“The theoretical case for routine cancer surveillance is strong. Tumor burden is expected to be lower before signs and symptoms of a recurrence develop; treatment is therefore expected to be more effective when a recurrence is identified by means of routine surveillance. [..]
The empirical case for routine cancer surveillance is weak. None of the 12 RCTs assessing imaging-based surveillance that were included in a 2021 systematic review revealed a statistically significant reduction in mortality associated with surveillance. [..] These findings suggest that detection and treatment of asymptomatic cancer recurrences offers no advantage over initiation of treatment only after symptoms develop.
The systematic review identified only one RCT of laboratory-based surveillance: a trial assessing surveillance in people with colorectal cancer with the use of carcinoembryonic-antigen testing that found no effect on the risk of death. Despite this lack of evidence, laboratory-based surveillance is common. Annual testing is routinely recommended after treatment for prostate cancer (assessment of prostate-specific antigen levels) and thyroid cancer (assessment of thyroglobulin levels), even though both cancers are associated with 20-year relative survival rates exceeding 95%. Laboratory-based surveillance is poised to become even more common, now that tests for blood-based biomarkers are being promoted for routine surveillance. Detection of ctDNA from an original tumor during surveillance will probably constitute a highly specific finding, providing powerful evidence that a cancer has come back. But whether detection of molecular recurrences helps patients live longer or live better is a question that has yet to be answered — and one that can be addressed only with an RCT.
Although it is tempting to believe that surveillance for metastases can only help people with cancer, the harms are evident. The costs of surveillance — including tests (such as computed tomography, magnetic resonance imaging, and positron-emission tomography), appointments, and follow-up evaluations for abnormal results — can be substantial and are increasingly borne by patients themselves. [..] Cancer surveillance is associated with worry and distress about what tests might find (so-called scanxiety) and may exacerbate people’s fear of recurrence. Abnormal, often incidental, findings can lead to additional tests and procedures — along with associated complications. Patients for whom treatment is recommended after a recurrence are subjected to the potentially toxic effects of therapy earlier than they would have been otherwise, at a time when they would have been asymptomatic and most likely had a higher quality of life.
In other cases, early detection of cancer recurrences may not influence therapy. Some oncologists don’t recommend treatment for asymptomatic metastatic disease, particularly in older patients or people with multiple coexisting conditions. Although forgoing treatment of metastases may be the best decision for patients, the implication in these cases is that testing only caused harm, since people whose cancers recur live longer with the knowledge that their cancer has come back.
Some advocates of routine surveillance argue that negative findings have important reassurance value. Although such findings may provide short-term reassurance for patients, evidence that surveillance improves people’s well-being is lacking. Conversely, there is strong evidence that surveillance-associated anxiety is common. Furthermore, we believe it’s unlikely that patients would opt to undergo routine surveillance and possibly receive a short-term reassurance benefit if they were informed about the absence of a mortality benefit — and about the potential for financial toxicity, the harms associated with radiation exposure for some types of testing, and the risk of incidental findings.
There are about 18 million cancer survivors in the United States, many of whom have undergone or are currently undergoing surveillance. Yet research generally suggests that routine surveillance finds more recurrences — leading to more surgery, chemotherapy, and radiation — without affecting the risk of death. The primary effect of surveillance is therefore to give patients bad news sooner. [..]
The existence of symptoms is an essential differentiator. Medical interventions implemented because of test findings can improve quality of life among patients who are currently sick and suffering. But for asymptomatic patients, the implied promise is that testing will help prevent deaths in the future. There is an ethical imperative to ensure that this implied promise is genuine.
Current data don’t support the validity of this promise, which highlights the need to implement strategies for mitigating the harms associated with surveillance. Primary among these strategies are those that involve reducing the use of routine surveillance. It will be important to address the various drivers of this practice: a genuine belief in the effectiveness of routine surveillance, guidelines based on the theoretical case for surveillance, and claims that surveillance constitutes “the standard of care.” It’s also important to acknowledge that surveillance-related testing may represent an important revenue stream for hospitals and cancer centers.
[..] Whereas billions of dollars are invested in RCTs of oncology drugs to support Food and Drug Administration approval, there has been little funding for trials assessing the value of routine surveillance. Trials should investigate not only the potential benefits of surveillance, but also the costs and potential harms. On the basis of existing data, however, less surveillance would be better for patients.”
Full editorial, HG Welch and LA Dossett, New England Journal of Medicine, 2025.4.26