Prognosis in pulmonary arterial hypertension (PAH) has significantly improved after the discovery of some of the pathophysiological bases of the disease and the development of drugs involving these pathways.1 Treatment strategy is no longer based solely on functional class, as it was when first vasodilatory drugs were commercialized. Prognostic stratification and treatment strategy is based now in multiparametric evaluations of several clinical, analytical, functional, and cardiac imaging parameters. Current clinical practice guidelines recommend the combination of variables to discriminate individual prognosis at diagnosis and during follow-up, counselling the intensification of treatment according to the individual prognosis.2 Despite those advances, long-term survival is still far from being perfect, and the patient's response to the available drugs seems to be different. In this setting, prognostic scores are evolving and adapting to the different PAH phenotypes,3 although several gaps remain unsolved to date.

First, the European guidelines of pulmonary hypertension have recently updated the recommendations on prognostic stratification at diagnosis, including some parameters of right ventricle (RV) function.2 This was a long-standing request in the field since the RV function determines symptoms and prognosis. In PAH, the RV adapts to the initial afterload increase by augmenting its contractility and cardiac mass. In case of failure of this adaptative mechanism, dilation of the right heart chambers and systemic congestion appear.4 The tricuspid annular plane systolic excursion (TAPSE) and the systolic pulmonary artery pressure (sPAP) are easily measurable load-dependent parameters of RV contractility and pressure overload, respectively. The TAPSE/sPAP emerges as a parameter of ventriculoarterial coupling, with lower values indicating a maladaptive RV response to the pressure overload, eliminating the influence of the volume overload in this equation. This ratio has been validated in multiple clinical scenarios, improving the mechanistic interpretation of the failing RV in PAH.5 Whilst cardiac magnetic resonance, now considered in the European guidelines,2 measures profoundly the complex RV structure and function, this technique is still used barely in clinical practice. Related with the prior considerations on RV function, some authors defend an aggressive afterload lowering as the main therapeutic target to ameliorate prognosis by improving the RV adaptation to the disease.6 On the other hand, the need for regular risk assessment and the daily clinical practice barriers to implement more complex risk scores – especially the lack of time – was one of the main motivations for introducing simpler risk stratification methods at follow-up. This strategy is based on the adequate applicability of only three widespread parameters (six-minute walk test distance, functional class, and cardiac biomarkers).7 Equally, the REVEAL Lite 2.0 score defend this easier approach, with the sum of other quickly measurable important prognostic parameters: renal function, blood pressure and heart rate.8 Whereas the simplification is now the rule, other risk markers undoubtedly related with prognosis like the diffusing capacity of the lung for carbon monoxide, or the presence of complications associated with a pulmonary artery aneurism are now neglected9,10 (Table 1).

Another important unmet question is the applicability of risk calculators in patients with comorbidities, the predominant PAH population. Recently, an analysis from the COMPERA registry identified that the 4-strata model is also useful in PAH with common cardiovascular risk factors (systemic hypertension, coronary artery disease, diabetes, and obesity), with similar precision for predicting mortality when compared with the remaining PAH population, also showing that the prognosis in the intermediate-low stratum was comparable with the low-risk stratum, hardly achieved in this population.11

Finally, it is important to note that novel molecular findings involved in the PAH pathophysiology could be interesting in terms of prognosis. The addition of the genetic results to the multiparametric evaluation could improve the identification of patients at higher risk strata.12 Concerning proteomics, some investigators have recently identified that a fully proteomic-based model (including SVEP1,PXDN, renin, neuropilin-1, thrombospondin-2, and peroxiredoxin-4) predicts very well the survival free of lung transplantation independent of common risk factors.13 Since these biomarkers relates with some vascular remodelling processes, angiogenesis, or fibrosis, they could predict outcomes before an overt RV failure develops. And others have created a cytokine-based model, both at diagnosis and during follow-up, suggesting the role of the pro-inflammatory hypothesis in the prognosis of this multidimensional disease.14

In conclusion, the prognostic evaluation in pulmonary arterial hypertension is an evolving complex scenario. The aim of this strategy is to treat depending on the individual risk of death. Despite recent important discoveries in terms of treatment of this rare disease,15 it is still necessary to design comprehensive prognostic tools to treat each patient more precisely, introducing novel mechanistic considerations in terms of right ventricle function and molecular discoveries, but also considering the applicability of this tools in clinical practice, as well as other currently neglected risk markers or frequent comorbidities.