6pLOH as a marker of immune-mediated bone marrow failure: Insights from a large single center experience

Abstract Copy-neutral loss of heterozygosity on chromosome 6p (6pLOH) is a recurrent acquired event in aplastic anemia (AA), detectable by SNP-array but not by conventional karyotyping. It includes the HLA locus and enables immune escape of hematopoietic stem cells lacking HLA class I, protecting them from T-cell–mediated attack. Therefore, 6pLOH, akin to PNH clone, is considered a molecular hallmark of immune-mediated bone marrow failure. We retrospectively analylzed patients with 6pLOH detected by SNP-array from a cohort diagnosed between Aug 1998 and Jun 2025 at a London tertiary center with AA, hypoplastic MDS (hMDS), inherited BMF syndromes (iBMF: telomeropathies, Fanconi anemia), PRCA, or acquired amegakaryocytic thrombocytopenia. We collected clinical, hematologic, cytogenetic, molecular (NGS-based myeloid gene panel, MGP), treatment and outcome data. Among 759 patients (median follow-up 38 months, IQR 13-105), 32 (4%) had 6pLOH at diagnosis or follow-up, 10 of which with failed or normal conventional karyotype. Median age was 43 years (10–77), with equal sex distribution. Most had AA (31/32, 97%): 13 (41%) non-severe, 17 (55%) severe, and 1 very severe; 1 had PRCA. No iBMF cases were observed. Autoimmune disorders were present in 7 (22%). Baseline median hematological parameters were Hb 90 g/L (60–140), reticulocytes 61×10⁹/L (10–204), platelets 24×109/L (2–185), neutrophils 1×109/L (0–3). In 25 patients (78%) 6pLOH was isolated; 7 had additional lesions, including monosomy 7, del7p, del13q, and CN-LOH 17p. MGP, available in 30, identified somatic mutations in 6 (20%): BCOR, CUX1, ETV6, DNMT3A, RUNX1+UBA1, DNMT3A+TP53, with a median VAF of 20% (6–46). A PNH clone was found in 18/32 (56%), median size 1% (0.2–12%). Twenty-nine received treatment (median 1 line, range 1–5); 20 received horse ATG + cyclosporine, with 13 (65%) responses; 3 underwent transplant. Seven patients (22%) progressed to MDS during follow-up. Among these, three harbored additional SNP-A lesions, including del(13q), monosomy 7, and CN-LOH 17p; the latter two also carried somatic mutations identified by MGP (TP53 + DNMT3A and RUNX1 + UBA1, respectively). One further patient who progressed had DNMT3A mutation. At last follow-up (median 24 months), 2 patients had died (6%). Compared to the remaining 727 patients, the 6pLOH group had a higher frequency of AA diagnosis (90% vs 72%, p=0.01), higher reticulocyte count (61 vs 36×10⁹/L, p=0.028), and more frequent PNH clones (56% vs 39%, p=0.04), though of smaller size (1, 0.2–12 vs 2, 0.1–99%, p=0.03). ATG-based treatment was more common (65% vs 50%, ns) and associated with higher, though non-significant response rate (65% vs 53%), while fewer patients proceeded to transplant (10% vs 24%, p=0.03). MDS/AML evolution was higher in the 6pLOH group (9 vs 21%, p=0.01), but mortality was significantly lower (6% vs 21%, p=0.01). When comparing 6pLOH patients to those with normal karyotype (n=441) results mirrored those from the overall cohort except for no difference in frequency of PNH clone. We finally compared 6pLOH patients to those with other cytogenetic abnormalities (n=134). The most common were chromosome 7 lesions (monosomy 7, del7p, CN-LOH 7p, i(7p), n=28), trisomy 8 (n=19), del13q (n=10), delY (n=9), 17p alterations (n=5), and del20q/trisomy 6 (n=4 each). Six patients had complex karyotypes (>3 abnormalities). Patients with other abnormalities more often had hMDS (19% vs 6%, p=0.04), showed a trend towards more frequent somatic mutations (38% vs 20%, p=0.07), and less frequently presented a PNH clone (41% vs 56%, p=0.08). MDS/AML progression occurred in 41/134 (30%), often in association with chromosome 7 (16/28), del20q (3/4), del13q (5/10), and 17p abnormalities (2/5). Mortality was significantly higher compared to the 6pLOH group (42% vs 6%, p=0.0002). 6pLOH is rare in bone marrow failure but enriched in acquired immune AA. It is typically isolated, and frequently associates with small PNH clones, supporting its role in immune-driven pathogenesis. Patients with 6pLOH respond well to ATG-based therapy, and show infrequent MDS progression, which is often linked to additional cyotgenetic or molecular abnormalities. Compared to other abnormalities, 6pLOH patients show lower mutational burden and more favorable outcomes. These findings support 6pLOH as a distinct clinical and biological subset of BMF with an immune signature, highlighting its value as a diagnostic and predictive marker in this setting.