Insights From: Thomas J. Kipps, MD, PhD, UC San Diego Moores Cancer Center
Thomas J. Kipps, MD, PhD: We have very sensitive tests that can measure for minimal residual disease. The most widely used test involves flow cytometry. What that means is that the cells are put through a machine where they’re excited by laser light, and you can then, with very sensitive gating on the cells, detect 1 cell in 100,000 quite easily. Some people say that you can detect 1 cell in a million, but I think I agree that 1 cell in 100,000 cells is probably the limit, or 1 cell in 10,000 cells is the practical limit with this technology. So, I would say it’s fairly easy with many different colors on the cells to detect at least 1 in 10,000 with pretty good assurance. With anything less than that, you have to really have some fine instrumentation to detect. So, 1 in 10,000 is the limit for that technology, and maybe pushed to 1 in 100,000 with the appropriate instrumentation.
There’s another technique, and that is to use next generation sequencing, where we actually sequence a genome. This can actually achieve a detection limit of 1 in a million. Another technique, which is to use little specific primers, can also detect 1 cell in a million. But there are some important exceptions in that sometimes the genetic sequence of the leukemic clone is such where you can’t get down to that level of sensitivity—maybe only 1 in 100,000. So, patients will vary, depending on the genetic sequence in their leukemia, with the sensitivity of that test.
The other problem we have with that test is that you really do need the leukemic cells that were there prior to treatment. Because, if you just had the blood sample of patients after treatment and you don’t have the ability to look at the genetic sequence of the cells prior to treatment, then you cannot really use these techniques very well. So, that takes a little bit of advanced planning. You have to bank away your cells before a treatment and then it has to be compared, making these less attractive for many patients. I think we still need to work out ways to make that more feasible for many patients.
These techniques are still very good, and I think we’re settling on a limit of 1 in 10,000 cells as being the cutoff. And in fact, even if we had a test that could detect 1 cell in a billion, for example, it gets to the problem of what I mentioned before, where you’re reaching into that drawer of socks. So, if I reached into the drawer and pulled out a handful of socks and looked at it, white socks, I can easily spot 1 red sock in a thousand socks. But my hand only contains about maybe 50 socks. So, the sensitivity of the test that I’m using cannot really be greater than the number of socks in my hand. And therefore, our sensitivity then is governed by a limitation of sampling. When we pull blood from the vein or get cells from the marrow, we’re only getting a finite number of cells, and the limit on the number of cells that we draw is like the handful of socks in our hand. And so, we’re reaching a limit of close to 1 in a million, but we couldn’t do 1 in, say, a hundred billion sensitivity.
But still, I think what it represents is a significant advance from if we’re just looking at it by a microscope. Particularly, if you look at the bone marrow and find no evidence of residual disease in the bone marrow, that’s an excellent outcome. And we have these criteria of response. For example, we have complete response and we have partial response, and what distinguishes a complete response from a partial response is that, for a complete response, one must not have any enlarged lymph nodes and the spleen should be normal in size.
We’re pretty strict about it. Namely, if your lymph node happens to be larger than 1.5 cm, which is only about a half an inch, then it’s not considered a complete remission by the current standards. I think we’re being a little bit too stringent, because you can have a lymph node that’s enlarged from scarring, and that would keep you from being called a complete response. If you look by, for example, a CT scan and find a lymph node that’s a little over a half an inch in diameter, that’s pretty small, but it would keep it from being called a complete response.
An interesting study was done by the German CLL study group, where they took 2 large trials comparing the regimen of FC versus FCR or comparing BR versus FCR, and it involved quite a lot of patients—over 1000, collectively, or at least close to 1000. And what they did was assess the response to treatment, whether they achieved a complete remission or a partial remission, and in some of the patients they were able to assess for complete responses with MRD testing.
It’s a subgroup of all the patients. But what was noted in that study was quite striking: Patients who were able to have eradication of minimal residual disease, such that you couldn’t detect it, did very well. They seemed to not fare better if they had a complete response versus a partial response. Those patients who achieved a partial response, if they had eradication of minimal residual disease, they actually did almost as well as patients who had complete responses—very little difference between those groups. And they did better than patients who had a complete response who had detectable minimal residual disease. So, it may sound great to have a complete response to therapy, but if you go in and take a look at the fine print—namely, whether you have a complete remission or not with minimal residual disease. Finding no evidence in minimal residual disease is a very good thing, even if you achieve what might be considered a partial remission.