2023 年 30 巻 12 号 p. 1917-1926
Aim: The aim of the present study was to clarify the association between the Janus kinase 2 (JAK2) V617F mutation and large cerebral artery disease (LCAD) in patients with myeloproliferative neoplasms (MPNs).
Methods: We retrospectively analysed patients diagnosed with MPNs between June 1992 and June 2022 who underwent brain magnetic resonance imaging. LCAD was defined as extracranial or intracranial large artery stenosis (≥ 50%) or occlusion on magnetic resonance angiography.
Results: A total of 86 patients (47 males; median age, 69 years old) were enrolled in this study. JAK2 V617F mutation was detected in 63 (73.3%) patients and LCAD in 35 (40.7%) patients. Univariate analysis showed that history of ischaemic stroke (LCAD, 62.9% vs. non-LCAD, 11.8%; P＜0.001), JAK2 V617F mutation (91.4% vs. 60.8%, P=0.002), and age ≥ 60 years (85.7% vs. 60.8%, P=0.016) were significantly associated with LCAD. Multiple logistic regression analysis showed that, in addition to ischaemic stroke, age ≥ 60 years and diabetes mellitus, JAK2 V617F mutation (odds ratio 29.2, 95% confidence interval 1.2–709.8, P=0.038) was independently associated with LCAD. LCAD was frequently observed in the intracranial carotid (14/35, 40.0%) and middle cerebral (13/35, 37.1%) arteries.
Conclusions: This study revealed a significant association between the JAK2 V617F mutation and LCAD in patients with MPNs. This suggests that the JAK2 V617F mutation may promote cerebrovascular atherosclerosis and could be very important in determining therapeutic strategies for patients with not only JAK2 V617F-mutated MPNs but also LCAD-related stroke.
Myeloproliferative neoplasms (MPNs) are clonal haematopoietic disorders characterised by a chronic increase in white blood cells, red blood cells, and platelets, and include polycythaemia vera (PV) and essential thrombocythaemia (ET). Thrombotic complications are very common in cases of MPNs and are the most serious causes of morbidity and mortality in patients with MPNs. Janus kinase 2 (JAK2) V617F somatic mutation is observed in more than half of patients with ET and 95% of those with PV, and has been recognised not only as an MPN driver gene mutation, but also as a strong risk factor for thrombosis, including cerebrovascular disease1, 2).
The aetiology of stroke in patients with MPNs is varied and complicated because the patients have heterogeneous characteristics in terms of cerebrovascular risk factors and the presence of gene mutations3-5). The results of recent studies suggest that the JAK2 V617F mutation could enhance platelet activity and promote atherosclerosis in mouse models6-9). Therefore, the JAK2 V617F mutation may increase susceptibility to atherosclerosis-related stroke in patients with MPNs.
The aim of this study was to identify the clinical characteristics of JAK2 V617F mutation in patients with MPNs and clarify whether the presence of JAK2 V617F mutation is independently related to cerebral artery stenosis in patients with MPNs, even after controlling for conventional cerebrovascular risk factors.
In this single-centre retrospective study, we enrolled patients who were diagnosed with MPNs and underwent magnetic resonance imaging (MRI) mainly for the detection of stroke or its differential diagnosis and blood tests for the detection of the JAK2 V617F mutation at the Kawasaki Medical Hospital between June 1992 and June 2022. This study was conducted according to the principles of the Declaration of Helsinki. The study protocol was reviewed and approved by the Ethics Committee of the Kawasaki Medical School and Hospital (Approval No: 3653-00). Consent for participation in this study was obtained using an opt-out approach.
Data CollectionWe retrospectively collected all relevant clinical information from the medical records of the patients. The data collected included information on stroke risk factors (age, sex, hypertension, diabetes mellitus, dyslipidaemia, atrial fibrillation, current smoking and alcohol consumption status); stroke and other thrombotic events (haemorrhagic and ischaemic stroke including stroke subtype, transient ischaemic attack [TIA], ischaemic heart disease, peripheral artery disease, pulmonary embolism, and deep vein thrombosis); vasomotor symptoms such as headaches, erythromelalgia, acroparesthesia, livedo reticularis, dizziness, tinnitus, and visual changes; the brain MRI findings on the nearest day from MPNs diagnosis; MPN subtype (PV, ET, or primary myelofibrosis [PMF]); main MPN driver mutations (JAK2 V617F, JAK2 exon 12, MPL, and CALR); MPN treatment received at the time MRI was performed and at stroke onset; and laboratory test results (white blood cell count, haematocrit, and platelet count) obtained during the first visit to the department of haematology in our hospital and at stroke onset. Ischaemic stroke was categorised into three subgroups based on the time of onset: past stroke, which denotes the onset of stroke more than 6 months before MPN diagnosis; pre-MPN stroke, which indicates the onset of stroke within 6 months before MPN diagnosis; and post-MPN stroke, which is the onset of stroke after MPN diagnosis. Furthermore, pre- and post-MPN strokes were defined as MPN-related strokes. We performed neck and head magnetic resonance angiography (MRA) and 12-lead electrocardiogram (ECG), 24-hour Holter ECG, or prolonged (≥ 3 days) ambulatory ECG monitoring for the evaluation of stroke subtypes, which were determined based on the Trial of Org 10172 in Acute Stroke Treatment classification10). All the subjects with missing data were excluded from the study. The categorisation of MPN subtypes and management of MPNs were performed by expert haematologists (S.Y., T.K., and H.W.). In addition, each patient was tested for JAK2, MPL, and CALR mutations as reported previously11).
Magnetic Resonance ImagingMRI, including diffusion-weighted imaging (DWI), fluid-attenuated inversion recovery (FLAIR), and time-of-flight (TOF) MRA, was performed using a 3T scanner (Ingenia 3.0T CX, Philips Medical Systems, Best, The Netherlands). The scan parameters for DWI were as follows: repetition time (TR), 3500 ms; echo time (TE), 75 ms; b values, 0 and 1000 s/mm2; field of view (FOV), 240×225 mm2; acquired voxel size, 1.88×1.58×5.00 mm3; reconstructed voxel size, 0.94×0.94×5.00 mm3; flip angle, 90 degrees. The scan parameters for FLAIR were as follows: TR, 10000 ms; TE, 120 ms; Inversion Time, 2700 ms; FOV, 240×192 mm2; acquired voxel size, 0.79×1.11×5.00 mm3; reconstructed voxel size,0.43×0.43×5.00 mm3; flip angle, 90 degrees. The scan parameters for brain TOF-MRA were as follows: TR, 25 ms; TE, 3.45 ms; FOV, 200×190 mm2; acquired voxel size, 0.52×1.02×1.05 mm3; reconstructed voxel size, 0.28×0.28×0.50 mm3; flip angle, 18 degrees. The scan parameters for neck TOF-MRA were as follows: TR, 20 ms; TE, 3.45 ms; FOV, 200×200 mm2; acquired voxel size, 0.66×1.03×1.60 mm3; reconstructed voxel size, 0.39×0.39×0.80 mm3; flip angle, 18 degrees. Brain and neck TOF-MRA were performed to evaluate the location and the degree of large cerebral artery disease (LCAD), which was defined as stenosis (≥ 50%) or occlusion of the internal carotid artery (ICA), vertebral artery, basilar artery, or proximal segment of the middle cerebral artery (MCA), anterior cerebral artery or posterior cerebral artery. Evaluation of LCAD was independently performed by two stroke specialists (N.O. and Y.Y.) who were blinded to clinical information. In cases of interrater differences, a final diagnosis of LCAD was made based on an agreement between the two.
Statistical AnalysisThe clinical characteristics of the patients with MPNs with or without LCAD were compared using Fisher exact test for categorical variables. The Mann-Whitney U test was used for the assessment of continuous variables in the univariate analysis. Categorical variables are expressed as numbers and percentages, whereas continuous variables are expressed as medians (interquartile ranges). The interrater reliability for LCAD diagnosis was validated using Cohen’s kappa coefficient.
The distribution of LCAD was assessed in this study. To investigate the contribution of the JAK2 V617F mutation to the development of LCAD, we performed a multiple logistic regression analysis that included JAK2 V617F mutation, conventional cerebrovascular risk factors (age ≥ 60 years, male sex, hypertension, diabetes mellitus, dyslipidaemia, and current smoking), and variables with a statistically significant association on the univariate analysis. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. The patients with and without the JAK2 V617F mutation were compared to clarify the clinical features of the mutation in patients with MPNs. To identify and clarify the characteristics of ischaemic stroke in patients with MPN, we assessed the time of stroke onset, stroke subtype, recurrence, and treatments received at stroke onset. All statistical analyses were conducted using JMP 14.0.0 software (SAS Institute Inc., Cary, NC, United States). A P value ＜0.05 was considered statistically significant.
A total of 86 patients with MPNs were enrolled in this study. Table 1 shows the demographic and clinical characteristics of the patients with MPNs with and without LCAD. The median age of the 86 patients was 69 (95% CI, 57–76) years, and 39 (45.3%) of them were female. Regarding MPN subtypes, 60 (69.8%) patients had ET, 19 (22.1%) had PV, and 7 (8.1%) had PMF.
| Variables | Total MPNs (N = 86) | LCAD | P value | |
|---|---|---|---|---|
| Yes (N = 35) | No (N = 51) | |||
| MPNs, n (%) | ||||
| ET | 60 (69.8) | 25 (71.4) | 35 (68.6) | 0.816 |
| PV | 19 (22.1) | 8 (22.9) | 11 (21.6) | 1.000 |
| PMF | 7 (8.1) | 2 (5.7) | 5 (9.8) | 0.696 |
| Genetic mutations, n (%) | ||||
| JAK2 V617F | 63 (73.3) | 32 (91.4) | 31 (60.8) | 0.002 |
| JAK2 exon12 | 0 (0) | 0 (0) | 0 (0) | |
| CALR | 14 (16.3) | 2 (5.7) | 12 (23.5) | 0.037 |
| MPL | 0 (0) | 0 (0) | 0 (0) | |
| triple negative | 9 (10.5) | 1 (2.9) | 8 (15.7) | 0.077 |
| Age, years, median (IQR) | 69 (57―76) | 71 (64―79) | 64 (54―75) | 0.017 |
| Age ≥ 60 years, n (%) | 61 (70.9) | 30 (85.7) | 31 (60.8) | 0.016 |
| Stroke risk factors, n (%) | ||||
| Male | 47 (54.7) | 20 (57.1) | 27 (52.9) | 0.826 |
| Hypertension | 59 (68.6) | 26 (74.3) | 33 (64.7) | 0.479 |
| Diabetes mellitus | 25 (29.1) | 14 (40.0) | 11 (21.6) | 0.091 |
| Dyslipidaemia | 40 (46.5) | 18 (51.4) | 22 (43.1) | 0.513 |
| Atrial fibrillation | 4 (4.7) | 1 (2.9) | 3 (5.9) | 0.643 |
| Current smoking | 12 (14.0) | 5 (14.3) | 7 (13.7) | 1.000 |
| Alcohol | 13 (15.1) | 7 (20.0) | 6 (11.8) | 0.364 |
| Stroke events, n (%) | ||||
| Ischaemic stroke | 28 (32.6) | 22 (62.9) | 6 (11.8) | ＜0.001 |
| TIA | 11 (12.8) | 5 (14.3) | 6 (11.8) | 0.752 |
| Haemorrhagic stroke | 3 (3.5) | 2 (5.7) | 1 (2.0) | 0.564 |
| Other thrombotic events, n (%) | ||||
| Ischaemic heart disease | 5 (5.8) | 3 (8.6) | 2 (3.9) | 0.393 |
| Peripheral artery disease | 5 (5.8) | 1 (2.9) | 4 (7.8) | 0.644 |
| Deep vein thrombosis | 7 (8.1) | 5 (14.3) | 2 (3.9) | 0.115 |
| Vasomotor symptoms, n (%) | 15 (17.4) | 4 (11.4) | 11 (21.6) | 0.261 |
| Treatments at MRI study, n (%) | ||||
| Antiplatelet therapy | 34 (39.5) | 10 (28.6) | 24 (47.1) | 0.117 |
| Anticoagulant therapy | 7 (8.1) | 4 (11.4) | 3 (5.9) | 0.436 |
| Cytoreductive agent | 41 (47.7) | 13 (37.1) | 28 (54.9) | 0.127 |
| JAK inhibitor | 3 (3.5) | 0 (0) | 3 (5.9) | 0.267 |
| Phlebotomy | 4 (4.7) | 1 (2.9) | 3 (5.9) | 0.643 |
| Statin | 18 (20.9) | 9 (25.7) | 9 (17.7) | 0.424 |
| No treatments | 29 (33.7) | 16 (45.7) | 13 (25.5) | 0.065 |
| Interval from MPNs diagnosis to MRI study (d), median (IQR) | 247 (-7.3 ― 1889) | 147 (-29 ― 630) | 327 (-2 ― 2116) | 0.139 |
| Blood count at MPNs diagnosis, median (IQR) | ||||
| WBC (×102/mm3) | 105.4 (77.2 ― 136.8) | 109.0 (79.2 ― 144.0) | 100.1 (69.0 ― 136.0) | 0.302 |
| Haematocrit (%) | 45.1 (40.8 ― 48.9) | 46.0 (42.7 ― 49.0) | 44.3 (40.0 ― 48.4) | 0.098 |
| Platelets (×104/mm3) | 79.0 (50.7 ― 105.3) | 81.9 (44.0 ― 105.1) | 77.6 (56.9 ― 105.7) | 0.833 |
| Site of LCAD, n (%) | ||||
| Extracranial ICA | 2 (2.3) | 2 (5.7) | ||
| Intracranial ICA | 14 (16.3) | 14 (40.0) | ||
| Middle cerebral artery | 13 (15.1) | 13 (37.1) | ||
| Anterior cerebral artery | 2 (2.3) | 2 (5.7) | ||
| Posterior cerebral artery | 0 (0) | 0 (0) | ||
| Vertebrobasilar artery | 4 (4.7) | 4 (11.3) | ||
MPNs indicates myeloproliferative neoplasms; LCAD, large cerebral artery disease; ET, essential thrombocythaemia; PV, polycythaemia vera; PMF, primary myelofibrosis; JAK, Janus kinase; TIA, transient ischaemic attack; MRI, magnetic resonance image; IQR, interquartile range; WBC, white blood cells; ICA, internal carotid artery
LCAD was observed in 35 (40.7%) patients and the JAK2 V617F mutation in 63 (73.3%). There was a good interrater agreement for LCAD diagnosis (Cohen’s kappa=0.85). In most cases (33/35 cases, 94.3%), LCAD was detected in intracranial large cerebral arteries, including the intracranial ICA (14/35, 40.0%) and MCA (13/35, 37.1%). Univariate analysis showed that patients with LCAD had a significantly higher probability of having the JAK2 V617F mutation than those without LCAD (LCAD: 91.4% vs. non-LCAD: 60.8%, P=0.002), whereas patients without LCAD had CALR mutations more frequently than those with LCAD (5.7% vs. 23.5%, P=0.037). In addition, LCAD was observed more frequently in older patients (71 years vs. 64 years, P=0.017) and patients with ischaemic stroke (62.9% vs. 11.8%, P＜0.001) than in younger patients and those without ischaemic stroke, respectively. There were no significant differences in other stroke risk factors, TIA, haemorrhagic stroke, other thrombotic events, vasomotor symptoms, MPN treatments, and complete blood count results between the two groups.
Association between the JAK2 V617F Mutation and Large Cerebral Artery DiseaseTo clarify the contribution of the JAK2 V617F mutation to the development of LCAD, we conducted a multiple logistic regression analysis adjusted for established stroke risk factors, ischaemic stroke, JAK2 mutations and variables with a statistically significant association on the univariate analysis. The results showed that JAK2 V617F mutation (OR, 29.2; 95% CI, 1.2−709.8; P=0.038), ischaemic stroke (OR, 22.0; 95% CI, 4.8−100.6; P＜0.001), age ≥ 60 years old (OR, 13.3; 95% CI, 2.4−73.1; P=0.003), and diabetes mellitus (OR, 6.5; 95% CI, 1.4−30.0; P=0.017) were independently associated with LCAD (Table 2).
| Variables | Odds Ratio (95%CI) | P value |
|---|---|---|
| JAK2 V617F | 29.2 (1.2 ― 709.8) | 0.038 |
| Ischaemic stroke | 22.0 (4.8 ― 100.6) | ＜0.001 |
| Age ≥ 60 years | 13.3 (2.4 ― 73.1) | 0.003 |
| Diabetes mellitus | 6.5 (1.4 ― 30.0) | 0.017 |
| Current smoking | 1.6 (0.2 ― 12.3) | 0.652 |
| Dyslipidaemia | 1.2 (0.3 ― 5.4) | 0.778 |
| Male | 0.7 (0.2 ― 2.6) | 0.552 |
| Hypertension | 0.6 (0.1 ― 2.9) | 0.537 |
| CALR | 0.1 (0.0 ― 3.1) | 0.185 |
CI indicates confidence interval; JAK, Janus kinase
We compared the clinical background and characteristics of patients with MPNs with and without the JAK2 V617F mutation (Table 3). Forty-one of the 60 (68.3%) patients with ET, all 19 (100%) patients with PV, and 3 of the 7 (42.9%) patients with PMF had the JAK2 V617F mutation. Significantly more patients with the JAK2 V617F mutation had ischaemic stroke events (39.7% vs. 13.0%, P=0.034) and LCAD, particularly MCA disease (20.6% vs. 0%, P=0.017), than those without the mutation. There were no significant differences in stroke risk factors, TIA, haemorrhagic stroke, other thrombotic events, or vasomotor symptoms between the two groups. Patients with the JAK2 V617F mutation had clearly increased leukocyte count and haematocrit compared with patients without the mutation; however, there was no difference in platelet count between the two groups.
| Variables | Total MPNs (N = 86) | JAK2 V617F mutation | P value | |
|---|---|---|---|---|
| Yes (N = 63) | No (N = 23) | |||
| MPNs, n (%) | ||||
| ET | 60 (69.8) | 41 (65.1) | 19 (82.6) | 0.184 |
| PV | 19 (22.1) | 19 (31.2) | 0 (0) | 0.002 |
| PMF | 7 (8.1) | 3 (4.8) | 4 (17.4) | 0.079 |
| Genetic mutations, n (%) | ||||
| JAK2 V617F, n (%) | 63 (73.3) | 63 (100) | 0 (0) | ＜0.001 |
| JAK2 exon12, n (%) | 0 (0) | 0 (0) | 0 (0) | |
| CALR, n (%) | 14 (16.3) | 0 (0) | 14 (60.9) | ＜0.001 |
| MPL, n (%) | 0 (0) | 0 (0) | 0 (0) | |
| triple negative, n (%) | 9 (10.5) | 0 (0) | 9 (39.1) | ＜0.001 |
| Age, median (IQR) | 69 (57―76) | 69 (55―75) | 67 (57―77) | 0.891 |
| Age ≥ 60 ys | 61 (70.9) | 45 (71.4) | 16 (69.6) | 1.000 |
| Stroke risk factors, n (%) | ||||
| Male | 47 (54.7) | 31 (49.2) | 16 (69.6) | 0.141 |
| Hypertension | 59 (68.6) | 42 (66.7) | 17 (73.9) | 0.607 |
| Diabetes mellitus | 25 (29.1) | 18 (28.6) | 7 (30.4) | 1.000 |
| Dyslipidaemia | 40 (46.5) | 29 (46.0) | 11 (47.8) | 1.000 |
| Atrial fibrillation | 4 (4.7) | 2 (3.2) | 2 (8.7) | 0.289 |
| Current smoking | 12 (14.0) | 8 (12.7) | 4 (17.4) | 0.726 |
| Alcohol | 13 (15.1) | 8 (12.7) | 5 (21.7) | 0.320 |
| Stroke events, n (%) | ||||
| Ischaemic stroke | 28 (32.6) | 25 (39.7) | 3 (13.0) | 0.021 |
| TIA | 11 (12.8) | 7 (11.1) | 4 (17.4) | 0.475 |
| Haemorrhagic stroke | 3 (3.5) | 3 (4.8) | 0 (0) | 0.561 |
| Other thrombotic events, n (%) | ||||
| Ischaemic heart disease | 5 (5.8) | 3 (4.8) | 2 (8.7) | 0.607 |
| Peripheral artery disease | 5 (5.8) | 5 (7.9) | 0 (0) | 0.318 |
| Deep vein thrombosis | 7 (8.1) | 6 (9.5) | 1 (4.4) | 0.669 |
| Vasomotor symptoms, n (%) | 15 (17.4) | 11 (17.5) | 4 (17.4) | 1.000 |
| Blood count at MPNs diagnosis, median (IQR) | ||||
| WBC (×102/mm3) | 105.4 (77.2 ― 136.8) | 112.9 (86.4 ― 144.0) | 74.9 (60.6 ― 100.1) | ＜0.001 |
| Haematocrit (%) | 45.1 (40.8 ― 48.9) | 45.9 (42.1 ― 50.2) | 40.6 (38.2 ― 44.6) | ＜0.001 |
| Platelets (×104/mm3) | 79.0 (50.7 ― 105.3) | 78.5 (47.9 ― 105.1) | 82.0 (62.7 ― 105.7) | 0.354 |
| LCAD, n (%) | 35 (40.7) | 32 (50.8) | 3 (13.0) | |
| Site of LCAD, n (%) | ||||
| Extracranial ICA | 2 (5.7) | 2 (6.3) | 0 (0) | 1.000 |
| Intracranial ICA | 14 (40.0) | 11 (34.4) | 3 (100.0) | 0.751 |
| Middle cerebral artery | 13 (37.1) | 13 (40.6) | 0 (0) | 0.017 |
| Anterior cerebral artery | 2 (5.7) | 2 (6.3) | 0 (0) | 1.000 |
| Posterior cerebral artery | 0 (0) | 0 (0) | 0 (0) | |
| Vertebrobasilar artery | 4 (11.3) | 4 (12.5) | 0 (0) | 0.570 |
MPNs indicates myeloproliferative neoplasms; JAK, Janus kinase; ET, essential thrombocythaemia; PV, polycythaemia vera; PMF, primary myelofibrosis; TIA, transient ischaemic attack; MRI, magnetic resonance image; WBC, white blood cells; IQR, interquartile range; IS, ischaemic stroke; CBC, complete blood count; ICA, internal carotid artery
Table 4 shows the details of ischaemic stroke history and subtype, and treatments received at the time of MPN-related stroke onset in patients with MPNs, most of whom (25/28, 89.2%) had the JAK2 V617F mutation. Half of the cases of ischaemic stroke occurred within 6 months before MPN diagnosis (pre-MPN stroke), whereas 11 (39.3%) patients had stroke after MPN diagnosis (post-MPN stroke). Twenty-three (82.1%) patients had MPN-related stroke. Nine (32.1%) patients had past stroke, which is stroke that occurred more than 6 months before MPN diagnosis. Of these, 6 patients had increased platelet counts (＞34.8×104/mm3), whereas the remaining 3 patients had a high-normal level of platelet count, with the minimum value being 31.0×104/mm3. Regarding stroke subtypes observed in patients with MPN-related stroke, large-artery atherosclerosis (12/23, 52.2%) was most frequently observed, followed by stroke of undetermined cause (10/23, 43.5%). Most of the patients (19/23, 82.6%) with MPN-related strokes received no treatment for MPNs at stroke onset.
| Number (%) | |
|---|---|
| Total Ischaemic stroke | 28 (100) |
| Past stroke | 9 (32.1) |
| with CBC abnormalities＊ | 6 (66.7) |
| without CBC abnormalities＊＊ | 3 (33.3) |
| Pre-MPNs stroke | 14 (50.0) |
| Post-MPNs stroke | 11 (39.3) |
| MPNs-related stroke (1st attack) | 23 (82.1) |
| Large artery atherosclerosis | 12 (52.2) |
| Cardioembolism | 0 (0) |
| Small-artery occlusion | 0 (0) |
| Other cause | 0 (0) |
| Undetermined cause | 10 (43.5) |
| Insufficient evaluation | 1 (4.3) |
| Repeated stroke | 9 (32.1) |
| Treatments at MPNs-related stroke (1st attack) | |
| Antiplatelet therapy | 1 (3.6) |
| Anticoagulant therapy | 0 (0) |
| Cytoreductive agent | 3 (10.7) |
| JAK inhibitor | 0 (0) |
| Phlebotomy | 0 (0) |
| Insufficient information | 1 (3.6) |
| No treatments | 19 (82.6) |
MPNs indicates myeloproliferative neoplasms; CBC, complete blood count; JAK, Janus kinase; min, minimum; IQR, interquartile range
＊Any value outside of the reference ranges for complete blood count was regarded as an abnormality.
＊＊Any patient had a high-normal level of platelet count, and the minimum value was 31×104/mm3.
The present study is the first to reveal a significant association between the JAK2 V617F mutation in patients with MPNs and LCAD, most of which were found in the intracranial cerebral arteries. This finding could be a promising novel clue for predicting the progression of cerebral artery atherosclerosis in patients with MPNs. In the present study, the JAK2 V617F mutation was associated with ischaemic stroke, which was observed more frequently in patients who received no medication for MPNs. This evidence may be useful in the treatment of patients with MPNs who have the JAK2 V617F mutation.
The results of several studies have suggested an association between MPNs and ischaemic stroke with large vessel disease. However, the estimation of stroke aetiology in patients with MPNs is often complicated. Kato et al. reported that of 10 consecutive patients with ischaemic stroke with ET analysed in their study, 2 (20%) had LCAD-related stroke4). Tanashyan et al. reported that 21% of the young (≤ 58 years old) patients with MPNs (21/102) in their study experienced acute cerebrovascular events, and 8 of them (38%) had thrombotic occlusion of one of the internal carotid arteries12). In their prospective study, Nagai et al. identified large vessel involvement in 5 of 23 (22%) patients with MPNs with ischaemic stroke3). In the present study, the frequency of the development of LCAD in patients with MPNs was approximately 40%, which is higher than the results reported in previous studies. This could be attributed to the differences in study design (prospective vs. retrospective) and patient characteristics (all vs. younger people; patients with MPNs vs. patients with ET only) between previous studies and the present study.
In this study, we analysed the correlation between JAK2 V617F gene mutation and LCAD in patients with MPNs. The results revealed an independent association between JAK2 V617F mutation status and LCAD in patients with MPNs. However, there was no significant difference in complete blood count results between patients with and without LCAD. This could be significant and useful for identifying high-risk patients with progressive cerebrovascular atherosclerosis. It is widely known that MPNs often induce thrombosis, such as arterial and venous thrombosis, accompanied with abnormal proliferation of complete blood count components. The JAK2 V617F mutation is known to be a critical risk factor for thrombosis in patients with MPNs, especially ET13); however, the pathophysiological mechanism underlying this correlation has not been fully elucidated yet. In a previous study, the JAK2 V617F mutation caused inherent changes in platelet differentiation and reactivity with increased platelet aggregation in an ET mouse model14). The results of more recent studies suggest that the JAK2 V617F mutation affects platelets and other cell types, including neutrophils, macrophages, erythrocytes, and endothelial cells, leading to atherosclerosis7-9, 15). Macrophages play important roles in the progression of atherosclerosis in Jak2 gene-modified mice by augmenting erythrophagocytosis, efferocytosis, DNA oxidative stress, and inflammation7, 8). Dotan et al. showed that acceleration of atherosclerosis owing to defects in cholesterol efflux from macrophages occurs in ApoE-null myeloid Jak2-specific knockout mice9). Hekimoglu et al. reported the induction of apoptosis and release of JAK2 V617F-positive microparticles in JAK2 V617F-positive endothelial cells15). These previous experimental studies suggested the association of JAK2 V617F mutation with not only platelet aggregation but also arterial atherosclerosis. Our study is the first to investigate the impact of JAK2 V617F mutation on cerebrovascular atherosclerosis in clinical practice and to demonstrate the significant association between LCAD and the JAK2 V617F mutation independent of complete blood count.
In this study, LCAD was more frequently observed in intracranial arteries than in extracranial arteries. Intracranial and extracranial LCAD was observed in 38% and 2% of the patients with MPNs, and in 48% and 3% of patients with the JAK2 V617F mutation, respectively. In general, intracranial LCAD is more frequently observed in Asians, Hispanics, and Africans than in Caucasians, possibly because of differences in genetic susceptibility and lifestyles between races16, 17). The prevalence of intracranial LCAD may be attributed to the lack of external elastic lamina and increased expression of proinflammatory proteasomes in intracranial arteries16-19). A results of a post-mortem pathological study of a patient with ET who had repeated strokes suggested that pathological platelet hyperaggregability and shear-induced platelet aggregation contribute to the development of atherosclerosis and in situ formation of occlusive platelet thrombi in intracranial artery stenosis20). The more curved and branched vessels in intracranial arteries than in extracranial arteries may be one of the reasons for shear-induced platelet activation and endothelial dysfunction in patients with MPNs, resulting in the formation of occlusive thrombi21, 22).
In the present study, most of the patients with MPN-related stroke received no treatment for MPNs at stroke onset. This means that stroke events were often the first presentation of MPNs. In fact, ischaemic stroke occurred just before MPN diagnosis in 50% of the cases. In a previous study, ischaemic events were the first manifestation of underlying MPNs in 58% of patients with MPN-related ischaemic stroke23). Interestingly, in the present study, high-normal platelet levels were observed in some patients at the onset of past stroke, which occurred more than 6 months before MPN diagnosis. Clonal haematopoiesis of indeterminate potential (CHIP) is a newly recognised condition characterised by expanded somatic blood cell mutations, such as DNMT3A, TET2, ASXL1, and JAK2, and is associated with development of atherosclerosis, cardiovascular diseases, and haematologic malignancies24, 25). A recent study demonstrated that a CHIP (DNMT3A and TET2 mutations) is associated with an increased risk of stroke26). Similarly, the JAK2 V617F mutation may promote cerebrovascular atherosclerosis before inducing overt blood count abnormalities. Evidence from previous studies and the results of the present study suggest that vigorous assessment of the JAK2 V617F mutation in patients with LCAD without overt haematological malignancies could be crucial for the prevention of ischaemic stroke.
This study has several limitations. First, we enrolled patients diagnosed with MPNs between 1992 and 2022. The classification and diagnostic criteria for MPNs have changed over time. Specifically, the World Health Organization made major revisions to the diagnostic criteria for MPNs in 2008 after the discovery of the JAK2 V617F allele in 2005, including lowering the platelet count threshold and adopting clonal markers such as JAK2 and MPL. The current diagnostic criteria for MPNs are based on those published in 2008; however, minor changes have been made several times. All patients enrolled in this study met the 2008 criteria for MPNs and were followed up until at least 2013. Therefore, we believe that the results of this study are relevant. Second, the interval between MPN diagnosis and MRI evaluation varied in this study; however, the difference in mean interval between the groups with and without LCAD was not significant. Cerebrovascular disease may progress markedly shortly after MPNs diagnosis in some cases. Third, stroke subtype was determined by evaluating at least the MRA and 12-lead ECG results of each patient, not the results of full studies, including transthoracic/transoesophageal echocardiogram and 24-hour Holter ECG. It has been reported that 0 to 10% of MPNs or ET patients with or without ischaemic stroke have concomitant cardioembolic diseases3-5, 27, 28). In the present study, approximately 5% of all the patients with MPNs had atrial fibrillation. This result is consistent with those of the previous studies. Finally, most of the patients in this study underwent MRIs to detect stroke or its differential diagnosis, which created a selection bias. Thus, the findings from our study cannot be generalised for all cases. Moreover, this was a retrospective single-centre study with a small sample size; therefore, further prospective studies with larger sample sizes in all patients with MPNs, those with ischaemic stroke, or the general population are needed to confirm our findings.
This study revealed a significant association between the JAK2 V617F mutation and LCAD, particularly intracranial LCAD, in patients with MPNs. The results of this study indicate that vigorous evaluation of not only cerebrovascular disease in patients with JAK2 V617F-mutated MPNs but also the JAK2 V617F mutation status in patients with LCAD-related ischaemia is essential. We strongly suggest good cooperation between neurologists and haematologists to improve the clinical management of patients with MPNs before the onset of stroke and patients with LCAD who have stroke.
We express our gratitude to Mari Okamoto for data collection and secretarial assistance. We also thank Editage (www.editage.com) for English language editing.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
The authors report no conflicts of interest.
