| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Medical Oncology (Drs. Ferretti, Bria, Carlini, Felici, Papaldo, Fabi, Ciccarese, Cuppone, Cecere, Nuzzo, Terzoli, and Cognetti), Regina Elena Cancer Institute, Rome; Biostatistics Unit (Dr. Giannarelli), Regina Elena Cancer Institute, Rome; and Division of Medical Oncology (Dr. Mandalà), Treviglio Hospital, Treviglio, Italy.
Correspondence to: Gianluigi Ferretti, MD, PhD, Division of Medical Oncology "A", Regina Elena Cancer Institute, Via Elio Chianesi 53, 00144 Rome, Italy; e-mail gia.fer{at}flashnet.it
Abstract
Purpose: To evaluate whether the incidence of recurrent venous thromboembolism (VTE) events after therapy differs for patients treated with long-term low-molecular-weight heparin (LMWH) or oral anticoagulant therapy (OAT).
Methods: All randomized studies were searched through computerized queries of MEDLINE, the Cochrane Controlled Trials Register, the American Society of Hematology abstract database, and the American Society of Clinical Oncology abstract database.
Results: Eleven studies including 2,907 patients were identified. Seven studies evaluated a period of 3 to 9 months after cessation of the allocated treatment: 5.4% of patients in the LMWH group vs 4% in the arm allocated to OAT had an episode of recurrent symptomatic VTE. Combined analysis showed a nonsignificant trend in lowering recurrent symptomatic VTE in favor of OAT (relative risk [RR], 1.29; 95% confidence interval [CI], 0.82 to 2.02; p = 0.27). By contrast, during active treatment, a statistically significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH over OAT was registered (RR, 0.63; 95% CI, 0.47 to 0.83; p = 0.001). Regarding cancer patients only, 37 of 569 patients (6.5%) in the LMWH group had recurrent symptomatic VTE vs 69 of 546 patients (12.6%) in the OAT group, with a statistically significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH (RR, 0.52; 95% CI, 0.35 to 0.76; p = 0.001).
Conclusions: Despite the significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH over OAT during treatment, patients treated with long-term LMWH do not seem to have more frequently recurrent VTE events compared with OAT after cessation of therapy. The significant difference favoring LMWH over OAT among all patients receiving treatment comes mostly from studies enrolling cancer patients.
Key Words: deep vein thrombosis • low-molecular-weight heparin • oral anticoagulant therapy • venous thromboembolism • vitamin k antagonists
The rebound of coagulation activation observed in patients with unstable angina treated with IV unfractionated heparin (UFH) following heparin treatment cessation has led to the theoretical concern of rebound following cessation of long-term low-molecular-weight heparin (LMWH) therapy. It has been reported12 that there is significant clustering of recurrent ischemic events within 24 h after cessation of both short-term UFH and enoxaparin treatment, and that patients should be carefully monitored during that period. Thus, it could be possible that the natural posttreatment history of recurrent venous thromboembolism (VTE) may differ for patients treated with long-term LMWH or oral anticoagulant therapy (OAT).
The purpose of this metaanalysis was to evaluate whether the incidence of recurrent VTE events after therapy differs for patients treated with LMWH or OAT (vitamin K antagonists [VKA]). The parallel aim was to address whether the rate of recurrent VTE events differed while patients were still receiving treatment.
Materials and Methods
Criteria for Considering Studies for Metaanalysis
The studies considered for this metaanalysis enrolled patients with objectively diagnosed symptomatic deep vein thrombosis (DVT), pulmonary embolism (PE), or both randomly allocated to long-term treatment with LMWH or OAT.
Outcome Measures
The primary outcome was the incidence of recurrent symptomatic VTE during the follow-up period and during treatment. The criteria accepted for the diagnosis of recurrent symptomatic VTE were those reported by Heijden et al.3 Patients with DVT and PE were not analyzed separately.
Selection of Studies
Randomized studies were searched through computerized queries of MEDLINE (1966 onwards, available at: www.ncbi.nlm.nih.gov/PubMed), the Cochrane Controlled Trials Register (available at: www.thecochranelibrary.com), the American Society of Hematology abstract database (available at: www.hematology.org), and the American Society of Clinical Oncology abstract database (available at: www.asco.org). For the search, we used key words "deep vein thrombosis," "low molecular weight heparin," "coumarins," and "randomized"; other key words used were "venous thromboembolism," "oral anticoagulant," "acenocoumarol," and "recurrent." Two reviewers reviewed and extracted data independently using a standard form.
Data Extraction
Eligible articles were independently reviewed, and summary information was independently extracted by two reviewers. The following information was sought: design of the study; number of patients with and without cancer; total number of patients; number of patients in each treatment arm; duration of therapy; type of anticoagulant (VKA or LMWH); and incidence of recurrent VTE. All calculations were independently performed by two different investigators.
Statistical Analysis
The incidence of recurrent symptomatic VTE for the different treatment arms were used to calculate the relative risk (RR) separately for each trial. These RRs were combined across the studies, giving weight to the number of events in each of the two treatment groups in each separate study using the inverse variance and the Mantel-Haenszel procedure method; both estimations were performed assuming a fixed-effect and a random-effect model.4 The heterogeneity between trials was tested with Q statistics, computing the square distance of each study from the combined effect and weighting these values with the inverse of variance of each study.5 Q statistics results were then compared with the 
2 distribution with k – 1 degree of freedom, where k was the number of studies. The significance of heterogeneity test suggests to prefer the random-effect estimation for a more appropriate evaluation of the results. All calculations were performed using statistical software (Comprehensive Meta-analysis, version 1.0.23; Biostat; Englewood, NJ).6
Primary analysis concerned all patients during the period of randomized treatment, and only patients recruited in those studies specifically reporting the aforementioned outcomes during the follow-up period. Patients with DVT and PE were not analyzed separately. Patients with and without malignancies were not analyzed separately, with the exception of recurrent VTE in those studies specifically reporting this outcome. All analyses were according to the intention-to-treat (ITT) analysis. When the individual studies did not use ITT analyses, the analyses of this review were on the basis of the data provided by the individual study.3
Description of Studies
Eleven randomized clinical trials7891011121314151617 were included in this metaanalysis (Table 1
). The progress through the stages of this metaanalysis is reported in a Quality of Reporting of Meta-Analysis diagram18 (Fig 1
). Two studies were excluded for non-English language.1920 The study by Hamann19 was a prospective, open randomized trial that recruited 200 patients (100 allocated to LMWH and 100 to VKA) with DVT confirmed with venography. The interventions used were phenprocoumon for 3 months or 6 months (international normalized ratio [INR], 2.0 to 3.0) compared with a 3- or 6-month course of subcutaneous dalteparin-sodium, 5,000 IU anti-Xa qd. Hamann19 did not use a clearly concealed randomization procedure, nor was blinded outcome assessment used in this study. The other study presented by the same author20 was a duplicate report.
|
|
One study,16 in a parallel-group comparison, randomly allocated patients to three treatment groups: initial UFH IV followed by VKA, initial LMWH followed by VKA, at initial LMWH followed by LMWH at a fixed maintenance daily dose. Another study17 showed a parallel design with patients allocated to one of three treatment groups: initial LMWH in the three groups, followed by a different dose of LMWH in two of them, and by VKA in the remaining group. Concerning the study by Kakkar et al,16 we compared group A and group B (initial UFH followed by VKA and initial LMWH followed by VKA) with group C (initial LMWH followed by fixed-dose maintenance LMWH); in the study by Deitcher et al,17 we compared group 1a and group 1b (initial LMWH followed by a different dose of LMWH) with group 2 (initial LMWH followed by warfarin).
All studies compared LMWH vs OAT administered for at least 3 months in the prevention of recurrent symptomatic VTE; but in some studies,811121314 treatment was prolonged for additional 3 to 9 months in an balanced/unbalanced fashion. In particular, in two studies,1213 anticoagulant treatment was prolonged for additional 6 months in a nearly balanced fashion; in two other studies,811 it was prolonged beyond 3 months in an unbalanced fashion, while Meyer et al14 did not report the type of prolonged anticoagulation by treatment arm (Table 1).
Quality of treatment with OAT was defined as an INR between 2.0 and 3.0. Data on the quality of OAT were provided in seven studies.78910121314 Only two studies714 specified the percentage of time during which the INR was in the therapeutic range (Table 1). Other studies91013 indicated the percentage of patients on target (INR, 2 to 3), or the percentage of patients with "good" anticoagulation (defined as at least 67% of INR values within the therapeutic range according to Pini et al,8 or > 75% of INR values within the therapeutic range according to Veiga et al12). The remainder studies11151617 did not report the quality of oral anticoagulation related to the INR target.
In the study by Meyer et al,14 the principal analysis of the primary outcome was conducted on 138 of 146 included patients, because 8 patients (4 in each group) were considered not evaluable by the outcome adjudication committee. The primary end point calculated by Meyer et al14 was a combined outcome event of treatment failure defined as symptomatic and objectively confirmed recurrent VTE and/or major bleeding within the 3-month treatment period. To evaluate the patients with recurrent VTE, we subtracted the number of patients with major bleeding from the number of patients with combined outcome events, since the authors14 reported that, during the 3-month treatment period, 15 patients assigned to receive warfarin had major hemorrhage or recurrent VTE compared with 7 patients assigned to receive enoxaparin. In the study by Lee et al,7 two patients in each group were excluded from the efficacy analysis because they did not have a qualifying thrombotic event, while three patients assigned to OAT did not receive the study drug and were excluded from the safety analysis.
Results
The 11 randomized clinical trials7891011121314151617 were published between 1994 and 2003 and included a total of 2,907 patients.
Recurrent Symptomatic VTE During Follow-up After Cessation of Active Treatment
Seven studies8101112131415 evaluated a period of 3 to 9 months after cessation of the allocated treatment. A total of 44 of 806 patients (5.4%) in the LMWH group vs a total of 33 of 806 patients (4%) in the arm allocated to OAT had an episode of recurrent symptomatic VTE. Combined analysis showed a not significant trend (p = 0.27) in lowering recurrent symptomatic VTE in favor of OAT (RR, 1.29; 95% confidence interval [CI], 0.82 to 2.026). The test for heterogeneity was not significant (p = 0.53) [Table 2
; Fig 2
].
|
|
Recurrent VTE During Active Treatment
Combining the 11 studies, a total of 78 of the 1,526 patients (5.1%) in the LMWH group had recurrent symptomatic VTE vs 118 of the 1,381 patients (8.5%) in the OAT group, showing a statistically significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH treatment (RR, 0.63; 95% CI, 0.47 to 0.83; p = 0.001). The test for heterogeneity was not significant (p = 0.38) [Table 2; Fig 3
].
|
|
In order to ascertain the presence of publication bias, we performed a funnel plot analysis. Each plot concerned those studies selected for assessing recurrent VTE during follow-up after cessation of active treatment (Fig 5 ), recurrent VTE during active treatment, recurrent VTE during active treatment in patients with cancer, and recurrent VTE during active treatment in patients with and without cancer. These funnel plots did not reveal evidence of important asymmetry.
|
It has been previously reported21 that patients treated with long-term LMWH have recurrent events earlier after cessation of therapy compared with warfarin, even though persistent excessive VTE events (true rebound) do not occur. Our results suggest that the natural posttreatment course of recurrent VTE could differ for patients treated with long-term LMWH or OAT. We propose that the incidence of recurrent VTE while receiving treatment is lower in patients receiving LMWH, but this advantage is lost after treatment discontinuation. Our findings showed a nonsignificant (p = 0.27) excess of recurrent symptomatic VTE outcomes during follow-up in the LMWH group compared with OAT (Table 2; Fig 2), despite the statistically significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH over OAT (p = 0.001) during active anticoagulant therapy (Table 2; Fig 3). In the results concerning VTE after therapy, the CI varied from a 20% reduction to a twofold increase (0.82 to 2.02); this means absence of evident excess of VTE recurrence with LMWH but not evidence of increased VTE recurrence with LMWH.
The presented results could be in part due to the different proportions of patients in the two groups who prolonged treatment beyond 3 months, without a standardization of treatment duration after the initial 3 months (Table 1). In some studies,811121314 treatment was prolonged for additional 3 to 9 months in an balanced/unbalanced fashion (see "Materials and Methods"). In a previous metaanalysis,22 it was not possible to perform an analysis of recurrence VTE after treatment withdrawal because of incompleteness of data. In our metaanalysis, based on abstracted data, patients receiving prolonged LMWH or OAT were not excluded from the analysis and were not considered starting from the effective date of treatment completion. A specific comment on the type of initial treatment in the OAT arm is needed as well. In a previous study by Hull et al,23 patients allocated to LMWH arm received OAT for additional 3 months, since an interim analysis showed higher recurrence rate in the LMWH group after the end of the study treatment period (as reported by Iorio et al22). In fact, patients treated with long-term LMWH had a significant excess of VTE recurrence (p = 0.002) after cessation of therapy when compared with those treated with LMWH as initial therapy, overlapped with warfarin continued to 84 days (home treatment protocol, 479 patients).21 Conversely, this excess of VTE recurrence in patients treated with long-term LMWH was attenuated in the comparison with patients treated with UFH as initial therapy, overlapped with warfarin for 84 days (initial in-hospital treatment protocol, 737 patients).21 Finally, by 1 year, VTE recurrence rate in LMWH patients showed no excess events, regardless of the type of initial therapy in the OAT arm.1521 As reported in other studies111314 enrolling patients with prolonged anticoagulation beyond 3 months (Table 1), those who received long-term LMWH did not experience, at the end of follow-up, a significant excess of VTE recurrence (Fig 2) in comparison with patients treated with LMWH as initial therapy, overlapped with warfarin. These results resembled those reported by Hull et al.21
Patients with malignancy treated with OAT have a higher rate of bleeding and an increased risk of recurrent VTE compared with patients without malignancy.2425 The observed statistically significant difference favoring LMWH over OAT among all patients receiving treatment comes mostly from studies7141517 that included cancer patients (Fig 4). In fact, excluding these studies enrolling only cancer patients, no significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH over OAT was observed (RR, 0.79; 95% CI, 0.52 to 1.21; p = 0.29). This fact, along with the observation that OAT control may have been suboptimal in some trials, means that clinicians should not yet conclude that LMWH is more effective than VKA in the treatment of noncancer patients with VTE. Conversely, in cancer patients, secondary prophylaxis with LMWH may be a more effective and practical alternative to OAT, considering also that VTE onset can increase comorbidities and worsen their quality of life. Concerning the biological effect of LMWH in cancer, Lee et al26 have hypothesized that, in patients with solid tumors without metastatic disease at the time of their thromboembolic event, impairing the establishment of tumor-related vasculature by a putative antiangiogenic agent, such as LMWH, could exert an inhibitory effect on tumor growth even beyond the time of drug exposure. The statistically significant improvement associated with dalteparin over OAT was demonstrated only in overall survival but not in cancer-related survival.27
Even though this meta-analysis addresses a still open question, it suffers the limited quality of the original data entered. During treatment, it is possible that LMWH provides an advantage, since the use of fixed doses guarantee a stable anticoagulation, while the maintenance of the therapeutic range with oral anticoagulants may be difficult. The quality of OAT was not homogeneously reported in the studies of this metaanalysis: poor INR control is a risk factor for recurrent VTE. Only 7 of the 11 trials analyzed actually assessed the event rates in a follow-up period after treatment cessation, and only a low percentage of patients remained on the therapeutic range while receiving OAT in the majority of the studies (< 50% in 4 studies, < 70% in 3 studies, not known in 4 studies). A single study,15 published only as an abstract, contributed approximately 40% of the patients to the analysis, and the rate of patients at target INR in this study remains unknown. Thus, the derived data could be affected by this variable. Unfortunately, newer studies71617 included in this metaanalysis, which were not included in previous such analyses,322 collected no information on recurrent VTE events during follow-up and, thus, did not help addressing the primary hypothesis of this metaanalysis.
By contrast, the possibility that the LMWH doses used in some trials for the long-term treatment of symptomatic VTE were too low should be taken into account as well. Across the studies, LMWH ranged from a prophylactic dosage (3,500 to 5000 UI qd89101216) to a full therapeutic dosage (150 to 200 UI/kg qd7141517 or divided in two doses,13 with one of the studies11 using an intermediate dose). A nonsignificant trend in favor of OAT for the prevention of recurrent symptomatic VTE used was observed in three8912 of the five studies employing low doses of LMWH. The metaregression analysis22 of seven studies showed a nonsignificant inverse relationship between the daily dose of LMWH used and the OR for VTE recurrences, indicating that a prophylactic LMWH dosage may be too low. Moreover, it might be highlighted that not all LMWHs are the same, since there are differences with regard to their biology, pharmacology, efficacy, and safety. In fact, there is variability in the anti-factor Xa to anti-factor IIa ratios among LMWHs and in their half-lives.2829 However, based on the pooled results coming from a metaanalysis by Dolovich et al,30 in which five different LMWH products were administered across 13 included studies, no major differences among LMWH preparations have been reported. Nevertheless, this conclusion is limited and must be tempered by the fact that the comparison of LMWH products is indirect; until studies that directly compare different LMWH preparations for the treatment of VTE are performed, it is impossible to make definitive conclusions about their relative safety and efficacy.
In addition, some methodologic differences between our study and the two previous metaanalyses322 on the same topic deserve some comments. Differently from the Cochrane metaanalysis3 but similarly to the metaanalysis by Iorio et al,22 we did not perform a quality scoring of the trials. In our study and in the study by Heijden et al,3 all analyses were according to the ITT principle. By contrast, in the study by Iorio et al,22 the ITT analysis was not applied to three studies,121314 which excluded some patients from the analysis. We addressed some issues not covered by the two aforementioned metaanalyses: the influence of LMWH treatment on recurrent VTE in cancer patients, the cancer-related mortality in cancer patients, and the overall mortality in mainly patients without cancer. Finally, these three metaanalyses reciprocally shares some important limitations: the studies included are not blinded by treatment, and the quality of OAT is poorly reported in some studies, so that the results may not apply to patients in whom an optimal management of OAT is achieved. In conclusion, despite the significant reduction of the risk of recurrent symptomatic VTE in favor of LMWH over OAT during treatment, patients treated with LMWH do not seem to have more frequently recurrent VTE events compared with OAT after cessation of therapy.
Acknowledgements
We thank Professor Holger Schunemann (Clinical Research Development and Information Translation, Unit Department of Epidemiology, Italian National Cancer Institute Regina Elena, Rome, Italy) for participating in the revision of this manuscript.
Footnotes
Abbreviations: CI = confidence interval; DVT = deep vein thrombosis; INR = international normalized ratio; ITT = intention to treat; LMWH = low-molecular-weight heparin; OAT = oral anticoagulant therapy; PE = pulmonary embolism; RR = relative risk; UFH = unfractionated heparin; VKA = vitamin K antagonists; VTE = venous thromboembolism
The authors have no conflicts of interest to disclose.
Received for publication February 16, 2006. Accepted for publication June 26, 2006.
References
This article has been cited by other articles:
|
|
 |
|
  J. F. Collen, J. L. Jackson, A. F. Shorr, and L. K. Moores Prevention of Venous Thromboembolism in Neurosurgery: A Metaanalysis Chest, August 1, 2008; 134(2): 237 - 249. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
