active therapy versus those who were not yet treated, physical WB (p<0.0001) functional WB (p=0.00014), and total FACT-G (p=0.0013) were lower in those on active treatment. All QOL measures but emotional WB were significantly associated with OS (all p<0.04), and all but LASA spiritual WB remained significant after adjusting for IPI and NHL subtype. Associations were stronger for OS than EFS. For OS, the strongest associations were with total FACT-G (adjusted HR=0.86, 95% CI: 0.79-0.94, p=0.00062) and functional WB (adjusted HR=0.88, 95% CI: 0.83-0.93, p<0.0001). The one-question QOL LASA was also associated with OS (adjusted HR=0.92, 95% CI: 0.87-0.97, p=0.0041). The strongest association with EFS was with functional WB (adjusted HR=0.92, 95% CI: 0.88-0.97, p=0.0015). Patients with a clinically deficient QOL (baseline overall QOL ≤50) had a median OS of 59 months compared to 121 months for patients with baseline QOL >50, p=0.0002, figure 1. In a sensitivity analysis, raw QOL was modeled for OS and EFS in the subset of patients with QOL assessed prior to treatment; hazard ratios were consistent with the adjusted QOL HR in all patients. Conclusions: In this large sample of patients with aggressive lymphoma, baseline QOL is predictive of OS and EFS, even after adjustment for known factors related to survival (IPI and histological subtype). This provides evidence that patient-reported outcomes are as important as other more objective previously confirmed IPI measures. QOL should be assessed at diagnosis as a prognostic factor in patients with aggressive lymphoma. Figure 1 Association of QOL with overall survival. Figure 1. Association of QOL with overall survival. Disclosures No relevant conflicts of interest to declare.

Abstract Abstract 684 Background. Obesity has been associated with an increased risk of developing of lymphoma, particularly DLBCL. Genetic variation in several energy balance genes, including leptin (LEP), its receptor LEPR, ghrelin (GHRL) and neuropeptide Y (NPY) have been shown to modulate risk of lymphoma, although their impact on survival is not known. In contrast, recent studies suggest that patients either overweight (Body Mass Index (BMI), 25 to<30 kg/m2) or obese (≥30 kg/m2) may have a superior prognosis in DLBCL, although its impact in FL and in the current treatment era is not known. We evaluated the association of BMI at the time of diagnosis, 2 years prior to diagnosis and at ages 18, 35 and 50 years, along with germline genetic variation in selected energy balance genes with event-free (EFS) and overall (OS) survival in DLBCL and FL. Methods. We evaluated these associations in a prospective cohort of newly diagnosed DLBCL (N=764) and FL (N=649) patients enrolled at the Mayo Clinic and University of Iowa from 2002–2009 as part of the Molecular Epidemiology Resource. Peripheral blood DNA, clinical data and self-reported historical weights were collected at enrollment. Patients received standard disease treatment strategies, and were systematically followed for disease progression, re-treatment, and death. Single nucleotide polymorphisms (SNPs) were genotyped as part of a custom Illumina iSelect panel. The most prevalent homozygous genotype was used as the reference group, and each SNP was modeled as having a log-additive effect in the regression model. Cox proportional hazards regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (CI) for the association of the obesity measures and SNPs with outcome, adjusted for treatment class and subtype specific prognostic factors (IPI, FLIPI). Results. For DLBCL patients (median age 63 years, range 18–92) the median follow-up was 47 months (range 0–101); 58% had an IPI 0–1 risk disease; and 87% received immunochemotherapy. For FL patients (median age 60 years, range 25–94), the median follow-up was 60 months (range 11–110); 24% had FLIPI score 3–5; and main initial treatments were observation (32%), rituximab (R)-monotherapy (12%), CVP +/−R (19%), and anthracycline +/−R (22%). The main results for BMI and prognosis are reported in the Table. For DLBCL, obese patients at diagnosis had better EFS and OS; this association was weaker for BMI at age 50 and 35 years. In contrast, for FL, overweight and obese patients at diagnosis had a poorer EFS and OS, and this was also observed for BMIs two years before diagnosis and at ages 50 and 35 years. In FL, the LEP SNP rs2167270 was associated with inferior EFS (HR=1.25; 95%CI 1.05–1.49) and OS (HR=1.28; 95%CI 0.90–1.83). The NPYSNP rs16147 was associated with inferior FL OS (HR=1.50; 95%CI 1.03–2.17) but was not associated with EFS. None of the energy balance and metabolism SNPs were associated with DLBCL EFS or OS. Conclusions. Higher BMI prior to and at diagnosis, as well as SNPs in LEP and NPY,were significantly associated with FL prognosis, with inferior outcomes in overweight and obese patients. These novel findings in the modern treatment era suggest an important role of obesity pathways in FL prognosis. In DLBCL, we confirm the recent findings of improved survival in DLBCL in obese patients, and extend this to the modern treatment era and BMI earlier in life. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 3637 Background: A recent large genome-wide association study confirmed the important role of the major histocompatibility complex in classical Hodgkin Lymphoma (cHL) susceptibility (rs6903608), but also identified significant associations with genetic variants outside of the HLA region at chromosomal regions 2p16.1 (REL, rs1432295), 8q24.21 (PVT1, rs2608053, rs2019960), and 10p14 (GATA3, rs501764, rs485411). These single nucleotide polymorphisms (SNPs) are localized in or near important genes involved in lymphoma pathogenesis: REL, which encodes for c-rel, is a member of NFkB pathway; PVT1, which is downstream of the MYC locus, encodes for non-coding microRNAs, some of which have oncogenic properties; and GATA3, which is a critical transcriptional factor for priming lymphoid T-cells for Th2 phenotype, is aberrantly expressed in HL. We investigated whether these six SNPs described in cHL etiology also influence the outcome of newly diagnosed cHL patients in two independent cohorts. Patients and Methods: The first cohort consisted of 342 cHL patients from the Groupe d'Etude des Lymphomes de l'Adulte (GELA) who were prospectively enrolled between 1998 and 2002. The second cohort consisted of 259 patients prospectively enrolled in the University of Iowa/Mayo Clinic SPORE Molecular Epidemiology Resource (MER). DNA for genotyping was extracted from a blood samples collected at diagnosis, and genotyping was conducted using Taqman technology (ABI Prism 7000, Applied Biosystems) for HLA-DRA, PVT1, GATA3 and PCR-based restriction fragment length polymorphism for REL in GELA cohort and Illumina iSelect for the 6 SNPs in MER cohort. Both cohorts were followed for disease progression, re-treatment and death. In the GELA series, the median age was 32 years (range, 15–93), 57% were male, and 72% had an Ann Arbor stage I–II. Histology was nodular sclerosis in 84%, and 29% of available cases (n=40/140) were EBV positive. In the MER series, the median age was 38 years (range, 18–89), 53% were male, and 53% had an Ann Arbor stage I-II. Nodular sclerosis was the most common subtype (62%) and 23 of 115 available cases were EBV positive (20%). The International Prognostic Score (IPS) was 0–1 in 46% and 52% of GELA and MER patients, respectively. ABVD regimen was used for 60% and 88% of the GELA and MER patients, respectively. The 5-year progression free survival (PFS) was 83% and 75% in GELA and MER cohorts, respectively. We estimated the prognostic value of each individual SNP for PFS for allelic and genotypic (dominant and recessive) modes of transmission in each series. In order to increase power, we then performed a meta-analysis of the two studies. Results: Consistent with prior reports, the minor allele frequencies for REL (rs1432295), HLA-DRA (rs6903608), PVT1 (rs2608053), PVT1 (rs2019960), GATA3 (rs501764) and GATA3 (rs485411) were 0.44, 0.50, 0.43, 0.29, 0.22, 0.28 in the GELA and 0.46, 0.41, 0.46, 0.27, 0.25, 0.31 in the MER series, respectively. In GELA, GATA3 (rs501767) was associated with PFS in the ordinal model (HR=1.66, 95%CI 1.11–2.49, P=.01), while there were suggestive associations for PVT1 (rs2608053) (HR=1.76, 95%CI 0.95–3.26, P=.07) and REL (rs1432295) (HR=0.63, 95%CI 0.37–1.06, P=.08) in the dominant model. In the MER, PVT1 (rs2608053) was associated with PFS in the ordinal model (HR=1.42, 95%CI 0.99–2.03, P=.06), while there was a suggestive association with HLA-DRA (rs6903608) in the dominant model (HR=0.64, 95%CI 0.39–1.05, P=.08) for whole cohort but also for EBV negative cHL (HR=0.41, 95%CI 0.17–0.99, P=.05). In a meta-analysis of the two studies, PVT1 (rs2608053) was significantly associated with PFS in the ordinal (HR=1.34, 95%CI 1.04–1.73, P=.02) and the dominant (HR=1.88, 95%CI 1.20–2.95, P=.006) models. Further adjustment for IPS had minimal impact on these associations (ordinal HR=1.30, 95%CI 1.00–1.70; dominant HR=1.78, 95%CI 1.13–2.83). Patients with PVT1(rs2608053) GG, GA, AA genotypes had a 5-year PFS rate of 83%, 74% and 73% in GELA and 86%, 70%, and 73% in the MER, respectively. Conclusions: The non-HLA cHL susceptibility locus PVT1 (rs2608053) was associated with PFS in two independent cohorts of cHL cases, and after adjustment for the IPS. While associations with GATA3 and REL were only observed in one of the studies, their evaluation in other studies is warranted. Functional studies of PVT1 in cHL pathogenesis should be pursued. Disclosures: No relevant conflicts of interest to declare.

Abstract In this month's CME article, an analysis of over 1300 patients with diffuse large B-cell lymphoma (DLBCL), Wang et al demonstrate that the approximately 13% of patients with concurrent indolent lymphoma at diagnosis had the same prognosis as those patients with isolated DLBCL.

PURPOSE In patients with diffuse large B-cell lymphoma (DLBCL), most relapses occur within the first 2 years of diagnosis. We sought to define the rate and outcome of late relapses that occurred after achieving event-free survival at 24 months (EFS24). METHODS We prospectively followed 1,324 patients with newly diagnosed DLBCL from 2002 to 2015 and treated with immunochemotherapy. Cumulative incidences of late DLBCL and indolent lymphoma relapses were analyzed as competing events. Postrelapse survival was defined as time from first relapse to death from any cause. RESULTS In 847 patients who achieved EFS24, the cumulative incidence of late relapse was 6.9% at 3 years, 9.3% at 5 years, and 10.3% at 8 years after EFS24. The incidence of DLBCL relapse was similar in patients with DLBCL alone at diagnosis (6.3% at 5 years), compared with patients with concurrent indolent lymphoma at diagnosis (5.2%; P = .46). However, the rate of indolent lymphoma relapse was higher in patients with concurrent indolent lymphoma (7.4% v 2.1% at 5 years; P < .01). In patients with DLBCL alone, the rate of DLBCL relapse was similar in the germinal center B-cell–like (GCB) (4.1% at 5 years) and non-GCB (4.0%; P = .71) subtypes, whereas the rate of indolent lymphoma relapse was higher in patients with the GCB subtype (3.9% v 0.0% at 5 years; P = .02). Postrelapse survival was inferior for patients who relapsed with DLBCL than for those who relapsed with indolent lymphoma (median 29.9 months v unreached; P < .01). CONCLUSION Patients with DLBCL with a concurrent indolent lymphoma and those with the GCB subtype had a higher rate of late relapse, owing to increased relapses with indolent lymphoma. Patients who relapsed with DLBCL had a worse prognosis than those who relapsed with indolent lymphoma.

SummaryThe diabetes mellitus (DM) drug metformin targets mechanistic/mammalian target of rapamycin and inhibits lymphoma growth in vitro. We investigated whether metformin affected outcomes of newly diagnosed diffuse large B‐cell (DLBCL, n = 869) and follicular lymphoma (FL, n = 895) patients enrolled in the Mayo component of the Molecular Epidemiology Resource cohort study between 2002 and 2015. Hazard ratios (HR) and 95% confidence intervals (CIs) adjusted for age, sex, body mass index, prognostic index and treatment were used to estimate the association of metformin exposure (No DM/No metformin; DM/No metformin; DM/Metformin) with event‐free (EFS), lymphoma‐specific (LSS) and overall (OS) survival. Compared to No DM/No metformin DLBCL patients, there was no association of DM/Metformin (n = 48; HR = 1·05, 95% CI 0·59–1·89) or DM/No metformin(n = 54; HR = 1·41, 95% CI 0·88–2·26) with EFS; results were similar for LSS and OS. Compared to No DM/No metformin FL patients, there was no association of DM/Metformin (n = 37; HR = 1·16, 95% CI 0·71–1·89) or DM/No metformin (n = 19; HR = 1·16, 95% CI 0·66–2·04) with EFS; results were similar for LSS. However, DM/Metformin was associated with inferior OS (HR = 2·17; 95% CI 1·19–3·95) compared to No DM/No metformin. In conclusion, we found no evidence that metformin use was associated with improved outcomes in newly diagnosed DLBCL and FL.

AbstractIndolent B‐cell lymphomas other than follicular lymphoma account for up to 10% of all B‐cell neoplasms. While they typically follow a slowly progressive course, some patients experience rapid progression and early mortality. Prognostic scoring systems have not been adopted, hindering the ability of clinicians or researchers to predict outcomes, or risk‐stratify patients during clinical trials. To address this, we evaluated the utility of existing prognostic indices and novel, early disease‐related outcomes, to predict subsequent long term survival. Baseline characteristics and outcomes data were generated from a longitudinal cohort study that prospectively enrolled 632 patients newly diagnosed with marginal zone lymphoma, lymphoplasmacytic lymphomas, or B‐cell lymphomas not otherwise specified, beginning in 2002. The International Prognostic Index (IPI), Follicular Lymphoma International Prognostic Index (FLIPI), and MALT International prognostic index (MALT‐IPI) demonstrated c‐statistics that ranged from 0.593‐0.612 for event‐free survival (EFS), and 0.683‐0.714 for overall survival (OS). Patients who attained event‐free survival at 12 months (EFS12) experienced similar mortality to the US general population (standardized mortality ratio [SMR] 1.19; 95% CI 0.95‐1.46). Patients who did not attain EFS12 had subsequent worse morality (SMR 3.14 (95% CI 2.05‐4.59). The MALT‐IPI demonstrated utility in predicting subsequent long‐term outcomes among patients with non‐follicular indolent B‐cell lymphomas. This index should be used by clinicians giving guidance to patients at the time of initial diagnosis, and risk stratification during clinical studies. The divergent long‐term outcomes experienced by patients who do or do not attain EFS12 suggest there exists a subset of patients who harbor high‐risk disease. Future research efforts should focus on methods to identify these patients at the time of diagnosis, in order to enable risk‐tailored therapy.