Low-Molecular-Weight Heparin versus Compression Stockings for Thromboprophylaxis after Knee Arthroscopy: A Randomized Trial

Camporese and colleagues (1) reported the results of the KANT trial and concluded that thromboprophylaxis with low molecular weight heparin (LMWH) should be offered to all patients undergoing knee arthroplasty. This recommendation differs from the American College of Chest Physicians (ACCP) evidence-based guidelines on the Prevention of Venous Thromboembolism (VTE) which recommend against the routine use of thromboprophylaxis following knee arthroscopy(2,3). I have a number of concerns about the KANT trial and my interpretation of the results differs from the authors.

First, the accuracy of DUS has been shown to be suboptimal for asymptomatic calf DVT and has never been validated as an outcome measure following arthroscopy(4,5,6).

Second, the distribution of calf DVT and muscular vein thrombi for each of the treatment groups is not presented. This is important because muscular vein thrombi are generally not even considered to be DVT; these thrombi are usually not included as outcomes in clinical trials; and their clinical relevance is uncertain at best. In KANT, 77% of the leg thrombi were distal (including true calf DVT and muscular vein thrombi of questionable significance). Providing the distribution of the types of thrombi is essential for readers since the difference between the groups is based on these distal thrombi.

Third, "symptomatic" calf thrombi are included in the primary outcome while "asymptomatic" calf thrombi are not. This distinction is not justified either in the publication or in the literature. Overall, 50% of all patients were "symptomatic". Symptoms were assessed in a manner unblinded to the intervention and none of the patients labeled as "symptomatic" were diagnosed at an unscheduled visit. There was also no relation between symptoms and leg thrombi. In fact, asymptomatic patients were twice as likely to have leg thrombi as symptomatic patients (3.6% vs. 1.8%).

Fourth, if we consider the hierarchy of thromboembolic outcomes based on clinical importance, there were no significant differences in death, symptomatic PE, symptomatic proximal DVT or PE, or all proximal DVT or PE. The differences between groups were based on distal thrombi, some of which were muscular vein thrombi of questionable or no clinical relevance and using an unjustified distinction between "symptomatic" and "asymptomatic" distal thrombi. When clinically important bleeding is added to the largely asymptomatic VTE outcomes, the relative benefits of LMWH are further reduced. Using the author's definition for the primary efficacy outcome, in the 7-day LMWH patients, there were 6 thromboembolic events (some asymptomatic and some of questionable significance) and 6 clinically important bleeding events - hardly a favorable tradeoff. Fifth, a previously published abstract described the KANT trial as a two-arm trial (7). Could the authors verify for interested readers the following: whether (and, if so, when) the trial design was changed from a two-arm to a three-arm trial; whether the sample size calculations reported in the published paper were those originally calculated for the three-arm trial; and whether the trial protocol a priori specified including muscular vein thrombi and only symptomatic (versus asymptomatic) calf DVT as components of the primary outcome measure?

Finally, before we can accept the authors' assertion about higher risk arthroscopy subgroups, all of the event rates need to be revealed to see if the conclusion still holds after removal of muscular vein thrombi, after combining symptomatic and asymptomatic calf DVT, and determining if the bleeding was also more common in patients having a menisectomy.

The authors claim that their study demonstrates that 7 days of LMWH reduces "clinically relevant venous thromboembolic events". Based on the data presented, this is not true. If orthopedic surgeons are faced with a decision to implement patient-administered subcutaneous injections of anticoagulant to a large number of outpatients that will not reduce the incidence of clinically-important thromboembolic events and are associated with a small increased risk of bleeding, it would be appropriate for them to reject this option except for the highest risk patients. This is the approach recommended by the guidelines.(2,3)

William Geerts, MD Sunnybrook Health Science Centre Toronto, ON, Canada M4N 3M5

References

1. Camporese G, Bernardi E, Prandoni P, Noventa F, Verlato F, Simioni P, et al. Low-molecular-weight heparin versus compression stockings for thromboprophylaxis after knee arthroscopy. A randomized trial. Ann Intern Med. 2008;149:73-82

2. Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell C, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126:338S-400S

3. Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, et al. Prevention of venous thromboembolism. American College of Chest Physicians Evidence-Based Practice Guidelines (8th Edition). Chest. 2008;133:381S-453S

4. Schellong SM, Beyer J, Kakkar AK, Halbritter K, Eriksson BI, Turpie AG, et al. Ultrasound screening for asymptomatic deep vein thrombosis after major orthopedic surgery: the VENUS study. J Throm Haemost. 2007;5:1431-7

5. Goodacre S, Sampson F, Thomas S, van Beek E, Sutton A. Systematic review and meta-analysis of the diagnostic accuracy of ultrasonography for deep vein thrombosis. BMC Med Imaging. 2005;5:1-13

6. Kassai B, Boissel J-P, Cucherat M, Sonie S, Shah NR, Leizorovicz A. A systematic review of the accuracy of ultrasound in the diagnosis of deep venous thrombosis in asymptomatic patients. Thromb Haemost 2004;91:655-66

7. Camporese G, Bernardi E, Prandoni P, et al. Graduated compression stockings (GCS) versus low-molecular-weight heparin (LMWH) for prevention of deep vein thrombosis (DVT) after knee arthroscopy (KA). A randomized study (KANT). J Thromb Haemost 2007;5(Suppl 2):O-T-052

We appreciate Dr. Geerts' efforts to evaluate our work. We are aware that our conclusions differ from the 8th ACCP guideline document about the prevention of venous thromboembolism after knee arthroscopy.(1) We also note that table 9 in those guidelines incorrectly lists the incidence of VTE found in our trial: the table's figures for our trial differ from those presented at the last ISTH Congress (Geneva, 2007) and from those presented in the published article. We offer the following responses to other points made by Dr. Geerts.

First, we agree that the accuracy of ultrasound in the setting of asymptomatic calf deep-vein thrombosis is suboptimal.(2) As stated in the methods of the published article, the primary efficacy end-point included only symptomatic calf DVT events, in accordance with the EMEA and the ACCP guidelines. Besides, ultrasound has been widely used in trials that evaluated thromboprophylaxis following knee (3-5) and ankle surgery (6).

Second, we strongly disagree with Dr. Geerts' statements that "muscular vein thrombi are generally not even considered to be DVT" and that "these thrombi are usually not included as outcomes in clinical trials". According to Gray's Anatomy, muscular veins are classified as deep as they are tributaries of the popliteal vein and accompany the branches of the popliteal artery.(7) As far as we are aware, none of the large trials of thromboprophylaxis after orthopedic surgery report "the distribution of calf DVT". Rather, most trials report "distal DVT" outcomes detected by either venography or ultrasonography (6,8,9). Regardless, for the sake of discussion, we report more detail regarding the site of asymptomatic and symptomatic thrombi below.

Third, we think that a distinction between symptomatic and asymptomatic calf DVT is justified based on the EMEA and ACCP requirements for ultrasonographic end-points. We could not afford a double-blind, double-dummy design. To minimize bias we employed an assessor-blind strategy. We agree with Dr. Geerts' statement that "asymptomatic patients were twice as likely to have leg thrombi as symptomatic patients." Many large trials that evaluate thromboprophylaxis report similar findings.

Fourth, it is not unusual for thromboprophylaxis trials to find statistically significant differences between groups in distal DVT, yet find no statistically significant differences between groups concerning less frequent outcomes such as death, symptomatic PE, symptomatic proximal DVT or PE, or all proximal DVT or PE. We note that the overall number of efficacy plus safety events in the 7-day LMWH group was about 50% less than the overall number of those events in the graduated compression stocking group (12 versus 23). Also, in the 7-day LMWH group, 4 of 6 safety events were non life-threatening haemarthroses only.

Fifth, as stated in the published article, the trial was designed as a three-arm trial and the data and safety monitoring board halted the 14-day heparin group after the second interim analysis. The sample size calculations reported in the published paper were those originally calculated for the three-arm trial. The trial protocol a priori specified only symptomatic distal DVT and not asymptomatic distal DVT as a component of the primary endpoint.

In conclusion, we believe our results are valid and we agree with Dr. Hull's editorial which says, "it is now time to change our clinical practice" and "to revisit policies that maximize bleeding avoidance and minimize thrombosis avoidance", as "sudden death due to massive pulmonary embolism may well be the dominant harm rather than fatal bleeding due to thromboprophylaxis".(10)

Giuseppe Camporese, MD; Enrico Bernardi, MD, PhD °; Franco Noventa, MD * Unit of Angiology and * Department of Clinical and Experimental Medicine, Clinical Epidemiology Group, at the University Hospital of Padua; and ° Department of Emergency and Accident Medicine, Hospital of Conegliano Veneto, Italy

References

1. Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, et al. Prevention of venous thromboembolism. American College of Chest Physicians Evidence-Based Practice Guidelines (8th Edition). Chest. 2008;133:381S-453S.

2. Kearon C, Julian JA, Newman TE, Ginsberg JS. Noninvasive diagnosis of deep venous thrombosis. McMaster Diagnostic Imaging Practice Guidelines Initiative. Ann Intern Med 1998;128(8):663-77

3. Wirth T, Schneider B, Misselwitz F, et al. Prevention of venous thromboembolism after knee arthroscopy with low molecular weight heparin (Reviparin): results of a randomized controlled trial. Arthroscopy 2001;17(4):393-9.

4. Michot M, Conen D, Holtz D, et al. Prevention of deep-vein thrombosis in ambulatory arthroscopic knee surgery: a randomized trial of prophylaxis with low-molecular weight heparin. Arthroscopy, 2002;18:257-263

5. Lassen MR, Ageno W, Borris LC, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty. N Engl J Med 2008;358(26):2776-86.

6. Lapidus L, de Bri E, Ponzer S, Elvin A, Norén A, Rosfors S. High sensitivity with color duplex sonography in thrombosis screening after ankle fracture surgery. J Thromb Haemost 2006;4(4):807-12.

7. Gray's Anatomy. Henry Gray (1821"“1865). Anatomy of the Human Body. 1918. (Accessed on Bartleby.com Aug 2008 6c under popliteal artery branches)

8. Leizorovicz A, Kassai B, Becker F, Cucherat M. The assessment of deep vein thromboses for therapeutic trials. Angiology. 2003;54(1):19-24

9. Kassaï B, Boissel JP, Cucherat M, Sonie S, Shah NR, Leizorovicz A. A systematic review of the accuracy of ultrasound in the diagnosis of deep venous thrombosis in asymptomatic patients.Thromb Haemost. 2004;91(4):655-66.

10. Hull R. Thromboprophylaxis in knee arthroscopy patients: revisiting values and preferences. Ann Intern Med, 2008;149:137-139.