Meta-Analysis on Incidence of Inhibitors in 1,945 Previously Untreated Patients Treated with Recombinant Factor VIII Products: Is There a Difference?

Introduction. Recent large cohort studies have shown data on inhibitor incidence in previously untreated patients (PUPs) with hemophilia A treated with recombinant factor VIII concentrates (rFVIII), which disputed previous reports or clinical trials. Aim of this meta-analysis was to assess the risk of inhibitor development in this patient population in order to clarify the relationship between the rFVIII product used and the development of FVIII inhibitors.

Methods. We carried out a systematic literature search in electronic databases (Medline through PubMed, EMBASE) for studies published from 1 January 1988 to 31 January 2015. We aimed to identify clinical studies (both prospective and retrospective) investigating the relationship between rFVIII product used and the development of FVIII inhibitors in PUPs and minimally treated patients (MTPs), with less than 5 previous exposure days, with severe (FVIII<1%) and moderate (FVIII1-5%) hemophilia. We conducted the systematic review in accordance with the MOOSE (Meta-analysis of Observational Studies in Epidemiology) guidelines.

The primary outcome measure of this meta-analysis was the development of clinically relevant inhibitory antibodies, while the secondary outcome measure was the development of a high-titer inhibitor, which was defined as a peak titer of at least 5 Bethesda units per millimeter (BU/mL) up to the 75th exposure day.

We computed pooled meta-analytic estimates according to the rFVIII product used by using the inverse-variance method, assuming a fixed, or a random-effects model if significant between-studies heterogeneity was present. Heterogeneity was measured using Cochrane Q statistics and Higgins I-squared statistics, i.e. the ratio of true heterogeneity between studies compared to the total observed variation.

Results. We identified 717 papers through database searches (538 from PubMed; 181 from EMBASE), of which 662 papers were excluded for the following reasons: duplicates (n=49); title and/or abstract not relevant for the endpoint of this study (n=231); review articles (n=131); letter to the editor (n=33); case-reports (n=98); animal studies (n=40); and others (n=80). Fifty-five (55) full-text papers were retrieved for detailed evaluation, plus eight additional papers identified through review of reference lists of the retrieved articles. Of these, 47 articles were excluded after full-text evaluation (incompleteness of data, multiple publications on the same cohort); sixteen (16) studies were included in the final meta-analysis.

The pooled estimate of all inhibitors (571 inhibitors out of 1,945 PUPs/MTPs), including high titer (>5 BU/mL) and low titer (≤ 5 BU/mL), in the overall population considered was 0.27 (95% CI: 0.23-0.31) with significant heterogeneity (I-squared: 68.9%, p<0.01). Considering only inhibitors in PUPs with severe hemophilia (353 inhibitors out of 1,223 PUPs), the pooled estimate of all inhibitors was 0.27 (95% CI: 0.23-0.32) with significant heterogeneity (I-squared: 64.9%, p<0.01). The heterogeneity observed was due to differences of inhibitor incidence among studies for the same product used. Pooled inhibitor incidence estimates among products used ranged from 0.23 to 0.42 with overlapping confidence intervals, in the absence of heterogeneity (p=0.28). Similar patterns were observed in subpopulations of patients with high titer inhibitors or low titer inhibitors.

Only few studies reported inhibitor hazard ratios with the different products used, adjusted to slightly different potential risk factors (Gouw et al, 2007, Gouw et al, 2013, Calvez et al, 2014, Collins et al, 2014). Meta-analysis of these studies showed PUPs/MTPs treated with Antihemophilic Factor (Recombinant) (Advate, Baxalta US Inc., Westlake Village, CA 91362) had a pooled inhibitor hazard ratio estimate of 0.63 (95% CI 0.48-0.83) as compared to Antihemophilic Factor (Recombinant) (Kogenate FS, Bayer HealthCare LLC, Whippany, NJ 07981). The pool estimates of other products used and tested were not significantly different.

Conclusions. The incidence of inhibitors in PUPs/MTPs included in this meta-analysis was 27%.  Although significant heterogeneity among studies was observed, no significant differences among products were found. Differences between products used were found only by considering hazard ratios in which potential risk factors were considered.