Thrombosis risk in patients with secondary myelofibrosis

Increased risk of thrombotic events is an established feature of Philadelphia-negative myeloproliferative neoplasms (MPN), including essential thrombocythemia (ET), polycythemia vera (PV), and primary myelofibrosis.¹ However, this increased risk of thrombotic events has not yet been recognized in secondary myelofibrosis (SMF) that occurs post PV (PPV-MF) or ET (PET-MF).

Establishing any increased risk of thrombotic events in patients with SMF would rationalize active treatment of the condition, given the significant morbidity and mortality that can accompany thrombotic events. However, as potential cytoreductive therapy carries its own risk of adverse events and toxicities, research into the area is needed to inform decisions about disease versus treatment risk in patients with SMF.¹

The MPN Hub has previously discussed a prognostic model of SMF from the Myelofibrosis Secondary to PV and ET (MYSEC) project. Here, we summarize findings relating to thrombosis risk from the MYSEC project database, the largest retrospective study to only include patients with SMF. Published in Leukemia in August 2022, Mora et al.¹ investigated the potential need to consider treatment of some patients with SMF to protect them from thrombotic events.

Study design

As of June 2021, the European MYSEC database had registered 12,58 patients with SMF who were diagnosed between 1981 and 2020 across 19 centers. Patient records were checked for thrombotic events, which were defined as:

• ischemic stroke;
• acute myocardial infarction;
• peripheral arterial thrombosis;
• venous thromboembolism, including deep vein thrombosis of the extremities;
• splanchnic vein thrombosis; or
• pulmonary embolism.

Using demographic and clinical data, univariate and multivariate analyses were performed to identify risk factors for thrombotic events, including venous and arterial, in patients with SMF.

Results

Among all patients, the median follow-up was 3.5 years (range, 1–21.5 years), and 615 had PPV-MF and 643 had PET-MF. The clinical characteristics of all patients are shown in Table 1. In total, 135 patients (10.7%) had experienced a thrombotic event and thrombotic events accounted for 2.5% of all reported causes of death. At the point of data cut-off, 48% of patients had died, with a mortality rate of 9.4% patients per year.

Table 1. Clinical characteristics of all patients included from the MYSEC database*

ET,essential thrombocythemia; Hb, hemoglobin; LCM, left costal margin; MYSEC-PM, Myelofibrosis Secondary to PV and ET prognostic model; PET-MF, post-essential thrombocythemia myelofibrosis; PLT, platelets; PPV-MF, post-polycythemia vera myelofibrosis; PV, polycythemia vera; SCT, stem cell transplant; SMF, secondary myelofibrosis; WBC, white blood cell. *Adapted from Mora, et al.1 †Chi-square or Fisher exact test for categorical variables, and T-test for continuous variables. ‡Wilcoxon rank sum test.
Characteristic, % (unless otherwise stated)	SMF (N = 1258)	PPV-MF (n = 615)	PET-MF (n = 643)	p value†
Median age at SMF (range), years	64 (20–95)	65 (33–95)	64 (20–93)	0.001
Median duration of PV/ET (range), years	11.2 (1–41.4)	11.4 (1–41.4)	11.0 (1–39.3)	0.66‡
Median Hb (range), g/dL	11.1 (4.2–16.0)	11.9 (6.3–16.0)	10.6 (4.2–16.0)	<0.0001‡
Hb <10 g/dL	30.3	22.4	37.7	<0.0001
Hb <8 g/dL,	5.2	2.9	7.3	0.001
Median PLT (range), ×109/L	327 (7–1,908)	271 (7–1,689)	375 (25–1,908)	<0.0001‡
PLT 50–100 × 109/L	6.6	8.9	4.4	0.002
PLT <50 × 109/L	2.9	4.3	1.5	0.003
Median WBC count (range), × 109/L	10.1 (0.4–121.5)	12.8 (0.6–121.5)	8.1 (0.4–90)	<0.0001‡
WBC count >11 × 109/L	45.9	57.7	34.6	<0.0001
Median spleen length (range), cm from LCM	6 (0–34)	8.0 (0–34)	4.0 (0–26)	<0.0001‡
Spleen length <5 cm from LCM	45.0	32.9	56.6	<0.0001
Spleen length 5–10 cm from LCM	30.1	34.8	25.7	0.0004
Spleen length >10 cm from LCM	24.9	32.4	17.7	<0.0001
Constitutional symptoms	39.0	44.7	33.4	<0.0001
JAK inhibitors at baseline	16.2	22.5	10.3	<0.0001
Anti-platelets at diagnosis	50.8	55.3	46.3	0.003
Previous thrombosis	22.2	25.7	18.8	0.003
Previous thrombosis; venous	48.7	63.5	36.5	0.02
Allogeneic SCT	12.2	10.0	14.4	0.05
MYSEC-PM low-intermediate 1	66.2	60.3	72.0	<0.0001
MYSEC-PM intermediate 2-high	33.8	39.7	28.0	<0.0001
Abnormal cytogenetics	36.7	41.9	31.9	0.02

The prevalence of overall, venous, and arterial thrombotic events in patients with SMF, PPV-MF, and PET-MF can be seen in Table 2. Of the thrombotic events recorded, 51.3% were arterial and 48.7% were venous. The most common venous thrombotic events were deep and superficial vein thrombosis (40.5% and 34.5%, respectively), with myocardial infarction (41.7%) and ischemic stroke (27%) the most common arterial thrombotic events.

Table 2. Incidence of thrombotic events in patients with SMF, PPV-MF and PET-MF (per year)*

Previous venous thrombotic events were more strongly associated with PPV-MF than PET-MF (63.5% vs 36.5%; p = 0.02). No thrombotic event progressed within 90 days of first recognition. Overall, 22.2% of patients had experienced a thrombotic event before evolving to SMF, of which 25.7% had PV and 18.8% had ET at the time of the event, the remainder having no hematologic disease.

Univariate and multivariate analyses demonstrated several factors predictive of first venous or arterial thrombotic event following progression to SMF (e.g., previous history of thrombotic event), while others were protective.

• Univariate analysis
  • Hemoglobin <11 g/dL (hazard ratio [HR],1.40; 95% confidence interval [CI], 0.99–1.97; p = 0.06)
  • Presence of constitutional symptoms (HR, 1.34; 95% CI, 0.96–1.89; p = 0.09)
  • Previous diagnosis of PV (HR, 1.38; 95% CI, 0.98–1.93; p = 0.07)
  • History of thrombotic event (HR, 1.51; 95% CI, 1.03–2.20; p = 0.03)
• Multivariate analysis
  • Hemoglobin <11 g/dL (HR, 1.70; 95% CI, 1.02–2.82; p = 0.04)
  • Cytoreductive therapy at baseline (HR, 0.46; 95% CI, 0.26–0.83; p = 0.009)

Regardless of death from any cause, cytoreductive therapy at baseline was protective against both arterial (subdistribution HR [sHR], 0.43; 95% CI, 0.36–5.55; p = 0.11) and venous (sHR, 0.49; 95% CI, 0.25–0.97; p = 0.04) thrombotic events. Conversely, previous history of venous thrombotic events during PV/ET was predictive of further venous thrombosis (sHR, 3.28; 95% CI, 1.58–6.78; p < 0.001), but not arterial thrombosis (p = 0.49).

Conclusion

Thrombosis is a real risk for patients with MPN. Previous history of thrombotic events is strongly predictive of further events in patients with MPN, and consideration of this risk factor must be addressed in the clinical assessment and history taking of this group of patients. This study has identified that patients receiving cytoreductive therapy at the point of evolving to SMF have fewer thrombotic events than patients not on treatment, suggesting that there is potential utility for the use of cytoreductive therapy for preventing thrombosis in SMF, alongside managing hemodynamic features of the disease, such as blood counts and splenomegaly.