Part 1: Abstracts From the NCCN 18th Annual Conference: Advancing the Standard of Cancer Care™

NCCN Oncology Research Program (ORP)-Funded Projects

AB2013-1. Phase I Study of Vorinostat With Concurrent Chemoradiotherapy (CRT) for Locally Advanced Nonsquamous Non-Small Cell Lung Cancer (NSCLC) (NCT01059552)

Ranee Mehra,^a Aruna Turaka,^a Joshua Meyer,^a Mary Pinder-Schenck,^b Keisuke Shirai,^c Neeta Somaiah,^d Brian Egleston,^a Hossein Borghaei,^a and George Simon^d

^aFox Chase Cancer Center; ^bH. Lee Moffitt Cancer Center; ^cHollings Cancer Center - Medical University of South Carolina; and ^dThe University of Texas MD Anderson Cancer Center

This study was approved and funded by NCCN from general research support provided by Merck & Co, Inc.

Background: Vorinostat (V) is a class I (histone deacetylase [HDAC] 1, 2, and 3) and II (HDAC 6) inhibitor of HDACs with IC50s in the nanomolar range. In A549 (non-small cell lung cancer [NSCLC]) and A375 (melanoma) cell lines (CLs), V accentuated radiotherapy (RT)-induced decrease in clonogenic survival, presumably via impaired DNA repair. In colorectal cancer CLs, V down-regulates thymidylate synthase, a target of pemetrexed (P), thus potentiating its antineoplastic effects. Long-term survival eludes patients with stage IIIA/IIIB NSCLC. Given this rationale, this study seeks to identify the maximum tolerated dose (MTD) of V with cisplatin (C), P, and RT in patients with nonsquamous NSCLC.

Methods: This is a phase I trial of V added to C, P, and RT for patients with stage IIIA/IIIB nonsquamous NSCLC. V is escalated in a standard phase I 3+3 design on a chemoradiation (CRT) platform of C (75 mg/m² q21d x 4), P (500 mg/m² q21d x 4 with folic acid, vitamin B₁₂, and steroid support), and RT (60 Gy). There is a small expansion cohort at the MTD. RT is administered with the first 3 cycles of C+P. V is taken orally once a day on Monday, Wednesday, and Friday during CRT and consolidative C+P. There are 3 dose levels (DLs): 100, 200, and 300 mg. All patients in a DL must complete CRT before accrual to the next DL. The primary end point is to determine the MTD. The protocol was amended to allow surgical resections after CRT if clinically feasible.

Results: To date, 8 patients (age, 51-77 years) with stage III nonsquamous NSCLC have been enrolled from May 2010 to October 2012. Four patients were enrolled at the first DL (V, 100 mg). Because the 3 evaluable patients did not experience a dose-limiting toxicity (DLT), enrollment proceeded onto DL 2 (V, 200 mg). At this DL, 1 patient had a DLT of grade 4 hyponatremia; thus, accrual is ongoing to a goal of 6 patients at DL 2. Common grade 1 or 2 toxicities included nausea, anorexia, dysphagia, dehydration, esophagitis, fatigue, constipation, and pain. Grade 3 toxicities included nausea, hyperglycemia, anemia, and leukopenia. One patient had grade 4 toxicity of hyponatremia. One patient on DL 1 had grade 4 thrombocytopenia, but this patient was treated when platelets were less than the parameters specified on protocol, and thus was not evaluable for DLT assessment. Of the 5 evaluable patients who completed all 4 cycles of therapy, all are alive to date. One patient developed chronic lymphocytic leukemia approximately 18 months after completion of therapy.

Conclusions: Preliminary results do not demonstrate unexpected toxicities with the addition of V to C, P, and RT. Enrollment is ongoing to obtain more safety data. In general, the degree of esophagitis was mild. Full assessment of survival and correlative studies will be analyzed at study completion.

AB2013-2. Interim Results for Neoadjuvant Bevacizumab With Weekly Nanoparticle Albumin Bound (nab)-Paclitaxel Plus Carboplatin Followed by Doxorubicin Plus Cyclophosphamide (AC) in Triple-Negative Breast Cancer

Jasgit C. Sachdev,^a Jessica Snider,^a Lee Schwartzberg,^b Robyn Young,^c Furhan Yunus,^a Jeffrey Allen,^a Carmel Verrier,^a Yasir Javed,^a and Mohammad Jahanzeb^a

^aUniversity of Tennessee Cancer Institute; ^bThe West Clinic; and ^cThe Center for Cancer and Blood Disorders

This study was approved and funded by NCCN from general research support from Celgene Corporation.

Background: Triple-negative breast cancer (TNBC) predominantly clusters with “basal-like” subtype on genomic profiling. Overexpression of secreted protein acidic and rich in cysteine (SPARC) is observed in basal-like breast cancer (Charaffe-Jaufrett et al. Oncogene 2006;2273-2284). Endothelial transcytosis of nab-paclitaxel via albumin-gp60-caveolin 1 interaction and entrapment of albumin by SPARC results in high intratumoral levels of nanoparticle albumin-bound (nab)-paclitaxel (Desai et al. Translational Oncology 2009;59-63). Exploiting this mechanism and the dysfunctional DNA repair in basal tumors, this study hypothesizes that nab-paclitaxel and the DNA damaging drug carboplatin would produce high pathologic responses in TNBC. Bevacizumab may enhance efficacy by blocking angiogenesis. In TNBC, pathologic complete remission (pCR) correlates with better disease-free survival (vonMinckwitz et al. J Clin Oncol 2012).

Methods: Patients with operable TNBC 2 cm or greater are eligible for this single-stage phase II trial. pCR (absence of invasive tumor cells) in the breast and in the breast plus the nodes are primary and secondary end points, respectively. Fifty-seven evaluable patients will provide 80% power for a type I error rate of 5%, assuming a pCR rate of 25% versus 40% for the null and alternate hypotheses, respectively. Patients receive carboplatin at area under the curve (AUC) 6.0 on day 1 plus nab-paclitaxel, 100 mg/m², on days 1, 8, and 15 every 28 days for 4 cycles, followed by doxorubicin and cyclophosphamide (AC) every 14 days for 4 cycles preoperatively. Bevacizumab at 10 mg/kg every 14 days is given with the first 6 cycles of preoperative chemotherapy and continued postoperatively to complete 1 year of treatment.

Results: Four patients came off study before surgery and are not in the efficacy population. Thirty patients are evaluable per protocol definition for the primary end point. Grade 3/4 hematologic adverse events (AEs) included neutropenia (grade 3: 69%; grade 4: 29%), febrile neutropenia (8%), thrombocytopenia (37%), and anemia (25%). Seven patients experienced treatment-related serious AEs, with 1 grade 5 event. Of the evaluable patients, 80% and 7% achieved clinical complete and partial responses, respectively. pCR rates are 55% (17/31) and 56% (17/30) in the breast and 52% (16/31) and 53% (16/30) in the breast plus nodes in the efficacy and evaluable populations, respectively. See Table 1.

AB2013-2. Table 1: Patient Characteristics (N=35)

View Table

Conclusions: Interim results show high pCR rates compared with those reported with anthracyclines/taxanes in TNBC. Pretreatment biopsies will be analyzed for predictive markers. An ongoing randomized study is evaluating the individual contributions of carboplatin and bevacizumab to pCR in TNBC.

Quality Improvement

AB2013-3. Medical Oncologists’ Assessment of Resectability in Patients With Liver-Limited Metastatic Colorectal Cancer

Michael A. Choti,^a Mark Green,^b Schiffon L. Wong,^c Kenzo Hirose,^a Matthew J. Weiss,^a Timothy M. Pawlik,^a Michael Eaddy,^b and Michele Thomas^b

^aJohns Hopkins University; ^bXcenda LLC; and ^cBristol-Myers Squibb

Background: Liver resection is an important strategy for improving survival in patients with liver-limited metastatic colorectal cancer (LL-mCRC). Understanding medical oncologists’ (MOs) views of potential candidates for surgical therapy can impact surgical referral patterns and the use of curative-intent therapy. This study explored MOs’ assessment of the technical aspects of resectability by using a standard clinical scenario to control for oncologic variability, while varying the anatomic parameters of metastatic size, number, and distribution of LL-mCRC. MOs’ resectability designations were compared with the designations of an expert surgical panel.

Methods: A total of 190 MOs practicing in the United States were surveyed using a Web-based tool. They were in practice for an average of 13.7 years (range, 5-35 years), with more than 90% of their time in direct patient care, and treated an average of 15.5 patients with colorectal cancer per month. MOs were queried regarding a single patient profile with 10 different sets of deidentified liver CT images representing a spectrum of extent LL-mCRC. On reviewing the images and a summary radiologic report, MOs were asked to assess resectability based on their expectation that all gross disease could be resected (R0). MOs designated each case as currently resectable, potentially convertible with chemotherapy response, or unresectable. MOs were also asked whether they believed surgical consultation was warranted. Through a modified Delphi approach, a panel of 3 expert hepatic surgical oncologists provided consensus regarding resectability. The congruency of the MOs’ and expert panel’s designations were compared.

Results: Among the 10 imaging scenarios, the expert panel designated 8 cases to be resectable, 1 case as potentially convertible, and 1 case as unresectable. The rank order of the MOs’ designations of resectability were aligned with the overall resectability rankings of the expert surgical panel. However, when comparing the designations of resectability status, MOs were congruent with the expert surgeons in 37% of cases. In those cases specifically designated as resectable by the expert panel, MOs concurred in 34% of cases. MOs would have recommended surgical consultation in 50% of the panel-designated resectable cases. The discordance of designations between the expert surgical panel and the MOs increased as the number (P<.0001), size (P=.002), and bilaterality (P<.0001) increased.

Conclusions: Although MOs in this study ranked their impression of the likelihood of resectability similar to that of surgical experts, they were less likely than the expert panel to consider a patient resectable and referred half of all potentially resectable cases for surgical consult. Future research should investigate reasons for discordance in assessing patient respectability and the ability of clear guidelines for surgical referral and multidisciplinary evaluation teams to influence outcomes in patients with LL-mCRC.

AB2013-4. Nurses’ Confidence Regarding Making Change, Facilitating Referral, and Eliciting Patient Questions

Susan T. Vadaparampil, Patricia Mullan, Nicole Hutchins, Richard Reich, and Gwendolyn Quinn

Moffitt Cancer Center

Background: Needs assessment is an important step in developing effective, relevant training opportunities for nursing professionals. These assessments provide an opportunity not only to document key knowledge gaps but also to explore attitudinal factors that may ultimately influence how and whether the information learned during training is integrated into clinical practice. This study explores 3 such factors: perceptions of ability to 1) make practice changes, 2) refer patients for additional care, and 3) elicit questions from patients with cancer that were collected in the context of a larger survey, to understand the needs of oncology nurses regarding communication with young patients with cancer about reproductive health.

Methods: As part of a 30-item Web-based survey administered to a select randomized, stratified list of members of the Oncology Nursing Society listserv (n=2000), 3 questions focused on nurses’ perceptions about ability to make change, refer patients, and elicit patient questions using a Likert-type response format (with scores ranging from 1, representing strongly disagree, to 4, representing strongly agree). Descriptive statistics were used to examine these responses according to nurses’ educational level.

Results: Of the 109 nurses who completed the survey, 31.0% were non-baccalaureate-prepared, 40.7% were baccalaureate-prepared, and 24.8% were graduate prepared. Approximately one-quarter (26.9%) of non-baccalaureate prepared nurses and 3.4% of graduate-prepared nurses did not feel confident in their ability to make changes in their practice setting. A small subset (6.7%) of graduate nurses and a larger proportion of baccalaureate-prepared (12.5%) and non-baccalaureate-prepared (30.8%) nurses expressed a lack of confidence in their ability to facilitate patient referral for additional clinical care. Most nurses (>84%) across all educational groups expressed confidence in their ability to elicit patient questions after a cancer diagnosis.

Conclusions: Results from the current study suggest that nurse training programs may be improved by developing targeted approaches based on nurse educational level to enhance nurses’ perceived abilities and confidence.

AB2013-5. CancerSupportSource: Validating a Web-Based Distress Screening Tool in the Community

Joanne Buzaglo, Melissa F. Miller, Christopher Gayer, Anne Morris, and Mitch Golant

Cancer Support Community Research and Training Institute

Background: The Institute of Medicine (IOM), NCCN, and the American College of Surgeons have all recognized that screening for psychosocial concerns is critical to ensuring quality cancer care for the whole patient. Few tools have been validated in the community in which up to 85% of patients with cancer are treated. CancerSupportSource (CSS) is a 25-item Web-based tool designed to screen patients for distress in the community and connect them to appropriate resources. CSS asks patients to rate their concerns across 25 items using a 5-point scale (ranging from 0, indicating not at all, to 4, indicating very seriously concerned). For each item, patients indicate whether they want to 1) talk with a health care team member; 2) receive print information; or 3) access online resources. The study objective was to validate psychometric properties of CSS.

Methods: A total of 239 members (90% female; median age, 57 years) of a community-based cancer support organization completed a Web-based Distress Thermometer (DT; a 0-10 visual analogue scale in the form of a thermometer) and CSS. A summary score (range, 0-25) was calculated as the total count of CSS items rated 2 or greater (somewhat to very seriously concerned). Receiver operating characteristic (ROC) curve analysis using the DT (≥4) as the criterion was used to explore the effect on sensitivity and specificity of the sum count of items rated greater than 2 (area under curve, 0.80).

Results: The median number of CSS items rated 2 or greater was 8. Forty-two percent of participants indicated a level of distress 4 or greater on the DT. The summary count of items rated greater than 2 (somewhat to very seriously concerned) was moderately correlated with the DT (R₂=0.38; P<.001). Individual items most strongly correlated with the DT included changes or disruptions in work, school, or home life (R₂=0.32; P<.001); feeling sad or depressed (R₂=0.29; P<.001); coping with one’s feelings (R₂=0.27; P<.001); worrying about the future and what lies ahead (R₂=0.25; P<.001); and feeling too tired to do the things one needs or wants to do (R₂=0.25; P<.001). Among those who rated their distress as 4 or greater on the DT (n=101), 91% also reported they were somewhat to very seriously concerned about 5 or more of the 25 CSS screening items (sensitivity, 91%; specificity, 49%).

Conclusions: CSS shows strong psychometric properties that can help screen for patients at high risk for distress and depression. Rating 5 or more of the CSS items as 2 or greater can correctly identify most (91%) patients who warrant further assessment and intervention for depression. Addressing the needs of patients at high risk for depression through appropriate screening, assessment, and intervention can lead to better patient-reported and cost outcomes.

Outcomes and Health Services Research

AB2013-6. Epidemiologic Modeling Estimating the Number of Patients With Relapsed Hodgkin Lymphoma After Autologous Stem Cell Transplant in 5 European Countries

Adeline Abbé,^a Andrew Lee,^a Alaa Hamed,^a Frank Neumann,^b Robert Olivares,^a and Andreas Engert^c

^aSanofi, Global Evidence and Value Development and ^bClinical Development; and ^cFirst Department of Internal Medicine, University Hospital of Cologne

This study was funded by sanofi-aventis.

Background: Hodgkin lymphoma (HL) is a lymphoid malignancy characterized by the presence of Reed-Sternberg cells in lymphoma tissues. High-dose chemotherapy followed by autologous stem cell transplant (ASCT) is the standard treatment for patients with relapsed/refractory HL. ASCT rates vary among European countries. The primary objective of this study was to estimate the percentage and number of patients with HL who experienced relapse after ASCT in France, Germany, Italy, Spain, and the United Kingdom.

Methods: The estimated number of patients with HL who experienced relapse after ASCT was based on the number of ASCTs performed for these patients and the relapse rate (RR) reported in the literature. The ASCT rate in patients with HL was extracted from the EU Group for Blood and Marrow Transplantation¹ registry (1990-2010) and extrapolated to the 5 European Union (EU) countries. A literature review using EMBASE identified the RR after ASCT for HL. An epidemiologic model was built by identifying the RR at 12, 24, 36, 48, and 60 months. If no RR was available at a time point, it was estimated by dividing the following data point by the period length. The difference in RR (eg, between 12 and 24 months) was used to estimate the RR for the second year. The same methods were used to calculate third-, fourth-, and fifth-year rates.

Results: The number of patients who underwent ASCT for HL in the 5 EU countries in 2012 was estimated to be 1098, ranging from 153 in Germany to 340 patients in Italy. Five studies providing post-ASCT RRs were identified. The estimated RR per year is presented in Table 1, showing that the estimated number of patients with relapsed HL post-ASCT in the 5 countries in 2012 was 200 patients. Over a 5-year period, the estimated cumulative RRs exhibit a large range, from 12% in France, the United Kingdom, and Germany, to 26% in Italy and Spain.

AB2013-6. Table 1: Estimated Number of Patients With HL Who Relapsed Post-ASCT in 2012 for 5 EU Countries

View Table

Conclusions: Using available data, a model was built to estimate the number of patients with HL experiencing relapse post-ASCT in 5 EU countries. Model estimates may be limited by the assumptions, but they are likely to be conservative and actual rate may be higher. Large variations in RR across countries in Europe are very likely because of the choice of first-line treatment, and suggest an opportunity for future studies and better understanding of HL, treatment options, and practices.

Best Practices in Implementation and Use of Clinical Practice Guidelines

AB2013-7. NCCN Symptom Guidelines Coupled With Nurse Practitioner Follow-Up Reduce Moderate to Severe Symptom Days by Half or Greater in Cancer Patients Receiving Outpatient Chemotherapy

William A. Dunson Jr,^a Kathi Mooney,^a,b Susan L. Beck,^a,b Bob Wong,^b and Debra Wujcik^c

^aHuntsman Cancer Institute; ^bCollege of Nursing, University of Utah Salt Lake City; and ^cVanderbilt University

Background: Supportive care guidelines provide direction for cancer symptom care during chemotherapy. Studies of the use of these guidelines and their impact on symptom outcomes are needed. The authors used the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Adult Cancer Pain, Cancer-Related Fatigue, Antiemesis, and Distress Management (depressed mood and anxiety) to guide nurse practitioner (NP) follow-up of moderate to severe symptoms of chemotherapy outpatients as part of a larger, NCI-funded study testing an automated remote symptom monitoring and treatment intensification system.

Methods: In this study, 358 enrolled participants were randomly assigned to either treatment (TC; n=180) or usual care (UC; n=178). Daily, both groups called and reported on presence and severity of symptoms and level of distress using an automated computer-based telephone remote monitoring system based on a scale ranging from 0 to 10. In the TC group, symptoms reported at 5 or greater generated an automated alert to a study NP, who responded by calling and using a case management system that integrated NCCN Guidelines to intensify symptom care. In addition, the TC group received automated, tailored symptom self-care messages during the daily calls.

Results: The average age of study participants was 55.8 years. Most were female (n=271, 75.7%), white/Caucasian (n=291, 83.0%), and had breast cancer (n=156, 43.6%) or lung cancer (n=61, 17.0%). No differences were seen between groups regarding all demographic or clinical variables. Of the 5 symptoms supported by NCCN Guidelines (fatigue, pain, nausea, mood, and anxiety), fatigue was the most prevalent symptom, with 83% of participants reporting moderate/severe-level days followed by pain, nausea, depressed mood, and anxiety (79%, 60%, 52%, and 49%, respectively). The average number of days on study was 83 and call adherence was 89.59% of expected call days, with no difference between TC and UC. Negative binomial regression was used to investigate treatment effects on the amount of moderate/severe days reported. Results showed that the TC group had fewer moderate/severe days than the UC group for all NCCN Guideline symptoms (estimated mean days reported for TC vs. UC, respectively: fatigue, 11.44 vs. 25.18; pain, 6.94 vs. 19.44; nausea, 2.63 vs. 8.31; depressed mood, 1.89 vs. 5.30; nervous/anxious, 1.31 vs. 3.92; all P values <.001).

Conclusions: Use of NCCN Guidelines for symptom management coupled with NP follow-up to intensify treatment for unrelieved symptoms at moderate to severe levels dramatically reduces these symptoms compared with usual care.

2013 Abstract Disclosure Index

View Table