Ventilation–Perfusion Scintigraphy Is More Sensitive than Multidetector CTPA in Detecting Chronic Thromboembolic Pulmonary Disease as a Treatable Cause of Pulmonary Hypertension

V/Q Scintigraphy

V/Q scintigraphy was performed using ^99mTc-labeled macroaggregated albumin (^99mTc-MAA) and ^81mKr gas (⁸¹Kr). Images were acquired using an ECAM dual-head γ-camera (Siemens Medical Solutions USA, Inc.). The perfusion images were acquired after intravenous administration of 100 MBq of ^99mTc-MAA with the patient in the supine position. The images were acquired in 4 standard projections (anterior, posterior, right posterior oblique, and left posterior oblique) with a medium-energy, general-purpose collimator. An energy window of 20% centered on the energy peak of ^99mTc (140 keV), and a matrix of 128 × 128 pixels was used and approximately 400,000 counts accumulated per view. Ventilation images using ^81mKr were acquired immediately after each perfusion view. ^81mKr was inhaled through a mouthpiece. The acquisition parameters were the same as those used for the perfusion study, but the 20% window was centered on the energy peak of ^81mKr (190 keV). Images were reported from a Hermes workstation (Nuclear Diagnostics) by experienced nuclear medicine physicians. Images were interpreted according to the modified PIOPED criteria (28,29). A high-probability scan was suggestive of chronic thromboembolic pulmonary disease, whereas a low-probability scan was suggestive of nonchronic thromboembolic pulmonary disease etiology. For intermediate-probability scans, 2 separate datasets were produced and analyzed. In one dataset, intermediate-probability scans were considered as suggestive of chronic thromboembolic pulmonary disease and, in the other dataset, the intermediate-probability scans were considered as not suggestive of chronic thromboembolic pulmonary disease. This was in accordance with previous similar studies (22).

Multidetector CTPA

CTPA was acquired with multidetector row CT scanners. The first 24 studies were acquired with a Somaton Four Plus CT scanner (Siemens Medical Solutions) using 4 × 3 mm collimation and 2.5-mm reconstruction. From March 2002, the remaining 203 studies were acquired on a Light Speed CT scanner (GE Healthcare), using 8 × 1.25 mm collimation with 1.25-mm reconstruction. A bolus injection of 100 mL of nonionic iodinated contrast medium (300 mg/mL) was administered intravenously via a peripheral cannula at 4 mL/s via a power injector with a variable-start delay. This was determined using automatic bolus trigger software with a circular region of interest positioned at the level of main pulmonary trunk. The images were acquired in the craniocaudal direction with the z-axis coverage and the field of view chosen to include the entire thorax, from the apex to the base of the lungs.

All CTPA was reported by experienced radiologists using PACS workstations (GE Healthcare) as well as Siemens and General Electric workstations to assess multiplanar reconstructions. A report was considered as suggestive of chronic thromboembolic pulmonary disease if it stated visualization of the thrombus, calcified thrombus, recanalization, sudden change in vessel caliber, strictures, poststenotic dilatation, webs, or perfusion abnormality (30–34). The presence of mosaic perfusion, defined as areas of increased and decreased attenuation and vascularity without evidence of destruction or displacement of pulmonary vessels, was also noted. However, its presence alone in the absence of other signs was not considered as diagnostic of chronic thromboembolic pulmonary disease (35–37).

Pulmonary DSA

The indications for referring patients for pulmonary DSA included (a) clinical suspicion of CTEPH, (b) at least one positive imaging modality for chronic thromboembolic pulmonary disease, and (c) whether the patient was being assessed for pulmonary thrombendarterectomy. Pulmonary DSA was performed in 61 patients by experienced vascular radiologists via a femoral venous approach using a 7-French Grollman pigtail catheter (William Cook Europe) placed in the main pulmonary trunk. More selective studies of the intrapulmonary artery branches were performed in some patients with a 7-French headhunter catheter (Cordis Europa). Images were obtained with an HDI 5000 fluoroscopy unit (Philips) after administration of 20 mL of nonionic iodinated contrast medium (300 mg/mL), per DSA run, at 7 mL/s via a pump injector or by hand injection with a maximum volume limit of 300 mL. The minimum projections per hemithorax included frontal and posterior oblique projections, with extra views obtained as required.

Patients with CTEPH: Group A

CTPA was reported as showing features suggestive of chronic thromboembolic pulmonary disease in 40 of 78 patients but was falsely negative in 38 of 78 cases, giving a sensitivity of 51%. The V/Q scintigraphy showed high probability in 75 patients, intermediate probability in 1 patient, and low probability in 2 patients. When the intermediate-probability scans were considered suggestive of chronic thromboembolic pulmonary disease (grouped with the high probability), the V/Q sensitivity was 97.4%, and when intermediate probability was considered as negative for chronic thromboembolic pulmonary disease (grouped with low probability), the sensitivity was 96% (Tables 1 and 2).

Patients with Non-CTEPH: Group B

CTPA did not show features suggestive of chronic thromboembolic pulmonary disease in 148 of 149 patients with one false-positive case. V/Q scintigraphy showed low probability in 134 patients, high probability in 8 patients, and intermediate in 7 patients.

The specificity of CTPA was 99%. The V/Q specificity was 90% when the intermediate-probability scans were analyzed as suggestive of chronic thromboembolic pulmonary disease and 95% when the intermediate scans were considered negative for chronic thromboembolic pulmonary disease (Tables 1 and 2).

Mosaic Perfusion

Mosaic perfusion was reported in 53 of the total 227 CTPA scans (23.34%). Among those, 43 (18.94%) were in group A and 10 (4.41%) were in group B.

Thus, the CTPA showed evidence of mosaic perfusion in 55% (43/78) of patients with CTEPH (group A).

False-Negative V/Q Scintigraphy

The 2 patients from group A who had their V/Q scintigraphy interpreted as low probability did not have the diagnosis of CTEPH confirmed on pulmonary DSA. The final diagnosis was made by CTPA, which showed calcified thrombus in one patient and narrowed but patent pulmonary arterial branches in the other patient. By corroborating the patient clinical history and the CT findings, it was concluded that these patients most likely had previous PE and the pulmonary vessels had recanalized.

V/Q scintigraphy that was interpreted as intermediate probability was confirmed on pulmonary DSA to be CTEPH. The CT was reported as showing severe emphysematous changes but failed to show signs of CTEPH. It was the V/Q result together with the clinical history that prompted the clinicians to request pulmonary DSA.

False-Positive V/Q Scintigraphy

Eight V/Q scans were reported as high probability and 7 were reported as intermediate probability among the non-CTEPH group. None of these patients had pulmonary DSA. The clinical diagnoses were idiopathic pulmonary arterial hypertension, pulmonary arterial hypertension associated with atrial septal defect, pulmonary venoocclusive disease, and PH related to parenchymal lung disease, such as emphysema or pulmonary fibrosis (Table 3).

Mosaic Attenuation Pattern

The combination of a mosaic pattern and disparity in the size of segmental vessels on HRCT was found in a study by Bergin et al. to reliably distinguish patients with CPTEH (33). Similarly, Schwickert et al. analyzed the CT findings in 75 patients diagnosed with CTEPH and found mosaic perfusion in 77% (34). However, the mosaic perfusion pattern can also be found in patients with PH due to other causes (31–37), and up to 12% of patients with idiopathic pulmonary arterial hypertension will have mosaic perfusion on CTPA (35). In our study, mosaic perfusion was present in only 55% of the CTPA performed in group A patients and 20% of group B patients. This confirms that, although the mosaic pattern is more common in CTEPH, it is not a specific feature.