Chest
Volume 157, Issue 1, January 2020, Pages 162-172
Journal home page for Chest

Pulmonary and Cardiovascular: Original Research
Retrospective Validation of the REVEAL 2.0 Risk Score With the Australian and New Zealand Pulmonary Hypertension Registry Cohort

https://doi.org/10.1016/j.chest.2019.08.2203Get rights and content

Background

Pulmonary arterial hypertension (PAH) prognosis has improved with targeted therapies; however, the long-term outlook remains poor. Objective multiparametric risk assessment is recommended to identify patients at risk of early morbidity and mortality, and for optimization of treatment. The US Registry to Evaluate Early and Long-Term PAH Disease Management (REVEAL) 2.0 risk score is a new model proposed for the follow-up of patients with PAH but has not been externally validated.

Methods

The REVEAL 2.0 risk score was applied to a mixed prevalent and incident cohort of patients with PAH (n = 1,011) from the Pulmonary Hypertension Society of Australia and New Zealand (PHSANZ) Registry. Kaplan-Meier survival was estimated for each REVEAL 2.0 risk score strata and for a simplified three-category (low, intermediate, and high risk) model. Sensitivity analysis was performed on an incident-only cohort.

Results

The REVEAL 2.0 model effectively discriminated risk in the large external PHSANZ Registry cohort, with a C statistic of 0.74 (both for full eight-tier and three-category models). When applied to incident cases only, the C statistic was 0.73. The three-category REVEAL 2.0 model demonstrated robust separation of 12- and 60-month survival estimates (all risk category comparisons P < .001). Although the full eight-tier REVEAL 2.0 model separated patients at low, intermediate, and high risk, survival estimates overlapped within some of the intermediate- and high-risk strata.

Conclusions

The REVEAL 2.0 risk score was validated in a large external cohort from the PHSANZ Registry. The REVEAL 2.0 model can be applied for risk assessment of patients with PAH at follow-up. The simplified three-category model may be preferred for clinical use and for future comparison with other prognostic models.

Section snippets

REVEAL Registry and REVEAL Models

The REVEAL Registry is a multicenter observational prospective registry that enrolled patients with WHO group 1 PAH at 55 US centers between 2006 and 2009.22,23 The original REVEAL risk model (N = 2,716) was developed to predict 1-year survival of incident and prevalent patients from time of registry enrollment based on 12 variables identified through multivariable Cox regression from the REVEAL cohort.7,9,14 In development of the new REVEAL 2.0 model, only patients who have survived at least 1

Study Population

Of 3,535 patients with pulmonary hypertension enrolled in the PHSANZ Registry, 2,736 adults were classified as WHO group 1 PAH and 1,632 survived at least 1 year postenrollment with follow-up data available for analyses. A further 556 patients (34%) were excluded because of absent right-sided heart catheter data or failure to meet hemodynamic definition of PAH (mean pulmonary arterial pressure ≥ 25 mm Hg and pulmonary arterial wedge pressure ≤ 15 mm Hg) at diagnosis. A total of 1,011 patients

Discussion

Multiple risk assessment tools have been proposed for prediction of mortality risk in PAH.5,9, 10, 11, 12, 13, 14 The original REVEAL risk score model was one of the first to be validated in the era of PAH-targeted therapies.9 The REVEAL 2.0 risk score modified the predictive parameters of REVEAL, and was derived from a baseline at 1 year postenrollment to allow the inclusion of all-cause hospitalization in the prior 6 months as a new parameter.24 Our study represents the first external

Conclusions

In a large external cohort, the REVEAL 2.0 risk model effectively discriminates risk in cohorts of both patients with previously diagnosed and incident PAH, supporting its use in wider populations. The three-category version of the REVEAL 2.0 risk score is particularly appealing because it conforms to the approach advocated by risk prognostication guidelines and may be easier for physicians to apply. Further work is needed to identify the optimal risk prognostication tool in PAH, and should

Acknowledgments

Author contributions: J. J. A. takes responsibility for the content of the manuscript, including the data and analysis. J. J. A., E. M. L., and G. S. conceived and designed the study. J. J. A. analyzed the data and prepared the manuscript with supervision from E. M. L. and G. S. Data were contributed by E. M. L., M. L, D. S. C., N. C., C. C., N. D., J. F., M. H., D. K., A. K., F. K., E. K., T. M., B. R., P. S., V. T., T. W., H. W., and J. P. W. All authors critically revised the manuscript for

References (24)

  • D.S. O'Callaghan et al.

    A critical analysis of survival in pulmonary arterial hypertension

    Eur Respir Rev

    (2012)
  • M. Humbert et al.

    Survival in incident and prevalent cohorts of patients with pulmonary arterial hypertension

    Eur Respir J

    (2010)
  • Cited by (23)

    • The evolving landscape of pulmonary arterial hypertension clinical trials

      2022, The Lancet
      Citation Excerpt :

      Risk scores also correlate with quality of life.110 Many risk scores have been validated retrospectively across global registries4,43,107,111 and in post-hoc analyses of clinical trials.112–114 Exploratory analyses using risk stratification have been performed in randomised control trials such as GRIPHON,15 PATENT-1,26 and AMBITION17 trials to identify subgroups of patients in whom treatment effects were more pronounced.

    • Does community size or commute time affect severity of illness at diagnosis or quality of care in a centralized care model of pulmonary hypertension?

      2021, International Journal of Cardiology
      Citation Excerpt :

      Inhaled therapies are not available in Canada. Recent studies have shown that integrating a combination of known prognostic variables into a risk score provides superior prognostication to using single variables alone [11–13]. Clinical practice guidelines now strongly recommend basing treatment decisions on risk scores [1,14].

    • Blood carbon dioxide tension and risk in pulmonary arterial hypertension

      2020, International Journal of Cardiology
      Citation Excerpt :

      The updated REVEAL 2.0 score was recently introduced and includes a set of 11 different parameters [15]. The prognostic value of the updated score was demonstrated in a registry study (Pulmonary Hypertension Society of Australia and New Zealand Registry) [22]. Integrating PaCO2 as an additional non-invasive criterion has the potential to improve individual risk assessment.

    • Risk assessment in pulmonary arterial hypertension: Insights from the GRIPHON study

      2020, Journal of Heart and Lung Transplantation
      Citation Excerpt :

      The prognostic value of the noninvasive French approach for transplant-free survival has been demonstrated in registry analyses of newly diagnosed patients,5,8 demonstrating the ability of this approach to identify patients at very low risk. The prognostic value of REVEAL 2.0 has been validated in a registry of predominantly prevalent patients.14 The association between risk and outcome has also been investigated using clinical trial data.

    View all citing articles on Scopus

    FUNDING/SUPPORT: Funding support for the Pulmonary Hypertension Society of Australia and New Zealand Registry was provided by Actelion Pharmaceuticals, Allied Healthcare, Bayer, GlaxoSmithKline, Novartis, and Pfizer. GlaxoSmithKline provided research funding support for the primary author.

    View full text