Cost-effectiveness analysis of escalating to natalizumab or switching among immunomodulators in relapsing-remitting multiple sclerosis in Italy

Multiple Sclerosis (MS) is a chronic condition, affecting young adults in the active working phase, with a significant economic and social burden [1, 2]. In most cases, patients with relapsing-remitting MS (RRMS) suffer from episodes of neurological deterioration (relapses), between periods of complete or partial remission. Transition to secondary progressive multiple sclerosis (SPMS), with or without superimposed relapses, can occur after an initial relapsing-remitting (RR) phase, leading to accumulation of irreversible disability [3]. Optimal treatment strategy then aims to minimize the occurrence of relapses and prolong the time to disability progression. As first-line treatment, RRMS patients can receive self-injectable disease-modifying treatments (DMTs), namely interferon beta (IFN) or glatiramer acetate (GA). However, a significant proportion of patients experience disease activity despite first-line DMT treatment [4]. At this stage, neurologists can decide whether switching treatment to another immunomodulator (i.e. from GA to IFN, or vice versa, or from a lower to a higher dose and/or more frequently administered IFN, hereafter called “switching strategy”) or initiating treatment with second-line therapy (i.e. from GA or IFN, to a high-efficacy DMT such as natalizumab, hereafter called “escalation strategy”) [5]. Several studies have shown that escalating to second-line therapy is more effective than switching [6, 7]. Although this evidence supports escalation from a clinical perspective, prescription of second-line treatment could lead to a relevant increase of therapeutic costs, being second-line options more expensive than first-line immunomodulators. Therefore, the economic investment for escalation should be measured against the incremental clinical benefit over switching. This analysis aims to evaluate cost-effectiveness of early escalation to natalizumab vs. switching among immunomodulators, followed by late escalation to natalizumab, in patients affected by RRMS who have failed first-line treatment with either IFNs or GA [6], adopting the Italian Societal perspective.

Table 4 shows the results of the analysis. Total lifetime social costs amounted to €699,676 and €718,604 in the ESC and SWI groups, respectively. In both groups, treatment costs and indirect costs were the most relevant drivers of expenditure, absorbing 85% of costs in the ESC group and 82% in the SWI group. The higher treatment and administration costs, monitoring costs and AE costs in the ESC group were offset by savings for the reduction of relapse burden and the delaying of disability progression, both reducing direct and indirect costs. The economic burden of AEs was negligible. Early adoption of escalation to natalizumab was more effective than switching among immunomodulators (IFNs and GA), leading to an increase of the discounted quality-adjusted survival (+ 1.52 QALYs; 11.19 in the ESC group, vs. 9.67 in the SWI group). The cost-effectiveness analysis showed that ESC dominated SWI in the societal perspective, being associated with lower costs and longer quality-adjusted survival. Early escalation of RRMS patients determined prolonged survival at a lower level of disability.

This result is shown in Fig. 2: after 15 years of observation, about 60% patients in the ESC group and 70% in the SWI group have an EDSS score greater than 5 (absolute reduction: − 10%). The clinical difference between the two groups reaches its maximum at around Year 30. Then the effect of mortality, similar in the two groups, progressively reduces the benefit. Similarly, the effect of relapse occurrence was reduced in the ESC group. The cumulative analysis conducted on relapses showed that, at the end of the observation, patients in the ESC group experienced fewer episodes than patients in the SWI group (8.84 vs. 12.09, respectively). For all tested scenarios, the results of the univariate sensitivity analysis confirmed early escalation to natalizumab being dominant over switching among IFNs and GA (Additional file 1: Table S7). Finally, probabilistic analysis (N = 1000 simulations) showed that ESC was cost-effective vs. SWI in 85.9% of cases, using a willingness to pay threshold of € 50,000 per QALY gained [51, 52]. In 54.4% of cases, the escalation strategy dominated the switching strategy (Fig. 3).

The broad experience achieved with natalizumab in pivotal clinical trials and in observational studies offers obust evidence of its effectiveness in RRMS patients (both naïve or who failed first-line DMT). Recently, the European Medicine Agency (EMA) approved the extension of natalizumab indication in patients with highly active RRMS despite a full and adequate course of treatment with at least one DMT, thus simplifying conditions for escalation [53]. Moreover, the recent review of the benefit-risk profile of natalizumab, conducted by EMA, led to the renewal of the marketing authorization with unlimited validity [54]. The present analysis provides evidence that escalation to natalizumab is a cost-effective option vs. switching among immunomodulators in RRMS patients who failed first-line therapy (IFNs, GA). To our knowledge, this is the first attempt of evaluating economic consequences of early escalation vs. switching in Italy. The clinical data used to conduct the present analysis were mainly derived from an observational analysis conducted in Italy. Each economic assessment is a dynamic process because: i) clinical efficacy data must be confirmed in real practice; ii) certain input data, such as prices of treatments, can change over time; iii) new technologies become available for patients. With this cost-utility analysis we were able to capture two relevant outcomes of the health technology assessment: costs and quality of life. Results of the analysis show that pharmacological expenditure is a relevant cost for healthcare services, but there are other non-negligible costs contributing to healthcare and social economic impact. In fact, although acquisition costs with natalizumab were substantially higher than IFNs and GA in our analysis, early escalation produced savings which entirely offset the investment with natalizumab, thus making escalation a dominant (i.e. less expensive and more effective) alternative vs. switching. Moreover, early escalation led to improved patient quality of life, prolonged survival with lower disability and slightly prolonged overall survival. Finally, these findings were confirmed in all tested scenarios (base-case, one-way sensitivity analysis, probabilistic analysis). In our base case analysis, we decided to adopt a societal perspective, as we aimed to capture the significant economic burden of multiple sclerosis, beyond costs sustained by our healthcare service. Indeed, in their Italian cost-of-illness assessment, Kobelt et al. [2] showed that direct healthcare costs of MS accounted for 28.6% of total per-patient costs, with total direct non-medical costs and productivity loss costs contributing for the remaining 42.3 and 29.1%, respectively.

Despite long-term modelling of costs and clinical consequences can be affected by some methodological limitations (iteration of clinical efficacy results, typically adopted in Markov models, provides a “proxy” of the real evolution of the disease) we believe this analysis is robust enough in terms of clinical efficacy model inputs (most of them were retrieved from published literature) and cost assumptions (retrieved from published literature and consistent with previous Italian publications on RRMS). Likely, the main limitation of this analysis concerns the use of data from a general RRMS population to estimate the natural history of the disease of a highly active RRMS population. Of course, the use of more specific data on highly active RRMS patients would have allowed a more accurate estimate of the clinical outcomes. Unfortunately, to our knowledge, such data are not available. However, we believe this bias disfavoured the option of escalating to natalizumab. Intuitively, the clinical benefits of a therapy reducing disability should be even more evident in a cohort of patients with highly active disease, characterized by faster disability progression. A second limitation of the study was that we did not use EDSS-dependant transition probabilities to model disease progression in the RRMS disease state. We adopted this approach as we did not have adequate sample size in the original study to calculate disability progression rates by EDSS state. Nevertheless, we believe that the benefit of using “real-data” on this specific target population would exceed the drawback of the lack of EDSS-dependant transition probabilities. Moreover, we tested the variability of this parameter in our sensitivity analyses, to get confirmation that the uncertainty associated with this variable would not determine a significant modification of the ICER estimates. Considering all these factors, we acknowledge that the present economic analysis does not represent a conclusive guidance for early treatment of natalizumab, which depends on several clinical factors. The findings of this analysis should be rather considered as a proof of the fact that the current economic evidence supports the escalation of natalizumab in the Italian setting, when this approach is considered appropriate from a clinical perspective. Furthermore, this economic analysis is adopting the Italian societal perspective. Therefore, caution should be taken around the validity of such conclusions in other countries.

The results of this study showed that early escalation to natalizumab in RRMS patients who do not respond adequately to conventional immunomodulators (IFNs, GA) led to both clinical and economic benefits, compared to switching among immunomodulators (IFNs, GA). Overall, the escalation approach: i) improved quality of life, ii) prolonged survival with lower disability, and iii) slightly prolonged overall survival. Although natalizumab acquisition costs were substantially higher than IFNs and GA, early escalation produced savings. Taking together, our findings showed that early escalation to natalizumab was a dominant option for the treatment of RRMS patients who do not respond adequately to conventional immunomodulators (IFNs, GA).