Longitudinal MRD Assessment in Real-World Multiple Myeloma Patients Using Next-Generation Sequencing (clonoSEQ® Assay)

Introduction Achieving minimal residual disease (MRD) negativity in multiple myeloma (MM) is positively correlated with prolonged PFS and OS (Landgren 2016, Munshi 2017), and longitudinal assessment of MRD can provide additional prognostication (Diamond 2021). This is substantiated by the inclusion of sustained MRD negativity (defined as two consecutive MRD negative results ≥12 months apart at 10 -5 threshold) within IMWG and NCCN guidelines as part of the response criteria for symptomatic MM. MRD has also recently been adopted as a potential primary endpoint in clinical trials, and there is increasing focus on the role of MRD in guiding treatment decisions, including discontinuation of therapy (Lahuerta 2021). Standardized, sensitive, and serial MRD assessment is useful to support both routine patient management consistent with guidelines, and clinical studies which incorporate MRD as a primary endpoint. NGS MRD (clonoSEQ Assay, Adaptive Biotechnologies, Seattle, WA) is currently the only FDA authorized test available for MRD assessment in patients with MM using bone marrow. The assay's standardization, sensitivity, and wide availability facilitates real-world evidence (RWE) analyses looking at MRD trends in MM patient cohorts over time. The aim of this abstract is to provide further insight into longitudinal MRD testing patterns and attainment of sustained MRD negativity in a real-world cohort of patients. Methods The study population includes a de-identified internal dataset of MM samples processed as part of Adaptive's clinical offering (excluding trials) to support routine patient management; patients were geographically dispersed and treated at a variety of institutions, and the analysis period was from January 2018 to July 2021. We examined sustained MRD negativity and other longitudinal patterns of MRD testing and negativity at deep MRD thresholds. Results In the analysis period, we identified 1,561 MM patients with at least two MRD tests following a trackable sequence from a baseline clonality (ID) test. Of those patients, 616 had ≥3 tests. The median testing interval for patients with ≥2 MRD tests was 195 days. The median age was 66 and 59% were male, which is generally consistent with epidemiological data. When examining MRD monitoring longitudinally, of the 1,561 patients that had ≥2 MRD tests, 383 (24.5%) and 231 (14.7%) patients had sustained MRD-negativity for at least 12 months at 10 -5 and 10 -6 thresholds, respectively. Of the 616 patients with ≥3 MRD tests, 200 (32.5%) and 102 (16.6%) has remained negative throughout the testing period, 191 (31.0%) and 181 (29.4%) had converted from positive to negative, and 76 (12.3%) and 61 (9.9%) had converted from negative to a positive state at the respective thresholds. The trends within subgroups (only positive, only negative, negative -> positive, and positive -> negative) vary based on a deeper sensitivity threshold. Conclusions Serial measurement of MRD can facilitate a dynamic assessment of risk for disease progression in patients with MM, and this real-world analysis of MRD clinical samples provides insight into serial testing patterns and sustained MRD-negativity using next-generation sequencing. Figure 1 Figure 1. Disclosures Jiang: AstraZeneca LLC: Current equity holder in publicly-traded company, Ended employment in the past 24 months; Adaptive Biotechnologies: Current Employment. Lee: Adaptive Biotechnologies: Current Employment. Eckert: Adaptive Biotechnologies: Current Employment, Current equity holder in publicly-traded company, Divested equity in a private or publicly-traded company in the past 24 months. Demaree: Adaptive Biotechnologies: Current Employment, Current equity holder in publicly-traded company, Divested equity in a private or publicly-traded company in the past 24 months. Hewitt: Adaptive Biotechnologies: Current Employment, Current equity holder in publicly-traded company, Divested equity in a private or publicly-traded company in the past 24 months.

Real-World Treatment Patterns and Outcomes of Patients with Chronic Lymphocytic Leukemia (CLL) Receiving First-Line (1L) Therapy in the United States (US)

Background: There have been many advances in CLL treatments over the past decade, with a number of novel agents targeting molecular pathways within CLL cells receiving approval from the US Food and Drug Administration. Here, we assessed the evolution of molecular testing patterns, treatment patterns, and clinical outcomes over time in patients receiving 1L CLL treatment in a real-world US database. Methods: This was a retrospective cohort study using the Flatiron Health database, a longitudinal database comprising de-identified, patient-level, structured and unstructured data, curated via technology-enabled abstraction. During the study period, the de-identified data originated from approximately 280 cancer clinics (~800 sites of care) in the US. Patients aged 18 years and older who were diagnosed with CLL and initiated 1L treatment between December 2015 and December 2020 were selected. Participants who took part in a clinical trial in any line of therapy, or who had any other primary cancer diagnosis, were excluded. Baseline characteristics, including testing patterns, at initiation of 1L treatment were assessed using descriptive statistics. Treatment patterns and outcomes, such as time to next treatment or death (TTNTD), were analyzed. Kaplan-Meier analysis was used to estimate TTNTD. Results: Among 3654 patients with treatment-naive CLL who were selected from the de-identified database, the mean age at 1L treatment initiation was 70 years (range, 29-85); 64.3% of patients were male; 72.1% were White, 8.2% Black, 3.9% Hispanic/Latino, 1.0% Asian, and 14.9% were of other ethnicity/race. Approximately one-third (34.7%) of patients had Rai stage 0-I disease, 6.9% had stage II, 6.3% stage III, 11.5% stage IV, and 40.6% had undocumented Rai stage. Testing patterns: The majority of identified patients (3202/3654; 87.6%) had undergone cytogenetic testing, fluorescence in situ hybridization, or IGHV mutation testing. Compared with 2015-2016, testing rates were higher in 2019-2020 for chromosome 17p deletion (del(17p); 36.1% vs 45.7%, respectively; p<0.001) and for IGHV mutation status (84.7% vs 89.2%, respectively; p=0.003). Overall, 11.0% of patients had del(17p). Of those tested for IGHV (1472/3654; 40.3%), 58.3% had unmutated IGHV. Treatment patterns: The 10 most commonly used 1L CLL treatments, which overall represented 91.8% of all 1L treatments, and their evolution over time, are reported in Table 1. Of the patients receiving these top 10 1L treatment regimens overall, 45.7% received regimens including novel targeted oral agents, 33.4% received chemo-immunotherapy (CIT), and 19.7% received anti-CD20 monotherapy. Evaluation of each 2-year period shows that treatment patterns for the top 10 1L treatment regimens shifted, with use of novel targeted oral agents increasing from 27.1% (2015-2016) to 63.8% (2019-2020) (p<0.001), while use of CIT and chemotherapy decreased over time (Table 2). Approximately 30.0% (1088/3654) of 1L-treated patients went on to receive second-line treatments. Outcomes: Median TTNTD was 34.4 months for all patients receiving 1L CLL treatment, and 36.5 months for patients who received the 10 most common 1L treatments across the 6-year study period (n=3360). Median TTNTD was 47.0 months for patients who received novel targeted oral agents and 41.5 months for patients who received CIT (unadjusted p=0.16). When evaluating outcomes in patients with high-risk cytogenetics, median TTNTD was 29.1 months for patients with del(17p) and 37.2 months for those with unmutated IGHV, but was longer in those patients who received treatment with novel targeted oral agents (median TTNTD of 43.9 and 46.7 months, respectively; Table 3). Conclusions: This analysis provides the current state of 1L CLL testing and treatment patterns and outcomes in the US from 2015 to 2020. As expected, the use of novel targeted oral agents increased over time, with a corresponding increase in TTNTD. Clinical outcomes were improved in patients receiving novel targeted oral agents, both overall and in high-risk subgroups. Following on from this, a comparative study of TTNTD for novel oral agents versus CIT, and analyses of outcomes of different sequencing of therapies, will be conducted. Figure 1 Figure 1. Disclosures Mato: Nurix: Research Funding; Johnson and Johnson: Consultancy, Research Funding; AbbVie: Consultancy, Research Funding; Acerta/AstraZeneca: Consultancy, Research Funding; DTRM BioPharma: Consultancy, Research Funding; Pharmacyclics LLC, an AbbVie Company: Consultancy, Research Funding; Adaptive Biotechnologies: Consultancy, Research Funding; BeiGene: Consultancy, Research Funding; MSKCC: Current Employment; Sunesis: Consultancy, Research Funding; AstraZeneca: Consultancy; TG Therapeutics: Consultancy, Other: DSMB, Research Funding; Genmab: Research Funding; LOXO: Consultancy, Research Funding; Genentech: Consultancy, Research Funding; Janssen: Consultancy, Research Funding. Ravelo: Genentech, Inc.: Current Employment; Roche Holdings: Current equity holder in publicly-traded company, Current holder of stock options in a privately-held company. To: Genentech, Inc.: Current Employment; F. Hoffmann-La Roche Ltd: Current equity holder in publicly-traded company, Divested equity in a private or publicly-traded company in the past 24 months. Schuldt: Genentech, Inc.: Current Employment; F. Hoffmann-La Roche Ltd: Current equity holder in publicly-traded company; Johnson & Johnson: Divested equity in a private or publicly-traded company in the past 24 months. Biondo: Genentech, Inc.: Current Employment; Roche: Current holder of individual stocks in a privately-held company.

658. Diagnostic Testing Among Patients with Suspected Recurrent Clostridioides difficile Infection (rCDI) in ECOSPOR III a Phase 3 Clinical Trial: Implications for Clinical Practice vs Clinical Trials

Background Accurate diagnosis of rCDI is challenging because of limitations in test performance and alternative causes of recurrent diarrhea, such as post-infectious irritable bowel syndrome (IBS). Stool enzyme immunoassay (EIA) toxin testing (TOX) is the best predictor of active disease, but may miss cases of CDI when toxins are below the limit of detection. In contrast, glutamate dehydrogenase (GDH) or PCR have high sensitivity but cannot differentiate colonization from infection, leading to possible overdiagnosis due to low specificity. In ECOSPOR III, SER-109, an investigational purified microbiome therapeutic, was superior to placebo in reducing rCDI (12.4% vs 39.8%, respectively; p-value < 0.001). We examined diagnostic testing patterns among screened subjects. Methods Patients with ≥2 prior episodes and ≥3 unformed bowel movements over 48 hours were screened. To ensure enrollment of patients with active CDI, toxin testing was required at entry via a local certified or central lab (Eurofins; Framingham, MA). Subjects with discordant GDH+/TOX- tests at the central lab had reflex confirmatory testing with a cell cytotoxicity neutralization assay (CCNA), considered the “gold standard” for toxin testing. Results The leading reason for screen failure among 281 subjects screened was a negative toxin test (50/99; 50.5%). Of 182 patients enrolled, 59 (32.4%) qualified with EIA TOX+ at the local lab (33 TOX+; 25 GDH+/TOX+) and 122 (67.0%) qualified by the central lab (Table 1). Of these 122 subjects, 87 qualified by GDH+/TOX+ but 35 required additional reflex testing by CCNA due to discordant GDH+/TOX-results; all 35 were positive. Diagnostic Testing for Qualifying C. difficile Episode in ITT Population Conclusion These diagnostic testing patterns suggest a subset of patients with suspected rCDI have toxin concentrations below the EIA threshold for detection or may have an alternative cause of diarrhea, such as post-infectious IBS. Thus, the limitations of EIA toxin testing need to be considered in clinical practice when evaluating patients with compatible symptoms of rCDI and a high prior probability of infection. In contrast, in trials of investigational agents, toxin testing assures enrollment of patients with active disease and accurate estimates of efficacy. Disclosures Matthew Sims, MD, PhD, FACP, FIDSA, Astra Zeneca (Independent Contractor)Diasorin Molecular (Independent Contractor)Epigenomics Inc (Independent Contractor)Finch (Independent Contractor)Genentech (Independent Contractor)Janssen Pharmaceuticals NV (Independent Contractor)Kinevant Sciences gmBH (Independent Contractor)Leonard-Meron Biosciences (Independent Contractor)Merck and Co (Independent Contractor)OpGen (Independent Contractor)Prenosis (Independent Contractor)Regeneron Pharmaceuticals Inc (Independent Contractor)Seres Therapeutics Inc (Independent Contractor)Shire (Independent Contractor)Summit Therapeutics (Independent Contractor) Sahil Khanna, MBBS, MS, Seres (Grant/Research Support) Darrell Pardi, MD, seres (Consultant)Vedanta (Consultant) Paul Feuerstadt, MD, FACG, Ferring/Rebiotix Pharmaceuticals (Consultant, Scientific Research Study Investigator, Speaker's Bureau)Finch Pharmaceuticals (Scientific Research Study Investigator)Merck and Co (Speaker's Bureau)SERES Therapeutics (Consultant, Scientific Research Study Investigator)Takeda Pharmaceuticals (Consultant) Elaine E. Wang, MD, Seres Therapeutics (Employee) Elaine E. Wang, MD, Seres Therapeutics (Employee, Shareholder) Barbara McGovern, MD, Seres Therapeutics (Employee, Shareholder) Lisa von Moltke, MD, Seres Therapeutics (Employee, Shareholder)

928. Clinical Characteristics and Microbiology Testing Patterns Among Transplant Recipients Admitted to Acute Care Hospitals for Suspected Infection

Background Solid organ transplant (SOT) is a growing option for patients with end-stage organ diseases. Immunosuppressive therapy (IT) is utilized in this population to minimize risk of allograft rejection, which increases infection risk particularly of atypical pathogens that can complicate the infection-related diagnostic journey. The purpose of this analysis was to evaluate baseline clinical characteristics and microbiological testing utilization patterns among a cohort of patients with a history of SOT and IT. Methods This retrospective cohort study utilized a US hospital-based, service-level database. Patients were selected from a subsample of database facilities utilizing plasma microbial cell-free DNA diagnostic assays. The study period was 1/1/2017-3/21/2020. Eligible patients were identified by 1st observation of SOT status and IT. Subsequent inpatient admissions for suspected infection were analyzed. Results We identified 749 patients with SOT history and use of IT, 56.4% were male, and the mean age was 52.8 (18.7) years. Kidney was the most prevalent transplant category (49.1%), followed by liver (14.1%), lung (10.9%), and heart (10.3%), and 9.7% were multi-organ. Patients experiencing multiple transplants had the most chronic conditions with a mean Elixhauser comorbidity score of 26.3 (14.7). The median length of stay was 4 [3-7] days. The median number of tests per encounter was 6 [IQR=3-11]. Culture was the most utilized test category (2 [1-4]). Blood culture was the highest utilized culture and overall test at 13.5% of all tests observed, while CMV PCR (7.8%) and multi-panel EIA (2.7%) were the most frequent molecular and antigen tests, respectively. Lung transplant recipients had the greatest utilization of tests overall (9 [3.5-17]) versus other transplant categories (6 [3-10]), consistent with the observed test rate in the 1st 48 hours of presentation (4 [1-7] vs. 2 [1-5]). Conclusion This analysis suggests that the infection-related diagnostic journey among patients with a history of SOT involves high utilization of microbiological testing, with greater utilization among lung transplant recipients versus other SOT recipients. Variation in clinical characteristics and microbiological testing patterns were observed across SOT categories. Disclosures T Matthew Hill, PharmD, PhD, Karius, Inc (Employee, Shareholder) Erick R. Scott, MD, MHS, Karius, Inc (Employee, Shareholder) Sivan Bercovici, PhD, Karius (Employee)

Biomarker testing patterns and actionability in advanced non-small cell lung cancer (aNSCLC) at OneOncology (OneOnc).

287 Background: Recent approvals of targeted treatments (tx) have improved personalized care in aNSCLC. Biomarker testing is crucial for patients (pts) to receive optimal tx expeditiously. This study examined aNSCLC biomarker testing and tx patterns at OneOnc. Methods: Pts diagnosed with aNSCLC (stage ≥ IIIb) from 1/1/2015 to 5/31/2020, aged ≥ 18 years, and with ≥ 1 visit ≤ 90 days of advanced (Adv) diagnosis (Dx) were retrospectively evaluated using the nationwide Flatiron Health electronic health record derived de-identified database from selected OneOnc sites. Descriptive analyses were conducted to evaluate testing patterns for ALK, BRAF, EGFR, KRAS, PD-L1, and ROS-1 biomarkers and actionable mutation tx pattern. Results: Overall 3,860 aNSCLC pts were included, median age was 69 years, 47% females, 66% non-squamous, 29% squamous, 4% histology NOS, and 23% with ECOG performance status 0-1. Of the 3,152 (82%) pts tested for any biomarker, 64% received next-generation sequencing (NGS) vs. 36% received other biomarker tests only. Testing rates varied by biomarker: EGFR (74%), ALK (72%), ROS-1 (66%), PD-L1 (57%), BRAF (56%), KRAS (54%). Pts who received all 6 biomarker tests increased from 12% (2015), 23% (2016), 40% (2017), 41% (2018), 48% (2019) to 56% (2020). Among the tested pts, the median time from Adv Dx to the first test result was 20 days (d) and from specimen collection after Adv Dx to the first test result was 12 d. Pts tested and treated before test result available declined from 28% (2015) to 16% (2020). Of 1,207 pts with actionable mutations, 390 (32%) received tx before the test result: 35% chemotherapy (chemo) only, 28% chemo + cancer immunotherapy (CIT), and 15% CIT only. After the test result, 26% to 81% of pts received no or other tx not specific to actionable mutations [Table]. Conclusions: Findings from this study demonstrated an increase in aNSCLC biomarker testing at OneOnc over time, while 44% pts in 2020 did not receive testing on all 6 biomarkers. Some pts had tx prior to the test result, but this trend appeared to decline. Further studies are warranted to better understand the reasons for pts receiving tx that were not specific to their actionable mutations.[Table: see text]

NGS testing patterns in advanced non-small cell lung cancer (aNSCLC) and metastatic breast cancer (mBC): OneOncology (OO) sites compared to Flatiron Health Nationwide (NAT).

288 Background: Personalized treatment (tx) decisions can be improved through diagnostic tests with NGS by detecting different actionable mutations. OO, a research-focused network of community practices, has a network-wide precision oncology initiative and has advocated for NGS testing in advanced cancers since 2019. This study evaluated NGS testing patterns in aNSCLC and mBC populations descriptively in OO community sites and Flatiron Health NAT. Methods: This study used the Flatiron Health EHR derived de-identified database from [1] four OO sites, and [2] NAT. Patients (pts) diagnosed (Dx) with aNSCLC (stage ≥ IIIb) or mBC from 1/1/2015 to 5/31/2020, aged ≥ 18 years, had ≥ 1 visit ≤ 90 days (d) of advanced or metastatic Dx, and had ≥ 1 biomarker test were included. NAT NGS was confirmed via abstraction from patient records. Descriptive analyses were conducted to assess NGS testing patterns and pts characteristics by tumor type. Results: Of biomarker tested pts at OO vs. NAT (community:academic: 90%:10% aNSCLC; 93%:7% mBC), 2,029 of 3,152 (64%) OO vs. 13,681 of 29,572 (46%) NAT in aNSCLC and 514 of 1,282 (40%) OO vs. 2,458 of 12,175 (20%) NAT in mBC received NGS ± other tests. Testing rate of all 5 aNSCLC biomarkers (ALK, BRAF, EGFR, ROS-1, and KRAS) was higher with NGS vs. other tests for OO (87% vs. 6%) and NAT (87% vs. 11%). In mBC, a higher testing rate of BRCA with NGS vs. other tests (OO: 68% vs. 26%, NAT: 71% vs. 28%) and similar testing rate on HER2 (OO: 98% vs. 98%, NAT: 100% vs. 99%). Median time from Dx to NGS test result at OO vs. NAT was 33 d vs. 32 d in aNSCLC and 70 d vs. 188 d in mBC. NGS testing rates increased over time, with higher rates at OO vs. NAT [Table]. Pts with NGS vs. other tests were slightly younger in aNSCLC (OO: 68 y vs. 70 y, p = 0.001; NAT: 69 y vs. 70 yr, p < 0.001) and mBC (OO: 61 y vs. 67 y, p < 0.001; NAT: 61 y vs. 66 y, p < 0.001), and slightly more commercially insured in aNSCLC (OO: 48% vs. 45%, p = 0.3; NAT: 37% vs. 33%, p < 0.001) and mBC (OO: 54% vs. 48% OO, p = 0.053; NAT: 42 % vs. 36 %, p < 0.001). Conclusions: The adoption of NGS differed by cancer type and NGS testing rates have increased over time in aNSCLC and mBC. While some pts may have received testing outside of the Flatiron network, OO had a higher NGS uptake than NAT, and had a shorter time to testing in mBC that was possibly related to a network wide strategy recommending testing at Dx of advanced disease. Future studies on tx pattern after NGS testing are warranted to improve the actionability of NGS to foster personalized tx. [Table: see text]

Real-world EGFR testing patterns among U.S. patients with advanced NSCLC.

298 Background: Guidelines recommend mutation testing prior to initiating first-line therapy (L1) for patients with advanced non-small cell lung cancer (aNSCLC), including testing for epidermal growth factor receptor (EGFR) mutations. Guidelines also recommend further testing after patients progress on targeted therapy to identify possible resistance mutations and inform selection of subsequent treatment. The present study characterized real-world (RW) EGFR testing patterns in aNSCLC patients. Methods: This retrospective observational cohort study evaluated an aNSCLC cohort (stage IIIB or higher) from the Flatiron Health database. Eligible patients were newly diagnosed, ≥18 years old, and received at least one line of therapy (LOT) during 2015-2020. Patients were followed to last activity, death, or end of the study period, whichever came first. Results: The final study cohort included a total of 22,726 patients. Of these, 17,165 (75.5%) had ≥1 EGFR test and 2,611 (15.2% of the tested study population) tested positive for EGFR mutations at any time during the study. The proportion of aNSCLC patients receiving EGFR testing prior to L1 initiation increased from 45.2% in 2015 to 56.6% in 2020. Among 11,791 patients who had an EGFR test prior to initiation of L1, 1,944 (16.5%) had positive results. Of these, the cumulative proportion of patients receiving a second EGFR test prior to L2, L3, and L4 (contingent on receiving the following LOT) was 36.3% (322 out of 888), 55.4% (241 out of 435), and 67.0% (146 out of 218), respectively. From 2015-2020, the median time from sample collection to EGFR test results decreased from 26 to 16 days for next generation sequencing (NGS) tests and from 17.5 to 12 days for polymerase chain reaction (PCR) tests. During this same period, the proportion of NGS EGFR tests increased from 27.6% to 75.0%. In addition, use of plasma samples for EGFR tests increased substantially (6.3% to 36.5%). Conclusions: The proportion of aNSCLC patients tested for EGFR mutations has increased over time (2015-2020). Median time to available EGFR test result has decreased for both NGS and PCR tests. NGS testing has become increasingly more common as has use of plasma samples. However, opportunities for improvement exist, including further increasing the proportion of patients tested prior to initiation of L1 therapy, as well as testing prior to the selection of subsequent therapy when indicated. As more targeted therapies emerge and resistance mutations to current therapies are further characterized, continued refinement of RW testing practices is required to optimize patient care.