Research Needs for Understanding the Biology of Overdiagnosis in Cancer Screening

Abstract

Many cancers offer an extended window of opportunity for early detection and therapeutic intervention that could lead to a reduction in cause-specific mortality. The pursuit of early detection in screening settings has resulted in decreased incidence and mortality for some cancers (e.g., colon and cervical cancers), and increased incidence with only modest or no effect on cause-specific mortality in others (e.g., breast and prostate). Whereas highly sensitive screening technologies are better at detecting a number of suspected "cancers" that are indolent and likely to remain clinically unimportant in the lifetime of a patient, defined as overdiagnosis, they often miss cancers that are aggressive and tend to present clinically between screenings, known as interval cancers. Unrecognized overdiagnosis leads to overtreatment with its attendant (often long-lasting) side effects, anxiety, and substantial financial harm. Existing methods often cannot differentiate indolent lesions from aggressive ones or understand the dynamics of neoplastic progression. To correctly identify the population that would benefit the most from screening and identify the lesions that would benefit most from treatment, the evolving genomic and molecular profiles of individual cancers during the clinical course of progression or indolence must be investigated, while taking into account an individual's genetic susceptibility, clinical and environmental risk factors, and the tumor microenvironment. Practical challenges lie not only in the lack of access to tissue specimens that are appropriate for the study of natural history, but also in the absence of targeted research strategies. This commentary summarizes the recommendations from a diverse group of scientists with expertise in basic biology, translational research, clinical research, statistics, and epidemiology and public health professionals convened to discuss research directions. J. Cell. Physiol. 231: 1870-1875, 2016. © 2015 Wiley Periodicals, Inc.

Fig. 1

Length-biased Sampling. Each cancer has a window of opportunity during which it can be detected by screening. However, screening tests are more effective at detecting slowly growing neoplasms than those that progress rapidly. In some cases, the neoplasms progress so slowly that the patient dies of undetected cancer or the patient never develops a consequential cancer and dies of unrelated causes (“overdiagnosis”). Critical features of length-biased sampling include optimization of timing for the window of opportunity to detect a rapidly progressing neoplasm before it progresses beyond the stage in which it can be cured and developing biomarker tests that distinguish rapidly evolving cancers from those that evolve slowly or not at all.

Fig. 2

Dynamics of Cancer Progression. Panel A. The linear model of disease has had a major influence on medicine for more than a century. In its recent versions, it is postulated that there is a slow, gradual, linear occurrence of molecular abnormalities before the development of cancer that provides a long “window of opportunity” for early detection. This model predicts that interrupting any event (e.g., B) in the linear pathway will prevent progression. Panel B. Recent advances in genome technologies have reported that cancers arise by “branched evolution.” In some cases, such as in Barrett’s esophagus, an early branch leads to a state in which the esophageal metaplasia can remain stable for long periods of time (A→B). However, in other cases, progression is branching. In this case inhibiting one step (e.g., C→D) will not necessarily block progression, which can proceed through C→E. Panel C. Progression of a neoplasm over time has typically been represented as a linear series of measures, such as tumor growth or size, that increase at different rates in a linear fashion. Panel D. Recent results from genome analyses of tumors have provided evidence that genomic alterations may occur at vastly different rates. For example, point mutations may occur slowly resulting in relatively gradual rates of progression (“gradual evolution”), or an “IDLE” condition, but exposure to environmental mutagens such as tobacco smoke or the presence of inherited conditions that increase mutation rate can lead to “accelerated evolution” of cancer. Finally, recent evidence from genomic studies of advanced cancers have reported evidence that some cancers may develop by “punctuated” jumps that alter large regions of chromosomes such as chromosome instability (an increase in the rate of gain or loss of whole chromosomes or large regions of chromosomes). Catastrophic evolution, including chromothripsis (chromosome shattering) and whole-genome doublings, may occur in a single event; for example whole-genome doublings have been observed in nearly 40% of cancers sequenced by TCGA. In some cases, a series of events may accelerate progression. For example, Barrett’s esophagus develops chromosome instability (“punctuated evolution”) within four years of the diagnosis of cancer and genome doublings (“catastrophic evolution”) within two years of cancer diagnosis. Multiple other cancers, including those of the breast, lung, colon and ovary, also undergo a similar sequence of events but the timing relative to the onset of cancer has not yet been determined. IDLE =indolent lesion of epithelial origin.