Citations

11034 Background: While resources exist to address the needs of graduate medical trainees, online resources to assist trainees with specialty-specific professional development have not been fully developed or described in the past. The Hem-Onc Fellows Network (@HemOncFellows) is an online community for hematology-oncology trainees in all stages of training, including medical students, residents, fellows, and doctoral students. The network's goals are to provide professional development in hematology-oncology, create a community and safe space for hematology-oncology trainees, and amplify the voices and needs of trainees. Methods: The @HemOncFellows Twitter account was created in February of 2021. The hashtag #HOFellows was registered and certified with healthcare Symplur. The network hosts Twitter Spaces (TS) with guests every two weeks and distributes a newsletter. Fifteen TS were held between July 2021 and February 2022. For each TS, a specific topic of professional development was discussed. At least one content expert was invited to each TS to provide additional commentary. Demographics of TS participants and newsletter subscribers were gathered and qualitative and quantitative analysis was performed. Results: Since its inception, the Hem-Onc Fellows Network has grown to more than 2200 followers. Symplur has over 500 #HOFellows tweets from November of 2021 to early February of 2022. The network organized 15 TS: 14 in professional development topics and one meet-and-greet for new matched fellows in December. TS had an average of 41 participants per event. Most participants were United States-based but there were participants from 21 countries worldwide. Forty-eight participants attended three or more spaces, including 15 hem-onc fellows and 17 internal medicine residents. The newsletter has 48 subscribers. Conclusions: We demonstrate that implementing a trainee network on Twitter attracted participation and engagement from a diverse pool of trainees worldwide. TS represents a novel educational tool for engaging trainees and facilitating professional development.

6552 Background: Significant amounts of expensive oncology drugs are discarded resulting in waste and financial consequences for patients. Beginning in 2023, the Infrastructure Investment and Jobs Act (IIJA) will require drug manufacturers to refund Medicare for any single-use drugs where greater than 10% of the drug is wasted. We explored potential IIJA-associated savings through assessment of patterns in spending for discarded oncology agents between 2017-2020. Methods: We utilized the Medicare Part B Discarded Drug Units database from 2017-2020. Using J codes, we extracted spending and reimbursement data for discarded medications for antineoplastic and classical hematology therapies. The primary outcome was identification of therapies with high percentage of waste and the economic impact of retroactive application of the IIJA for discarded therapies. Results: Medicare Part B utilization data was extracted for a median 77 (64-152) oncologic therapies per year with median reimbursement for discarded treatments of $590 M ($566-616). From 2017-2020, bortezomib, romiplostim, and nab-paclitaxel were consistently among the top 5 therapies by value of wasted product. Pembrolizumab waste by value declined significantly after 2017 following transition to fixed dosing. In 2020, the top 5 therapies by value of wasted product were bortezomib, nab-paclitaxel, trastuzumab, romiplostim, and cabazitaxel with a value of $278 M. In 2020, the agents with the highest percentage of wasted product were bortezomib (27%), cabazitaxel (28%), decitabine (23%), topotecan (23%), and azacitidine (22%) with a collective discarded value of $160 M. Notably, these agents have weight-based dosing and single-use vials. Based upon the average number of drug claims per recipient and volume of product discarded, eliminating drug waste would allow an additional 6564 patients to receive treatment with discarded bortezomib. The number of additionally treated patients for the other top 5 agents: cabazitaxel (1216), decitabine (1509), topotecan (291), and azacitidine (2682). Retroactive application of IIJA policy would have impacted a median 20 (12-26) oncology agents resulting in median annual cost savings of $153 M (142-157) and the median annual spending on wasted oncology therapies would decline to $440 M (410-466). Conclusions: The IIJA would have yielded a median annual savings of $153 M in Medicare spending if applied retroactively. Given a 10% waste cutoff, a limited number of agents would be targeted by this legislation. In our study, we observe that drugs associated with weight-based dosing and single-use drug vials account for most drug waste. Strategies to minimize discarded units (through storage of unused product, approval of multi-dose vial formulations, or fixed-dose therapies) could yield substantial savings while enabling health equity in collaboration with communities with difficult access to these drugs.

6514 Background: Tobacco use is a modifiable risk factor for adverse outcomes among patients diagnosed with cancer. Despite ASCO’s recommendation for assessment and treatment of tobacco use, integration into cancer care is suboptimal. Socioeconomic contexts influence access and utilization of tobacco treatment, but little is known about the relationship between neighborhood socioeconomic disadvantage (NSD) and tobacco assessment, assistance, and cessation among cancer patients enrolled in clinical trials. Methods: The NCI Cancer Patient Tobacco Use Questionnaire (C-TUQ) was centrally administered to participants enrolled in 10 ECOG ACRIN clinical trials (9 therapeutic, 1 imaging). We examined associations of NSD with patient-reported rates of receiving brief tobacco cessation support (i.e., Ask, Assist (counseling)) and cessation (past 30d quit attempts and duration). NSD was measured using the national Area Deprivation Index (ADI) based on participant’s zip code. Associations between ADI (low, intermediate, and high) and tobacco variables were evaluated using logistic regression and ANOVA. Results: 740 patients, completing the C-TUQ between June 2017-October 2021, can be classified as 402 (54%) never smokers, 81 (11%) current smokers, and 257 (35%) former smokers. Patients were 70% male; 94% white; 3% Hispanic; mean age 58.8 (SD 9.0). Cancer diagnoses were 36% leukemia; 19% lymphoma, 18% prostate, 11% breast; 9% melanoma, 7% myeloma, and 0.5% head and neck. Patients were categorized into high (33%), intermediate (34%) and low (33%) disadvantaged neighborhoods. Patients in high (vs. low) disadvantaged neighborhoods were more likely to report being asked about smoking (OR = 3.90; 95% CI (confidence interval), 1.61 to 9.46; p = 0.0062) but less likely to report receiving counseling to help quit smoking (OR = 0.20; 95% CI, 0.06 to 0.73; p = 0.0234). Patients from high disadvantaged neighborhoods had the shortest quit duration, followed by patients from intermediate and low disadvantaged neighborhoods (mean = 145.78, 187.66, and 210.98 months, respectively, p = 0.0372). Conclusions: Greater socioeconomic neighborhood disadvantage was associated with increased assessment of tobacco use but decreased tobacco treatment referral, and the shortest quit duration. More research is needed to promote increased referral to tobacco treatment for individuals with cancer from disadvantaged neighborhoods to promote and sustain cessation.

4145 Background: Pancreatic ductal adenocarcinoma (PDAC) is the 4th leading cause of cancer death. Outcomes remain poor, due to irresectability at diagnosis for many and sub-optimal responses to systemic therapy. Cytotoxic chemotherapy remains the standard of care. High microsatellite instability (MSI-H) predicts response to immune checkpoint inhibition (ICI) in many cancers. Detecting high MSI is rare in PDAC (incidence <2%), but case reports demonstrate potential therapeutic benefit with ICI. Here, we present multi-institutional data to characterize the clinical behavior of MSI-H PDAC, with special attention to response to ICI. Methods: Cases of MSI-H PDAC were obtained by reviewing data of all PDAC patients from our tertiary cancer center who had undergone genomic sequencing by one commercially available platform. The resulting cohort was supplemented with MSI-H PDAC cases identified by GI oncology specialists at multiple institutions. De-identified patient data were compiled and analyzed. Results: 15 MSI-H PDAC patients were identified. 20% had stage II disease at diagnosis, 27% stage III, and 53% stage IV. 73% of patients received ICI (n=11); 40% as 1st line and 33% as 2nd line. These patients demonstrated 100% overall response rate; 45% complete response (1 pathologic CR, 4 radiographic CR) and 55% partial response. No patient that received ICI had lost response or died after a median follow-up of 18 months (range 6-89 mos). 1 patient had oligoprogression of a single hepatic lesion after 7 mos that was then irradiated; this patient retained radiographic CR for 17 subsequent mos (ongoing). In this cohort, we observe poor responses to cytotoxic chemotherapy. In total, 12 regimens were trialed among 9 patients. Overall response rate was 0%. 42% achieved disease stability, with median duration of response of 2 mos; only 2 cases maintained disease stability for >5 mos. 4 patients did not receive ICI; all patients died, with a median survival of 7.5 mos. Conclusions: MSI-H PDAC represents a rare but important subtype of PDAC with unique clinical behavior. Given its rarity, large-scale analyses and trials are unlikely to be performed, making case series such as ours crucial. In our cohort, we observe impressive, durable responses to ICI, along with very poor responses to cytotoxic chemotherapy. Our data argues for consideration of ICI in any patient presenting with MSI-H PDAC, including in the first-line and neoadjuvant settings.

4130 Background: KRAS is mutationally activated in over 90% of pancreatic ductal adenocarcinoma (PDAC). Compared to pts with KRAS mutation, KRAS wild-type (wt) PDAC seem to have better response to therapy and may harbor potentially actionable molecular alterations. Here, we analyze the molecular profile and clinical outcome of a cohort of pts with KRAS wt PDAC. Methods: A retrospective review was conducted on pts with PDAC who underwent CLIA-certified Next Generation Sequencing (NGS) testing at Mayo Clinic between December 1, 2018 and December 1, 2021. Pts with KRAS wt PDAC with available reports were included. Their genomic drivers, RNA expression, demographics, disease characteristics, therapies offered, and clinical outcome data were collected. The study was approved by the institutional IRB. Results: Of the 241 eligible pts, 8% (19) has KRAS wt PDAC. Among those, 2 pts had no mutation identified by the gene/molecular panel used. Of the 17 pts (89%) with identified alterations, mutations found in ³ 2 pts were TP53 (53%), CDKN2A (16%) and CDKN2B/ERBB2/PTEN/MSH3/RNF43/FBXW7/KMT2D/GNAS (11% each). Chromosomal rearrangements were identified in 5 (26%): CADPS2-BRAF, GP2-ERBB2, PTPRK-RSPO3, EML4-NTRK3 and TFG-MET. RNA expression results were available in 12 pts: common overexpression were ERBB2 (27%) and MET/NRAS/MYC/CCDN1/CCNE1/AR (18% each); and the under-expression MGMT (18%). Among the 13 pts with available MSI status via NGS, 2 (11%) were MSI-high (both had high TMB [28.4 and 23.7 m/MB]) while all others were TMB < 10 m/MB. The median age at diagnosis was 61 years (68% males). 8/19 (42%) were Stage IV at diagnosis with 15/19 (79%) pts ultimately diagnosed with metastatic disease. Among metastatic pts, median lines of treatment received was 2.5 (range:0-4). 4 pts received FOLFIRINOX (FFX), 2 gemcitabine/ nab-paclitaxel (GP) and tumor response were comparable to previously reported results. 1 received 1st-line pembrolizumab and remained on therapy at the time of analysis. The median length of follow up from diagnosis was 29 months. A patient with TFG-MET re-arrangement previously progressed on FFX and GP was treated using a MET inhibitor, and achieved significant CA19-9 drop and pancreas tumor shrinkage at 1st restaging, with ongoing response. Conclusions: The molecular profile of KRAS wt PDAC is highly heterogeneous and difficult to generalize. Novel approaches (e.g., basket trials) are needed to develop therapy for this rare PDAC subgroup.

Background:Research on the utilization and effectiveness of antitumor necrosis factor (TNF) biologics in children with very early onset inflammatory bowel disease (VEOIBD) is urgently needed. Here we describe anti‐TNF use and durability in a multicenter cohort.Methods:We performed a retrospective cohort study of patients diagnosed with VEOIBD (<6 years) between 2008 and 2013 at 25 North American centers. We performed chart abstraction at diagnosis and 1, 3, and 5 years after diagnosis. We examined the rate of initiation and durability of infliximab and adalimumab and evaluated associations between treatment durability and the following covariates with multivariate Cox proportional hazard regression: age at diagnosis, sex, disease duration, disease classification, and presence of combined immunomodulatory treatment versus monotherapy.Results:Of 294 children with VEOIBD, 120 initiated treatment with anti‐TNF therapy and 101 had follow‐up data recorded [50% Crohn disease (CD), 31% ulcerative colitis (UC), and 19% IBD unclassified (IBD‐U)]. The cumulative probability of anti‐TNF treatment was 15% at 1 year, 30% at 3 years, and 45% at 5 years from diagnosis; 56 (55%) were treated between 0 and 6 years old. Anti‐TNF durability was 90% at 1 year, 75% at 3 years, and 55% at 5 years. The most common reason for discontinuation of anti‐TNF were loss of response in 24 (57%) children. Children with UC/IBD‐U had lower durability than those with CD (hazard ratio [HR] 0.17; 95% confidence interval [CI], 0.06–0.51; P = 0.001).Conclusions:Utilization and durability of anti‐TNF in VEOIBD is relatively high and comparable with older children. Having Crohn disease (compared with UC/IBD‐U) is associated with greater durability.

PURPOSE: Precision oncology promises improved outcomes but the cost-effectiveness and accessibility of targeted therapies is debatable. We report price change patterns from 2015 to 2019 for several oral anticancer medications for common solid tumor malignancies. METHODS: We collected provider utilization and payment data from the public Medicare Part D database and extracted drug price information for commonly prescribed targeted oral anticancer agents for lung, breast, and prostate cancer. We then calculated median Pearson correlation coefficient values for various drugs (containing more than two data points) within each therapeutic class. We also calculated compound annual growth rates (CAGRs) for medication costs within each class and compared them with the consumer price index (CPI). RESULTS: Our study included six epidermal growth factor receptor inhibitors (EGFRi; one generic), five anaplastic lymphoma kinase inhibitors (ALKi), two B-Raf inhibitors (BRAFi), three hormonal agents (one generic), three cyclin-dependent kinases 4 and 6 inhibitors (CDK4/6i), two poly-ADP-ribose inhibitors (PARPi), and seven antiandrogen agents (two generic). The median (range) Pearson correlation coefficient values for cost of drugs within each therapeutic class were 0.967 (0.915-0.978) for EGFRi, 0.981 (0.966-0.989) for ALKi, 0.996 for BRAFi, 0.994 (0.992-0.999) for CDK4/6i, 0.855 for PARPi, and 0.442 (–0.522 to 0.962) for antiandrogens. Therapies with two or fewer data points (generic erlotinib, dacomitinib, abiraterone, apalutamide, and darolutamide) were excluded. The median CAGRs in costs over the 5-year period were 4.56% (EGFRi), 6.40% (ALKi), 2.58% (BRAFi), 5.48% (hormonal agents), 5.21% (CDK4/6i), 27.29% (PARPi), and 34.8% (antiandrogens). The CPI over 5 years was 2.26%/year, and the average inflation rate was 1.90%/year. CONCLUSION: The median CAGR in costs for modern oral precision-driven cancer therapeutic classes mostly outpaced CPI and the average inflation. Increase in cost within the same class should be weighed against incremental clinical benefit for the patients to ensure that rising costs do not limit access to targeted therapies.

Objective: Trauma-induced coagulopathy (TIC) is provoked by multiple mechanisms and is perceived to be one driver of massive transfusions (MT). Single laboratory values using prothrombin time (INR) or thrombelastography (TEG) are used to clinically define this complex process. We used a proteomics approach to test whether current definitions of TIC (INR, TEG, or clinical judgment) are sufficient to capture the majority of protein changes associated with MT. Methods: Eight level I trauma centers contributed blood samples from patients available early after injury. TIC was defined as INR >1.5 (INR-TIC), TEG maximum amplitude <50 mm (TEG-TIC), or clinical judgment (Clin-TIC) by the trauma surgeon. MT was defined as >10 units of red blood cells in 24 hours or >4 units RBC/hour during the first 4 hours. SomaLogic proteomic analysis of 1305 proteins was performed. Pathways associated with proteins dysregulated in patients with each TIC definition and MT were identified. Results: Patients (n = 211) had a mean injury severity score of 24, with a MT and mortality rate of 22% and 12%, respectively. We identified 578 SOMAscan analytes dysregulated among MT patients, of which INR-TIC, TEG-TIC, and Clin-TIC patients showed dysregulation only in 25%, 3%, and 4% of these, respectively. TIC definitions jointly failed to show changes in 73% of the protein levels associated with MT, and failed to identify 26% of patients that received a massive transfusion. INR-TIC and TEG-TIC patients showed dysregulation of proteins significantly associated with complement activity. Proteins dysregulated in Clin-TIC or massive transfusion patients were not significantly associated with any pathway. Conclusion: These data indicate there are unexplored opportunities to identify patients at risk for massive bleeding. Only a small subset of proteins that are dysregulated in patients receiving MT are statistically significantly dysregulated among patients whose TIC is defined based solely on laboratory measurements or clinical assessment.

Background: Automated insulin delivery (AID) systems have proven effective in increasing time-in-range during both clinical trials and real-world use. Further improvements in outcomes for single-hormone (insulin only) AID may be limited by suboptimal insulin delivery settings. Methods: Adults (≥18 years of age) with type 1 diabetes were randomized to either sensor-augmented pump (SAP) (inclusive of predictive low-glucose suspend) or adaptive zone model predictive control AID for 13 weeks, then crossed over to the other arm. Each week, the AID insulin delivery settings were sequentially and automatically updated by an adaptation system running on the study phone. Primary outcome was sensor glucose time-in-range 70–180 mg/dL, with noninferiority in percent time below 54 mg/dL as a hierarchical outcome. Results: Thirty-five participants completed the trial (mean age 39 ± 16 years, HbA1c at enrollment 6.9% ± 1.0%). Mean time-in-range 70–180 mg/dL was 66% with SAP versus 69% with AID (mean adjusted difference +2% [95% confidence interval: −1% to +6%], P = 0.22). Median time <70 mg/dL improved from 3.0% with SAP to 1.6% with AID (−1.5% [−2.4% to −0.5%], P = 0.002). The adaptation system decreased initial basal rates by a median of 4% (−8%, 16%) and increased initial carbohydrate ratios by a median of 45% (32%, 59%) after 13 weeks. Conclusions: Automated adaptation of insulin delivery settings with AID use did not significantly improve time-in-range in this very well-controlled population. Additional study and further refinement of the adaptation system are needed, especially in populations with differing degrees of baseline glycemic control, who may show larger benefits from adaptation.