Haemophilia A is an inherited bleeding disorder caused by a deficiency of coagulation factor VIII (FVIII) due to F8 gene mutations, leading to impaired thrombin generation and unstable clot formation. The severity is classified by plasma FVIII activity as severe (<1%), moderate (1%–5%), or mild (5%–40%), with a reference range of 50%–150% [1]. The primary goal of treatment is to prevent bleeding episodes and long-term joint damage [2].
Current guidelines recommend prophylactic (PPX) coagulation factor replacement as the standard of care for severe haemophilia, but frequent intravenous administration poses challenges, including venous access difficulties, prolonged preparation, and patient discomfort [3,4]. Alternative therapies, such as antibodies that mimic FVIII or inhibit endogenous coagulation regulators, have emerged. Concizumab, a humanised monoclonal antibody targeting tissue factor pathway inhibitor (TFPI), is designed for subcutaneous prophylaxis in haemophilia A or B, with or without inhibitory alloantibodies to FVIII or FIX. By inhibiting TFPI, concizumab restores haemostatic balance, potentially reducing bleeding episodes and improving adherence compared to intravenous therapies.
Early-phase clinical trials have demonstrated concizumab's ability to enhance thrombin generation in a dose-dependent manner. However, its safety profile, particularly concerning thrombotic risk and immunogenicity, remains under investigation [8,9]. This systematic review and meta-analysis consolidates existing evidence on concizumab's safety, evaluating adverse events, thrombosis risk, and immunogenic responses in randomised controlled trials. The findings aim to inform clinical decision-making, optimise haemophilia care, and identify gaps for future research.
A systematic search was performed across PubMed, Cochrane Library, Scopus, Google Scholar, and ClinicalTrials.gov from inception to February 15, 2025. The search strategy incorporated both controlled vocabulary (MeSH and Supplementary Concepts) and free-text terms to identify relevant studies on the safety of concizumab in haemophilia. The following search string was applied:
(“Hemophilia A”[Mesh] OR “Hemophilia B”[Mesh] OR hemophilia[tiab]) AND (“Concizumab”[Supplementary Concept] OR Concizumab[tiab]) AND (“Adverse Events”[Mesh] OR “Drug-Related Side Effects and Adverse Reactions”[Mesh] OR “Safety”[tiab] OR “Serious Adverse Events”[tiab] OR “Upper Respiratory Tract Infections”[Mesh] OR “Hemarthrosis”[Mesh])
This strategy ensured a comprehensive retrieval of studies assessing concizumab's safety profile, focusing on adverse events, serious complications, and trial-based evidence.
Studies were included if they met the following criteria: (i) randomised controlled trials or clinical trials comparing concizumab with placebo or standard therapy, (ii) enrolled patients with haemophilia A or B, (iii) reported safety outcomes, including overall adverse events, serious adverse events, upper respiratory tract infections, or haemarthrosis, (iv) published in English, and (v) provided sufficient data for extraction. Studies were excluded if they were observational, review articles, case reports, or involved concizumab in combination with other interventions that could confound safety outcomes.
Two reviewers independently screened titles and abstracts to identify relevant studies, followed by a full-text assessment to confirm eligibility. Any differences were settled through discussion or by consulting a third reviewer. Extracted data included study design, sample size, patient demographics, intervention details (concizumab dose and duration), comparator information, and reported safety outcomes.
The primary outcomes assessed were safety-related parameters, including the incidence of adverse events, serious adverse events, upper respiratory tract infections, and joint bleeding episodes. Data was extracted according to the follow-up periods reported in each study.
Two reviewers independently extracted data using standardised forms, capturing study characteristics, patient demographics, intervention details, comparator information, and safety data (e.g., event counts and total participants per group). Discrepancies were resolved through consensus.
Two reviewers independently assessed the methodological quality of included studies using Version 2 of the Cochrane Risk of Bias tool (RoB 2) for randomised trials [6]. This evaluation covered biases related to randomization, allocation concealment, blinding, incomplete outcome data, and selective reporting. Any disagreements were addressed through discussion or, if needed, consultation with a third reviewer.
The meta-analysis followed the Cochrane Handbook for Systematic Reviews of Interventions and adhered to PRISMA guidelines [7]. Statistical analyses were performed using RevMan 5.4 (Cochrane Management System). Risk ratios (RR) with 95% confidence intervals (CI) were used to analyse dichotomous outcomes [8]. Heterogeneity was assessed using the I2 statistic, with values over 50% indicating significant heterogeneity. A p-value of <0.05 was considered statistically significant unless specified otherwise. Analyses were conducted based on intention-to-treat populations whenever possible.
Our comprehensive electronic database search yielded a total of 50 records. After subsequent screening of abstract, titles and full-text evaluation, four studies [9,10,11,12] were included in our analysis. The PRISMA flowchart for inclusion process is shown in Figure 1.
PRISMA flow diagram for the systematic review
The PRISMA flow diagram illustrates the study selection process for this systematic review and meta-analysis, showing the identification, screening, eligibility, and inclusion phases. Numbers at each step represent studies identified through database and other sources, duplicates removed, studies screened, full-text articles assessed for eligibility, and studies included in the final analysis, providing an overview of how studies were selected and reasons for exclusion.
Table 1 summarises the characteristics of the included studies.
Characteristics of studies included in the systematic review and meta-analysis
| STUDY | STUDY DESIGN | STUDY LOCATION | TREATMENT DURATION | INCLUSION CRITERIA | MEAN AGE | INTERVENTION GROUP | CONTROL GROUP | OUTCOMES MEASURED |
|---|---|---|---|---|---|---|---|---|
| Chowdary 2015 [12] | Prospective, multi-centre, open-label, randomised phase 3a trial | 69 investigational sites in 31 countries |
|
| 25.3 ± 3.1 | Group 2, who received concizumab prophylaxis | Group 1, who received no prophylaxis and continued on-demand clotting factor treatment |
|
| Shapiro 2019 [9] |
|
| At least 24 weeks for those receiving concizumab (in both trials) |
| N/A | Participants who received concizumab prophylaxis (both trials) | explorer4 (inhibitor trial): participants with haemophilia A or B with inhibitors, who received on-demand treatment with eptacog alfa activated (rFVIIa) instead of concizumab prophylaxis |
|
| Chowdary 2024 [11] | First human dose, phase 1, multi-centre, randomised, double-blind, placebo-controlled, single-dose, dose-escalation trial | Multiple sites in 9 countries: Austria, Denmark, Germany, Malaysia, South Africa, Spain, Switzerland, Thailand, UK | 43 days |
| 30–36 years | Participants who received concizumab | Participants who received placebo, with 1:3 ratio of placebo to concizumab within each dose cohort |
|
| Matsushita 2023 [10] | Prospective, multi-centre, open-label, phase 3a trial, comparing concizumab prophylaxis with no prophylaxis | N/A |
|
| N/A |
| Group 1, who received no prophylaxis (on-demand treatment with bypassing agents). |
|
Product comparison data for three patients who received a SHL FIX product for a prior surgery
| CERTAINTY ASSESSMENT | PATIENTS (N;%) | EFFECT | CERTAINTY | IMPORTANCE | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| STUDIES (N) | STUDY DESIGN | RISK OF BIAS | INCONSISTENCY | INDIRECTNESS | IMPRECISION | OTHER CONSIDERATIONS | CONCIZUMAB | STANDARD | RELATIVE (95% CI) | ABSOLUTE (95% CI) | ||
| Bleeding joint | ||||||||||||
| 2 | RCT | NS | NS | NS | NS | publication bias strongly suspecteda | 15/30 (50.0%) | 10/13 (76.9%) | RR 0.66 (0.45 to 0.96) | 262 fewer per 1,000 (from 423 fewer to 31 fewer) |
| |
| Adverse events | ||||||||||||
| 4 | RCT | S | NS | NS | NS | none | 74/99 (74.7%) | 21/38 (55.3%) | RR 1.17 (0.89 to 1.54) | 94 more per 1,000 (from 61 fewer to 298 more) |
| |
| Upper respiratory tract infection | ||||||||||||
| 2 | RCT | S | NS | NS | NS | publication bias strongly suspecteda | 4/69 (5.8%) | 2/28 (7.1%) | RR 0.75 (0.15 to 3.85) | 18 fewer per 1,000 (from 61 fewer to 204 more) |
| |
| Serious adverse events | ||||||||||||
| 3 | RCT | S | NS | NS | NS | none | 7/81 (8.6%) | 6/32 (18.8%) | RR 0.46 (0.06 to 3.53) | 101 fewer per 1,000 (from 176 fewer to 474 more) |
| |
RCT: Randomised controlled trial
NS: Not serious
S: Serious
CI: Confidence interval
RR: Risk ratio
Since a funnel plot could not be generated for the outcome, publication bias is expected
Since the analysis included a study with a high risk of bias, a strong bias is expected
The quality of the included studies was assessed using the Cochrane Risk of Bias 2.0 Tool [6], which evaluates key aspects such as how participants were randomly assigned, whether outcome data were complete, and whether all planned results were fully reported. The assessment revealed varying levels of bias across studies. Matushita 2023 [10] and Chowdary 2024 [11] (phase III trials) were classified as high risk of bias, mainly due to missing outcome data and selective reporting of results. Missing data and selective reporting can potentially skew the study findings, leading to overestimation or underestimation of the true effect. In contrast, Chowdary 2015 [12] and Shapiro 2019 [9] were deemed low risk, reflecting a more robust study design with minimal bias concerns. These findings highlight potential limitations in some studies that may affect the overall reliability of the evidence. A detailed summary of the risk of bias assessment is presented in Figure 2. Inverted funnel plots for the primary outcomes were obtained, as illustrated in Figure 3. The assessment of these plots, along with Egger's regression tests, found no significant indication of publication bias.
Risk of bias summary for included studies using Cocharane Risk of Bias 2.0 tool
Funnel plot for outcome ‘adverse events’
The funnel plot shows evaluation of potential publication bias for the outcome ‘adverse events’. The X-axis represents the risk ratios (RR) between concizumab and placebo; the Y-axis represents the standard error of log risk ratios (SE(log[RR])). Each open circle corresponds with an individual study. The vertical line indicates the null hypothesis (RR=1). The triangle represents the expected 95% confidence interval in the absence of publication bias. Symmetry around the vertical line suggests a low likelihood of publication bias, while asymmetry may indicate potential reporting or small-study effects.
The evaluation of evidence was conducted using GRADEpro Guideline Development Tool [software] [13] A detailed summary of the evidence certainty for each outcome is shown in Table 2.
For the outcome of mean annualised bleeding ratio, three studies involving 105 participants were included. Treatment with concizumab significantly reduced annualised bleeding rates compared with placebo, with an overall mean difference (MD) of −15.79 (95% CI, −21.28 to −10.31; p<0.00001). Heterogeneity across studies was low (I2 = 0%). Subgroup analysis showed a similar effect in patients with inhibitors (two studies; MD=−15.52, 95% CI, −21.87 to −9.18) and in those without inhibitors (1 study; MD=−16.60, 95% CI, −27.55 to −5.65), indicating consistent benefit across both groups. (Figure 4a).
Forest plots highlighting aspects of the included studies
Figure 4a. Mean annualised bleeding rates (ABR) across the included studies
For the outcome of number of patients experiencing bleeding episodes (defined as the number of participants with at least one bleeding episode during follow-up), four studies were included, with two studies each in the inhibitor and non-inhibitor subgroups. Overall, concizumab was associated with a lower, but not statistically significant, risk of experiencing a bleeding episode compared with placebo (RR=0.69, 95% CI, 0.38 to 1.28), although substantial heterogeneity was observed (I2=79%, p=0.24). Subgroup analysis showed a non-significant reduction in patients with inhibitors (RR=0.62, 95% CI, 0.28 to 1.37), while results in patients without inhibitors were inconclusive due to wide confidence intervals (RR=2.03, 95% CI, 0.10 to 42.74). Sensitivity analysis identified Shapiro 2019 [9] and Chowdary 2024 [11] as main contributors to heterogeneity. Excluding these studies revealed a consistent and statistically significant effect, with fewer patients experiencing bleeding episodes in the concizumab group compared with placebo (RR=0.46, 95% CI, 0.31 to 0.67; I2=0%, p<0.0001) (Figure 4b).
Proportion of patients experiencing one or more bleeding episodes
For the outcome of joint bleeding (defined as the number of patients experiencing joint bleeding), two studies including 43 participants were analysed. Overall, treatment with concizumab significantly reduced the risk of bleeding joints compared with placebo (RR=0.66, 95% CI, 0.45 to 0.96; I2=0%, p=0.03). Subgroup analysis showed a significant reduction in patients with inhibitors (one study; RR=0.64, 95% CI, 0.43 to 0.94), whereas in patients without inhibitors (one study), the effect was uncertain due to a wide confidence interval (RR=1.33, 95% CI, 0.20 to 8.71) (Figure 4c).
Frequency of bleeding events involving joints
For the outcome of adverse events (defined as the number of patients experiencing any adverse event), four studies including 137 participants were analysed. The pooled analysis showed no significant difference between concizumab and placebo (RR=1.17, 95% CI, 0.89 to 1.54; I2=0%, p=0.25). Subgroup analysis revealed similar findings in patients with inhibitors (two studies; RR=1.14, 95% CI, 0.83 to 1.58) and in those without inhibitors (two studies; RR=1.25, 95% CI, 0.75 to 2.08), indicating no clear subgroup effect (Figure 4d).
Incidence of all reported adverse events
For the outcome of serious adverse events, three studies including 113 participants were analysed. Overall, there was no significant difference between concizumab and placebo (RR=0.46, 95% CI, 0.06 to 3.53), with moderate heterogeneity (I2=56%, p=0.46). Subgroup analysis showed a non-significant reduction in patients with inhibitors (two studies; RR=0.27, 95% CI, 0.01 to 7.31), whereas the single study in patients without inhibitors reported no clear effect (RR=1.15, 95% CI, 0.06 to 23.88). Sensitivity analysis identified Shapiro 2019 [9] as the main contributor to heterogeneity. Excluding this study resulted in more consistent findings, showing no significant difference between groups (RR=1.15, 95% CI, 0.36 to 3.70; I2=0%, p=0.81). (Figure 4e).
Occurrence of serious adverse events
For the outcome of upper respiratory tract infections, two studies including 97 participants were analysed. The pooled results showed no significant difference between concizumab and placebo (RR=0.75, 95% CI, 0.15 to 3.85). No subgroup analysis was performed for this outcome.
Our meta-analysis demonstrates that concizumab consistently reduces annualised bleeding rates and joint bleeding events in people with haemophilia A and B, both with and without inhibitors. These efficacy findings are directly aligned with the pharmacologic mechanism of concizumab as a TFPI blocker. By inhibiting TFPI, concizumab enhances thrombin generation, thereby improving hemostasis without directly replacing deficient clotting factors [14]. This mechanism helps explain the observed reductions in bleeding episodes across trials [10].
Safety outcomes in the included RCTs showed no statistically significant increase in overall or serious adverse events compared with placebo. While this suggests that short-term TFPI modulation is generally well tolerated [10], it is important to contextualise these findings with the limitations of small sample sizes, short follow-up durations, and the lack of real-world pharmacovigilance data. Rare but clinically important events, including thrombotic complications, may not be captured in these controlled trial populations [15].
These results are consistent with prior non-randomised and early-phase studies, which also reported favourable safety and efficacy profiles, though with similar limitations regarding sample size and follow-up [14,15]. Mechanistically, TFPI inhibition carries a theoretical risk of hypercoagulability, highlighting the importance of ongoing monitoring in larger patient populations. Moreover, immunogenicity trends remain incompletely characterised beyond the short-term follow-up reported in most trials.
Integrating mechanistic understanding with clinical outcomes provides valuable insight for clinicians considering non-factor prophylactic therapies. While concizumab shows promising efficacy and short-term tolerability, caution is warranted regarding rare or delayed adverse events, and these findings should inform careful patient selection and monitoring strategies. Future studies incorporating larger populations, longer follow-up, and real-world data are essential to fully define the safety profile and guide clinical use [16].
Immunogenicity is an important consideration for therapeutic antibodies such as concizumab, as the development of anti-drug antibodies (ADAs) could theoretically reduce efficacy or alter safety profiles. None of the included randomised trials systematically reported ADA incidence or its potential impact. Furthermore, all trials had relatively short follow-up durations, which limits the ability to detect delayed immunogenicity or long-term effects on efficacy and safety [10,15]. Future studies and post-marketing surveillance are therefore needed to evaluate whether immunogenicity may influence clinical outcomes during extended treatment.
While this meta-analysis focused on concizumab, it is important to contextualise its safety profile relative to other emerging non-factor therapies, including emicizumab and fitusiran. Emicizumab, a bispecific antibody bridging factors IXa and X, has demonstrated a low incidence of thrombotic events in clinical trials, although rare thrombotic microangiopathy has been reported in combination with activated prothrombin complex concentrate [16,17]. Fitusiran, an RNA interference (RNAi) therapeutic targeting antithrombin, carries theoretical and observed risks of thrombosis, prompting careful monitoring in ongoing studies [16]. Direct comparisons with concizumab are not available; however, differences in mechanism of action may influence both efficacy and safety profiles. These considerations underscore the need for continued pharmacovigilance and real-world studies to better understand the comparative safety of novel non-factor therapies [16,17].
Regulatory agencies, including the European Medicines Agency (EMA) and the United States Food and Drug Administration (FDA), conduct post-marketing and post-trial safety reviews that can provide additional insights beyond published clinical trials. While concizumab has demonstrated short-term tolerability in the included RCTs, these regulatory assessments can capture rare adverse events and emerging safety signals, including thrombotic events, immunogenicity, and effects of concomitant therapies. Incorporating regulatory safety data into future reviews will be important for a comprehensive understanding of concizumab's risk–benefit profile [18].
A key limitation of this meta-analysis is the combination of studies including people with haemophilia both with inhibitors and without inhibitors, which may have introduced clinical heterogeneity and limited the ability to draw definitive subgroup-specific conclusions. Another limitation is the relatively small number of participants included in the analysed studies, which may reduce the statistical power and the generalisability of the findings.
This meta-analysis included only four RCTs encompassing 137 participants. The small number of studies and limited sample size reduce the statistical power of pooled analyses and restrict the ability to detect infrequent but clinically relevant adverse events such as thrombosis or immunogenicity. As a result, while the available data suggest that concizumab is generally well tolerated in the short term, larger phase 3 studies and post-marketing surveillance data are needed to validate the safety profile and assess long-term risks.
Another key limitation is the absence of real-world data. Randomised controlled trials, while methodologically rigorous, typically include small, highly selected populations and relatively short follow-up durations. As a result, rare but clinically important safety events such as thrombosis or immunogenicity may not be captured, increasing the risk of a type II error. Incorporating real-world evidence and post-marketing pharmacovigilance data will therefore be essential to validate these findings and better define the true incidence of infrequent adverse events associated with concizumab.
Subgroup analyses by inhibitor status (patients with vs. without inhibitors) were performed for both efficacy and safety outcomes. These analyses indicate that the treatment effect on annualised bleeding rates and adverse events was generally consistent across subgroups. However, due to limited data, further subgroup analyses by trial phase or follow-up duration were not possible, which remains a limitation.
Publication bias assessment was limited by the small number of included trials, and for two outcomes, formal evaluation was not possible. Consequently, these analyses may not reliably detect reporting bias, and the possibility of unreported adverse events cannot be excluded. The interpretation of pooled estimates should therefore be considered cautiously.
The included trials provided limited information on concomitant therapies administered alongside concizumab, such as factor replacement or bypassing agents. The use of these therapies could influence the incidence and severity of adverse events, including thrombotic events or bleeding episodes. Due to insufficient reporting, we were unable to perform analyses accounting for concomitant treatments, which represents a limitation. Future studies should systematically report concomitant therapy use to allow a more precise assessment of concizumab's safety profile in real-world clinical practice.
Adverse events of special interest (AESI), including thrombosis, transaminitis, and injection site reactions, are important considerations for concizumab therapy. The included randomised trials did not systematically report these specific outcomes, limiting the ability to assess their incidence in this review. Mechanistically, concizumab enhances thrombin generation through TFPI inhibition, which could theoretically increase thrombotic risk. Injection site reactions and transient liver enzyme elevations have been reported with other non-factor therapies, although comprehensive data for concizumab are lacking. Future studies and post-marketing surveillance are needed to characterise the incidence and clinical significance of these AESI during long-term therapy.
This review places the observed efficacy and safety of concizumab into mechanistic and clinical context. Concizumab acts by inhibiting TFPI, thereby enhancing thrombin generation and improving haemostasis without replacing missing clotting factors. The reduction in annualised bleeding rates and joint bleeding aligns with this mechanism. Safety outcomes suggest short-term tolerability; however, the limited number of trials, small sample sizes, and short follow-up periods restrict insights into rare or delayed adverse events, including thrombotic events and potential immunogenicity. Comparative insights from other non-factor therapies, such as emicizumab and fitusiran, indicate that differences in mechanism can influence both efficacy and risk, highlighting the importance of individualised treatment strategies. Taken together, integrating mechanistic understanding with trial data and early observational evidence provides a more nuanced interpretation of concizumab's safety profile and helps inform clinical decision-making while identifying key gaps for future research.
In conclusion, concizumab demonstrates promising efficacy in reducing bleeding, particularly joint bleeds, with a favourable safety profile. The modest sample sizes and limited long-term data highlight the need for larger, well-designed trials to confirm these findings and establish its role in haemophilia prophylaxis. Long-term, closely monitored follow-up using national or international registries may provide more feasible and informative data on the long-term safety and efficacy of concizumab, given the challenges of conducting large RCTs in rare diseases.