Cancer screening continues to be a major challenge, with reliable evidence-based tests only being available for breast, lung, cervical, and colorectal cancer. These screening strategies have made great strides in reducing the cancer burden, morbidity, and mortality of these disease sites; however, in 2023 roughly 60% of the new cancer cases and 60% of cancer-related deaths in the United States are due to types of cancer for which we do not have reliable screening methods (1).

Using the advances of machine learning and high-throughput genomics, Multi-Cancer Early Detection (MCED) tests are being developed that allow for rapid screening of multiple cancer types concurrently through blood tests. These tests are designed on the principle that all tumors have shared biological features with a common set of cellular products accessible through the circulatory system (2). The PATHFINDER study was a prospective return-of results MCED study that enrolled healthy participants over the age of 50 years to assess the safety and feasibility of the Galleri (GRAIL LLC, Menlo Park, CA, USA) MCED cell-free DNA methylation testing for cancer screening. If a cell-free DNA methylation signature was indicative of cancer, up to 2 cancer signal origin(s) informed further clinician diagnostic assessment (3).

Our institution was the highest recruiting center on the PATHFINDER study, enrolling 1735 participants (26% of total trial enrollment). Here, we report our single institution experience participating in the PATHFINDER study and provide interesting case reports. We aim to illustrate the clinical scenarios and describe potential challenges and considerations in interpreting MCED test results to provide insights into its role in the future of cancer care.

Under IRB approval, a retrospective chart review of participants enrolled in the PATHFINDER study was performed. The MCED test used in this case series is based on targeted methylation analysis of cell-free DNA in peripheral blood, using next general sequencing to identify methylation patterns characteristic of cancer and predictive of cancer types. A machine learning classifier determines the presence of a cancer signal and predicts the likely cancer signal origin (CSO), guiding further diagnostic evaluation (4). Cancer risk factors, outcomes of tests and studies performed due to MCED signal positive, time to diagnostic resolution, and treatment outcomes were obtained from chart-review. Cases were selected to illustrate a range of clinical scenarios encountered during the early implementation of MCED testing at our institution.

From January 2020 to December 2020, our institution enrolled 1735 participants (26% of total study enrollment), 27 of which returned a signal positive for cancer suspicion (1.6%), and ultimately 12 diagnosed cancers (true positives) for a positive predictive value of 44% ( Table 1 ). Two of 12 were recurrent cancers in participants > 3 years from cancer therapy. There were 15 signal positives without cancer diagnoses (false positives, Table 2 ). Six of 15 false positive results correlated to monoclonal B-cell lymphocytosis (CLL precursor). During the course of 12-month follow-up for signal negatives (n=1708 participants), 19 were diagnosed with a cancer (site-specific sensitivity: 39%, specificity: 99.1%).

Our single-institution experience highlights one of the first comprehensive applications of multi-cancer early detection testing within the framework of the PATHFINDER study, offering key insights into clinical utility, diagnostic challenges, and patient impact. Through the evaluation of 27 signal positive patient cases, we observed a spectrum of outcomes, illustrating key aspects of MCED testing’s potential integration into clinical care, underscoring potential benefits of early detection, and highlighting challenges of interpreting results.

There is a paucity of literature detailing patient-specific pathways for those with an MCED signal positive test (3, 5). Among our cases was a true positive detection of stage 2 prostate cancer and relapsed acute myeloid leukemia after bone marrow transplant. Given the early-stage diagnosis in a clinically asymptomatic patient, the patient underwent successful and localized radiation with no evidence of disease for 4 years following treatment. This case exemplifies the power of MCED testing in its ability to detect subclinical relapse before overt clinical symptoms or major hematologic abnormalities. Our two additional true positive cases with stage I diagnoses of Hodgkins Lymphoma and HPV-related head and neck squamous cell carcinoma demonstrate the value of early detection with early intervention leading to patients experiencing fewer complications and maintaining a longer period of disease control.

While MCED testing identified true positive cases leading to successful early intervention, one true positive detection of colon cancer in a 90-year-old woman with complex past medical history raised ethical and clinical considerations regarding screening in patients who, even if diagnosed with cancer, may not be candidates for definitive treatment. In this patient, significant comorbidities and advanced age yielded challenges in aggressive therapy and they ultimately chose not to proceed with curative treatment. Another particularly challenging case involved a false positive for an 83-year-old female who underwent extensive workup, including CT, PET-CT, biopsy, mammography, and a breast MRI. Her workup was complicated by the multiple cancer signal origin signals provided, and while no malignancy was found, the psychological distress and financial impact is non-trivial. These types of challenges with medical workup and psychosocial distress from diagnostic workups have been previously reported (6, 7).

Overall, our case series contributes to the growing body of literature on MCED implementation by providing a nuanced perspective on potential benefits and considerations. Future research should focus on establishing clear clinical pathways to manage indeterminate or false-positive results. As MCED testing continues to evolve, its integration into routine cancer screening will require careful calibration to maximize benefits while minimizing potential harms.