Switching from current factor VIII (FVIII) to longer acting FVIII concentrates – what is the real potential benefit?

As Charles Dickens wrote ‘it was the best of times, it was the worst of times’. Dickens was of course writing about the French Revolution; a time of tremendous change and upheaval with much uncertainty and little awareness of what the future would be like.

Today the haemophilia community similarly finds itself in a time of change and uncertainty. Factor concentrates have existed for over 50 years. Despite advances, particularly in the area of viral safety, all factor VIII (FVIII) concentrates until now, have had similar pharmacokinetics. However, now things are changing with the introduction of extended half-life (EHL) concentrates. How will we use these products, what will be the implications of different ways of using them and will the results be better or worse is the subject that Gringeri et al. 1 are ultimately trying to address in their modelling study published in this issue of the journal.

Gringeri and colleagues did a simulation exercise of 1000 patients ≥12 years of age (median age and weight: 25.4 years and 72.4 kg) and compared theoretical factor levels patients would have if they received every other day (EOD) prophylaxis using rAHF-PFM (Recombinant anti-hemophilic factor- protein free manufacture) (subsequently referred to as rFVIII) to what they would have if they were on four different prophylaxis regimens (different doses and frequencies) using rFVIIIFc, the first EHL FVIII to be licensed. In doing so they made certain assumptions regarding the pharmacokinetics of these concentrates and patient adherence. Given their assumptions they showed that a regimen of 30 IU kg⁻¹ of rFVIII given EOD would result in patients spending the least amount of time with FVIII:C levels <3% (3 IU dL⁻¹) in comparison to different regimens using rFVIIIFc: 30 IU kg⁻¹ every third day (ETD); 50 IU kg⁻¹ every 4 and 5 days and 65 IU kg⁻¹ every 7 days and similarly the 30 IU kg⁻¹ EOD rFVIII regimen would result in more time spent weekly with FVIII:C levels >10% (98.9 h) than the every 4, 5 or 7 day rFVIIIFc 50 or 65 IU kg⁻¹ regimens. However, it should be pointed out that the regimen that led to the maximum time with FVIII:C >10% was the rFVIIIFc 30 IU kg⁻¹ ETD. It also needs to be pointed out that the difference in the median time that patients would spend with FVIII:C levels <3% between the 30 IU kg⁻¹ rFVIII EOD regimen and the 30 IU kg⁻¹ rFVIIIFc ETD regimen was 0 vs. 0.05 h, that is 3 min!

Gringeri and colleagues pointed out the potential benefits as well as the potential risks associated with the use of rFVIIIFc in comparison to rFVIII. They showed that with the 30 IU kg⁻¹ ETD rFVIIIFc regimen that many patients were able to go ≥150 h per week with FVIII:C >10%. This was something that did not happen with the EOD rFVIII regimen. The corollary is that many other patients on this rFVIIIFc regimen experienced much less time with FVIII:C >10% in comparison to what they would if they had received rFVIII EOD. And overall fewer patients (41.1%) on the 30 IU kg⁻¹ ETD rFVIIIFc regimen maintained at all times a FVIII:C level >3% in comparison to the 30 IU kg⁻¹ EOD rFVIII prophylaxis regimen.

The findings of Gringeri and colleagues are thought provoking but not really unexpected. EHL FVIII concentrates unlike EHL FIX concentrates have half-lives that are only modestly prolonged (on average 1.5 fold). As such it is not a surprise that giving EHL FVIII concentrates every 4, 5 or 7 days will not result in superior factor levels than what we can obtain with current rFVIII. The reality is that with a 1.5 fold half-life extension the expectation would be that EHL FVIII given ETD would result in comparable amounts of time that patients would spend with FVIII:C <3 and >10% as they currently do with EOD rFVIII which is essentially what Gringeri and colleagues showed.

There are a number of issues pertaining to this study that merit discussion.

Assumptions in the model

When conducting a modelling study, the assumptions used in the model are critical. Change the assumptions and usually the results of the model change. Gringeri et al. assumed that patients would not miss any doses and that the mean incremental recovery (IR) and mean residence time (MRT) of rFVIII would be 2.3 and 16.1 h respectively while the corresponding figures for rFVIIIFc would be 1.9 and 27 h, values based on two reported studies 2, 3. The assumptions regarding MRTs are in keeping with the findings of Mahlangu et al. who undertook sequential pharmacokinetic studies on 28 patients, first with rFVIII and then with rFVIIIFc. In contrast the assumptions regarding IR can be challenged. Powell reported IR on 16 patients who received both rFVIII and rFVIIIFc (in both cases either 25 or 65 U kg⁻¹) 4. IR with rFVIII was 2.56 and 2.0 with 25 and 65 U kg⁻¹ doses and 2.44 and 1.8 for similar doses of rFVIIIFc. Similarly, Mahlangu et al. 5 determined an IR of 2.4 for rFVIII and 2.2 for rFVIIIFc in their study. Collectively these studies would suggest an IR of about 2.3–2.4 for rFVIII and about 2.1–2.2 for rFVIIIFc and not 1.9 as used by Gringeri et al. in their model. The results of the model might have been somewhat different if the model had used an IR of 2.1–2.2 for rFVIIIFc. The assumption of perfect adherence is discussed later.

Modelling factor levels vs. reality of bleeding

While on prophylaxis FVIII:C levels are an important, but not the only relevant variable, that contributes to bleeding. Gringeri and colleagues inferred from their study that given the longer amounts of time that patients would spend with FVIII:C levels <3% with every 4th, 5th and 7th day dosing that these regimens should be used with caution – implying that they may place patients at higher risk of bleeding. Although intuitively this makes sense it has been demonstrated in many clinical trials of both conventional FVIII and FIX as well as EHL FVIII concentrates that unexpectedly despite factor levels that would suggest that patients would not do well that many patients still do quite well on these less frequent infusion regimens. For example, in the study of Valentino et al. 6 a prophylaxis regimen of 100 IU kg⁻¹ of rFIX resulted in a mean annualized bleeding rate of only 4.6 while in the study of Mahlangu et al. 5 the mean annualized bleeding rate in patients placed on a regimen of 65 IU kg⁻¹ once per week of rFVIIIFc was only three. The modelling study by Gringeri et al. would have predicted much worse outcomes with these prophylaxis regimens.

Doing better than a regimen would predict is likely attributed to patients adjusting their behaviour to the prophylaxis regimen that they are on. They can do this in a number of ways: selecting the best days to administer factor; infusing in the mornings and being more adherent. Given that patients may be more/less active on certain days they can adjust their activities or their prophylactic regimen such that on the 1–3 days prior to their next infusion knowing that their factor levels are likely to be quite low they may be much less active thereby mitigating the risk inherent in having factor levels below a certain arbitrary FVIII:C threshold.

The time at which an infusion is done represents not only when peaks occur but also when troughs occur. Given that humans in general are sleeping for about one-third of their lives (8 h daily) it means that the impact of low FVIII:C levels are likely to be much less in the hours when patients are asleep in contrast to when they are awake and active. Consequently it is possible that when factor is given in the mornings that regimens associated with more time spent with factor levels <3% may actually be more or just as protective as regimens associated with less time spent with FVIII:C levels <3% but where trough levels occur at times when patients are active. Finally a missed prophylactic dose negates the theoretical benefits of higher factor levels associated with a regimen. Although not proven, experience suggests that regimens that involve more frequent infusions are likely in the ‘real’ world to lead to more missed doses.

In the ‘real’ world patients may not always infuse on the best days of the week, they may not necessarily infuse in the mornings and they may not be perfectly adherent – they seldom are. All of these behaviours may lead to many patients doing well (those who adjust their lifestyle to their prophylaxis regimen, who infuse in the mornings and who are very adherent) despite their predicted FVIII:C levels associated with their regimen being ‘inferior’ while other patients may not do as well as their regimen would predict – on the basis of time below and above certain FVIII:C levels.

Fewer infusions with same trough levels

Gringeri et al. give little emphasis to the fact that one argument for using EHL FVIII is not that they will further reduce bleeding rates which for many patients already with current factor concentrates is ≤1 annual bleed but simply that with EHL FVIII patients may be able to achieve the same bleeding rates with less infusions. A reduction from EOD (182 annual infusions) to ETD (121 annual infusions) infusions represents a reduction of 61 infusions annually – a considerable number.

Issue of adherence

Adherence is crucial to prophylaxis regimens. In a number of studies, investigators showed that reduced adherence contributed to a substantial proportion of bleeds in patients on EOD prophylaxis with rFVIII (the same rFVIII used in the study by Gringeri et al.) 7. So although the reader might agree with Gringeri et al. that 1 year of EOD prophylaxis with rFVIII might achieve slightly overall better factor levels than ETD with rFVIIIFc this would only occur if patients were completely adherent to both regimens – an unlikely occurrence. The question of whether adherence in the real world (as opposed to clinical trials) with EHL FVIII with more time between infusions will be better than with rFVIII remains to be evaluated.

Conclusion

The haemophilia community needs to learn how to use these newer EHL FVIII concentrates. We learn from papers like that of Gringeri et al. but such simulations will not replace direct observation gathered from ‘real world’ use. As haemophilia patients start using these EHL FVIII concentrates they and their physicians will begin to learn how to best use these products and tailor them to specific patients. Some patients might do well with significantly less frequent infusions (e.g. once per week) being able to align their activities to their prophylaxis regimens. Others might need ETD infusions while others might consider staying on EOD infusions to achieve much higher trough levels then they currently do with rFVIII in order to perhaps participate in activities that until now they have avoided. For now the haemophilia community has to move with appropriate caution recognizing that a 1.5 fold extension of FVIII half-life will not have a huge impact on prophylaxis regimens. Years from now when the haemophilia community looks back on the time before the introduction of these EHL FVIII concentrates they will not necessarily see it as the worst of times and after their introduction it will not necessarily be the best of times. However, these EHL FVIII products do appear to represent an incremental improvement.

Disclosures

Manuel Carcao has received Honoraria for speaking or participating in scientific advisory boards or symposia from Baxter, Bayer, Biogen, Biotest, CSL Behring, Novo Nordisk, Octapharma and Pfizer. He has also received research support from Baxter, Bayer, Biogen, Grifols, Novo Nordisk and Pfizer.