Search

PURPOSE For many patients with differentiated thyroid cancer, use of radioactive iodine (RAI) does not improve survival or reduce recurrence risk. Yet there is wide variation in RAI use, emphasizing the importance of understanding patient perspectives regarding RAI decision making. PATIENTS AND METHODS All eligible patients diagnosed with thyroid cancer from 2014 to 2015 from the Georgia and Los Angeles SEER registries were surveyed (N = 2,632; response rate, 63%). Patients in whom selective RAI use is recommended were included in this analysis (n = 1,319). Patients were asked whether they felt like they had a choice to receive RAI (yes or no), how strongly their physician recommended RAI (5-point Likert-type scale), whether they received RAI (yes or no), and how satisfied they were with their RAI decision (more [score of 4 or greater] v less). Multivariable, weighted logistic regression with multiple imputation was used to assess the associations between patient characteristics and perception of no RAI choice and between perception of no RAI choice with receipt of RAI and decision satisfaction. RESULTS More than half of respondents (55.8%) perceived they did not have an RAI choice, and the majority of patients (75.9%) received RAI. The odds of perceiving no RAI choice was greater among those whose physician strongly recommended RAI (adjusted odds ratio [OR], 1.56; 95% CI, 1.13 to 2.17). Patients who perceived they did not have an RAI choice were more likely to receive RAI (adjusted OR, 2.50; 95% CI, 1.64 to 3.82) and report lower decision satisfaction (adjusted OR, 2.31; 95% CI, 1.67 to 3.20). CONCLUSION Many patients did not feel they had a choice about whether to receive RAI. Patients who perceived they did not have a choice were more likely to receive RAI and report lower decision satisfaction, suggesting a need for more shared decision making to reduce overtreatment.

PURPOSE Genetic testing for cancer risk has expanded rapidly. We examined clinical genetic testing and results among population-based patients with breast and ovarian cancer. METHODS The study included all women 20 years of age or older diagnosed with breast or ovarian cancer in California and Georgia between 2013 and 2014 and reported to the SEER registries covering the entire state populations. SEER data were linked to results from four laboratories that performed nearly all germline cancer genetic testing. Testing use and results were analyzed at the gene level. RESULTS There were 77,085 patients with breast cancer and 6,001 with ovarian cancer. Nearly one quarter of those with breast cancer (24.1%) and one third of those with ovarian cancer (30.9%) had genetic test results. Among patients with ovarian cancer, testing was lower in blacks (21.6%; 95% CI, 18.1% to 25.4%; v whites, 33.8%; 95% CI, 32.3% to 35.3%) and uninsured patients (20.8%; 95% CI, 15.5% to 26.9%; v insured patients, 35.3%; 95% CI, 33.8% to 36.9%). Prevalent pathogenic variants in patients with breast cancer were BRCA1 (3.2%), BRCA2 (3.1%), CHEK 2 (1.6%), PALB2 (1.0%), ATM (0.7%), and NBN (0.4%); in patients with ovarian cancer, prevalent pathogenic variants were BRCA1 (8.7%), BRCA2 (5.8%), CHEK2 (1.4%), BRIP1 (0.9%), MSH2 (0.8%), and ATM (0.6%). Racial/ethnic differences in pathogenic variants included BRCA1 (ovarian cancer: whites, 7.2%; 95% CI, 5.9% to 8.8%; v Hispanics, 16.1%; 95% CI, 11.8% to 21.2%) and CHEK2 (breast cancer: whites, 2.3%; 95% CI, 1.8% to 2.8%; v blacks, 0.1%; 95% CI, 0% to 0.8%). When tested for all genes that current guidelines designate as associated with their cancer type, 7.8% of patients with breast cancer and 14.5% of patients with ovarian cancer had pathogenic variants. CONCLUSION Clinically-tested patients with breast and ovarian cancer in two large, diverse states had 8% to 15% prevalence of actionable pathogenic variants. Substantial testing gaps and disparities among patients with ovarian cancer are targets for improvement.

10518 Background: Germline genetic testing is widespread after breast cancer diagnosis and increasingly informs treatment decisions; however, guidelines do not advise selecting chemotherapy regimens based on genetic testing results. It is unknown whether women with pathogenic variants (PVs) in BRCA1, BRCA2 ( BRCA1/2) or other cancer risk genes receive different chemotherapy regimens than women with negative genetic testing results. Methods: We linked Surveillance, Epidemiology and End Results (SEER) registry records from Georgia and California to clinical germline genetic testing results from four participating laboratories (Ambry, Bioreference/GeneDx, Invitae, and Myriad). For this analysis, we included patients who: 1) were diagnosed with stages I-III breast cancer, either hormone receptor-positive and HER2-negative (HR+HER2-) or triple-negative, in Georgia or California from 2013-2017; 2) received chemotherapy based on SEER records; and 3) linked to a genetic testing result. We further selected cases by genetic testing results: 50% PVs in BRCA1/2 or another cancer risk gene, 25% variant of uncertain significance (VUS) only and 25% negative. We extracted details of chemotherapy regimens from SEER text fields completed by registrars. We categorized regimens by drug classes reported (anthracycline, taxane, platinum, nitrogen mustard, other). We used multivariable models that controlled for age, race/ethnicity, stage, grade, surgical procedure, radiotherapy receipt and geographic site to test whether PV carriers received a more intensive chemotherapy regimen. For HR+HER2-, a more intensive regimen was defined as at least three drugs including an anthracycline and for triple-negative, as at least four drugs including an anthracycline and a platinum (versus fewer drugs). Results: 2,293 women were included, 1,451 with HR+HER2- and 842 with triple-negative disease. On multivariable analysis, receipt of a more intensive chemotherapy regimen was associated with having a BRCA1/2 PV among women with HR+HER2- disease (odds ratio 1.22, p = 0.036), but not among women with triple-negative disease. Moreover, platinum use was elevated in BRCA1/2 PV carriers with HR+HER2- disease (from an adjusted model: BRCA1/2 PV 10.9%, other PV 3.6%, VUS 5.6%, negative 5.7%), while in BRCA1/2 PV carriers with triple-negative disease, platinum use did not vary significantly by genetic results ( BRCA1/2 27.7%, other PV 27.7%, VUS 20.9%, negative 20.7%; p = 0.025 for interaction between genetic result and subtype). Conclusions: Compared to women with negative genetic testing results, women with BRCA1/2 PVs more often received a platinum and/or an anthracycline in chemotherapy regimens for early-stage, HR+HER2- breast cancer. This suggests potential over-treatment. No differences in chemotherapy regimen by genetic testing result were observed in triple-negative disease.

Purpose Little is known about the extent to which genetic counseling is integrated into community practices for patients newly diagnosed with breast cancer. We examined the receipt of clinically indicated genetic counseling in these patients. Patients and Methods We surveyed 5,080 patients between the ages of 20 and 79 years, diagnosed from July 2013 to August 2015 with early-stage breast cancer and reported to the SEER registries of Georgia and Los Angeles County. Surveys were linked to SEER clinical data and genetic test results. The study sample (N = 1,711) comprised patients with indications for formal genetic risk evaluation. Results Overall, 47.4% did not get tested, 40.7% tested negative, 7.4% had a variant of uncertain significance only, and 4.5% had a pathogenic mutation. Three quarters (74.6%) received some form of genetic counseling (43.5%, formal counseling and 31.1%, physician-directed discussion). Virtually all tested patients (96.1%) reported some form of genetic discussion (62.2%, formal counseling and 33.9%, physician-directed discussion). However, only one half (50.6%) of those not tested received any discussion about genetics. Younger women were more likely to report some type of counseling, controlling for other factors: odds ratio, 4.5 (95% CI, 2.6 to 8.0); 1.9 (95% CI, 1.1 to 3.3); and 1.5 (95% CI, 1.0 to 2.3) for women younger than 50 years of age, 50 to 59 years of age, and 60 to 69 years of age versus those 70 years of age and older. Patients’ assessments of the amount of information they received about whether to get tested were similarly high whether they were counseled by a genetics expert or by a physician only (80.8% v 79.4% stated information was just right, P = .59). Conclusion Less than one half (43.5%) of patients with clinical indications received formal genetic counseling. There is a large gap between mandates for timely pretest formal genetic counseling in higher-risk patients and the reality of practice today.

Purpose To determine the demographic, clinical, decision-making, and quality-of-life factors that are associated with treatment decision regret among long-term survivors of localized prostate cancer. Patients and Methods We evaluated men who were age ≤ 75 years when diagnosed with localized prostate cancer between October 1994 and October 1995 in one of six SEER tumor registries and who completed a 15-year follow-up survey. The survey obtained demographic, socioeconomic, and clinical data and measured treatment decision regret, informed decision making, general- and disease-specific quality of life, health worry, prostate-specific antigen (PSA) concern, and outlook on life. We used multivariable logistic regression analyses to identify factors associated with regret. Results We surveyed 934 participants, 69.3% of known survivors. Among the cohort, 59.1% had low-risk tumor characteristics (PSA < 10 ng/mL and Gleason score < 7), and 89.2% underwent active treatment. Overall, 14.6% expressed treatment decision regret: 8.2% of those whose disease was managed conservatively, 15.0% of those who received surgery, and 16.6% of those who underwent radiotherapy. Factors associated with regret on multivariable analysis included reporting moderate or big sexual function bother (reported by 39.0%; OR, 2.77; 95% CI, 1.51 to 5.0), moderate or big bowel function bother (reported by 7.7%; OR, 2.32; 95% CI, 1.04 to 5.15), and PSA concern (mean score 52.8; OR, 1.01 per point change; 95% CI, 1.00 to 1.02). Increasing age at diagnosis and report of having made an informed treatment decision were inversely associated with regret. Conclusion Regret was a relatively infrequently reported outcome among long-term survivors of localized prostate cancer; however, our results suggest that better informing men about treatment options, in particular, conservative treatment, might help mitigate long-term regret. These findings are timely for men with low-risk cancers who are being encouraged to consider active surveillance.

10517 Background: Breast and ovarian cancer patients increasingly undergo germline genetic testing. While studies suggest a greater chemotherapy benefit for carriers of BRCA1/2 pathogenic variants, little is known about whether pathogenic variants in other genes are associated with cancer mortality. Methods: Georgia and California Surveillance, Epidemiology and End Results (SEER) registry records of women diagnosed with breast cancer or ovarian cancer from 2013-2017 were linked to results of clinical germline genetic testing from four participating laboratories. Patients were included if they linked to a genetic result, had stages I-III breast cancer or I-IV epithelial ovarian cancer and received chemotherapy. Multivariable Cox proportional hazard models were used to examine the association of genetic results, demographic and clinical factors with cancer-specific mortality. Results: 21,348 breast and 4,320 ovarian cancer patients were analyzed with median follow-up of 41 months. Pathogenic variants were present in 12% of patients with estrogen and progesterone receptor-positive, HER2-negative breast cancer, 9% with HER2-positive breast cancer, 17% with triple-negative breast cancer and 18% with ovarian cancer. Pathogenic variants were most common in BRCA1/2, CHEK2, PALB2, ATM and BRIP1. Among triple-negative breast cancer patients, mortality was lower with pathogenic variants in BRCA1 (hazard ratio (HR) 0.27, 95% confidence interval (CI) 0.17-0.45) and genes other than BRCA1/2 (HR 0.33, CI 0.13-0.81) versus no pathogenic variant. Genetic results were not associated with mortality in other breast cancer subtypes. Among ovarian cancer patients, mortality was lower with pathogenic variants in BRCA2 (HR 0.36, CI 0.26-0.49) and in genes other than BRCA1/2 (HR 0.48, CI 0.33-0.70). Conclusions: Among breast and ovarian cancer patients treated with chemotherapy, those with germline pathogenic variants in several cancer-associated genes had equivalent or lower short-term mortality than those testing negative. These results may guide patient counseling and clinical trial design.

e14028 Background: While prior research has examined how primary care providers (PCPs) can help care for breast and colon cancer survivors, little is known about the role of PCPs in thyroid cancer survivorship. We surveyed PCPs to understand their involvement and confidence in providing several aspects of thyroid cancer survivorship care. Methods: Thyroid cancer patients (N = 2632) diagnosed in 2014-2015 and identified from the Georgia and LA SEER registries, were surveyed, and identified their PCPs, who then were also surveyed (N = 162, 56% response rate). PCPs were asked to report their: 1) involvement in thyroid cancer long-term surveillance (not involved: never/rarely vs. involved: sometimes/often/almost always involved); and 2) confidence in handling four aspects of thyroid cancer survivorship care (low: not at all/a little/somewhat vs. high: quite/very confident). We examined PCP factors (individual, practice, attitudes and beliefs) associated with PCP involvement in long-term surveillance using multivariable logistic regression. We then examined associations between PCP involvement in long-term surveillance and PCP confidence in handling the four aspects of thyroid cancer survivorship care using Chi-squared tests. Results: The majority of PCPs (76%) reported involvement in long-term surveillance. Adjusted odds of PCPs reporting involvement was greater among PCPs who somewhat/strongly believed that PCPs have the skills necessary to provide survivorship care (OR 4.33; 95% CI 1.38-15.37), and PCPs who noted clinical guidelines as being the most influential in guiding treatment (OR 4.48; 95% CI 1.67-13.07). PCP reports of high confidence in handling the four aspects of survivorship care were consistently less than 50%: when to refer patient to specialist (40%), role of neck ultrasound (36%), role of random thyroglobulin levels (27%), and when to end long-term surveillance (14%). PCPs who reported involvement in long-term surveillance were more likely to report high confidence in discussing the role of random thyroglobulin levels (33.3% vs. 7.9% for those not involved, p < 0.01) but not the other aspects of survivorship care. Conclusions: Though many PCPs reported they were involved in thyroid cancer long-term surveillance, gaps remain in their confidence in handling several aspects of thyroid cancer survivorship care. Guidelines were noted as important in influencing management; therefore, developing guidelines that specifically delineate PCP roles in thyroid cancer survivorship may present one opportunity to increase PCPs’ confidence about their participation.

Purpose We examined racial and ethnic disparities in patient-provider communication (PPC), perceived care quality, and patient activation among long-term cancer survivors. Methods In 2005 to 2006, survivors of breast, prostate, colorectal, ovarian, and endometrial cancers completed a mailed survey on cancer follow-up care. African American, Asian/Pacific Islander (Asian), Hispanic, and non-Hispanic white (white) survivors who had seen a physician for follow-up care in the past 2 years (n = 1,196) composed the analytic sample. We conducted linear and logistic regression analyses to identify racial and ethnic differences in PPC (overall communication and medical test communication), perceived care quality, and patient activation in clinical care (self-efficacy in medical decisions and perceived control). We further examined the potential contribution of PPC to racial and ethnic differences in perceived care quality and patient activation. Results Compared with white survivors (mean score, 85.16), Hispanic (mean score, 79.95) and Asian (mean score, 76.55) survivors reported poorer overall communication (P = .04 and P < .001, respectively), and Asian survivors (mean score, 79.97) reported poorer medical test communication (P = .001). Asian survivors were less likely to report high care quality (odds ratio, 0.47; 95% CI, 0.30 to 0.72) and reported lower self-efficacy in medical decisions (mean score, 74.71; P < .001) compared with white survivors (mean score, 84.22). No disparity was found in perceived control. PPC was positively associated with care quality (P < .001) and self-efficacy (P < .001). After adjusting for PPC and other covariates, when compared with whites, Asian disparities remained significant. Conclusion Asian survivors report poorer follow-up care communication and care quality. More research is needed to identify contributing factors beyond PPC, such as cultural influences and medical system factors.

Purpose To evaluate the financial experiences of a racially and ethnically diverse cohort of long-term breast cancer survivors (17% African American, 40% Latina) identified through population-based registries. Methods Longitudinal study of women diagnosed with nonmetastatic breast cancer in 2005 to 2007 and reported to the SEER registries of metropolitan Los Angeles and Detroit. We surveyed 3,133 women approximately 9 months after diagnosis and 4 years later. Multivariable models evaluated correlates of self-reported decline in financial status attributed to breast cancer and of experiencing at least one type of privation (economically motivated treatment nonadherence and broader hardships related to medical expenses). Results Among 1,502 patients responding to both surveys, median out-of-pocket expenses were ≤ $2,000; 17% of respondents reported spending > $5,000; 12% reported having medical debt 4 years postdiagnosis. Debt varied significantly by race: 9% of whites, 15% of blacks, 17% of English-speaking Latinas, and 10% of Spanish-speaking Latinas reported debt (P = .03). Overall, 25% of women experienced financial decline at least partly attributed to breast cancer; Spanish-speaking Latinas had significantly increased odds of this decline relative to whites (odds ratio [OR], 2.76; P = .006). At least one privation was experienced by 18% of the sample; blacks (OR, 2.6; P < .001) and English-speaking Latinas (OR, 2.2; P = .02) were significantly more likely to have experienced privation than whites. Conclusion Racial and ethnic minority patients appear most vulnerable to privations and financial decline attributable to breast cancer, even after adjustment for income, education, and employment. These findings should motivate efforts to control costs and ensure communication between patients and providers regarding financial distress, particularly for vulnerable subgroups.

52 Background: Team-based cancer care models promote primary care providers taking on a larger role in survivorship care. However, little is known about PCP attitudes toward and propensity to order non-recommended surveillance testing. Methods: A stratified random sample of PCPs identified by early-stage breast cancer patients diagnosed in 2013-15 who participated in iCanCare Study (Georgia and Los Angeles SEER registries) were surveyed about their experiences caring for cancer patients (N = 519, 58% response rate). PCPs were asked in a clinical vignette whether they would order non-recommended bone scans, other imaging (i.e. PET) or tumor marker testing (i.e. CA-125) in an asymptomatic, early-stage breast cancer survivor. A composite score was created by averaging their responses to the individual items, and categorized by tertiles into low, selective and high propensity to order non-recommended testing. PCP confidence in their knowledge about appropriate testing was also measured (5 pt. Likert-type scale; not all all-very confident) and compared with propensity to order. Multivariable, weighted, multinomial logistic regression was used to then evaluate PCP-reported factors associated with high and selective propensity to order non-recommended testing (vs. low). Results: In this sample, 32% of PCPs had a low propensity, 40% selective, and 28% a high propensity to order non-recommended surveillance tests. Of the 80% of PCPs who reported they were confident in their knowledge about appropriate testing, 27% had a high propensity to order. PCPs practicing in staff-model HMOs were less likely to have a high or selective propensity to order (vs. low) when compared to PCPs in private practice. (high aOR: 0.4, 95%CI: 0.2-0.7; selective aOR: 0.3, 95%CI: 0.2-0.6). Conclusions: Over a quarter of PCPs had a high propensity to order non-recommended surveillance testing for early-stage breast cancer patients, yet the majority reported they were confident in their knowledge about appropriate surveillance testing. Efforts to increase PCP knowledge about the specifics of breast cancer surveillance may be warranted to reduce the overuse of non-recommended testing during survivorship.

157 Background: Thyroid cancer is one of the most common cancers in young adults, age 18-39 years. Since this population has distinct medical and developmental needs, it is important to understand age differences in worry about quality of life. Methods: Thyroid cancer patients diagnosed between 2014-15 were accrued from the Surveillance, Epidemiology and End Results Program (SEER) registries of Georgia and Los Angeles County and surveyed between 2017-18. The primary outcome was patient reported worry about quality of life. Covariates for this analysis included age (categorized 18-25, 26-39, 40-64, and 65-79), involvement in surgical decision making (low, just right, high) and amount of information received on treatment side effects (low, just right, high). Race/ethnicity, sex, and tumor characteristics were also evaluated. Multivariable analyses were used to assess associations between worry about quality of life and independent variables. Results: In the cohort of 2,632 patients, 39.8% were somewhat to very much worried about quality of life not being the same as before thyroid cancer diagnosis, 15.1% reported low involvement in surgical decision making, and 32.6% reported low amount of information regarding side effects. Younger age was associated with high worry about quality of life (p < 0.001), low involvement in decision making (p < 0.001) and low amount of information on treatment side effects (p < 0.001). In multivariable analyses, high worry was significantly associated with female sex (Odds ratio [OR]: 1.29 95% confidence Interval [CI]:1.04-1.59), non-white, non-Hispanic race/ethnicity (OR: 1.64 CI: 1.38-1.95), positive lymph nodes (OR: 1.41 CI: 1.14-1.75) and younger age (ages 18-25 years OR: 2.34 CI: 1.45-3.76; ages 26-39 years OR: 2.26 CI: 1.68-3.05; and ages 40-64 years OR:1.94 CI 1.50-2.51). Age differences remained significant after adding either involvement in decision-making or information received. Conclusions: Young adults report higher worry about quality of life. Controlling for decision-making involvement or information received about side effects did not eliminate age differences in worry. Ongoing psychosocial support addressing quality of life worries among younger survivors is needed.

560 Background: Increasing use of germline genetic testing may have unintended consequences on breast cancer treatment. We do not know whether treatment deviates from guidelines for women with pathogenic variants (PV) in cancer susceptibility genes. Methods: SEER data for all women aged ≥20 years, diagnosed with breast cancer in 2014-15 and reported to Georgia and California registries (N = 77,588) by December 1, 2016 were linked to germline genetic testing results from 4 laboratories that did nearly all clinical testing. We examined first course of therapy (before recurrence or progression) of stage < IV patients who linked to a genetic test: bilateral mastectomy (BLM) in candidates for surgery (unilateral, stages 0-III); post-lumpectomy radiation in those with an indication (all but age ≥70, stage I, hormone receptor (HR)-positive and HER2-negative); and chemotherapy in those without a definitive indication (stage I-II, HR-positive, HER2-negative and 21-gene recurrence score < 30). We report the percent treated based on multivariable modeling, adjusted for age, race, stage, grade, insurance and socioeconomic status. Results: The table shows that 9% of patients who linked to a genetic test result had a PV (N = 1,283). Compared to women with negative results,women with BRCA1/2 PVs were more likely to receive BLM, more likely to receive chemotherapy without definitive indication, and less likely to receive indicated radiation in their first course of therapy. Lower-magnitude effects were seen with other PVs but not variants of uncertain significance (VUS). Conclusions: In a population-based setting, women with PVs in BRCA1/2 or other cancer susceptibility genes may have a higher risk of receiving locoregional and systemic treatment that does not follow guidelines. [Table: see text]

Purpose Adolescents and young adults (AYAs) diagnosed with cancer face numerous physical, psychosocial, and practical challenges. This article describes the health-related quality of life (HRQOL) and associated demographic and health-related characteristics of this developmentally diverse population. Patients and Methods Data are from the Adolescent and Young Adult Health Outcomes and Patient Experience (AYA HOPE) study, a population-based cohort of 523 AYA patients with cancer, ages 15 to 39 years at diagnosis from 2007 to 2009. Comparisons are made by age group and with general and healthy populations. Multiple linear regression models evaluated effects of demographic, disease, health care, and symptom variables on multiple domains of HRQOL using the Pediatric Quality of Life Inventory (PedsQL) and the Short-Form Health Survey 12 (SF-12). Results Overall, respondents reported significantly worse HRQOL across both physical and mental health scales than did general and healthy populations. The greatest deficits were in limitations to physical and emotional roles, physical and social functioning, and fatigue. Teenaged patients (ages 15 to 17 years) reported worse physical and work/school functioning than patients 18 to 25 years old. Regression models showed that HRQOL was worse for those in treatment, with current/recent symptoms, or lacking health insurance at any time since diagnosis. In addition, sarcoma patients, Hispanic patients, and those with high school or lower education reported worse physical health. Unmarried patients reported worse mental health. Conclusion Results suggest that AYAs with cancer have major decrements in several physical and mental HRQOL domains. Vulnerable subgroups included Hispanic patients, those with less education, and those without health insurance. AYAs also experienced higher levels of fatigue that were influenced by current symptoms and treatment. Future research should explore ways to address poor functioning in this understudied group.

Purpose Disparities in receipt of adjuvant chemotherapy may contribute to higher breast cancer fatality rates among black and Hispanic women compared with non-Hispanic whites. We investigated factors associated with receipt of chemotherapy in a diverse population-based sample. Patients and Methods Women diagnosed with breast cancer between August 2005 and May 2007 (N = 3,252) and reported to the Detroit, Michigan, or Los Angeles County Surveillance, Epidemiology, and End Results (SEER) registry were recruited to complete a survey. Multivariable analyses examined factors associated with chemotherapy receipt. Results The survey was sent to 3,133 patients; 2,290 completed a survey (73.1%), and 1,403 of these patients were included in the analytic sample. In multivariable models, disease characteristics were significantly associated with the likelihood of receiving chemotherapy. Low-acculturated Hispanics were more likely to receive chemotherapy than non-Hispanic whites (odds ratio [OR], 2.00; 95% CI, 1.31 to 3.04), as were high-acculturated Hispanics (OR, 1.43; 95% CI, 1.03 to 1.98). Black women were less likely to receive chemotherapy than non-Hispanic whites, but the difference was not significant (OR, 0.83; 95% CI, 0.64 to 1.08). Increasing age (even in women age < 50 years) and Medicaid insurance were associated with lower rates of chemotherapy receipt. Conclusion In this population-based sample, disease characteristics were strongly associated with receipt of chemotherapy, indicating that clinical benefit guides most treatment decisions. We found no compelling evidence that black women and Hispanics receive chemotherapy at lower rates. Interventions that address chemotherapy use rates according to age and insurance status may improve quality of systemic treatment.

Purpose To examine the impact of cancer on work and education in a sample of adolescent and young adult (AYA) patients with cancer. Patients and Methods By using the Adolescent and Young Adult Health Outcomes and Patient Experience Study (AYA HOPE)—a cohort of 463 recently diagnosed patients age 15 to 39 years with germ cell cancer, Hodgkin's lymphoma, non-Hodgkin's lymphoma, sarcoma, and acute lymphocytic leukemia from participating Surveillance, Epidemiology, and End Results (SEER) cancer registries—we evaluated factors associated with return to work/school after cancer diagnosis, a belief that cancer had a negative impact on plans for work/school, and reported problems with work/school after diagnosis by using descriptive statistics, χ2 tests, and multivariate logistic regression. Results More than 72% (282 of 388) of patients working or in school full-time before diagnosis had returned to full-time work or school 15 to 35 months postdiagnosis compared with 34% (14 of 41) of previously part-time workers/students, 7% (one of 14) of homemakers, and 25% (five of 20) of unemployed/disabled patients (P < .001). Among full-time workers/students before diagnosis, patients who were uninsured (odds ratio [OR], 0.21; 95% CI, 0.07 to 0.67; no insurance v employer-/school-sponsored insurance) or quit working directly after diagnosis (OR, 0.15; 95% CI, 0.06 to 0.37; quit v no change) were least likely to return. Very intensive cancer treatment and quitting work/school were associated with a belief that cancer negatively influenced plans for work/school. Finally, more than 50% of full-time workers/students reported problems with work/studies after diagnosis. Conclusion Although most AYA patients with cancer return to work after cancer, treatment intensity, not having insurance, and quitting work/school directly after diagnosis can influence work/educational outcomes. Future research should investigate underlying causes for these differences and best practices for effective transition of these cancer survivors to the workplace/school after treatment.

159 Background: For many patients with differentiated thyroid cancer, use of radioactive iodine (RAI) does not improve survival or reduce recurrence risks. Yet, most patients continue to receive RAI suggesting the need for a better understanding of patient perspectives regarding RAI decision making. Methods: All eligible sequentially diagnosed patients with thyroid cancer in 2014-15 from the Georgia and Los Angeles SEER registries were surveyed (N = 2097, current response rate: 63%). Patients with intermediate risk thyroid cancer where selective RAI use is recommended were included in this analysis (N = 1357). Patients were asked whether or not they felt they had a choice to receive RAI (yes/no), how strongly their physician recommended initial RAI treatment (5 point Likert responses: Strongly against RAI--Strongly recommended RAI), whether they received RAI (yes/no), how satisfied they were with their RAI decision (5-pt Likert-type scale), categorized as more (score ≥4) vs. less satisfied. Multivariable logistic regression was used to assess 1) association between patient characteristics and perception of no RAI choice, 2) perception of no RAI choice with receipt of RAI and 3) perception of no RAI choice with decision satisfaction. Results: In this sample,over half (57%) of the respondents perceived they did not have a RAI choice, and the majority of them (76%) received RAI. The odds of perceiving no RAI choice was greater among Hispanic and Asian patients (Hispanic OR: 1.4, 95%CI: 1.0, 1.9, Asian OR: 1.9, 95%CI: 1.2, 2.9), and those whose physician strongly recommended RAI (OR: 1.9, 95%CI: 1.4, 2.6). Patients who perceived they did not have a RAI choice were more likely to receive RAI (Adjusted OR: 3.2, 95%CI: 2.1, 4.9) and report lower decision satisfaction (Adjusted OR: 2.7, 95%CI: 1.9, 3.7). Conclusions: Many patients in whom selective RAI use is recommended, particularly those of more vulnerable groups, did not feel they had a choice about whether or not to receive RAI. Patients who perceived they did not have a RAI choice were more likely to receive RAI and report lower decision satisfaction, suggesting a need for more shared treatment decision making to reduce overtreatment.

1578Background: Genetic testing for cancer risk has expanded rapidly, with more genes tested. Little is known about test use or pathogenic variant (PV) prevalence among population-based cancer pati...

10080Background: Little is known about whether growing awareness of the financial toxicity of cancer diagnosis and treatment has led to increased clinician engagement nor the status of current pati...

127 Background: Combined with RT, ADT is a highly effective and utilized treatment for men with localized PC, but some studies suggest that use of ADT leads to increased CVD. We explored the association between ADT and CVD in men receiving RT for localized PC in the National Program for Cancer Registry’s (NPCR) Breast and Prostate Cancer Patterns of Care (POC) study. Methods: From 7 population-based cancer registries, we constructed a sample of men with localized PC treated with definitive RT, stratified by race/ethnicity. Cases diagnosed in 2004 were followed through 2009. Comorbidity, at or before diagnosis, was quantified using the Adult Comorbidity Evaluation 27; subcategories were combined into CV disease (CVD = myocardial infarction, coronary artery disease, congestive heart failure, arrhythmia), CVD-equivalents (CVE = peripheral artery disease and stroke) and CV-risk factors (CVRF = hypertension, diabetes, obesity). Cause of death was determined from the National Death Index data and linkage with vital statistics data. Rates of CV death, in those receiving RT alone versus RT+ADT, were compared in univariate and multivariable analyses. Results: The sample included 2,413 men with mean age 67.7 years (range 39-94), 54.5% white non-Hispanic, 997 received RT alone and 1,416 received RT+ADT. Five-year CVD was 2.3% with RT alone and 3.4% with RT+ADT. In univariate analysis, the following predicted higher CVD with ADT: age < 60 (OR 1.42, p = 0.04), white (OR 2.06, p = 0.001), divorced/separated/widowed (OR 1.29, p = 0.02), insured by Medicare (OR 1.64, p = 0.04), living in mixed urban-rural area (OR 1.87, p = 0.03), higher education level (OR 1.56, p = 0.05), high socioeconomic status (OR 1.70, p = 0.03), no or mild comorbidity level (OR 2.14, p = 0.05 and OR 1.66, p = 0.02, respectively), low PSA (OR 5.44, p = 0,02), and Gleason score 3 (OR 2.12, p = 0.02). In multivariate analysis, use of ADT did not significantly predict hazard of death from heart disease (HR 1.21, p = 0.28) or PC (HR 0.76, p = 0.34). Conclusions: After controlling for confounding variables, use of ADT was not associated with increased risk of CVD in men receiving radiation as definitive therapy for localized PC.

Purpose There is concern that minority women have limited access to breast reconstruction. We described patterns of use, experiences with clinicians, and patients' satisfaction with treatment decisions for women of different race/ethnicities. Methods A total of 3,252 patients with breast cancer from Los Angeles and Detroit Surveillance, Epidemiology, and End Results registries were surveyed near the time of diagnosis (n = 2,260, response rate 72.2%). The primary outcomes were receipt of reconstruction, access to information about reconstruction, and decisional satisfaction. The primary independent variable was race/ethnicity (white, African American [AA], highly acculturated Latina [Latina-high], and less acculturated Latina [Latina-low]). Control variables included other sociodemographic and clinical factors. χ2 and multivariate logistic regression were used for the analyses. Results Receipt of reconstruction varied significantly by patient race/ethnicity—40.9% of whites, 33.5% of AAs, 41.2% of Latina-high, and only 13.5% of Latina-low (P < .001)—and persisted when we controlled for demographic and clinical factors. Minority women were significantly less likely than whites to see a plastic surgeon before initial surgery and were more likely to desire more information about reconstruction (17.0% of whites v 27.0% of AAs, 30.0% of Latina-high, and 55.9% of Latina-low; P < .001). Decisional satisfaction was lowest among minority women without reconstruction (P < .001). Conclusion Minority women, particularly less acculturated Latinas, had low receipt of breast reconstruction, which may be related to limited information about the procedure and less access to plastic surgeons. Greater desire for information and lower satisfaction with surgical decisions among these patients motivate greater attention to treatment support for these patients.

Purpose We investigated racial/ethnic differences in functional outcomes up to 5 years after diagnosis among men with aggressively treated localized prostate cancer. Patients and Methods Patients were from the Prostate Cancer Outcomes Study, a population-based cohort study that surveyed patients at 6, 12, 24, and 60 months after diagnosis. Analyses were stratified by primary treatment. Racial/ethnic differences at each time point were assessed using Generalized Estimating Equations, adjusting for pretreatment function, age at diagnosis, secondary treatment, and other confounders. An adjusted summary score for each functional domain was calculated for each time period. Results Patients included 1,475 non-Hispanic white, 321 African-American, and 279 Hispanic prostate cancer patients. After 60 months, among prostatectomy patients, African-Americans had significantly higher sexual function scores than non-Hispanic whites (43.9 v 36.1; P = .02), but were more likely to have a moderate to big problem with sexual function (50.6% v 44.4%; P = .04). African-Americans also had higher urinary function scores at 5 years than non-Hispanic whites (78.5 v 72.4; P = .04) and were less likely to have problems with incontinence. Changes in sexual and bowel function after radiotherapy showed no significant racial/ethnic differences. Conclusion This long-term cohort study found that, among prostatectomy patients, African-Americans had better recovery of sexual and urinary function at 60 months after diagnosis that was likely to be of mild clinical significance, despite reporting more problems with sexual function than non-Hispanic whites. More study is necessary to understand reasons for these differences. In contrast, no racial/ethnic differences in recovery from radiotherapy were found.

10052 Background: Many patients with breast cancer work for pay at time of diagnosis, and the treatment plan may threaten their livelihood. Given rapidly evolving policies, evidence, and treatment options, we evaluated work experiences in a contemporary population-based sample of breast cancer patients to inform initiatives to reduce the burden of cancer care. Methods: We surveyed women aged 20-79 years diagnosed with stages 0-II breast cancer as reported to the SEER registries of Georgia and Los Angeles in 2014-15. Of 3672 eligible women, 2502 responded (68%); we analyzed 1006 who reported working prior to diagnosis. Multivariable models evaluated correlates of missing > 1 month and stopping work altogether vs missing ≤1 month. Results: In this diverse sample (48% white, 19% black, 20% Latina, 11% Asian), most pts (62%) received lumpectomy; 16% had unilateral mastectomy (8% with reconstruction); 23% had bilateral mastectomy (19% with reconstruction). One third (33%) received chemotherapy. The vast majority (84%) worked full time at diagnosis, but only 50% had paid sick leave, 39% disability benefits, and 38% flexible work schedules. Surgical treatment was strongly associated with missing > 1 month of work (OR 7.8 for bilateral mastectomy with reconstruction vs lumpectomy) and with stopping altogether (OR 3.1 for bilateral mastectomy with reconstruction vs lumpectomy). Chemotherapy receipt (OR 1.3 for missing > 1 month; OR 3.9 for stopping altogether) and race (OR 2.0 for missing > 1 month and OR 1.7 for stopping altogether, blacks vs whites) also correlated. Those with paid sick leave were less likely to stop working (OR 0.5), as were those with flexible schedules (OR 0.3). Those with disability benefits were more likely to stop working (OR 1.6) or miss > 1 month of work (OR 2.7). Conclusions: Working patients who received more aggressive treatments, particularly surgery, were much more likely to experience substantial employment disruptions. Given the growing choice of bilateral mastectomy by patients seeking peace of mind, particularly among young women with years of potential employment ahead, these findings suggest the importance of discussing impact of treatment decisions on employment. Funded by NCI P01CA163233.

e18012 Background: Little is known about the size and characteristics of informal decision support networks of women diagnosed with breast cancer and whether involvement of informal decision supporters (DSP) influences breast cancer treatment decisions. Methods: A population-based sample of newly diagnosed breast cancer patients reported to the Georgia and Los Angeles SEER registries in 2014-15 were surveyed approximately 6 months after diagnosis (N = 2,502, 68% response rate). Network size was estimated by asking women to list up to 3 of the most important DSPs who helped them with locoregional therapy decisions. For each individual DSP listed, respondents reported how important each DSP’s opinion was in treatment decision making, and how satisfied they were with their involvement (5 pt. scales, “not at all” to “very”). Decision deliberation was measured using 5-items assessing degree patients thought through the decision, with higher scores reflecting more deliberative treatment decisions. We compared network size (0-3 or more) across patient-level characteristics and estimated the association between network size and deliberation using multivariable linear regression. Results: Of the 2,502 women in this analysis, 51% reported having 3 or more DSPs, 20% reported 2, 18% reported 1, and 11% reported not having any DSPs. Married/partnered women, those younger than 45 years old, and black women were all more likely to report larger networks (all p < 0.001). Partnered women most often reported their partner as their main DSP (37.9%), whereas not partnered/unmarried women most often reported children (38.4%). The majority of women were highly satisfied with their DSP being involved in their decisions (76.5%) and 68.6% felt their DSP was very important in their decision making. Larger support networks were associated with more deliberative surgical treatment decision-making (p < 0.001). Conclusions: Most women engaged multiple DSPs in their treatment decision making, including spouses, children, and friends. Involving more DSPs was associated with more deliberative treatment decisions. Future initiatives to improve breast cancer treatment decision making should acknowledge and engage informal DSPs.

541 Background: Advances in tumor genomic profiling enable increasingly precise estimates of the benefit of adjuvant chemotherapy in early-stage breast cancer. However, little is known about how chemotherapy use, medical oncologists’ (MO) perspectives and recommendations have changed in recent years, particularly in key clinical subgroups such as node-negative and node-positive. Methods: We surveyed 5,080 women (70% response rate), newly diagnosed with breast cancer in 2013-2015 and accrued through two population-based SEER registries (Georgia and Los Angeles), about their MOs’ chemotherapy recommendations and whether they received chemotherapy. Using patient report, we identified 470 attending MOs and surveyed them (n=310, 66% response) about approaches to chemotherapy recommendation, using node-negative and node-positive case scenarios. We evaluated factors associated with chemotherapy receipt over time using multi-level logistic regression. Results: The analytic sample was 2,926 patients with stages I-II, estrogen receptor-positive, HER2-negative breast cancer. Chemotherapy use declined to 21% from 34% during the study period (2013-2015, p

157 Background: Prior studies suggest that women often involve a network of family and friends in their cancer treatment decision making, yet very little is known about the size and characteristics of these decision support networks and whether their involvement leads to high-quality breast cancer treatment decisions. Methods: A weighted random sample of patients newly diagnosed with breast cancer in 2014-15 as reported to the Georgia and Los Angeles SEER registries were surveyed approximately 6 months after diagnosis (N = 2,502, 70% response rate). Network size was estimated by asking women to list up to 3 of the most important decision support people (DSP) who helped them make their locoregional therapy decisions. Decision deliberation was measured using 4-items assessing degree to which patients thought through the decision, with higher scores reflecting more deliberative breast cancer treatment decisions. We compared the size of the network (0-3+ people) across patient-level characteristics and estimated the adjusted mean deliberation scores across levels of network size using multivariable linear regression. Results: Of the 2,502 women included in this analysis, 51% reported having at least 3 DSPs, 20% reported 2, 18% reported 1, and 10% reported not having any DSPs. Among women who were not partnered (N = 961), 51% had 3 DSPs, 18% had 2, 16% had 1 and 16% had 0 DSPs. Of the DSPs that the respondents identified, the majority were children (30%), followed by partners/spouses (23%), friends (15%), siblings (10%), other family members (6%), and parents (5%). Married/partnered women (p < 0.001), those younger than 45 years old (p < 0.001), those with more than 1 comorbidity (p < 0.001), and black women (p = 0.02) were all more likely to report larger networks on average. Larger support networks were associated with more deliberative surgical decisions (p < 0.001). Conclusions: In this population-based sample, the majority of women engaged DSPs in their treatment decision making and for non-partnered patients, DSPs still played a key role in decision making. Larger size decision support networks were associated with higher quality decisions, underscoring the importance of efforts to identify and engage DSPs in the breast cancer decision making process.

158 Background: Prior studies have shown that worry about recurrence is a common problem during cancer treatment and survivorship and may be associated with symptom response and surveillance after primary treatment. However, whether worry about recurrence influences patient preferences for which provider to see for their continuing care remains unknown. Methods: A random sample of patients newly diagnosed with breast cancer in 2014-15 as reported to the Georgia and Los Angeles SEER registries were surveyed approximately 6 months after diagnosis (N = 2,502, 70% response rate). Frequency of worry about recurrence was defined by asking women to indicate on a 5-pt scale how often they worried about their cancer coming back in the past month (not at all-always) and was then dichotomized as frequent worry (sometimes/often/almost always) vs. less worry (almost never/rarely). Patient preferences for which provider manages certain aspects of care after treatment were ascertained for: follow-up mammograms, screening for other cancers, general preventive care, and treatment of comorbidities. Response categories included primary care clinician (PCP), cancer doctors, either or both. The associations between patient-reported worry about recurrence with preferences for provider roles were assessed using multinomial logistic regression. Results: In this sample, 37% of women reported worrying frequently about recurrence. Controlling for patient and clinical factors, women who reported more frequent worry (vs. less worry) were more likely to prefer to see both clinicians (vs. PCP only) for mammograms (OR: 2.3, 95%CI: 1.5, 3.6), screening for other cancers (OR: 2.3, 95%CI: 1.5, 3.5), general preventive care (OR: 1.6, 95%CI: 1.1, 2.3) and comorbidity care (OR: 1.5, 95%CI: 1.03, 2.2). Conclusions: Frequent worry about recurrence was common in this sample of women with favorable prognosis breast cancer. More frequent worry about breast cancer recurrence was associated with stronger preferences for seeing both PCPs and oncologists for continuing care after treatment. Assessing and managing worry about recurrence early in survivorship may improve collaborative cancer care and reduce duplicated services after treatment.

197 Background: Because childhood cancer survivors (CCS) report a high number of unmet cancer-related information needs, knowing where CCS access information may provide insights into appropriate delivery systems to address information deficits. Active seeking of health information among CCS is associated with positive health behaviors, including adherence to follow-up care. This study examined patterns of cancer-related health information seeking as an indicator of health care engagement. Methods: Participants (N = 193) were young adult CCS diagnosed with any cancer type in Los Angeles County, 54% Hispanic, with a mean age of 19.87, and at least two years from treatment. CCS were asked where they accessed health information related to their cancer with 8 response options which were categorized into four information domains: hospital resources, social media, other survivors, and family members. Multivariable logistic regression was used to assess correlates of each information domain, including sociodemographic information, post-traumatic growth, and health care engagement, controlling for age, ethnicity, sex, education, and health insurance status. Results: Hospital resources were the most frequently endorsed information domain (65.3%), and Hispanic CCS (vs. non-Hispanic) were more likely to access this source. Online sources were more likely accessed by female CCS. Seeking information from other cancer survivors was associated with follow-up care and post-traumatic growth after cancer. Hispanic CCS were less likely to seek information from other survivors and their family than non-Hispanics. Conclusions: While CCS obtain information from a variety of sources, hospital resources were the most commonly accessed, particularly for Hispanics. Information sharing between survivors may promote positive health care engagement; however, Hispanics may be less likely to pursue this resource. Future work should address barriers Hispanic CCS face in information sharing with other cancer survivors. [Table: see text]

99 Background: Prior studies have explored barriers to collaborative cancer care and suggest a need for greater clarity about clinician roles. However, little is known about patient’s preferences and expectations for which clinicians handle various aspects of care after primary cancer treatment. Methods: A weighted random sample of patients newly diagnosed with breast cancer in 2014-15 as reported to the Georgia and Los Angeles SEER registries were surveyed approximately 6 months after diagnosis (N = 2,342, 70% response rate). Patient preferences for which clinician handles the following aspects of care after treatment were ascertained: follow-up mammograms, screening for other cancers, general preventive care, treatment of comorbidities and reassurance about their breast cancer. Response categories included primary care clinician (PCP), cancer doctors, either or both. The associations between patient demographic and clinical factors with preferences for clinician roles were assessed using multinomial logistic regression. Results: The majority of women preferred to see oncologists for mammograms (65%), other cancer screening (64%), and reassurance about their cancer (69%), and PCPs for general preventive care (76%) and comorbidity care (79%). Black women and Hispanic women (vs. White) were more likely to prefer to see both clinicians (vs. PCP) for mammograms (Black OR: 2.8, 95%CI: 1.5, 5.1; Hispanic OR: 1.9, 95%CI: 1.1, 3.3), screening for other cancers (Black OR: 3.3, 95%CI: 1.8, 6.1; Hispanic OR: 1.8, 95%CI: 1.0, 3.1), general preventive care (Black OR: 3.5, 95%CI: 2.0, 6.0; Latina OR: 2.5, 95%CI: 1.4, 4.3) and comorbidity care (Black OR: 2.1, 95%CI: 1.3, 3.6; Hispanic OR: 2.0, 95%CI: 1.2, 3.3). Less than a high school education was also associated with a preference for seeing both clinicians for mammograms. Conclusions: While patients report clear preferences for which clinicians handle various aspects of their collaborative survivorship care, variation exists by race and education. Targeting these patients to clarify clinician roles may be particularly effective in interventions to improve collaborative cancer care.

PURPOSE: To compare health-related quality-of-life outcomes after primary androgen deprivation (AD) therapy with orchiectomy versus luteinizing hormone-releasing hormone (LHRH) agonists for patients with prostate cancer. PATIENTS AND METHODS: Men (n = 431) newly diagnosed with all stages of prostate cancer from six geographic regions who participated in the Prostate Cancer Outcomes Study and who received primary AD therapy but no other treatments within 12 months of initial diagnosis were included in a study of health outcomes. Comparisons were statistically adjusted for patient sociodemographic and clinical characteristics, timing of therapy, and use of combined androgen blockade. RESULTS: More than half of the patients receiving primary AD therapy had been initially diagnosed with clinically localized prostate cancer. Among these patients, almost two thirds were at high risk of progression on the basis of prognostic factors. Sexual function outcomes were similar by treatment group both before and after implementation of AD therapy. LHRH patients reported more breast swelling than did orchiectomy patients (24.9% v 9.7%, P < .01). LHRH patients reported more physical discomfort and worry because of cancer or its treatment than did orchiectomy patients. LHRH patients assessed their overall health as fair or poor more frequently than did orchiectomy patients (35.4% v 28.1%, P = .01) and also were less likely to consider themselves free of prostate cancer after treatment. CONCLUSION: Most endocrine-related health outcomes are similar after surgical versus medical primary hormonal therapy. Stage at diagnosis had little effect on outcomes. These results provide representative information comparing surgical and medical AD therapy that may be used by physicians and patients to inform treatment decisions.

6501Background: The 21-gene assay enables oncologists to tailor chemotherapy in early-stage, estrogen-receptor positive breast cancer. Few studies have studied the impact of test results on chemoth...

e22542Background: Well differentiated / dedifferentiated (WD / DD) liposarcoma is one of the most common subtypes of soft tissue sarcomas. There are no known risk factors and a genetic predispositi...

177 Background: Contralateral prophylactic mastectomy (CPM) use is increasing in women who are not at increased risk of contralateral cancer development and will experience no survival benefit from the more morbid procedure. Little is known about treatment decision-making or provider interactions. Methods: We surveyed a weighted random sample of newly diagnosed patients with early-stage breast cancer who were treated in 2013-14, identified through the population-based SEER registries of Los Angeles and Georgia about 3 months after surgical treatment, and merged with SEER data (N=2632, RR=70%), to determine receipt of diagnostic tests and factors related to the decision about surgery (including knowledge and perceived physician recommendation). Results: Nearly half of 2,436 respondents with unilateral non-metastatic cancer considered CPM (25% strongly). Only 37% of those who considered CPM knew that it does not improve survival for all women with breast cancer (24% believed it does, 39% didn’t know). Among women receiving CPM, 37% believed it generally improves survival. Ultimately, 1,464 (60%) received BCS and 972 (40%) mastectomy (of whom 438, or 18% overall, received CPM). On multivariable analysis, pts who received CPM were younger, more likely to be white, and more likely to have a family history, private rather than Medicaid insurance, and received MRI. Even among pts without a deleterious genetic mutation or family history in multiple relatives (2,303), 400 (17%) received CPM. CPM was uncommon among pts who reported that their surgeons recommended against it (2.0% [17/832]) but much higher among those who reported no surgeon recommendation regarding CPM (21.3% [229/1,077]), and among those who perceived their surgeons to have recommended it (55.4% [147/265]). Conclusions: Many patients consider CPM, but knowledge is low. Use of CPM is substantial among patients without clinical indications but is low when patients report their surgeon recommended against it. In the context of shared decision-making, surgeon recommendations against CPM might help reduce potential overtreatment.

1541 Background: Breast cancer pts at high risk for pathologic genetic mutations should consider testing. Even those without elevated risk may have concerns about genetic predisposition. Little is ...

1011 Background: Contralateral prophylactic mastectomy (CPM) use is increasing in women who are not at increased risk of contralateral cancer development and will experience no survival benefit fro...

6575 Background: The purpose of this study was to investigate the association between area-level socioeconomic status (SES) and breast cancer biology controlling for demographic factors and comorbi...

31 Background: Androgen deprivation therapy (ADT) has been associated with an increased risk of developing diabetes (DM) and cardiovascular disease (CVD), though this is controversial, particularly for CVD. We prospectively assessed the relationship between ADT and incident DM and CVD in the Prostate Cancer Outcomes Study (PCOS), a population-based cohort of prostate cancer survivors followed longitudinally for 15 years from diagnosis. Methods: We identified men in the PCOS with non-metastatic prostate cancer diagnosed from 1994 to 1995 and followed through 2009 to 2010. We used multivariable logistic regression models to compare groups receiving short-term ADT (less than 2 years), prolonged ADT (2 years or more) and no ADT to assess the relationship between ADT exposure and subsequent diagnoses of DM and CVD (determined by patient report and cause of death data). We evaluated the effects of age at diagnosis, race, stage, and comorbidity on the development of DM and CVD. Results: Among 3,526 men with comorbidity and treatment data, 2,985 men without baseline DM and 3,112 men without baseline CVD constituted the DM and CVD cohorts, respectively. Regardless of duration of ADT exposure, there was not an increased risk of DM or CVD in men younger than 70 at diagnosis. Compared to no ADT exposure, prolonged ADT was associated with an increased risk of DM and CVD that increased steadily over age 76 at diagnosis for DM (OR 2.11 at age 74, 95% CI 1.02 – 4.36; OR 2.65 at age 80, 95% CI 1.09 – 6.47) and age 74 at diagnosis for CVD (OR 1.89 at age 74, 95% CI 1.02 - 3.49; OR 3.19 at age 80, 95% 1.25 – 8.17). Increasing comorbidity burden modified risk of DM and CVD (for 3 or more comorbidities vs. no comorbidities; for DM, OR 4.25, 95% CI 2.3 - 7.9; and for CVD, OR 8.1, 95% CI 4.3 -15.5 P

e20532 Background: To identify the prevalence and predisposing, enabling and need for care characteristics of long-term adult cancer survivors who use complementary and alternative medicine (CAM) modalities. Methods: We analyzed data from the Follow-up Care Use among Survivors (FOCUS) Survey, a cross-sectional population-based investigation of the follow-up care experiences among racially and ethnically diverse long-term (i.e., five years or more post-diagnosis) cancer survivors recruited from California NCI SEER registries. Ever and last year use of CAM modalities, as well as reasons for use, were assessed. Multivariable logistic regression models were applied to identify if predisposing, enabling and need for care characteristics described in the CAM Healthcare Model were associated with CAM use in the past year. Results: Long-term cancer survivors in the study (N=1,666) were predominately female (62%) and older (mean age=69.5), with breast (24%), prostate 26(%), colorectal (25%), ovarian (13%) and endometrial cancers (12%). Forty-two percent of survivors used CAM in the past year to relieve stress (28%), treat or prevent cancer from coming back (22%), relieve cancer-related symptoms (18%) and help deal with another medical condition (18%). Certain predisposing (i.e., female gender, higher optimism) and need for care factors (i.e., experienced cancer-related symptoms in the past two years, ever had depression/anxiety) were significantly associated with CAM use (p-values

e20511 Background: Breast cancer survivors remain “lost in transition” after active treatment due to changing needs and fragmented care. We sought to describe the patterns of breast cancer survivorship care in a diverse sample to identify factors associated with receipt of survivorship care by medical oncologists (med onc). Methods: We conducted a mailed survey to 2,290 women who resided in the Los Angeles and Detroit SEER registry catchment areas with a confirmed invasive breast cancer case from 6/05-2/07. Women were then surveyed approximately 4 years after diagnosis (n=1,536). On the follow-up survey, women were asked to identify their main provider of breast cancer follow up care: med onc, surgeon, primary care physician (PCP)/other physician. We examined the relationship between report of med onc as main provider of survivor care and: age, race/ethnicity, insurance, tumor stage, receipt of chemotherapy, and care by med onc reported on the baseline survey. Bivariate analyses and logistic regression were used to examine med onc-led survivorship care and the factors above. Results were weighted to account for sample selection and non-response. Results: 858 women had invasive disease at time of diagnosis: 65.2% reported med oncs as the main provider of survivorship care, followed by PCP/other physicians (24.3%) and surgeons (10.5%). Of the women who did not receive chemotherapy, 56% reported med onc-led survivorship care, compared with 79% of chemotherapy recipients. In multivariable logistic regression, black women were less likely than white women to report med onc-led survivorship care (OR 0.49, 95% CI 0.34-0.71). Privately-insured patients were more likely than Medicaid recipients to report med onc-led care (OR 2.00, 95% CI 1.16-3.45). Women with higher disease stage, those who received chemotherapy, and those who saw a med onc at baseline were significantly more likely to report med onc-led survivorship care. Conclusions: The oncology workforce shortage compels the identification of optimal and efficient survivorship care models. A mismatch may exist between chemotherapy receipt and med onc-led care. Our data suggest clear guidelines are needed to direct patients to the providers most appropriate to manage their care.

9601 Background: Concerns exist regarding the employment and financial experiences of cancer survivors and whether they differ by race/ethnicity. Methods: In a longitudinal survey of women reported to the Los Angeles and Detroit SEER registries for nonmetastatic breast cancer, we compared experiences of 4-year survivors by race/ethnicity. Results: Overall, 31% of 1,536 respondents (68% response rate) felt their financial status was worse since diagnosis (63% attributed this to breast cancer). This varied by race/ethnicity: 41% of Spanish-speaking Latinas (SSL), 33% English-speaking Latinas (ESL), 23% blacks (B), and 29% whites (W), p$5000. 12% had medical debt 4 yrs post-diagnosis: 17% of ESL, 14% B, 10% SSL, and 9% W (p=0.01). Minority respondents were more likely to report foregoing medical care due to cost and other privations due to their medical expenses (Table). Overall, 14% felt their employment status was worse since diagnosis, and 61% of these attributed this to breast cancer. 755 worked for pay some time after diagnosis, of whom 56% said it was at least somewhat important to work to keep health insurance (55% of SSL, 65% ESL, 65% B, 50% W, p=0.03); 24% would look for a new job if assured of comparable benefits (45% of SSL, 29% ESL, 22% B, 17% W, p

106 Background: Bone complications of androgen deprivation therapy (ADT) have been described using administrative databases and smaller prospective studies with relatively short follow-up. Prospectively acquired data with long term follow-up are lacking. We assessed the relationship of patient-reported bone health and ADT exposure in a population-based prospective cohort of prostate cancer survivors followed longitudinally for 15 years from diagnosis. Methods: Using PCOS 15 year patient-reported survey data, we identified men with non-metastatic prostate cancer diagnosed from 1994-1995 and followed through 2009-2010. We evaluated subgroups receiving >1 yr, ≤1 yr, and no ADT. We then queried participants regarding the development of osteoporosis, fracture, and use of osteoporosis medications. We assessed the relationship between duration of ADT and bone health outcomes using univariable logistic regression models, and evaluated the association between ADT and osteoporosis medications using multivariable logistic regression adjusted for PSA and geographic location. Results: Among 961 men who completed 15 year surveys, 157 received >1 yr ADT, 120 received ≤1 yr ADT, and 684 did not receive ADT. During the study period, 12 men developed bone metastases and were excluded from the fracture analysis (7 received >1 yr ADT, 2 received ≤1 yr ADT, and 3 did not receive ADT). Men receiving >1 yr ADT were more likely to report osteoporosis (OR 4.29, 95% CI 2.38-7.71) or fracture (OR 1.73, 95% CI 1.04-2.89) than men not receiving ADT. When compared with men not receiving ADT, men receiving >1 yr ADT were more likely to report bone mineral density testing (OR 4.59, 95% CI 3.09-6.83), and bone medication use (OR 3.22, 95% CI 2.19-4.72). Half (50.3%) of men treated for >1 yr ADT reported bone medication use. Among men who reported use of bone medications, 94% reported calcium or vitamin D use and 6% reported bisphosphonate use. Conclusions: Men treated with prolonged ADT (>1 yr) reported higher rates of osteoporosis and fracture at 15 year follow up. Accordingly, they reported more frequent screening and treatment for osteoporosis, with 50% of men receiving prolonged ADT reporting use of bone medications.

9130 Background: Several studies have documented the negative impact of chemotherapy on patients’ cognitive functioning. However, this concern has been understudied among non-Hodgkin Lymphoma (NHL) survivors. Methods: We analyzed survey data from a population-based study of survivors diagnosed with aggressive NHL in Los Angeles County 2-5 years prior to the study; data were collected between 2003-2005. We assessed prevalence of cognitive symptoms (CS) by a single item screener that asked if in the past 6 months survivors experienced problems with memory, attention, or concentration; if yes, we asked if they discussed the problem with a doctor. Frequency of CS was assessed by a 4 item scale that measured how often survivors experienced different CS in the past 4 weeks (none, a little, some, most, all the time). Logistic regression analyses identified correlates of reporting CS (yes/no) as well as discussion with a doctor (yes/no). Analyses were based on data from 358 survivors who had received chemotherapy and saw a doctor for follow-up care in the past year. Results: 134/358 survivors (37%) reported CS, many reporting difficulty some, most, or all the time with memory (64%), attention (58%), concentration (57%), and problem solving (46%). Survivors with higher education, who were not married/partnered, and who had more comorbidities were more likely to report CS (p

6023 Background: Incorporating partners into treatment decision making is an important element of patient-centered care, yet little is known about the role of partners in the decision process. Methods: We surveyed 503 partners of a population-based sample of breast cancer survivor 4 years after diagnosis (RR= 76%, N=382).The outcome was partners’ reports of decision regret. Independent variables included decision making process measures (partners’ reports of sufficient treatment information receipt and sufficient involvement in decision making), race/ethnicity, age, education and income. Multivariable logistic regression was used to assess associations between decision regret and race/ethnicity, controlling for other variables. Results: 49% of partners were white, 14% African American, 15% more-acculturated Latino, and 18% less-acculturated Latino. One quarter (26%) of partners reported that they received insufficient information and one third (35%) desired more involvement in decision-making. Compared to whites, less-acculturated Latino partners more often reported that they received insufficient information (41% vs. 18%, p

9007 Background: Missed paid work during breast cancer treatment is common but less is known about long term consequences of treatment on paid work among survivors. Methods: We surveyed a diverse, ...

e19512 Background: With the growing debate about who should oversee cancer survivors’ medical care, information on who survivors see post-treatment is needed. This study describes non-Hodgkin’s Lym...

510 Background: We examined patterns of endocrine therapy and correlates of non-persistence in a population-based sample of women with breast cancer. Methods: We surveyed 2245 women newly diagnosed...

6032 Background: Little is known about the extent nonclinical patient factors are considered by oncologists when deciding to recommend adjuvant breast cancer chemotherapy and whether oncologist and...

9098 Background: Worry about recurrence (worry) is a persistent concern of women with breast cancer; however, little is known about racial/ethnic differences or mechanisms in the health care experi...

6016 Background: There has been no research evaluating provider level variation in patient appraisal of 3 key indicators of quality breast cancer care defined by ASCO: timely access, care coordinat...