A Somatic Variant at the Edge of Detection in a Myeloid Neoplasm
A 67-year-old man with pancytopenia and 4% blasts in the marrow undergoes a 40-gene myeloid next-generation sequencing panel on bone marrow. The report identifies a JAK2 V617F at 1.8% VAF (alternate reads 9 of 512; limit of detection 2% VAF at ≥500x) and a DNMT3A R882H at 31% VAF. The JAK2 result is below the validated LOD. The hematologist asks whether to treat for a myeloproliferative neoplasm. A repeat marrow 3 months later shows JAK2 V617F at 3.1% VAF with rising platelets.
1. How should the sub-LOD JAK2 result be interpreted, and what does the follow-up result add?
Read the worked answer
A JAK2 V617F at 1.8% VAF with only 9 alternate reads is below the assay's validated LOD and near the background error floor; it should be reported as an indeterminate/low-level detection with a caveat, not as a definitive positive. The follow-up sample at 3.1% VAF, above the LOD, with rising platelets, provides confirmatory evidence of a clonal JAK2-mutant population. The trajectory (increasing VAF and rising platelet count) supports a myeloproliferative neoplasm-most consistent with essential thrombocythemia or early primary myelofibrosis depending on marrow morphology. The teaching point is that low-level calls require orthogonal confirmation (e.g., a more sensitive assay, droplet digital PCR, or repeat testing), and that serial VAF trends, interpreted with clinical and morphologic data, carry more weight than a single borderline result.
2. What analytical factors could produce a false-positive low-level JAK2 call, and how would you confirm it?
Read the worked answer
Possible causes: (1) sequencing/PCR error and index hopping, especially at low alternate read counts; (2) FFPE or storage-related artifacts (less likely in fresh marrow); (3) contamination or carryover from a prior positive sample; (4) misalignment in a homologous region; (5) true low-level clonal hematopoiesis. Confirmation strategy: repeat on an independent draw with a sensitive orthogonal method such as allele-specific qPCR or droplet digital PCR, which can reliably quantify below 1% VAF; review read-level metrics (strand bias, mapping quality, base quality, position within read); and check for contamination using a sample-specific identifier or negative controls. Only after orthogonal confirmation should a sub-LOD result be treated as actionable.
3. How do the DNMT3A and JAK2 findings interact in interpretation?
Read the worked answer
DNMT3A R882H at 31% VAF is a common age-related clonal hematopoiesis (CHIP) variant and, by itself, is not diagnostic of a myeloid neoplasm. Its presence indicates an expanded clone, but CHIP is common in older adults and does not establish malignancy. JAK2 V617F, in contrast, is a disease-defining driver for myeloproliferative neoplasms. The combination-a high-VAF DNMT3A clone plus a lower-VAF JAK2 clone-suggests clonal evolution with a JAK2 subclone, or two independent clones. Interpretation should integrate marrow morphology, blood counts, and serial VAFs. The DNMT3A finding should not be over-interpreted as diagnostic; the JAK2 result, once confirmed, is the actionable driver. This distinction matters for avoiding both under-treatment of a true MPN and over-treatment of CHIP.