Cancer Found Accidentally (Incidentalomas)

An incidentaloma is a mass found by accident — on a scan done for abdominal pain, a chest CT after a car accident, an MRI for headaches. The scan answers the question it was asked, and reports something else as well. The adrenal glands, kidneys, liver, thyroid, lungs, pituitary, ovaries and pancreas are the usual sites.

They are common. In the USPSTF's review of lung cancer screening, the proportion of participants with incidental findings deemed significant or requiring further evaluation ranged from 4.4 percent to 40.7 percent depending on the trial and how findings were counted — coronary artery calcification, aortic aneurysms, emphysema and masses in various organs. That range is wide because there is no universal agreement on which findings deserve a mention. The headline point stands: scan enough of anything and you will find something.

The great majority of incidental findings are benign. In the National Lung Screening Trial, cancers of the kidney, thyroid or liver were ultimately diagnosed in 0.39 percent of participants — meaning that of the many incidental findings reported, a very small fraction were malignancies. Small adrenal nodules, simple kidney cysts, hepatic haemangiomas and small thyroid nodules are so common that they are better understood as normal variation than as disease.

This is why the wording of radiology reports is so unnerving and so uninformative on its own. "Cannot exclude malignancy" is a statement about the limits of the image, not an estimate of your risk.

Guidelines exist precisely so that a chance finding does not trigger an open-ended investigation. The American College of Radiology's Incidental Findings Committee publishes algorithm-driven white papers built by multidisciplinary panels, with the explicit aim of identifying significant conditions early while "avoiding unnecessary testing and treatment of low-risk incidental findings." Published and in-development topics cover organ-specific findings including adrenal masses, fatty liver disease, parathyroid lesions and pineal cysts.

In practice the ladder runs roughly:

1. Characterise what already exists. Size, density, margins, enhancement, and — often decisive — comparison with any older scan. A lesion unchanged over several years is almost always benign, and the single most useful next step is frequently to find a prior image rather than order a new one.
2. Dedicated protocol imaging. A scan designed for the specific question: adrenal washout CT, multiphase liver MRI, thyroid ultrasound, contrast-enhanced kidney imaging.
3. Laboratory tests where the organ is hormonally active. Adrenal incidentalomas, for example, are routinely assessed for hormone excess as well as for malignancy, because a functioning benign tumour can matter more than a non-functioning one.
4. Interval imaging. Repeat at a defined point to test for growth. Stability is evidence.
5. Biopsy or resection, last, and only where the preceding steps leave real doubt or the risk is high enough to justify it.

Investigation is not free, and the cost is paid mostly by people who did not have cancer.

In NLST, 24.2 percent of low-dose CT screens were positive across three rounds, and the USPSTF review reported false-positive rates of 26.3 percent at baseline. Invasive procedures followed in 1.7 percent of screened participants, with complications in 0.1 percent. In the broader review, complications from needle biopsy ran 0.03 to 0.07 percent and surgical procedures followed in 0.5 to 1.3 percent. Repeated CT adds radiation and contrast exposure. Findings recorded in your record can affect insurance. And the anxiety of an unresolved finding under surveillance is a real, sustained harm that does not show up in complication tables.

NCI's screening guidance names the same problems: false positives "usually followed by more tests and procedures, which also have risks", procedural complications, and overdiagnosis — cancers found "that don't matter because they would have gone away on their own or never caused any symptoms", with no way to know at the individual level whether treatment helped.

The useful question is not "could this be cancer" — almost anything could — but "what does the guideline for this specific finding, at this specific size, in someone with my risk factors, actually recommend?" That question has a documented answer for most common incidentalomas, and asking for it converts an open-ended worry into a defined plan with an endpoint.

• USPSTF — Lung Cancer Screening recommendation (harms, incidental findings)
• NCI PDQ — Cancer Screening Overview (harms of screening)
• NCI — National Lung Screening Trial