C4OH is a Potential Screening Marker - A Multicenter Retrospective Study of Patients with Beta-Ketothiolase Deciency in China

Background: Beta-ketothiolase deciency (BKTD) is an autosomal recessive disorder caused by biallelic mutations in ACAT1 that affects both isoleucine catabolism and ketolysis. Scant information is available regarding the incidence, newborn screening (NBS), and mutational spectrum in China. Methods: We collected NBS, biochemical, clinical, and ACAT1 mutation data from 18 provinces or municipalities in China between January 2009 and May 2020, and systematically assessed all available published Chinese BKTD patients data. Results: Totally 16,088,190 newborns were screened and 14 were identied through NBS, with an estimated incidence of 1 per 1 million newborns in China. Twenty-nine patients were genetically diagnosed as BKTD and 12 patients were newly identied. Most patients showed typical blood acylcarnitine and urinary organic acid proles. In particular, almost all patients (15/16, 94%) showed elevated C4OH levels. Eighteen patients presented acute metabolic decompensations and displayed variable clinical symptoms. The acute episodes of 9 patients were triggered by infections, diarrhea, and vaccination. About two thirds of patients have favorable outcomes, one showed developmental delay, while three had died. Twenty-seven distinct variants were identied in ACAT1, among which 5 were found to be novel. Conclusion: This study presented the largest series of BKTD cohort in China. Our results indicated that C4OH is a useful marker for the detection of BKTD. The performance of BKTD NBS could be improved by adding C4OH to the current panel of C5OH and C5:1 markers in NBS. The mutational spectrum and molecular proles of ACAT1 in Chinese population were expanded with 5 newly identied variants. are c.1119dup (p.V374Sfs*86), c.631C > A (p.Q211K), c.1154A > T (p.H385L), c.401T > C (p.M134T), and c.481T > C (p.Y161H). And all novel variants were predicted to be pathogenic by in silico analysis, expanding the molecular proles of ACAT1. C4OH to C5OH and C5:1 for the combination in NBS. The mutational spectrum of ACAT1 in Chinese population was established. Previously unreported ve novel variants were identied, expanding the molecular proles of ACAT1.

BKTD is included in newborn screening (NBS) programs in many countries, and both 3-hydroxyisovalerylcarnitine (C5OH) and tiglylcarnitine (C5:1) are primarily utilized as screening markers [4]. BKTD patients commonly have elevated levels of C5OH and C5:1. However, normal acylcarnitine pro les were reported in some patients even during acute metabolic crises [5,6]. Therefore, NBS for BKTD can be challenging as some patients may not be identi ed, indicating that only two markers C5OH and C5:1 are not enough for BKTD NBS.
Since the rst description of BKTD in 1971, approximately 250 patients have been detected worldwide [1]. While several retrospective studies investigating BTKD patients in various ethnic backgrounds have been reported [7][8][9][10], scant information is available regarding the incidence, NBS, and mutational spectrum in China.
We encountered a case of genetically diagnosed BKTD with increased C4OH only but no abnormal C5OH and C5:1 pro le at screening and no increased 2M3HB in urine even during acute metabolic disorder. To further evaluate the signi cance of C4OH in BKT, we launched a multicenter national cohort study on the platform of Zhejiang Neonatal Disease Screening Center, a unit of China Neonatal Screening Group. The speci c objectives of the study were: (a) to investigate the baseline level of amino acids and acylcarnitines in BKT; (b) to evaluate the importance of C4OH as well as C5OH and C5:1 in BKT screening; (c) to further understand the incidence, clinical features, genetic features and prognosis of BKT.
We systematically reviewed the available previous literatures of clinical reports of BKT in Chinese population and retrospectively analyzed the biochemical, clinical, and molecular features of total 29 Chinese BKTD patients from our NBS and selective metabolic screening (SMS). Notably, 3-hydroxybutyrylcarnitine (C4OH) was primarily used to evaluate the metabolic pro le of BKTD in previous studies, whereas elevated levels of C4OH can often be observed in our patients during NBS. Thus, we propose that C4OH is a potential marker for BKTD screening.
Meanwhile, we identi ed 5 novel variants among total 27 distinct variants of ACAT1 in Chinese population.

BKTD Special Case Report
A female newborn (patient #4 in Supplemental Table S1), now 23 months old, was found with increased C4OH (2.19 µmol/L, 4 times more than normal) only at NBS with tandem mass spectrometry (MS/MS). Her C4OH level returned to normal after 2 weeks when recalled and thus released with the regular process of NBS. She showed up at 11 months old with fever, severe acidosis, and drowsiness after injection of meningococcal vaccine and admitted to pediatric ICU. Her C4OH level at that time was 1.28 µmol/L (2 times more than normal) and only increased 3-hydroxybutyric acid (ketone body) was found with urine organic acid analysis. Genetic diagnosis revealed that her genotype ACAT1 variants are compound heterozygous c.163T > A (p. 555L) and c.1119dup (p.V374Sfs*86), respectively. At 1-year-old, she was hospitalized again due to fever, acidosis, somnolence, and metabolic disorder with increased C4OH level at 1.18 µmol/L. Although her condition became stable after emergency treatment, she had irreversible mental and motor retardation. In this case, regardless at the time of NBS or during acute metabolic crises, only increased C4OH but not characteristic index of BKT (C5OH and C5:1 or corresponding increase of 2M3HB in urine) was found. This special case trigged us to further evaluate whether C4OH is a potential marker to be used like C5OH and C5:1 for BKT screening.

Participating Study Centers
Participating centers from 18 provinces and municipality cities were selected nationally with the strict criteria of the China Neonatal Screening Group. The selected centers cover seven national administrative regions, including Northwest, Northeast, East, Middle, South, Southwest, and North China. Each center had more than 15,000 accumulated newborns screened by MS/MS. Internal quality controls were used in each sample batch. The results were evaluated by adding two in-house quality control specimens per 96-well microplate for the MS/MS analysis. External quality control programs from the China National Center for Clinical Laboratories (NCCL) and Centers for Disease Control and Prevention (CDC) were also included for yearly evaluations. This study was approved by the Ethical Committee of Children's Hospital, Zhejiang University School of Medicine (reference number: 2018-IRB-077) and was performed in accordance with the Declaration of Helsinki. Written informed consent was signed by and obtained from the parents of all infants.

Study Population
The study cohort covered NBS and reports between January 2009 and May 2020 (Fig. 1). Patients genetically diagnosed as BKTD (compound heterozygous or homozygous for ACAT1 variants) were included. All cases of genetically con rmed Chinese BKTD patients published previously were reviewed and included in the study.

Genetic Diagnosis and Data Analysis
The incidence was calculated by dividing the number of BKTD patients diagnosed via NBS by the total number of screened newborns. The patients' acylcarnitine pro les, clinical information, biochemical and genetic testing results were collected for analysis. For new identi ed patients, genetic testing was performed by the Genetic Diagnostic Laboratory at Children's Hospital, Zhejiang University School of Medicine (Hangzhou, Zhejiang, China). Target next-generation sequencing (NGS) was done as previously described [11], all potentially pathogenic variants identi ed through NGS were validated by Sanger sequencing. One hundred healthy newborns who were screened negative for BKTD were selected to assess variant frequencies in normal controls. The pathogenicity of novel variants was assessed using several in silico tools, including SIFT, PolyPhen-2, PROVEAN, and MutationTaster.

BKTD NBS and acylcarnitine analysis
During the study period, a total of 16,088,190 newborns were screened, six of the 18 provinces or municipalities have detected BKTD, and 14 patients eventually diagnosed with BKTD through NBS. The overall incidence of BKTD was 1 in 1,149,156 births (Table 1). In total, 29 Chinese patients were genetically diagnosed as BKTD in this cohort, 17 had been previously reported [12][13][14][15][16]. The median C4OH concentration was 1.38 ± 0.94 µmol/L (range 0.26-3.58 µmol/L, reference value 0.02-0.3 µmol/L), almost all patients (15/16, 94%) showed elevated C4OH levels except for one patient (Supplemental Table S1 showed elevated C5OH and C5:1 levels (Supplemental Table S1). Of particular notes, patients No. 4 and 11 had increased C4OH only but not C5OH and C5:1, even at the time of metabolic disorders. However, as described in the above case report, C4OH level in patient No. 4 returned to normal when recalled after 2 weeks of NBS, indicating that C4OH could be normal in the stable phase. In contrast, patient No. 10 had increased C5OH and C5:1 but not C4OH. All others have both increased C4OH as well as C5OH and C5:1 at the same time of NBS in the current available data. were not detected in the control group. All variants were not present or have extremely low allelic frequencies in the dbSNP, ExAC, 1000 Genome, and GnomeAD databases. In silico analysis suggested that all novel variants were potentially pathogenic (Supplemental Table S2). The most common variant in this cohort was c.622C > T (p.R208*) with a frequency of 17.2%, followed by c.1006-1G > C (8.6%) and c.1124A > G (p.N375S) (8.6%). In addition, c.419T > G (p.L140R) and c.997G > C (p.A333P) were relatively common ( Table 2).

Discussion
To our knowledge, so far this is the largest BKTD cohort studied in China. The large number of screened newborns in this study provided a more comprehensive perspective on the incidence of BKTD detected via NBS in China. Biochemical, clinical, and molecular features of Chinese BKTD patients were summarized, contributing to NBS, early diagnosis and timely treatment of this rare disease.
Few studies regarding the incidence of BKTD have been reported. The incidence is estimated at 1:190,000 in northern Vietnam [10], 1:313,000 in North Carolina and 1:232,000 in Minnesota in the United States [5,17]. Our study investigated for the rst time more than 16 million newborns with 14 identi ed BKTD patients for an incidence of approximately 1:1,000,000 in China. The incidence is lower than those reported in other studies [5,10,17], possibly due to differences in ethnic backgrounds, screening experience and awareness of the disease. It is noteworthy that the actual incidence of BKTD may be higher in China because some early neonatal deaths and patients who suffered mild episodes may not be identi ed.
BKTD is an ideal disease for NBS from a clinical perspective. However, false-negatives (FN) has been reported in several NBS programs [5,6,18]. Although both C5OH and C5:1 acylcarnitines are well-known markers for BKTD screening, these two markers are not necessary high during NBS and even at acute metabolic crisis. In our study, two patients (No. 4 and 11) had increased C4OH only but not C5OH and C5:1, even in the period of acute decompensation. However, we also found one patient (No. 10) showed increased levels of C5OH and C5:1 but no increase in C4OH. Notably, C4OH level in patient No. 4 was signi cantly high during NBS but returned to normal when recalled, then C4OH increased again at 11 months old after injection of meningococcal vaccine, indicating that C4OH is sometimes variable and could be normal in a stable condition. Beta-ketothiolase not only acts in ketone body utilization (ketolysis) by catalyzing thiolytic cleavage of acetoacetyl-coenzyme A to produce 2 molecules of acetyl CoA in extrahepatic tissues, but also catalyzes conversion of 2-methylacetoacetyl-coenzyme A in isoleucine catabolism. D-3-hydroxybutyrate ketone body can be converted into D-3-hydroxybutyrylcarnitine (C4OH) in vivo and in vitro [19]. It is therefore reasonable that there is not only an increase in C5OH and C5:1 but also an increase in C4OH in BKTD patients. Taking together, we speculate that some BKTD patients may be missed if either one of the three acylcarnitines (C5OH, C5:1 and C4OH) is used independently as a screening marker. Conversely, the FN results can be reduced by using several markers and/or their combinations. Our study therefore strongly suggests that C4OH is a very useful and powerful marker for the detection of BKTD. The performance of BKTD NBS may be improved by adding C4OH to C5OH and C5:1 combination in NBS.
In this study, despite almost all patients (except patient No. 4) exhibited the characteristic increase of urinary 2M3HB and TIG, both urinary markers were slightly elevated or undetectable even during acute crisis. Sometimes the analysis of urinary organic acid pro les is not typical, as observed in patient No. 4. Thus the diagnosis of BKTD should be con rmed by genetic test. In addition, elevated excretion of urinary 2M3HB and TIG indicates not only BKTD but also HSD10 mitochondrial disease (HSD10MD, OMIM #300438), which is a rare X-linked recessive disorder caused by a hemizygous or heterozygous mutation in the HSD17B10 gene [20]. Fukao et al. reported a 6year-old Japanese boy who was initially diagnosed with BKTD based on metabolic pro les, however, enzyme activity assays and mutation analysis nally con rmed the patient had HSD10MD [21]. Grunert et al. recently described two patients who may actually have HSD10MD but were misdiagnosed as BKTD in earlier reports [1]. Thus the diagnosis of BKTD cannot be based solely on metabolite data. Given the confusing blood acylcarnitine and urinary organic acid pro les between the two disorders, enzyme activity assays or mutation analysis are essential for differential diagnosis.
Most symptomatic patients in this cohort presented with acute metabolic decompensations or displayed neurologic impairment. Similar to previous study, neonatal presentation was very rare in this cohort and only two patients with neonatal onset [1]. About two thirds of patients have a favorable outcome. However, three clinical patients who did not undergo NBS presented with acute metabolic decompensations and died early, highlighting the importance of NBS for BKTD. NBS may be the only method for early detection of BKTD, and severe metabolic crises could be avoided if patients are properly managed. It is well known that acute episodes of most BKTD patients were associated with infections. Consistent with previous studies, six of our patients were triggered by respiration tract infections or diarrhea [4]. Notably, this study reported two patients developed severe metabolic crises triggered by vaccination and one had died. The patient who died experienced acute metabolic crises at 8 months that were triggered by febrile reaction to inactivated Japanese encephalitis vaccine, indicating that the risk of metabolic decompensation should be considered and special caution should be taken for BKTD patients before and after the injection of the vaccine.
At least 105 ACAT1 variants associated with BKTD have been described so far [22]. Most are private variants, only four variants have been identi ed in more than six families. The most frequent variant, c.622C > T (p.R208*), was found in 28 families and most of them are Vietnamese origin [22][23][24]. This variant was detected in 6 families and 10 individuals in our patient cohort, and counted for 17.2% of all the mutant alleles identi ed in Chinese patients. This is consistent with previous studies that it is the most common variant. The second most common variant was c.1006-1G > C, a splice site variant associated with exon 11 skipping, that was detected in 13 families and most of them are also Vietnamese [22,25].

Consent for publication
We con rm that the family has signed a written informed consent for publication of their own and their children's genetic data, clinical details, and/or any accompanying images.
Availability of data and materials The datasets used and/or analysed during the current study can be obtained from the corresponding author upon a reasonable request.

Con ict of Interest
The authors declare no con icts of interest.  concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.