CACHEXIO: Evaluation of body composition changes and immunotherapy in patients with metastatic cancer

209

Background: Given body composition predicts toxicity for patients receiving cytotoxic chemotherapy, we explored changes in body composition and biomarkers as predictors of immune-related adverse events (irAEs) and health care utilization. Methods: We conducted a longitudinal study of patients with metastatic solid tumor receiving immunotherapy (07/2014-10/2017). Eligible patients had a computed tomography (CT) scan prior to first-line immunotherapy with at least two additional CT scans at three, six or nine months after immunotherapy initiation. We analyzed body composition using cross-sectional CT scans at the third lumbar vertebra. We utilized mixed effect linear regression models to assess longitudinal changes in body composition (weight, skeletal muscle, total adipose). We examined associations of baseline body composition and biomarkers (albumin, neutrophil-lymphocyte ratio (NLR)) with the incidence of irAEs and healthcare utilization (hospitalizations/ED visits) using logistic regression. Results: Of 140 patients treated with immunotherapy, 61 met inclusion criteria. The majority (80%) received prior chemotherapy and the most common malignancies included lung (26%), head and neck (20%), and melanoma (20%). We found that one-third (n=19) experienced an irAE and colitis (53%) was the most common irAE. Patients experienced substantial weight loss over time (B= -1.88, p<0.001) with a decrease both in skeletal muscle (B= -3.08, p=0.001) and total adipose tissue (B =-50.44, p<0.001). We found greater skeletal muscle at baseline was associated with lower risk of health care utilization (OR 0.98, 95% CI: 0.965-0.998, p=0.03). We observed no association with biomarkers and/or body composition variables with irAEs or health care utilization. Conclusions: Patients with metastatic cancer receiving immunotherapy lose weight including skeletal muscle and adipose tissue. Aside from higher baseline skeletal muscle predicting less health care utilization, we did not observe any other associations between body composition changes and irAEs or health care utilization.

Efficacy of intravenous (IV) NEPA, a fixed NK1/5-HT3 receptor antagonist combination, for prevention of emesis following highly emetogenic chemotherapy (HEC)

194

Background: An IV formulation of NEPA (fixed combination of fosnetupitant 235 mg and palonosetron 0.25 mg) was recently approved offering clinicians an additional antiemetic treatment option. Approval of IV NEPA was based on showing pharmacokinetic bioequivalence and comparable safety to oral NEPA. This post-hoc analysis presents the efficacy of IV NEPA relative to that of oral NEPA and other NK1 RAs in the HEC setting. Methods: Data is compiled from 3 pivotal NEPA registration studies in a total of 951 adult chemotherapy-naïve patients with solid tumors undergoing predominantly cisplatin-based HEC. All studies had similar inclusion/exclusion criteria. IV NEPA was administered as a single 30-min infusion and a single capsule of oral NEPA was given prior to HEC. All patients received dexamethasone (DEX) on Days 1-4. Data is also compiled from registration HEC trials on other NK1 RA (aprepitant, fosaprepitant, rolapitant) regimens. No emesis rates are summarized for the overall phase (0-120h) of the initial cycle of HEC for each study and also pooled for oral NEPA. No formal statistical comparisons were performed. Results: The overall no emesis rate was 84.2% for IV NEPA and 81.6% for oral NEPA for the pooled studies. The no emesis rates for aprepitant, fosaprepitant and rolapitant regimens ranged from 66.0% to 77.7%. Clinical trial information: NCT02517021. Conclusions: Both IV and oral formulations of NEPA represent effective and convenient single-dose prophylactic antiemetics targeting two distinct CINV pathways, offering comparable options for clinicians/patients in different settings.[Table: see text]

What the HEC? Physician variation and attainable compliance targets in antiemetic prophylaxis

74

Background: U.S. National Antiemetic Guidelines recommend upfront triple prophylaxis (NK1 receptor antagonist (RA) + 5HT3 RA + dexamethasone) for patients receiving highly emetogenic chemotherapy (HEC), including carboplatin AUC ≥ 4 per 2017 guidelines. While existing data show gaps in guideline compliance, variation between individual physicians is less studied, and a realistic target compliance rate remains unknown. Methods: In a large electronic health record database (IBM Explorys), we identified HEC courses of therapy initiated from 2012 to 2017. Guideline compliance was defined as triple prophylaxis at chemotherapy initiation. Patient courses for ≥ 7 day cycles of cisplatin or anthracycline + cyclophosphamide (AC), or carboplatin (≥ 14 day cycles as a proxy for AUC ≥ 4) were ascribed to oncologists based on encounter frequency. We then ranked physicians treating ≥ 5 HEC courses and evaluated guideline compliance and individual physician variation. Results: In total, 10,074 HEC courses were identified and attributed to 451 unique physicians. Overall antiemetic guideline compliance with cisplatin and AC averaged 68% and 81% respectively. When ranked by compliance, the top 20% of physicians were 2.5 - 1.5 times as compliant as the bottom 20% (cisplatin 100% vs 40%; AC 100% vs 67%). For cisplatin, 32% of physicians had > 90% compliance; the remaining 68% were evenly distributed from 0 - 90%. For AC, 56% of physicians had > 90% compliance, and another 14% had 80 - 90%; the remaining 30% were evenly distributed. For carboplatin, 62% of physicians had ≤ 10% compliance, and another 17% had 11 - 20%; however, the majority of these data preceded guideline inclusion of carboplatin AUC ≥ 4 as HEC. Rates were independent of course volume per physician. Conclusions: Considerable physician-level variation exists in triple antiemetic prophylaxis guideline adherence for HEC. Hundreds of physicians had > 90% compliance with guidelines, suggesting 90% is a realistic target. However, the majority exhibited substantial gaps in NK1 RA use in HEC, placing patients unnecessarily at risk for CINV. Interventions are needed to bolster triple antiemetic prophylaxis in HEC, perhaps especially for carboplatin.

FIT: Functional and imaging testing for patients with metastatic cancer

208

Background: Despite multiple cancer cachexia (CC) trials evaluating novel agents, the FDA has not approved a single drug to date. One key challenge in CC trials is selection of endpoints. The aim of this study was to explore changes in body composition and associations with functional and patient-reported outcomes (PROs) to clarify CC trial endpoint selection. Methods: We identified metastatic solid tumor cancer patients receiving cancer-directed therapies at a single cancer center (2016-2018). Patients completed all assessments at study enrollment and 3 months from enrollment. We analyzed body composition utilizing cross-sectional computed tomography (CT) scans at the third lumbar vertebra. Functional assessments included the 6-minute walk test (6MWT), Timed Up-and-Go (TUG) test, and Short Physical Performance Battery (SPPB). PROs included the Functional Assessment of Anorexia/Cachexia Therapy (FAACT) and Functional Assessment of Cancer Therapy Fatigue (FACT-F). We examined changes in body composition and functional assessments from enrollment to 3 months using paired t-tests. We utilized linear regression models to assess the relationship between changes in body composition and changes in functional assessment adjusting for age and sex. Results: A total of 57 patients completed baseline assessments; 19 patients did not complete 3-month assessments (5 died, 1 hospice, 13 withdrew). Of the 38 patients with complete data (mean age 61.8 years, 47% female, 71% GI malignancy), 50% received chemotherapy, 16% immunotherapy, and 34% combination therapy. From enrollment to 3 months, we observed an increase in total adipose tissue (16.9±52.4 cm2, 95% CI -33.79-0.63; p = 0.059), but not weight or skeletal muscle. Greater losses in skeletal muscle were associated with greater declines in 6MWT (B = 0.036, p = 0.014) and SBBP (B = 2.444, p = 0.002), but not the TUG. We observed no association with change in weight with all functional outcomes or PROs. Moreover, we found no association with body composition and PROs from enrollment to 3 months. Conclusions: In future CC trials, changes in longitudinal body composition rather than weight should be utilized. Furthermore, changes in skeletal muscle and the 6MWT and/or SBBP may be preferred endpoints.

The impact of emergency general surgery on end-of-life care among older patients with metastatic cancer

56

Background: Despite high post-discharge mortality among older patients with metastatic cancer who undergo emergency general surgery (EGS), little is known about the impact of EGS on the type of end-of-life care received. We sought to examine the association between EGS and established markers of high intensity or poor quality end-of-life care for cancer patients. Methods: This retrospective cohort study used 2001-2013 Surveillance, Epidemiology, and End Results-Medicare to identify beneficiaries 65 years or older, diagnosed initially with stage IV cancer (lung, colorectal, breast, ovarian, pancreatic, or melanoma), who received one of the seven highest-burden EGS operations, and died within 180 days of surgery. Non-EGS controls were exact-matched by age, sex, race, cancer type, and cancer diagnosis date then assigned a pseudo-exposure date corresponding to the EGS date. Conditional logistic regression adjusting for region and Charlson score was performed among pairs discharged alive to compare location of death (facility or home/hospice), healthcare utilization (hospitalization, intensive care unit (ICU) stay, emergency department (ED) visit) in the last 30 days of life, and hospice use (death in hospice, hospice enrollment less than three days from death). Results: Among 1,129 matched pairs, EGS patients had higher odds of death in facility (OR [95% CI]: 1.29 [1.05 - 1.58]) as well as hospitalization (1.83 [1.54 - 2.18]), ICU stay (2.05 [1.66 - 2.53]) or ED visit (1.76 [1.47 - 2.10]) in the last 30 days of life compared to non-EGS patients. EGS patients had higher odds of dying in hospice (1.22 [1.02 - 1.45]), but also experienced higher odds of hospice enrollment less than three days from death (1.72 [1.20 - 2.46]). Conclusions: Older patients with metastatic cancer who survive EGS experienced higher intensity end-of-life care than similar non-EGS patients. Such EGS patients may benefit from targeted interventions during the emergent hospitalization to improve the end-of-life care received.

More opioids, more constipation? Evaluation of longitudinal total oral opioid consumption and self-reported constipation in patients with cancer

211

Background: Constipation is a distressing physical symptom inadequately assessed in patients with cancer. Total opioid consumption may impact opioid-induced constipation (OIC) prevalence. This study examined total daily opioid consumption on self-reported constipation in patients with cancer. Methods: We performed a retrospective analysis in patients who had a histologically-confirmed cancer diagnosis and completed at least 2 outpatient PC visits at an academic palliative care (PC) clinic. At each PC visit, we collected self-reported constipation scores (11-point scale, 0 [no symptom] to 10 [worst symptom]) and 24-hour oral morphine equivalents (OME) for up to 5 visits. Spearman correlation examined the association between OME and presence of constipation rated ≥3 and Wilcoxan-Mann-Whitney examined OME between patients with or without constipation. Univariable logistic regression examined associations of independent variables on constipation. Results: We identified 404 patients with cancer (mean age 55.5±14.3 years, 61.9% women, 89% metastatic disease). The most common cancer types were gastrointestinal (n = 115; 31%) and gynecologic (n = 47; 13%). Stimulant laxative use was observed in 172 (44.9%) patients. Across all visits, we observed a weak association between OME and patient-reported constipation score (r = 0.16-0.24, p < 0.05 for 4 of 5 visits). Mean OME increased, while mean constipation score decreased across visits. At the first visit, higher mean OME was seen in patients who self-reported constipation (128.8 vs 92.3; p < 0.05). In contrast, for each subsequent clinic visit there was no difference in mean OME in patients with or without constipation. Age, sex, metastatic disease, and stimulant laxative use were not associated with constipation. Conclusions: A weak association was observed between OME and self-reported constipation in cancer patients across multiple palliative care visits. Except for the first clinic visit, there was no difference in OME and self-reported constipation scores. These results suggest a lack of a clear association between total opioid consumption and patient-reported constipation.

NCCN Guidelines Insights: Antiemesis, Version 2.2017

The NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Antiemesis address all aspects of management for chemotherapy-induced nausea and vomiting. These NCCN Guidelines Insights focus on recent updates to the NCCN Guidelines for Antiemesis, specifically those regarding carboplatin, granisetron, and olanzapine.

The development of a prediction tool to identify cancer patients at high risk for chemotherapy-induced nausea and vomiting

Background

Despite the availability of effective antiemetics and evidence-based guidelines, up to 40% of cancer patients receiving chemotherapy fail to achieve complete nausea and vomiting control. In addition to type of chemotherapy, several patient-related risk factors for chemotherapy-induced nausea and vomiting (CINV) have been identified. To incorporate these factors into the optimal selection of prophylactic antiemetics, a repeated measures cycle-based model to predict the risk of ≥ grade 2 CINV (≥2 vomiting episodes or a decrease in oral intake due to nausea) from days 0 to 5 post-chemotherapy was developed.

Patients and methods

Data from 1198 patients enrolled in one of the five non-interventional CINV prospective studies were pooled. Generalized estimating equations were used in a backwards elimination process with the P-value set at <0.05 to identify the relevant predictive factors. A risk scoring algorithm (range 0–32) was then derived from the final model coefficients. Finally, a receiver-operating characteristic curve (ROCC) analysis was done to measure the predictive accuracy of the scoring algorithm.

Results

Over 4197 chemotherapy cycles, 42.2% of patients experienced ≥grade 2 CINV. Eight risk factors were identified: patient age <60 years, the first two cycles of chemotherapy, anticipatory nausea and vomiting, history of morning sickness, hours of sleep the night before chemotherapy, CINV in the prior cycle, patient self-medication with non-prescribed treatments, and the use of platinum or anthracycline-based regimens. The ROC analysis indicated good predictive accuracy with an area-under-the-curve of 0.69 (95% CI: 0.67–0.70). Before to each cycle of therapy, patients with risk scores ≥16 units would be considered at high risk for developing ≥grade 2 CINV.

Conclusions

The clinical application of this prediction tool will be an important source of individual patient risk information for the oncology clinician and may enhance patient care by optimizing the use of the antiemetics in a proactive manner.

Evaluation of daily breakthrough chemotherapy-induced nausea and vomiting (CINV) rates in a phase III study of NEPA versus an aprepitant (APR)/granisetron (GRAN) regimen

120

Background: Antiemetic trials typically evaluate CINV control during acute (Day 1), delayed (Days 2-5), and overall (Days 1-5) phases post-chemotherapy; the daily course of events is often unstudied in the delayed phase. In a head-to-head study evaluating NK1RA-containing regimens, a single dose of NEPA, an oral fixed combination of the NK1RA netupitant and 5-HT3RA palonosetron, was non-inferior to a 3-day regimen of APR and GRAN for overall complete response (no emesis/no rescue use) rates (74% NEPA vs 72% APR/GRAN) in 828 patients receiving cisplatin-based highly emetogenic chemotherapy (HEC). This secondary analysis explores daily rates of breakthrough CINV in this study. Methods: This was a double-blind study in chemotherapy-naïve patients with solid tumors. Daily rates of breakthrough CINV, defined as the % of patients with emesis and/or rescue use were calculated with differences between treatment groups evaluated by the Cochran-Mantel-Haenszel method stratified by sex. Results: While daily rates of patients with breakthrough CINV remained stable between 13%-15.1% for APR/GRAN, they declined from 15.5% to 8% over the 5 days for NEPA, with the difference between groups reaching statistical significance on Day 5 (Table). Percent of total patient days with CINV events were 11.7% [NEPA] and 14.0% [APR/GRAN]. The % of patients with ≥ 3 days of breakthrough CINV were 8.5% and 12.3%, respectively. Conclusions: In this study evaluating guideline-recommended antiemetic regimens for HEC, CINV was prevented in most patients during the overall phase yet breakthrough CINV on individual days differed between treatment groups. APR/GRAN showed a relatively constant rate over time, while NEPA rates decreased and patients had fewer total days with breakthrough. This suggests the need for close monitoring of CINV events during the delayed phase. [Table: see text]

Advance care planning in Hispanic/Latino and non-Hispanic white patients with cancer

10

Background: Advance care planning (ACP) is a process whereby patients communicate their end-of-life treatment preferences. Despite a growing US Hispanic/Latino population, much remains unknown regarding ACP for Hispanic/Latinos. This study examined the association between ethnicity and ACP, including individual surrogate markers—code status documentation, advance directive (AD)/physician order for life-sustaining treatment (POLST) completion, and/or palliative care (PC) consultation—in deceased Hispanic/Latino vs white non-Hispanic (WNH) cancer patients. Methods: A retrospective analysis was performed in randomly-selected, matched pairs of deceased (2011-2016) Hispanic/Latino and WNH cancer patients at an NCI-designated cancer center. Pairs were matched based on sex, age (at diagnosis/death), and cancer type. Conditional logistic regression was used to assess ethnicity (Hispanic/Latino vs WNH) and the presence any ACP (yes/no). Secondary aims examined the association between ethnicity and the presence of individual ACP surrogate markers using separate logistic regression models. All analyses were completed using SAS 9.4. Results: 152 eligible matched pairs were analyzed with no significant differences in presence of any ACP, code status documentation, or PC consultation. Cancer patients with AD/POLST completion were 58% less likely to be Hispanic/Latino than WNH. Conclusions: Historically, ACP in cancer care is difficult to implement with 20-30% penetrance across all racial/ethnic groups. This study suggests even lower rates of AD/POLST completion in Hispanic/Latino patients consistent with prior studies. However, based on the small sample size, these results are not generalizable and requires further evaluation. This study affirms the need for continued efforts to improve AD/POLST completion in Hispanic/Latinos. [Table: see text]

Hospitalization costs for nausea and vomiting: A savings opportunity

155

Background: Nausea and vomiting (NV) are tracked by CMS as 2 of 10 top drivers of potentially avoidable hospital admissions for patients with cancer. These admissions pose a savings opportunity, yet limited literature exists on the costs of chemotherapy-induced NV (CINV). Burke’s 2010 assessment of 19,139 patients from 2004-2007 found CINV hospitalizations cost $7,448 and comprised the majority of overall CINV events and costs vs ambulatory events. Rashid (2016) evaluated 1,682 patients with metastatic breast cancer from 2007-2011 and found CINV admissions cost $10,074 in 2013 US dollars. Using data from the US Agency for Healthcare Research and Quality ’s Healthcare Cost and Utilization Project website (HCUPnet), and after applying adjustment factors from recent literature, we evaluated the cost of NV hospitalizations. Methods: Data on US hospital discharges for NV (based on primary discharge diagnosis) were obtained for 2014 from HCUPnet. Charges were adjusted using Smith’s (2015) model showing commercial insurers paid 48.7% of their hospital charges. They were further modified to include professional fees equaling 26.4% of facilities paid amounts, as reported by Peterson (2015). The CPI-Medical rate was used to adjust costs to 2016 dollars. A US payer perspective was used. The proportion of total NV discharges due to CINV is not known; it is also unknown whether charges for CINV differ from those for NV overall. Results: HCUPnet reported 37,730 hospital discharges for NV, with mean charges of $23,603 per event. The mean payment to hospitals, after adjustment to reflect amounts insurers actually paid, was estimated at $11,232 per event. Adding physicians’ professional fees incurred in the hospital setting yet normally charged separately yielded a total hospitalization cost of $14,197 ($15,120 in 2016 US dollars). Conclusions: Hospitalization for NV is common and costly; adjusted paid amounts averaged > $15,000 per discharge, a level consistent with the CINV literature. This economic impact, in addition to the consequences for patients’ quality of life, suggests the need for continued advances in preventing CINV and optimizing compliance with national antiemetic guidelines, particularly for chemotherapy with high emetogenic potential.

Economic impact of palliative care among elderly cancer patients

91

Background: Randomized trials among advanced cancer patients demonstrate that early palliative care integration into usual oncology care reduces symptom burden, improves quality of life and caregiver outcomes, and may improve survival. The impact of palliative care on health economics remains poorly defined and reported cost savings are an unintentional consequence of providing care aligned with patient goals. This study determined the impact of palliative care on healthcare costs among elderly patients with advanced cancer. Methods: We conducted a matched case-control study among Medicare beneficiaries with metastatic lung, colorectal, breast and prostate cancers. We matched patients who received a palliative care consultation to similar patients who did not receive a palliative care consultation. To determine the economic impact of a palliative care consultation we compared costs between cases and controls before and after the palliative care intervention. Costs included inpatient, outpatient, home health care, hospice, and medical equipment, and were adjusted to 2011 dollars. Results: Among the 2,576 patients in this study the total healthcare costs per patient in the 30 days before palliative care consultation was balanced between palliative care ($12,881) and non-palliative care control patients ($12,335). Palliative care intervention reduced total healthcare costs after the intervention. The total cost of care per patient in the 120 days after palliative care exposure was $6,880 compared to $9,604 for controls (28% decrease; p < 0.001). The economic effect of palliative care depended on timing of the consult. Palliative care consultation within 7 days of death decreased healthcare costs by $975, whereas palliative care consultation more than 4 weeks from death decreased costs by $5,362. Conclusions: This study demonstrates that palliative care has the capacity to substantially reduce healthcare expenditures among advanced cancer patients. Furthermore, the cost reduction depends on timing of the palliative care consult.

Next-generation sequencing ordering trends in the cancer trajectory

10104

Background: Next-generation sequencing (NGS) molecular tumor profiling is increasingly being ordered for advanced cancer patients to evaluate non-traditional therapeutic options. The timing of when NGS is ordered relative to date of diagnosis, palliative care (PC) consultation, and death remains unknown. The primary objective of this study was to examine NGS ordering patterns among cancer patients. Methods: This was a single center, retrospective data analysis in cancer patients at our institution between January 2011 and February 2016. Cancer patients ≥16 yrs of age were identified from a tumor registry and matched to an existing NGS tumor profiling database. Additional data were collected from an electronic medical record and compiled into a single database. Differences in the date of when NGS was ordered compared to date of diagnosis, PC consultation, and/or date of death were determined. A Mann-Whitney rank sum test examined differences in patients where NGS was ordered relative to the date of PC consultation. Logistic regression examined variables possibly associated with PC consultation. Results: Analysis included 1596 (807 women) cancer patients. Mean±SD age was 55.5±15.2 years, 30.8% (n = 492) of patients had metastatic disease, with breast and lung the most common cancers. The difference between date of cancer diagnosis and date of NGS order was 1053.6±1568.5 days (n = 1546). The difference between date of NGS order and date of death was 221.2.4±186.6 days. Two-hundred and fifty-one patients (15.7%) received a PC consultation, of which 82 patients had a NGS order before the PC consultation and 169 patients had a NGS order after the PC consultation. The mean difference in number of days between a NGS order before versus after a PC consultation was 147.3±216.8 vs. 179.8±169.7 days (p < 0.005). Four-hundred and sixty-six (29%) patients have died with 121 receiving a PC consultation. Metastatic disease, but not age and sex, was associated with PC completion (OR 1.7; 95%CI 1.27-2.21). Conclusions: NGS was frequently ordered near the time of death. PC consultations were completed in a minority of patients. NGS ordering in advanced cancer patients may serve as a trigger for PC consultation.

Administration of the ASCO Pain Survey to identify patient education deficiencies in patients with cancer

68

Background: The Doris Howell Palliative Care (PC) Outpatient Consultation Service consists of a multidisciplinary team. Reasons for consultation include pain management, treatment of other physical symptoms, education about non-drug options for pain, and advanced care planning. To assess methods used for patient education, the PC team administered the ASCO pain survey to those prescribed opioid pain medications. Methods: The survey was a 26-item questionnaire evaluating patient education content communicated by a provider to a patient during a routine clinic visit. Twenty surveys were completed anonymously. After survey review, the outpatient PC team developed an intervention to address patient education concerns. The intervention consisted of written and verbal instructions on pain medication management reviewed by the PC nurse with the patient and/or caregiver at the end of the outpatient visit. Another twenty surveys were completed anonymously with patients after intervention implementation. Time required to complete the intervention was less than two minutes per patient. The Fisher’s exact test was used to analyze the differences between surveys completed with or without the intervention. Results: The majority of patients stated that his/her provider adequately explained the use (n = 40, 100%), side effects (n = 37, 93%), and storage (n = 32, 80%) of pain medications, regardless of the intervention. Both groups also indicated that the provider and/or nurse adequately explained the risks associated with medical history (n = 34, 85%) and other medications while taking pain medications (n = 33, 83%). The intervention did increase understanding to avoid sharing pain medications (75% vs. 85%, p < 0.05) and to use a lock box to secure pain medications (45% vs. 60%, p < 0.05). Additionally, asking patients if family members have a history of alcohol or substance abuse was also significant (p < 0.05). Conclusions: Written and verbal instructions as an intervention improved patient understanding to avoid sharing pain medications and to secure pain medications in a lock box.

Palliative care pharmacists as advocates to conduct advance care planning

73

Background: Advance care planning (ACP) is a process whereby individuals consider their end-of-life (EoL) treatment preferences and make them known to caregivers and clinicians in the event of decisional incapacity. We previously reported a focused ACP intervention led by a clinical social worker (J Palliat Med 2016). Based on this trial’s initial success, we aimed to evaluate this approach when led by a pharmacist. To date, no data exists regarding pharmacist-led ACP discussions. Methods: Advanced stage cancer patients were screened at a single, academic oncology palliative care clinic. Training of the pharmacist included observation of ACP interventions conducted by the clinical social worker, didactic lectures, and role playing. Subjects engaged in a 1-hour ACP intervention with a palliative care pharmacist within 4 weeks of consent. Details of the ACP intervention were documented in the electronic medical record and AD/POLST completion was encouraged but not required. The study outcome was the identification of an informed proxy. After the patient’s death, proxies were contacted to determine if EoL wishes were achieved. Descriptive analyzes were performed. Results: Patient demographics were 22 patients, who were mostly woman (n = 13), Caucasian (n = 17), married (n = 14), with a mean age ± SD of 60.3 ± 10.5 years and with gastrointestinal (n = 5) and genitourinary (n = 5) as the most common primary cancers. After the ACP intervention by the pharmacist, 21 of 22 patients identified an informed proxy. The most commonly identified proxy was a spouse (n = 13). Fourteen (64%) subjects completed an AD/POLST after the ACP intervention. As of October 2016, 27% (6/22) of patients have died. Four out of 6 patients died in a setting consistent with their EoL wishes. Mean time to death from ACP intervention was 133±132 days. Sixteen subjects remain in surveillance. Conclusions: Preliminary results suggest that a pharmacist can conduct ACP discussions to identify an informed HCP. To date, the majority of patients achieved a death concordant with their EoL wishes.

Blood transfusions at end of life for stem cell transplant patients

115

Background: Transfusions are an essential palliative tool in the stem cell transplant (SCT) population. Limited data exist regarding transfusion practices at end-of-life for SCT patients and whether these practices may limit enrollment in hospice. Methods: A retrospective chart review was conducted of deceased patients with hematologic malignancies who underwent SCT at an academic medical center from 2011 to 2015. The primary objective was to determine the difference between the dates of last transfusion and death in patients enrolled and not enrolled in hospice. A secondary objective was evaluation of the number of transfusions between groups. Data were compiled from a single electronic medical record. Descriptive analyses were performed. Days to last transfusion were analyzed using the Wilcoxon-Mann-Whitney test. Number of packed red blood cell (PRBC) transfusions and platelets transfusions on the last day were analyzed using Fisher and chi-squared tests, respectively. Results: A total of 633 SCT were performed from 2011 to 2015 including 39% (n = 245) allogeneic and 61% (n = 388) autologous transplants (n = 29 patients had 2 transplants). Mean ± SD age of SCT patients was 55 ± 13 years. As of January 2016, 20% (n = 119) of these SCT patients have died. Of those that died, 15% (n = 18) were enrolled in hospice. For SCT patients enrolled in hospice, the mean ± SD time of last blood transfusion from death was 42.3 ± 63.4 days, with mean ± SD 0.67 ± 0.77 units of PRBC’s and 0.72 ± 0.75 units of platelets administered. For SCT patients not enrolled in hospice, the mean ± SD time of last blood transfusion from death was 14.2 ± 47.9 days, with mean±SD total 0.69 ± 1.03 units of PRBC’s and 1.14 ± 1 units of platelets administered. Hospice patients had a statistically significant longer number of days until last blood transfusion compared to non-hospice patients (p < 0.001). There was no difference between SCT patients enrolled in hospice and not enrolled in PRBC transfusions (p = 0.069), but there was a significantly higher amount of platelet transfusions in patients not enrolled in hospice (p < 0.005). Conclusions: This data suggests that time to last transfusion may be a significant obstacle for SCT patients when enrolling in hospice, but requires further validation.

Advance care planning and palliative care consultation for stem cell transplant patients

113

Background: Advance care planning (ACP) in stem cell transplantation (SCT) is particularly challenging given the potential for cure for patients with blood cancers despite an increased risk of suffering and even death. Data regarding ACP and palliative care (PC) integration in SCT is limited. Methods: A retrospective chart review was conducted of patients with hematologic malignancies who underwent SCT at UCSD from January 2011 to December 2015. The primary objective was to determine the medical discipline of the initial and last code status documentation. Secondary objectives included evaluation of AD and/or POLST completion, PC consultation, hospice enrollment, and location of death. Data were compiled from a single electronic medical record and descriptive statistical analyses performed. Results: A total of 633 SCT were performed from 2011 to 2015 including 39% (n = 245) allogeneic and 61% (n = 388) autologous transplants (n = 29 patients had 2 transplants). Mean age of SCT patients was 55 years (±13). All but one (n = 632) had code status documentation, and 0.8% (n = 5) were initially DNR. The initial code status was documented outpatient for 3% (n = 17), and by the primary SCT physician for 1 patient. The final code status was documented outpatient for 22% (n = 14), and by the primary SCT physician for 0.9% (n = 6). Nearly half (44%, n = 279) had an AD and/or POLST completed. PC consultation occurred for 19% (n = 121) with the majority (83%, n = 101) completed inpatient. PC consultation requests were made by the primary SCT physician (18%, n = 22), inpatient SCT team (68%, n = 82), critical care (8%, n = 10), or other (5%, n = 6).The most common reason for consultation was symptom management (80%, n = 94). As of January 2016, 20% (n = 127) of SCT patients died with a mean time from transplant of 312 days (± 317). Of those that died, the majority (83%, n = 106) died in the hospital, 15% (n = 19) were full code, 48% (n = 62) had an AD and/or POLST, and 14% (n = 18) were enrolled in hospice. Conclusions: These single center data suggest ACP and PC integration in SCT is limited. Opportunities exist to expand integration to the outpatient setting and earlier in the course of illness.

Palliative care and health care utilization at the end of life

152

Background: Palliative care’s role in oncology has expanded, but its impact on aggressiveness of care at the end of life has not been characterized at the population level. Methods: This matched retrospective cohort study examined the effect of an encounter with palliative care on healthcare utilization at the end of life among 6,580 Medicare beneficiaries with advanced prostate, breast, lung, or colorectal cancer. We compared healthcare utilization before and after palliative care consultation to a matched non-palliative care cohort. Results: The palliative care cohort had higher rates of healthcare utilization in the 30 days prior to palliative care consult compared to the non-palliative cohort, with higher rates of hospitalization (risk ratio [RR] 3.33; 95% CI 2.87-3.85), invasive procedures (RR 1.75; 95% CI 1.62-1.88), and chemotherapy administration (RR 1.61; 95% CI 1.45-1.78). The opposite pattern emerged in the interval from palliative care consultation through death, where the palliative care cohort had lower rates of hospitalization (RR 0.53; 95% CI 0.44-0.65), invasive procedures (RR 0.52; 95% CI 0.45-0.59), and chemotherapy administration (RR 0.46; 95% CI 0.39-0.53). Patients with early palliative care consultation had larger absolute reductions in healthcare utilization compared to those with palliative care consultation closer to the end of life. Conclusions: This population-based study found that palliative care substantially decreased healthcare utilization among Medicare beneficiaries with advanced cancer. Given the increasing number of elderly patients with advanced cancer, this study emphasizes the importance of early integration of palliative care alongside standard oncologic care.

Molecular tumor profiling ordering trends in cancer patients

165

Background: Molecular tumor profiling may provide information as to whether to initiate or not initiate a targeted therapy. As to the timing of when the tumor profiling is ordered relative to date of diagnosis, date of death, and palliative care (PC) consultations are unknown. The primary objective of this study was to examine molecular tumor profiling ordering trends in the course of cancer illness. Methods: A preliminary, retrospective chart review was conducted in a cohort of patients with a confirmed diagnosis of cancer at an academic, NCI-designated comprehensive cancer center. Patients were identified from a tumor registry and then matched to a next generation sequencing molecular tumor profiling database. The date of palliative care consultation was collected from the electronic medical record. Differences in the date of when tumor profiling was ordered and date of diagnosis, date of PC consultation, and/or date of death were determined. Data were compiled into a single database and descriptive statistical analyses were performed. Results: A cohort of 397 (205 women) cancer patients was included. Metastatic disease was present in 108 (27.2%) patients, with mean±SD age of 58.7 ± 13.5 yrs. One-hundred and nine (27.6%) patients received a PC consultation (n=60 inpatient, n = 49 outpatient). As of February 2016, 119 (30%) patients died, with 58 (48.7%) out of 119 receiving a PC consultation. The difference between date of cancer diagnosis and date of tumor profiling ordered was 2467.4 ± 6865.7 days (n = 376), while the difference between date of tumor profiling ordered and date of death was 229.1 ± 185.7 days (n = 111). The difference between date of cancer diagnosis and date of death was 1507.5 ± 2002.1 days (n = 119). In patients were the tumor profiling was ordered before the PC consultation (n = 29), the difference between date of PC consultation and date tumor profiling ordered was 157.3 ± 258.1 days. In contrast, in patients were the tumor profiling was ordered after the PC consultation (n = 76), the difference was 194.6 ± 168 days. Conclusions: This analysis suggests that molecular tumor profiling is ordered at the end and not at the beginning of a cancer illness. PC consultations are not routinely performed in patients who participate in tumor profiling.