Cardiopulmonary Exercise Testing to Detect Chronic Thromboembolic Pulmonary Hypertension in Patients with Normal Echocardiography

Matthias Held a Maria Grün a Regina Holl a Gudrun Hübner a Ralf Kaiser c
Sabine Karl b Martin Kolb e Hans Joachim Schäfers d Heinrike Wilkens c
Berthold Jany a
a Department of Internal Medicine, Medical Mission Hospital, Academic Teaching Hospital, and b Institute of
Mathematics, Julius Maximilian University of Würzburg, Würzburg , c Department of Internal Medicine V, Respiratory
and Critical Care Medicine, and d Clinic for Thoracic and Cardiovascular Surgery, University Hospital Homburg
Saar, Homburg , Germany; e Department of Medicine, Pathology and Molecular Medicine, Firestone Institute for
Respiratory Health, McMaster University, Hamilton, Ont. , Canada

Respiration 2014;87:379–387

Abstract
Background: Chronic thromboembolic pulmonary hypertension
(CTEPH) is a serious complication of pulmonary embolism
(PE). Taking into account the reported incidence of
CTEPH after acute PE, the number of patients with undiagnosed
CTEPH may be high.
Objectives: We aimed to determine
if cardiopulmonary exercise testing (CPET) could serve
as complementary tool in the diagnosis of CTEPH and can
detect CTEPH in patients with normal echocardiography.
Methods: At diagnosis, we analyzed the data of CPET parameters
in 42 patients with proven CTEPH and 51 controls, and
evaluated the performance of two scores.
Results: V E /V CO 2
slope, EQ O 2 , EQ CO 2 , P(A-a) O 2 , end-tidal partial pressure of CO 2
at anaerobic threshold (PET CO 2 ) and capillary to end-tidal carbon
dioxide gradient [P(c-ET) CO 2 ] were significantly different
between patients with CTEPH and controls (p < 0.001). P(c-ET) CO 2 was
the single parameter with the highest sensitivity
(85.7%) and specificity (88.2%). A score combining V E /V CO 2
slope, P(A-a) O 2 , P(c-ET) CO 2 , PET CO 2 [4-parameter-CPET (4-PCPET)
score] reached a sensitivity of 83.3% and a specificity of
92.2% after cross-validation. In 42 patients with CTEPH, echocardiography
identified PH in 29 patients (69%), but it was
normal in 13 patients (31%). All patients with normal or unmeasurable
right ventricular systolic pressure had a pathological
CPET. Twelve of the 13 patients (92%) were detected
by both CPET scores.
Conclusion: CPET is a useful noninvasive
diagnostic tool for the detection of CTEPH in patients with
suspected PH but normal echocardiography. The 4-P-CPET
score provides a high sensitivity with the highest specificity.