Full title: Cost-effectiveness of neonatal screening for spinal muscular atrophy and severe combined immunodeficiency

Authors: García Pérez L, Guirado Fuentes C, Velázquez Perelló C, Valcárcel Nazco C, Herrera Ramos E, Cazaña Pérez V

Contact person: Lidia García Pérez (lidia.garciaperez@sescs.es)

ABSTRACT

Introduction

Spinal Muscular Atrophy

Spinal muscular atrophy (SMA) [ICD-11] is a rare genetic neuromuscular disorder characterized by the irreversible loss or degeneration of lower motor neurons in the spinal cord. This disease leads to muscle weakness and muscle mass loss and, in more severe forms, can affect the patient’s ability to feed and breathe.

SMA is caused by pathogenic variants in genes that encode the survival motor neuron (SMN) protein. While most functional SMN is produced through the SMN1 gene, several different forms of the SMN protein are also produced by the SMN2 gene. The number of SMN2 gene copies is generally associated with disease severity; typically, a higher number of copies correlates with milder forms of the disease. Clinically, SMA is classified into five distinct subtypes (Type 0 to Type IV) based on age of onset, clinical severity, and symptom appearance (the most severe forms usually present in early infancy, while milder ones appear in adulthood).

SMA is considered the leading genetic cause of infant mortality. Its global incidence is estimated, with some uncertainty, at 1:12,500 live births. In Spain, it is estimated that between 1:6,000 and 1:11,000 newborns will present with SMA, corresponding to approximately 33 to 60 new cases per year, with an average of 40 new cases annually.

In the absence of newborn screening, clinical suspicion of SMA is based on symptom presentation (such as hypotonia and progressive muscle atrophy). Diagnosis is confirmed through quantitative genetic testing of the SMN1 gene. Once SMA is confirmed, quantification of SMN2 copy number provides prognostic information and supports therapeutic decision-making.

To date, the European Medicines Agency (EMA) has approved three drugs for the treatment of SMA: nusinersen (Spinraza®), onasemnogene abeparvovec (Zolgensma®), and risdiplam (Evrysdi®). These drugs differ in their mechanisms of action, treatment duration, and administration protocols. Nusinersen (administered by lumbar puncture) and risdiplam (oral solution) can be administered indefinitely as long as the patient continues to benefit from the treatment. Onasemnogene abeparvovec (OA) is a gene therapy delivered as a single intravenous infusion, designed to replace the deletion or dysfunction of the SMN1 gene.

Severe Combined Immunodeficiency

Severe combined immunodeficiency (SCID) [ICD-10: D81.0, D81.1, D81.2] encompasses a group of rare, inherited, monogenic diseases characterized by a profound reduction in functional peripheral T lymphocytes. SCID is typically asymptomatic at birth, with clinical symptoms usually appearing within the first few months of life. Patients have increased susceptibility to opportunistic infections from bacteria, fungi, viruses, and protozoa.

The global incidence is estimated at around 1:50,000–1:60,000 live births, with regional differences and higher incidence among populations with high rates of consanguinity. ADA-SCID, caused by adenosine deaminase (ADA) deficiency, is one of the most prevalent variants. In Spain, the exact incidence is unknown; however, an unpublished retrospective study in Catalonia estimated it at 1:57,000 and another in Andalusia at 1:34,000. More recent unpublished data from Catalonia suggest an incidence of 1:60,000.

Diagnosis includes, among other tests, flow cytometry and genetic testing. Treatment consists of continuous antimicrobial prophylaxis and immunoglobulin replacement therapy, among others. Early detection is important to avoid exposure to viruses and live attenuated vaccines. Definitive cure requires allogeneic hematopoietic stem cell transplantation (HSCT). Gene therapy has proven effective, particularly in patients with ADA deficiency.

Without treatment, SCID typically leads to severe infections, autoimmune phenomena, and death, usually within the first year of life. Infection-free transplanted patients have a 95% survival rate. The highest survival is observed in patients transplanted at 3.5 months of age or younger.

Newborn screening for spinal muscular atrophy and severe combined immunodeficiency

Currently, a kit is available that enables the simultaneous newborn screening of both SMA and SCID. It uses quantitative polymerase chain reaction (qPCR) technology to analyze DNA from dried blood spot samples routinely collected at birth. The aim of SMA screening is the detection of homozygous deletion of the SMN1 gene, followed by quantification of SMN2 gene copies. For SCID, screening involves quantifying T-cell receptor excision circles (TREC). Other pilot screenings are quantifying T-cell lymphopenia through TREC and B-cell lymphopenia through kappa-deleting recombination excision circles (KREC).

Currently, both SMA and SCID are included in the screening programs of the Canary Islands, Galicia, and Madrid. Other regions have included one or the other disease in their screening programs.

Objectives

Main objective:

• To assess the efficiency of population-based newborn screening for SMA and SCID jointly.

Secondary objectives:

• To analyze the cost-effectiveness of the SMA newborn screening program based on published studies.

• To analyze the cost-effectiveness of joint newborn screening for SMA and SCID based on published studies.

• To analyze the cost-effectiveness of joint newborn screening for SMA and SCID from the perspective of the Spanish National Health System (NHS).

• To estimate the budget impact of jointly incorporating these two diseases into the newborn screening program

Methodology

Systematic Review of Cost-Effectiveness Studies

The systematic review of cost-effectiveness studies was based on the adaptation and update of the systematic review carried out by the Health Technology Assessment Agency of Ireland (HIQA) on neonatal screening for SMA. In January 2023, they conducted a search in MEDLINE, Embase, and the Cochrane Library, supplemented by a gray literature search. Our update of the search was conducted between January 2023 and July 2024 using the same search strategy but adding the Cost-Effectiveness Analysis Registry database. Two reviewers independently selected the studies. Full economic evaluations were selected in which population screening for SMA or the combined SMA and CIDG screening were compared to no screening in neonates. To be included, outcome measures had to be the incremental cost-effectiveness ratio (ICER) or the costs and effectiveness of the alternatives in comparison. The methodological quality of the included studies was assessed using the CHEC questionnaire. To analyze the transferability of the only identified economic evaluation of the combined SMA and CIDG screening, the Welte et al. tool was used, and its methodological quality was assessed using the 10 criteria from Drummond et al.

Economic Evaluation

A full economic evaluation was carried out in which the costs and health outcomes of the combined neonatal screening for SMA and CIDG were compared to no screening. This was done by merging the results of two separate decision models for each disease, based on common characteristics defined for both models. For the modeling of neonatal screening for SMA, a new model consisting of a decision tree and a Markov model was designed. In the absence of a single base case, different scenarios were analyzed based on different treatment strategies and the duration of the treatment effects. For CIDG, the results of a previous model for Spain, carried out by SESCS-RedETS, were updated, which was in turn an adaptation of the model provided by Dr. Chilcott from the University of Sheffield. The study population corresponds to all neonates born in Spain in one year. The primary time horizon, common to both diseases, was the entire patient lifetime (maximum 100 years), but other time horizons (2, 5, and 60 years) were also tested. A 3% discount rate was applied to both costs and effects. The analysis was performed from the perspective of the Spanish National Health System (SNS), where only direct healthcare costs were included, expressed in 2022 euros. The outcome measures were life years gained (LYG) and quality-adjusted life years (QALYs). The results were combined using the ICER, which was compared to the cost-effectiveness threshold for Spain estimated by Vallejo et al. between 20,000 and 25,000 €/QALY. For the combined SMA and CIDG screening analysis, the results were merged in absolute values (costs and life years gained, adjusted or not by quality of life), not average values, and the common screening costs were considered to avoid cost duplication. Deterministic sensitivity analyses were performed on the results of the most favorable scenario for neonatal SMA screening. Additionally, a 5-year budget impact analysis was conducted, informing the cost for SNS of incorporating SMA and CIDG into the national neonatal screening program in Spain.

Results

Systematic Review of Cost-Effectiveness Studies

The search identified and included seven economic evaluations addressing neonatal SMA screening and one addressing the combined SMA and CIDG neonatal screening. All included studies were model-based economic evaluations, with three funded by disease-modifying treatment manufacturers. The healthcare system perspective was the most common, although two studies reported results from a societal perspective.

The studies addressing neonatal SMA screening generally showed heterogeneity in terms of the SMA genotype in the screened cohort, the health states considered, and the treatments assumed for the disease. In this sense, the treatment strategy had a considerable influence on the ICER results. The identified studies also had limitations regarding the long-term treatment efficacy evidence.

Among the studies reporting results in terms of cost per QALY gained, three European studies funded by treatment manufacturers reported dominance (cost savings and greater effect) of screening compared to no screening. Another European study reported an ICER below the threshold or dominance in favor of screening for the base case and most scenarios analyzed, except one (equal treatment distribution across alternatives and healthcare system perspective). A study conducted in the US, which only treated patients with nusinersen, reported very high ICERs, above 150,000 €/QALY. A study conducted in Australia reported ICERs ranging from cost-ineffective to cost-saving depending on the treatment strategy. When comparing screening versus no screening, treating both with nusinersen, the ICER was over 350,000 €/QALY. When comparing screening and treating with OA versus no screening and treating with nusinersen, cost savings were reported.

The only study identified that addressed the combined neonatal screening for SMA and CIDG was of acceptable quality and conducted in Australia from the payer perspective (Australian government), with a time horizon of 5 and 60 years. The SMA model part was based on a previously published model by the authors, which is included in this systematic review of economic evaluations. The main limitation, acknowledged by the authors in the discussion, is the lack of scientific evidence on the long-term effect of SMA treatments, assuming that the short-term effects remain in the long run.

For a 5-year time horizon, the result was cost-ineffective with an ICER (expressed in 2022 euros) of 338,800 €/QALY, but with high uncertainty reflected in a confidence interval ranging from dominance to a very high ICER of 651,342 €/QALY. For a 60-year time horizon, screening was dominant, with a confidence interval between dominance and 31,967 €/QALY. The parameters most affecting the results were the incidence of the two diseases and the cost of SMA treatment.

The transferability analysis conducted for this Australian study indicates that the conclusions of its economic evaluation cannot be considered directly transferable to the Spanish context. This is due to the uncertainty related to one of the key factors, the costs and relative prices presented in the model compared to those in Spain. In particular, the costs by type of SMA and the health states associated with CIDG, as included in the model, are not available for Spain, which would require using foreign data and incorporating uncertainty into the resulting cost-effectiveness outcomes.

Cost-Effectiveness and Budget Impact Analysis

Both the results of the economic evaluation for neonatal SMA screening and CIDG (separately) show that screening is more costly and more effective than no screening, in terms of LYG and QALYs. The estimated ICERs are 50,540 €/QALY for SMA and 29,210 €/QALY for CIDG, both above the cost-effectiveness threshold of 25,000 €/QALY. The results of the deterministic sensitivity analysis suggest that if a discount rate for effects of 1.5% is assumed, as recommended by some authors/institutions for discounting future benefits in public health interventions, the ICER comparing screening for CIDG versus no screening would be below the willingness-to-pay threshold of 25,000 €/QALY.

The ICER for the combined neonatal screening for SMA and CIDG remains above the threshold with a value of 42,938 €/QALY. It is worth noting that the result presented for SMA, used in the combined screening analysis, corresponds to the ‘reference case,’ which is the most favorable to screening among the realistic scenarios analyzed for this disease.

The sensitivity analysis shows that the cost of the screening kit has little effect on the ICER, with the cost of SMA treatments having a greater impact, although most results remain above the cost-effectiveness threshold.

Among the main limitations of the cost-effectiveness analysis conducted are the various possible treatment strategies with some of the disease-modifying drugs approved for SMA, the use of clinical effectiveness measures based on different outcome measures reported in trials for each SMA treatment, uncertainty regarding the long-term effect of SMA treatments, uncertainty regarding the costs associated with both diseases, and the different structure of the SMA and CIDG models, which could affect the combined results.

The budget impact analysis, which shares some of the limitations of the cost-effectiveness analysis, shows that the net cost per newborn after 5 years is estimated at 17.33 €, equivalent to 28.5 million euros for the total number of neonates born in Spain during that period.

1. Complete report: