Advertisement
The Journal of Heart and Lung Transplantation
International Society for Heart and Lung Transplantation.
Research Article| Volume 32, ISSUE 3, P293-298, March 2013

Download started.

Ok

Performance of the REVEAL model in WHO Group 2 to 5 pulmonary hypertension: Application beyond pulmonary arterial hypertension

Published:December 20, 2012DOI:https://doi.org/10.1016/j.healun.2012.11.012

      Background

      The majority of patients with pulmonary hypertension (PH) have non-pulmonary arterial hypertension PH (non-PAH PH) or multifactorial PH. The REVEAL score was designed to predict 1-year survival in patients with pulmonary arterial hypertension (PAH) only. It is unknown whether this model is applicable to a more general population of PH patients.

      Methods

      Both newly diagnosed and previously diagnosed patients with PH of any etiology (n = 200) were enrolled in an observational cohort between the years 2003 and 2009. REVEAL scores were assessed for the ability to predict 1-year survival in the following groups: (1) PAH; (2) non-PAH PH; (3) multifactorial PH; and (4) the entire cohort.

      Results

      Of the 200 patients, 126 (63%) had PAH, 32 (16%) had non-PAH PH and 42 (21%) had multifactorial PH. The concordance indices for the model when applied to the various groups were: PAH, 0.72; non-PAH PH, 0.97; multifactorial PH, 0.77; and entire cohort, 0.775. Observed and predicted survivals of the entire cohort according to model-assigned risk strata were not statistically different from one another (p = 0.60), suggesting adequate model calibration.

      Conclusions

      The REVEAL survival prediction model for PAH has comparable performance when applied to a broad population of PH patients. The data suggest that the model may have utility in PH patients in general, subject to validation in a larger cohort.

      Keywords

      Pulmonary hypertension (PH) is a marker of poor prognosis, regardless of etiology.
      • Burrows B.
      • Kettel L.J.
      • Niden A.H.
      • et al.
      Patterns of cardiovascular dysfunction in chronic obstructive lung disease.
      • Grigioni F.
      • Potena L.
      • Galie N.
      • et al.
      Prognostic implications of serial assessments of pulmonary hypertension in severe chronic heart failure.
      • Lettieri C.J.
      • Nathan S.D.
      • Barnett S.D.
      • et al.
      Prevalence and outcomes of pulmonary arterial hypertension in advanced idiopathic pulmonary fibrosis.
      • Oudiz R.J.
      Pulmonary hypertension associated with left-sided heart disease.
      • Traver G.A.
      • Cline M.G.
      • Burrows B.
      Predictors of mortality in chronic obstructive pulmonary disease. A 15-year follow-up study.
      Most studies to date, however, have focused on pulmonary arterial hypertension (PAH), a rare diagnosis.
      • Humbert M.
      • Sitbon O.
      • Chaouat A.
      • et al.
      Pulmonary arterial hypertension in France: results from a national registry.
      In 2009, a task force assessed the state of the research in the broader PH field.
      • Hoeper M.M.
      • Barbera J.A.
      • Channick R.N.
      • et al.
      Diagnosis, assessment, and treatment of non-pulmonary arterial hypertension pulmonary hypertension.
      At this meeting, the term “non-PAH PH” was introduced, encompassing World Health Organization (WHO) Groups 2 through 5; that is, those forms of PH associated with left heart disease, chronic lung disease, chronic thromboembolic pulmonary hypertension (CTEPH) and other diseases. Currently, no survival prediction model exists for these PH groups.
      A contemporary prediction model does exist, however, for PAH. This model assigns a 1-year survival probability and was developed in 2010 from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Disease Management (REVEAL).
      • Benza R.L.
      • Miller D.P.
      • Gomberg-Maitland M.
      • et al.
      Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL).
      This prediction model may aid in selecting patients for advanced therapies,
      • Badagliacca R.
      • Pezzuto B.
      • Poscia R.
      • et al.
      Prognostic factors in severe pulmonary hypertension patients who need parenteral prostanoid therapy: the impact of late referral.
      • Galie N.
      • Manes A.
      • Negro L.
      • et al.
      A meta-analysis of randomized controlled trials in pulmonary arterial hypertension.
      timing of listing for lung transplantation when applicable,
      • Benza R.L.
      • Miller D.P.
      • Frost A.
      • et al.
      Analysis of the lung allocation score estimation of risk of death in patients with pulmonary arterial hypertension using data from the REVEAL Registry.
      and managing patients’ expectations. It is unknown whether this model has predictive ability in patients with other classes of PH.
      We assessed whether the REVEAL prediction model could be applied to a broader group of PH patients, including those with multifactorial PH and non-PAH PH. As the model was designed to assess disease severity, we also investigated whether the model could predict the composite end-point of 1-year survival or freedom lung transplant (FLT).

      Methods

      Study cohort

      The cohort was comprised of 200 patients with pulmonary hypertension who were consecutively enrolled in a University of California, San Francisco (UCSF) genetic study between the years 2003 and 2009. The source of patients for this cohort was the UCSF PH clinic, which screens patients prior to formal consultation for the likelihood of PH. The cohort was overall enriched with PAH patients and included both newly diagnosed and established patients. Informed consent was obtained from all participants and the committee on human research at UCSF approved the study.
      Patients were included in this analysis if they met hemodynamic criteria by right heart catheterization (RHC) of a mean pulmonary arterial pressure of≥25 mm Hg. Subgroups were defined as follows: PAH patients were defined as those with idiopathic or familial PAH, or PAH associated with collagen vascular disease, congenital systemic-to-pulmonary shunts, portal hypertension, drugs or toxins or human immunodeficiency virus (HIV) infection. In addition, the pulmonary capillary wedge pressure (PCWP) had to be≤15 mm Hg. If patients in this group had any other diagnosis believed to contribute to pulmonary hypertension (e.g., pulmonary fibrosis, or anti-phospholipid antibody syndrome with documented pulmonary emboli), then they were placed in the multifactorial group. Non-PAH PH was defined for those patients in WHO Group 2 through 5 PH. Patients were classified as multifactorial pulmonary hypertension if there was more than one potential etiology of pulmonary hypertension (e.g., ejection fraction of 35% and severe chronic obstructive pulmonary disease [COPD] on oxygen therapy, or scleroderma with concomitant pulmonary fibrosis). Two further examples of multifactorial patients with out-of-proportion PH to another diagnosis are described here for further clarity of the classification system. A patient with a mean PA pressure of 50 mm Hg and a forced expiratory volume in 1 second (FEV1) of 65% of predicted (PH out of proportion to the degree of lung disease) would be labeled as having multifactorial PH (Groups 1 and 3) for this analysis. A patient with rheumatoid arthritis, diastolic dysfunction, a pulmonary capillary wedge pressure (PCWP) of 18 mm Hg, a transpulmonary gradient>15 mm Hg and pulmonary vascular resistance>3 Wood units would also be labeled as having multifactorial PH (Groups 1 and 2).

      Baseline variables

      To calculate REVEAL 1-year survival probabilities, the following clinical and demographic criteria data were collected from the time of enrollment: PAH subgroup; age; gender; modified WHO functional class; heart rate and systolic blood pressure; 6-minute walk distance (6MWD); brain natriuretic peptide (BNP) levels; creatinine; height and weight; presence or absence of pericardial effusion by echocardiogram; percent predicted carbon dioxide diffusion capacity (DLco); and RHC data, including right atrial (RA) pressure and pulmonary vascular resistance (PVR). For patients who had multiple diagnostic tests, the test closest to the time of enrollment was chosen. Renal insufficiency was defined as a glomerular filtration rate (GFR) of<60 ml/min/m2 as calculated by the MDRD equation.
      • Giavarina D.
      • Cruz D.N.
      • Soffiati G.
      • et al.
      Comparison of estimated glomerular filtration rate (eGFR) using the MDRD and CKD-EPI equations for CKD screening in a large population.
      Right heart catheterization and pulmonary function data were included in the analysis if performed within 1 year from the time of enrollment, otherwise the data were left as missing.

      Reveal survival probability calculation

      A detailed description of the REVEAL cohort and the methods used to derive the model coefficients have been described previously.
      • Benza R.L.
      • Gomberg-Maitland M.
      • Miller D.P.
      • et al.
      The REVEAL registry risk score calculator in patients newly diagnosed with pulmonary arterial hypertension.
      • McGoon M.D.
      • Krichman A.
      • Farber H.W.
      • et al.
      Design of the REVEAL registry for US patients with pulmonary arterial hypertension.
      The cohort was comprised of 2,716 patients with both newly diagnosed and previously diagnosed PAH. Variables in the model that were associated with a poor prognosis included: PAH associated with portal hypertension; PAH associated with connective tissue disease (CTD); familial PAH; male gender; age>60 years; WHO Functional Class III or IV; renal insufficiency; resting systolic blood pressure<110 mm Hg; heart rate>92 beats/min; mean RA pressure>20 mm Hg; 6MWD<165 m; brain natriuretic peptide (BNP)>180 pg/ml; PVR>32 Wood units; percent predicted DLco≤32%; and presence of a pericardial effusion by echocardiogram. Four variables were associated with an increased 1-year survival: modified WHO Functional Class I; 6MWD≥440 m; BNP<50 pg/ml; and percent predicted carbon monoxide diffusing capacity≥80%. An individual’s predicted survival from the model is calculated by:
      S0(1)exp(Zβγ)


      where S0(1) = 0.9698, γ = 0.939 and Z′β is the sum of the patient’s individual characteristics multiplied by the β coefficients for each of the 19 parameters in the model.
      • Benza R.L.
      • Miller D.P.
      • Gomberg-Maitland M.
      • et al.
      Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL).
      Missing data were handled as a “0” for each missing variable (treated as though the predictor was absent), as was done in the derivation cohort. It should be noted that, for the non-PAH patients, the PAH subtype was left as “0” or missing, as this part of the score did not apply. The maximum number of negative predictors available, therefore, was different in the non-PAH PH group (12 as compared with 15 in the PAH and multifactorial PH groups).

      Outcome data

      Survival. Vital status and date of death (all-cause mortality) was obtained using the Social Security Death Index. Patients who underwent lung or liver transplant during follow-up were censored at that time of transplant, which was obtained by chart review.
      Losses to follow-up. Although vital statistics should be accurate for all patients enrolled, we could be certain that patients who were eligible for lung transplant but lost to follow-up were not transplanted at other centers. For both outcomes (survival or composite survival or FLT), transplantation at another center represents a competing risk in this analysis. For this reason, patients who were transplant-eligible (defined as age<65 years, on infusion or inhaled therapies with continued WHO Functional Class III or IV symptoms) and were lost to follow-up before the 1 end-point were censored at the time of the of the last UCSF clinic visit or date of the last hospital discharge. All other losses were assumed as alive if not registered by the death index.

      Statistical analysis

      All statistical analyses were performed using STATA, version 12. Continuous variables are presented as mean±SD. Analysis of variance (ANOVA) was used to compare continuous variables between the subgroups, when appropriate. Categorical variables are presented as number and percent and comparisons and were obtained using the chi-square test.
      Survival estimates of the entire cohort and the cohort stratified by subgroup were obtained using the Kaplan–Meier method. Survival curves for 1- and 5-year follow-up were compared using the log-rank test.
      Discrimination of the REVEAL model was assessed using the concordance (c)-index (c-index). The c-index is a mathematical measure of area under the receiver operating curve (ROC), which can range from 0 to 1, with 1 representing perfect discrimination and 0.5 reflecting discrimination that would be achieved by chance alone.
      To assess REVEAL model calibration of the entire cohort, patients were assigned to one of five risk strata (low, average, moderately high, high, very high risk) based on the calculated 1-year survival probabilities from the model. The observed and expected survivals were then compared with a Hosmer–Lemeshow goodness-of-fit test.
      For all analyses, p<0.05 was considered statistically significant.

      Results

      Clinical characteristics

      The cohort included 200 patients with pulmonary hypertension who were enrolled between July 2003 and December 2008. The baseline characteristics of the cohort by subgroup are displayed in Table 1. The PAH patients had a higher proportion of women, were younger overall, and had a higher PA pressure and pulmonary vascular resistance than the non-PAH PH and multifactorial PH patients. The non-PAH PH group and multifactorial PH patients had a lower DLco than the PAH patients.
      Table 1Baseline Characteristics of the Cohort by Subgroup
      CharacteristicPAHNon-PAH PHMultifactorial PHp-value
      n1263242
      Age (years)46.7±12.361.0±13.661.4±12.9<0.001
      Female89 (70.6)20 (62.5)13 (31.0)0.67
      Race
       White88 (69.8)23 (71.9)24 (57.1)0.11
       Black4 (3.2)5 (15.6)4 (9.5)
       Asian or Pacific Islander14 (11.1)1 (3.1)3 (7.1)
       Other20 (15.9)3 (9.4)11 (26.2)
      Ethnicity
       Hispanic14 (11.1)3(9.4)5 (11.9)
      Methamphetamine27 (21.4)6 (14.3)<0.05
      HIV11 (8.7)1 (2.4)0.09
      WHO functional class
       I8 (6.4)8 (6.4)00.17
       II62 (44.3)3 (9.4)25 (59.5)0.08
       III42 (33.3)13 (40.6)19 ( 45.2)0.38
       IV8 (6.4)6 (18.8)3 (7.1)0.08
       Indeterminate/missing13 (9.3)2 (6.3)3 (7.1)
      RA mean (mm Hg)10.7±6.58.8±4.99.9±5.60.38
      PA mean (mm Hg)50.5±14.534.7±9.043.8±6<0.001
      PVR (Wood units)11.6±6.34.6±2.96.3±3.9<0.001
      Serum BNP (pg/ml)178±204268±301191±263
      Renal insufficiency24 (19.1)12 (37.5)9 (21.4)<0.05
      HR (beats/min)79±1774±1279±160.14
      SBP (mm Hg)118±17125±18124±200.33
      Pericardial effusion25 (19.8)6 (18.8)12 (28.6)0.40
      DLco (% predicted)62±1545±2347±21<0.001
      Values are expressed as mean±SD or number (%). BNP, brain natriuretic peptide; DLco, diffusion capacity; HIV, human immunodeficiency virus; HR, heart rate; non-PAH PH, non-pulmonary arterial hypertension pulmonary hypertension; PA, pulmonary arterial; PAH, pulmonary arterial hypertension; PH, pulmonary hypertension; PVR, pulmonary vascular resistance; RA, right atrial; SBP, systolic blood pressure; WHO, World Health Organization.
      Subgroups are listed in Table 2. Of the total 200 patients, there were 126 PAH, 32 non-PAH PH and 42 multifactorial PH patients. There was a large proportion with methamphetamine- (19.3%), HIV- (8.7%) and portopulmonary-associated (16.4%) PAH compared with other published series.
      • Benza R.L.
      • Miller D.P.
      • Gomberg-Maitland M.
      • et al.
      Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL).
      • Kane G.C.
      • Maradit-Kremers H.
      • Slusser J.P.
      • et al.
      Integration of clinical and hemodynamic parameters in the prediction of long-term survival in patients with pulmonary arterial hypertension.
      Among the non-PAH PH patients, 8 (25%) cases were due to left-sided heart disease and 18 (56.3%) were due to underlying lung disease.
      Table 2Etiologies of Pulmonary Hypertension by Subgroup
      Type of PHNumber (%)
      PAH (n = 126)
       Idiopathic PAH77 (55.0)
       Familial PAH2 (1.4)
       Associated PAH
        Connective tissue disease38 (27.1)
        Methamphetamine27 (19.3)
        HIV11 (8.7)
        Portopulmonary23 (16.4)
        Congenital heart disease3 (2.4)
      Non-PAH PH (n = 32)
       Group 2 (left-sided heart disease)8 (25.0)
       Group 3 (lung disease)18 (56.3)
       Group 4 (thromboembolic disease)3 (16.7)
       Group 5 (miscellaneous)3 (16.7)
      Multifactorial PH (n = 42)
       Groups 1 and 320 (47.6)
       Groups 1 and 29 (21.4)
       Groups 2 and 37 (16.7)
       More than 3 contributing etiologies6 (14.3)
      Values are expressed as number and percentage within each subgroup. HIV, human immunodeficiency virus; PAH, pulmonary arterial hypertension; non-PAH PH, non-pulmonary arterial hypertension pulmonary hypertension.
      The mean length of follow-up for the entire cohort was 3.6 years (±2.6) and 4.1 years (±2.7) for survivors. During the first year of follow-up there were 16 deaths (8.0%) and 8 lung transplants (4.0%). Of lung transplant-eligible patients, there were 2 patients (1.0%) lost to follow-up over the first year. As stated in the Methods, these patients were censored at the time of the last clinic visit or hospitalization.
      Of the 200 patients in the cohort, 161 (80.5%) had right heart catheterization data, 194 (97.0%) had echocardiographic data, 138 (69.0%) had BNP measurements, and 158 (79.0%) had creatinine measured within 1 month of enrollment. The average number of days from right heart catheterization to enrollment was 116±110. Although all patients had pulmonary function tests as part of their initial evaluation of PH, DLco data were used for REVEAL score calculation only if performed within 1 year of enrollment (99 patients, 49.5%).

      Model discrimination

      The ROC curves for the REVEAL model to predict 1-year survival by patient subgroup are displayed in Figure 1. The c-index for PAH patients was 0.677 (95% CI 0.421 to 0.932), 0.966 for non-PAH PH (0.898 to 1.0) and 0.835 (0.694 to 0.976) for multifactorial patients. This was compared with 0.744, which was the optimum bias-corrected c-index reported in the original cohort. The model was then assessed as a predictor of composite 1-year survival or freedom from lung transplant. The resultant c-indices by subgroup are displayed in Figure 2. For the combined end-point, the c-index for PAH patients was 0.760 (95% CI 0.586 to 0.934), 0.931 (0.826 to 1.0) for non-PAH PH and 0.778 (0.609 to 0.947) for multifactorial patients.
      Figure thumbnail gr1
      Figure 1REVEAL model discrimination as determined by a receiver operator curve analysis for 1-year survival by subgroup. PAH, pulmonary arterial hypertension; non-PAH PH, non-pulmonary arterial hypertension pulmonary hypertension; PH, pulmonary hypertension.
      Figure thumbnail gr2
      Figure 2Comparison of the REVEAL model c-indices is shown by patient group for 1-year survival or combined 1-year survival or freedom from lung transplant. FLT, freedom from lung transplant; PAH, pulmonary arterial hypertension; non-PAH PH, non-pulmonary arterial hypertension pulmonary hypertension; PH, pulmonary hypertension.

      Model calibration

      The calculated 1-year survival probabilities from the entire cohort were then used to separate patients into one of five risk groups, as previously described (low risk:≥95% 1-year survival; average: 90% to<95%; moderate: 85% to<90%; high: 70% to<85%; very high:<70%). The observed survivals were obtained from the Kaplan–Meier method and the predicted vs observed were not statistically different according to the Hosmer–Lemeshow goodness-of-fit test (p = 0.60; Figure 3). Model calibration was not assessed by subgroup given the small number of patients per strata in the non-PAH PH and multifactorial subgroups.
      Figure thumbnail gr3
      Figure 3Predicted vs observed survival at 1 year for the entire cohort. Comparison of the observed 1-year survival to predicted survival by the model according to model assigned risk group.

      Survival of the cohort

      The calculated overall survival of the cohort was 91.3% at 1 year, 63.8% at 5 years and 39.7% at 8 years of follow-up. The 1-year survival by subgroup was 94.1% for PAH, 93.3% for non-PAH PH and 81.4% for multifactorial PH, which trended toward a statistically significant difference between the groups (p = 0.065). At 5 years, survival by subgroup was 68.6% for PAH, 64.2% for non-PAH PH and 46.2% for multifactorial PH patients (p = 0.053; Figure 4).
      Figure thumbnail gr4
      Figure 4Five-year Kaplan–Meier survival curves of patients in the validation cohort stratified by patient subgroup. PAH, pulmonary arterial hypertension; non-PAH PH, non-pulmonary arterial hypertension pulmonary hypertension; PH, pulmonary hypertension.

      Discussion

      Our results show that the REVEAL survival prediction model for PAH had comparable performance in both discrimination and calibration when applied to a broad group of PH patients, including those with non-PAH PH and multifactorial PH. This may expand the clinical utility of this model. As this model was designed to predict disease severity, lung transplant was included as outcome. This decision was based on the fact that patients who underwent lung transplant would most likely not have been long-term survivors, and the time to transplant was an easily defined outcome that could be assessed. The model discrimination for this combined outcome was similar to the model discrimination for survival alone.
      One explanation as to why this model performed well in the broader PH population is that there are only a few variables in the original model that are specific to PAH. Virtually all negative predictive variables included in the model, such as elevations in BNP, RA pressure and reduced WHO functional class, have been shown to be markers of poor prognosis in heart failure from other etiologies.
      • Campana C.
      • Pasotti M.
      • Klersy C.
      • et al.
      Baseline and 6-month B-type natriuretic peptide changes are independent predictors of events in patients with advanced heart failure awaiting cardiac transplantation.
      • Damman K.
      • Voors A.A.
      • Hillege H.L.
      • et al.
      Congestion in chronic systolic heart failure is related to renal dysfunction and increased mortality.
      • Levy W.C.
      • Mozaffarian D.
      • Linker D.T.
      • et al.
      The Seattle Heart Failure Model: prediction of survival in heart failure.
      The fact that the model may not be specific for pulmonary vascular disease is both a strength and weakness of the model.
      With regard to REVEAL model performance in PAH patients, Figure 1 shows that the model discrimination was poor in this population, which is the group for which the model was designed. The model tended to underestimate the risk among patients with higher predicted survival and to overestimate the risk among patients with intermediate predicted survival. The small sample size of the PAH cohort may have contributed to this result, which is suggested by the fact that model performance improved when the event rate increased in the combined death or lung transplant outcome.
      Although the REVEAL model has added an important clinical tool for PAH patients, there are also limitations to the model that are suggested with the present data. Some of the decisions made to make the REVEAL model usable may have been at the cost of model performance. For example, the dichotomization of all continuous predictors in the model (RA mean>20 mm Hg or PVR≥32 Wood units), which allowed for the conversion of the model into an easily calculated score, likely led to loss of information. This point is underscored by a recent re-derivation of the NIH survival equation for idiopathic, familial and anorexigen-associated PAH, which used only three hemodynamic variables (RA pressure, mean PA pressure and cardiac index, all in continuous form), and demonstrated reasonable calibration in a contemporary validation set.
      • Thenappan T.
      • Shah S.J.
      • Rich S.
      • et al.
      Survival in pulmonary arterial hypertension: a reappraisal of the NIH risk stratification equation.
      In addition, although the REVEAL model is designed to handle missing data, the inclusion of 19 predictors in the model, many of which are not measured on routine follow-up,
      • McLaughlin V.V.
      • Archer S.L.
      • Badesch D.B.
      • et al.
      ACCF/AHA 2009 expert consensus document on pulmonary hypertension a report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents and the American Heart Association developed in collaboration with the American College of Chest Physicians; American Thoracic Society, Inc.; and the Pulmonary Hypertension Association.
      may lead to patient misclassification. In fact, in the original publication, the c-index fell to 0.69 when applied to patients with≤13 predictors measured.
      • Benza R.L.
      • Miller D.P.
      • Gomberg-Maitland M.
      • et al.
      Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL).
      Despite these potential limitations of the current REVEAL model, the model has been proven in several validation sets and has also recently been shown to hold prognostic information when calculated serially. With the data from this analysis, it appears that there may be a role for use of the model in the broader PH population, although this will need to be validated in a larger data set.
      Our study has several limitations. We assessed a small, single-center cohort, and the confidence intervals around the estimates are wide, reflective of the small sample size. The non-PAH and multifactorial PH patients retained and managed at the UCSF PH clinic likely represent a more complex cohort than the larger population of non-PAH PH and multifactorial PH patients. This may impact the generalizability of our results. The c-index for the non-PAH PH group should be interpreted with caution because this group was small and there were only 2 deaths over 1 year. Model calibration was not assessed by subgroup given the small number of patients that would be assigned to each strata in the non-PAH PH and multifactorial PH groups. Finally, although the inclusion of lung transplant as an outcome did increase the event rate in this small cohort, the results should be interpreted with caution because the timing of this end-point can vary.
      In conclusion, the REVEAL survival prediction model for PAH has comparable performance when applied to a broad population of PH patients. These data provide preliminary support that the model can be used in PH patients in general, although this remains to be validated in a larger cohort.

      Disclosure statement

      R.S. receives grant support from the NIH; D.M. serves as a consultant for Actelion, Gilead and United Therapeutics, and receives grant support from United Therapeutics; and T.D. serves as a consultant for Actelion, Gilead and United Therapeutics, and receives grant support from Novartis and United Therapeutics.
      The remaining authors have no conflicts of interest to disclose.

      References

        • Burrows B.
        • Kettel L.J.
        • Niden A.H.
        • et al.
        Patterns of cardiovascular dysfunction in chronic obstructive lung disease.
        N Engl J Med. 1972; 286: 912-918
        • Grigioni F.
        • Potena L.
        • Galie N.
        • et al.
        Prognostic implications of serial assessments of pulmonary hypertension in severe chronic heart failure.
        J Heart Lung Transplant. 2006; 25: 1241-1246
        • Lettieri C.J.
        • Nathan S.D.
        • Barnett S.D.
        • et al.
        Prevalence and outcomes of pulmonary arterial hypertension in advanced idiopathic pulmonary fibrosis.
        Chest. 2006; 129: 746-752
        • Oudiz R.J.
        Pulmonary hypertension associated with left-sided heart disease.
        Clin Chest Med. 2007; 28: 233-241
        • Traver G.A.
        • Cline M.G.
        • Burrows B.
        Predictors of mortality in chronic obstructive pulmonary disease. A 15-year follow-up study.
        Am Rev Respir Dis. 1979; 119: 895-902
        • Humbert M.
        • Sitbon O.
        • Chaouat A.
        • et al.
        Pulmonary arterial hypertension in France: results from a national registry.
        Am J Respir Crit Care Med. 2006; 173: 1023-1030
        • Hoeper M.M.
        • Barbera J.A.
        • Channick R.N.
        • et al.
        Diagnosis, assessment, and treatment of non-pulmonary arterial hypertension pulmonary hypertension.
        J Am Coll Cardiol. 2009; 54: S85-S96
        • Benza R.L.
        • Miller D.P.
        • Gomberg-Maitland M.
        • et al.
        Predicting survival in pulmonary arterial hypertension: insights from the Registry to Evaluate Early and Long-Term Pulmonary Arterial Hypertension Disease Management (REVEAL).
        Circulation. 2010; 122: 164-172
        • Badagliacca R.
        • Pezzuto B.
        • Poscia R.
        • et al.
        Prognostic factors in severe pulmonary hypertension patients who need parenteral prostanoid therapy: the impact of late referral.
        J Heart Lung Transplant. 2012; 31: 364-372
        • Galie N.
        • Manes A.
        • Negro L.
        • et al.
        A meta-analysis of randomized controlled trials in pulmonary arterial hypertension.
        Eur Heart J. 2009; 30: 394-403
        • Benza R.L.
        • Miller D.P.
        • Frost A.
        • et al.
        Analysis of the lung allocation score estimation of risk of death in patients with pulmonary arterial hypertension using data from the REVEAL Registry.
        Transplantation. 2010; 90: 298-305
        • Giavarina D.
        • Cruz D.N.
        • Soffiati G.
        • et al.
        Comparison of estimated glomerular filtration rate (eGFR) using the MDRD and CKD-EPI equations for CKD screening in a large population.
        Clin Nephrol. 2010; 74: 358-363
        • Benza R.L.
        • Gomberg-Maitland M.
        • Miller D.P.
        • et al.
        The REVEAL registry risk score calculator in patients newly diagnosed with pulmonary arterial hypertension.
        Chest. 2012; 141: 354-362
        • McGoon M.D.
        • Krichman A.
        • Farber H.W.
        • et al.
        Design of the REVEAL registry for US patients with pulmonary arterial hypertension.
        Mayo Clin Proc. 2008; 83: 923-931
        • Kane G.C.
        • Maradit-Kremers H.
        • Slusser J.P.
        • et al.
        Integration of clinical and hemodynamic parameters in the prediction of long-term survival in patients with pulmonary arterial hypertension.
        Chest. 2011; 139: 1285-1293
        • Campana C.
        • Pasotti M.
        • Klersy C.
        • et al.
        Baseline and 6-month B-type natriuretic peptide changes are independent predictors of events in patients with advanced heart failure awaiting cardiac transplantation.
        J Cardiovasc Med (Hagerstown). 2009; 10: 671-676
        • Damman K.
        • Voors A.A.
        • Hillege H.L.
        • et al.
        Congestion in chronic systolic heart failure is related to renal dysfunction and increased mortality.
        Eur J Heart Fail. 2010; 12: 974-982
        • Levy W.C.
        • Mozaffarian D.
        • Linker D.T.
        • et al.
        The Seattle Heart Failure Model: prediction of survival in heart failure.
        Circulation. 2006; 113: 1424-1433
        • Thenappan T.
        • Shah S.J.
        • Rich S.
        • et al.
        Survival in pulmonary arterial hypertension: a reappraisal of the NIH risk stratification equation.
        Eur Respir J. 2010; 35: 1079-1087
        • McLaughlin V.V.
        • Archer S.L.
        • Badesch D.B.
        • et al.
        ACCF/AHA 2009 expert consensus document on pulmonary hypertension a report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents and the American Heart Association developed in collaboration with the American College of Chest Physicians; American Thoracic Society, Inc.; and the Pulmonary Hypertension Association.
        J Am Coll Cardiol. 2009; 53: 1573-1619