Recently, patients with glycogen storage disease (GSD) have been described as having gut dysbiosis, lower fecal pH, and an imbalance in SCFAs due to an increase in acetate and propionate levels. Here, we report the fecal measurement of bacterial-related metabolites formic, acetic, lactic, propionic, and succinic acid, a key metabolite of both host and microbiota, on a previously described cohort of 24 patients (GSD Ia = 15, GSD Ib = 5, 1 GSD III = 1 and GSD IX = 3) and 16 healthy controls, with similar sex and age, using the high-performance liquid chromatography technique. The succinic acid levels were higher in the GSD patients than in the controls (patients = 38.02; controls = 27.53; p = 0.045), without differences between the groups for other metabolites. Fecal pH present inverse correlation with lactic acid (R = −0.54; p = 0.0085), while OTUs were inversely correlated with both lactic (R = −0.46; p = 0.026) and formic (R = −0.54; p = 0.026) acids. Using two distinct metrics of diversity, borderline significance was obtained for propionic acid, affecting the microbial structure on Euclidean basis in 8% (r² = 0.081; p = 0.079), and for lactic acid, affecting 6% of microbial structure using Bray–Curtis distance (r² = 0.065; p = 0.060). No correlation was found between SCFAs and total carbohydrate consumption among the participants or uncooked cornstarch consumption among the patients.

Currently, the diagnosis is usually provided by genetic analysis of the FBP1 gene, which is located on chromosome 9q22.32. The gene spans approximately 31 kb and includes seven exons, that are translated into an enzyme comprising 338 amino acids and are expressed in the liver, intestine and kidneys (^{El-Maghrabi et al., 1995}). The enzyme is modulated by cofactors (mono and divalent cations) and inhibitors (AMP and fructose-2,6-phosphate) that stabilize the conformation active or inactive of the protein, respectively (^{Shi et al., 2013}). The Human Gene Mutation Database (HGMD^®, http://www.hgmd.cf.ac.uk/ac/gene.php?gene=FBP1) (^{Stenson et al., 2017}) describes 59 different variants involved in the pathogenesis of this disorder, including the NM_000507.4:c.472C>T, NP_000498.2:p.(Arg158Trp); NM_000507.3:c.958G>A, NP_000498.2:p.(Gly320Arg) and NM_000507.3:c.986T>C, NP_000498.2:p.(Leu329Pro) variants detected in Brazilian patients (^{Pinheiro et al., 2019}). The variant p.(Arg158Trp) presents a frequency of 0.00003668 in gnomAD database (^{Karczewski et al., 2020}). However, other variants detected in Brazilian patients are absent of gnomAD (Karczewski et al., 2020) and AbraOm (^{Naslavsky et al., 2017}) databases.

This study sought to analyze whether an accurate diagnosis of the type and subtype of hepatic Glycogen Storage Diseases (GSDs) could be performed based on general clinical and biochemical aspects via comparing the proposed diagnostic hypotheses with the molecular results. Twelve physicians with experience in hepatic GSDs reviewed 45 real cases comprising a standardized summary of clinical and laboratory data. There was no relation between the hit rate and the time since graduation, the time of experience in GSD, and the number of patients treated during their careers. The average assertiveness was 47%, with GSD Ia and Ib being the best-identified types, while no expert correctly identified GSD IXc. Underage investigation for later manifestations, incomplete clinical description, and complementary analysis, the overvaluation of a specific clinical finding (“false positive”) or the discarding of the diagnosis in the absence of it (“false negative”), as well as the lack of knowledge of the rarest GSD types, may have impacted the accuracy of the assessment. This study emphasized that characteristics considered as determinants in identifying the specific types or subtypes of GSD are not exclusive, thus becoming factors that may have induced the evaluators to misdiagnose.

Massive sequencing platforms allow the identification of complex clinical phenotypes involving more than one autosomal recessive disorder. In this study, we report on an adult patient, born to a related couple (third degree cousins), referred for genetic evaluation due to ectopia lentis, deafness and previous diagnosis of juvenile idiopathic arthritis. He was biochemically diagnosed as having Classic Homocystinuria (HCU); Sanger sequencing of the CBS gene showed the genotype NM_000071.2(CBS):c.[833T>C];[833T>C], compatible with the diagnosis of pyridoxine-responsive HCU. As he also had symptoms not usually associated with HCU, exome sequencing was performed. In addition to the variants found in the Sanger sequencing, the following variants were identified: NM_001256317.1(TMPRSS3):c.[413C>A];[413C>A]; and the NM_005807.6(PRG4):c.[3756dup]:[3756dup], confirming the diagnosis of autosomal recessive nonsyndromic deafness and Camptodactyly-Arthropathy-Coxa Vara-Pericarditis Syndrome (CACP), respectively. Genomic analysis allowed the refinement of the diagnosis of a complex case and improvement of the patient’s treatment.