Citation
Gadisseur, A. P. A. (2006, June 21). Improving the quality of oral anticoagulant therapy.
Retrieved from https://hdl.handle.net/1887/4455
Version:
Corrected Publisher’s Version
License:
Licence agreement concerning inclusion of doctoral thesis in the
Institutional Repository of the University of Leiden
THERAPEU TIC Q U AL ITY CO N TRO L
O F O RAL AN TICO AG U L AN T THERAPY
CO M PARIN G THE S HO RT-ACTIN G
ACEN O CO U M ARO L AN D THE L O N G -ACTIN G PHEN PRO CO U M O N
Ala in P.A. G a d isse u r1,3, F e lix J.M . v a n d e r M e e r1,3, He n k J. Ad ria a n se n4, S te p h a n D . F ih n2, F rits R. Ro se n d a a l1, 2.
D e p a rtm e n t o f He m a to lo g y1 D e p a rtm e n t o f Clin ic a l Ep id e m io lo g y2
L e id e n U n iv e rsity M e d ic a l Ce n te r (L U M C), L e id e n , Th e N e th e rla n d s.
L e id e n An tic o a g u la tio n Clin ic3, L e id e n , Th e N e th e rla n d s.
Ro o se n d a a l An tic o a g u la tio n Clin ic 4 F ra n c isc u s Ho sp ita l, Ro o se n d a a l, Th e N e th e rla n d s.
Summary
In The Netherlands tw o coumarins are routinely used for oral anticoagulant therapy (OAT), the short-acting acenocoumarol (t 1/2 = 11 hrs), and the long-acting phenpro-coumon (t 1/2 = 140 hrs). To investigate if the different pharmacokinetics of these coumarins result in a different q uality of anti-coagulation w e studied patients from the Leiden anticoagulation clinic treated w ith acenocoumarol or phenprocoumon in 1998 -1999 w ith a treatment duration of more than 16 w eeks. Out of 1368 eligib le patients 228 pairs could b e closely matched for indication for anticoagulant therapy, age, sex and date of start of treatment. As a validation test a similar analysis w as carried out on 5 1 matched patient pairs at a second anticoagu-lation clinic.
45 6 patients w ith 7245 INR checks yield-ed 230 patient-years. The q uality of OAT cal-culated over the w hole treatment period w as higher in patients treated w ith phenpro-coumon as expressed b y numb er of INR checks in the therapeutic range (phenpro-coumon: 42.7% , acenocoumarol: 36.5 % , difference: 6.1% , CI95 of the difdifference: 3.0 -9.3% ) and b y time in range (phenprocoumon: 46.6% , acenocoumarol: 41.6% , difference: 5 .0% , CI95 of the difference: 1.3 – 8 .6% ). After the initial 6 w eeks of OAT, w hen a more sta-b le effect should have sta-b een reached, the dif-ferences b ecame more pronounced (differ-ence: 6.1% , CI95 : 1.8 – 10.4% ).
The incidence of severe b leeding compli-cations w as similar (phenprocoumon: 0.04/ patient/ year vs. acenocoumarol: 0.03/patient/year) w ith a slight excess of minor b leeds w ith phenprocoumon (0.19/ patient/ year vs. 0.06/ patient/ year). There w as 1 non-fatal pulmonary emb olism
in the phenprocoumon group (0.009/patient/year) against 2 fatal pulmonary emb oli in the acenocoumarol group (0.017/patient/year).
Introduction
Coumarins are widely used as oral antico-agulants. They inhibit the production by the liver of the vitamin K -dependent coagulation factors (Factor II, V II, IX and X ). A number of studies have suggested that longer-acting coumarins give a more stable anticoagulation than short-acting coumarins (Ensor et al, 1957; Simson et al, 1962 ; Rodman et al, 1964; Sullivan et al, 1965; Breed et al, 1969; Fekkes et al, 1971; Pattacini et al, 1994), which is thought to be associated with fluctuations in the plasma levels of Factor V II induced by the shorter-acting drugs (Loeliger et al, 1963; K az mier et al, 1965; Thijssen et al, 1988). Detailed comparisons of different coumarins are rare, which is probably due to the geo-graphic distribution of their use. A large and recent analysis compared acenocoumarol with warfarin (Pattacini et al, 1994). Those studies comparing phenprocoumon and acenocoumarol were performed 30 years ago and involved few patients (Breed et al, 1969; Fekkeset al, 1971).
In the Netherlands two coumarin deriva-tives are licensed for oral anticoagulation: acenocoumarol (Sintrom Mitis® ) which has a half-life of 11 hours (Hemker et al, 1985), and phenprocoumon (Marcoumar® ) which has a half-life of 140 hours (Hemker et al, 1985; Thijssen et al, 1988; K elly et al, 1979). Acenocoumarol is prescribed in about 75% of cases, whereas only 25% of the patients are treated with phenprocoumon. In The Netherlands the administration of oral antico-agulant therapy (OAT) is highly structured with a system of regional anticoagulation clin-ics (van den Besselaar et al, 1988; Loeliger et al, 1984). These anticoagulation clinics are responsible for the determination of the pro-thrombin time and INR values, the dosing of
the medication and the prescription of vita-min K in case of excessively high INR values or bleeding complications.
The most important side effect of OAT is the occurrence of bleeding complications, even at INR values within the therapeutic range. An increase of bleeding complications is seen with increasing INR values and in older age ranges (van der Meer et al, 1993; Palareti et al, 1996; Harder et al, 1996; Cleophas et al, 1993; Fihn et al, 1993; Fihn et al, 1996; Cannegieter et al, 1995; van der Meeret al, 1996). It is therefore important to restrict the number of INR values above the therapeutic range while ensuring a good ther-apeutic effect, i.e. to aim for constant INR val-ues within the therapeutic range.
Patients, materials
and methods
Study design
The Leiden Anticoagulation Clinic follows around 10,000 patients per year, with around 5,000 patients actively on OAT at any given time. The relevant administrative, clinical and laboratory data are held in computerised files. Around 75% of the patients are treated with phenprocoumon (Marcoumar®) and around 25% with acenocoumarol (Sintrom Mitis ®). The choice of anticoagulant drug is essential-ly made by the referring physician mainessential-ly based on his or her experience with one of the drugs in the past. Laboratory checks and consequently adjustments of the dosing schedules occur at intervals of 1-6 weeks. Patients on long-term OAT (estimated dura-tion of more than 16 weeks), using aceno-coumarol and starting treatment between 01-01-98 and 01-01-2000, were matched to patients using phenprocoumon for the indica-tion for the OAT, age, sex and date of start of treatment (in this order of priority). Through this matching process any bias through pref-erence for one of the drugs in certain indica-tions or at a certain patient age was ruled out. Out of 515 patients on acenocoumarol 228 could be matched absolutely for indication and age (+ /- 10 years) with patients on phen-procoumon. Out of these, 64/228 were also matched for sex and the starting date of the treatment (+ /- 100 days), 94/228 were matched for sex but not for the starting date of OAT, and 37/228 were matched for the starting date of OAT but not for sex. 33/228 were not matched for sex nor starting date of treatment (mean difference 257 days, CI95= 234-280 days). For the 228 patient pairs the
data concerning the monitoring of the OAT during the period 01-01-1998 to 01-04-2000 were analysed. The remainder of the patients on acenocoumarol could not be matched.
Because the Leiden Anticoagulation Clinic has a preponderance of patients on phenpro-coumon (75% phenprophenpro-coumon vs. 25% acenocoumarol) it is conceivable that the higher degree of experience with the former drug could bias the outcome of the compari-son between the two coumarins. In order to rule out this bias we conducted a similar analysis as a validation test at another Dutch anticoagulation clinic (Roosendaal) that has a nearly reverse ratio in the use of phenpro-coumon (20%), and acenocoumarol (80%).
Measurement
of the Prothrombin Time
Venous blood was collected in 3.2% sodi-um citrate and plasma was obtained by cen-trifugation at 2800 g for 10 minutes. The Prothrombin Time (PT) was measured in the plasma using RecombiPlastin® reagent (Ortho Diagnostic Systems, Rariton, NJ, USA) on an Elektra 1000 or 1800 automated coagu-lometer (MLA, Pleasantville, NY, USA). PT val-ues were expressed in INR according to inter-national convention.
Determination of OAT dosage
schedule
Determination of the dosage of the OAT (acenocoumarol and phenprocoumon) was done by 5 physicians routinely working in the Leiden Anticoagulation Clinic with the aid of a computerised dosing program (TRODIS, Infotrom, Leiden, The Netherlands). This pro-gram evaluates the stability of the PT/INR val-ues and proposes dosing schedules in about 50% of patients that are then checked by the physicians. Around 15-20% of these comput-er-generated proposals is changed by the physicians. In the other 50% of patients no dosage proposal can be generated and dosing is done completely by the physicians. Details of the dosing algorithm have been published previously (Wiegman et al, 1977).
Therapeutic INR target ranges are defined for all patients on OAT based on their indica-tion for the treatment. Two therapeutic ranges are routinely used depending on the indica-tion for the OAT: low level anticoagulaindica-tion at INR 2.5-3.5, and high level at INR 3.0-4.0. The same method of dosing and therapeutic target ranges are used by the Roosendaal Anticoagulation Clinic where dosing was done by 2 physicians.
Study Endpoints
The quality of the OAT was calculated in several ways:
1. INR checks in range: percentage of all INR values within the therapeutic target range. 2. Time in range: estimated time spent by the patient within the therapeutic target range based on the method of linear inter-polation. This method of approximation of the time in range has been published previously (Rosendaal et al, 1993; van der
Meer et al, 1996).
This analysis was done for the whole treatment course within the study period and also for the initial 6-7 weeks of OAT contrast-ed to the subsequent period of prolongcontrast-ed treatment.
Bleeding complications were classified as major if they were fatal, or necessitated hos-pitalisation, or blood transfusions, or if they concerned intra-articular, intra-muscular, retro-peritoneal, gastro-intestinal (overt melaena) or intra-cranial bleeding. Minor bleeding complications were defined as all other overt bleeding, with ecchymoses only being counted if more than 10cm diameter and epistaxis if the duration exceeded 30 minutes. Data concerning haemorrhagic or thrombo-embolic complications are routinely collected by the nurses at the anticoagulation clinic through direct observation of the patients, or through information relayed by the patients or treating physicians (oral or written communications). Causes of death are collected from medical correspondence. Cardiovascular deaths are not counted as thrombo-embolic complications but as a sep-arate category because of their multi-causal aspect.
Statistical analysis
Values for the percentage of INR checks within the target range and the time in range are given with 95% confidence intervals, the mean difference, and the 95% confidence interval of the difference.
Results
As the patient selection was done by a process of matching, the two patient groups (n=228) were comparable as to indication for OAT, age, sex, date of start of treatment, and duration of treatment (table 1). Overall the mean interval between INR measurements was similar (d = 0.07 weeks, CI95 = -0.01 – 0.14) for phenprocoumon (1.64 weeks) and acenocoumarol (1.57 weeks) but in the
patients in 3.0-4.0 INR target range there was a small difference (d = 0.23 weeks, CI95 = 0.05 – 0.42) where acenocoumarol patients were evidently checked more frequently.
Complete information to the end-of-fol-low-up was available for 422/456 patients (92.5%): most patients (399/456, 87.5%) com-pleted their treatment during the study, and 23/456 patients (5.0%) continued with the OAT past the end-of-follow-up. The data were incomplete in 34/456 patients (7.5%)
Ta b le 1 . c h a r a c te r is tic s o f th e p a tie n t g r o u p s
P h e n p r o c o u m o n A c e n o c o u m a r o l All patients
No. of patients 228 228 Male/Female 121/107 107/121 Mean age in years, (range) 63.8 (19-89) 64.2 (19-94) Mean follow-up in weeks, (SD) 25.1 (16.8) 27.3 (18.5) Mean interval between checks in weeks, (SD) 1.64 (0.46) 1.57 (0.38)
INR range 2.5-3.5
No. of patients 168 164 Male/Female 74/94 67/97 Mean age in years, (range) 63.8 (19-89) 64.2 (19-94) Mean follow-up in weeks, (SD) 21.0 (13.6) 23.6 (14.9) Mean interval between checks in weeks, (SD) 1.58 (0.37) 1.57 (0.39) (postoperative proph y lax is (n = 1 3 6 ) , atrial fib rillation (n = 6 0 ) , pu lm on ary em b olu s (n = 5 6 ) , d eep ven ou s th rom b osis (n = 5 0 ) , n on -su rg ic al proph y lax is (n = 2 4 ) , C A B G (n = 2 ) , c ereb rovasc u lar in su ffic ien c y (n = 2 ) , vasc u lar prosth etic d evic es (n = 2 ) )
INR range 3.0-4.0
No. of patients 60 64 Male/Female 47/13 40/24 Mean age in years, (range) 63.7 (29-87) 64.2 (25-85) Mean follow-up in weeks, (SD) 32.9 (21.3) 33.4 (24.5) Mean interval between checks in weeks, (SD) 1.80 (0.62) 1.56 (0.37) (vasc u lar prosth etic d evic es (n = 3 6 ) , C A B G (n = 3 0 ) , m y oc ard ial in farc tion (n = 3 0 ) , periph eral vasc u lar in su ffic ien c y (n = 1 0 ) , artific ial h eart valves (n = 8 ) , an g in a pec toris (n = 2 ) , arterial em b olism (n = 2 ) , c ereb ral em b olism (n = 2 ) ,
because the patients had moved out of the area. OAT had to be stopped prematurely in 7/456 patients (1.5%) because of complica-tions or an acquired contra-indication and 38/456 patients (8.3%) died during the treat-ment (19/228 in each of the treattreat-ment groups).
Analysed over the full follow-up period from the start of the OAT, 42.7% (CI95= 40.4 - 44.9%) of the INR checks in the phenpro-coumon group were within the therapeutic
range, compared to 36.5% (CI95= 34.3 -38.8%) in the acenocoumarol group (δ=6.1%, CI95= 3.0 - 9.3%). Over the same period the percentage of time in range was 46.6% (CI95= 44.0 - 49.2%) in the phenprocoumon group versus 41.6% (CI95= 39.1 - 44.1%) in the acenocoumarol group (δ=5.0%, CI95= 1.3 -8.6%) (table 2).
Table 2. Overall stability of OAT
Phenprocoumon Acenocoumarol d ifference mean C I9 5 mean C I9 5 mean C I9 5 All patients
INR checks in range 42.7% 40.4 - 44.9% 36.5% 34.3 -38.8% 6.1% 3.0 - 9.3% Time in range 46.6% 44.0 - 49.2% 41.6% 39.1 - 44.1% 5.0% 1.3 - 8.6%
INR 2.5-3.5
INR checks in range 44.0% 41.3 - 46.6% 39.5% 36.8 - 42.2% 4.4% 0.6 - 8.3% Time in range 47.6% 45.6 - 51.0% 44.5% 41.4 - 47.6% 3.1% -1.3 - 7.5%
INR 3.0-4.0
INR checks in range 39.0% 35.3 - 42.9% 28.7% 25.1 - 32.3% 10.4% 5.1 - 15.7% Time in range 43.8% 39.2 - 48.5% 34.2% 30.5 - 37.8% 9.7% 3.8 - 15.6%
When we looked at the individual per-centage of time in range, 47.4% of the patients on phenprocoumon were at least 50% of the time within their target INR range while this was true for only 32.2% of the patients on acenocoumarol.
This analysis was repeated for the two main therapeutic INR target ranges. In the INR 2.5-3.5 group the difference between phen-procoumon and acenocoumarol was 4.4%
When the analysis was restricted to the first 6 weeks of the OAT during which the anticoagulation is unstable, no difference was observed for the number of INR checks in range or time in range between phenpro-coumon and acenocoumarol. When the analysis was confined to OAT treatment after
the initial period of 6 weeks there was a clear difference of 8.9% (CI95= 4.6 -13.1%) in the number of INR checks in the target range for phenprocoumon users, mirrored by a differ-ence of 6.1% (CI95= 1.8 - 10.4%) in the per-centage of the time in the target range (table 3).
Table 3. Stability of OAT during start up phase (first 6 weeks) and after 6 weeks.
Phenprocoumon Acenocoumarol difference mean CI95 mean CI95 mean CI95 All patients
Start up phase
INR checks in range 33.1% 29.9 - 36.3% 30.9% 27.6 - 34.1% 2.3% -2.3 - 6.8% Time in range 32.3% 28.8 - 35.7% 31.5% 27.8 - 35.1% 0.8% -4.2 - 5.9% After 6 weeks
INR checks in range 50.5% 47.5 - 53.4% 41.6% 38.5 - 44.6% 8.9% 4.6 - 13.1% Time in range 51.3% 48.2 - 54.4% 45.1% 42.1 - 48.1% 6.1% 1.8 - 10.4% I N R 2 .5 -3 .5
Start up phase
INR checks in range 34.3% 30.5 - 38.1% 32.6% 28.8 - 36.4% 1.7% -3.7 - 7.1% Time in range 33.4% 28.2 - 36.5% 32.4% 27.9 - 36.9% 1.0% -5.1 - 7.2% After 6 weeks
INR checks in range 52.1% 48.5 - 55.6% 46.1% 42.4 - 49.8% 5.9% 0.8 - 11.1% Time in range 52.7% 49.0 - 56.5% 49.1% 45.4 - 52.8% 3.6% -1.6 - 8.9% I N R 3 .0 -4 .0
Start up phase
INR checks in range 29.8% 24.0 - 35.5% 26.4% 20.6 - 32.2% 3.4% -4.9 - 11.7% Time in range 29.1% 22.9 - 35.3% 29.1% 23.1 - 35.1% -0.03% -8.8 - 8.7% After 6 weeks
This difference was again most striking for the patients with a target INR of 3.0-4.0 at 12.2% (CI95= 5.6-18.7%) and less so for those with a target range of 2.5-3.5 INR at 3.6% (CI95= -1.6-8.9%). The percentages of INR checks in range show a similar picture.
The superiority of phenprocoumon in yielding more stable OAT was present for all indications, and the difference was most strik-ing for vascular prosthesis and vascular sur-gery (δ=12.9%, CI95= -1.1 - 26.9%), CABG (δ=12.8%, CI95= 5.1 - 20.5%) and in atrial fib-rillation (δ=8.2%, CI95= -1.6 - 18.0).
The overall proportion of time above an INR of 5 was 3.5% (CI95= 2.5 - 4.4%) in the phenprocoumon group (CI95= 3.7 - 5.7%) compared with a similar 2.9% (CI95= 2.0 -3.7%) in the acenocoumarol group (δ=0.6%, CI95= -0.7 – 1.8%). During the initial 6 weeks of treatment relatively more time was spent above an INR of 5 in the phenprocoumon group at 4.1% (CI95= 1.0 – 7.2%) against 2.4% (CI95= 0.0 - 4.9%) in the acenocoumarol group (δ=1.6%, CI95= -0.4 – 3.7%). After the first 6 weeks both patient groups spent an equal time (δ=0.3%, CI95= -1.3 – 1.8%) above an INR of 5 with phenprocoumon at 3.3% (CI95= 1.3 - 5.4%) and acenocoumarol at 3.1% (CI95= 0.9 – 5.2%). The time spent above an INR of 6.0 was also similar for both drugs.
Looking at the overall period of time spend below an INR of 2, i.e. at
subtherapeu-tic levels, this was clearly lower (δ= -4.9%, CI95= -7.9 – -2.0%) in the phenprocoumon group at 10.3% of the time (CI95= 0.8 – 12.2%), than in the acenocoumarol group at 15.2% (CI95= 13.0 – 17.4%). This difference was again largest during the first 6 weeks of the treatment at -7.2% (CI95= -12.5 – -1.9%) against -2.9% (CI95= -5.7 – -0.05%) for the OAT after the initial 6 weeks.
The same methods for patient selection, matching and analysis were used for the analysis of patients in the Roosendaal Anticoagulation Clinic. Because of a change in the computer dosing program during the period between 01-01-1998 and 01-01-2000 only 378 patients (320 on acenocoumarol and 58 on phenprocoumon) were eligible for inclusion in the study. 51 patient pairs could be matched (table 4). Results for the Roosendaal Anticoagulation Clinic are given in table 5. Overall phenprocoumon scored 49.8% (CI95= 44.4 - 53.3%) of the INR checks and 53.0% (CI95= 47.5 - 58.5%) of the time in target range, while acenocoumarol only scored 41.1% (CI95= 36.1 - 46.2%) of the INR checks and 47.1% (CI95= 41.0 - 53.2%) of the time in target range, the differences being 7.7% (CI95= 0.9 -14.5%) for the percentage of INR checks in range and 5.9% (CI95= -2.4 – 14.2%) for time in range. These results con-firmed our findings in the Leiden Anticoagulation Clinic.
Table 4 . R elevant characteristics of the two patient groups in the R oosendaal Anticoagulation Clinic
Phenprocoumon Acenocoumarol All patients
Table 6. hemorrhagic and thromboembolic complications during OAT
Phenprocoumon Acenocoumarol RR difference Leiden Anticoagulation Clinic
All bleeding complications (/patient/year) 26 (0.24) 12 (0.10) 2.40 1.21 - 4.76 Minor bleeding complications (/patient/year) 21 (0.19) 7 (0.06) 3.17 1.35 - 7.45 Major bleeding complications (/patient/year) 4 (0.04) 4 (0.03) 1.33 0.33 - 5.33 Fatal bleeding complications (/patient/year) 1 (0.01) 1 (0.01) 1.00 0.06 - 15.99 Bleeding complications below range (/patient/year) 1 (0.01) 1 (0.01) 1.00 0.06 - 15.99 within range (/patient/year) 6 (0.06) 5 (0.05) 1.20 0.38 - 3.78 above range (/patient/year) 19 (0.17) 6 (0.05) 3.40 1.36 - 8.51 All thrombo-embolic complications (/patient/year) 1 (0.01) 2 (0.02) 0.53 0.05 - 5.84 Fatal thrombo-embolic complications (/patient/year) - 2 (0.02)
Cardiovascular deaths (/patient/year) 7 (0.06) 7 (0.06) 1.10 0.39 - 3.13
Roosendaal Anticoagulation Clinic
Total bleeding complications (/patient/year) 5 (0.18) 7 (0.26) 0.69 0.22 - 2.18 Minor bleeding complications (/patient/year) 4 (0.15) 5 (0.19) 0.79 0.21 - 2.94 Major bleeding complications (/patient/year) 1 (0.04) 2 (0.08) 0.50 0.04 - 5.51 Fatal bleeding complications (/patient/year) -
-All thrombo-embolic complications (/patient/year) - -Table 5. Stability of OAT in the Roosendaal Anticoagulation Clinic
Phenprocoumon Acenocoumarol difference
mean CI95 mean CI95 mean CI95 CI8 0 All patients
INR checks in range 48.9% 44.4 - 53.3% 41.1% 36.1 - 46.2% 7.7% 0.9 - 14.5% 3.3 - 12.2% Time in range 53.0% 47.5 - 58.5% 47.1% 41.0 - 53.2% 5.9% -2.4 - 14.2% 0.5 - 11.3% Star t up ph ase ( f ir st 6 w eek s)
INR checks in range 40.8% 33.9 - 47.8% 36.5% 29.4 - 43.5% 4.4% -5.7 - 14.4% -2.1 - 10.9% Time in range 37.9% 30.5 - 45.3% 36.9% 29.0 - 44.9% 1.0% -10.1- 12.0% -6.2 - 8.1% Stab ility ph ase ( af ter 6 w eek s)
The number of minor bleeding complica-tions in the Leiden patients (table 6) was higher in the phenprocoumon group (21 minor bleeding episodes = 0.19 haemor-rhages/ patient/year) than in the aceno-coumarol group (7 minor bleeding episodes = 0.06 haemorrhages/patient/year) (RR= 3.17, CI95= 1.35 – 7.45). There was no difference in major bleeding complications (phenpro-coumon: 4 episodes = 0.04 bleedings/patient/year; acenocoumarol: 4 episodes = 0.03 bleedings/patient/year) (RR= 1.33, CI95= 0.33 – 5.33). Most of the bleeding complications occurred at INR values above the target range (phenprocoumon 19/28, acenocoumarol 6/12, RR= 3.4, CI95= 1.36 -8.51). In contrast to our findings the Roosendaal data showed less bleeding com-plications with phenprocoumon than with acenocoumarol (table 6).
Discussion
Our results show a clear advantage of phenprocoumon over acenocoumarol in the quality of OAT, especially for treatment extending beyond 6 weeks and for indica-tions with a higher therapeutic INR target range. Currently, 75% of the patients on OAT in the Netherlands are treated with the short-acting acenocoumarol and only 25% with the long-acting phenprocoumon.
During the start-up phase of OAT and for short-term anticoagulation a small effect was already visible, which became more pro-nounced during long term treatment. A high-er numbhigh-er of patients had INR values in the target range for a longer period of time with phenprocoumon than with acenocoumarol. This difference did not translate itself into much less frequent INR checks for phenpro-coumon patients than for acenocoumarol patients. In view of the difference in the quality of OAT between phenprocoumon and acenocoumarol more frequent INR checks would seem advisable in acenocoumarol patients.
Because of the possibility that our results were influenced by a “learning effect“ because of the preponderance of phenpro-coumon patients in the Leiden Anticoagulation Clinic, the same analysis was carried out on the records of another Dutch anticoagulation clinic (Roosendaal Anticoagulation Clinic) which has a prepon-derance of acenocoumarol patients. The results were largely similar, confirming supe-riority of the longer-acting drug.
The increased frequency of minor bleed-ing complications that was observed in the phenprocoumon group in the Leiden clinic can not be easily explained. As the occur-rence of bleeding complications is correlated
with the time the patients spent above their target range, it was unexpected to find that the coumarin with the most time within the target INR range showed more minor bleed-ing complications. More than 50% of the bleeding complications in our patients occurred at INR values above the therapeutic target range and there was no difference between the drugs in the time spent above an INR of 5. It has to be stressed that the differ-ence occurred mainly in the number of mild bleeding complications, as there was little dif-ference in major bleeding complications between the two different coumarins. In the Roosendaal data no difference in bleeding complications, minor or major, between the different coumarins was observed but patient numbers were low. The methods used for the surveillance of bleeding complications were the same for both coumarins and a pharma-cological cause for the difference is hard to imagine. Therefore the significance of this finding in the Leiden patients is unclear, although the existence of a trend towards more bleeding complications with phenpro-coumon has been reported by us in previous studies and as such can not be discounted.
compli-cations between the two coumarins were it not for the fact that difference is seen in the period of time the patients spent above the target range and this is similar with both drugs.
Our findings confirm the previous obser-vations that longer-acting coumarins provide a higher quality of oral anticoagulant therapy than short-acting coumarins. In particular, phenprocoumon should be preferred over acenocoumarol especially for extended OAT. In terms of clinically important bleeding com-plications there is little difference between phenprocoumon and acenocoumarol, but there is a tendency towards more minor bleeding episodes. In contrast, there is a ten-dency towards more thrombo-embolic com-plications with acenocoumarol where patients spent longer periods of time below their target INR range.
cerning the stability of the anticoagulant effect of acenocoumarol and phenprocoumon. Acta Medica Scandinavica 1969; 186:283-8.
Cannegieter SC, Rosendaal FR, Wintzen AR, van der Meer FJ, Vandenbroucke JP, Briet E. Optimal oral anticoag-ulant therapy in patients with mechanical heart valves. New England Journal of Medicine 1995; 333:11-7. Cleophas TJ, Tavenier P, Niemeyer MG. The risk of
emer-gency intestinal bleeding among users of aceno-coumarin: a population-based cohort study. Angiology 1993; 44:85-92.
Ensor R, Peters H. Experience with the anticoagulant Marcumar. Annals of Internal Medicine 1957; 47:731-43.
Fekkes N, Jonge Hd, Veltkamp JJ, Bieger R, Loeliger EA. Comparative study of the clinical effect of aceno-coumarol (Sintrom) and phenprocoumon (Marcoumar) in myocardial infarction and angina pec-toris. Acta Medica Scandinavica 1971; 190:535-40. Fihn S, McDonell M, Martin D, et al. Risk factors for
com-plications of chronic anticoagulation. Annals of inter-nal medicine 1993; 118:511-520.
Fihn S, Callahan C, Martin D, McDonell M, Henikoff J, White R. The risk for and severity of bleeding compli-cations in elderly patients treated with warfarin. Annals of internal medicine 1996; 124:970-979. Harder S, Thurmann P. Clinically important drug
interac-tions with anticoagulants. An update. Clinical Pharmacokinetics 1996; 30:416-44.
Hemker HC, Frank HL. The mechanism of action of oral anticoagulants and its consequences for the practice of oral anticoagulation. Haemostasis 1985; 15:263-70. Kazmier F, Spittell J, Thompson J, Owen C, Minn R. Effect
of oral anticoagulants on factors VII, IX, X and II. Archives of Internal Medicine 1965; 115:667-673. Kelly J, O’Malley K. Clinical pharmakokinetics of oral
anti-coagulants. Clinical Pharmakokinetics 1979; 4:1-15. Loeliger E, van der Esch B, Matternand M, den Brabander
A. Behaviour of Factors II, VII, IX and X during Long-Term Treatment with Coumarin. Thromb Diath Haemorrh 1963; 9:74-89.
Loeliger E, van Dijk-Wierda C, van den Besselaar A,
John Wiley and Sons Ltd, 1984:135-77.
Palareti G, Leali N, Coccheri S, et al. Bleeding complica-tions of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. The Lancet 1996; 348:423-8.
Pattacini C, Manotti C, Pini M, Quintavalla R, Dettori AG. A comparative study on the quality of oral anticoagulant therapy (warfarin versus acenocoumarol). Thrombosis and Haemostasis 1994; 71:188-91.
Rodman T, Pastor B, Resnick M. Phenprocoumon, Diphenadione, Warfarin and Bishydroxycoumarin: a comparative study. American Journal of Medical Science 1964; 247:655-664.
Rosendaal FR, Cannegieter SC, van der Meer FJ, Briet E. A method to determine the optimal intensity of oral anti-coagulant therapy. Thrombosis and Haemostasis 1993; 69:236-9
Simson G. Five-year study of anticoagulants. Angiology 1962; 13:226-30.
Sullivan P, Kozonis M. How effective is clinical control of anticoagulant therapy? Postgraduate Medicine 1965; 38.
Thijssen HH, Hamulyak K, Willigers H. 4-Hydroxycoumarin oral anticoagulants: pharmacoki-netics-response relationship. Thrombosis and Haemostasis 1988; 60:35-8.
Van den Besselaar AM, van der Meer FJ, Gerrits-Drabbe CW. Therapeutic control of oral anticoagulant treat-ment in The Netherlands. American Journal of Clinical Patholology 1988; 90:685-90.
Van der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briet E. Bleeding complications in oral anticoagulant therapy. An analysis of risk factors. Archives of Internal Medicine 1993; 153:1557-62.
Van der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briet E. Assessment of a bleeding risk index in two cohorts of patients treated with oral anticoagulants. Thrombosis and Haemostasis 1996; 76:12-6.
