Strong association between reduction of late-stage cancers and reduction of cancer-specific mortality in meta-regression of randomized screening trials across multiple cancer types.

BACKGROUND: Late-stage cancer incidence has been proposed as an early surrogate for mortality in randomized controlled trials (RCTs) of cancer screening; however, its validity has not been systematically evaluated across screening RCTs of different cancers. METHODS: We conducted a meta-regression analysis of cancer screening RCTs that reported both late-stage cancer incidence and cancer mortality. Based on a systematic literature review, we included 33 RCTs of screening programs targeting seven cancer types, including lung (n = 12), colorectal (n = 8), breast (n = 5), and prostate (n = 4), among others. We regressed the relative reduction of cancer mortality on the relative reduction of late-stage cancer incidence, inversely weighted for each RCT by the variance of estimated mortality reduction. RESULTS: Across cancer types, the relative reduction of late-stage cancer incidence was linearly associated with the relative reduction of cancer mortality. Specifically, we observed this association for lung (R2 = 0.79 and 0.996 in three recent large trials), breast (R2 = 0.94), prostate (R2 = 0.98), and colorectal cancer (R2 = 0.75 for stage III/IV cancers and 0.93 for stage IV cancers). Trials with a 20% or greater reduction in late-stage cancers were more likely to achieve a significant reduction in cancer mortality. Our results also showed that no reduction of late-stage cancer incidence was associated with no or minimal reduction in cancer mortality. CONCLUSIONS: Meta-regression of historical screening RCTs showed a strong linear association between reductions in late-stage cancer incidence and cancer mortality.

Impact of screening participation on modelled mortality benefits of a multi-cancer early detection test by socioeconomic group in England.

BACKGROUND: Cancer burden is higher and cancer screening participation is lower among individuals living in more socioeconomically deprived areas of England, contributing to worse health outcomes and shorter life expectancy. Owing to higher multi-cancer early detection (MCED) test sensitivity for poor-prognosis cancers and greater cancer burden in groups experiencing greater deprivation, MCED screening programmes may have greater relative benefits in these groups. We modelled potential differential benefits of MCED screening between deprivation groups in England at different levels of screening participation. METHODS: We applied the interception multi-cancer screening model to cancer incidence and survival data made available by the National Cancer Registration and Analysis Service in England to estimate reductions in late-stage diagnoses and cancer mortality from an MCED screening programme by deprivation group across 24 cancer types. We assessed the impact of varying the proportion of people who participated in annual screening in each deprivation group on these estimates. RESULTS: The modelled benefits of an MCED screening programme were substantial: reductions in late-stage diagnoses were 160 and 274 per 100 000 persons in the least and most deprived groups, respectively. Reductions in cancer mortality were 60 and 99 per 100 000 persons in the least and most deprived groups, respectively. Benefits were greatest in the most deprived group at every participation level and were attenuated with lower screening participation. CONCLUSIONS: For the greatest possible population benefit and to decrease health inequalities, an MCED implementation strategy should focus on enhancing equitable, informed participation, enabling equal participation across all socioeconomic deprivation groups. TRIAL REGISTRATION NUMBER: NCT05611632.

Clinical performance and utility: A microsimulation model to inform the design of screening trials for a multi-cancer early detection test.

OBJECTIVES: Designing cancer screening trials for multi-cancer early detection (MCED) tests presents a significant methodology challenge, as natural histories of cell-free DNA-shedding cancers are not yet known. A microsimulation model was developed to project the performance and utility of an MCED test in cancer screening trials. METHODS: Individual natural history of preclinical progression through cancer stages for 23 cancer classes was simulated by a stage-transition model under a broad range of cancer latency parameters. Cancer incidences and stage distributions at clinical presentation in simulated trials were set to match the data from Surveillance, Epidemiology, and End Results program. One or multiple rounds of annual screening using a targeted methylation-based MCED test (GalleriⓇ) was conducted to detect preclinical cancers. Mortality benefit of early detection was simulated by a stage-shift model. RESULTS: In simulated trials, accounting for healthy volunteer effect and varying test sensitivity, positive predictive value in the prevalence screening round reached 48% to 61% in 6 natural history scenarios. After 3 rounds of annual screening, the cumulative proportions of stage I/II cancers increased by approximately 9% to 14%, the incidence of stage IV cancers was reduced by 37% to 46%, the reduction of stages III and IV cancer incidences was 9% to 24%, and the reduction of mortality reached 13% to 16%. Greater reductions of late-stage cancers and cancer mortality were achieved by five rounds of MCED screening. CONCLUSIONS: Simulation results guide trial design and suggest that adding this MCED test to routine screening in the United States may shift cancer detection to earlier stages, and potentially save lives.

Avoiding lead-time bias by estimating stage-specific proportions of cancer and non-cancer deaths.

PURPOSE: Understanding how stage at cancer diagnosis influences cause of death, an endpoint that is not susceptible to lead-time bias, can inform population-level outcomes of cancer screening. METHODS: Using data from 17 US Surveillance, Epidemiology, and End Results registries for 1,154,515 persons aged 50-84 years at cancer diagnosis in 2006-2010, we evaluated proportional causes of death by cancer type and uniformly classified stage, following or extrapolating all patients until death through 2020. RESULTS: Most cancer patients diagnosed at stages I-II did not go on to die from their index cancer, whereas most patients diagnosed at stage IV did. For patients diagnosed with any cancer at stages I-II, an estimated 26% of deaths were due to the index cancer, 63% due to non-cancer causes, and 12% due to a subsequent primary (non-index) cancer. In contrast, for patients diagnosed with any stage IV cancer, 85% of deaths were attributed to the index cancer, with 13% non-cancer and 2% non-index-cancer deaths. Index cancer mortality from stages I-II cancer was proportionally lowest for thyroid, melanoma, uterus, prostate, and breast, and highest for pancreas, liver, esophagus, lung, and stomach. CONCLUSION: Across all cancer types, the percentage of patients who went on to die from their cancer was over three times greater when the cancer was diagnosed at stage IV than stages I-II. As mortality patterns are not influenced by lead-time bias, these data suggest that earlier detection is likely to improve outcomes across cancer types, including those currently unscreened.

Potential for Cure by Stage across the Cancer Spectrum in the United States.

BACKGROUND: Cure fraction-the proportion of persons considered cured of cancer after long-term follow-up-reflects the total impact of cancer control strategies, including screening, without lead-time bias. Previous studies have not reported stage-stratified cure fraction across the spectrum of cancer types. METHODS: Using a mixture cure model, we estimated cure fraction across stages for 21 cancer types and additional subtypes. Cause-specific survival for 2.4 million incident cancers came from 17 US Surveillance, Epidemiology, and End Results registries for adults 40 to 84 years at diagnosis in 2006 to 2015, followed through 2020. RESULTS: Across cancer types, a substantial cure fraction was evident at early stages, followed by either a sharp drop from stages III to IV or a steady decline from stages I to IV. For example, estimated cure fractions for colorectal cancer at stages I, II, III, and IV were 62% (95% confidence interval: 59%-66%), 61% (58%-65%), 58% (57%-59%), and 7% (7%-7%), respectively. Corresponding estimates for gallbladder cancer were 50% (46%-54%), 24% (22%-27%), 22% (19%-25%), and 2% (2%-3%). Differences in 5-year cause-specific survival between early-stage and stage IV cancers were highly correlated with between-stage differences in cure fraction, indicating that survival gaps by stage are persistent and not due to lead-time bias. CONCLUSIONS: A considerable fraction of cancer is amenable to cure at early stages, but not after metastasis. IMPACT: These results emphasize the potential for early detection of numerous cancers, including those with no current screening modalities, to reduce cancer death.

Performance of a Cell-Free DNA-Based Multi-cancer Detection Test in Individuals Presenting With Symptoms Suspicious for Cancers.

PURPOSE: A multi-cancer detection test using a targeted methylation assay and machine learning classifiers was validated and optimized for screening in prospective, case-controlled Circulating Cell-free Genome Atlas (ClinicalTrials.gov identifier: NCT02889978) substudy 3. Here, we report test performance in a subgroup of participants with symptoms suspicious for cancer to assess the test's ability to potentially facilitate efficient diagnostic evaluation in symptomatic individuals. METHODS: We evaluated test performance (sensitivity, specificity, and accuracy of cancer signal origin [CSO] prediction accuracy) in participants with clinically presenting cancers (CPCs) and noncancer with underlying medical conditions and among two subgroups (65 years and older and GI cancers). Overall survival (OS) of participants who had a cancer signal detected/not detected was compared with SEER-based expected survival. RESULTS: A total of 2,036 cancer and 1,472 noncancer participants were included. Specificity was high in all noncancer participants (99.5% [95% CI, 98.4 to 99.8]). In participants with CPCs, the overall sensitivity was 64.3% (95% CI, 62.2 to 66.4) and the overall accuracy of CSO prediction in true positives was 90.3%. For GI cancers, the overall sensitivity was 84.1% (95% CI, 80.6 to 87.1). In participants 65 years and older, test performance was similar to that of all participants. Individuals with cancers not detected had a significantly better OS than that expected from SEER (P < .01). CONCLUSION: This test detected a cancer signal with high specificity and CSO prediction accuracy and moderate sensitivity in symptomatic individuals, with especially high performance in participants with GI cancers. The survival analysis implied that the cancers not detected were less clinically aggressive than cancers detected by the test, providing prognostic insights to physicians. This multi-cancer detection test could facilitate efficient workup and stratify cancer risk in symptomatic individuals.

Modelled mortality benefits of multi-cancer early detection screening in England.

BACKGROUND: Screening programmes utilising blood-based multi-cancer early detection (MCED) tests, which can detect a shared cancer signal from any site in the body with a single, low false-positive rate, could reduce cancer burden through early diagnosis. METHODS: A natural history ('interception') model of cancer was previously used to characterise potential benefits of MCED screening (based on published performance of an MCED test). We built upon this using a two-population survival model to account for an increased risk of death from cfDNA-detectable cancers relative to cfDNA-non-detectable cancers. We developed another model allowing some cancers to metastasise directly from stage I, bypassing intermediate tumour stages. We used incidence and survival-by-stage data from the National Cancer Registration and Analysis Service in England to estimate longer-term benefits to a cohort screened between ages 50-79 years. RESULTS: Estimated late-stage and mortality reductions were robust to a range of assumptions. With the least favourable dwell (sojourn) time and cfDNA status hazard ratio assumptions, we estimated, among 100,000 screened individuals, 67 (17%) fewer cancer deaths per year corresponding to 2029 fewer deaths in those screened between ages 50-79 years. CONCLUSION: Realising the potential benefits of MCED tests could substantially reduce late-stage cancer diagnoses and mortality.

Surgical resection as a predictor of cancer-specific survival by stage at diagnosis and cancer type, United States, 2006-2015.

BACKGROUND: When solid tumors are amenable to definitive resection, clinical outcomes are generally superior to when those tumors are inoperable. However, the population-level cancer survival benefit of eligibility for surgery by cancer stage has not yet been quantified. METHODS: Using Surveillance, Epidemiology and End Results data allowing us to identify patients who were deemed eligible for and received surgical resection, we examined the stage-specific association of surgical resection with 12-year cancer-specific survival. The 12-year endpoint was selected to maximize follow-up time and thereby minimize the influence of lead time bias. RESULTS: Across a variety of solid tumor types, earlier stage at diagnosis allowed for surgical intervention at a much higher rate than later-stage diagnosis. At every stage, surgical intervention was associated with a substantially higher rate of 12-year cancer-specific survival, with absolute differences of up to 51% for stage I, 51% for stage II, and 44% for stage III cancer, and stage-specific mortality relative risks of 3.6, 2.4, and 1.7, respectively. CONCLUSIONS: Diagnosis of solid cancers in early stages often enables surgical resection, which reduces the risk of death from cancer. Receipt of surgical resection is an informative endpoint that is strongly associated with long-term cancer-specific survival at every stage.

Modeled residual current cancer risk after clinical investigation of a positive multicancer early detection test result.

BACKGROUND: Positive results of a multi-cancer early detection (MCED) test require confirmatory diagnostic workup. Here, residual current cancer risk (RR) during the process of diagnostic resolution, including situations where the initial confirmatory test does not provide resolution, was modeled. METHODS: A decision-tree framework was used to model conditional risk in a patient's journey through confirmatory diagnostic options and outcomes. The diagnostic journey assumed that cancer signal detection (a positive MCED test result) had already led to a transition from screening to diagnosis and began with an initial positive predictive value (PPV) from the positive result. Evaluation of a most probable (top) predicted cancer signal origin (CSO) and then a second-most probable predicted CSO followed. Under the assumption that the top- and second-predicted CSOs were each followed by a targeted confirmatory test, the RR was estimated for each subsequent scenario. RESULTS: For an initial MCED test result with typical performance characteristics modeled (PPV, 40%; top-predicted CSO accuracy, 90%), after a negative initial confirmatory test (sensitivity, 70%, 90%, or 100%) the RR ranged from 6% to 20%. A second-predicted CSO (accuracy, 50%), after a negative second confirmatory test, still provided a significant RR (3%-18%) in comparison with the National Institute for Health and Care Excellence-recommended cancer risk threshold warranting investigation in symptomatic individuals (3%). With a 40% PPV for an MCED test and 90% specificity for a confirmatory test, the risk of incidental findings after one or two confirmatory tests was 6% and 12%, respectively. CONCLUSIONS: These results may illustrate the impact of a positive MCED test on follow-up decision-making.

Performance of a targeted methylation-based multi-cancer early detection test by race and ethnicity.

Disparities in cancer screening and outcomes based on factors such as sex, socioeconomic status, and race and ethnicity in the United States are well documented. A blood-based multi-cancer early detection (MCED) test that detects a shared cancer signal across multiple cancer types and also predicts the cancer signal origin was developed and validated in the Circulating Cell-free Genome Atlas study (CCGA; NCT02889978). CCGA is a prospective, multicenter, case-control, observational study with longitudinal follow-up (overall N = 15,254). In this pre-specified, exploratory, descriptive analysis, test performance was evaluated among racial and ethnic groups. Overall, 4077 participants comprised the independent validation set with confirmed cancer status (cancer: n = 2823; non-cancer: n = 1254). Participants were stratified into the following racial/ethnic groups: Black (non-Hispanic), Hispanic (all races), Other (non-Hispanic), Other/unknown and White (non-Hispanic). Cancer and non-cancer participants were predominantly White (n = 2316, 82.0% and n = 996, 79.4%, respectively). Across groups, specificity for cancer signal detection ranged from 98.1% [n = 103; 95% CI: 93.2-99.5%] to 100% [n = 85; 95% CI: 95.7-100.0%]. The sensitivity for cancer signal detection across groups ranged from 43.9% [n = 57; 95% CI: 31.8-56.7%] to 63.0% [n = 192; 95% CI: 56.0-69.5%] and generally increased with clinical stage. The MCED test had consistently high specificity and similar sensitivity across racial and ethnic groups, though results are limited by sample size for some groups. Results support the broad applicability of this MCED test and clinical implementation on a population scale as a complement to standard screening.

Multi-cancer early detection test sensitivity for cancers with and without current population-level screening options.

There are four solid tumors with common screening options in the average-risk population aged 21 to 75 years (breast, cervical, colorectal, and, based on personalized risk assessment, prostate), but many cancers lack recommended population screening and are often detected at advanced stages when mortality is high. Blood-based multi-cancer early detection tests have the potential to improve cancer mortality through additional population screening. Reported here is a post-hoc analysis from the third Circulating Cell-free Genome Atlas substudy to examine multi-cancer early detection test performance in solid tumors with and without population screening recommendations and in hematologic malignancies. Participants with cancer in the third Circulating Cell-free Genome Atlas substudy analysis were split into three subgroups: solid screened tumors (breast, cervical, colorectal, prostate), solid unscreened tumors, and hematologic malignancies. In this post hoc analysis, sensitivity is reported for each subgroup across all ages and those aged ⩾50 years overall, by cancer, and by clinical cancer stage. Aggregate sensitivity in the solid screened, solid unscreened, and hematologic malignancy subgroups was 34%, 66%, and 55% across all cancer stages, respectively; restricting to participants aged ⩾50 years showed similar aggregate sensitivity. Aggregate sensitivity was 27%, 53%, and 60% across stages I to III, respectively. Within the solid unscreened subgroup, aggregate sensitivity was >75% in 8/18 cancers (44%) and >50% in 13/18 (72%). This multi-cancer early detection test detected cancer signals at high (>75%) sensitivity for multiple cancers without existing population screening recommendations, suggesting its potential to complement recommended screening programs.Clinical trial identifier: NCT02889978.

Evaluation of cell-free DNA approaches for multi-cancer early detection.

In the Circulating Cell-free Genome Atlas (NCT02889978) substudy 1, we evaluate several approaches for a circulating cell-free DNA (cfDNA)-based multi-cancer early detection (MCED) test by defining clinical limit of detection (LOD) based on circulating tumor allele fraction (cTAF), enabling performance comparisons. Among 10 machine-learning classifiers trained on the same samples and independently validated, when evaluated at 98% specificity, those using whole-genome (WG) methylation, single nucleotide variants with paired white blood cell background removal, and combined scores from classifiers evaluated in this study show the highest cancer signal detection sensitivities. Compared with clinical stage and tumor type, cTAF is a more significant predictor of classifier performance and may more closely reflect tumor biology. Clinical LODs mirror relative sensitivities for all approaches. The WG methylation feature best predicts cancer signal origin. WG methylation is the most promising technology for MCED and informs development of a targeted methylation MCED test.

Racial/Ethnic Differences in Cancer Diagnosed after Metastasis: Absolute Burden and Deaths Potentially Avoidable through Earlier Detection.

BACKGROUND: Racial/ethnic disparities in cancer mortality are well described and are partly attributable to later stage of diagnosis. It is unclear to what extent reductions in the incidence of late-stage cancer could narrow these relative and absolute disparities. METHODS: We obtained stage- and cancer-specific incidence and survival data from the Surveillance, Epidemiology, and End Results Program for persons ages 50 to 79 years between 2006 and 2015. For eight hypothetical cohorts of 100,000 persons defined by race/ethnicity and sex, we estimated cancer-related deaths if cancers diagnosed at stage IV were detected earlier, by assigning them outcomes of earlier stages. RESULTS: We observed a 3-fold difference in the absolute burden of stage IV cancer between the group with the highest rate (non-Hispanic Black males, 337 per 100,000) and the lowest rate (non-Hispanic Asian/Pacific Islander females, 117 per 100,000). Assuming all stage IV cancers were diagnosed at stage III, 32-80 fewer cancer-related deaths would be expected across subgroups, a relative reduction of 13%-14%. Assuming one third of metastatic cancers were diagnosed at each earlier stage (I, II, and III), 52-126 fewer cancer-related deaths would be expected across subgroups, a relative reduction of 21%-23%. CONCLUSIONS: Across population subgroups, non-Hispanic Black males have the highest burden of stage IV cancer and would have the most deaths averted from improved detection of cancer before metastasis. IMPACT: Detecting cancer before metastasis could meaningfully reduce deaths in all populations, but especially in non-Hispanic Black populations. See related commentary by Loomans-Kropp et al., p. 512.

Prognostic Significance of Blood-Based Multi-cancer Detection in Plasma Cell-Free DNA.

PURPOSE: We recently reported the development of a cell-free DNA (cfDNA) targeted methylation (TM)-based sequencing approach for a multi-cancer early detection (MCED) test that includes cancer signal origin prediction. Here, we evaluated the prognostic significance of cancer detection by the MCED test using longitudinal follow-up data. EXPERIMENTAL DESIGN: As part of a Circulating Cell-free Genome Atlas (CCGA) substudy, plasma cfDNA samples were sequenced using a TM approach, and machine learning classifiers predicted cancer status and cancer signal origin. Overall survival (OS) of cancer participants in the first 3 years of follow-up was evaluated in relation to cancer detection by the MCED test and clinical characteristics. RESULTS: Cancers not detected by the MCED test had significantly better OS (P < 0.0001) than cancers detected, even after accounting for other covariates, including clinical stage and method of clinical diagnosis (i.e., standard-of-care screening or clinical presentation with signs/symptoms). Additionally, cancers not detected by the MCED test had better OS than was expected when data were adjusted for age, stage, and cancer type from the Surveillance, Epidemiology, and End Results (SEER) program. In cancers with current screening options, the MCED test also differentiated more aggressive cancers from less aggressive cancers (P < 0.0001). CONCLUSIONS: Cancer detection by the MCED test was prognostic beyond clinical stage and method of diagnosis. Cancers not detected by the MCED test had better prognosis than cancers detected and SEER-based expected survival. Cancer detection and prognosis may be linked by the underlying biological factor of tumor fraction in cfDNA.

Modeled Reductions in Late-stage Cancer with a Multi-Cancer Early Detection Test.

BACKGROUND: Cancer is the second leading cause of death globally, with many cases detected at a late stage when prognosis is poor. New technologies enabling multi-cancer early detection (MCED) may make "universal cancer screening" possible. We extend single-cancer models to understand the potential public health effects of adding a MCED test to usual care. METHODS: We obtained data on stage-specific incidence and survival of all invasive cancers diagnosed in persons aged 50-79 between 2006 and 2015 from the US Surveillance, Epidemiology, and End Results (SEER) program, and combined this with published performance of a MCED test in a state transition model (interception model) to predict diagnostic yield, stage shift, and potential mortality reductions. We model long-term (incident) performance, accou.

Projected Reductions in Absolute Cancer-Related Deaths from Diagnosing Cancers Before Metastasis, 2006-2015.

BACKGROUND: New technologies are being developed for early detection of multiple types of cancer simultaneously. To quantify the potential benefit, we estimated reductions in absolute cancer-related deaths that could occur if cancers diagnosed after metastasis (stage IV) were instead diagnosed at earlier stages. METHODS: We obtained stage-specific incidence and survival data from the Surveillance, Epidemiology, and End Results Program for 17 cancer types for all persons diagnosed ages 50 to 79 years in 18 geographic regions between 2006 and 2015. For a hypothetical cohort of 100,000 persons, we estimated cancer-related deaths under assumptions that cancers diagnosed at stage IV were diagnosed at earlier stages. RESULTS: Stage IV cancers represented 18% of all estimated diagnoses but 48% of all estimated cancer-related deaths within 5 years. Assuming all stage IV cancers were diagnosed at stage III, 51 fewer cancer-related deaths would be expected per 100,000, a reduction of 15% of all cancer-related deaths. Assuming one third of metastatic cancers were diagnosed at stage III, one third diagnosed at stage II, and one third diagnosed at stage I, 81 fewer cancer-related deaths would be expected per 100,000, a reduction of 24% of all cancer-related deaths, corresponding to a reduction in all-cause mortality comparable in magnitude to eliminating deaths due to cerebrovascular disease. CONCLUSIONS: Detection of multiple cancer types earlier than stage IV could reduce at least 15% of cancer-related deaths within 5 years, affecting not only cancer-specific but all-cause mortality. IMPACT: Detecting cancer before stage IV, including modest shifts to stage III, could offer substantial population benefit.

High-intensity sequencing reveals the sources of plasma circulating cell-free DNA variants.

Accurate identification of tumor-derived somatic variants in plasma circulating cell-free DNA (cfDNA) requires understanding of the various biological compartments contributing to the cfDNA pool. We sought to define the technical feasibility of a high-intensity sequencing assay of cfDNA and matched white blood cell DNA covering a large genomic region (508 genes; 2 megabases; >60,000× raw depth) in a prospective study of 124 patients with metastatic cancer, with contemporaneous matched tumor tissue biopsies, and 47 controls without cancer. The assay displayed high sensitivity and specificity, allowing for de novo detection of tumor-derived mutations and inference of tumor mutational burden, microsatellite instability, mutational signatures and sources of somatic mutations identified in cfDNA. The vast majority of cfDNA mutations (81.6% in controls and 53.2% in patients with cancer) had features consistent with clonal hematopoiesis. This cfDNA sequencing approach revealed that clonal hematopoiesis constitutes a pervasive biological phenomenon, emphasizing the importance of matched cfDNA-white blood cell sequencing for accurate variant interpretation.

Alignment-free filtering for cfNA fusion fragments.

MOTIVATION: Cell-free nucleic acid (cfNA) sequencing data require improvements to existing fusion detection methods along multiple axes: high depth of sequencing, low allele fractions, short fragment lengths and specialized barcodes, such as unique molecular identifiers. RESULTS: AF4 was developed to address these challenges. It uses a novel alignment-free kmer-based method to detect candidate fusion fragments with high sensitivity and orders of magnitude faster than existing tools. Candidate fragments are then filtered using a max-cover criterion that significantly reduces spurious matches while retaining authentic fusion fragments. This efficient first stage reduces the data sufficiently that commonly used criteria can process the remaining information, or sophisticated filtering policies that may not scale to the raw reads can be used. AF4 provides both targeted and de novo fusion detection modes. We demonstrate both modes in benchmark simulated and real RNA-seq data as well as clinical and cell-line cfNA data. AVAILABILITY AND IMPLEMENTATION: AF4 is open sourced, licensed under Apache License 2.0, and is available at: https://github.com/grailbio/bio/tree/master/fusion.