BACKGROUND:Blood-based multicancer early detection tests have potential for clinical benefit, but key performance metrics for cancer screening, such as sojourn time and sensitivity of these tests for detecting preclinical disease, are not yet known. METHODS:In a retrospective analysis of stored plasma samples in the American Cancer Society Cancer Prevention Study 3 (CPS3), GRAIL's MCED test was evaluated for detectability prior to cancer diagnosis. Classical state-transition models for cancer screening were modified to characterize the natural history of ctDNA-shedding cancers. The sensitivity estimand for detecting preclinical cancers by the MCED test was proposed in the context of retrospective testing, and a Bayesian likelihood method was developed to estimate preclinical detectable duration and sensitivity. RESULTS:Analysis of CPS3 data showed that for the 12 prespecified cancers that represent two-thirds of cancer deaths in the United States, the test had 64% estimated overall sensitivity across all stages, and 43% sensitivity during an average 1.36-year preclinical detectable window before metastasis. Untestable assumptions on state transition and sensitivity are discussed, along with limitations related to using stored plasma samples. CONCLUSIONS:State-transition models were developed for retrospective analysis of plasma samples from the CPS3. Estimates for the length of the preclinical detectable window and the sensitivity at screening support annual MCED screening to intercept late-stage cancers. IMPACT:Retrospective analysis of plasma samples from the CPS3 supports the potential of GRAIL's MCED test to detect cancers early in the preclinical state. See related In the Spotlight, p. 1484.

INTRODUCTION:Previous studies have estimated the mean sojourn and dwell times within stages for commonly screened cancer types. However, little is known about the preclinical detection window of circulating tumor DNA (ctDNA) (ie, ctDNA positivity), which is important for understanding multicancer early detection. METHODS:The duration of preclinical detectability and prognostic value of ctDNA detection was estimated from patients with cancer in two biobank studies: CPS-3, where cancer was diagnosed (n = 1064) within 3 years of a prior blood draw (2006-2013), and CCGA3 (NCT02889978), with a blood draw (2016-2019) concurrent with clinical diagnosis (n = 2604). To infer these quantities, Bayesian models were used for detection rates as a function of time, as well as to infer prognostic effects. RESULTS:Median [credible interval] sojourn times were 0.75 [0.47, 1.30], 0.89 [0.61, 1.33], 1.2 [0.84, 1.67] years for cancers diagnosed at local, regional, and distant stage, respectively, and ranged by type from pancreas, 0.49 [0.26, 0.88] to lymphoma, 2.45 [1.14, 4.87]. The extrapolated effect of ctDNA positivity at clinical diagnosis versus negative in CPS-3 cancer cases was a relative hazard ratio of 1.98 [1.08-4.22] for mortality. CONCLUSIONS:These results provide estimates for average ctDNA detectable sojourn time in tumors across multiple cancer sites and stages and can inform the design of future screening studies for multicancer early detection.

Background Cancer biobank studies can inform the natural history of cancer and possibilities for earlier detection. The authors investigated the detectability of a cell-free DNA methylation-based cancer signal in biobanked plasma samples from the large American Cancer Society Cancer Prevention Study 3.Methods In total, 303,692 participants provided a blood sample at enrollment (2006-2013) and were followed for cancer incidence and mortality through state cancer registries. Samples were obtained from participants with and without a diagnosis of cancer (up to 3 years after enrollment), and their prediagnostic plasma samples were tested for cancer signals.Results For 1371 participants without cancer, zero cancer signals were detected. For 1383 participants who were diagnosed with cancer within a 3-year period, cancer signal detection depended on the time between blood draw and cancer diagnosis. Among all cancers, detectability was 15.2%, 6.0%, and 2.3% for samples collected 1, 2, and 3 year(s) before diagnosis, respectively. Among cancers that went on to be fatal, detectability was 55.1%, 21.6%, and 5.1%, respectively; among nonfatal cancers, detectability was 10.2%, 4.1%, and 1.8%, respectively. Cancer signal detection occurred, on average, 323 days before clinical diagnosis, with 40% of signal detections occurring greater than 1 year prior, including some cases up to 3 years prior.Conclusions Cancer signal detection rates change rapidly as lead time approaches clinical diagnosis. The average lead time of approximately 1 year supports annual screening intervals.

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

In recent years, there has been a surge in the development of new, blood-based, single- and multi-cancer detection tests (SCD and MCD), which can detect cancer signals prior to the onset of symptoms or clinical diagnosis of cancer. Recognizing the need for consensus definitions and standardized evidence development frameworks for these new types of blood tests, the Early Detection and Screening Working Group of the Blood Profiling Atlas in Cancer Consortium, a collaborative initiative dedicated to advancing standards and best practices, developed and published a lexicon for liquid biopsy-based SCD and MCD tests. During the preparation of the lexicon, the group recognized challenges with regard to the definitions of key terms and concepts describing absolute and RR assessment of intended use populations for cancer screening tests. This article captures the working group's discussions on (i) risk assessment including considerations for adapting historical SCD risk terminology like "average risk" and "elevated risk" to MCD tests, (ii) the implications of this terminology for describing intended use populations, and (iii) the existing gaps in evidence for determination of absolute risks.

OBJECTIVE:Multi-cancer early detection (MCED) tests are novel technologies that detect cancer signals from a broad set of cancer types using a single blood sample. The objective of this study was to estimate the effect of screening with an MCED test at different intervals on cancer stage at diagnosis and mortality endpoints. DESIGN:The current model is based on a previously published state-transition model that estimated the outcomes of a screening programme using an MCED test when added to usual care for persons aged 50-79. Herein, we expand this analysis to model the time of cancer diagnosis and patient mortality with MCED screening undertaken using different screening schedules. Screening intervals between 6 months and 3 years, with emphasis on annual and biennial screening, were investigated for two sets of tumour growth rate scenarios: 'fast (dwell time=2-4 years in stage I) and 'fast aggressive' (dwell time=1-2 years in stage I), with decreasing dwell times for successive stages. SETTING:Inputs for the model include (1) published MCED performance measures from a large case-control study by cancer type and stage at diagnosis and (2) Surveillance, Epidemiology and End Results (SEER) data describing stage-specific incidence and cancer-specific survival for persons aged 50-79 in the US for all cancer incidence. OUTCOME MEASURES:We used the following outcome measures: diagnostic yield, stage shift, and mortality. RESULTS:Annual screening under the fast tumour growth scenario was associated with more favourable diagnostic yield. There were 370 more cancer signals detected/year/100,000 people screened, 49% fewer late-stage diagnoses, and 21% fewer deaths within 5 years than usual care. Biennial screening had a similar, but less substantial, impact (292 more cancer signals detected/year/100,000 people screened; 39% fewer late-stage diagnoses, and 17% fewer deaths within 5 years than usual care). Annual screening prevented more deaths within 5 years than biennial screening for the fast tumour growth scenario. However, biennial screening had a higher positive predictive value (54% vs 43%); it was also more efficient per 100,000 tests in preventing deaths within 5 years (132 vs 84), but prevented fewer deaths per year. CONCLUSION:Adding MCED test screening to usual care at any interval could improve patient outcomes. Annual MCED test screening provided more overall benefit than biennial screening. Modelling the sensitivity of outcomes to different MCED screening intervals can inform timescales for investigation in trials.

ABSTRACT Background Blood‐based tests present a promising strategy to enhance cancer screening through two distinct approaches. In the traditional paradigm of “one test for one cancer”, single‐cancer early detection (SCED) tests a feature high true positive rate (TPR) for individual cancers, but high false‐positive rate (FPR). Whereas multi‐cancer early detection (MCED) tests simultaneously target multiple cancers with one low FPR, offering a new “one test for multiple cancers” approach. However, comparing these two approaches is inherently non‐intuitive. We developed a framework for evaluating and comparing the efficiency and downstream costs of these two blood‐based screening approaches at the general population level. Methods We developed two hypothetical screening systems to evaluate the performance efficiency of each blood‐based screening approach. The “SCED‐10” system featured 10 hypothetical SCED tests, each targeting one cancer type; the “MCED‐10” system included a single hypothetical MCED test targeting the same 10 cancer types. We estimated the number of cancers detected, cumulative false positives, and associated costs of obligated testing for positive results for each system over 1 year when added to existing USPSTF‐recommended cancer screening for 100,000 US adults aged 50–79. Results Compared with MCED‐10, SCED‐10 detected 1.4× more cancers (412 vs. 298), but had 188× more diagnostic investigations in cancer‐free people (93,289 vs. 497), lower efficiency (positive predictive value: 0.44% vs. 38%; number needed to screen: 2062 vs. 334), 3.4× the cost ($329 M vs. $98 M), and 150× higher cumulative burden of false positives per annual round of screening (18 vs. 0.12). Conclusions A screening system for average‐risk individuals using multiple SCED tests has a higher rate of false positives and associated costs compared with a single MCED test. A set of SCED tests with the same sensitivity as standard‐of‐care screening detects only modestly more cancers than an MCED test limited to the same set of cancers.

Relationship between PPV and diagnostic tests to save a life for post-CSO-directed workups, shown stratified by cancer signal origin, colored by sex

Overall PPVs modeled for each CSO prediction by age. Cancer incidence increases with age, and in this model, FPs are constant across age groups, so the PPV increases with age. Importantly, despite minor changes in relative incidence of cancer types with age, the PPV still indicates a noticeable risk of cancer within each age range. The dashed line represents a threshold PPV of 7%, typically justifying workup.

PURPOSE:Cancer survivors are at risk for recurrence and second primaries, yet often lack clear guidance for long-term surveillance. METHODS:We assessed the performance of a blood-based multicancer early detection (MCED) test in 1,609 survivors who participated in PATHFINDER, a prospective study of adults without current suspicion of cancer. RESULTS:Previous cancers included breast (47%), melanoma (10%), prostate (9%), colorectal (4%), and lymphoma (4%). Average time since diagnosis was 11.2 years, and a cancer signal was detected in 1.2% (20/1,609). Ten new cancer diagnoses occurred: 5 second primaries (stage I uterine, stage II sarcoma, stage III ovarian, stage IV lymphoma, and stage IV colorectal) 8-15 years after original diagnosis and five recurrences (breast cancer) 4-11 years after original diagnosis. Test performance metrics were similar in those with and without previous cancer history. CONCLUSION:These findings highlight the potential of MCED tests to address a significant unmet need in long-term surveillance of survivors.

Diagnostic tests per lives saved for CSO-directed workups, age bands covering 50-80 years, incidence as default for SEER (“any” smoking status as smoking status is unknown in SEER)

Introduction: National surveillance efforts have reported rural-urban disparities in childhood vaccination coverage by metropolitan statistical area designations, measured at the county level. This study's objective was to quantify vaccination trends using more discrete measures of coverage and rurality than prior work. Methods: Serial, cross-sectional analyses of National Immunization Survey-Child restricted-use data collected in 2015-2021 for U.S. children born 2014-2018 were conducted. ZIP code of residence was merged with rural-urban commuting area codes. Vaccination coverage and patterns, including on-time receipt of recommended vaccines, were assessed using vaccinations recorded from birth through age 23 months. To determine whether trends differed by rurality, an interaction between birth year and RUCA was tested in multivariable regression models. Analyses were conducted in November 2023-January 2024. Results: In nationally representative analyses of N=59,361 children, 87.7%, 7.1%, and 5.3% lived in urban, large rural, or small town/rural areas, respectively. Among children born in 2018, coverage for the combined 7-vaccine series was 71.2% (95% CI=69.6%, 72.9%) in urban, 64.9% (95% CI=58.8%, 71.0%) in large rural, and 62.6% (95% CI=56.2%, 68.9%) in small town/rural areas. There was a positive trend in on-time vaccination in urban areas (adjusted prevalence ratio [aPR] for birth year=1.06; 95% CI=1.05, 1.08). While the trend did not significantly differ for large rural versus urban areas (interaction aPR=1.02; 95% CI=0.96, 1.08), there was less improvement in on- time vaccination in small town/rural areas (interaction aPR=0.93; 95% CI=0.88, 0.99). Conclusions: Increased efforts are needed to eliminate disparities in routine and on-time vaccination for rural children.

Extrapolated prevalence performance numbers

State transition diagram for the interception model, expanded to show the five possible trajectories of cancer detectability by stage, and the potential detection by MCED or by usual care

Background It is unclear what proportion of the population cancer burden is covered by current implementation of USPSTF A/B screening recommendations.Objective We estimated the proportion of all US cancer deaths caused by cancer types not covered by screening recommendations or cancer types covered but unaddressed by current implementation.Methods We used 2018-2019 National Center for Health Statistics mortality data, Surveillance, Epidemiology, and End Results registries incidence-based mortality data, and published estimates of screening eligibility and receipt.Results Of approximately 600,000 annual cancer deaths in the US, 31.4% were from screenable cancer types, including colorectal, female breast, cervical, and smoking-associated lung cancers. Further accounting for the low receipt of lung cancer screening reduced the proportion to 17.4%; accounting for receipt of other screening reduced it to 12.8%. Thus, we estimated that current implementation of recommended screening may not address as much as 87.2% of cancer deaths-including 30.4% from individually uncommon cancer types unlikely ever to be covered by dedicated screening.Conclusions The large proportion of cancer deaths unaddressed by current screening represents a major opportunity for improved implementation of current approaches, as well as new multi-cancer screening technologies.

Overall PPVs modeled for each cancer signal origin prediction by smoking status, accounting for uncertainty in sensitivity, specificity, and cancer signal origin assignment

Breakdown of total numbers of tests by type under either a strategy using CSO-directed workups followed by post-CSO non-CSO-directed workups

Multi-cancer early detection (MCED) tests may detect a broad spectrum of cancer types, including uncommon types that lack recommended screening. After a cancer signal is detected by an MCED test, some diagnostic process must definitively confirm the patient's cancer status. A commercially available blood-based MCED test detects a cancer signal and then predicts an anatomic location, a cancer signal origin (CSO), to guide the diagnostic process. We extended a preexisting model for MCED cancer screening, adding predicted CSO categories and a simple model of the diagnostic chain. We then predicted outcomes of the diagnostic chain for each predicted CSO and in populations with differing clinical risk factors. Typical positive predictive values were>40%, and using a minimal sufficient level of positive predictive value (>7%), (i) diagnostic workup based on any CSO was generally warranted, and (ii) continued workup for cancers in locations other than the CSO was justifiable. The benefit of prediction-directed workups was also observed via estimated clinical utility metrics, such as lives saved per diagnostic test, and remained applicable in populations with varying cancer risk, such as lung cancer prediction-directed workups in never-smokers. CSO predictions may enable most true-positive cases to be resolved by short and efficient diagnostic processes. The model predicted a large enough conditional benefit to warrant diagnostic workup based on any CSO prediction from an MCED test, assuming late-stage reduction by MCED leads to mortality reduction, which remains to be demonstrated. SIGNIFICANCE:MCED tests may detect a signal from many cancers. Predicting an anatomic location from which the cancer signal may originate allows effective, usual diagnostic workup. In this study, we show that these predictions are beneficial to physicians choosing a diagnostic path, even for uncommon cancer types and among populations with differing cancer risks.

Diagnostic tests per lives saved for post-CSO-directed workups, age bands covering 50-80 years, incidence as default for SEER

All PPVs in the modeled diagnostic chain for 65- to 69-year-old females with any smoking status, accounting for uncertainty in sensitivity, specificity, and cancer signal origin assignment