Early, remote evaluation of therapeutic efficacy and disease worsening in patients with pulmonary arterial hypertension

Background:
In patients with pulmonary arterial hypertension (PAH), risk stratification is used to aid clinical decision making and guide treatment, categorising patients as low- (<5%); low-intermediate (5-10%); high-intermediate (10-20%); or high (>20%) at 1-year. The parameters used for risk stratification require hospital-based assessment. I therefore sought to identify mortality-associated parameters that may be remote monitored to facilitate early evaluation of clinical efficacy following treatment escalation (TE) or indicate clinical deterioration, thereby permitting early intervention.

Methods:
Patients with all forms of pulmonary hypertension (PH) were identified from the ASPIRE (6/YH/0352) registry. Univariate and multivariate Cox Regression analysis were undertaken to identify parameters that may be remote monitored that were also associated with mortality. Mortality weighted z-scores of age, incremental shuttle walk test (ISWT), heart rate reserve (HRR) and total pulmonary resistance (TPR) were summed to give an individual physiological risk score value. The relationship between remote-monitored physiology in the time preceding and following clinically-indicated TE and clinical worsening events (CWE) was also examined.

Results:
Age, ISWT, HRR and TPR had statistically significant relationships to mortality. Survival analysis showed increased mortality with each decile of baseline risk score (OR5.5,95%CI4.3-7.0, each decile p<0.01). Comparison of the ASPIRE registry PAH group stratified by COMPERA 2.0 risk model and LOESS-derived thresholds for physiological risk were well-matched at 1,3 and 5 years, Cohens-weighted Kappa score showed good agreement (0.61). Following TE, remote monitored mean PAP and TPR were reduced, and CO and physical activity increased compared to baseline (p<0.05), remote physiological risk score also improved compared to control (p<0.0001). There was an increase in remote monitored mean PAP and TPR and reduced CO and physical activity at 10 days prior to a CWE (p<0.05). Following treatment with an experimental therapy remote monitoring parameters and physiological risk score identified change in advance of the traditional 6-month clinical study endpoint used for putative anti-remodelling therapies.

Conclusion:
A physiological risk score accurately categorised patients as low, intermediate, or high risk. Remote risk scores were found to deteriorate pre-CWE and improve post-clinically-indicated TE highlighting that a remote physiological risk score may be applied to provide real-time risk stratification.