AUTHORS
Min Wu, MD, PhD, Han Yan2, Yongmei Zhu3, Gang Cai4, Jianqing Mi, MD,PhD5, Huijin Zhao6
AFFILIATIONS
1 Shanghai Institute of Hematology, State Key Laboratory for Medical Genomics, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, China
2 Shanghai Institute of Hematology, State Key Laboratory for Medical Genomics, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, China
3 Shanghai Institute of Hematology, State Key Laboratory for Medical Genomics, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, China
4 Department of Laboratory Medicine, Ruijin Hospital, Shanghai JiaoTong University School of Medicine, Shanghai, China, Shanghai, China
5 Shanghai Institute of Hematology, State Key Laboratory for Medical Genomics, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, China
6 Shanghai Institute of Hematology, State Key Laboratory for Medical Genomics, Ruijin Hospital,Shanghai Jiao Tong University School of Medicine, Shanghai, China
ABSTRACT
Background
Somatic PIGA variants underlie the glycosylphosphatidylinositol-anchor deficiency characteristic of paroxysmal nocturnal hemoglobinuria (PNH). Although peripheral blood (PB) flow cytometry remains the reference method for detecting and quantifying PNH clones, PIGA variants are increasingly identified during routine bone marrow (BM) next-generation sequencing (NGS) performed for diverse hematologic disorders. Their clinical significance, relationship with flow-cytometric PNH clone size, and distribution between BM and PB remain incompletely defined. We characterized the pan-hematologic spectrum and molecular context of PIGA variants, evaluated the relationship between PIGA variant allele frequency (VAF) and PNH clone size, and explored the feasibility of PB-targeted sequencing using paired BM and PB samples.
Methods
We retrospectively reviewed 5,200 consecutive BM NGS tests performed at our center between October 2020 and June 2026. Forty-four patients harboring PIGA variants were identified. Clinical diagnoses, PNH classifications, flow-cytometric clone sizes, PIGA variant profiles, VAFs, concurrent somatic variants, and treatment data were systematically reviewed. Maximum single-variant VAF and cumulative PIGA VAF were calculated at the patient level. Their correlations with granulocyte and monocyte PNH clone sizes were assessed using Spearman rank correlation. A sensitivity analysis was performed in patients with approximately contemporaneous BM sequencing and flow-cytometric assessment. Four patients with paired BM and PB targeted deep sequencing were analyzed separately.
Results
PIGA variants were identified in 44 of 5,200 BM NGS tests (0.85%). The clinical spectrum was heterogeneous: 28 patients (63.6%) had classical PNH, aplastic anemia, or otherwise unclassified cytopenias, whereas 11 (25.0%) had primary hematologic malignancies, including acute myeloid leukemia, acute lymphoblastic leukemia, and myelodysplastic neoplasms. The remaining 5 patients had other hematologic disorders. Concurrent non-PIGA somatic variants, including variants involving DNMT3A, ASXL1, RUNX1, and other recurrent hematologic genes, were identified in 25 patients (56.8%), indicating that PIGA variants may occur within genetically complex hematopoietic clones. Twenty-three patients were clinically classified as classical PNH, bone marrow failure-associated PNH, or subclinical PNH and had evaluable granulocyte flow-cytometric data. Fourteen patients had a single PIGA variant and 9 had multiple variants. Both maximum PIGA VAF and cumulative PIGA VAF correlated positively with granulocyte PNH clone size (Spearman ρ=0.555, P=0.006 and ρ=0.593, P=0.003, respectively). In a sensitivity analysis restricted to 16 patients with approximately contemporaneous sequencing and flow-cytometric assessment, the correlations were stronger (ρ=0.609, P=0.012 for maximum VAF and ρ=0.694, P=0.003 for cumulative VAF). Similar correlations were observed with monocyte PNH clone size (ρ=0.533, P=0.009 and ρ=0.569, P=0.005, respectively). In four patients with paired BM and PB sequencing, at least one identical dominant PIGA variant was detected in both compartments. The median patient-level maximum VAF was 30.87% in BM (range, 14.82%-41.50%) and 9.36% in PB (range, 5.86%-27.04%), with a higher BM VAF in all four cases. At the variant level, 5 of 6 BM PIGA variants were detected in PB; the only discordant variant was a minor BM subclone with a VAF of 0.95%.
Conclusions
PIGA variants detected during routine BM NGS occur across a broad spectrum of hematologic conditions and are not, in isolation, diagnostic of PNH. In clinically established PNH, PIGA mutational burden correlated with granulocyte and monocyte clone sizes, with cumulative VAF showing a numerically stronger association than maximum single-variant VAF. Exploratory paired sequencing demonstrated that PB NGS consistently identified dominant PIGA variants but yielded lower VAFs and missed one minor BM subclone. These findings support integrating incidental PIGA variants with clinical phenotype and high-sensitivity flow cytometry. PB-targeted sequencing may serve as a minimally invasive molecular adjunct when repeated BM sampling is impractical, but it should not replace flow cytometry or BM assessment. Larger longitudinal studies are required to establish its sensitivity and utility for clonal monitoring.
DISCLOSURES
Min Wu, MD, PhD: No relevant conflicts of interest to declare.