For patients with chronic lymphocytic leukemia (CLL), minimal residual disease (MRD) has emerged as a highly sensitive indicator of disease burden during and at the end of treatment, and has correlated with time-to-event outcomes. Use of MRD has shown promise for guiding duration of treatment for CLL patients on time-limited therapy.
Undetectable MRD (uMRD) -- defined as less than one CLL cell in in the peripheral blood and bone marrow at the end of treatment -- is assessed by multicolor flow cytometry, real-time quantitative polymerase chain reaction, and high throughput sequencing. Recent, more sensitive analysis can detect MRD at thresholds of 10-5 and even 10-6.
MRD has correlated with favorable progression-free survival (PFS) and overall survival (OS) after chemoimmunotherapy. Given the utility in evaluating depth of response, determining uMRD status is becoming a focus of outcomes in CLL trials, particularly of fixed-duration therapy.
Recently updated results of the showed that treatment-naive patients given fixed-duration venetoclax (Venclexta) and obinutuzumab (Gazyva) exhibited detectable MRD at a later timepoint than those given chlorambucil plus obinutuzumab, with only 7.9% in the venetoclax arm requiring a next line of therapy more than 3 years after treatment cessation. Estimated 4-year PFS and OS were also higher in the venetoclax arm.
Significantly, regular MRD assessments over time demonstrated deep clearance of CLL cells in the peripheral blood of venetoclax-obinutuzumab recipients, testifying to its potential as an end-of-treatment prognostic parameter. MRD in this study was detected at a sensitivity level of 10-6.
"MRD measurement is not necessarily standard of care at the moment, but it will continue to emerge as an important goal of treatment in fixed-duration CLL regimens," said Catherine Coombs, MD, of the University of North Carolina School of Medicine in Chapel Hill. "Studies in venetoclax-based treatment have found achieving negative MRD levels to be prognostic of favorable outcomes with pronounced better PFS and OS."
Daniel Persky, MD, of the University of Arizona College of Medicine in Tucson, agreed:
"Assessment of MRD is becoming an important and actionable test in the care of CLL patients. We use it to make decisions about discontinuing CLL treatment. Data show that patients with MRD-negative complete remissions may consider discontinuing some CLL treatments."
So how low do uMRD levels have to go to predict a good outcome?
"Classically, we used to take 10-4, but we now have the ability to detect levels of 10-6, and a lower level is likely better," Coombs said.
For patients on continuous therapy with Bruton's tyrosine kinase (BTK) inhibitors, however, MRD does not preclude prolonged PFS. For example, earlier this year a reported that MRD was not associated with improved PFS in patients on indefinite ibrutinib (Imbruvica)-based therapy. PFS, however, was significantly longer in patients with undetectable MRD who received six cycles of fludarabine, cyclophosphamide, and rituximab (Rituxan).
As to whether MRD levels are ready to guide clinical decision-making, Coombs was cautious:
"I don't think we're there yet. That's still up for debate, and the big question is whether MRD results are actionable. It's pretty clear that patients with undetectable MRD do better, and I think MRD measurement will be the way of the future."
The speed with which uMRD is achieved may be an important factor, according to Jacob Soumerai, MD, of Massachusetts General Hospital Cancer Center in Boston, and Andrew Zelenetz, MD, PhD, of Memorial Sloan Kettering Cancer Center in New York City.
At last year's American Society of Hematology (ASH) virtual annual meeting, their group presented a looking at MRD-driven time-limited therapy with zanubrutinib (Brukinsa), obinutuzumab, and venetoclax in previously untreated CLL.
The combination achieved frequent, durable uMRD responses, with 89% of patients achieving uMRD in both the peripheral blood and the bone marrow; therapy was discontinued after a median of 10 months.
"Importantly, the MRD kinetics data we presented at ASH suggest that it's not just whether or not you achieve uMRD, but rather your early MRD response kinetics might predict MRD outcomes and define biologic differences," Soumerai told .
In the study, 60% of patients exhibited rapid MRD response kinetics (ΔMRD400 achievers) and these patients had rapid bone marrow uMRD (100% within 8 months) and more durable uMRD of <10-5. In contrast, 40% of patients did not achieve ΔMRD400. These had more delayed bone marrow uMRD, and early detectable MRD of ≥10-5, despite longer treatment courses owing to the trial's MRD-driven treatment duration.
"These data suggest that early dynamic MRD response assessment is a prognostic biomarker for MRD outcomes and should be explored as a predictive biomarker for guiding treatment," Zelenetz said.
So how often should MRD be measured?
"This is an important question," Soumerai said. "I assess MRD at end of treatment, typically using the approved clonoSEQ assay, but I do not use MRD as a monitoring tool after treatment discontinuation."
For a patient who develops recurrent detectable MRD without clinical evidence of disease, this does not currently impact management. "I continue to monitor this patient at the same interval of time and only initiate therapy if the patient develops an indication for treatment," Soumerai said.
Recurrent lymphocytosis or adenopathy is typically detected in asymptomatic patients by physical exam and routine laboratories.
"The question is whether some patients might benefit from early therapy in the relapsed/refractory setting, but this should only be done in the context of a clinical trial," Zelenetz said.
MRD kinetics appear to identify two groups of patients, said Soumerai: "Patients who achieve ΔMRD400 reach early uMRD, which appears to be durable. The patients with slower responses are at risk for early recurrence of MRD and probably should have longer treatment duration."
As for MRD's role in other CLL treatments such as BTK and PI3Kdelta inhibitors, "since these regimens rarely achieve uMRD and are given continuously irrespective of the depth of response, I do not routinely assess MRD for patients on these therapies," Soumerai noted.
Coombs said that although MRD levels are not yet routinely used across the country to alter the length of treatment with fixed-duration venetoclax-based regimens, "I think further trials will bring MRD-adaptive therapy into play and alter the length of treatment based on MRD response, but that has to be hashed out in future studies."
Persky foresees a time when MRD will routinely guide decisions about duration of therapy and may become a new goal of treatment.
"It could also potentially be used to restart treatment on someone whose CLL has converted from MRD-negative to MRD-positive status," he said.
But while MRD testing is becoming more sensitive and able to detect smaller and smaller amounts of residual disease, "before we agree on uMRD becoming a new goal, we need to agree on a standard way of measuring it," he added.