https://orcid.org/0000-0002-3216-2966
Lineage-specific assessment of residual disease could identify patients who can safely cease therapy.
Detection of BCR::ABL1 DNA in granulocytes and T cells at TKI cessation is a better predictor of relapse than those in total leukocytes.
Patients with chronic myeloid leukemia who are eligible for treatment-free remission (TFR) may still relapse after tyrosine kinase inhibitor (TKI) cessation. There is a need for accurate predictors of outcome to enable patients with a favorable profile to proceed while avoiding futile attempts. Sensitive detection of residual disease in total leukocytes at treatment cessation is associated with relapse but is not highly discriminatory, likely because it is a composite measure of residual leukemia derived from different cell lineages, whereas only some lineages are relevant for relapse. We prospectively measured BCR::ABL1 DNA as a predictive yes/no binary test in 5 cellular fractions from 48 patients meeting conventional criteria for TKI discontinuation. The median BCR::ABL1 DNA level was higher in granulocytes and T cells, but not in other lineages, in patients who relapsed. Among the 40 patients undergoing their first TFR attempt, we defined 3 groups with differing relapse risk: granulocyte-positive group (100%), granulocyte-negative/T-cell–positive group (67%), and granulocyte-negative /T-cell–negative group (25%). These data show the critical importance of lineage-specific assessment of residual disease in the selection of patients who can attempt to achieve TFR with a high expectation of success and, concurrently, defer patients who have a high probability of relapse.
Comments
Response to Machova Polakova et al.
1,2
In a prior work Machova Polakova and colleagues used paired analysis of BCR::ABL1 DNA and RNA and showed that patients with undetectable BCR::ABL1 using both measurements had the lowest relapse risk and those double-positive had the highest relapse risk3. They pose the question whether the two studies are reporting the same phenomenon: i.e. does the detection of BCR::ABL1 DNA without RNA reflect the presence of BCR::ABL1-positive T lymphocytes with absent RNA expression, and therefore an intermediate risk of relapse.
At diagnosis BCR::ABL1 is expressed in both B and T cells.2 The median fraction of BCR::ABL1 DNA positive lymphocytes was around 1% and the median expression ratio was similar, suggesting that the relative expression per cell is similar to other lineages that were tested, including granulocytes.
An alternative explanation for the pattern reported by Machova Polakova is that sensitive RNA-based methods have a background false positive rate.4,5 Since the odds of having a false positive result in both RNA and DNA MRD tests is lower, the finding of concordant values may reveal a higher-level of MRD (double-positive) or a stringent lower-level of MRD (double-negative), with discordant values representing an intermediate level. Rare transcripts from B lymphocytes could represent a kind of biological false positive, i.e. a technical true positive, but not associated with relapse risk.
Although BCR::ABL1 DNA PCR on sorted cells is technically demanding, lineage-specific MRD provides greater biological insights than bulk MRD into the cell types responsible for relapse risk.
References
1. Pagani, I.S., Shanmuganathan, N., Dang, P., Saunders, V.A., Grose, R., Kok, C.H., James, J., Tolland, M., Braley, J.A., Altamura, H.K., et al. (2023). Lineage-specific detection of residual disease predicts relapse in patients with chronic myeloid leukemia stopping therapy. Blood 142, 2192-2197. 10.1182/blood.2023021119.
2. Pagani, I.S., Dang, P., Saunders, V.A., Grose, R., Shanmuganathan, N., Kok, C.H., Carne, L., Rwodzi, Z., Watts, S., McLean, J., et al. (2020). Lineage of measurable residual disease in patients with chronic myeloid leukemia in treatment-free remission. Leukemia 34, 1052-1061. 10.1038/s41375-019-0647-x.
3. Machova Polakova, K., Zizkova, H., Zuna, J., Motlova, E., Hovorkova, L., Gottschalk, A., Glauche, I., Koblihova, J., Pecherkova, P., Klamova, H., et al. (2020). Analysis of chronic myeloid leukaemia during deep molecular response by genomic PCR: a traffic light stratification model with impact on treatment-free remission. Leukemia 34, 2113-2124. 10.1038/s41375-020-0882-1.
4. Franke, G.N., Maier, J., Wildenberger, K., Cross, M., Giles, F.J., Muller, M.C., Hochhaus, A., Niederwieser, D., and Lange, T. (2020). Comparison of Real-Time Quantitative PCR and Digital Droplet PCR for BCR-ABL1 Monitoring in Patients with Chronic Myeloid Leukemia. J Mol Diagn 22, 81-89. 10.1016/j.jmoldx.2019.08.007.
5. Branford, S. (2020). Why is it critical to achieve a deep molecular response in chronic myeloid leukemia? Haematologica 105, 2730-2737. 10.3324/haematol.2019.240739.
DNA based MRD analysis and treatment free remission in CML
These findings closely align with our published data (Machova Polakova et al. Leukemia 2020), which explored molecular relapse-free survival (MRFS) based on the detection of BCR::ABL1 DNA and RNA in total leukocytes before TKI cessation. Using a very similar traffic light system to Pagani et al, we identified three distinct patient groups with significant differences in MRFS probability (P= 0.0005). The green group (n=11, RNA negative/DNA negative) exhibited 100% MRFS at 18 months, the yellow group (n=16, MRD RNA negative/DNA positive) showed 63% MRFS, and the red group (n=15, double positive) had 20% MRFS.
Earlier work by Pagani et al. (Leukemia 2020) indicated that in patients in TFR with undetectable BCR::ABL1 transcripts, BCR::ABL1 DNA positivity was confined to the lymphoid compartment. Combining these findings, it appears that T cells may be BCR::ABL1 positive without expressing BCR::ABL1, and this likely explains the similarities in the TFR stratification models between both studies. We suggest that the approach taken in our study is easier to implement in routine practice since it does not require analysis of cell fractions.
References:
1. Pagani IS, Shanmuganathan N, Dang P, et al. Lineage-specific detection of residual disease predicts relapse in patients with chronic myeloid leukemia stopping therapy. Blood 2023; 142(25): 2192-2197
2. Machova Polakova K, Zizkova H, Zuna J, et al. Analysis of chronic myeloid leukaemia during deep molecular response by genomic PCR: a traffic light stratification model with impact on treatment-free remission. Leukemia 2020; 34:2113-2124.
3. Pagani IS, Dang P, Saunders VA, et al. Lineage of measurable residual disease in patients with chronic myeloid leukemia in treatment-free remission. Leukemia 2020; 34:1052-1061