Search

543 Background: Early validation studies using the Oncotype DX recurrence score (RS) in NSABP B20 demonstrated that women with node negative breast cancer and RS >31 had significant survival benefit from the addition of adjuvant chemotherapy to endocrine therapy (CET). Consequently, in the prospective TAILORx trial, node negative women with RS >26 received CET. These studies did not clearly delineate the magnitude of benefit of adjuvant chemotherapy for post-menopausal node negative women. A recently published well-designed adjuvant trial (RxPONDER) demonstrated that adjuvant chemotherapy was not beneficial in post-menopausal pts with ER+/HER2- breast cancer, 1-3 positive nodes, and RS 26 compared to endocrine therapy (ET) alone, given that CET is more beneficial in women 50 with ER+/HER2- pT1-2N0M0 breast cancer with RS >26, to assess real world utilization. We separated women into two groups based on adjuvant treatment: ET alone or CET. Chi-square and logistic regression analysis determined difference between different systemic treatment groups. OS was analyzed using a multivariable Cox model. Results: A total of 16,745 eligible women who underwent surgery and received ET were identified in the NCDB—4,740 (28.3%) received ET alone and 12,005 (71.7%) received CET. We observed that CET use increased over time. Women were more likely to receive CET if their tumors were moderately differentiated (OR = 1.853, p < 0.001), poorly/undifferentiated tumors (OR = 3.875, p < 0.001), or associated with lymph-vascular invasion (OR = 1.206, p = 0.001). After accounting for demographic and oncologic factors, 5-year OS rates in this cohort were significantly superior in women receiving CET compared to ET alone (95.4% vs 92.0%, Hazard Ratio = 0.680, p < 0.001). Conclusions: Utilizing the NCDB to represent real world outcomes, we observed that women > 50 years with pT1-2N0M0 ER+/HER2- breast cancer, and RS > 26 had a significantly superior 5-year OS when receiving adjuvant chemotherapy provides a measurable OS benefit for post-menopausal women in this setting and should be discussed with patients.

1100 Background: EMERALD demonstrated significantly prolonged progression-free survival (PFS) and a manageable toxicity profile for elacestrant vs standard of care endocrine therapy (SOC) in patients with ER+/HER2- mBC following progression on prior endocrine and CDK4/6 inhibitor therapy. Benefit was observed in the overall study population and in patients with ESR1 mutations (mESR1). Here, we report a subgroup analysis from EMERALD in patients with no prior chemotherapy. Methods: EMERALD (NCT03778931) is a randomized, open-label, phase 3 trial that enrolled patients with ER+/HER2− mBC who had 1–2 lines of endocrine therapy, mandatory pretreatment with a CDK4/6 inhibitor, and ≤1 chemotherapy. Patients were randomized 1:1 to elacestrant (400 mg orally daily) or SOC (investigator’s choice of fulvestrant or aromatase inhibitor). Primary endpoints were PFS in all patients and patients with mESR1. In this analysis, we compared PFS between elacestrant and SOC in patients without prior chemotherapy. Results: Among the 477 patients enrolled in the trial, 77.8% (n = 371) had not received prior chemotherapy for mBC (median age = 64). Among patients without prior chemotherapy, treatment with elacestrant was associated with significantly prolonged PFS compared to SOC in both the overall population (hazard ratio [HR] = 0.68 [95% CI, 0.52-0.89] P = 0.004; median PFS 3.7 vs 2.0; 6-mo PFS 38% vs 23%; 12-mo PFS 27% vs 12%), and patients with mESR1 (HR = 0.54 [95% CI, 0.36-0.80] P = 0.002; median PFS 5.3 vs 1.9; 6-mo PFS 44% vs 24%; 12-mo PFS 31% vs 12%). Key treatment-related adverse events (AEs) in the no prior chemotherapy elacestrant group were nausea (25.9%), fatigue (12.7%), and hot flush (11.1%). There were no treatment-related deaths in either group. Conclusions: Among patients with ER+/HER2− mBC without prior chemotherapy, elacestrant significantly prolonged PFS compared to SOC endocrine therapy and showed favorable outcomes in this subgroup. Clinical trial information: NCT03778931.

540 Background: Based on the results of the RxPonder trial, post-menopausal women over age 50 with estrogen receptor (ER)+ breast cancer, 1-3+ nodes, and a 21-gene Oncotype DX recurrence score (RS) score of 20 being associated with a significantly inferior overall survival (OS) (P value range: < 0.001-0.019). In women with RS of 20-25, CET was associated with a significant improvement in OS compared to ET alone, regardless of age (age < = 50: HR = 0.334, P = 0.002; age > 50: HR = 0.521, P = 0.019). Conclusions: Among women with ER+/HER2- breast cancer with 1–3 positive nodes, and RS of 20-25, in contrast to the RxPonder trial we observed that CET was associated with an OS benefit in women regardless of age.

TPS10618 Background: Coronavirus disease 2019 (COVID-19), caused by betacoronavirus SARS-CoV-2, is associated with an increased risk of severe infection or death in cancer patients compared to the general population. The CANVAX trial recently demonstrated that short term immune responses to SARS-CoV-2 vaccines are modestly impaired in patients with cancer— particularly those who receive myelosuppressive chemotherapy. Because little is known regarding longitudinal antibody or T-cell responses in cancer patients who receive cytotoxic chemotherapy or non-myelosuppressive targeted systemic therapy, the aim of this longitudinal study is to assess immune B and T cell responses to SARS-CoV-2 over a 12-month period in solid tumor patients who receive chemotherapy or non-immunosuppressive therapy compared to healthy individuals without cancer. Methods: This is an ongoing prospective non-interventional clinical trial (NCT05238467). Approximately 100 patients will be enrolled into three different arms. Accrual began in May 2021 and 37 patients have been enrolled. Eligible patients must not have prior COVID-19 infection < 6 months from study enrollment and have a diagnosis of a solid tumor (breast, genitourinary, or gastrointestinal cancers), who either: received myelosuppressive chemotherapy within 60 days prior to initial or booster COVID vaccination, or who started on chemotherapy within 30 to 60 days after the initial or booster COVID vaccination (Arm A); or received non-immunosuppressive treatments (Arm B); or have no history of cancer or prior history of cancer but beyond 12 months from completion of curative cancer treatment (Arm C, control cohort). Whole blood will be collected in accordance with standard operating procedures. Blood samples analyzed for the presence of antibodies against the major antigenic components of SARS-CoV-2 including the spike glycoprotein (S), receptor binding domain (R) and nucleocapsid phosphoprotein (N). Antibody levels will be quantified utilizing quantitative ELISA. T-cell responses will also be quantified. The primary endpoint is seroprotection rate with an antibody titer protective (1:40) at any point: baseline, 2, 6, and 12 months. The secondary endpoint is to evaluate differences in longitudinal immunological responses to SARS-CoV-2 over a 12-month period. The difference of the seroprotection rate among 3 cohorts of participants will be examined using chi-square test. Moreover, the effect of treatment (chemotherapy, endocrine, TKIs) on seroprotection will be estimated using multivariable logistic regression controlling the effects of confounders, such as age, gender and cancer type. COVID antibody titers measured over time (baseline, 8 weeks, 6, 9, 12 months after the second vaccination) will be analyzed using mixed-effect models. Clinical trial information: NCT05238467.

584 Background: Approximately 30% of all breast cancers are characterized as triple-negative (TNBC). TNBC typically occurs in younger women and is associated with a poorer prognosis relative to other breast cancer subtypes. High levels of tumor-infiltrating lymphocytes (TILs) in residual disease after Neoadjuvant chemotherapy (NACT) have previously been shown to be associated with better prognosis in TNBC. In this work, we sought to evaluate the prognostic value of computationally derived measures of TIL spatial architecture in residual TNBC after NACT. Methods: H&E-stained samples from 92 patients (pts) with TNBC (41 died, 45 had disease recurrence) and residual disease after NACT were retrospectively collected from 2 sites: Instituto Nacional de Enfermedades Neoplásicas (S1) and University Hospitals (S2). 45 pts (16 deaths, 23 recurrences) from S1 formed the training set and 47 pts (25 deaths, 22 recurrences) from S2 formed the independent validation cohort. Samples were digitized at 20x. Computerized algorithms automatically identified 2 types of nuclei (TILs and non-TILs) and built clusters for each nuclei type based on cell proximity. The spatial arrangement of these clusters was then quantified using network graph metrics. The top 5 features, determined by least absolute shrinkage and selection operator, were used to train a Cox regression model that assigned a risk of death and recurrence to each patient on the training set. The percentile 33 risk score was used as a threshold for stratifying pts on the validation set as either low or high risk. For comparison, we also employed a model based on TIL density alone. Survival analysis was used to evaluate the performance of both approaches on disease-free survival (DFS) and overall survival (OS). Results: Pts in S2 (n=47) identified as “high risk” by the model based on spatial architecture of residual TILs had a significantly shorter survival time. The median OS for pts at high risk was 25 months vs. 55 months for low-risk pts. The median DFS for pts at high risk was 32 months vs 51 months for low-risk pts. Univariable analysis showed this model was prognostic for both OS (Hazard Ratio (HR) = 2.57, 95% Confidence Interval (CI): 1.07-6.16, p=0.03) and DFS (HR=2.38, CI: 1.01-5.62, p=0.04). In contrast, the model based on TIL density was not prognostic for OS (HR=1.24, CI: 0.33-4.63, p=0.73) nor DFS (HR=1.19, CI: 0.32-4.34, p=0.78). Conclusions: A computerized image analysis model based on measurements of spatial arrangement of residual TILs and surrounding cells was found to be prognostic in TNBC pts who received NACT. This method appears to be more prognostic than TIL density alone. Additional multisite validation and multivariable analysis is needed to further establish the independent prognostic utility of TIL based image biomarkers in the post-NACT TNBC.

Purpose Patients who are diagnosed with stage IV metastatic melanoma have an estimated 5-year relative survival rate of only 17%. Randomized controlled trials of recent US Food and Drug Administration–approved immune checkpoint inhibitors—pembrolizumab (PEM), nivolumab (NIVO), and ipilumumab (IPI)—demonstrate improved patient outcomes, but the optimal treatment sequence in patients with BRAF wild-type metastatic melanoma remains unclear. To inform policy makers about the value of these treatments, we developed a Markov model to compare the cost-effectiveness of different strategies for sequencing novel agents for the treatment of advanced melanoma. Materials and Methods We developed Markov models by using a US-payer perspective and lifetime horizon to estimate costs (2016 US$) and quality-adjusted life years (QALYs) for treatment sequences with first-line NIVO, IPI, NIVO + IPI, PEM every 2 weeks, and PEM every 3 weeks. Health states were defined for initial treatment, first and second progression, and death. Rates for drug discontinuation, frequency of adverse events, disease progression, and death obtained from randomized phase III trials were used to determine the likelihood of transition between states. Deterministic and probabilistic sensitivity analyses were conducted to evaluate model uncertainty. Results PEM every 3 weeks followed by second-line IPI was both more effective and less costly than dacarbazine followed by IPI then NIVO, or IPI followed by NIVO. Compared with the first-line dacarbazine treatment strategy, NIVO followed by IPI produced an incremental cost effectiveness ratio of $90,871/QALY, and first-line NIVO + IPI followed by carboplatin plus paclitaxel chemotherapy produced an incremental cost effectiveness ratio of $198,867/QALY. Conclusion For patients with treatment-naive BRAF wild-type advanced melanoma, first-line PEM every 3 weeks followed by second-line IPI or first-line NIVO followed by second-line IPI are the most cost-effective, immune-based treatment strategies for metastatic melanoma.

7057 Background: Cancer patients (pts) are at high risk of unplanned hospital readmissions. Predicting which cancer patients are at higher risk of readmission would improve post-discharge follow-up/navigation, decrease cost, and improve pt outcomes. Methods: We conducted a retrospective cohort study of non-surgical cancer pts hospitalized at our center between 12/2014 to 7/2018. A machine learning algorithm was trained on 348 medical, sociodemographic and cancer-specific variables with a total of 1,801,944 data points. The cohort was randomly divided into training (80%) and validation (20%) subsets. Prediction performance was measured by area under the receiver operator characteristic curve (AUC). Results: A total of 5,178 hospitalizations were included, of which 45.1% were women, and 27.6% experienced an unplanned readmission within 30 days. The most frequently represented cancers were hematologic malignancies (30.5%), followed by GI (18.1%), lung (13.7%), and GU (10.9%). Significant variables that impacted the algorithm decision are ranked from the most to the least important, including: days from last admission; planned index chemotherapy admission; number of vascular access lines, drains, and airways in use; length of stay; cancer diagnosis; total ED visits in past 6 months; age; discharge lab values (sodium, albumin, alkaline phosphatase, bilirubin, platelets); number of prior admissions; and discharge disposition. The AUC for the validation subset was 0.80. To ease the translation of this model into the clinic, we developed a web application whereby users can supply the aforementioned variables to the model and receive a personalized prediction that highlights those variables most affecting a subject’s readmission risk status: www.Cancer-Readmission.com. Conclusions: A cancer-specific readmission risk model with high AUC for 30-days unplanned readmission has been developed. The model is embedded in a freely available web application that provides personalized, patient-specific predictions. Programs that integrate this model can identify cancer patients with a greater risk for unplanned hospital readmission, thus providing a personalized approach to prevent future unplanned readmissions.

64 Background: In the era of increasing therapeutic options and complexity of disease, some cancer patients (pts) continue aggressive treatment even within days of death. Previous studies report 30-66% of pts do not receive hospice or palliative services in the last month of life and many are enrolled in hospice < 3 days before death. Quality measures are endorsed by ASCO, National Quality Forum, and Oncology Care Model to increase hospice utilization. It is shown that pts enrolled in hospice have increased survival time with reduced in-hospital resources. These benefits increase the longer pts are in hospice. We hypothesize early identification of high risk pts by a multidisciplinary group and formulation of a care plan will prompt early discussion for hospice eligibility, increasing referrals to hospice and the number of days spent in hospice. Methods: As reported (ASCO 2018, Abst 6547), an Interdisciplinary Care Team (ICT) was created with palliative medicine and oncology physicians, nurses, and social workers. Twice monthly pts with high utilization over a 60-day period were identified. Care plans (CP) were created using a team based approach with parallel input from outpatient teams. CP was communicated back to the primary team. Results: 112 pts were discussed over 24 months; 39 pts died with a solid tumor malignancy and this was our study cohort. 85% pts (33/39) were referred to or had a hospice discussion and 82% pts (27/33) enrolled. 6 pts declined and 6 acutely died. Of the 27 pts that entered hospice 78% (21/27) were enrolled > 3 days and 22 % (6/27) < 3 days. Average number days in hospice was 19.7 (median 11) for all who entered hospice. In the subgroup that were enrolled for > 3 days, average number days was 25 (median 21). 62% pts (17/27) entered hospice within 60 days of ICT meeting and CP. Conclusions: Early identification of high-utilizing cancer patients along with review by ICT may correlate with early recognition of hospice eligibility, enrollment, and therefore greater number days spent in hospice. This increases hospice utilization allowing patients and families to experience the full benefit of hospice-directed care. Further interventions should be explored in optimizing transitions of care.

e18184 Background: Venous thromboembolism (VTE) complicates several anti-cancer regimens including chemotherapy and anti-angiogenic agents. Cyclin-dependent kinase 4/6 inhibitors (CDKIs) are a new approach for hormone receptor positive (HR+) metastatic breast cancer (mBC). Reported VTE rates in randomized trials range from 0.6% for ribociclib (MONALEESA-2) to 2% for palbociclib (PALOMA-3) and 5% for abemaciclib (MONARCH-3) but these may underestimate actual rates compared to patients in clinical practice who are generally older and have greater comorbidities. Little is known about real world incidence or prevalence of VTE with CDKIs in mBC. The aim of this study was to evaluate rates of VTE in clinical practice and association with outcomes in mBC patients on CDKIs. Methods: We conducted a retrospective cohort study at Cleveland Clinic Taussig Cancer Institute, approved by the institutional review board. We identified consecutive mBC patients who received any of three FDA-approved CDKIs (palbociclib, ribociclib, abemaciclib) from 1/2015 through 12/2017. VTE including deep venous thrombosis (DVT) and pulmonary embolism (PE) were identified by electronic medical record review. Overall survival (OS), progression free survival (PFS) and time to VTE were estimated by the Kaplan-Meier method and evaluated for association with VTE using Cox proportional hazard regression. Results: The study population included 424 patients, with a median age at diagnosis of 54.76 yrs, (range 27 -85). Palbociclib was the most commonly used CDKI (n = 390, 91.8%); 27 patients (6.3%) received more than one drug. VTE during CDKI therapy occurred in 9% of patients (n = 38), including DVT in 52.6% (n = 20), PE in 18.5% (n = 7) and visceral vein thrombosis (VVT) in 15.8% (n = 6). Median time to VTE was 314 days, and 6-months rate was 4.1%. VTE was associated with numerically worse PFS and OS, but this was not statistically significant (PFS [HR 1.25, 95% CI 0.73 – 2.14, p = 0.42], OS [HR 1.60, 95% CI 0.89 – 2.87, p = 0.12]). Conclusions: VTE affected nearly 10% of breast cancer patients receiving CDKIs, 2- to 5-fold greater than reported in registration trials. Further work is necessary to identify pathophysiology, risk factors and benefit of thromboprophylaxis.

e12515 Background: Women with stage IV HER2+ breast cancer typically have longer overall survival (OS) compared to other breast cancer subsets due to the effectiveness of dual anti HER-2 antibody therapy. The role of mastectomy remains controversial. Methods: We reviewed patients who were diagnosed with stage IV HER2+ breast cancer between 2/2015 and 12/2017 at Cleveland Clinic. Overall survival (OS) was estimated by the Kaplan-Meier method, and compared by the log-rank test. Univariable and multivariable analysis were performed using Cox regression to identify predictors of OS. Results: We identified 47 patients, with a median age of 58 (range: 22 – 87). Twenty-eight (60%) and 14 (30%) patients had ER+ and PR+ disease, respectively. Four patients had brain metastasis at time of stage IV diagnosis. All patients received systemic therapy. 17 (36%) patients underwent mastectomy after diagnosis of stage IV breast cancer,. Of the 30 (64%) patients who did not undergo mastectomy, 24 (80%), 2 (7%), and 4 (13%) were treated with both chemotherapy and HER2-directed therapy, chemotherapy alone, and HER2-directed therapy alone, respectively. Breast radiotherapy was performed on 9 (53%) and 8 (27%) patients in mastectomy and no mastectomy cohorts, respectively. Median follow-up time was 22 months . The two-year OS for mastectomy and no mastectomy cohorts were 94% (95% CI: 83 – 100) and 50% (95% CI: 33 – 76), respectively (p=0.009). On univariable analysis, only mastectomy vs no mastectomy (HR: 0.18, 95% CI: 0.04 – 0.80, p=0.025) predicted OS. On multivariable analysis, mastectomy vs no mastectomy has remained to be statistically significant predictor of OS (HR: 0.08, 95% CI: 0.01 – 0.66, p=0.019), whereas age, chemotherapy, HER2-directed therapy, and breast radiation were not independent predictors of improved OS (p>0.05). Conclusions: In our cohort, mastectomy was an independent predictor of longer OS in women with stage IV HER2+ breast cancer.

e12538 Background: CDK inhibitors (CDKi), in combination with aromatase inhibitors (AI), are approved for the treatment of hormone receptor positive (HR+) Her2 negative metastatic breast cancer (MBC). The effectiveness of continuing CDKi beyond first disease progression is not known. This study evaluated real world evidence and assessed the impact of continuation of CDKi beyond first disease progression in combination with endocrine therapy. Methods: This is a retrospective, single institution review of HR+ MBC patients treated with CDKi from 2015-2018 who continued CDKi after initial progression. The primary outcome was progression-free survival (PFS) beyond first disease progression, as assessed by the clinician based on radiological and/or clinical criteria. Overall survival (OS) – defined as date of initial CDKi treatment to date of death or last follow up – was a secondary outcome. Results: 30 women with HR+/HER2- MBC, median age 47.5 years (range: 31 – 81), sequentially continued on CDKi beyond first progression were identified from a database of patients treated with Palbociclib. Median and average follow up times on CDKi were 27.18 and 24.53 months, respectively. Initial endocrine/CDKi regimen received included: palbociclib (PA)/letrozole (LTZ) [67%], PA/fulvestrant (FULV) [23%], and PA/other AI [10%]. Prescribed combinations beyond 1st progression were: PA/FULV [56.7%], PA/LTZ [16.7%], and PA/other AI [20%], abemaciclib plus LTZ or FULV [6%]. As of 1/31/2019, 25 patients (83.3%) were still alive, and 19 (63%) had undergone a second progression on CDKi. The estimated median PFS for the entire duration while on CDKi was 23.5 months (95% CI 12.8 – 27.8), of which 11.8 months (95% CI 5.34 – 13.13) was the median PFS beyond first progression. The estimated median OS was 45.4 months. Conclusions: Among a small cohort of HR+ MBC patients, in a non-clinical trial setting, continuation of palbociclib plus endocrine therapy beyond first progression was associated with a median PFS of approximately 11 months. Formal clinical evaluation of continuation of CDK inhibitor plus endocrine therapy beyond first progression is warranted.

1027Background: T-DM1 is an antibody-drug conjugate composed of trastuzumab and the maytansinoid antimicrotubule, DM1. T-DM1 was granted FDA approval in 2nd-line MBC after prior trastuzumab (T) and...

e13034Background: Palbociclib (PA), a CDK4/CDK6 inhibitor, has been shown in clinical trials to prolong progression free survival (PFS) in HR+/Her-2 negative MBC when combined with an aromatase inh...

6547Background: Due to complexity of disease and treatments, oncology patients (pts) have among the highest hospitalization rate. In our cancer institute, just 6% of all discharged pts accounted fo...

171 Background: Due to complexity of disease and treatments, oncology patients have among the highest hospitalization rate, especially towards End of Life (EOL). In our cancer institute, just 6% of all discharged patients accounted for >40% of unplanned readmissions, and continue to be highest risk of future admissions, ICU stay, ED visits, overuse of chemotherapy and under use of hospice care. We hypothesized that developing individualized care plans (ICP) for this high-utilization group will provide guidance in the complex care they require to reduce unnecessary and aggressive medical services. Methods: An Interdisciplinary Care Team (ICT) was created consisting of palliative medicine and oncology physicians, social workers, care coordinators, and nurses. On a bimonthly basis, patients with at least two unplanned hospital readmissions over the last 60 days were identified. ICPs were created using a team-based approach with parallel input from patient’s primary outpatient providers. Results: A total of 36 patients, 226 hospitalizations, and 163 ED visits were evaluated over a 6-month period, with an average number of hospitalizations of 1.08 per patient month (ppm). After implementation of ICP, hospitalizations decreased to 0.23 ppm, with an average length of stay decrease from 7.17 to 4.06 days per admission. Average ED visits decreased from 0.58 to 0.34 ppm, and the average number of unplanned readmissions decreased from 0.43 to 0.13 ppm. Of the 10 patients expired since creation of ICP, 8 utilized hospice care, while 2 patients died in an ICU. Average time to death from creation of ICP was 72 days among this cohort, while time to death from last exposure to chemotherapy was 58 days. Conclusions: Creation of individualized care plans for high-utilizing cancer patients decreased number of hospitalizations, ED visits, unplanned readmissions, and length of stay. A dedicated focus from a team of experts, beyond disease biology, on a unique patient situation may result in improved patient experience with decreased aggressiveness of care at EOL and overall resource utilization.

522 Background: Breast cancer risk stratification with the 70-gene signature (70-GS) provides a binary low risk (LR) or high risk (HR) result; by contrast the 21-gene assay (21-GA) provides LR, intermediate (IR), and HR results. Results from these two assays were compared for 769 patients from 5 institutions. Methods: The study included patients from McGill University (n = 86), UPMC (n = 437), USF (n = 135), Morton Plant North Bay Hospital (n = 79), and Cleveland Clinic (n = 32, all 21-GA IR). Results: With the 70-GS, 487 (63%) patients had a LR and 282 (37%) patients had a HR result. Excluding 32 cases selected for 21-GA IR results (n = 737), the 21-GA gave 369 (50%), 250 (34%), and 118 (16%) patients with LR, IR, and HR scores, respectively. Using the TAILORx cutoff, there were 134 (18%), 432 (59%), and 171 (23%) patients with LR, IR, and HR scores, respectively. There were 329 (45%) and 486 (66%) patients who were not classified in the same risk category by both assays using the clinical and TAILORx cutoffs for IR, respectively. Conclusions: In a large multi-institutional study the 70-GS and 21-GA results were discordant in 45-66% of patients, and the proportion of patients with a 21-GA score in the IR range varied from 34-59%. The 70-GS provided clinically actionable results for all patients. [Table: see text]

56 Background: MDSCs are a phenotypically diverse population of bone-marrow-derived cells that play an important role in tumor progression based on their immunosuppressive and pro-angiogenic properties. Mobilization of MDSCs from the bone marrow into the blood stream is dependent on tumor derived factors. Therefore, it is likely that levels of circulating MDSCs in cancer patients are reflective of tumor activity. However, the clinical applicability of MDSCs as a predictive biomarker in cancer has been hindered by a lack of consensus on the surface markers that defines them. Although there have been several attempts at harmonizing the nomenclature and characterization standards of MDSCs, there is still a lot of confusion particularly in the detection of MDSCs in unfractionated peripheral blood. To address this issue we have devised a simple flow cytometric test that reliably detects in whole blood the MDSCs subtypes that have been consistently reported to have clinical relevance in cancer. Methods: Whole blood from cancer patients and normal volunteers was stained with Lin cocktail, HLADR, CD33, CD15, CD14, and CD11b antibodies. After staining, red blood cells were lysed, samples were washed and analyzed by flow cytometry. Results: This approach identifies monocytic MDSCs (Linlo/CD33+/HLADR-/CD14+/CD15-), granulocytic or polymorphonucleated polymorphonuclear MDSCs (Linlo/CD33+/HLADR-/CD14-/CD15+), and promyelocytic or immature MDSCs (Linlo/CD33+/HLADR-/CD14-/CD15-), as well as Linlo/HLADR-/CD33+/CD11b+ MDSCs. This latter subpopulation has been reported by us and others to correlate with disease stage and tumor burden in several solid malignancies. This assay also allows the accurate determination of absolute numbers of each MDSC population when white blood cell counts are taken into consideration. Conclusions: Our proposed simple and rapid flow cytometry-based assay allows the measurement of frequency and absolute numbers of circulating MDSC subpopulations in unfractionated blood, and could provide a reference for their further characterization as predictive biomarkers for solid malignancies.

148 Background: Hospitalized patients (pts) on solid tumor oncology (STO) services have palliative needs including pain management. This quality improvement project sought to establish a new STO-specific co-rounding PM consult service, evaluate the use of consult criteria for STO inpatients, assess the impact/interest in an embedded service, improve access to PM for STO pts and improve palliative education to STO teams. Methods: During October 2015 to January 2016, a new PM consult service was established for the 2 STO inpatient services at Cleveland Clinic. The PM attending physician (MD) rounded with each of the STO teams twice a week. On weekdays, the PM MD chart-screened all STO pts for palliative needs such as uncontrolled pain (2 pain scores ≥ 6 out of 10 in 24 hours), unplanned readmission within 30 days or contact with a PM MD as an outpatient. Other PM needs were assessed on rounds. PM consults were offered for pts who screened positive and were performed if approved by the STO team. STO MDs were surveyed anonymously regarding acceptance of the embedded service. Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) data regarding pain management during the pilot period was reviewed. Results: Average daily census for the 2 STO teams was 28 pts. There were 282 positive palliative screens in 4 months, 119 of whom were seen in consultation (42%.) The embedded service saw 42-45 new consults per month. The PM team followed 22-35% of all STO inpatients. 14-18 pts/month new to PM were referred to the outpatient PM clinic after discharge. STO MDs indicated strong acceptance of the embedded PM team for pain management, STO team education and coordination of care. All STO MDs wanted the service to continue. HCAHPS pain scores for the entire STO floor improved from a baseline 39th percentile to 98thpercentile. Conclusions: PM was integrated successfully into daily hospital care of STO pts at our institution using a co-rounding model and consult criteria. The service was busy and well received by STO MDs. Continuity with outpatient PM was provided. HCAHPS pain scores improved for the entire STO floor, including pts not directly seen by PM.

594Background: HER-2 overexpression confers a more aggressive tumor phenotype in breast cancer, even in stage 1 disease. The large adjuvant trastuzumab trials excluded patients with node-negative tumors smaller than 1 cm. Subset analysis of one trial suggested patients with tumors 1–2 cm in size derived at least as much clinical benefit from adjuvant trastuzumab as did the overall cohort.It is currently unknown whether patients with T1a-T1bN0 HER-2+ tumors should routinely receive adjuvant chemotherapy and trastuzumab. Our goal was to examine clinical outcomes in patients with small HER2+ tumors treated at Cleveland Clinic (CC) to help guide treatment decisions. Methods: Cancer data registry was utilized to identify patients with stage I HER2+ cancer treated with or without trastuzumab and chemotherapy at CC after 2006. Patient charts were reviewed for recurrence and survival information. Disease-free survival (DFS) was defined as time from date of diagnosis to relapse, second primary cancer or death, whi...

e18217Background: Use of staging imaging (SI), including CT, PET and bone scans, is not routinely recommended in stage I-IIA breast cancer patients (ESBCP) in the absence of signs or symptoms of me...

e18185Background: Reducing unplanned hospital readmissions is a national policy priority aimed at improving health care quality. It is unclear if readmissions are preventable in cancer patients wit...

6607Background: Patients (pts) diagnosed with unresectable stage III or IV melanoma have 5-year survival rates of 15-20%. Recent FDA approved treatment options include pembrolizumab (PEM), nivoluma...

140 Background: The quality of care transfers is known to influence clinical outcomes. In an inpt onc setting at a major tertiary care referral center, patient (pt) adm originate from many different areas and times. A detailed evaluation of onc adm by source of transfer, admission time, and provider type, may identify opportunities to improve inpt clinical outcomes. Methods: We retrospectively reviewed all adm to the inpt solid tumor onc service from July - December 2014 from CCF regional hospital emergency departments (ED), outside hospital (OSH) ED, OSH inpt services, and CCF outpt clinics. Pts were excluded if the adm was planned or if admitted from the CCF Main Campus ED. Data collected included pt and encounter characteristics and provider type (house-staff or nocturnal hospitalist). Clinical outcomes, including activation of the adult medical emergency team (AMET), ICU transfers, length of stay (LOS), and in-hospital mortality were compared using chi-squared test; ECOG PS and LOS with the Kruskal-Wallis tests and Wilcoxon rank sum test. Results: A total of 413 unique pt admissions were reviewed. 213 were included after exclusion criteria were applied. The probability of AMET activation, mortality, and LOS differed by origin of transfer. Pts admitted from CCF regional EDs had the lowest median LOS and no deaths. OSH int transfers demonstrated significantly higher mortality vs other origins of transfer. Pts whose first orders were placed after 5pm had no significant differences in AMET activation, ICU transfers, LOS, or mortality vs daytime adm. There were no differences in adverse outcomes by the type of admitting provider. Conclusions: Onc inpts transferred from an outside healthcare setting were at highest risk for adverse outcomes (AMETs, increased LOS, and mortality) include those originating from OSH inpt services. Process and communication interventions focused on transfers from outside inpt facilities may improve safety and outcomes in this population. [Table: see text]

123 Background: The Cleveland Clinic is utilizing disease specific teams to craft care paths for value based disease management. The report by Groheux (JNCI 2012; 104:1879-1887), reported 18 FDG-PET-CT detected distant metastasis in 10.7, 17.5 and 36.5% of stage IIB- IIIB breast cancer patients. Based on this report, the breast team incorporated 18 FDG-PET- CT imaging for initial staging of IIB-IIIB patients. Methods: Treatment algorithms were developed by a multi-disciplinary breast cancer team. The content was reviewed by stakeholders throughout the health system, and feedback was incorporated into the care paths as appropriate. Content was communicated using physician meetings, electronic communication, tumor boards, and operational pilots. The treatment algorithm, along with the supporting narrative, was placed on the cancer center intranet, accessible to all practicing physicians. Additional intranet analysis was performed using Google Analytics, which identifies the site location and frequency of document downloads. We retrospectively examined the impact of this posting after one year (7/2014-7/2015). Manual chart review identified new patient consults for Stage IIB-IIIB breast cancer at our main campus and highest volume regional sites. Results: PET scans were ordered in 36 stage IIB, 19 stage IIIA, and 3 stage IIIB patients. Of the 36 IIB patients scanned, occult metastatic disease was noted in 1, and 2 patients migrated to IIIA without a change in treatment. 19 IIIA patients had no change in stage, although an asymptomatic second primary lung cancer was discovered in 1 .Three stage IIIB patients were without staging change. Conclusions: PET/CT imaging did not contribute to patient management for stage IIB-IIIB breast cancer patients at our institution, unlike the results reported elsewhere. Analysis of care path metrics allows us to expediently review and adjust recommendations to affiliated physicians.

589 Background: Brain metastasis (BM) is a serious complication of HER2+ breast cancer (BC). We evaluated prognostic factors for overall survival (OS) in a contemporary cohort of patients (pts) wit...

TPS1110 Background: The internalizable transmembrane glycoprotein NMB (gpNMB) is overexpressed in 20% of BC, including 40% of TNBC (Yardley JCO, in press), where it is a poor prognostic marker (Ros...

531 Background: Pertuzumab is FDA approved for use in combination with trastuzumab and chemotherapy as neoadjuvant therapy in women with non-metastatic HER2+ breast cancer. The TRYPHAENA trial repo...

260 Background: American Society for Gastrointestinal Endoscopy guidelines recommend endoscopic metal stent placement for pancreatic carcinoma patients with biliary obstruction and estimated life expectancy of >6 months. Because life expectancy of many such patients has until now been rd party payer perspective in 2012 U.S. dollars/QALY. In sensitivity analyses, overall survival was varied from 6-24 months to assess the impact of uncertainty in estimates on model outcomes. Results: Patients with metal stents had lower costs and greater overall and quality-adjusted survival. Placement of metal stents saved approximately $1,500 per patient over a lifetime, improving OS by 0.07 months and quality-adjusted survival by 0.10 months. These findings were robust in sensitivity analyses varying the length of survival for patients with pancreatic cancer. Conclusions: This model demonstrates that placement of metal biliary stents at initial onset of obstructive jaundice in patients with stage III/IV pancreatic adenocarcinoma is cost saving and improves survival when compared with use of plastic stents.

e22135 Background: Clinical studies have shown that circulating tumor cell (CTC) detection via the CELLSEARCH assay is a validated prognostic marker that is a predictor of overall survival (OS) in metastatic breast cancer (mBC) patients. By analyzing real-world observational data in one center, we evaluated the impact of CTC testing on predicting OS. Methods: Electronic medical records (EMR) were used to identify patients who were: (i) women > 18 years diagnosed with mBC on or after their date of registration, (ii) registered to the EMR system on or after 4/1/2010, and (iii) followed for at least 2 months from the time of metastases. Patients who had CTC testing were required, on average, to have at least 1 valid CTC test result every 3 months. OS from the date of initial diagnosis of metastases among mBC patients with CTC testing was compared to a cohort of patients without CTC testing (non-CTC) using a Cox model controlling for covariates such as age, race, hormone receptor status, and scanning frequency. Results: Mean (SD) ages of CTC (N=79) and non-CTC (N=1,146) patients were 65 (13) and 64 (14) years, respectively. ER or PR+ patients comprised 67% of the CTC population, and 64% of the non-CTC population. The majority of patients in both cohorts were Caucasian (89% - CTC, 57% - non-CTC). The Cox model showed that the group where CTC testing occurred had a statistically significant 40% reduction in risk of death (HR=0.60, p-value=0.03); this reduction was even greater among ER or PR+ patients who received CTC testing compared to those who did not (51%, HR=0.49, p-value=0.03). Conclusions: This retrospective analysis shows that the CTC population had a better prognosis than the non-CTC population. Thus, the assessment of CTCs could be associated with improved monitoring and prediction of cancer progression, and improved evaluation of response to therapy in mBC patients, possibly guiding more informed care decisions. The reliance on a single center and the relatively small number of CTC patients limits the ability to rule out other potential confounders in the analysis. Further studies using larger CTC samples and data from other centers with CTC testing are needed to verify and better understand this result.

6548 Background: Everolimus in combination with exemestane is approved for the treatment of postmenopausal women with hormone-receptor (HR) positive, HER2-negative metastatic breast cancer (MBC). The BOLERO-2, a randomized phase 3 trial, demonstrated a significantly improved progression free survival (PFS) with everolimus plus exemestane compared to exemestane alone in patients previously treated with non-steroidal aromatase inhibitors. In order to better inform U.S. policymakers, this study aimed to assess the cost-effectiveness, from a payer perspective, of everolimus in combination with exemestane. Methods: We created decision analytical and Markov models using published data from the BOLERO-2 trial. Utilities were derived from available literature. Costs were obtained from the Center for Medicare Services drug payment table and physician fee schedule and were represented in 2012 U.S. dollars. The quality-adjusted life-years (QALY) and incremental cost-effectiveness ratio (ICER) were calculated. One way and probabilistic sensitivity analyses were performed. Results: Everolimus added 0.42 years of progression-free survival (PFS) by central radiographic assessment with an incremental cost of $33,103, an overall cost of $62,751.54 per year of PFS gained, and an ICER of $79,376/QALY. By local assessment, everolimus added 0.29 years PFS years with an incremental cost of $31,873, an overall cost of $83,222 per year of PFS gained, and an ICER of $108,131/QALY. The results of the model were robust in sensitivity analyses. The primary drivers in this model were found to be: PFS duration, progression free probability on therapy, and overall everolimus cost. Conclusions: Everolimus plus exemestane appears to be cost-effective in the treatment of metastatic breast cancer. Based on efficacy and value, this newly approved combination should be considered to be a viable option in treating patients with HR+/HER2- MBC upon progression on non-steroidal aromatase inhibitors.

e11591 Background: Emerging resistance to single agent aromatase inhibitors (AI) or fulvestrant as first line treatment for postmenopausal women with advanced hormonal receptor positive breast cancer calls for alternative therapeutic options. This meta-analysis studies the effectiveness of combination therapy with fulvestrant and an AI, as compared to an AI alone in first line treatment of postmenopausal patients with hormonal receptor positive relapsed or metastatic breast cancer. Methods: Literature search was performed using PubMed, Google Scholar, Embase, ASCO and ESMO to search for studies published during the last 10 years using relevant keywords. Two prospective randomized clinical trials were found to fulfill the search criteria for combination of fulvestrant + AI vs. AI alone (both studied anastrozole in combination with fulvestrant). Meta-estimates were calculated by combining study estimates using the DerSimonian and Laird random effects model. The linear mixed-effects model was used to generate 95% prediction intervals for study-specific hazard and odds ratios. Results: Pooled hazard ratio for progression free survival was 0.88 (95% CI 0.72-1.09, 95% prediction interval [PI] 0.65-1.21). Pooled HR for overall survival was 0.88 (95% CI 0.72-1.08, 95% PI 0.68-1.14). Pooled odds ratio for response rate was 1.13 (95% CI 0.79-1.63, 95% PI 0.78-1.65). Conclusions: Pooled results showed a small, non-statistically significant, improvement in progression-free and overall survival. These results do not support the use of combination therapy with fulvestrant and anastrozole in the first line treatment of postmenopausal patients with hormonal receptor positive relapsed or metastatic breast cancer.

TPS11124 Background: Triple-negative breast cancer (TNBC) is a subtype of breast cancer that is clinically negative for expression of estrogen and progesterone receptors (ER/PR) and human epidermal growth factor receptor-2 (HER2). It is characterized by its unique molecular profile, aggressive behavior, distinct patterns of metastasis, and lack of commercially available targeted therapies. Chemotherapy has been the mainstay of treatment for women with TNBC, but this current standard-of-care is suboptimal. AEZS-108 is an LHRH-cytotoxic hybrid drug whose rational design covalently couples the carrier D-Lys6-LHRH (an LHRH agonist) to the cytotoxic doxorubicin radical. Because LHRH receptors are expressed in a majority of TNBC, AEZS-108 represents a novel way to selectively deliver cytotoxic chemotherapy in patients with TNBC via a new therapeutic target. Methods: In this open label randomized two-arm multicenter phase II study, patients will be randomized in a 1:1 ratio into one of the two treatment arms: AEZS-108 (267 mg/m2 every 21 calendar days) [Arm A] or SSC (standard single agent cytotoxic chemotherapy [Arm B]) at discretion of treating oncologist cycled every 21 calendar days. Stratified randomization will be used with number of prior lines of therapies (1-2 vs. >2), ECOG performance status 0-1 vs. 2, and liver metastases (absent vs. present). Analysis of the main study endpoint, TTP, will follow a group sequential design with two interim analyses, including the final analysis. O’Brien Fleming stopping boundaries will be used. The primary endpoint is to evaluate the median time to progression (TTP) of AEZS-108 in patients with chemotherapy resistant advanced TNBC treated with AEZS-108 in relation to patients receiving standard single agent cytotoxic chemotherapy. Secondary endpoints include: overall response, clinical benefit, duration of response, overall survival, toxicity profile and quality of life (QoL). Clinical trial information: NCT01698281.

138 Background: Perioperative chemotherapy (chemo) with ECF (epirubicin/cisplatin/5-fluorouracil) plus surgery improved survival over surgery alone in GC in the MAGIC trial. Herein we report our experience using DCF in the perioperative setting in patients (pts) with locally advanced GC. Methods: We conducted a retrospective IRB-approved study of pts with potentially resectable locally advanced GC who were treated with DCF with neoadjuvant intent. Pts received 3 cycles of preoperative (pre-op) DCF every 3 weeks, followed by surgery, then 3 cycles of postoperative (post-op) DCF. Patients with a poor pathologic response could be changed to radiation (RT) or an alternate chemo regimen postop. Results: A total of 41 pts were identified, 24 with gastric and 17 with GEJ adenocarcinoma. All pts received at least 1 cycle of DCF and 78% received at least 3 cycles pre-op. Five pts progressed during neoadjuvant DCF, 4 were unresectable by CT after neoadjuvant DCF and 2 were lost to follow-up. The remaining 30 pts had surgery with curative intent. Post-op, 2 pts were lost to follow-up, 12 received DCF (with 6 of these also receiving RT), 11 received a different chemo regimen due to a poor response to neoadjuvant DCF (including 6 pts who also received RT). Two pts received post-op RT only. The median PFS was 16.8 months (95% CI 7.7 - 25.9) and the median OS was 26.9 months (95% CI 18.7–35.1). The PFS was longer for pts who had a radiological or pathological response to neoadjuvant DCF (log rank p = 0.005 and 0.02 respectively) and for pts who received DCF post-op (log rank p = 0.005). Among pts who did not receive DCF post-op, there was no survival difference between the pts who were switched to an alternative chemo or chemoRT regimen post-op compared to those who received no further therapy. The most common chemo-related adverse events were anemia (27% grade 3 or 4), nausea/vomiting (17% G3 or 4), and febrile neutropenia (12%). Conclusions: The DCF regimen is well tolerated in locally advanced GC. Patients who do not have a good response (either radiologic or pathologic) to pre-op DCF appear to have a poor prognosis regardless of the post-op treatment given.

9565 Background: Cure rates have reached a plateau for many pediatric solid tumors. Identifying new therapeutic targets, biomarkers of response and prognostic indicators should improve clinical outcomes. Accumulation of myeloid derived suppressor cells (MDSCs) is an important mechanism of tumor mediated immune evasion. Increased levels of MDSCs in adult cancer patients correlate with a worse prognosis, and decrease in levels with treatment is associated with benefit. Little is known about MDSCs in childhood, or in children with cancer. This pilot measured levels of MDSCs in pediatric patients with cancer and healthy children. Methods: We enrolled subjects using an Institutional Review Board approved protocol after obtaining informed consent. Blood was obtained from 14 children with newly diagnosed or recurrent solid tumors at start of therapy. In 10 of these patients, levels were also drawn after therapy. Blood was obtained once from 6 healthy children in a primary care office. Samples were obtained concurrently with complete blood counts. MDSCs were measured on fresh whole blood and were defined as Lin-1+/HLADR-/CD 33+/CD11b+/ by flow. Total MDSC numbers were then calculated. Results: The mean total number of MDSCs was 596 cells/ul at diagnosis for the 14 children with cancer and 170 cells/ul for the 6 healthy children (t(18) = 3.02, p = .0073). For the 10 children with cancer who had repeat values measured after treatment, MDSCs decreased in 4 and increased in 6. The mean initial MDSC count for these children was 494 cells/ul, and the mean post treatment count was 1716 cells/ul (t(9) = 1.81, p = .1036). Larger change scores tended to be associated with children treated with alkylator therapy followed by G-CSF. The results for percent MDSC in white cells mirrored those of total number. Conclusions: Circulating MDSCs were higher in pediatric cancer patients than healthy controls. Cancer treatment did not reliably reduce MDSC levels. After some treatments, the levels increased, potentially increasing immune tolerance. Further research is needed to determine if circulating MDSCs are a reliable predictive or prognostic marker in pediatric cancer, and whether they represent a potential target for therapeutic intervention.

e15107 Background: Treatment with abiraterone improves overall survival (OS), time to prostate-specific antigen progression and radiologic progression-free survival when added to prednisone and best supportive care in patients with advanced castrate-resistant prostate cancer (CRPC) who previously received docetaxel. Little is known about its cost-effectiveness in the United States. Methods: In order to raise awareness of pharmacoeconomics concepts and inform policy-makers in the US, this study aimed to assess the cost-effectiveness of abiraterone in the treatment of advanced CRPC patients, from a payer perspective. We created a decision-analytical model using clinical data from the pivotal phase III trial: COU-AA-301. Health utilities were derived from the available literature. Costs for drug acquisition, physician visits and laboratory tests were obtained from the Center for Medicare Services Drug Payment Table and Physician Fee Schedule and are represented in 2011 US dollars. Life-years saved (LY), Quality-adjusted life years (QALY) and Incremental Cost Effectiveness Ratio (ICER) were calculated using updated survival data presented at the 2011 ASCO meeting. Other main scenarios used the original median survival data published in the New England Journal of Medicine article and adjusted median OS to represent an overall survival hazard ratio of .66. Sensitivity analyses were performed using the confidence intervals for median OS in the pivotal study as well as by halving or doubling all other model inputs. No discounting was used due to the short time-horizon. Results: Abiraterone added 0.38 LY and 0.30 QALY with an incremental cost of $39,320 and therefore a cost of $102,600/LY and an ICER of $129,000/QALY. The main drivers of the model were drug acquisition cost, median OS, and health utility values. The results of the model were robust in different scenarios and sensitivity analyses. Conclusions: Using commonly accepted willingness-to-pay thresholds, such as those suggested by the World Health Organization, treatment of patients with advanced CRPC patients with abiraterone is likely to be cost-effective in the US.

e11038 Background: NACT is used for the treatment of locally advanced breast cancer. Evaluation of breast cancer subtypes and their responses to specific chemotherapy regimens is important. Methods: Retrospective review of medical records from University of Miami/Sylvester Comprehensive Cancer Center/Jackson Memorial Hospital from 1998-2011. Complete pathologic response (pCR) was classified as absence of malignant tissue both in the breast and the lymph nodes in the surgical specimen. Results: One hundred ninety-one women with ER-positive breast cancer received platinum based NACT. Mean age was 52 years (range 31-83). One hundred eleven (58.1%) women were premenopausal and 80 (41.9%) postmenopausal. The clinical stage at presentation: IIA (6.3%), IIB (26.7%), IIIA (31.9%), IIIB (30.9%), IIIC (4.2%). Sixty one (31.9%) pts had her-2/neu positive tumors. pCR occurred in 17 patients (8.9%), premenopausal vs. postmenopausal (9.9% vs. 7.5%, p=0.56), white race vs. black (9.9% vs. 4.9%, p=0.32), Hispanic vs. non-Hispanic (9.4% vs. 6.8%, p=0.6), tumor size 4 cm (14.7% vs. 7.6%, p=0.19), her-2/neu positive vs. her-2/neu negative (13.1% vs. 6.9%, p=0.16), T4 vs. T1-T3 (4.8% vs. 10.9%, p=0.19). Conclusions: pCR rates with platinum based NACT in pts with hormone receptor-positive tumors were low. The pCR rate did not appear to be influenced by her2/neu, tumor size, race, ethnicity or menopausal status.

4043 Background: A Phase I/II trial evaluated single and repeated cycles of fractionated radioimmunotherapy (RAIT) with 90Y-labeled humanized mAb (90Y-hPAM4) plus Gem as first-line therapy in Stage 3-4 APC. Methods: Cycles of Gem once-weekly x 4 with 90Y-hPAM4 on wks 2, 3 and 4 were repeated until progression, withdrawal or unacceptable toxicity. In Part I, 90Y doses were escalated with Gem fixed at 200 mg/m2. In Part II, Gem was increased up to 1000 mg/m2, with 90Y fixed at 12 mCi/m2 for cycle 1 and lowered for retreatment. Results: Of 100 pts entered, 10 withdrew early, while 90 (73 stage IV) received 1-4 cycles. In Part I, 38 pts received 90Y-hPAM4 weekly x 3 at 6.5, 9, 12, or 15 mCi/m2, with the same cycle repeated 1-3 times in 13 pts. By CT-RECIST criteria, 6 pts (16%) had PRs and 16 (42%) had stabilization as best response (58% disease control). After cycle 1, 52% (13/25) with PET-avid images had >25% SUV reduction, and 33% (9/27) with elevated CA19-9 levels decreased by >50%. The median OS was 7.7 mo., but 11.8 mo. for retreated pts [46% (6/13) survived ≥1 yr.], and with improved efficacy at higher 90Y doses. NCI-CTCv3 Grade 3-4 platelets or ANC developed in 20/38 (53%) after cycle 1 (all reversible to Grade 1) and in all retreated pts (irreversible in 4/9 pts at 12 or 15 mCi/m2). In Part II, 52 pts received increased Gem without evidence of improved efficacy, while 13 pts were retreated with more acceptable toxicity at lower 90Y doses of 6.5 or 9 mCi/m2. Treatment was well tolerated with no infusion reactions. Infections requiring IV antibiotics occurred at a low rate and responded to appropriate coverage (bacteremia/sepsis, 7%; febrile neutropenia, 4%; ascending cholangitis, 3%; pneumonia, 2%; others 1%). One case of bleeding occurred, due to rectal tumor invasion. Anecdotal reports of good performance and decreased pain medication requirements require further validation. Conclusions: Fractionated RAIT with 90Y-hPAM4 combined with low-dose 200 mg/m2 GEM appears promising as a treatment regimen for APC. Hematologic toxicity was dose limiting. A 90Y-hPAM4 dose of 12 mCi/m2 for cycle 1 and 6.5 mCi/m2 for cycle 2 have been selected as suitable for further clinical development in the first-line setting.

e16525 Background: Eribulin was FDA approved in 2010 for pts previously receiving >2 prior chemotherapeutic regimens for MBC. This approval was based on the phase 3 trial (EMBRACE) which demonstrated that E significantly improved median overall survival (OS) relative to different TPC by 2.5 months [HR 0.81; p=0.041]. Methods: The aim of this study was to assess the cost-effectiveness of E versus the 3 most commonly selected TPC in EMBRACE. We also evaluated the cost effectiveness of E compared to other approved drugs for MBC. We created a decision-analytical model using clinical data from the EMBRACE trial. Health utilities were derived from the published literature. Costs for drug acquisition, physician visits, and laboratory tests were obtained from Medicare Services Drug Payment Table and Physician Fee Schedule and are represented in 2011 USD. Life-years saved (LY), Quality-adjusted life years (QALY), and Incremental Cost Effectiveness Ratio (ICER) were calculated using the EMBRACE median OS data. TPC cost was calculated with 3 most commonly used drugs: vinorelbine (V), gemcitabine (G), and capecitabine (X), comprising 60% of pts in the TPC arm. Other drugs analyzed included liposomal doxorubicin (D), nab-paclitaxel (A), and ixabepilone (I). Greater GCSF use with E vs. TPF was also accounted for in our model. Results: E added 0.208 LY and 0.119 QALY with an incremental cost over TPC of $24,035; and therefore a cost of $115,369/LY and an ICER of $201,790/QALY. The main drivers of the model were drug acquisition cost, OS, and health utility values. The results of the model were robust in sensitivity analyses. Because of the very low cost of V and G, we looked at the cost-effectiveness of E relative to several other drugs commonly used in this setting, but used in fewer pts in the pivotal trial. Relative to ixabepilone, liposomal doxorubicin, nab-paclitaxel, and capecitabine the ICER for E was $85,130, $98,538, $119,029, and $154,591, respectively. Conclusions: Using commonly accepted willingness-to-pay thresholds, E appears to be cost effective in the treatment of MBC relative to D, A, X, and I; with marginal cost effectiveness for less expensive drugs such as V and G.

227 Background: A Phase I/II trial was undertaken to evaluate repeated cycles of 90Y-labeled anti-mucin humanized mAb (90Y-hPAM4) plus Gem as first-line therapy in Stage 3-4 APC. Methods: Pts received Gem once-weekly x 4 with 90Y-hPAM4 on wks 2, 3 and 4, with cycles repeated until progression or unacceptable toxicity. In Part I, pts were treated in cohorts with escalating 90Y doses and Gem fixed at a low 200 mg/m2 dose for radiosensitization. In Part II, the Gem doses were increased up to standard levels, with 90Y doses fixed for first cycle, but decreased for subsequent cycles. Tumor responses were assessed by CT, FDG/PET and serum CA19-9; safety by NCI-CTCv3. Results: Of 100 untreated pts enrolled, 10 withdrew early, while 90 (73 stage IV) received 1-4 cycles. In Part I, 38 pts received 90Y-hPAM4 weekly x 3 at 90Y doses of 6.5 (N=4), 9 (N=12), 12 (N=17) or 15 (N=5) mCi/m2, with the same cycle repeated 1-3 times in 13 pts. Grade 3-4 platelets or ANC developed in 20/38 (53%) after cycle 1 (all reversible to Grade 1) and in all retreated pts (irreversible in 4/9 pts at 12 or 15 mCi/m2). There were 3 febrile neutropenias, 4 other infections treated with IV antibiotics, but no major bleeding or other AEs. By CT-RECIST criteria, 6 pts (16%) had PRs and 16 (42%) had stabilization as best response (58% disease control). After cycle 1, 52% (13/25) with PET-avid images became negative or had >25% SUV reduction, and 33% (9/27) with elevated CA19-9 levels decreased by >50%. The median overall survival was 7.7 mo., but 11.8 mo. for retreated pts [46% (6/13) survived ≥1 yr.], with improved efficacy at higher 90Y doses. In Part II, 52 pts received 12 mCi/m2 90Y-hPAM4 x 3 with Gem doses of 200 (N=17), 600 (N=8) or 1000 mg/m2 (N=27), with 13 pts now retreated at 90Y doses of 6.5 or 9 mCi/m2. Results so far indicate no advantage to giving higher doses of Gem with RAIT. Toxicity, response and survival data for this group will be presented at the conference. Conclusions: Fractionated RAIT with 90Y-hPAM4 combined with low-dose gemcitabine appears to be a manageable and active first-line therapy for APC. It may provide comparable efficacy yet less toxicity compared to other regimens.

3067 Background: Navitoclax (ABT-263), a novel BH3 mimetic, binds with high affinity and inhibits multiple antiapoptotic Bcl-2 proteins. Preclinical studies show synergy between navitoclax and G in...

6060 Background: Bevacizumab in combination with chemotherapy increases progression-free survival (PFS) but not overall survival (OS) when compared to chemotherapy alone in the treatment of metasta...

240 Background: The90Y-labeled anti-mucin humanized mAb, clivatuzumab tetraxetan (90Y-hPAM4), is in clinical development in APC. A phase I/II trial of 90Y-hPAM4 with low-dose radiosensitizing Gem has now concluded 90Y-dose escalation. Methods: Pts with untreated, inoperable, stage 3-4 APC received 200 mg/m2 Gem once-weekly x 4 with 90Y-hPAM4 on wks 2-4, and with 90Y escalated in cohorts by 3+3 design. Tumor responses were assessed by CT, FDG/PET and serum CA19.9, with cycles repeated until progression or unacceptable toxicity. Results: Of 42 pts (40-87 yrs, ECOG PS 0-1, 36 stage 4), 4 withdrew early while 38 received weekly x 3 90Y doses of 6.5 (N=4), 9 (N=12), 12 (N=17) and 15 (N=5) mCi/m2. Treatment was well-tolerated with few non-hematologic side-effects, including 13 pts retreated with 1-3 additional cycles. CTCv3 grade 3-4 plts or ANC developed in 21/38 (55%) pts after cycle 1 and all (100%) retreated pts after last cycle. Escalation reached limits on radiation doses to the marrow, but hematologic suppression was reversible without major infections or bleeding events, except for 3 pts after repeated cycles, one with extensive marrow tumor infiltration. By CT, the overall disease control rate was 55%, including 6 pts (16%) with partial responses (PRs) by RECIST criteria and 15 pts (39%) with stabilization as best response. After cycle 1, 43% (10/23) improved by PET studies (negative or >25% reduced uptake), and 36% (9/25) with elevated CA19.9 levels had >50 decreases. With 26% (10/38) of pts still in follow-up, 55% (21/38) have now achieved survival of ≥ 6 months [18% (7/38) ≥ 1 yr]. Treatment outcome may increase with 90Y dose, since pts treated at 3 x ≥12 mCi/m2 vs ≤9 mCi/m2 had 19% vs 6% PRs by CT, 47% vs 22% CA19.9 decreases, 63% vs 25% PET improvement, and 64% vs. 44% survival ≥ 6 months. Anecdotally, PS and pain level improved, which needs validation. Updated survival will be presented at the meeting. Conclusions: Fractionated 90Y-hPAM4 plus low-dose Gem showed encouraging therapeutic activity with manageable hematological toxicity. The 12-mCi/m2 dose level was selected for continued dose exploration now underway, involving standard Gem doses and adding maintenance Gem. [Table: see text]

e16036 Background: To correlate serum cytokine and angiogenic factor (CAF) levels and overall survival (OS) in metastatic renal cell cancer (mRCC) treated with interferon-alpha (IFN-α). Methods: Serum CAF levels were measured in 103 patients treated on a randomized trial with IFN-α 0.5 million units (MU) twice daily or 5 MU daily. Concentrations of 17 analytes including vascular endothelial growth factor (VEGFA) and several cytokines were determined by multiplex bead immunoassays or ELISA (basic fibroblast growth factor). We used proportional hazards models to evaluate the effect of CAF levels and clinical factors on OS. Results: Pretreatment serum interleukin (IL)-5, IL-12p40, VEGFA, and IL-6 levels, and Memorial Sloan-Kettering Cancer Center risk grouping were independently correlated with OS, with hazard ratios of 2.33, 2.00, 2.07, 1.82 and 0.39, respectively (concordance index = 0.69 for the combined model versus 0.60 for the CAF model versus 0.52 for the clinical model). Based on an index derived from these five risk factors (RF), patients with 0–2 RF had a median OS time of 32 months, versus 9 months for patients with 3–5 RF (p < 0.0001). Conclusions: Serum CAF profiling contributes to prognostic evaluation in mRCC and helps to identify a subset of patients with 20% 5-year OS. No significant financial relationships to disclose.

4046 Background: Adjuvant chemotherapy (AC) in patients with stage III colon adenocarcinoma prevents recurrences and improves survival. Although most trials mandate initiation of AC within 8 weeks of resection, the impact of timing of AC is still not clear with few studies reporting conflicting results. We hypothesized that AC commenced within 60 days of resection would increase survival in patients with stage II and III colon cancer. Methods: Patients with newly diagnosed stage II or III colon adenocarcinoma who received fluoropyrimidine based AC in two centers (a private cancer center and a large community hospital) between 2000 and 2007 were included into analysis. Time to adjuvant chemotherapy (TTC), overall survival (OS), and relapse-free survival (RFS) were calculated from the surgery date. Patients were dichotomized into early- (group 1) and late-treatment (group 2) groups using the TTC of 60 days. The demographic, clinical, and laboratory characteristics of patients in two groups were compared using chi-square and t-test. Kaplan-Meier survival curves were constructed employing univariate log-rank test to assess the effects of demographic and clinical characteristics on OS. Then the impact of TTC on OS and RFS was analyzed using a Cox proportional hazard model incorporating the significant factors found in the univariate analysis. Results: 190 patients were eligible for the study (median age: 57 yrs [range 14–84]). 116 patients (61%) were female and 35 patients (18%) had stage II disease. Median TTC was 49 days (range 24–196) and median follow-up was 143 weeks (range 9–451). 134 patients (70%) received AC within 60 days of surgery (group 1) and 56 (30%) received after (group 2). The only difference between the two groups was the higher N stage in group 1. The treating hospital and the N stage were found to be the factors affecting the OS in univariate analysis. Five-year OS for group I was 75.2% as compared to 61.3% for group II (HR 2.11, CI: 1.00–4.45, p=0.049). Five-year RFS for group I was 65.7% as compared to 59.0% for group II (HR: 1.19, CI: 0.65–2.20, p=0.570). Conclusions: Delay of AC more than 60 days after resection is associated with inferior survival in stage II/III colon cancer. No significant financial relationships to disclose.

15506 Background: The treatment of patients with locally advanced pancreatic cancer remains challenging with median survival in the range of 10 months. This study seeks to establish whether neoadju...

14503 Background: IFNa may prolong survival in MRCC patients (pts) due to stimulation of cell-mediated immunity. We hypothesized that IFNa exerts an anti-tumor effect by upregulation of Th1 cytokines and that patients (pts) with elevated serum levels of Th1 cytokines either at baseline (BL) or after treatment with IFNa would have a superior clinical outcome. Methods: Cytokine profiling was performed on 104 pts with MRCC treated in a randomized phase III trial with IFNa 0.5 million units (MU) subcutaneously (SC) twice daily or 5 MU SC daily. Serum samples were collected at BL (n = 104) and after 8 weeks of IFNa therapy (C1) (n = 89). Cytokine concentrations were determined using a 16-plex immunoassay. The linear mixed-effect model was fit to assess the change of cytokine levels from BL to C1. Cox proportional hazards model was fit to evaluate the effect of BL cytokine levels or change of cytokine levels from BL to C1 on the risk of death. Results: Of 16 cytokines evaluated (IL-1b, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12 p40, IL-13, IL-15, IL-17, IFNa, IFNg, GM-CSF, TNFa, VEGF), lower BL TNFa, IL-8 (Th1), and IL-13 (Th2) levels were associated with clinical benefit (major response or progression-free status at 6 months) (p = 0.01 and 0.03, respectively). By multivariate analysis, only extremely low or high levels of IFNa (p = 0.02) and IL-12 (p = 0.002) at BL were associated with an increased risk of death. IFNa therapy after C1 produced higher levels of several Th1 cytokines (IL-8, IL-12 p40, IL-15) (p < .001) and lower levels of Th2 cytokines (IL-4, IL-13). Unexpectedly, there were significantly lower levels of TNFa and GM-CSF (Th1) and higher levels of IL-10 (Th2) with IFNa. Only an increase of IL-2 levels from BL to C1 (RR 1.45; p = 0.05) correlated with an increased risk of death. Conclusions: Lower BL serum levels of TNFa, IL-8, and IL-13 were associated with clinical benefit to IFNa. Although IFNa therapy favored a shift towards a Th1 response, this effect alone did not correlate with clinical outcome. No significant financial relationships to disclose.

13051 Background: The ubiquitin-proteosome pathway plays an important role in cell cycle regulation, neoplastic growth, and metastasis. Nuclear factor-kB (NF-kB) is a key transcription factor in cancer, primarily regulated by proteosome-mediated degradation of the inhibitor protein I kappa B alpha-associated protein kinase (IkBa). Interruption of this degradative pathway, with the 26S proteosome inhibitor bortezomib, has displayed significant clinical anti-tumor activity, in multiple myeloma and lymphomas. The role of bortezomib in the treatment of advanced solid tumors, however, is unclear at present. Preclinical work in multiple myeloma cell lines has demonstrated synergy in cytotoxicity and growth inhibiton with the combination of bortezomib and the cyclooxygenase (COX)-2 inhibitors. We therefore hypothesized that the combination of the COX-2 inhibitor celecoxib and bortezomib is a potentially active combination in the treatment of advanced solid tumors. Towards this aim, we are conducting a phase I study to determine the maximum tolerated dose (MTD) and toxicity profile of bortezomib and celecoxib. Methods: Patients with advanced solid tumors and an ECOG PS of 0–2 received escalating doses of bortezomib (1.0–1.6 mg/m2), either weekly for 5 of 6 weeks or twice weekly for 2 of 3 weeks, and celecoxib (200–400 mg twice daily). Accrual was planned to a maximum of 6 cohorts, each with a minimum of 3 patients. Only those dose-limiting toxicities (DLTs) occurring during the first cycle were used to define the MTD. DLTs included any grade 4 hematologic toxicity related to therapy, or any grade 3 or 4 non-hematologic toxicity. Results: To date, 10 patients (median age of 63y) have been treated. Represented tumor types include stage IV: colorectal, pancreatic, bladder, leiomyosarcoma, head and neck squamous cell carcinoma, hepatocellular, and renal cell. A median number of 2 cycles were administered. Thus far, no DLTs have been observed. No cumulative toxicities have been noted. No objective tumor response has been observed, however, stable disease was maintained in one heavily pre-treated patient with metastatic papillary renal cell cancer for 5 months. Conclusions: The combination of bortezomib and celecoxib is safe and well tolerated, with no dose-limiting toxicities thus far. [Table: see text]

17077 Background: The overall 5-yr survival for the treatment of lung cancer patients is less than 20% due to the inability to control metastatic disease. Methods: To identify genes that promote the development of metastases in NSCLC and other solid tumors, we first performed three separate microarray analyses using Affymetrix U133A chips whereby expression values of a pool of normal lymph nodes was compared to: lung cancer cell lines (n = 4), and metastatic lymph nodes from breast (n = 3) and pancreatic (n = 3) cancer patients. Separate lists of the 35 most highly overexpressed genes for each cancer type were compiled. Results: We observed that each list contained EpCAM, XAG, CK19, and CK8 (p = 1.1E-18). To search for genes that might regulate expression of these four genes, we queried the CGAP NCI60 gene expression database with the 87 genes contained on the three lists and constructed a connectivity map such that the presence of a gene on the map required: 1) high correlation (p < 8.0E-6) with at least two other genes, and 2), direct or indirect contact to EpCAM or XAG. The map contained two gene clusters (XAG cluster = 6 genes; EpCAM/CK19/CK8 cluster = 7 genes) that were connected to, and potentially regulated by, the Ets transcriptional factor Esx. Genes from both clusters, as well as Esx, were highly overexpressed in metastatic mediastianal lymph nodes obtained from NSCLC patients. To investigate whether Esx might regulate expression of one or more genes in the two clusters, we transfected an NSCLC cell line derived from lymph node metastases with siRNA to Esx and observed: 1) a reduction in expression of EpCAM (4-fold), XAG (7-fold), and the orphan nuclear receptor ESRRα (1,000-fold), and 2), an inhibition of cell growth. Conclusions: Based on our ability to simultaneously inhibit cell growth and expression of multiple metastasis-associated genes with a single siRNA, we conclude that Esx is a major regulator of lymph node metastasis. No significant financial relationships to disclose.

656 Background: We aimed at determining the pathologic features of metastatic breast lesions to the brain, and their correlation with the primary tumor, and their effect on postneurosurgical survival (PNS). Methods: Tissue samples from 123 pts who underwent surgical resection of BM at MDACC between 1984–2004, were examined for estrogen receptor-α (ER) by IHC and HER2 by FISH or IHC. Pts’ medical records were retrospectively reviewed for CC. Results: Median (med) age at BM diagnosis was 51 years. The med time from primary diagnosis to BM was 33 months (Range (R)0–304). Infiltrating ductal cancers were found in 89%. ER was negative and HER2 amplified in 73% and 37% of BM, respectively; ER-/Her2- was seen in 43%, ER-/Her2+ in 30%, ER+/Her2+ in 6%. Concordance of ER and HER2 status between primary and BM were both 88% (p=NS, for discordance). ER positivity was associated with a longer primary diagnosis to BM interval [med 39 months (R, 0–213) vs. 29 months (R, 0–304) in the ER- pts; p=0.02]; HER2 amplification was not (p=0.82). The med PNS was 11.2 months (R, 9.0–13.5) for 112 pts with no pre- craniotomy brain radiation. Pts with ER+ metastases experienced apparent longer PNS survival than ER- pts [med 18.9 months (95% CI, 4.8–33.1) vs. 9.9 months (95% CI, 7.6–12.1 months), respectively, p=0.06], as did those with HER2-amplified BM [med 13.8 months vs. 9.9 months, respectively; p=0.10]. Pts aged 45 to No significant financial relationships to disclose.