Robert B Tompkins MDa, Vaughn Harris MDa, Clayton Brown, MD, David E Griffith MDb
Correspondence to Robert B Tompkins, MD. Email: [email protected]
SWRCCC 2014;2(8):5-16
doi: 10.12746/swrccc2014.0208.097
...................................................................................................................................................................................................................................................................................................................................
We recently identified four patients with pulmonary embolism (PE) who presented with progressive, but nonspecific, respiratory symptoms and were initially diagnosed with an exacerbation of established or presumed airflow obstruction. No specific explanation for the apparent respiratory exacerbation was found and each patient failed to respond to bronchodilator and/or antibiotic therapy, thereby triggering an evaluation for PE. Pulmonary embolism occurs with disturbing frequency in the setting of an apparent chronic airflow obstruction exacerbation without a clear explanation. Suspicion of PE by the clinician is essential as patient symptoms and routine laboratory evaluation do not always differentiate PE from exacerbation of airflow obstruction. D-dimer determination is a reasonable first diagnostic step with subsequent testing directed by the D-dimer result in the context of an individualized assessment of PE risk. The diagnosis of PE in this setting is critically important for avoiding the attendant morbidity and mortality associated with untreated PE.
...................................................................................................................................................................................................................................................................................................................................
The classic clinical presentation of acute pulmonary embolism (PE) is familiar to most clinicians. When a patient with known risk factors for thromboembolic disease, such as malignancy, recent surgery, or immobilization, presents with unilateral lower extremity edema and/or inflammation, acute shortness of breath, pleuritic chest pain, and tachycardia, PE quickly rises to the top of the differential diagnosis list, and the appropriate workup is initiated. Unfortunately, these ideal clinical circumstances are frequently not present in patients with PE, especially in those with suspected or diagnosed underlying lung disease. Consequently, PE may be initially overlooked in favor of empiric treatment for an exacerbation of underlying lung disease. In fact, in patients with chronic obstructive pulmonary disease (COPD) who present with a presumed COPD exacerbation, there are no significant differences in symptoms between patients with and without PE.1-3 Given these difficulties, PE should be included in the differential diagnosis for COPD patients when the clinical presentation is not readily explained by common factors that exacerbate airflow obstruction, such as bronchospasm, pneumonia, or heart failure, or for patients who do not readily improve with targeted therapy for these common exacerbating factors. The timely diagnosis of PE is further complicated by a paucity of tests that are sensitive and specific for diagnosing PE in patients with chronic airflow obstruction. The consequences of missed diagnosis of PE for patients with chronic lung disease can be dire and include increased risk for recurrent PE, increased short and long term mortality, aggravated pulmonary hypertension, and right ventricular dysfunction.4-11
We recently encountered four non-smoking patients with previously diagnosed or suspected airflow obstruction who were admitted with respiratory deterioration and were ultimately diagnosed with PE. The major objective of this manuscript is to review circumstances in which acute and/or recurrent PE might occur in patients with established or suspected airflow obstruction. These circumstances notably include an apparent exacerbation of the patient’s airflow obstruction with no identifiable exacerbating factor(s) or poor response to therapy directed at exacerbating factors for airflow obstruction. We also suggest diagnostic strategies for PE under these circumstances.
TABLE 1: CASE SUMMARIES
Pulmonary embolism in patients with COPD is not rare. The reported incidence from older studies of PE found post mortem in COPD patients ranged from 28% to 51%, .12-13 More recent studies suggest that the true frequency of PE in COPD patients in whom PE is clinically suspected ranges from 19% to 29%.1-3 Lesser and colleagues examined the characteristics of 108 patients with COPD and suspected PE, of whom 19% (21/108) were diagnosed with PE.14 In addition, Tille-Leblond, et al. found that in a series of 197 consecutive patients with COPD referred for severe exacerbation of unknown origin, 25% had PE on CT.2 Risk factors and clinical symptoms, such as change in dyspnea, pleuritic pain, hemoptysis, tachycardia and lower extremity edema, were no more associated with PE than with COPD without PE. The level of hypoxemia, the level of pCO2, and the alveolar-arterial (A-a) gradient were also no more associated with PE than with COPD without PE. However, a decrease in PCO2 of at least 5 mm, previous thromboembolic disease, and malignant disease were associated with PE. Because the symptoms of PE and COPD exacerbation are essentially identical, it is not surprising that the possibility of PE in patients with COPD is sometimes over-looked. Patients with chronic airflow obstruction are at risk for other manifestations of venous thromboembolic disease. Shetty, et al. found that patients admitted for acute COPD exacerbation were at increased risk for DVT because of immobilization, advanced age, smoking, underlying malignancy, and RV failure. 15 They identified 668 patients with COPD and DVT and compared them with 4000 patients with DVT but without COPD. Patients with COPD and DVT were likely to be older, male, inpatient in the ICU on mechanical ventilation, with an IVC filter, and with greater medical acuity. Stein, et al. found that VTE (PE and DVT) is frequently under-diagnosed in patients with COPD and that the relative risk is higher in younger patients compared to older patients who often have multiple other risk factors for VTE.16
Previously published identifiable risk factors for PE include immobilization, surgery within three months, stroke, paresis, paralysis, central venous instrumentation within three months, malignancy, chronic heart disease, autoimmune diseases, history of venous thromboembolism, obesity in women, cigarette smoking, and hypertension.17-19 Chronic obstructive pulmonary disease is not included on that list, although patients with chronic airflow obstruction with hypoxia may have subclinical or overt pulmonary hypertension and RV dysfunction, which then promotes venous stasis and thromboembolism. Similarly, obstructive sleep apnea (OSA), a disease associated with chronic hypoxia and common to the patients in this report, has recently been associated with increased risk of venous thromboembolic disease and hypercoagulability.20-21 The apparent association of venous thromboembolic disease and OSA is particularly worrisome due to the frequent under-diagnosis of OSA.
At some point during the hospitalization of a patient with an exacerbation of diagnosed or presumed airflow obstruction, the question must be asked: is there an explanation for this respiratory exacerbation and is the patient responding (improving) appropriately to the therapeutic regimen? For the majority of patients presenting with apparent airflow obstruction exacerbation, the answers will be affirmative. However, for patients without an identifiable etiology for airflow obstruction exacerbation, or for those who are not responding appropriately to therapy directed at the assumed exacerbating factor, consideration must be given to the possibility of PE. While there is literature about the occurrence of PE with underlying COPD, there is little published about PE in the presence of other chronic lung diseases, such as pulmonary fibrosis or bronchiectasis. Presumably, the same considerations would be pertinent for those disorders as well; that is, PE should be suspected in the absence of a clear exacerbating factor or with poor response to usual or ”standard” therapy.
There are significant consequences for missing the diagnosis of PE. Untreated PE is associated with a mortality rate of approximately 30% with recurrent PE the most common cause of death.5-7 Pulmonary embolism also increases the rate of death from COPD at one year.22 Persistent RV dysfunction is also more likely associated with a recurrent PE, recurrent DVT, and a two fold increase in PE-related mortality compared to patients with no RV dysfunction or RV dysfunction with regression. 8-9
The diagnosis of an acute pulmonary embolism is a clinical challenge in the setting of underlying lung disease because many of the symptoms are vague and non-specific, physical findings can be lacking, and diagnostic tests have limited sensitivity and specificity.1-3 It cannot be overstated that the most important aspect of PE diagnosis in the setting of established or even presumed underlying lung disease is suspicion of the diagnosis. Routine tests such as arterial blood gases (ABGs), electrocardiograms (ECGs), chest radiographs, complete blood counts (CBC), and chemistry profiles are usually not helpful in the diagnosis of PE. If prior or baseline ABGs are available, a reduction in the PCO2 level of > 5mm is significantly associated with PE in COPD patients.2
Prior to the advent of helical-CT angiography, ventilation-perfusion (V/Q) lung scanning was the only non-invasive radiographic study for diagnosing pulmonary embolism. Ventilation-perfusion scans are helpful for excluding PE in patients with low pre-test probability of PE and low-probability scan results. Alternatively, they are also helpful in diagnosing patients with high pre-test probability of PE and high-probability scan results.23 However, it is well-recognized that the V/Q scan is sometimes limited in usefulness for patients with underlying lung diseases that are associated with severe and sometimes diffuse V/Q mismatching, resulting in V/Q scan abnormalities that cannot be distinguished from those caused by PE. A normal V/Q scan virtually excludes PE. However, that is a rare finding in patients with underlying lung disease.
Doppler-ultrasound evaluation of the lower extremities for evidence of DVT can give additional evidence to help diagnose or exclude PE. In patients with symptoms, elevated D-dimer and an ultrasound-Doppler positive for DVT, the diagnosis of PE is quite probable and treatment can be started. An ultrasound-Doppler that is negative for DVT does not exclude the diagnosis of PE in patients with symptoms or risk factors for PE. In one series, lower extremity ultrasound identified 51% of patients with PE in the setting of unexplained COPD exacerbation by diagnosing a lower extremity DVT.2
The blood test for fibrin split products, D-dimer, is usually elevated in PE with sensitivities ranging from 85-98% but with low specificity in the range of 40%.24-26 Evidence indicates that a D-dimer level <500 ng/mL is sufficient to exclude PE in patients with a low clinical pretest probability of PE. A negative D-dimer is not sufficient to exclude PE in patients with a high pretest probability of PE and has variable performance for those with a moderate pretest probability of PE.24-25 A positive result requires further assessment in order to diagnose or rule out PE. D-dimer levels can be high among hospitalized patients, especially those with malignancy or recent surgery.26-27 Table 2 is a partial list of disorders associated with increased plasma levels of fibrin D-dimer. Quantitative ELISA and semi-quantitative latex agglutination (“sensitive D-dimer testing”) are preferred because their high negative predictive value indicates that they are superior to erythrocyte agglutination for excluding PE.28 There is no evidence COPD, per se, alters D-dimer determination.29
TABLE 2: CONDITIONS ASSOCIATED WITH ELEVATED D-DIMER LEVELS
Brain natriuretic peptide (BNP) levels are typically increased in patients with PE compared to patients without PE. However, many patients with PE do not have elevated BNP, and there are many alternative causes of an elevated BNP.30-32 BNP levels do predict RV dysfunction and mortality. In one study, short-term mortality was increased six times among patients with BNP > 100 pg/ml and 16 times among patients with an NT—pro BNP > 600 pg/ml.32
Helical CT scanning is now the most frequently used test for evaluating PE in patients with normal renal function. It can quickly evaluate for intraluminal filling defects within the pulmonary arterial tree with high sensitivity for PE in the lobar or main pulmonary arteries.33-35 This test may be non-diagnostic when the PE involves segmental or sub-segmental arteries. A normal helical CT does not completely exclude the diagnosis of PE and is similar to a “low-probability” V/Q scan. Numerous studies have shown, however, that there is a low risk of PE following a negative CT-PA.33-36 While imperfect, this test is currently considered the “gold standard” for radiographic confirmation or exclusion of PE. Chronic airflow obstruction also does not affect the diagnostic performance of CTA or pulmonary angiography. 28 A coronal cut from the CTA of patient B showing a large proximal right main pulmonary artery thromboembolism is shown in Figure 1.
FIGURE 1: CT PULMONARY ANGIOGRAM STUDY SHOWING LARGE PROXIMAL RIGHT MAIN PULMONARY ARTERY THROMBOEMBOLISM
Pulmonary angiography was previously the “standard” for diagnosis of pulmonary embolism but is a costly, invasive procedure with many potential serious side effects and a reported mortality rate of approximately 0.5%.37 It is contraindicated in patients with renal impairment or a known allergy to the contrast material, and many health care institutions do not have access to this test. Pulmonary angiography for the diagnosis of PE has been largely supplanted by the widespread availability of CT-PA.
A summary of advantages and disadvantages of the available diagnostic studies for PE in patients with suspected or previously diagnosed airflow obstruction is shown in Table 3, and a suggested diagnostic algorithm for PE in these patients is shown in Figure 2.
TABLE 3: ADVANTAGES AND DISADVANTAGES OF DIAGNOSTIC STUDIES FOR PULMONARY EMBOLISM IN PATIENTS WITH UNDERLYING LUNG DISEASE
FIGURE 2 ALGORITHM FOR EVALUATION AND TREATMENT OF PULMONARY EMBOLISM IN PATIENTS WTIH EXACERBATION OF PRESUMED OR PREVIOUSLY DIAGNOSED AIRFLOW OBSTRUCTION.
Our cases highlight several important points for diagnosing PE in patients either diagnosed with or suspected of having underlying airflow obstruction. All four patients presented with a presumed exacerbation of airflow obstruction as a consequence of bronchospasm and/or infection. Pulmonary embolism was suspected only after each patient failed to improve clinically, despite initial therapy directed at the presumed exacerbating factor(s). It is instructive that three of the patients were non-smokers, and two did not have a previous diagnosis of obstructive lung disease. In spite of the lack of an established diagnosis of airflow obstruction, these two patients were presumed to have exacerbation of airflow obstruction at presentation, underscoring the non-specific nature of PE symptoms and the similarity between PE symptoms and symptoms of airflow obstruction. Three of the four patients had an abnormal D-dimer that helped direct further evaluation for PE. Although the fourth patient had a normal D-dimer level, the clinical suspicion for PE was sufficiently high in this patient to continue an aggressive diagnostic strategy for PE. This patient illustrates another important point, namely that the aggressiveness with which the diagnosis of PE is pursued is always driven by the underlying clinical suspicion for PE. The presence of OSA in our patients also likely increased their risk for PE and should be considered an additional risk factor for PE.19-20 This association adds one more reason to an already long list for carefully assessing all patients for OSA.
There are multiple published guidelines for treating PE.38-40 Therapy begins with timely evaluation of the airway, breathing, and circulation in conjunction with provision of supplemental oxygen. An initial bolus of intravenous fluid should be used for patients with hypotension (SBP < 90mmHg or a drop in SBP >40mmHg from baseline), and IV vasopressors may be needed if hypotension persists despite the initial fluid resuscitation.
Anticoagulant therapy is the mainstay of treatment for venous thromboembolism including PE and produces a passive reduction in thrombus size. Many pharmaceutical agents are available, including unfractionated heparin, low-molecular weight heparin, and warfarin, as well as newer agents, including dabigatran, rivaroxaban, apixaban, and fondaparinux.38-40 For patients with suspected or known heparin-induced thrombocytopenia, a non-heparin based anticoagulant should be used such as lepirudin, argatroban, or bivalirudin.40-41 A recent meta-analysis of clinical and safety outcomes with various venous thromboembolism therapies suggests that an unfractionated heparin/vitamin K antagonist regimen is less effective, while rivaroxaban and apixaban regimens are associated with a lower risk of bleeding.42
In contrast to anticoagulant therapy, thrombolytic therapy actively reduces thrombus size by increasing the conversion of plasminogen into plasmin which then cleaves fibrin molecules.43 Thrombolysis is usually reserved for clinically unstable patients since it is associated with a significant risk of bleeding events which may be fatal or cause significant morbidity. Studies have shown that treatment with fibrinolysis produces faster restoration of lung perfusion and, thereby, provides a more rapid improvement in clinical symptoms.44-46
Surgical embolectomy remains an alternative to pharmaceutical thrombolysis in clinically unstable patients with massive or sub-massive PE and evidence of RV dysfunction when either thrombolysis fails or is contraindicated.47-49
Finally, for patients who are not candidates for surgical embolectomy, anticoagulation, thrombolysis, or have had previous bleeding events or recurrent VTE while on anticoagulation, inferior vena cava filters may be used. Notably, IVC filter use has increased in recent decades50-51 while the benefits of IVC filters remain controversial.50-53 The PREPIC Trial showed that IVC filters reduced the incidence of recurrent PE at 12 days and eight years, but were also associated with increased recurrent DVT at two years with no overall effect on mortality.52
Preventing lower extremity DVT is the most important measure we can undertake to prevent PE. In hospitalized patients, anticoagulant DVT prophylaxis is recommended for acutely-ill medical patients at increased risk of thrombosis.54 The level of risk is based on a score derived from a risk assessment model that does not include either COPD or OSA. Age ≥ 70 years and obesity (BMI ≥ 30), risk factors commonly found in COPD and OSA patients, are included in this model, as is respiratory failure, but the combination of these three risk factors does not produce a score deemed “high risk.” Due to the inherent limitation in mobility associated with COPD and OSA, we suggest that these co-morbid conditions be considered in the assessment of risk for venous thromboembolism and that DVT prophylaxis is indicated for hospitalized COPD and OSA patients
Patients with presumed or established airflow obstruction frequently present with worsening respiratory symptoms attributed initially to exacerbation of the underlying airflow obstruction. For patients without a confirmed etiology for the exacerbation, such as bronchitis, pneumonia, or bronchospasm, or for those who do not respond favorably to therapy directed at the presumed exacerbating factor, PE will be a common finding as these patients frequently do not present with specific subjective or objective clues to the diagnosis of PE. While initial clinical evaluation may reasonably begin with obtaining a D-dimer and/or Doppler-USG of the legs, it is the clinical suspicion of PE by the health care provider that remains the critical element for diagnosis of PE. Prompt diagnosis and treatment of PE can improve the patient’s short and long term prognosis.
Key words- COPD, pulmonary embolus, diagnosis.
...................................................................................................................................................................................................................................................................................................................................
Received: 09/21/2014
Accepted: 09/29/2014
Reviewers:Gilbert Berdine MD, Kenneth Nugent MD
Published electronically: 10/15/2014
Conflict of Interest Disclosures: none