AUTHORS
Katharine Moosic, PhD, Tess M Deddens2, Omar Elghawy, MD3, Timothy William Gaul4, Elizabeth Dudley5, Prashant Chaulagain, MBBS6, Matthew Schmachtenberg7, Bryna Shemo, MA8, Kusuma Ananth, PhD9, Su-Fern Tan, PhD10, Aakrosh Ratan, PhD11, Felipe Andrade, MD, PhD12, Erikah Darrah, PhD13, David J Feith, PhD14, Thomas P Loughran, MD15
AFFILIATIONS
1 University of Virginia, Charlottesville, Virginia, United States
2 University of Virginia, Charlottesville, Virginia, United States
3 Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania, United States
4 University of Virginia, Charlottesville, Virginia, United States
5 University of Virginia, Charlottesville, Virginia, United States
6 University of Virginia, Charlottesville, Virginia, United States
7 University of Virginia, Charlottesville, Virginia, United States
8 University of Virginia, Charlottesville, Virginia, United States
9 The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
10 University of Virginia, Charlottesville, Virginia, United States
11 University of Virginia School of Medicine, Charlottesville, Virginia, United States, University of Virginia, Charlottesville, United States
12 The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
13 The Johns Hopkins University School of Medicine, Baltimore, Maryland, United States
14 University of Virginia School of Medicine, Charlottesville, Virginia, United States, University of Virginia, Charlottesville, Virginia, United States
15 University of Virginia School of Medicine, Charlottesville, Virginia, United States, University of Virginia, Charlottesville, Virginia, United States
ABSTRACT
T-Large granular lymphocyte leukemia (LGLL) is a rare chronic hematologic malignancy characterized by clonal CD8+ T cell expansion. Patients exhibit frequent autoimmune manifestations like rheumatoid arthritis (RA), cytopenias like anemia and neutropenia, and somatic activating STAT3 mutations. These STAT3 mutations frequently correlate with clinical features and favorable therapeutic response. Patients with STAT3 mutations are also more likely to have RA, however, mutation rates within this subpopulation of LGLL have not been well-characterized. Droplet digital (dd)PCR is a relatively new technology that can be harnessed to detect low-level STAT3 mutations (approximate limit of detection 0.007% variant allele frequency, VAF) compared to other sequencing options previously utilized in the context of LGLL.
Due to the frequency of STAT3 mutations and their association with clinical features in LGLL, we hypothesized that ddPCR analysis of common STAT3 mutations would find previously undetected mutations in LGLL and LGLL/RA patients and that these low-level mutations would correlate to clinical features.
Peripheral blood mononuclear cells (PBMCs), clinical parameters, and demographic data were collected from 93 LGLL patients and 90 LGLL patients with concomitant RA (LGLL/RA) enrolled in the University of Virginia LGL Leukemia Registry. PBMCs were also collected from 42 age-, race-, and sex-matched healthy donors (HD). DNA was isolated from patient PBMCs and analyzed for STAT3 mutations by both Sanger sequencing and ddPCR. The ddPCR assays were targeted to the most common mutations in STAT3, Y640F and D661Y, whereas Sanger sequencing detected any mutation in the STAT3 SH2 domain that exceeded ~15% VAF. Samples were then divided into STAT3 wild-type (WT) and mutant categories, and clinical parameters were correlated to mutational status.
The ddPCR assay detected many previously un-annotated low VAF (<15% VAF) STAT3 mutations in LGLL patient PBMCs, with 55/93 (59%) exhibiting Y640F and D661Y STAT3 mutations compared to 24/76 (32%) in those patients with Sanger sequencing available. 14% of LGLL patients (13/93) exhibited double mutations (concomitant Y640F and D661Y mutation), only one of which was identified by Sanger sequencing. A significantly higher rate of STAT3 mutation (75/90, 83%, p = 0.001) as well as double mutation (40/90, 44%, p = 0.003) was detected in the LGLL/RA population compared to those with LGLL alone. STAT3 mutations were also highly specific to LGLL, with only 2/42 HD samples exhibiting STAT3 mutations (both <0.04% VAF) by ddPCR. Of note, the presence of STAT3 mutation in the LGLL cohort was significantly associated with cytopenias including neutropenia (WT 43%, STAT3 mut 67%, p = 0.036) and anemia (WT 41%, STAT3 mut 72%, p = 0.002). Only the anemia-STAT3 association preserved when analysis was performed dividing solely by Y640F and D661Y mutations detected by Sanger sequencing (WT 48%, STAT3 mut 83%, p = 0.004). Logistic regression models show an increased probability of either neutropenia or anemia with increasing VAF. Similarly, ANCs were significantly lower in LGLL patients with STAT3 mutation vs. WT (median: STAT3 mut 1054 vs WT 1880 cells/µL, p = 0.009), and red cell distribution width (RDW) was significantly increased in patients with STAT3 mutation vs. WT (STAT3 mut 14.9% vs WT 13.8% p = 0.01). Both hemoglobin (Hgb) and hematocrit (Hct) trended lower in patients with STAT3 mutations (Hgb: STAT3 mut 11.8 g/dL vs WT 13.1 g/dL, p = 0.1; Hct: STAT3 mut 35.6% vs WT 39.45%, p = 0.06), although neither reached statistical significance. None of these parameters were significantly different when STAT3 mutant samples were divided by Sanger sequencing status.
This study demonstrates that ddPCR more effectively detects STAT3 mutations in LGLL compared to Sanger sequencing, and correlation with cytopenias and numerous clinical blood counts is revealed upon inclusion of these low VAF mutations. The discovery of increased prevalence of STAT3 mutations, especially double mutations in LGLL/RA populations, suggests differential disease pathogenesis with potential implication for additional study of STAT3 therapeutic targeting in this disease subset. The correlation of STAT3 with many clinical parameters only when low VAF mutations are included suggests that even low-level mutations have a clinical impact, opening up a much larger population for JAK/STAT targeted therapies in the future.
DISCLOSURES
Katharine Moosic, PhD: No relevant conflicts of interest to declare.