Health

Screening can lead to early detection of cancers and save lives, but sometimes innocuous abnormalities are picked up. The potential for harm in treating these ‘indolent’ cancers needs to be balanced against the benefits of diagnosis. By Sophia Auld.

Diagnosing cancer

The chief executive of Cancer Council Australia, Professor Sanchia Aranda.
Credit: Cancer Council Australia

Imagine going through cancer treatment and discovering it might have been unnecessary. According to research released in early 2020, this happens to about 30,000 Australians each year.

The Medical Journal of Australia study explored lifetime risks for five types of cancer – prostate, breast, renal and thyroid cancers and melanoma. By comparing the difference between 1982 figures (when there were no screening programs) and those from 2012 (when screening and early detection were common), researchers concluded about 11,000 cancers in women and 18,000 in men may be overdiagnosed each year.

Study author Associate Professor Katy Bell explains overdiagnosis means being diagnosed with a harmless cancer. “It’s a cancer that would never grow or grow very slowly,” she says. “It wouldn’t have spread or caused problems if it wasn’t detected.”

With figures from an Australian Institute of Health and Welfare report estimating 150,000 new Australian cancer cases in 2020, getting a diagnosis might seem to make sense. However, overdiagnosis can cause anxiety and expose people to harm from unnecessary surgeries and treatments, says Bell. The University of Sydney researcher and member of the Wiser Healthcare group – a research collaboration for reducing overdiagnosis and overtreatment – uses thyroid cancer to illustrate her point. “Of the three in four patients overdiagnosed with thyroid cancer in 2012, almost all would have had their thyroids completely removed. This entails risks from surgical complications, plus having to take thyroid medication for life.

“There’s a similar problem with prostate cancer,” Bell continues. “Radical prostatectomy can have significant consequences for a man’s quality of life, including impotence and incontinence.”

Likewise, mastectomy for breast cancer, including for ductal carcinoma in situ or DCIS (abnormal cells in the milk ducts that haven’t spread into breast tissue), is major surgery. “For aggressive cancer, we may be prepared to accept the risks because of the potential benefit early treatment may have in saving lives. But if it’s a cancer that won’t cause problems, then surgery cannot benefit but may still cause harm,” Bell says.

The dilemma lies in determining which cancers will be aggressive versus those that scientists call “indolent”, or in layman’s terms those that wouldn’t become symptomatic in a patient’s lifetime or contribute to death. “Cancer is a single word,” says Bell, “but it’s used to describe a wide spectrum of diseases, ranging from slow-growing or dormant clumps of abnormal cells right through to very aggressive tumours that can be fatal in a few months. Over the last few decades, we’ve learnt that the dormant and slow-growing types are more common than we thought. But we don’t really know what to do about them.”

She adds that distinguishing accurately between the two – particularly those on the interface between malignant and benign – can be very difficult, even for highly trained and skilled pathologists. Consequently, “we have to treat everyone to try and save the few people who have aggressive disease, even though a lot of people may be getting treatment for a disease that probably doesn’t matter.”

Aside from the physical harms, significant psychological distress can be associated with a cancer diagnosis. For example, a 2010 study of 342,497 men, published in the Journal of the National Cancer Institute, found an increased suicide risk within the year following a prostate cancer diagnosis.

Bell says another issue is the opportunity costs associated with funding and resourcing early detection. For clinicians, spending time and energy on people with borderline lesions means less for those who could benefit most from their expertise and treatment.

At a health policy level, reducing overdiagnosis would free up resources for finding new treatments, optimising use of existing ones and better management for cancer patients, including palliative care. Moreover, resources could drive more primary prevention programs, such as the successful sun awareness campaigns that “are actually of measurable benefit in lowering risk of melanoma in Australia”, Bell says.

The ramifications of overdiagnosis have another side, says Professor Sanchia Aranda, the chief executive of Cancer Council Australia. “We do need to inform people [about overdiagnosis] but we need to talk about the harms and benefits in a balanced way. The danger is that people will step away from early diagnosis because of concerns about it.”

She explains this happened in Britain, when changed information about breast screening meant fewer women took up the service. Breast screening reduces mortality from breast cancer by between 30 and 40 per cent, she says. “Is that a sufficient benefit to offset the diagnoses that wouldn’t have become problematic?”

Another consideration is that harms are not all equal, she says. “The psychological harm of having a diagnosis that wouldn’t become problematic is a big harm, but I would argue that a death from cancer that could have been avoided is a bigger one.” She adds that community juries in the Choosing Wisely program – a national initiative aimed at reducing unnecessary tests, treatments and procedures – when presented with the harms and benefits have voted to continue screening.

Furthermore, treatments have become more tailored, she says. For example, every woman with breast cancer would once have had a mastectomy, but now might have a lumpectomy. Men with low-risk prostate cancer might go into active surveillance programs rather than having surgery or radiotherapy. Research focused on genetic changes associated with various cancers may help predict the ones that won’t become problematic, enabling doctors to offer people better advice about treatment options, Aranda says.

Bell recognises the tension between overdiagnosis and finding clinically relevant cancers. She says the Medical Journal of Australia study acknowledges some level of overdiagnosis is unavoidable. “In trying to detect the small number of clinically important cancers, you’re bound to pick up a few harmless cancers in your net,” she says. But targeting tests to higher-risk populations more likely to have clinically aggressive cancers – known as risk-stratified screening – will help retain potential benefits of early detection while minimising potential harms from overdiagnosis.

Aranda agrees risk-stratified screening may be a way forward and says national screening programs may become increasingly refined. “In breast cancer, for example, as we understand more about the role of things like breast density, alcohol intake and the age of puberty onset, there may be some risk adjustments to the harm–benefit equation. But we’re not there yet.”

She highlights how the risk of overdiagnosis can be limited through only implementing screening programs with robust supporting evidence, as done in Australia, by citing an example from South Korea. There, population-level screening for thyroid cancer led the incidence to “go through the roof, while the death rate stayed exactly the same. That’s a real indication those were overdiagnosed cancers and they’ve had to roll that program back.”

Aranda also cautions against basing any policy changes on tests with limited accuracy. “PSA [prostate-specific antigen] testing, for example, does lead to a situation where we’re overdiagnosing cancers. There’s lots of reasons why PSA goes up. But that’s not a nationally sanctioned program.”

Another concern is direct-to-consumer blood tests for circulating tumour cells. “The promoters are saying they’re a non-invasive test, but they lack precision,” she says. “They may have been researched in terms of accuracy of finding cells, but there’s no evidence having those tests leads to better diagnosis or survival. And they lead people to having all these other tests, particularly CT scans, that are going to find innocuous lesions that never become problematic.”

Aranda notes that Australia’s health system is unsustainable if spending keeps increasing at current levels, but believes optimising cancer treatment is more efficient than reducing overdiagnosis. “It’s the treatment costs that really blow out, particularly as we see more very expensive drugs for advanced disease. That makes early diagnosis even more important, because it’s cheaper to treat people at an earlier stage. In melanoma, for example, the cost of a simple excision is a couple of thousand dollars, whereas the treatment of advanced melanoma is a couple of hundred thousand.”

Katy Bell stresses the study’s findings don’t relate to people with cancer symptoms. “The work we’re doing is about the problem of people who feel healthy and have no symptoms or signs of cancer,” she says. “We want to emphasise that people shouldn’t ignore any warning signs such as a breast lump, bleeding from the bowel or mole changes. Get to the GP straight away if you notice anything like that.”

She advises those diagnosed with cancer to ask their doctor whether it’s a low-risk type and, if so, to get a second opinion on pathology results and alternative management options. “We are hoping that the research produced by the Wiser Healthcare group can support citizens to be more active participants in decisions about their health,” Bell says, “so they can choose whether to have a medical test, and how a low-risk cancer could be best managed for them.”

This article was first published in the print edition of The Saturday Paper on Jul 11, 2020 as "Risk analysis".

A free press is one you pay for. In the short term, the economic fallout from coronavirus has taken about a third of our revenue. We will survive this crisis, but we need the support of readers. Now is the time to subscribe.

Sophia Auld is a freelance writer and editor based on the Sunshine Coast.