Could the Cochin risk prediction score be applied in daily practice to predict pulmonary hypertension in systemic sclerosis? Comment on the article by Meune et al


Could the Cochin Risk Prediction Score be Applied in Daily Practice to Predict Pulmonary Hypertension in Systemic Sclerosis? Comment on the Article by Meune et al

To the Editor:

In a recent article in Arthritis & Rheumatism, Meune et al (1) proposed a score to estimate the likelihood of precapillary pulmonary hypertension (PH) developing in patients with systemic sclerosis (SSc). It was proposed that age, diffusing capacity for carbon monoxide/alveolar volume (DLCO/VA), and forced vital capacity (FVC) were independent risk factors for developing precapillary PH within 3 years. The proposed “Cochin” risk prediction score (RPS) is calculated as follows: RPS = 0.0001107(age) + 0.0207818(150 − FVC) + 0.04905(100 − DLCO/VA). According to this formula, the older the patient and the lower the FVC and DLCO/VA, the higher the risk of developing precapillary PH. Meune and colleagues applied this formula in a validation sample of 443 consecutive patients presenting with “PH-free” SSc at baseline, among whom 17 developed precapillary PH verified by right-sided heart catheterization. Meune et al reported that the patients in the fifth quintile of the Cochin RPS had a 35.96-fold risk of developing precapillary PH compared with patients in the first quintile. Moreover, a cutoff value of 2.73 had 89.5% sensitivity and 74.1% specificity for the detection of patients developing precapillary PH during followup.

We applied the Cochin RPS in the independent Itinerair-Sclérodermie cohort of patients recruited between September 2002 and July 2003 (2, 3). Briefly, an algorithm for the detection of pulmonary arterial hypertension (PAH), based on the presence of dyspnea and findings of Doppler echocardiographic evaluation of the velocity of tricuspid regurgitation (VTR), was applied to 384 patients without evidence of PAH at the first visit. Patients with a VTR ranging from 2.8 meters/second to 3.0 meters/second with unexplained dyspnea or patients with a VTR of >3.0 meters/second underwent right-sided heart catheterization. Of note, we excluded SSc patients with known severe lung fibrosis (FVC or total lung capacity <60% of predicted), which poses a higher risk of development of PH, and patients with known severe cardiac disease, who have a higher risk of developing postcapillary PH. The characteristics of our population did not differ from those of the population studied by Meune et al with regard to age (mean 53 years versus 55 years), sex (87% female versus 81% female), antitopoisomerase antibody positivity (28% versus 25%), and anticentromere antibody positivity (47% versus 41%). Over a mean ± SD followup period of 41.03 ± 5.66 months, PAH was diagnosed in 8 patients, postcapillary PH was diagnosed in 8 patients, and precapillary PH was diagnosed in 2 patients due to lung fibrosis (incidence of precapillary PH was 0.76 per 100 patient-years). Interestingly, the incidence of PH due to lung fibrosis was similar to that reported by Meune et al.

We used the formula and the classification thresholds for PH risk proposed by Meune and colleagues to analyze our cohort. Among 346 patients (missing data did not allow for the calculation in 38 patients), 207 patients (59.83%) were at low risk, 108 (31.21%) were at intermediate risk, and 31 (8.96%) were at high risk of developing PH during the followup. Of the 8 patients diagnosed as having PAH during followup, 5 were classified as high risk, 1 as intermediate risk, and 2 as low risk. Of the 2 patients with PH associated with severe fibrosis, 1 was classified as intermediate risk and the other as low risk.

Moreover, among the factors age, FVC, and DLCO/ VA, only the latter was significantly associated with PAH (P = 0.01), presumably because we excluded patients with severe lung fibrosis at baseline. Applying the proposed cutoff value of 2.73 in our cohort, we found 50.0% sensitivity and 86.0% specificity, yielding a 9.6% positive predictive value and a 98% negative predictive value to predict precapillary PH. The likelihood ratio allows us to calculate pretest and posttest probability of precapillary PH using Bayes theorem. For 346 patients, the 3-year incidence of precapillary PH was 2.9% in the Itinerair-Sclérodermie cohort and 3.9% in the Meune et al cohort. These values can be considered as the pretest probability of precapillary PH in these populations. Knowing the results of the Cochin RPS would lead to a posttest probability of 9.64% in the Itinerair-Sclérodermie cohort and 12.66% in the Meune cohort. Clearly, knowing the results of the Cochin RPS had a modest but not meaningless effect in both cohorts. Being classified in the low-risk subset did not exclude the risk of PH occurrence within a 3-year period. Nevertheless, such an approach is helpful for identification of patients who need closer monitoring and for early detection of the disease, which may result in better long-term outcomes (4). In addition, it reinforces the message that SSc patients with low DLCO (which is highly correlated to DLCO/VA) in the absence of lung fibrosis are at risk of developing PAH (5), and that SSc patients with a low FVC due to severe lung fibrosis are at risk of developing precapillary PH (6). The Cochin RPS may be helpful in clinical practice for close monitoring of high-risk SSc patients.


Dr. Humbert has received speaking fees and honoraria from Actelion, Bayer, GlaxoSmithKline, Lilly, Novartis, Pfizer, and United Therapeutics (less than $10,000 each).

Eric Hachulla MD, PhD*, Pierre Clerson MD†, Marc Humbert MD, PhD‡, * Hôpital Claude Huriez, Université de Lille 2, Lille, France, † Orgamétrie, Roubaix, France, ‡ Université Paris-Sud 11, INSERM U999 and Hôpital Antoine Béclère, AP-HP, Clamart, France.