1 Recommendations

The marketing authorisation for rozanolixizumab is for an add-on to standard treatment for AChR or MuSK antibody-positive gMG. In its submission, the company positioned rozanolixizumab for a narrower population – people with refractory AChR or MuSK antibody-positive gMG – based on the following criteria:

• the disease is classified as Myasthenia Gravis Foundation of America (MGFA) class II to IVa

• the disease is uncontrolled after 2 or more previous therapies, excluding anticholinesterase inhibitors

• an additional therapy, such as IVIg or PLEX, is being administered or considered.

The final scope issued by NICE listed the following comparators:

• efgartigimod (subject to NICE evaluation)

• zilucoplan (subject to NICE evaluation)

• ravulizumab (now terminated)

• standard of care without rozanolixizumab (including immunosuppressive therapies [including rituximab] with or without IVIg or PLEX).

MycarinG was a phase 3, randomised, multicentre, double-blind, placebo-controlled trial. It recruited adults with gMG with positive serology for anti-AChR or anti-MuSK antibodies, with an MGFA class of II to IVa, an MG-ADL score of 3 or more and a Quantitative Myasthenia Gravis (QMG) score of 11 or more, and who were being considered for additional treatment (such as IVIg or PLEX). Of the 200 people included in the 3 trial arms, 66 were randomised to the licensed weight-based dose of around 7 mg/kg and 67 to the placebo arm (the 67 people who were randomised to the higher unlicensed dose of around 10 mg/kg are not relevant to this evaluation). People in both arms also continued to have standard treatment with corticosteroids and immunosuppressants. MycarinG consisted of a 6-week treatment phase, followed by an 8-week observation period in which people did not have rozanolixizumab. In the treatment phase, people had either 6 once-weekly subcutaneous infusions of either rozanolixizumab or placebo as an add-on to standard care, which comprised 1 treatment cycle. The post-hoc refractory subgroup of relevance to this evaluation was defined as uncontrolled disease despite standard of care treatment, specifically 2 or more previous gMG therapies, excluding acetylcholinesterase inhibitors. The primary outcome was change in MG-ADL score (a higher MG-ADL score shows more severe symptoms) from baseline to day 43. From baseline to day 43 rozanolixizumab was associated with a statistically significant greater reduction in MG-ADL score than placebo in the overall trial population (−3.370 compared with −0.784, least squares mean difference −2.586 [95% confidence interval −4.091 to −1.249; p<0.001]). MycarinG also reported the number of people who had an MG-ADL response, defined as a 2-point or more improvement in MG-ADL score, as a secondary outcome. At day 43 and among the whole-trial population, more people who had rozanolixizumab had an MG-ADL response than those who had placebo (68.2% compared with 28.4% [odds ratio 5.77; 95% confidence interval 2.10 to 14.88; p<0.001]), and this was statistically significant. The treatment effect of rozanolixizumab on these outcomes is largely in the same direction for the post-hoc refractory subgroup, but this is considered confidential by the company so cannot be reported here. The anti-AChR and anti-MuSK antibody-positive subgroups were prespecified. The EAG noted that the results for change in MG-ADL in these 2 prespecified subgroups were also generally consistent with those of the overall study population. The committee noted that evidence showed that rozanolixizumab was associated with a reduction in mean MG-ADL scores in people with gMG from baseline to day 43 within 1 treatment cycle. But the treatment effect may be overestimated because the chosen timing of outcomes assessment in MycarinG may have been the optimal time to assess the best response. Also, people in MycarinG only had 6 weeks' treatment, so the treatment effect of rozanolixizumab in the longer term is uncertain. The committee concluded that rozanolixizumab as an add-on to standard treatment is more effective at improving MG-ADL score than standard treatment alone, but the treatment effect may be overestimated and is uncertain in the longer term.

MG0007 was an open-label extension trial that included people who had participated in other rozanolixizumab trials. People could enter MG0007 if they had:

• entered or completed the observation period of MycarinG

• required rescue therapy (limited to IVIg or PLEX) during the observation period of MycarinG, or

• completed at least 6 visits in MG0004 (a discontinued extension study of MycarinG).

In its submission, the company positioned rozanolixizumab for people with refractory gMG who tested positive for either anti-AChR or anti-MuSK antibodies. The EAG noted that people with refractory gMG were only a subgroup of the MycarinG population. It was concerned that the outcomes observed in the whole MycarinG population may not be generalisable to the refractory population that would have rozanolixizumab in the NHS. But clinical advice to the EAG suggested that the baseline characteristics of the whole MycarinG population approximated the baseline characteristics of the refractory population in the NHS that would be considered for IVIg or PLEX. This is because the MycarinG eligibility criteria included having or being considered for IVIg or PLEX, meaning that the overall trial population is likely to reflect a refractory population. The clinical experts also considered that refractory gMG is expected to respond as well as non-refractory gMG. This is because treatments like rozanolixizumab have a novel mechanism of action, which people with refractory gMG will not have previously tried and their gMG may respond to. The EAG explained that MycarinG primarily included people who tested positive for anti-AChR antibodies, with only a minority who tested positive for anti-MuSK antibodies (placebo, n=8, 11.9%; rozanolixizumab around 7 mg/kg, n=5, 7.6%). The EAG also explained that while the overall trial population approximates the relative proportions of people who test positive for anti-AChR antibodies or anti-MuSK antibodies in the NHS, there is uncertainty about the efficacy outcomes for people who test positive for anti-MuSK antibodies because of the very small number of people in this subgroup. The clinical experts explained that there are some differences in the way that AChR antibody-positive and MuSK antibody-positive gMG respond to treatments. In particular, MuSK antibody-positive gMG responds better to treatment with PLEX. The clinical experts also explained that it is reasonable to assume that people with MuSK antibody-positive gMG might respond slightly better to rozanolixizumab, but this was uncertain because of the very small number of people who test positive for anti-MuSK antibodies. The committee concluded that the outcomes of the whole-trial population in MycarinG could be representative of the refractory gMG population in the NHS. It also concluded that the outcomes of the whole-trial populations in MycarinG could be generalised to those with MuSK antibody-positive gMG, because it is reasonable to assume that rozanolixizumab is at least as effective as in people with AChR antibody-positive gMG.

The company did network meta-analyses (NMAs) and matching-adjusted indirect comparisons (MAICs) to estimate the comparative effectiveness of rozanolixizumab with the comparators. NMAs were done for the outcomes of change from baseline to day 43 in MG-ADL score (defined as 2 or more in MG‑ADL score), and MG-ADL response, but only for the comparisons with zilucoplan and efgartigimod. For the comparison with efgartigimod, the company also did an anchored MAIC on the outcomes of MG-ADL score and MG-ADL response. Neither zilucoplan nor efgartigimod was a relevant comparator at the time of evaluation (see section 3.6). The company also did an unanchored MAIC for the comparison between rozanolixizumab and IVIg. But this was limited because it did not analyse the outcomes of change from baseline to day 43 in MG-ADL score or MG-ADL response because data were not available. The results of the indirect treatment comparisons are considered confidential by the company so cannot be reported here. The EAG explained that for this unanchored MAIC the company matched the rozanolixizumab arm of the MycarinG trial to the IVIg arm of the trial reported by Barth et al. (2011). But this study did not report on MG-ADL outcomes. The company instead reported QMG response and QMG change from baseline for this unanchored MAIC analysis. The EAG also explained that QMG outcomes do not inform the economic model, and the company had not reported QMG outcomes for any other NMA or MAIC analyses. So, this unanchored MAIC was not informative. No indirect comparisons were provided for PLEX. The EAG explained that it had requested that the company provide a systematic literature review to investigate whether any single-arm studies or phase 2 trials on IVIg or PLEX could be included in the unanchored MAIC to enable the comparisons between rozanolixizumab and IVIg or PLEX for the MG-ADL or any other relevant outcomes. But the company did not provide this. The clinical experts noted that there may be very few randomised controlled trials on IVIg or PLEX that could be used to inform the networks for indirect treatment comparisons, but agreed with the EAG that it might be useful to include phase 2 trials or single-arm studies. The EAG suggested that it may also be possible to explore real-world evidence sources and cohort studies reporting outcomes for IVlg or PLEX. The EAG also noted that imputation techniques could be used to map the results from different reported outcomes to the MG-ADL outcomes relevant to this evaluation. The committee concluded that the company's indirect treatment comparisons did not adjust for heterogeneity or baseline risk of populations across studies, so were not appropriate for decision making. The committee requested that the company:

• do a systematic literature review(s) that includes single-arm trials, phase 2 studies and real-world evidence, including observational studies and registries that assess the treatment effect of IVIg or PLEX on MG-ADL or other relevant outcomes

• explore the feasibility of other indirect treatment comparison methods, such as multivariate NMAs to borrow strength across different outcomes reported in the studies or NMAs adjusted for baseline risks with an informative prior on the between-study variability. This approach may provide more information on the relative treatment effects between relevant comparators and may adjust for potential placebo effects observed in some studies.

The company used a cohort state transition model to estimate the cost effectiveness of rozanolixizumab against the comparators. The model included 7 health states. People start in the 'uncontrolled' health state and transition to the 'response' health state if they meet the treatment response criteria (a decrease of 2 or more in MG‑ADL score) at the response assessment timepoint. Responders are further divided into 3 health substates:

• 'stable response' (MGADL score remains stable after time of response assessment)

• 'loss of response'

• 'continued response' (MG‑ADL score continues to improve after time of response assessment).

Over time, people in the model return to the 'uncontrolled' health state and have only corticosteroids and immunosuppressants. The company's model does not account for any future use of IVIg or PLEX for people who stop either rozanolixizumab or the comparator treatments. The EAG considered it likely that if gMG does not respond or loses response to a targeted treatment, people with the condition would change to an alternative treatment. The committee noted statements from the patient experts and clinical experts that gMG requires lifelong management. It also recalled the discussion about whether rituximab might be used as a subsequent treatment for refractory gMG or only for people with refractory MuSK antibody-positive gMG (see section 3.4). The committee agreed that it was implausible that someone with refractory gMG would stop rozanolixizumab and never have another treatment except corticosteroids and immunosuppressants. The clinical experts noted that they would consider IVIg or PLEX for people who stop rozanolixizumab. They explained that if someone's refractory gMG did not previously respond to a particular treatment they would not use it again. So, there may be differences in the choice and proportion of subsequent treatments in the rozanolixizumab and comparator arms. The committee concluded that it would like to see the company include subsequent treatments in the economic model.

The company used the NMA results to estimate the MG-ADL response rates for rozanolixizumab, zilucoplan and efgartigimod. There were differences in placebo responses across the trials included in the company's NMAs. To adjust for differences in placebo response, the company converted the odds ratios of each of these treatments from the NMAs compared with placebo into relative risks. Then, the relative risks were applied to a referent response rate. The referent response rate was calculated by running a baseline random effects model using all the placebo response rates from studies identified in the NMA. The company considers the response rates for rozanolixizumab, zilucoplan and efgartigimod, and the referent response rate, confidential so they cannot be reported here. The EAG noted the uncertainties with the company's NMAs. It also considered the company's referent response rate implausible because the placebo response rates in MycarinG, RAISE and ADAPT were much higher or lower. IVIg or PLEX response rates in the company model were based on data from Barth et al. (2011), a Canadian randomised controlled trial of 84 people with gMG who had either IVIg or PLEX. The calculated response rates were 51% (IVIg) and 57% (PLEX). The EAG noted the following limitations with using data from Barth et al.:

• the population was not explicitly defined as refractory

• MGADL data was not available, so the response was defined as a 3-point or more improvement in QMG

• no confidence intervals or standard errors were provided with the response rates.

The company selected the response assessment timepoints from the zilucoplan, efgartigimod and rozanolixizumab trials (12, 10 and 6 weeks, respectively) and used an assumption for IVIg and PLEX (6 weeks). The EAG noted that clinical advice suggested it would be reasonable to assess all interventions at 6 weeks, so it chose to use the response assessment timepoint of 6 weeks for all treatments in its analysis. The clinical experts explained that assessment of treatment response for IVIg and PLEX is typically done much earlier than 6 weeks, usually at 2 or 3 weeks, and that 3 weeks would be more appropriate to assess for a response. The committee noted that there appeared to be inconsistencies in clinical opinion on the most appropriate timepoint for the response assessment of IVIg and PLEX. It concluded that a response assessment timepoint of 3 weeks reflected NHS practice for IVIg and PLEX, but an assessment timepoint of 6 weeks was appropriate for rozanolixizumab.

The company's model applied treatment costs for IVIg every 3 weeks and for PLEX every 4 weeks. The EAG received clinical advice that, in the NHS, IVIg and PLEX are typically given every 4 to 8 weeks, with the interval between treatments sometimes extended to 12 weeks or, rarely, 16 weeks. The clinical experts at the committee meeting noted that treatment intervals of 8 weeks or longer are not common and that 4 weeks is more typical. The committee noted that IVlg and PLEX might be expected to be given every 6 weeks in the NHS and this is included in the EAG's base case. But it concluded that IVIg and PLEX costs should be applied every 4 weeks for consistency with the evaluation of zilucoplan.

The committee noted the high level of uncertainty in the evidence and modelling, specifically that:

• the treatment effect of rozanolixizumab may be overestimated and its treatment effect in the longer term is uncertain (see section 3.7)

• the model does not account for subsequent treatments (see section 3.12)

• the comparative effectiveness of rozanolixizumab against IVIg and PLEX is highly uncertain, but this is not reflected in the model (see sections 3.10 and 3.13)

• there may be uncaptured benefits of rozanolixizumab that the committee would like the company to try to account for (see sections 3.15 and 3.21).

The committee's preferred assumptions included:

• The comparators should be modelled as a 'basket' of standard care, with some people having IVIg, some having PLEX, and some having neither. Everyone should have corticosteroids and immunosuppressants. Zilucoplan and efgartigimod should not be included as comparators (see section 3.6).

• The results of the whole-trial populations of MycarinG can be generalised to those with refractory gMG in the NHS (see sections 3.7 to 3.9).

• The committee would prefer an indirect comparison that incorporates data from all available studies, includes IVIg and PLEX, and adjusts for the placebo response. Also, any uncertainty from indirect comparisons should be incorporated in the model (see sections 3.10 and 3.13).

• The response assessment timepoint should be 3 weeks for IVIg and PLEX, but 6 weeks is more appropriate for rozanolixizumab (see section 3.14).

• The costs of IVIg and PLEX should be applied every 4 weeks, and the NHS reference cost should be used for PLEX administration (see section 3.16).

• There may be uncaptured benefits of rozanolixizumab that may affect the utility of people who have it. The committee would prefer the company to present scenario analyses that incorporate some of these uncaptured benefits in the modelling (see sections 3.15 and 3.22).

NICE's manual on health technology evaluations notes that, above a most plausible ICER of £20,000 per quality-adjusted life year (QALY) gained, judgements about the acceptability of a technology as an effective use of NHS resources will take into account the degree of certainty around the ICER. The committee will be more cautious about recommending a technology if it is less certain about the ICERs presented. But it will also take into account other aspects, including uncaptured health benefits. Because of confidential commercial arrangements for rozanolixizumab and some of the comparators, the exact cost-effectiveness results are confidential and cannot be reported here. Although some of the company's base-case incremental cost-effectiveness ratios (ICERs) were within the range NICE normally considers to be a cost-effective use of NHS resources, they did not include the committee's preferred assumptions. The EAG's base-case ICER was substantially above this range.

The committee noted the variation in access to IVIg and PLEX across the NHS. Some centres may exclusively use IVIg, some may use a mixture of IVIg and PLEX, and some may not have access to either. The committee also considered that gMG may have a different burden on women than men. gMG is more prevalent in women, women are typically younger at disease onset, and women typically have higher mortality. Also, pregnancy may contraindicate some types of treatment. Sex is a protected characteristic under the Equality Act 2010. But, because its recommendation does not restrict access to treatment for some people over others, the committee agreed this was not a potential equality issue.

The committee considered whether rozanolixizumab was innovative. The patient experts clearly noted that treatment with IVIg or PLEX was time-consuming and required regular hospital stays. They thought that rozanolixizumab, as a short-duration subcutaneous infusion, would be more convenient and could improve adherence. The clinical experts noted how resource intensive IVIg and PLEX are to administer. They also explained that people who have rozanolixizumab may be able to reduce their corticosteroid dose. This could lead to fewer corticosteroid-related adverse effects. Both the patient and clinical experts considered rozanolixizumab to have advantages for patients, carers and healthcare professionals. But the committee noted that similar QALYs were generated by each treatment in the model. So, the committee concluded that some benefits of rozanolixizumab may not be captured in the modelling. The committee asked the company to present scenario analyses that account for some of these benefits.

The committee considered that the cost-effectiveness estimates presented by the company and EAG were highly uncertain. Given the uncertainty, it would like to see additional analyses. But the committee considered that, given its preferred assumptions and based on the analysis it had seen, the cost-effectiveness estimates were highly likely to be above the range that NICE considers a cost-effective use of NHS resources. The committee concluded that rozanolixizumab could not be recommended for treating refractory gMG in adults who test positive for anti-AChR or anti-MuSK antibodies.