Abstract
PAH patients reaching a low risk profile with targeted therapies have an excellent long-term survival http://ow.ly/RBjo30i6N57
To the Editor:
In its August 2017 issue, the European Respiratory Journal published two papers on risk assessment in pulmonary arterial hypertension (PAH), one coming from the French registry [1], the other from COMPERA, a European pulmonary hypertension (PH) registry [2]. Both groups utilised abbreviated versions of the risk assessment strategy proposed in the current European PH guidelines [3, 4] and both included basically the same variables, i.e. World Health Organization functional class (FC), 6-min walking distance (6MWD), cardiac index, right atrial pressure and serum levels of brain natriuretic peptide (BNP) or the N-terminal fragment of its propeptide (NT-proBNP). In both studies, the cut-off values used to determine risk were those proposed in the European guidelines [3, 4]. However, the strategy to determine individual risk differed in the two studies: COMPERA calculated the average individual risk by assigning a score of 1, 2 or 3 to each criterion (1: low risk; 2; intermediate risk; 3: high risk) and rounding to the mean of the available variables, as had been previously proposed by a group from Sweden [5]. The French group, in contrast, looked at the number of variables that met the low-risk criteria. Despite these differences, the main results of both studies were similar, especially 1) that the risk stratification tool proposed by the European guidelines adequately predicted mortality, 2) that follow-up risk assessment on treatment was a more reliable predictor of survival than the initial risk assessment at the time of diagnosis, and 3) that FC, 6MWD and BNP/NT-proBNP at follow-up were independent predictors of outcome. However, there also were important differences in the two analyses: in the French study, 19% of the patients with idiopathic PAH (IPAH) met all three noninvasive low-risk criteria at follow-up, i.e. FC I or II, 6MWD >440 m, BNP <50 ng·L−1 or NT-proBNP <300 ng·L−1; these patients had a 1-year survival of 100% and a 5-year survival of 97%. In COMPERA, 21% of the IPAH patients were classified as low risk at follow-up; these patients had a 1-year survival of 96% but the 5-year survival of these patients was only 72% (data available in the online supplement of [2]). We wondered whether the differences in long-term survival were due to the fact that the low-risk definition was less strict with the COMPERA strategy than with the French strategy or whether differences in the two patient cohorts were responsible for these findings (patients in French cohort were younger than the patients in COMPERA and combinations of PAH drugs were used more frequently in the French series than in the COMPERA cohort).
In order to address these questions, we revisited the COMPERA cohort that was analysed for the recent risk assessment paper [2] and applied the French risk assessment strategy [1] to 579 IPAH patients for whom complete data sets of FC, 6MWD and BNP/NT-proBNP were available at first follow-up (median 4.6 months, interquartile range 3.1–9.5 months) after treatment initiation. This cohort represented 77% of the 756 IPAH patients from the original manuscript for whom follow-up data were available. The baseline characteristics and treatment patterns of this subgroup were identical to the entire cohort [2]. The three dichotomised low-risk criteria remained independently associated with survival in a multivariate Cox model adjusted for age and sex. The hazard ratios (95% CI) were as follows. FC I/II: 0.48 (0.30–0.75), p=0.001; 6MWD >440 m: 0.34 (0.16–0.69), p=0.003; NT-proBNP <300 ng·L−1 or BNP <50 ng·L−1: 0.54 (0.33–0.88), p=0.013.
When the French noninvasive risk assessment strategy was applied to the COMPERA cohort, the estimated survival rates at 1, 3 and 5 years of patients meeting all three low-risk criteria at follow-up were 100%, 100% and 95%, respectively. The corresponding survival rates were 99%, 95% and 76% for patients meeting two low-risk criteria; 93%, 75% and 64% for patients meeting one low-risk criterion; and 90%, 61% and 43% for patients who had no low-risk criteria at follow-up (p<0.001 by log-rank test; figure 1). These figures were comparable with the corresponding survival rates observed in the French cohort [1].
In the present analysis, at baseline, three, two, one and no low-risk criteria were present in 3%, 8%, 21% and 69% of the patients, respectively. From baseline to first follow-up, 5% of the patients worsened, 64% remained unchanged and 31% improved their number of low-risk criteria. At first follow-up, three, two, one and no low-risk criteria were found in 9%, 15%, 24% and 52% of the patients, respectively. The corresponding numbers in the French cohort were 19%, 24%, 29% and 28%, respectively. In other words, only 48% of the COMPERA patients reached at least one noninvasive low-risk criterion at first follow-up compared to 72% of the patients in the French registry. Differences in treatment strategies may have contributed to the better long-term survival in the French cohort, where early combinations of PAH drugs were used more frequently than in the COMPERA cohort [1, 2]. However, it is likely that the age differences between the two groups contributed to the survival differences as well. The average age at inclusion was 64±16 years in COMPERA versus 57±17 years in the French cohort. In the present analysis, there was a tight relationship between the likelihood of reaching low-risk criteria and age at time of diagnosis: the mean±sd age of patients meeting all three noninvasive low-risk criteria at follow-up was 45±18 years; the respective age of patients meeting two, one or no low-risk criteria was 57±16, 65±15 and 68±14 years. Still, the observed survival differences between the risk strata remained statistically significant (p<0.001 in Cox regression analysis) when adjusted for age and sex.
In summary, the present data support the notion that the French risk assessment strategy provides a more accurate identification of patients with an excellent long-term survival than the approach of averaging risk scores as originally applied by the Swedish group [5] and the COMPERA investigators [2]. Patients meeting low-risk criteria for FC, 6MWD and BNP/NT-proBNP at first follow-up on therapy had a 5-year survival rate of 97% in the French registry and of 95% in COMPERA. However, only 19% of the French patients and 9% of the COMPERA patients met these criteria at first follow-up. Our results indicate that the likelihood of reaching a low-risk profile declines with increasing age, but it is unclear how much this is attributable to changes in PAH phenotypes, comorbidities or a tendency to use less aggressive PAH treatment strategies in elderly patients with IPAH [6].
Acknowledgements
The authors are indebted to the COMPERA investigators.
Footnotes
Conflict of interest: M.M. Hoeper reports receiving personal fees from Actelion, Bayer, GSK, Pfizer, Gilead and Merck, during the conduct of the study.
Conflict of interest: D. Pittrow reports receiving personal fees from Actelion, Aspen, Bayer, Boehringer Ingelheim, Daiichi Sankyo, Novartis, Pfizer and AstraZeneca, outside the submitted work.
Conflict of interest: C. Opitz reports receving personal fees from Actelion, Bayer, GSK, Pfizer and Novartis, during the conduct of the study.
Conflict of interest: J.S.R. Gibbs reports receiving personal fees from Actelion, Bayer, GSK, Pfizer, MSD, Bellerophon and Arena, during the conduct of the study.
Conflict of interest: S. Rosenkranz reports receving grants and personal fees from Actelion, Bayer, Pfizer, Novartis and United Therapeutics, and personal fees from GSK, Gilead and MSD, during the conduct of the study.
Conflict of interest: E. Grünig reports receiving personal fees from Actelion, Bayer, GSK, Pfizer, United Therapeutics and MSD, outside the submitted work.
Conflict of interest: K.M. Olsson reports receiving personal fees from Actelion, Bayer, GSK, Pfizer and United Therapeutics, during the conduct of the study.
Conflict of interest: D. Huscher reports receiving personal fees from Actelion outside the submitted work.
Support statement: This work was supported by the DZL and the Deutsche Forschungsgemeinschaft (to M.M. Hoeper, grant number HO 1599/2-1). COMPERA is funded by unrestricted grants from Actelion Pharmaceuticals, Bayer and GSK. These companies were not involved in data analysis or the writing of this manuscript. Funding information for this article has been deposited with the Crossref Funder Registry.
- Received December 15, 2017.
- Accepted January 2, 2018.
- Copyright ©ERS 2018