Part 2: Abstracts From the NCCN 18th Annual Conference: Advancing the Standard of Cancer Care™

The following abstracts were accepted for presentation at the NCCN 18th Annual Conference: Advancing the Standard of Cancer Care™ inaugural General Poster Session at the Westin Diplomat in Hollywood, Florida, on March 14–15, 2013. NCCN considered original abstracts from investigators in the oncology community related to Best Practices in Implementation and Use of Clinical Practice Guidelines, Quality Improvement, Outcomes and Health Services Research and NCCN Oncology Research Program (ORP)-funded Projects. This is the second in a 2-part series highlighting the abstracts presented at the conference. The first installment appeared in the March issue of JNCCN.

Quality Improvement

AB2013-8. Patient-Reported Outcomes Measurement Information System (PROMIS) Global Health Instrument as a Distress Screen in Routine Cancer Care.

Kelly Bugos, Oxana Palesh, Kavitha Ramchandran, Kathryn Gail Bailey, Blake Herring, Julie Kuznetsov, and Douglas W. Blayney

Stanford Cancer Institute, Stanford Women's Cancer Center

Background: Stanford Cancer Institute's goal to integrate NCCN standards of care for distress management into routine practice started with a multidisciplinary distress screen committee. The committee selected the PROMIS Global Health (GH) patient-reported outcomes measurement information system to recognize, monitor, document and address psychosocial health at all stages of cancer illness. PROMIS GH was tested in the Stanford Women's Cancer Center, a center for women with breast and gynecologic cancer.

Method: PROMIS GH, developed and funded by the NIH, consists of 10 items to measure quality of life and mental and physical health on a Likert-type scale. The questionnaire consisted of this instrument along with the question “Would you like help with these issues?” The paper form was added to the routine check-in process during September 2012. A medical assistant screened all responses. Any response of “fair to poor” or “often to always” was assessed by a nurse coordinator and discussed with the medical team, and relevant psychosocial service referrals, including palliative care, were offered.

Results: A total of 1350 patients seen in 19 clinic days received the questionnaire. A convenience sample of 304 questionnaires showed a high acceptability rate: 96% were fully complete. Patient ratings of mental health, including mood and ability to think, were “excellent” (18%), “very good” (26%), “good” (33%), “fair” (20%), and “poor” (3%). Patient ratings of emotional problems, such as feeling anxious, depressed, or irritable, were “never” (11%), “rarely” (29%), “sometimes” (39%), “often” (17%), and “always” (3%). These requests were for help (12%), among which were for management of distress (30%), pain (30%), physical issues (16%), sleep disturbance (8%), fatigue (5%), cognitive changes (3%), and a general request for help (8%). Clinic staff said the questionnaire was valuable in “taking better care of the patient.”

Conclusions: PROMIS GH was well accepted by patients and staff as a screen for emotional distress during this month-long pilot project. Including the questionnaire in the check-in process enabled universal administration. The authors were successful in routinely screening for distress in female patients with cancer using a validated instrument in busy clinics with existing personnel. Future plans include adding a spiritual question, adding the questionnaire to the electronic medical record, and expanding its use to the entire Stanford Cancer Institute.

AB2013-9. Meeting Patient-Centered Standards: CancerSupportSource—A Community-Based Distress Screening Program

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

Cancer Support Community Research and Training Institute

Background: The Cancer Support Community (CSC) has developed CancerSupportSource (CSS), an evidence-based, psychosocial distress screening, referral, and follow-up program designed to meet screening standards established by the American College of Surgeons Commission on Cancer. CSS was designed to target 7 key areas of psychosocial needs identified in the IOM Report (2007) and integrates a 25-item self-report measure with automated links to information and referral for support services. CSC represents a network of more than 50 nonprofit affiliates dedicated to providing support and education to people affected by cancer. Although distress screening is currently being integrated into care delivery, little research has been performed to examine the impact of screening, referral, and follow-up on patient-reported outcomes. The objective of the present study was to demonstrate the feasibility and effectiveness of implementing CSS in 2 CSC sites. Screening and data collection are ongoing.

Methods: Patients are asked to complete CSS at baseline and are rescreened at 30 to 45 days. Participants rate their current concerns and identify specific types of assistance they desire for each concern (eg, talk to a staff member, online resources, print information). CSS is Web-based and can be completed in the clinic or at home. Reports are generated in real time for both patients and clinicians, allowing for efficient follow-up and referral. A Personal Support Care Plan based on the patient's responses to CSS is printed and/or given to the patient. For the health care team, a Distress Management Report for each patient is automatically generated.

Results: Twenty-three patients (100% female; 83% white; 33% income <$40K; 65% breast cancer; 9% colorectal) provided Internal Review Board–approved consent for data to be accessed for research. The top 4 problems rated as moderate to very serious concerns at baseline and rescreening, respectively, were eating and nutrition (65% and 74%), sleep problems (48% and 44%), worrying about the future (61% and 35%), and ability to exercise and be physically active (48% and 35%). Feeling nervous or afraid was a top concern at baseline (48%), but the proportion who endorsed it as a top concern significantly decreased between baseline and rescreening from 48% to 17% (P=.016). A significant decrease was seen in the sum of the 25 problem ratings (mean ± standard error), from 50.6 ± 2.9 to 45.6 ± 2.9 (P=.01), and in the summary count of items rated moderate or very serious concerns, from 7.3 ± 1.0 to 5.5 ± 1.0 (P=.03).

Conclusions: Preliminary findings suggest automated distress screening can lead to positive patient-centered outcomes, including enhanced communication with the oncologist and improved access to psychosocial services within a community-based setting.

AB2013-10. Utilization of Advanced Practice Providers Caring for Patients Undergoing Blood and Marrow Transplantation

Marc Matsui, F. Marc Stewart, and William Levy

Fred Hutchinson Cancer Research Center

Background: Fred Hutchinson Cancer Research Center (FHCRC), in collaboration with Seattle Cancer Care Alliance (SCCA), University of Washington Medical Center (UWMC), and Seattle Children's Hospital (SCH), performed 485 blood/marrow transplants (BMT) from July, 2011 through June, 2012. A total of 422 adult patients and 63 pediatric patients received a transplant. Recent Center for International Blood & Marrow Transplant Research analyses show that survival for patients treated at FHCRC consistently exceed expected 1-year rates. Outpatients were cared for at SCCA with an average daily census of 134.2 and 11.8 new patients per week. Adult inpatients were cared for at UWMC with an average daily census of 27.8 and total inpatient days of 10,185. Pediatric inpatients were cared for at SCH with an average daily census of 14.1 and total inpatient days of 5234. This study reviews utilization patterns of BMT-trained advanced practice providers (APP) consisting of physician assistants and nurse practitioners during this time frame.

Methods: Patients undergoing BMT received day-time care, defined as 8:00 am to 5:00 pm, and overnight care, defined as 5:00 pm to 8:00 am. Attending physicians supervised provider teams consisting of staff APPs or rotating physicians in training. APP utilization data was converted to daily equivalents and then weighted for weekdays and weekends to calculate a daily rate.

Results: Daytime adult inpatient care consisted of teams totaling of 2 attendings daily, with 6 providers on weekdays or 4 providers on weekends. APPs covered 75% of these shifts. Daytime weekday adult outpatient care consisted of teams totaling 5 attendings with 12 providers daily. Weekend care consisted of 2 providers. APPs covered 71% of these shifts. Daytime pediatric inpatient care consisted of 1 attending daily, with 3 providers on weekdays or 2 providers on weekends. APPs covered 80% of these shifts. Daytime pediatric outpatient care consisted of 1 attending and 2 providers daily. APPs covered 88% of the shifts. Weekend care was provided on the inpatient unit. In total, APPs covered 74% of the daytime shifts for all patients. Overnight adult inpatient care consisted of 1 physician or APP hospitalist. APPs covered 67% of the overnight shifts. Overnight pediatric inpatient care was provided by 1 physician hospitalist 100% of the time. Attendings were generally not present on the unit.

Conclusions: Under the supervision of physician faculty, BMT-trained APPs provide a significant amount of high-quality care for patients undergoing BMT at FHCRC.

AB2013-10. Table 1:

Average Daily Advanced Practice Providers Coverage, July 2011–June 2012

AB2013-10. Table 1:

Outcomes and Health Services Research

AB2013-11. The Economic Value of a Cancer Case Management Program

Charlotte Wu, Wade Bannister, Hungching Chan, Arthur Rossof, Pamela Schumacker, Michael Rosen, and Kelly Blair

OptumHealth, UnitedHealth Group

Background: The Cancer Support Program (CSP) of Optum provides patients and their families and caregivers with telephonic case management (education, support, referrals) by experienced oncology nurses and social workers, supported by a staff of appropriately trained physicians. These services are intended to help patients better comply with treatments and more effectively manage symptoms. As a result, patients are less likely to have cancer-related complications and will experience a better quality of life. Since the inception of the program, patients have reported extremely high satisfaction with the program. The intent of this study is to further estimate the financial impact of the CSP.

Methods: This case control study compared participants and nonparticipants of the Optum CSP. Propensity score weighting was applied to study subjects to adjust for baseline differences between participant and nonparticipant groups. The study population (N=8833) comprises patients receiving qualified referrals from July 2009 through June 2011. Participants are those actively participating in the CSP afterwards. The 6 months before referral was used as the baseline period to construct the propensity score model. To account for potential mortality bias, separate models were developed for survivors and decedents; for survivors (N=7445) the study period was the 6 months after program qualification, and for decedents (N=1388) it was the time after program qualification through the date of death. The primary outcome is the cost difference between participants and nonparticipants. Hospice use during end of life for decedents was also examined. Outcomes were estimated using generalized linear models.


  • Cancer-related cost savings: for program participants who survived through at least 6 months after the study period and under active treatment, the cancer-related cost per patient per month during the 6-month study period was reduced by 9.8% (95% CI, 3.7%, 15.6%; P=.002), or $1128. The decedent's cost during the last 3 months of the life was reduced by 19.9% (95% CI, 2.8%, 34.1%; P=.025), or $3999.

  • End-of-life hospice use: average hospice days among all deceased CSP participants increased by 42.3% (95% CI, 13.9%, 77.8%; P=.002), or 0.56 days.

Conclusions: The Optum CSP is effective in reducing cancer-related health care costs and promoting the use of hospice in patients with terminal cancer by extending hospice length of stay.

AB2013-12. Evidence-Based Diagnosis and Treatment Patterns of Colon Cancer in Canada

Julia Klein-Geltink, Rami Rahal, Tonia Forte, Carolyn Sandoval, Gina Lockwood, Heather Bryant, in collaboration with the System Performance Steering Committee and Technical Working Group

Canadian Partnership Against Cancer

Background: More than 9000 people in Canada die from colorectal cancer each year, and approximately 70% of those cases involve tumors of the colon. Clinical practice guidelines recommend a minimum of 12 nodes be removed to reduce the chance of understaging. For patients with node-positive disease (stage III), delivery of chemotherapy after resection has been shown to improve outcomes. Measuring the concordance of pan-Canadian treatment patterns with these 2 practice guidelines can help inform opportunities for quality improvements.

Methods: The System Performance initiative at the Canadian Partnership Against Cancer (CPAC) collaborated with provincial cancer agencies and programs to collect population level administrative data (registry and treatment) from 8 provincial cancer registries to examine the percentage of all colon resections with 12 or more nodes removed and examined in each of 2007, 2008, and 2009. Administrative data from 5 provincial cancer registries was used to examine the percentage of all patients with stage III colon cancer diagnosed in each of 2007, 2008, and 2009 who received postoperative chemotherapy. Results were examined by province, year, sex, and patient age at diagnosis.

Results: Substantial interprovincial variation was seen in the percentage of colon resections with 12 or more nodes removed and examined in 2009—from 59% to 89%—with little variation by age or sex. Concordance rates increased from 2007 to 2009 in 5 of the 7 provinces submitting more than 1 year of data for this indicator. Interprovincial variation was seen in the percentage of resected stage III colon cancer cases treated with adjuvant chemotherapy in 2009, ranging from 56% to 82%. Adjuvant chemotherapy rates seemed to decrease from 2007 to 2009 in 4 of the 5 provinces that submitted more than 1 year of data for this indicator. The adjuvant chemotherapy rate in 2009 was 90% for patients younger than 60 years of age compared with 20% for patients aged 80 and older. Differences were seen by sex for this indicator, but only among patients aged 70 years and older, with a treatment rate of 38% for women and 46% for men.

Conclusions: The analysis suggests that interprovincial and demographic variations exist in the evidence-based diagnosis and treatment of colon cancer. Although no national performance targets or benchmarks are currently available for either of these indicators in Canada, efforts are underway within the initiative to develop targets for several System Performance indicators, including these. Measuring these indicators against targets will inform future quality improvement efforts.

AB2013-13. Risk Factors for Chemotherapy-Induced Nausea and Vomiting in US Community Oncology Practice After Moderately or Highly Emetogenic Chemotherapy: INSPIRE Study

Nancy W. Peacock, James W. Gilmore, Anna Gu, Stephen Szabo, Melissa Rammage, Joyce Sharpe, Sally Haislip, Toni S. Perry, Tim Lee Boozan, Katherine Meador, Xiting Cao, Thomas A. Burke

Tennessee Oncology

Background: Electronic health records (EHRs) are a valuable source to evaluate the quality of oncology care, particularly when combined with patient outcomes data. The primary objective of this study was to evaluate the impact of NCCN Guideline–consistent/inconsistent chemotherapy prophylaxis (GCCP/GICP) on the incidence of no chemotherapy-induced nausea and vomiting (No CINV) after cycle 1 of highly emetogenic chemotherapy (HEC) or moderately emetogenic chemotherapy (MEC). An exploratory objective was to investigate factors associated with No CINV.

Methods: INSPIRE (Impact of NCCN Antiemesis Guideline uSage on PatIent Reported Emesis) was a prospective, observational, multicenter study that enrolled chemotherapy-naïve adults initiating single-day HEC or MEC in cycle 1. Results from the Multinational Association of Supportive Care in Cancer (MASCC) Antiemesis Tool, administered 5 to 8 days after HEC/MEC, were merged with EHR data. A logistic regression analysis was used to evaluate the effect of guideline consistency and other potential risk factors (ie, age, gender, cancer site, chemotherapy, secondary antiemetics, calendar month, and interaction terms) on No CINV (ie, no emesis and no clinically significant nausea). Variables were selected for inclusion in the model using stepwise selection, with a P value threshold of less than 0.10 for inclusion and less than 0.20 to stay in the model.

Results: A total of 1295 patients (mean age, 59.3 years; 30.0% male; 35.5% HEC) were enrolled from 4 large community oncology practices. The prevalence of GCCP was 57.3% (28.7% HEC; 73.1% MEC). GCCP-treated patients were older (P=.010), more likely to be receiving HEC (P<.0001), and had a different distribution of primary cancer sites (P<.0001) and practice sites (P<.0001) than GICP-treated patients. The percentage with No CINV was significantly higher for patients receiving GCCP (53.4% vs. 43.8%; P=.0006). GCCP, non–cisplatin-containing regimens, underweight, younger age, and male gender were associated with No CINV (Table 1).

Conclusions: Antiemesis guideline adherence, non–cisplatin-containing regimens, underweight, younger age, and male gender were independently associated with No CINV. Adherence to NCCN Clinical Practice Guidelines in Oncology for Antiemesis is an important potential quality-of-care measure.

AB2013-13. Table 1:

Risk Factors for No Chemotherapy-Induced Nausea and Vomiting

AB2013-13. Table 1:

NCCN Oncology Research Program (ORP)-Funded Projects

AB2013-14. A Phase I Trial of Vorinostat in Combination With Standard Chemoradiation Therapy in Advanced Oropharyngeal Squamous Cell Carcinoma: Preliminary Results

Theodoros N. Teknos, Matthew Old, Enver Ozer, Amit Agrawal, Ricardo Carrau, Panos Savvides, Bo Chao, Bhavna Kumar, Jharna Datta, Quintin Pan, John Grecula, and Robert Baiocchi

The Ohio State University Comprehensive Cancer Center - James Cancer Hospital and Solove Research Institute

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

Background: Because of the human papillomavirus (HPV), the incidence of oropharyngeal squamous cell carcinoma (OPSCC) is increasing at an alarming rate. Recent investigations have shown that concurrent chemoradiation therapy (CRT) is a highly successful treatment approach for HPV-positive OPSCC. However, the cure rates for HPV-negative tumors remain poor, and CRT regimens are highly toxic and immunosuppressive. As a result, novel therapies are required to minimize toxicity, modulate immunosuppression, and further improve outcomes. Vorinostat, with its ability to regulate cancer gene transcription, induce tumor apoptosis, protect normal mucosa, and modulate tumor immunity, may provide great synergy with current CRT regimens.

Methods: The primary objective of this phase I study is to identify the maximum tolerated dose of vorinostat in combination with standard, concurrent CRT in stage III/IV unresectable OPSCC. Secondary objectives include 1) identifying the effects of vorinostat on adaptive and innate immunity, and 2) identifying changes in the microRNA (miR) expression patterns. To date, 11 patients with stage IVa OPSCC have been enrolled in this trial and 9 have completed therapy. The schema involves a 1-week run-in of vorinostat followed by standard cisplatin/intensity-modulated radiotherapy, in combination with vorinostat. Blood is drawn at baseline, after vorinostat run-in, at study conclusion, and 3, 6, and 12 months after completion to determine laboratory correlates of adaptive and innate immunity.

Results: Five patients were enrolled at dose level 1 (100 mg, qdx3, weekly), and 2 developed dose-limiting toxicities (DLTs). These DLTs included deep venous thrombosis/pulmonary embolism and grade IV febrile neutropenia in patient 1, and grade III febrile neutropenia and Clostridium difficile colitis in patient 5. Six subsequent patients have been enrolled at dose-1 (100 mg, qodx3, weekly). One patient in this second cohort experienced the DLT of a grade II decrease in absolute neutrophil count, which did not respond to a 2-week vorinostat treatment break. All other patients on the trial (n=6) have completed the regimen, and 2 are currently undergoing therapy and tolerating the regimen well. No significant adverse events were reported in 5 patients, and 1 patient had a grade III toxicity (dehydration and nausea, dose 1). Notably, mucositis scores have been exceptionally low (7/9 grade I; 2/9 grade II). All patients remain disease-free and were gastronomy tube–free at last follow-up (median follow-up, 15 months; range, 4 weeks to 26 months).

Conclusions: The preliminary findings of this phase I study combining vorinostat with CRT in the treatment of OPSCC are encouraging. Vorinostat has illustrated promising antitumor effects both in vivo and in vitro, with a surprisingly good toxicity profile, especially with regard to mucositis.

AB2013-15. Pralatrexate (P) Plus Oxaliplatin (O) in Advanced Esophagogastric Cancer: Preliminary Report of the Safety Lead-in Cohort in a Phase II Trial

Usha Malhotra, Austin Miller, Araba Adjei, Sai Yendamuri, Peter Loud, and Nikhil I. Khushalani

Roswell Park Cancer Institute

This trial was approved and funded by NCCN from general research support provided by Allos, Inc.

Background: Advanced esophagogastric cancer (ECG) is an aggressive cancer with a poor prognosis. Given the plateau in the efficacy of existing chemotherapy regimens, novel agents merit investigation. In preclinical studies, the antifolate pralatrexate (P) has demonstrated increased cellular uptake and polyglutamylation compared with other antifolates, and has shown synergy with platinum. This trial investigates the safety and efficacy of the combination of P and oxaliplatin (O) in advanced EGC, along with pharmacogenomics (polymorphisms in genes of the folate pathway) and pretherapy tissue microRNA (miRNA) profiling to identify signatures predictive of toxicity and response.

Methods: Therapy-naïve patients with EGC with metastatic/unresectable disease, adequate organ function and performance status, and less than grade 2 peripheral neuropathy are eligible. Treatment consists of biweekly O (85 mg/m2) in combination with P (dose 1, 120 mg/m2; dose-1, 100 mg/m2). Patients also receive folate and vitamin B12 commensurate with P labeling instructions. Because of lack of phase I data for this combination, a safety lead-in cohort of 6 patients was planned at the starting dose, with provisions for dose de-escalation for dose-limiting toxicity (DLT). Using an exact 2-stage design, 17 patients will be enrolled in stage 1; if more than 4 patients experience a RECIST-defined response, 16 more patients will be enrolled. The primary end point is response rate. Secondary clinical end points include toxicity and survival. Molecular correlates are as defined earlier.

Results: Thirteen patients have been enrolled. Two of 4 patients at dose 1 developed DLTs (1 grade 3 diarrhea, 1 grade 3 mucositis), prompting a dose-level decrease to be investigated as planned. Seven patients have been treated at dose-1, for a total of 11 patients in the safety cohort. This includes 1 patient who was replaced because of the need for palliative radiation. No DLTs were identified at dose -1. Median age of this cohort is 62 years (range, 39–83 years), most are male (n=9), and most had adenocarcinoma (n=10). Grade 4 toxicities include neutropenia (n=1) and myocardial infarction (n=1). The most common grade 3 events have been mucositis (n=3), diarrhea (n=2), nausea (n=2), anemia (n=2), aspartate aminotransferase increase (n=1), dehydration (n=1), hypersensitivity (n=1), neuropathy (n=1), and aspiration pneumonia (n=1). Including grade 1 and 2 events, mucositis and nausea have been the most common toxicities seen so far (9 events each), which would be expected.

Conclusions: The safety lead-in cohort in this phase II trial has established the dose of P at 100 mg/m2 to be tolerable in combination with O at 85 mg/m2 every 2 weeks, and all future patients will be enrolled at this dose level. The planned interim analysis will determine the preliminary efficacy of this novel combination in advanced EGC.

Best Practices in Implementation and Use of Clinical Practice Guidelines

AB2013-16. Results of NCCN Guideline–Compliant Care for Management of Fibroepithelial Lesions of the Breast

Andrew Van Osdol, Jeffrey Landercasper, Jeremiah Andersen, Richard Ellis, Erin Gensch, Jeanne Johnson, Brooke De Maiffe, Kristen Marcou, Mohammed Al-Hamadani, and Choua Vang

Norma J. Vinger Center for Breast Care, Gundersen Health System

Background: Subtle histopathologic differences between fibroadenoma (FA) and phyllodes tumors of the breast sometimes make these fibroepithelial lesions (FELs) difficult to differentiate on percutaneous core needle biopsy (CNB). NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) recommend selective, not routine, surgical excision of FELs. The authors hypothesized the NCCN Guidelines on selective surgical excision of FELs on CNB would result in an acceptable low rate of missed phyllodes.

Methods: A retrospective chart review of a breast center prospective patient data registry for patients with ambiguous FEL on CNB from 2001 to 2006 was performed. The study date ended in 2006 to allow a 5-year follow-up for missed lesions. Patients with definitive phyllodes on CNB were excluded. Care pathways for FELs of the breast were compliant with the NCCN Guidelines for Breast Cancer Screening and Diagnosis. The following comparison variables were studied to determine 1) why providers selected certain lesions to undergo excision and 2) which excised lesions were at higher risk for upgrade to phyllodes: study year, patient age, detection method, lesion palpability, CNB type, CNB size, lesion size, BI-RADS assessment, radiologist, surgeon, pathologist, and pathologist “comment of concern.” Chi-Square, Fisher exact, t test, and Wilcoxon rank sum tests were used for data analysis.

Results: A total of 313 patients met inclusion criteria. Of these patients, 261 (83%) underwent observation only with a mean follow-up of 6.1 years, and of these patients, 3 (0.9%) had benign phyllodes at follow-up. Of the 313 total patients, 52 (17%) underwent surgical excision, with 18 (35%) lesions upgraded to phyllodes. Factors associated with excision instead of observation were younger age, palpability, clinical detection, smaller biopsy needle size, lesions with larger size, radiology provider, and pathologist “comment of concern” regarding the lesion (all P<.05). Factors associated with upgrade from FEL on CNB to phyllodes on excision were older patient age (P=.004) and pathologist “comment of concern” (P<.001). Lesion BI-RADS assessment category approached but did not achieve statistical significance for upgrade (P=.06).

Conclusions: Clinical, imaging, and pathologic factors affected the likelihood of excision, whereas patient age and pathologist opinion predicted upgrade to phyllodes. This finding underscores the importance of breast radiologist and pathologist concordance assessment and emphasizes the importance of an interdisciplinary concept of patient management. With a missed lesion rate of less than 1%, this study supports the recommendation in the NCCN Guidelines of selective, not routine, excision of ambiguous FELs, with scheduled follow-up for patients undergoing observation.


The NCCN Foundation, through private philanthropy and grants, advances the mission of NCCN to improve the quality and effectiveness of care to patients with cancer. The 2011 NCCN Foundation Young Investigator Awards were made possible through support from the following companies: Amgen Inc., Genentech, Inc., Millennium Pharmaceuticals, and Pfizer Oncology.

AB2013-17. Adherence to National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology Improves Survival in Veterans With Diffuse Large B-Cell Lymphoma

Kenneth Carson, Suhong Luo, Katiuscia O'Brian, and Graham Colditz

St. Louis VA Medical Center and Washington University School of Medicine

Background: Twelve years ago, the NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for diffuse large B-cell lymphoma (DLBCL) were developed to help ensure that patients receive care that will optimize clinical outcomes. Since then, they have been updated roughly on an annual basis. Although guidelines are based on evidence from clinical trials and expert opinion, scant evidence demonstrates that guideline adherence enhances survival in a real-world practice setting.

Methods: Using data from the Veterans Health Administration (VHA) central cancer registry, the authors assembled a cohort of 3227 patients diagnosed with DLBCL between October 1, 1998 and December 31, 2008. The cohort was limited to patients with systemic DLBCL diagnosed and treated within the VHA system after June 1, 2000, the release date for the first DLBCL guidelines. This resulted in an initial cohort of 1826 patients. Guideline adherence was based on first treatment administered (similar to intention to treat) compared with the treatment recommended by guidelines in effect at the time of DLBCL diagnosis. Overall survival was measured from diagnosis to death. Cox modeling was used to control for potential confounders in survival analyses.

Results: Baseline comparison showed that patients commencing NCCN Guideline–adherent therapy were significantly younger (65 vs. 82 years; P<.001) and had lower mean Charlson comorbidity scores (2.1 vs. 3.3; P<.001) than those undergoing noadherent therapy. Age-adjusted 3-year survival was 62% and 41% (P<.001) in the adherent and nonadherent groups, respectively. Cox analysis showed increased mortality associated with age (hazard ratio [HR], 1.03; 95% CI, 1.02–1.03), stage (HR, 1.62; 95% CI, 1.38–1.91), B symptoms (HR, 1.38; 95% CI, 1.12–1.60), comorbidities (HR, 1.08; 95% CI, 1.05–1.11), and lactate dehydrogenase level (HR, 2.02; 95% CI, 1.72–2.38). Use of granulocyte stimulating factors was associated with a trend toward reduced mortality (HR, 0.88; 95% CI, 0.76–1.03). After controlling for these factors, starting with guideline-adherent therapy was associated with reduced mortality (HR, 0.61; 95% CI, 0.51–0.72). Notably, only 847 of 1506 patients (56%) completed the appropriate number of treatment cycles and/or radiation to finish in a guideline-adherent fashion.

Conclusions: After controlling for potential confounders, starting with NCCN Guideline–adherent therapy was associated with significantly reduced mortality in DLBCL. This finding suggests that adherence could be used as a quality measure that will correlate with survival. A major limitation of this observational study may be unmeasured variables contributing to residual confounding. Additional work is underway to fine-tune these results and understand the factors associated with beginning and completing treatment in an adherent versus nonadherent fashion.

This work was funded by the NCCN Foundation. Any opinions, findings, and conclusions expressed in this material are those of the author(s) and do not necessarily reflect those of the National Comprehensive Cancer Network or the NCCN Foundation.

AB2013-18. Does Disease Severity Impact Distress? Comparison of Acute Leukemia and Breast Cancer Survivors

Joanne Lester, Robin Stout, Kara Crosthwaite, Joseph Flynn, Chris Holloman, and Barbara Andersen

The Ohio State University Comprehensive Cancer Center - James Cancer Hospital and Solove Research Institute

Background: Distress is a common response to challenges experienced during treatment and into early survivorship. Resolution of distress may positively impact patient outcomes and improve overall quality of life.

Methods: A multigroup cross-sectional design was used; the setting was a Midwestern comprehensive cancer center. Acute leukemia (ALL; n=100) and breast cancer (n=100) survivors completed self-report instruments at 1 of 4 time points: during treatment, at its end, or at 3 or 6 months after end of treatment. Instruments included the NCCN Distress Thermometer (0–10 analog scale) and NCCN distress symptom profile (38 items) with 5 subscales: practical, family, emotional, spiritual, and physical. Participants listed their top 3 sources of distress and bothersome symptoms.

Results: Breast cancer and ALL survivors had similar mean distress scores: 4.27 (standard deviation, 3.07) and 4.32 (standard deviation, 2.83), respectively. A statistical difference was noted at 6 months posttreatment (P<.00934), presumably as the patients with leukemia underwent bone marrow transplant. Symptom mean scores were similar in breast cancer and ALL groups, respectively, for the physical (4.41 vs. 6.53), emotional (2.17 vs. 2.62), practical (0.77 vs. 0.92), family (0.56 vs. 0.51), and spiritual subscales (0.03 vs. 0.77). Significant differences were noted in breast cancer survivors between time points and distress the areas of eating (P<.04), fatigue (P<.01), fevers (P<.02), getting around (P<.05), mouth sores (P<.05), congestion (P<.02), sleep (P<.02), fears (P=.05), nervousness (P=.04), physical appearance (P<.04), bathing, (P=.05), and work schedule (P<.012). Factors that were significantly associated included age (P<.004), perceived health status (P<.04), income before diagnosis (P<.05), caregivers working outside home (P<.04), living alone (P<.03), and employment status (P<.02).

Significant differences were noted in ALL survivors between time points and distress in the areas of treatment decisions (P<.04), fevers (P<.004), getting around (P<.03), mouth sores (P<.05), and sexuality (P<.02). Factors that were significantly associated included income before diagnosis (P<.04), level of education (P<.04), perceived health status (P<.05), treatment decisions (P<.04), and type of insurance (P<.02). The top 3 narrative sources of distress were the same in breast cancer and ALL survivors: 1) practical, 2) physical, and 3) family. The top most bothersome symptoms in breast cancer and ALL survivors, respectively, were 1) fatigue/fatigue, 2) pain/gastrointestinal, 3) emotional/emotional, and 4) integumentary/integumentary.

Conclusions: The cause of distress is multifocal and may be dependent on the disease process, although the severity of disease does not seem to have a significant effect on distress symptoms or scores. Survivors have common sources of distress that deserve practitioners' attention and evidence-based interventions.

This work was funded by the NCCN Foundation. Any opinions, findings, and conclusions expressed in this material are those of the author(s) and do not necessarily reflect those of the National Comprehensive Cancer Network or the NCCN Foundation.

AB2013-19. Geriatric Assessment and Telephone Intervention in Elderly Women with Ovarian Cancer: Preliminary Findings

William P. Tew, Heidi M. Yulico, Arti Hurria, Howard T. Thaler, Oana Orodel, Dennis S. Chi, Ginger J. Gardner, Jason Konner, Carol A. Aghajanian, Stuart Lichtman, and Beatriz Korc-Grodzicki

Memorial Sloan-Kettering Cancer Center

Background: More than half of all patients diagnosed with ovarian cancer are older than 65 years. Because of inadequate screening tools and nonspecific symptoms, patients present in advanced stages of disease, and curative treatment requires both aggressive surgery and chemotherapy. Older patients are less likely to be offered standard treatments, have poorer outcomes, and develop higher toxicity. The principles of geriatric assessment (GA) are not commonly addressed in gynecologic oncology despite its proven importance in predicting outcomes in other subspecialties. The Cancer and Aging Research Group (CARG) GA predicted grades 3 to 5 toxicity in patients with various cancer types through measuring prechemotherapy functional status, comorbidities, cognition, emotional, social support, and nutrition factors (Hurria et al, J Clin Oncol 2011; 29:3457-3465).

Methods: This study is being conducted at Memorial Sloan-Kettering Cancer Center as a proof-of-concept and feasibility study for cancer-specific GA and geriatric nursing telephone interventions in a gynecologic oncology practice. Patients aged 65 years and older with stage I–IV ovarian cancer complete 4 GAs over 6 months: before initiation and on completion of a chemotherapy regimen and surgical procedure. GA feasibility and clinical predictors of toxicity to chemotherapy and surgery will be described. A weekly telephone call (random selection 1:1) from geriatric nurse practitioner reinforces medication compliance, offers psychosocial support, and facilitates appropriate referrals. Toxicity, quality of life, use of care, and satisfaction parameters will be compared with those experienced by the patients receiving standard oncology care alone.

Results: Since November, 2011 a total of 23 patients have been enrolled, with 13 patients (57%) completing the study. Further accrual continues (goal, 50 patients). The mean age is 75 years (range, 65–91 years). Nineteen patients (82%) had ovarian cancer and 4 (18%) had fallopian tube cancer. Sixteen patients (70%) received neoadjuvant chemotherapy and 7 (30%) underwent primary surgical debulking. Regimens include platinum/taxane intravenous (80%), platinum/taxane intraperitoneal (10%), or single-agent carboplatin (10%). Baseline dose reduction was needed in 12 patients (57%). Baseline GA variables included a mean physician-rated Karnofsky performance score of 70% (50%–90%); 36% of patients having 2 or more comorbidities (range, 0–8); 14% of patients having a recent fall history; 32% of patients experiencing unintentional weight loss (>5%); 23% of patients with a body mass index of less than 22; 9% of patients having a significant cognitive impairment (Blessed Orientation-Memory-Concentration test score >10); and 36% of patients having high anxiety/depression (Hospital Anxiety and Depression score >14). Twelve patients (52%) were randomized to telephone intervention.

Conclusions: A cancer-specific GA and weekly geriatric nursing telephone contact is feasible. Comorbidity, nutritional loss, cognitive impairment, and depression/anxiety are detected with GA. Most older patients accrued to this study were referred for neoadjuvant chemotherapy. Accrual to the study continues with final results expected in 2014.

This work was funded by the NCCN Foundation. Any opinions, findings, and conclusions expressed in this material are those of the author(s) and do not necessarily reflect those of the National Comprehensive Cancer Network or the NCCN Foundation.

AB2013-20. Attitudes Toward and Use of Cancer Management Guidelines in a National Sample of Medical Oncologists and Surgeons

Reshma Jagsi, Grace Huang, Kent Griffith, Brian Zikmund-Fisher, Nancy Janz, Jennifer Griggs, Steven Katz, and Sarah Hawley

University of Michigan Comprehensive Cancer Center

This abstract was previously published in the Journal of Clinical Oncology.

Huang G, Hawley ST, Zikmund-Fisher B, et al. Attitudes toward and use of cancer management guidelines in a national sample of medical oncologists and surgeons. ASCO MEETING ABSTRACTS Dec 4, 2012:163. Available at:

This work was funded by the NCCN Foundation. Any opinions, findings, and conclusions expressed in this material are those of the author(s) and do not necessarily reflect those of the National Comprehensive Cancer Network or the NCCN Foundation.

AB2013-21. Impact of Age on the Efficacy of Newer Adjuvant Therapies in More Than 14,500 Patients With Stage II/III Colon Cancer: Findings From the ACCENT Database

Nadine J. McCleary, Jeffrey Meyerhardt, Erin Green, Greg Yothers, Aimery de Gramont, Eric Van Cutsem, Michael O'Connell, Christopher Twelves, Leonard Saltz, Daniel Haller, and Daniel Sargent

Dana-Farber Cancer Institute

This abstract was submitted for publication in the Journal of Clinical Oncology.

McCleary NJ, Meyerhardt JA, Green E, et al; for the ACCENT collaborative group. Impact of age on the efficacy of newer adjuvant therapies in 14,528 patients with stage II / III colon cancer: findings from the ACCENT Database. J Clin Oncol, in review.

This work was funded by the NCCN Foundation. Any opinions, findings, and conclusions expressed in this material are those of the author(s) and do not necessarily reflect those of the National Comprehensive Cancer Network or the NCCN Foundation.

2013 Abstract Disclosure Index

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 501 72 2
PDF Downloads 177 78 5
EPUB Downloads 0 0 0