Colorectal cancer is the third most common cancer worldwide with nearly 1.4 million new cases every year¹. Around half of all patients develop metastatic disease². Survival of patients with metastatic colorectal cancer (mCRC) has improved during the last 30 years, mainly due to improved and individualised surgical and oncological treatment strategies. Still, patient care pathways are heterogeneous in terms of treatment responses, quality of life and survival. This stresses the need for reliable biomarkers that can aid in clinical decision making; both identifying patients that will likely benefit from treatment upfront, for the timely management of subsequent treatment lines and the discontinuation of treatment.

DNA contains the recipe for making you unique. It can be compared to a library of about 5000 books with a total of 1 million pages, all stacked within each cell of the body. If a cell becomes cancerous, the letters, words or even pages of the books in the library may be erroneously edited, moved, deleted or torn out. DNA is normally located within the cell nucleus, but as DNA from dead cells are cleared from the body, short DNA fragments may temporarily be shed into the blood stream. These DNA fragments are called circulating cell-free DNA (cfDNA), and the subset derived from cancer cells may be termed circulating tumour DNA (ctDNA). ctDNA can be detected and characterised from a simple blood test by using appropriate sensitive methodologies.

In this study, we asked the following question; what is the clinical utility of ctDNA in predicting response to treatment or survival in patients with mCRC?

We performed a systematic review identifying more than 2000 papers on this topic, and 71 studies were included in the final review. Through this work, we describe four scenarios where ctDNA could be of clinical relevance;

1. Predicting treatment response/prognosis prior to treatment
2. Predicting treatment response/prognosis after treatment initiation
3. Early identification of progression (molecular biological progression)
4. Identification of novel actionable targets

We found that high levels of ctDNA prior to systemic anti-cancer treatment were associated with poor prognosis. Equally, a small or no early decrease in ctDNA levels during treatment was associated with both poor survival and predicted inferior treatment response. 

The prognostic impact seemed strong, and was confirmed in a subset meta-analysis including twenty-four studies. However, it is still unclear as to what degree ctDNA can complement or replace existing prognostic markers in mCRC.

Less data was available on molecular biological progression and the identification of novel actionable targets. A minority of studies reported on lead-time between molecular biological progression and radiological progression, and the mean lead-time ranged from 0 to 3.6 months. Although some studies indicate a utility in this regard, it is yet unknown if clinical outcomes will improve given that treatment is guided by ctDNA rather than or in addition to radiological assessments.

ctDNA may be more than just a prognostic marker. It can also be a tell-tale biomarker, revealing snippets of the cancer recipe along the patient care pathway. This information may be of potential relevance for monitoring changes in the disease biology, identifying novel actionable targets for treatment, and giving timely therapy advice to patients. Our review reveals that this potential has only partially been unravelled in this patient group, and prospective trials with companion diagnostics are urgently needed to move this closer to the clinic. Some clinical applications include, but are not limited to, the identification of patients eligible for re-challenge with anti-EGFR therapy, BRAF-targeted therapy, HER2-related therapy or immune checkpoint inhibition (i.e., MSI-H or high tumour mutational burden).

When working with this review and meta-analysis, we have identified a massive research interest within the field of ctDNA and mCRC. There has been a close-to exponential increase in relevant studies from the mid-2010s. Yet, our impression is that most studies/analyses have been conducted retrospectively, and analytical methodologies still vary greatly. As a result, most of the included studies had high risk of bias in at least one domain. Although recent efforts have been made to standardising pre-analytical sample handling, much work lies ahead in optimising for instance quality analyses, marker selection, cut-off levels and timing of sampling during treatment.

Advances in methodology have partly paved the way for much of the retrospective and explorative ctDNA research up until now. However, future ctDNA research should be strongly guided by the current and future clinical landscape of mCRC treatment.

We urge for prospective studies; investigating the role of ctDNA as companion diagnostics; comparing ctDNA with current clinical tools; relating the use of ctDNA with patient reported outcome measures; and exploring the cost-effectiveness of ctDNA.

 References:

1. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015 Mar 1;136(5):E359–86.
2. van der Geest LGM, Lam-Boer J, Koopman M, Verhoef C EM and de WJ, LG van der G, J L-B, M K, C V, MA E, et al. Nationwide trends in incidence, treatment and survival of colorectal cancer patients with synchronous metastases. Clin Exp Metastasis. 2015 Jun 10;32(5):457–65.