Comment on “The lingering mysteries of metastatic recurrence in breast cancer”

We have read with interest the excellent commentary on “The lingering mysteries of metastatic recurrence in breast cancer” [1]. The article carefully and thoroughly reviews many explanations for metastatic recurrence. However, the primary emphasis is on what we would refer to as cancer cell-intrinsic pathways. By this, we mean that the article has, as its primary focus, alterations that are primarily restricted to the breast cancer cells themselves, which either contribute to their metastatic potential or their ability to overcome so-called dormancy. While these explanations are unquestionably important, we believe they fail to fully recognise critical non-tumour cell intrinsic (tumour microenvironment) or non-tumour extrinsic (host) aspects of the control of metastatic recurrence in women with breast cancer.

Firstly, it is critical to recognise several remarkable features of breast cancer recurrence, particularly of the oestrogen receptor-positive (ER+) subtypes. While oestrogen receptor-negative (ER−) breast cancers initially tend to have an earlier likelihood of recurrence, eventually, ER+ breast cancers catch up. In fact, multiple studies have shown that even following 5 years of disease-free survival and optimal administration of prophylactic adjuvant endocrine therapy, the recurrence of ER+ breast cancer is essentially a straight line for the next 25 years [2]. This implies that, if it were not for other inevitable causes of mortality, ER+ breast cancer would essentially be incurable. Multiple studies have shown that circulating breast cancer cells present many years following apparently curative therapy [3]. Studies of bone marrow sampling show that, even within 5 mL of sampled marrow, 30–40% of women can be shown to have viable breast cancer cells [4]. While this is a negative prognostic variable, within the timeframe of 5–10 years, many women with positive marrows still have not recurred. Clearly, this persistent appearance of metastatic disease over time must be considered in any analysis of drivers of metastasis.

Secondly, while we do not dispute cancer cell-intrinsic pathways contributing to metastatic dissemination, non-tumour cell “host” factors have an unquestionable contribution to breast cancer recurrence. Non-tumour cells that affect recurrence (and not limited to) include various immune, endothelial, fibroblast, and adipocyte cells. Although the authors correctly point out the “varied and unpredictable nature of metastatic recurrence”, the distant tissue site in breast and other metastatic cancers is not varied and unpredictable and is driven by tumour–host interactions. First described in 1889 by Stephen Paget [5], breast and other tumour cells (seed) disseminate to distinct and highly specific organ sites (soil), showing a clear role of the non-tumour host mechanisms in metastasis. Indeed, in breast cancer which lung, liver, bone, and brain are common sites for the dissemination of the primary tumour, breast cancer subtype displays distinct organ tropisms to bone (ER+), lung, and brain (triple-negative) and brain (HER2+) [6]. These specific host-driven sites of metastasis have clear implications for both dormancy and recurrence of specific subtypes of breast cancer as reflected by our comments above on ER+ breast cancer recurrence.

Thirdly, we agree with the authors when they conclude about finding ways of “identifying dangerous versus indolent DTCs”, the “development of more clinically relevant models”, and “avoiding over-treatment”. However, we suggest that these major drivers and risk factors for metastatic recurrence can be driven by a variety of host and non-tumour-derived contributors that are largely preventable and/or treatable. These include host factors including overweight/obesity, metabolic syndrome, diabetes, clinical depression, serious distress, and other inflammatory comorbidities that are highly prevalent in the ageing breast cancer patient. Indeed, obesity is projected to surpass smoking as the top preventable cause of cancer-related death in western countries including the UK in the next decade [7]. That these contribute to causation cannot be disputed, as clinical trials of weight reduction, glucose control, and management of stress and depression all reduce the likelihood of recurrence both in clinical studies and preclinical models. What has remained mysterious is how to link these non-tumour extrinsic factors that interact with the tumour cell-intrinsic pathways and eventually decrease recurrence. One exciting avenue that we have pursued is signalling through the Receptor for Advanced Glycation End-products (RAGE) as a final pathway mediating these non-tumour environmental effects within the tumour and its microenvironment [8]. RAGE is a receptor for multiple pro-inflammatory ligands upregulated by the tumour/tumour-host response, including for AGEs, S100s and HMGB1, and has profound effects on tumour progression [8]. We and others have shown that a series of syngeneic breast cancer models that rapidly disseminate in rodents and cause lethal metastatic disease have a remarkable reduction in tumour progression to metastasis when inoculated into RAGE knockout mice [8, 9]. These effects can be reproduced using several RAGE-specific drugs that can antagonise RAGE signalling. Taken together, these experiments unequivocally show that manipulation of the host potently impacts the progression of metastatic disease independent of the genetic makeup or intrinsic lethality of the breast cancer cells themselves. Indeed, multiple preclinical studies have shown a clear role for RAGE signalling in driving pathophysiological effects in diabetes, obesity, and other inflammatory non-tumour states [8]. It is particularly noteworthy that multiple human studies have revealed that elevations in RAGE ligands are hallmarks of clinical conditions and increased rates of breast cancer relapse, including obesity, diabetes, and depression. Clinical interventions such as exercise, cognitive behavioural therapy, and control of obesity and diabetes, already known to reduce breast cancer occurrence and recurrence, may well be signalling, at least in part, through reductions in circulating RAGE ligands [10]. Therefore, we propose RAGE as a major targetable mechanism linking tumour cell-intrinsic and non-tumour cell-extrinsic recurrence in breast cancer.

Without lessening the value of studies devoted to cancer cell-intrinsic pathways, we believe, particularly for ER+ breast cancer, that greater attention is needed to the role of non-tumour-host mechanisms that drive breast cancer recurrence and may therefore reveal tractable means of lessening the burden of breast cancer.

Author contributions

BIH drafted and edited the manuscript. MEL drafted and edited the manuscript.

Data availability

Not applicable.

Competing interests

The authors declare no competing interests.