Distinguishing Germline from Acquired Pathogenic CSF3R Variants: Key Diagnostic Features and Clinical Implications

Background

Neutrophilia is a common clinical finding. Conversely, chronic neutrophilic leukemia (CNL) is a rare myeloproliferative neoplasm associated with mutations in colony stimulating factor receptor (CSF3R), classically CSF3R T618I (Maxson, 2013). Increased genomic testing has yielded higher rates of identification of CSF3R mutations. Novel germline variants have been described, some predisposing to hematologic malignancies (Trottier, 2020), though data on the clinical significance of non-canonical CSF3R subtypes, including germline (G-CSF3R), are limited. We aimed to characterize phenotypes and outcomes in a CSF3R-mutated series to expose the distinct clinical impact of individual variants.

Methods

This multicenter retrospective/prospective study evaluated patients with a CSF3R mutation identified between 2019-2022 (centralized laboratory in Quebec). Testing was performed using Sanger sequencing or NGS (42-gene myeloid panel). Clinical data was abstracted from patient records (IRB approval). Conventional statistical methods were used (JMP® Pro 14.1.0; SAS Institute, NC, USA).

Results

Between 2019 and 2022, 35 of 323 (10.8%) patients who underwent CSF3R mutational testing resulted positive. Of those, n=17 had informative laboratory and clinical data available. The cohort included: T618I (n=8; 47%), E835K (4; 23%), G683R (2; 12%) and M696T (3; 18%). Of the 9 non-T618I mutations, 6 were identified as germline via buccal swab sampling, and an additional 3 were strongly suspected germline on account of VAF exactly 50% with long-standing neutrophilia. Reasons for testing: G-CSF3R: 33% neutrophilia; 67% thrombocytosis or incidental; T618I: 100% neutrophilia and/or other. Seven of 8 patients with T618I mutations (88%) had an ICC-defined diagnosis of CNL, compared to none of those with G-CSF3R (0%).

Phenotypes: G-CSF3R vs CSF3R-T618I

Subjects with G-CSF3R mutations vs T618I presented a distinct phenotype: absence of splenomegaly (0 vs 50%; p=0.006) and constitutional symptoms at diagnosis (0 vs 75%; p=0.0003). They also had a significantly lower median neutrophil count (10.1 vs 26.7 x 10⁹/L; p=0.002) and lower neutrophil-to-lymphocyte ratio (4.6 vs 16.4; p=0.009); categorically none had leukocytes > 20 x 10⁹/L (0 vs 63%; p=0.002). In addition, patients with G-CSF3R had lower median ferritin levels (54 vs 533 ug/L; p=0.001) and significantly fewer presented abnormal LDH levels compared to T618I-positive counterparts (11 vs 63%; p=0.02). Median VAF in the G-CSF3R cohort was strictly 50% (vs 46% in CSF3R-T618I; p=0.03). There were no significant differences in hemoglobin (median 138 vs 120 g/L; p=0.16) or platelet counts (median 304 vs 319 x 10⁹/L; p= 0.7) respectively, between G-CSF3R vs T618I groups.

Clinical trajectories and management: G-CSF3R vs CSF3R-T618I

Over a median follow-up of 32.2 months (range 12-64), none of the G-CSF3R-mutated cases presented thromboembolic events or progressed to acute leukemia or another hematologic malignancy (vs 1 each in the CSF3R-T618I group; p=0.21). Interestingly, none of the 17 patients had a hemorrhagic event. Regarding management, none of the G-CSF3R-mutated cases required any treatment; all were managed expectantly, vs 67%/56% of CSF3R-T618I cases necessitating 1/2 lines of therapy, respectively (p=0.001). Notably, Kaplan-Meier survival analysis disclosed significantly more favorable overall survival in G-CSF3R vs T618I patients, with median not reached vs 32.2 months, respectively (p=0.04).

Conclusion

This is one of the largest CSF3R-mutated cohort series worldwide. An unexpectedly high proportion of tested patients resulted CSF3R-positive, contrasting with the rarity of CNL, suggesting not all mutated cases harbor this disease. Previous work has demonstrated aggressive clinical course in CSF3R-T618I patients. Reciprocally, we thoroughly characterized patients with non-T618I CSF3R germline/suspected germline variants, who presented an appreciably more indolent phenotype and favorable outcomes. Clinicians should suspect a germline mutation in appropriate clinical context (family history, absence of splenomegaly/symptoms, lower neutrophil count, normal LDH) and confirm with germline tissue sampling. Although germline CSF3R variants appear to confer a more indolent clinical profile, their long-term impact is undetermined; we therefore recommend long-term monitoring in these patients.