There are growing concerns that the explosion in screening for disease and illness has led to an unhealthy increase in overdiagnosis and overtreatment, especially in regard to diseases such as cancer. While early diagnosis was supposed to reduce the number of deaths related to cancer, recent studies have shown that has not happened. Instead, as the US National Cancer Institute noted in April 2012, we are now confronting the problem of overdiagnosis leading to overtreatment of cancers. Dr. Barry Kramer, director of NCI’s Division of Cancer Prevention, warned that “Overdiagnosis from cancer screening is one of the most pressing clinical issues in the field of cancer screening”.
So what is the difference between diagnosis and overdiagnosis?
A diagnosis is what names a sickness; that is, it describes the nature of an illness or disease. For example, the label “influenza” describes the set of headaches, muscle pain and fever that are characteristic of this illness, and suggests its severity and future course. While the influenza virus can produce symptoms ranging from mild infection to serious complications and even death, when we hear that someone has the flu, we usually think of an illness of mild severity that usually lasts a total of about seven days and heals without causing a lot of damage. A diagnosis enables a prognosis and allows us to balance benefits and harms of medical interventions with the potential loss of quality of life and of life expectancy.
Cancer overdiagnosis, on the other hand, refers to the detection of tumours that, left untreated, would be unlikely to cause health problems or death.
What does a diagnosis of cancer mean?
The Greek word “karkinos” (carcinoma) describes a malignant tumour as a crab. Like the crab, cancer has many legs, a group of diseases involving unregulated cell growth. The causes of cancer (known medically as a malignant neoplasm) are diverse and include genetics, environmental pollutants, radiation exposure, hormones and viruses. Its mechanism of action is only partly understood.
What we do know is that cancer cells can divide and grow uncontrollably. Sometimes these cells invade other parts of the body, a process known as metastasis. When cancer begins it invariably produces no symptoms with signs only appearing as the malignant cells continue to grow. Cancer can be detected in a number of ways, including screening or testing “healthy” persons to detect tumours before the presence of any signs and symptoms. Without treatment, cancer can proliferate, resulting in metastasis and death.
The diagnosis of cancer, then, describes a very serious disease with an ill-defined cause that may lead to death unless prompt treatment is employed. However, as in the case of influenza, cancer is a disease with extreme heterogeneity – that is, it can develop very differently in different people and for reasons that are not clearly understood. Sometimes it is found in persons who, without any symptoms or signs of cancer, have died of other causes (like “a sleeping crab”, a type of cancer that is histologically malignant but biologically benign). Cancer can also be extremely aggressive and rapidly fatal (a kind of “monster crab”, histologically and biologically malignant). There is no true gold standard we can use to accurately predict the outcome and course that a diagnosis of cancer will produce.
Cancer overdiagnosis refers to the diagnosis of a cancer that would otherwise not go on to cause symptoms or death. It is an error of prognosis rather than of diagnosis. Overdiagnosis can increase the impact of cancer on the quality of life and longevity because it leads to overtreatment, exposing patients to potential harms without offering any benefits.
An example of overdiagnosis leading to overtreatment occurred with the rise of screening infants for neuroblastoma, the most common malignant solid tumour during childhood. Neuroblastoma often spreads to other parts of the body before any symptoms are apparent (up to 60% of all children with neuroblastoma present with metastases). This disease is one of the human malignancies known to demonstrate spontaneous regression from an undifferentiated state to a completely normal and benign cellular appearance.
Austria, Canada, Germany and Japan began screening (via urinalysis) asymptomatic infants at three weeks, six months, and one year in the 1980s. Screening was halted in 2004 after studies showed there was no reduction in the rate of death due to neuroblastoma. However, the overdiagnosis of neuroblastoma – a type of cancer that likely would have resolved without treatment – subjected those infants to unnecessary surgery and chemotherapy.
Breast cancer, cervical cancer, ovarian cancer, prostate cancer, lung cancer, thyroid cancer, colon cancer and melanoma screening and overdiagnosis
The overdiagnosis of cancer has become a costly public health problem, one fuelled by increased screening and the use of diagnostic tests in general. Unfortunately, clinical trials seldom quantify the harms associated with cancer screening – including overtreatment and false positives.
Overdiagnosis may also be contributing to public perceptions that the rate of death from cancer – especially of the breast, cervix, ovaries, prostate, lungs, thyroid, colon and skin (melanoma) – is increasing at a greater pace than is in fact the case. For example, the incidence of thyroid cancer in the United States has tripled in the past 30 years but the vast majority of these cancers are small low-risk papillary thyroid cancers that are unlikely to ever progress enough to cause symptoms or death.
For melanoma, the rate of diagnosis has also almost tripled but again, the rate of death is generally stable, with little change in the past 15 years. We have do not have an epidemic of melanomas but of unnecessary diagnosis of melanoma (overdiagnosis)
This is also the case also for breast cancer, ovarian cancer, prostate cancer and lung cancer. Overdiagnosis is often ignored for cervical cancer and colon cancer, while adding human papillomavirus (HPV) testing, for example, to Pap smear testing causes more overdiagnosis without any providing any benefits.
So more screening means huge increases in the number of diagnoses and treatments – with with almost no decreases in specific cancer mortality. That is, more patients were overdiagnosed and received treatment as if the cancer they had resembled the “monster crab” (biological cancer) instead of the “sleeping crab” or even the crab that self-resolved (histological cancer).
Unnecessary treatment can result in harm without benefits – condemning millions of people to live the rest of their lives with a fake Sword of Damocles hanging over their futures. There is a vast army of “survivors of cancer” who are actually survivors of cancer overdiagnosis. For example, a review of the data for the past 30 years (1976-2008) found that breast cancer was overdiagnosed in 1.3 million U.S. women during the period.
Patients need and would like more information about overdiagnosis. Sadly: “Physicians’ counselling on screening does not meet patients’ standards. Most individuals desired information about screening harms, which was not given, and attested that this knowledge would matter to them: 69% of the sample indicated that they would not start screening if overdiagnosis was as high (i.e. ≥10 cases per 1 life saved) as it is in mammography (breast cancer) and PSA (prostate) testing”.
As the authors of a 2012 paper in the British Medical Journal observed, “Medicine’s much hailed ability to help the sick is fast being challenged by its propensity to harm the healthy”. To reverse the epidemic of overdiagnosis the practice of medicine should take place under the motto “less is better”.
About the Contributor
Juan Gérvas, MD, PhD, is a retired general practitioner, Equipo CESCA (Madrid, Spain) and an honorary professor of public health at the Autonomous University, Madrid. He is a visiting professor of international health at the National School of Public Health, Madrid.