Circulating Tumor DNA May Be Able To Predict Breast Cancer Metastases

Lao Saal, M.D. Interview with:
Lao Saal, M.D. Ph.D.
Head, Translational Oncogenomics Unit
Assistant Professor
Department of Oncology and Pathology
Lund University Cancer Center
Lund, Sweden

Medical Research: What is the background for this study?

Dr. Saal: About a quarter of women diagnosed with primary (non-metastatic) breast cancer will unfortunately progress and later be found to have metastatic spread, which can occur many years to even a decade or more after the first diagnosis. At this point, the prognosis after identification of metastatic breast cancer is very poor. Metastatic disease is typically diagnosed only after it has grown large enough to cause symptoms, be noticed on exam, or be detectable by imaging. It is thought that early detection of metastasis has the potential to lead to better outcomes because therapies could be modified when the metastasis is still very small. Moreover, a very sensitive and specific test that could identify patients who appear “cancer-free” could also be useful. Essentially all cancers have unstable genomes, where chromosomes physically break and are reassembled incorrectly and thus the normal sequence is altered. Importantly, DNA material from cancer cells can be found in the blood circulation and therefore this circulating tumor DNA has the potential to be a cancer biomarker.

Medical Research: What are the main findings?

Dr. Saal: Eleonor Olsson, a PhD student in my lab who defends her thesis next week, and Christof Winter, a postdoc bioinformatician in my group, were the first authors of the paper. In our study we tested whether periodic monitoring of circulating tumor DNA (ctDNA) in blood plasma samples, taken before surgery for primary breast cancer and at multiple timepoints after surgery, could identify the metastastic spread, and whether the quantity of ctDNA was associated to patient outcome. We analyzed a retrospective cohort of 20 patients, who had enrolled many years ago in a separate epidemiological study run by Helena Jernström in our department, wherein the appropriate blood plasma samples had been biobanked and tumor tissue was available and we had long-term clinical follow-up information.

As far as we are aware, our study is the first to show the potential for serial ctDNA monitoring in the context of primary breast cancer. We found that our ctDNA blood tests could discriminate patients with eventual metastasis from those with long‐term disease‐free survival with 93% sensitivity and 100% specificity. Furthermore, ctDNA‐based detection of metastatic disease preceded clinical detection for 86% of patients by an average 11 months and in some cases by 3 years. In all of the patients who had long-term disease-free survival, we did not detect any ctDNA in any timepoints after surgery. Lastly, the measured quantity of ctDNA was a significant predictor of outcome: for each doubling of the ctDNA level, the odds ratio for metastasis was 2.1 and the odds ratio for death was 1.3. An interesting anecdote — one patient we studied had bilateral breast cancer and we found that it was the right-side tumor (which actually had more favorable clinical characteristics) that gave rise to the metastasis and not the left-side tumor.

I should note that our approach was, in the first step, to perform low-pass whole-genome sequencing of the primary tumors that is sufficient to identify chromosomal rearrangements that are specific to each patient’s tumor. We developed our own software SplitSeq to improve the reconstruction of the exact breakpoint sequence. These rearrangements serve as genetic “fingerprints” for the tumor. We then designed personalized assays for a panel of 4-6 rearrangements for each patient and used droplet digital PCR to measure the quantity of the rearranged ctDNA in each blood sample (93 samples in all). In the paper we show the method to have a limit of detection at least down to 0.01% fraction of ctDNA, and we know this can go even further. Our method is an improvement upon an earlier method called PARE, which was developed by Victor Velculescu and colleagues at Johns Hopkins University. Importantly, our method is very specific and we have yet to encounter a false-positive rearrangement in an analysis of over 2.5 million normal haploid genome equivalents; so it compares very well to other methods. For serial monitoring of many timepoints per patients, our approach is more economical than others that apply next-gen sequencing to each timepoint, and once the personalized assays are ready, the actual ctDNA analyses can be performed with a very short turnaround time.

Medical Research: What should clinicians and patients take away from your report?

Dr. Saal: Our study shows proof-of-concept that ctDNA is a robust cancer biomarker that could be used in the context of early breast cancer to detect metastatic disease earlier than current clinical practice. In addition, the quantity of ctDNA appears to be clinically relevant in this patient group.

Medical Research: What recommendations do you have for future research as a result of this study?

Dr. Saal: We believe that ctDNA analysis holds much promise and we endeavor to advance this work so that one day it can benefit patients. Of course, further studies need to be performed, in larger patient groups both retrospectively and prospectively, to validate our findings and provide evidence that ctDNA analyses can lead to better clinical outcomes. The methods also apply to most cancer types, and there are a number of different approaches to analyze ctDNA including analyzing for gene mutations. My group is pursuing such studies and we are very interested in collaborating with other researchers, biotech, and pharma. We believe our results and the work of others suggest further studies to evaluate the clinical utility of serial ctDNA measurements for monitoring response to treatment, identifying metastatic relapses earlier, and for steering therapy decisions. Another interesting avenue is whether ctDNA analysis can be used to help avoid the overtreatment of patients.


Eleonor Olsson, Christof Winter, Anthony George, Yilun Chen, Jillian Howlin, Man‐Hung Eric Tang, Malin Dahlgren, Ralph Schulz, Dorthe Grabau, Danielle van Westen, Mårten Fernö, Christian Ingvar, Carsten Rose, Pär‐Ola Bendahl, Lisa Rydén, Åke Borg, Sofia K Gruvberger‐Saal, Helena Jernström, Lao H Saal. Serial monitoring of circulating tumor DNA in patients with primary breast cancer for detection of occult metastatic disease. EMBO Molecular Medicine, May 2015

[wysija_form id=”3″] Interview with: Lao Saal, M.D. Ph.D. (2015). Circulating Tumor DNA May Be Able To Predict Breast Cancer Metastases

Last Updated on May 19, 2015 by Marie Benz MD FAAD