This retrospective cohort study conducted in a Portuguese referral center for pulmonary hypertension included all patients hospitalized with PE between 2014 and 2019, identified according to the International Classification of Diseases, Ninth (ICD-9) or Tenth (ICD-10) Revision, applying ICD-9 codes for “pulmonary embolism” (code 415.1) for the period between January 1, 2014 and December 31, 2015 and ICD-10 codes (I26.92, I26.99, I26.02, or I26.09) for patients hospitalized between January 1, 2016 and December 31, 2019. Only patients with a PE diagnosis confirmed by imaging techniques or autopsy were included in the study.

Through individual examination of each of the 969 clinical records identified with a PE diagnosis (including admission transthoracic echocardiography or CTPA and troponin levels when applicable), patient risk was stratified according to the criteria proposed by the current European Society of Cardiology and European Respiratory Society guidelines,3 except that the pulmonary embolism severity index (PESI) score was not considered in this stratification, and only hemodynamic stability, right ventricular dysfunction and troponin levels were used.

To calculate the incidence of episodes of hospitalization with PE, estimates of the resident population provided by Statistics Portugal (INE; www.ine.pt) of the corresponding geographic areas were obtained for the years under study. The incidence of PE in our population was estimated as the number of estimated new PE episodes per year and was expressed as 100000 population/year calculated as a six-year average.

Intermediate-high and high-risk PE were further analyzed for the purposes of this study, including estimation of CTEPH prevalence and determination of CTEPH predictors. Patients who died or were lost to follow-up within three months of the acute PE event were excluded from this further analysis. All patients included in this analysis had at least one follow-up appointment and were managed according to current guidelines.

Hospital clinical records (from hospital or emergency room admissions) and hospital or primary care appointments accessible from the national electronic registry (Registo de Saúde Eletrónico) provided data on demographics, comorbidities, clinical presentation, laboratory and imaging findings, treatment of PE and CTEPH, clinical course, mortality and risk factors for CTEPH. Follow-up was analyzed in all patients until the last available record or date of death.

The study protocol was approved by the institutional ethics committee. The protocol was established in accordance with the 1975 Helsinki Declaration of the World Medical Association. The corresponding authors are entirely responsible for the integrity of data and data analysis. After data collection individual participants were anonymized.

The primary purpose of this study was to estimate the prevalence of CTEPH following a symptomatic intermediate-high or high-risk PE in a Portuguese population.

CTEPH was considered probable in patients, irrespective of symptoms, in the presence of persistent mismatched perfusion defects on ventilation/perfusion (V/Q) scan and intermediate- or high-probability pulmonary hypertension on echocardiography but not confirmed by RHC.2

CTEPH was considered confirmed when, in addition to the suggested changes on the V/Q scan, RHC established the diagnosis and excluded other forms of pulmonary hypertension. CTEPH was defined hemodynamically according to current guidelines1,2 as precapillary pulmonary hypertension with a mean pulmonary arterial pressure (mPAP) >20 mmHg, pulmonary arterial wedge pressure (PAWP) ≤15 mmHg and pulmonary vascular resistance (PVR) >2 Wood units, after at least three months of therapeutic anticoagulation.

In the present study, suspected CTEPH included both confirmed and probable CTEPH.

Secondary purposes were to identify CTEPH predictors and to analyze CTPA at the time of the initial PE event in order to identify patients with unknown pre-existing CTEPH at the time of the acute PE event (acute-on-chronic PE).

Potential predictors of CTEPH considered in the analysis included the following: age, PE admission risk, cardiovascular comorbidities, hormone therapy, previous deep venous thrombosis (DVT), previous PE, active malignancy, recent major surgery, immobility, prolonged travel, idiopathic PE, recurrent PE, symptom onset >2 weeks before PE diagnosis, thyroid hormone replacement, splenectomy, pacemaker lead infection, ventriculoatrial shunt, antiphospholipid syndrome, myeloproliferative disorder, varicose veins in lower limbs, PE presentation, PESI score, peak N-terminal pro-brain natriuretic peptide (NT-proBNP) at the time of the PE event, echocardiographic findings at PE diagnosis and during follow-up, CTPA findings at PE diagnosis, and PE treatment, particularly reperfusion.

The CTEPH prediction score18 was applied to the 129 patients included in the analysis. Analysis of its accuracy in excluding CTEPH in our population was further undertaken.

The CTPA of patients with suspected CTEPH at the initial event was retrospectively examined by an expert radiologist to identify findings suggesting the presence of unidentified CTEPH at the time of the initial PE event. These findings included the presence of mural thrombi lining the pulmonary vascular walls, eccentric wall-adherent filling defects, intravascular webs or bands, mosaic parenchymal perfusion patterns, right ventricular hypertrophy or right atrial dilatation, pericardial effusion, dilatation of the main pulmonary artery (>29 mm in men, >27 mm in women) and dilated bronchial arteries (≥1.5 mm).4,17,19,20

Continuous variables were presented as mean±standard deviation or median and interquartile range (IQR) and were compared between groups using the independent samples t test or the Mann–Whitney U test, according to distribution. Normality was tested with the Shapiro–Wilk test and homogeneity of variances with Levene's test. Categorical variables were presented as frequencies and percentages and were compared using the chi-square test or Fisher's exact test, as appropriate. Logistic regression analysis was used to assess predictors of the presence of CTEPH during follow-up using the enter method. The Hosmer–Lemeshow test was used to calibrate the regression model. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. The discriminative ability of the previously published CTEPH prediction score10 to predict CTEPH in our population was assessed using receiver operating characteristic (ROC) curve analysis. All reported p values were two-sided and a p-value <0.05 was considered statistically significant. Statistical analysis was performed with IBM SPSS software, version 26.0 (IBM Corp., Armonk, NY, USA).

Between January 2014 and December 2019, 969 patients were hospitalized with PE, as depicted in the study flowchart (Figure 1). The mean annual incidence was estimated at 46/100000 population. Of these, 477 (49.2%) were stratified as low-risk patients, 298 (30.8%) as intermediate-low risk, 142 (14.7%) as intermediate-high risk and 52 (5.4%) as high-risk patients.

Figure 1.

Study flowchart. * Risk stratification at admission following the current ESC guidelines, although PESI score was not considered for this categorization: low risk (absence of right ventricular [RV] dysfunction and troponin elevation); intermediate-low risk (RV dysfunction on transthoracic echocardiogram [TTE] or computed tomography pulmonary angiography [CTPA]; or elevated troponin levels); intermediate-high risk PE (RV dysfunction on TTE or CTPA and elevated troponin levels); high-risk (cardiac arrest: need for cardiopulmonary resuscitation; obstructive shock: hypotension (systolic blood pressure [BP] <90 mmHg) or vasopressors required to achieve BP >90 mmHg despite adequate filling status) and end-organ hypoperfusion (altered mental status; cold, clammy skin; oliguria/anuria; increased serum lactate); persistent hypotension (systolic BP <90 mmHg or systolic BP drop >40 mmHg, lasting longer than 15 min and not caused by new-onset arrhythmia, hypovolemia, or sepsis). CTEPH: chronic thromboembolic pulmonary hypertension; PE: pulmonary embolism; RHC: right heart catheterization.

(0.46MB).

Of the 194 patients included in the study with intermediate-high or high-risk PE, 54 were excluded due to death in the 90 days following the PE event. Of the 140 patients alive at least 90 days after the PE event, 11 were not included due to loss of clinical follow-up in the first three months. Therefore, the final study population included 129 patients, with a median age of 69.0 years (56.0–79.0) at the time of the first PE event. The median follow-up was 41.0 (24.0–58.5) months.

The baseline characteristics of the study population are summarized in Table 1.

During follow-up of the 129 patients diagnosed with intermediate-high or high-risk acute PE, four patients were diagnosed with CTEPH confirmed by RHC. Thus, a prevalence of 3.1% (n=4) of confirmed CTEPH following an intermediate-high or high-risk PE was found in our population. Additionally, four patients with compatible clinical presentation were found to have intermediate- or high-probability pulmonary hypertension on echocardiography and persistent mismatched perfusion defects on V/Q scan compatible with a diagnosis of CTEPH, but RHC was not performed for the following reasons: malignancy with poor prognosis, comorbidities with poor prognosis, advanced dementia or frailty. Therefore, a prevalence of suspected CTEPH of 6.2% (n=8) was estimated in our population following an intermediate-high or high-risk PE. The median time until diagnosis of CTEPH was 9.5 (6.0–21.0) months. Follow-up was completed at two years in 89.9% of patients.

Hemodynamic and clinical characteristics of patients with suspected CTEPH are summarized in Table 2.

Among patients with confirmed CTEPH, all were anticoagulated (three on direct oral anticoagulants and one on a vitamin K antagonist). One underwent pulmonary thromboendarterectomy (PEA) and another underwent balloon pulmonary angioplasty, both with significant clinical and hemodynamic improvement after the procedures. The other two patients refused PEA. No patient is currently medicated with pulmonary vasodilators.

The comparison between patients with and without suspected CTEPH is outlined in Table 1.

Older age was found in patients with CTEPH, with a median age of 73.5 years (IQR 15.0) in CTEPH patients compared to 68.0 years (IQR 24.0) in patients without CTEPH. Patients with CTEPH were all admitted with intermediate-high risk PE, whereas 16.5% of patients not diagnosed with CTEPH were stratified as having high-risk PE (p=0.357).

Analysis of CTEPH predictors revealed some significant differences. A higher incidence of varicose veins in the lower limbs was described in patients who developed CTEPH (37.5%) than in those not diagnosed with CTEPH (p=0.027). Idiopathic PE was more frequently associated with CTEPH, although not significantly (p=0.146). Antiphospholipid syndrome was diagnosed in one patient (12.5%) with CTEPH, while no patients without CTEPH had this comorbidity (p=0.062). Median pulmonary artery systolic pressure (PASP) in patients diagnosed with CTEPH was 82.0 mmHg (IQR 45.0), compared to 53.0 mmHg (IQR 19.0) in patients without CTEPH (p=0.004). Increased PASP at admission (OR 1.12; 95% CI 1.04–1.22; p=0.005) and the presence of varicose veins in the lower limbs (OR 7.47; 95% CI 1.53–36.41; p=0.013) were predictors of CTEPH in multivariate analysis.

ROC curve analysis (Figure 2A) demonstrates the good discriminative ability of PASP at acute PE admission to predict CTEPH during follow-up (area under the ROC curve 0.85; 95% CI 0.76–0.92; p=0.001). PASP >60 mmHg at admission identified patients with CTEPH during follow-up with sensitivity and specificity of 83.3% and 76.3%, respectively. PASP >79 mmHg at the index event identified CTEPH with sensitivity of 66.7% and specificity of 98.7%.

Figure 2.

Receiver operating characteristic curve analysis demonstrating the discriminative ability of PASP at admission for acute PE (A) and the CTEPH prediction score (B) to predict the presence of CTEPH in follow-up. AUC: area under the curve; CTEPH: chronic thromboembolic pulmonary hypertension; PASP: pulmonary artery systolic pressure.

(0.19MB).

The CTEPH prediction score in CTEPH patients was a median 6.5 compared to 5.0 in patients not diagnosed with CTEPH (p=0.290). This score (Figure 2B) did not show a good discriminative ability in our population to predict CTEPH in the follow-up (area under the ROC curve 0.61; 95% CI 0.52–0.69; p=0.137).

No significant difference was found between fibrinolysis and percutaneous thrombectomy in the two groups.

At follow-up, higher PASP levels were found in patients with CTEPH, with a median of 70.5 mmHg (IQR 40.0), compared to a median of 30.0 mmHg (IQR 13.0) in patients with no CTEPH (p<0.001). Additionally, a right ventricular/left ventricular basal diameter/area ratio >1 in the follow-up was also statistically significant in predicting CTEPH development. Four of the patients with CTEPH (50%) had dilated right heart chambers compared to six (7.1%) of those not diagnosed with CTEPH (p=0.004).

Multidetector CTPA findings from the index PE in patients with suspected CTEPH were analyzed by two radiologists (Table 3). In all patients at least two signs of CTEPH were detected on the multidetector CTPA and in five of them there were at least three findings suggestive of CTEPH.

Despite the limitations inherent to a retrospective analysis, this is the first study to estimate the prevalence of CTEPH after a PE event confirmed by RHC in a Portuguese population. However, we recognize some limitations in our study that should be addressed.

The main limitation of our work was the small number of patients with CTEPH, which precluded the identification of predictors for this entity. Due to this small sample, multivariate analysis did not show statistical differences in multiple predictors that in other studies were associated with increased risk of developing CTEPH. Further studies are required for additional research into predictors of CTEPH following a more severe form of PE.

Secondly, it was not possible to confirm the diagnosis of CTEPH by RHC in all patients with suspected CTEPH, due to the comorbidities present in some patients that prevented them from undergoing this invasive procedure. For this reason, in individuals unable to undergo the invasive procedure, CTEPH was defined as the presence of echocardiographic findings suggestive of pulmonary hypertension (including peak tricuspid regurgitation velocity >2.9 m/s) and positive findings on the V/Q scan. Even though this alternate diagnosis is not the gold standard for diagnosis and therefore not as accurate as RHC, the exclusion of patients who were unable to undergo RHC due to comorbidities would greatly underestimate the prevalence of CTEPH following acute PE.

Thirdly, only CTEPH patients had their CTPA analyzed by expert radiologists, which precludes multivariate analysis of radiological predictors of CTEPH, which have been undertaken in other studies.