One Drop at a Time

Liquid biopsy is a hot topic in pathology at the moment – but, like many hot topics, we must delve deeper to discover the true pros and cons, and to uncover the most effective methods of accessing the benefits while avoiding pitfalls. In the case of liquid biopsy, it’s a minimally invasive technique that can provide rapid results. Droplet digital PCR (ddPCR) introduces speed and sensitivity, as well as absolute quantification to the liquid biopsy analysis. Alexander Dobrovic and Cloud Paweletz discuss their laboratories’ experiences with ddPCR and how other pathologists can achieve the same results.

Where do current methods of cancer characterization fall short?

Alexander Dobrovic: Current methods of cancer characterization based on tissue biopsies are often compromised by long turnaround times. In contrast, liquid biopsy samples can be taken as soon as the treating physician orders a test and sent directly to the laboratory. 

Of course, every approach has its strengths and limitations. Tissue biopsy remains the method of choice in many situations, especially in initial diagnosis, where anatomical pathology is a key part of the tumor evaluation. The ability to use circulating tumor DNA as a biopsy tool clearly depends on the presence of a detectable amount of tumor DNA in a relatively small (10–20 mL) blood draw, which is often not the case, particularly in early-stage tumors.

ddPCR is arguably the best approach to liquid biopsy for several reasons. Based on enumerating single molecules, it’s a technique that enables absolute quantification, making it very efficient at detecting even miniscule amounts of tumor DNA. ddPCR is highly sensitive and specific. Unlike massively parallel sequencing, there’s no need to batch samples for analysis, and results can be obtained in as little as six hours from the drawing of a blood sample in cases where an urgent result is needed.

Cloud Paweletz: I think it is worth mentioning that tissue biopsies are a fundamental part of cancer care and will not be eliminated. One has to remember that invasive biopsies are still the gold standard for making diagnoses, to clarify a diagnosis, to stage and re-stage, and to perform molecular testing. A good liquid biopsy can replace some inconvenient biopsies and open new opportunities that tumor biopsies may not offer. Philosophically, I see invasive and liquid biopsies as complementary.

So how can pathologists fit the technique into existing workflows?

AD: ddPCR precisely atomizes a single PCR reaction into approximately 20,000 micro-droplets as a stable emulsion in oil. These droplets act as micro-PCR reaction chambers. Only some will contain the DNA target of interest – so PCR amplification will only proceed in those droplets. Once the reaction is complete, amplification is detected by increased fluorescence using either intercalating dyes or probes; positive droplets are counted by flow cytometry.

Our experience using ddPCR has taught me how critical it is to test each assay to determine the false positive rate. This is particularly true when the main purpose of the assay is to detect minimal residual disease (MRD), because it determines the threshold above which one can make a confident positive call. It is reasonable to consider that the better we become at detecting emerging residual cancer early and accurately, the better our patients’ outcomes will be – so a thoroughly tested ddPCR assay whose evaluation of MRD can be trusted is an invaluable resource.

CP: We have to separate liquid biopsy from ddPCR; they are not synonymous. Liquid biopsy is the sampling and analysis of non-tissue samples. ddPCR is a technique to analyze samples, which can be based on tumors or plasma cell-free DNA), and next generation sequencing (NGS) is another such technique. 

ddPCR takes advantage of recent developments in microfluidics and surfactant chemistries. Conventional digital PCR involves diluting input DNA into individual wells for analysis; ddPCR emulsifies the input DNA into thousands of droplets that are PCR amplified and fluorescently labeled, then read in an automated droplet flow cytometer. Each droplet is individually assigned a positive or negative value based on its fluorescent intensity. A flow cytometer reads the number of positive and negative droplets, which is used to calculate the concentration of either wild-type or mutant allele (see Figure 1).

Any good (molecular) pathology laboratory that practices good PCR techniques and follows CLIA GLP/GDP guidelines will have no problem fitting ddPCR into their workflow. In essence, it adds two extra steps – droplet generation and automated droplet reading.

What obstacles remain for ddPCR’s widespread implementation?

AD: ddPCR needs a cancer-specific marker to function. However, this is also the downfall of the technique. In many cases, individual patients may need a bespoke marker – something that may not be financially or practically viable. And that’s why, at the moment, such assays remain boutique tests for the most part.

In my laboratory, we have always been aware of potential problems with PCR contamination – an issue with which every lab should be more than familiar. We now take even more precautions, using physical separation of laboratory areas to eliminate potential contamination, because even a miniscule amount of template can sabotage MRD detection.

The most useful ddPCR assays are based on driver mutations that can be used for both diagnosis and monitoring of response to treatment – and the more common the mutation, the more likely that test is to enter routine use in the clinic. Examples of tests that have made the leap are BCR-ABL for chronic myelogenous leukemia, JAK2 V617F for myelodysplasia, and BRAF V600E for melanoma.

CP: The main advantage and the pitfall of a liquid biopsy is that it promises to offer compelling insights at our fingertips without the hassles of obtaining a tumor biopsy – the complex scheduling, the risk for the patient, and the fact that they are not always successful. It’s important to understand how to interpret liquid biopsy results (whether NGS or ddPCR) so that the community doesn’t let a bad liquid biopsy supplant an inconvenient tumor biopsy.

I am agnostic to the method of analyzing cell-free DNA. In some cases, ddPCR may be the approach of choice – for instance, if one needs a rapid test to study a single variant of a single gene. In other cases, more comprehensive analysis using NGS may be more appropriate.

How did you make a business case for ddPCR?

AD: My colleagues and I made a business case to our hospital for a trial to diagnose and monitor disease in melanoma and lung cancer patients. Although our arguments were sound and our evidence base solid, it was critical to have the hospital’s medical oncologists support our case. We had already been providing this service to them from our research funds, a solution that was unsustainable in the long term. During this trial period, the value of ctDNA monitoring became apparent.

Liquid biopsy is an exciting space to be working in at the moment. As thrilling as it is to be at the cutting edge of new technologies, the most gratifying aspect is the positive contributions it makes to patient management.