Risk-based approach must guide VTE prophylaxis for ambulatory patients with cancer

Individuals with cancer are more than four times as likely as those in the general population to develop venous thromboembolism.

Chemotherapy receipt increases that risk to more than sixfold, a major reason why VTE complications are the second leading cause of death for those with cancer behind only the malignancy itself.

Patients who require chronic anticoagulation are at increased risk for bleeding, cancer therapy delays, VTE recurrence, impaired quality of life, increased health care utilization, and increased morbidity and mortality.

Thromboprophylaxis for hospitalized patients with cancer is universally recommended; however, there has been little guidance on primary VTE prophylaxis in the ambulatory setting, despite a better understanding of risk factors that predispose those with cancer to VTE.

Risk factors

The Khorana Risk Score has been shown to predict VTE among ambulatory patients with cancer.

The risk assessment model — validated in more than 15,000 patients — incorporates clinical variables routinely collected in practice, such as tumor site, BMI, leukocyte count, hemoglobin and platelet count. This makes integration into daily workflow feasible.

A real-world analysis of 6,194 people newly diagnosed with cancer initiating therapy confirmed 12-month cumulative VTE rates were significantly different based on whether patients had Khorana Risk Scores of 0 (2.8%), 1 (4.6%), 2 (7.9%), or 3 or higher (13.2%). More than 25% of patients had scores of 2 or higher, and they were three times more likely than others to develop VTE after cancer diagnosis.

Some groups have suggested improved predictive models that incorporate D-dimer or soluble P-selectin levels, or even genetic risk factors that predispose people to VTE.

The Khorana Risk Score remains the most extensively validated model, but there is room for improvement.

The debate of simplistic vs. complex models and sensitivity thresholds to predict VTE will impact adoption in clinical practice. Regardless, improving the ability to predict VTE risk is of substantial importance and will provide guidance for thromboprophylaxis.

ASCO guidelines do not recommend routine thromboprophylaxis for ambulatory patients with cancer, although it may be considered for “highly select high-risk patients.”

Further, patients with multiple myeloma who are receiving immunomodulatory (IMiD) agents plus dexamethasone should receive prophylaxis with either low-molecular-weight heparin (LMWH) or aspirin to prevent VTE.

Lastly, the use of novel oral anticoagulants (NOACs) is not recommended to treat or prevent VTE among patients with cancer.

Primary VTE prevention among patients with cancer

Prior randomized trials that included mixed cancer populations evaluated extended prophylaxis with LMWH.

Although some identified statistical reductions in VTE events compared with placebo (eg, SAVE-ONCO trial, 1.2% vs. 3.4%, P < .001; PROTECHT trial, 2% vs. 3.9%, P = .02), the overall event rates and absolute benefit were low, resulting in questionable benefit over risk.

More recently, groups have capitalized on using a risk-based approach to widen the margin of benefit. They also have utilized NOAC therapy, which may be a more favorable option than LMWH.

The randomized, placebo-controlled, double-blind AVERT trial assessed the efficacy and safety of apixaban (Eliquis; Bristol-Myers Squibb, Pfizer) 2.5 mg twice daily for thromboprophylaxis of ambulatory patients with cancer who were at intermediate to high risk for VTE (Khorana Risk Score 2) and initiating chemotherapy.

In the modified intention-to-treat population, which included 563 patients who received at least one dose of the study drug, VTE rates were 4.2% in the apixaban group and 10.2% in the placebo group (HR = 0.41; P < .001). Major bleeding rates were 3.5% in the apixaban group and 1.8% in the placebo group (HR = 2; P = .046).

CASSINI investigators conducted a similar randomized, placebo-controlled, double-blind trial using rivaroxaban (Xarelto, Janssen) 10 mg daily for ambulatory patients with cancer who had Khorana Risk Scores of 2 or higher and were initiating chemotherapy.

In the intention-to-treat population (n = 841), researchers reported VTE rates of 5.95% in the rivaroxaban group and 8.79% in the placebo group (HR = 0.66; 95% CI, 0.4-1.09). Importantly, nearly 40% of events occurred after discontinuation of the study drug.

In a prespecified analysis of all randomly assigned patients during the on-treatment period, VTE occurred at rates of 2.62% in the rivaroxaban group and 6.41% in the placebo group (HR = 0.4; P = .007).

Event rates were notably lower in the CASSINI trial than the AVERT trial, which may be due in part to protocol-mandated screening ultrasound of the lower extremity, excluding 4.5% of patients with asymptomatic VTE at baseline.

The most common tumor types for patients enrolled in CASSINI were pancreatic, gastrointestinal and lung, whereas AVERT primarily enrolled patients with gynecologic cancer, lymphoma and pancreatic cancer.

Approximately 40% of patients in both studies discontinued prophylaxis prior to completing 180 days. However, there was no difference in discontinuation rate between active therapy and placebo, nor any difference in discontinuation due to drug tolerability.

CASSINI and AVERT, for the first time, demonstrated a wider margin of benefit using NOACs to prevent VTE among high-risk ambulatory patients with cancer.

Although CASSINI did not meet its primary endpoint, the findings underscore the importance of continued drug administration during the high-risk 6-month period.

Further, CASSINI demonstrated a high incidence of asymptomatic VTE at baseline, raising the question of whether patients with a high Khorana Risk Score should be screened by ultrasound prior to starting chemotherapy.

Using a risk-based approach, both trials clearly showed a lower number needed to treat and larger number needed to harm (ie, bleeding) than prior studies with LMWH that did not use a risk-based approach, favoring clinical benefit over risk for toxicity.

VTE prevention in multiple myeloma

Use of IMiD drugs — such as thalidomide (Thalomid, Celgene), lenalidomide (Revlimid, Celgene) or pomalidomide (Pomalyst, Celgene) — increases VTE risk for patients with multiple myeloma.

According to International Myeloma Working Group (IMWG), risk is further increased by the addition of dexamethasone, as well as the following risk factors: BMI, history of VTE, central venous catheter, cardiac disease, chronic kidney disease, diabetes, acute infection, immobilization, surgery, anesthesia, trauma, erythropoietin-stimulating agent, clotting disorders, hyperviscosity and multiagent chemotherapy.

IMWG guidelines recommend thromboprophylaxis based on risk assessment with either aspirin, LMWH or warfarin. VTE rates from the Myeloma XI trial were high (11.8%) despite protocol-directed use of guidelines for VTE prophylaxis.

Sanfilippo and colleagues further demonstrated that aspirin may be inadequate as thromboprophylaxis for patients with multiple myeloma receiving IMiDs.

IMWG guidelines were last updated in 2010 and require investigation into validated risk factors and adequate thromboprophylaxis.

Several abstracts presented in December at ASH Annual Meeting and Exposition attempted to address VTE risk factors in multiple myeloma.

Sanfilippo and colleagues developed the IMPEDE VTE score based on an analysis of 4,448 veterans with multiple myeloma. The score is calculated based on several factors, including IMiD use (3 points), BMI (1 point), pathologic fracture of pelvis or femur (2 points), use of erythropoietin-stimulating agent (1 point), dexamethasone use (4 points for high dose, 2 points for low dose), doxorubicin (2 points), Asian ethnicity/race (–3 points), history of VTE (3 points), central venous catheter (2 points), use of therapeutic anticoagulation with warfarin or LMWH (–5 points), or use of prophylactic anticoagulation with LMWH or aspirin (–2 points).

VTE rates within the first 6 months of starting chemotherapy were 3.1% for patients with scores of 3 or less, 7.5% for those with scores of 4 to 6, and 13.3% for patients with scores of 7 or higher.

Li and colleagues analyzed 2,397 patients with multiple myeloma receiving IMiDs. They identified several factors associated with VTE risk: history of VTE, recent surgery, cytotoxic chemotherapy, higher-dose dexamethasone and older age. Race also appeared to be a factor, with black individuals demonstrating higher risk and Asian patients demonstrating lower risk.

VTE incidence was greater in the very high-risk group than standard-risk group at 3 months (9.5% vs. 3.7%) and 6 months (16.3% vs. 6.3%).

Researchers at Levine Cancer Institute at Atrium Health performed an institutional analysis of 373 patients with multiple myeloma receiving IMiDs plus dexamethasone. They identified no significant difference in incidence of thromboembolism — venous and arterial combined — between those who received aspirin prophylaxis or a stronger anticoagulant (eg, LMWH, warfarin or NOAC). However, those who received stronger anticoagulants had more baseline comorbidities.

The overall rate of thromboembolism was 8.3%, and those with at least one comorbidity identified by IMWG had a significantly higher risk for an event. Researchers observed no association between thromboembolism risk and disease burden as measured by bone marrow biopsy plasma cell percentage.

These data suggest an urgent need to update guidelines regarding risk-based thromboprophylaxis for patients with multiple myeloma, as well as the importance of further investigation into the use of stronger anticoagulants in this setting.

Early data from two studies show encouraging efficacy and safety results regarding the use of apixaban to prevent VTE among patients with multiple myeloma receiving IMiD therapy. Randomized studies are needed to confirm NOAC benefit and stratified thromboprophylaxis based on risk assessment (eg, those with at least one comorbidity receive NOAC and all others receive aspirin).

Future implications

Data from CASSINI and other studies suggest a high rate of asymptomatic baseline VTE among patients with cancer, particularly those with high Khorana Risk Scores.

These incidental events are associated with similar recurrence and mortality rates as symptom-detected events, underscoring their clinical significance and the need for earlier screening or closer monitoring.

Ideally, those with high Khorana Risk Scores are referred to a nonmalignant hematology specialist and considered for baseline ultrasound screening and primary thromboprophylaxis.

Clinical decision support tools embedded within the electronic medical record that automatically calculate Khorana Risk Score and trigger referrals would improve screening and reduce VTE complications downstream.

The question of appropriate thromboprophylaxis therapies is becoming clearer.

Traditionally, LMWH and warfarin were the mainstay therapies, with warfarin falling out of favor over the years due to increased monitoring, drug interactions and side effects.

NOACs have been used in the past despite a lack of data, but they are now increasingly utilized to both treat and prevent VTE based on newer compelling trial results.

The ADAM VTE, Hokusai-VTE and Select-D trials demonstrated at least noninferiority of NOACs compared with LMWH for VTE treatment.

Studies have suggested patients tend to prefer NOACs to LMWH due to their oral administration rather than a subcutaneous route, and this can increase compliance. Further, major bleeding rates are low, alleviating the concern of monitoring or requiring a reversal agent, although andexanet alfa (Andexxa, Portola Pharmaceuticals) is now commercially available for this indication.

The last few years have provided us with compelling data to suggest drastic changes to VTE guidelines are needed, particularly regarding risk assessments, primary prevention and use of NOACs.

A concerted effort should be made earlier in the cancer care continuum to reduce VTE incidence and its associated morbidity and mortality.

References:

Agnelli G, et al. N Engl J Med. 2012;doi:10.1056/NEJMoa1108898.

Agnelli G, et al. Lancet Oncol. 2009;doi:10.1016/S1470-2045(09)70232-3.

Bradbury CA, et al. Abstract 331. Presented at: ASH Annual Meeting and Exposition; Dec. 9-12, 2017; Atlanta.

Heit JA, et al. Arch Intern Med. 2000;160(6):809-815.

International Myeloma Working Group. IMWG guidelines for the prevention of thalidomide- and lenalidomide-associated thrombosis in myeloma. Available at: imwg.myeloma.org/imwg-guidelines-for-the-prevention-of-thalidomide-and-lenalidomide-associated-thrombosis-in-myeloma. Accessed on Jan. 20, 2019.

Khorana AA, et al. Blood. 2008;doi:10.1182/blood-2007-10-116327.

Knight R, et al. N Engl J Med. 2006;354(19):2079-2080.

Lyman GH, et al. J Clin Oncol. 2015;doi:10.1200/JCO.2014.59.7351.

Pegourie B, et al. J Clin Oncol. 2017;doi:10.1200/JCO.2017.35.15_suppl.8019.

Prandoni P, et al. Blood. 2002;100(10):3484-3488.

Raskob GE, et al. N Engl J Med. 2018;doi:10.1056/NEJMoa1711948.

Sanfilippo KM, et al. Abstract 3419. Presented at: ASH Annual Meeting and Exposition; Dec. 9-12, 2017; Atlanta.

Silverstein MD, et al. Arch Intern Med. 1998;158(6):585-593.

Young A, et al. Blood. 2017;130:625.

The following were presented at ASH Annual Meeting and Exposition; Dec. 1-4, 2018; San Diego:

Cornell RF, et al. Abstract 1233.

Khorana AA, et al. Abstract LBA-1.

Li A, et al. Abstract 144.

McBane RD, et al. Abstract 421.

Patel JN, et al. Abstract 4814.

Sanfilippo KM, et al. Abstract 141.

For more information:

Jai N. Patel, PharmD, BCOP, is chief of pharmacology research and associate professor in the division of hematology/oncology at Levine Cancer Institute at Atrium Health. He also is a HemOnc Today Editorial Board Member. He can be reached at jai.patel@atriumhealth.org.

Disclosure: Patel reports serving on the CASSINI steering committee, receiving research funding and consultant fees.