Considerations in the Implementation of Multicancer Early Detection Tests

Editor’s Note: This post first appeared in Future Oncology on September 5, 2022. It was reprinted under the open-access, creative commons license.

Over the last several decades, cancer screening efforts in the USA have led to substantial reductions in cancer mortality. Data from the Surveillance, Epidemiology, and End Results (SEER) program show that the mortality rate for the most deadly cancers has improved, but progress has not been uniform across all cancer types [1]. Mortality improvements in kidney and lung cancer have been substantial, while pancreatic and urinary cancers have seen no improvements [2]. One aspect underlying this clinical reality is that survival rates are higher when cancer is diagnosed at an earlier stage [3]; accordingly, increases in late-stage cancer diagnoses can be expected to be associated with a corresponding uptick in cancer mortality. A 2019 population-based registry study (n = 10,944 cancer deaths) demonstrated that metastasis present at the time of diagnosis led to nearly two-thirds of cancer deaths in solid tumors [4].

Inequities in cancer screening, diagnosis & outcome

In addition to unequal gains in mortality rates among cancer subtypes, disparities based upon a complex interplay of other behavioral, social, racial, ethnic and economic factors remain [5,6]. The underlying factors that contribute to variation in cancer incidence and/or mortality are difficult to disentangle, but racial and ethnic differences have been associated with inequity in cancer burden [6]. For example, African–Americans have higher cancer mortality rates than any other racial group, consistent with the observation that they are more likely to be diagnosed with a more advanced cancer compared with other racial groups [6,7]. A recent study utilized stage- and cancer-specific incidence and survival data from SEER in cohorts defined by race and ethnicity and sex to estimate the mortality impact of diagnosing stage IV cancers earlier. Consistent with prior findings, non-Hispanic Black males showed the highest burden of stage IV cancer and would have the most deaths avoided with improved cancer detection prior to metastasis [8].

Sign up for Schaeffer Center news

Geographic region is also a factor, with rural counties showing an 8% higher all-cancer mortality rate, and more dramatic gaps exist for certain cancers, such as lung cancer (18–20% higher), which is closely tied to smoking [9]. The direct role that gaps in health insurance coverage plays in these observed disparities is also well documented [10]. Taken together, the literature paints a picture that in the USA the color of your skin, where you live and the quality of your health insurance are significant predictors of outcome. The COVID-19 pandemic has likely widened these inequities. Recent data found a dramatic reduction in cancer screening during the pandemic, leading some oncologists to fear a new wave of late-stage cancer diagnoses [11].

Can innovation in cancer screening narrow the gap?

Given the observed disparity in cancer outcome that is outside biological differences, the need for effective early cancer detection raises the question “Could innovation in cancer screening narrow these persistent gaps?” Our understanding of cancer biology has evolved dramatically from histologic stratification (e.g., in lung cancer) to genetic subtyping. There is a growing recognition that all cancers are unique based on their genetic, transcriptional and phenotypic variations. Differences exist not only across subtypes but also between patients with exactly the same subtype of cancer and even within a single tumor (intertumoral and intratumoral heterogeneity). The therapeutics community has leveraged this understanding to design targeted therapies – almost to the point of saturation [12].

Meanwhile, innovation in cancer screening and diagnostic tools has not kept pace. The United States Preventive Services Task Force (USPSTF) currently relies on long-standing technology such as ionizing radiation (mammogram, low-dose computed tomography), infrequent direct visualization (colonoscopy), microscopy (Pap smear) or low-specificity historical biomarkers (prostate-specific antigen). Some important developments have taken advantage of technological advances, including the use of DNA testing of human papillomavirus to identify strains with the highest carcinogenic risk and the use of artificial intelligence to interpret radiologic findings [13,14]. Even so, we still lack a comprehensive approach to cancer screening, and only a few, select cancers (breast, cervical, colorectal and lung) have level A or B USPSTF-recommended screening protocols [15].

New technologies, powerful tools

How do we harness the principles of cancer biology, DNA sequencing and machine learning to transform how we detect cancer? Designing blood-based multicancer screening that recognizes cancer-specific genomic signatures could enable effective and efficient early identification of cancer. Several multicancer early detection (MCED) tests are in development, and two are either approaching commercial readiness or have recently (June 2021) become available. Because the two technologies apply different genomic approaches, we consider both in turn.

The CancerSEEK screening test developed by researchers at Johns Hopkins University, and now under clinical investigation (Thrive, an Exact Sciences Company), uses a combination of protein biomarkers plus gene mutations to detect eight to ten cancers. This test has received breakthrough designation status from the US FDA for detection of ovarian and pancreatic cancer. A case-control study of 1800 participants generated excitement in the field, reporting 99% specificity, 69–99% sensitivity for eight cancers and localization of tumors in 83% of cases [16]. The group has now studied this approach in a 10,000-person prospective, interventional DETECT-A study [17].

The CancerSEEK test has since evolved with regard to the detection of different proteins, matched white blood cell sequencing, absence of machine learning for localization and addition of diagnostic positron emission tomography-computed tomography (PET-CT) as the final step. The current configuration is based on a four-step process: baseline blood test, confirmatory blood test, formal multidisciplinary tumor board and diagnostic PET-CT scan. This approach utilizes an assay that is not locked, which can hamper one’s ability to draw meaningful conclusions. This newer version of the blood test does not include localization capability and reports a higher specificity for multicancer detection (i.e., a desired false positive rate of <1%) [17].

Combining multiple blood tests and diagnostic PET-CT could introduce complexity in terms of practical implementation, risk, cost and access. The DETECT-A study showed that the four-step combined-modality approach can be implemented in a clinical practice setting, and importantly its use did not reduce adherence to other screening modalities. With confirmed blood testing alone (i.e., clonal hematopoiesis of indeterminate potential was excluded by the confirmation test), the specificity was 98.9% and the positive predictive value (PPV) was 19.4%. The addition of PET-CT to the blood test increased the specificity to 99.6% and the PPV to 28.3% but could limit the accessibility of the test, particularly in rural areas.

Another MCED test – developed by GRAIL, LLC, a subsidiary of Illumina, Inc., currently held separate from Illumina, Inc., under the terms of the Interim Measures Order of the European Commission dated 29 October 2021 – received breakthrough designation status with the FDA in May 2019 [18]. An earlier evaluation of cfDNA approaches for multicancer early detection found that whole-genome methylation was most promising and informed subsequent development of the GRAIL targeted-methylation MCED test. It is a blood-based MCED test utilizing cfDNA sequencing with machine learning to detect methylation-based cancer signals across multiple cancer types and to predict cancer signal origin with high accuracy [19,20]. This localization is possible because aberrant methylation of DNA is a hallmark of cancer, and methylation signals also contain a signature of the tissue of origin, as these methyl groups control gene regulation that influences cell differentiation. The MCED test became available in June 2021 (Galleri®) [18].

Validation results for the MCED test from prespecified substudies of the Circulating Cell-free Genome Atlas study have been published [19,20]. The third and final Circulating Cell-free Genome Atlas substudy revealed 99.5% specificity and 51.5% sensitivity for cancer signal detection across 50+ cancer types [19]. Among 12 prespecified high-signal cancers that account for two-thirds of cancer mortality, sensitivity for cancer signal detection was 76.3%. When projecting these findings to the SEER cancer registry, the PPV for this MCED test was estimated at 44.4% (adjusted to SEER incidence). The validation cohort reported 88.7% accuracy in predicting the origin of the cancer signal [19]. Long-term follow-up of a Circulating Cell-free Genome Atlas cohort also suggested that cancer detection with an MCED test using cfDNA methylation patterns is prognostic, in that cancers detected by the methylation-based MCED test required treatment, whereas those cancers not detected were more likely to be indolent and have a better prognosis [21]. This makes biological sense, as shedding of methylated DNA is associated with cell division rate.

The methylation-based MCED test was evaluated in a target population (adults over the age of 50 years) enrolled within 7 healthcare systems across the USA. PATHFINDER (NCT04241796) is an interventional study that returns test results to healthcare providers and evaluates how test results affect cancer diagnostic and care pathways. Interim data (6629 participants with analyzable results through 19 March 2021) have been presented [22], demonstrating a 1.4% (92/6629) cancer signal detection rate and a preliminary PPV of 44.6% (nearly a third of those with a cancer signal detected have achieved diagnostic resolution to date). The source of the cancer signal was predicted with high accuracy (96.3% accuracy for the first or second cancer signal origin prediction in those with cancer diagnosis).

Some other MCED blood tests beyond the early stages of development include PanSeer (Singlera Genomics), a not-yet-named test (Burning Rock) and PanTum Detect (RMDM Diagnostics). PanSeer utilizes ctDNA methylation patterns to detect five common cancer types at a specificity of 96% [23]. Burning Rock’s test has been studied in lung, colorectal and liver cancer patients with specificities ranging from 96% to 99% [24]. This test also utilizes methylation patterns of ctDNA to detect cancer, while PanTum uses monocyte biomarkers to detect oral squamous cell, cholangiocellular, pancreatic and colorectal carcinomas [25].

Putting theory into practice

Regardless of which approach is taken, new frameworks are needed to assess the benefits, harms and adoption of MCED tests. Genomics and machine learning are now challenging our existing assumptions and heuristics for how to assess MCED tests. For example, the healthcare system thus far has been operating under a single-cancer screening paradigm, in which clinicians, regulators and payers have come to expect decades-long, government-funded research programs to define cancer-specific survival in single cancers. We have unassailable randomized controlled trial (RCT) evidence that early cancer detection saves lives across many solid tumors (e.g., lung, breast, colon and cervical) and other trial or observational evidence for prostate, esophageal and pancreatic cancers, among others [26–28]. To date, findings reported for MCED tests have been based on interventional or observational trials with large patient populations enriched to increase the number of observed cancers (e.g., due to advanced age [NCT04241796, NCT05155605] or in those already undergoing cancer screening [NCT03085888]).

Screening for a shared cancer signal across multiple cancers broadly as opposed to the current approach of single-cancer screening requires a new mindset. This includes acknowledging that lengthy, separate RCTs that were conducted to demonstrate the benefit of single-cancer screening are not realistic in the evaluation of MCED tests, particularly given unavoidable challenges in achieving sufficient enrollment for rare cancers as well as the time required to establish mortality benefits for individual cancers. As one solution, a large RCT (NHS-Galleri; ISRCTN 91431511) is assessing MCED test performance and clinical utility collectively across cancers when added to standard of care by comparing the reduction in absolute numbers of stage III and IV cancers in the intervention versus the control arm. The use of stage shift as a surrogate for mortality has been challenged [29], though these findings were based on a limited set of single cancers (bladder, liver, lung, lymphoma, ovary, pancreas and stomach), and neither aggregated end points nor other types of cancer were considered. Others have argued that traditional RCTs (e.g., the UKCTOCS trial and the PLCO trial) that lasted decades were still susceptible to external confounding and suggested that trials using late-stage incidence as a surrogate could reduce these challenges [30]. Since MCED testing is a new technology, we would argue that it is not yet possible to define the ideal trial scenarios, though groups such as the MCED Initiative have been formed to promote a better understanding of how to evaluate clinical utility and establish appropriate care pathways when using this technology [31]. The attributes that are typically associated with recommended screening tests (e.g., high sensitivity requirements, compromise between high false positive rates and low PPVs, high costs and an inability to discern aggressive from indolent cancers) have resulted in a healthcare system wherein only four cancers have recommended screening in the USA. The traditional single-cancer screening model has been critical for improving cancer detection; however, after accounting for the respective rates of compliance and test performance for each single-cancer screening test, only approximately 15% of the 1.3 million cancers diagnosed each year among those 50–79 years of age are detected early [32]. If multicancer early detection tests were used to complement existing screening, the net benefits to society – even after accounting for screening costs – would be substantial.

Real-world challenges

As is the case for all cancer screening tools, the risk of false positives, leading to unnecessary imaging or invasive procedures, is a valid clinical concern. MCED tests have prioritized specificity to limit the rate of false positives, but since this approach is so new, there is limited clinical guidance to date on the appropriate steps if diagnostic work-up does not confirm the initial positive MCED test result. It is also unknown if some positive MCED test results that are not confirmed upon imaging or biopsy (and are therefore considered false positive results) are actually a reflection of the molecular signature of a cancer signal that is not yet detectable by traditional imaging or biopsy procedures. And, of course, since the MCED test approach, by definition, is casting a wide net across many cancers (both screened and unscreened), it is also difficult to provide guidance in broad strokes, since the potential impact of MCED testing may be quite different based on many cancer type-specific factors (e.g., whether the cancer already has a screening modality available, individual cancer prevalence, the likelihood of an indolent malignancy/overdiagnosis, the level of complexity typically required for diagnostic resolution and whether treatments are available) [33]. Finally, as with any new technology, it is not yet possible to establish effectiveness relative to reduced morbidity, mortality or cost, since it is not currently confirmed whether these benefits will be realized by earlier detection across a population-scale implementation of such a test.

Key next steps & outstanding questions

Despite the challenges and unknown factors noted above, the new technology allowing for genomic MCED testing should be leveraged with the goal of reducing overall cancer burden. The understanding of cancer physiology at the molecular level has started to leave its mark in tailored cancer treatments, and it is expected that within the next 5–10 years, this knowledge will continue to be successfully applied to cancer screening and detection, including with blood-based MCED tests. Ultimately, the hope is that the availability of easy-to-use, blood-based MCED tests will reduce the disparities in the overall cancer experience that are based on nonbiological factors, including racial, ethnic, geographic and socioeconomic circumstances that correlate with poor access to cancer screening. For example, administration of a blood-based test does not require sophisticated equipment or repeated, arduous patient visits, which may help close the gap for underserved communities commonly associated with receiving a lower quality of care. However, it is critical to note that achievements in cancer management (e.g., available biomarker screening or novel treatments) have historically exacerbated disparities based on racial/ethnic, socioeconomic and geographic differences [34], so much work is left to be done.

Real change in downstream processes also will be needed to ensure that patients with a signal-detected result are able to access available diagnostic work-up procedures and curative treatments. It is important to view this in the context that disparities from many sources will impact outcome, including a higher rate of more aggressive tumor types among Black and Hispanic patients [35].

With all of this in mind, over the next decade, more work needs to be done to validate these genomic-based multicancer tests in real-world studies and to understand the health economics and societal value of these different approaches. Strong test performance in clinical validation will be required for regulators, and there are proposed intermediate end points such as reductions in late-stage cancer diagnoses and stage shift that can serve in assessments of clinical utility.

Regulators, payers and experts will play a key role in helping to define new frameworks to advance these new technological approaches without waiting 10–20 years for government-sponsored survival studies. Postapproval, real-world evidence generation may provide a partial solution to this challenge, including interventional studies to evaluate the potential for harm due to unnecessary invasive procedures and to characterize the patient perspective, particularly for those categorized as false positives. A quantitative approach that uses modeled assumptions may be employed as well, relying on published examples such as a recent description of a quantitative framework to evaluate potential benefits and harms of MCED testing [33]. In the meantime, government policy and legislative efforts, such as the Multi-Cancer Early Detection Screening Coverage Act of 2021, authorizing Medicare and Medicaid coverage of blood-based MCED tests [36], should continue to be supported.

Once real-world data are available, it also will be possible to better outline for physicians specifically how MCED tests should fit into routine care – something that is not currently known for an approach that represents a major shift from single-cancer screening alone. What workflows should be adapted to successfully integrate these tests into existing cancer screening paradigms for the largest impact to patients? Are there specific accommodations that need to be made within certain medical specialties to maximize the benefit? How do we ensure that physicians are equipped to direct subsequent care with the information they receive from test results? When implementing a test that can potentially identify a shared cancer signal from dozens of cancers, how do we manage expectations and communicate the relative risks of false positives to our patients? Are we seeing equitable use of MCED tests across populations? How do we maximize access to these tests, particularly to underserved populations, to ensure the largest impact of the test? Ideally, the simplicity of the blood draw used for MCED testing could be leveraged to allow for increased access across socioeconomic groups.

There are a number of questions that the next decade of research and clinical experience will seek to answer. These answers are worth pursuing. If we want to bend the curve on cancer mortality and address the inequities in cancer care, we need policies that enable society to embrace innovative technological approaches that may provide more universal access to comprehensive MCED testing.

Author contributions

L Nadauld and DP Goldman substantially contributed to the content and focus of the article and interpreting of the relevant literature and revised it critically for important intellectual content. Both authors approved the final manuscript and take full responsibility for the content included.

Financial & competing interests disclosure

L Nadauld declares no potential conflicts of interest; During the past two years, DP Goldman has received research support from the following sources: Amgen, Blue Cross Blue Shield of Arizona, Bristol Myers Squibb, Cedars-Sinai Health System, Edwards Lifesciences, Gates Ventures, Genentech, Gilead Sciences, GRAIL,LLC, Johnson & Johnson, Kaiser Family Foundation, National Railway Labor Conference, National Institutes of Health, Novartis, Pfizer, Roche, and Walgreens Boots Alliance. He serves as a paid scientific advisor for Biogen, GRAIL,LLC and the National Railway Labor Conference. He holds equity in EntityRisk. He has received travel support from The Aspen Institute. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Medical writing and editorial assistance were provided by J Hepker from Prescott Medical Communications Group and ProEd Communications, Inc., and were funded by GRAIL, LLC, a subsidiary of Illumina, Inc.GRAIL, LLC is currently held separate from Illumina, Inc., under the terms of the Interim Measures Order of the European Commission dated 29 October 2021.

References

  1. Seabury SA, Goldman DP, Gupta CN et al. Quantifying gains in the war on cancer due to improved treatment and earlier detection. Forum Health Econ. Policy 19(1), 141–156 (2016).CrossrefMedlineGoogle Scholar
  2. Howlader N, Noone AM, Krapcho M, Miller D, Brest A, Yu M. SEER Cancer Statistics Review, 1975–2017. National Cancer Institute, MD, USA (2020).Google Scholar
  3. Canary Foundation. Early detection facts and figures (2021). www.canaryfoundation.org/wp-content/uploads/EarlyDetectionFactSheet.pdfGoogle Scholar
  4. Dillekas H, Rogers MS, Staume O. Are 90% of deaths from cancer caused by metastases? Cancer Med. 8(12), 5574–5576 (2019).CrossrefMedlineGoogle Scholar
  5. National Minority Quality Forum. Late stage diagnosis of unscreened cancer: a health disparity (2021). https://static1.squarespace.com/static/5be307ae5b409bfaa68b1724/t/6089696d2a2ef65d997f4145/1619618157361/NMQF-GRAIL+Late-Stage+Diagnosis+of+Unscreened+Cancer+-+Paper+4+26+21.pdfGoogle Scholar
  6. Zavala VA, Bracci PM, Carethers JM et al. Cancer health disparities in racial/ethnic minorities in the United States. Br. J. Cancer 124(2), 315–332 (2021).CrossrefMedlineGoogle Scholar
  7. American Cancer Society. Cancer Facts & Figures for African Americans 2019–2021. GA, USA, 8 (2019).Google Scholar
  8. Clarke CA, Patel AV, Kurian AW, Hubbell E, Gomez SL. Racial/ethnic differences in cancer diagnosed after metastasis: absolute burden and deaths potentially avoidable through earlier detection. Cancer Epidemiol. Biomarkers Prev. 31(3), 521–527 (2022).CrossrefMedlineGoogle Scholar
  9. Singh GK, Williams SD, Siahpush M et al. Socioeconomic, rural-urban, and racial inequalities in US cancer mortality: part I – all cancers and lung cancer and part II – colorectal, prostate, breast, and cervical cancers. J. Cancer Epidemiol. 10.1155/2011/107497 (2019).Google Scholar
  10. Ko NY, Hong S, Winn RA, Calip GS. Association of insurance status and racial disparities with the detection of early-stage breast cancer. JAMA Oncology 6(3), 385–392 (2020).CrossrefMedlineGoogle Scholar
  11. de Pelsemaeker MC, Guiot Y, Vanderveken J, Galant C, Van Bockstal MR. The impact of the COVID-19 pandemic and the associated Belgian governmental measures on cancer screening, surgical pathology and cytopathology. Pathobiology 88(1), 46–55 (2021).CrossrefMedlineCASGoogle Scholar
  12. Hsiehchen D, Hsieh A. Nearing saturation of cancer driver gene discovery. J. Hum. Genet. 63(9), 941–943 (2018).CrossrefMedlineCASGoogle Scholar
  13. Chrysostomou AC, Kostrikis LG. Methodologies of primary HPV testing currently applied for cervical cancer screening. Life (Basel) 10(11), 290–301 (2020).CASGoogle Scholar
  14. Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts H. Artificial intelligence in radiology. Nat. Rev. Cancer 18(8), 500–510 (2018).CrossrefMedlineCASGoogle Scholar
  15. US Preventive Task Force Service, Grade A and B recommendations for cancer screening. www.uspreventiveservicestaskforce.org/uspstf/topic_search_results?topic_status=P&grades%5B%5D=A&grades%5B%5D=B&category%5B%5D=15&searchterm= Google Scholar
  16. Cohen JD, Li L, Wang Y et al. Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science (New York, NY) 359(6378), 926–930 (2018).CrossrefCASGoogle Scholar
  17. Lennon AM, Buchanan AH, Kinde I et al. Feasibility of blood testing combined with PET-CT to screen for cancer and guide intervention. Science (New York, NY) 369(6499), eabb9601 (2020).CrossrefCASGoogle Scholar
  18. GRAIL, LLC. www.galleri.com Google Scholar
  19. Klein EA, Richards D, Cohn A et al. Clinical validation of a targeted methylation-based multi-cancer early detection test using an independent validation set. Ann Oncol. 32(9), 1167–1177 (2021).CrossrefMedlineCASGoogle Scholar
  20. Liu MC, Oxnard GR, Klein EA, Swanton C, Seiden MV. Sensitive and specific multi-cancer detection and localization using methylation signatures in cell-free DNA. Ann Oncol. 31(6), 745–759 (2020).CrossrefMedlineCASGoogle Scholar
  21. Chen X, Dong Z, Hubbell E et al. Prognostic significance of blood-based multi-cancer detection in plasma cell-free DNA. Clin. Cancer Res. 27(15), 4221–4229 (2021).CrossrefMedlineCASGoogle Scholar
  22. Beer TM, McDonnell CH, Nadauld L et al. Interim results of PATHFINDER, a clinical use study using a methylation-based multi-cancer early detection test. J. Clin. Oncol. 39(Suppl. 15), S3010 (2021).CrossrefGoogle Scholar
  23. Chen X, Gole J, Gore A et al. Non-invasive early detection of cancer four years before conventional diagnosis using a blood test. Nature communications 11(1), 3475 (2020).CrossrefMedlineCASGoogle Scholar
  24. Li B, Wang C, Xu J et al. Multiplatform analysis of early-stage cancer signatures in blood. Proceedings of: The AACR Special Conference on Advances in Liquid Biopsies. FL, USA, 13–16 January 2020.Google Scholar
  25. Saman S, Stagno MJ, Warmann SW, Malek NP, Plentz RR, Schmid E. Biomarkers Apo10 and TKTL1: epitope-detection in monocytes (EDIM) as a new diagnostic approach for cholangiocellular, pancreatic and colorectal carcinoma. Cancer Biomark. 27(1), 129–137 (2020).CrossrefMedlineCASGoogle Scholar
  26. Smith RA, Andrews KS, Brooks D et al. Cancer screening in the United States, 2019: a review of current American Cancer Society guidelines and current issues in cancer screening. CA Cancer J. Clin. 69(3), 184–210 (2019).CrossrefMedlineGoogle Scholar
  27. Visaggi P, Barberio B, Ghisa M et al. Diagnosis of early esophageal cancer: from blood biomarkers to advanced endoscopy and artificial intelligence. Cancers 13(13), 3162 (2021).CrossrefMedlineCASGoogle Scholar
  28. Gheorghe G, Bungau S, Ilie M et al. Early diagnosis of pancreatic cancer: the key for survival. Diagnostics (Basel) 10(11), 869 (2020).CrossrefCASGoogle Scholar
  29. Owens L, Gulati R, Etzioni R. Stage shift as an endpoint in cancer screening trials: implications for evaluating multi-cancer early detection tests. Cancer Epidemiol. Biomarkers Prev. 31(7), 1298–1304 (2022).CrossrefMedlineGoogle Scholar
  30. Raoof S, Lee RJ, Jajoo K, Mancias JD, Rebbeck TR, Skates SJ. Multicancer early detection technologies: a review informed by past cancer screening studies. Cancer Epidemiol. Biomarkers Prev. 31(6), 1139–1145 (2022).CrossrefMedlineCASGoogle Scholar
  31. Multicancer Early Detection Consortium Initiative. Establishing the foundation for a multicancer early detection consortium (2021). www.tapestrynetworks.com/sites/default/files/publication_pdf/MCED%20-%20Summary%20of%20Themes%20Final_0_0.pdfGoogle Scholar
  32. The cancer detection rate – a public health approach to early detection. Cancer Letter 46(37), (2020). https://cancerletter.com/articles/20201002_7/Google Scholar
  33. Jiao B, Gulati R, Katki HA, Castle PE, Etzioni R. A quantitative framework to study potential benefits and harms of multi-cancer early detection testing. Cancer Epidemiol. Biomarkers Prev. 31(1), 38–44 (2022).CrossrefMedlineGoogle Scholar
  34. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics 2022. CA: A cancer journal for clinicians (2022). https://doi.org/10.3322/caac.21708CrossrefGoogle Scholar
  35. Sexton Ward A, Van Nuys K, Lakdawalla DN. Reducing racial disparities in early cancer diagnosis with blood-based tests (2021). https://healthpolicyuscedu/research/reducing-racial-disparities-in-early-cancer-diagnosis-with-blood-based-tests/Google Scholar
  36. Ezell S. Seizing the transformative opportunity of multi-cancer early detection (2021). https://itif.org/publications/2021/04/19/seizing-transformative-opportunity-multi-cancer-early-detectionGoogle Scholar