An update on prevention of venous thromboembolism in hospitalized acutely ill medical patients

Both the recently updated consensus guidelines published by the American College of Chest Physicians, and the International Union of Angiology recommend thromboprophylaxis with either low-molecular-weight heparin (LMWH) or unfractionated heparin (UFH) in medical patients at risk of VTE. However, no guidance is given regarding the appropriate dosing regimens that should be used for thromboprophylaxis in this patient group. LMWH (enoxaparin and dalteparin) and UFH have been shown to be effective for thromboprophylaxis in at-risk hospitalized medical patients. Although LMWH once daily (o.d.) has been shown to be as effective as UFH three times daily (t.i.d.) for thromboprophylaxis in at-risk medical patients, there are no data to show that UFH twice daily (b.i.d) is as effective as either LMWH o.d. or UFH t.i.d. On the basis of currently available evidence, the LMWHs enoxaparin and dalteparin are more attractive alternatives to UFH for the prevention of VTE in hospitalized medical patients because of their convenient once-daily administration and better safety profile, demonstrated in terms of reduced bleeding, HIT, and other adverse events.


Introduction
In the absence of thromboprophylaxis, the incidence of venous thromboembolism (VTE) ranges from 10-20% in general medical patients to 80% in trauma patients, spinal cord injury patients, and patients in the critical care unit [1,2]. Despite evidence from large, randomized clinical studies demonstrating the benefits of providing thromboprophylaxis for hospitalized medical patients at risk of VTE [3][4][5], thromboprophylaxis is not currently prescribed to the extent that might be expected in this patient population, leaving many patients exposed to significant risk of acute thrombotic complications and their long-term consequences [6-9].
Consensus guidelines published by the American College of Chest Physicians (ACCP) and the International Union of Angiology (IUA) recommend assessment of all hospitalized medical patients for the risk of VTE and the provision of appropriate thromboprophylaxis [1,2]. Furthermore, simple and clinically-relevant risk assessment models (RAMs) are available to facilitate VTE risk assessment [10,11]. A recently published evidence-based RAM, developed specifically for hospitalized medical patients, should provide additional guidance to physicians in this patients group [12]. This RAM integrates patient VTE risk level with appropriate thromboprophylactic strategies in the form of a management algorithm. Computerised reminders have also been shown to be valuable for improving prophylaxis prescribing rates [13], and an electronic risk assessment tool has recently been developed for use in medical patients as well as surgical patients [14].
The recently updated ACCP consensus guidelines give a grade 1A recommendation for thromboprophylaxis using either low-molecular-weight heparin (LMWH) or lowdose unfractionated heparin (UFH) in medical patients with congestive heart failure (CHF) or severe respiratory disease, or in medical patients who are confined to bed and have one or more risk factors for VTE, such as active malignancy, acute neurological disease, inflammatory bowel disease, previous VTE, or sepsis [1]. LMWHs, unlike UFH, have greater bioavailability and consistent anticoagulation effects, and the benefit of once-daily dosing [15,16].
The aim of this manuscript is to review recent advances in thromboprophylaxis in hospitalized medical patients, and discuss them in light of the recently updated ACCP consensus guidelines.
The efficacy of LMWH thromboprophylaxis in hospitalized medical patients has been investigated in several large, randomized, placebo-controlled clinical studies (Table 1). Compared with placebo, thromboprophylaxis with LMWH reduces the risk of VTE by 45-66% [3,4,[24][25][26]. In the international Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) study, two doses of enoxa- parin (20 mg and 40 mg s.c. o.d.) were compared with placebo in acutely ill medical patients [3]. The MEDENOX study, in which deep-vein thromboses (DVT) were confirmed using venography, showed a significant reduction in the incidence of VTE when 40 mg enoxaparin was used compared with placebo for 6-14 days (relative risk, 0.37; 97.6% confidence interval (CI), 0.22-0.63; p < 0.001), but not when 20 mg enoxaparin was used ( Table 1). The benefit with 40 mg enoxaparin was maintained at 3month follow-up (relative risk, 0.41; 95% CI, 0.25-0.68; p < 0.001). Moreover, Kaplan-Meier plots of the probability of 3-month survival rates suggest that the overall mortality was lower in patients receiving 40 mg enoxaparin than those receiving placebo, although differences in mortality rates between the three groups did not reach statistical significance. A retrospective multihospital analysis of data from hospitalized acutely ill medical patients in the USA also showed that thromboprophylaxis with enoxaparin resulted in a significant reduction in the incidence of VTE compared with no thromboprophylaxis (1.9% vs 6.2%, p = 0.023) [28].
The Prospective Evaluation of Dalteparin Efficacy for Prevention of VTE in Immobilized Patients Trial (PREVENT) later confirmed the benefits of LMWH prophylaxis in hospitalized medical patients. This study, which enrolled a patient population with a slightly lower thrombotic risk than those in MEDENOX and used compression ultrasound to confirm DVT, showed that a high thromboprophylactic dose of 5,000 IU dalteparin significantly reduced the incidence of VTE in acutely ill medical patients compared with placebo (relative risk, 0.55; 95% CI, 0.38-0.80; p = 0.0015) [4]. However, in contrast to the findings of the MEDENOX study [3], the reduced incidence of VTE was not statistically significant at the 3month follow-up (relative risk, 0.70; 95% CI 0.36-1.35) and a trend towards reduced mortality was not observed in the dalteparin group compared with the placebo group [4]. It is difficult to conclude whether these differences between the findings of the MEDENOX and PREVENT studies were a result of different pharmacological properties of the two LMWHs [29] or differences in the methodology and patient characteristics of the two studies.
Several randomized studies have compared the efficacy of LMWH and UFH as thromboprophylaxis in hospitalized medical patients at risk of VTE ( It should be noted that this meta-analysis did not include data from the MEDENOX and PREVENT studies, and it also has some limitations in terms of differences in study designs and the small size of some of the studies included. Nevertheless, it does provide an important insight into the impact of heparin therapy on VTE risk in medical patients. Notably, a recent post-hoc analysis demonstrated that asymptomatic proximal DVT, which was more commonly diagnosed than symptomatic DVT in acutely ill medical patients in the major clinical studies mentioned above [3,4], is associated with increased mortality [36]. This adds weight to the clinical opinion that asymptomatic proximal DVT is a relevant surrogate endpoint for symptomatic thromboembolic disease.  (Figure 1). Venous thromboembolic events occurred in 9.7% of the patients with CHF receiving enoxaparin 40 mg compared with 16.1% of those receiving UFH, although the patient numbers were too small to show a statistically significant difference between groups [34]. In patients with CHF in the MEDENOX study [3], which did not have a UFH arm, 40 mg enoxaparin resulted in a significant reduction in the incidence of VTE compared with placebo (relative risk, 0.29; 95% CI, 0.10-0.84; Figure 1) [38]. The lower incidence of VTE in CHF patients receiving enoxaparin 40 mg compared with either placebo or UFH warrants further investigation in a randomized clinical trial with sufficient power to show superior efficacy of enoxaparin, 40

Thromboprophylaxis with other anticoagulants
Although most studies on thromboprophylaxis in acutely ill medical patients have investigated the efficacy of LMWHs, a placebo-controlled trial of the synthetic factor Xa inhibitor fondaparinux (the Arixtra for Thromboembolism Prevention in a Medical Indications Study [ARTEMIS]) has recently been published [5]. The ARTEMIS study, which used venography to confirm DVT, showed a significant reduction in the incidence of VTE in elderly acutely ill medical patients receiving fondaparinux, 2.5 mg o.d., compared with placebo (5.6% vs. 10.5%, respectively, p = 0.029; relative risk reduction 46.7%; 95% CI 7.7%-69.3%) ( Table 1) [5]. There was also a significant reduction in the incidence of symptomatic fatal or non-fatal pulmonary embolism in the fondaparinux group compared with the placebo group (1% vs. 3%, p = 0.029). Patient inclusion criteria in ARTEMIS differed from those used in the MEDENOX and PREVENT studies: patients were older (>60 years compared to >40 years) and were immobilised for longer (>4 days compared with ≤ 3 days) [3][4][5]. This may explain the high mortality rate in the placebo group of the ARTEMIS study; almost twice as many deaths occurred in the placebo arm compared with the fondaparinux arm (6.0% vs 3.3%, p = 0.06) [5]. Considering that the efficacy of thromboprophylaxis with LMWH or UFH is now firmly established in medical patients at risk of VTE, and that thromboprophylaxis with LMWH or low-dose UFH has a grade 1A recommendation from the ACCP consensus guidelines, the ethics of including a placebo arm in a study of this patient group is now questionable. Currently, only UFH, enoxaparin and dalteparin are approved for use in the prevention of VTE in medical patients. Further studies comparing fondaparinux with UFH or LMWH are awaited.

Safety of thromboprophylaxis
Low-molecular-weight heparin o.d. has similar efficacy to UFH t.i.d. when used as thromboprophylaxis. However, several randomized clinical studies have shown that LMWH has a better safety profile than UFH (  (Table 3).
In the previously described meta-analysis of randomized studies that showed equivalent thromboprophylactic efficacy when directly comparing LMWH and UFH [35], use of LMWH also reduced the relative risk of major haemorrhage by 52% compared with UFH (relative risk, 0.48; 95% CI, 0.23-1.00; p = 0.049) [35]. Another, more recent, analysis pooled safety data from 2,346 patients from the MEDENOX, THE-PRINCE, and THE PRIME studies to determine the risk of haemorrhage following thromboprophylaxis with UFH or enoxaparin [39]. While the incidence of major haemorrhage was similar in patients given enoxaparin, UFH, or placebo, the incidence of minor bleeding associated with UFH was significantly greater than that associated with enoxaparin (relative risk, 1.7; 95% CI, 1.3-2.2; p = 0.0001).
The ARTEMIS study [5] reported a low risk of bleeding complications when elderly acutely ill medical patients were given thromboprophylaxis with fondaparinux 2.5 mg o.d.: major bleeding occurred in one patient in the fondaparinux group (0.2%) and one in the placebo group (0.2%). Minor bleeding occurred in 11 patients (2.6%) in the fondaparinux group and four in the placebo group (1.0%) [5].
The risk of heparin-induced thrombocytopenia (HIT) should also be considered when providing patients with thromboprophylaxis. The risk of HIT has been shown to be significantly lower in surgical patients receiving LMWH prophylaxis than in those receiving UFH [40]. Recent studies, however, suggest that the incidence of HIT in UFH-treated medical patients may be lower than in surgical patients [31,41]. Nevertheless, HIT remains an important medical issue because of the associated risk of thromboembolic events. A recent prospective cohort study reported that the incidence of HIT in hospitalized medical patients receiving subcutaneous UFH was less than 1%, but HIT was associated with a high incidence (60%) of thromboembolic events [41].

Cost-effectiveness of thromboprophylaxis
Several economic studies have examined the cost-effectiveness of thromboprophylaxis in medical patients [42][43][44][45]. Three studies based on data from the MEDENOX study [3] showed that thromboprophylaxis with enoxaparin (40 mg s.c. o.d.) was cost-effective compared with placebo when examined from French, Canadian and Spanish cost perspectives [42][43][44]. A UK study of combined data from the MEDENOX study and a meta-analysis by Mismetti et al [35] showed that enoxaparin (40 mg s.c. o.d.) was associated with fewer VTE events and lower costs compared with no thromboprophylaxis [45] and was cost neutral compared with UFH.

Conclusion
Evidence from clinical studies has shown the benefits of thromboprophylaxis with LMWH (enoxaparin and dalteparin) or UFH for the prevention of VTE in at-risk medical patients. Thromboprophylaxis with LMWH is as effective as UFH at preventing VTE in this group of patients, but has a significantly better safety profile. If UFH is used, physicians should choose a t.i.d. dose regimen, as there are no data to show that UFH b.i.d. is as effective as LMWH o.d. or UFH t.i.d. dosing. Recent data suggest that fondaparinux may be an effective and safe option in high-risk elderly medical patients, but further studies are awaited before general recommendations can be given regarding its use in all acutely ill medical patients. On the basis of currently available evidence, it appears that, while low-dose UFH is also recommended by the ACCP [1] although without specifying a dose regimen, the LMWHs enoxaparin and dalteparin may be more attractive alternatives for the prevention of VTE in medical patients because of their convenient once-daily administration and better safety profile, in terms of fewer bleeding complications, and a lower risk of HIT and other adverse events.