Prophylaxis use of clotting factor replacement products in people with non‐severe haemophilia: A review of the literature

Abstract

Introduction

People with non‐severe haemophilia appear to be under‐treated in many countries, and this may lead to joint damage and worsen quality of life.

Aim

To review literature for clotting factor replacement prophylaxis in people with non‐severe haemophilia A and B (HA/HB) in relation to long‐term outcomes to support clinical decision‐making.

Methods

A targeted literature search was performed to identify studies published between 2000 and 2021 that included prophylaxis in people with non‐severe HA/HB and long‐term outcomes, including annualized bleeding rates, joint health and quality of life.

Results

Although eligible articles included 2737 and 2272 people with mild or moderate HA, respectively, only 22% (n = 609) and 29% (n = 668) reported treatment regimens. A total of 549 people with moderate HA were treated with factor replacement prophylaxis and were from high‐income countries. On the contrary, nearly all people with mild HA received desmopressin (n = 599). Details of treatment regimens for women with haemophilia and people with HB were sparse. Three studies provided long‐term outcomes for people with moderate haemophilia who received prophylaxis with factor concentrate, supporting early prophylaxis in people with a frequent bleeding phenotype regardless of their endogenous clotting factor level to preserve joint health.

Conclusion

There remain large knowledge gaps when considering how to provide optimal treatment for people with non‐severe haemophilia. Nonetheless, there is a strong rationale that prophylaxis should be considered early in life according to similar strategies as for severe haemophilia for those with a frequent severe bleeding phenotype.

1. INTRODUCTION

People with non‐severe haemophilia A (HA) and haemophilia B (HB) are defined as having endogenous clotting Factor VIII (FVIII) or Factor IX (FIX) concentrations between .01– < .4 IU/mL (1– < 40%) of normal levels. ¹ Despite detectable FVIII/FIX activity levels, a subgroup of these people experience frequent bleeds. ² , ³ Historically, it was thought that 1–3% of factor levels are enough to protect from frequent bleeding and joint damage; however, higher levels may be required to prevent joint damage and poor quality of life (QoL). ² , ³ , ⁴ , ⁵ Recent advances in care mainly impacted the management of the severe disease (FVIII < 1% and FIX < 2%) leaving behind non‐severe patients despite their bleeding phenotype. Hence, for this patient population, several unmet needs remain, including an unacceptable frequency of bleeds, poor joint health, work/school absenteeism, challenges on personal relationships and a negative impact on their participation in recreational activities. ³ , ⁶ , ⁷ Treatment goals for people with haemophilia (PWH) have expanded from reducing annual bleeding rates (ABRs) to allow for full participation in social activities (leisure, sports, travelling, and physical activities) through higher protection against joint damage and better QoL. ⁸ Nordic countries recommend prophylaxis for PWH with baseline FVIII/FIX ≤.03 IU/mL from early ages, prior to the onset of joint disease, but in most countries, prophylaxis is not standard practice for these PWH. ⁹ This paper reviews prophylactic use of replacement factor products in people with non‐severe HA and HB in relation to long‐term outcomes to support clinical decision‐making when choosing treatment options in relation to bleeding phenotype.

2. METHODS

2.1. Literature search

Relevant articles in English on the use of prophylaxis in non‐severe HA published between January 2000 and June 2021 were searched in the PubMed database (see Table 1 for search strings). Each article's abstract was manually screened for any mention of prophylaxis, report of treatment outcomes and risk factors for inhibitor development. Single patient case reports and congress abstracts were not included. Separate searches were conducted for prophylaxis use in women with symptomatic non‐severe HA or HB, and people with non‐severe HB. Additional relevant studies were manually added if they had not been identified using the original search strategy.

TABLE 1.

Search terms used for literature searches

	Search string
Non‐severe haemophilia	(“moderate haemophilia”[title] or “moderate hemophilia”[title] or “mild haemophilia”[title] or “mild hemophilia”[title] or “non‐severe hemophilia”[title] or “non‐severe haemophilia”[title]) AND ((”2000“[Date ‐ Publication] : ”3000″[Date ‐ Publication]) NOT (“case report” or “case study”[title/abstract]))
Women with haemophilia	(“haemophilia”[title]) or (“hemophilia”[title]) AND (“treatment”[title/abstract]) AND ((“women”[title]) OR (“female”[title]) OR (“carrier”[title])) AND (“2000”[Date ‐ Publication] : “3000”[Date ‐ Publication]) NOT (“case report” or “case study”[title/abstract])
Non‐severe haemophilia B	(“mild haemophilia b”[title]) or (“mild hemophilia b”[title]) or (“moderate haemophilia b”[title]) or (“moderate hemophilia b”[title]) or (“non‐severe haemophilia b”[title]) or (“non‐severe hemophilia b”[title]) AND (“2000”[Date ‐ Publication] : “3000”[Date ‐ Publication]) NOT (“case report” or “case study”[title/abstract])

2.2. Analysis

For each eligible article, information about study type, study objective, total number of PWH enrolled, treatment interventions, severity of haemophilia, age of participants, endogenous factor procoagulant activity, outcome measures, numbers of treated PWH and details of their treatment regimens were summarized. Each study was graded according to its quality of evidence (Table S1). ¹⁰

3. RESULTS

3.1. Identified citations

The initial literature search identified 236 unique articles, of which 91 were considered eligible for further screening (Figure 1). Twenty‐seven articles mentioned prophylaxis treatment for people with non‐severe HA, of which 15 were prospective ³ , ⁵ , ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ and 12 retrospective studies ⁹ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ (cohort studies, n = 10 ³ , ¹¹ , ¹³ , ²⁴ , ²⁶ , ²⁷ , ²⁸ , ³² , ³³ , ³⁴ ; observational studies, n = 11 ¹² , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ²² , ²³ , ²⁵ , ²⁹ , ³⁰ , ³¹ , ³⁴ ; cross‐sectional studies, n = 2, ⁵ , ⁹ other study types, n = 4 ¹⁴ , ¹⁹ , ²⁰ , ²¹ ). Most of the studies (19 out of 27) provided Level IV evidence (well‐designed case‐controlled or cohort studies). In total, these studies included 2737 people with mild HA (endogenous FVIII activity range > .05–.5 IU/mL as per study definition) and 2272 people with moderate HA (endogenous FVIII activity range > .01–.05 IU/mL). Two studies included 66 people with moderate HB. ⁹ , ²⁴ Two articles were identified from the additional searches on women with haemophilia: they included 115 women with non‐severe HA, 25 with non‐severe HB. ³⁵ , ³⁶ Three additional articles identified outside the web‐based search strategy were included in the literature review (these articles did not use the terms ‘mild’, ‘moderate’ or ‘non‐severe’ or ‘treatment’ in the either the title or abstract). ³⁷ , ³⁸ , ³⁹ Ten articles were identified that described risk factors for developing inhibitors in people with non‐severe HA, ³ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ and one article was manually added as it did not mention ‘treatment’ in the title or abstract. ⁴⁹

FIGURE 1.

Flowchart for citations identified, screened and included in literature review

3.2. Treatment regimens reported for people with non‐severe haemophilia A or B

Among eligible studies, treatment regimens were reported for 631/2769 (22.8%) and 676/2291 (29.5%) with mild or moderate HA, respectively. ⁹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ⁵⁰ A total of 550 of people with moderate HA were treated with factor replacement prophylaxis, although the dose, frequency and duration of the prophylaxis regimen was not always mentioned; the vast majority came from western Europe ³ , ⁹ , ¹³ , ¹⁶ , ¹⁷ , ¹⁸ , ²⁴ , ⁵⁰ and Australia. ²⁸ One study reported short‐term (6–12 weeks) low‐dose prophylaxis in 66 children with HA and arthropathy, of whom 32 children had moderate disease severity. ²² No people with mild HA were reported to have received long‐term prophylaxis factor replacement.

Two studies reported prophylaxis use among women with haemophilia but were unclear about how many women received prophylaxis and for how long. ³⁵ , ³⁶ One retrospective study of 47 women (HA, n = 37; HB, n = 10) from three haemophilia treatment centres (HTCs) in the United States mentioned the use of FVIII prophylaxis in two women with mild HA who had excessive menstrual bleeding (but did not report whether prophylaxis was continuous or during menstruation only), four women who received factor replacement in preparation of birth and one woman who underwent prophylaxis at time of surgery. ³⁵ In this cohort, women, who were obligate or potential carriers, had history of excessive bleeding and could have normal factor levels. ³⁵ Another study reported intermittent prophylaxis for one woman (out of eight) with moderate haemophilia without reporting whether this woman had HA or HB. ³⁸

Desmopressin (administered either nasally, subcutaneously or intravenously) was reported for 599 people with mild HA (i.e. nearly all for whom the treatment regimen was reported) and 83 people with moderate HA. Desmopressin is considered the main treatment for PWH with mild HA in those who have been shown to respond. One study investigated the use of recombinant interleukin 11 in five people unresponsive to desmopressin ²⁰ and another investigated ex vivo clotting function with emicizumab in people with mild (n = 6) or moderate (n = 10) HA. ¹⁹

Details of treatment regimens for people with HB were sparse. In the MoHem study, 21/56 (38%) people with moderate HB received prophylactic FIX replacement, starting at the median age of 10 years (interquartile range: 6–37 years). ⁹ Another study did not provide details of prophylaxis FIX replacement in participants with HB. ²⁴ In the Bridging Hemophilia B Experiences, Results and Opportunities into Solutions (B‐HERO‐S) survey, the majority of adult people with mild (63% [47/74]) or moderate (86% [162/189]) HB received some form of FIX prophylaxis. ³⁷

3.3. Outcomes reported for people with mild haemophilia A or B

None of the identified studies reported data on arthropathy, joint bleeds or long‐term prognosis in people with mild HA or HB in relation to their treatment regimen or bleeding tendency. For mild HA, most studies that reported desmopressin treatment were investigating predictors of response (such as FVIII gene [F8] mutations), ¹¹ , ²⁵ , ²⁹ , ³¹ , ³² , ³⁴ efficacy (resolution of a bleeding event) depending on mode of administration or clinical setting, ¹² , ¹⁴ , ¹⁵ , ²⁶ , ³⁰ , ³³ or combined desmopressin treatment with FVIII concentrate replacement therapy in non‐responders to desmopressin. ²¹ , ²⁵

Another study reported 26 out of 186 pregnant women with HA, HB or von Willebrand disease who received prophylactic replacement in preparation for child birth. ³⁶ The authors observed higher incidence of postpartum haemorrhage in women who received prophylaxis and had third trimester FVIII/FIX levels < .50 IU/mL, suggesting insufficient replacement factor treatment prior to birth. ³⁶ Current treatment guidelines recommend that factor levels should be ≥1.0 IU/mL at parturitian. ⁵¹ , ⁵²

Although not identified using our original search strategy, a birth‐cohort analysis of a large United States database reported that between .2% and 1.3% of people with mild HA/HB used prophylaxis. ³⁹ All PWH with mild disease reported between 2.3 and 7.2 target joints. Older PWH with mild disease were more likely to self‐report as disabled compared with younger PWH (13% of the cohort born before 1958; 2% of the cohort born 1983–1992). ³⁹

3.4. Outcomes reported for people with moderate haemophilia A or B

Identified studies included outcomes of bleeding, joint health, factor consumption, orthopaedic outcomes and QoL, ³ , ⁹ , ²⁴ pharmacokinetics, ¹⁷ short‐term bleeding outcomes for low‐dose prophylaxis, ²² the effect of non‐neutralizing allo‐antibodies (inhibitors), ¹³ PWH preferences regarding on‐demand versus prophylaxis treatment in males aged 13–23 years, ⁴ and use of clotting factors in people with moderate HA. ²⁸

3.4.1. Prophylaxis use of factor concentrates in moderate haemophilia

Three studies provided long‐term outcomes for people with moderate haemophilia who received prophylaxis with factor concentrate (Table 2), ³ , ⁹ , ²⁴ reporting the limitations of current (or recent) practices in the Netherlands, the United Kingdom (UK) and Nordic countries.

TABLE 2.

Prophylaxis use of factor replacement treatment in moderate haemophilia

	den Uijl et al. 2014 24	THUNDER 3	MoHem 9
Study type	Retrospective single‐centre cohort study (The Netherlands)	Prospective, multicentre cohort study (United Kingdom)	Prospective (joint health assessments) multicentre, multinational (Nordic countries) cross‐sectional study; medical history was assessed retrospectively
Study objective	To assess short‐ and long‐term outcomes, including its association with treatment, in patients with moderate haemophilia	To describe the current standard of care and to evaluate unmet need in people living in the UK with severe HA and moderate HA, by age and inhibitor status in the whole UK Haemtrack population	To evaluate the joint health in Nordic people with moderate HA/HB in relation to their treatment modality
Number of patients with non‐severe haemophilia	67 patients with moderate HA and 8 patients with moderate HB (n = 75)	163 patients with moderate HA were included in bleeding analysis (self‐reported via Haemtrack)	145 patients, either moderate HA (n = 89) or moderate HB (n = 56)
Baseline endogenous factor activity, median (IQR)	FVIII/FIX:C = .03 (.02–.04) IU/mL	FVIII:C = .01 (.01–.02) IU/mL	FVIII:C = .03 (.02–.04) IU/mL FIX:C = .02 (.01–.02) IU/mL
Treatment groups	All patients were receiving secondary prophylaxis factor replacement therapy	On‐demand/episodic: n = 49 Regular prophylaxis: 106/154 (68.8%) patients without inhibitors and 8/9 (88.8%) of patients with inhibitors	HA: 34 (38%) HA: 21 (38%)
Inhibitor	No history of inhibitors	Current, n = 9	History of inhibitor, n = 3 (2%)
Annual factor consumption	Median 148 IU/kg (range, 0–2903 IU/kg)	Not reported	HA: 333 (18–2600) IU/kg HA: 88 (0–2109) IU/kg
Age of first joint bleed, median (IQR)	4.8 (3.5–8.5) years	Not reported	HA: 5 (3–7) years HA: 7 (5–12) years
Age of starting prophylaxis treatment, median	14 years (range, 1–59 years)	Not reported	HA: 10 years (IQR, 4–22 years) HA: 10 years (6–37 years)
ABR, median (IQR)	2.0 (.8–3.7)	Without inhibitors On‐demand/episodic: 11.0 (4.8–20.3) Prophylaxis: 3.0 (1.0–7.0) With inhibitors On‐demand/episodic: not reported Prophylaxis: 2.0 (.5–4.5)	HA: 0 (0–1) HA: 0 (0–0) (no difference between treatment modalities)
AJBR, median (IQR)	Prior to prophylaxis: 4.1 (2.5–6.5) After prophylaxis: .0 (.8–3.7)	Without inhibitors On‐demand/episodic: 5.0 (2.0‐15.3) Prophylaxis: 2.0 (.0‐5.0) With inhibitors Not reported	HA: 0 (0–1) HA: 0 (0–0) (no difference between treatment modalities)
PWH who were bleed‐free, %	Not reported	Without inhibitors On‐demand/episodic: 10% Prophylaxis: 24%	Not reported
PWH who were joint bleed‐free, %	Not reported	Without inhibitors On‐demand/episodic: 15% Prophylaxis: 30%	Not reported
HJSH score*	82% scored < 10 out of 128 Median (IQR) for: Early onset joint bleeding (≤5 years; n = 15): 8 (0–12) Late onset joint bleeding (>5 years; n = 60): 3 (0–8)	Score not reported for moderate haemophilia but reported to be similar pattern as in severe HA (median [IQR] baseline FVIII:C for the 122 people with moderate haemophilia who had HJSH performed was .02 [.01–.03] IU/mL). Median HJHS increased incrementally with age, both in people using on‐demand/episodic and those using prophylaxis treatment	HA: 2 (0–10) HA: 4 (1–9) (n = 135) Not reported separately for different treatment modalities
HEAD‐US	Not reported	Not reported	0/48 points for both HA and HA (n = 118) Not reported separately for different treatment modalities
Orthopaedic surgeries	13/75 participants; five people had undergone minor orthopaedic surgery, such as synovectomy or excision of cysts, three people had undergone at least one ankle arthrodesis and six people had undergone additional joint replacements	Not reported	HA: 16 (18%) HA: 6 (11%) Orthopaedic surgery was more frequent in the prophylaxis group (75% [6/8] vs. 17% [3/18]) (P < .01) in the age group 35–54 years. Ten participants had undergone surgery in more than one joint. Knee arthroplasties accounted for 18 joints in 14 participants, followed by ankle arthrodesis in 12 joints in 11 participants
Quality of life measures	Minimal physical limitations regardless of age of onset of joint bleeding. Participants with FVIII/FIX .01 IU/mL had lower vitality, bodily pain and general health subscores of short‐form 36	Not reported	Not reported
Main study limitations	This was a single centre, observational study, with relatively small number of participants, and did not analysis people with HA separately from those with HB. Due to technical limitations of factor activity testing, all participants with activity levels of .005–.0099 IU/mL were excluded from the study, meaning that few participants were included with an activity level of .01 IU/mL	The people with moderate HA who self‐reported outcomes and treatment via Haemtrack were biased towards those who had a high bleeding risk and relatively low FVIII:C.	The study population was skewed towards younger participants; older people were less likely to enrol often due to distance to travel to the clinic being too great. Baseline FVIII/FIX:C was performed at local laboratories at different times using different methods. Bleed history was collected retrospectively, and records for older participants were sparse
Main study interpretation	Some people with moderate HA or HB have high bleeding frequencies despite prophylactic treatment and tend to start prophylactic treatment later than patients with severe disease. Some people with moderate disease require additional treatment to reduce frequency of bleeding events. The authors recommend primary prophylaxis in all PWH who experience their first joint bleed before 5 years old	Some people with moderate HA and high bleeding risk experience arthropathies from early adulthood onwards	The study demonstrated a strong correlation between HEAD‐US and HJHS, validating both for evaluating joint assessment. Although most people with moderate HA or HA had good joint health, a subset had severe arthropathy. Baseline FVIII/FIX:C ≤.03 IU/mL was associated with a younger age at first joint bleed and higher HJHS. MoHem study indicates a need for more extended use of prophylaxis among people with moderate haemophilia and a high bleeding risk prior to joint disease
	All studies show that non‐severe PWH should be better monitored, and that prophylaxis should be started very early in life according to similar strategies as for severe haemophilia. Factor levels at or below 3% or people with bleeding risk should be sufficient reasons to start prophylaxis treatment with factor replacement therapy.

^*Studies used different variations of the HJSH score; version 10, ²⁴ version not reported for THUNDER. ^{3 .}

ABR, annualized bleed rate; AJBR, annualized joint bleed rate; FVIII:C, Factor VIII clotting activity; FIX:C Factor IX clotting activity; HA, haemophilia A; HB, haemophilia B, HJSH, haemophilia joint health score; HEAD‐US, haemophilia early arthropathy detection with ultrasound; IQR, interquartile range; PWH, people with haemophilia.

The study by den Uijl et al., ²⁴ assessed short‐ and long‐term outcomes in relation to treatment in 75 people with moderate HA or HB born prior to 2000 and who were receiving secondary prophylaxis (defined as regular continuous prophylaxis initiated after two or more joint bleeds but before onset of joint disease ⁵³ ). Median (IQR) age was 37 (23–52) years and 89% had HA. Median (IQR) age at first joint bleed was 4.8 (3.5–8.5) years. Participants experienced a median of 4.1 (2.5–6.5) joint bleeds before starting secondary prophylaxis. Prophylaxis was defined as at least one regular infusion per week for at least 45 weeks per year. ²⁴ This cohort had a high bleeding risk (endogenous factor activity ranged from .01 to .05 IU/mL) and 29% (n = 22) had a history of primary prophylaxis because of a high bleeding frequency, but the duration of either primary or secondary prophylaxis was not reported. The median ABR was 2.0 after commencing secondary prophylaxis, including a median of 0 joint bleeds/year (range 0–8.8 bleeds/year). A total of 33 PWH (44%) had not experienced any joint bleeds during the preceding 5 years. ²⁴ Most participants (50/60 [82%]) had HJHS < 10/126 points indicating good joint health status. A history of primary prophylaxis was associated with younger age at first joint bleed (P < .01) but not baseline factor activity (P = .12). Orthopaedic outcomes were reported for 13 people, all born prior to 1965, 10 of whom underwent orthopaedic surgery and had not always used prophylaxis treatment in the past. Overall, QoL, as measured by the short‐form 36 (SF‐36) questionnaire, was similar to the general population but people with low endogenous FVIII (.01 IU/mL) scored lower on vitality, bodily pain and general health SF‐36 subscores. Median haemophilia activities list (HAL) score was 96 out of 100 (IQR, 83–100), indicating minimal physical limitations. The study did not report QoL HB outcomes separately from HA. ²⁴

Results from the THUNDER study, ³ a study of the UK national haemophilia database that included people with moderate HA, showed that some people with moderate HA have a high bleeding risk and experience arthropathies from early adulthood onwards. The study described current standard of care and evaluated unmet needs among people with severe or moderate HA using Haemtrack, an internet‐ and smartphone‐based system for people to self‐report treatment and symptoms, which integrates with the UK National Haemophilia Database (NHD) and electronic patient records. The study included 864 people with moderate HA who were registered with the NHD during 2015; however, only 163 of these people had completed enough Haemtrack entries to be considered compliant with regards to their bleeding and treatment history. People with moderate HA who were Haemtrack compliant had a high risk of bleeding and a median endogenous FVIII activity of .01 IU/mL, representing a subset of people with moderate HA. ³ A total of 106/154 (68.8%) had no inhibitors reported regular prophylaxis. For those receiving prophylaxis, ABRs/AJBRs were higher than people with severe HA (median [IQR] ABR/AJBR for moderate HA: 3.0 [1.0–7.0]/2.0 [.0–5.0; severe HA: 2.0 [.0–7.0]/1.0 [.0–4.0]). Joint assessments using Haemophilia Joint Health Score (HJHS) were made for 122 PWH and median scores increased incrementally with age regardless of treatment modality, but the authors did not report age of first joint bleed nor age prophylaxis started. ³

The third study, the MoHem study looked at joint health of Nordic people with moderate HA or HB in relation to their treatment regimen. ⁹ A total of 145 PWH were evaluated, comprising 89 and 56 people with HA and HB, respectively. ⁹ Median baseline FVIII/FIX activity was .02 IU/mL (interquartile range [IQR], .02–.04 IU/mL). Median age of first joint bleed was 5 years (IQR, 3–7 years) for people with HA and 7 years (5–12) for people with HB. A total of 38% (HA: 34/89; HB: 21/56) received prophylaxis and the median age at start of prophylaxis was 10 years (IQR, 4–24 years), indicating a delay between first joint bleed and starting prophylaxis. Younger PWH were more likely to receive prophylaxis treatment, supporting the trend towards using prophylaxis for younger people with moderate disease in Nordic countries. HJHS was higher (indicating more joint problems) for people treated with prophylaxis compared with those treated with on‐demand/episodic factor concentrates, and scores increased with age. It is likely that people who received prophylaxis were at higher risk of bleeding and had experienced more joint bleeds prior to starting prophylaxis treatment, particularly, as there was a delay between the first joint bleed and the initiation of prophylaxis. ⁹ Orthopaedic surgery was also more frequent in the prophylaxis group versus on‐demand group among the age of 35–54 years (75% [6/8] vs. 17% [3/18]; P < .01), again indicating that higher bleeding risk is associated with longer‐term joint health issues compared with lower bleeding risk.

A more recent publication described the first magnetic resonance imaging (MRI) study of joints in 51 people with non‐severe HA, of whom only 2 out of 19 people with moderate HA received prophylaxis. ⁵⁰ The median total HJHS score was 3 (IQR 2–7) for the total cohort and 7 (IQR 4–11) for those with moderate HA. Ankle joints were the most frequently and structurally affected with haemosiderin deposits seen in 34% of ankles, 3% of knees and 8% of elbows, and 14% of bleed‐free joints were observed to have haemosiderin deposits. ⁵⁰ In comparison, people with moderate HA from the MoHem study had a lower median total HJHS score of 2 (IQR 0–10), ⁹ although the MoHem cohort was younger (median age 28 and 43 years, respectively). ⁹ , ⁵⁰

3.5. Risk of inhibitors in non‐severe haemophilia

The risk of inhibitor development has been reported as ∼4% ³ in moderate HA as opposed to ∼30% ⁵⁴ in severe HA. Overall, the risk of inhibitors is expected to be lower in non‐severe HA, but there is limited evidence about the risk of inhibitors in cohorts of mild or moderate HA patients exposed to factor VIII. Also, the strength of the association between life‐long risk in non‐severe HA and cumulative exposures to FVIII concentrates (and thus age) is largely unknown. People with non‐severe HA continue to be at risk beyond 50 or 100 exposure days, ⁴⁰ , ⁴⁹ whereas the risk is highest after 10–15 exposure days in severe HA. ⁵³ Intensive FVIII treatment is associated with inhibitor development in people with non‐severe HA; for example, when a PWH undergoes major orthopaedic surgery or experiences a massive bleed. ⁴³ , ⁴⁹ , ⁵⁵ Moreover, for people with mild HA the administration of FVIII concentrates through continuous infusion has been associated with inhibitor development. ⁴⁴ , ⁴⁶ , ⁵⁶ Specific missense mutations in the F8 predispose inhibitor development (e.g. Arg593Cys) and could be used to identify PWH at risk prior to planned surgery. ⁴⁴ , ⁴⁵ , ⁴⁹ Nonetheless, inhibitor development in non‐severe PWH is often missed even when their F8 mutation is known to be high risk. ⁴⁸ If intensive treatment with FVIII is required for surgery, monitoring for inhibitor development and additional treatment with desmopressin should be considered as the latter does not have an inhibitor risk.

3.6. Knowledge gaps in our understanding of non‐severe haemophilia

Despite advances in care for people with severe haemophilia, there remain many knowledge gaps, particularly when it comes to providing care for people with non‐severe disease (Table 3). These gaps are due, in part, to people with milder symptoms not receiving documented treatment, ⁷ , ⁵⁷ or when treatment regimens are documented, outcomes are not consistently reported. ³⁹ Conversely, some cohort studies report outcomes, but not in respect to treatment regimens. ⁶ , ³⁷ , ³⁹ A recent literature review on mild haemophilia has also highlighted a dearth of studies, but noted that mild disease has a measurable impact on QoL and productivity. ⁵⁷ Although the cross‐sectional PROBE study has provided insights into health‐related QoL in people with mild (n = 102) or moderate (n = 134) haemophilia, showing that they have significant reductions in health‐related QoL in terms of pain, reduced mobility and fewer activities of daily living compared with age‐ and sex‐matched people without haemophilia (n = 173), little is known about the long‐term health status of people with non‐severe haemophilia. ⁵ Longitudinal studies are required to understand how to optimise care of people with non‐severe haemophilia. ⁵

TABLE 3.

Evidence limitations for non‐severe haemophilia

Limitation	Implication
Prevalence of mild disease	An unknown number of people may be at risk of bleeding events and long‐term consequences of disability and pain in later life
No data on arthropathy, joint bleeds, pain or long‐term prognosis specifically for people with mild haemophilia who have been treated with on‐demand/episodic desmopressin	It is not known whether on‐demand/episodic desmopressin treatment, the most common treatment for people with mild haemophilia A, is sufficient to stop people with mild disease experiencing joint bleeds that lead to arthropathies later in life. There is some evidence that for many people with joint or muscle bleeds that desmopressin is not sufficient treatment prevent these types of bleeds and these people require additional factor replacement therapy to prevent joint bleeding event 25
The lack of knowledge around non‐severe disease is compounded by these PWH making fewer visits to haemophilia centres	There is a lack of medical evidence available to clinicians for making treatment decisions for people with mild disease
It is not yet known whether ongoing long‐term prophylaxis will prevent arthropathy in PWH who have already experienced joint bleeds in PWH with non‐severe disease. The prevalence of comorbid joint disease from other causes is unknown	Although there is some evidence to suggest that secondary prophylaxis helps prevent joint bleeds, people still experience arthropathies in later life if at high risk of bleeding
There is limited evidence about the timescale of joint complications in people with mild or moderate haemophilia compared with severe disease	As the natural history of joint disease in people with non‐severe haemophilia is unknown, it is hard to determine optimal preventive treatment
Disease natural history and aging with non‐severe disease is not known	Treatment decisions for elderly people with non‐disease haemophilia may be complicated by age‐associated comorbid conditions
Minimal information about using FVIII prophylaxis use in people with mild haemophilia	Using prophylaxis factor replacement in people with high risk of bleeding but with endogenous factor activity > .03 IU/mL may benefit some people with high bleeding risk due to their preferred level of physical activity, concomitant conditions or other factors
Limited data specifically about prophylaxis factor replacement in women with symptomatic non‐severe HA or HB, or people with non‐severe haemophilia B	Unknown number of people who could benefit from factor replacement prophylaxis
Limited evidence about mental health issues around prophylaxis treatment for people with non‐severe haemophilia, such as treatment anxiety	Factor replacement prophylaxis in people with non‐severe haemophilia will add burden of care to the PWH or carers
Most patient‐reported outcomes or bleeding risk instruments have been developed for severe haemophilia	Some haemophilia‐specific instruments may need to be re‐evaluated for people with non‐severe disease, particularly if treatment response is being assessed
Limited evidence about whether the risk of developing inhibitors is different for people with mild disease depending on whether they receive prophylaxis factor replacement or on‐demand/episodic treatment	Choosing a treatment regimen that reduces the risk of inhibitor development may prevent inhibitor‐related complications

HA, haemophilia A; HB, haemophilia B; PWH, people with haemophilia.

It is also uncertain how many people are living with non‐severe haemophilia worldwide. One estimate, calculated from a meta‐analysis of registry data, showed that although there were approximately 200,000 people living with haemophilia in 2017 (as reported by the World Federation of Haemophilia Annual Global Survey), the expected global prevalence of non‐severe disease in males is approximately 700,000, a number which increases to 1.125 million when people with severe disease are included. ⁵⁸ As such, the vast majority of PWH remain undocumented. Moreover, there remains a lack of information about women with haemophilia.

3.6.1. Mortality risk in people with non‐severe haemophilia

Overall life‐expectancy of people with non‐severe haemophilia with access to treatment is considered to be similar to people without haemophilia. ⁵⁹ The INSIGHT study showed that fatal intracranial bleeding was 3.5‐fold higher among its 2709 non‐severe HA cohort (mild, n = 1990; moderate, n = 719) compared with healthy aged‐matched males. ²⁷ The median cumulative number of days of exposure to FVIII concentrates was 13.5 (IQR 6–33), indicating that the cohort had not received prophylaxis treatment, ²⁷ which has been shown to protect against intracranial haemorrhage in severe haemophilia. ⁶⁰ Overall, there is insufficient evidence to know whether prophylaxis treatment would help to reduce mortality risk in non‐severe haemophilia.

3.7. Prophylaxis use of clotting factor replacement products in people with non‐severe haemophilia

Clinicians should consider the treatment most likely to achieve a zero‐bleeding risk in people with non‐severe haemophilia and tailor dosing according to bleeding phenotype and the person's lifestyle. This is especially important in children, as physically activity with no major restrictions is important for musculoskeletal development.

There is difficulty in defining disease severity by endogenous factor levels, as these do not necessarily correspond with bleeding risk as evidenced by some people with higher endogenous FVIII concentrations having a higher bleed risk than others with lower concentrations. ²⁵ Although distinction between moderate and mild haemophilia remains valid, categorization may be less clear cut due to inter‐PWH variability related to variables beyond factor levels. This underlines the need to consider prophylaxis particularly if joint bleeds are reported during early childhood. ⁸ , ⁶¹ Indeed, den Uijl et al. recommend that prophylaxis should be considered for all children who experience a joint bleed before the age of 5 years. ²⁴ The authors of the THUNDER study recommend that prophylaxis be individualized not only based on bleeding risk but also pharmacokinetics. ³ As there is an association between a delay in starting prophylaxis after the first joint bleed and people experiencing worse joint health in adulthood, ⁹ , ²⁴ there is a strong rationale to start prophylaxis factor replacement after the first joint bleed, particularly for people with endogenous factor levels .01–.03 IU/mL; however, in accordance with WFH treatment guidelines primary prophylaxis should be considered if FVIII/FIX is less than .3 IU/mL prior to first joint bleed or if a person has a severe bleeding phenotype. ⁵³

In order to reduce treatment burden, an appropriate extended half‐life FVIII product administered once weekly may be sufficient to reduce bleeding risk to zero for people with HA who could benefit from prophylaxis. Interim results of the HAVEN 6 study show that emicizumab may also be an effective treatment to prevent bleeds for people with non‐severe HA. ⁶² Regardless, clinicians need to align treatment goals and the individual needs. For some people with non‐severe haemophilia, prophylaxis may increase FVIII/FIX to desired trough levels to provide protection against spontaneous bleeds allowing no restrictions on physical or social activities.

3.8. Limitations

The main limitation of our review is the paucity of studies that report factor replacement prophylaxis among people with non‐severe haemophilia. Most studies did not specify treatment type, and the studies that did report treatment regimens were mostly located in high‐income countries, including Nordic countries, that typically have a higher use of prophylaxis among PWH with non‐severe disease than other countries. As such, we are unable to determine the proportion of people with non‐severe haemophilia who receive factor replacement prophylaxis. Another limitation is that our review was not comprehensive, as the search terms did not identify publications that use uncommon terminology or that have multiple cohorts not focused on mild or moderate haemophilia.

3.9. Discussion of clinical implications

There is the need for individualized treatment regimens for all PWH, regardless of disease severity as defined by residual plasma clotting factor levels, based on individual bleeding tendency, personal preference, lifestyle and the person's pharmacokinetic profile. ⁸ Recent therapeutic advances has significantly ameliorated the management of people with severe haemophilia, rendering people with moderate haemophilia the ‘new severe’ and highlighting the heterogeneity of mild patient population (due to the wide range of baseline FVIII/FIX levels). Furthermore, the type assay used to assess endogenous factor activity may impact treatment decisions particularly for people classified as having mild HA. ²³ , ²⁹ This highlights the importance of assessing each person's bleeding risk based on individual circumstances, including women with haemophilia who experience debilitating heavy menstrual or gynaecological bleeding, rather than classifying their treatment needs based on their endogenous factor activity. Any PWH with a high bleeding risk may benefit from prophylactic treatment, giving them the FVIII/FIX levels needed to protect their joints, as well as during periods of physical activity to protect against spontaneous bleeds. In summary, prophylaxis in people with non‐severe haemophilia should principally be addressed similarly to what is seen with people with severe haemophilia, although decisions about start and dosing differ.

AUTHOR CONTRIBUTIONS

Study conception, design and supervision of data acquisition: EB

Data analysis and interpretation: All authors

Manuscript preparation first draft: EB

Critical review of multiple full drafts: All authors

Approval of final manuscript draft: All authors

CONFLICT OF INTEREST STATEMENT

AI, CK, MTR, DR, MS, EB are members of the Council of the Hemophilia Community (CHC), which is organized and funded by Bayer.

AI's institution has received project‐based funding via research or service agreements with Bayer, Biomarin, CSL, Freeline, Grifols, NovoNordisk, Octapharma, Pfizer, Roche, Sanofi, Sobi, Spark, Takeda, and Uniqure.

MEM has acted as paid consultant, advisor or speaker for Bayer, Biomarin, CSL Behring, Grifols, Kedrion, LFB, Novo Nordisk, Octapharma, Pfizer, Roche, Sanofi, Sobi, Spark Therapeutics, Takeda and UniQure.

DR has been a paid consultant to Bayer but all payments go to National Hemophilia Foundation.

MWS's institution has received research funding from BioMarin, Freeline, Roche, Takeda, uniQure; and fees for advisory board or educational presentations from: Bayer, BioMarin, Pfizer (DMC), Roche/Genentech, Sanofi, Spark (DMC), Takeda.

MTR has acted as a paid consultant for advisory boards or speaker bureaus for Bayer, BioMarin, CSL Behring, HemaBiolgics, Novo Nordisk, SanofiEnzyme, Takeda and his institution has received research support from Bayer and BioMarin.

CK has acted as a paid advisor or speaker for Bayer, CSL Behring, Novo Nordisk, Roche/Chugai, Sobi/Sanofi and Takeda and his institution has received research funding from Bayer, Biotest, EB has acted as paid consultant to Bayer, CSL Behring, Octapharma, Sobi, Takeda, and has received funding for research from Bayer, CSL Behring, Shire/Takeda, Sobi and Bioverativ.

CSL Behring, Intersero, Novo Nordisk, Pfizer, Sobi/Sanofi and Takeda.

Medical writing support was provided by Ingrid Koo, PhD, and Celia J. Parkyn, PhD, from Fleishman Hillard, and was funded by Bayer.

Supporting information

Supplementary Information

Iorio A, Königs C, Reding MT, et al. Prophylaxis use of clotting factor replacement products in people with non‐severe haemophilia: A review of the literature. Haemophilia. 2023;29:33–44. 10.1111/hae.14676

DATA AVAILABILITY STATEMENT

All data included in this review are in the public domain.