CALR: Calreticulin; DVT: Deep Vein Thromboses; EPO: Erythropoietin; ET: Essential Thrombocythemia; MPN: Myeloproliferative Neoplasm; PE: Pulmonary Embolism; PMF: Primary Myelofibrosis; PV: Polycythemia Vera

Philadelphia-negative myeloproliferative neoplasms (MPNs) are clonal hematopoietic disorders, including polycythemia Vera (PV), Essential Thrombocythemia (ET) and Primary Myelofibrosis (PMF). Three driver mutations have been identified as inducers of these diseases: JAK2V617F (95-97% of PV and 50-60% of ET and PMF patients), Calreticulin (CALR) and MPL mutations (20-30% and 2-5% of ET or PMF patients, respectively). JAK2 is a tyrosine-kinase bind to the intracellular part of the EPO-receptor. JAK2V617F is a somatic acquired mutation, located in its pseudo-kinase domain which transforms JAK2 in a gain-of-function enzyme, by continuous activation of the kinase domain despite the low level of circulating EPO. It induces an independent growth from cytokines (i.e., EPO for immature red cells) and a hypersensibility to these cytokines (if they are added to cells cultures) [1]. The result is an increase of mature cells in bone marrow and blood, inducing hyperviscosity and thromboses which are the main complications (20 to 50% of the patients). Ten to fifteen percent of MPNs are diagnosed at the time of a thrombotic event. Venous thromboses represent one-third of these events [2].

The clinicians (mostly non-hematologists) who treat people experiencing thromboses are constantly searching for the etiology, to adapt the treatment to each case and to prevent recurrence. If arterial thrombotic events are mostly due to arteriosclerosis (cardio-vascular risk factors should be identified, i.e., hypertension, diabetes mellitus, hypercholesterolemia…), venous thrombotic events are more relating to extrinsic compression (i.e., solid or hematological cancers) or perturbation of the bloodstream (i.e., hyperviscosity by an increase of circulating cells or pro-coagulant status like inflammation...).

In the general population, deep vein thrombosis with or without pulmonary embolism (DVT/PE) are the most frequent venous thrombotic sites. The diagnoses can be made with a good clinical exam evaluating the Wells score (or assimilated) plus quantification of D-dimers and realization of duplex ultrasound for lower legs DVT and/ or CT angiography for PE [3]. These sites have been widely studied for the presence of hidden or obvious cancers, such as MPNs. The research of those mutations is quite easy because it just requires blood sample but could be time consuming with a usual timing of two to four weeks to receive the results.

Because, a multidisciplinary approach may be then necessary to decide the best treatment (i.e., direct oral anti-coagulants have not been approved for use in cases of cancer whereas cytoreductive drugs have been approved to reduce the risk of thrombotic recurrence in MPNs), screening for MPNs in these situation need to be discussed.

The purpose of this review is to analyze available data from studies referenced on PubMed website regarding only the screening of MPNs mutations in cases of DVT/PE, to summarize what we can learn from them and to propose an appropriate method for this screening, to reduce time and costs.

Since 2005 when the JAK2V617F mutation was identified, sixteen studies of a series of patients have been published on this topic. Only two were related to the screening of CALR mutations and none addressed MPL mutations. All the results are listed in Table 1.

Table 1: Screening for JAK2V617F and CALR mutations among patients who experienced idiopathic deep vein thrombosis and/or pulmonary embolism. DVT: Deep Vein Thrombosis; nb: Number; nd: Not Done; PE: Pulmonary Embolism; VTE: Venous Thrombotic Event

At least, 2907 patients have been screened for JAK2 or CALR mutations. Only 39 (1.3%) of these patients have been identified as being positive for JAK2. In cases that begin with idiopathic thrombosis, biologists identified only 27/2338 (1.15%) positive patients. In contrast, 12/569 (2.1%) patients with recurrent DVT/PE were identified as positive (in studies that clearly reported the recurrence status).

Unfortunately, of 394 patients with one thrombotic event and 372 patients with recurrence, none of the patients was positive for CALR mutations.

It may not be easy to determine the precise localization of venous events and the status of recurrence because the combination of DVT and PE was not systematically and clearly separated from DVT alone or PE alone. However, when this information was provided, the occurrence of DVT or PE did not change the prevalence of positivity. Additionally, the classification of an idiopathic event could also differ from one study to another, thus altering the rate of positivity.

Minimal data are available for 34 (87%) patients (Table 2). The sex ratio is 0.36 due to the presence of 25 women. The median age of the patients is 70 years (33-87), with 26 patients (76.5%) who are older than 60 years. From the hemogram, twelve (35.3%) patients had blood cell counts greater than the normal range (8 with leukocytosis, 6 with thrombocytosis and 2 with polycythemia). The median values were 9.3 giga/l (3.7-32.8) for leukocytes, 13.9 g/dl (8.9-17.7) for the hemoglobin level and 304 giga/l (154-1180) for platelets. The allele burden was quantified in 21 cases, with a median value of 3.4% (0.75-32). This level is low and more commonly associated with ET than with other MPNs [1].

Table 2: Characteristics of the JAK2V617F-positive patients identified during the screening of DVT/PE occurrences. ET: Essential Thrombocythemia; F: Female; M: Male; m: Months; nd: Not Done; PV: Polycythemia Vera; uMPN: Unspecified Myeloproliferative Neoplasm

Two patients were previously followed for MPNs at the time of their thrombosis and eight patients were secondary diagnosed as MPN (6 ET, 3 PV and 1 unspecified MPN) during the first year of followup. Interestingly, only patients with thrombocytosis or polycythemia were diagnosed for MPN (all but one with thrombocytosis). The occurrence of leukocytosis in some patients could only be related to the inflammatory state of the thrombosis (in most cases, the values spontaneously decrease into normal range). Thus, despite the presence of the JAK2V617F mutation, 70.6% of the patients failed to be diagnosed with MPN, even if the follow-up could be considered sufficiently long (median of 42 months, 0.25-112). Insufficient information has been obtained to determine whether these patients have been effectively seen by hematologists to confirm or rule out the MPN status.

Another interesting point is the recurrence rate: 19/28 (68%) of the patients who were positive for a JAK2 mutation experienced more than one thrombotic event regardless of the follow-up timing. Even if most of the patients did not exhibit MPN features, the presence of the JAK2V617F mutation could probably be considered as a thrombophilic state, which should not be minimized. These patients should be carefully followed once they are identified.

For the first time, a complete analysis of published studies concerning the screening for MPNs mutations in case of DVT and/or PE has been made here. Salient points could be resumed as:

1) Ten to fifteen percent of the MPNs diagnoses are made after a thrombotic event and one-third of the events are venous sites; 2) DVT and/or PE are the most frequent venous thrombotic sites in general population but the screening for MPNs mutations in these situations is unproductive with less than 1.5% of positivity; 3) a complete and systematical screening should be avoid in this thrombotic situation, in order to reduce costs and time wasting because only JAK2V617F have been found so, only patients with a proliferative profile on the hemograms should be screened to identify a MPN (<30% of the positive cases or 0.3% of all tested cases) enhancing the value of this simple and cheap exam; 4) all positive cases should be systematically investigated for MPN with a bone marrow trephine biopsy and/or red mass cells to change the way of treating them (systematical prescription of cytoreductive drugs); 5) but, even in case of non-proven MPNs, all the patients also need to be carefully followed to reduce thrombotic recurrence observed for 68% of them (i.e., long-lasting prescription of vitamin-K inhibitors like for patients with solid cancer ?).

The same analysis could be done for all thrombotic sites because most of them have been explored for MPNs mutations but, few data are available on the rare positive cases, reducing the value of this analysis. Far less frequent than DVT, thromboses of cerebral or retinal veins are also rarely found positives for JAK2V617F (5.2% and 0%, respectively), with no study for CALR mutations. Among arterial thrombotic events, only 0.6% of the screened patients were positive for JAK2V617F [20,21].

On the other hand, and despite their rarity, the most relevant sites to screen for MPNs mutations are splanchnic vein thromboses. In the principal published series, the authors have observed a positivity of JAK2V617F in 34% of portal vein thromboses and 47% of Budd-Chiari syndrome, but less than 1% of CALR positive cases [22]. Interestingly, in these thrombotic sites, most of the patients have normal blood counts due to a typical congestive splenomegaly. So, the patients with such an idiopathic thrombosis are the only ones who should be systematically investigated for MPNs mutations added to a bone marrow trephine biopsy (whatever was the results of the screening) and/or a red mass cells (in case of JAK2V617F positivity) to assess the presence of a MPN.

This review focused on the screening for classical MPN mutations among patients who experienced DVT and/or PE. This screening should not be systematically performed in these situations but only in case of thrombocytosis or polycythemia on hemograms and only for JAK2V617F mutation. All the positive cases should be discussed with hematologists after a bone marrow trephine biopsy to properly diagnose and treat a MPN.