Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

Predictive Value of Non-Invasive Tests for the Diagnosis of Scleroderma-Associated Pulmonary Hypertension: PHAROS Registry.

Khanna4,  Dinesh, Saggar4,  Rajeev, Furst4,  Daniel E., Allanore3,  Yannick, Seibold5,  James R., Clements4,  Philip J., Saggar4,  Rajan

Georgetown University
Northwestern Univ
Paris Descartes University
Univ of Connecticut


Pulmonary hypertension (PH) is a major cause of death in systemic sclerosis (SSc). The Pulmonary Hypertension Assessment and Recognition of Outcomes in Scleroderma (PHAROS) is a prospective longitudinal study of patients at risk of developing PH and those who have definite PH.


Our objective was to assess the performance of non-invasive tests (echo RVSP, PFTs, serum BNP) for the diagnosis of PH (defined as mPAP>=25 mmHg on right heart catheterization [RHC]; includes WHO Groups1–3) in SSc patients.


Entry criteria for patients "at risk" for PH included a DLCO<=55% predicted, a FVC%/DLCO%>=1.6 or a PASP on echo >= 40 mmHg. RHC are performed based on the clinical judgment of the physician. We performed 3 analyses: 1) assessed individual positive (PPV) and negative predictive value (NPV) for echo RVSP, FVC/DLCO, DLCO% predicted, and serum BNP compared to initial RHC mPA; 2) analyzed discriminatory power of echo RVSP, FVC%/DLCO%, DLCO% predicted, and BNP using classification and regression tree (CART) analysis; and 3) calculated what proportion of patients are captured by serum BNP and PFTs who were missed by different RVSP cut offs.


We analyzed 209 SSc patients who underwent RHC after enrollment into PHAROS; 177 had the cath done at visit#1 and 32 had it done during follow-up visits. 143 (68%) had PH on RHC. PPV for RVSP (35 –50 mmHg) ranged from 0.78 to 0.91; FVC%/DLCO% (1.4 to 2.0) ranged from 0.67 to 0.78; and DLCO% cut offs (50%–70%) ranged from 0.68 to 0.73, and serum BNP (>100) was 0.94. Combination of RVSP >50 mmHg and DLCO< 50% had greatest PPV (0.95). CART analysis showed that RVSP >50 mmHg had greatest discrimination for diagnosing PH followed by serum BNP >100 units Figure. However, 10% to 40% (Echo cutoff>50 mmHg) of patients with PH were missed with RVSP > 35 and > 50mmHg respectively (Table). In those missed, 2–5% were captured by serum BNP> 100, 5–29% by DLCO< 60% and 3–23% by FVC/DLCO <= 1.6. (Table).

Table. Performance of various non-invasive tests

Echo cut offPH missed by Echo RSVPAdditional patients captured with RHC-defined PH
  Serum BNP >100*DLCO <60%FVC/DLCO >=1.6
Echo >35 mm Hg10%2%5%3%
Echo >40 mm Hg20%3%14%10%
Echo >45 mm Hg30%4%22%16%
Echo >50 mm Hg40%5%29%22%
*only 67 patients had serum BNP reported


Echo RVSP > 50 mmHg has the greatest discriminatory power to detect PH in SSc in this enriched cohort. However, at this cut-off, serum BNP and PFT captures additional 27% of patients with PH. Serum BNP and PFTs complement Echo RVSP and should be included as part of work up of PH in SSc. This data needs to be confirmed in another real-life cohort.

To cite this abstract, please use the following information:
Khanna, Dinesh, Saggar, Rajeev, Furst, Daniel E., Allanore, Yannick, Seibold, James R., Clements, Philip J., et al; Predictive Value of Non-Invasive Tests for the Diagnosis of Scleroderma-Associated Pulmonary Hypertension: PHAROS Registry. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :729
DOI: 10.1002/art.28497

Abstract Supplement

Meeting Menu