Anaplastic large cell lymphoma (ALCL), ALK negative (ALK−) is an aggressive lymphoproliferative disorder of mature T lymphocytes characterised by hallmark cells, CD30 positivity and lacking ALK protein expression. ALCL, ALK− has to be differentiated from peripheral T-cell lymphoma-not otherwise specified and classical Hodgkin’s lymphoma. ALK− anaplastic large cell leukaemia should be considered in a patient with a history of ALCL, ALK− presenting with new leukaemia. We report a rare presentation of relapsed ALCL, ALK− with leukaemia after autologous stem cell transplantation in a 57-year-old male. Leukaemia, either as primary presentation or secondary transformation confers worse prognosis in ALCL, ALK− with very few cases reported so far. Emergency resuscitation with leukapheresis and treatment of tumour lysis syndrome along with supportive care should be followed by combination chemotherapy. Brentuximab vedotin and stem cell transplantation are the backbone of treatment for relapsed/refractory disease.

Background: Essential thrombocythemia (ET) is a chronic myeloproliferative neoplasm which is associated with an increased risk of thrombohemorrhagic complications as well as progression to myelofibrosis and frank leukemia. Patients with ET are at an elevated risk for stroke. However, studies of prevalence and outcomes of stroke in hospitalized patients with ET have been limited to case series. The median survival of patients with ET is comparable to normal population but the quality of life may be significantly altered due to the occurrence of thrombotic events in the cerebrovascular and cardiovascular systems. By conducting a retrospective analysis of nationwide data from hospitalized ET patients between the years of 2006 and 2014, we sought to identify if there are any statistically significant associations between stroke and/or in-hospital mortality with respect to patients' gender, age group, race, and comorbidities like hypertension, diabetes atrial fibrillation and chronic kidney disease. Methods: Data from the National Inpatient Sample was utilized to identify outcomes in hospitalized patient with ET who were admitted for stroke. The National Inpatient sample is a database maintained by the Agency for Healthcare Quality and Research. Utilizing the current procedural terminology code (CPT) for ET, outcomes of patients with ET who were hospitalized with stroke were studied for the year 2006 to 2014. Patient demographics of age, gender and race were collected and hospital characteristics of location and size were correlated to outcomes. The extent of common medical comorbidities such as hypertension, diabetes, chronic kidney disease and atrial fibrillation was studied in ET patients who died with and without stroke. Chi square test was used to determine odds ratios and multiple logistic regression was used to determine independent predictors of mortality. Results: Between 2006 to 2014, 552422 hospitalizations involved patients with a diagnosis of ET of which 20650 hospitalizations were due to stroke. Of these patients with stroke there was a preponderance of prevalence in females (13400 vs. 7251). The percentage of stroke in these hospitalizations varied between 3.64 to 4.29 over 15 years and mortality in these patients did not significantly change during this time period. The prevalence of stroke was highest amongst Asians and Caucasians (4.7% and 3.86%) with a statistically significant difference (p=0.0000). The age group of 80+ years and the difference in prevalence between different age groups (18-34 vs. 35-49 vs, 50-64 vs. 75-79) was statistically significant (p=0.0000) with Medicare being the insurance for most of these patients (p=0.000)). Notably, mortality was highest in the same group but was not significantly different from other age groups. Large sized hospitals were noted to have a higher proportion of ET patients with stroke compared to smaller and medium sized hospitals (p=0.0002). No difference in such proportions was noted in hospitals varying by region (Northeast vs. Midwest vs. South vs. West). Burden of medical comorbidities as measured by Charlson's comorbidity index was noted to be in the 4-6 range. Similarly, hypertension, hyperlipidemia, diabetes, atrial fibrillation, smoking status were also found to be more frequent in ET patients with stroke. A majority of ET patients with stroke were discharged to skilled nursing facilities. Multiple regression showed that female gender, atrial fibrillation, stroke, higher Charlsons comorbidity score and 80+ age were independent predictors of mortality (OR: 0.75, 1.35, 1.8, 2 to 5.7, 13.9 respectively). Conclusions: Patients with ET who are hospitalized with stroke have significantly worse outcomes. This study demonstrated that a statistically significant difference exists among different age groups of patients with ET and stoke who died during hospitalization when stratification is made using age groups and Charlson Score. This study may serve as an initial point to include new risk factors for further risk stratification. Early identification of patients at higher risk may reduce the incidence and decrease the morbidity of stroke in patients with ET. Disclosures Kota: BMS: Honoraria; Novartis: Honoraria; Xcenda: Honoraria; Incyte: Honoraria; Pfizer: Honoraria.

e21504 Background: Patients with NSCLC with liver metastasis tend to have significantly reduced overall survival (OS) compared to NSCLC with metastasis to other sites. We aimed looking at the patients with lung cancer at our institution over a five year period and assess the outcome of NSCLC with liver metastasis based on different treatment modalities. Methods: This is a retrospective study of patients with NSCLC and liver metastasis over the past five years. Patient data was securely collected and stratified based on demographics, histology, driver mutation analysis, date of diagnosis, place of metastasis, date of treatment, types of treatment received, and date of death, as applicable. Results: Eight of 132 patients had liver metastasis. Out of which one had SCLC and was treated with carboplatin-etoposide-atezolizumab with OS of 17 months. Remaining seven were NSCLC where five patients had adenocarcinoma and 2 had squamous histology. Pembrolizumab was part of treatment for all 4 patients with adenocarcinoma. 5th patient was treated with chemotherapy and had OS of 2 months. Female to male ratio 3:2 for adenocarcinoma and 0:2 for squamous histology with equal ethnic distribution. Two patients had squamous histology, both males and had average OS of 4.5 months. Conclusions: Patients with liver metastasis regardless of the histology, have short overall survival. However it was observed that such patients when treated with immunotherapy as part of their treatment modality do much better compared to those who did not. Specifically immunotherapy benefits are more pronounced in patients with adenocarcinoma histology compared to squamous. However further studies are indicated to evaluate personalized approach to therapy in NSCLC with liver metastasis. Our study was limited based on our limited number of patients.

e17533 Background: Squamous cell carcinoma of the nasopharynx, oropharynx and hypopharynx constitutes a majority of head neck malignancies. The incidence-based mortality across different races has been noted to be divergent. This study analyzes the trend in incidence-based mortality from the years 2000 to 2014 amongst both the genders in Caucasian/White and African American/Black patients. Methods: The Surveillance, Epidemiology, and End Results (SEER) Database was queried to conduct a nation-wide analysis for the years 2000 to 2014. Incidence-based mortality for all stages of nasopharyngeal, oropharyngeal and hypopharyngeal cancer was queried and the results were grouped by race (Caucasian/White, African American/Black, American Indian/Alaskan native and Asian/Pacific Islander) and gender. All stages and ages were included in the analysis. T-test was used to determine statistically significance difference between various subgroups. Linearized trend lines were used to visualize the mortality trends of all sub groups. Results: Incidence-based mortality rates (per 1000) for nasopharyngeal, oropharyngeal and hypopharyngeal cancer all races and both the genders is shown in the table below. The male to female gender disparity in mortality is~ 1:3 in patients with nasopharynx across all races and becomes worse to ~1:4 and ~1:5 for patients with oropharyngeal and hypopharyngeal cancers respectively. Notably the highest incidence based mortality for nasopharyngeal cancers is seen in Asian/pacific Islander males and a similar peak is noted for hypopharyngeal cancers in African American/Black males. Conclusions: A significant gender disparity exists in all there pharyngeal cancers across all races. It is unclear if female gender is protective but further study is warranted in a stage- specific and age-specific manner to better understand this disparity.[Table: see text]