Alloantibodies in previously untreated hemophilia A patients: the role of environmental factors

Abstract

The therapy of hemophilia A has nowadays reached a high degree of quality, being probably the most efficacious and safe treatment available among other monogenic disorders. In this context, the most challenging complication of therapy has become the development of inhibitory alloantibodies against FVIII. These inhibitors, which develop in up to 30% of severe hemophilia A patients, render replacement therapies ineffective, limit patient access to a safe and effective standard of care, and predispose them to an increased risk of morbidity and mortality. Several possible risk factors, both genetic and environmental, for inhibitor development have been identified in previously untreated patients, confirming that the etiology of this phenomenon is multifactorial. In this narrative review, we will focus on acquired risk factors, discussing the pathogenic and clinical data available from the literature analysis.

Keywords:

• Hemophilia A
• Therapy
• Inhibitors
• Alloantibodies
• Bleeding

Introduction

The development during the last 25 years of viral inactivation procedures and of recombinant products has greatly improved the replacement therapy of hemophilia A with a positive impact on the health status of patients, due to the widespread diffusion of prophylaxis to prevent arthropathy, and consequently on their quality of life and life expectancy. In this context, nowadays the most significant debilitating therapeutic complication of hemophilia A, at least in the developed world, has become the development of inhibitors and a continuously increasing number of articles have been published on this topic during the last few years. Inhibitory alloantibodies develop in approximately 20–30% of severe hemophilia A patients, usually within the first 10 exposure days, and render ineffective FVIII substitutive treatment precluding the access of hemophiliacs to a safe and effective standard of care and predisposing to an unacceptably high risk of morbidity and mortality.^1,2

The research in this area has been mainly directed at the identification of possible risk factors contributing to inhibitor development, and previously untreated patients (PUPs) have been considered the ideal population for studying the etiology of this phenomenon.³ These studies have revealed the importance of genetic (e.g. ethnicity, FVIII gene mutations (large deletions, nonsense mutations), major histocompatibility complex (MHC) genotype, polymorphisms of immune-responses genes (interleukin-10, tumor necrosis factor-a, cytotoxic T-lymphocyte antigen-4)) and environmental (e.g. the number of FVIII exposure days, age at first exposure of FVIII concentrate, type of FVIII concentrate administered, and modality of treatment) risk factors in the development of inhibitors.^4–6 All these lines of evidence document that inhibitor formation in hemophilia is a complex multifactorial process.

In this narrative review, I will focus on the role of non-genetic factors in determining the risk of inhibitor development in PUPs with severe hemophilia A and, in particular, I will analyze the published literature data on three of them, which are actually still controversial and debated: age and intensity at first treatment, prophylaxis, and type of FVIII product.

Age and intensity at first FVIII exposure

During the last few years, there has been growing interest in the identification of potentially modifiable factors predisposing the patient to inhibitor development.⁷ The role of these non-genetic factors has been increasingly supported by evolving concepts of the immune response based on the ‘danger model’⁸: the immune system is activated by alarm signals arising from the injured tissues to a greater extent than by the recognition of non-self. In this respect, the presentation of the exogenous FVIII may not be sufficient for initiating an immune response. In the presence of danger conditions (i.e. severe bleeds, trauma, or surgery with major tissue injury), the foreign protein is intensively presented (high-dose and/or prolonged treatment) in association with signals that up-regulate the cellular T- and B-lymphocyte response.

This hypothesis is supported by a number of studies showing a key role of the intensity and age of treatment at the first FVIII exposure (Table 1).^9–16 For instance, in a Spanish cohort of 62 children with severe hemophilia A, 24% of whom developed persistent inhibitors, a statistically significant difference (P = 0.03) in the cumulative incidence of laboratory-detected inhibitor formation based on age at first exposure was demonstrated.⁹ A cohort study from the Netherlands of 81 children with severe hemophilia A confirmed this trend, assessing the incidence at 100 exposure days.¹⁰ An Italian case–control study of children treated exclusively with recombinant FVIII (rFVIII) compared 60 children with inhibitors to 48 inhibitor-free controls. Although there seemed to be an increased trend for inhibitor formation if patients were treated at an early age, this did not hold true after adjustment for genetic factors.¹¹ In a retrospective French cohort study, children treated for the first time after 12 months of age showed, at a multivariate analysis, 0.3-fold fewer inhibitors than those treated before 6 months of age.¹² In another retrospective cohort study from the United Kingdom of 348 children with severe hemophilia A, 20% developed an inhibitor and 10% of these were high titer. When correlating this to age at first exposure, there was a significant decrease in inhibitor development when children were treated after they were 18 months old (20–26% at ages 0–18 months and 9% at age >18 months).¹³ In the retrospective CANAL (Concerted Action on Neutralizing Antibodies in Severe Hemophilia A) study, which investigated 366 consecutive PUPs born between 1990 and 2000 from 14 centers in Europe and Canada,¹⁴ and in the combined analysis of data on 236 patients from the four rFVIII registration PUP studies,¹⁵ surgical and prolonged (≥5 days) FVIII exposure at first treatment, and high FVIII dose (≥50 IU/kg) during the first 50 exposure days were associated with a two-fold to three-fold increase in the risk of inhibitor development. In addition, an association between age at the first treatment (<1 month versus >18 months) and inhibitor development was found at univariate but not at multivariariate analyses.¹⁴ In a recent retrospective case–control study, only high-intensity treatments (treatment peaks of 5 exposure days or more and high-frequency treatment) were found to be an independent risk factor for inhibitor development.¹⁶ Accordingly, an intensive treatment at initial exposure has been considered the most significant determinant (three points) in the prognostic score, which combined genetic- and treatment-related risk factors, determined from the CANAL data.¹⁷ Finally, a recent systematic review by Eckhardt et al.¹⁸ on surgery and inhibitor development in hemophilia A patients found that intensive FVIII treatment for surgery at first exposure leads to a higher inhibitor risk compared with intensive treatment for bleeding or prophylaxis (OR 4.1; 95% CI 2.6–6.5).

Table 1. Age and intensity at first FVIII exposure and inhibitor development in severe hemophilia A patients: analysis of the most important published studies

Author, year (reference)	Study design (no. patients)	Main findings
Lorenzo, 2001^9	Retrospective cohort study (62)	Cumulative inhibitor incidence at 3 years from the first FVIII exposure was 41% in patients starting therapy before the age of 6 months, 29% in patients starting therapy between 6 and 12 months of age, and 12% in those starting therapy beyond 1 year (P = 0.03). At multivariate analysis, the age at initial treatment was an independent variable.
Van der Bom, 2003^10	Retrospective cohort study (81)	Cumulative inhibitor incidence at 100 exposure days was 34% in patients starting therapy before the age of 6 months, 20% in patients starting therapy between 6 and 12 months of age, 13% in those starting therapy between 1 and 1.5 years, and 0% in those starting therapy beyond 1.5 years (P = 0.03).
Santagostino, 2005^11	Case–control study (108)	A trend was found for an increased risk (OR 2.8; 95% CI 1.0–7.6) of inhibitor development in children first treated at a young age (≤11 vs. >16 months); however, this was not confirmed after adjusting for genetic factors.
Goudemand, 2006^12	Retrospective cohort study (148)	At multivariate analysis, children treated for the first time after 12 months of age showed fewer inhibitors (RR 0.3; 95% CI 0.1–0.7) than those treated before 6 months of age.
Chalmers, 2007^13	Retrospective cohort study (348)	At univariate analysis, a significant difference in inhibitor development and age at first exposure across all age groups was observed (20–26% at ages 0–18 months vs. 9% at ages >18 months, P = 0.018). At multivariate analysis, only FVIII gene mutation remained significant.
Gouw, 2007^14	Retrospective cohort study (366)	At univariate analysis, but not at multivariate analysis, age at first treatment (<1 month versus >18 months) was associated with an increased RR of inhibitor development (2.7 [95% CI 1.2–5.9] for all inhibitors and 2.4 [95% CI 1.0–5.6] for high-titer inhibitors). Surgical procedures and peak treatment moments at the start of treatment increased inhibitor risk (RR 3.7; 95% CI 2.0–7.1 and 3.3; 95% CI 2.1–5.3, respectively).
Gouw, 2007^15	Combined analysis of four registration studies (236)	Age at first exposure was not associated with inhibitor development. Peak treatment moments and surgical procedures were related to an increased inhibitor risk (RR 1.6; 95% CI 1.0–2.6 and 2.7 95% CI 1.3–5.7, respectively)
Maclean, 2011^16	Retrospective case–control study (78)	No association was found between inhibitor development and age at first FVIII exposure. High-intensity treatment increased inhibitor risk around 2.5 folds.

RR, relative risk; OR, odds ratio.

On the other hand, it is likely that the conflicting data from the literature on this issue could be attributable to the presence of several genetic and non-genetic confounders, such as FVIII mutation, product type, ethnicity, family history, etc., which further testify the complex multifactorial nature of inhibitor formation.

Prophylaxis

According to the danger model, a number of studies have assessed whether regular exposure to lower doses of antigen in the absence of danger signals, which occurs during prophylaxis, may induce the tolerization of the foreign protein.

After some indirect data or non-controlled studies,^19,20 a protective effect of prophylaxis against the development of inhibitors was shown at multivariate analysis in a case–control Italian study,¹¹ which reported a 70% reduction of inhibitor risk in children starting prophylaxis at a median age of 35 months, compared with children receiving on-demand treatment. Similar findings came from the large multinational CANAL study, in which early regular prophylaxis (introduced at a median age of 20 months) was an independent predictor associated with a 60% reduction of risk of inhibitor development as compared with on-demand treatment.¹⁴

Starting from these pathophysiological and clinical observations, Kurnik et al.²¹ demonstrated that the incidence of inhibitors was markedly reduced by early prophylaxis with once-weekly low-dose FVIII during the first 20 exposure days, thereby minimizing danger signals related to the need of intensive exposure to FVIII. Interestingly, only one of 26 patients receiving such a new regimen developed inhibitors compared to 14 of 30 children retrospectively analyzed treated with standard prophylaxis, resulting in a >90% reduction of risk. However, preliminary results from 606 patients enrolled in the RODIN (Research Of Determinants of INhibitor development) study, recently published in abstract form showed that regular prophylaxis prevented the development of late inhibitors (after about 20 exposure days) for both all the inhibitors and high responder inhibitors.²²

Type of FVIII product

The role of FVIII replacement therapy (i.e. plasma-derived (pd) versus recombinant products) in the likelihood of developing inhibitors has been explored by a number of observational multicenter PUP studies during the last two decades and some investigators have postulated that pdFVIII concentrates might have a lower inhibitor incidence thanks to the protective effect of their von Willebrand factor (VWF), which would mask the epitope sites of inhibitors on FVIII molecule or would protect against FVIII by dendritic cells, thus preventing presentation to CD4⁺ cells and subsequent antibody production.²³ However, several criticisms can be raised against those in vitro studies supporting the first mechanism due to the difficulty in reproducing in the laboratory the physiological in vivo conditions (where circulating VWF is in great excess compared to FVIII and immediately binds exogenous FVIII molecules).²⁴ Furthermore, the fact that VWF attached to FVIII masks the C2 domain on the FVIII molecule, and are identified to be significant in antibody production, does not account for alloantibodies directed to the A2 or A3 domains.²⁵ Contradictory studies have recently been published on the second mechanism.^26,27

Clinical studies have also yielded conflicting results.²³ A prospective study was conducted by Mauser-Bunschoten et al.²⁸ that compared the incidence of inhibitors in 59 severe hemophilia A PUPs, who were initially treated with cryoprecipitate or intermediate purified FVIII concentrates, with that in 22 patients exclusively treated with monoclonally purified FVIII and rFVIII products. The inhibitor incidence rate was very similar between the two groups (24 versus 23%, respectively). In another prospective, comparative study of PUPs with moderate to severe hemophilia A, Kreuz et al.²⁹ detected inhibitors in 18/51(35%) patients treated with pdFVIII and 4/21(19%) patients treated with rFVIII. In the severe subgroup, inhibitors developed in 46% of pdFVIII- and 36% of rFVIII-treated patients. No difference was observed in the development of high-titer inhibitors among the severe group treated with pdFVIII (13/35; 37%) compared with rFVIII (4/11; 36%). In the prospective German Thrombosis and Haemostasis-Previously untreated patients (GTH-PUP) study, a preliminary evaluation in 2003 demonstrated a non-statistically significant difference in the cumulative inhibitor incidence in 57 severe hemophilia A patients treated with one pdFVIII concentrate and 47 patients using one rFVIII product (21 versus 36%, P = 0.08).³⁰

Goudemand et al.¹² reported a more than doubled risk of inhibitors in a historical cohort of 86 severe hemophilia A PUP patients treated with a full-length rFVIII product (inhibitor risk, 31%) relative to a similar cohort of 62 patients treated with a single pdFVIII product containing large amounts of VWF (inhibitor risk 11%; adjusted relative risk (RR) 2.4; 95% CI 1.0–5.8; P = 0.049). When only high-titer inhibitors were considered at multivariate analysis, the statistical evidence disappeared (cumulative inhibitor incidence; 5% with pdFVIII versus 15% with rFVIII, P = 0.157). Similarly, in a retrospective cohort study on 348 severe hemophilia A children, Chalmers et al.¹³ observed a higher risk of inhibitors in patients initially treated with rFVIII products than in those treated with pdFVIII concentrates (27 versus 14%, respectively; P = 0.009); however, this difference was no longer statistically significant when high-titer inhibitors were considered (15 versus 10%, P = 0.173) or at multivariate analysis including other potential risk factors (OR 1.83, 95% CI 0.9–3.72). In the CANAL study,³¹ the risk of inhibitors was not lower in the patients receiving pdFVIII compared with those treated with rFVIII (RR for all inhibitors, 0.8; 95% CI 0.5–1.3; P = 0.34; RR for high-titer inhibitors, 0.9; 95% CI 0.8–2.5; P = 0.72) and no difference was detectable when low- or high-VWF content products were separately considered. Furthermore, the switching between FVIII products did not increase the risk of inhibitors (RR 1.1; 95% CI 0.6–1.8).

In 2003, Wight and Paisley³² published a systematic review investigating the association of FVIII product type with inhibitor formation. Fifty relevant retrospective or prospective studies were identified in the literature. A comparison of 13 studies on PUPs found that patients treated with pdFVIII had a lower cumulative inhibitor incidence than those treated with rFVIII. In patients treated with pdFVIII, the cumulative inhibitor incidence ranged from 0 to 12.4% (weighted mean: 6.8%) for all inhibitors and from 0 to 2.5% (weighted mean: 1.4%) for high responders (>5 Bethesda Units, BU). In contrast, patients treated with rFVIII reported cumulative inhibitor rates between 36 and 38.7% (weighted mean: 37.5%) for all patients; for high responders, the inhibitor incidence ranged from 11.3 to 18% (weighted mean: 15.1%). However, several methodological criticisms have been raised against this review which compared very heterogeneous trials in terms of study design (e.g. prospective/retrospective, the frequency and method of inhibitor testing, length of follow-up) and study populations (ethnicity, type of gene mutation, definition of disease severity, age at first exposure to FVIII, the number of exposure days, treatment regimen, etc.),³³ making it impossible to draw any conclusion based on the comparison of inhibitor incidence of the different products across studies.

The more recent meta-analysis by Iorio et al.³⁴ identified 2094 patients, from 24 retrospective and prospective studies, developing 420 inhibitors. The pooled incidence inhibitor rate was 14.3% for pdFVIII concentrates and 27.4% for rFVIII products (P < 0.001), which however lost statistical significance at multivariate analysis, where the study design, study period, testing frequency, and median follow-up were found to explain most of the variability. Similar results were observed when the analysis was restricted to the 19 prospective studies (9.1% for pdFVIII concentrates and 23.7% for rFVIII products, P < 0.001). However, the statistical significance disappeared again when only high-titer inhibitors were considered (6.0% with pdFVIII versus 19.4% with rFVIII products, P = 0.195).

In an attempt to control as much as possible the variables of the different studies, to render the data more homogeneous, we have conducted a systematic review and meta-analysis using selective criteria (i.e. only prospective studies, PUPs, and severe hemophilia A patients).³⁵ The study included data from 800 PUPs with severe hemophilia A enrolled in 25 prospective studies and found no statistically significant difference in inhibitor rate between patients treated with plasma-derived and recombinant FVIII products (21 versus 27%). Similarly, high-titer inhibitors did not differ significantly between the two groups (14% with plasma-derived versus 16% with recombinant FVIII products). Table 2 summarizes the results of the most important comparative studies and of the three published systematic reviews. Interestingly, comparing the results of our meta-analysis³⁵ with those of Iorio et al.³⁴, it emerges that while the rate of inhibitor (overall and high-responding) development with rFVIII products is very similar between the two studies, the difference becomes substantial when pdFVIII products are analyzed. This is due, in our opinion, to the lower quality of the pdFVIII studies as compared with those more recent evaluating rFVIII products. This was clearly outlined during the quality assessment (using the Newcastle-Ottawa Scale (NOS) and Strengthening the Reporting of Observational studies in Epidemiology (STROBE)) and with an ad hoc score system (Table 3, unpublished data), which evaluated seven items (population size, median follow-up, exposure days, definition of severity of hemophilia, inhibitor lower limit, pharmacokinetic study, inhibitor screening frequency). In particular, the latter technical quality score, which was created in order to evaluate some specific characteristics not considered in NOS and STROBE checklists, documented a statistically significant superiority of studies on rFVIII products as compared with studies on pdFVIII concentrates (median values 9 versus 14; P = 0.01).^36–58 This difference, probably due to the fact that rFVIII studies were conducted more recently and thus used better designed protocols than pdFVIII studies, could explain the lower inhibitor incidence, although not statistically significant, found in the latter studies.

Table 2. Summary of the main results of the published comparative studies and systematic reviews comparing the inhibitor development with the use of rFVIII or pdFVIII concentrates in PUPs with severe hemophilia A

First author, year^ref. (study design – no. of pts./study)	Type of FVIII	Overall inhibitor incidence mean % (range), P		High-responding inhibitor mean % (range), P
Comparative studies
Mauser-Bunschoten, 2001^28 (PS, 81)	CP-pdFVIII	23.7	NS	NA	NA
	MoFVIII-rFVIII	22.7		NA
Kreuz, 2002^29 (PS, 72)	pdFVIII	45.7	NS	37.1	NS
	rFVIII	36.4		36.4
GTH-PUP, 2003^30 (PS, 112)	pdFVIII	21	NS	NA	NA
	rFVIII	36		NA
Goudemand, 2006^12 (RS, 148)	pdFVIII	11.3	0.049	4.8	NS
	rFVIII	31.4		15.1
Chalmers, 2007^13 (RS, 348)	pdFVIII	13.6	0.009	9.8	NS
	rFVIII	27.3		15.1
CANAL, 2007^31 (RS, 316)	pdFVIII	21.5	NS	17.8	NS
	rFVIII	29.3		23.8
Systematic reviews
Wight, 2003^32 (8 RS, 13 PS)	pdFVIII	6.8 (0–12.4)	NA	1.4 (0–2.5)	NA
	rFVIII	37.5 (36–38.7)		15.1 (11.3–18)
Iorio, 2010^34 (5 RS, 19 PS)	pdFVIII	14.3 (10.4–19.4)	<0.001	9.3 (6.2–13.7)	0.004
	rFVIII	27.4 (23.6–31.5)		17.4 (14.2–21.2)
Franchini, 2012^35 (25 PS)	pdFVIII	21 (14–30)	NS	14 (8–25)	NS
	rFVIII	27 (21–33)		16 (13–20)

NA, nota available; RS, retrospective; PS, prospective; NS, not significant; CP, cryoprecipitate; Mo, monoclonally purified.

Table 3. Ad hoc quality score of the studies included in the meta-analysis by Franchini et al.³⁵ (unpublished data)

Author, year (reference)	Population size (n)	Follow-up Median (range)	Exposure days Median (range)	Definition of severe hemophilia	Inhibitor lower limit (BU)	Pharmacokinetic study	Inhibitor screening frequency	Total score
Plasma-derived FVIII concentrates
Lusher, 1990, 1991^Citation36,Citation37	25	3 y	NA	<1%	NA	NA	6 months	7
Ehrenforth, 1992^38	27	8 (0.08–15.25) y	NA	<1%	NA	NA	Every 20th ED	8
Ljung, 1992^39	77	2–22 y	NA	<1%	NA	NA	NA	8
Addiego, 1992^40	30	29 (9–48) m	NA	<1%	NA	5 patients	3 months	9
De Biasi, 1994^41	48	8.5 (0.8–15.5) y	97 (2–753)	<1%	0.8	NA	>annually	10
Yee, 1997^42	37	83 (15–113) m	200 (40–700)	<2%	NA	NA	>annually	10
El Alfy, 2000^43	25	3 y	NA	<1%	1.0	NA	6 months	8
Mauser-Bunschoten, 2001^28	59	14.5 (1.8–23.4) y	NA	<1%	1.0	NA	0.7–2 per year	10
Kreuz, 2002^29	35	23 y	270 (1–3051)	<1%	0.6	NA	Every 20th ED	13
Recombinant FVIII concentrates
Kreuz, 2002^29	11	23 y	270 (1–3051)	<1%	0.6	NA	Every 20th ED	13
Bray, 1994^44	72	7.3 y	94	<2%	0.6	All patients	3 months	15
Gruppo, 1998^45
Goodeve, 2000^46
Rothschild, 1998^7	50	32 (9–74) m	64 (2–840)	<1%	0.6	NA	>6 months	10
Lusher, 1993, 2004^Citation48,Citation49	65	8 y	126 (1–712)	<2%	0.6	38 patients	3 months	15
Yoshioka, 2003^50	31	49.2 (11–80) m	NA	<1%	NA	NA	3 months	8
Giangrande, 2002^51	19	2 y	85 (5–272)	<1%	0.6 (Nijmegen)	All patients	3–4 ED unitl 20 ED, then every 10 ED until 50 ED, then every 3 months	15
Kreuz, 2005^52	37	3.5 y	114 (4–478)	<2%	0.6	All patients	3-4 ED unitl 20 ED, then every 10 ED until 50 ED, then every 3 months	15
Oldenburg, 2006^53
Musso, 2008^54	13	3 y	187 (121)*	<2%	NA	NA	NA	7
Lusher, 2003,2005^Citation55,Citation56	101	6 y	197 (1–1466)	<2%	0.6 (Nijmegen)	11 patients	3 months	16
Courter, 2001^57
Pollmann, 2007^58	16*	24.7 (16.4) m*	142 (93–180)	<2%	0.6	NA	NA	9

m, months; y, years; ED, exposure days; NA, not available; BU, Bethesda Units.

*Mean (SD)

The seven-item score includes: (a) Population size (1: 10 – <20 patients; 2: 20–50 patients; 3: >50 patients), (b) Median follow-up (1: <2 years; 2: 2–5 years; 3: >5 years), (c) exposure days (1: <50 days; 2: 50–100 days; 3: >100 days), (d) definition of severe hemophilia (1: <2% FVIII:C; 2: <1% FVIII:C), (e) inhibitor detection lower limit (1: 0.8 BU or more; 2: 0.6 BU; 3: 0.6 BU [using Nijmegen modification]), (f) pharmacokinetic study (FVIII recovery and half-life). (1: In a subgroup of patients; 2: All patients), (g) inhibitor screening frequency (1: >3 months [≥20 ED]; 2: every 3 months; 3: every 5 ED until the 20th ED, every 10 ED until the 50th ED, then every 3 months.). The minimum score was 0 and the maximum score was 19. A score between 13 and 19 points identified good quality studies; a score between 7 and 12 points identified medium quality studies while low-quality studies were those that reached a score between 0 and 6 points.

Conclusions

The analysis of the current literature data further confirms the complex multifactorial nature of inhibitor development in severe hemophilia A. However, while genetic inhibitor determinants have been well characterized in studies and meta-analysis, less uncertainty still exists as regards environmental factors. Indeed, there is some concern that factor VIII exposure at a younger age, especially less than 6 months of age, contributes to alloantibody formation, although the data are quite conflicting. More convincing data have been published on the role of intensity of treatment (i.e. high dose and/or prolonged treatment), especially in the presence of danger conditions such as surgery, and inhibitor development.

The published literature data suggest a possible protective role for prophylaxis against inhibitor development, especially for late inhibitors (≥20 exposure days).

Finally, there is no conclusive evidence that pdFVIII and rFVIII products are associated with different inhibitor risks. To further address these issues, a number of prospective studies and registries have been started, such as SIPPET, PEDNET, RODIN, GTH-PUP Study and EUHASS. The completion of them will probably help us to answer this issue. For instance, a preliminary 2-year analysis of the data from the prospective registry EUHASS (European Haemophilia Safety Surveillance System) found almost the same inhibitor incidence with pdFVIII and rFVIII products (27 versus 25%).

In conclusion, the studies on the environmental risk factors for inhibitor development in PUPs are very important because while they contribute to elucidate the complex pathophysiology of inhibitor formation, on the other hand they are modifiable and thus represent the ideal target for prevention strategies with the aim of modifying the natural history of alloantibodies, especially in patients with high-risk molecular defects.