CLL: deep dive for residual cells by NGS matters

Thompson et al show that a substantial fraction of patients with chronic lymphocytic leukemia (CLL) in complete remission still harbors leukemic cells by using a next-generation sequencing (NGS) technique for detection of minimal residual disease (MRD). This NGS-based technology is able to detect 1 CLL cell out of 1 million leukocytes, thus having the potential of identifying patients with CLL at high risk for clinical relapse and allowing preparation of therapeutic rescue strategies as early as possible.¹ 

In CLL, impressive long-term remissions, including complete remissions with undetectable minimal residual disease (U-MRD), can be induced by chemoimmunotherapy based on fludarabine, cyclophosphamide, and rituximab (FCR).^2,3  Nevertheless, patients with CLL with specific high-risk features, such as an unmutated IGHV status, unfortunately still relapse after this aggressive chemotherapy despite some testing negative for MRD by classic flow cytometry–based assays.

The work by Thompson and colleagues tries to provide a better understanding for patients that are at true risk for a clinical relapse while tested negative for MRD in their blood or bone marrow by flow cytometry. Although the latter assay can only detect 1 positive CLL cell out of 10 000 normal cells (U-MRD4), an NGS-based technique lowers the detection sensitivity to the order of 1 positive cell out of 1 million cells (U-MRD6). Therefore, patients with undetectable disease by flow cytometry are rendered positive by this highly sensitive NGS-based method. In 62 patients with undetectable disease by flow cytometry (U-MRD4) at end of treatment with FCR, only 1 of 4 patients was documented to be MRD negative by NGS (U-MRD6). This significant difference between techniques is even more apparent for patients with CLL with an unmutated IGHV, a group considered to be at high risk for relapse. Only 13% of patients considered MRD negative by flow cytometry had undetectable disease by NGS. This observation has a major clinical impact because, unfortunately, most of these patients will relapse sooner or later. The median progression-free survival (PFS) was significantly shorter for patients with detectable disease by NGS at end of treatment.

The data on MRD assessment by NGS in this paper by Thompson et al were based on patients that were treated with a chemoimmunotherapy (FCR) that is nowadays often supplanted by novel targeted drugs like the Bruton tyrosine kinase inhibitor ibrutinib or the BCL2 inhibiting drug venetoclax.^4,5  Although 27.4% of patients showed undetectable disease by NGS after therapy with FCR, frontline therapy with venetoclax plus obinutuzumab is able to induce MRD-negative disease in a higher proportion (ie, 42% of patients) at this sensitivity level, as has been recently shown in the context of the CLL14 trial of the German CLL Study Group. In this trial, the former standard for elderly nonfit patients with CLL (ie, chlorambucil plus obinutuzumab) showed U-MRD in only 7% of patients, resulting in significantly shorter PFS compared with venetoclax plus obinutuzumab (hazard ratio 0.35). Therefore, in this trial, U-MRD by NGS has been shown to be associated with a survival advantage in a head-to-head comparison (see table).

The question arises whether NGS-based assays are ready for daily practice and might guide treatment decisions in the near future. Currently, a molecular relapse would not necessitate an immediate intervention, especially not after a time-limited frontline chemotherapy with FCR. Many treatment options for such patients, like ibrutinib or venetoclax plus rituximab, are available and approved at a clinical relapse. Not many laboratories would be able to fulfill the sophisticated standards needed to provide reliable MRD data at such a high sensitivity level, making sample shipping to experienced reference centers mandatory. However, MRD-guided stopping and reinitiation of therapy could be of major interest despite these hurdles, when prescribing more and more continuous or prolonged treatments with novel drugs to our patients. A time-limited treatment with a nonchemotherapy oral drug combination would be the future goal for the majority of our patients with CLL. Although a combination of ibrutinib and venetoclax can induce U-MRD in up to 69% of patients based on flow cytometry (U-MRD4), NGS-based assays might even better guide us in the future to define the subpopulation of patients with residual disease that would benefit from a prolonged combination therapy or an extended maintenance phase.⁶  Similar MRD-tailored treatment concepts in the frontline and relapsed setting have been introduced also by other research groups.^7,8  The better and more sensitive our MRD detection tool is, the better and more precise our treatment strategy will be for the individual patient with CLL. NGS-based MRD assessment is currently also developed for molecular monitoring of other leukemias (eg, chronic myeloid leukemia). Therefore, the data by Thompson et al in this issue of Blood are an important step forward that might guide us in the near future to treat our patients with different types of leukemia as much as possible, but as little as necessary.