PURPOSE:Oral adjuvant endocrine therapy (AET) reduces the risk of cancer recurrence and death for women with hormone receptor-positive (HR+) breast cancer. Because of adverse symptoms and socioecologic barriers, AET adherence rates are low. We conducted post hoc analyses of a randomized trial of a remote symptom and adherence monitoring app to evaluate characteristics associated with higher app use, satisfaction, and how app use was associated with AET adherence. METHODS:Patients prescribed AET were randomly assigned to receive one of three intervention conditions: app, app + feedback, or enhanced usual care. Baseline and 6-month follow-up surveys, app use, and pillbox-monitored AET adherence data for app and app + feedback participants were used. Logistic regression evaluated the association between sociodemographic/clinical characteristics and app utilization and satisfaction, and how app use was associated with AET adherence (>80%). RESULTS:Overall, 163 women with early-stage HR+ breast cancer were included; 35.0% had high app use (≥75% of weeks enrolled). No sociodemographic characteristics were associated with app use. Satisfaction with the app was higher among those who were younger (88.9% for age 31-49 years v 54.9% for age 65+ years, P < .001), identified as White (76.8% v 60.1% for Black, P = .045), had lower health literacy (85.4% v 68.2% with higher health literacy, P = .017), or were nonurban residents (85.7% v 68.6% for urban, P = .021). Most participants (90.3%) with high app use were AET-adherent compared with 66.8% for those with lower app use (P < .001). CONCLUSION:Use of a remote monitoring app was similar across sociodemographic characteristics, and more frequent app use was associated with a higher likelihood of 6-month AET adherence. Encouraging women to monitor medication adherence and communicate adverse symptoms could improve AET adherence.

ImportanceAdjuvant endocrine therapy (AET) use among women with early-stage, hormone receptor-positive breast cancer reduces the risk of cancer recurrence, but its adverse symptoms contribute to lower adherence. ObjectiveTo test whether remote monitoring of symptoms and treatment adherence with or without tailored text messages improves outcomes among women with breast cancer who are prescribed AET. Design, Setting, and ParticipantsThis nonblinded, randomized clinical trial (RCT) following intention-to-treat principles included English-speaking women with early-stage breast cancer prescribed AET at a large cancer center with 14 clinics across 3 states from November 15, 2018, to June 11, 2021. All participants had a mobile device with a data plan and an email address and were asked to use an electronic pillbox to monitor AET adherence and to complete surveys at enrollment and 1 year. InterventionsParticipants were randomized into 3 groups: (1) an app group, in which participants received instructions for and access to the study adherence and symptom monitoring app for 6 months; (2) an app plus feedback group, in which participants received additional weekly text messages about managing symptoms, adherence, and communication; or (3) an enhanced usual care (EUC) group. App-reported missed doses, increases in symptoms, and occurrence of severe symptoms triggered follow-ups from the oncology team. Main Outcomes and MeasuresThe primary outcome was 1-year, electronic pillbox-captured AET adherence. Secondary outcomes included symptom management abstracted from the medical record, as well as patient-reported health care utilization, symptom burden, quality of life, physician communication, and self-efficacy for managing symptoms. ResultsAmong 304 female participants randomized (app group, 98; app plus feedback group, 102; EUC group, 104), the mean (SD) age was 58.6 (10.8) years (median, 60 years; range, 31-83 years), and 60 (19.7%) had an educational level of high school diploma or less. The study completion rate was 87.5% (266 participants). There were no statistically significant differences by treatment group in AET adherence (primary outcome): 76.6% for EUC, 73.4% for the app group (difference vs EUC, -3.3%; 95% CI, -11.4% to 4.9%; P = .43), and 70.9% for the app plus feedback group (difference vs EUC, -5.7%; 95% CI, -13.8% to 2.4%; P = .17). At the 1-year follow-up, app plus feedback participants had fewer total health care encounters (adjusted difference, -1.23; 95% CI, -2.03 to -0.43; P = .003), including high-cost encounters (adjusted difference, -0.40; 95% CI, -0.67 to -0.14; P = .003), and office visits (adjusted difference, -0.82; 95% CI, -1.54 to -0.09; P = .03) over the previous 6 months compared with EUC participants. Conclusions and RelevanceThis RCT found that a remote monitoring app with alerts to the patient's care team and tailored text messages to patients did not improve AET adherence among women with early-stage breast cancer; however, it reduced overall and high-cost health care encounters and office visits without affecting quality of life. Trial RegistrationClinicalTrials.gov Identifier: NCT03592771

45 Background: Despite the efficacy of oral adjuvant endocrine therapy (AET) for improving survival among women with hormone receptor-positive early-stage breast cancer, adherence rates remain low. Our THRIVE study (NCT03592771) investigated the effectiveness of a mobile health remote monitoring app with and without tailored educational messages on AET adherence among women with breast cancer and found no statistically significant treatment effects on 1-year AET adherence overall. Health literacy contributes to patients’ understanding of their treatment plans. We conducted a post hoc analysis of the THRIVE study to examine if health literacy moderates the effectiveness of this intervention. Methods: This non-blinded randomized controlled trial included women with early-stage breast cancer prescribed AET at a large cancer center with 14 clinics across three states. Participants used a pillbox to electronically monitor AET adherence for 1 year and completed surveys at enrollment. Consented participants were randomized into (1) “App”, receiving access to the study adherence and symptom monitoring app for 6 months, with increasing/severe symptoms and missed doses reported in the app triggering follow-ups from the oncology team; (2) “App+Feedback”, receiving additional weekly text messages about managing symptoms, adherence, and communication for 6 months; or (3) “Enhanced Usual Care (EUC).” The primary outcome was 1-year AET adherence captured with the pillbox (≥80% Proportion of Days Covered [PDC] vs. < 80%). The enrollment survey captured participant’s sociodemographic characteristics, including race and ethnicity, education, household income, and health literacy. We used a linear probability model to measure the interaction between the study arm and health literacy on AET adherence. Results: Among 304 women randomized (104 EUC, 98 App, and 102 App+Feedback), the 12-month follow-up retention rate was 88% (n = 266) and 19.4% reported low health literacy at enrollment. Low health literacy was more prevalent among Black (29.4%) vs. White (13.4%) participants (p < 0.001), those with incomes below the federal poverty level (34.4% vs. 17.2% of those with higher incomes, p = 0.02), and those with only a high school degree or lower education (31.7%) vs. with those with some college or higher education (16.4%, p < 0.01). In the low health literacy group, 80.0% of App+Feedback were AET adherent vs. 42.1% of EUC, a 37.9 percentage point (ppt) difference (95% CI: 4.1 to 71.7, p = 0.03); in higher literacy group adherence, 47.1% of App+Feedback vs. 59.0% of EUC were adherent, a -11.8 ppt difference (95% CI: -27.9 to 4.3, p = 0.15). Conclusions: A remote monitoring app with tailored educational text messages led to higher 1-year AET adherence among participants with low health literacy, but not for those with high health literacy. Clinical trial information: NCT03592771 .

Tables of ICD9 and ICD10 codes used.

PDF file - 46K, Overall Survival in Prespecified Subgroups

PDF file - 55K, Study Drug Dosing (Safety Population)

PDF file - 44K, Sensitivity Analyses of Progression-free Survival

6528 Background: Symptom burden may contribute to racial differences in cancer treatment adherence and survival. Evidence on changes in symptom burden during chemotherapy and whether these differ by race is scarce. We used patient reported outcomes data collected before and after breast cancer chemotherapy initiation to compare symptom burden by race. Methods: Using electronic medical records of a large cancer center in the southern region of the US, we identified Black and White women diagnosed with stage I-III, hormone-receptor positive breast cancer from January 2007 to December 2015. A tablet-based platform [ConcertAI] was used to collect patient reported symptoms at the point of care. We included patients with at least one completed symptom report before and during chemotherapy. We focused on two standardized composite scores – physical symptoms and treatment side-effect (mean of 50 and standard deviation of 10), and calculated changes in symptoms using the closest report before chemotherapy and the most severe score reported during chemotherapy. Patients with a 10-point increase were classified as having a clinically meaningful increase in symptom burden. We used Oaxaca-Blinder decomposition to quantify racial differences in symptom burden change explained by baseline characteristics. These included baseline symptom scores, sociodemographic characteristics (age, regional level household income and education, state) and clinical characteristics (cancer stage and primary chemo regimen). Results: Among 1,167 included patients, Black women (30%) were younger (52 vs. 55 years old, p<.001), more likely to live in areas with lower median household income and less education, and reported most severe scores about 2 weeks later than White women ( p<.05). They were also more likely to report a 10-point increase in symptom burden for physical (68.5% vs. 61.2%, p=.017) and side-effects symptoms score (49.0% vs. 41.4%, p=.015). This was driven by larger increases in selected individual symptoms among Black women, such as sweating, itching, and numbness (under physical symptom score), and hair loss and taste change (under side-effect score). Decomposition analyses showed that baseline characteristics (especially primary chemo regimen) explained 79.2% ( p=.002) and 35.2% ( p=.131) of the increased probability of Black women reporting a 10-point increase in physical symptom and side-effects scores respectively. Conclusions: Black women with early-stage breast cancer were more likely to report a clinically meaningful increase in treatment side-effects and physical symptoms during chemotherapy compared to White women. Differences by race in physical symptoms scores were mostly explained by baseline characteristics. Future studies should examine whether racial differences in symptom burden translate into differences in treatment adherence and mortality.

Importance:Adjuvant endocrine therapy (AET) reduces breast cancer recurrence, but symptom burden is a key barrier to adherence. Black women have lower AET adherence and worse health outcomes than White women.Objective:To investigate the association between symptom burden and AET adherence differences by race.Design, Setting, and Participants:A retrospective cohort study using electronic health records with patient-reported data from a large cancer center in the US. Patients included Black and White women initiating AET therapy for early-stage breast cancer from August 2007 to December 2015 who were followed for 1 year from AET initiation. Sixty symptoms classified into 7 physical and 2 psychological symptom clusters were evaluated. For each cluster, the number of symptoms with moderate severity at baseline, and symptoms with 3-point or greater increases during AET were counted. Adherence was measured as the proportion of days covered by AET during the first-year follow-up. Multivariable regressions for patients' adherence adjusting for race, symptom measures, sociodemographic characteristics, and clinical characteristics were conducted. Kitagawa-Blinder-Oaxaca decomposition was used to quantify racial differences in adherence explained by symptoms and patient characteristics. Analyses were conducted from July 2021 to January 2022.Exposures:Physical and psychological symptoms at baseline and changes during AET.Results:Among 559 patients (168 [30.1%] Black and 391 [69.9%] White; mean [SD] age 65.5 [12.1] years), Black women received diagnoses younger (mean [SD] age at diagnosis, 58.7 [13.7] vs 68.5 [10.0] years old) than White women, with more advanced stages (30 Black participants [17.9%] vs 31 White participants [7.9%] had stage III disease at diagnosis), and lived in areas with fewer adults attaining high school education (mean [SD], 78.8% [7.8%] vs 84.0% [9.3%]). AET adherence in the first year was 78.8% for Black and 82.3% for White women. Black women reported higher severity in most symptom clusters than White women. Neuropsychological, vasomotor, musculoskeletal, cardiorespiratory, distress, and despair symptoms at baseline and increases during the follow-up were associated with 1.2 to 2.6 percentage points decreases in adherence, which corresponds to 4 to 9 missed days receiving AET in the first year. After adjusting for psychological symptoms, being Black was associated with 6.5 percentage points higher adherence than being White.Conclusions and Relevance:In this cohort study, severe symptoms were associated with lower AET adherence. Black women had lower adherence rates that were explained by their higher symptom burden and baseline characteristics. These findings suggest that better symptom management with a focus on psychological symptoms could improve AET adherence and reduce racial disparities in cancer outcomes.

PURPOSE: Cancer prevalence and outcomes data, necessary to understand disparities in transgender populations, are significantly hampered because gender identity data are not routinely collected. A database of clinical data on people with cancer, CancerLinQ, is operated by the ASCO and collected from practices across the United States and multiple electronic health records. METHODS: To attempt to identify transgender people with cancer within CancerLinQ, we used three criteria: (1) International Classification of Diseases 9/10 diagnosis (Dx) code suggestive of transgender identity; (2) male gender and Dx of cervical, endometrial, ovarian, fallopian tube, or other related cancer; and (3) female gender and Dx of prostate, testicular, penile, or other related cancer. Charts were abstracted to confirm transgender identity. RESULTS: Five hundred fifty-seven cases matched inclusion criteria and two hundred and forty-two were abstracted. Seventy-six percent of patients with Dx codes suggestive of transgender identity were transgender. Only 2% and 3% of the people identified by criteria 2 and 3 had evidence of transgender identity, respectively. Extrapolating to nonabstracted data, we would expect to identify an additional four individuals in category 2 and an additional three individuals in category 3, or a total of 44. The total population in CancerLinQ is approximately 1,300,000. Thus, our methods could identify 0.003% of the total population as transgender. CONCLUSION: Given the need for data regarding transgender people with cancer and the deficiencies of current data resources, a national concerted effort is needed to prospectively collect gender identity data. These efforts will require systemic efforts to create safe healthcare environments for transgender people.

Importance:Race disparities persist in breast cancer mortality rates. One factor associated with these disparities may be differences in symptom burden, which may reduce chemotherapy tolerance and increase early treatment discontinuation.Objectives:To compare symptom burden by race among women with early-stage breast cancer before starting chemotherapy and quantify symptom differences explained by baseline characteristics.Design, Setting, and Participants:A cross-sectional analysis of symptom burden differences by race among Black and White women with a diagnosis of stage I to III, hormone receptor-positive breast cancer who had a symptom report collected before chemotherapy initiation in a large cancer center in the southern region of the US from January 1, 2007, through December 31, 2015. Analyses were conducted from November 1, 2019, to March 31, 2021. Blinder-Oaxaca decomposition was used, adjusting for baseline sociodemographic and clinical characteristics.Main Outcomes and Measures:Four symptom composite scores with a mean (SD) of 50 (10) were reported before starting chemotherapy (baseline) and were derived from symptom items: general physical symptoms (11 items), treatment adverse effects (8 items), acute distress (4 items), and despair (7 items). Patients rated the severity of each symptom they experienced in the past week on a scale of 0 to 10 (where 0 indicates not a problem and 10 indicates as bad as possible).Results:A total of 1338 women (mean [SD] age, 54.6 [11.6] years; 420 Black women [31.4%] and 918 White women [68.6%]) were included in the study. Before starting chemotherapy, Black women reported a statistically significantly higher (ie, worse) symptom composite score than White women for adverse effects (44.5 vs 43.8) but a lower acute distress score (48.5 vs 51.0). Decomposition analyses showed that Black patients' characteristics were associated with higher symptom burden across all 4 scores. However, these differences were offset by relatively greater, statistically significant, unexplained physical, distress, and despair symptom reporting by White patients.Conclusions and Relevance:In this study, before starting chemotherapy, Black patients with early-stage breast cancer reported significantly higher burden for symptoms that may be exacerbated with chemotherapy and lower distress symptoms compared with White patients. Future studies should explore how symptoms change before and after treatment and differ by racial/ethnic groups and how they are associated with treatment adherence and mortality disparities.

132 Background: Performance status is used to characterize patient ability to tolerate chemotherapy and as a selection criterion for clinical research. Poor performance status can exclude patients from clinical trial participation. Further, African American cancer patients are underrepresented in cancer clinical trials. The study purpose was to document performance status at the initial patient visit to a community oncology practice and to explore racial disparities between White and Black patients. Methods: This study used a retrospective, observational design with ePRO collected via the Patient Care Monitor™ (PCM). All study data were collected as part of routine clinical care at a community oncology practice during 1/2019–11/2019. An Eastern Cooperative Oncology Group (ECOG) score was automatically calculated after patients at an initial clinic visit completed a 1-item question that assessed performance status via e-tablet. Results: 6,613 patients completed the PCM survey (mean age 59; 33% male/67% female; 55.4% White, 38% Black). Cancer type was known for a subset of patients (22% breast, 9% hematologic, 4% lung, 5% colorectal, 3% prostate, 11% other types). The average ECOG score for the total sample was 0.97. 50% indicated they were able to complete their normal daily activities without any restriction; 26.9% were able to complete their normal daily activities and some light work. In contrast, 10.3% indicated they could take care of themselves, but could not work and are in bed/chair less than half the day. 10.3% could take care of themselves sometimes but could not work and are in bed/chair more than half the day. 4.5% indicated they could not take care of themselves and were in bed/chair almost always. When assessing racial differences between those self-identifying as White or Black/African American, average ECOG score was higher in Black patients [Mean(SD) = 1.03(1.24)] when compared to White patients [Mean(SD) = 0.93(1.14)] (p = 0.003). We observed a higher percentage of Black patients reported not being able to take care of themselves (51.9% Black v. 41.0% White). In contrast, a higher percentage of White patients reported being able to complete all daily activities without restriction (38.3% Black v. 54.5% White). Conclusions: This study shows significant racial disparities in performance status among patients seen at a community oncology practice with Black patients exhibiting significant worse performance status than White patients. These findings have implications for disparities in treatment outcomes and racially biased access to clinic trials.

e19191 Background: ASCO has implemented the Quality Oncology Practice Initiative (QOPI), a certification program established to evaluate oncology practice performance. Also, a growing number of accreditation (JCAHO) and merit-based organizations (MIPS) maintain falls risk assessment standards. Practices often lack the necessary resources to comply with required metric reporting standards. The study purpose was to document the effectiveness of using an electronic Patient Reported Outcome (ePRO) system to facilitate compliance with a core QOPI standard, documentation of smoking status by second office visit, and with JCAHO and MIPS falls risk assessment. Methods: This study used a retrospective, observational design with ePRO collected via the Patient Care Monitor (PCM), a web-based ePRO system linked to electronic medical record data. All study data were collected as part of routine clinical care at a community oncology practice during an 11-month interval (1/2019–11/2019). Patients at an initial clinic visit completed a tobacco usage survey and a brief falls risk survey on the PCM platform via a handheld e-tablet. Results: Overall, 6,613 unique patients completed the PCM survey (mean age 59; 33% male/67% female; 55.4% White, 38% Black). Cancer type was known for a subset of patients (22% breast, 9% hematologic, 4% lung, 5% colorectal, 3% prostate, 11% other types). Across the collected PRO measures, there was an over 98% completion rate with only 1-2% missing data. A relatively significant proportion (51%) indicated they had never used tobacco products and 15% indicated that they were current users. Among patients who ever used tobacco products, 34% indicated they smoked cigarettes, 4% smoked cigars, and 3% used electronic cigarettes. Over a fifth of patients (22%) indicated they had at least one fall over six months; 10% indicated having experienced one fall; 6% indicated two falls; 6% indicated 3 falls or more. 17% indicated they use an ambulatory aid and 12% reported a recent fall within the past 3 months. Conclusions: This study demonstrates that using an ePRO system is an effective way to screen for tobacco usage and falls risk and can be used to: 1) monitor health-related behaviors to enhance physician-patient communication; 2) provide an audit trail for QOPI, JCAHO, MIPS and other quality metric reporting. Automated collection of PRO data allows the healthcare team to focus their clinical time on patients showing increased risk. Overall, an ePRO system contributes to creating a culture of excellence at community oncology practices.

e19318 Background: ECOG PS is a prognostic indicator of outcomes, and scores of 0-1 (good ECOG PS) are often required for clinical trial enrollment. Patients treated in non-trial settings often lack ECOG PS scores limiting the ability of Real World Data from these patients to be used in external control arms (ECAs) or to provide optimal specificity for clinical effectiveness research. Machine Learning can be used to impute ECOG PS scores from other clinical data at various points during treatment. Methods: We developed a series of models using logistic regression (LR) or XGBoost (XGB) that impute ECOG PS at initial diagnosis, metastatic diagnosis and final evaluation using a curated Non-Small Cell Lung Cancer cohort of 31,425 patients with at least one ECOG PS score. Results: AUC-ROC values of up to 0.81 could be obtained for imputing a patient’s final ECOG PS, with lower AUC values when imputing ECOG PS at initial and metastatic diagnosis using large numbers (i.e. thousands) of features. We developed more interpretable models with 110 or 40 features with reduced but still satisfactory AUC, with accuracy of predicting good ECOG PS scores of around 80%. Key features were obtained from lab tests, physical exams, comorbidities, medications, age and metastatic status. The table below shows the results of several of these models. Where the models misclassify ECOG PS, the error was rarely greater than 1 grade. Conclusions: ECOG PS is subjective, suggesting that ML based cohort assignment will be sufficiently accurate to support their use in research. Further work will be required to assess if the ML predicted cohorts have different outcomes. [Table: see text]

171 Background: ASCO has implemented the Quality Oncology Practice Initiative (QOPI), a certification program established to evaluate oncology practice performance. However, practices often lack the necessary resources to meet the required standards and fulfill required metric reporting. The study purpose was to document the effectiveness of using an electronic Patient Reported Outcome (ePRO) system to facilitate compliance with two QOPI standards: pain assessment and depression screening. Methods: This study used a retrospective, observational design with ePRO and clinical data collected via the Patient Care Monitor (PCM), a web-based ePRO system linked to electronic medical record data. All study data were collected as part of routine clinical care at a community oncology practice during a 4-month interval (1/1–4/26/19). Patients at an initial oncology clinic visit completed the PCM Core Symptom survey and the PHQ2/PHQ9 to screen for pain and depression respectively via a handheld e-tablet. Results: Overall, 10,449 patients completed the PCM survey (mean age 63; 30% male/70% female; 62% White, 34% Black). Cancer type was known for 53% of all patients (22% breast, 9% hematologic, 4% lung, 5% colorectal, 3% prostate, 11% other types). Fifteen percent of patients (n = 1584) met the pain threshold (scored 7 or above on a 0-10 scale) that signaled an alert to the physician for further evaluation. About 3% of patients (n = 413) skipped the pain question on the PCM Core Symptom survey. With respect to depression, 5% (n = 559) reported a PHQ2 threshold level that triggered the remaining PHQ9 items. Among those at higher risk for depression, 16% (n = 92) met criteria for moderate to severe depression warranting further clinical evaluation and the PCM signaled a real-time alert to the physician in advance of the patient consultation. Approximately 4% (n = 413) skipped the PHQ2. Conclusions: This study demonstrates that using an ePRO system is an effective way to screen for pain and depression and can be used to: 1) monitor symptoms to enhance physician-patient communication during the consultation; 2) provide an audit trail for QOPI metric reporting. Overall, an ePRO system contributes to creating a culture of excellence at community oncology practices.

9110 Background: Immune Checkpoint Inhibitors (ICIs) were first approved for the treatment of aNSCLC in 2014, and since this time have seen rapid adoption in the marketplace. We sought to describe the characteristics of patients with aNSCLC receiving ICIs in the real-world, as well as to examine treatment patterns and outcomes in the time since initial ICI approval. Methods: We conducted a retrospective, observational cohort study using statistically de-identified data from January 2011 to November 2018 in CancerLinQ, ASCO’s real-world oncology database. Adult patients with a curated diagnosis of Stage III or IV NSCLC who received ≥1 dose of an ICI and had ≥2 clinical visits were eligible for inclusion. Stage III patients were excluded if they received any local therapy < 1 year prior to receiving ICI. Patients were also excluded if they received ICI prior to the first FDA approval date. Demographic and clinical characteristics of aNSCLC patients receiving ICI are reported. Outcomes including time to treatment discontinuation (TTD), time to next treatment (TTNT), real-world progression free survival (rwPFS) and overall survival (OS) were examined via the Kaplan Meier method. Results: Among 2,425 aNSCLC ICI patients included in this analysis, median age was 68.0 years (IQR 60.7, 75.2], 54% were male and 73% of patients were white. Non-squamous histology accounted for 64% of aNSCLC ICI users, and 81% had Stage IV disease. Eastern Cooperative Oncology Group (ECOG) performance status was 0-1 in 77% and 2+ in 23% of patients, and 70% were current or former smokers. The majority (75%) of patients received ICI as second-line or later therapy. Treatment outcomes and survival are reported in the Table. Conclusions: This analysis demonstrates that aNSCLC patients receiving ICI therapy in the real-world are older than what was reported in some clinical trials, though survival outcomes were similar. Further research to examine impact of covariates on outcomes is warranted. [Table: see text]

6583 Background: Although pts with AD are routinely excluded from ICI clinical trials, evidence suggests they may be receiving ICI therapy once approved. We sought to understand the prevalence of AD among all pts receiving ICIs in real world clinical care, as well as in advanced non-small cell lung cancer (aNSCLC) alone, and to describe the characteristics of ICI pts with and without evidence of AD. Methods: We conducted a retrospective, observational cohort study using statistically de-identified data from January 2011 to November 2018 in CancerLinQ, ASCO’s real-world oncology database. Adult pts who received ≥ 1 dose of an ICI and had ≥ 2 clinical visits were eligible for inclusion. A sub-analysis examining only aNSCLC pts was also carried out. To reduce the likelihood of capturing pts who may have been on a clinical trial, pts were excluded if they received the ICI prior to its first FDA approval date. AD status was determined by the presence of select ICD-9/ICD-10 codes or a medication used to treat autoimmune disease (including steroids) prior to ICI treatment start date. Symphony claims data were linked to CLQ via tokenization to build out cohorts. Characteristics of pts with and without autoimmune disease were compared using Chi-square or Fisher’s exact tests. Results: Prevalence of AD was 23% (538/2425 pts) in the aNSCLC population and 27% (3407/12712 pts) in the all ICI patient population. Median age did not differ between AD pts and those with no evidence of AD (All ICI: 67.6 v 67.3 years; aNSCLC: 68.5 v 67.9). AD pts were more likely to be female (All ICI: 46% v 40%, p < 0.001; aNSCLC: 55% v 44%, p < 0.001). Among all ICI pts, AD pts were less likely to be Stage IV (62% v 65%) or to have melanoma (4.6% versus 8.7%) compared to pts with no evidence of AD. The most common ADs among all ICI and aNSCLC patients were glucocorticoid deficiency (6.3% and 3.9%), rheumatoid arthritis (4.2% and 5.8%), and sacroiliitis (2.7% and 3.9%), respectively. Conclusions: This analysis of real-world data finds that a large proportion of pts receiving ICI may have pre-existing AD. Further examination is warranted to examine how AD status may impact outcomes.

5 Background: The American Society of Clinical Oncology recommends that providers encourage early advance care planning with their patients; yet, many cancer patients do not have advance directives (ADs). A potential reason for low AD rates is inadequate communication between the provider and patient. To address this gap, we developed an outpatient clinic AD initiative for any stage cancer patients via an ePRO system. The AD module was designed to ensure patients are aware of ADs and to assess whether or not the patient had ADs, specifically a Living Will and/or a durable power of attorney and their interest in receiving information about ADs. The study purpose was to assess patient AD status at the patient’s initial visit to an oncology clinic. Methods: This study used a retrospective, observational design that involved use of PRO data collected via the Patient Care Monitor (PCM), a web-based ePRO system linked to electronic medical record data. All study data were collected as part of routine clinical care. All patients at an initial visit to an oncology clinic completed the PCM survey, including the AD module, via a handheld e-tablet. Results: Overall, 18,239 patients completed the AD module (mean age 62; 32% male/68% female; 55% married; 60% White, 36% Black). One third of all patients (30%) reported having an AD at the time of the initial visit to the oncology clinic, specifically 11% indicated having a Living Will and 19% a durable power of attorney. The remaining two-thirds (70%) indicated either NOT having an AD (58%) or not knowing if they had an AD (12%). Patients with ADs were more likely to be older (M= 69 v. 59 yrs) and White (77% v 20% Black) ps < .0001. Of the patients without ADs, 10% requested more information on ADs. These patients were more likely to be older (M= 61 v. 58 yrs), female (68% v 32%), Black (54% v 40% White) and non-married/non-partnered (53% v 47%); ps < .005. Conclusions: This study demonstrates that the majority of patients do not have ADs at the time of an initial visit to a community-based oncology practice. Using an ePRO system can be an innovative and non-threatening way to identify unmet needs of patients and link them to resources for developing advance directives.

The clinical benefit of radiation therapy for patients with locally advanced stage III (N2) Non-Small Cell Lung Cancer (NSCLC) treated with surgery remains unclear due to a paucity of level 1 evidence. This study examined real-world evidence using a cohort of NSCLC patients from a multi-institution dataset to evaluate outcomes in patient treated with surgery alone, neoadjuvant or adjuvant radiation therapy. We extracted a sample of 415 fully-abstracted patients with pathologic stage III (N2) NSCLC defined at the initial surgical intervention performed within 6 months of initial diagnosis. Two- and five-year survival outcomes were evaluated for patients who received radiation neoadjuvant, adjuvant (PORT), or no radiation therapy. Sub-analyses and normalization of treatment groups were performed including use of chemotherapy, targeted therapy and IO therapy as well as age and gender using Cox proportionate hazards. Of the 415 patients evaluated, 54% (n=206) were treated with surgery alone, 16% (n=66) neoadjuvant radiation therapy and 34% (n=143) adjuvant radiation therapy. Chemotherapy use for patients across groups was balanced, 68% (n=140), 83% (n=56), 72% (n=103) for patient treated with surgery, neoadjuvant or adjuvant radiation therapy, respectively. Patients who received neoadjuvant radiation therapy had a 45% reduction in 2-year mortality risk (HR = .55, 95% CI 0.35 - 0.89), compared to patients treated with surgery alone (HR = 0.59, 95% CI .37 - 0.93). Radiation therapy did not improve 5-year mortality risk except in female patients, who show a 48% reduction in mortality risk at 5-years from radiation therapy pre- or post-operatively (HR = .52, 95% CI .29 - .94). There was no statistically significant difference in survival outcomes for patients treated with adjuvant radiation therapy, compared to surgery alone. Analysis of 415 abstracted patients this dataset suggests that the addition of preoperative radiation therapy improves 2 and 5-year survival for patients with pathologic stage III (N2) NSCLC, compared to patients treated with surgery alone. A significant survival benefit was not observed for adjuvant radiation therapy, however surgical margin status, performance status, and incidence of comorbidities was not included in the analysis due to the lack of granularity of the retrospective data.Abstract 3222; Table 1Surgery Only / No Radiation TherapyNeoadjuvant Radiation TherapyAdjuvant Radiation TherapyAll PatientsTotal (n)2-Year OS5-year OSTotal (n)2-Year OS5-Year OSTotal(n)2-Year OS5-Year OSTotal (n)2-Year OS5-Year OSAll patients20677%67%6692%86%14384%76%41582%71%Age >= 6015177%66%4994%78%10083%75%30082%71%Age < 605578%67%1712%71%4386%77%11583%71%Male10977%64%3090%70%7077%66%20979%66%Female9777%69%3694%81%7390%85%20685%77%Chemotherapy14076%74%5591%73%10380%70%29880%68%Neoadjuvant chemotherapy3093%87%4489%76%1090%80%8590%81%No chemotherapy6680%71%11100%91%4095%90%11787%79% Open table in a new tab