INTRODUCTION
Pulmonary embolism ranks as the third most common cardiovascular emergency after ischemic heart disease and stroke2,5,22. The incidence of pulmonary embolism increases markedly with increasing age thereby making the disease a major health problem, especially among the elderly2,5,22.
Estimates of survival after acute pulmonary embolism have been widely reported in the literature1,3,6,7,10,21,26. Most of the reported studies, however, lacked a comparable reference group of individuals without pulmonary embolism1,3,6,10,21. This precluded the evaluation of pulmonary embolism as an independent predictor of survival.
We conducted a prospective study that was aimed at evaluating the short- and long-term survival in a sample of consecutive patients who were referred to our institution for suspected acute pulmonary embolism. Survival rates in patients with proven pulmonary embolism were compared with those in whom the disease had been excluded. In multivariate analysis, we estimated the adjusted odds ratios of death for patients with pulmonary embolism as a function of time since diagnosis and of the extent of pulmonary vascular obstruction at baseline, using patients without pulmonary embolism as the reference category. We also evaluated the degree of restoration of pulmonary perfusion in the patients who survived a full year after acute pulmonary embolism, the cumulative incidence of recurrent pulmonary embolism, and the incidence of chronic thromboembolic pulmonary hypertension.
PATIENTS AND METHODS
Sample
The study sample consisted of 913 consecutive patients who were referred to the Institute of Clinical Physiology, Pisa, Italy, for suspected acute pulmonary embolism, between April 1, 1996, and January 31, 2000. The suspicion of pulmonary embolism had been raised based on the following criteria: presence of pertinent symptoms such as unexplained dyspnea, chest pain, or syncope; electrocardiographic or echocardiographic signs of acute right ventricular overload; arterial hypoxemia with respiratory alkalosis; chest radiographic abnormalities such as focal oligemia, enlargement of the descending pulmonary artery with sign of vascular amputation, or lung consolidations suggestive of infarction16.
Seventy-nine patients (9%) were excluded from the study because of inability to obtain informed consent (n = 15), contraindication to pulmonary angiography (hypersensitivity to contrast medium or renal failure, n = 10), expected survival of less than 1 month (n = 49), or geographic inaccessibility precluding follow-up (n = 5). The 834 other patients were examined uniformly according to a standardized protocol which included clinical evaluation, perfusion lung scanning, and pulmonary angiography15-18. None of the patients had undergone any objective testing for pulmonary embolism before study entry. All the procedures (including pulmonary angiography) were carried out in a dedicated diagnostic unit at our institution. The protocol was approved by the institutional review board for human studies.
Clinical Evaluation
Clinical evaluation was carried out by 1 of 12 chest physicians and included history taking, physical examination, interpretation of the electrocardiogram and chest radiograph, and measurement of the partial pressure of oxygen (PaO2) and carbon dioxide (PaCO2) in arterial blood. Arterial blood samples were obtained in all patients while they were breathing room air. Clinical and laboratory data were recorded by the physicians on a standard form before any further objective testing.
Immobilization for longer than 3 consecutive days, any major surgical procedure, or any bone fracture of the lower extremities were considered risk factors if they occurred within 4 weeks before study entry. History of lower limb deep vein thrombosis, or any prior episode of pulmonary embolism, was recorded if the patient had, at any time, documented episodes of deep vein thrombosis or pulmonary embolism that required anticoagulant therapy. Estrogen use was defined as use of estrogen-containing drugs within the past 3 months. Postpartum period was defined as the presence of pregnancy within the past 3 months.
Cardiovascular (coronary artery disease, systemic arterial hypertension, heart failure, left heart valvular disease, cerebrovascular disease), pulmonary (chronic obstructive pulmonary disease, asthma, interstitial lung disease), endocrine (diabetes mellitus, thyroid dysfunction of any cause), or any other kind of nonmalignant diseases were recorded if documented any time before study entry. Neoplastic disease was recorded if there was evidence of clinically active malignancy with pathologic diagnosis within the past 3 months.
Perfusion Lung Scanning
Perfusion lung scans were obtained after intravenous injection of human serum albumin microspheres labeled with 99m Technetium (1.8 × 108 Bq), taking care to inject the radioactive bolus with the patient held as closely as possible to the sitting position in order to preserve the effect of gravity on the regional distribution of pulmonary blood flow. Lung scans were acquired by means of a large field gamma camera equipped with a high resolution, parallel-hole collimator, using a 20% symmetric window set over the 140 KeV photopeak. Images consisted of anterior, posterior, both lateral, and both posterior oblique views, with 500,000 counts per image.
Lung scans were independently attributed to 1 of 4 predetermined categories15: normal (no perfusion defects); near-normal (impressions caused by enlarged heart, hila, or mediastinum are seen on an otherwise normal scan); abnormal, suggestive of pulmonary embolism (single or multiple wedge-shaped perfusion defects); abnormal, not suggestive of pulmonary embolism (single or multiple perfusion defects other than wedge-shaped).
Pulmonary Angiography
Pulmonary cineangiograms were obtained according to standardized procedures within 24 hours of study entry15. Before angiography, informed written consent was obtained. Initial filming was in the anteroposterior view, after having advanced the catheter into the main pulmonary artery of the lung which showed the greatest perfusion abnormalities on lung scanning. If there was a doubt about the presence of filling defects, the appropriate vessel was selectively entered with a balloon-tipped catheter and angiograms were repeated by manual injection of contrast material. Pulmonary angiograms were examined by experienced physicians who were blind to clinical information. Angiographic criteria for diagnosing pulmonary embolism included the identification of an embolus obstructing a vessel or the outline of an embolus within a vessel. In patients who died before angiography (n = 7), the diagnosis was established at autopsy.
Diagnostic Criteria
The diagnosis of pulmonary embolism was based on angiographic or autopsy documentation of pulmonary emboli. Criteria for excluding pulmonary embolism were a normal pulmonary angiogram, absence of pulmonary emboli at autopsy, or a normal perfusion scan. Performing pulmonary angiography in patients with normal scans was deemed unethical because available data indicate that such a scintigraphic pattern, in itself, rules out clinically significant pulmonary embolism8,9,25.
Patients with pulmonary embolism received anticoagulant therapy. The form and duration of treatment were left to the discretion of the attending physicians. Usually, however, treatment consisted of a 1-week heparin infusion followed by oral anticoagulation for 1 year.
Follow-Up
All patients were followed until death or January 31, 2001, whichever came first, by a team of 5 physicians. Information regarding inpatients was obtained from the attending physicians. Patients who were discharged from the hospital were interviewed at 3-month intervals. Whenever required, their family physicians were also contacted. Critical events recorded by the physicians in charge of follow-up included: hospital readmissions for any cause, recurrent episodes of pulmonary embolism, and death.
In patients with pulmonary embolism, a clinical and scintigraphic follow-up was pursued at 1 week, 1 month, and 1 year of study entry to assess the restoration of pulmonary perfusion. Patients who showed a persistence of large bilateral perfusion defects in sequential lung scans were evaluated by transthoracic echocardiography to look for signs of right ventricular overload. Pulmonary hypertension was indicated by an estimated pulmonary artery systolic pressure by echocardiography of at least 36-50 mm Hg, or a resting tricuspid regurgitant velocity of 2.8-3.4 m/s, (assuming a normal right atrial pressure of 5 mm Hg)24. If echocardiographic findings were suggestive of pulmonary hypertension, a right heart catheterization was performed to ascertain whether the patient met the hemodynamic criteria for chronic thromboembolic pulmonary hypertension (see below). Echocardiographic studies were not obtained in patients who showed a nearly complete or complete restoration of pulmonary perfusion, because this in itself makes the possibility of postembolic pulmonary hypertension very unlikely.
Outcomes
Time to death was measured for patients with pulmonary embolism and those without. The cause of death was established by an independent panel of 3 physicians who reviewed clinical files, autopsy findings, or death certificates.
Recurrent pulmonary embolism was diagnosed if the patients had symptoms that prompted a new investigation, and met any one of the following criteria: new segmental perfusion defects on the lung scan, a positive angiogram, or evidence of fresh pulmonary emboli at autopsy.
In patients with pulmonary embolism, we estimated the extent of scintigraphically detectable pulmonary vascular obstruction as an index of the severity of the disease. This analysis was carried out by 2 independent raters, who were blind to clinical information, according to a method previously described14. Briefly, each lobe was attributed a weight according to regional blood flow as follows: right upper lobe: 0.18; right middle lobe: 0.12; right lower lobe: 0.25; left upper lobe: 0.13; lingula: 0.12; left lower lobe: 0.20. The perfusion of each lobe was estimated visually by means of a 5-point score: 0, 0.25, 0.5, 0.75, 1, where 0 means "not perfused" and 1 means "normally perfused." Visual estimates of perfusion were based on the combined evaluation of 6 scintigraphic views (anterior, posterior, both lateral, and both posterior oblique). Each lobar perfusion score was obtained by multiplying the weight assigned to the lobe by the estimated perfusion of that lobe. The overall perfusion score was the sum of the perfusion scores of the 6 lobes, and the percentage of pulmonary vascular obstruction was calculated as (1 − overall perfusion score) × 100. The percentage of pulmonary vascular obstruction was estimated on the lung scans obtained at the time of diagnosis and at 1 week, 1 month, and 1 year since diagnosis to assess the degree of restoration of pulmonary perfusion. The average of the 2 independent ratings was used for the analysis throughout.
Chronic thromboembolic pulmonary hypertension was considered to be present if the following criteria were met: angiographic evidence of vascular narrowing or occlusion; persistence of multiple segmental or lobar perfusion defects in sequential lung scans; a mean pulmonary artery pressure >25 mm Hg with a mean pulmonary artery wedge pressure <15 mm Hg at right heart catheterization.
Statistical Analysis
The baseline characteristics of patients with and without pulmonary embolism were compared by contingency tables for categorical variables. For the continuous variables, differences between the 2 diagnostic groups were tested for by Wilcoxon rank-sum nonparametric procedure. P values <0.05 were considered statistically significant throughout. Kaplan-Meier survival curves were used to estimate the cumulative probability of mortality in the 2 diagnostic groups, and the cumulative incidence of recurrent pulmonary embolism. Time to death was graphically presented by Kaplan-Meier survival curves separately for patients without pulmonary embolism, patients with pulmonary embolism and vascular obstruction <50%, and those with vascular obstruction ≥50%. The difference between the survival curves was tested by both the log-rank and Wilcoxon (Breslow) tests. The latter was applied given that the survival curves crossed one another, suggesting lack of proportional hazard functions4. P values were based on the chi-square asymptotic approximation.
The hazard functions based on Kaplan-Meier estimates were graphically shown within the first month from the event separately for patients without pulmonary embolism, patients with pulmonary embolism and obstruction <50%, and those with obstruction ≥50%. They were smoothed using a triangular kernel function with half-width of 10 days.
Initially, we developed multivariate Cox proportional-hazard models. No reasonably simple model, however, appeared to account satisfactorily for the lack of proportionality of the hazard functions in the 3 groups of patients. Since there were no censored observations within a year of study entry, we preferred to model the probability of surviving at different time points separately. For each patient, 5 dichotomous variables were defined according to whether the patient survived 1 day, 1 week, 1 month, 3 months, and 1 year. Five separate logistic regressions were then estimated, introducing each of the dichotomous variables as the dependent variable. The independent variables were selected for each regression model separately. Continuous variables were categorized as follows: time elapsed from symptoms onset was dichotomized (1 day, more than a day); age was split in 3 categories based on the tertiles (15-64, 65-74, 75-97 years); and the percentage of pulmonary vascular obstruction, in patients with pulmonary embolism, was split in 2 based on the 50% cutoff value. All the patients' characteristics were first considered in univariate analysis. Categorical variables entered the models by means of indicator variables. The variables that showed some association with the dependent variable were then introduced in a multivariate regression model. Besides the 2-level categorical variable associated with the severity of pulmonary embolism, which was included in all the models, only the variables that were significant or appeared to be confounders were kept in the models. Confounder was defined as a variable whose removal from a model caused changes in the other variables' coefficients greater than 10%. Pair-wise interaction terms were tested and left out of the models when not significant. The following variables remained in all the regression models: age, male sex, prolonged immobilization, recent surgery, chronic diseases other than cardiopulmonary, and malignancy. For patients with pulmonary embolism, the adjusted odds ratios of death and their 95% confidence intervals (95% CI) were reported using patients without pulmonary embolism as the reference category. Statistical analysis was performed using Stata statistical software (StataCorp, College Station, TX).
RESULTS
The prevalence of pulmonary embolism in the study sample was 38.4% (320/834). The baseline characteristics of patients with and without pulmonary embolism are summarized in Table 1. In patients with pulmonary embolism, the median extent of scintigraphically detectable pulmonary vascular obstruction at diagnosis was 42.8% (range, 4.5%-81.8%). Pulmonary vascular obstruction was <50% (median, 32%) in 200 patients, and ≥50% (median, 60%) in 120.
;)
TABLE 1: Patients' Baseline Characteristics
Patients who were excluded from the study were managed according to a diagnostic protocol based on combining clinical probability with perfusion lung scan results16. Such a protocol had been previously validated by direct comparison with pulmonary angiography16. Pulmonary embolism was diagnosed in patients who had a high clinical probability paired with an abnormal scan suggestive of pulmonary embolism (wedge-shaped perfusion defects). Pulmonary embolism was deemed absent on the basis of a normal scan or a low clinical probability paired with an abnormal scan not suggestive of pulmonary embolism (perfusion defects other than wedge-shaped). Whenever required, lower limb ultrasonography was used to look for signs of deep vein thrombosis.
Survival
Follow-up was completed in all patients, and had a median duration of 2.12 years (range, 0-4.79 yr). Of the 834 patients, 271 (32.5%) died, providing a total of 1870 person-years of follow-up. In-hospital deaths accounted for 59.8% of all deaths.
The causes of death in patients with and without pulmonary embolism are reported in Table 2. With the exception of pulmonary embolism, there was no statistically significant difference between the 2 groups as regards the various causes of death. As shown in Figure 1, most of the patients with pulmonary embolism who died within the first month did so as a consequence of the initial or a recurrent episode of embolism. At a later stage, the underlying comorbid conditions were the causes of most deaths in patients with pulmonary embolism.
TABLE 2: Causes of Death in Study Sample
FIGURE 1: Cumulative number of deaths as a function of time since diagnosis of pulmonary embolism. Solid bars are deaths due to pulmonary embolism; open bars are deaths due to causes other than pulmonary embolism.
The patients who died from pulmonary embolism within 1 day of study entry had a median vascular obstruction of 66.9% (range, 43%-79%) that was significantly greater (p < 0.003) than that in patients who died between day 2 and 7 (median, 42%; range, 28%-77%) and in those who died between day 8 and 30 (median, 43%; range, 12.5%-80%). There was no significant difference between the groups as to age, prevalence of cardiovascular diseases, pulmonary diseases, or malignancy.
Kaplan-Meier survival functions in patients with and without pulmonary embolism are reported in Table 3, and are graphically displayed in Figure 2. Within the first 6 months of follow-up, the survival function in patients with pulmonary embolism and vascular obstruction ≥50% declined at a faster rate than in patients having vascular obstruction <50% and those without pulmonary embolism. After the second year, the distance between the curves decreased, and in the final year of follow-up they overlapped.
TABLE 3: Kaplan-Meier Survival Functions*
FIGURE 2: Kaplan-Meier survival curves for patients with pulmonary embolism and vascular obstruction ≥50% (solid line), patients with pulmonary embolism and vascular obstruction <50% (dashed line), and patients without pulmonary embolism (dotted line). Tests for equality of survival functions: log-rank p value: 0.2289; Wilcoxon (Breslow) p value: 0.0381.
Figure 3 shows the smoothed hazard functions for patients with pulmonary embolism and vascular obstruction ≥50%, patients with pulmonary embolism and vascular obstruction <50%, and for patients without pulmonary embolism. At any given time, the hazard represents the risk of dying for someone who has survived until then. In patients without pulmonary embolism, the risk of dying was slight and constant over time. In patients with pulmonary embolism and vascular obstruction ≥50%, the hazard was elevated right after the event and rapidly declined in the first few days. In patients with obstruction <50%, the hazard in the first few days was slightly, but not significantly, greater than in patients without pulmonary embolism, although much smaller than in patients with major obstruction. The patients with embolism who survived the first week had about the same risk of dying within 1 month from pulmonary embolism as from other causes, regardless of their initial vascular obstruction.
FIGURE 3: Smoothed hazard curves, based on Kaplan-Meier estimates, for patients with pulmonary embolism and vascular obstruction ≥50% (solid line), patients with pulmonary embolism and vascular obstruction <50% (dashed line), and patients without pulmonary embolism (dotted line).
The adjusted odds ratios of death for patients with pulmonary embolism, as a function of time since diagnosis and of the extent of pulmonary vascular obstruction at baseline, are reported in Table 4. It should be remembered that these data, as opposed to those reported in Figure 3, are adjusted for the potential confounding effect of comorbidities and the other modifying covariates, and provide information about the cumulative, not instantaneous, risk of death.
TABLE 4: Estimated Odds Ratios of Death in Patients With Pulmonary Embolism
In patients with pulmonary embolism and vascular obstruction ≥50%, the odds of dying within a day or within a week were, respectively, 8-fold and 4-fold higher than in patients without pulmonary embolism. At 1 month from the incident event, the adjusted cumulative risk of death for these patients was twice as much that of patients without pulmonary embolism (see Table 4). Thereafter, the excess risk of death tapered off over time such that, by 1 year of diagnosis, the risk of death among patients with pulmonary embolism was no higher than among patients without embolism.
For patients with pulmonary embolism who had a baseline vascular obstruction <50%, the adjusted odds of death were not significantly different from those of patients without pulmonary embolism (see Table 4).
Independent predictors of reduced survival in the 5 logistic regression models were male sex, age ≥75 years, prolonged immobilization, chronic diseases other than cardiopulmonary, and malignancy. For patients with active malignancy at presentation, the adjusted odds ratio of death was 3.8 (95% CI, 2.0-7.2) at 1 month, and 9.2 (95% CI, 5.8-14.6) at 1 year. Conversely, recent surgery was an independent predictor of increased survival with an adjusted odds ratio of death of 0.33 (95% CI, 0.12-0.87) at 1 week, and 0.26 (95% CI, 0.16-0.44) at 1 year.
Recurrent Pulmonary Embolism
Recurrent pulmonary embolism was diagnosed in 30 (9.4%) of the 320 patients with pulmonary embolism at inclusion. Twenty-eight patients had a single episode of recurrence, and 2 patients had 3 recurrences each. Overall, there were 34 recurrences, of which 9 were fatal. The actual distribution of the episodes of recurrent pulmonary embolism was as follows: 6 recurrences within 1 week (of which 2 were fatal), 9 within 2 weeks (1 fatal), 7 within 1 month (3 fatal), 5 within a year (1 fatal), and 7 beyond 1 year (2 fatal). The cumulative incidence of recurrent pulmonary embolism was 2.0% (95% CI, 0.9%-4.3%) at 1 week, 6.4% (95% CI, 4.0%-9.8%) at 1 month, 8.9% (95% CI, 6.1%-12.8%) at 1 year, and 10.2% (95% CI, 7.2%-14.3%) at 2 years. None of the patients without pulmonary embolism at presentation had symptomatic episodes of pulmonary embolism during follow-up.
Restoration of Pulmonary Perfusion
Of the 244 patients who survived a full year after acute pulmonary embolism, 235 (96.3%) completed the 1-year scintigraphic follow-up. At the time of diagnosis, these patients had a median age of 71 years (range, 16-95 yr), a median pulmonary vascular obstruction of 42.3% (range, 8.2%-72.9%), and a median PaO2 of 64 mm Hg (range, 33-107 mm Hg). For most patients (216/235, or 91.9%), treatment consisted of a 1-week heparin infusion followed by oral anticoagulation for 1 year. Of the 19 other patients, 1 received thrombolytic therapy, and 18 were given low molecular weight heparins followed by 1-year oral anticoagulation.
As shown in Figure 4A, the extent of scintigraphically detectable pulmonary vascular obstruction decreased progressively over time. At 1 month of diagnosis, 90% of the patients had a residual vascular obstruction ≤30%. After 1 year, the residual pulmonary vascular obstruction was ≤15% in 90% of the patients, and ≤5% in 75%. In 153 (65.1%) of the 235 patients, the lung scan was rated normal consistently by 2 independent observers. Restoration of pulmonary perfusion was associated with a considerable improvement in arterial oxygenation (Figure 4B). At 1 year, the median PaO2 was 84 mm Hg, with 75% of the patients having a PaO2 >78 mm Hg.
FIGURE 4: Box-and-whiskers plots of extent of scintigraphically detectable pulmonary vascular obstruction (A), and partial pressure of oxygen in arterial blood (B) in 235 patients with pulmonary embolism evaluated at diagnosis and after 1 week, 1 month, and 1 year of diagnosis. Line in box: 50th percentile (median); limits of the box: 25th and 75th percentile; whiskers: 10th and 90th percentile.
Chronic Thromboembolic Pulmonary Hypertension
Four (1.3%) of the 320 patients with pulmonary embolism at presentation met the criteria for chronic thromboembolic pulmonary hypertension. In 3 patients the diagnosis was established within 6 months, and in 1 within a year of study entry. At the time of inclusion, these patients had a median age of 58 years (range, 55-71 yr), and a median extent of pulmonary vascular obstruction of 59.7% (range, 52.3%-74.5%). The latter did not appreciably change in lung scans taken during follow-up, and there was no evidence of recurrent pulmonary embolism. An example of persistent perfusion defects in a patient with postembolic pulmonary hypertension is given in Figure 5. Figure 6 shows, for comparison, the time course of perfusion restoration after acute pulmonary embolism. Most likely these 4 patients had had prior episodes of pulmonary embolism that remained undiagnosed for the lack of a prompt suspicion of the disease. As a matter of fact, 3 of them had had documented episodes of deep vein thrombosis, but none of them had undergone any objective testing for suspected pulmonary embolism before entering the study. One patient died of heart failure 18 months after inclusion, and 3 were censored. One of them underwent bilateral pulmonary thromboendarterectomy within a year of study entry.
FIGURE 5: Perfusion lung scans of a patient with postembolic pulmonary hypertension. Images are (from left to right) anterior, right lateral, and left lateral. Row A shows the images taken on study entry, and row B, those taken 6 months later. The persistence of large bilateral perfusion defects, virtually unchanged over time, is evident. Age of emboli at study entry is a likely explanation.
FIGURE 6: Perfusion lung scans of a patient with acute pulmonary embolism. Images are (from left to right) anterior, right lateral, and left lateral. Row A features the images taken on study entry, showing bilateral wedge-shaped perfusion defects and adjacent areas of overperfusion featuring a wedge-shaped configuration. In row B are the images taken 1 month later, showing a complete restoration of pulmonary perfusion and a physiologic base-to-apex distribution of blood flow.
DISCUSSION
In the present study, we followed over time a sample of consecutive patients suspected of having pulmonary embolism, and we compared the survival rates of patients with proven pulmonary embolism with those who had the diagnosis excluded. The 2 diagnostic groups featured similar characteristics as regards age, gender, and location at the time of symptoms onset. Even though the patients with pulmonary embolism were fewer in our sample than in other reports6,7, they were representative of a broad spectrum of the disease severity (from minor to massive). We, therefore, modeled the probability of surviving in patients with pulmonary embolism as a function of the extent of pulmonary vascular obstruction at baseline. In estimating the severity of vascular obstruction, we applied a scintigraphic method12, originally validated against pulmonary angiography, that has been used in other broad prospective studies19,23,27.
We found that massive pulmonary embolism (vascular obstruction ≥50%) is a risk factor for mortality within the first few days after onset, but subsequently has no significant effect on survival. However, the cumulative risk of death for patients with massive pulmonary embolism remained substantial for up to 1 month after onset (adjusted odds ratio, 2.08) because it was affected by the higher proportion of deaths occurring shortly after the incident event. As shown in Table 4, nearly 1 year was required for the cumulative risk of death to be cleared from the effect of the initial higher mortality rate. By contrast, the adjusted risk of death for patients with minor or moderate pulmonary embolism (vascular obstruction <50%) was no higher than in patients without pulmonary embolism at any time after onset.
In the International Cooperative Pulmonary Embolism Registry (ICOPER), 405 (39%) of 1035 patients with pulmonary embolism, who underwent transthoracic echocardiography within 24 hours of diagnosis, showed signs of right ventricular hypokinesis13. In these patients, the 30-day survival rate was 83.7%, a figure that compares favorably with our 30-day survival rate of 85.9% in patients with massive pulmonary embolism. In multivariate analysis, right ventricular hypokinesis remained an independent predictor of 30-day mortality (hazard ratio, 1.94).
Only a minority of the patients in our sample underwent transthoracic echocardiography at the time of their enrollment in the study, so we could not assess the prognostic value of right ventricular hypokinesis. However, we found that, among patients with pulmonary embolism, the extent of scintigraphically detectable pulmonary vascular obstruction at baseline was significantly greater in those who had echocardiographic findings of right ventricular dysfunction than in those without17.
In the current study, the 1-year cumulative incidence of recurrent pulmonary embolism (8.9%) was similar to that reported by others3,19. Most often, pulmonary embolism recurred within the first month of diagnosis. In a few patients, however, it continued to recur for up to 2 years from the initial event. Patients with late recurrences often had predisposing risk factors for pulmonary embolism such as older age, disabling neurologic disorders, or active cancer.
In our sample, the rate of mortality from pulmonary embolism was nearly 3-fold that reported by Carson et al3. In that study, 9 of the 10 patients whose death was attributed to pulmonary embolism died as a consequence of a recurrent episode of embolism. In our study, most of the patients who died from pulmonary embolism (28 of 31) did so within a month of study entry. Among them, 6 had documented episodes of recurrent fatal pulmonary embolism (2 within a week and 4 within a month). Of the 22 other patients whose death was attributed to the initial episode of pulmonary embolism, 16 had massive pulmonary embolism (pulmonary vascular obstruction ≥50%), which, in itself, may have caused their death within a few days after onset. Since we did not obtain autopsy verification in all patients whose death was attributed to pulmonary embolism, we cannot absolutely exclude the possibility of having missed occult cases of recurrence. Yet, such underestimation ought to be small. Indeed, as reported above, the 1-year cumulative incidence of recurrent pulmonary embolism in the current study was very close to that reported by Carson et al3, and Pengo et al19.
A further difference between the study by Carson et al3 and the current study is the rate of mortality due to the underlying comorbidities. In the study by Carson et al3, the most frequent causes of death were cancer (34.7%), infection (22.1%), and cardiac diseases (16.8%). Overall, they accounted for nearly 74% of all deaths. In the current study, instead, cancer, infection, and cardiac diseases accounted for some 57% of all deaths in patients with pulmonary embolism (p < 0.01 by chi-square test). Such a difference is remarkable inasmuch as the data reported by Carson et al refer to a 1-year follow-up, whereas our data span a much longer period of follow-up. Thus, it appears that in Carson's study the underlying diseases, rather than pulmonary embolism itself, caused most of the reported deaths.
Chronic thromboembolic pulmonary hypertension was diagnosed in only 1% of the patients with pulmonary embolism in our sample. All cases were identified within a year of study entry. In the long-term prospective study by Pengo and coworkers19, 7 (3%) of 223 patients, who reportedly had a first episode of symptomatic pulmonary embolism, developed chronic thromboembolic pulmonary hypertension during follow-up. The cumulative incidence of such complication was 3% at 1 year and close to 4% at 2 years of the initial event, with no subsequent increase in incidence. In that study, however, patients with pulmonary embolism featured a 1-year mortality rate that was substantially lower than our patients' (13.4% versus 24%). This may, in part, explain the different incidence of chronic thromboembolic pulmonary hypertension in the 2 studies.
In the study by Pengo el al19, however, no lung scan follow-up was pursued, at predefined time intervals, to assess the restoration of pulmonary perfusion in patients who allegedly had a first episode of pulmonary embolism. This precluded the possibility of identifying patients with persistent perfusion defects in sequential scans likely reflecting old emboli from a previous unrecognized episode of pulmonary embolism.
It is unlikely that we missed cases of chronic thromboembolic pulmonary hypertension among patients who survived a full year after pulmonary embolism because most of these patients showed a nearly complete restoration of pulmonary perfusion, with 65% having no residual perfusion defects on the lung scan. Recovery of pulmonary perfusion was accompanied by a substantial improvement in pulmonary gas exchange. However, we cannot exclude the possibility that we failed to identify patients who may have developed chronic thromboembolic pulmonary hypertension at a later stage.
In summary, we found that pulmonary embolism featuring a vascular obstruction ≥50% is a strong, independent predictor of reduced short-term survival. It should be considered, however, that our results refer to a sample of patients most of whom were hospitalized at the time of study entry, so they may not apply to other clinical settings (for example, the community, nursing home, or other chronic care facilities) where pulmonary embolism may occur. Nevertheless, our findings, in conjunction with others'11-13,20, underscore the need for risk stratification of patients presenting with an acute episode of pulmonary embolism. This may prove useful in probing more effective strategies of treatment. In particular, our data emphasize the need for most accurate clinical identification of patients at risk of pulmonary embolism in order to establish a prompt diagnosis or exclusion of the disease.
Once a diagnosis of pulmonary embolism is established, monitoring the resolution of pulmonary emboli by means of a suitable imaging technique is, in our experience, an effective strategy for identifying patients with persistent large perfusion defects who may be at risk of chronic thromboembolic pulmonary hypertension.
ACKNOWLEDGMENTS
The authors thank the following physicians who took part in the study: Germana Allescia, Laura Carrozzi, Giosuè Catapano, Giorgio Di Ricco, Erica Filippi, Bruno Formichi, Carlo Marini, Massimo Pistolesi, Renato Prediletto, Luigi Rizzello.
REFERENCES
1. Alpert JS, Smith R, Carlson J, Ockene IS, Dexter L, Dalen JE. Mortality in patients treated for pulmonary embolism.
JAMA. 1974;229:1606-1613.
2. Anderson FA Jr, Wheeler HB, Goldberg RJ, Hosmer DW, Patwardhan NA, Jovanovic B, Forcier A, Dalen JE. A population-based perspective on the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism: the Worcester DVT Study.
Arch Intern Med. 1991;151:933-938.
3. Carson JL, Kelley MA, Duff A, Weg JG, Fulkerson WJ, Palevsky HI, Schwartz JS, Thompson BT, Popovich J, Hobbins TE, Spera MA, Alavi A, Terrin ML. The clinical course of pulmonary embolism.
N Engl J Med. 1992;326:1240-1245.
4. Collett D.
Modelling Survival Data in Medical Research. London: Chapman & Hall, 1994.
5. Giuntini C, Di Ricco G, Marini C, Melillo E, Palla A. Pulmonary embolism: epidemiology.
Chest. 1995;107:3S-9S.
6. Goldhaber SZ, Visani L, De Rosa M, for ICOPER. Acute pulmonary embolism: clinical outcomes in the International Cooperative Pulmonary Embolism Registry (ICOPER).
Lancet. 1999;353:1386-1389.
7. Heit JA, Silverstein MD, Mohr DN, Petterson TM, O'Fallon WM, Melton LJ III. Predictors of survival after deep vein thrombosis and pulmonary embolism. A population-based cohort study.
Arch Intern Med. 1999;159:445-453.
8. Hull RD, Raskob GE, Coates G, Panju AA. Clinical validity of a normal perfusion scan in patients with suspected pulmonary embolism.
Chest. 1990;97:23-26.
9. Kipper MS, Moser KM, Kortman KE, Ashburn WL. Long term follow-up of patients with suspected pulmonary embolism and a normal lung scan.
Chest. 1982;82:411-415.
10. Kniffin WD Jr, Baron JA, Barrett J, Birkmeyer JD, Anderson FA Jr. The epidemiology of diagnosed pulmonary embolism and deep vein thrombosis in the elderly.
Arch Intern Med. 1994;154:861-866.
11. Kucher N, Wallmann D, Windecker S, Meier B, Hess OM. Incremental prognostic value of troponin I and echocardiography in patients with acute pulmonary embolism.
Eur Heart J. 2003;24:1651-1656.
12. Kucher N, Goldhaber SZ. Cardiac biomarkers for risk stratification of patients with acute pulmonary embolism.
Circulation. 2003;108:2191-2194.
13. Kucher N, Rossi E, De Rosa M, Goldhaber SZ. Prognostic role of echocardiography among patients with acute pulmonary embolism and systolic arterial pressure of 90 mmHg or higher.
Arch Intern Med. 2005;165:1777-1781.
14. Meyer G, Collignon MA, Guinet F, Jeffrey AA, Barritault L, Sors H. Comparison of perfusion lung scanning and angiography in the estimation of vascular obstruction in acute pulmonary embolism.
Eur J Nucl Med. 1990;17:315-319.
15. Miniati M, Pistolesi M, Marini C, Di Ricco G, Formichi B, Prediletto R, Allescia G, Tonelli L, Sostman HD, Giuntini C. Value of perfusion lung scan in the diagnosis of pulmonary embolism: results of the Prospective Investigative Study of Acute Pulmonary Embolism Diagnosis (PISA-PED).
Am J Respir Crit Care Med. 1996;154:1387-1393.
16. Miniati M, Prediletto R, Formichi B, Marini C, Di Ricco G, Tonelli L, Allescia G, Pistolesi M. Accuracy of clinical assessment in the diagnosis of pulmonary embolism.
Am J Respir Crit Care Med. 1999;159:864-871.
17. Miniati M, Monti S, Pratali L, Di Ricco G, Marini C, Formichi B, Prediletto R, Michelassi C, Di Lorenzo M, Tonelli L, Pistolesi M. Value of transthoracic echocardiography in the diagnosis of pulmonary embolism: results of a prospective study in unselected patients.
Am J Med. 2001;110:528-535.
18. Miniati M, Monti S, Bottai M. A structured clinical model for predicting the probability of pulmonary embolism.
Am J Med. 2003;114:173-179.
19. Pengo V, Lensing AWA, Prins MH, Marchiori A, Davidson MPH, Tiozzo F, Albanese P, Biasolo A, Pegoraro C, Iliceto S, Prandoni P. Incidence of chronic thromboembolic pulmonary hypertension after pulmonary embolism.
N Engl J Med. 2004;350:2257-2264.
20. Schoepf UJ, Kucher N, Kipfmueller F, Quiroz R, Costello P, Goldhaber SZ. Right ventricular enlargement on chest computed tomography: a predictor of early death in acute pulmonary embolism.
Circulation. 2004;110:3276-3280.
21. Siddique RM, Siddique MI, Connors AF Jr, Rimm AA. Thirty-day case-fatality rates for pulmonary embolism in the elderly.
Arch Intern Med. 1996;1564:2343-2347.
22. Silverstein MD, Heit JA, Mohr DN, Petterson TM, O'Fallon WM, Melton LJ III. Trends in the incidence of deep vein thrombosis and pulmonary embolism: a 25-year population-based study.
Arch Intern Med. 1998;158:585-593.
23. Simonneau G, Sors H, Charbonnier B, Page Y, Laaban JP, Azarian R, Laurent M, Hirsch JL, Ferrari E, Bosson JL, Mottier D, Beau B, for the THESEE Study Group. A comparison of low-molecular-weight heparin with unfractionated heparin for acute pulmonary embolism.
N Engl J Med. 1997;337:663-669.
24. The Task Force on Diagnosis and Treatment of Pulmonary Arterial Hypertension of the European Society of Cardiology. Guidelines on diagnosis and treatment of pulmonary arterial hypertension.
Eur Heart J. 2004;25:2243-2278.
25. van Beek EJR, Kuijer PMM, Schenk BE, Brandjes DPM, ten Cate JW, Buller HR. A normal perfusion lung scan in patients with clinically suspected pulmonary embolism: frequency and clinical validity.
Chest. 1995;108:170-173.
26. van Beek EJR, Kuijer PMM, Buller HR, Brandjes DPM, Bossuyt PMM, ten Cate JW. The clinical course of patients with suspected pulmonary embolism.
Arch Intern Med. 1997;157:2593-2598.
27. Wartski M, Collignon MA. Incomplete recovery of lung perfusion after 3 months in patients with acute pulmonary embolism treated with antithrombotic agents. THESEE Study Group. Tinzaparin ou Heparin Standard: Evaluation dans l'Embolie Pulmonaire Group.
J Nucl Med. 2000;41:1043-1050.
