Relation between quality of anticoagulant treatment and the development of the postthrombotic syndrome
Carlo J. J. van Dongen, Department of Clinical Epidemiology and Biostatistics, Room J2-204, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands.
Tel.: +31 20 566 8975; fax: +31 20 691 2683; e-mail: firstname.lastname@example.org
Summary. Background: About 30% of patients with an episode of adequately treated deep venous thrombosis (DVT) develop the postthrombotic syndrome (PTS) within 2 years. During treatment with vitamin K antagonists (VKA) patients spend only 60% of time between an International Normalized Ratio (INR) of 2.0 and 3.0. We hypothesized that patients who spend a large amount of their time beneath this range will have an increased risk of the PTS. Objective: To investigate the relation between the quality of anticoagulant therapy with VKA and the risk of the development of the PTS. Methods: The time spent beneath the therapeutic range was calculated for patients with a first episode of DVT, who were treated with VKA for at least 3 months. At follow-up assessments for a maximum of 5 years, presence and severity of signs and symptoms of PTS were recorded. Results: A total of 244 patients, with a median duration of follow-up of 4.9 years were included for analysis. Of these, 81 patients (33%) developed the PTS. The multivariate model showed that patients who spend more than 50% of their time beneath an INR level of 2.0 are at higher risk for PTS [odds ratio (OR): 2.71, 95% CI: 1.44–5.10]. Conclusions: Low quality treatment with VKA, which is a common condition, is related to the occurrence of the PTS in patients with DVT. Strategies aimed at improving the quality of long-term anticoagulation might have the potential to reduce the incidence of this complication.
Following a first episode of deep vein thrombosis (DVT) of the lower extremities, almost 20% of the patients suffer a recurrent thromboembolic event and 20–50% develop the postthrombotic syndrome (PTS) within 2 years [1,2]. The PTS is a chronic complication and patients suffer of various symptoms of which the most prevalent are pain, swelling, pruritus and cramp of the affected leg. In severe cases even ulceration of the leg may result . In contrast with the extensive evaluation of risk factors of recurrent venous thromboembolism, little is known about predictors of the development of the PTS.
Although the pathophysiology of the PTS is controversial, persistent venous obstruction, valve damage, and an impaired microcirculation in the veins are thought to play a role [3,4]. Inadequate dissolution of the thrombus might lead to increased venous obstruction and valve damage. The usual treatment of DVT is anticoagulation, which is started with unfractionated or low-molecular-weight heparin followed by vitamin K antagonists (VKA). In several studies it has been shown that patients, even in the setting of randomized clinical trials, spend on average only 60% of their time in the therapeutic range during treatment with VKA [5,6]. This suboptimal quality might be related to inadequate clot dissolution. Therefore, we hypothesized that patients who spend a large amount of their time beneath the therapeutic range while on oral anticoagulation, will have an increased risk for the development of the PTS. This hypothesis was tested in a cohort of patients with an episode of DVT who were treated with VKA for 3 months and were followed for a maximum of 5 years.
Consecutive out-patients with an objectively confirmed symptomatic episode of proximal DVT presenting at the department of Internal Medicine of the University Hospital of Padua (Italy) between March 1993 and January 1998 were potentially eligible if they met the following criteria: treatment with unfractionated or low-molecular-weight heparin for a minimum of 5 days; treatment with VKA which was started either directly or the next day and continued for at least 3 months with a target International Normalized Ratio (INR) range between 2.0 and 3.0; INR monitoring in the University Hospital of Padua and no history of ipsilateral DVT. Patients were excluded from the analysis if the follow-up period was not long enough (< 6 months) to record the development of postthrombotic manifestations or if they had refused to give informed consent. Parts of our study population were originally included in two randomized clinical trials to the initial treatment of venous thromboembolism, which have been published before [5,6].
All patients were instructed to wear below-knee compression stockings (30–40 mmHg at the ankle) for at least 2 years. They were seen 3 and 6 months after the initial referral and thereafter returned to the hospital every 6 months for follow-up assessments, or earlier if complications occurred between two visits. The maximum duration of follow-up was 5 years. If patients were suspected of a recurrent episode of DVT this was confirmed or rejected using an objective diagnostic work-up in order to distinguish recurrences from the PTS.
Definition of postthrombotic syndrome
At each of the follow-up visits, the presence and severity of postthrombotic signs and symptoms in the ipsilateral leg were scored by physicians using a validated scoring system. This score has been shown to have a good reproducibility, and correlates well with the patient's perception of the interference of their leg complaints with daily life . For this score five subjective symptoms (pain, cramps, heaviness, pruritus, paresthesia) and six objective signs (induration of the skin, edema, hyperpigmentation, redness, pain during calf compression, new venous ectasia) were considered. For each item a score was given ranging from 0 (not present or minimal) to 3 (severe). In addition, the presence or absence of ulceration of the leg was assessed. A total score of 15 or more per visit on two consecutive visits or the presence of a venous ulcer indicated a severe PTS, while a total score of 5–14 per visit on two consecutive visits indicated a mild PTS. In all patients in whom the PTS was diagnosed recurrent DVT was ruled out as the cause of the complaints.
Estimation of time spent in the therapeutic range
The percentage time spent in the therapeutic INR range (INR 2.0–3.0) was calculated using linear interpolation [8,9]. This method assumes that INR values between two consecutive measurements vary linearly from the first to the second measurement. If the number of days between two consecutive measurements exceeded 28, the INR was considered not predictable for the middle part of this interval. For example, if the time between two consecutive INR measurements was 32 days, the INR level in the 4 days in the middle of this interval was not interpolated and considered to be missing. If the number of days spent in the target INR range could not be calculated at all, patients were excluded from analysis.
The relationship between the occurrence of the PTS and the percentage time spent in the lowest INR category (INR < 2.0) was explored using logistic regression analysis. Using multivariate models, adjustments were made for those variables that were statistically significant in the univariate analyses (P-value < 0.10) and those that were supposed to be clinically significant. Statistical analyses were computed with SAS version 8.02 (SAS Institute Inc, Cary, NC, USA).
A total of 300 patients was found eligible for inclusion. Of these, 56 were excluded because the follow-up period was shorter than 6 months. Characteristics of the remaining 244 patients are shown in Table 1. The median duration of follow-up was 4.9 years ranging from 6 months to 5.2 years.
Table 1. Patients characteristics (n = 244)
|Mean age at diagnosis DVT ± SD||62.4 ± 16.3|
|Age > 65 years (%)||126 (52)|
|Male (%)||106 (43)|
|Mean BMI ± SD||26.0 ± 4.1|
|Presence of cancer (%)||44 (18)|
|Ipsilateral recurrence* (%)||16 (7)|
|Mean duration of VKA therapy in days ± SD||87 ± 12|
|Mean percentage time in target range ± SD||58.9 ± 29.2|
Percentage time in the therapeutic range
Overall, approximately 60% of time was spent in the therapeutic range and 30% was spent beneath this range. A total of 69 of the 244 patients (28%) spent more than 50% of their time in the lowest INR category.
Of the 244 patients, 81 patients (33%) developed any PTS. Of these, eight were severe including five who had an ulcer. Of those patients who developed a PTS, 43 already had complaints at the follow-up visit at 6 months, 15 patients between 6 months and 1 year, 14 patients between 1 and 2 years and the remaining nine patients between 2 and 4 years. No further PTS was appreciated after the completion of 4 years of follow-up in any of the remaining patients.
Predictors of the postthrombotic syndrome
The results of the univariate and multivariate analyses are depicted in Table 2. In the univariate analysis age (> 65 years vs. ≤ 65 years), gender, ipsilateral recurrence, body mass index (BMI) (> 25 vs. ≤ 25), and time spent in the lowest INR range (> 50% vs. ≤ 50%) showed statistical significant association with the development of the PTS. Also cancer was explored univariately but was not significantly associated with the development of the PTS. When introduced as continuous variables, time spent in the lowest INR range, BMI and age were statistically significant related to the occurrence of the PTS. Correspondingly, the odds ratios (OR) remained statistically significant, independent of the cut-off level for the percentage of time spent in the lowest INR range, BMI and for age. The multivariate analysis revealed that, adjusted for all other variables, the occurrence of an ipsilateral recurrence (OR: 9.57, 95% CI: 2.64–34.7; BMI: OR: 1.14, 95% CI: 1.06–1.23), more than 50% of the time spent beneath an INR level of 2.0 (OR: 2.71, 95% CI: 1.44–5.10) and age above 65 years (OR: 2.56, 95% CI: 1.39–4.71) were related to the occurrence of the PTS. When patients with recurrent DVT were excluded from the multivariate analyses, the OR for patients who spent more than 50% of time beneath the INR range, adjusted for age above 65, BMI, gender and cancer was 2.60 (95% CI: 1.34–4.93).
Table 2. Estimated risk for development of the postthrombotic syndrome (univariate and multivariate analysis)
|Age > 65 year||2.33||1.34–4.05||2.56||1.39–4.71|
|Female vs. male||1.74||1.00–3.01||1.65||0.90–3.02|
|Cancer vs. no cancer||1.05||0.53–2.09||1.17||0.58–2.69|
|Percentage of time in lowest INR range*||2.42||1.36–4.32||2.71||1.44–5.10|
The OR for different cut-offs in percentage of time of an INR below 2.0 are shown in Table 3. There was a clear tendency for a higher risk of the PTS when the cut-off level for time spent in the lowest INR range was increased.
Table 3. Risk of occurrence of PTS for different cut-offs in percentage of time spent in the lowest INR range
|> 30% vs. ≤ 30%||1.89||1.10–3.23|
|> 40% vs. ≤ 40%||2.12||1.22–3.67|
|> 50% vs. ≤ 50%||2.43||1.36–4.32|
|> 60% vs. ≤ 60%||2.55||1.38–4.70|
|> 70% vs. ≤ 70%||3.01||1.46–6.20|
|> 80% vs. ≤ 80%||2.89||1.25–6.69|
|> 90% vs. ≤ 90%||3.69||1.29–10.53|
In this cohort study undertreatment with VKA, ipsilateral recurrence, high BMI and old age were found to be significantly related to the development of postthrombotic manifestations in patients suffering a first episode of DVT. Although ipsilateral recurrence gave a higher risk for the PTS than undertreatment with VKA, this complication was only present in 7% of the patients, while undertreatment with VKA was much more frequent, occurring in almost one third of patients. A comparable significant relationship between poor therapeutic quality and the occurrence of the PTS was observed when patients with recurrent DVT were excluded from the analysis.
Oral anticoagulants act by decreasing the blood tendency to coagulate. Although adequate anticoagulation with VKA cannot dissolute clots by itself, it can facilitate endogenous thrombolysis by preventing further thrombus growth, which might lead to improved recanalization and better preservation of venous valves. When intensity of treatment with VKA is beneath the target range for a large period of time, the potential for this beneficial effect is reduced. The increased incidence of PTS in patients with low treatment intensity indicates that such mechanisms might play a role in the pathophysiology of PTS.
Our results are consistent with those previously obtained in cohorts of patients with DVT who had a long-term observation [1,10]. In both studies recurrent ipsilateral DVT was significantly related to the occurrence of the PTS. The latter study also reported on the quality of anticoagulation and the risk for the PTS. Also in that study an increased risk of the PTS was recorded in patients who had not received strict anticoagulation. As we could carefully quantify the duration of undertreatment and calculate the associated risk for the PTS, the results of our investigation fully support the hypothesis that a low anticoagulation quality is a predictor of late postthrombotic sequelae. We also observed a positive association between BMI and the occurrence of the PTS, which has been previously described by others [11,12].
Some methodological issues about our study require comment. As the underlying hypothesis was conceived in retrospect, follow-up data of 56 patients could not be retrieved and therefore these patients were excluded. A very poor condition, as well as feeling very good, are plausible explanations for why patients did not return to the hospital. Unfortunately, we can only speculate about the reasons for these patients not to show up at follow-up visits. However, the incidence of the PTS in patients with an episode of venous thromboembolism observed in our study (33%) is comparable with the incidence observed in other studies [1,2]. Therefore, it is not likely that not enrolling these patients has influenced the observed relationship between quality of anticoagulation and the development op the PTS.
Also, the issue of compliance needs to be addressed. The higher risk of the occurrence of the PTS can also be because of a temporary underuse of elastic stockings. Wearing these stockings reduces the risk of the PTS by about 50%. It could be hypothesized that patients who did not wear the compression stockings were also the patients who did not take the medication very regularly, which might result in a poor quality of anticoagulation. However, it is unlikely that those patients who did not wear the compression stockings very regularly were the same patients who inadequately ingested their anticoagulant medication as mechanisms behind the lack of compliance for stockings and drugs are likely to be quite different. While the incorrect intake of tablets is largely related to forgetfulness, wearing compression stockings can be a real hassle in daily life because of discomfort, warmth, transpiration and cosmetics. In addition, recent data show that poor compliance is not associated with the quality of oral anticoagulant therapy . Other factors, such as co-medication are probably more likely to influence therapeutic quality. Underuse of elastic stockings is not related to a poor quality of treatment and therefore it is not likely that it does influence the relationship between quality of therapy and the development of the PTS.
What do our results imply? As improving the quality of oral anticoagulant treatment has the potential to prevent the occurrence of the PTS, the anticoagulation level should be monitored more frequently and carefully as currently done. In conclusion both patients and their attending physicians should be alerted about the increased risk of late postthrombotic manifestations following a low-quality treatment with VKA after an episode of DVT.