Although clinical trials have always been the gold standard of proof regarding effectiveness and safety of new drugs, there is nowadays a great interest in real-world research since they represent patients in real-life settings. To our knowledge, FIRE was the largest French real-word study that assessed the effectiveness and safety of ibrutinib, in accordance with the French marketing authorization in 2016, for the treatment of CLL/SLL in patients who received at least one prior line of treatment, or who were previously untreated and had a del17p and/or TP53 mutation unsuitable for chemoimmunotherapy. In this extensive study, set up in 65 centres, 388 CLL/SLL patients (194 retrospective and 194 prospective) were included in the effectiveness population and followed-up for five years.
Our results showed a median PFS of 53.1 months for retrospective patients and of 52.9 months for prospective patients with one-year PFS survival rates of 93.2% and 83.5% and 5-year rates of 45.5% and 45.1% respectively. The median OS was not reached for both groups. The OS rates were 97.9% for retrospective patients and 87.6% for prospective patients at one year, and 64.5% and 63.3%, respectively, at 5 years. Our results are consistent with previous effectiveness findings [16, 17]. In a real-world multicenter retrospective study conducted on 205 CLL patients treated with ibrutinib, the 12-months PFS and OS rates were 86.3% and 88.8% respectively [16]. In another study on long-term efficacy and safety with a median follow-up of 5 years, in which 31 treatment-naïve and 101 R/R patients were included, the median PFS in R/R patients was 51 months with a 5-year PFS rate of 44% [17]. The median OS was not reached and the OS rate at 5 years was 60%. In a UK/Ireland-based study, the one-year OS was 83.8% [18]. In the clinical trial RESONATE, only R/R CLL patients were included. When comparing the results at similar timepoints between RESONATE and FIRE, the one-year PFS and OS rates in RESONATE (84% and 90% respectively) as well as the 5-year PFS (40.0%) were similar to those of FIRE [7, 8, 19]. The ORR was also similar: 91% in RESONATE vs. 96.8% and 96.6% for retrospective and prospective patients, respectively, in FIRE [8] (Online Resource 6). However when comparing median PFS and OS, those in RESONATE were lower: 44.1 (95% CI: 38.5–56.2) months for the median PFS and 67.7 (95% CI: 61.0 – not reached) months for the median OS [8] (Online Resource 6). One explanation could be the longer follow-up period in RESONATE (6 years vs. 5 years). However, taking the fact that our results are included in the confidence intervals of the PFS and OS of RESONATE, our finding are consistent. Therefore, although the FIRE population is slightly different than the population in RESONATE (e.g. age, ECOG PS, mutations status, number of prior therapies), it is reassuring to see that our effectiveness results are similar to the results of clinical trials.
The safety profile of ibrutinib and dose reductions were also assessed. Our results showed that among patients who had a dose modification, 80.5% (70/87) of the retrospective patients and 88.9% (72/81) of the prospective patients had at least one dose reduction, mainly due to toxicity. Efficacy results showed that patients with at least one dose reduction had a similar PFS than patients with no dose reduction, supporting the fact that CLL patients in France are well managed, follow-up and treated. Our results not only confirm those of previous real-world studies [18, 20, 21] but also encourage the idea that ibrutinib can still be administrated to patients presenting AEs. Therefore, if physicians need to modify the dose because of an AE, dose reduction may be the best option. Suggesting dose reduction to patients in need of dose modification will thus reduce treatment discontinuation, increase patient adherence, improve patients outcome and on a long-term strategy decrease financial and economic burden. However, to obtain the best benefit from ibrutinib, it is important to promptly identify and manage AEs, and understand specific AEs that can lead to dose reductions. Identifying specific AEs and other factors that could potentially be associated with the need for dose reduction could be the next step in order to well maintain ibrutinib treatment for patients with AEs at the correct reduced dose.
The median time to dose reduction as the first dose modification was 7.4 (range: 0.4–60.9) months for retrospective patients and 9.3 (range: 0.4–57.4) months for prospective patients. One retrospective chart review on first line and R/R CLL patients treated with ibrutinib either in academic practice or community network that assessed the median time to first dose reduction found an overall median time of 3.6 months [22]. However, when the overall result was stratified by academic/community settings and by first line and R/R patients, the median time to first dose reduction for first line patients treated in academic settings increased to 16.6 months versus 3.6 months for R/R academic-treated patients. Furthermore, a review on ibrutinib dose modifications in the management of CLL mentioned that in real-world settings, dose reductions over the first year was often noticed [23]. However, addressing the question of time in dose reduction still remains rare and unclear. Therefore, further research on this topic is necessary in order to better understand the role of the time in dose reductions and ibrutinib outcome.
Furthermore, the results showed that among patients who discontinued ibrutinib, toxicity was the main reason for 43.5% of the retrospective patients and 42.0% of the prospective patients. These results were similar to the one found in a Swedish retrospective study: 40.4% (19/47) [20]. However, a Danish retrospective study showed a higher rate: 54.7% (47/86) [16]. In RESONATE, the discontinuation rate due to toxicity was much lower 21.1% (32/152) (Online Resource 6). One explanation to this lower discontinuation rate due to AEs compared to the FIRE study could be that RESONATE is a clinical trial with eligibility criteria which promote inclusion of selected patient. Although our results on discontinuation rates due to AEs differed from the one found in RESONATE, they illustrate the need of real-world research on long-term safety on heterogeneous population.
Among adverse events noticed in our study, patients reported low rate of major bleeding events (2.0% retrospective and 8.2% prospective). This rate was five times less for retrospective patients but similar for prospective patients than the rate reported in RESONATE (10.0%) [8]. Of note, in FIRE, more patients had a bleeding / major bleeding event when they were under anti-thrombotic treatment. Explanations could be that bleeding events are side effects of such treatments, and in RESONATE, patients under anticoagulation containing warfarin were excluded. In addition, the rate of atrial fibrillation was similar for patients in the two studies (FIRE: 8.1% for retrospective and 11.2% for prospective patients; RESONATE: 12.0%) but the rate of hypertension was lower in FIRE than in RESONATE (FIRE: 14.6% for retrospective and 14.8% for prospective; RESONATE: 21.0%) [8]. Nevertheless, it is reassuring to see that there was no new AE observed and that the safety profile of ibrutinib in our study seems to be consistent with previous studies.
Our study has several strengths and limitations. First, FIRE was an extensive research and the largest French real-word study on the effectiveness and safety of ibrutinib. Second, because of its real-world design, effectiveness and safety parameters were presented through descriptive data in a “real-life condition”, and therefore, our results complement those of clinical trials. Moreover, all consecutive patients who met the eligibility criteria and who had therapy-demanding disease were considered for inclusion in order to reduce selection bias. However, there might have been a bias in effectiveness results between retrospective and prospective patients since retrospective patients who died before enrolment were not included. Therefore, retrospective patients who were included in the study should be considered in “better” health than prospective patients. However, it is reassuring to see that the results between the two groups are quite similar. In addition, because of the exclusion of retrospective patients who died before enrolment, it was difficult to pull data of retrospective patients together with the data of prospective patients. Moreover, the number of adverse events for retrospective patients have been underestimated since TEAEs that occurred before inclusion and that were not related to ibrutinib were not collected for these patients. Finally, our focus was on the effectiveness and safety profile of ibrutinib. Therefore, other aspects such as the quality of life of patients under ibrutinib were not considered in this article. Although such data might have been informative and complement the findings of this article, this whole topic could be discussed in a further paper.