Weight-adjusted dosing of tinzaparin for thromboprophylaxis in obese medical patients

Background The optimal dose of tinzaparin for prophylaxis in obese medical patients is not well defined. Objectives To evaluate the anti-Xa activity in obese medical patients on tinzaparin prophylaxis adjusted for actual bodyweight. Methods Patients with a body mass index of ≥30 kg/m2 treated with 50 IU/kg tinzaparin once daily were prospectively included. Anti-Xa and anti-IIa activity; von Willebrand factor antigen and von Willebrand activity; factor VIII activity; D-dimer, prothrombin fragments; and thrombin generation were measured 4 hours after subcutaneous injection between days 1 and 14 after the initiation of tinzaparin prophylaxis. Results We included 121 plasma samples from 66 patients (48.5% women), with a median weight of 125 kg (range, 82-300 kg) and a median body mass index of 41.9 kg/m2 (range, 30.1-88.6 kg/m2). The target anti-Xa activity of 0.2 to 0.4 IU/mL was achieved in 80 plasma samples (66.1%); 39 samples (32.2%) were below and 2 samples (1.7%) above the target range. The median anti-Xa activity was 0.25 IU/mL (IQR, 0.19-0.31 IU/mL), 0.23 IU/mL (IQR, 0.17-0.28 IU/mL), and 0.21 IU/mL (IQR, 0.17-0.25 IU/mL) on days 1 to 3, days 4 to 6, and days 7 to 14, respectively. The anti-Xa activity did not differ among the weight groups (P = .19). Injection into the upper arm compared to the abdomen resulted in a lower endogenous thrombin potential, a lower peak thrombin, and a trend to a higher anti-Xa activity. Conclusion Dosing of tinzaparin adjusted for actual bodyweight in obese patients achieved anti-Xa activity in the target range for most patients, without accumulation or overdosing. In addition, there is a significant difference in thrombin generation depending on the injection site.


| I N T R O D U C T I O N
The prevalence of obesity (body mass index [BMI] ≥ 30 kg/m 2 ) is rapidly increasing worldwide [1]. About 18% of all adults in Germany already belong to this vulnerable group [2]. Obesity is associated with a chronic inflammation and oxidative stress impairing the protective endothelial function. As a result, obese subjects have higher levels of procoagulant tissue factor, factor (F)VIII, and von Willebrand factor (VWF) as well as an impaired fibrinolysis and a higher platelet activity [3]. Consequently, patients with obesity have a significantly higher risk of venous thromboembolism (VTE) [4]. Therefore, sufficient thromboprophylaxis is particularly important in this patient cohort because adequate thromboprophylaxis is associated with a nearly 90% lower risk of VTE complications [5].
Patients with obesity show differences in the absorption, distribution, metabolism, and elimination of drugs. Barrett et al. found that the clearance of tinzaparin decreases by approximately 25% in obese patients [6]. In addition, absorption may be prolonged after a s.c. injection, which was shown by Sanderink et al. with enoxaparin in obese volunteers [7]. The volume of distribution of low molecular weight heparins (LMWHs) is mainly intravascular, which can be overestimated with a bodyweight-adapted dosage and may lead to an accumulation in obese patients [8]. However, whether a high fixed dose or a bodyweight-adapted dosage is more appropriate in patients with obesity is still a matter of debate. Most studies evaluating thromboprophylaxis with LMWHs in patients with obesity have been carried out with enoxaparin [9][10][11][12][13]. Tinzaparin has different pharmacologic properties compared to enoxaparin and other LMWHs. Tinzaparin has the lowest anti-Xa/anti-IIa activity ratio of all LMWHs and shows the highest inhibition of thrombin generation (TG) at similar anti-Xa activity compared to other LMWHs [14]. Because tinzaparin has a higher molecular weight, the proportion of nonrenal elimination is greater than that of other LMWHs [15]. Therefore, tinzaparin does not lead to a significant accumulation of the anti-FXa activity in elderly patients with impaired renal function [16][17][18].
In a pharmacokinetic study in healthy obese volunteers, tinzaparin at a prophylactic dose of 75 IU/kg and a therapeutic dose of 175 IU/kg per actual bodyweight resulted in a comparable anti-Xa activity as in nonobese volunteers [19]. A recent case series examined obese medical patients receiving 175 IU/kg tinzaparin per actual bodyweight. Only 13.3% of them had anti-Xa activities below the target range and no overdosing occurred up to a bodyweight of 222 kg [20]. However, there are very few studies focusing on thromboprophylaxis with tinzaparin in obese patients and most of them have been carried out in patients undergoing bariatric surgery, [21][22][23]  The exclusion criteria were as follows: age < 18 years, no informed consent, surgery within 14 days prior to inclusion, active malignancy, glomerular filtration rate (GFR) < 20 mL/min, liver cirrhosis child B or C, current pregnancy and lactation, mechanical heart valve, allergy to tinzaparin, or a history of heparin-induced thrombocytopenia.

Essentials
• The optimal prophylactic dose of tinzaparin in obese medical patients is not well defined.
• The target anti-Xa activity was achieved in 66%, whereas 32% were below the target range.
• Anti-Xa activity did not differ among weight groups and showed no accumulation for over 14 days.
Blood samples were drawn between day 1 and day 14 after initiation of tinzaparin prophylaxis 4 hours after the s.c. administration of tinzaparin to measure anti-Xa activity; anti-IIa activity; FVIII, VWF antigen, and VWF activity; prothrombin fragment and D-dimer activity; and TG. The following variables were collected from patient records: age, sex, weight, height, diagnosis, and comorbidities. In addition, standard laboratory results including hematology, renal, and liver parameters were taken from the clinical routine. Bleeding events were defined according to the International Society on Thrombosis and Haemostasis nomenclature [26,27].

| Blood sampling and laboratory measurement
Blood samples were collected into 3 mL tubes (Sarstedt) containing 3.2% citrate. The samples were processed within 1 hour after the blood collection, centrifuged at 2000 g for 20 minutes to prepare platelet-poor plasma, aliquoted, and immediately stored at −80 • C.
The aliquots were thawed directly before the analysis. Anti-Xa activity was determined with a chromogenic substrate and a bovine FXa (Hyphen BioMed), and a chromogenic anti-IIa assay with human thrombin (Hyphen BioMed) was used to measure anti-IIa activity. The

| Statistical analysis
Descriptive statistics are reported for quantitative variables as either mean ± SD for normally distributed data or as median with IQR otherwise. Qualitative data are given as numbers (percentages).
Intergroup comparisons were performed using either t-test, if the data are normally distributed, or the Mann-Whitney U-test or Kruskal-Wallis test otherwise, and all plasma samples from all patients were analyzed. Only the first plasma sample from every group was taken into account for the comparison of time-dependent variables and Friedman's test was used. Adjustment for multiple testing was performed using the Bonferroni correction. Spearman-Rho correlation coefficients were calculated to assess the effects between the different laboratory parameters and bodyweight. A P value of <.05 was considered statistically significant. Statistical analysis was conducted using the program SPSS version 27 (SPSS Inc) and GraphPad Prism version 9.4.0 (GraphPad Software).      Table 2.

| Ethical considerations
Patients were classified into 5 weight groups according to the tinzaparin dose administered, as shown in Table 1 There was no significant difference in the distribution of the anti-Xa activity among the 5 weight groups (P = .19) when plasma samples taken from all days were considered together (part A of Figure).
No accumulation of the anti-Xa activity occurred between days 1 to 3, days 4 to 6, and days 7 to 14 (P = .42; part B of Figure).
The anti-Xa activity and the achievement of the anti-Xa target range among the different weight groups are summarized in Table 3.

| Anti-IIa activity
There was a significant difference in the distribution of the anti-IIa

| The influence of body mass index and bodyweight on other coagulation parameters
There was a significant, but weak correlation between BMI and D- The lowest D-dimer activity was found in patients with a BMI of <40 kg/m 2 , followed by those with a BMI of >50 kg/m 2 (P = .09) and those with a BMI of 40 to 50 kg/m 2 (P = .03). There was no change in D-dimer activity between days 1 to 3, days 4 to 6, and day 7 to 15 (P = .74). The anti-IIa activity was not different between these 3 BMI groups (P = .06). There was a significant difference in the distribution of the VWF antigen over the 3 BMI groups, with highest values in patients with a BMI of 40 to 50 kg/m 2 (P = .02). Apart from that, no differences in the activity of VWF, FVIII, and prothrombin fragments were found between the BMI groups. The distribution of coagulation parameters among different BMI groups is summarized in Table 4.
There was a significant difference in the distribution of the Ddimer among the weight groups (P = .03). D-dimer were significantly higher in patients weighing >200 kg than those with a bodyweight of 101 to 119 kg (P = .03). VWF antigen was significantly lower in the group with ≤100 kg than in the group with 120 to 159 kg (P = .01) or in the group with ≥200 kg (P = .006). Otherwise, no significant differences in prothrombin fragment, anti-Xa, FVIII, and VWF activity or TG parameters were found between the bodyweight groups. The distribution of laboratory parameters among bodyweight groups is summarized in Table 5.

| Correlation of anti-Xa and anti-IIa activity with TG parameters
There was a moderate but significant correlation between the factor anti-IIa and anti-Xa activity with a correlation coefficient of 0.692 (P <

| The influence of the injection site on the laboratory results
Tinzaparin was injected into the upper arm in 102 plasma samples (84.3%), whereas 2 injections were administered into the thigh and 17 into the abdominal wall. We found a significant difference in the distribution of all of the 6 TG parameters over the 2 groups. As shown in Table 6, there were a significantly lower ETP, a significantly lower peak thrombin, and a trend to a higher anti-Xa activity in plasma samples taken after the injection of tinzaparin into the upper arm compared to the abdominal wall, but bodyweight and BMI were not different. The T A B L E 3 Anti-Xa activity and target range among weight groups and day of tinzaparin application  -5 of 10 differences between the injection into the upper arm and that into the abdomen are summarized in Table 6.

| The influence of the glomerular filtration rate on the laboratory results
There was no significant difference in the anti-Xa activity (P = .81) or the anti-IIa activity (P = .18) between patients with a GFR of ≤50 mL/min   subjects of normal weight [19,28]. We found that no accumulation occurred after repeated dosing for more than 7 days. This is consistent with the study by Mahe et al. [18] in patients with renal failure, which showed no accumulation of anti-Xa activity by tinzaparin after 8 days. Tseng et al. [23] investigated the safety of weight-adjusted prophylaxis with 75 IU/kg tinzaparin after bariatric surgery and concluded that it appears to be a safe strategy with few major bleeding events (1.6%) and a low rate of venous thromboembolism (0.5%) within 30 days after bariatric surgery. Unfortunately, the anti-Xa activity was not monitored in that study. Most data exist only on the use of enoxaparin for thromboprophylaxis in bariatric surgery.
These studies conclude that patients receiving a weight-based enoxaparin prophylaxis are more likely to have their anti-Xa activity in the target range than with a fixed-dose regimen, whereas thrombotic complications are rare [29][30][31][32][33][34].
Nevertheless, whether the anti-Xa activity correlates with the clinical outcome, particularly venous thromboembolism or bleeding, is still a matter of debate [35]. Bara et al. [36] examined 440 patients who underwent total hip replacement surgery and found no correlation between the anti-IIa or anti-Xa activity and the clinical outcome.
Cooper et al. reported that not weight-adjusted tinzaparin dosing but rather the addition of acetylsalicylic acid results in persistent wound drainage after knee and hip arthroplasty [37]. In a recently published study on patients with gynecologic cancer receiving prophylactic treatment with tinzaparin, the anti-Xa activity was significantly lower in patients who developed VTE than in those without a thrombotic event, and the anti-Xa activity was significantly higher in patients receiving a weight-adjusted prophylaxis compared to a fixed dose of 4500 U of tinzaparin [38]. According to these findings, some current guidelines and expert opinion suggest a bodyweight-adjusted VTE prophylaxis in obese patients [39][40][41]. However, most international guidelines do not address dosing in patients with obesity because of the lack of evidence [42][43][44][45][46][47]. For high-risk bariatric surgery patients, the measurement of the anti-Xa activity should be considered, [48] as there is a 28-fold increase in the mortality risk once venous thromboembolism occurs postoperatively [49]. As the target range of anti-Xa was only achieved in two-thirds of the patients, we suggest that its measurement is warranted especially in the context of high mortality with thromboembolism.
The absorption of LMWH in obese patients has not been fully understood. We found a trend to a higher anti-Xa activity after the administration of tinzaparin into the upper arm compared to that into the abdominal wall. In addition, injection into the abdominal wall led to significantly lower ETP and peak thrombin values. This might be clinically associated with less effective thromboprophylaxis. However, this observation might be biased by the uneven sample distribution (abdominal wall: n = 17; upper arm: n = 102), so further confirmation studies are needed. In addition, the difference could be caused by delayed absorption after injection into the abdominal wall because of lower venous return, [50] so that the plasma peak occurs 4 hours after injection. Local skin movement is limited in the abdomen compared to the upper arm region, which impacts s.c. blood flow and absorption [51]. Studies focusing on the differences between injection sites, especially in obese people, are very limited. Sanderink et al. [7] examined the pharmacokinetic parameters on days 1 and 4 after daily s.c. enoxaparin administration in 24 nonobese and obese volunteers.
They found that after s.c. injection at day 4, the absorption time is Obesity seems to be associated with a higher risk of VTE [52].
This obesity-induced prothrombotic state results from the interplay of physical immobility and proinflammatory and hypofibrinolytic effects [53]. Adipocytes secrete inflammatory cytokines that stimulate the endothelium and platelets, leading to increased tissue factor expression and TG [54,55] and an impaired fibrinolysis [3]. TG is known to return to lower values after weight-loss following bariatric surgery [56]. As C-reactive protein decreases too, fewer proinflammatory stimuli are a possible explanation [57]. With regard to FVIII and VWF, we confirmed that BMI seems to be associated with higher VWF and FVIII levels as demonstrated by Atiq et al [58]. Finally, D-dimer-likely inflammatory-driven-were elevated in our patient cohort. This confirms the data of Franco et al., [59] who reported elevated D-dimer in obese patients in correlation to the waist-to-hip ratio. Elevated Ddimer have been reported to predict VTE in acutely ill medical patients [60][61][62]. In the MEDENOX trial that included nonobese patients, 40mg of enoxaparin but not 20-mg of enoxaparin or placebo resulted in a small but significant reduction in the elevated baseline D-dimer PFREPPER ET AL.
-7 of 10 activity after 10 days [62]. VTE was documented only among patients with elevated D-dimer at baseline that persisted after 10 days. In contrast, D-dimer did not decrease over time in our cohort of obese medical patients despite a weight-adapted thromboprophylaxis. This might be explained by the proinflammatory status associated with obesity. Considering that one-third of our patients did not achieve the targeted anti-Xa activity, this might justify an even higher dose of thromboprophylaxis. However, our cohort was too small to draw definite conclusions and was not powered for endpoints such as VTE occurrence.
With regard to TG, we found higher correlations between the TG parameters and the anti-IIa compared with the anti-Xa activity. The fact that TG measurements are more influenced by the anti-IIa activity than by the anti-Xa activity of LMWH has been described previously [14,63]. As a result, tinzaparin has a higher effect on TG than enoxaparin, nadroparin, or dalteparin because of the lower anti-Xa/anti-IIa ratio. However, whether this effect is clinically relevant or just an effect of the more downstream inhibition of TG by the anti-IIa activity or whether a low anti-IIa activity may indicate a less efficient prophylaxis sooner remains unclear. Therefore, our data cannot be generalized to other LMWHs.

L I M I T A T I O N S
This study has several limitations because of the relatively small cohort of patients and the uneven distribution among the different bodyweight and BMI groups. Therefore, despite that correction for multiple testing was made, our results should be interpreted with caution and confirmed in a larger cohort. In addition, the observation period was short, and no follow-up regarding thrombotic complications and no duplex sonography were performed to exclude occult thrombosis. Therefore, no final conclusion can be drawn regarding the effectiveness of the dosing regimen. Nevertheless, this is, to our knowledge, the first study that examined the bodyweight-adjusted dosing of tinzaparin in obese medical patients by collecting anti-Xa and anti-IIa peak values.

S U M M AR Y
Our data suggest that bodyweight-adjusted dosing of tinzaparin results in an effective thromboprophylaxis in obese medical patients, with the majority of patients achieving the anti-Xa target range. No accumulation of the anti-Xa activity occurred over a period of 14 days.
Furthermore, we found evidence that s.c. injection into the abdominal wall may lead to a less effective inhibition of TG than the administration into the upper arm. Therefore, we recommend administration of weight-adjusted tinzaparin into the upper arm in patients with obesity. Nevertheless, further research is required to prove our findings.

FUNDING
The study was supported by an unrestricted grant from LEO Pharma.

AUTHOR CONTRIBUTIONS
Leo Pharma supported the study but did not contribute as an author.