Inebilizumab for neuromyelitis optica spectrum disorders in Italy: a budget impact model

ABSTRACT

Background

The Italian National Health Service (INHS) has recently reimbursed the monoclonal antibody inebilizumab as a second line monotherapy after rituximab (RTX) use for neuromyelitis optica spectrum disorders (NMOSD) patients ≥ 18 years anti-aquaporin 4 antibody-immunoglobulin G positive, who experienced a relapse in the last year or cannot receive RTX, if incident patients. Other INHS-reimbursed drugs for NMOSD treatment are satralizumab, eculizumab and, off-label, besides RTX, ocrelizumab, tocilizumab, and immunosuppressants.

Research design and methods

A 3-year (2023–2025) prevalence-based budget impact model following the INHS viewpoint compared the costs and the NMOSD attacks without (1st scenario) and with inebilizumab (2nd scenario). The epidemiology of NMOSD, and the INHS-funded healthcare resources (drugs and their administration; specialist visits; hospitalizations due to drug-related adverse events and NMOSD attacks) were obtained from the literature. One-way, threshold value and scenario sensitivity analyses investigated the robustness of the baseline findings.

Results

During 2023–2025 inebilizumab saves the INHS €8,373,125.13 (1st scenario: €176,770,028.63; 2nd scenario: €168,396,903.50) and 12.74 NMOSD attacks (1st scenario: 213.94; 2nd scenario: 201.19). Sensitivity analyses confirmed the robustness of the baseline results.

Conclusion

Inebilizumab reduces the INHS expenditure for NMOSD drugs. Future research should explore the cost-effectiveness of inebilizumab vs other NMOSD-targeting drugs in Italy.

KEYWORDS:

• Neuromyelitis optica spectrum disorders
• epidemiology
• inebilizumab
• budget impact model
• Italy

1. Introduction

Neuromyelitis optica spectrum disorders (NMOSD) is an autoimmune disease of the central nervous system that affects preferentially the optic nerve and spinal cord [1,2].

NMOSD prevalence is estimated in 0.54–6/100,000 individuals worldwide, with higher number of cases in South America and Asia [3,4].

NMOSD is associated primarily with the anti-aquaporin 4 (AQP4) antibody. Positivity to AQP4-immunoglobulin G (IgG) test helps neurologists with differentiating between NMOSD and multiple sclerosis and provides the specialists with a clinical rationale for starting systemic immunosuppressants (including steroids; azathioprine; mycophenolate and rituximab [RTX]) [5–8].

Without effective therapy, ≥50% of NMOSD patients become clinically blind or permanently disabled/unable to walk independently [9–12], and nearly 80%-90% of them experience recurrent NMOSD relapses (also known as attacks), such as optic neuritis, transverse myelitis, or a combination of the symptoms of both, often with a poor recovery after the attack [10,13].

NMOSD patients report remarkable consumption of healthcare resources, for both acute (e.g. high dose corticosteroids and plasma exchange) and long-term therapies (e.g. disease modifying therapies), neuroimaging, specialist visits, and comorbidities, as well as a worsened health-related quality of life [14,15].

Therefore, NMOSD is associated with relevant annual healthcare costs, that yielded €52,000 per patient in US in 2021 [15,16].

Inebilizumab (Uplizna®) (Horizon Therapeutics, Deerfield, IL) is a monoclonal antibody used as monotherapy to treat NMOSD patients aged ≥18 years who tested AQP4-IgG+ [17].

In the N-Momentum trial 21 out of 174 (12%) of the NMNSOD patient on inebilizumab vs. 22 out of 56 (39%) in the control arm experienced an attack, corresponding to a 73% reduction in adjudicated relapse risk over the trial duration (hazard ratio 0.27; 95% CI 0.15–0.50; p < 0.001) [18]. A further analysis showed a reduction of relapse risk by 77% among the AQP4-IgG+ NMOSD patients on inebilizumab vs. their AQP4-IgG+ counterparts who received placebo.

The Italian National Health Service (INHS) has recently reimbursed the monoclonal antibody inebilizumab as a second line monotherapy after RTX use for NMOSD patients ≥18 years anti-aquaporin 4 antibody-immunoglobulin G positive, who experienced a relapse in the last year or cannot receive RTX, if incident patients [17].

The aim of this research is to create a 2-scenario budget impact model (BIM) and compare the costs and the NMOSD attacks following the INHS viewpoint before (1st scenario) and after (2nd scenario) the availability of inebilizumab on the Italian market.

All the parameters that populate the BIM are detailed in the Supplementary Information (SI) Tables SI1-SI14.

2. Patients and methods

2.1. Patient population

The number of NMOSD patients was calculated by projecting onto the Italian population [19–21] the prevalence and the incident rates obtained from the NMOSD registry maintained by the Lombardy region (North-West part of the country; 9.94 million inhabitants in 2022) [22,23], subtracted NMOSD-related and age and gender-specific all-cause mortality as reported by Italian life tables [24,25] (Tables SI1-SI4).

The gender-specific prevalence rates for NMOSD were stratified into four age classes and expressed as cases per 100,000 inhabitants: 18–39 years (female: 3.227; male: 1.396); 40–64 years (female: 4.482; male: 1.903); 65–79 years (female: 3.114; male: 1.540); 80–109 years (female: 0.708; male: 1.350). The incident rate, fixed for age and gender, was 0.160 cases per 100,000 inhabitants per year [22] (Table SI3). To give a more realistic view, the NMOSD incidence rate was adjusted for age and gender (Table 1).

Table 1. Base case analysis – Methods – BIM main parameters – I.

	Years
	2023	2024	2025	Source
Epidemiology of NMOSD
Prevalent NMOSD patients
Male 18–39 years	98	98	98	[19–23]
Female 18–39 years	215	214	213	[19–23]
Male 40–64 years	205	204	203	[19–23]
Female 40–64 years	499	496	492	[19–23]
Male 65–79 years	69	70	71	[19–23]
Female 65–79 years	161	163	166	[19–23]
Male 80–109 years	23	24	24	[19–23]
Female 80–109 years	20	20	20	[19–23]
Total prevalent NMOSD patients (A)	1291	1293	1296
Incident NMOSD patients
Male 18–39 years	12	12	12	[19–23]
Female 18–39 years	12	12	12	[19–23]
Male 40–64 years	20	20	20	[19–23]
Female 40–64 years	25	25	24	[19–23]
Male 65–79 years	8	8	8	[19–23]
Female 65–79 years	9	9	10	[19–23]
Male 80–109 years	3	3	3	[19–23]
Female 80–109 years	5	5	5	[19–23]
Total incident NMOSD patients (B)	94	94	93
NMOSD-related death (prevalent patients only)
Male 18–39 years	1	1	1	[24]
Female 18–39 years	2	2	2	[24]
Male 40–64 years	2	2	2	[24]
Female 40–64 years	4	4	4	[24]
Male 65–79 years	1	1	1	[24]
Female 65–79 years	1	1	1	[24]
Male 80–109 years	0	0	0	[24]
Female 80–109 years	0	0	0	[24]
Total NMOSD-related death (C)	11	11	11
All-cause-related death different from NMOSD (prevalent patients only)
Male 18–39 years	0	0	0	[25]
Female 18–39 years	0	0	0	[25]
Male 40–64 years	11	11	11	[25]
Female 40–64 years	13	13	13	[25]
Male 65–79 years	14	14	14	[25]
Female 65–79 years	20	20	20	[25]
Male 80–109 years	12	12	12	[25]
Female 80–109 years	12	12	12	[25]
Total all-cause-related death different from NMOSD (D)	80	80	80
Total NMOSD-related + all-cause mortality	91	91	92
Total (prevalent + incident) NMOSD patients alive (E)	1293	1296	1297
Testing AQP4-IgG and eligibility to inebilizumab
Patients AQP4-IgG tested (F)^a	1007	1009	1010	[26]
Patients AQP4-IgG+ given test (G)^b	799	800	801	[26]
Patients eligible to therapy given AQP4-IgG+ (H)^c	719	720	721	[26]
Incident NMOSD patients ineligible to RTX given AQP4-IgG+ (J)^d	11	11	11	[26]
Prevalent NMOSD patients switching from RTX given AQP4-IgG+ (K)^e	382	383	383	[27]
Total (prevalent + incident) NMOSD patients eligible to NMOSD therapy different from RTXf	393	394	394

Abbreviations: AQP4, Aquaporin-4; BIM, budget impact model; IgG, immunoglobulin G; NMOSD, neuromyelitis optica spectrum disorders; RTX, rituximab.

^a(E) x 0.779, where 0.779 is the probability of being tested for AQP4-IgG [26].

^b(F) x 0.793, where 0.793 is the conditional probability of resulting positive given AQP4-IgG test [26].

^c(G) x 0.900, where 0.900 is the conditional probability of being eligible to therapy given AQP4-IgG+ [26].

^d(B)/[(A)+(B)] x (H) x 0.233, where 0.233 is the conditional probability of not being eligible to RTX given eligibility to therapy [26].

^e[(H)-(J)] x 0.540, where 0.540 is the conditional probability of switching from RTX (for lack of effectiveness or safety) given eligibility to therapy [27].

^f[(J)+(K)].

Only prevalent patients were assumed to die. To avoid double-counting, the annual probability of NMOSD-related death (0.015) [24] and age and gender-specific all-cause mortality [25] were considered as competing risks [28]. Consistent with the way prevalence rates for NMOSD were reported [22], age and gender-specific all-cause mortality retrieved from the Italian life tables [25] were stratified in four age classes [28]. The half-cycle correction (that is, the assumption that a patient passes away at six months instead of at the beginning or at the end of the year) was applied [29] (Table SI2).

2.2. Design of the budget impact model

Following the international and the Italian Medicines Agency (Agenzia Italiana del Farmaco – AIFA) guidelines [30,31], a prevalence-based BIM was developed in Excel Windows® 2010 (Microsoft, Redmond, WA, US).

Both BIM scenarios, which stretch over a 3-year time horizon (2023–2025), included NMOSD patients who tested AQP4-IgG+ and were considered eligible to receive inebilizumab according to the reimbursement conditions negotiated with AIFA [17,26,27] (Table 1; Table SI4).

As the patient population is forecast to remain stable during 2023–2025 [19–21], no variation in the number of NMOSD AQP4-IgG+ patients was expected.

In the 1st scenario, that does not include inebilizumab, incident NMOSD patients who were not eligible to receive RTX, and treatment experienced NMOSD patients who suffered a relapse in the last year were administered satralizumab, eculizumab, ocrelizumab, tocilizumab or immunosuppressants (mycophenolate mofetil, azathioprine or prednisone, as per neurologist’s choice). While ocrelizumab and tocilizumab are used as off-label, second line therapies in NMOSD after RTX, their beneficial effects on NMOSD patients’ health state have been proved [32–34].

In the 2nd scenario inebilizumab is the new entry that broadens the therapeutic options for NMOSD patients in Italy included in the 1st scenario.

When rounded up to the nearest integer, the number of patients eligible for NMOSD therapies different from RTX ranged between 393 (2023) and 394 (2024 and 2025) per year in both scenarios (Table 1). The variation in INHS cost due to inebilizumab availability on the Italian market was calculated as the difference between the costs totalized in the two scenarios.

As the BIM is not populated with real patients’ data, no Ethics Committee approval was required by the current Italian legislation [35].

2.3. Market shares

The market shares for all the drugs included in the 1st scenario were retrieved from a recent market research study [36] (Table 2; Table SI7).

Table 2. Base case analysis – Methods – BIM main parameters – II.

	Drugs
	Inebilizumab		Satralizumab		Eculizumab		Ocrelizumab		Tocilizumab		Immunosuppressants^a
Healthcare resources consumption	Non monetary value	Unit cost	Non monetary value	Unit cost	Non monetary value	Unit cost	Non monetary value	Unit cost	Non monetary value	Unit cost	Non monetary value	Unit cost	Source
Drug units per year 2023	9	€15,838.88	15	€7310.25	108	€4142.75	3	€5640.63	39	€320.79	1460^b;1095^c;365^d	€0.80	[33, 34, 37–41]
Drug units per year 2024–2025	6	€15,838.88	13	€7310.25	104	€4142.75	2	€5640.63	39	€320.79	1460^b;1095^c;365^d	€0.80	[33, 34, 37–41]
Drug administrations per year 2023	3	€11.62	15	€0.00	28	€11.62	3	€11.62	13	€11.62	365	€0.00	[33, 34, 37–42]
Drug administrations per year 2024–2025	2	€11.62	13	€0.00	28	€11.62	2	€11.62	13	€11.62	365	€0.00	[33, 34, 37–42]
Office visits per year 2023 (1^st visit)	0	-	1	€20.66	0	-	0	-	0	-	1	€20.66	ra, [42],
Office visits per year 2023 (Follow-up visit)	0	-	1	€20.66	0	-	0	-	0	-	1	€20.66	ra, [42],
Office visits per year 2024–2025 (1^st visit)	0	-	1	€20.66	0	-	0	-	0	-	1	€20.66	ra, [42],
Office visits per year 2024–2025 (Follow-up visit)	0	-	1	€20.66	0	-	0	-	0	-	1	€20.66	ra, [42],
AAR 2023–2025	0.052	€2126.97	0.080	€2126.97	0.020	€2126.97	0.157	€2126.97	0.157	€2126.97	0.404	€2126.97	[32, 43–45,46, 47]
AE – Annual probability of anaphylaxis year 2023–2025	0.000	-	0.000	-	0.000	-	0.005	€898.88	0.005	€898.88	0.000	-	[26, 46, 47]
AE – Annual probability of arthralgia year 2023–2025	0.011	€4516.45	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of renal abscess year 2023–2025	0.000	-	0.000	-	0.003	€4516.45	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of bacteremia year 2023–2025	0.000	-	0.000	-	0.000	-	0.002	€5781.92	0.002	€5781.92	0.000	-	[26, 46, 47]
AE – Annual probability of bronchitis year 2023–2025	0.000	-	0.000	-	0.003	€2021.56	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of cellulitis year 2023–2025	0.000	-	0.000	-	0.003	€456.23	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of acute cholangitis year 2023–2025	0.011	€2370.05	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of acute cholecystitis year 2023–2025	0.011	€2370.05	0.000	-	0.003	€2370.05	0.000	-	0.000	-	0.000	-	[18, 43, 46, 47]
AE – Annual probability of diarrhea year 2023–2025	0.011	€765.98	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of dyspnea year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€2101.24	[18, 46, 47]
AE – Annual probability of impairments of the nervous system year 2023–2025	0.000	-	0.005	€1587.81	0.000	-	0.000	-	0.000	-	0.000	-	[46–48]
AE – Annual probability of hemo-lymphopoietic cell system year 2023–2025	0.000	-	0.005	€2609.02	0.000	-	0.000	-	0.000	-	0.000	-	[46–48]
AE – Annual probability of upper abdomen pain year 2023–2025	0.000	-	0.000	-	0.003	€1315.38	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of migraine year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€1310.77	[18, 46, 47]
AE – Annual probability of gallbladder empyema year 2023–2025	0.000	-	0.000	-	0.003	€2370.05	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of fever year 2023–2025	0.000	-	0.000	-	0.003	€2279.91	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of atrial fibrillation year 2023–2025	0.000	-	0.000	-	0.003	€1187.65	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of liver impairment year 2023–2025	0.011	€2170.65	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of infection year 2023–2025	0.000	-	0.042	€5781.92	0.000	-	0.005	€5781.92	0.005	€5781.92	0.000	-	[26, 46–48]
AE – Annual probability of hypoglycemia year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€2153.84	[18, 46, 47]
AE – Annual probability of injury, drug poisoning or procedure-related complication year 2023–2025	0.000	-	0.021	€1411.73	0.000	-	0.000	-	0.000	-	0.000	-	[46–48]
AE – Annual probability of viral meningitis year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€2629.00	[18, 46, 47]
AE – Annual probability of myelitis year 2023–2025	0.011	€1922.70	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of neutropenia year 2023–2025	0.000	-	0.000	-	0.000	-	0.036	€2609.02	0.036	€2609.02	0.000	-	[26, 46, 47]
AE – Annual probability of deep venous thromboembolism year 2023–2025	0.000	-	0.000	-	0.010	€2265.22	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of pneumonia year 2023–2025	0.011	€3164.75	0.000	-	0.005	€3164.75	0.000	-	0.000	-	0.035	€3164.75	[18, 43, 46, 47]
AE – Annual probability of allergic reaction year 2023–2025	0.000	-	0.000	-	0.000	-	0.010	€898.88	0.010	€898.88	0.000	-	[26, 46, 47]
AE – Annual probability of sepsis year 2023–2025	0.000	-	0.000	-	0.003	€5781.92	0.002	€5781.92	0.002	€5781.92	0.000	-	[26, 43, 46, 47]
AE – Annual probability of third-degree burn year 2023–2025	0.011	€3547.15	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	[18, 46, 47]
AE – Annual probability of pleuritis year 2023–2025	0.000	-	0.000	-	0.003	€3854.34	0.000	-	0.000	-	0.000	-	[43, 46, 47]
AE – Annual probability of chest pain year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€1326.18	[18, 46, 47]
AE – Annual probability of visual impairment year 2023–2025	0.000	-	0.000	-	0.000	-	0.000	-	0.000	-	0.035	€1966.54	[18, 46, 47]
MS scenario without inebilizumab year 2023	-	-	0.283	-	0.212	-	0.032	-	0.047	-	0.426	-	[36]
MS scenario without inebilizumab year 2024	-	-	0.348	-	0.271	-	0.015	-	0.030	-	0.335	-	[36]
MS scenario without inebilizumab year 2025	-	-	0.407	-	0.292	-	0.003	-	0.011	-	0.286	-	[36]
MS scenario with inebilizumab year 2023	0.040	-	0.272	-	0.196	-	0.028	-	0.042	-	0.422	-	Elaboration on [36]
MS scenario with inebilizumab year 2024	0.110	-	0.322	-	0.228	-	0.007	-	0.021	-	0.313	-	Elaboration on [36]
MS scenario with inebilizumab year 2025	0.200	-	0.331	-	0.235	-	0.003	-	0.001	-	0.230	-	Elaboration on [36]

Abbreviations: AAR, annual attack rate; AE, adverse event; BIM, budget impact model; MS, market share; ra, research assumption.

^aThe BIM assumes that three drugs can be prescribed as immunosuppressants: mycophenolate mofetil (conditional probability 2023–2025 of mycophenolate mofetil given immunosuppressants = 0.330); azathioprine (conditional probability 2023–2025 of azathioprine given immunosuppressants = 0.330); prednisone (conditional probability 2023–2025 of prednisone given immunosuppressants = 0.340). These conditional probabilities hold for both scenarios.

^bDrug units per year 2023–2025 mycophenolate mofetil.

^cDrug units per year 2023–2025 azathioprine.

^dDrug units per year 2023–2025 prednisone.

The availability of inebilizumab was assumed to cause a redistribution of the market shares of NMOSD-targeting drugs that was captured in the 2nd scenario. The variation in the market shares was estimated by means of reasonable forecasts based on the market shares that populated the 1st scenario of the BIM [36] (Table 2; Table SI9).

In both scenarios during the 3-year timespan the conditional probability of receiving mycopehate mofetil, azathioprine or prednisone given immunosuppressant therapy was estimated to be constant at 0.330, 0.330 and 0.340, respectively (Table 2; Tables SI8 and SI10).

2.4. Healthcare resources

For inebilizumab, satralizumab and eculizumab the annual posology and the number of administrations were obtained from the summary of product characteristics [37–39]. The same parameters were retrieved from the literature for ocrelizumab [33], tocilizumab [34], immunosuppressants (mycophenolate mofetil and azathioprine) [40] and prednisone [41] (Table 2; Table SI2).

The route of administration (RoA) was subcutaneous (sc) for satralizumab and per os for immunosuppressants, whereas for the other drugs the RoA was infusional (Table SI2).

A first and a follow-up specialist visits per year were assumed for all the drugs that were administered sc or per os, whereas patients receiving infusion therapy were hypothesized to be assessed upon drug administration without the need of further specialist visits (Table 2; Table SI11).

As reported by other authors [49] patients’ perfect medication compliance to drug prescription was assumed in both scenarios.

The drug-specific annual attack rate (AAR) was obtained from randomized clinical trials for inebilizumab, satralizumab, eculizumab, and immunosuppressants [32,43–45] (Table 2; Tables SI3 and SI14).

Ocrelizumab and tocilizumab were assumed to have the same AAR of RTX (0.157) [32] for two different reasons. Ocrelizumab and RTX share the same mechanism of action [50], whereas there is no evidence that tocilizumab is more effective than RTX or ocrelizumab.

In addition, the AAR estimate of tocilizumab was conservative, as the only available study on this drug was conducted on high-activity NMOSD patients whose AAR was 0.4 [34] instead of 0.157 [32].

The probability of NMOSD drugs-related level ≥ 3 [51] adverse events (AEs) was provided by literature [18,26,43,48] (Table 2; Tables SI5 and SI6).

For all the drugs the AAR as well as the probability of experiencing a given NMOSD drug-related AE were kept constant during the 3-year timespan the BIM stretches over (Table 2; Tables SI3, SI5 and SI6).

No healthcare resource concerning NMOSD diagnosis and Emergency Room visits were considered, as their consumption was not assumed to differ among the drugs included in the BIM.

2.5. Cost

Consistent with the BIM viewpoint, only INHS-funded healthcare resources were costed (Table 2; Tables SI12 and SI13). Drugs were valued at their net price that is, ex-factory price x 0.95^{^2} [17,52–57]. Hidden discount (that is, the mandatory reduction of the ex-factory price reported in the AIFA decree issued after the reimbursement decision. The hidden discount should be acknowledged by drug manufacturer to public hospitals and sometimes also to private hospitals endorsed by INHS as healthcare services providers) was not considered, as the percentage of the hidden discount on the ex-factory price is kept confidential between AIFA and the manufacturer [58].

Infusion drug administrations, as well as specialist visits, were valued according to the version of the INHS tariffs for outpatient procedures currently in force [42] (Table 2 ; Table SI12).

Per os and sc RoAs were not costed as patient self-administration was assumed.

As NMOSD patients are oftentimes severely ill [14,15], NMOSD attacks and NMOSD drugs-related AEs were assumed to be managed in hospital setting and costed via the last updated Diagnosis Related Groups (DRGs) tariffs adopted by the INHS [46].

As the Italian DRGs tariffs for inpatient and day-hospital stay differ, a weighted DRG tariff for AAR and each NMOSD drug-related AE was calculated, the weight being the last available national number of medical and surgical DRG-specific discharges from inward and day-hospital settings with and without complications or comorbidities for patients aged ≥18 years, as reported by the Italian Ministry of Health [47] (Table 2 ; Tables SI12 and SI13).

Costs were expressed in €2023 values.

Following the recommendations of the literature on BIM [30] costs were not discounted.

2.6. Statistical analysis

A parametric 95% confidence interval (95% CI) was calculated for all the BIM parameters (167/244 = 68.44%) whose uncertainty could be modeled via a theoretical probability distribution [29,59,60]. Whenever the sample size (N) was available, the standard error (SE) of the sampling distribution of the parameter was calculated analytically. A coefficient of variation was imposed on the parameter point estimate to determine the SE of the mean if N could not be collected from the data [29,59,60].

The Gamma distribution was applied to NMOSD prevalence and incidence rates as well as AARs (Table SI3).

The Beta distribution was assigned to binomial data (e.g. probability for a NMOSD patient to be tested for AQP4-IgG) (Tables SI4-SI6).

A Dirichlet distribution was fitted to multinomial data (e.g. market shares of the different drugs in both scenarios the BIM is composed of) (Tables SI7-SI 10).

If different from drugs, the uncertainty concerning the volume and the unit cost of healthcare resources consumed by notional NMOSD patients was modeled via a Gamma and a Normal distribution, respectively (Tables SI11-SI 13).

Drug posology, administrations and cost are exogenous parameters [59] as their value is approved or set by international or national regulatory agencies (e.g. European Medicines Agency; AIFA). Therefore, their values are certain (Table 2; Table S2).

Sample estimates and ranges for age-specific forecasts of the Italian population were retrieved from the same official sources [19–21] (Table S1).

2.7. Sensitivity analyses

The robustness of the base case results was checked via a comprehensive set of sensitivity analyses [61,62].

2.7.1. One-way sensitivity analysis

The way each one of the 244 parameters included in the BIM contributed to the cost difference for the INHS between the 1st and 2nd scenarios was first investigated via one-way sensitivity analysis (OWSA). In OWSA the bounds of the range or 95% CI replaced the sample estimate of each parameter at a time while keeping the other ones unchanged [30,31,61,62].

2.7.2. Scenario sensitivity analysis

The reliability of the base case results was also explored via five scenario sensitivity analyses (SSA) [30,31,61,62].

A first SSA tested if a reduction of the BIM time horizon (from 1 up to 2 years) changes the base case findings due to a progressive but incomplete redistribution of the market shares after the entry of inebilizumab in the Italian market.

A second SSA assumed that only patients belonging to one of the four age class (18–39; 40–64; 65–79; 80–109 years) entered the BIM at a time.

In a third SSA the baseline prevalence and incidence rates were replaced by the lower and upper bounds of their 95% CIs.

The same replacement was made for AAR in the fourth SSA.

The fifth SSA investigated three different hypotheses of extreme redistribution of the market shares of inebilizumab, satralizumab and eculizumab in the 2nd scenario of the BIM in which inebilizumab was available. Each drug was assumed to capture the market shares of the remaining two and the subsequent effect on the INHS saving was calculated. The other medications maintained the market shares estimated in the 2nd scenario of the BIM (Table 2).

2.7.3. Threshold value analyses

Threshold value analysis (TVA) aims at determining the critical value of a given parameter (or a set of parameters) after which the baseline results reverse (e.g. the base case saving becomes an additional cost for the third-party payer) [61–63].

Both TVAs refer to the 2nd scenario of the BIM as they assumed the availability of inebilizumab.

The first TVA investigated the effect of a variation (±10% at a time) in the 3-year average market share for inebilizumab (from −100% to + 100% vs baseline BIM) counterbalanced by an equivalent variation in the market shares of satralizumab and eculizumab.

The second TVA explored the impact of reducing or increasing (±5% at a time) the net price of inebilizumab (from −100% to + 100% vs baseline BIM).

3. Results

3.1. Base case analysis

The annual and 3-year per patient costs for the drugs included in the BIM are reported separately (Tables SI15-SI 23) and, unlike the BIM scenarios, they are calculated without taking the respective market shares into account. During the 3-year time horizon their amounts vary from €5995.85 (immunosuppressants) (Table SI23) to €1,310,512.56 (eculizumab) per patient (Table SI17).

As far as the BIM is concerned, the overall cost for the INHS yields €176,770,028.63 (1st scenario; inebilizumab unavailable on the Italian market) and €168,396,903.50 (2nd scenario; inebilizumab available on the Italian market). For both scenarios, the cost for the INHS is led by the cost of drugs that totaled €175,802,696.50 and €167,480,869.43 (or 99.45% and 99.46% of the overall cost), respectively (Tables 3 and 4). Conversely, the office visits showed the smallest contribution to the overall cost (0.02%), with €33,943.82 and €30,755.18 in the 1st and 2nd scenario, respectively.

Table 3. Base case analysis – Results – Cost and saving for the INHS – I – Scenario without inebilizumab (€2023)^a.

	Years
	2023	2024	2025	Overall
Satralizumab (cost per patient)
Cost of drug	€109,653.75	€95,033.25	€95,033.25	€299,720.25
Cost of drug administration	€0.00	€0.00	€0.00	€0.00
Cost of office visits	€41.32	€41.32	€41.32	€123.96
Cost of adverse events	€291.67	€291.67	€291.67	€875.00
Cost of NMOSD attacks	€170.16	€170.16	€170.16	€510.47
Total cost per patient	€110,156.89	€95,536.39	€95,536.39	€301,229.68
Total (prevalent + incident) NMOSD patients eligible to Satralizumab	111	137	161
Eculizumab (cost per patient)
Cost of drug	€447,416.54	€430,845.56	€430,845.56	€1,309,107.66
Cost of drug administration	€325.36	€302.12	€302.12	€929.60
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€115.90	€115.90	€115.90	€347.69
Cost of NMOSD attacks	€42.54	€42.54	€42.54	€127.62
Total cost per patient	€447,900.34	€431,306.11	€431,306.11	€1,310,512.56
Total (prevalent + incident) NMOSD patients eligible to Eculizumab	83	107	115
Ocrelizumab (cost per patient)
Cost of drug	€16,921.88	€11,281.25	€11,281.25	€39,484.38
Cost of drug administration	€34.86	€23.24	€23.24	€81.34
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€163.23	€163.23	€163.23	€489.68
Cost of NMOSD attacks	€333.04	€333.04	€333.04	€999.11
Total cost per patient	€17,453.00	€11,800.75	€11,800.75	€41,054.50
Total (prevalent + incident) NMOSD patients eligible to Ocrelizumab	12	6	1
Tocilizumab (cost per patient)
Cost of drug	€12,510.95	€12,510.95	€12,510.95	€37,532.85
Cost of drug administration	€151.06	€151.06	€151.06	€453.18
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€163.23	€163.23	€163.23	€489.68
Cost of NMOSD attacks	€333.04	€333.04	€333.04	€999.11
Total cost per patient	€13,158.27	€13,158.27	€13,158.27	€39,474.82
Total (prevalent + incident) NMOSD patients eligible to Tocilizumab	19	12	5
Immunosuppressants (cost per patient)
Cost of drug	€773.25	€484.38	€484.38	€1742.01
Cost of drug administration	€0.00	€0.00	€0.00	€0.00
Cost of office visits	€41.32	€41.32	€41.32	€123.96
Cost of adverse events	€518.13	€518.13	€518.13	€1554.40
Cost of NMOSD attacks	€858.49	€858.49	€858.49	€2575.48
Total cost per patient	€2191.20	€1902.32	€1902.32	€5995.85
Total (prevalent + incident) NMOSD patients eligible to Immunosuppressants	168	132	113
Overall INHS – Cost – Scenario without inebilizumab
Cost of drug	€50,131,892.75	€60,114,815.76	€65,555,987.99	€175,802,696.50
Cost of drug administration	€30,401.59	€34,795.67	€35,750.51	€100,947.77
Cost of office visits	€11,523.71	€11,119.68	€11,300.43	€33,943.82
Cost of adverse events	€134,013.40	€123,754.46	€119,637.91	€377,405.77
Cost of NMOSD attacks	€176,634.56	€147,320.40	€131,079.82	€455,034.77
Overall INHS cost	€50,484,466.01	€60,431,805.96	€65,853,756.66	€176,770,028.63

Abbreviations: INHS, Italian National Health Service; MS, market share; NMOSD, neuromyelitis optica spectrum disorders.

^aFor each drug the number of patients per year is calculated by multiplying the number of annual patients (Table 1) by the annual MS without inebilizumab (Table 2). Values are rounded up to the nearest integer.

Table 4. Base case analysis – Results – Cost and saving for the INHS – II – Scenario with inebilizumab (€2023)^a.

	Years
	2023	2024	2025	Overall
Inebilizumab (cost per patient)
Cost of drug	€142,549.88	€95,033.25	€95,033.25	€332,616.38
Cost of drug administration	€34.86	€23.24	€23.24	€81.34
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€207.42	€207.42	€207.42	€622.26
Cost of NMOSD attacks	€110.60	€110.60	€110.60	€331.81
Total cost per patient	€142,902.76	€95,374.51	€95,374.51	€333,651.79
Total (prevalent + incident) NMOSD patients eligible to Inebilizumab	16	43	79
Satralizumab (cost per patient)
Cost of drug	€109,653.75	€95,033.25	€95,033.25	€299,720.25
Cost of drug administration	€0.00	€0.00	€0.00	€0.00
Cost of office visits	€41.32	€41.32	€41.32	€123.96
Cost of adverse events	€291.67	€291.67	€291.67	€875.00
Cost of NMOSD attacks	€170.16	€170.16	€170.16	€510.47
Total cost per patient	€110,156.89	€95,536.39	€95,536.39	€301,229.68
Total (prevalent + incident) NMOSD patients eligible to Satralizumab	107	127	131
Eculizumab (cost per patient)
Cost of drug	€447,416.54	€430,845.56	€430,845.56	€1,309,107.66
Cost of drug administration	€325.36	€302.12	€302.12	€929.60
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€115.90	€115.90	€115.90	€347.69
Cost of NMOSD attacks	€42.54	€42.54	€42.54	€127.62
Total cost per patient	€447,900.34	€431,306.11	€431,306.11	€1,310,512.56
Total (prevalent + incident) NMOSD patients eligible to Eculizumab	77	90	93
Ocrelizumab (cost per patient)
Cost of drug	€16,921.88	€11,281.25	€11,281.25	€39,484.38
Cost of drug administration	€34.86	€23.24	€23.24	€81.34
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€163.23	€163.23	€163.23	€489.68
Cost of NMOSD attacks	€333.04	€333.04	€333.04	€999.11
Total cost per patient	€17,453.00	€11,800.75	€11,800.75	€41,054.50
Total (prevalent + incident) NMOSD patients eligible to Ocrelizumab	11	3	1
Tocilizumab (cost per patient)
Cost of drug	€12,510.95	€12,510.95	€12,510.95	€37,532.85
Cost of drug administration	€151.06	€151.06	€151.06	€453.18
Cost of office visits	€0.00	€0.00	€0.00	€0.00
Cost of adverse events	€163.23	€163.23	€163.23	€489.68
Cost of NMOSD attacks	€333.04	€333.04	€333.04	€999.11
Total cost per patient	€13,158.27	€13,158.27	€13,158.27	€39,474.82
Total (prevalent + incident) NMOSD patients eligible to Tocilizumab	17	8	0
Immunosuppressants (cost per patient)
Cost of drug	€773.25	€484.38	€484.38	€1742.01
Cost of drug administration	€0.00	€0.00	€0.00	€0.00
Cost of office visits	€41.32	€41.32	€41.32	€123.96
Cost of adverse events	€518.13	€518.13	€518.13	€1554.40
Cost of NMOSD attacks	€858.49	€858.49	€858.49	€2575.48
Total cost per patient	€2191.20	€1902.32	€1902.32	€5995.85
Total (prevalent + incident) NMOSD patients eligible to Immunosuppressants	166	123	91
Overall INHS – Cost – Scenario with inebilizumab
Cost of drug	€48,994,057.39	€56,782,130.05	€61,704,681.99	€167,480,869.43
Cost of drug administration	€28,514.22	€30,016.52	€30,411.47	€88,942.21
Cost of office visits	€11,280.54	€10,330.61	€9144.04	€30,755.18
Cost of adverse events	€133,915.19	€122,032.07	€112,457.53	€368,404.79
Cost of attacks	€174,912.11	€139,717.87	€113,301.91	€427,931.89
Overall INHS cost	€49,342,679.44	€57,084,227.12	€61,969,996.94	€168,396,903.50
Overall INHS – Saving – Scenario with inebilizumab
Cost of drug	€1,137,835.37	€3,332,685.70	€3,851,306.00	€8,321,827.07
Cost of drug administration	€1887.37	€4779.15	€5339.05	€12,005.56
Cost of office visits	€243.17	€789.07	€2156.39	€3188.63
Cost of adverse events	€98.22	€1722.39	€7180.38	€9000.98
Cost of NMOSD attacks	€1722.45	€7602.53	€17,777.90	€27,102.88
Overall INHS savingb	€1,141,786.57	€3,347,578.84	€3,883,759.72	€8,373,125.13

Abbreviations: INHS, Italian National Health Service; MS, market share; NMOSD, neuromyelitis optica spectrum disorders.

^aFor each drug the number of patients per year is calculated multiplying the number of annual patients (Table 1) by the annual MS with inebilizumab (Table 2). Values are rounded up to the nearest integer.

^bThe overall INHS saving is the difference between the overall INHS cost without and with inebilizumab (Tables 3 and 4).

The availability of inebilizumab on the Italian market reduces the number of NMOSD attacks from 213.94 to 201.19, with a difference of 12.74 over the 3-year time horizon (Figure 1; Table SI24), whereas the cost of managing the NMOSD relapses does not exceed 0.26% (€455,034.77) and 0.24% (€427,931.89) of the 3-year INHS cost in the 1st and 2nd scenario, respectively (Tables 3 and 4).

Figure 1. Base case analysis - Number of NMOSD attacks - Scenario without inebilizumab (Panel A), scenario with inebilizumab (Panel B) and difference between the two scenarios (Panel C)

Abbreviations: NMSOD, neuromyelitis optica spectrum disorders.

The availability of inebilizumab on the Italian market generates a €8,373,125.13 overall saving for the INHS, that was mainly due to a reduction in public expenditure for NMOSD-targeting drugs (€8,321,827.07, or 99.39% of the INHS total saving) (Table 4).

3.2. Sensitivity analyses

3.2.1. One-way sensitivity analysis

All the OWSA findings are consistent with the baseline results in showing a 3-year saving for the INHS due to inebilizumab entering the Italian market (Figure 2).

Figure 2. One-way sensitivity analysis - Results concerning the first 10 parameters of the BIM that causes the largest variations in the base case INHS saving due to inebilizumab availability (€8,373,125.13) (€2023). ^a

Abbreviations: BIM, budget impact model; CI, confidence interval; INHS, Italian National Health Service; LL, lower limit; NMSOD, neuromyelitis optica spectrum disorders; UL, upper limit; w, with; wo, without. aY and X-axes intersect at the baseline INHS saving.

Interestingly, the market shares that populated the 1st and the 2nd BIM scenarios represent 9 out of the 10 most relevant OWSA parameters.

The overall INHS saving calculated in the base case BIM is most sensitive to variations in the market share for eculizumab in 2025 in the 1st scenario (from −89.01% to + 94.75% vs baseline results).

Conversely, changes in the satralizumab market share in 2025 (1st scenario) have a mild leveraging on the 3-year INHS saving (from −18.85% to + 19.12% vs baseline results).

3.2.2. Scenario sensitivity analysis

The first SSA highlights the role of the increasing market share for inebilizumab in the 2nd scenario, as 46.38% of the overall saving for the INHS is generated during the 3rd year (Table 5).

Table 5. Scenario sensitivity analysis – Results (€2023).

	Saving for the INHS due to inebilizumab availability on the Italian market (per year and overall)
Parameter variation	2023	2024	2025	Overall
Base case analysis	€1,141,786.57	€3,347,578.84	€3,883,759.72	€8,373,125.13
SSA 1 – BIM time horizon
1 year	€1,141,786.57	-	-	€1,141,786.57
Δ vs base case analysis	0.00%	-	-	−86.36%
2 years	€1,141,786.57	€3,347,578.84	-	€4,489,365.41
Δ vs base case analysis	0.00%	0.00%	-	−46.38%
SSA 2 - Patients age class when entering the BIM
18–39 years only	€296,618.60	€864,949.73	€999,593.77	€2,161,162.10
Δ vs base case analysis	−74.02%	−74.16%	−74.26%	−74.19%
40–64 years only	€633,291.64	€1,846,831.62	€2,124,990.27	€4,605,113.53
Δ vs base case analysis	−44.54%	−44.83%	−45.29%	−45.00%
65–79 years only	€186,558.70	€553,158.67	€652,145.47	€1,391,862.85
Δ vs base case analysis	−83.66%	−83.48%	−83.21%	−83.38%
80–109 years only	€24,133.39	€70,968.91	€82,376.54	€177,478.85
Δ vs base case analysis	−97.89%	−97.88%	−97.88%	−97.88%
SSA 3 – LL and UL 95% CI prevalence and incidence NMOSD
LL	€395,200.34	€1,156,749.42	€1,340,540.28	€2,892,490.04
Δ vs base case analysis	−65.39%	−65.45%	−65.48%	−65.46%
UL	€2,019,416.69	€5,910,822.60	€6,849,967.37	€14,780,206.66
Δ vs base case analysis	76.86%	76.57%	76.37%	76.52%
SSA 4 – LL and UL 95% CI AAR
LL	€1,140,712.35	€3,342,837.46	€3,872,672.40	€8,356,222.22
Δ vs base case analysis^a	−0.09%	−0.14%	−0.29%	−0.20%
UL	€1,143,376.49	€3,354,596.40	€3,900,169.73	€8,398,142.61
Δ vs base case analysis^b	0.14%	0.21%	0.42%	0.30%
SSA 5 – Variations in inebilizumab MS
Inebilizumab MS captures satralizumab and eculizumab MS^c,d	€21,153,964.13	€32,446,262.89	€34,823,436.93	€88,423,663.95
Δ vs base case analysis	1752.71%	869.25%	796.64%	956.04%
Satralizumab MS captures inebilizumab and eculizumab MS^e,f	€27,697,766.50	€34,373,181.59	€36,172,973.68	€98,243,921.77
Δ vs base case analysis	2325,83%	926,81%	831,39%	1073,32%
Eculizumab MS captures inebilizumab and satralizumab MS^g,h	-€39,795,527.93	-€52,933,737.40	-€65,368,633.41	-€158,097,898.74
Δ vs base case analysis	−3585,37%	−1681,25%	−1783,13%	−1988,16%

Abbreviations: AAR, annual attack rate; BIM, budget impact model; CI, confidence interval; INHS, Italian National Health Service; LL, lower limit; MS, market share; NMOSD, neuromyelitis optica spectrum disorders; sensitivity analysis; UL, upper limit.

^aThis corresponds to an overall number of 4.80 NMOSD attacks saved thanks to inebilizumab availability (2023: 0.30; 2024: 1.35; 2025: 3.15), or −62.32% vs base case results.

^bThis corresponds to an overall number of 24.50 NMOSD attacks saved thanks to inebilizumab availability (2023: 1.56; 2024: 6.87; 2025: 16.07), or + 92.31% vs base case results.

^cMS inebilizumab (2023: 0.508; 2024: 0.659; 2025: 0.766); MS satralizumab and eculizumab (2023–2025: 0.00); MS other drugs (2023–2025: MS scenario with inebilizumab, as per Table 1).

^dThis extreme MS redistribution corresponds to an overall number of 14.64 NMOSD attacks saved (2023: 1.34; 2024: 4.25; 2025: 9.05), or + 14.91% vs base case results.

^eMS satralizumab (2023: 0.508; 2024: 0.659; 2025: 0.766); MS inebilizumab and eculizumab (2023–2025: 0.00); MS other drugs (2023–2025: MS scenario with inebilizumab, as per Table 1).

^fThis extreme MS redistribution corresponds to an overall number of 6.69 NMOSD attacks in excess (2023: −4.26; 2024: −3.02; 2025: 0.59), or −152.49% vs base case results.

^gMS eculizumab (2023: 0.508; 2024: 0.659; 2025: 0.766); MS inebilizumab and satralizumab (2023–2025: 0.00); MS other drugs (2023–2025: MS scenario with inebilizumab, as per Table 1).

^hThis extreme MS redistribution corresponds to an overall number of 39.01 NMOSD attacks saved (2023: 7.73; 2024: 12.56; 2025: 18.72), or + 206.22% vs base case results.

The results of the second SSA are consistent with the epidemiology of NMOSD in Italy. Since the highest prevalence of the disease is reported for the 40–64 age class, populating the BIM solely with this stratum of patients causes the lowest reduction in the overall INHS saving vs baseline analysis (−45.00%).

The third SSA shows that replacing the point estimate of age and gender-specific prevalence and incidence rates with the lower and the upper bounds of their respective 95% CIs, produces remarkable departures from the overall INHS saving estimated in the base case BIM (−65.46% and + 76.52%, respectively).

The fourth SSA proves the impact on the 3-year INHS saving due to the same replacement for AAR to be negligible (−0.20% and + 0.30% vs baseline analysis, respectively), while sparing 4.80 and 24.50 attacks (−62.32% and + 92.31% vs base case findings, respectively).

The findings of the fifth SSA confirms the cost-saving profile of inebilizumab. When inebilizumab captures the market shares of satralizumab and eculizumab, the overall saving for the INHS reaches €88,423,663.95 with 14.64 NMOSD attacks spared (+956.04% and +14.91% vs base case results, respectively), whereas it raises to €98,243,921.77, but with 6.69 more NMOSD attacks (+1073.32% and −152.49% vs baseline BIM, respectively) when both inebilizumab and eculizumab are totally replaced by satralizumab. Conversely, when the market shares of inebilizumab and satralizumab are assumed to drop to zero in favor of eculizumab, the overall saving for the INHS becomes negative. Put differently, this last hypothesis creates an expected INHS overall cash outlay of €158,097,898.74 while saving 39.01 NMOSD attacks (−1988.16% and + 206.22 vs base case analysis, respectively).

Interestingly, even under these extreme redistribution of market shares, inebilizumab only increases the INHS saving and reduces NMSOD attacks when compared to the baseline findings.

3.2.3. Threshold value analyses

The first TVA shows that the 3-year saving for the INHS becomes < 0 only if the baseline 3-year mean market share of inebilizumab drops by ≥ 30% (Figure 3). Doubling the average market share of inebilizumab (that is, from 0.117 to 0.233) during the timespan the BIM stretches over, has a 4.61-time multiplicative effect on the overall INHS savings calculated in the base case analysis (€38,594,983.99 vs €8,373,125.13).

Figure 3. Threshold value analysis – I - Effect of inebilizumab market share variation on the base case INHS saving due to inebilizumab availability (€8,373,125.13) (€2023).^a

Abbrevitations: INHS, Italian National Health Service; TVA, threshold value analysis. a-/+ variations in inebilizumab base case market share are counterbalanced by +/- variations in satralizumab and eculizumab base case market shares.

As shown by the second TVA, the overall saving for the INHS proves to be robust against extreme increases in the inebilizumab net price, as the 3-year saving for the INHS becomes < 0 only if inebilizumab net price increases by ≥ 50% (Figure 4).

Figure 4. Threshold value analysis – II - Effect of inebilizumab net price variation on the base case INHS saving due to inebilizum abavailability (€8,373,125.13) (€2023).

Abbreviations: INHS, Italian National Health Service; TVA, threshold value analysis.

4. Discussion

This research reports on a two-scenario BIM aimed at investigating the variations in INHS cost caused by the availability of inebilizumab as a new treatment for NMOSD in Italy.

Inebilizumab entry displaces mainly the use of other drugs also approved for the treatment of NMOSD, such as satralizumab and eculizumab [17,37,39,54,55], and also prevents a remarkable number of NMOSD attacks.

According to the baseline results and the findings of the comprehensive set of sensitivity analyses inebilizumab is expected to reduce by €1.14-€88.42 million the cost of managing NMOSD patients for the INHS, in addition to saving 4.80–24.50 NMOSD attacks.

No previous researchers investigated the impact of inebilizumab on the costs borne by third-party payers of other jurisdictions. In addition, BIMs concerning other drugs targeting NMOSD are poorly represented in the health economic literature.

According to the results of a Markov model-supported BIM performed in Thailand, when prescribed to 84 out of 280 NMOSD notional patients who were resistant to azathioprine, the 3-year BIM for mycophenolate mofetil and RTX biosimilar with CD27+ memory B cell monitoring ranged between 33,000–198,000 US dollars (USD), 2019 values [64].

However, the literature provides cost descriptions [62] on NMOSD that can be considered for comparing the results of the BIM detailed at the previous paragraphs.

A retrospective cost description that took into account 2014–2018 insurance claims concerning 162 US NMOSD patients retrieved from the IBM MarketScan US database, revealed that the cost for drugs was, on average, USD 6349 per year, or 10.12% of the mean overall cost per NMOSD patient [65], which is way lower than the 99.45% reported in our research.

Provided that comparison of costs totaled by patients with the same disease but managed in different healthcare systems brings about some methodological issues [66], it is also worth mentioning that these insurance claims date back to an era of cheaper drugs, when eculizumab, inebilizumab and satralizumab were not yet available for NMOSD AQP4+ patients [65].

A US cross-sectional survey that enrolled 193 NMOSD patients showed the cost for drugs (mainly RTX [60.6%], prednisone/corticosteroid [20.2%], mycophenolate mofetil [17.1%], and azathioprine [14.5%]) to sum up to USD262,598 (year unspecified), that is the 40.83% of the overall healthcare costs (USD643,103) [67]. Interestingly, and consistent with two of the assumptions made in our research, less than one half of the patients enrolled in the survey (82, or 42.4%) were examined every six months and those on RTX were assessed by physicians just before the drug infusion, with no need for further follow-up visits during the year.

A rapid review reported on the cost of managing a NMOSD AQP4+ patient during a 3-year timespan with eculizumab in US to be twice as costing as with inebilizumab (USD 2,100,000 vs USD 917,000) [68], but also highlighted that inebilizumab is 25–40 times more expensive than RTX. As far as the Italian setting is concerned, RTX is not a competitor of inebilizumab, as the place in therapy for inebilizumab is a second line treatment for NMOSD AQP4+ patients who switched from RTX for effectiveness and/or safety reasons [17].

The main limitation of our research rests on the recent entry of inebilizumab in the Italian market. Therefore, this 2-scenario BIM is basically a forecasting exercise that is assumed to be informative for the INHS decision makers as well as for the neurologists dealing with NMOSD patients in Italy. While it is apparent that the redistribution of the market shares across the two scenarios because of inebilizumab drives the saving for the INHS, it remains to be seen how the Italian market for NMOSD drugs will react to the availability of inebilizumab in the years to come.

In addition, future empirical economic evaluations performed alongside clinical trials comparing inebilizumab vs other NMOSD drugs in the real-world setting should focus on the difference, in terms of cost and saving for the INHS, between active and non-active phases of the disease [69], a topic that our research does not cover.

Besides, NMOSD attacks and NMOSD drugs-related AEs management was costed according the Italian DRG tariffs that reimburse hospitals via an all-inclusive perspective lump sum [46]. This approach is in line with the INHS perspective adopted in the BIM [31] and the research assumption that, being oftentimes severely ill [14,15], NMOSD patients are cared for in inpatient setting. However, the DRG scheme does not allow to single out the healthcare services that contribute to the production of a hospitalization episode [62], making any micro-costing exercise unfeasible and double-counting (that is, costing the same healthcare resource twice) prone. Hence, it remains to be seen whether the DRG tariffs-based valuation of healthcare resources is a fair and true proxy for the real cost borne by hospitals for treating NMOSD attacks and NMOSD drugs-related AEs [70]. That said, it would be interesting if future research following the hospital or the societal perspective [62] could derive an average empirical cost for managing NMOSD patients in Italy in inpatient and outpatient settings, as already reported for Germany [15,69].

Eventually, we assumed that the NMOSD notional patients included in the BIM were totally compliant to their therapies [49]. The plausibility of this research hypothesis in the real world setting and its bearing on the INHS saving due to the availability of inebilizumab on the Italian marker should be investigated in future clinical and health economic studies.

5. Conclusion

In conclusion, according to our base case assumptions the reimbursement of inebilizumab is expected to generate about €8 million saving for the INHS during a 3-year time horizon. While this should be seen as an outstanding achievement for INHS decision makers who manage the public expenditure for drugs, future complete economic evaluation of healthcare programs [62] should also explore the cost-effectiveness [62,71] of inebilizumab when compared to other drugs targeting NMOSD in Italy.

Declaration of interest

C Lazzaro has received an unconditional research grant from Horizon Therapeutics S.r.l., Milan, Italy. Outside this research, in the past three years C Lazzaro has received research grants, speaker or consultancy fees from AstraZeneca S.p.A, Ipsen S.p.A., Janssen-Cilag SPA, Roche S.p.A., Roche Diagnostics S.p.A., Sanofi s.r.l., Santen GmbH. N A Mazzanti and S Rossi are employees and equity owners of Horizon Therapeutics S.r.l., Milan, Italy. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Declaration of generative artificial intelligence in scientific writing

Authors declare that no section of this manuscript, including tables, figures and supplemental information was supported by Artificial Intelligence.

Author contributions

C Lazzaro: Conceptualization, Data curation, Formal analysis, Manuscript preparation, Manuscript review. N A Mazzanti: Conceptualization, Data curation, Manuscript review. S Rossi: Conceptualization, Data curation, Manuscript review. F Parazzini: Conceptualization, Expert suggestions, Manuscript review. All authors read and approved the final version of the manuscript for publication.

Reviewer disclosures

A reviewer on this manuscript has disclosed receiving consulting and board fees from Horizon therapeutics. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Complying with ethics of experimentation

Ethical approval was not needed.

Consent

Participatory consent was not needed.

SUPPLEMENTARY MATERIAL

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14737167.2023.2267176

Additional information

Funding

This research was funded by Horizon Therapeutics S.r.l., Milan, Italy.