An adaptive-design dose-ranging study of PD 0348292, an oral factor Xa inhibitor, for thromboprophylaxis after total knee replacement surgery

Authors


  • Manuscript handled by: M. Levi
  • Final decision: F. R. Rosendaal, 14 June 2013

Correspondence: Raphael Pak, Clinical Sciences, Primary Care Business Unit, 445 Eastern Point Road, Groton, CT 06340, USA.

Tel.: +1 734 8878011; fax: +1 860 7159926.

E-mail: raphael.pak@pfizer.com

Summary

Background

PD 0348292 is an oral, selective, direct and reversible factor Xa inhibitor. This was an adaptive dose-ranging study evaluating a 100-fold PD 0348292 dose range in subjects undergoing total knee replacement (TKR).

Objective

To assess the efficacy and safety of a dose range of PD 0348292 relative to enoxaparin for the prevention of venous thromboembolism (VTE).

Methods

Extensive dose-response modeling and trial simulations were used to select the PD 0348292 dose range for the Phase 2 study. Subjects were randomized to a blinded PD 0348292 dose (0.1 mg qd to 10 mg qd) or open-label enoxaparin (30 mg bid) for 6–14 days after TKR surgery. Efficacy was assessed by mandatory bilateral venography. Results were analyzed using a dose-response modeling approach.

Results

Observed VTE frequency ranged from 1.4–37.1% across PD 0348292 doses and was 18.1% for enoxaparin. The PD 0348292 dose-response relationship for VTE was statistically significant (< 0.0001). The dose of PD 0348292 equivalent to enoxaparin 30 mg bid for VTE prevention was estimated to be 1.16 mg (95% CI = 0.56 mg, 2.41 mg) qd. Total bleeding ranged from 4.9% to 13.8% across PD 0348292 doses and was 6.3% with enoxaparin. The dose-response relationship for total bleeding was not statistically significant (= 0.2464). Overall, PD 0348292 and enoxaparin were well tolerated.

Conclusion

Characterization of the dose-response relationship for VTE and bleeding using an adaptive Phase 2 study design provided a strong quantitative basis for Phase 3 dose selection.

Introduction

Total knee replacement (TKR) surgery predisposes subjects to venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE) [1]. Short-term thromboprophylaxis with low-molecular-weight heparins, fondaparinux or vitamin K antagonists (VKAs) after TKR are effective, but each has significant inconvenient and costly drawbacks [2]. Factor Xa (FXa) is an attractive target for novel anticoagulants because of its specific role in the conversion of prothrombin to thrombin in the coagulation cascade, with little activity outside of coagulation [3]. Several direct FXa inhibitors approved or in clinical development have emerged as oral anticoagulants poised to address many shortcomings of prior therapy [4].

PD 0348292 is an oral, selective, direct and reversible FXa inhibitor [5, 6] that was safe and well tolerated at a wide range of doses studied in early clinical single- and multiple-dose trials [7]. The pharmacokinetics of PD 0348292 are linear and time-independent over a dose range of 0.l mg to 10 mg and the terminal half-life is approximately 10 h. The objectives of this dose-ranging study were to characterize the dose-response relationship of PD 0348292 with respect to VTE prevention and bleeding endpoints in subjects undergoing an elective, unilateral TKR and to estimate the dose equivalent to enoxaparin 30 mg bid for efficacy.

Methods

Study design

This was a randomized, active-controlled, parallel-group, adaptive dose-ranging, multicenter study. Subjects were randomized to one of a number of double-blind PD 0348292 dose groups, or an open-label enoxaparin control group. A total of seven oral doses of PD 0348292 were studied: 0.1, 0.3, 0.5, 1.0, 2.5, 4 and 10 mg qd. Initially subjects were randomized to one of the five lowest doses (0.1, 0.3, 0.5, 1.0, or 2.5 mg qd) of PD 0348292 or subcutaneous enoxaparin 30 mg bid, in a 1:1:1:1:1:2 ratio. Discontinuation of lower doses or addition of higher doses was based on prespecified dose decision analyses (described below in Statistical Analysis). Randomization was stratified by region (North America or ex-North America).

Subjects began treatment 6–8 h postoperatively with PD 0348292, or 12–24 h postoperatively with enoxaparin, and continued treatment for 6–14 days. Intermittent pneumatic compression devices were prohibited during treatment, but elastic compression stockings were allowed. A mandatory bilateral venography of the lower extremities was performed at the end of the treatment. Safety assessments were performed after surgery, at Study Day 3, at the end of study treatment, and at the follow-up clinic visit, which occurred 32 ± 4 days after the final treatment dose. The study was approved by the institutional review boards at each site.

Study population

Subjects enrolled were at least 18 years of age and were due to undergo unilateral TKR. Women were either postmenopausal or surgically sterilized. All randomized subjects signed an informed consent document. Exclusion criteria included: low body weight (< 45 kg for women, < 57 kg for men); DVT or PE within the past year, or suspected post-thrombotic state; significant bleeding; anemia (hemoglobin level < 10.0 g dL−1); thrombocytopenia (platelet count < 100 × 103 μL−1); uncontrolled hypertension; creatinine clearance < 30 mL min−1 (based on the Cockcroft-Gault formula); active hepatic disease; cancer; an immunocompromised status; ischemic stroke or myocardial infarction within 3 months; surgery/trauma within 6 months; persistent immobilization; current use of anticoagulants, platelet aggregation inhibitors, or any other drug influencing coagulation; and history of significant adverse reaction to heparin.

Study outcomes evaluation

The primary efficacy endpoint was the incidence of VTE (proximal or distal DVT or PE), occurring at any time during the study treatment period, determined by a central independent imaging core laboratory (Perceptive Informatics, Inc., East Windsor, NJ, USA). Secondary efficacy endpoints were incidence of proximal DVT or PE, incidence of proximal DVT, incidence of distal DVT and incidence of PE.

Subjects were examined for signs and symptoms of DVT and PE throughout the study. Suspected VTE was confirmed by imaging (compression ultrasound [CUS], venography and ventilation-perfusion [VQ] scans), which was objectively adjudicated by the independent blinded central imaging core laboratory. The adjudicated results were used for outcome analysis.

The primary safety endpoint was the incidence of total bleeding, described as major or minor bleeding in accordance with the International Society on Thrombosis and Haemostasis definition [8]. Major bleeding was defined as fatal bleeding, clinically overt bleeding causing a fall in hemoglobin ≥ 20 g L−1, clinically overt bleeding leading to transfusion of ≥ 2 units of whole blood or red cells, or symptomatic bleeding in areas of special concern (e.g. intracranial, retroperitoneal, intraocular, intraspinal, pericardial, intramuscular with compartmental syndrome or intra-articular). Minor bleeding was defined as bleeding that did not meet the definition of major bleeding.

Statistical analyses

Datasets

Subjects included in the efficacy analysis were all randomized subjects who received at least one dose of study medication and:

  1. had bilateral venograms adequate for evaluation as determined by the central imaging core laboratory; or
  2. had symptomatic, objectively confirmed DVT or PE; or
  3. were asymptomatic and for whom bilateral venography could not be performed, but unilateral venography demonstrated thrombus.

All randomized subjects who received at least one dose of study medication were included in bleeding and other safety analyses.

Efficacy

Analysis of the primary efficacy endpoint was based on the treatment period (from the first dose until 12 h after the last dose of enoxaparin or 24 h after the last dose of PD 0348292). The dose of PD 0348292 equivalent to enoxaparin 30 mg bid for total VTE (Deq292) and its 95% confidence interval (CI) were estimated using the following logistic regression model:

display math

In this equation: p is the probability of a VTE; E0 is the intercept, reflecting the response of 1 mg PD 0348292; α is the treatment effect (difference from 1 mg PD 0348292 on the logit scale) of enoxaparin 30 mg bid; Dose292 is the dose of PD 0348292; β is the slope of the log(dose) response relationship for PD 0348292; and λ is the treatment effect of region (difference between North America and ex-North America on the logit scale). The shape of the VTE dose-response relationship was based on dose-response data of other anticoagulants [9]. The parameters E0, α, β and λ were estimated and the dose of PD 0348292 equivalent to 30 mg bid enoxaparin (Deq292) was calculated as:

display math

To aid in the characterization of the dose-response relationship of PD 0348292 for the primary and secondary efficacy endpoints, the following descriptive statistics (non-model-based estimates and CIs) were provided: the estimated incidence and associated 95% CI for each PD 0348292 dose arm and enoxaparin, and the estimated risk ratio and associated 95% CI for each PD 0348292 dose arm relative to enoxaparin.

Safety

The dose-response relationships for PD 0348292 and the safety endpoints of total bleeding incidence and major bleeding incidence were estimated using the following equation:

display math

In this equation: P is the probability of a bleeding event; E0 is the intercept, reflecting placebo response; α is the treatment effect (difference from placebo on the logit scale) of enoxaparin 30 mg bid; Dose292 is the dose of PD 0348292; β is the slope of the dose-response relationship for PD 0348292; and the other parameters are as defined above for the efficacy analysis. The shape of the bleeding dose-response relationship was based on dose-response data of other anticoagulants [9].

In addition, for the endpoints of total, major and minor bleeding, non-model-based estimates and associated 95% CIs of incidence for each PD 0348292 dose arm and enoxaparin and risk ratio for each PD 0348292 dose arm relative to enoxaparin were provided, based on all subjects who received at least one dose of study medication.

Dose decision analysis and sample size

In order to protect subjects from excessive VTE and major bleeding, and to allow for investigation of higher doses of PD 0348292 (i.e. 4 and 10 mg qd), predefined dose decision analyses (DDA), to be assessed by an unblinded, independent Data Monitoring Committee (DMC), were included in the study protocol [9]. An unblinded statistician not directly involved with the study performed the DDA and provided the results to the DMC.

A DDA was to be performed using the available data and the logistic regression equations described above after approximately every 147 subjects had either successfully completed the bilateral venograms, or had symptomatic VTE that was objectively confirmed (maximum of four DDAs). In addition to the model parameters (E0, α, β and λ), the statistics used to guide the dose decisions (e.g. incidence point estimates, odds ratios [ORs], CIs) were estimated.

At each DDA the dose range of PD 0348292 could be modified according to prespecified rules. The lowest dose arm would be discontinued if the one-sided 90% lower confidence bound of estimated OR for VTE was > 1.5. Any dose arm would be discontinued if the one-sided lower confidence bound of estimated major bleeding incidence was > 5%. A higher dose would be added if the next lower dose did not meet the criterion described above, and the estimated incidence of major bleeding was acceptable (the extrapolated point estimate for major bleeding was < 5%). No more than five dose arms of PD 0348292 could be active after each DDA.

Sample size was based on estimating the dose of PD 0348292 equivalent to enoxaparin 30 mg bid for reduction of total VTE [9]. A fixed total sample size of 735 subjects with bilateral venograms adequate for evaluation would provide at least 92% probability (power) that the selected dose of PD 0348292 had true total VTE and major bleeding incidences within a factor of 1.3 of enoxaparin 30 mg bid. Assuming a 40% non-evaluable/missing venogram rate, randomization of at least 1225 subjects was planned.

Results

Study population

A total of 1411 subjects were randomized to treatment at 99 centers in Australia, Europe, North and South America and South Africa. The protocol was amended at the discretion of the DMC to increase the sample size by approximately 200 to allow for adequate sample size in the highest dose group. Three DDAs were performed during the study: the three lowest dose groups of PD 0348292 (0.1, 0.3 and 0.5 mg qd) were discontinued early from the study, and the two higher dose groups (4 mg and 10 mg qd) were added. Approximately 98% of the randomized subjects (N = 1389) were treated with at least one dose of the study drug. They were included in bleeding and other safety analyses. The rate of non-evaluable venograms was 47% in this study, which was higher than the expected 40% non-evaluability rate. Nonetheless, a total of 749 subjects with evaluable venograms constituted the dataset for efficacy analysis, providing > 92% power. Ten of the 749 subjects had major protocol violations (i.e. used concomitant non-study anticoagulant, missed > 2 consecutive days of study treatment, were treated with < 6 days of study drug, or not treated within 48 h post-surgery). The overall subject disposition is summarized in Fig. 1.

Figure 1.

Disposition of study population. BVG = bilateral venogram.

The demographic and baseline characteristics of the subjects in this study were comparable across the treatment groups, as summarized in Table 1. Approximately 64% of the randomized subjects were women, the mean age of all subjects was 67 years and > 90% were white. The medical history and the median duration of treatment for the PD 0348292 dose groups and enoxaparin (ranging from 8 to 10 days) were also comparable. The median time between surgery and first dose of study treatment ranged from 6.8 to 7.3 h for PD 0348292 dose groups and was 19.3 h for enoxaparin.

Table 1. Demographic and baseline characteristics of all randomized subjects
Charac-teristicPD 0348292Enoxaparin 30 mg bid N = 397
0.1 mg qd N = 610.3 mg qd N = 1410.5 mg qd N = 1831 mg qd N = 2022.5 mg qd N = 2004 mg qd N = 14010 mg qd N = 65
  1. bid, twice daily; BMI, body mass index; N, number of treated subjects; qd, once daily; SD, standard deviation.

Gender
Male N (%)25 (41.0)44 (31.2)59 (32.2)73 (36.1)75 (37.5)51 (36.4)28 (43.1)140 (35.3)
Female N (%)36 (59.0)97 (68.8)124 (67.8)129 (63.9)125 (62.5)89 (63.6)37 (56.9)257 (64.7)
Age (years)
Mean (SD)67.3 (9.6)67.0 (8.5)68.4 (8.0)66.6 (9.5)67.2 (8.9)65.8 (9.4)65.8 (9.7)66.8 (9.3)
Range47–8643–8649–8538–8742–8745–8428–8728–88
Race, N (%)
White55 (90.2)131 (92.9)163 (89.1)186 (92.1)181 (90.5)125 (89.3)59 (90.8)362 (91.2)
Black4 (6.6)8 (5.7)8 (4.4)5 (2.5)8 (4.0)5 (3.6)2 (3.1)21 (5.3)
Asian1 (1.6)08 (4.4)2 (1.0)5 (2.5)2 (1.4)1 (1.5)1 (0.3)
Other1 (1.6)2 (1.4)4 (2.2)9 (4.5)6 (3.0)8 (5.7)3 (4.6)13 (3.3)
Weight (kg)
Mean (SD)84.9 (18.5)86.3 (19.3)85.0 (17.7)85.7 (17.6)86.2 (18.4)83.8 (18.1)85.8 (20.1)85.0 (17.8)
Range49.0–131.554.0–144.046.0–154.249.0–134.746.0–147.046.7–140.646.0–157.847.6–171.0
Body mass index (kg/m2)
Mean (SD)30.3 (6.2)31.7 (5.8)30.9 (5.3)31.1 (5.7)31.2 (5.8)30.0 (4.7)30.5 (5.9)30.8 (5.8)
Range17.0–49.519.4–46.119.7–47.819.1–52.119.7–52.717.2–42.618.9–55.50.5–54.0

Efficacy

The observed incidence of VTE during the study treatment period ranged from 1.4% to 37.1% across the PD 0348292 doses (Table 2). There were no apparent reasons for the seemingly low incidence of total VTE observed in the 4 mg qd dose group. The observed incidence for enoxaparin was 18.1%. Similar results were obtained using a dataset that excluded the 10 subjects with major protocol violations (data not shown). The PD 0348292 dose-response relationship for incidence of total VTE was statistically significant (P < 0.0001) (Fig. 2). Based on this dose-response model for efficacy, the dose of PD 0348292 equivalent to enoxaparin 30 mg bid for VTE prevention in TKR subjects was estimated to be 1.16 mg qd (95% CI = 0.56 mg, 2.41 mg).

Table 2. Summary of the observed frequency of venous thromboembolic events during the study treatment period
EventPD 0348292Enoxaparin 30 mg bid N = 188
0.1 mg qd N = 350.3 mg qd N = 890.5 mg qd N = 1041 mg qd = 1202.5 mg qd N = 1124 mg qd N = 7410 mg qd N = 27
  1. bid, twice daily; CI, confidence interval; N, number of subjects for efficacy analysis; n, number of subjects with event; qd, once daily; RR, relative risk of event for dose of PD 0348292 versus enoxaparin; VTE, venous thromboembolic event. *VTE = proximal or distal deep vein thrombosis (DVT) or pulmonary embolism (PE).

VTE* n (%)13 (37.1)33 (37.1)30 (28.8)23 (19.2)16 (14.3)1 (1.4)3 (11.1)34 (18.1)
95% CI21.1, 53.227.0, 47.120.1, 37.612.1, 26.27.8, 20.800.0, 4.000.0, 23.012.6, 23.6
RR (95% CI)2.05 (1.21, 3.48)2.05 (1.36, 3.08)1.60 (1.04, 2.45)1.06 (0.66, 1.71)0.79 (0.46, 1.36)0.07 (0.01, 0.54)0.61 (0.20, 1.86) 
Proximal DVT w/wo distal DVT n (%)2 (5.7)3 (3.4)4 (3.8)4 (3.3)2 (1.8)01 (3.7)4 (2.1)
Distal DVT n (%)12 (34.3)31 (34.8)29 (27.9)21 (17.5)13 (11.6)1 (1.4)3 (11.1)30 (16.0)
PE n (%)0001 (0.8)1 (0.9)003 (1.6)
Figure 2.

Predicted dose-response relationship for total venous thromboembolism (VTE) and total bleeding during the study treatment period. Solid lines are dose-response curves for PD 0348292 and effect of enoxaparin 30 mg bid estimated by logistic regression models. Dashed lines are lower and upper 95% CIs. Squares are observed frequencies for total VTE and circles are observed frequencies for total bleeding.

The majority of VTEs that occurred during the study were asymptomatic distal DVTs, with an incidence of 1.4% to 34.3% across the PD 0348292 dose groups, and 16.0% for the enoxaparin group. The number of proximal DVT events was small during the treatment period, with an incidence of 0% to 5.7% in the PD 0348292 dose groups and 2.1% in the enoxaparin group. Two subjects developed symptomatic DVT, one in the PD 0348292 0.5 mg qd group and one in the enoxaparin group. Incidence of the composite endpoint of proximal DVT or PE was comparable across all PD 0348292 dose groups and the enoxaparin group. One case of PE occurred in each of the 1 mg and 2.5 mg qd PD 0348292 dose groups (representing an incidence of 0.8% and 0.9%, respectively). The observed incidence of PE in the enoxaparin group was 1.6%. There were no fatal VTEs. During the 30-day post-treatment follow-up period, one subject in the PD 0348292 0.5 mg qd group developed symptomatic distal DVT, and one subject in the enoxaparin group developed PE with proximal DVT.

Safety

The incidence of total bleeding during treatment ranged from 4.9% to 13.8% in the PD 0348292 dose groups (Table 3). The observed incidence for enoxaparin was 6.3%. The PD 0348292 dose-response relationship for total bleeding during the treatment period was not statistically significant (P = 0.2464) (Fig. 2).

Table 3. Summary of the observed frequency of bleeding events during the study treatment period
 PD 0348292Enoxaparin 30 mg bid N = 397
0.1 mg qd N = 610.3 mg qd N = 1410.5 mg qd N = 1831 mg qd N = 2022.5 mg qd N = 2004 mg qd N = 14010 mg qd N = 65
  1. qd, once daily; bid, twice daily; RR, relative risk of event for dose of PD 0348292 versus enoxaparin; CI, confidence interval; N, number of treated subjects; n, number of subjects with event. *Major bleeding = fatal bleeding, or bleeding causing a fall in hemoglobin ≥ 20 gL−1, or bleeding requiring transfusion of 2 units of blood, or symptomatic bleeding in areas of special concern. †Minor bleeding = bleeding not meeting the criteria of major bleeding. ‡Total bleeding = major and/or minor bleeding.

Major bleeding*n (%)002 (1.1)1 (0.5)1 (0.5)01 (1.5)3 (0.8)
95% CI RR (95% CI)  00.0, 2.6 1.45 (0.24, 8.58)00.0, 1.5 0.66 (0.07, 6.26)00.0, 1.5 0.66 (0.07, 6.32) 00.0, 4.5 2.04 (0.22, 19.28)00.0, 1.6
Minor bleeding n (%)3 (4.9)8 (5.7)18 (9.8)15 (7.4)16 (8.0)15 (10.7)8 (12.3)22 (5.5)
95% CI RR (95% CI)00.0, 10.3 0.89 (0.27, 2.88)1.9, 9.5 1.02 (0.47, 2.25)5.5, 14.2 1.77 (0.98, 3.23)3.8, 11.0 1.34 (0.71, 2.53)4.2, 11.8 1.44 (0.78, 2.69)5.6, 15.8 1.93 (1.03, 3.62)4.3, 20.3 2.22 (1.03, 4.78)3.3, 7.8
Total bleeding n (%)3 (4.9)8 (5.7)19 (10.4)16 (7.9)16 (8.0)15 (10.7)9 (13.8)25 (6.3)
95% CI RR (95% CI)00.0, 10.3 0.78 (0.24, 2.51)1.9, 9.5 0.90 (0.42, 1.95)6.0, 14.8 1.65 (0.93, 2.92)4.2, 11.6 1.26 (0.69, 2.30)4.2, 11.8 1.27 (0.69, 2.32)5.6, 15.8 1.70 (0.92, 3.13)5.4, 22.2 2.20 (1.08, 4.50)3.9, 8.7

The number of major bleeding events was very low, with an observed incidence of major bleeding during treatment of 0% to 1.5% for the PD 0348292 dose groups, compared with 0.8% for enoxaparin (Table 3). The PD 0348292 dose-response relationship for major bleeding was not statistically significant (P = 0.6040). Three of the five major bleeding events in the PD 0348292 groups and two of the three major bleeding events in the enoxaparin group occurred at the surgical site. The incidence of minor bleeding for PD 0348292 (4.9% to 12.3% across doses) also did not show a significant dose-dependent relationship (Table 3).

Overall, PD 0348292 and enoxaparin were well tolerated. No deaths or unexpected adverse events were reported. The incidence of treatment-related adverse events across all PD 0348292 dose groups ranged from 11.5% to 20.0%, compared with 15.9% in the enoxaparin group. The incidence of adverse events did not exhibit a dose-dependent trend among subjects treated with PD 0348292. Treatment-related serious adverse events with PD 0348292 occurred only in the 0.3 mg, 0.5 mg and 10 mg dose groups (in 1.4%, 1.1% and 3.1% of subjects, respectively). In the enoxaparin group, 0.8% of subjects reported treatment-related serious adverse events.

Discussion

Critical to the success of a new anticoagulant is finding a dose that provides the appropriate balance between efficacy and safety. Given the potentially serious consequences of under- or overdosing, Phase 2 dose-finding studies of anticoagulants present special challenges. Published results of conventional Phase 2 studies of marketed [10] and investigational [11] FXa inhibitors for prophylaxis of VTE following total hip or knee replacement have demonstrated that quite often doses found to be well tolerated in Phase 1 studies [12, 13] must be discontinued during Phase 2 studies because of an unacceptably high risk of major bleeding. Thus, conventional Phase 2 studies that base dosing upon the maximum tolerated dose in Phase 1 may predispose subjects in the study to increased risk of major bleeding and premature study withdrawal, resulting in insufficient data to allow identification of the lowest dose that is both safe and effective.

An adaptive design was chosen for this Phase 2 dose-finding study of PD 0348292 in order to maximize the probability of finding a dose with efficacy and safety equivalent to that of enoxaparin 30 mg bid, while at the same time minimizing the number of subjects requiring treatment exposure as well as the risk to subjects of either VTE or bleeding. Extensive dose-response modeling and simulation were used to leverage prior knowledge of other anticoagulant compounds as well as of PD 0348292 and support the dose selection and study design [9]. An important factor in the choice of study doses was how well the prespecified rules for dropping or adding doses (DDAs) were expected to protect subjects from excessive VTE and major bleeding. The prespecified rules were determined using clinical trial simulations that estimated the overall incidence of VTE and major bleeding for PD 0348292 vs. that for enoxaparin 30 mg bid. The results indicated that the overall risk to subjects from exposure to PD 0348292 using an adaptive design that incorporated these rules was expected to be no greater than that with enoxaparin.

The efficacy of PD 0348292 doses was benchmarked against enoxaparin 30 mg bid administered 12–24 h post-surgery, which has been used as the active control in several recent Phase 2 and Phase 3 TKR trials [14-18]. The observed incidence of total VTE, 18.1%, for enoxaparin 30 mg bid in this study was comparable to that observed in the Phase 2 dose-ranging studies of fondaparinux [14], ximelagatran [15] and apixaban (ranged from 16.5% to 27.8%) [17], but was considerably lower than that reported in the ODIXa-knee study (44.3%) [16]. The incidence of proximal DVT + PE for enoxaparin in this study (3.7%) was comparable to that across all these studies (~3% to ~6%).

In conclusion, PD 0348292 and enoxaparin were well tolerated in this study. The adaptive design enabled comparison of a 100-fold dose range of PD 0348292 with enoxaparin 30 mg bid in TKR surgical subjects. Through dose-response modeling of the efficacy data, the dose of PD 0348292 equivalent to enoxaparin 30 mg bid for the prevention of total VTE in TKR subjects was estimated with good precision while at the same time minimizing both the number of subjects requiring treatment exposure as well as the risk to subjects of either VTE or bleeding. Although this compound did not progress beyond Phase 2 after co-development of another FXa inhibitor was initiated, characterization of the dose-response relationship for both efficacy and bleeding would have provided a strong quantitative basis for Phase 3 dose selection.

Addendum

A. T. Cohen performed the study, analyzed and interpreted data, and critically reviewed the manuscript. R. A. Boyd designed the study, analyzed and interpreted data, and wrote the manuscript. J. Mandema designed the study, analyzed and interpreted data, and critically reviewed the manuscript. L. DiCarlo designed the study, analyzed and interpreted data, and critically reviewed the manuscript. R. Pak designed the study, analyzed and interpreted data, and wrote the manuscript. All authors gave their consent to the final version of the manuscript.

Acknowledgements

The authors would like to thank all participating patients and all study nurses for their work on enrolling patients. Statistical analysis was supported by S. McBride, who was a Pfizer employee at the time this study was conducted.

Funding

The study was supported by funding from Pfizer, Inc.

Disclosures of Conflict of Interest

A. T. Cohen received honoraria as an investigator. R. A. Boyd and R. Pak are employees of Pfizer, Inc. L. DiCarlo was also a Pfizer employee when this study was conducted. J. Mandema is an employee of Quantitative Solutions and was a paid consultant to Pfizer, Inc. for dose-response modeling and in connection with the development of this article.

Appendix

Study Investigators. Australia: R. Baker, B. Chong, A. Gallus, E. Gan, H. Salem, C. Ward. Canada: D.R. Armstrong, L.K. Bredo, M.A. Crowther, C. Demers, G. Dervin, J. Kassis, W.A. Pisesky, D. Puskas, D. Stevens, M.E. Zarnett. Chile: S. Bittelman. Colombia: G. E. Gomez, A. L.Solano, P. J. Llinas, R. Botero. Czech Republic: J. Deniger, M. Fila, V. Karjagin, K. Karpas, I. Kofranek, J. Kubes, M. Pink, T. Pink, J. Sedivy, O. Stano, T. Trc. Denmark: P. H. Gebuhr, N. K. Jensen, P. S. Joergensen, S. Mikkelsen, C. H. Ryge, S. A. Schmidt. France: P. Coriat, E. Gaertner, Y. Hemon. Italy: B. Borghi, F. M. Franchin. Poland: J. Deszczynski, T. Gadzik, K. Kwiatkowski, J. Skowronski, A. Wall, J. Widuchowski, Z. Wieczorek. Portugal: A. Freitas, F. Judas, C. Ribeiro. Russian Federation: I. Belenkiy, N. N. Kornilov, G. Kuropatkin, V.V.Yakusevich. Slovakia: G. Vasko. South Africa: C. T. Frey, B. F. Jacobson, C. Spargo. Spain: L. L. Duran, I. Gomez, P. Hinarejos, A. C. Munoz, A. N. Quilis, L. Peidro. United Kingdom: A. T. Cohen, R. E. Field. United States: D. G. Bramlet, W. J. Bryan, S. Burton, A. Corces, R. J. Friedman, D. C. Han, P. B. Hanson, G. S. Gill, M. J. Jazayeri, R. R. King, L. M. Kwong, G. C. Landon, J. M. Lipke, K. A. Martin, H. S. Minkowitz, H. B. Morris, J. Muntz, R. M. Murray, J. Papilion, G. S. Patel, P. C. Peters, Jr., H. F. Riegler, D. R. Schmidt, J. R. Schwappach, S. K. Singla, M. J. Skyhar, S. L. Sledge, A. I. Spitzer, F. A. Stanziola, R. E. Strain, Jr., R. A. Wagner, J. Ward.

Ancillary