To the editor:
In chronic myelogenous leukemia (CML) and Philadelphia chromosome–positive (Ph+) acute lymphoblastic leukemia (ALL), extensive data are available on the development of the 15 to 20 most common BCR-ABL1 kinase domain (KD) mutations arising after imatinib treatment and their differential response to other tyrosine kinase inhibitors (TKIs). This letter reports on less common KD mutations and suggests reasons for their rarity.
Over the past 5 years, we performed KD mutational analysis covering codons 221 to 500 of the ABL kinase on RNA extracted from blood or bone marrow leukocytes in patients with persistent/recurrent Ph+ ALL (n = 113) or TKI-resistant CML (n = 870), per published criteria.1,2 Mutation screening was performed after nested polymerase chain reaction with a sensitivity of 10%-20% mutation-bearing cells.3 Single nucleotide polymorphisms (noncoding changes observed involving codons 240, 249, 300, 320, 348, 354, and 499 as well as K247R4 ) were excluded, as were mutations occurring at common sites (codons 244, 250, 252, 253, 255, 276, 299, 311, 315, 317, 351, 355, 359, 387, 396, 453, and 459) whose features have been previously described.3,5-7 We noted 35 different uncommon point mutations arising at 30 codons in 41 patients (supplemental Table 1, available on the Blood Web site; see the Supplemental Materials link at the top of the online article). All of these mutations were unequivocally present by bidirectional sequencing and confirmed to be associated with TKI treatment by demonstration of their absence in previous and/or subsequent samples from the same patient, except for N331S, which was fully predominant on interferon therapy before imatinib. PCR artifact was excluded in those mutations that were not predominant over the unmutated sequence by repeat PCR from cDNA.
For CML, the median time to development of these uncommon KD mutations (41 months of TKI treatment) was similar to that for more common mutations. 4 mutations (L298V, L364I, E450G, and F486S) were present at high levels with another mutation (T315I in 3 and V299L in 1) in 1 case each, consistent with tandem occurrence in the same cells, with 3 occurring after sequential imatinib and dasatinib therapy.8 Most of the remaining mutations in CML arose in patients with variable imatinib resistance who often had interruptions and dose variations in imatinib due to toxicity or patient noncompliance (Table 1). Mutations noted at time of relapse of ALL/CML-blast phase were associated with prior or concurrent use of imatinib or dasatinib as maintenance therapy.
Clinical feature . | Mutations . |
---|---|
Mutations arising in CML chronic or accelerated phase | |
Arising with intermittent imatinib treatment | M237V, Y257C, L273M, E279K, E292K, E292V, I293V, S348L, L364I, V422I, W423R, Y435C, E450K, F486S |
Arising after nilotinib (no prior imatinib)* | K378R |
Arising after dasatinib (no prior imatinib)* | R473Q |
Mutation regressed followed imatinib dose escalation (or major molecular response) | V260A, E292V, I293V, G303E, W423R, Y435C |
Mutation regressed after switch of TKI (or major molecular response) | E292K, E292Q, E292V, L298V, V304A, L364I, M388L, L411P, I418T, V422I, A433T, L452P, F486S, T495R, M496I |
No regression after switch to a new TKI (see supplemental Table 1) | N311S (bosutnib + nilotinib), S348L (dasatinib) |
Mutations arising in CML blast phase or relapsed Ph+ ALL | |
At relapse of Ph+ ALL | V338G, E450G (along with T315I) |
At CML blast transformation | E279K (myeloid blasts), A337P (T-cell blasts), L364I (along with T315I, B-cell blasts), V379I |
Clinical feature . | Mutations . |
---|---|
Mutations arising in CML chronic or accelerated phase | |
Arising with intermittent imatinib treatment | M237V, Y257C, L273M, E279K, E292K, E292V, I293V, S348L, L364I, V422I, W423R, Y435C, E450K, F486S |
Arising after nilotinib (no prior imatinib)* | K378R |
Arising after dasatinib (no prior imatinib)* | R473Q |
Mutation regressed followed imatinib dose escalation (or major molecular response) | V260A, E292V, I293V, G303E, W423R, Y435C |
Mutation regressed after switch of TKI (or major molecular response) | E292K, E292Q, E292V, L298V, V304A, L364I, M388L, L411P, I418T, V422I, A433T, L452P, F486S, T495R, M496I |
No regression after switch to a new TKI (see supplemental Table 1) | N311S (bosutnib + nilotinib), S348L (dasatinib) |
Mutations arising in CML blast phase or relapsed Ph+ ALL | |
At relapse of Ph+ ALL | V338G, E450G (along with T315I) |
At CML blast transformation | E279K (myeloid blasts), A337P (T-cell blasts), L364I (along with T315I, B-cell blasts), V379I |
Both mutations regressed in retesting at 6 months after continued nilotinib or dasatinib therapy.
Overall, rarely observed BCR-ABL1 KD mutations were detected in 3 contexts: (1) in suboptimally treated patients with intermittent TKI therapy; (2) soon after switch from imatinib to another TKI (often regressing with continued therapy with the same TKI); or (3) in lymphoid blasts after maintenance TKI therapy. In repeat testing for partial or resistant disease after 6 to 12 months, all but 5 CML-associated mutations (L273M, E279K, L364I, N331S, and S348L) regressed with imatinib dose escalation or after switch to a new TKI. These results suggest that the range of BCR-ABL1 KD mutations is broad (similar to what is observed with in vitro mutagenesis),9 but that the common mutations10 out-compete the others and usually predominate at the time of overt TKI resistance when testing is commonly performed.
Authorship
Contribution: D.J. designed the study, analyzed data, and wrote the manuscript; S.S.C. performed experiments and analysis and reviewed the final manuscript; E.J. analyzed data, provided materials, and reviewed the manuscript; M.B.R. reviewed data and the manuscript; H.K. contributed materials and reviewed the manuscript; and J.C. contributed materials, reviewed the data, and wrote the manuscript.
Conflict-of-interestdisclosure: The authors declare no competing financial interests.
Correspondence: Dan Jones, MD, PhD, Quest Diagnostics Nichols Institute, 14225 Newbrook Dr, Chantilly, VA 20153; e-mail: dan.jones@questdiagnostics.com.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal