J.V. Scott, L.C. Lohmueller, Kraisangka J, M.K. Kanwar , R.L. Benza. Chest, Volume 156, Issue 4, Supplement, 2019, Pages A1173-A1174, ISSN0012-3692
PURPOSE: Invasive hemodynamics are a valuable component of risk assessment in pulmonary arterial hypertension (PAH). Although several commonly obtained variables are highly associated with survival, the full breath of alternative hemodynamic assessments (HA) and their prognostic implications may be underutilized. We assessed both the statistical and clinician determined importance of these alternative HA along with standard HA to survival in a large cohort of PAH patient who participated in several clinical trials.
METHODS: Ten PH clinical experts were asked to share their ‘clinical gestalt’ about hemodynamic variables they felt had an impact on PAH patient survival, based on their experience. Concurrently, Cox regression analyses was performed on 2802 PAH patients for baseline hemodynamic parameters from 5 clinical trials (SERAPHIN, BREATHE-1, PATENT, CHEST, and the pivotal subcutaneous treprostinil trial). The p-values of each hemodynamic variable were aggregated across the clinical trials using Stouffer’s method, which weights each p-value by the number of patients used in the derivation.
RESULTS: 26 hemodynamic variables were considered in this analysis, including 14 mentioned by clinical experts (Table 1.) The parameters most mentioned by clinical experts were cardiac index, right atrial pressure, and mean pulmonary artery pressure (aggregated p-value 0.0007, 0.0003, and 0.073 respectively). However, the parameters with the most significant aggregated p-value based on clinical trial data were stroke volume, stroke work index, and stroke work (all < 0.000).
[Expert mentions], Hemodynamic Variable (n), p-value[1], Acute Hemodynamic index (1200) p=0.35
[1], Cardiac efficiency (2476) p<0.0001
[10], Cardiac index (2551) p=0.0006
[1], Cardiac power output (3818) p<0.0001
[0], Diastolic BP (2757) p=0.27
[1], Diastolic PAP (2604) p=0.003
[7], Mean PAP (2573) p=0.07
[1], PA compliance (642) p=0.03
[1], PA elastance (2473) p < 0.0001
[1], PCWP (2497) p=0.11
[6], PVR (2486) p<0.0001
[7], RAP (2544) p=0.0003
[2], RVSW and RVSWI (1202) p=0.05
[2], Stroke volume and SVI (2479) p<0.0001
[1], Stroke work (2473) p<0.0001
[0], Systolic PAP(2552) p= 0.29
[3], TPR (2553) p < 0.0001
CONCLUSIONS: This analysis presents the most comprehensive assessment of baseline hemodynamic variables for predicting survival in PAH patients. Clinical gestalt alone may underestimate evidence-based hemodynamics in predicting PAH patient mortality risk. Therefore, the role of methodical risk stratification based on evidence-based risk assessment cannot be overemphasized. This analysis will be used to select features for derivation of a machine-learned PAH patient risk stratification tool.
CLINICAL IMPLICATIONS: Collection of hemodynamic patient data beyond the may be important for assessing patient risk. This analysis will be used to select features for derivation of machine-learned PAH patient risk stratification models.