The D-KAF (Dalteparin in Knee-to-Ankle Fracture) Trial

RESEARCH QUESTIONS

1. What is the incidence of clinically-important venous thromboembolism (VTE) following surgically-repaired, isolated fractures of the tibia, fibula or ankle?

2. In patients with surgically-repaired, isolated fractures of the tibia, fibula or ankle, is simple anticoagulant thromboprophylaxis effective, safe, and cost-effective?

   BACKGROUND AND RATIONALE

   Isolated below-knee fractures are very common injuries seen in all emergency departments and general hospitals. Venographic studies suggest that the incidence of overall deep vein thrombosis (DVT) following such fractures is in the range of 20-50%. However, the prospective studies that have screened patients with these fractures for DVT are few in number and have methodologic limitations. In addition, most of the thrombi detected are distal and small and, therefore, of uncertain clinical significance. Although patients with leg fractures are at risk for symptomatic DVT, PE, and occasionally fatal PE, there is a paucity of reliable data about the incidence of, and risk factors for, the development of clinically-important VTE following contemporary treatment of these fractures.

   Furthermore, data related to thromboprophylaxis in these patients are insufficient to recommend prophylaxis, even in patients whose fractures are managed surgically. There is considerable variation in the use of prophylaxis for such patients across North America; the majority of orthopedic surgeons use no prophylaxis, a small proportion use in-hospital prophylaxis only, and an even smaller proportion use therapeutic levels of anticoagulation until the patient is fully mobile without a cast. The routine use of prophylaxis for patients with below-knee fractures may not be necessary, may be associated with adverse effects related to the prophylaxis, and may not be cost-effective. A methodologically rigorous study is required to address all of the above issues.

   The D-KAF trial will determine the incidence of, and risk factors for, clinically-important VTE in patients with isolated fractures of the tibia, fibula or ankle, who are managed surgically. At the same time, D-KAF will assess the role, if any, of routine thromboprophylaxis in such patients. The study has the potential to significantly impact on the care of these patients irrespective of whether the result is "positive" or "negative." The use of clinically relevant endpoints will be of great importance in the design of future prophylaxis trials in this and other patient groups.

   OBJECTIVES

   Among patients with isolated, lower extremity fractures of the tibia, fibula or ankle, treated surgically, to determine:

   1. The incidence of clinically-important VTE over the first 3 months after injury in patients receiving usual care (no thromboprophylaxis);
   2. The predictors of clinically-important VTE in usual care patients;
   3. The effectiveness, compared to placebo, of a LMWH, dalteparin, given as 5,000 U once daily for up to 14 days, in the prevention of clinically-important VTE;
   4. The safety of this prophylaxis regimen with respect to clinically-important bleeding;
   5. The cost-effectiveness of the active intervention;
   6. The predictors of prophylaxis failure (clinically-important VTE and adverse events);
   7. The incidence of clinically-important VTE from the time that LMWH is discontinued to week 12; this objective seeks to determine if prophylaxis beyond the first 2 weeks is indicated.

   STUDY DESIGN

   This is a randomized, double-blind, multi-center comparison of the LMWH, dalteparin, versus placebo in patients with isolated fractures of the tibia, fibula or ankle, who undergo surgical repair. It will assess the epidemiology of VTE following these injuries and determine the effectiveness, safety and cost-effectiveness of dalteparin in preventing clinically important VTE.

   INTERVENTIONS

   All eligible patients who have provided informed consent, are randomized to receive:

   dalteparin, 5000 anti-Xa units OR saline placebo given once daily by subcutaneous injection using identical pre-loaded syringes

   • The first dose of study drug commences as soon as possible after the fracture and always within 72 hours of injury.
   • Non-study anticoagulant prophylaxis and mechanical prophylaxis (graduated compression stockings, pneumatic compression pumps or venous foot pumps) is not allowed. Use of aspirin, antiplatelet agents and NSAIDS is discouraged unless the indication is for cardiac or stroke prophylaxis, the management of arthritis, or short-term postoperative analgesia.

   PATIENT ASSESSMENT

   Baseline Assessment:

   At baseline, each patient has:

   • History and physical examination
   • Completion of baseline case report form which summarizes past medical and surgical history and details of the current injury and its management
   • CBC, INR, PTT, creatinine

   Follow-up Assessments:

   • All inpatients are seen daily to monitor adherence to the protocol as well as symptoms and signs of DVT, PE, clinically significant bleeding and other adverse events.
   • If discharged prior to 14 days, patients are taught to self-administer the study injections at home by a research nurse. Patients are called at home every 3 days to ensure compliance with the study medication and to address any questions or concerns related to the study.
   • All patients are seen in follow-up 14 ± 2 days after injury. At this visit, a focussed history and physical examination is performed and a CBC obtained. All patients are screened for proximal DVT using bilateral duplex ultrasound examination.
   • Patients with normal DUS have the study medication discontinued and no further DVT prophylaxis is given for the remainder of the study period.
   • All patients receive a telephone follow up at 6 ± 1 week and 3 months ± 1 week after randomization to assess their clinical status. They are asked specifically about symptoms and signs of DVT and PE, hospitalizations, and co-intervention (administration of anticoagulants).
   • Clinical suspicion of possible DVT or PE during the 3-month study period is investigated in a standardized manner using objective diagnostic tests and pre-specified diagnostic algorithms.
   • All patients who develop objectively confirmed DVT or PE, either symptomatic or asymptomatic (detected on the day 14 screening ultrasound), are treated with full-dose anticoagulation (generally LMWH followed by warfarin) for at least 3 months.

   OUTCOME MEASURES

   Primary Outcome = Clinically important venous thromboembolism in the two study groups. This is a composite outcome measure that includes symptomatic DVT, symptomatic PE, and fatal PE during the 3 months following fracture plus asymptomatic proximal DVT at the end of the prophylaxis phase.

   Symptomatic DVT: is defined as clinically suspected DVT, confirmed by positive duplex ultrasound or venography.

   Symptomatic PE: is defined as clinically suspected PE, confirmed by high probability V/Q scan or positive leg imaging for DVT, spiral CT or pulmonary angiogram.

   Fatal PE: as confirmed by autopsy.

   Asymptomatic proximal DVT: by screening ultrasound examination at 14 +2 days.

   Secondary Outcomes:

   Clinically important VTE during the prophylaxis phase (day 1 to discontinuation of study medication - this usually corresponds to the end-of-prophylaxis phase DUS or day 14 + 2).

   Clinically important VTE during the post-prophylaxis phase (usually day 14 to 3 months).

   Bleeding (modified International Bleeding Classification):

   Major - defined as overt bleeding that is life threatening or requires surgical intervention, the transfusion of 1 or more units of red blood cells or is associated with a drop in hemoglobin of ³ 20 g/L or development of hemarthrosis or compartment syndrome.

   Minor - overt bleeding other than that defined as major. Bleeding index: This validated measure of bleeding takes into account both change in hematocrit and the amount of blood transfused.

   Costs associated with the prophylaxis and with the relevant outcomes.

   ANALYSES

   Baseline characteristics of the two study groups will be compared for patients randomized and for those who successfully complete the study using appropriate parametric or nonparametric statistics.

   Primary analysis will be based on an intention-to-treat principle and will compare the proportion of patients in each group that develop the primary outcome, objectively-confirmed, clinically-important VTE, over the 3 month follow-up period, using the Yates-corrected chi-square test of two independent proportions.

   Secondary analyses will compare the rates of clinically important VTE for the two study phases, the prophylaxis phase (randomization to day 14 + 2) and the post-prophylaxis phase (day 14 to 3 months).

   A safety analysis will compare bleeding and other adverse events in the two groups over the prophylaxis phase and the entire study period.

   Compliance with the assigned intervention will be compared.

   Cost-effectiveness will be determined prospectively.

   Predictors of thrombosis in patients with lower extremity fractures (from the placebo group) and predictors of prophylaxis failures (from the active treatment group) will be explored using logistic regression analysis (although we recognize that the number of events will likely be low). The factors to be considered include: age, gender, obesity (BMI), smoking, varicose veins, oral contraceptive or hormone replacement use, number of risk factors, type of fracture, method of fracture treatment, delay to surgery, OR time, time after injury to initiation of prophylaxis, compliance, mobility at the end of the prophylaxis phase, and fracture complications (including infection, reoperation, non-union).

   SAMPLE SIZE ESTIMATION

   1. Assumptions used: a = 0.05 (1-tail), power = 80%, rate of clinically important VTE in the control group = 4%; risk reduction with prophylaxis = 75%, yielding a Number Needed to Prophylax = 33.
   2. Assuming a 5% post-randomization attrition rate, the estimated sample size will be 350 patients per group; a total of 700 patients.
   3. A formal, blinded interim analysis will be performed after 500 patients have completed follow-up.