From a clinical perspective, metastasis is the most critical aspect of tumorigenesis, because over 90% of cancer mortality is caused by metastasis.
Is metastasis really the most critical aspect for cancer treatment?
The review above seems to think so, and has amassed over 2000 citations to date, so it must be right… right?
A handful of the papers referring to this paper carry forward the above statement unquestioned.
Fan et al. (2013) in Protein Cell write:
In clinic, metastasis is the most critical step of tumorigenesis, because over 90% of cancer mortality is caused by metastasis.
They then go on to reuse pretty much the next paragraph of Grivvenikov et al. 🐸☕
Zhang et al. (2012) in International Journal of Biological Sciences don’t address the clinic, but also state that:
Metastasis is the most critical aspect of tumourigenesis, and more than 90% of cancer mortality is caused by metastasis.
Kadomatsu et al. (2014) describe in Trends in Endocrinology and Metabolism how:
Tumor metastasis is the most common cause of death in cancer patients.
You get the idea. Others describe metastasis as an important predictor of patient prognosis (e.g. Zbytek 2008), though there’s the subtle difference in this phrasing: it removes the implied causation, leaving the possibility that metastasis is more an effect, or symptom to be targetted.
In 2010, the American Association for Cancer Research staged a series of reviews on the occasion of its centenary. In one, Talmadge and Fidler made the familiar claim that metastasis is the ‘primary clinical challenge’,
as it is unpredictable in onset and it exponentially increases the clinical impact to the host.
In many patients, metastasis has occurred by the time of diagnosis, although this may not be apparent clinically. Although tumor metastasis can occur early in tumor progression when the primary tumor is small or even undetectable, most occurs later when the primary tumor is larger. This pattern is supported by the observation that surgical excision of smaller lesions is often curative and forms the basis for tumor, nodes, metastasis (TNM) staging. When metastasis has already occurred, strategies targeting tumor cell invasion and other early aspects of the metastatic process may not be relevant to outcome. The suggestion, based on microarray analysis, that all tumor cells within a malignant tumor are equally metastatic is likely to be incorrect. This controversial conclusion is associated with the low sensitivity of microarray data that support the similarity of primary and metastatic tumor expression signatures, which are similarly due to the activation of a metastatic genetic program in early progenitors. It has been suggested that the overgrowth and dominance within primary and secondary lesions by a single tumor cell population, with a uniform metastatic signature, is associated with early metastasis. An extension of this hypothesis is that the late emergence of metastatic clones will result in divergent expression patterns between primary tumors and metastases, secondary to the masking of metastatic signatures in the primary tumor, by persisting non-metastatic clones. However, studies of clonal cell lines derived from late-stage human carcinomas have provided direct evidence that individual cancer cells, coexisting within a tumor, differ in their metastatic capability, including ones that are nonmetastatic, confirming the tumor heterogeneity shown in preclinical studies with murine, as well as human tumors. Furthermore, the expression signatures of tumors derived from cloned weakly and/or nonmetastatic human cell lines and from their isogenic metastatic counterparts from the same patient tumors differ, although the expression signature of metastases and that of their corresponding primaries are similar.
Aita et al. (2015) is one paper among a raft of recent research that seeks to destroy the idea that metastasis spreads like wildfire. Rather, there’s pattern to the process in terms of tissue invasion events, analogous to gene-level findings published in February that the order of (JAK and TET2) mutation affects cancer behaviour.
I wrote last month of the pair of prostate cancer studies that had traced the lineage of metastatic cancers through 'ultra-deep sequencing’, and surprisingly unearthed primary clones in the bloodstream.
These cells were said to have persisted in 'clinically occult sites’, which may complicate the use of metastasis as clinical target. The role these primary clones were playing was not commented on by the authors in February, but there’s an unavoidable possibility in their observation that this further complexity, previously undetectable with microarray or only moderately deep sequencing in a small number of samples, may be common in other cancers than prostate.
These primary clones would arguably support the theory touted in 2006 that cancer may be a disease of 'self-seeding’ (a reference to Paget’s 1889 ’seed and soil’ hypothesis).
In this view, the world of tumour size as 'predictive’ of metastasis [in turn predictive of patient prognosis] is turned upside down. Large tumour size is only linked to metastasis-enhancing gene signatures in 'retrospect’, i.e. causality is mistaken, since metastases communicate 'back’ into primary tissue thus leading to larger tumours through 'reseeding’. The larger tumours are therefore only as a result of metastasis, rather than tumour size being a causative factor towards its metastatic potential.
Joyce and Pollard noted such 'backcommunication’ in passing within their 2009 review:
If this is the case, nothing is currently known about the underlying mechanisms.
Last month, Arnal et al. (preprint via ResearchGate ahead of publication in Evolutionary Applications) wrote of how Norton’s self-seeding hypothesis is not so much a challenge to metastatic views of cancer, but provides further importance to the role of the tumour microenvironment.
Studies have been arguing that the production and dissemination of metastatic cells should be counter selected at the initiation and early stages of tumours due to local resource availability (the selection should favour cells resistant to anoikis (programmed cell death) and contact inhibition, but with no migratory potential (Gatenby and Gillies 2008). At later stages when damage to the tumor accommodating organ significantly restricts resource availability, tumor cells with increased motility should have selective advantage (and higher fitness) despite the cost of most migrating cells dying without establishing a new colony (Merlo et al. 2006). However, recent studies challenge the traditional view of a late acquisition of metastatic potential and instead propose that tumour cells acquire the motile phenotype early in tumorigenesis (Eyles et al. 2010) as a result of selection favouring expansion of primary tumours.
Pathologic cell mobility could indeed contribute significantly not only to metastasis but also to primary tumour growth (cancer self –seeding theory (Norton and Massagué 2006)), but the pathways to the self-seeding to the primary tumour will take depends on the cues and concomitant selective forces of tumour microenvironment (welcoming nutrient rich, or hostile depleted primary tumour site will result in different outcomes (i) dislodging, then reattaching in/at the primary site, (ii) dislodging, circulation in blood stream then reattachment in/at the primary site, (iii) dislodging, circulation in blood stream then reattachment in a novel (metastatic) site (Norton and Massagué 2006)).
There’s only one other review that to my memory has presented a direct challenge to considerations of cancer therapy in recent years, and from discussions with others I’m hesitant to align myself too strongly with it (the editorial has gained only 2 citations to date, one of which was self-citation). In 2013, Kulkarni et al. maligned the 'deterministic view’ of cancer phenotypic change, in Molecular Cancer.
They noted Felsher’s 2004 report on the reversal of tumorigenesis through oncogene inactivation, ’uncover[ing] the pluripotent capacity of tumours to differentiate into normal cellular lineages and tissue structures, while retaining their latent potential to become cancerous’. It’s described as “landmark” and was recommended on the Faculty of 1000 platform.
Conventional wisdom suggests that cancer cells contain so many mutations that their reversal to normalcy is unlikely. This belief has led to the development of treatments aimed at killing cancer cells – an ambitious aim that has been difficult, if not impossible, to achieve for most cancers. However, in 2004, a landmark paper by Felsher and coworkers startled the cancer world by demonstrating that indeed, rogue tumor cells can be reformed. By conditionally turning on the oncogene c-Myc in mice hepatocytes with the Tet system, the authors induced hepatocellular carcinoma in these transgenic animals. They then turned off Myc expression in these animals and surprisingly observed that Myc inactivation resulted en masse in tumour cells differentiating into hepatocytes and biliary cells forming bile duct structures.
As a result, the authors of the 2013 paper suggested a shift from gene-centric focus on mutations that arise, to protein interaction network stochasticity.
It is now widely accepted that stochasticity or noise in gene expression can give rise to phenotypic variations among clonal cells in homogeneous environments. Thus, in response to the same stimulus, two genetically identical cells can display very different phenotypes, and it has been argued that this inherent stochasticity can serve as a key driving force for tumorigenesis. However, in addition to noise in gene expression, emerging evidence indicates that the information transduced in cellular signaling pathways is also significantly affected by noise. It has been proposed that, noise in these pathways may be generated by the interconnected and promiscuous nature of protein interactions that are necessary to transduce signals. However, how this noise arises and what consequences it has on cell fate is poorly understood.
