Search

The US veteran population has a high proportion of chronic lymphocytic leukemia (CLL) risk factors. Using the Veterans Health Administration (VHA) population, we conducted a retrospective chart review of 1205 CLL patients who initiated treatment with a novel oral agent. For 1L ibrutinib, 33% (n = 107) discontinued therapy during the study, of which 64% discontinued due to adverse events (AEs). For relapsed/refractory (R/R) ibrutinib, 35% (n = 262) discontinued therapy, of which 63% discontinued due to AEs. For R/R venetoclax, 31% (n = 27) discontinued therapy, of which 41% were due to AEs. For idelalisib, 84% (n = 41) discontinued therapy, of which 54% were due to AEs. This real-world study suggests that AEs play an important role in dose reductions and discontinuations; however, physician inexperience in using these drugs when they were first introduced could be part of what is leading to these negative outcomes.

8556 Background: Durvalumab is an FDA-approved immunotherapy for the treatment of adults with UnResectable stage III non-small cell lung cancer (UR-NSCLC) without disease progression following concurrent chemoradiotherapy (CRT). There are limited real-world data regarding Durvalumab treatment initiation delays (TIDs) and reasons for them in the UR-NSCLC population. Methods: Patients with stage III UR-NSCLC receiving consolidation Durvalumab at the Veterans Health Administration (VHA) between January 1, 2017 and June 30, 2020 were selected from the VHA database using ICD-10, HCPCS, and J codes. All had the opportunity to be treated for 12 months and were followed from Durvalumab initiation through the earliest of their last VHA visit, loss to follow up, death, or the study’s end (and excluded if Durvalumab therapy was ongoing at the study’s end). Trained data abstractors determined the occurrence and reasons for TIDs (> 6 weeks from end of CRT to initiation of Durvalumab as in the PACIFIC trial) by chart review. Results: 935 patients were eligible for analysis (median age = 69 years; 95% males; 16% with ECOG performance status >1). TIDs occurred in 39% of the patients (Table). Durvalumab was initiated 61 days (median) from the end of CRT in TID patients vs. 31 days for those without TIDs. There were no significant (α

6526 Background: Racial and ethnic disparities in access to care and outcomes are well-established and are critical issues across several malignancies, including DLBCL. Previous studies from national registry datasets have shown racial disparities in DLBCL disease characteristics, treatment and outcomes. The VHA is an equal access system providing a unique environment to investigate cancer disparities across the disease continuum. Methods: This is a retrospective chart review of 4033 randomly selected patients with an ICD code for lymphoma treated within the VHA between 01/01/2011 and 12/31/2017. Data abstractors collected baseline patient and disease characteristics and treatment responses for those with an initial diagnosis of DLBCL in that time frame. Survival time was determined via electronic health record query on 11/30/2021. Chi-square tests were used to analyze relationship between race and variables of interest. Cox proportional hazards model was used to estimate hazard ratios (HR) for race and controlling factors. Results: 2141 DLBCL patients met our inclusion criteria. 97% were male. Majority were Non-Hispanic Whites (NHW 75%) followed by Non-Hispanic Blacks (NHB 12.5%), Hispanics (H 5.7%) and others (O 6.8%). NHB were diagnosed at younger median age (63 years) when compared to the NHW, H and O (68 years). There was no statistically significant difference in stage at diagnosis, IPI score, cell of origin (COO) and hit status amongst racial subgroups. Outcomes analysis (Table) revealed similar treatment and response rates, median OS, 1- and 2- year survival across all racial subgroups. However, after adjusting for age, IPI, COO, and exposure to agent orange, and including up to 10-years of survival data, H had 36% lower risk of death (HR=0.64, 95% CI 0.44-0.93) than NHW, while NHB and O had similar outcomes to NHW. Conclusions: This large retrospective study is a continuation of our group’s work (Williams et al, 2020) that doubles the cohort size and confirms that when standard of care therapy is given with equal access to care, short-term treatment and survival outcomes are same for all races. Further studies are needed to analyze risk factors associated with differences in long term outcomes.[Table: see text]

8554 Background: The PD-1/PD-L1 pathway is a mechanism of immune evasion and disruption of this pathway with immune checkpoint inhibitors (ICIs) has shown clinical benefit in multiple malignancies. Based on results from the PACIFIC trial, durvalumab is approved as consolidation therapy in patients (pts) with stage III unresectable non-small cell lung cancer (UR-NSCLC) without progression following concurrent chemoradiotherapy (cCRT). Durvalumab has been used extensively in Veterans Health Administration (VHA) facilities, providing an opportunity to evaluate durvalumab treatment interruptions (TI), treatment discontinuations (TD), and the reasons for these on a national scale. Methods: Patients with stage III UR-NSCLC receiving durvalumab consolidation immunotherapy at the VHA between January 1, 2017 and June 30, 2020 with a minimum follow up for 12 months were included using ICD-10, HCPCS, and J codes and followed from their durvalumab start date through the earliest of last VHA visit, loss to follow up, death, or end of study (excluded if durvalumab therapy was ongoing at the end of the study, because the full treatment course could not be determined). TI were defined as durvalumab infusions separated by >28 days. Reasons for TI and TD are presented descriptively. Durations are reported using medians and interquartile ranges (IQR). Results: 935 pts were included (median age = 69 years; 95% males; 96% current or former smokers; 70% with COPD; histologies [squamous (50%), non-squamous (43%), other/missing (7%)]; and 77% with carboplatin-paclitaxel as their platinum-based CRT). Durvalumab TI were experienced by 19% of pts (median [IQR] number of TI = 1 [1-1], median [IQR] TI duration = 53 days [39-90]). The main reasons for TI were toxicity (8%) and social reasons (3%) (Table). The median duration of treatment (DoT) with durvalumab (TI included) was 9.0 months (IQR 2.9-11.8). Durvalumab TD occurred in 59% of pts. Top reasons for discontinuation across all 935 pts included disease progression (24%) and toxicity (18%) (Table). Conclusions: In this real world analysis of national VHA data, durvalumab DoT was similar to PACIFIC despite having a patient population with worse prognostic factors (e.g. more males, squamous, COPD) with 8% of VHA pts experiencing TI and 18% TD due to toxicity. Patients could benefit from additional efforts to prevent, identify, and manage toxicities in the UR-NSCLC population [Table: see text]

8526 Background: Evidence from the PACIFIC study and real-world data highlight the benefit of durvalumab in patients with stage III unresectable non-small cell lung cancer (UR-NSCLC). However, limited literature exists regarding disparities in durvalumab treatment patterns such as treatment initiation delays (TID), treatment interruptions (TI), number of doses, duration of therapy (DOT), adverse effects (AEs), and treatment discontinuation (TD) in minority populations. Methods: Patients with stage III UR-NSCLC and a self-reported racial identity of Black or White treated with durvalumab following chemoradiotherapy (CRT) at any Veterans Health Administration (VHA) facility from January 1, 2017 to June 30, 2020 were included. Patients were followed from their date of durvalumab initiation through the earliest of their last VHA visit, loss to follow up, death, or end of the study; therefore, all patients had the opportunity to be treated for 12 months. Patients were excluded if durvalumab therapy was ongoing at the end of the study. Patient charts were retrospectively reviewed for baseline characteristics and durvalumab treatment patterns including TID (>42 days from end of CRT to durvalumab start), TI (>28 days between doses), number of doses, DOT, AEs, and TD. Nominal variables were compared using chi-square/Fisher’s exact tests. Continuous variables were compared using Student’s t-tests/Wilcoxon Rank Sum tests. Results: Among 924 patients, Black patients were younger than White patients (median age 67 years [IQR, 63-71] vs. 70 years [IQR, 65-73]; p

Introduction Since the introduction of NAs in 2013, the treatment paradigm for CLL has changed significantly with the increased uptake of NAs for first line (1L) and refractory CLL. NA have introduced a personalized approach to CLL treatment that considers patient baseline characteristics, individual risk factors and preferences. 1 Despite improvement in survival outcomes with CLL, Black patients with CLL have demonstrated inferior overall survival as compared to White patients. 2 Studies have shown a difference in disease biology between Black and White patients. 3 Differences in disease biology and access to medical care and treatment have contributed to racial disparities. The purpose of this study was to assess the uptake of NAs in the VHA where access to care is equal for both Black and White patients. Methods: A retrospective study was conducted which observed adult patients with an ICD code for CLL managed in the VHA from 2014 to 2017. Electronic and manual data was extracted to review start and stop dates for 1L CLL therapies. Data up to twenty years prior to the CLL treatment initiation date was reviewed to define the baseline variables. Descriptive statistics were used to summarize the data, and chi-square and Student's t-tests were used to compare NA use, baseline characteristics, healthcare resource use, and complications. A two-step nominal logistic regression model was used to determine the effect of race on receipt of NAs. Baseline patient characteristics were compared with bivariable statistical tests and variables with a p-value less than 0.10 were subsequently entered into the multivariable nominal logistic regression model. P-values less than 0.05 were considered statistically significant Results: The study included 565 patients; 86% were White and 14% were Black. Black patients were younger than White patients with a median age of age 66 and 69 respectively, p< 0.0001 (Table 1). Overall, Black patients were less likely to receive NAs than White patients (14% vs. 26%, p=0.0165) (Table 1). However this difference narrowed over the study period. (Figure 2) Traditional chemotherapy/chemoimmunotherapy (CT/CIT), was more common in Black patients. White and Black patients experienced similar treatment outcomes and health care utilization (Table 2). When all baseline variables listed in Table 1 were compared for Black and White ptients, the following variables had p-values less than 0.10 and were subsequently entered into a multivariant model: patient age (p Conclusions: In this retrospective cohort study, there was a statistically significant difference in the use of NAs between Black and White patients with CLL in the VHA for the study period. However, when NA use was examined by year, the disparity was largest in the early study years with reduced differences in NA utilization over time. Potential limitations of this study include the smaller number of patients, biologic factors that were not captured, and potential confounding factors such as physician experience with NA use. With the introduction of more NAs, the VHA should pay close attention to ensure all patients receive equal access to new therapies. We hope to implement interventions to ensure that patients of all races have equal access to NAs as soon as they are introduced. References 1. Burger, J, O'Brien, S. Evolution of CLL treatment - from chemoimmunotherapy to targeted and individualized therapy. Nat Rev Clin Oncol. 2018 Aug;15(8):510-527. 2. Sikander Ailawadhi, Dongyun Yang, Nidhi Jain, Mie Mie Thinn, Wendy Cozen, Asher Chanan-Khan; Ethnic Disparities in Chronic Lymphocytic Leukemia Survival: A SEER Database Review. Blood 2012; 120 (21): 757. 3. Flowers, RC, Po, B. Racial differences in chronic lymphocytic leukemia. Digging Depper. 2013 Oct 15;119(20):3593-5. Figure 1 Figure 1. Disclosures Obodozie-Ofoegbu: University of Texas at Austin: Ended employment in the past 24 months. Nooruddin: AstraZeneca: Research Funding. Ryan: AstraZeneca: Current Employment, Current equity holder in publicly-traded company. Jones: AstraZeneca: Research Funding. Frei: AstraZeneca: Research Funding.

Introduction Around 2-5% of patients with diffuse large B-cell lymphoma (DLBCL) will experience central nervous system (CNS) relapse resulting in a poor prognosis. The Central Nervous System International Prognostic Index (CNS-IPI) is a validated risk model used to help identify DLBCL patients receiving R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone) chemotherapy who are at risk for CNS relapse. CNS prophylaxis is recommended for those with high-risk CNS-IPI scores; however, the role of CNS prophylaxis has been called into question given recent large datasets showing no impact on CNS relapse. The purpose of this study is to evaluate the characteristics of Veterans who experienced CNS relapse within the Veterans Healthcare Administration (VHA) and to validate the CNS-IPI risk stratification within this population. Methods Trained abstractors performed a retrospective chart review of 3287 lymphoma patients seen in the VHA nationwide between 01/01/2011 and 12/31/2017. Figure 1 describes the selection of the study cohort. We evaluated baseline patient and disease characteristics including CNS-IPI score, performance of diagnostic lumbar puncture (LP) and first-line chemotherapy regimen. Pathology reports to identify cell of origin (COO), and additional risk factors for CNS relapse present at the time of original diagnosis including HIV associated lymphoma, testicular lymphoma, and high-grade B-cell lymphomas (HGBLs) defined as double or triple-hit DLBCL were evaluated. We also assessed response to first-line treatment, type of CNS prophylaxis used, including number of doses, and time to CNS relapse. Results A total of 1621 patients met criteria for analysis. Patients were predominately male, white, had a median age of 67, and presented with advanced disease (Table 1). At the time of diagnosis, 81% of the cohort had an ECOG performance status of 0-2, 73% received a CHOP based regimen, and 52% were designated as having a high-risk CNS-IPI score. Patients were about twice as likely to have a germinal center B-cell (GCB) COO rather than activated B-cell (ABC), but the COO was unavailable for almost 30% of the cohort. About 6% of the patients were known to have HGBLs, but the "hit status" of around 65% of the patients was unknown. Diagnostic LP was performed in 10%, 14%, and 19% of patients in the low-, intermediate-, and high-risk CNS-IPI groups respectively. The median follow-up time for the subjects in the study was 44 months. The low-risk group (12% of all patients analyzed), the intermediate-risk group (36%) and the high-risk group (52%) showed rates of CNS relapse of 1% (with a 95% CI: 0% to 2.4%), 2.4% (CI: 1.2% to 3.7%), and 2.4% (CI: 1.3% to 3.4%) respectively with no statistically significant difference across the risk groups (p=0.30). In patients with CNS relapse, only 35% of patients had a diagnostic LP. More than 90% of patients deemed high-risk for CNS relapse did not receive CNS prophylaxis. Among the 36 patients with CNS relapse, only 3 were given CNS prophylaxis at baseline (Table 2). Of the patients with CNS relapse, 75% of patients achieved a complete response with initial treatment. When categorized by the CNS-IPI score, there is no significant difference between intermediate- and high-risk based on type of CNS relapse, type of CNS prophylaxis used, response to first-line therapy, median time to relapse, median survival time, or median time from relapse to death. Those with CNS relapse had a shorter survival time compared to those with systemic relapse or no relapse (Figure 2). Conclusions When compared to the previously validated CNS-IPI study, there were fewer instances of CNS relapse in our patient population in the intermediate- and high-risk groups, however, about a quarter of the patients in these groups did not receive CHOP based therapy. Although there is published data demonstrating COO and HGBLs as contributing factors to CNS relapse, our data did not show any statistically significant difference in relapse rates. Potential limitations include that the study is a retrospective chart review of a predominately male veteran population. Our data suggests the CNS-IPI may not identify patients at risk for CNS relapse with adequate accuracy. Figure 1 Figure 1. Disclosures Nooruddin: AstraZeneca: Research Funding.

Introduction Diffuse large B cell lymphoma (DLBCL) is the most common type of non-Hodgkins lymphoma (NHL), constituting 25% of NHL cases. (CA Cancer J Clin PMID: 27618563) Patients with aggressive hematological malignancies (HM) including lymphomas and leukemias have substantial palliative care needs that entail symptoms management and Hospice as an essential component of end of life (EoL) care. Despite significant improvement in treatment options, patients with relapsed refractory (RR) HM have suboptimal EoL care, (Blood PMID: 28774879 and Hematology Am Soc Hematol Educ Program PMID: 33275720) compared to patients with solid malignancies. (J Clin Oncol PMID: 25154824) The purpose of this study was to assess for Hospice enrollment in patients with DLBCL within the Veterans Health Administration (VHA). Methods Trained abstractors performed a retrospective chart review of 3274 randomly selected patients with an ICD code for lymphoma treated within the VHA between 01/01/2011 and 12/31/2017. During the study period, 1699 veterans were confirmed to have a diagnosis of DLBCL, out of which 830 deceased patients were analyzed for this study. (Figure 1) Rates of palliative care consults, Hospice discussions, and Hospice enrollment were compared by age, race and outcomes of first line treatments using the Chi-Squared test of independence. Overall survival (OS) and time from Hospice to death were compared for the same variables with the Wilcoxon-Mann-Whitney test. A significance level of 0.05 was used for this study and all tests were two-tailed. Results The median age at diagnosis was 70 years. The predominant population was white males. At the time of diagnosis, approximately three fourths had an ECOG performance status of 0 to 2. Over 50% were diagnosed at stage III-IV and had an IPI score of 4 or 5. (Table 1) Within the study, 63%, 60% and 52% of patients were introduced to palliative care, had Hospice discussions, and enrolled in Hospice respectively. The median OS from date of diagnosis to death was 461 days, and the median time from Hospice enrollment to death for those with a known Hospice enrollment date was 12 days. (Table 2) There was no statistically significant difference in palliative care or Hospice usage, median OS or median time in Hospice to death, by age or race. Patients with complete response (CR) to first line treatment had longer median OS in comparison to the four other groups: patients who received no first line treatments, who began a first line treatment which was interrupted by death, who were RR after first line treatment and those with an unknown treatment response. Patients who did not receive any first line treatments, RR patients or those with an unknown treatment response were more likely to have a palliative care consult, discuss Hospice with a provider and enroll in Hospice (p-value Conclusion In this retrospective study, more than half of deceased patients with DLBCL within the VHA utilized Hospice services. There was no difference in Hospice use based on age or race, indicating equal access to care. Limitations of this study include a predominant male population and equal health care access that may limit generalizability of our finding to the non-Veterans population. Early recognition of high-risk patients may correlate with earlier enrollment to Hospice and increased lengths of stay in Hospice for patients with aggressive HM. Future educational interventions are needed to improve Hospice utilization for EoL care in patients with HM. Figure 1 Figure 1. Disclosures Nooruddin: AstraZeneca: Research Funding.

Introduction With an incidence of 5.6 per 100,000 per year, diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma (Teras, 2016) and has a 5-year relative survival of 63.8% (SEER Statistics, 2018). While there is not currently a universally accepted timeline for initiation of therapy in DLBCL, retrospective studies have demonstrated that delays in initiation of treatment can lead to adverse outcomes in DLBCL, such as decreased overall survival (Phipps, 2018). We performed a retrospective review of Veterans Affairs (VA) patients nationwide to assess how frequently delays occurred in treatment initiation and we analyzed the impact of time to treatment on response rates to first-line chemotherapy and on survival outcomes. Methods We performed a retrospective chart review of 2036 randomly selected records of patients seen within the VA nationwide who were diagnosed with lymphoma between 1/1/2011 and 12/31/2017. We included patients diagnosed with DLBCL. We excluded other types of lymphoma, patients whose workup and treatment were outside of the VA system, and patients with primary central nervous system (CNS) lymphoma. We determined the time from diagnosis to treatment (TDT), defined as the number of days from the date of pathology results to the date of initiation of treatment for each patient, and divided these into two-week blocks. The Wilcoxon-Mann-Whitney test was used to compare median overall survival between the groups. Results 971 patients were included in the study. Patients were predominantly male (96.2%), with a median age of 67 (Table 1). The median TDT was 20 days. Those with TDT of 0-14 days, 15-28 days, 29-42 days, and 43+ days had an objective response rate (ORR) of 70.9%, 84.5%, 82.9%, and 77.6%, respectively (Table 2). The same groups had a 2-year overall survival (OS) rate of 61.6%, 77.6%, 80.7%, and 72.7%, respectively. They demonstrated median OS of 36.1 months, 46.9 months, 46.4 months, and 47.1 months, respectively. The 2-year OS rates were significantly lower for the 0-14 day group compared to the 15-28 days group (P = 0.0001), 29-42 group (P Although the rates of high International Prognostic Index (IPI) scores and advanced stage disease were highest in the 0-14 day groups, we found that when we assessed only the patients with IPI 3-5, the 2-year OS rates, ORR, and median OS trends were similar, with inferior outcomes in the earlier treatment initiation groups (Table 3). When we analyzed only the patients with stage III-IV disease, as well as only those who were outpatient at time of diagnosis, we again found similar trends of inferior outcomes in those who began treatment within 0-14 days. The Cox proportional hazards model was employed to explore factors associated with survival time (Figure 1). An increase in age, a higher IPI score, and being an inpatient at the time of diagnosis were statistically significant factors that explained a lower OS in the 0-14 day TDT subgroup. When these were included in the model, TDT no longer made a significant difference in OS. Conclusion Our study demonstrated that shorter TDT did not improve survival outcomes for this aggressive lymphoma, contrary to what we expected. In fact, patients who had shorter TDT had statistically significant lower survival rates at 2 years, lower median OS, and lower ORR. We theorized that there was a propensity for patients who have more high-risk disease to be started on treatment more quickly because they are more often symptomatic or gravely ill at presentation. We found that when we controlled for inpatient status, high IPI, and advanced stage disease individually, the survival outcomes remained poorest in the earliest treatment initiation groups. However, when we controlled for all of these variables, there was no longer a statistically significant difference in OS based on TDT. Future studies are needed to better understand why our outcomes differed from prior studies which showed improved survival and response in shorter TDT subgroups, and to control for other possible confounders, such as treatment regimens and patients' comorbid conditions. Disclosures No relevant conflicts of interest to declare.

Introduction Diffuse large B-cell lymphoma (DLBCL) is the most common type of non-Hodgkin lymphoma (NHL), constituting 25% of NHL cases (Teras, 2016). Although survival rates have improved, with a 5-year relative survival of 63.8% in the United States (SEER Cancer Statistics, 2018), outcomes in DLBCL remain heterogeneous with inferior survival amongst some patient subgroups. Racial and ethnic disparities in access to care and outcomes are well-established and are critical issues across a number of malignancies, including NHL (Shenoy, 2011; Griffiths, 2010). The purpose of this study was to assess for racial and ethnic differences in patient and disease characteristics at diagnosis, and in outcomes for patients diagnosed with DLBCL within the Veterans Health Administration (VHA), where access to care may be less susceptible to other socioeconomic factors. Methods Trained abstractors performed a retrospective chart review of 2036 randomly selected patients seen in the VHA nationwide who were diagnosed with lymphoma between 01/01/2011 and 12/31/2017. We included patients diagnosed with DLBCL and excluded patients based on the criteria in Figure 1. We evaluated baseline patient and disease characteristics, including Eastern Cooperative Oncology Group (ECOG) performance status, stage at diagnosis, International Prognostic Index (IPI) score, pathology reports to identify high-grade lymphomas, and response to first-line treatment. Results A total of 971 patients met inclusion criteria for analysis. Patients were predominantly male, white, had a median age of 67, and presented primarily with advanced disease (Table 1). Patients in each subgroup presented with similar rates of stage III and IV disease, with no statistically significant difference in stage at presentation amongst each racial subgroup (white vs black, P=0.85; white vs Hispanic, P=0.30; white vs other, P=0.11). Most patients in each racial/ethnic group had a good performance status at diagnosis, with ECOG 0-2 in 75.4 - 82.5% of patients in each subgroup. The entire study population had an objective response rate (ORR) of 87.4% (complete response (CR) rate 66%) (Table 2). Response rates were similar across the 4 subgroups, with the majority of patients achieving a complete response (CR) after first-line therapy (66.7%, 68.9%, 65.3%, and 70% for black, Hispanic, white, and other/unknown patients, respectively). There were no statistically significant differences in ORR amongst subgroups (white vs black, P=0.28; white vs Hispanic, P=0.75; white vs other, P=0.75). Median overall survival (OS) from the time of diagnosis was 40.5 months for the entire study population (Table 2). OS rates were similar regardless of race with a median OS of 43 months for black patients, 49.2 months for Hispanic patients, 40.5 months for white patients, and 33.3 months for other/unknown patients (Figure 2). There was no statistically significant difference in median OS between subgroups (white vs black, P= 0.84; white vs Hispanic, P=0.39; white vs other, P=0.18). The 1-year survival rates were similar at 75.8%, 72.1%, 76.4%, and 71.7% for black, Hispanic, white, and other/unknown patient subgroups, respectively. Between 60 - 68.5% of patients in each subgroup remained alive at 2 years, with no significant differences in survival rates at 1 or 2 years. Conclusions In this retrospective study of patients diagnosed with DLBCL in the VHA nationwide, we found that there were no statistically significant differences in baseline patient characteristics at diagnosis or in response rates to first-line chemotherapy, 1- and 2-year OS rates, or median OS amongst each racial subgroup. Potential limitations of this study include that the population is predominantly male and therefore, may not be applicable to the female population, and that there was missing/incomplete data for pathologic assessment of high grade lymphoma in 68.5% of our population, which could provide important data about expected outcomes. Further studies with a longer follow-up period are needed to help characterize potential differences in outcomes and relapse rates. Our data suggest that when standard of care therapy is given equally to patients with DLBCL, similar outcomes occur for black, Hispanic, and white patients. The development of interventions to address healthcare disparities and to ensure access to appropriate and timely care for all patient populations is of paramount importance. Disclosures No relevant conflicts of interest to declare.

Introduction: Diffuse large B-cell lymphoma (DLBCL), is predominantly diagnosed in the elderly with a median age of diagnosis approaching the seventh decade of life. The prognosis for DLBCL has improved for all age groups, but less so for those over 75 years of age. The VHA (Veterans Health Administration) is one of the largest integrated provider of cancer care in the United States and its population, on average, is older than the general US population, giving us an unique advantage in exploring outcomes in elderly population. Methods: We reviewed 2036 medical records on patients that had a diagnosis code of some form of lymphoma between 2011 to 2017. Patients were randomly selected. Inclusion criteria was any veteran with atleast one entry in their records of selected ICD-9 and ICD-10 codes for lymphoma. Exclusion criteria included primary CNS lymphoma, other low grade variant lymphomas, diagnosis and treatment outside VHA, inability to determine if patient record met study criteria or inability to fill in critical fields of interest for the study. 971 patients were included. Data was manually abstracted by trained researchers. The Wilcoxon-Mann-Whitney test was used to compare median Overall Survival (OS) between two groups. All statistical comparisons were made using the R statistical analysis package. A level of significance of 0.05 was used for all the tests. Results: Of the 971 patients that met the inclusion criteria, 739 (76.1%) were 3 (50.1 % in Patients in the For those who received chemotherapy, 82.4% in Conclusions: Nearly 1/4th of the patient population were older than 75 years, and despite receiving comparable lines of chemotherapy, had lower survival when compared to patients who were Disclosures No relevant conflicts of interest to declare.

252 Background: Previously, we reported that adverse events (AEs) were the top reason for dose reductions in VHA patients on novel agents. Some AEs might be problematic but might not directly lead to dose reduction; however, these concomitant AEs should not be overlooked. This study describes AE clusters that occurred with the primary AE that led to dose reduction in VHA CLL patients receiving ibrutinib. Methods: This is a retrospective chart review study of CLL patients treated with ibrutinib in the VHA from October 2013 to March 2018. Variables included the presence of dose reduction, the reasons for dose reduction, primary AEs leading to dose reduction, and concomitant AEs present at the time of dose reduction. Descriptive statistics were used to summarize AE clusters. Results: Out of 1069 CLL patients on ibrutinib, 285 patients experienced dose reduction due to AEs and were included in this analysis. The most common AEs leading to dose reduction were: musculoskeletal (11%), bleeding (11%), fatigue (10%), infection (9%), atrial fibrillation (8%), diarrhea (8%), and rash (8%). Fatigue was the leading concomitant AE present at dose reduction among patients who were dose reduced due to musculoskeletal (18%), infection (12%), diarrhea (18%), and rash (25%). Fatigue was also the second-leading or third leading concomitant AE among those who were dose reduced due to atrial fibrillation (13%) or bleeding (10%). Musculoskeletal was the leading or second-leading concomitant AE present at dose reduction among those who were dose reduced due to fatigue (17%), infection (12%), or diarrhea (14%). Bleeding was the leading concomitant AE among those who were dose reduced due to atrial fibrillation (17%) and vice versa. Conclusions: This study provides evidence that fatigue and musculoskeletal AEs are problematic in CLL patients on ibrutinib in the VHA. These were not only among the most common primary AEs but also the leading concomitant AEs present at dose reduction. Although not directly leading to dose reduction, these ‘nuisance’ AEs can greatly affect quality of life among CLL patients and warrant more attention from clinicians. A better understanding of all AEs present at dose reduction may help clinicians better manage patients on novel agents. These data also highlight the unmet need for novel agents with a ‘cleaner’ safety profile.

e19339 Background: The first novel agent for use in CLL was approved in 2014; however, the extent of novel agent uptake in the VHA is largely unknown. Objective: This study described the pharmacoepidemiology of three novel agents (ibrutinib, idelalisib, venetoclax), and traditional chemotherapies/chemoimmunotherapies (CT/CIT) in the VHA. Methods: This was a retrospective study of 26,879 adults with CLL in the VHA from 10/01/2013 to 5/31/2018. All were followed for at least 6 months. Data were extracted from the VHA electronic health record. Patients came from all 18 Veterans Integrated Service Networks, spanning all 50 states and US territories. Descriptive statistics were used to summarize baseline characteristics, CLL treatments, next therapies, and secondary complications. Results: A total of 3670 patients received at least one of 12 CLL therapies of interest. Patients had a median age of 69 years (47% were 65+ and 26% were 75+), a median age-adjusted Charlson comorbidity score of 6, and 6% had a history of exposure to Agent Orange. Ibrutinib accounted for 89% of the novel agent use. Ibrutinib use across all lines of therapy (LOTs) increased sevenfold over the study period (Table). Venetoclax (42%) and idelalisib (30%) were the most common therapies for the next LOT after ibrutinib. Across all LOTs, traditional CT/CIT use declined steadily over the study period. However, in fiscal year (FY) 2018, there were still 17% of patients receiving CT/CIT. Ibrutinib was the most common therapy for the next LOT in these patients (43–74%). Incidence of diffuse large B cell lymphoma post-index was 2–6 times higher in patients on CT/CIT than those on ibrutinib. Other secondary complications were similar between ibrutinib and CT/CIT. Conclusions: To our knowledge, this is the largest study looking at CLL treatment patterns among VHA patients in the real world. There has been a major shift in the treatment of CLL, with fast adoption of novel agents in the VHA from 2013 to 2018. The impact of this shift on healthcare resource use and cost burden in the VHA will need to be examined. [Table: see text]

CNS involvement by CLL is a rare occurrence, usually happening in the context of a transformation to a more aggressive lymphoma in what is known as Richter’s transformation. We report a patient with active CLL who developed confusion and was found to have West Nile encephalitis that initially mimicked CNS involvement by CLL. The patient recovered with supportive treatment and later restarted ibrutinib therapy. This case illustrates the importance of maintaining a broad differential among cancer patients with new onset confusion as well as that of questioning malignant infiltration of CNS when there is concomitant active CNS infection.

In 2008, the World Health Organization (WHO) recognized the BCR-ABL1 negative myeloproliferative neoplasm (MPN) chronic neutrophilic leukemia (CNL) as a distinct entity based on the features of an increased white blood cell count (WBC) with a preponderance of neutrophils, no increase in blasts, no

Approximately 20%–25% of breast cancers overexpress human epidermal growth factor receptor 2 (HER2).1,2 In the last 15 years, multiple targeted agents directed at HER2 have emerged, leading to improved progression-free and overall survival.3,4 Recently, trastuzumab emtansine (T-DM1) was approved by the FDA for metastatic HER2-overexpressing (HER2+) breast cancers previously treated with trastuzumab and a taxane. This approval came after a phase III trial showed a 5-month improvement in median overall survival and an objective response rate of 43.6% with T-DM1 compared with 30.8% on lapatinib/capecitabine.5 Toxicity, including grade 3, was lower with T-DM1 and included fatigue, nausea, elevated liver enzymes, gastrointestinal symptoms, and cytopenias. Among patients with HER2+ breast cancer, up to 55% will develop brain metastases (BM),6 with studies suggesting up to a 4 times greater incidence relative to patients with non-HER2–expressing cancers.7 This increased incidence has been attributed to an “unmasking” effect of excellent systemic control prolonging survival, with poor drug CNS penetration. Preclinical data also suggest that the quantitative HER2 protein expression may be directly related to the development of BM, suggesting a possible causative mechanism for increased BM in the setting of HER2 overexpression.8 Many patients with metastatic HER2+ breast cancer undergo CNS radiation, including stereotactic radiosurgery (SRS), at some point in their treatment course. The interaction between SRS and these newer targeted agents is unknown and previously unreported. Since the FDA approval of T-DM1, we have observed an unexpected reaction among patients treated with T-DM1 who received SRS for BM within 15 months of undergoing T-DM1 therapy. These 4 patients, all with HER2+ metastatic breast cancer, presented shortly after a T-DM1 infusion with clinically significant increased edema as evidenced by neurologic changes and MRI findings of increased T2 signal in the area of previously treated BM (Fig. 1). Fig. 1. Image series from one patient who received a total of 2 infusions of T-DMI. The 4 patients, of median age 56.5 years (range, 37y–57y), had received SRS to one or more lesions at a median of 8.5 days (range: 3d–449d) prior to a T-DM1 infusion. Two patients experienced symptoms immediately after the first infusion; one patient received 5 infusions on trial (NCT01276041) and then became symptomatic after the second infusion off trial; one patient developed symptoms on her fifth infusion. Symptoms included headaches, nausea/vomiting, speech impairment, short-term memory deficits, imbalance, gait disturbance, and visual deficits. For all 4 patients, fusion of MRI with the SRS treatment plan revealed that the area of T1-contrast enhancement was entirely encompassed by the radiation prescription isodose line, suggesting no evidence of tumor progression. All 4 patients initiated steroids with eventual clinical and radiographical improvement over the weeks to months following steroid initiation. Three patients stopped T-DM1 due to neurologic symptoms; the patient who continued developed worsening symptoms and required steroid resumption. Due to severity of symptoms, this latter patient was taken to the operating room for resection of a single posterior fossa metastasis. Pathology revealed severe radionecrosis with no viable tumor cells identified. For comparison, we then evaluated all patients with metastatic breast cancer and brain metastases who were treated with SRS at our institution during the 2-year time period when these 4 patients had been treated. Overall, a total of 13 breast cancer patients received SRS. Of these, only the 4 patients previously reported developed clinically significant radiation necrosis. Three additional HER2+ breast cancer patients were treated with T-DM1 and SRS and did not develop radiation necrosis, for an overall rate of clinically significant radiation necrosis among this treatment group of 57% (Table 1). Table 1. Clinical and treatment characteristics for 7 patients with HER2+ breast cancer treated with SRS and T-DM1 over a 2-year period

Introduction : CLL is the most common chronic leukemia in the US, with nearly 20,000 new cases expected annually. Novel agents, such as ibrutinib, idelalisib, and venetoclax, have been approved in recent years and provide oral options for CLL. However, there is limited data regarding real-world treatment patterns with these novel agents. This study describes dose reduction and discontinuation rates, reasons for both, and outcomes, including overall survival (OS) and duration of therapy (DOT), in CLL patients treated with novel agents. Methods : This is an analysis of a large retrospective cohort study of adult patients (≥ 18 years old) with CLL, treated with novel agents in the VHA from 10/01/2013 to 3/31/2018. Historical data were examined for up to 20 years prior to the enrollment period (10/01/1993 to 9/30/2013). Index date was defined as the date of novel agent initiation. The follow-up period was a minimum of 6 months post index date. Variables, collected via a structured EMR database, included patient demographics, and clinical and treatment characteristics. CLL diagnosis, molecular profiles, and reasons for dose reduction and discontinuation were abstracted by chart review. Descriptive statistics were used to summarize baseline characteristics, treatment patterns, and outcomes. Results: A total of 1205 CLL patients were included in this analysis. Of these, in first observed line, 328 (27%) patients received ibrutinib; in relapsed/refractory observed line (r/r), 741(62%) patients received ibrutinib, 49 (4%) patients on idelalisib, and 87 (7%) patients on venetoclax. Ibrutinib patients in first observed line had a median (range) age of 73 (48-96) years and a median follow-up of 23 (3-54) months after treatment initiation. Dose reduction (n=83, 25%) and discontinuation (n=108, 33%) were frequently due to adverse events (AEs) (93% and 64%). Median DOT to ibrutinib discontinuation was 8 months. The most common AEs leading to dose reduction were major bleed (15%) and rash (15%). The most common AEs leading to discontinuation were atrial fibrillation (20%), major bleed (19%), and infection (11%). The calculated median OS from initiation was 31 (14-49) months. R/R ibrutinib patients had a median age of 72 (45-96) years and had 31 (2-85) months of follow-up after treatment initiation. Dose reduction (n=242, 33%) and discontinuation (n=263, 35%) were frequently due to AEs (89% and 63%). Median DOT to ibrutinib discontinuation was 12 months. The most common AEs leading to dose reduction were thrombocytopenia (13%), arthralgia/myalgia (13%), and infection (12%). The most common AEs leading to discontinuation were atrial fibrillation (19%), infection (15%), and major bleed (11%). the calculated median OS from initiation was 39 (9-57) months. R/R idelalisib patients (n=49) had a median age of 72 (55-93) years and had 27 (3-53) months of follow-up after treatment initiation. Dose reduction (n=8, 16%) and discontinuation (n=41, 84%) were frequently due to AEs (100% and 54%). Median DOT to idelalisib discontinuation was 5 months. The most common AE leading to dose reduction was neutropenia (50%). The most common AEs leading to discontinuation were infection (27%) and pneumonia (18%). R/R venetoclax patients (n=87) had a median age of 72 (47-90) years and had 9 (0-35) months of follow-up after treatment initiation. Dose reduction (n=24, 28%) and discontinuation (n=27, 31%) were frequently due to AEs (100% and 41%). Median DOT to venetoclax discontinuation was 5 months. The most common AEs leading to dose reduction were neutropenia (27%) and thrombocytopenia (27%). The most common AEs leading to discontinuation were neutropenia (36%), thrombocytopenia (18%), and infection (18%). There was not enough follow-up time to have a meaningful OS in this cohort. Conclusions: To our knowledge, this is the largest EMR/chart review study among CLL patients initiating treatments in the real-world setting. This study provides evidence regarding patient characteristics, treatment patterns, and outcomes among patients initiating novel agents for the treatment of CLL in the national VHA population. Dose reduction and discontinuation were frequent across all novel agents, with AEs as the most common reason. These data highlight the significant difference in real world data compared with clinical trial data and indicate the unmet need for more tolerable treatment options for CLL patients. Disclosures Frei: AstraZeneca: Research Funding. Le:AstraZeneca: Employment, Other: Stocks. McHugh:AstraZeneca: Employment. Elesinmogun:AstraZeneca: Employment, Equity Ownership. Obodozie-Ofoegbu:UT Austin: Employment.

BACKGROUND: Diffuse Large B cell Lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma (NHL). Rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) remains the standard of care for most advanced stage DLBCL. There is limited outcome data in DLBCL in Veteran population. Having unique health care access, our goal was to study basic characteristics of Veterans with DLBCL and to assess and compare treatment responses to similar patient population. METHODS: With IRB approval, we performed a retrospective analysis of DLBCL patients (n=55) who received R-CHOP at the VA Audie Murphy Hospital in San Antonio, Texas from 2007-2017. We compared the Veteran patients with DLBCL who received R-CHOP to Insured Community Setting ICS (n=58) and uninsured patients UCS (n=24) from 2007-2017 in the same zip code. Only patients who started and completed their care with respective institutions were included in this study. Rates of neutropenia, and use of hematopoietic support either pre-emptively or after infection were measured for VA patients. Other variables studied included LDH, stage, HIV status, age, time from definitive diagnosis to first R-CHOP treatment (TDT), treatment response (CR, PR, PD), relapse date, and morbidity/mortality. Similar variables were compared for patients at neighboring community stetting. Outcomes were compared based on health care access - Veterans (VA) vs Insured Community Setting (ICS) vs Uninsured Community Setting (UCS). Frequency among variables and predictive analysis were evaluated by the FREQ and NPAR1WAY procedures in SAS 9.4 using the Chi-squared Test of Independence and the Cochran-Mantel-Haenszel test. RESULTS The median age of diagnosis was 64.5 years (yrs.), 57 yrs. and 58 yrs. for the VA patients, ICS and UCS patients respectively. TDT for VA vs ICS vs UCS was 20 days vs. 22 days vs. 21.5 days (p=0.66). VA patients had a higher rate of stage III/IV disease at 78% compared to ICS 53% and UCS 43% (p=0.0033). Accounting for the higher rate of Stage III/IV cancers in the VA population, there was a higher failure rate of 40 % among VA patients vs 18 % for ICS vs 20 % for UCS to initial chemotherapy(p=0.034). CONCLUSION Our comparison data of VA patients with DLBCL receiving R-CHOP to a neighboring community setting show similar TDT. Response to initial treatment was inferior in Veteran population as compared to patient population with similar demographics and having similar TDT. Also, in comparing VA patients to uninsured patients, VA patients still have worse outcomes for treatment of DLBCL with R- CHOP. Older patient population, advanced stage and multiple co-morbidities are the possible contributing factors for difference in treatment outcomes for DLBCL in the VA community. Disclosures No relevant conflicts of interest to declare.

Background : Post Transplant Lymphoproliferative disorder (PTLD) represents a distinct and rare complication following solid organ transplantation (SOT). Insight into the biology of this disorder is limited to retrospective reviews and case series. In one of the first reports for post-transplant Diffuse Large B cell Lymphoma (PT-DLBCL) cases, we demonstrated a higher incidence and improved outcomes in PT-DLBCL non-germinal center B-cell (non-GCB) subtype compared to PT-DLBCL germinal center B-cell (GCB) Subtype. Published data suggests immunocompetent DLBCL non-GCB subtypes are less common and fare worse than immunocompetent GCB DLBCL. The reason for this unexpected finding in our PT-DLBCL pts is not fully understood. Recently Kiyasu and colleagues demonstrated that PD-L1 overexpression was significantly associated with non-GCB DLBCL, EBV virus positivity and poor prognosis in immunocompetent DLBCL samples. Therefore based on this we hypothesized that PT-DLBCL non-GCB subtype may have negative PD-L1 overexpression thus possibly accounting for improved outcomes compared to their immunocompetent counterparts. Hence we sought to test PD-L1 expression in our samples with PT-DLBCL. Methods: With IRB approval, we retrospectively identified PT- DLBCL patients treated at the University of Colorado between Jan 1989 to April 2015. We retrieved formalin fixed paraffin embedded PT-DLBCL tissue specimens and determined cell of origin by the Hans Algorithm. We assessed PD-L1 expression by immunohistochemistry. PD-L1 positive PT-DLBCL was defined as 30% of more of the lymphoma cells showing distinct membranous and or cytoplasmic staining. In addition EBER-ISH was performed to assess EBV status. Results: 86 adult SOT pts with PTLD were treated at our institution. 75 of 86 pts (87%) had monomorphic histology. Among monomorphic PTLD, 64% (48 of 75) had DLBCL. The median age at transplantation was 49.5 yrs (5-74 yrs). Median time from SOT to PTLD was 37 mos (1.4-499). The most common transplanted organ included kidney (40%), liver (38%), lung (13%) and heart (9%). 31% had early PTLD (12mos of SOT). 60% were EBV positive. 77% with early PTLD and 49% with late PTLD were EBV positive. Due to a paucity of archived tissue blocks, IHC staining was applied to 32/48 samples with DLBCL. Non-GCB subtype was identified in 75% (24/32) samples and GCB subtype in 25% (8/32) samples. Of the 48 pts with PT-DLBCL histology, PD-L1 stain was performed on 18 samples. Of the 18 PT-DLBCL samples, 77% (14/18) had non-GCB subtype and 16% (3/18) had GCB subtype. PD-L1 expression was negative in 78% (11/14) and positive in 22% (3/14) of non-GCB DLBCL samples. PD-L1 expression was negative in 100% (3/3) of GCB DLBCL samples. The sample size was too small to effectively describe the survival experience of pt subsets. Using Fisher's exact test we found no evidence to support an association between EBV Status and PDL1 expression (p-value 0.316). Conclusions: We previously reported in our consecutive series of PTLD after SOT an increased incidence and improved survival in pts with PT-DLBCL non-GCB subtype (ASH 2015) compared to PT-DLBCL GCB subtype. The reason for this is not fully understood. However, our limited series reveals that a majority of pts with PT- DLBCL non-GCB subtype was negative for PD-L1 overexpression. This might explain the improved outcomes in the PT-DLBCL non-GCB population. Despite a small sample size it is also interesting to note that 100% pts with PT-DLBCL GCB subtype were negative for PD-L1 overexpression. In the era of immunotherapy further studies in larger patient cohorts are warranted in order to understand the unique biology and outcomes of PT-DLBCL since it may have therapeutic implications. Disclosures Pollyea: Celgene: Other: advisory board, Research Funding; Ariad: Other: advisory board; Alexion: Other: advisory board; Pfizer: Other: advisory board, Research Funding; Glycomimetics: Other: DSMB member. Kamdar:Seattle Genetics: Speakers Bureau.

e19078Background: Primary central nervous system lymphomas (PCNSL) are a rare type of b cell lymphoma, confined to the CNS, the majority of which are Diffuse Large B-Cell Lymphomas (DLBCL). Despite...

Post Transplant Lymphoproliferative Disorder (PTLD) represents one of the most fatal complications following solid organ transplants (SOT). A multicenter analysis in 2010 reported a 3yr overall survival (OS) of 62%. In 2014 a single institutional study reported an improved median OS of 8 years in the rituximab era. Clinical studies have attempted to subdivide post transplant diffuse large B cell lymphomas (PT-DLBCL) into immunophenotypic subgroups with most reviews demonstrating a higher incidence of activated B-cell (ABC) subtype. Because of the rarity of PTLD, insight into disease development and progression is currently limited. We examined the clinical features and outcomes based on immunohistochemical (IHC) analysis in consecutive PTLD pts treated at the University of Colorado over a 25-year period. Methods: This retrospective analysis focused on 52 adult SOT pts who were treated for PTLD at our institution between Jan 1989 to April 2015. Results: 52 pts were analyzed. The median age at transplantation was 47.5 years (5-73yrs) and the median time from SOT to PTLD was 52.5 months (2.6 to 251). The most common transplanted organ included liver (46%), kidney (40%), heart (10%), lung (2%) and [kidney pancreas] (2 %). The grafted organ was involved in 10 pts (19%) Males comprised 67%, 63% were Stage III-IV, 69% were EBV positive, 23% had extranodal involvement, 19% had bone marrow involvement and 21% had CNS involvement. Of the 52 patients, 46 (88 %) had monomorphic histology. Amongst monomorphic PTLD 31 pts (60%) had DLBCL, 6 pts (14%) had Burkitt Lymphoma, 4 pts (9%) had CNS lymphoma, 1 pt (2%) had plasmablastic lymphoma, 1 pt (2%) had NK/T cell lymphoma and 1 pt (2%) had aggressive B cell lymphoma unclassifiable. Of the analyzed pts 85% were CD20 positive, 77 % had elevated LDH and 49% had an International Prognostic Index (IPI) > 1. Thirty Eight pts (73%) were on >1 immunosuppressive agent with a median immunosuppressive treatment duration of 36 months (range 0.25 to 23 yrs). All pts (100%) diagnosed within 12 months of SOT were EBV positive. Similarly all pts (100%) diagnosed 12 months after were EBV negative. Initial treatments included reduction in immunosuppression in 63% pts, rituximab monotherapy in 28% pts and chemotherapy in 53% pts. 67% of pts received rituximab either alone or in combination with chemotherapy. 48 of 52 pts were assessable for response; 4 pts were lost to follow up. 64% achieved a complete remission after first line therapy while 12% were refractory to front line therapy. 7 of 48 pts (14%) pts relapsed with a median time of 60 months (2yrs-15yrs). IHC staining utilizing the Hans Algorithm was applied on 20/31 pt samples with DLBCL. ABC subtype was identified in 70% (14/20) and GC (germinal center B-cell) subtype in 30% (6/20) samples. Median overall survival was 69 months (Figure 1). On univariate analysis CNS disease (HR 2.82, 95% CI 1.25- 6.3, p= 0.01), pts who received chemotherapy (HR 6.65, 95% CI 1.63-27.04, p= 0.008), graft involvement with PTLD (HR 2.63, 95% CI 1.01-6.86, p=0.04) and IPI > 1 (HR: 3.55; 95% CI: 1.47 - 8.57, p= 0.005) were associated with poor survival whereas EBV positivity (HR .25, 95% CI 0.10-0 .64, p= 0.005) was associated with improved survival. Advanced stage (Stage III-IV), high LDH, monomorphic histology, ABC DLBCL subtype, more than 1 immunosuppressive agents and rituximab treatment were not statistically significant parameters for survival. In multivariate analysis, EBV positivity was predictive of improved survival (p=0.006) while graft organ involvement (p=0.02) and chemotherapy administration (0.013) was associated with poor overall survival. Conclusion: In this consecutive series of PTLD after SOT from a single institution over a 25 year period, EBV positivity was associated with improved survival whereas graft organ involvement and chemotherapy was associated with poor outcomes. Similar to previously published data, in PT-DLBCL cases we were not able to demonstrate a definite association between histogenetic phenotype and prognosis. However we found 70% of PT-DLBCL samples to have an ABC subtype (Figure 2). The reason for this increased predilection is not fully understood. ABC subtype is known to have an inferior overall survival compared to GCB DLBCL which could explain the relatively poor outcomes in PT-DLBCL patients. This merits further studies in larger patient cohorts due to potential therapeutic implications. Figure 1. Figure 1. Figure 2. Figure 2. Disclosures Pollyea: Ariad: Consultancy; Agios: Consultancy, Membership on an entity's Board of Directors or advisory committees; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; GlycoMimetics: Other: Member of data safety monitoring board; Karyopharm: Consultancy; Pfizer: Consultancy.