Impact of early detection on cancer curability: A modified Delphi panel study

Abstract

Expert consensus on the potential benefits of early cancer detection does not exist for most cancer types. We convened 10 practicing oncologists using a RAND/UCLA modified Delphi panel to evaluate which of 20 solid tumors, representing >40 American Joint Committee on Cancer (AJCC)-identified cancer types and 80% of total cancer incidence, would receive potential clinical benefits from early detection. Pre-meeting, experts estimated how long cancers take to progress and rated the current curability and benefit (improvement in curability) of an annual hypothetical multi-cancer screening blood test. Post-meeting, experts rerated all questions. Cancers had varying estimates of the potential benefit of early cancer detection depending on estimates of their curability and progression by stage. Cancers rated as progressing quickly and being curable in earlier stages (stomach, esophagus, lung, urothelial tract, melanoma, ovary, sarcoma, bladder, cervix, breast, colon/rectum, kidney, uterus, anus, head and neck) were estimated to be most likely to benefit from a hypothetical screening blood test. Cancer types rated as progressing quickly but having comparatively lower cure rates in earlier stages (liver/intrahepatic bile duct, gallbladder, pancreas) were estimated to have medium likelihood of benefit from a hypothetical screening blood test. Cancer types rated as progressing more slowly and having higher curability regardless of stage (prostate, thyroid) were estimated to have limited likelihood of benefit from a hypothetical screening blood test. The panel concluded most solid tumors have a likelihood of benefit from early detection. Even among difficult-to-treat cancers (e.g., pancreas, liver/intrahepatic bile duct, gallbladder), early-stage detection was believed to be beneficial. Based on the panel consensus, broad coverage of cancers by screening blood tests would deliver the greatest potential benefits to patients.

Fig 1. Ratings of cancer curability today and the estimated benefit from a hypothetical screening blood test.

The x-axis represents expert ratings on how likely they believe a cancer could be “cured” today, defined as the receipt of effective treatment such that a population of individuals who are “cured” would have the same life expectancy as a population that never had the cancer being considered. Curability was rated on a scale of 1 (extremely unlikely to be cured) to 9 (extremely likely to be cured). The y-axis represents expert ratings on the estimated benefit from a hypothetical screening test, defined as the improvement in curability as a result of the test. Improvement in cure was rated on a scale of 1 (not at all likely to increase) to 9 (increase a great deal). Cancers in blue are those experts believed could potentially benefit the most from a hypothetical screening test (those with high curability today in earlier stages and likely to progress). Experts believed cancers in purple may show some benefit from a hypothetical screening test (lower curability today in earlier stages and likely to progress). Cancers in orange were rated as least likely to benefit from a hypothetical screening test (high curability in earlier stages today and progress slowly).

Fig 2. Estimated benefit from a hypothetical screening blood test with typical versus best-available treatment.

This figure illustrates expert ratings on the estimated benefit from a hypothetical screening test, defined as the improvement in curability as a result of the test on a scale of 1 (not at all likely to increase) to 9 (increase a great deal), when considering typical versus best available treatment. Typical treatment was defined as the care provided to the population as a whole; best available treatment was defined as guideline-concordant care. Items in grey represent no difference in ratings for typical versus best available treatment. Items in blue represent a difference in ratings for typical versus best available treatment, with the dark blue representing typical treatment and light blue representing best available treatment.