Limits...
Detection of minimal residual disease in NPM1-mutated acute myeloid leukemia by next-generation sequencing.

Salipante SJ, Fromm JR, Shendure J, Wood BL, Wu D - Mod. Pathol. (2014)

Bottom Line: Next-generation sequencing was precise and semiquantitative over four orders of magnitude.Further, in one-third of patients, sequencing detected alternate NPM1 mutations in addition to the patient's index mutation, consistent with tumor heterogeneity.This approach may complement current technologies to enhance patient-specific clinical decision-making.

View Article: PubMed Central - PubMed

Affiliation: 1] Departments of Laboratory Medicine, University of Washington, UW Hematopathology Laboratory at SCCA, Seattle, WA, USA [2] Genome Sciences, University of Washington, Seattle, WA, USA.

ABSTRACT
Detection of minimal residual disease predicts adverse outcome in patients with acute myeloid leukemia. Currently, minimal residual disease may be detected by RQ-PCR or flow cytometry, both of which have practical and diagnostic limitations. Here, we describe a next-generation sequencing assay for minimal residual disease detection in NPM1-mutated acute myeloid leukemia, which encompasses ∼60% of patients with normal karyotype acute myeloid leukemia. Exon 12 of NPM1 was PCR amplified using sequencing adaptor-linked primers and deep sequenced to enable detection of low-prevalence, acute myeloid leukemia-specific activating mutations. We benchmarked our results against flow cytometry, the standard of care for acute myeloid leukemia minimal residual disease diagnosis at our institution. The performance of both approaches was evaluated using defined dilutions of an NPM1 mutation-positive cell line and longitudinal clinical samples from acute myeloid leukemia patients. Using defined control material, we found this assay sensitive to approximately 0.001% mutant cells, outperforming flow cytometry by an order of magnitude. Next-generation sequencing was precise and semiquantitative over four orders of magnitude. In 22 longitudinal samples from six acute myeloid leukemia patients, next-generation sequencing detected minimal residual disease in all samples deemed negative by flow cytometry. Further, in one-third of patients, sequencing detected alternate NPM1 mutations in addition to the patient's index mutation, consistent with tumor heterogeneity. Next-generation sequencing provides information without prior knowledge of NPM1 mutation subtype or validation of allele-specific probes as required for RQ-PCR assays, and without generation and interpretation of complex multidimensional flow cytometry data. This approach may complement current technologies to enhance patient-specific clinical decision-making.

Show MeSH

Related in: MedlinePlus

Evidence for NPM1-mutation heterogeneity in clinical samples(A) Multiple alignment of presentation (index) and variant sub-clone NPM1 mutations detected in Patients-1 and -2. Bold indicates NPM1 stop codon. (B) Detection of sub-clones in Patient-1 showing decline of the index clone and subsequent rise of secondary clone (Type A mutation). (C) Detection of sub-clones in Patient-2, showing development and low-level persistence of secondary Type A mutation. Heterogeneity is also demonstrated in the index case by the presence of a Type R allele. (D). Flow cytometry immunophenotype for Patient-1 at time-points: 0 (index), top row, day 1162, middle row, and then one additional time-point, approximately 2.5 months after day 1162, bottom row. The original index clone was an expanded monoblast population, (top row, 2nd column, asterisk) without a substantial myeloid blast population component (top row, 3rd and 4th columns). The alternate clone in subsequent samples (bottom 2 rows) no longer had a monoblastic immunophenotype but rather had an abnormal myeloid blast population with aberrant lymphoid antigen expression and accounted for 0.3% and then 1.8% of total white cells, respectively (bottom two rows, 3rd and 4th columns, arrows)
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4201902&req=5

Figure 2: Evidence for NPM1-mutation heterogeneity in clinical samples(A) Multiple alignment of presentation (index) and variant sub-clone NPM1 mutations detected in Patients-1 and -2. Bold indicates NPM1 stop codon. (B) Detection of sub-clones in Patient-1 showing decline of the index clone and subsequent rise of secondary clone (Type A mutation). (C) Detection of sub-clones in Patient-2, showing development and low-level persistence of secondary Type A mutation. Heterogeneity is also demonstrated in the index case by the presence of a Type R allele. (D). Flow cytometry immunophenotype for Patient-1 at time-points: 0 (index), top row, day 1162, middle row, and then one additional time-point, approximately 2.5 months after day 1162, bottom row. The original index clone was an expanded monoblast population, (top row, 2nd column, asterisk) without a substantial myeloid blast population component (top row, 3rd and 4th columns). The alternate clone in subsequent samples (bottom 2 rows) no longer had a monoblastic immunophenotype but rather had an abnormal myeloid blast population with aberrant lymphoid antigen expression and accounted for 0.3% and then 1.8% of total white cells, respectively (bottom two rows, 3rd and 4th columns, arrows)

Mentions: In Patient-1, an alternate clone was clearly identified. In the index sample, only one NPM1 mutation (NM_002520.6:c.[864G>C;869delinsTCCTA]) was detected. However, in four subsequent samples spanning 38 months, next-generation sequencing detected increasing levels of the Type A allele that were quantified significantly above the empiric error frequency (>3.5 standard deviations), with a concurrent decrease of the index allele (Figure 2A, B). At days 1127 and 1162, the read counts of Type A allele surpassed those of the index allele. In Patient-2, we similarly detected heterogeneity (Figure 2A, C), harboring a Type B index mutation (NM_002520.6:c.863_864insCATG). A low-prevalence, but significant level (>900 standard deviations above the empiric error frequency) of Type R allele (NM_002520.6:c.863_864insTATG) was found in the index case, but became undetectable after initialization of treatment. By day 37, a low-prevalence Type A clone was identified at significant levels (>25 standard deviations above the empiric error frequency), which became dominant at day 110. At ∼1 year, the amount of Type A allele was surpassed by the index clone, but nevertheless persisted.


Detection of minimal residual disease in NPM1-mutated acute myeloid leukemia by next-generation sequencing.

Salipante SJ, Fromm JR, Shendure J, Wood BL, Wu D - Mod. Pathol. (2014)

Evidence for NPM1-mutation heterogeneity in clinical samples(A) Multiple alignment of presentation (index) and variant sub-clone NPM1 mutations detected in Patients-1 and -2. Bold indicates NPM1 stop codon. (B) Detection of sub-clones in Patient-1 showing decline of the index clone and subsequent rise of secondary clone (Type A mutation). (C) Detection of sub-clones in Patient-2, showing development and low-level persistence of secondary Type A mutation. Heterogeneity is also demonstrated in the index case by the presence of a Type R allele. (D). Flow cytometry immunophenotype for Patient-1 at time-points: 0 (index), top row, day 1162, middle row, and then one additional time-point, approximately 2.5 months after day 1162, bottom row. The original index clone was an expanded monoblast population, (top row, 2nd column, asterisk) without a substantial myeloid blast population component (top row, 3rd and 4th columns). The alternate clone in subsequent samples (bottom 2 rows) no longer had a monoblastic immunophenotype but rather had an abnormal myeloid blast population with aberrant lymphoid antigen expression and accounted for 0.3% and then 1.8% of total white cells, respectively (bottom two rows, 3rd and 4th columns, arrows)
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4201902&req=5

Figure 2: Evidence for NPM1-mutation heterogeneity in clinical samples(A) Multiple alignment of presentation (index) and variant sub-clone NPM1 mutations detected in Patients-1 and -2. Bold indicates NPM1 stop codon. (B) Detection of sub-clones in Patient-1 showing decline of the index clone and subsequent rise of secondary clone (Type A mutation). (C) Detection of sub-clones in Patient-2, showing development and low-level persistence of secondary Type A mutation. Heterogeneity is also demonstrated in the index case by the presence of a Type R allele. (D). Flow cytometry immunophenotype for Patient-1 at time-points: 0 (index), top row, day 1162, middle row, and then one additional time-point, approximately 2.5 months after day 1162, bottom row. The original index clone was an expanded monoblast population, (top row, 2nd column, asterisk) without a substantial myeloid blast population component (top row, 3rd and 4th columns). The alternate clone in subsequent samples (bottom 2 rows) no longer had a monoblastic immunophenotype but rather had an abnormal myeloid blast population with aberrant lymphoid antigen expression and accounted for 0.3% and then 1.8% of total white cells, respectively (bottom two rows, 3rd and 4th columns, arrows)
Mentions: In Patient-1, an alternate clone was clearly identified. In the index sample, only one NPM1 mutation (NM_002520.6:c.[864G>C;869delinsTCCTA]) was detected. However, in four subsequent samples spanning 38 months, next-generation sequencing detected increasing levels of the Type A allele that were quantified significantly above the empiric error frequency (>3.5 standard deviations), with a concurrent decrease of the index allele (Figure 2A, B). At days 1127 and 1162, the read counts of Type A allele surpassed those of the index allele. In Patient-2, we similarly detected heterogeneity (Figure 2A, C), harboring a Type B index mutation (NM_002520.6:c.863_864insCATG). A low-prevalence, but significant level (>900 standard deviations above the empiric error frequency) of Type R allele (NM_002520.6:c.863_864insTATG) was found in the index case, but became undetectable after initialization of treatment. By day 37, a low-prevalence Type A clone was identified at significant levels (>25 standard deviations above the empiric error frequency), which became dominant at day 110. At ∼1 year, the amount of Type A allele was surpassed by the index clone, but nevertheless persisted.

Bottom Line: Next-generation sequencing was precise and semiquantitative over four orders of magnitude.Further, in one-third of patients, sequencing detected alternate NPM1 mutations in addition to the patient's index mutation, consistent with tumor heterogeneity.This approach may complement current technologies to enhance patient-specific clinical decision-making.

View Article: PubMed Central - PubMed

Affiliation: 1] Departments of Laboratory Medicine, University of Washington, UW Hematopathology Laboratory at SCCA, Seattle, WA, USA [2] Genome Sciences, University of Washington, Seattle, WA, USA.

ABSTRACT
Detection of minimal residual disease predicts adverse outcome in patients with acute myeloid leukemia. Currently, minimal residual disease may be detected by RQ-PCR or flow cytometry, both of which have practical and diagnostic limitations. Here, we describe a next-generation sequencing assay for minimal residual disease detection in NPM1-mutated acute myeloid leukemia, which encompasses ∼60% of patients with normal karyotype acute myeloid leukemia. Exon 12 of NPM1 was PCR amplified using sequencing adaptor-linked primers and deep sequenced to enable detection of low-prevalence, acute myeloid leukemia-specific activating mutations. We benchmarked our results against flow cytometry, the standard of care for acute myeloid leukemia minimal residual disease diagnosis at our institution. The performance of both approaches was evaluated using defined dilutions of an NPM1 mutation-positive cell line and longitudinal clinical samples from acute myeloid leukemia patients. Using defined control material, we found this assay sensitive to approximately 0.001% mutant cells, outperforming flow cytometry by an order of magnitude. Next-generation sequencing was precise and semiquantitative over four orders of magnitude. In 22 longitudinal samples from six acute myeloid leukemia patients, next-generation sequencing detected minimal residual disease in all samples deemed negative by flow cytometry. Further, in one-third of patients, sequencing detected alternate NPM1 mutations in addition to the patient's index mutation, consistent with tumor heterogeneity. Next-generation sequencing provides information without prior knowledge of NPM1 mutation subtype or validation of allele-specific probes as required for RQ-PCR assays, and without generation and interpretation of complex multidimensional flow cytometry data. This approach may complement current technologies to enhance patient-specific clinical decision-making.

Show MeSH
Related in: MedlinePlus