Psychometric Evaluation Of Health-Related Quality Of Life Assessments From The Along and Blong Hemophilia Clinical Trials

Hemophilia patients on prophylactic treatment can experience an improvement in health-related quality of life (HRQoL) through better protection from breakthrough bleeds and the reduction in long-term joint damage complications compared to episodic treatment. In two recent Phase 3 clinical trials with new, longer lasting factor concentrates (ALONG: rFVIIIFc; BLONG: rFIXFc), the benefit of HRQoL was assessed as a secondary endpoint in addition to primary objectives of safety, tolerability, and efficacy in adolescents and adults with severe hemophilia A and B, respectively, who received prophylactic or episodic regimens. HRQoL was assessed in the adult population via the EuroQoL-5 Dimension (EQ-5D) and the 46-item Haem-A-QoL questionnaire, yielding 10 domain scores and a total score ranging from 0 (best HRQoL) to 100 (worst HRQoL). Incorporating the hypothesis of improved HRQoL for patients receiving prophylaxis treatment, the goal of this analysis was to use these trials’ data to understand the Haem-A-QoL’s reliability, validity, and sensitivity to change over time in an adult male hemophilia population from six continents.

In ALONG, 133 adult patients (mean age: 36.0 years) had at least one Haem-A-QoL domain score at baseline, with the highest HRQoL mean impairments at baseline in the ‘Sport & Leisure’ (52.4) and ‘Physical Health’ (41.9) domains. In BLONG, the 73 adult patients (mean age: 33.6 years) with at least one baseline domain score also had their highest baseline mean scores in these two domains (56.2 and 42.8, respectively). Internal consistency reliability as assessed by Cronbach’s alpha (α) was generally adequate (α > 0.70) for all 10 Haem-A-QoL domains and for the total score in both trials.

Comparisons of mean Haem-A-QoL baseline domain and total scores for each of the EQ-5D item’s Level 1 responders (no problem) to those at Levels 2 or 3 (some problems or unable) in ALONG demonstrated the validity of most scores (p <0.05). Comparisons between subjects with and without bleeds in the past 12 months also showed important differences (p <0.05) in most Haem-A-QoL baseline domain and the total scores. Convergent validity was tested correlating baseline scores of the Haem-A-QoL with the EQ-5D utility index and the modified Hemophilia Joint Health Score (HJHS). These analyses revealed a strong negative correlation between the two instruments’ total scores (r = -0.63), and between the EQ-5D index and the ‘Physical Health’ (r = -0.63) and ‘Feeling’ (r = -0.62) domains of the Haem-A-QoL. Moderate negative relationships were observed between the EQ-5D scores and the ‘View,’ ‘Sports & Leisure,’ ‘Work & School,’ ‘Treatment,’ ‘Future,’ and ‘Relationships & Sexuality’ domains. The moderate correlations between the ‘Physical Health,’ ‘Feelings,’ ‘Sports & Leisure,’ ‘Work & School,’ and ‘Future’ domain scores and the Haem-A-QoL total score when compared to the HJHS at baseline in the ALONG trial further support the instrument’s construct validity (p <0.001 for all reported correlations).

Change score correlations (baseline to 28 weeks) with the EQ-5D were moderate in magnitude for the Haem-A-QoL total score, and ‘Physical Health’ and ‘Feeling’ domains, demonstrating the sensitivity to change for these outcome measures in the ALONG data (p <0.02). Although differences were not statistically significant (p >0.05), mean change score comparisons between the ALONG treatment arms demonstrated a trend supporting the expectation that improved mean change in the HRQoL total scores (-3.2 and -3.4) were similar for the tailored prophylaxis and weekly dosing arms, respectively, compared to very limited change in the on-demand group (-1.1). This pattern was also observed for the BLONG treatment arms at 26 weeks, where the individualized treatment arm had improved mean change total scores (-4.0) similar to the fixed weekly interval treatment arm (-6.5).

The use of rFVIIIFc and rFIXFc during study was associated with an overall HRQoL improvement from baseline, in addition to demonstrating differences between patients receiving prophylactic or episodic treatment. These psychometric analyses using the ALONG and BLONG clinical trial data underscores this finding by providing a substantial body of evidence that is pertinent to the reliability, validity, and ability to detect change of this instrument, and provide further evidence to support the usefulness of the Haem-A-QoL in an adult multinational population.