THromboprophylaxis In Sickle Cell Disease with central venous catheters (THIS): an internal pilot randomised controlled trial protocol

Abstract

Introduction

Individuals with sickle cell disease (SCD) and central venous catheters (CVC) are at high risk for venous thromboembolism (VTE). Minimal data exist regarding the use of anticoagulation as thromboprophylaxis of VTE in this demographic, and as a result, clinical equipoise exists. Prophylactic dose rivaroxaban, a direct oral anticoagulant, is efficacious and safe as thromboprophylaxis in other demographics, and may be an optimal agent in SCD with CVC. Prior to conducting a full clinical trial to assess rivaroxaban as thromboprophylaxis in SCD with CVC, a pilot study is needed to gauge its feasibility.

Methods and analysis

THromboprophylaxis In Sickle Cell Disease pilot trial is an investigator-initiated, multicentre, double-blinded, randomised controlled trial (RCT) assessing if it is feasible and safe to conduct an adequately powered RCT comparing rivaroxaban to matching placebo as thromboprophylaxis in those with SCD and CVC. Fifty adult patients with SCD and CVC will be randomised to receive either rivaroxaban 10 mg daily or matching placebo for the duration of the CVC in situ for up to 1 year. After randomisation, follow-up visits will occur every 3 months. The primary outcomes pertain to the feasibility of a full trial and include numbers of eligible and recruited participants. Exploratory outcomes include overall incidence of VTE and bleeding complications, as well as quality of life. If the full trial is feasible, blinding will be maintained and patients in the pilot study will be included in the full trial.

Ethics and dissemination

The trial was initially approved by the University Health Network Research Ethics Board (REB) in Toronto, Canada. All sites will obtain approval from their respective REB prior to commencement of study activities. Study results will be disseminated through presentations at medical conferences and peer-reviewed publications.

Trial registration number

NCT05033314.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Investigator-initiated, multicentre, double-blinded, randomised internal pilot clinical trial.

• Data are urgently needed to resolve clinical equipoise regarding the use of anticoagulation as thromboprophylaxis in those with sickle cell disease and central venous catheters.

• The pilot feasibility study alone cannot provide definitive evidence for the efficacy and safety of rivaroxaban as thromboprophylaxis in those with sickle cell disease and central venous catheters.

• Conducting a clinical trial in the sickle cell disease population may be challenging as previous studies have been limited by low enrolment and high loss to follow-up.

Introduction

Sickle cell disease (SCD) is considered to be a prothrombotic state. The sickling and haemolysis of red blood cells cause ischaemia-reperfusion injuries, oxidative stress and inflammation which in turn lead to activation of the coagulation cascade and platelets, damage to the vascular endothelium and chronic vasoconstriction with decreased blood flow.¹ In addition to these intrinsic processes, individuals with SCD can be exposed to external risk factors that further increase the risk of venous thromboembolism (VTE),² such as central venous catheters (CVC), which are often required for exchange transfusion, difficult intravenous access or long-term antibiotics. The risk of VTE with CVC in patients with SCD is estimated to be as high as 1 event per 1000 catheter days.² The consequences of VTE put these patients at risk for: (1) major and fatal bleeding with therapeutic anticoagulation treatment; (2) permanent loss of vascular access at a venous access site where a VTE occurred, eliminating means by which these patients can be treated; (3) long-term morbidity, including post-thrombotic syndrome and chronic thromboembolic pulmonary hypertension and (4) mortality, with a hazard of death 2.5-fold that of an individual with SCD without VTE.³

It is possible that prophylactic anticoagulation can reduce the incidence of VTE in those with SCD and CVC. Rivaroxaban, a direct oral anticoagulant (DOAC), is administered at fixed doses without need for routine monitoring. In large randomised controlled trials (RCTs), low-dose rivaroxaban has proven to be efficacious and safe in decreasing the incidence of VTE as primary prevention in posthip arthroplasty and high-risk cancer populations and secondary prevention in those with prior VTE.^4–6 While those with SCD and CVC are well acknowledged to have a high VTE risk, due to limited data, guidelines regarding thromboprophylaxis in this group are ambivalent or absent and clinical equipoise is present.^7–9 In ambulatory cancer patients, who are similarly at high risk for VTE, those with a VTE risk of 10% or greater at 6 months are considered for thromboprophylaxis.¹⁰ There remains strong potential for clinical benefit with thromboprophylaxis in SCD with CVC, as a recent retrospective multicentre observational study found lack of thromboprophylaxis compared with thromboprophylaxis with any antithrombotic agent, was associated with a rate ratio for VTE of 4.0 (95% CI 1.2 to 12.6).¹¹ However, there is also potential for harm, as SCD also has a predisposition to bleeding, including haemorrhagic stroke, vitreous haemorrhage and haematuria. In a population-based study, individuals with SCD had a cumulative incidence of bleeding requiring emergency department or hospitalisations of 21% (95% CI 19.8% to 22.3%) by age 40.¹²

To assess the efficacy of rivaroxaban as thromboprophylaxis in patients with SCD with CVC, if the untreated rate of VTE is 0.21 per year, a trial with 3 years of accrual and 2 years of additional follow-up would need 181 participants to have 80% power to detect an HR of 0.5 with treatment, using alpha of 0.05 and assuming 3.5% annual loss to follow-up. However, SCD is an orphan disease and as such, a trial of this size would represent a large commitment of resources. Given previous concerns about slow enrolment and high loss to follow-up in trials in SCD, we propose a pilot trial to determine the feasibility of this full-scale RCT.

Methods and analysis

Study overview

THromboprophylaxis In Sickle Cell Disease with CVC (THIS) is an investigator-initiated pilot, double-blinded, multicentre RCT to assess if it is feasible and safe to conduct an adequately powered RCT comparing rivaroxaban to matching placebo as thromboprophylaxis in adult SCD participants with CVC. The study is funded by the Paul B. Helliwell Foundation Hematology Research Innovation Fund and is registered on clinicaltrials.gov (NCT05033314).

The trial will be conducted at three adult centres in Canada: the University Health Network (UHN) in Toronto, Ontario, the University of Alberta Hospital in Edmonton, Alberta, and the London Health Sciences Centre in London, Ontario. Drs. Jameel Abdulrehman and Kevin H.M. Kuo are coprincipal investigators. The trial study will be conducted as an internal pilot study, in that if the full trial is feasible, patients in the pilot study will continue to be followed, their treatment assignment will remain concealed, and they will be included in analyses of the full study.

Participants

The trial will include adult participants with documented SCD, with new or pre-existing CVC planned for long-term (at least 6 months) use, who are able to provide consent. Participants will be excluded for previous VTE within the past 3 months, current use of anticoagulation for an indication other than CVC thromboprophylaxis, and concerns pertaining to potential safety risks of anticoagulation with rivaroxaban such as concomitant dual antiplatelet therapy, creatinine clearance <30 mL/min (Cockcroft-Gault equation), or cirrhosis with Child-Pugh score B or C. Given the limited data for bleeding risk associated with rivaroxaban 10 mg daily in participants with SCD, the exclusion of participants with additional risk factors for bleeding, such as moyamoya, severe vasculopathy or routine non-steroidal anti-inflammatory drugs (NSAID) use will be at the discretion of the treating physician based on their holistic bleeding risk assessment. Please see figure 1 for full list of exclusion criteria. Females of childbearing potential must agree to avoid pregnancy during use of study drug and for 30 days after, either through non-oestrogen-based contraception or abstinence.

Figure 1.

THIS pilot study flow diagram. CVC, central venous catheter; SCD, sickle cell disease; THIS, THromboprophylaxis In Sickle Cell Disease; VTE, venous thromboembolism.

Interventions

Recruited participants will be block-randomised with 1:1 allocation ratio, stratified by type of CVC (peripherally inserted central catheter or other) and whether the CVC is new or pre-existing, to either daily oral rivaroxaban 10 mg or matching placebo for the duration of the CVC in situ, for up to 1 year. A centralised blinded study coordinator will have a document with the randomisation sequences generated by the study’s statistician comprising a list of alphanumeric codes. The coordinator will receive emails from the respective sites requesting randomisation, and a return email will include the next available alphanumeric code to provide to the investigational pharmacy. The investigational or clinical trials’ pharmacist will have a list which connects each code to the treatment arm assignment and will dispense the patient’s investigational product accordingly. Investigational pharmacies at the respective participating centres will package and dispense study drug and placebo in identical containers. Study participants, treating physicians, study investigators, adjudicators and outcome assessors will be unaware of treatment arm assignment.

Rivaroxaban was chosen as the active intervention for the following reasons: it is a once-a-day fixed dose oral medication without the need for routine monitoring, it has a rapid onset of effect and a short half-life and was found to be efficacious at the low dose of 10 mg daily as primary prevention in the setting of posthip arthroplasty and high-risk patients with cancer and secondary prevention in those with prior VTE.4–6 13 14 Rivaroxaban decreased the risk of VTE with absolute risk reduction of 1.7% (95% CI 1.0% to 2.5%) compared with enoxaparin in posthip arthroplasty participants, with HR of 0.40 (95% CI 0.20 to 0.80) compared with placebo in high-risk cancer participants, and with HR of 0.26 (95% CI 0.14 to 0.47) compared with aspirin 81 mg daily in participants with prior VTE.^4–6 Compared with traditional oral anticoagulation with vitamin K antagonists (VKAs), the DOACs have a lower risk of major, fatal, intracranial, clinically relevant non-major and total bleeding.¹⁵ Compared with other DOACs, apixaban has twice daily dosing, which may decrease adherence, and less information is available regarding the use of dabigatran and edoxaban as thromboprophylaxis.¹⁶ Despite the effectiveness of low-molecular-weight heparin in decreasing the incidence of CVC-related VTE in patients with cancer, it was not selected as the intervention as it is expensive, burdensome and uncomfortable to administer as a subcutaneous injection.¹⁷

Placebo was chosen as the control arm rather than active control since there is no approved drug for primary thromboprophylaxis in patients with SCD with CVC and there is clinical equipoise as to whether anticoagulation should be used at all for this indication.^18–22 Furthermore, there is no universal standard of care as no clinical trials have assessed the efficacy and safety of anticoagulation compared with no anticoagulation as thromboprophylaxis in this demographic.⁷ The study placebo is manufactured by the Toronto Institute of Pharmaceutical Technology (Toronto, Ontario, Canada).

Outcomes of the trial

The primary outcomes of the pilot trial are the following:

1. Enrolment of an eligible patient.

2. Indications for CVC (chronic transfusion therapy, poor venous access, chronic antibiotics or other).

3. Duration of CVC insertion prior to enrolment.

4. Adherence to the study drug (percentage of administered drugs not returned, using pill counts).

5. Adherence to study procedures (attend all required study visits).

6. Loss to follow-up over the year after randomisation.

If the pilot trial establishes feasibility, the following outcomes will be summarised in aggregate. If the full trial is not feasible, these outcomes will be compared between study arms.

1. VTE (including catheter-related thrombosis (CRT), right atrial or ventricular thrombus, proximal deep vein thrombosis (DVT), segmental pulmonary embolus (PE), unusual site VTE and unexplained death in which PE could not be ruled out).

2. Major bleeding or clinically relevant non-major bleeding (CRNMB) (defined as per International Society on Thrombosis and Haemostasis (ISTH) criteria).²³

3. Major bleeding (defined as per ISTH criteria).²³

4. CRNMB, defined as per ISTH criteria.²³

5. One-year event-free survival (defined as free from VTE or major bleeding events).

6. Arterial thrombotic events (ATE) (including myocardial infarction, ischaemic stroke (both symptomatic and silent) and systemic embolism).

7. CVC change due to thrombosis or catheter occlusion not amenable to infusion of thrombolytics.

8. Quality of life using the modified Duke Anticoagulation Satisfaction Scale (see online supplemental figure 1).²⁴

9. The extent of bruising using a Likert scale (see online supplemental figure 2).

Description of trial procedures

Participants in the pilot study will be followed for up to 12 months while the CVC remains in situ with assessments at baseline, and months 3, 6 and 12 and a safety visit at 30 days±7 days (table 1). Participants are free to withdraw their consent and discontinue study drug and withdraw from follow-up at any time without giving a reason. The investigator will also have the right to discontinue study drug in the event of illness or other reasons concerning the health or well-being of the participant, or in case of lack of cooperation with study procedures. Should a participant discontinue study drug early, an early discontinuation visit takes place.

Table 1.

This pilot trial: schedule of events

	Screening	Dosing	Months 3, 6	Month 9	Month 12 (EOS)	Safety visit	Early discontinuation visit
	Day −7 to −1	Day 1	±15 days	±15 days	±15 days	(30±7 days after last dose)
Informed consent	X	-	-	-	-	-	-
Inclusion/exclusion criteria and eligibility	X	-	-	-	-	-	-
Baseline demographics	X	-	-	-	-	-	-
Concurrent medications	X	-	X	-	X	-	X
Weight	X	-	X	-	X	-	X
Complete blood count	X	-	X	-	X	-	X
Creatinine, AST ALT	X	-	X	-	X	-	X
Beta-hCG	X	-	-	-	-	-	-
Drug dispensation	-	X	X	X	-	-	-
Adherence assessment	-	-	X	X	X	-	X
HrQOL assessment	-	-	X	-	X	-	X
AE monitoring	-	-	X	-	X	X	X

AE, adverse event; ALT, alanine transaminase; AST, aspartate aminotransferase; EOS, end of study; hCG, human chorionic gonadotropin; HrQOL, health-related quality of life.

Study participants will be requested to monitor and record any new symptoms or perceived adverse events (AEs) and will be screened for AE at follow-up visits at months 3, 6 and 12, the safety visit and the early discontinuation visit. AE is defined as any undesirable event occurring to a participant during the study, whether or not it is considered to be related to study drug. All AE will be assigned a severity based on the Common Terminology Criteria for Adverse Events version 5.0, and an association to the study drug as determined by study investigators.²⁵

The study database will use Medidata software and will be managed by Ozmosis Research. Data will be collected and entered into the database by study staff at the respective sites through the use of electronic case report forms. Ozmosis Research will be responsible for data monitoring and queries will be resolved by the staff at the respective sites. Once data collection is complete, the database will be locked for statistical analysis.

An independent data safety and monitoring board (DSMB) made up of 3 members with appropriate expertise (two haematologists and a biostatistician) will be appointed to oversee the safety of trial participants and meet as required per the DSMB charter. The DSMB will review AEs trends and make recommendations regarding continuation, modification, suspension or termination of the study.

Assessment of outcomes

Primary and secondary outcomes

Primary outcomes related to enrolment will be recorded by study personnel at time of referral and recruitment. Other primary outcomes and all secondary outcomes will be captured by study personnel at regularly scheduled follow-up visits.

VTE outcomes

CRT, proximal DVT and unusual site VTE must be confirmed on either Doppler ultrasound demonstrating a non-compressible vein, or alternatively CT venography or MRI venography demonstrating an intraluminal filling defect. Right atrial or ventricular thrombus must be confirmed on either MRI, CT scan or echocardiography. PE must be confirmed on CT pulmonary angiography demonstrating a filling defect in a segmental or larger vessel, V/Q scan demonstrating a high probability of a PE in patients with non-low clinical probability of PE, or pulmonary angiography demonstrating in intraluminal filling defect in two or more views, or a new sudden contrast cut-off of one or more vessels of a diameter more than 2.5 mm. Unexplained death in which PE cannot be ruled out must have either no documented attributable cause to death and PE cannot be ruled out, or PE has to be diagnosed by autopsy.

A committee composed of VTE experts blinded to study drug assignment will adjudicate VTE outcomes based on clinical reports, laboratory data and diagnostic imaging reports. Each outcome must be adjudicated by two adjudicators who must reach consensus. If consensus is not reached, a third adjudicator will be brought in to reach a majority decision. This method of using clinical adjudicators in the assessment of VTE endpoints has been found to be reliable and is used in most large VTE trials.^{13 26–29}

Bleeding outcome

Major and CRNMB is defined by the ISTH.²³ A committee composed of VTE experts blinded to study drug assignment will adjudicate bleeding outcomes based on clinical reports, laboratory data and diagnostic imaging reports. Each outcome must be adjudicated by two adjudicators who must reach consensus. If consensus is not reached, a third adjudicator will be brought in to reach a majority decision.

Sample size and power calculations

The aim is to enrol 50 patients in the pilot trial. The expected number of patients that will need to be assessed to identify 50 patients ranges from 63 to 84 across participation percentages of 60%–80%. The resulting percentage of participants can be estimated with a 95% CI of approximately ±10%. The estimate from the pilot will be applied to patient counts at potential future study sites to (a) assess feasibility of randomisation of approximately 200 patients across all sites and (b) calculate the number of sites and duration of accrual needed for the full trial.

In a meta-analysis of studies by Jeng et al,² Abdul-Rauf et al,³⁰ Phillips et al³¹ and Shah et al,³² the estimated pooled rate of VTE events is 0.21 per year (corresponding to a median time to thrombosis without prophylaxis of 3.3 years³³). Combining 50 patients from the pilot (with a maximum of 1 year of follow-up) with an additional 160 patients in the full trial (with 3 years of accrual and 2 years of follow-up) gives 80% power to detect a HR for treatment of 0.5, using an alpha of 0.05 and an assumed 3.5% loss to follow-up per year. The estimated baseline rate will be revised based on any new data available before the full trial commences, in particular the estimated baseline rate from the pilot, calculated from the blinded pilot data. The assumed drop-out rate will also be updated with pilot trial data.

Statistical analysis

Analyses of the pilot study data will involve estimation, rather than hypothesis testing.

Baseline demographics

All demographics and baseline characteristics will be summarised using descriptive statistics as applicable. The baseline characteristics of the study drug arm and the placebo arm will be presented together. If a full study is deemed not feasible, the baseline characteristics of the study drug arm and the placebo arm will be presented separately.

Primary and secondary endpoints (feasibility)

Binary feasibility outcomes will be summarised using frequency counts and percentages with associated 95% CIs. Percentage drug adherence and duration of CVC use prior to enrolment will be summarised by medians and IQRs, and also categorised, at a threshold of 80% for adherence and as new versus pre-existing for CVC.

The pilot trial will be deemed to have demonstrated feasibility success and will proceed to a full trial under the current protocol if all the criteria below are met (based on point estimates):

• Participant recruitment is >60%.

• Participant adherence to study drug >80%.

• Compliance with study procedures >80%.

• Proportion of participants loss to follow-up <10%.

The feasibility trial will be deemed a success, but further modifications to the protocol will be required before proceeding to a full trial if outcomes meet a combination of the criteria above and the following criteria:

• Participant recruitment is 40%–60%.

• Participant adherence to study drug 60%–80%.

• Compliance with study procedures 60%–80%.

• Proportion of participants loss to follow-up 10%–25%.

The feasibility trial will not be deemed successful, and there will be no plans to pursue a full trial if two or more of the following criteria are met.

• Participant recruitment is <40%.

• Participant adherence to study drug <60%.

• Compliance with study procedures <60%.

• Proportion of participants loss to follow-up >25%.

Secondary outcomes

The full trial is designed to show the superiority of study drug over placebo in decreasing the incidence of composite VTE events. However, for the pilot study, to maintain blinding of the full trial and because of low power for showing differences in clinical endpoints in the pilot, the clinical endpoints will only be presented using descriptive statistics, aggregating both study arms. Only if a full study is deemed not feasible will the outcomes of the study drug arm and the placebo arm will be presented separately for the pilot study. In that case, the VTE, bleeding and event-free outcomes will be compared between the study drug arm and the placebo arm using a stratified log-rank test. In addition, for both arms combined, Kaplan-Meier survival curves will assess the composite VTE events, the composite of major and CRNMB, and event-free time from time of randomisation to event. Study participants who do not have a VTE or bleeding outcome prior to discontinuation of CVC will be censored at time of removal of CVC. Participants who withdraw consent, are loss to follow-up or died from reasons other than composite VTE, who did not have a composite VTE event will be censored at the last day the participant was assessed for outcomes. Quality of life outcomes will be summarised using descriptive statistics.

Planned subgroup analyses

No subgroup analyses are planned for this pilot study.

The proposed frequency of analyses

No formal interim analysis is planned for this pilot study.

Success of the feasibility trial

If the feasibility trial is deemed a success, a full trial will be pursued. Study participants will be enrolled in the full trial, and recruitment of new study participants will resume. The rules above were adopted as the criteria to decide whether to proceed to the full trial.

Patient and public involvement

Patients and the public were not involved in the development of the research question, study design, study recruitment or study conduct. The results of the study will be disseminated to study participants and the public through SCD patient advocacy groups.

Ethics and dissemination

The protocol was approved by the UHN Research Ethics Board (REB) (Study ID 20-5846.0). Prior to commencement of study activities, each site will require approval by their respective REB. Written informed consent will be sought from all study participants. Study recruitment began at the primary site at the UHN in Toronto in May 2022. Study results will be disseminated through presentations at medical conferences and peer-reviewed publications.

Discussion

As the population of individuals with SCD and CVC increases, data are urgently needed to guide management decisions regarding thromboprophylaxis. In developed countries, the prevalence of SCD is increasing due to an increase in immigration from endemic countries, and improving mortality associated with advancements in supportive care.⁷ Only 60 years ago, SCD was thought of as a ‘disease of childhood’ as few would survive to adulthood, but today in well-resourced countries, individuals with SCD can live well into their 50s and 60s.^{34 35} Furthermore, the use of CVCs in those with SCD is likely to increase as studies continue to demonstrate the benefit of chronic transfusion therapy for prevention of stroke and end-organ damage.^36–39

A considerable challenge in conducting a trial in patients with SCD is the risk of low enrolment and poor follow-up. A prior RCT, investigating the effect of VKA compared with placebo on pulmonary artery systolic pressure in patients with SCD with mild pulmonary hypertension, with planned follow-up of 12 months, was terminated early due to poor enrolment (NCT01036802). While the study targeted a total of 20 patients, from December 2009 to September 2012, only 3 patients were successfully enrolled. To maximise enrolment in our study and increase generalisability, we are enrolling at multiple sites, and we have broad inclusion criteria to include all types of CVCs, whether new or pre-existing, and for all indications. Furthermore, patients may be more likely to participate in an anticoagulation study with a DOAC rather than a VKA given its greater convenience and safety profile.¹⁵ A more recent RCT in patients with SCD assessing the effect intravenous crizanlizumab compared with placebo on sickle cell-related pain crises, with 12-month treatment phase, was successfully completed, but had a high rate of early study discontinuation.⁴⁰ Of the 198 patients randomised to study drug, 69 (35%) failed to complete the study. Nonetheless, we expect that our rate of early study discontinuation will be much lower, as our study evaluates an oral medication rather than an intravenous one, and the majority of our study demographic is estimated to be dependent on chronic transfusion therapy which requires regular in-person visits, making study follow-up more likely. However, to accurately gauge if a full trial is feasible, the proposed pilot study is necessary.

Compared with a full trial, pilot studies are reduced in scope and size, and although they can provide insight into the efficacy and safety of an intervention, their findings are rarely conclusive.⁴¹ Rather, the goals of pilot studies are to evaluate trial processes, financial and time resources, data management and scientific inquiries on a smaller scale, prior to committing to an expensive and resource intensive full clinical trial.⁴² THIS pilot study will allow us to accurately estimate participant recruitment, participant adherence to study drug, compliance with study procedures and the proportion of participants loss to follow-up, and if modifications are required to the study protocol for the full trial to be successful. Moreover, the knowledge gained from the pilot study trial will help decide which centres would be able to recruit participants, and how many centres would be required to successfully complete the trial.

In conclusion, THIS pilot study will evaluate the feasibility to conduct an adequately powered full RCT comparing rivaroxaban to placebo as thromboprophylaxis in patients with SCD with CVCs. Individuals with SCD and CVC are at high risk of VTE, but little data exist to guide if thromboprophylaxis should be considered. The full trial could provide strong evidence to support or refute the use of low dose rivaroxaban as thromboprophylaxis in this demographic and solve the clinical equipoise.

Ethics statements

Patient consent for publication

Not applicable.