Not so exclusive: Co-mutations in JAK2, MPL and CALR define distinct hematologic and clonal signatures

Abstract

Abstract Background: The canonical driver mutations in JAK2, CALR, and MPL are typically regarded as mutually exclusive in myeloproliferative neoplasms (MPNs). However, co-occurrence is increasingly recognized, raising questions about their clinical and clonal implications. Such co-mutations may contribute to disease heterogeneity and influence hematologic phenotype, clonal fitness, and progression risk. The extent to which these co-occurring drivers shape disease biology and evolution remains unclear and warrants systematic investigation. Methods: We retrospectively identified patients with co-mutations involving any two of the canonical MPN drivers (JAK2, CALR, and MPL)—including biallelic or dual mutations within the same gene (e.g., JAK2–JAK2, CALR–CALR, MPL–MPL)—from a precision molecular hematopathology database. All available NGS and clinical data were systematically reviewed. Serial next-generation sequencing (NGS) was available for a subset of patients to evaluate clonal evolution and infer clonal competition. Given the limited sample size, analyses were primarily descriptive; hematologic parameters were compared across mutational subgroups using the Kruskal–Wallis test. Results: A total of 32 patients were included (median age 68.5 years, range 28–95; 63.5% male). The median follow-up was 28.8 months (range 1.6–377.9), during which 11 experienced disease progression and 10 patients died. Diagnoses included primary myelofibrosis (PMF, 34.4%), essential thrombocythemia (ET, 25.0%), MDS/MPN overlap syndromes (21.9%), MDS (9.4%), AML (6.3%), and polycythemia vera (3.1%). The most frequent co-mutation patterns were JAK2-MPL (37.5%), followed by JAK2–CALR (21.9%), JAK2–JAK2 (18.8%), CALR–CALR (15.6%), CALR–MPL (3.1%), and MPL–MPL (3.1%). The ET cohort was dominated by JAK2–JAK2 (50%), whereas PMF showed greater mutational diversity. JAK2 V617F was detected in 78.1% of patients (median VAF 0.194, range 0.03–0.844). Among these, 48.0% had co-mutations in MPL (median VAF 0.182), and 28.0% in CALR (median VAF 0.362). Additionally, 24.0% harbored non-V617F JAK2 variants (median VAF 0.159). In CALR–MPL co-mutant cases, CALR typically represents the dominant clone (median VAF: 0.36), whereas MPL exhibits a markedly lower median VAF (≈ 0.13), consistent with a clearly subclonal architecture relative to CALR. Hematologic profiles varied by genotype. In the MDS/MPN group, JAK2–MPL co-mutations were associated with higher hemoglobin (median 11.2 g/dL) and platelet counts (995 ×10⁹/L) compared to other patterns (p=0.05). Among PMF cases, the JAK2–CALR group exhibited a hyperproliferative phenotype with the highest median WBC (25.6 × 10⁹/L), ANC (20.6 × 10⁹/L), and hemoglobin (14.1 g/dL). The CALR–CALR group had elevated WBC (21.0 × 10⁹/L) but the lowest hemoglobin (7.8 g/dL), while JAK2–JAK2 cases showed the lowest WBC (3.3 × 10⁹/L) and platelet count (170 × 10⁹/L). MPL–MPL patients had intermediate blood counts but the highest monocytes (1.12 × 10⁹/L), with platelet counts trending lower than other genotypes. Serial NGS was available for 9 patients, revealing three patterns of clonal evolution: (1) stable clonal dominance without driver switching, (2) competitive takeover, and (3) parallel co-expansion. Seven patients (77.8%) maintained the same top driver, while two (22.2%) exhibited driver competition—CALR → JAK2 and JAK2 → MPL—though neither experienced disease progression. Among six patients with stable clonal expansion, three progressed. Gene-specific trends showed that JAK2 rose in three (including one overtaking CALR), remained stable in one, and declined in two; CALR expanded in two, remained stable in one, and declined in three; MPL rose in two, fell in two, and remained flat in one—suggesting a context-dependent trajectory. Conclusions: Co-mutations among JAK2, CALR, and MPL are more frequent than previously recognized and associate with distinct hematologic phenotypes. Clonal evolution patterns reveal that JAK2 often asserts late dominance, CALR remains relatively stable, and MPL displays context-dependent behavior. Notably, stable clonal expansion—rather than driver switching—was more frequently linked to disease progression, suggesting that clonal stability may reflect greater biological fitness. These findings underscore the importance of longitudinal molecular profiling in understanding clonal dynamics and refining risk assessment in myeloid neoplasms.