Intended for healthcare professionals

Rapid response to:

Head To Head

Should we rename low risk cancers?

BMJ 2019; 364 doi: (Published 23 January 2019) Cite this as: BMJ 2019;364:k4699

Rapid Response:

Should we rename low risk cancers? – Author’s reply

I thank the readers for their interest in our paper and take this opportunity to respond to some of the rapid responses.

Dr Spitzer suggests that the “no” argument was written by a pathologist who is oblivious to issues around patient communication, but this discussion was much more nuanced than a simple “yes” versus “no” debate. Although each author focussed on the points for and against renaming low risk tumours respectively, both of us are aware of good arguments in favour as well as against this proposition. I am not opposed to renaming renaming low-risk cancer per se but argue that this must be done judiciously. Terms such as “tumour of uncertain malignant potential” may be misunderstood by patients and hence I suggest that lay persons are involved in shaping new terminology.

Professor Baum highlights the issue of overdiagnosis resulting from cancer screening programmes. The pros and cons of screening have been extensively debated in various forums. However, “to screen or not to screen” may be the wrong question. While some patients may be harmed by overdiagnosis resulting from cancer screening, others could have their lives saved by early detection. A more appropriate question could be “who should we screen and for what?”. Screening could be more effective if limited to individuals at higher risk of clinically significant disease and designed to detect only higher risk lesions. For example, it is recognised that low-grade ductal carcinoma in situ (LGDCIS) and high-grade DCIS (HGDCIS) have very different malignant potentials with the former being less likely to progress into invasive malignancy. Moreover, LGDCIS is generally a precursor for less aggressive low-grade invasive cancer. Hence, the diagnostic threshold for the diagnosis of LGDCIS could be raised so that patients at the lower end of the spectrum (lesions that are also most prone to interobserver variation) are not overtreated as these would now be diagnosed as benign (atypical ductal hyperplasia). Similarly, a case could be made for not reporting high-grade prostatic intraepithelial neoplasia (PIN) in prostate biopsies. These suggestions would not eliminate the risk of overdiagnosis but could help reduce it while retaining the utility of screening to pick more clinically significant lesions. Ultimately, the key is to educate the public that reduction of risk to zero is simply unattainable. Even a simple mastectomy would not completely eliminate the risk of developing breast cancer as all breast tissue cannot be removed during the surgery.

Dr Kenwright refers to the wide biological spectrum of cancer and fears that renaming low risk cancers may emphasise the pessimism for “real cancer”. The very nature of cancer has changed over the decades. In the 1960s cancer was diagnosed in patients who generally presented with symptoms associated with advanced disease (weight loss, fractures etc) and hence the disease was often lethal. Subsequently, increased awareness led to patients presenting earlier with clinically evident organ confined tumours that could be effectively treated. However, growing use of investigations (particularly radiological) and screening programmes have led to an increasing number of cancers are being diagnosed as incidental findings and a significant number of these that may not have progressed in the patient’s life time would represent overdiagnosis. Hence, it is entirely reasonable to discuss whether we need to change our “language” to reflect this changing landscape.

Dr Sundar suggests that renaming low risk cancer would be a cosmetic exercise and that the underlying issue of over-investigation must be addressed. I totally agree that over-zealous use of investigations should be discouraged, and that patient education is the key. However, renaming low risk cancer is not a purely cosmetic exercise. Benign and very low risk cancer are not biologically distinct diseases but a continuous spectrum of increasing risk with subjective cut-offs. This is analogous to the diagnosis of hypertension where the biological significance of a systolic BP of 140mmHg (hypertension) is not different from that of 139mmHg (prehypertension). Similarly, the prognosis of very low risk cancer is not significantly different from that of a benign tumour and the determination of the minimum risk of harm required for a tumour to be labelled as malignant is entirely subjective. A very low diagnostic threshold could result in overdiagnosis and overtreatment of large number of patients while using a higher threshold would risk underdiagnosis and undertreatment. Raising the subjective threshold slightly would result in very low risk tumours being re-categorised as benign precluding the need for anxiety inducing alternative terminology and protracted follow-up of these patients who are least likely to come to harm.

Competing interests: No competing interests

28 January 2019
Murali Varma
Consultant Histopathologist
University Hospital of Wales
Heath Park, Cardiff, UK