Prospective 12-Year Follow-up Study of Clinical and Hemodynamic Sequelae After Deep Vein Thrombosis in Low-Risk Patients (Zürich Study)
Background No prospective study of the long-term sequelae of more than 10 years after acute deep vein thrombosis exists so far. Therefore, 58 low-risk patients with DVT were included in a prospective study to evaluate the natural history of postthrombotic syndrome.
Methods and Results Clinical and hemodynamic examinations were performed at the time of admission; after 3, 6, and 12 months; after the 2nd, 3rd, 4th, 5th years; and finally after the 12th year. All patients received heparin initially and oral anticoagulants subsequently. After 12 years, 64% of the patients exhibited normal findings. Mild skin changes were found in 28%, marked trophic changes in 5%, and only one venous ulcer occurred. Regular use of compression stockings was reported by 54% of the patients with multilevel thrombosis. Although mean maximum venous outflow was significantly reduced from the acute event to 12 years later (P<.003) compared with the contralateral leg, a significant (P<.05) improvement was observed 6 months later. Recanalization of calf vein thrombosis was detected by Doppler sonography after 3 months. Sixty-four percent of the multilevel thromboses were recanalized completely or in part after 1 year; in 69%, valvular incompetence was found.
Conclusions In contrast to earlier reports, this prospective study up to 12 years after deep vein thrombosis demonstrates a low incidence of postthrombotic syndrome by administration of oral anticoagulants and regular compression therapy. However, the adverse clinical event rate (mortality 14%) and a recurrence rate of 24% show that the prognosis after deep vein thrombosis does not appear favorable even in low-risk patients.
Despite the great impact of postthrombotic syndrome on patient welfare and on healthcare expenditure, no long-term prospective follow-up lasting more than 7 years, which analyzes hemodynamic changes and the incidence of chronic venous insufficiency after conventional anticoagulant treatment of deep vein thrombosis (DVT), is reported in the literature. A few prospective studies1 2 3 4 provide follow-up data for 3 years and others for 6 and 7 years.5 6 7 Comparison between retrospective studies after conservative, thrombolytic, or surgical therapy8 9 10 cannot be made because of different study designs and treatment modalities. Moreover, an accurate description of the initial clinical venous status is missing in some studies.8 9 10 11 12 13 14 15 16
In this prospective study, 58 patients of a low-risk group with acute DVT confirmed by phlebography, who all received conservative treatment with heparin and coumarins, were followed for 12 years. Investigations comprised a thorough clinical examination; plethysmographic evaluation of arterial inflow, venous volume, and maximal venous outflow; and conventional Doppler sonography for detecting venous obstructions and valvular insufficiency. The clinical fate of the patients and the incidence of recurrent thrombotic episodes were analyzed. Compliance with compression therapy was evaluated.
Fifty-eight patients (14 women, 44 men) with acute DVT all verified by ascending contrast phlebography were included in the study from 1979 to 1982. The mean age of the patients was 50.5±14.4 years (25 to 78 years). The patients were referred to the outpatient clinic of the Angiology Division, Department of Medicine, University Hospital Zürich, by their external physicians. The duration of symptoms did not exceed 5 days, which placed these patients at low risk for adverse clinical events.
Patients were only enrolled if there was no history of preceding DVT and pulmonary embolism (PE), thrombophilia (protein C, protein S, and antithrombin III deficiency), or evidence of preexisting skin changes caused by chronic venous insufficiency. Further exclusion criteria were cardiorespiratory and rheumatic diseases requiring extensive medical therapy, known malignant tumors, pregnancy, and other contraindications for oral anticoagulation therapy (severe arterial hypertension, severe hepatic failure, recent peptic ulcer disease, and so forth). The fact that the patients had no previous DVT, PE, and hypercoagulable state determined them as a low-risk population.
Site of Thrombosis
The localization and extent of thrombotic changes are listed in the Table⇓. The lower limb veins were divided into three segments: the calf veins; the thigh veins, including popliteal and superficial femoral veins; and the pelvic veins, including the common femoral and iliac veins. Delineation of the pelvic veins by ascending phlebography permitted a sufficient judgment in all cases. Seventy-seven percent of the patients had combined thromboses at two or three levels. In 23% only one level was involved.
All patients received an initial intravenous bolus dose of 5000 IU unfractionated heparin followed by a continuous intravenous heparin infusion for 10 days. The heparin dose was adjusted according to the results of the activated partial thromboplastin time (1.5 to 2.5 times the normal control value of 30 seconds). During the first 5 days, the patients had bed rest with elevation of the affected limb. The heparin therapy was followed by initially overlapping oral anticoagulation with coumarins for 6 months in patients with calf vein thrombosis and for 12 months in proximal and multilevel thrombosis. Oral anticoagulation was started on the second day after admission. The prothrombin time was monitored by the Quick-test and adjusted to 20% to 25% (INR, 3.0 to 3.5).
After the acute leg swelling had subsided, the patient received a thigh or calf compression stocking of compression class III (Sigvaris, Ganzoni & Cie), which results in a compression pressure of approximately 40 mm Hg. The stockings were adapted for each patient by an experienced professional fitter. At each visit throughout the study period, the patients were instructed on the importance of daily compression therapy. When patients with proximal thrombosis could not tolerate thigh stockings, calf stockings were prescribed instead.
The compliance with compression therapy was classified as regular when the patients stated that they used the graded elastic stockings on the average of at least 5 days per week during the daytime. The treatment was considered “not regular” or sporadic if the stockings were applied for less than 5 days.
All patients had a clinical workup at the time of admission during the acute stage, 3 months, 6 months, and 1 year later. Follow-up visits then were performed at yearly intervals up to the 5th year, with a last visit at the 12th year. The mean follow-up time was 11.6 years (11.0 to 12.4 years). The physical examination was performed throughout the follow-up period by the same four physicians (A.B., U.K.F., K.J., and E.S.). The clinical inspection focused on signs of chronic venous insufficiency (CVI) as manifested by swelling, hyperpigmentation, trophic changes of the skin, lipodermatosclerosis, and distended superficial venous collaterals.
CVI was classified in three stages17 18 : CVI stage I, mild to moderate lower limb edema and mild skin changes; CVI stage II, marked trophic changes with hyperpigmentation in the gaiter area, subcutaneous fibrosis, eczematoid changes, and/or edema; and CVI stage III, acute venous ulceration or healed venous ulcer.
Ascending contrast phlebography was performed in all patients only at study entry and in 3 patients with recurrent thrombosis during the study.
In patients who died during the 12-year period, the cause of death was obtained by consultation with the relevant family physician.
Doppler Ultrasound Examination
Doppler ultrasound examinations19 20 were carried out at all visits by the same four physicians. The sites of recordings were the femoral, popliteal, posterior tibial, and greater saphenous veins bilaterally. A hand-held, bidirectional, continuous-wave Doppler with 8.2- and 5.3-MHz probes (Parks Electronic Lab) was used. The patients were examined in the supine position after at least 10 minutes of rest. The popliteal veins were studied with the patient in the prone position with the knee slightly flexed and the feet supported by a pillow. The criteria used for the Doppler evaluation19 20 were the following: Normal finding was defined as physiological respiratory modulation of spontaneous Doppler signals with increase of the venous flow velocity during expiration and cessation during inspiration. Venous obstruction was diagnosed in the absence of a spontaneous Doppler signal and when the augmentation signal after distal compression was absent or diminished. The criterion for collateral veins was a continuous, high-frequency spontaneous signal. When the Doppler signal was modulated by deep breathing only, the augmentation diminished, and in addition there was new reflux compared with the previous examination, partial recanalization was diagnosed. The criteria for valvular incompetence was reflux during proximal compression and/or Valsalva maneuver.
Venous occlusion (strain gauge) plethysmography (Periquant, Gutmann) was performed throughout follow-up by the same vascular technician (I.S.). Measurements were made at the calf level with the cuffs around the mid thigh. The measurements were performed after 15 minutes with the patients in the supine position and elevation of the legs (30 cm above heart level). Arterial inflow (mL/100 mL per minute), venous volume (mL/100 mL), and maximum venous outflow (mL/100 mL per minute) were measured after 4 minutes of venous occlusion with a cuff pressure of 60 mm Hg.21 22
Data are expressed as mean values with standard deviations. Statistical analysis was performed using repeated-measures ANOVA (Scheffé’s F test) for serial data, and the paired two-tailed Student’s t test or the χ2 test was applied when appropriate. Statistics were performed (J.G.) at the Division of Informatics, Department of Medicine, University Hospital Zürich, and at the Computer Center of the University of Zürich. Differences were considered to be statistically significant at P<.05.
Oral informed consent was obtained from all patients. The study was approved by the Ethical Committee of the Medical Polyclinic, Department of Medicine, University Hospital Zürich.
Dropout and Mortality Rate
Twelve years after enrollment, 7 of 58 patients (12%) were lost because of migration (n=3) or unwillingness to participate further in the study (n=4). The times of dropout were in 3 patients the first month and in 4 patients in months 4, 12, 13, and 38 after the acute event. Four patients did not participate further in the study because of new concomitant diseases (peripheral arterial disease in 1 patient, diabetes in 1 patient, Parkinson’s disease in 1 patient, and senility in 1 patient). The time of these dropouts was between years 5 and 12. Four patients were not included in the final analysis because they did not regularly attend follow-up visits. However, at the 12-year control, they had no postthrombotic syndrome.
The mortality rate was 14% (8 of 58 patients). The causes of death were vascular events in 5 patients (1 myocardial infarction, 1 pulmonary edema, 2 PEs, 1 stroke) and nonvascular events in 3 patients (1 bronchial small cell carcinoma, 1 terminal renal failure, 1 murder). The times of death varied between 21 and 89 months after study entry.
A complete follow-up over the entire 12 years could be obtained in 39 patients (8 women, 31 men) with 13 calf and 26 multilevel and/or proximal thromboses.
During the study period, recurrent thrombosis developed in 24% (14 of 58 patients; 2 women, 12 men). Diagnosis was confirmed in 3 patients by ascending phlebography and in 11 patients by Doppler ultrasound and strain gauge plethysmography. In 4 patients the ipsilateral leg and in 10 patients the contralateral leg was affected. The recurrent thrombosis occurred 6 to 58 months after study entry and in all patients after discontinuation of coumarin treatment. The 2 patients who died of PE were included in this group with recurrent DVT.
Twelve years after the acute event, normal clinical findings of the lower limbs were present in 64% of the 39 patients; mild CVI of stage I with only discrete edema and signs of venous congestion at the foot and no trophic lesions was found in 28% of these patients. Moderate CVI stage II characterized by marked peripheral venous congestion with trophic changes such as edema, hyperpigmentation, and lipodermatosclerosis but without venous ulceration was present in 5% of all patients available at the 12-year follow-up examination. Severe CVI stage III with venous ulceration was found in only 1 of the 39 patients (3%) 12 years after the acute thrombosis. All the skin changes developed between the second and fourth years, and thereafter no significant alterations in clinical status occurred.
In Fig 1⇓, the incidence of CVI at the 12th year in 13 patients with calf vein thrombosis and 26 patients with multilevel or proximal DVT is shown. After isolated calf vein thrombosis, no trophic skin lesions were detected. Mild CVI developed in only 2 patients (1 compliant with compression therapy and 1 noncompliant). Among the 26 patients with multilevel or proximal thrombosis, no postthrombotic changes were found in 54%. Thirty-four percent had only minor skin changes with edema and 8% had marked postthrombotic changes. Only 1 patient developed a venous ulcer after the 4-year visit.
After 12 years, only 23% of patients with isolated calf vein thromboses used regular compression therapy, whereas 77% wore the stockings only occasionally or never. However, in multilevel and proximal thromboses, the compliance was significantly better (P<.01). Regular compression therapy was performed in 54% and sporadic or no compression in 46% of the patients. Only 4 patients with multilevel and proximal thromboses tolerated thigh compression stockings; however, their compliance was excellent.
In patients with multilevel or proximal thrombotic disease, the effect of regular compression therapy was evident. All patients who used regular compression therapy had only mild or no postthrombotic symptoms after 12 years. No relevant trophic skin lesions were present in these highly compliant patients, whereas the only venous ulcer observed in this study developed in a patient who used compression stockings only occasionally. Two additional patients with poor compliance showed marked trophic changes of the lower leg skin.
The mean values and standard deviations of the hemodynamic data during the 12-year study period are shown in Fig 2⇓.
During the observation period, no significant differences of the mean arterial inflow were found between the diseased leg and the contralateral leg (Fig 2⇑). The ANOVA for repeated measures revealed no significant differences during follow-up.
Mean venous volume of legs with thrombosis was significantly reduced from the acute thrombotic event up to the third year of follow-up (P<.05). Venous volume showed a clear tendency to increase between the first and the second visit, but this did not reach statistical significance (P>.05). However, the increase in venous volume between visits 1 and 3 was significant (P<.05). At the last three visits, no significant differences of the thrombosed leg compared with the control leg could be detected (Fig 2⇑).
Maximal Venous Outflow
Mean maximum venous outflow was significantly decreased (P<.0001) compared with the contralateral leg throughout the study (Fig 2⇑). However, a significant (P<.05) improvement in venous outflow was observed 6 months after the acute DVT compared with the first visit. After the third examination 6 months after the acute event, maximum venous outflow did not change (P>.05) any more during the 12-year follow-up, demonstrating a remarkable stability.
Venous Doppler Examination
The results of the continuous-wave Doppler ultrasound examinations in patients with calf and multilevel thrombosis during the 12-year follow-up period are summarized in Figs 3⇓ and 4⇓. In patients with phlebographically proven thrombosis of the calf veins involving the posterior tibial veins, new evidence of reflux was detected by Doppler ultrasound as early as 3 months after the acute thrombotic event. Six months after calf vein thrombosis, 50% of the patients exhibited recanalization with valvular incompetence of the posterior tibial veins in 20% and partial recanalization in 17%. At the 12-year follow-up, 54% had normal Doppler findings and 46% valvular insufficiency. Partial or complete recanalization was diagnosed in 7% and 39%, respectively.
Sixty-four percent of the patients with proximal and multilevel thrombosis had partially or completely recanalized deep veins 1 year after the acute event. At the 5-year examination, 93% of multilevel thromboses had a normal Doppler examination or one suggesting partial or complete recanalization. The final examination after 12 years revealed normal findings, with competent valves in 27%. Only 4% had Doppler findings suggesting obstruction without reflux. Fifteen percent and 54%, respectively, had partially and completely recanalized deep veins with valvular incompetence.
In this observational study, all treatment decisions were always at the discretion of the treating physician. However, no discrepancies in treatment modalities between study protocol and actual therapy were found during the follow-up. Limitations also include the relatively small number of patients and the observational design, which limits conclusions regarding treatment outcome.
The patients studied form a selected group according to inclusion and exclusion criteria, respectively. The strict criteria were necessary to obtain a homogenous group of patients with DVT and to ensure a low dropout rate.
In the present study, the relatively low incidence of moderate and severe postthrombotic syndromes after acute lower limb DVT is striking. Only 1 of 39 patients of a low-risk population with a follow-up to 12 years after the acute DVT developed a venous ulcer. Marked trophic changes were present in only 5% and minor skin changes in 28%. Almost two thirds of the patients had no clinical signs of CVI at all. In contrast to the low incidence of severe postthrombotic skin changes, the adverse clinical event rate was relatively high, with two fatal PEs. Thus, the prognosis of DVT does not appear favorable, even in this low-risk group.
Detailed analysis shows that the incidence of postthrombotic changes after calf vein thrombosis is especially low. Eighty-five percent had normal findings at the end of the follow-up period and 15% only presented minor symptoms. In patients with multilevel thrombosis, severe CVI was found in 12%, including one ulcer, and almost 90% had only minor clinical signs or were normal without symptoms. The trophic skin alterations developed between the second and fourth years after the acute event. During the period from years 5 to 12, no additional changes were observed.
Most of the recent short- and middle-term studies1 2 3 4 5 6 support our findings that the incidence of the postthrombotic syndrome with venous ulceration occurring in only 4% of the patients with multilevel thrombosis exclusively is lower than initially suggested.23 The data from this study confirm the results of studies from the United States,3 6 Sweden,4 5 Norway,7 and the United Kingdom2 showing that over 50% of patients with calf and multilevel thromboses have only mild or no symptoms at 3 to 7 years after DVT.
These results suggest that calf vein thrombosis without involvement of the popliteal vein has an even more benign long-term clinical course of CVI, if adequate conservative therapy with immediate intravenous heparin followed by oral anticoagulation and compression therapy is performed. Minor postthrombotic changes develop in only 15%. Others2 4 5 8 25 26 27 have described similar findings with no venous ulceration developing after distal thrombosis. However, it must be emphasized that immediate administration of intravenous standard heparin or subcutaneous low-molecular heparin treatment is necessary to prevent further extension of the thrombus into the popliteal and femoral veins. In 1979, when this study was started, the duration of coumarin treatment for uncomplicated DVT lasted longer than nowadays, in which 3 months of oral anticoagulation after uncomplicated acute thrombosis appears to have an acceptable benefit-risk ratio.24 There are two ongoing studies in North America and at least one in Europe in which patients with DVT are being randomized to either 3 months or to 1 to 3 years of oral anticoagulation. These studies will show the optimal duration of oral anticoagulation.
The recurrence rate of lower limb thrombosis was 24% in this low-risk patient population. This indicates that a preceding thrombosis is an important risk factor for further thrombosis. However, rethrombosis not only concerned the initially affected side (4 of 14 recurrences) but was even more frequent on the contralateral limb (10 of 14). The present recurrence rate confirms the results of other prospective studies with shorter duration, in which rethrombosis occurred in between 6% and 28%4 5 30 31 ; in calf vein thrombosis, the recurrence rate is lowest with 6%.4 Rethrombosis occurred 6 to 58 months after study entry and in all patients after discontinuation of coumarin treatment.
This study was not designed to prove the efficacy of compression therapy in the prevention of the postthrombotic syndrome in a randomized fashion. During the follow-up visits, we always stressed the importance of wearing the compression stockings. However, comparing the outcome of the compliant and noncompliant patients, the positive effect of a regular compression therapy is obvious. Calf vein thrombosis was rarely associated with symptoms such as leg swelling or subjective complaints by the patients. Therefore, patients with distal thrombosis most likely had the worst compliance. Recently Büller et al28 presented preliminary data of the prospective, randomized Amsterdam study regarding the prevention of the postthrombotic syndrome by compression therapy with graded elastic stockings. After a 5-year follow-up, they concluded that the use of stockings could reduce the incidence of mild to moderate and severe postthrombotic changes by approximately 50%. These and our data demonstrate that compression therapy with elastic stockings presently represents the only proven effective prevention available for the postthrombotic syndrome. Although the mechanism of compression therapy is not yet completely understood, the reduction in transmural capillary pressure during walking appears to be the most important factor with respect to the nutritive function of the vascular system.29
Conventional continuous-wave Doppler ultrasound is a useful method to document venous obstruction and valvular incompetence in the deep and superficial venous system.19 32 However, we are aware of the difficulty that collateral veins in the vicinity of an occluded segment may be interpreted occasionally as a recanalized vein25 or that an incomplete thrombosed segment may suggest partial recanalization. On the other hand the Doppler examinations were performed exclusively by four experienced angiologists. Presently, duplex scanning is available, which was not the case 12 years ago and is of course superior to the hand-held venous Doppler.
The posterior tibial vein appears to be the most important of the calf veins regarding the development of the postthrombotic syndrome because of its role for drainage during calf muscle pumping and its communication with the posterior arch vein. In all patients with calf vein thrombosis, phlebography showed thrombosed posterior tibial veins and thrombosis of at least one of the other calf vein groups; isolated calf vein thrombosis was not present. Thromboses of the posterior tibial vein started to recanalize relatively fast with an approximately 35% reflux rate after 3 months. After 6 months, 50% were completely recanalized, with 20% valvular incompetence and partial recanalization occurring in 17%. After 12 years, 54% of the patients had normal Doppler findings and the rest had valvular incompetence. Our early results obtained with the continuous-wave Doppler are confirmed now by the very similar outcome of ultrasonic duplex scanning with a median follow-up time of 33 months.33 34
Recanalization of popliteal, superficial femoral, and iliac vein thromboses lasts longer than that of smaller calf veins. One year after the acute event, 64% of the proximal thromboses were recanalized completely or in part as suggested by the Doppler examination. This indicates an approximate 6-month delay compared with calf vein recanalization. The final visit after 12 years revealed normal Doppler findings without reflux in 27%; 15% were partially recanalized and 54% completely recanalized with reflux. This shows that even in proximal thrombosis, recanalization does not necessarily result in detectable valvular incompetence of the initially completely thrombosed segment.
The study was supported in part by a grant from the Swiss Phlebology Society. We are indebted to Prof J.E. Tooke, Exeter, UK, for revision of the manuscript.
- Received May 25, 1995.
- Revision received July 18, 1995.
- Accepted August 16, 1995.
- Copyright © 1996 by American Heart Association
Lagerstedt C, Olsson CG, Fagher B, Norgren L, Tengborn L. Recurrence and late sequelae after first-time deep vein thrombosis: relationship to initial signs. Phlebology. 1993;8:62-67.
Widmer LK, Zemp E, Widmer MT, Schmitt HE, Brandenberg E, Voelin R, Biland L, da Silva A, Maggs M. Late results in deep vein thrombosis of the lower extremity. Vasa J Vasc Dis. 1985;14:264-268.
Partsch H, Weidinger P, Mostbeck A, Olbert F, Denck H. Funktionelle Spätergebnisse nach Thrombektomie, Fibrinolyse und konservativer Therapie von Bein-Beckenvenenthrombosen. Vasa J Vasc Dis. 1980;9:53-61.
Partsch H, Leitner G, Mostbeck A. Klinisches Bild und venöse Funktion 5 Jahre nach isolierter Unterschenkelvenenthrombose. Phlebol Proctol. 1981;10:9-13.
Thiele C, Theiss W, Kurfürst-Seebauer R. Langzeitergebnisse nach fibrinolytischer Behandlung tiefer Venenthrombosen im Becken-Bein-Bereich. Vasa J Vasc Dis. 1989;18:48-55.
Browse NL, Clemenson G, Thomas ML. Is the postphlebotic leg always postphlebotic? Relation between phlebographic appearances of deep-vein thrombosis and late sequelae. Br Med J. 1980;281:1167-1170.
Widmer LK, Stähelin HB, Nissen C, da Silva A. Venen-, Arterienkrankheiten, koronare Herzkrankheit bei Berufstätigen: Prospektiv-epidemiologische Untersuchung: Basle Study I-III. Bern, Switzerland: Huber Verlag; 1981.
Barnes RW. Doppler ultrasonic diagnosis of venous disease. In: Bernstein EF, ed. Noninvasive Diagnostic Techniques in Vascular Disease. St Louis, Mo: CV Mosby Co; 1985:724-729.
Bollinger A. Was leistet die Doppler-Ultraschall-Technik in der Diagnose der tiefen Venenthrombose? Vasa J Vasc Dis. 1975;4:16-20.
Sumner DS. Mercury strain-gauge plethysmography. In: Bernstein EF, ed. Noninvasive Diagnostic Techniques in Vascular Disease. St Louis, Mo: CV Mosby Co; 1985:133-150.
Bauer G. Roentgenological and clinical study of the sequelae of thrombosis. Acta Chir Scand. 1942;86(suppl 74):1-110.
Strandness DE, Langlois Y, Cramer M, Randlett A, Thiele BL. Long-term sequelae of acute venous thrombosis. JAMA. 1983;250:1289-1292.
Büller HR, Brandjes DPM, ten Cate JW. The post-thrombotic syndrome prevention by graded elastic stockings. Proceedings of the 20th Congress of the International Union of Angiology, European Chapter, Beaune, France. October 6-8, 1993. Abstract.
Brakkee AJM, Kuiper JP. The influence of compressive stockings on the hemodynamics in the lower extremities. Phlebology. 1988;3:147-153.
Bollinger A, Mahler F, Sépibus de G. Diagnostik peripherer Venenerkrankungen mit Doppler-Strömungsdetektoren. Dtsch Med Wochenschr. 1968;46:2197-2201.