. 2021 May 21;35(8):2166–2181. doi: 10.1038/s41375-021-01290-6

Table 1.

Selected practically relevant studies focusing on management issues in JAK2 unmutated erythrocytosis.

Condition predisposing to erythrocytosis (Incidence of erythrocytosis)	Study/Reference	Noteworthy findings
COPD (5.9–18.1%)	Prognostic value of hematocrit in severe COPD receiving long-term oxygen therapy (retrospective observational study) Chambellan et al., Chest [125]	- 2524 COPD pts with mean HCT 45.9 ± 7.0% in men and 43.9 ± 6.0% in women. - HCT an independent predictor of survival, hospital admission rate, and cumulative duration of hospitalization. - 3-year survival 24% when HCT < 35%, 59% with HCT 50–54%; and 70% when HCT ≥ 55% (p < 0.001).
Hemoglobin levels and long-term survival in COPD. (retrospective study) Kollert et al. [107]	- 309 COPD pts with chronic respiratory failure under optimized therapy, including oxygen and non-invasive ventilation. Excluded pts on phlebotomy. - 46 (14.9%) anemic (Hb < 12 g/dL in females, Hb < 13 g/dL in males), 56 (18.1%) polycythemic (Hb ≥ 15 g/dL in females, Hb ≥ 17 g/dL in males), or 207 (67%) normocythemic with median survival 29, 112 and 51 months (p = 0.008). (values determined prior to oxygen therapy). - Hb ≥ 14.3 g/dL in females, and ≥15.1 g/dL in males independently associated with longer survival.
Pulmonary embolism in chronic hypoxemic patients. (prospective study) Ristic et al., Med Glas (Zenica) [126]	- 362 pts with severe COPD exacerbation or raspatory failure, with D-dimer ≥500 µg/l referred for Doppler ultrasound and CT. - Group 1 (100 pts with erythrocytosis) vs group II (262 without erythrocytosis). - Higher pulmonary embolism in group 1 vs 2 (39% vs 11.1%)
Prevalence of VTE in COPD with erythrocytosis. (retrospective case-control study) Nadeem et al. Clinical and Applied Thrombosis/Hemostasis [101]	- COPD with erythrocytosis, HCT ≥ 50 (n = 86) vs age, sex-matched COPD without erythrocytosis, HCT < 50 (n = 86). - No difference in VTE ;17 (19.8%) cases vs 12 (14%) controls, p = 0.42 - Trend toward higher incidence of idiopathic VTE in cases (n = 10 vs 4, p = 0.16) - Similar BMI, smoking, cancer, OSA with differences in hematocrit, oxygen use, pulmonary HTN. Phlebotomy data not provided
Incidence of cardiovascular and thrombotic events in secondary polycythemia. (retrospective study) Mao et al. [109]	- Impact of phlebotomy on prevalence of arterial/venous thrombosis in COPD with erythrocytosis (n = 115). - Thrombosis in 11/35 (31.4%) phlebotomized vs 26/116 (22.4%) in non-phlebotomized (p = 0.28). - Amongst phlebotomized pts, thrombosis in 4/16(25%) with HCT < 52% vs 7/19(36.8%) in those with higher HCT (p = 0.45)
Secondary polycythemia and perioperative hemorrhage or thrombosis. (retrospective case-control study Lubarsky et al. [102]	- COPD with erythrocytosis, Hb > 16 g/dl (16.2–20.1 g/dl) (n = 100) vs age, sex, surgery, ASA physical status-matched controls without erythrocytosis. - No thrombosis in erythrocytosis group vs 3 events in controls. - Significantly less transfusion requirement in erythrocytosis group vs controls (p < 0.025).
Symptomatic and pulmonary response to acute phlebotomy in secondary polycythemia. (Double-blind clinical trial) Dayton et al. [110]	- 11 pts with hypoxic lung disease associated erythrocytosis, HCT > 54 underwent phlebotomy vs 8 pts as controls (sham procedure). - 8/11 phlebotomized patients with symptom (dyspnea, fatigue, headache) improvement within 24 h and lasted > 7 days in 5 pts, particularly if HCT > 60 vs no improvement in controls (p < 0.005). - No objective improvement in airway obstruction, gas exchange or exercise tolerance.
Cyanotic congenital heart disease	Risk of stroke in adults with cyanotic congenital heart disease. (observational study) Perloff et al. [104]	- 112 adults with cyanotic heart disease excluding those with independent risk factors for stroke were observed for 748 pt-years. - 2 groups, (i) compensated (HCT 46–72%, absent/mild symptoms, iron replete with phlebotomy at intervals > 1 yr if symptoms) (n = 101) vs (ii) decompensated (HCT 61.5–75%), iron deficiency, marked/severe symptoms with phlebotomy every 3–6 months (n = 11). - None of the pts in group (i) had cerebral arterial thrombosis, vs one pt in group (ii) with amaurosis fugax.
Cerebrovascular events in adults with cyanotic congenital heart disease (retrospective study) Ammash et al. [104]	- 162 adults with cyanotic congenital heart disease; Group I (n = 140) no history of cerebrovascular event after age 18 years, Group II (n = 22) well documented cerebrovascular event (TIA (n = 19), reversible ischemic neurologic deficit n = 4, infarct, n = 6). - 46/162 (28.4%) underwent phlebotomy, with increased risk of events after phlebotomy (35/140 in Group I vs 11/22 in Group 2, p = 0.016). - Strong association of iron deficiency and/or microcytosis with cerebrovascular events (11/41; p = 0.004).
Hydroxyurea for secondary erythrocytosis in cyanotic congenital heart disease. (case series) Reiss et al. [113]	- 4 pts with symptomatic secondary erythrocytosis and cyanotic congenital heart disease. - 2 pts with recent TIA or stroke, one with extreme fatigue/dyspnea and one with extreme exhaustion following phlebotomy. - Median hydroxyurea use: 15 months; symptomatic improvement, but myelosuppression in 2 pts and thrombocytopenia in 1 pt which required dose reduction. 1 pt with TIA on hydroxyurea. Minimal increase in HbF.
Post-renal transplant erythrocytosis (8-15%)	Enalapril for erythrocytosis after renal transplant. (randomized double-blind study) Beckingham et al. [116]	- 2.5 mg of enalapril daily (n = 15) or placebo (n = 10) for 4 months. - Hematocrit decreased from 52.7 to 47.1 at 1 month and 46.1 after 4 months with enalapril, no change with placebo (p = 0.004). No change in erythropoietin level. - No phlebotomy or thrombosis.
Effects of theophylline on erythrocytosis after renal transplant. (prospective study) Bakris et al. [118]	- Pts with post-renal transplant erythrocytosis (n = 8) vs normal controls (n = 5). - 8 weeks of theophylline, EPO significantly reduced in transplant pts (60 to 9 units; p < 0.05) and controls (6.9 to 4.7 units; p < 0.05). - Hematocrit reduced in transplant pts (0.58 to 0.46; p < 0.05) and controls (0.43–0.39; p < 0.05). Requirement of weekly phlebotomy eliminated.
Comparison of enalapril and losartan on post-renal transplant erythrocytosis. (prospective randomized study) Yildiz et al. [117]	- 27 pts treated with enalapril 10 mg/day (n = 15), vs losartan 50 mg/day (n = 12) for 8 weeks. - Hemoglobin significantly decreased with both losartan (17.1–15.9 g/dl, p = 0.01) and enalapril (17.4–14.9 g/dl, p = 0.001). greater decrease with enalapril (−3.26 vs −1.70, p = 0.05). - Among the responders who discontinued treatment, there was a trend for longer time to relapse with losartan (7.38 months) compared with enalapril (2.75 months) (p = 0.11).
Testosterone therapy and erythrocytosis	Prevalence and management of secondary erythrocytosis in transgenders on testosterone. (retrospective study) Oakes et al. [122]	- 234 pts, mean pre-testosterone hemoglobin 13.5 g/dL and hematocrit 40.3%. Mean hemoglobin peak 15.7 g/dl, hematocrit 47.2% at an average of 21 months post therapy. 23.5% pts with hematocrit > 50%, and 8.5% hemoglobin > 17.5 g/dL. Only one thrombotic event. - Dose reduction in 14.5% with erythrocytosis, no phlebotomy. - 88.9% of patients with erythrocytosis had received testosterone cypionate
Chuvash polycythemia (CP)	Thrombosis in CP. (prospective study) Sergueeva et al. [105]	- CP pts matched by age, sex and place of residence (n = 128). - CP pts with lower blood pressure, body mass index, white count, but more smokers than controls. - New thrombosis in CP pts 0.031 events/pt/year (34 arterial+venous events vs 3 arterial events in controls), history of phlebotomy but not hematocrit predictor of thrombosis. - 9 (7.0%) deaths in CP pts (median age; 54 years) compared to 2 (1.6%) controls (median age at death; 81 years) (p = 0.058). All deaths related to thrombosis.
Congenital erythrocytosis	Thrombotic risk in congenital erythrocytosis. (prospective study) Gordeuk et al. [106]	- CP pts and matched controls (n = 155). - 40 thrombotic events in CP at enrollment (n = 27) with 37 new events (n = 33) including 9 fatal events during 11-year observation. Thrombosis in 3 controls at enrollment with 5 new events. - In multivariate analysis in CP pts, age and past thrombosis but not hematocrit were independent predictors of new events. Phlebotomy associated with increased thrombosis (HR 1.9, p = 0.028). None were on anticoagulation at second event; aspirin 75 mg/day was not protective. - HIF2A p.M535V variant six-generation pedigree (8 subjects), arterial/venous thrombosis in 5 of 8 vs none in 17 HIF2A wild-type pts (p = 0.001). - Thrombotic events despite phlebotomy with HCT < 45%.

Condition predisposing to erythrocytosis (Incidence of erythrocytosis)

Study/Reference

Noteworthy findings

COPD (5.9–18.1%)

Prognostic value of hematocrit in severe COPD receiving long-term oxygen therapy

(retrospective observational study)

Chambellan et al., Chest [125]

- 2524 COPD pts with mean HCT 45.9 ± 7.0% in men and 43.9 ± 6.0% in women.

- HCT an independent predictor of survival, hospital admission rate, and cumulative duration of hospitalization.

- 3-year survival 24% when HCT < 35%, 59% with HCT 50–54%; and 70% when HCT ≥ 55% (p < 0.001).

Hemoglobin levels and long-term survival in COPD.

(retrospective study)

Kollert et al. [107]

- 309 COPD pts with chronic respiratory failure under optimized therapy, including oxygen and non-invasive ventilation. Excluded pts on phlebotomy.

- 46 (14.9%) anemic (Hb < 12 g/dL in females, Hb < 13 g/dL in males), 56 (18.1%) polycythemic (Hb ≥ 15 g/dL in females, Hb ≥ 17 g/dL in males), or 207 (67%) normocythemic with median survival 29, 112 and 51 months (p = 0.008). (values determined prior to oxygen therapy).

- Hb ≥ 14.3 g/dL in females, and ≥15.1 g/dL in males independently associated with longer survival.

Pulmonary embolism in chronic hypoxemic patients.

(prospective study)

Ristic et al., Med Glas (Zenica) [126]

- 362 pts with severe COPD exacerbation or raspatory failure, with D-dimer ≥500 µg/l referred for Doppler ultrasound and CT.

- Group 1 (100 pts with erythrocytosis) vs group II (262 without erythrocytosis).

- Higher pulmonary embolism in group 1 vs 2 (39% vs 11.1%)

Prevalence of VTE in COPD with erythrocytosis.

(retrospective case-control study)

Nadeem et al. Clinical and Applied Thrombosis/Hemostasis [101]

- COPD with erythrocytosis, HCT ≥ 50 (n = 86) vs age, sex-matched COPD without erythrocytosis, HCT < 50 (n = 86).

- No difference in VTE ;17 (19.8%) cases vs 12 (14%) controls, p = 0.42

- Trend toward higher incidence of idiopathic VTE in cases (n = 10 vs 4, p = 0.16)

- Similar BMI, smoking, cancer, OSA with differences in hematocrit, oxygen use, pulmonary HTN. Phlebotomy data not provided

Incidence of cardiovascular and thrombotic events in secondary polycythemia.

(retrospective study)

Mao et al. [109]

- Impact of phlebotomy on prevalence of arterial/venous thrombosis in COPD with erythrocytosis (n = 115).

- Thrombosis in 11/35 (31.4%) phlebotomized vs 26/116 (22.4%) in non-phlebotomized (p = 0.28).

- Amongst phlebotomized pts, thrombosis in 4/16(25%) with HCT < 52% vs 7/19(36.8%) in those with higher HCT (p = 0.45)

Secondary polycythemia and perioperative hemorrhage or thrombosis.

(retrospective case-control study

Lubarsky et al. [102]

- COPD with erythrocytosis, Hb > 16 g/dl (16.2–20.1 g/dl) (n = 100) vs age, sex, surgery, ASA physical status-matched controls without erythrocytosis.

- No thrombosis in erythrocytosis group vs 3 events in controls.

- Significantly less transfusion requirement in erythrocytosis group vs controls (p < 0.025).

Symptomatic and pulmonary response to acute phlebotomy in secondary polycythemia.

(Double-blind clinical trial)

Dayton et al. [110]

- 11 pts with hypoxic lung disease associated erythrocytosis, HCT > 54 underwent phlebotomy vs 8 pts as controls (sham procedure).

- 8/11 phlebotomized patients with symptom (dyspnea, fatigue, headache) improvement within 24 h and lasted > 7 days in 5 pts, particularly if HCT > 60 vs no improvement in controls (p < 0.005).

- No objective improvement in airway obstruction, gas exchange or exercise tolerance.

Cyanotic congenital heart disease

Risk of stroke in adults with cyanotic congenital heart disease.

(observational study)

Perloff et al. [104]

- 112 adults with cyanotic heart disease excluding those with independent risk factors for stroke were observed for 748 pt-years.

- 2 groups, (i) compensated (HCT 46–72%, absent/mild symptoms, iron replete with phlebotomy at intervals > 1 yr if symptoms) (n = 101) vs (ii) decompensated (HCT 61.5–75%), iron deficiency, marked/severe symptoms with phlebotomy every 3–6 months (n = 11).

- None of the pts in group (i) had cerebral arterial thrombosis, vs one pt in group (ii) with amaurosis fugax.

Cerebrovascular events in adults with cyanotic congenital heart disease

(retrospective study)

Ammash et al. [104]

- 162 adults with cyanotic congenital heart disease; Group I (n = 140) no history of cerebrovascular event after age 18 years, Group II (n = 22) well documented cerebrovascular event (TIA (n = 19), reversible ischemic neurologic deficit n = 4, infarct, n = 6).

- 46/162 (28.4%) underwent phlebotomy, with increased risk of events after phlebotomy (35/140 in Group I vs 11/22 in Group 2, p = 0.016).

- Strong association of iron deficiency and/or microcytosis with cerebrovascular events (11/41; p = 0.004).

Hydroxyurea for secondary erythrocytosis in cyanotic congenital heart disease.

(case series)

Reiss et al. [113]

- 4 pts with symptomatic secondary erythrocytosis and cyanotic congenital heart disease.

- 2 pts with recent TIA or stroke, one with extreme fatigue/dyspnea and one with extreme exhaustion following phlebotomy.

- Median hydroxyurea use: 15 months; symptomatic improvement, but myelosuppression in 2 pts and thrombocytopenia in 1 pt which required dose reduction. 1 pt with TIA on hydroxyurea. Minimal increase in HbF.

Post-renal transplant erythrocytosis

(8-15%)

Enalapril for erythrocytosis after renal transplant.

(randomized double-blind study)

Beckingham et al. [116]

- 2.5 mg of enalapril daily (n = 15) or placebo (n = 10) for 4 months.

- Hematocrit decreased from 52.7 to 47.1 at 1 month and 46.1 after 4 months with enalapril, no change with placebo (p = 0.004). No change in erythropoietin level.

- No phlebotomy or thrombosis.

Effects of theophylline on erythrocytosis after renal transplant.

(prospective study)

Bakris et al. [118]

- Pts with post-renal transplant erythrocytosis (n = 8) vs normal controls (n = 5).

- 8 weeks of theophylline, EPO significantly reduced in transplant pts (60 to 9 units; p < 0.05) and controls (6.9 to 4.7 units; p < 0.05).

- Hematocrit reduced in transplant pts (0.58 to 0.46; p < 0.05) and controls (0.43–0.39; p < 0.05). Requirement of weekly phlebotomy eliminated.

Comparison of enalapril and losartan on post-renal transplant erythrocytosis.

(prospective randomized study)

Yildiz et al. [117]

- 27 pts treated with enalapril 10 mg/day (n = 15), vs losartan 50 mg/day (n = 12) for 8 weeks.

- Hemoglobin significantly decreased with both losartan (17.1–15.9 g/dl, p = 0.01) and enalapril (17.4–14.9 g/dl, p = 0.001). greater decrease with enalapril (−3.26 vs −1.70, p = 0.05).

- Among the responders who discontinued treatment, there was a trend for longer time to relapse with losartan (7.38 months) compared with enalapril (2.75 months) (p = 0.11).

Testosterone therapy

and erythrocytosis

Prevalence and management of secondary erythrocytosis in transgenders on testosterone.

(retrospective study)

Oakes et al. [122]

- 234 pts, mean pre-testosterone hemoglobin 13.5 g/dL and hematocrit 40.3%. Mean hemoglobin peak 15.7 g/dl, hematocrit 47.2% at an average of 21 months post therapy.

23.5% pts with hematocrit > 50%, and 8.5% hemoglobin > 17.5 g/dL. Only one thrombotic event.

- Dose reduction in 14.5% with erythrocytosis, no phlebotomy.

- 88.9% of patients with erythrocytosis had received testosterone cypionate

Chuvash polycythemia (CP)

Thrombosis in CP.

(prospective study)

Sergueeva et al. [105]

- CP pts matched by age, sex and place of residence (n = 128).

- CP pts with lower blood pressure, body mass index, white count, but more smokers than controls.

- New thrombosis in CP pts 0.031 events/pt/year (34 arterial+venous events vs 3 arterial events in controls), history of phlebotomy but not hematocrit predictor of thrombosis.

- 9 (7.0%) deaths in CP pts (median age; 54 years) compared to 2 (1.6%) controls (median age at death; 81 years) (p = 0.058). All deaths related to thrombosis.

Congenital erythrocytosis

Thrombotic risk in congenital erythrocytosis.

(prospective study)

Gordeuk et al. [106]

- CP pts and matched controls (n = 155).

- 40 thrombotic events in CP at enrollment (n = 27) with 37 new events (n = 33) including 9 fatal events during 11-year observation. Thrombosis in 3 controls at enrollment with 5 new events.

- In multivariate analysis in CP pts, age and past thrombosis but not hematocrit were independent predictors of new events. Phlebotomy associated with increased thrombosis (HR 1.9, p = 0.028). None were on anticoagulation at second event; aspirin 75 mg/day was not protective.

- HIF2A p.M535V variant six-generation pedigree (8 subjects), arterial/venous thrombosis in 5 of 8 vs none in 17 HIF2A wild-type pts (p = 0.001).

- Thrombotic events despite phlebotomy with HCT < 45%.