Next-generation sequencing in the lymphoid malignancies
Dr Aaron Logan - UCSF, San Fransisco, USA
Surprisingly, most of the decisions that we make clinically about whether a patient has responded adequately to a therapy occur within a very small dynamic range of the disease burden that a patient can have. These are all old definitions that are based on the ability of a pathologist to see residual disease by light microscopy. So that has a sensitivity only down to a few percent residual malignant cells and so for many years the field has been looking at ways to look at deeper levels of disease response than the typical complete remission.
There are a number of technologies available for different diseases to quantify the amount of residual disease that’s present after initial or subsequent therapy. These include things like flow cytometry or a specific quantitative PCR called allele-specific oligonucleotide PCR and then more recently next generation sequencing that uses a consensus primer so that there is no need to customise the assay for each individual patient like needs to be done for ASO PCR. So using any of these technologies it’s been shown in several lymphoid malignancies that the achievement of a deep disease response, meaning less than 10-4 or 0.01% disease, is a prognostic for both disease free survival and, in many cases, overall survival.
There are some problems with wide implementation of flow cytometry because there is a lot of variability between labs and lack of standardisation so the molecular techniques potentially have the ability to standardise. Unfortunately the quantitative PCR method that I mentioned called ASO PCR does require patient specific primers which in the United States presents a barrier to moving it forward as a companion diagnostic. Now, we’re in contrast here in the US to the European Union where there’s actually very extensive use of ASO PCR through the EuroMRD consortium. However, that’s just not really going to work here in the United States. So the next generation sequencing approach for genetic MRD quantification has the benefit of using standardised primers and a standardised platform for all patients and so there’s no customisation. This is a platform that can potentially move forward as a companion diagnostic as an indication for additional therapy if a patient fails to have an adequate disease response.
What information does this technology produce?
This technology actually does not sequence the entire genome. This uses primers that bind to regions of the immunoreceptors that are rearranged during the natural process of development in B- and T-lymphoid cells. So using consensus primers for specific loci such as the immunoglobulin heavy chain or the immunoglobulin light chain or the various T-cell receptors such as beta, delta and gamma, you can amplify the entire repertoire of that specific gene rearrangement within a sample and that sample could be peripheral blood, it could be bone marrow, it could be a piece of lymphoma tissue. By doing that at diagnosis you can identify the dominant clone, or clonotype as we call it, and that represents the patient’s malignancy. It’s basically a naturally occurring bar code within the cancer cells that arose because of the natural process of immunoreceptor gene rearrangement. Once you know that specific bar code for a patient from using the next gen approach at diagnosis you can then take subsequent samples after treatment and quantify that specific clonotype sequence down to a very deep level and quantify residual disease down to one in a million leukocytes, so 10-6 sensitivity which is two logs more sensitive than flow cytometry and a log more sensitive than typical ASO PCR.
It’s very targeted high throughput parallel amplification of a specific locus so you’re not doing genome wide amplification although in the future there may be applications where you could do whole genome sequencing, still capture the immunoreceptors and have a way to quantify those. But currently this platform just amplifies specific immunoreceptors.
In lymphoid malignancies we have the advantage that these B- and T-lymphoid cells undergo a natural process of immunoreceptor rearrangement so essentially these diseases come to us already bar-coded. There’s no other disease that’s typically going to have this other than this class of lymphoid malignancies. So we can take advantage of this as a biomarker of the patient’s disease to quantify it with a very high sensitivity and very high specificity.
Are you finding any generalisations?
The generalisation is that within specific disease types we can make more or less summary statements about the meaning of specific disease burdens after induction therapy and then at subsequent time points through the course of therapy. Where I think the field needs to go is to start intervening on low levels of MRD progression as opposed to wait for patients to be significantly debulked and then relapse, be filled up with leukaemia cells again, be having symptoms from that. Typically they don’t have other blood counts so they’re really in clinical distress and if we can move to a point where we use a platform like the next gen platform for MRD quantification in serial fashion we’ll actually be able to capture the recurrence of disease at a very low level and potentially intervene at that time point. There are many therapies, including immunotherapies, that have a higher likelihood of working when there is a small amount of progressive disease as opposed to full clinical relapse.