Gene addition for hemophilia gene therapy involves adeno-associated viral (AAV) vectors, small viruses that target the liver for endogenous factor expression. AAV vectors are advantageous because they are non-pathogenic. They transduce dividing and non-dividing cells but have low integration rates. They are administered by IV infusion.

Following a single IV infusion over 1–3 h, the vector particles are picked up by liver cell receptors, and taken up into the cell, where the vector particle uncoats and delivers the DNA to the nucleus of the cell. Vector DNA forms stable extrachromosomal episomes, which form concatemeric episomes in cell nuclei. Genetic elements that accompany the gene allow for efficient expression and ultimate secretion of FVIII or FIX protein into the plasma, ultimately reaching a steady state between secretion and clearance that is represented by a measurable factor level [13].

Current gene therapies for hemophilia use AAV vectors to target their transgene to hepatocytes. Specifically, both valoctocogene roxaparvovec and etranacogene dezaparvovec-drlb use the AAV serotype 5 (AAV5) [14, 15]. AAV5 was selected from AAV serotypes 1–10 because it is immunologically distinct from other serotypes, has low seroprevalence, provokes minimal cross-reactivity against common pre-existing AAV2 neutralizing antibodies, and efficiently transduces hepatocytes [2, 16]. One disadvantage of AAV is that naturally occurring immunity, in the form of neutralizing antibodies, is common among the general population and can hamper the efficacy of gene therapies. The relatively low seroprevalence and cross-reactivity of AAV5 compared to other AAVs help circumvent this potential challenge [2, 17].

2.1 Select Phase 3 Gene Therapy Trials for Hemophilia2.1.1 Valctocogene Roxaparvovec GENEr8-1 Phase 3 Trial for Hemophilia A

Two gene therapies for hemophilia received approval for clinical use in 2022. Valoctocogene roxaparvovec was approved in Europe in August 2022. Etranacogene dezaparvovec-drlb received FDA approval in the USA in November 2022 and European approval in February 2023 [14, 15, 18].

Valoctocogene roxaparvovec was developed by BioMarin Pharmaceutical Inc. and is indicated for the treatment of severe hemophilia A in adult patients without a history of FVIII inhibitors and without detectable antibodies to AAV5 (Table 1) [14]. The GENEr8-1 Phase 3 clinical trial (NCT03370913) leading to this approval is part of an ongoing, open-label, single-arm trial. The study population included adult males with severe hemophilia A (FVIII ≤ 1 IU/dL) who had been on prophylactic FVIII replacement for at least 12 months prior to enrollment. Participants had at least 150 exposure days of treatment with FVIII concentrates or cryoprecipitate and could not have a history of FVIII inhibitors (Table 1) [14, 19]. Participants were excluded if they had detectable levels of antibodies to the AAV5 capsid, significant liver dysfunction, an additional bleeding disorder, a platelet count below 100 x 109/L, or creatinine levels ≥ 1.5 mg/dL (Table 1) [14, 19]. After a study amendment, participants with human immunodeficiency virus (HIV) were excluded. Upon recommendation from the FDA, the primary endpoint for the GENEr8-1 trial was changed from FVIII activity levels (by chromogenic substrate assay) to annualized bleed rate (ABR) at year 1 after infusion of the AAV5-hFVIII-SQ vector.

Table 1 Phase 3 trials for approved multinational gene therapies in hemophilia

The total intention-to-treat (ITT) population included 134 participants. The modified intention-to-treat (mITT) population included 132 HIV-negative participants. A rollover population included 112 HIV-negative participants with pre-infusion bleed and treatment data collected for comparison to post-infusion data. One- and 2-year data have been published [19, 20]. At the time of the 2-year data cut-off, one participant was lost to follow-up and one participant with a history of depression died of suicide [20]. Seventeen participants had 3-year data available.

At year 1, mean FVIII activity was 42.9 IU/dL, reflecting a substantial increase from baseline (Table 2) [19]. Median FVIII activity at year 1 was 23.9 IU/dL (Table 2) [19]. FVIII activity trended down over time (Table 2) [20]. Statistical modeling predicted a transgene half-life of 123 weeks, projecting mean and median FVIII activity levels to remain in the mild hemophilia range over 5 years.

Table 2 Efficacy and safety in Phase 3 trials for gene therapies in hemophilia

Overall, there was a substantial decrease in ABR and all bleeding events, and based on the year 1 analysis of ABR in the rollover population, gene therapy was superior to FVIII prophylaxis (p < 0.001). Low bleeding rates were maintained at 2 years post infusion [20], though, as expected, lower mean FVIII activity was associated with more treated spontaneous bleeding and traumatic bleeding events. Modeling of bleeding events and FVIII activity suggested that participants with FVIII 3–5% (moderate range) had a bleeding phenotype more consistent with hemophilia (Table 2) [20].

Factor consumption decreased due to discontinuation of prophylaxis and reduction in bleeding. In the rollover population at week 5 and beyond, the use of factor VIII for usual prophylaxis decreased by 99.6% (Table 2) [20]. At year 2, 128 of 134 participants in the ITT group did not resume prophylaxis, whereas six participants resumed prophylaxis (five used FVIII clotting factor and one used emicizumab), including five participants with FVIII activity < 5 IU/dL by chromogenic substrate assay (CSA) [20]. Overall, participants had lower bleeding rates compared to baseline both before and after resuming prophylaxis.

Safety outcomes were measured in the ITT population (n = 134) (Table 2) [14, 19]. Twenty-two participants (16.4%) reported serious adverse events, and 3.7% reported treatment-related serious adverse events [19, 20]. The most common treatment-related adverse events were increased alanine aminotransferase (ALT) [85.8%], elevated aspartate aminotransferase (AST) (29.1%), and nausea (23.1%) (Table 2) [19, 20]. Treatment-related serious adverse events at year 1 included ALT increase, anaphylactic reaction, hypersensitivity, maculopapular rash, and presyncope. No treatment-related serious adverse events emerged at year 2, and no new safety signals were reported [20]. At year 2, ALT increase was reported as an adverse event of special interest in 29.1% of participants [20]. No participants developed FVIII inhibitors during this trial. Two participants developed malignancies, but both were determined to be unrelated to the treatment [21,22,23]. There were no deaths or withdrawals due to adverse events at year 1, but at year 2 there was one death from suicide that was determined to be unrelated to the treatment [19, 20].

Of all participants who experienced elevated ALT (85.8%), the median time to first elevation was 8.0 weeks, and the median elevation duration was 15 days (Table 2) [20]. ALT elevations grade 3 or higher occurred in 8.2% of patients at year 1 and 0.7% of participants at year 2 [20]. Increases in ALT were managed primarily with glucocorticoids. All were resolved without escalating to grade 4 elevations or drug-induced liver injury (DILI) [19]. Of the participants, 79.1% received glucocorticoids, and the median treatment duration with glucocorticoids was 230 days. At the 2-year data cutoff, 96.2% of ALT elevation events were resolved, two were still resolving, nine were unresolved, and one was unknown [20]. In addition, 29.1% of participants were administered other types of immunosuppressants due to contraindications, adverse effects of glucocorticoids, or a poor response to glucocorticoid management (Table 2) [19, 20]. These patients received budesonide, tacrolimus, mycophenolate, or methylprednisolone [20]. The ongoing GENEr8-3 trial (NCT04323098) will further explore relationships between ALT elevations, FVIII expression, and the use of glucocorticoids and/or other immunosuppressants (Table 2) [14, 19].

In summary, challenges of valoctocogene roxaparvovec treatment include waning of gene expression over years and the need for prolonged use of immunosuppression with corticosteroids for many patients [8]. Data continue to accrue regarding predictability and durability of response and these data will be important to guide patient decision making about gene therapy. Control over bleeding remained strong at 2 years post-treatment, but longer-term data will inform about durability of gene expression and bleeding phenotype [20]. Since some participants had low expression levels, and others had high expression levels, it has become clear that it is not possible to predict expression level, which is a significant drawback that impacts the decision to pursue gene therapy. Additionally, protocols to minimize elevated liver enzymes and loss of gene expression should be optimized. Patients with pre-existing neutralizing antibodies were excluded from this study. This study excluded female patients, children, and males who had a history of FVIII inhibitors [19]. Future studies may improve this therapeutic option by exploring attempts to fill these gaps.

In March 2021, the FDA granted valoctocogene roxaparvovec a Regenerative Medicine Advanced Therapy (RMAT) designation by the FDA [24]. RMAT is a program that seeks to accelerate development and review of regenerative therapies. It received conditional approval by the European Medicines Agency in 2022 and is pending regulatory review in the USA.

2.1.2 Etranacogene Dezaparvovec-drlb HOPE-B Phase 3 Trial for Hemophilia B

For hemophilia B, the gene therapy etranacogene dezaparvovec-drlb was developed by uniQure and CSL Behring. The AAV vector-based gene therapy with a hyperfunctional F9 variant (AAV5-hFIXco-Padua) is indicated for the treatment of adults with hemophilia B who currently use FIX prophylaxis therapy, have current or prior life-threatening hemorrhage, or have repeated, serious spontaneous bleeds (Table 1) [15, 25]. The HOPE-B Phase 3 clinical trial (NCT03569891) is an ongoing multinational, open-label single-arm trial with 54 participants. The trial included adult males with severe or moderately severe hemophilia B, defined as a FIX activity level of ≤ 2 IU/dL. Participants must have been on FIX prophylaxis with at least 150 prior days of exposure to FIX treatment. Participants with or without pre-existing neutralizing antibodies to AAV5 were included. Patients with FIX inhibitors prior to or at screening were excluded (Table 1) [15, 25].

The primary efficacy endpoint in the HOPE-B trial was non-inferiority of the ABR during months 7–18 after infusion compared with the 26-week lead-in period. The ABR during the lead-in period was 4.19 (95% confidence interval (CI) 3.22, 5.45) and the ABR during months 7–18 decreased to 1.51 (95% CI 0.81, 2.82) consistent with non-inferiority of gene therapy to factor replacement prophylaxis (Table 2) [25]. Overall, all bleeding events decreased and this decrease was sustained over time.

At baseline 81% of participants had FIX activity less than 1 IU/dL. Six months post-treatment, FIX activity (least squares mean) increased to 39.0 ± 18.7 IU/dL, and the increased level of FIX activity was sustained at 12 and 18 months. In most gene therapy trials, participants are excluded if they have pre-existing AAV neutralizing antibodies as it is anticipated that the antibodies will block transgene efficacy. However, in HOPE-B, antibodies were measured but not used as a trial exclusion. At month 18, mean FIX levels were 31.1 IU/dL for participants with neutralizing antibodies against AAV5 (n = 21) and 39.9% for those without (n = 33). No correlation between neutralizing antibody titer and FIX level was observed [25].

Factor consumption decreased due to reduction in prophylaxis and treated bleeds. Ninety-six percent of participants discontinued their use of FIX replacement prophylaxis between post-treatment day 21 and month 18 [25]. During the lead-in period, participants used a mean of 257,339 ± 149,013 IU of FIX per year. Between the lead-in period and the post-treatment period, FIX use decreased by a mean of 248,825 IU/year per participant (Table 2).

Quality of life was examined as an exploratory endpoint using the International Physical Activity Questionnaire (iPAQ) and the EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L), and at 12 months post-treatment, participants demonstrated improved QoL compared to baseline based on the Hem-A-QoL total scores [25].

Safety outcomes from the HOPE-B trial included adverse events that occurred or worsened during or after treatment, abnormalities in liver function, vector shedding, and an immune reaction to the AAV5 vector or transgene. All participants experienced adverse events that occurred or worsened during or after treatment. The most common treatment-related adverse events were ALT elevation (17%), headache (15%), influenza-like illness (13%), and AST elevation (9%) (Table 2) [25]. Seventeen percent of participants received glucocorticoid treatment for elevated ALT. The mean duration of glucocorticoid treatment for elevated ALT was 79.8 ± 26.6 days, and no adverse events related to the use of corticosteroids were reported (Table 2).

One participant developed hepatocellular carcinoma that was determined to be unrelated to the AAV5 vector. At 18 months post-treatment, clearance of vector DNA was observed in semen specimens from 61% of participants and blood specimens from 46% of participants. No patients developed FIX inhibitors during the trial [25]. Adverse events were similar among participants with or without AAV5 neutralizing antibodies.

2.1.3 Comparison of Hemophilia A versus Hemophilia B Gene Therapy

Compared to the GENEr8-1 study for hemophilia A, the HOPE-B study for hemophilia B showed etranacogene dezaparvovec-drlb provides a more sustained response with less hepatotoxicity. While FVIII levels decrease over time after treatment with valoctocogene roxaparvovec, results from the Phase 2b etranacogene dezaparvovec-drlb trial showed stable and durable FIX levels 3 years post-treatment [26], and gene therapy recipients in other hemophilia B gene therapy trials have maintained their response for nearly a decade [8, 19, 27]. Difficulty in predicting expression levels is an important consideration for patients considering therapy, but if there is lack or loss of response, patients may safely resume prophylactic therapy [19, 25].

Valoctocogene roxaparvovec is associated with higher rates of transaminitis and consequently associated immunosuppressive treatments than etranacogene dezaparvovec-drlb. This is an important point when balancing risks and benefits of gene therapy. Therefore, liver health is essential for all AAV gene therapy and patients must understand key liver health exclusions and precautions, including avoiding alcohol in excess and hepatotoxic medications.

The HOPE-B trial included patients with pre-existing neutralizing antibodies to AAV5, while the GENEr8-1 trial did not [19, 25]. Patients with hemophilia A who are considering gene therapy must understand that they may be excluded due to pre-existing antibodies and all patients should understand that with current strategies re-dosing will not be possible since everyone mounts an antibody response to AAV after treatment.

2.2 Remaining Questions for Gene Therapies in Hemophilia

Many of the questions about gene therapies for hemophilia focus on long-term efficacy. Although the expectation is a single infusion with durable efficacy, the actual duration of treatment effectiveness is largely unknown [3, 28]. Notably, the durability of valoctocogene roxaparvovec therapy is shorter than the durability of etranacogene dezaparvovec-drlb therapy [19, 20, 25, 26]. Reasons are unclear but may be inherent to the fact that the native site of FIX production is the hepatocyte, which coincides with the AAV vector target cell. In contrast, the native site for FVIII production is liver sinusoidal endothelial cells, which is different from the AAV target cell. The difference in cell type for FVIII could lead to stress on the endoplasmic reticulum, potentially reducing the protein expression over time [29, 30]. As the trials are still ongoing, the continued collection of efficacy data will be important to increase knowledge about how to use these gene therapies to provide maximum benefit for patients. Long-term follow-up is essential for all gene therapy recipients, including those who receive commercial product, and patients will be enrolled into gene therapy registries.

In addition to examining questions about efficacy over time, it is critical to consider concerns about long-term safety, especially with regard to liver health [3, 28]. Prolonged immunosuppression with corticosteroids and other immunosuppressants causes undesirable side effects but is an important tool for managing ALT elevations. Long-term effects on liver health are currently unknown, and it is unclear if risks will be different in patients with HIV or other immunodeficiency. Clearly, patients undergoing gene therapy for hemophilia will need to actively maintain their liver health by avoiding excessive alcohol intake and hepatotoxic medications. Maintaining liver health will be necessary to maintain gene expression [3, 28].

There are further questions surrounding the use of immunosuppressants. Specifically, it would be beneficial to gain knowledge about when to initiate immunosuppressants. Would a patient benefit most from immunosuppressant initiation during prophylaxis, immediately before beginning gene therapy, immediately after the gene therapy infusion, or later? Answers to these questions will optimize the treatment protocol to help ease the burden of undergoing gene therapy [3].

Further, cancer was reported at rates seen in the general population in both Phase 3 trials, and the cases were determined by investigators to be unrelated to the gene therapy [19, 20, 25, 26]. Nonetheless, gene therapy that uses liver-targeted AAV vectors must be carefully considered for patients with pre-existing risk factors for hepatocellular carcinoma.

Additional potential safety concerns include thrombosis with supraphysiologic factor levels and genot