We recently had a chance to interview one of the speakers at GTCbio’s upcoming Molecular Diagnostics Summit 2015. Dr. Mark Erlander, Chief Science Officer of Trovagene, will be giving a presentation entitled “Quantitating ctDNA Mutations from Patient Urine as an Early Pharmacodynamic Biomarker for Response to Targeted Therapy”. We caught up with him recently to ask about the liquid biopsy market and how next generation sequencing is making an impact.
AllSeq – What are the key advantages of liquid biopsies over more traditional diagnostics?
Dr. Erlander – My prediction is that liquid biopsies will be transformational in the diagnostic and monitoring space of oncology. This is really taking hold because of some fundamental advantages over traditional techniques. Liquid biopsies are useful both as diagnostics and for patient monitoring.
First, there’s the question of tumor biology. In terms of diagnostics, liquid biopsies offer a ‘whole body’ assay. In stage IV disease and metastatic patients the tumors are heterogeneous, so the tumor biology is much more complex. The treating physician has to ask “what metastatic lesion do I biopsy to determine what’s going on with this patient’s cancer and how do I treat it?” There’s an issue of both inter-tumor and intra-tumor heterogeneity – they correspond to different types of mutations. By performing a liquid biopsy (as opposed to the traditional method of biopsying a particular tissue), you get a much more complete, comprehensive view of the tumor.
Second, there’s the ‘non-invasive’ aspect of liquid biopsies. For example, in lung cancer there’s a significant risk of patient morbidity with CT-guided biopsies where a lung could collapse as well as other risks. Related to this is the idea of patient compliance where some patients say “I don’t want to do that!” Liquid biopsies don’t have these issues.
The third advantage is that of cost. There’s substantial cost associated with biopsies. In lung cancer, a CT-guided biopsy costs upwards of $4,000 and when you add in the complications seen in 20-30% of the patients, you’re more in the $15,000 range. In addition to the initial diagnosis, because of the lower costs and non-invasive nature, liquid biopsies are much more amenable to patient monitoring. Monitoring patients, like they were not able to do before, gets us into the other area and that’s CT scans. With CT scans, if a patient is getting a cancer therapy they’re then coming in every 6 weeks to get a scan to see if the tumor has shrunk, progressing or stabilized. These kinds of non-molecular assessments don’t tell you the whole story. If you can have a liquid biopsy that says “No, hang on. They have an EGFR-activating mutation – it’s gone down. The number of detected fragments has gone down so they’re responding to the therapy”. The thing about the CT scan is that you’ve got to have an appointment. It takes the whole day to get the scan and then you’ve got to wait for the results. By measuring the number of copies of a somatic mutation you can tell very quickly whether or not the patient is responding to therapy. What that means is that liquid biopsies allow you to give a much more near real-time assessment and one doesn’t have to wait for a scan.
AllSeq – What makes Trovagene unique in this market?
Dr. Erlander – All of the other liquid biopsy companies out there work only with blood. Ours are the only assays that work with both blood and urine. Urine is the best specimen for a variety of reasons. First, it’s a lot easier to get a urine sample than any other sample. It doesn’t require any health professional – it can be done with what we call ‘home sampling’. Urine is stable for two weeks so it’s not an issue – they can just send it to us at room temperature. Also, there are some advantages to urine just at the level of non-invasiveness and the level of practicality. We give a lot greater options to the treating physician for their patients.
The second thing is you can get a lot more urine than you can get blood. We get 10x more volume than blood and we get 10x more DNA, as the concentration of DNA is roughly equivalent between the two. Although there is some variation in urine, when we’ve looked over hundreds of patients, we get on average about 10x more DNA than from blood. Having more DNA enables us to do more assays and it enables us to be more sensitive. While tumors shed apoptotic cells, it’s in a sea of huge amounts of wild-type DNA from other, normal cells that go through apoptosis – their genomic DNA is in blood and in urine as well. So you’re always dealing with a sensitivity issue with liquid biopsies because your background is probably 100-fold greater than your target.
In a normal tissue biopsy people talk about having a sensitivity of 1%, so out in 1 out of 100 tumor cells you should be able to see a mutation. And beyond that, clinically we don’t know what it means if you’re more sensitive. Now, in liquid biopsies, you’re not looking at a tumor – you’re looking at the whole sea of wild-type DNA for an entire organism from both normal cells and tumor cells. With the large amounts of DNA in urine our assays have detection limits in the range of anywhere from 1:20,000 to 1:100,000.
Getting back to the idea of monitoring, working with urine allows you to have daily collections. We just had an oral presentation at the European Lung Cancer conference in Geneva with our collaboration with UCSD that showed that for a lung cancer patient who gets an EGFR inhibitor, we can monitor daily and have a baseline for the first seven days. And at some point in the first seven days we see massive peaks of T790M or EGFR activating mutation which is indicative of tumor cell kill by the drug. That peak is indicative of the fact that the drug is doing what it’s supposed to do to those tumor cells. When you kill the tumor cells, the genomic DNA is going to end up in the urine. The urine allows daily collection, something that you can’t really do with blood. So the urine is a very simple way of determining right away if the drug is doing what it’s supposed to be doing. It really opens up the horizons significantly and that’s why we’re so excited about it.
AllSeq – How are you using NGS in your assays?
Dr. Erlander – We use standard sequencing with an upfront enrichment for the mutations out of the wild-type background. We do targeted sequencing and we block wild-type amplification, significantly reducing the wild-type background, enabling the detection of the sequences with the somatic mutation. This gives us much higher sensitivity. We use NGS to count the number of sequences we see along with controls so that we can extrapolate back to an actual quantitation of how much was present in the initial sample.
AllSeq – What improvements in NGS are you looking for?
Dr. Erlander – We don’t need longer reads because we look at very short sequences. Error rates are always an issue. We control for those, but the lower the error rate the less the background and the greater the sensitivity. We’re currently using the MiSeq platform, but we may switch to a HiSeq just to get the deeper sequencing for added sensitivity. I would love to have the sequencing be cheaper. Because it’s expensive, we’re forced to look at other platforms as well.
But a lot of the improvements will likely have to come from our proprietary upfront processing – the building of the libraries. You have to manipulate upfront from the sequencer at this point. Sequencing is the workhorse, but to have the sensitivities that we’re talking about, enrichment for the mutations prior to sequencing is required.