BackgroundCancer in adolescents and young adults (AYAs) (15–39 years) is increasingly recognized as a distinct clinical and biological entity. Cancer of unknown primary (CUP), a disease traditionally presenting in older adults with a median age of 65 years, poses several challenges when diagnosed in AYA patients. This study describes clinicopathological features, outcomes and challenges in caring for AYA-CUP patients.MethodsA retrospective review of 47 AYAs diagnosed with CUP at MD Anderson Cancer Center (6/2006–6/2013) was performed. Patients with favorable CUP subsets treated as per site-specific recommendations were excluded. Demographics, imaging, pathology and treatment data was collected using a prospectively maintained CUP database. Kaplan-Meier product limit method and log-rank test were used to estimate and compare overall survival. The cox-proportional model was used for multivariate analyses.ResultsMedian age was 35 years (range 19–39). All patients underwent comprehensive workup. Adenocarcinoma was the predominant histology (70%). A median of 9 immunostains (range 2–29) were performed. The most common putative primary was biliary tract based on clinicopathological parameters as well as gene profiling. Patients presented with a median of 2 metastatic sites [lymph node (60%), lung (47%), liver (38%) and bone (34%)]. Most commonly used systemic chemotherapies included gemcitabine, fluorouracil, taxanes and platinum agents. Median overall survival for the entire cohort was 10.0 (95% confidence interval (CI): 6.7–15.4) months. On multivariate analyses, elevated lactate dehydrogenase (Hazard ratio (HR) 3.66; 95%CI 1.52–8.82; P = 0.004), ≥3 metastatic sites (HR 5.34; 95%CI 1.19–23.9; P = 0.029), and tissue of origin not tested (HR 3.4; 95%CI 1.44–8.06; P = 0.005) were associated with poor overall survival. Culine’s CUP prognostic model (lactate dehydrogenase, performance status, liver metastases) was validated in this cohort (median overall survival: good-risk 25.2 months vs. poor-risk 6.1 months).ConclusionsAYA-CUP is associated with a poor prognosis. In the current “-omics” era collaborative research efforts towards understanding tumor biology and therapeutic targets in AYA-CUP is an unmet need, necessary for improving outcomes in young CUP patients.

Background: Thrombocytosis can be a result of a reactive process such as acute blood loss, infections, iron deficiency anemia (IDA) or a clonal disorder such as Essential Thrombocythemia. The challenge of correctly identifying the etiology of thrombocytosis in an individual patient becomes particularly essential when the clinician is confronted with decisions regarding further workup, follow up as well as preventing future complications such as thrombotic events. Prior data linking iron deficiency anemia and thrombocytosis has been concluded from studies with small sample size, and to a large part, this association is based on anecdotal evidence. Our study examines the association between IDA and platelet counts and to the best of our knowledge is the largest study to do so. Methods: We performed a retrospective chart review of patients seen at our outpatient clinic from 1st January to 31st December 2017. We defined Iron deficiency as serum ferritin <30 mcg/dl, anemia in males as hemoglobin (Hb) <13.1 g/dl, anemia in females as Hb <11.7g/dl and thrombocytosis as platelet count >450,000/ul. Patient demographics, as well as hematological data, were collected and compared between three groups: Iron deficiency anemia, Iron deficiency without anemia, and control. We defined controls as subjects with neither iron deficiency nor anemia. Bivariate analysis using the Chi-square test for categorical variables, and a One-Way ANOVA for continuous variables were used to determine the association between the three groups of interest and the covariates. The relationship between platelet counts and other hematological parameters was evaluated using Pearson's correlation test. Results: Of the 4896 subjects included in our study, 1225 (25.02%) had iron deficiency anemia, 699 (14.28%) had iron deficiency without anemia, and 2972 (60.70%) were controls. The mean age of the cohort was 55.8 (SD = 18.31) years, and 73.9% of the subjects were females. In the IDA group, the mean hemoglobin was 10.21 (SD = 1.39) g/dl, the mean platelet count was 319.24 (SD = 106.92) k/uL, and 127 (10.37%) subjects had thrombocytosis whereas 1098 (89.63%) had normal platelet counts. When compared to the control group, 32 (1.08%) subjects had thrombocytosis, and 2940 ( 98.92%) had normal platelet counts. The difference in numbers of subjects with thrombocytosis between iron deficiency anemia, iron deficiency without anemia and control group was statistically significant as indicated by P<0.001 on the Chi-Square test. Furthermore, we found an inverse relationship between platelet count and ferritin as well as hemoglobin level with a p < 0.0001 across all groups. Conclusions: Our study demonstrated a significant association between iron deficiency anemia and thrombocytosis when compared to the control group however at a lower frequency than seen in the previously published studies. Further studies are needed to study causality and to elucidate the mechanism by which iron deficiency anemia leads to thrombocytosis. Table. Table. Disclosures No relevant conflicts of interest to declare.

A 54-year-old female presented with shortness of breath and cyanosis. Work up with chest X-ray and subsequent echocardiogram revealed an intracardiac bi-atrial mass leading to emergent cardiothoracic resection. Pathology was consistent with a primary cardiac high-grade osteosarcoma. Post-resection staging positron emission tomography-computed tomography (PET-CT) showed hypermetabolic mixed lytic and sclerotic lesion of T10 concerning for metastasis. She received five cycles of adriamycin and ifosfamide chemotherapy before discontinuation due to systolic dysfunction. Nine months later, she developed a high tumor burden with progressive disease and was treated with second-line gemcitabine/docetaxel with disappointing results. She is currently on treatment with cyclophosphamide and topotecan as third-line treatment with an excellent clinico-radiographic response. Osteosarcomas are aggressive with a high incidence of recurrence and metastasis. Fewer than 50 cases of primary cardiac osteosarcomas have been reported in the literature. Even though complete resection can be achieved in some cases, long-term results are usually poor. No standard therapy has been established.

e12506 Background: Breast cancer accounts for a million new cases annually worldwide with 40,000 deaths reported in 2020 in the United States. Treatment includes surgery, adjuvant chemotherapy, radiotherapy and endocrine therapy. Oncotype recurrence score (RS), a 21-gene expression assay, is a tool to predict the benefit of adjuvant chemotherapy in hormone receptor (HR) positive breast cancer. High RS suggests benefit with chemotherapy. The goal of this study is to evaluate the relationship of the Oncotype RS with traditional prognostic/predictive factors. Methods: Retrospective IRB approved study of all patients from January 1, 2017, to December 31, 2019 with invasive breast cancer and an Oncotype RS in a community hospital setting. Study parameters included patient age, grade, tumor size, histologic subtype, HR status, number positive (0-4) lymph nodes (LN), extra nodal extension (ENE), lymphovascular invasion (LVI), and RS. Statistical analysis was done using linear regression and one-way ANOVA followed up with Tukey’s procedure. Results: A total of 470 patient charts evaluated and 22 patients excluded due to incomplete data with the following breakdown. LN positive = 14% (61/450). Oncotype score low (RS < 11) = 25% (114/448), intermediate (RS 11-25) = 59% (265/448) and high (RS > 25) = 15% (69/448). Grade 1 = 34% (153/448), Grade 2 = 50% (225/448) and Grade 3 = 15% (70/448). There was statistical significance in the mean difference in RS between the levels of grade and the levels of HR. The mean differences in RS between Grades 3 versus 1 was 10.9 (p < 0.001); between Grades 3 versus 2 was 8.3 (p < 0.001); and between Grades 2 versus 1 was 2.6 (p = 0.026). The mean Oncotype RS for grades 1,2,3 was 13.87 vs 16.33 vs 28.24 respectively. HR status was categorized as 51-100% (strong HR+), 11-50% (intermediate), or 1-10% (weak). Comparing strong HR+ versus intermediate/weak, the mean differences in RS were -41.4 and -35.7, respectively (p < 0.001). There were no significant differences in RS evaluated against the following parameters, patient age, tumor size, number of positive lymph nodes (0-4), LVI, or ENE. Conclusions: Tumor grade and HR status had strong association with the Oncotype RS as expected. Grade 3 tumor associated with high RS (mean 28). Intermediate HR (11-50%) associated with much higher RS = 52 similar to weak HR (1-10%) = 58. We report that RS had no correlation to patient age, tumor size or even LN status. The LN and RS association may have been impacted by low (14%) LN positive status (reportable in the modern era). Our study adds to the data that biology trumps size (traditional risk factors) while posing the question does intermediate HR status (11-50%) deserve recognition as a more substantial risk factor than previously considered.

5580 Background: NCCN guidelines recommend consideration of neoadjuvant chemotherapy (NACT) for poor surgical candidates with bulky stage III or IV ovarian, fallopian tube, or primary peritoneal cancer. However, new convincing evidence favoring NACT instead of primary tumor reductive surgery (PTRS) has resulted in shifting practices among gynecologic oncologists. This study compares operative and post-operative outcomes between patients receiving NACT and those undergoing PTRS. Methods: After IRB approval, patients who received NACT or PTRS were identified through the tumor registry and surgical database at a single institution from 2008-2012. Statistical analyses included Wilcoxon Mann-Whitney, Chi square and Fisher’s exact test. Results: Of 163 patients, 109 (67%) received NACT and 54 (33%) underwent PTRS. The majority in both groups was Caucasian (82%) and had ovarian cancer (85%). The median age of all patients was 62 years. There was no difference in median age between groups. High-grade serous histology was most common. In the PTRS group, 72% were stage IIIC. During cytoreductive surgery, NACT cases had significantly shorter total operative and anesthesia time compared to PTRS cases (p = 0.005). NACT patients had significantly less blood loss (p=0.002) than PTRS patients. Based on available data, there was no difference in intraoperative transfusion rate (p = 0.098). However, PTRS cases had a higher rate of postoperative transfusion compared to NACT cases (p = 0.043). There was no difference in the proportion of optimal surgical cytoreduction between groups (p = 0.422). Available data demonstrated that PTRS patients had significantly more complications and intensive care unit (ICU) admissions compared to those receiving NACT (p = 0.024, p = 0.002 respectively). Conclusions: NACT offers advantages regarding operative and anesthesia time, blood loss, postoperative complications, and ICU admissions. NACT may offer a valuable alternative to PTRS. Further studies are warranted to better define the role of NACT in the upfront treatment of advanced ovarian cancer.

No abstract

e12554 Background: Breast cancer is a frequently diagnosed malignancy in the United States accounting for approximately 266,000 cases and 40,000 deaths annually. Human epidermal growth factor receptor 2 (HER2) is a surface protein shown to affect cell proliferation, migration and survival - the hallmarks of neoplastic cells. HER2 is overexpressed in 20-35% of breast cancer cases and has a strong prognostic and therapeutic significance. This discovery led to development of therapeutic agents targeting HER 2 receptor. In 2013, College of American Pathologists (CAP) updated their recommendations for HER2 testing by incorporating an algorithm with immunohistochemistry (IHC) and In situ hybridization (ISH). The primary objective of our study was to assess for variance in HER2 interpretation by applying the updated 2013 criteria to all diagnosed cases in our hospital for the immediate previous three years. Methods: We retrospectively reviewed charts of patients diagnosed with invasive breast carcinoma from January 1, 2010 to December 31, 2012 at our hospital. The new American college of pathology criteria assessing FISH HER2 copy number and HER2/CEP ratio for each patient was compared to the old criteria to assess for variance in HER2 interpretation. Categorical variables are presented as percentages and kappa statistic was used to assess for variance. Results: We reviewed 526 charts with an invasive breast cancer diagnosis over a 3-year period. 10.83% of patients were HER2 amplified, 86.12% were non-amplified and 3.04% were equivocal. Applying the updated 2013 CAP criteria, we found that 11.21% were amplified, 74.33% were non amplified, and 14.44% remained equivocal. Kappa statistic indicated substantial agreement (0.804) between the old and new criteria. A review of our cohort revealed an overall lower level of HER2 positive cases at 10.83% (old criteria) and 11.21% (updated criteria) when compared to historical studies. The increase in equivocal cases from 3.04% to 14.44% with the new criteria suggests under diagnosis of HER2 positive cases. Due to the lack of IHC performed (as per old standards), further classification of equivocal cases (based on new criteria) was not possible. While we saw some movement in our data from non-amplified to equivocal, our analysis indicated substantial statistical agreement between the old and the new diagnostic criteria Conclusions: Our study did not show a variance in HER2 interpretation between the 2007 and 2013 CAP criteria which is reassuring. An increase of 11.4% cases in the equivocal category was noted and may reveal a significant difference if further delineation is feasible. This variance could be important given the prognostic and therapeutic implications in HER2 positive breast cancer.[Table: see text]

No abstract