**9. Special population**

### **9.1 Obesity**

Anticoagulant dosing of obese patients with PE remains to be an area that has not been well studied [71]. The bulk of the data comes from pharmacokinetic/ pharmacodynamic (PK/PD) studies and subgroup analyses of premarketing trials comparing obese patients to patients with normal body weight [29, 32, 71, 72]. Furthermore, warfarin appears to have the most robust data available in this population. Warfarin pharmacokinetics have been compared in various studies of obese patients with a BMI > 30 and 40 kg/m<sup>2</sup> to normal body weight subjects. These studies found that obesity was associated with a greater delay and dosage to achieve a therapeutic INR. Data associated with the risk of bleeding, however, have been conflicting, indicating that obesity may or may not increase the risk [71]. While warfarin has the most robust data in this population, the usage of warfarin appears to be on the decline [71].

Dabigatran is a direct thrombin inhibitor that sets it apart from other DOACs. Subgroup analyses of Phase 3 trials for VTE (RE-LY) and atrial fibrillation (RECOVER) suggested no significant differences in the efficacy and safety outcomes of obese patients in comparison with those with lower body weights [29, 32]. In contrast, treatment failures have been documented in morbidly obese patients using standard doses. The authors reported that standard doses failed to achieve "therapeutic levels," suggesting a higher volume of distribution (Vd) and a higher clearance [27, 28].

The factor X inhibitors, rivaroxaban, apixaban, and edoxaban also have some data in obese populations. Rivaroxaban was evaluated in a small PK/PD study involving a heterogeneous group of subjects including those in the obese range (>120 kg). The half-life, Vd, and clearance declined slightly with increasing body weight. The authors felt the declines were not clinically relevant [30]. A separate PK/PD analysis using pooled data from the phase ll EINSTEIN DVT and ODIXa-DVT trials reported similar findings. These studies enrolled some patients with BMIs >35 kg/m2 and found clinical benefits comparable across all weight groups [36]. The authors of this analysis also concluded that standard doses should be sufficient to treat obese patients; however, a review questioned whether the data were robust enough to draw that conclusion [36, 72].

In the apixaban's phase 1 PK/PD study, the pharmacokinetic data were somewhat different compared to those for rivaroxaban. The overall Vd was higher in subjects >120 kg or BMI > 30 kg/m2 than normal body weight subjects. The half-life declined almost proportionately to the increase in Vd (27% ↓ vs. 24% ↑ for half-life and Vd, respectively). Peak level and area under the plasma drug concentration-time curve (AUC) were also reduced in the higher body weight subjects. The authors concluded the differences were unlikely to be of clinical significance [47]. The phase 3 trials, ARISTOTLE and AMPLIFY, contained a significant proportion of patients with body weights >100 kg and BMIs above 30 kg/m2 . Subgroup analyses found no differences in efficacy; however, more bleeding episodes were reported in the ARISTOTLE trial. Whether the increase in hemorrhagic complications between these trials was due to age, renal clearance, or other patient-specific factors is unclear [48–50].

Edoxaban has less data than other DOACs. Phase 3 Hokusai-VTE enrolled a large group of patients >100 kg. There was no difference in efficacy and safety compared to groups with other weights [51].

A recent international retrospective study of LMWHs examined dosing in obesity regarding capped (<18,000 IU/d) and uncapped dosages (>18,000 IU/d). The data were obtained from the RIETE Registry, a large prospective case series of patients with VTE. LMWHs included enoxaparin, dalteparin, and tinzaparin. The authors reported that the results may be subject to selection bias despite attempts to control for potential confounders in multivariable analysis. Nevertheless, they found that after adjustment for multiple potential confounders, patients with obesity (>100 kg or BMI >30 kg/m2 ) who received capped doses were at a lower risk of having the composite outcome of VTE recurrences, major bleeding, or all-cause death at 15 and 90 days. Bleeding was also reduced with capped dosages [52].

A recent retrospective study of UFH for acute venous thromboembolism (VTE) compared three body mass index (BMI) cohorts: (i) non-obese (less than 30 kg/m2 ), (ii) obese (30 to 39.9 kg/m2 ), and (iii) morbid obesity (⩾40 kg/m2 ). The dosing employed was based on actual body weight. The median times to therapeutic aPTT were reported as 16.4, 16.6, and 17.1 h in each of the three cohorts [53].

Obese patients on warfarin may require a higher dose or more time to achieve a therapeutic INR. Bleeding risk may or may not be greater. Available data for DOACs other than dabigatran suggested usual doses may not negatively impact efficacy in obese patients. Apixaban studies reported conflicting results on bleeding risk. However, there are probably insufficient data for DOACs to suggest that usual doses would be adequate in the subgroup of morbidly obese patients. Obese patients with capped doses of LMWHs may have better efficacy and safety outcomes. Capped dosages are conditionally recommended in the 2018 American Society of Hematology guidelines [54].

### **9.2 Renal dysfunction**

Anticoagulants have been evaluated in CKD and ESKD patients in various PK/PD and clinical efficacy trials [55]. The PK/PD of warfarin in CKD and ESKD is not completely understood [55, 56]. Official dosing guidelines do not recommend an alteration of dose [55]. Warfarin is extensively metabolized by the cytochrome P450 type 2C9 (CYP2C9) enzyme [73]. Although not removed by dialysis, there are no data evaluating whether this procedure alters its pharmacokinetics and pharmacodynamics [55]. One study comparing individuals with a GFR of 30–59 mL/min and healthy controls reported a shorter half-life and increased clearance. Other data showed that CKD, especially GFRs <30 mL/min per 1.73 m<sup>2</sup> or ESKD, complicates warfarin therapy [58, 59]. These data reported that lower doses were required to maintain therapeutic INR with greater fluctuations in INR values and higher risks of major bleeding events for any given INR value. A recent meta-analysis of ESKD patients with atrial fibrillation found that warfarin had no benefit in reducing ischemic stroke incidence. The authors concluded that the drug appeared to be associated with a significantly higher risk of hemorrhagic stroke but no increased risk of other types of major bleeding. They also found no change in mortality [71, 74]. How all these data apply to other indications such as PE is unknown.

The various DOACS have also been evaluated in renal disease. A small study of dabigatran using reduced doses (150 mg daily for CKD and 50 mg daily for ESKD) examined the effect of CKD and ESKD on pharmacokinetic parameters. Subjects with a creatinine clearance <30 mL/min demonstrated a 6.5-time increase in the AUC with a doubling of the half-life compared to normal controls [75].

Rivaroxaban has somewhat conflicting data in CKD and ESKD. In a phase 3 trial subgroup analysis of the Rivaroxaban Once Daily Oral Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF), the authors reported that reduced renal function (creatinine clearances <80 mL/min) had no impact on rivaroxaban's effectiveness and safety [76]. In contrast, a PK/PD study comparing dialysis patients to normal controls reported a 56% increase in AUC when 15 mg doses

were administered after dialysis [77]. The AUC was decreased by only 5% when administered pre-dialysis suggesting dialysis is ineffective in clearing the drug [77]. Another PK/PD study comparing three ranges of renal insufficiency to normal controls reported AUCs of 1.4-, 1.5-, and 1.6-fold higher in cases of creatinine clearance concentrations of 50–80, 30–50, and < 30 mL/min, respectively [78].

Pharmacokinetic data for apixaban appear similar to those of rivaroxaban, except that the incremental upsurges in AUC are somewhat lower in magnitude. A small PK/PD study using a 5 mg dose in ESKD patients and normal controls found a 36% increase in AUC in the ESKD group compared to controls. The apixaban dose was administered pre-dialysis [79]. Another small PK/PD study used a single 10 mg dose and compared subjects with varying degrees of CKD to normal controls [8]. Compare to the controls, the AUCs increased by 16%, 29%, and 38% in the CKD cohorts with creatinine clearances of 50–80, 30–50, and < 30 mL/min, respectively. The mean half-life was only slightly increased in the total CKD population (17 h) compared to the controls (15 h) [79]. The overall results for both studies were not unexpected, as apixaban is metabolized more extensively and demonstrates less unchanged renal elimination than other DOACs [1]. Again, the drug appears to be poorly dialyzable [55].

Edoxaban was evaluated in a small PK/PD study comparing CKD subjects to normal controls. The data, available as an abstract only, showed that the mean AUCs increased by 32%, 74%, and 72% with creatinine clearances of 50–80, 30–50, and < 30 mL/min (not on dialysis), respectively [55]. The mean AUC for subjects on peritoneal dialysis was 93% [55]. Similar to other Factor Xa inhibitors, edoxaban is poorly cleared by hemodialysis [80].

Probably, the best data for LMWH come from a subgroup analysis of the ExTRACT TIMI 25 trial. This study compared a 1 mg/kg per day dose of enoxaparin to UFH in ST-elevation myocardial infarction (STEMI) patients with CrCLs less than 30 mL. There was no significant difference in mortality or recurrent MI (33 vs. 37.7%) and in bleeding risk (5.7 vs. 2.8%) for enoxaparin and UFH, respectively. Mortality increased for both drugs as renal function declined [81].

UFH is rapidly eliminated by the reticular endothelial system and, to a lesser extent, the kidney. However, the degree to which the RES functions in this role varies from person to person. Nevertheless, lower doses in VTE treatment are recommended to reduce bleeding risk. An advantage for UFH over enoxaparin is the ability to reverse the former easier with protamine [11, 82].

All DOACs increase AUC with the magnitude inversely proportional to renal function. These drugs may be the anticoagulants of choice for stages 1, 2, and 3 CKD [83]. Some sources recommend apixaban as an alternative to warfarin in selected patients with ESKD [55]. Warfarin, although predominately metabolized in the liver, may require a lower dose to reach a therapeutic INR. Historically, it is the oral anticoagulant of choice in ESKD [11, 82]. However, how does warfarin's potential lack of benefit in ESKD patients with AF apply to PE? Are LMWHs better than UFH? Both were associated with similar mortality increases in STEMI patients with declining kidney function. Should these data also apply to renal failure patients and PE?

### **9.3 Liver dysfunction**

Patients with cirrhosis typically have elevated INR and aPTT. As a result, there have been misconceptions that these patients are "autoanticoagulated." However, patients with liver disease may have a substantially increased risk of VTE (RR 1.74 (95% CI, 1.54–1.95) for liver cirrhosis and 1.87 (95% CI, 1.73–2.03) for noncirrhotic liver disease) [83]. A meta-analysis in patients with the liver disease found


**Table 3.**

*FDA recommendations of DOACs in liver dysfunction.*

the incidence of PE from nine studies to be 0.28% (95% CI 0.13–0.49%) and the prevalence of PE from two studies to be 0.36% (95% CI 0.13–0.7%) [84].

Determining the ideal anticoagulant to use after a PE in a patient with liver cirrhosis may be challenging. Warfarin can be safely used in patients with liver dysfunction. Elevations in baseline INR due to liver disease, however, may lead to unclear INR targets during warfarin therapy [85, 86]. LMWHs have a good safety profile with liver dysfunction; however, subcutaneous administration may limit compliance, and lower anti-Xa levels in liver dysfunction limit efficacy monitoring [87, 88]. Finally, there are a lack of data on using DOACs in patients with liver disease. Clinical trials with DOACs excluded patients with liver disease as most DOACs are predominantly cleared hepatically (apixaban 75%, rivaroxaban 65%, edoxaban 50%, and dabigatran 20%) [73, 89]. Therefore, dosing recommendations are derived from pharmacokinetic studies (**Table 3**).

Regardless of which anticoagulant is chosen, the risk of bleeding should be thoroughly evaluated. In patients with liver disease, bleeding can be due to varices, portal hypertensive gastropathy, peptic ulcer disease, and arteriovenous malformations. A small, 45-patient cohort study, comparing DOACs to warfarin/LMWH found no difference in thrombotic events. Still, there were significantly fewer major bleeding episodes in the DOAC group (1 patient [4%] vs. 5 patients [28%], *p* = 0.03) [90].

### **9.4 Cancer patients**

Cancer patients have a five- to sevenfold increased risk for VTE within the first year of diagnosis [91]. Additionally, VTE is considered an independent predictor of mortality in these patients [92, 93]. The pathophysiological and epidemiological association between PE and cancer is well established. For decades, LMWHs were considered a first-line therapy for cancer-associated PEs, as knowledge of the efficacy and safety of DOACs in cancer patients was lacking [77, 94, 95]. Since then, four large randomized control trials have been published comparing DOACs with LMWH that have highlighted the utilization of most DOACs for the treatment of PEs [77, 81, 96, 97].

The Hokusai VTE Cancer trial randomized 1050 cancer patients with acute VTE to either edoxaban, an oral direct factor Xa inhibitor, or dalteparin, an LMWH. The trial found that edoxaban was non-inferior to dalteparin for the primary outcome of recurrent VTE or major bleeding during a follow-up period of 12 months (95% confidence interval 0.70 to 1.36; *p* = 0.006) [98]. There was a lower rate of VTE, however, nonsignificant risk difference of −3.4 (−7.0 to 0.2), but the major bleeding rate was significantly higher (risk difference of 2.9 (0.1 to 5.6)) [98]. A nonsignificant lower VTE rate was seen, but the major bleeding rate was significantly higher in the edoxaban group. Major bleeding events were frequently observed in the subgroup with upper gastrointestinal tract neoplasms [98].

SELECT-D randomized 406 patients with cancer and acute VTE to oral rivaroxaban, a factor, or dalteparin for a treatment duration of 6 months [99]. SELECT-D was a 6-month open-label, pilot, randomized control study that compared rivaroxaban 15 mg BID for 3 weeks then 20 mg once daily to dalteparin (200 IU/kg once daily for 1 month, then 150 IU/kg once daily) in patients with VTE and solid or hematologic cancers [99]. The trial found that the VTE recurrence rate was 4% with rivaroxaban and 11% with dalteparin (HR 0.43, 95% CI 0.19 to 0.99) [99]. Major bleeding occurred in 4% with dalteparin and 6% with rivaroxaban (HR 1.83, 95% CI 0.68 to 4.96) and clinically relevant, nonmajor bleeding occurred in 4% with dalteparin and 13% with rivaroxaban (HR 3.76, 95% 1.63 to 8.69) [99].

CARAVAGGIO was an open-label, non-inferior study that randomized 1170 patients to apixaban (10 mg twice daily for 7 days, then 5 mg twice daily) or dalteparin (200 IU/kg once daily for 1 month, then 150 IU/kg once daily) for 6 months of treatment [100]. The trial resulted in a VTE recurrence rate of 5.6% with apixaban and 7.9% with dalteparin (risk difference − 2.3%; HR 0.63, 95% CI 0.37 to 1.07, *p* < 0.001) [100]. Additionally, major bleeding occurred in 3.8% in the apixaban group versus 4% in the dalteparin group (risk difference − 0.2%; HR 0.82, 95% CI 0.4 to 1.69, *p* = 0.60) [100].

Given that these oral agents have shown efficacy and provide a more convenient medication dosage form, we have already begun to see a shift in the way PEs are treated in cancer patients [77, 81, 96, 97]. However, since many of these studies have excluded patients with GI tumors or a history of GI bleeding, it is still recommended to continue using LMWH in these patients due to the higher GI bleeding risk seen with DOACs [77, 81, 96, 97]. If a DOAC is to be utilized, the choice of which oral anticoagulant to be used in a cancer patient should be made on an individual basis with considerations of drug interactions with chemotherapy agents and the type of cancer.

### **9.5 Pregnancy**

UFH and LMWH remain the only currently available choices for the safe management of VTE including PE during pregnancy and puerperium. UFH provides the least exposure to the fetus and risk to the mother, since it does not cross the placenta, is not distributed into breast milk, and the dose is easily titrated [101]. LMWHs have a similar safety profile as UFH for the fetus since they also do not cross the placenta. Concerns about distribution into breast milk have been raised [101, 102]. Most studies have not demonstrated that the LMWH levels in breast milk are high enough to cause coagulopathy when using standard prophylaxis doses. Further studies are needed for a complete safety profile.

The ability to easily monitor aPPT levels for UFH and anti-Xa levels for LMWHs provides clinicians with the ability to maintain therapeutic concentrations throughout pregnancy. Therapeutic-level monitoring provides safety and efficacy for both the fetus and the mother, since changes in the drug volume of distribution and clearance progressively increase throughout pregnancy [23, 101, 103–105]. Additionally, UFH and LMWH have short half-lives, allowing for anticoagulation to be maintained within 24 h of delivery and permitting the safe use of neuroaxial anesthesia if needed [102, 106]. Restarting LMWH within 6 h postdelivery in patients without bleeding concerns has been shown to be safe. Anticoagulation should be maintained during puerperium [101]. Thromboprophylaxis for future pregnancies with UFH or LMWH is recommended for women at risk [107, 108].

The safety profile and pharmacokinetics of DOACs currently do not favor their use during pregnancy or puerperium, since they cross the placenta and distribute into breast milk [101, 109]. The use of DOACs may be considered in women who are not or are no longer breastfeeding. The inability to readily monitor DOAC levels limits the clinician's ability to maintain therapeutic levels throughout pregnancy and puerperium due to the pharmacokinetic changes listed above. Maintaining a safe and therapeutic efficacious anticoagulation regimen up to and through the time of delivery requires a multidisciplinary effort and needs to be closely coordinated [101, 102, 106].

### **9.6 Elderly**

The risk of thromboembolic disease, including PE, increases with age [110–112]. The trauma to the vascular endothelium caused by a thromboembolic event places the patient at a higher risk of recurrence, requiring prolonged anticoagulation [110, 113]. Comorbidities accompanying aging, especially cardiovascular and pulmonary disease, add to the complexity of treatment of thromboembolic disease [114]. The increased risk for bleeding further complicates the safe management of an acute venous thromboembolic event, and prolonged anticoagulation is required to prevent future events [4].

PE in elderly patients may not present with the typical symptoms seen in younger patients, making early diagnosis more challenging in this population. Several studies have found syncope to be the most frequent presenting symptom in elderly patients, versus pleuritic chest pain in a younger population [115–117]. Early pharmacotherapy intervention is critical to prevent further thrombus formation, regardless of the patient's age [110].

Recommendations for initial treatment for PE in a stable patient regardless of the prognosis include DOACs (apixaban or rivaroxaban) or initial LMWH, followed with dabigatran or edoxaban. LMWH or UFH and warfarin may be preferred in patients with reduced renal function (CrCL <30 mL/min) [118].

Stable patients with a good prognosis may be managed in an outpatient setting. The current recommendations for stable patients with a poor prognosis require hospitalization [118]. A thorough medication history is important prior to starting a DOAC due to the potential drug–drug interactions seen with this class of anticoagulants. P-glycoprotein inhibitors or CYP3A4 inhibitors or inducers should be avoided, since they can alter the plasma of DOACs [118]. The unstable patient requires hospitalization and may be a candidate for thrombolytic therapy, preferably by catheter-directed thrombolysis [118].

### **9.7 COVID-19 patients**

The emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from Wuhan, Hubei province, China, in December 2019 has adversely impacted the world. Among the several complications related to SARS-CoV-2 is VTE [119], a systematic review of 3487 COVID-19 patients from 30 studies demonstrated VTE incidence to be 26%, 12% for PE with or without DVT, and 14% for DVT alone.

The etiology of PE in COVID-19 is secondary to the well-known VTE risk factors, with indirect aspects of the severity of illness and direct effects of SARS-CoV2 viral infection. These patients are at VTE risk due to the recurrent use of intravascular access devices, sedation, and vasopressors, within the intensive care unit (ICU) setting. These factors promote stasis. Respiratory failure, hypoxia, comorbidities, multi-organ failure, obesity, and prolonged immobility are other elements. Additional confounding factors include a history of VTE and or neoplasm, sepsis, surgery, trauma, or stroke [120, 121].

COVID-19 is associated with a profound early response of proinflammatory cytokines. This may result in a cytokine storm, increased risk of vascular hyperpermeability, multi-organ failure, and death [122, 123]. While thrombin's primary role is to accelerate clot formation *via* platelet activation and the conversion of fibrinogen to fibrin, thrombin equally exerts multiple cellular effects. It could further enhance inflammation through proteinase-activated receptors (PARs). Thrombin production is highly regulated *via* negative feedback mechanisms and physiological anticoagulants, such as the protein C system, antithrombin III, and tissue factor pathway inhibitor [124]. These three control mechanisms may become impaired throughout the inflammatory process, resulting in decreased anticoagulant concentrations secondary to reduced production and increased consumption. This impaired procoagulant– anticoagulant equilibrium results in disseminated intravascular coagulation (DIC), microthrombosis, multi-organ failure, and elevated d-dimer levels [91, 125, 126]. Hypercoagulable state, endothelial dysfunction, injury, and viral-induced procoagulant effect all also play an immense role in COVID-19-associated PE [127–129].

Both LMWH and UFH are recommended for VTE prophylaxis and PE treatment in COVID-19 patients. The recommended dose of the LMWH enoxaparin for VTE prophylaxis is 40 mg (4000 IU) daily. An intermediate dose of LMWH enoxaparin 40 mg subcutaneously every 12 h is recommended in the critically ill, the obese, and those patients with multiple VTE risk factors. A UFH dose of 5000 IU Q8H is recommended for VTE prophylaxis, and doses of up to 7500 IU have been employed in critically ill patients [98–100, 130, 131]. DOACs are recommended in stable patients and in the post-acute phase of PE and outpatient settings when the benefits from their use outweigh the risks. Renal function monitoring with dosage modifications is recommended when LMWH, UFH, and DOACs are employed. Additionally, Anti-Xa monitoring is recommended in patients requiring the therapeutic anticoagulation with LMWH, UFH, and DOAC therapy. DOAC plasma level monitoring is also recommended [101, 131]. Therapeutic anticoagulation should always be considered first; thrombolytic therapy is recommended in patients who go to develop sub-massive or massive PE. An inferior vena cava filter may be employed in at-risk patients; extracorporeal membrane oxygenator (ECMO) is an option, in conjunction with surgical embolectomy or catheter-directed management [1, 102, 103]. Several weeks of therapy is recommended with LMWH or DOACs, post-hospitalization for COVID-19 patients [132].

### **10. Conclusions**

PE is a medical emergency that affects thousands of Americans each year. Thousands of Americans die from this condition annually. The therapy for PE has evolved over the years. Traditional therapies such as UFH, VKA, and warfarin are being abandoned by clinicians in favor of LMWH and DOACs. Reversal agents such as 4-factor prothrombin complex concentrate, andaxanet alfa, and the monoclonal antibody idarucizumab have allowed clinicians to push the boundaries of PE management with confidence, particularly in the outpatient setting. Special populations, such as obese, renal dysfunction, liver impairment, cancer, pregnancy, and COVID-19 patients with PE, pose a tremendous therapeutic burden and challenge to clinicians. Despite these challenges, tremendous progress has been made, with demonstrated improved patient outcomes in PE treatment over the last three decades.

### **Conflict of interest**

The authors declare no conflict of interest.

*Anticoagulants in the Management of Pulmonary Embolism DOI: http://dx.doi.org/10.5772/intechopen.100471*
