2.1 Search Strategy

An electronic search was conducted in the following databases on 31 May 2023: MEDLINE (to present) and MEDLINE In-Process Citations, through Pubmed.com interface; Cochrane Library (to present): including Central Register of Controlled Trials (CENTRAL) and Database of Systematic Reviews (CDSR) according to PRISMA. For details of the search strategy see Tables S1–S6.

2.2 Study Selection

Studies reporting efficacy, safety or consumption/dosing of pdVWF/FVIII in inherited VWD of all types, in patients of all ages who received treatment with any treatment regimen (OD, intermittent, LTP and SP) were included. More details on the inclusion/exclusion criteria are described in Table S1. Results from evidence synthesis were presented as ranges. Where haemostatic efficacy data were comparable, these were pooled across studies for OD, LTP and SP.

2.3 Pharmacovigilance Data

Post-marketing surveillance (PMS) data were collected for the period 7 June 1982–31 May 2023. Only adverse events (AEs) where a causal relationship between pdVWF/FVIII [7] and the reported AE was considered possible (adverse drug reactions [ADRs]) were included. Historically, pdVWF/FVIII was used to treat patients with haemophilia A, however, in recent years it is primarily used in patients with VWD [8]. No distinction was made between ADRs reported for either VWD or haemophilia A indications during PMS reporting, hence the pharmacovigilance data includes ADRs for both indications (for other pharmacovigilance methodology please refer to Supplemental Materials).

3.1 Systematic Review

From a total of 164 records identified, 144 were included in the full-text screening level. Of these 144, 124 articles had exclusion criteria, resulting in the inclusion of 20 articles. Additionally, 28 records were identified in the grey literature review (Figure 1). Following eligibility screening, 18 records reporting data from 15 individual studies were reviewed (Table S7).

Characteristics of the population in these 15 studies are presented in Table 1. Most studies included both paediatric (<18 years) and adult (≥18 years) patients: one study only paediatric [10], and three only adult patients [11-13]. One study reported evidence in women with VWD admitted for childbirth [13]. For each study, a breakdown of number of patients receiving OD treatment, LTP and SP, has been included if available. For some studies, patients receiving OD or LTP treatment may have also been included in the SP population subgroup if they also underwent surgeries/invasive procedures.

Treatment protocols differed between studies; in the interventional studies, protocols ranged from conventional dosing regimens (per product labelling) [7] to tailored pharmacokinetic (PK)-guided dosing based on VWF levels. Data are reported as median or mean with corresponding ranges in the tables.

3.2 OD Treatment With pdVWF/FVIII

Five studies evaluated the efficacy of pdVWF/FVIII for OD treatment of bleeding events (Table 2) [11, 14–18]. Four studies reported bleeding pattern data, with epistaxis and GI bleeds as the main causes for OD treatment [11, 14, 15, 17]. Overall, the control of bleeding events with pasteurized pdVWF/FVIII was rated excellent/good in 95%–98% of treated bleeds, with only a few as a moderate/poor response [14, 15, 17, 18]. The latter cases tended to be patients with type 3 and type 2B VWD (Supplemental Material). Pooled haemostatic efficacy data for 1351 treated bleeds from 264 patients across four studies (Figure 2A), showed that 95.7% of bleeds (N = 1293) resolved with excellent/good efficacy and 3.9% (N = 53) with moderate/poor efficacy or no response [14, 15, 17, 18].

Five studies reported dosing and consumption data for pdVWF/FVIII for OD treatment (Table S8) [11, 14–17]. In the four observational studies, pdVWF/FVIII treatment posology was at the discretion of the treating physician and varied greatly, making comparison difficult [11, 14–16].

Five studies assessed the safety of pdVWF/FVIII for OD treatment, although the outcomes of interest varied between studies (Table 3) [11, 14–17]. The methodologies varied in terms of the definition of AE and ADR and duration of follow-up [16]. Two of the studies reported no AEs [11, 15], whereas Gill et al. reported a single possibly related AE of a mild allergic reaction, and two serious adverse events (SAEs) considered unrelated to study treatment; one menorrhagia and anaemia, and one haemorrhage [17].

Thromboembolic event (TE) incidence was the safety outcome most frequently assessed, with all five studies reporting the absence of any TEs following pdVWF/FVIII treatment [11, 14-17]. FVIII or VWF inhibitor formation was only investigated in two studies [11, 14] with no cases being reported [11, 14].

3.3 LTP With pdVWF/FVIII

For efficacy of pdVWF/FVIII in LTP treatment, only studies reporting efficacy outcomes as changes in annualized bleeding rate (ABR) or efficacy rated by the treating physician, were evaluated in the analysis (Table 2) [15, 16, 18–20]. In particular, Abshire et al. reported outcome data for patients who were treated prophylactically with a variety of VWF/FVIII concentrates, with pdVWF/FVIII being the most frequently used (77% of patients) [20]. Therefore, data interpretation/comparison may warrant caution.

The most frequent indications for prophylactic treatment initiation were joint bleeds, GI bleeds and epistaxis [15, 19, 20].

Two studies reported haemostatic efficacy for LTP rated per cycle by the treating physician [15, 16], with various definitions of rating categories and units of analysis. In the study by Federici et al., although the duration of a cycle was not defined, patients were treated with multiple LTP sessions (17 cycles per 12 patients; n = 9 type 3 VWD) [15]. Overall rating of excellent/good was reported in 100% (37/37) of prophylactic cycles as assessed by Federici et al. [15] and Lillicrap et al. [16]. Efficacy of pdVWF/FVIII was reported in Lethagen et al. as reduced ABRs, with a median of 11 and one bleeding event(s) before and after prophylaxis initiation, respectively [19]. ABR data were stratified per bleeding type in the study by Abshire et al. with the most pronounced reduction in ABR for joint bleedings (AjBR); AjBR was 15.6 before and 1.3 during prophylaxis [20].

Four studies included dosing and consumption data of pdVWF/FVIII during LTP (Table S8) [14–16, 20]. Data were not homogeneously reported, with only two studies providing a median treatment duration per patient of 201 and 803 days [15, 20]. An LTP regimen of 2–3 weekly infusions was reported in three studies [14, 15, 20], with one also specifying dosing of 20 IU/kg FVIII in about 90% of cases [14]. In one of the studies, the median number of weekly infusions was reported per prophylactic indication, with epistaxis, GI bleeds and menorrhagia as indications with the highest median of three infusions [20].

Six studies evaluated the safety of pdVWF/FVIII in LTP (Table 3), with limited data on outcomes of interest and varying exposure time and safety assessment methodology. Two studies [11, 15] observed no AEs during follow-up, and another reported no treatment-related AEs [14]. SAEs or TEs did not occur during patient follow-up in any of the studies. Lethagen et al. reported one case of antibodies to VWF in a patient with type 3 VWD during LTP treatment with pdVWF/FVIII [19]. In another study, antibodies were detected in one patient who then discontinued treatment [20]. No cases of VWF or FVIII inhibitors were reported in the studies by Castaman et al. [14] and Miesbach et al. [11].

3.4 SP With pdVWF/FVIII

Thirteen studies evaluated the efficacy of pdVWF/FVIII in SP (Table 4) with various definitions of major surgery. All but one study reported the types of procedure [16, 18]. The study by Carcao et al. [10] included only dental procedures, whilst the remaining studies included major (6%–74% of all procedures), minor (9%–57%), minimally invasive (12%–26%) and dental procedures (2%–52%). The proportion of procedures for which the overall haemostatic efficacy was rated excellent/good, ranged from 94% to 100%. Pooled haemostatic efficacy rating for 420 procedures in 339 patients across eight studies showed that 98.1% of surgical bleeds (N = 412) were resolved with excellent/good efficacy and 1.4% (N = 6) had moderate/poor efficacy or no response (Figure 2C) [14, 15, 21–26]. One study [16, 18] could not be included in the pooled analysis as the number of patients receiving specific treatment regimens were not reported. One study reported the expected intra-operative blood loss, and all procedures had less than the expected volume according to the surgeon [25].

Twelve studies reported dosing and consumption data for pdVWF/FVIII in SP (Table S9). Treatment protocols and variables of interest varied across the studies. Three studies reported the median number of infusions, ranging from 3 to 6 administrations per procedure [14, 15, 24]. Stratified data per procedure type were reported in Federici et al. [15], Lethagen et al. [26], Franchini et al. [23] and the CSL-4002 study [25], with a median of 7–10 infusions and 3–9.5 infusions in major and minor procedures, respectively. Median treatment duration ranged between 2.0 and 6.4 days [16, 21, 24, 26], varying according to procedure type (see more details in Supplemental Materials) [15].

Twelve studies evaluated the safety of pdVWF/FVIII for SP (Table 3). Both CSL-4001 and -4002 trials reported similar safety evaluations, including AE monitoring during the surgery phase, from loading dose to post-operative day 14 or, until the end of treatment [25, 26]. In the CSL-4001 and -4002 trials, 75% (n = 21) and 86% (n = 30) of patients reported at least one AE, of which 19% (n = 5) and 9% (n = 3) were considered treatment-related, respectively [25-27]. These included headache, nausea, itching, dizziness, thrombophlebitis of the leg, pulmonary embolism, vomiting, rash and an increase in alanine aminotransferase [25-27]. Thompson et al. reported safety data per procedure, with AEs reported in 57% (n = 24) of patients, of which 17% (n = 7) were classified as treatment-related [24]. Gill et al. reported SAEs in 9% (n = 3) of patients [25] and Thompson et al. in 7% (n = 3) of patients, all considered unrelated to treatment [24]. Lethagen et al. reported a single SAE, a pulmonary embolism, possibly related to treatment (details in Supplemental Results) [26]. In another study, one patient with type 2B VWD discontinued due to a pseudo-thrombocytopenia [24]; the patient was not withdrawn from the study and treatment efficacy was rated good to excellent [24].

All studies reported the incidence of TEs ranging from 0% to 7% of patients; three cases of TEs of interest are discussed in supplemental results. No cases of inhibitor development occurred in the four studies that reported this safety outcome [11, 14, 21, 26].

3.5 Pharmacovigilance Data of pdVWF/FVIII

From 7 June 1982 to 31 May 2023 a total of 3227 ADRs in 1396 case reports were identified from post-marketing sources pertaining specifically to pdVWF/FVIII; 1201 of these were spontaneous and 195 were solicited (Table 5).

Over 18 billion IU VWF of pdVWF/FVIII were distributed worldwide from the year 1990 to 31 May 2023 as indicated from available CSL commercial records. Key ADRs associated with the product included the development of FVIII/VWF inhibitors, TEs, hypersensitivity reactions (including anaphylaxis) and suspected transmission of infectious agents.

3.6 Inhibitor Development

The number of case reports associated with the development of neutralizing FVIII/VWF inhibitors was 134 (9.6% of all cases) and described a total of 148 ADRs (146 serious, two non-serious), including inhibitors (both low- and high-titers) against FVIII (n = 83) and VWF (n = 47). Due to the nature of PMS reporting, we could not discern whether the inhibitors were from patients with haemophilia A or VWD.

3.7 Thromboembolic Events

A total of 64 cases (4.6% of all ADR cases) reporting 76 TEs, all considered serious ADRs, were retrieved. Most commonly reported TEs were pulmonary embolism (n = 19) and deep vein thrombosis (n = 14).

3.8 Hypersensitivity Reactions

Hypersensitivity reactions (including anaphylaxis) were identified from 214 cases (15.3% of all ADR cases) describing a total of 321 ADRs (185 serious, 136 non-serious). Among ADRs relating to hypersensitivity reactions, hypersensitivity (n = 92) and anaphylactic reactions (n = 57) were most common.

3.9 Transmission of Infectious Agents

Suspected transmission of infectious agents was identified from 142 cases (10.2% of all ADR cases) reporting 184 ADRs (168 serious, 16 non-serious). These also included reports describing the laboratory identification of virus antibodies without acute disease and cases of other infectious agents not known to be transmitted via blood products such as influenza or COVID-19. None of the suspected virus transmission cases were confirmed to have been associated with pdVWF/FVIII.