Neuromyelitis optica spectrum disorder (NMOSD) is a chronic and devastating autoimmune disease of the central nervous system characterized by inflammatory lesions mainly of the optic nerve and the spinal cord (1). Despite being considered as a rare disease (2), it is a serious condition, as permanent disability or even death are driven by relapses (3–5). Thus, the main therapeutic goal for patients with NMOSD is to reduce the frequency and severity of these episodes with the aim of avoiding long-term disability accumulation (6, 7).

Historically, NMOSD has been treated with off-label treatments, mainly immunosuppressants (ISTs), oral corticosteroids (OCs) or rituximab. However, in recent years, targeted high-efficacy therapies with a positive safety profile have been approved for its management (8). These new treatments have broadened the landscape of available options, but in turn may have complicated the decision-making process (9, 10). A consequence of this difficulty might be the presence of therapeutic inertia (TI) in decision-making, defined as the lack of treatment initiation or intensification when therapeutic goals are not met (11). In fact, TI has been shown to be present in more than half and up to 96% of treatment decisions in a demyelinating disease with multiple treatment choices such as multiple sclerosis (MS) (12). Several factors have been related to TI prevalence in MS, some influenced by neurologist’s professional experience, such as a lower volume of patients, not being an MS specialist, and fewer years of practice; and others by personality-related traits, namely an aversion to ambiguity and lower tolerance to uncertainty (13). However, what might be influencing TI is the complexity of decisions in MS due to the increasing availability of numerous different treatments and the limited training physicians receive in terms of decision-making processes and risk management in complex environments with multiple options and uncertainty (10). Furthermore, complicated treatment monitoring and fear of side effects have been related to TI as well (10, 14).

In a varying context in NMOSD aquaporin-4 positive (AQP4+) with a broadened therapeutic landscape, the lack of head-to-head trials to facilitate decisions, and no clear management guidelines, understanding the decision-making process and its contributors may be important to provide information on how these decisions are made and reach the best patient care approach. Thus, the aim of the present study was to assess the prevalence of TI and associated factors in the clinical decisions of neurologists caring for patients with NMOSD in Spain.

Understanding the current treatment decision-making process in NMOSD is of significant relevance, as there are few consensus or management guidelines and new therapies have been approved for this condition in the last few years (6, 15, 29, 30). Due to the severity of relapses in this condition, which result in disability accumulation and can even lead to death (83% of patients with partial or no recovery from relapses and mortality rates vary from 3.3–25%) (3–5, 31, 32), we classified TI as not pursuing high-efficacy treatments when therapeutic goals are not met. We found that TI is a common phenomenon in the management of NMOSD patients, observed in more than one-third of participating neurologists. Factors associated with TI were working in a low complexity hospital and giving higher importance to patient tolerability/safety when choosing a treatment.

The management of patients with NMOSD has become more challenging with the approval of the first new therapies. Despite these advances in the last few years, there is little information on treatment decisions including new agents. In this study, we specifically assessed TI in NMOSD with unbranded treatments mimicking the available ones. Min et al. performed a clinical record review in six countries worldwide and found that relapses in NMOSD do not always lead to immediate initiation of maintenance therapy or a switch from off-label OC/ISTs to higher efficacy monoclonal antibodies (33). For newly diagnosed patients, they found that 47% did not receive maintenance treatment within 2 months of diagnosis for reasons such as disease stability, patient refusal and cost/access restrictions, being the severity of relapse a factor related to immediate treatment initiation (33). They also found that 32% of neurologists strongly agreed on initiating OC/IST as first-line therapy in patients with moderate-to-severe symptoms, and 38% agreed if symptoms were mild. When exploring treatment escalation, lack of efficacy, relapse severity and insufficient recovery from relapse drove 54% of all switches (33). However, almost half of the patients most recently seen by the participating neurologists had a relapse and did not change their treatment, mainly for reasons such as relapse mildness, patient stability, or receiving good treatment already (33). Furthermore, a proportion of 42.5% of patients were switched between different OC/ISTs when they had a relapse instead of escalating to a high-efficacy monoclonal antibody, even though half of those relapses were moderate-to-severe in nature (33). This would imply a similar-to-higher percentage of TI than in our study, although new agents were not available in all countries at the time of survey completion. Similar to our results, Thon et al. found that 40% of neuroimmunologists (n = 10/25) would switch none or up to 25% of their patients to one of the novel NMOSD treatments following a relapse (34). The authors hypothesized that the reasons behind this decision could be due to insurance and cost-related barriers (34). However, these factors did not apply in our study as treatments were unbranded and only efficacy/safety data was provided to respondents.

Another survey about treatment choices performed in Korea before new treatments’ approval found that ISTs were prescribed by 85% of the 27 participating neurologists as first-line therapies, and 70% of them would switch to rituximab as second-line therapy if there was a relapse (35). Participants treating a higher number of NMOSD patients annually were more likely to prescribe rituximab as second-line therapy (35). Similarly, a higher volume of patients managed by the responding neurologist has been associated with lower TI in other studies in MS, as well as more years of experience or being an MS specialist (12, 13). However, in our study, the only professional characteristic found as a predictor of TI was working in a low complexity hospital. Additionally, although we found a higher prevalence of TI among neurologists with a lower degree of empathy, and those who gave higher importance to treatment ease of use and their own experience rather than research or scientific evidence, the only predictor of TI in the multivariate analysis in this matter was giving higher importance to patient tolerability/safety when choosing a treatment. We hypothesized that all these factors could be strategies adopted by the respondent in an attempt to reduce the ambiguity and risk-aversion that has been related with TI in previous studies (13).

Uncertainty due to the limited experience with new agents and their differences, the rarity of the disease, the lack of clinical management guidelines including the increasing therapeutic options, together with long experience with off-label treatments, might be affecting the decision-making process (36), giving importance to relapse severity or lesion location as a driver of switches (33, 37). However, there is no clear consensus on how relapse severity is defined in clinical practice and current mildness does not necessary reflect the same in subsequent relapses, the prevention of all relapses being the therapeutic goal regardless of their severity (6, 7, 38, 39). In fact, patients describe disease stability as the absence of any relapse, since they focus on the impact relapses have on daily life and wellbeing (40, 41). Thus, the relapsing nature of the disease together with the importance of a patient-centered approach with shared decision-making should lead to a shift in the treatment paradigm to early use of high-efficacy therapies when treating this condition, as it is evolving in the MS field (42). This is a matter of importance in NMOSD due to the severity of relapses, poor recovery, and associated disability accumulation (5, 22, 39, 43). New therapeutic landscapes might facilitate reaching this ambition, as recent studies have shown an increase in the use of high-efficacy monoclonal antibodies with better outcomes at the expense of IST/OCs (7, 22, 44). However, as shown in this study, there is still a need to improve therapeutic decisions in order to reduce the prevalence of TI and its magnitude, as new highly effective therapies may be an advantage, but the choice between multiple options might lead to suboptimal decisions. For this reason, it could be helpful to implement continuous updated medical education and training with innovative therapeutic interventions that facilitate the decision-making process, with the aim to achieve better patient outcomes (45–47). As NMOSD is a rare disease, patients also could benefit from being treated at referral centers, where continuous training and a greater number of patients assisted is usually more common, thus providing the center with higher expertise and specialization.

Several limitations should be mentioned. First, the cross-sectional design did not allow us to assess changes or causal relationships in neurologists’ TI over time, as the study consisted of a single online survey. Furthermore, we acknowledge it would had been better to perform a prospective or two stage study with training implementation and subsequent measurement of results. Second, there could be potential selection bias in including people with a greater interest in collaborating or who are actively involved with the society. Third, case scenarios did not include all the details (e.g., MRI lesions) or all patient situations, such as adolescents or patients with concurrent autoimmune conditions, but they do contain the minimal information in order to make proper decisions and not develop cognitive fatigue, which could affect the results. Fourth, the study was only performed in Spain, and generalizability of results might not apply to other countries. However, country-specific factors that could affect decisions such as price or insurance cover were removed from the case scenarios. Fifth, we only explored TI in NMOSD AQP4+, without including seronegative patients, as new treatments are approved only in seropositive population. Future studies should be performed to fill these gaps.

In conclusion, TI was detected in 4 out of 10 participating neurologists when initiating or switching treatments in NMOSD AQP4+, a severe condition with accumulating disability led by potentially devastating relapses. This study uncovers the need to challenge the therapeutic status quo in NMOSD, pursuing high-efficacy treatments from the start, and developing specific intervention strategies to ensure optimal therapeutic decisions and patient care. Future longitudinal studies should investigate this matter and strategies to reduce TI in NMOSD, including patient perspectives.

Qualified researchers may request access to individual level data through the clinical study data request platform (https://vivli.org/). Further details on Roche’s criteria for eligible studies are available here (https://vivli.org/members/ourmembers/). For further details on Roche’s Global Policy on the Sharing of Clinical Information and how to request access to related clinical study documents, see here https://www.roche.com/innovation/process/clinical-trials/data-sharing.

The studies involving humans were approved by the research ethics board of Hospital Universitario Clínico San Carlos, Madrid, Spain. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

ÁC-C: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. RG-B: Conceptualization, Writing – original draft, Writing – review & editing. AO: Data curation, Writing – review & editing. MD: Data curation, Writing – review & editing. SB: Data curation, Writing – review & editing. MA-V: Data curation, Writing – review & editing. MS: Data curation, Writing – review & editing. PR: Conceptualization, Writing – review & editing. PL-L: Conceptualization, Writing – review & editing. JM: Conceptualization, Writing – original draft, Writing – review & editing. NTL: Conceptualization, Data curation, Writing – original draft, Writing – review & editing.