Cardiopulmonary exercise testing for early detection of pulmonary hypertension in primary myelofibrosis

Pulmonary Hypertension (PH) is a frequently under-recognized complication of primary myelofibrosis (PMF); cardiopulmonary exercise testing (CPET) proved to be an essential and non-invasive diagnostic method in PH. We aimed to identify CPET parameters useful to earlier PH detection in PMF. Patients (pts) with diagnosis of PMF (acc. WHO 2008), age >18 years and fit to perform CPET were included. Main exclusion criteria were pulmonary or cardiac disease, portal hypertension and clinical history of PH. At enrollment (t0) pts underwent echocardiography, CPET, global spirometry and diffusing capacity of the lung for carbon monoxide (DLCO) analysis; echocardiographic assessment was repeated after 6 months (t1). During CPET we monitored ECG and beat-to-beat arterial pressure; we also used ventilatory expired-gas and transcutaneous PCO2/PO2 analyzer. Statistics was performed using T-test unpaired, T-test paired and Pearson’s correlation. Twenty-three pts were compared to age and gender-matched controls. Median age was 68 years; 10 pts carried JAK2 V617F mutation, 11 a CARL mutation (7 type-1, 4 type-2), 2 were triple-negative. DIPSS was low or intermediate-1 in 9 and 14 cases, respectively. Fourteen pts were receiving hydroxyurea and 17 antiplatelets; 18 pts had a prefibrotic-stage MF, while 5 had grade 2-3 bone marrow fibrosis. During CPET PMF pts compared to controls demonstrated raised minute ventilation carbon dioxide production relationship at anaerobic threshold (AT) (38.3+3.3 vs 28.5+2.9; p<0.05); high PaCO2 at rest (44.2+2.1 vs 39.4+1.8 mmHg; p<0.02) and high difference between artery and tele-expiratory CO2-pressure [dP(a-et)CO2] at AT (+4+1.5 vs -5+1.8 mmHg; p<0.02). By echocardiography PMF showed slightly elevated pulmonary arterial pressures (PAPs) (34+4.2 vs 23+4.5 mmHg; p<0.05), short pulmonary valve acceleration time (PVAccTime) (88+11 vs 122+9 msec; p<0.01), mildly increased left atrial volume (42+7 vs 29+5 ml/mq, p<0.02). Global spirometry and DLCO analysis were normal. At t1 PMF demonstrated a significant increase of PAPs and reduction of PVAccTime when compared with t0 (37+ 5.6 vs 34+4.2 mmHg; 80+14 vs 88+11 msec; p<0.05). A positive relationship between dP(a-et)CO2 and both dPAPs and dPVAccTime was identified (Rsquare 0.45 and 0.39; p 0.025 and 0.04). This study defines dP(aet) CO2 as an early predictor of pulmonary circulation impairment, useful for early identification of PMF most at risk of developing clinically significant PH.