NGS and emerging molecular scenarios in chronic myeloproliferative neoplasms: a case of double mutated polycythemia vera

Abstract

Myeloproliferative neoplasms (MPNs) are clonal hematopoietic disorders driven by mutually exclusive JAK2, CALR, MPL mutations. NGS has revealed rare cases with coexisting driver mutations, challenging classical diagnostic paradigms, highlighting clonal complexity. We describe a 71-year-old man with JAK2-positive polycythemia vera (PV) whose hematologic parameters remained suboptimally controlled under hydroxycarbamide treatment. Concern for disease progression prompted NGS analysis, surprisingly revealing a dominant CALR type 2 mutation alongside non-driver DNMT3A variants, while the JAK2 V617F clone was below the NGS limit of detection. Retrospective analysis confirmed dynamic clonal competition between the dominant CALR clone and the minor low VAF JAK2 clone, whereas histopathology remained consistent with PV, supporting an integrated diagnosis of MPN–NOS. This case demonstrates that additional mutated clones can significantly influence phenotype and disease evolution, highlighting the limitations of stepwise molecular testing. Comprehensive profiling of clonal architecture is essential to refine diagnosis and prognosis, in the NGS era.

Introduction

Myeloproliferative Neoplasms (MPNs) are clonal diseases originating from a single hematopoietic stem cell characterized by excessive production of mature blood cells.

According to the 2022 World Health Organization (WHO) and International Consensus Classification (ICC) criteria, the MPNs group encompasses three main entities: Polycythemia Vera (PV), Essential Thrombocythemia (ET), and Primary Myelofibrosis (PMF) [1, 2].

Based on molecular characterization, the onset of MPNs is driven and promoted by acquired gain-of-function mutations in three commonly recognized driver genes: JAK2 (Janus Kinase 2), MPL (MPL proto-oncogene thrombopoietin receptor) and CALR (Calreticulin) genes.

The JAK2 V617F mutation is detected in > 95% of patients with PV and in approximately 50% of those with ET or PMF. In contrast, CALR and MPL mutations define distinct molecular subsets of ET and PMF but are rarely encountered in PV.

The JAK2 V617F assay represents the first step in molecular diagnosis. Since driver mutations are traditionally considered mutually exclusive, JAK2 V617F positive patients are usually excluded from CALR and MPL mutation analysis in clinical practice.

In recent years, however, the MPNs mutational landscape has become increasingly complex. The identification of additional mutations in the myeloid context has expanded our understanding of the molecular mechanisms underlying the disease heterogeneity, and the use of multigene sequencing panels in next-generation sequencing (NGS) era has revealed a growing number of patients carrying double driver mutations [3]. Although such co-occurrence was initially considered unlikely, it has now been documented, and preliminary observations suggest that these combinations may be associated with distinct clinical phenotypes and differing prognostic implications.

In this context, we report a case of PV characterized by JAK2 V617F mutation coexisting with a CALR type 2 variant, revealing the presence of two independent clones both probably contributing to the phenotypic expression.

Case description

A 71-year-old man with a history of ischemic heart disease, previously treated with angioplasty, stent placement, and subsequent coronary bypass surgery, was referred to our hematology clinic in 2022 because of leukocytosis with associated erythrocytosis and thrombocytosis identified during routine laboratory testing.

At the first hematology consultation, the patient showed mildly elevated white blood cell count (WBC 14.710/uL) and thrombocytosis (PLT 1.106.000/uL), with increased hemoglobin (Hb 17.4 g/dL) and Hematocrit (Hct 54%). Serum erythropoietin levels were found to be low-normal; LDH and spleen dimensions were within the normal range.

Given the clinical suspicion of MPNs onset, a qualitative assessment of JAK2 V617F variant was performed on peripheral blood (PB) sample using the Ipsogen JAK2 MutaScreen assay, which yielded a positive result.

Concurrently, bone marrow (BM) trephine biopsy revealed histopathological features of age-adjusted hypercellularity, panmyelosis with erythroid islands expansion and polymorphic megakaryocytes forming loose clusters; blast count by CD34 IHC was assessed as 4% and reticulin stain showed mild and partial bone marrow fibrosis (MF-1) (Fig. 1A, B, C). Those findings were deemed consistent with a JAK2 mutated PV.

Fig. 1.

First bone marrow trephine biopsy: (A) Hematoxylin-Eosin stain shows hypercellular bone marrow adjusted for age, with panmyelosis. Erythroid islands appear expanded and all lineages show terminal differentiation. (B) CD61 stain shows sharp increase in megakaryocytes, which appear polymorphic and have a moderate tendency to form loose clusters. No significant atypia is noted. (C) Reticulin stain shows a slight and partial increase in marrow fibrosis (MF-1). Second bone marrow trephine biopsy: D) Hematoxylin-Eosin stain still shows hypercellular bone marrow adjusted for age, with panmyelosis. Erythroid islands appear slightly less prominent than before, probably as a therapy-related effect. E) CD61 stain shows a moderate decrease in megakaryocytic population, with polymorphic appearance and occasional loose clusters formation. F) Bone marrow fibrosis is still low-grade, albeit slightly increased compared to previous sample

Based on JAK2 V617F positivity and BM biopsy, PV was diagnosed and hydroxycarbamide therapy was started.

Although the patient initiated cytoreductive therapy, hematocrit and white blood cell count remained suboptimally controlled, necessitating a gradual escalation of the hydroxycarbamide dose. Given concern for possible myelofibrotic transformation or the emergence of additional somatic mutations during the disease course, a histologic and molecular re-evaluation was undertaken with a new BM biopsy and targeted NGS (t-NGS) analysis.

t-NGS was performed on DNA extracted from BM cells obtained via Buffy-Coat procedure. The Sophia DDM™ Dx Myeloid Solution panel (Sophia Genetics, Saint Sulpice, Switzerland) was used to screen 30 genes commonly mutated in myeloid malignancies, applying a variant allele frequency (VAF) cut-off of 1% and a minimum coverage of 1000x; however, longitudinal monitoring of the JAK2 V617F mutation included manual interrogation of low-confidence variant calls (VAF < 1%) to optimize analytical sensitivity. All procedures followed the manufacturer’s protocols. The comprehensive gene panel is detailed at [https://www.arrowdiagnostics.it/download/all/MYS_CEIVD_FS_GL-GN-2200014-r2.pdf].

Surprisingly, the t-NGS results highlighted the presence of a classical CALR type 2 mutation and the absence of the JAK2 V617F variant. In addition, a variant in DNMT3A p.(Gln816*) and a single nucleotide polymorphism (SNP) in ZRSR2 were identified. Mutational profile details are reported in Table 1 and illustrated in the fish plot in Fig. 2. Based on these unexpected findings, retrospective t-NGS analysis of the diagnostic PV sample revealed a low-VAF JAK2 V617F mutation and confirmed the presence of CALR and DNMT3A mutations as well as the ZRSR2 SNP, verifying sample identity; additionally, a second low-VAF DNMT3A p.(Tyr735Cys) mutation was identified (Table 1; Fig. 2).

Table 1.

Targeted NGS results obtained from all analyzed time points, applying a variant allele frequency (VAF) cutoff of 1% and a minimum sequencing depth of 500×; Reference Genome: GRCh37/hg19

	Gene	Chr	Type of alteration	Alteration result	cDNA	Protein	Details	VAF (%)	Coverage	ACMG/AMP classification
PV onset (PB sample)	JAK2	9	SNV	Missense	c.1849G > T	p.(Val617Phe)	V617F	2.7	2854	P
	CALR	19	INDEL	Frameshift	c.1154_1155insTTGTC	p.(Lys385Asnfs*47)	K385Nfs*47	37.0	2941	P
	DNMT3A	2	SNV	Nonsense	c.2446 C > T	p.(Gln816*)	Q816*	33.3	2553	LP
	DNMT3A	2	SNV	Missense	c.2204 A > G	p.(Tyr735Cys)	Y735C	1.6	2776	LP
	ZRSR2	X	INDEL	Inframe_6	c.1338_1343dup	p.(Ser447_Arg448dup)	S447_R448dup	89.6	1254	LB
PV Reassessment (BM sample)	CALR	19	INDEL	Frameshift	c.1154_1155insTTGTC	p.(Lys385Asnfs*47)	K385Nfs*47	46.6	5621	P
	DNMT3A	2	SNV	Nonsense	c.2446 C > T	p.(Gln816*)	Q816*	46.3	5148	LP
	ZRSR2	X	INDEL	Inframe_6	c.1338_1343dup	p.(Ser447_Arg448dup)	S447_R448dup	97.7	2436	LB
PV last follow-up (PB sample)	CALR	19	INDEL	Frameshift	c.1154_1155insTTGTC	p.(Lys385Asnfs*47)	K385Nfs*47	41.3	1800	P
	DNMT3A	2	SNV	Nonsense	c.2446 C > T	p.(Gln816*)	Q816*	46.0	1707	LP
	ZRSR2	X	INDEL	Inframe_6	c.1338_1343dup	p.(Ser447_Arg448dup)	S447_R448dup	94.9	822	LB

PV Polycythemia Vera, PB Peripheral Blood, BM Bone Marrow, SNV Single Nucleotide Variant, Chr Chromosome, P: Pathogenic; LP Likely Pathogenic; VUS Variant of Uncertain Significance, LB Likely Benign (LB)

Fig. 2.

Fishplot depicting clonal architecture of the patient from polycythemia onset to hematologic re-evaluation. The plot illustrates the relative contribution of each clone over time. At diagnosis, a low-VAF JAK2 V617F and DNMT3A p.Tyr735Cys clone coexisted with a dominant CALR type 2 and DNMT3A p.Gln816* clone. At hematologic re-evaluation, the CALR/DNMT3A p.Gln816* clone remained dominant, while JAK2 V617F persisted at very low VAF, illustrating dynamic clonal evolution. The ZRSR2 p.Ser447_448dup SNP confirmed patient identity

To rule out a potential bias, a more sensitive quantitative assessment of JAK2 V617F was performed in both samples using the Ipsogen JAK2 MutaQuant assay. The allelic burden was 2.5241% at diagnosis and 0.2893% at hematologic re-evaluation; these results are consistent with t-NGS results since 0.2893% is lower than detection limit of t-NGS justifying its negativity (Fig. 2). The presence of CALR type 2 was also confirmed in both samples using the qualitative Ipsogen CALR RGQ PCR Kit CE Real-Time assay (Fig. 2).

To exclude differences between the diagnostic PB sample and the BM re-evaluation sample, the full molecular panel was repeated on a newly collected PB sample. Results were fully concordant with the BM analysis and aligned with published data showing no significant differences between PB and BM derived molecular findings in MPNs (Table 1).

Histologically, the re-evaluation of BM biopsy showed an overall similar picture compared to previous sample, with mild reduction of overall cellularity, erythroid islands and megakariocytes, probably as a result of therapy. Of note, erythroid islands were occasionally located near bone trabeculae. There was no increase in blasts or overt fibrosis compared to the prior biopsy, confirming blast at 4% and MF-1 (Fig. 1D, E, F).

The detection of a dominant CALR mutated clone alongside a minor JAK2 V617F population leads to an integrated diagnosis of MPN-NOS, highlighting the diagnostic challenges in double mutated cases. To exclude a misdiagnosis, the initial diagnostic bone marrow biopsy was reviewed. Despite the newly obtained molecular data, the histopathological features remained consistent with a PV phenotype.

Consideration and discussion

The introduction of NGS has greatly expanded our understanding of the MPNs molecular landscape, but it has also highlighted several “controversial” aspects of these diseases. Although the discovery that the three canonical driver mutations converge on the JAK-STAT signalling pathway partially explains how distinct mutations can give rise to similar disease phenotypes (e.g., ET or MF), the mechanisms by which MPNs arise in the absence of these mutations (triple-negative cases) remain incompletely understood [4].

Similarly, how the same molecular mutation can produce widely variable clinical phenotypes is still unclear, although the timing of mutation acquisition may play a critical role. The MPNs molecular complexity is mirrored by their clinical heterogeneity. These disorders represent a highly specific group in which distinct clinical and prognostic features coexist with overlapping characteristics, sometimes necessitating a compromise diagnosis, as in MPN-NOS, where the criteria for a classic MPNs are not fully met [5].

Advances in molecular characterization have enabled the identification of more specific disease subsets. For example, CALR mutated ET patients display a distinct phenotype compared with JAK2 mutated ET, including higher platelet counts, lower white cell counts, and a reduced thrombotic risk; consequently, CALR mutated ET is generally considered a more benign entity [6, 7]. Furthermore, the presence of non-driver mutations, particularly high molecular risk (HMR) variants such as ASXL1, EZH2, IDH1, IDH2, and SRSF2, is associated with a markedly worse prognosis and increased risk of leukemic transformation [8–10].

The progression from clonal haematopoiesis to overt MPNs is influenced not only by the acquisition of driver and non-driver mutations but also by factors governing clonal expansion, including hereditary predisposition, additional somatic mutations, and inflammatory stimuli [11].

Despite these advances, our understanding of the full mutational landscape and the relative impact of specific mutations remains incomplete.

Several reports challenge long-held assumptions regarding driver mutations; Some question their presumed mutual exclusivity, while others challenge the strict association of specific mutations with particular phenotypes.

In this context, it is not surprising that the coexistence of multiple driver mutations within a single patient can influence both clinical and histopathological features, a phenomenon clearly illustrated by the present case.

The patient’s clinical history of arterial thrombosis, polycythemia with trilinear hyperplasia on peripheral blood counts, and panmyelosis on bone marrow biopsy is consistent with PV, a diagnosis further supported by the identification of a second driver mutation through t-NGS. Based on current data, the individual contribution of each clonal mutation to the disease phenotype cannot be definitively determined. Despite extensive molecular analyses and multiple bone marrow evaluations, it is not possible to establish whether PV was driven primarily by the minor JAK2 V617F clone, the dominant CALR clone, or the combined activity of both.

This case raises several intriguing questions. One key issue is how the patient’s complex clonal architecture informs the PV diagnosis. Does the small JAK2 V617F clone act as the true driver, or does this represent one of the exceedingly rare cases of PV associated with a CALR mutation? In the latter scenario, could the JAK2 clone represent a concurrent clonal hematopoiesis of indeterminate potential (CHIP)?

Notably, at the time of diagnosis, the JAK2 VAF was > 2%, exceeding typical CHIP thresholds. It is possible that the JAK2 clone was higher at an earlier time point (e.g., during prior thrombotic events) and was subsequently suppressed by the expanding CALR clone.

Evidence from the literature supports the pathogenic potential of CALR in PV. Broseus et al. described two PV patients negative for JAK2 but positive for CALR mutation, highlighting the potential, albeit rare, association between CALR and PV phenotype [12]. In the same year, retrospective analyses identified concurrent JAK2 V617F and CALR exon 9 mutations in MF and ET patients [13]. Subsequent studies suggest that double driver mutations have likely been underestimated, with accumulating evidence indicating that JAK2/CALR co-mutations are more frequent than JAK2/MPL or CALR/MPL co-mutations [14].

Double mutations are most commonly reported in ET and MF, with very few documented cases in PV. There is currently no consensus regarding the interpretation of double mutations in MPNs, nor whether these cases should be considered a distinct clinical and prognostic entity. Some reports, including the present case, suggest increased clinical aggressiveness and poorer treatment responses in patients with double mutations. Additionally, these concurrent mutations are often associated with a low JAK2 V617F VAF (< 2%), suggesting that a pre-existing CHIP clone may acquire a second oncogenic event (CALR or MPL), giving rise to an overt MPN phenotype [15, 16].

These findings underscore the importance of analyzing CALR and MPL not only in JAK2 negative patients but also in patients with low JAK2 VAF, as such co-mutations may influence clinical phenotype, disease aggressiveness, and treatment response.

In conclusion, the mutational landscape of MPNs is far more complex than previously recognized, and the presence of multiple coexisting clones can give rise to diseases with distinct clinical and prognostic characteristics. Therefore, our case report underscores the importance of comprehensive clonal architecture analysis in the era of NGS. This approach is essential for achieving accurate diagnosis and prognostic stratification, particularly in cases with significant diagnostic uncertainty.

Future studies are needed to clarify how these molecular interactions should inform diagnosis, prognosis, and therapeutic decisions.

Author contributions

EC clinically evaluated the patient and wrote the manuscript; AC performed the molecular analysis and contributed to writing the manuscript; FV, FRI, RMB clinically evaluated the patient; FLG performed the biopsy analysis; MVR, MDC, LR and SA critically reviewed the manuscript. All authors read and approved the final version of the manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

The study was conducted in accordance with the Declaration of Helsinki.

Consent for publication

Written informed consent was obtained from the patient prior to study participation.

Competing interests

The authors declare no competing interests.