Fibrinogen Levels and Bleeding Risk in Patients Undergoing Ultrasound-Assisted Catheter-Directed Thrombolysis for Submassive Pulmonary Embolism

Objectives. We sought to test the hypothesis that patients undergoing ultrasound-assisted catheter-directed thrombolysis (USAT) with standard alteplase and heparin dosing would not develop significant depletion of systemic fibrinogen, which may account for the lower risk of bleeding seen in contemporary trials. We also sought to compare the relative outcomes of individuals with submassive pulmonary embolism (PE) undergoing USAT and anticoagulation alone. Methods. Utilizing a single-center prospective registry, we identified 102 consecutive adult patients with submassive PE who were considered for USAT based on a standardized treatment algorithm between November 2016 and May 2019. Patients not receiving USAT therapy were treated with anticoagulation alone. Results. Baseline characteristics were generally similar between groups (n = 51 in each group). Major bleeding rates were not significantly different between groups (2.0% vs 5.9% in USAT vs control, respectively; P=.62). Notably, no USAT patient experienced clinically significant hypofibrinogenemia (mean trough fibrinogen, 369.8 ± 127.1 mg/dL; minimum, 187 mg/dL). The mean trough fibrinogen of patients experiencing any bleeding event (major or minor) was 306.6 mg/dL (SE, 23.9 mg/dL) vs 380.3 mg/dL (SE, 20.4 mg/dL) in those without a bleeding event (P=.02). Conclusions. In this cohort analysis of patients undergoing USAT, there was no evidence for clinically significant depletion of fibrinogen or intracranial hemorrhage. Although our data suggest an association between lower fibrinogen levels and bleeding events, our results are not clear enough to suggest a clinically useful fibrinogen cut-off value. Further study is needed to determine the utility of routine fibrinogen monitoring in this population.

J INVASIVE CARDIOL 2021;33(9):E702-E708. Epub 2021 June 10.

Key words: PE, PESI score, submassive pulmonary embolism, ultrasound-assisted catheter-directed thrombolysis

Ultrasound-assisted catheter-directed thrombolysis (USAT) for the treatment of submassive pulmonary embolism (PE) has been shown to improve right ventricular (RV) strain and hemodynamics with low risk of bleeding. The ULTIMA study, the first published research on the clinical efficacy of USAT, included only 30 patients in the intervention arm and demonstrated significant improvement in the RV to left ventricular (LV) diameter ratio with USAT compared with the anticoagulation (AC)-only control group.¹ The follow-up SEATTLE II trial included 119 patients with submassive PE in a single-arm study and demonstrated similar improvement in RV:LV diameter ratio with very low bleeding/complication rates.² Although a recent meta-analysis identified 24 studies related to the clinical application of catheter-directed interventions for PE, only 6 assessed the efficacy of USAT.³ In the included studies, pooled mortality was 4% (95% confidence interval [CI], 1-11).⁴ Conflicting findings were found subsequently in a meta-analysis that included 20 studies and 945 patients with submassive PE who were treated with catheter-directed therapies.⁴ They found significant improvements in RV function, a 1.4% risk of major bleeding, and 0% mortality at 30 days.⁴ To date, only 1 paper has investigated risk factors for bleeding events in USAT, showing that in the SEATTLE II trial, massive PE and multiple venous access attempts were independently associated with a major risk of bleeding.⁵ Also of note, the OPTALYSE trial, investigating 4 different dosing regimens of tissue plasminogen activator (tPA), found higher rates of bleeding in the highest-dose arm, suggesting a dose-response effect.⁶ 

Fibrinogen is one of a number of proteins important for hemostasis, specifically providing a structural matrix supporting clot formation.⁷ While past studies have demonstrated fibrinogen as a predictor of bleeding complications in thrombolytic therapy for ischemic stroke, myocardial infarction, and peripheral ischemia, only 1 retrospective study with a small number of patients has investigated fibrinogen levels in catheter-directed therapy for PE.⁸ In the 2 largest studies of peripheral thrombolytic therapy, hypofibrinogenemia was associated with an increased risk of bleeding.^9,10 Two smaller studies did not find an association, perhaps as a result of being underpowered.^11,12 The precise level of serum fibrinogen at which a patient is exposed to an increased risk of bleeding is not well established. In clinical practice, a fibrinogen level falling below 150 mg/dL is most often the point at which physicians adjust the dosing of fibrinolytic or AC therapy.¹³ This cut-off was utilized in a number of studies to define hypofibrinogenemia and its association with bleeding.^9,11,14 Based on a number of epidemiologic studies, normal fibrinogen levels are thought to vary between approximately 163 mg/dL and 468 mg/dL, increasing with age and in conditions such as pregnancy.^7,15,16

In this analysis, we sought to test the hypothesis that USAT would not be associated with hypofibrinogenemia and a corresponding increase in the risk of bleeding complications. Likewise, we sought to compare whether fibrinogen level (above the threshold of hypofibrinogenemia) would be correlated with risk of bleeding complications. 

Study design and data collection. This study was conducted at the University of North Carolina Hospitals in Chapel Hill, North Carolina. The study was approved by the University of North Carolina at Chapel Hill institutional review board (IRB). Informed consent was obtained for all patients undergoing USAT. Because this research project involved data collection only and no experimental procedures, IRB approval included a waiver of  informed consent for electronic medical record (EMR) data collection.

At the time of introduction of USAT therapy at our institution in November 2016, a standardized treatment algorithm for PE was developed that included ultrasound-guided fibrinolytic therapy. A prospective registry of all patients with submassive PE considered for catheter-directed therapy was created. For USAT patients, serum fibrinogen was monitored post procedure, generally at 6-hour intervals. We included only patients with submassive PE, as defined by American Heart Association practice guidelines. These guidelines define submassive PE as an acute PE with evidence of myocardial necrosis or right heart dysfunction.¹⁷ The criteria for right heart dysfunction or myocardial necrosis were RV dilation (RV:LV diameter ratio >0.9) on echocardiography or computed tomography, elevation of brain natriuretic peptide (>90 pg/mL), elevation of N-terminal pro- brain natriuretic peptide (>500 pg/mL), elevation of troponin (troponin I >0.4 ng/mL or troponin T >0.1 ng/mL), or electrocardiographic changes.¹⁷ We excluded patients with hemodynamic instability as defined by systolic blood pressure <90 mm Hg for at least 15 minutes or requiring inotropic support, pulselessness, or persistent bradycardia (<40 beats/minute).¹⁸ Patients were also excluded if they received systemic fibrinolytic therapy or if they underwent a non-USAT catheter-based therapy. AC included low-molecular-weight heparin, vitamin K antagonists, direct-acting oral anticoagulants, and fondaparinux. Although patients with recurrent PE were included, we only included 1 index admission, which could have been for the initial PE event or recurrent PE event, if the initial event occurred before the creation of the registry.

Clinical outcomes were determined by individual chart review. Diagnostic imaging reports were reviewed for language confirming RV dysfunction. Echocardiographic reports demonstrating RV dilation and dysfunction were subcategorized by whether the dilation or dysfunction was “mild,” “moderate,” or “severe.” If a patient underwent USAT, we collected the transduced pulmonary artery (PA) systolic, diastolic, and mean pressures at the time of placement and removal.  All data were prospectively entered into a REDCap database (REDCap Consortium) via a standard case report form. 

PE therapies at our institution are guided by a standard algorithm, adopted to ensure the uniform application of therapies in appropriately selected patients. Patients are stratified based on the type of PE (massive, submassive, or neither) and any additional risk factors. Only those with a submassive PE and without contraindications to fibrinolytic therapy were eligible for USAT. Although the algorithm guides therapy, shared decision making — including patient preference, ongoing symptoms, presence of major comorbidities, and the resulting multidisciplinary discussion — mean that not all patients with a submassive PE received USAT. 

USAT procedural details. All patients included in the intervention arm of our study received catheter-based therapy using the EKOS Endovascular System (BTG International Ltd). Patients in the intervention cohort (n = 51) received between 8 and 24 mg of tPA infused via USAT catheter. Our institutional protocol called for the infusion of 12 mg tPA via single USAT catheter for those with a unilateral PE or 24 mg tPA via 2 USAT catheters for those with bilateral or saddle PE. For each catheter, the infusion consisted of a 2 mg bolus followed by 0.75 mg/hr for a total of 12 mg/catheter (24 mg total dose for bilateral intervention). Deviation from this protocol only occurred once, with 1 patient receiving a lower dose based on preplacement bleeding concern. Heparin infusion was continued during tPA infusion, typically at a rate of approximately half the therapeutic infusion rate.

Outcome measurement. We compared study outcomes including length of stay, intensive care unit (ICU) length of stay, readmission, bleeding complications, episodes of hypotension, directly transduced PA pressure, echocardiogram-based RV dilation and dysfunction, discharge on oxygen, and mortality between the cohort receiving USAT and the cohort receiving AC alone. Length of stay was measured based on an inclusive number of days from admission through discharge. ICU length of stay was based on the number of days when an ICU progress note was available in the EMR. We identified bleeding events that occurred during the index admission according to the Bleeding Academic Research Consortium (BARC) criteria. In-hospital mortality was based on a manual chart review of the EMR. Fibrinogen was measured through routinely collected laboratory fibrinogen assays. Pulmonary embolism severity index (PESI) scores were calculated based on medical history and vital sign data collected via chart review. 

Statistical analysis. All data were exported from the collection database to JMP Pro 14 (SAS Institute). We used P-value threshold of <.05 to indicate significance. Missing data were excluded from the analysis. Variables are reported as mean with standard error (SE) or ± standard deviation. We used a Fisher’s Exact test to assess categorical variables. We compared continuous variable means with 1-sided and 2-sided t-tests. Furthermore, we conducted a backward stepwise multivariable logistic regression model to identify predictors of treatment with USAT. 

Data for RV dilation and dysfunction and directly transduced PA pressures were only included if there was a pre- and postintervention value available for the patient (dilation/dysfunction n = 19; PA mean n = 42; PA systolic n = 30). Means for pre- and postintervention PA pressures were calculated and compared via paired 2-sample t-tests in JMP Pro 14. 

Baseline patient characteristics. Based on the inclusion criteria, we identified 102 consecutive adult patients with submassive PE who were considered for USAT between November 2016 and May 2019. Of those, 51 patients received USAT therapy while 51 received AC alone. Baseline characteristics are shown in Table 1. Patients in the intervention group were younger  (mean age, 57.3 years [SE, 1.93] in the USAT group vs 64.5 years [SE, 2.03] in the AC group; P=.01), and more likely to be male (51.0% males in USAT group vs 47.1% in the AC group; P=.84). Comorbidities were largely similar between the 2 groups (Table 1); however, malignancy was more common in patients treated with AC alone (9.8% in the USAT group vs 33.3% in the AC group; P<.01). Mean PESI scores were similar (109.4 [SE 5.1] for the USAT group vs 117.1 [SE, 4.6] for the AC group; P=.26). A backward stepwise logistic regression of demographic and risk factors found that patients were less likely to be treated with USAT if they had a prior malignancy (odds ratio [OR], 0.25; 95% CI, 0.08-0.79; P=.01) and were age 80 or older (OR, 0.10; 95% CI, 0.012-0.91; P=.04), while patients with a period of prolonged immobility preceding presentation were more likely to be treated with USAT (OR, 2.24; 95% CI, 1.27-12.31; P=.08).

Clinical and physiologic outcomes. There was no difference in hospital length of stay (5.9 days [SE, 1.0 days] in the USAT group vs 5.7 days [SE, 0.8 days] in the AC group; P=.45). The length of ICU stay was longer for USAT patients (2.9 days in the USAT group [SE, 0.4 days] vs 1.8 days in the AC group [SE, 0.4 days]; P=.02), as ICU admission is part of the USAT protocol at our institution. All but 1 of the USAT patients spent time in an ICU while 20 patients in the AC category did not require ICU level of care. There was a higher rate of patients being discharged on supplemental oxygen in the USAT group (31.4% in the USAT group vs 14.3% in the AC group; P=.04); however, patients receiving USAT were more like to be on supplemental oxygen on admission (84.3% in the USAT group vs 60.8% in the AC group; P=.01). Two patients in the AC group died during admission, compared with none in the USAT group (P=.25). There was 1 BARC 2-5 event in the USAT group vs 3 BARC 2-5 bleeding events in the AC group (P=.31). The single major bleeding event post USAT involved a patient who developed heparin-induced thrombocytopenia (HIT), and subsequently, multiple abdominal hematomas. In the USAT cohort, there were 6 BARC 1 bleeding events (non-actionable events defined as bleeding events not requiring further studies or treatment) (Table 2).¹⁹  These 6 non-actionable bleeds included 4 episodes of self-limited hematomas at the site of catheter insertion and 2 episodes of self-limited epistaxis.

Among patients receiving USAT, RV dilation and RV dysfunction on echocardiography significantly improved immediately after receiving therapy, with the percentage of patients with moderate or severe RV dilation falling from 68.4% to 15.8% (P<.01) and moderate or severe RV dysfunction also falling from 68.4% to 15.8 (P<.01) (Table 3). In patients undergoing USAT, mean PA pressure decreased by an average of 6.5 mm Hg (34.4 mm Hg to 27.9 mm Hg; 95% CI change, 4.14-8.77) (Table 4). Systolic PA pressure fell by an average of 16.5 mm Hg (58.4 mm Hg to 41.2 mm Hg; 95% CI change, 12.7-20.4) (Table 4). Because follow-up echocardiograms and right heart catheterization were not generally performed on patients in the AC group, no physiologic outcomes were reported in this group. 

Fibrinogen levels measured during and after the infusion of tPA via USAT had a mean nadir of 369.7 ± 127.1 mg/dL (Figure 1). There were no postprocedure fibrinogen levels collected on 2 patients, neither of whom had adverse outcomes, including bleeding events. The lowest recorded fibrinogen level was 187 mg/dL. Studying patients who had a fibrinogen drawn at 6, 12, and 18 hours post procedure (n = 24), the lowest mean fibrinogen levels occurred at 18 hours at 416.5 ± 157.4 mg/dL vs a mean of 465.3 ± 178.6 mg/dL at 6 hours and 419.1 ± 166.1 mg/dL at 12 hours (Figure 2). In the patients who had bleeding complications, the mean trough fibrinogen (lowest at hour 6, 12, or 18) was 306.6 mg/dL (SE, 23.9 mg/dL) as compared with 380.3 mg/dL (SE, 20.4 mg/dL) in those without bleeding complications (P=.02) (Table 5). Mean fibrinogen levels (mean of 6-, 12-, and 18-hour readings for each patient, with the mean then taken for the entire group) were 343.6  mg/dL (SE, 25.4 mg/dL) in the group with bleeding complications vs 412.2 mg/dL (SE, 22.0 mg/dL) in the group without bleeding complications (P=.03).

The results of our study provide further support for the relative safety of USAT utilizing a standard treatment algorithm, with only 1 major bleed and no mortality events for patients in the intervention arm. Our study of real-world practice shows that USAT, when used in appropriate patients with a standardized treatment algorithm, was associated with hemodynamic improvements (improved RV function and decreased pulmonary artery pressure) similar to those demonstrated in earlier studies. Most significantly, patients treated with USAT did not experience clinically significant hypofibrinogenemia. This stands in contrast to an earlier, more limited study (n = 30) that did demonstrate fibrinogen levels as low as 37 mg/dL in patients undergoing USAT for PE.⁸ This study, however, included alteplase infusion rates as high as 2.0 mg/hr, which is a higher rate than utilized in our protocol. The low rate of major bleeding events and absence of hypofibrinogenemia in the USAT group provides support for the safety of this therapy in appropriately selected patients. Our reporting of fibrinogen levels in USAT provides new insight into the safety of catheter-directed therapies for submassive PE.  

Older patients and those with recent cancer were less likely to undergo USAT. Other studies, such as ULTIMA, also included cancer patients.¹ SEATTLE II included cancer patients but did not have a control.² Some studies have excluded patients with malignancy.²⁰ Similar cohort studies have included cancer patients, achieving similar numbers in control and intervention groups.¹⁸ Other single-arm studies included patients with malignancy in similar proportion to our study.^21,22 Patients in the AC group did not undergo USAT most commonly because of clinical improvement or stability in the setting of mild symptoms (19 patients; 37%), followed by a history of bleeding or perceived increased bleeding risk secondary to comorbidities, eg, brain metastases (13 patients; 25%), patient or family preference (12 patients; 24%), or vascular anatomy and a PE location that were not amenable to USAT (2 patients; 4%). For 5 patients (10%), the reason for not undergoing USAT was not specified in the chart.  Larger, randomized trials are needed to understand the relative benefit of USAT in younger and healthier patients vs older patients with multiple comorbidities. It is possible, for instance, that older patients with conditions such as cancer and higher clot burden could achieve greater symptomatic and physiologic benefit compared with younger patients, given the more immediate physiologic improvement seen with USAT compared with AC alone. 

Our study demonstrated that the hemodynamic improvement noted in clinical trials can be replicated in real-world circumstances with good safety outcomes. Physiologic outcomes in our study were comparable to previous research.^1,20,21,23 It remains plausible that the acoustic streaming provided by USAT, combined with a significantly lower dose of thrombolytic agent, could lead to improved clot dissolution and early benefit for the pulmonary microvasculature, while minimizing potential harm of bleeding through lower thrombolytic dose. Additional clinical trial data  are needed to determine whether long-term outcomes for USAT are similar or superior to those for systemic thrombolysis. 

There was only 1 major bleeding event (BARC 2-5) in the USAT group, likely the result of concomitant HIT, rather than thrombolysis-related coagulopathy. Notably, this patient had a fibrinogen level of 374 mg/dL, above the mean of 369.7 mg/dL. There were 3 major bleeds reported in the control group, each of which occurred in a patient under consideration for USAT, with the bleeding event occurring before, and therefore precluding the patient from receiving interventional therapy. No fibrinogen levels fell below 150 mg/dL in the USAT group; this level is noted in previous studies as a threshold for increased bleeding risk.^9,11,14 Because of the catheter-directed delivery and slower infusion rates in our institutional protocol, total tPA exposure may remain below the level required to induce hypofibrinogenemia. Without hypofibrinogenemia, other studies have shown bleeding risk to be minimal.^3,19,23 These findings, especially regarding major bleeding risk and fibrinogen, further validate the safety of USAT under a standard protocol and provide some insight into the etiology of the intervention’s favorable safety profile. 

Study limitations. Our study should be considered in the context of its limitations. While we utilized a standardized treatment algorithm, fibrinogen levels were not collected on all USAT cohort patients at all protocol time intervals. Despite this, we had at least 1 fibrinogen measure on 49 patients in the 18 hours after placement of the USAT catheter, with completely missing fibrinogen data on only 2 patients in the study period. All patients with a bleeding complication had at least 2 fibrinogen measures. Finally, our data reflect the experience of a single center utilizing a single treatment algorithm. Nonetheless, we believe our data offer further insight into the emerging body of evidence for the safe implementation of USAT protocols and procedures.

Patients treated with USAT at a single academic center had lower pulmonary pressures and improved cardiac function following USAT. There was only 1 major bleeding event and there were no episodes of intracranial hemorrhage or hypofibrinogenemia in the USAT cohort. These data support the general efficacy and safety of USAT when using a standardized treatment algorithm, as measured through improved physiologic outcomes and low adverse event rates. Although no patients reached the critical threshold of hypofibrinogenemia, those with any bleeding complication had generally lower fibrinogen levels, suggesting a link between fibrinogen levels and bleeding risk in USAT patients. Further study is needed to understand factors associated with bleeding and the development of hypofibrinogenemia and to better define clinically useful thresholds of systemic fibrinogen level at which bleeding risk may become markedly increased. Nonetheless, we believe that our study provides further evidence that USAT is generally safe with regard to fibrinogen as a proxy for bleeding risk.