Renaming low risk conditions labelled as cancerBMJ 2018; 362 doi: https://doi.org/10.1136/bmj.k3322 (Published 12 August 2018) Cite this as: BMJ 2018;362:k3322
- Brooke Nickel, PhD candidate12,
- Ray Moynihan, senior research fellow13,
- Alexandra Barratt, professor of public health1,
- Juan P Brito, assistant professor of medicine4,
- Kirsten McCaffery, professorial research fellow12
- 1Wiser Healthcare, Sydney School of Public Health, University of Sydney, New South Wales, Australia
- 2Sydney Health Literacy Lab, Sydney School of Public Health, University of Sydney, New South Wales, Australia
- 3Centre for Research in Evidence Based Practice, Bond University, Queensland, Australia
- 4Knowledge and Evaluation Research Unit, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, MN, USA
- Correspondence to: K McCaffery
Evidence is mounting that disease labels affect people’s psychological responses and their decisions about management options.1 The use of more medicalised labels can increase both concern about illness and desire for more invasive treatment. For low risk lesions where there is evidence of overdiagnosis and previous calls to replace the term cancer,2345 we consider the potential implications of removing the cancer label and how this may be achieved.
Our changing understanding of the prognosis of cancers
Some cancers are non-growing or so slow growing that they will never cause harm if left undetected.6 A prime example is low risk papillary thyroid cancer. Autopsy studies show a large reservoir of undetected papillary thyroid cancer that never causes harm,7 and the incidence of thyroid cancer has risen substantially in many developed countries. This rise has been predominantly driven by an increase in small papillary thyroid cancers, with mortality remaining largely unchanged.8 These small papillary thyroid cancers are increasingly being detected because of new technologies, increased access to health services, and thyroid cancer screening.4 Studies show that rates of metastases, progression to clinical disease, and tumour growth in patients with small papillary thyroid cancer who receive immediate surgery are comparable with those in patients who follow active surveillance.910
Likewise, for both low risk ductal carcinoma in situ (DCIS) and localised prostate cancer, detection strategies have become controversial as long term outcomes for both conditions have been shown to be excellent1112 and there is evidence and concern about overdiagnosis and overtreament.6 Given the potential harms of overtreatment of DCIS, active surveillance is now being trialled internationally as an alternative approach.131415 Active surveillance is already recognised as a safe and desirable option for localised prostate cancer, although invasive procedures such as a prostatectomy remain the most common choice for men with localised disease.
In addition to these cancers, there is some evidence and informed speculation that melanoma in situ, small lung cancers, and certain small kidney cancers may be considered low risk and subject to similar overdiagnosis and overtreatment.6161718
The cancer label
For decades cancer has been associated with death. This association has been ingrained in society with public health messaging that cancer screening saves lives. This promotion has been used with the best of intentions, but in part deployed to induce feelings of fear and vulnerability in the population and then offer hope through screening (box 1).30
Effect of the cancer label
Enthusiasm for cancer screening
Landmark US survey found that 87% of adults believe routine cancer screening is almost always a good idea and 74% of adults said that finding cancer early (most or all of the time) saves lives19
A British survey of 2024 men and women aged 50-80 years found that nearly 90% of people believe that screening is “almost always a good idea” and 49% said that they would be tested for cancer even it if was untreatable20
In studies on breast and cervical cancer women are often highly resistant to the idea of less intensive screening, with concerns about the frequency of screening intervals and that the changes are being made to save money rather than because of improved evidence about managing the cancer in question212223
Interviews with more than 10 000 Europeans show that 92% of women and 89% of men overestimate (or do not know) the mortality benefit of breast and prostate cancer screening24
Desire for surgery
A study of healthy US adults found that when treatment was framed as harmful, participants were significantly more inclined to opt for surgery than medication (65% v 38%, Χ2=11.40, P=0.001), even though doing so may increase their chance of death25
A study of 394 women found that when ductal carcinoma in situ (DCIS) was described as a non-invasive cancer, 47% women preferred surgery over non-surgical treatment options such as medication or active surveillance, whereas only 34% preferred surgery when it was described as a “breast lesion” and 31% when it was described as “abnormal cells” (P≤0.001)26
Uncertainty about active surveillance
In a five year nationwide follow-up study, 23% of men discontinued active surveillance for low to intermediate risk prostate cancer for non-biological reasons (20% patient preference and 3% other reasons)27
Across a sample of 1521 men with localised prostate cancer, those who were more emotionally distressed at the time of diagnosis were more likely to choose surgery over active surveillance (relative risk reduction=1.07; 95% confidence interval 1.01 to 1.14; P=0.02).28
A population based prospective cohort study of 341 men showed that at 9-11 years after diagnosis men with low risk localised prostate cancer who started active surveillance or watchful waiting had higher levels of distress and hyperarousal than men who had radiation or high dose brachytherapy (adjusted mean difference 5.9 (95% CI 0.5 to 11.3) and 5.4 (95% CI 0.2 to 10.5), respectively) and higher levels of distress and avoidance than men who had low dose brachytherapy (5.3 (95% CI 0.2 to 10.3) and 7 (95% CI 0.5 to 13.5), respectively)29
Although conservative management approaches such as active surveillance are becoming an option for some patients with cancer, a strong perception remains that aggressive treatments are always required.25 Recent studies in men with localised prostate cancer have found that the emotional distress of the diagnosis may motivate them to choose more aggressive treatment.28 However, not treating prostate cancer and following active surveillance also increases men’s levels of anxiety, rates of depression, and fear of cancer recurrence.29 Importantly, almost a quarter of men who initially choose to manage their prostate cancer with active surveillance opt for surgery or radiation therapy within five years for non-biological reasons.27
One potential strategy to calibrate expectation and to avoid unnecessary testing and treatment for these low risk cancers is to remove the cancer label from conditions unlikely to cause harm if left untreated. This strategy has been proposed by several international experts,2345 including a National Institutes of Health state of the science conference panel and a National Cancer Institute working group. Notwithstanding the challenges, we agree there is now a clear need to relabel some precancerous conditions and low risk cancers (table 1).
Evidence supports change
Evidence from several studies shows that describing a condition using more medicalised labels, including the term “cancer,” can lead to an increased preference for more invasive management options (table 2).1 This supports calls to remove the cancer label, when appropriate. The increased desire for more invasive management may be particularly important to consider in cancers that have a high public profile such as DCIS and prostate cancer. In DCIS it has been shown that women are increasingly opting for more aggressive treatments such as mastectomy and bilateral mastectomy rather than lumpectomy,3637 even though these treatments do not improve breast cancer specific survival. .38 Other outcomes such as the rate of local recurrence or a preference for reconstruction may be driving treatment preferences, although women with a DCIS diagnosis have been shown to have exaggerated and persistent fears of breast cancer recurrence and death.39 Similarly, in localised prostate cancer, for which active surveillance has been a recommended management option for several years, studies have shown that most men still prefer to opt for radical prostatectomy or radiation therapy.2740
How clinicians categorise conditions and recommend treatments may also be influenced by labels.4142 Several factors may drive them to overdiagnose and overtreat, albeit unconsciously. According to a recent review of the literature,43 potential drivers of overdiagnosis include fear of litigation or missing disease, an overemphasis on the need to diagnose, a lack of awareness of potential iatrogenic harms, and the challenge of doing nothing rather than something. Removing the cancer label from low risk conditions may help shift clinicians’ perspectives and enable them to feel more comfortable recommending less invasive options to patients.
Examples where cancer label has been removed
The cancer label has already been removed from other tumours that have been clearly shown to be largely indolent and unlikely to cause harm (table 3). An early example was the World Health Organization and International Society of Urological Pathologists’ joint decision to rename bladder tumours. A multidisciplinary group of experts agreed to reclassify papilloma and grade 1 carcinoma of the bladder as papillary urothelial neoplasia of low malignant potential.44 Similarly, a change in the description of cervical abnormalities found on smear testing from cervical intraepithelial neoplasia to squamous intraepithelial lesions, using the Bethesda system, has helped support more women to follow active surveillance. This change reflected important advances in the biological understanding of cervical neoplasia as well as advances in cervical screening technology and was driven by a motivation to help provide more uniform, evidence based, clearer, and less anxiety provoking terminology.45
A more recent example is the renaming of non-invasive encapsulated follicular variant of papillary thyroid carcinoma (EFVPT) as non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP).46 An Endocrine Pathology Society working group comprising international experts reviewed hundreds of cases of patients who had been followed for at least 10 years. It found that none of the patients whose tumours stayed within their capsules had any evidence of cancer, and this resulted in agreement to change the label. The decision, which aimed to decrease unnecessary treatment and reduce the psychological and financial burden for patients,4748 was endorsed by numerous leading professional societies internationally.
Although these changes have almost certainly been important, we were unable to find any formal evaluation of their effect on practice, clinician behaviour, or patient outcomes.
Relabelling low risk conditions
To help make progress on removing the cancer label from potential low risk conditions, box 2 suggests actions within clinical practice, medical education, and research. To start the major reform process of removing the cancer label, we propose an initial global round table meeting including cancer classification and staging groups such as the WHO Classification of Tumours Group, the International Collaboration on Cancer Reporting, and the American Joint Committee on Cancer, as well as government health agencies, leading professional cancer societies, and, importantly, public and patient representatives. In line with contemporary community expectations of independence, those formulating recommendations for reform must be free of conflicts of interest.
Actions to help make progress on removing the cancer label
Clinicians should initiate discussions about the likely benign nature of low risk conditions, the possibility of overdiagnosis and overtreatment, and the option of less invasive management such as active surveillance, both before and after diagnostic interventions
Clinicians should convey risk information using event rates (or absolute risks) to show the long term outcomes for people with low risk conditions, for both active surveillance and immediate treatment, over relevant timeframes such as 10 or 20 years49
New medical education curriculums can help students and clinicians gain a deeper understanding of overdiagnosis and strategies to communicate about low risk conditions
Information should be designed and widely promulgated for the public about overdiagnosis and the benign nature of some low risk conditions
Calculation of precise estimates on the proportions of patients affected by a change in nomenclature
More studies of long term outcomes of less invasive management options for low risk conditions
Testing possible alternative labels
Removing the cancer label from low risk candidates that display evidence of invasion under the microscope (such as low risk thyroid cancer and localised prostate cancer) may be more difficult than for those that display no invasive elements (eg, DCIS). A broadly representative multidisciplinary group, such as the one we propose, could start by considering these challenges, as well as current uncertainties and disagreement. The group should review current evidence on the risk of progression of each of the identified low risk conditions, establish standardised agreement in pathology reporting and diagnostic criteria across each condition, and then identify (where appropriate) an alternative label to describe the biological and clinical characteristics of the lesion.
Change and innovation in medicine are often resisted.505152 Changing something as fundamental as our shared understanding of the nature and meaning of cancer will therefore face many challenges and barriers, making a multistakeholder process essential. For example, recent qualitative evidence suggests clinicians treating papillary thyroid microcarcinomas do not see the merits of removing the cancer label.42 Similarly, focus groups with a random sample of community members found some participants were resistant to removing the cancer label, although others expressed a strong openness.53
It is also important to consider the potential of relabelling to cause harm. A label may provide beneficial effects, including an explanation and symptom validation.5455 There may also be implications for receiving benefits within the healthcare system, making some people ineligible for certain forms of support from government or health insurers.
A collective approach that includes informed citizens and consumers will provide insights into how a new label might help recalibrate expectations for detection, follow-up, and treatment. Any relabelling process needs to consider the effect not only on new patients but also on those already diagnosed with the condition. Removing cancer from a condition’s label may lead patients to reconsider the nature and extent of follow-up and question the need for additional treatments, potentially reducing overtreatment and any associated harmful psychological effects.1 On the other hand, patients might perceive that the new label undermines their current care, including changing the support they can access.56 Once labelled with cancer, people become part of a wider community of cancer survivors. Removing the cancer label could mean many patients perceive that they have been falsely classified, are no longer cancer survivors, and may have potentially received unnecessary treatments. This may cause psychological distress and confusion. As a patient advocate recently suggested, discussing why diagnostic terminology has changed with their doctor may help patients accept it.56
Various names have been proposed to help convey the favourable prognosis of low risk lesions, including indolent lesions of low malignant potential (IDLE), abnormal cells, and microtumour.24 Although the label needs to be biologically accurate, it also needs to be something patients can understand and that will not induce disproportionate concern. Civil society and consumer involvement in the relabelling process will help to ensure that any new labels will be understood and supported by the broader community.
Ultimately removing the cancer label will create controversy and take time. If done through a broad multistakeholder process, however, it should help ensure appropriate evidenced based care for future and current patients. Safe and effective reform also requires learning from past examples and formal evaluation of the practice implications and patient outcomes of any changes. Although it remains unclear exactly how best to move forward, we cannot continue to tell many people they have cancer when that label may be doing them more harm than good.
Labels used to describe medical conditions can influence treatment decisions and psychological responses
Removing the cancer label from low risk conditions that are unlikely to cause harm if left untreated has been proposed to reduce overdiagnosis and overtreatment
Change requires discussions between key cancer classification and staging groups, health agencies, cancer societies, and citizens and consumer groups
Formal evaluation of practice implications and patient outcomes is vital to evaluate the benefits and harms and ensure future safety
We thank Paul Glasziou for his helpful comments on the manuscript.
Competing interests: We have read and understood BMJ policy on declaration of interests and declare that RM and AB have been cochairs of the scientific committee for the Preventing Overdiagnosis conference, which is supported by The BMJ.
Contributors and sources: All authors are involved in research to understand and reduce overdiagnosis and overtreatment. All authors contributed to the concepts and structure of this manuscript. KMMcC is the guarantor.
Provenance and peer review: Not commissioned, externally peer reviewed.