The first in the Research Committee's new 'Paper of the Month' series comes from the high-impact basic science journal Science, July 2017. [1]

Click here to read this paper

What is known on the subject?

The spread of cancer cells from primary tumours to regional lymph nodes is often associated with reduced survival. One prevailing model to explain this association proposes that fatal, distant metastases are seeded by lymph node (LN) metastases. This view provides a mechanistic basis for the TNM staging system and is often the rationale for surgical resection of tumour-draining lymph nodes. However (remarkably), it is still unclear whether a single metastatic subclone evolves in the primary tumour, subsequently spreading to LNs and distant sites, or whether multiple subclones in the primary tumour independently seed LN and distant metastases.

What this study adds

This collaborative study between Harvard and major cancer centres in Holland and the UK addressed the above question using human colorectal cancer as a model. The team systematically reviewed 1373 patient records and diagnostic materials to derive 17 patients with sufficient pathological i.e. multiple primary tumour, lymph node metastatic, liver metastatic and normal tissue (germline) samples (total: 239 samples including 111 positive nodes and organ metastases). Automation of a relatively new (2014) methodology traced the evolution of these cancers. In brief, the methodology (which was not entirely straightforward to understand – at least not to me) leverages indel mutations in hypermutable, noncoding polyguanine repeats that accumulate during tumour evolution. Mathematical algorithms then generate genetic distances between tumour primaries, nodal deposits, and metastases, allowing construction of phylogenetic trees that visually depict tumour evolution.

The main finding was that in 65% of cases (11 out of 17), lymphatic and distant metastases arose from independent subclones in the primary tumour, whereas in 35% they shared a common subclonal origin. These findings were not influenced by common clinicopathological variables e.g. treatment history.

Implications for colorectal practice

The findings of this study will not immediately lead to a difference in colorectal surgical practice. As surgeons, I think all are well aware that there is a discordance between LN and liver metastases due in part to direct portal venous spread (extramural vascular invasion), and that LN harvest has a prognostic benefit as an indicator for adjuvant treatment that goes beyond simple disease elimination. There may also be differences between current findings (that were limited to the liver) and lung metastases, especially for rectal cancer (there were only 3 rectal cancers in the series). Nevertheless, I think the study is attractive not least to be able to provide the keen student with some sort of evidence-based answer to this fundamental question. For the interested reader, the perspective by Markowitz [2] provides greater discussion of the findings.


  1. Naxerova K, Reiter JG, Brachtel E, Lennerz JK, van de Wetering M, Rowan A, Cai T, Clevers H, Swanton C, Nowak MA, Elledge SJ, Jain RK. Origins of lymphatic and distant metastases in human colorectal cancer. Science. 2017;357(6346):55-60.
  2. Markowitz SD. Cancer bypasses the lymph nodes. Science. 2017;357(6346):35-36.
ESCP Affiliates