Predicting early relapse in multiple myeloma an ‘area of unmet need’
Click Here to Manage Email Alerts
As oncologists currently lack a cure for multiple myeloma, most patients will relapse at some point. The ability to predict when this will occur, however, may lead to more personalized therapy.
Although the heterogeneous nature of multiple myeloma has complicated attempts at developing prognostic tools, studies exploring different methods and markers for predicting patient outcomes, such as liquid biopsy and assessment of minimal residual disease, are underway.
In an interview with Healio, Shambavi Richard, MD, associate professor of medicine and co-director of CAR T and Stem Cell Transplant in the Multiple Myeloma Program at Tisch Cancer Institute and Icahn School of Medicine at Mount Sinai, discussed some of the factors associated with relapse in multiple myeloma, potential ways to monitor treatment and predict patient outcomes, and more.
Healio: What percentage of patients with multiple myeloma relapse after successful treatment?
Richard: Multiple myeloma is not curable. Most people who are newly diagnosed with myeloma will respond to first-line treatment. However, almost all will relapse and will need subsequent treatment. About a couple of decades ago, the median survival of patients who were newly diagnosed with multiple myeloma was only about 3 years. Since then, significant therapeutic advances have been made, resulting in the approval of several new drugs. Nevertheless, even though this has improved the outlook for myeloma, the percentage of patients who live beyond 5 years is still just over 50%. Although we have made good progress, we have still not been able to consistently cure myeloma, and it is only the occasional patient who may be considered cured.
Healio: Are certain patient or disease characteristics more predictive of relapse in multiple myeloma?
Richard: Myeloma is a very heterogeneous disease, and there have been several attempts at developing prognostic scores. We know that some patients may achieve a long-lasting remission with first-line regimens including proteasome inhibitors and immunomodulatory agents, followed by transplant and maintenance. However, early relapse is still considered an independent risk factor for resistance to subsequent treatments and shortened OS.
Individual treatment response duration is influenced by several disease-specific and patient-related features. Some patient-related features include age, sex — males seem to do less well than females — depth of response achieved with the treatment and inadequate response to primary treatment. We also now recognize that achievement of minimal residual disease (MRD) negativity is an important clinical goal to achieve with upfront treatments. Also, sustainability of MRD negativity is significant: the longer the patient stays in an MRD-negative state, the better the outcome.
Other patient characteristics, such as the presence of renal failure at diagnosis, can affect decision-making with respect to treatment in two ways. First, it may limit the choice and doses of drugs we use. Second, myeloma has a tendency to affect the kidneys, and this may remain a feature of the disease presentation at various relapse timepoints as well.
Additionally, other comorbidities that a patient may have may affect the ability to receive various treatments or their ability to participate in clinical trials. The effective use of maintenance therapy post-transplant is very important to prolong DFS. In fact, use of high-dose chemotherapy and transplant has been shown in several clinical trials to improve PFS and delay relapse. Therefore, that may be predictive as well.
In terms of disease-related factors, disease stage, as measured by International Staging System or the Revised International Staging System; lactate dehydrogenase level; cytogenetic abnormalities at diagnosis or at various relapse timepoints are very important. Certain high-risk genetic abnormalities, such as 17p deletions; 1q gain; translocations such as t(4;14), t(14;16) or t(14;20); and presence of more than one cytogenetic abnormality or hypolodiploidy all will elevate the risk for disease resistance or relapse. Also, presentation of the disease with extramedullary involvement is predictive of poor outcome.
The kinetics of presentation at any timepoint, meaning the speed of disease progression, the type of progression, whether it is biochemical or clinical, etc., may be predictive as well. Additionally, a high tumor burden with plasma cells more than 60% of the bone marrow or the presence of circulating plasma cells may also be predictive of earlier relapses.
Relapses that occur early — within 12 to 18 months of primary diagnosis — are highly predictive of an overall poor outcome.
Healio: Can physicians use other measures to predict relapse?
Richard: Predicting early relapse continues to be an area of unmet clinical need. There have been a lot of attempts to better characterize these patients. We know there is considerable evidence that MRD change from a negative to positive is highly predictive of clinical relapse. Bone marrow-based MRD assessment is something we are starting to use a lot more now, but it has several limitations because of the fact that the disease may be heterogeneously distributed in the marrow. It is also worth noting that sometimes, the samples may be hemodiluted. We may have a false-negative result with few plasma cells seen in the bone marrow, but patients may have extramedullary relapse. Therefore, using PET scan imaging would be a very important part of assessing this. Of course, bone marrow biopsies are invasive and uncomfortable. It is not any patient's favorite. Because of these issues, trying to frequently monitor MRD becomes very challenging.
The quest for an easily available and reproducible biomarker that predicts relapse is, of course, an area of much research interest. These are called liquid biopsies, which involve the detection of circulating tumor cells or circulating tumor DNA in the peripheral blood of patients. Circulating tumor DNA can be found either within the tumor cell or as cell-free DNA. These noninvasive biomarkers can be use to see if they can predict relapse in multiple myeloma. They are easily measurable in the peripheral blood and seem promising in some retrospective studies. However, we should remember it is probably unlikely that one single biomarker will be sufficient to be completely predictive of clinical outcomes, so we may need to integrate other parameters, such as protein expression, genomics, transcriptomics, etc., for more accurate predictions, but these are research-worthy areas that are of considerable interest.
Healio: How would these measures be used in clinical practice?
Richard: Although these have the potential to predict clinical outcomes, none of these are actually ready for primetime yet.
The goal would be to incorporate these technologies in clinical practice, both for therapeutic decision-making as well as for predicting treatment outcomes. Several studies, which are in full swing in the United States and Europe, are now looking at MRD-guided therapy, including the MASTER trial, the DRAMMATIC study, the AURIGA study, the OPTIMUM/MUK nine trial and the GMMG-CONCEPT trial. The myeloma community is recognizing the importance of being able to more accurately guide therapy instead using a one-size-fits-all approach. Ideally, the goal would be to eventually use these technologies both in newly diagnosed patients and in the relapse setting.
Healio: Are researchers looking into other ways to predict relapse?
Richard: The field is now starting to move more toward precision medicine or personalized medicine. This means managing an individual patient according to that patient's tumor biology or their risk characteristics. The goal in personalized medicine would be to select therapies based on the presence of these specific mutations, gene expression profiles, protein expression, etc. For instance, this is being studied in the Multiple Myeloma Research Foundation’s MyDRUG trial. Then, because there is a huge interplay between the patient's immune system and the tumor microenvironment, immune profiling of patients also is being explored to see how that influences treatment and outcome.
Finally, another approach being studied for predicting early relapse is machine learning, which is another big area of research.
Healio: Do you have a take-home message for our readers?
Richard: Ultimately, the only way we can make headway with a lot of these is just to enroll in clinical trials. Although we are doing better, particularly in terms of enrolling more diverse patient populations, more progress needs to be made in this area. Industries are working on innovative technologies, but physicians should enroll patients in clinical trials so we can get more information. That would be the way to partner in this approach.
Collapse
Healio Interview
Click Here to Manage Email Alerts