Among the 26 individuals included in the study (as shown in Table 1), 13 were male, and 13 were female. The range of age at initiation varied widely between 5 days and 9.50 years, with a central age of 4.09 years. Among all patients, only 2 (7.7%) were younger than 1. Sixteen (61.5%) patients were aged between 1 and 6 years, and 8 (30.8%) patients were aged between 6 and 10 years. The diagnostic age ranged from 1 to 12 years, with the middlemost age being 5.92 years. There were no patients younger than 1 year. Among the 26 patients, 16 (61.5%) were between the ages of 1 and 6, and the remaining 10 (38.5%) were 6–12 years old. Seven individuals (26.9% of the total) had a genetic background of elevated lipid levels in the blood.

Among the 23 individuals who showed symptoms, the majority (19, 82.6%) had xanthomas as the most prevalent manifestation. Among a total of 19 patients, 9 (47.4%) had xanthomas of the knee joint. There were 7 (38.9%) xanthomas in the elbow joint and buttock, 5 (27.8%) in the heel, and 4 (22.2%) each in the arm, leg and ankle. Less common locations included the popliteal fossa, back of the hand, and wrist, with 2 (11.1%) xanthomas each. Remarkably, one (5.6%) lesion was buried within the muscle tendon (Fig. 1A-H). Among the 19 individuals with xanthomas, 11 (57.9%) had fusion, and nearly half (9, 47.4%) had mysterious linear golden rashes. Xanthomas were observed in patients who were as young as 5 days and as mature as 9 years, with an average age of 3 years. Several other indications were observed, including joint pain in 7 patients (26.9%) and stunted growth in 4 patients (15.4%).

Images of xanthomas in different positions on the body of pediatric patients

Among the 26 patients, 57.7% (15 patients) had previously undergone testing for sitosterol and campesterol in their serum either before beginning medication or after starting food control. The serum sitosterol levels varied greatly, ranging from a minimum of 156.30 to a maximum of 4753.28 μmol/L; the average value was high at 764.98 ± 1193.

The concentration of campesterol exhibited a wide range, with values fluctuating between 47.52 and 963.17 μmol/L. The average value was an impressive 371.82 ± 328.88 μmol/L. prior to treatment, the recorded cholesterol levels varied greatly, with the lowest at 4.9 mmol/L and the highest at 27.15 mmol/L. The average level was a staggering 13.36 ± 6.03 mmol/L, showing a wide range. The LDL-C levels were equally variable, ranging from 2.81 to 12.08 mmol/L, with an average of 5.98 ± 2.53 mmol/L. Among the 12 individuals examined, there was a wide range of Lp(a) levels, from 7.02 to 1020.00 mmol/L. Interestingly, two of these patients (16.7%) had extremely high Lp(a) levels exceeding 300 mmol/L. Among the 18 patients, only 2 (11.1%) had hemoglobin levels less than 110 g/L (< 6 years), and 2 others (11.1%) had hemoglobin levels ranging from 110 to 120 g/L (for ages 6 and over), indicating the presence of anemia. A substantial majority of 14 individuals (77.8%) had normal hemoglobin levels. Before treatment, we examined the platelet levels of 18 patients. Only one patient (5.6%) had a lower than expected platelet count of 90 × 109/L, and the remaining 17 patients (94.4%) had a normal platelet count. The results for thyroid activity and liver function tests were within the optimal range.

Each patient received an abdominal ultrasound, as well as a cardiac and carotid artery ultrasound. In addition, fundus examinations were conducted on 14 patients. Two patients (7.7%) were diagnosed with aortic regurgitation, and the remaining 24 patients (92.3%) had no signs of abnormalities. Furthermore, there was thickening in the intima media of the bilateral carotid artery in 4 of the patients (15.4%), and in one patient (3.9%), there was thickening specifically in the right carotid artery. Notably, thickening of the carotid artery walls and plaque buildup in the unnamed right artery were observed in a 12-year-old patient (patient 3). None of the patients had splenomegaly or fundus vascular sclerosis.

Among the 24 individuals (92.3%) who underwent genetic testing, 19 (79.2%) were found to have compound heterozygous variants, and 5 (20.8%) had homozygous variants. According to the genetic report, a single individual possessed a perplexing combination of compound heterozygous variants (c.1189C > T (p.Gln397Ter277), c.1240C > G (p.Leu414Val) and c.63 + 217C > T) in the ABCG8 gene, along with a single heterozygous variant (c.904 + 5G > C) in the ABCG5 gene. This patient was ultimately categorized as having an ABCG8 variant. Among all 24 patients, the majority (15, 62.5%) had ABCG5 variants, and the remaining patients (9, 37.5%) had ABCG8 variants. Interestingly, among these patients, the highest proportion (7, 29.2%) had both nonsense and missense variants, and the rest had a mix of other variant types, such as missense variants (6, 25.0%), nonsense variants (3, 12.5%), splicing and missense variants (3, 12.5%), splicing and nonsense variants (1, 4.2%), splicing, nonsense and missense variants (1, 4.2%), missense and frameshift variants (1, 4.2%), missense and synonymous variants (1, 4.2%) and splicing variants (1, 4.2%). This demonstrates the complex and diverse nature of the genetic variations present in this group of patients, showing the highest level of intricacy.

Among the 15 patients with type 2 sitosterolemia and ABCG5 variants, the variant c.1166G > A (p.Arg389His) was the most common and was present in 10 patients (66.7%). The second most common variant was c.1336C > T (p. Arg446*), which was present in 7 patients (46.7%), followed by c.751C > T (p. Gln251*), which was identified in 4 patients (26.7%) (Fig. 2A). Among the 8 patients who were diagnosed with type 1 sitosterolemia, c.788G > A (p.Arg263Gln) and c.694 + 5G > C were the most commonly observed ABCG8 variants, with a total of 3 (37.5%) patients (Fig. 2B). A visual representation of all variant types can be found in Fig. 3.

Variants in patients with sitosterolemia A Different types of variants in ABCG5. B Different types of variants in ABCG8. C Variant ratios of nonsense variants, missense variants, splicing variants, frameshift variants and synonymous variants in ABCG5 and ABCG8 patients

Different variant locations in the ABCG5 (A) and ABCG8 (B) genes

Among the 15 individuals who were diagnosed with type 2 sitosterolemia and who carried ABCG5 variants, a large majority (11, 73.3%) had developed xanthomas, and a significant number (5, 33.3%) were affected by arthritis. Additionally, a minority (3, 20.00%) experienced growth restriction, and one individual (6.7%) suffered from thrombocytopenia. Furthermore, numerous patients (3, 20%) reported experiencing anemia, and a small percentage (2, 13.3%) displayed high levels of Lp(a). In addition, 2 individuals (13.3%) exhibited cardiac ultrasound abnormalities, and 5 individuals (33.3%) had thickening of the bilateral carotid arteries. Among the 8 patients who carried ABCG8 variations, xanthomas were present in the majority (6, 66.7%) of individuals, and arthritis, growth restriction, and local thickening of the bilateral carotid arteries were found in 2 (22.2%) patients each. The research did not present any information regarding conditions such as low platelet counts, low red blood cell counts, abnormally high levels of Lp(a) (over 300 mg/L) or anomalous results from cardiac echocardiography. Further analysis, as shown in Table 2, did not indicate any significant differences in these traits among patients with ABCG5 and ABCG8 variations. We examined the differences in variant types between the ABCG5 and ABCG8 genes (Fig. 2C). Nonsense variants were more prevalent in ABCG5 (P = 0.09) than in ABCG8, and splicing variants were more frequently found in ABCG8 (P = 0.01). There was no significant difference in the frequency of missense variants, frameshift variants or synonymous variants (P > 0.05).

Table 3 shows a comparison between the clinical information of patients who had the most common harmful variation, c.1166G > A (p.Arg389His), and that of patients who had different variations. An astonishing 80.0% of patients with that variant exhibited xanthomas. Additionally, 40.0% presented with arthritis, and 30.0% presented with growth restriction. Additionally, 10.0% of patients experienced thrombocytopenia, 20.0% had high Lp(a) levels, another 20.0% exhibited abnormalities in their cardiac ultrasound results, and a striking 30.0% showed thickening of both carotid arteries. Among the patients included in the study with other variant, 60.0% had xanthomas, 20.0% had arthritis, 20.0% had anemia, and 40.0% had thickening of the bilateral carotid arteries. The occurrence of growth restriction, thrombocytopenia, high Lp(a) levels, and abnormalities in cardiac ultrasound results were not documented. It is intriguing to note that individuals with the variant c.1166G > A (p.Arg389His) exhibited a greater likelihood of having increased levels of Lp(a), as well as experiencing xanthomas, joint pain, stunted growth, and elevated levels of sitosterol and campesterol. The disparities, although observable, did not show statistical significance (P > 0.05). Among all patients, only 3 harbored with the common pathogenic variant c.751C > T (p.Gln251*), and a slightly greater number of 4 patients harbored the variant c.788G > A (p.Arg263Gln). As a result, a single statistical comparison was not conducted. The variant c.1336C > T (p. Arg446*) was frequently observed together with c.1166G > A (p. Arg389His); therefore, it was not subjected to statistical comparison with any other variants.

Most patients (11, 42.3%) primarily received ezetimibe, either alone or in conjunction with other lipid-lowering medications, for treatment. Another group (13, 50.0%) focused solely on basic dietary management, and a select few used either simvastatin (1, 3.9%) or cholestyramine (1, 3.9%) alone. The levels of total cholesterol and LDL-C were assessed before and after treatment. For most patients, blood lipid levels can be controlled within 3 to 6 months. Following a change in diet or a combination of ezetimibe, the treatment led to a marked reduction in the levels of cholesterol and low-density lipoprotein compared to the levels seen before the treatment (P < 0.0001), as shown in Fig. 4.

Total cholesterol (TC) and low-density lipoprotein cholesterol (LDL-C) levels were measured before and after treatment