Original Title: Efectividad clínica del cribado neonatal de errores congénitos del metabolismo mediante MS/MS. Actualización y análisis del estudio piloto: Enfermedad de la orina con olor a jarabe de arce; Acidemia isovalérica; Homocistinuria

Authors' objectives: Update the clinical effectiveness of neonatal screening by tandem mass spectrometry (MS/MS) of MSUD, IVA and HCY (CBS) and analysis of data collected in the pilot study

Authors' results and conclusions: The initial search resulted in 728 articles, 8 of which met the inclusion criteria, all of an observational and descriptive nature and referring to screening programs. Seven government screening programs were found through the general internet search, mostly European, 4 of which were included, making a total of 12 studies and programs included in this work. The quality of evidence was evaluated using QUADAS-2 tool, obtaining a low risk of bias and a high applicability. In order to estimate the detection rate in the pilot study, the longest time intervals reported by the Autonomous Regions have been taken into account, while to estimate the test performance, the time interval for carrying out the pilot study have been considered (2016 - 2017). Maple Syrup Urine Disease (MSUD): is a rare inherited disorder of branched-chain amino acids (BCAAs) metabolism characterized by a decreased activity of the BCKDC complex, involved in the degradation of BCAAs: leucine (Leu), isoleucine (Ile) and valine (Val). Consequently, the BCAAs (Leu, Ile and Val) and their intermediates accumulate in blood, urine and tissues, causing mainly neurological damage. Five phenotypes are known: classic or severe neonatal, intermediate, intermittent, thiamine-sensitive and dihydroprolil dehydrogenase deficient. The classical neonatal form is the most frequent and severe. The latency period is short, and can appear in the first few days of life of newborns (encephalopathy, vomiting, lethargy) which if untreated result in coma and death. Newborn screening is performed by MS/MS, measuring the BCAA profile Leu, Ile and Val in a dried blood spot. The current MS/MS technology is not able to differentiate Leu, Ile, aloisoleucine (Allo-Ile) and hydroxyproline (OHPro) because they have the same mass/load ratio, so the result is the sum of all of them (Xleu). To improve diagnostic differentiation, it is recommended to assess ratios based on Xleu and presence of Alo-Ile in a second tier test. For the screening programmes included in this review, 82 cases of MSUD were detected by MS/MS. In Spain, between 2012 and 2017, 3 asymptomatic cases were detected at diagnosis (two of them with classical phenotype) and 2 that appeared before screening (both with the classic phenotype). Sensitivity was close to 100% in two programs, and in three, it ranged between 66% and 90% due to false negatives and differences in screening protocols. Specificity was close to 100%, only reported by two programmes. These results come from a small number of studies representing the 34% of the population included in this review. The recalled rate for doubtful or positive specific results for MSUD was 0.025%, and in the pilot study it represented 0.071%. For the group of studies and programmes included, the false positives rate (FP) in any of the screening stages (FP1 + FP2) was 0.010%. In the pilot study, the FP rate was 0.079%, mostly on first sample (FP1) and the FP of the confirmation process (FP2) represented 0.002%. Potential sources of FP are increases in OHPro due to benign hydroxyprolinemia or Leu elevations from ketotic neonates affected by galactosemia and other liver diseases. In order to reduce the FP, it is recommended to assess ratios based on Xleu and Aile as a second tier test. PPV was very low (close to 7%) and heterogeneous between studies (data from the pilot study and 4 programmes). Based on the SICN definition, the PPV obtained in the pilot study was less than 1%. With regard to false negatives results, for the whole of the included programmes 4 FN were recorded, one of which could have been avoided by lowering the cut-off limit for Xleu (no information is available for the remainder). In the pilot study, 60% (3/5) of the cases detected by screening were asymptomatic at diagnosis (2 presented the classic phenotype) and 40% (2/5) debuted before the test results were obtained (both with the classic phenotype). Treatment started between 8 and 19 days of age, and allowed normal development and good health in all newborns. Regarding to morbidity, there is direct evidence of low-moderate quality, which seems to indicate that screening reduces neurological damage, especially intellectual development, and prevents the appearance of encephalopathy crises. No direct evidence that MSUD screening reduces mortality has been found. Isovaleric Acidemia (IVA): is characterized by a defect in the catabolism of the Leu, which means that when food containing Leu is ingested, there is an accumulation in the body of different compounds and derivatives that cause damage to the brain and nervous system. The clinical presentation is variable, from acute to chronic intermittent form, and can first appear at any moment between the first week of life and childhood. The acute neonatal or classic form is the most frequent, in which the RN appear in the first week of life with ketoacidosis and encephalopathy that can progress to coma and death. An average or benign phenotype is also known, in which they remain asymptomatic without treatment. Isovaleryl/2-methylbutyryl carnitine (C5) is used as the primary biomarker for screening the disease. Different C5-based fractions are commonly used as secondary biomarkers (C5/C0, C5/C2, C5/C3, C5/C4, C5/C8). One limitation of MS/MS technology is that it does not make it possible to differentiate isovalerylcharnitine, 2-methylbutyrilcarnitine (2-methylbutyric aciduria marker, 2MBCD), pivaloylcarnitine or valerylcarnitine, so it is necessary to make a differential diagnosis, especially of 2MBCD, the presence of pivaloylcarnitine, glutaric acidaemia type II, or ethyl-malonic encephalopathy. To improve the test performance, second tier tests have been developed, such as organic acids in urine (3,4-OH-isovaleric, isovaleryl glycine (IVG) or 2-methylbutylglycine). For the group of screening programs evaluated, 81 cases of IVA were detected by MS/MS, 9 of which were asymptomatic at diagnosis, and no deaths were recorded. In Spain, in the last 7 years (2012 - 2018), 6 cases were asymptomatic at diagnosis. Sensitivity was close to 100% as none registered FN, and specificity was also high and close to 100% (data from 34.4% of the total population screened). The overall recall rate was 0.023%, and in the pilot study it represented 0.12%. For the group of studies and programmes included, the FP rate in any of the screening stages was 0.018%, with high variability (0.0008% - 2.4%). In the pilot study it reached 0.120%, of which 96% were on first sample (FP1). The differences observed in the screening protocols could justify these differences, although the studies yielded contradictory results. Potential sources of FP include exposure to nipple care creams or antibiotics containing pivalic acid. In order to reduce FP, it is recommended to assess organic acids in urine, especially IVG as it is pathognomic. Based on the available data and for the group of studies, the PPV was 5.6% when classic neonatal phenotypes (22) and variants (10) were included in the estimation. If only the neonatal phenotype is considered, PPV would be reduced to 4% or 1.5%. In the pilot study, the PPV did not exceed 1%. There is uncertainty about the benefit of the screening obtained. There is no evidence that screening improves prognosis or clinical outcomes (cognitive, intellectual or developmental function), or that it reduces early mortality of severe neonatal forms. On the other hand, an overdiagnosis is suspected, since the phenotypic spectrum detected by screening includes a larger number of benign cases. Homocystinuria HCY (CBS): is the most frequent type of homocystinuria, and is due to a deficiency of cystathionine beta synthase (CβS). This is a multisystemic disease that mainly affects the ocular, skeletal, central nervous, and vascular systems. There are two phenotypes of the disease depending on the response to pyridoxine (vitamin B6): the non-sensitive or resistant (more severe) form, and the sensitive or dependent (mild) form. The condition has a long latency period, and the clinical manifestation becomes evident from the first year of life, with ectopia lens, myopia, osteoporosis, long extremities, alterations in the vascular system (thromboembolisms) and impairment of the central nervous system (intellectual disability). Its heterogeneity in the clinical spectrum (severe - asymptomatic), age of presentation (childhood - adolescence) and progression of the disease, means that diagnosis is made late, when the damage is irreversible. Screening is carried out with MS/MS by quantifying the Met concentration. However, the sensitivity of Met as a primary biomarker is limited and highly variable, depending on the cut-off value. To improve the test performance it is recommended to use Met/Phe as a secondary biomarker, and total homocysteine (tHcy) as a second tier. In patients sensitive to B6, the Met concentration at the time of screening is usually not high enough, and despite lowering the cut-off value, with the current methodology, this phenotype is not frequently detected, so patients who are not sensitive or resistant to B6 would benefit the most in a screening program. In the total number of programmes, 17 cases of MSUD were detected by MS/MS, of which the phenotype of 41% (7/17) is known, which was not sensitive or resistant to B6 in any of them. In Spain, seven cases of HCY (CBS) were detected in the last seven years, and the phenotype and clinical state at the time of diagnosis were unknown. The specific recall rate was 0.097% (data from the pilot study). The FP rate was 0.0184% with major variations, which increases the imprecision of the PPV, which was close to 2%. In the pilot study, the FP rate was 0.095%, 94% of the first sample. Potential sources of FP include liver disease caused by galactosemia or tyrosinemia I, parenteral nutrition, prematurity at birth or MAT deficiency. In addition, tHcy increases can be observed in MTHRF deficiency, cobalamin metabolism defects, and maternal vitamin B12 deficiency. Early diagnosis and treatment in the asymptomatic phase prevents associated morbidity, especially ocular and vascular sequelae, but does not reverse the damage that has already been caused. The evidence does not demonstrate that screening provides a benefit in terms of reduced mortality. Conclusions: • The evidence on the effectiveness of the MSUD, IVA and HCY (CBS) neonatal screening programmes evaluated in this review is of moderate-low quality and is based on observational studies, and at times on pilot programmes. • In MSUD screening there is some variability in the biomarkers used (primary and secondary), the cut-off values, and the implementation of Aile as a second tier, which has a significant impact on the test performance. The FP rate seems acceptable, but it reduces the PPV to 7% and less than 1% in the pilot programme. With respect to meeting the requirements for implementation in a screening programme, MSUD does not have a sufficiently long latency period in more than 80% of the detected cases for the programme to achieve the expected benefit. The test performance is low, sensitivity is highly variable and dependent on the protocol used, PPV is very low, and FN are not known exactly. Neither has direct evidence been found of the impact of screening on mortality. • The uniformity that exists in the screening algorithms for isovaleric acidaemia, especially in secondary biomarkers and second tier testing, limits test performance. The FP rate is acceptable but very heterogeneous between the studies, which increases the imprecision in PPV, also influenced by the phenotype considered in its estimation, ranging from 5.6% (classics and variants) and 4% - 1.5% (classics). IVA would not comply with some of the implantation requirements, such as the existence of a sufficient latency period or a clear benefit in improving prognosis or reducing early mortality. On the other hand, a possible overdiagnosis of the mild forms is suspected with screening, and there are uncertainties about the performance of the test. • In HCY (CBS) screening, there is significant heterogeneity in the screening algorithms, especially in the use of secondary biomarkers (Met/Phe) and tHcy as a second tier. The sensitivity of Met is limited and highly variable depending on the cut-off value. One limitation of the method is the difficulty of detecting the phenotype sensitive to B6. HCY (CBS) has a sufficient latency period and a treatment which, when administered early in the latency phase, reduces the risk of thromboembolic, ocular, and intellectual complications. However, the test performance is low and shows no benefit in terms of reduced mortality. • Agreement must be reached on an adequate and specific screening protocol for each condition, which optimises the test performance and improves the comparability of the results, as well as standardising the registry of the health status of the cases identified at the moment of diagnosis and the phenotype. It is recommendable to define indicators for health results, and to define optimum and acceptable levels of some process indicators, as well as the FP rate or age at the start of treatment. All of this information will help to measure the degree to which the objectives have been achieved, in the decision-making process.

Authors' methods: Systematic review of the literature in the main biomedical databases: Medline, Embase, Cochrane Library, HTA, DARE, and INAHTA, among others. The starting point was an HTA report carried out by Avalia-t in 2012. The different search strategies were updated taking into account the time limit used (2012-present) and carrying out periodic updates. The process was completed with a manual review of the bibliography of the articles included, as well as a general internet search on official pages of screening programmes, organisations and/or scientific societies. Two reviewers working independently and in duplicate reviewed all the abstracts and selected full-text articles according to previously established inclusion criteria. Disagreements were resolving by consensus. The pilot study information was provided by the Screening Programmes Unit of the Ministry of Health, through data provided by the Autonomous Regions to the Neonatal Screening Programme Information System (SICN)

Project Status: Completed

URL for project: https://avalia-t.sergas.gal/Paxinas/web.aspx?tipo=paxtxt&idLista=4&idContido=840&migtab=840

Year Published: 2020

URL for published report: https://avalia-t.sergas.gal/DXerais/840/avalia-t201803CribadoEndocrinoMetabol.pdf

English language abstract: An English language summary is available

Publication Type: Mini HTA

Country: Spain

Organisation Name: Scientific Advice Unit, avalia-t; The Galician Health Knowledge Agency (ACIS)

Contact Address: Conselleria de Sanidade, Xunta de Galicia, San Lazaro s/n 15781 Santiago de Compostela, Spain. Tel: 34 981 541831; Fax: 34 981 542854;

Contact Name: avalia-t@sergas.es

Contact Email: avalia-t@sergas.es

Copyright: <p>Galician Agency for Health Technology Assessment (AVALIA-T)</p>