Assessment of erythrocyte transketolase, whole blood thiamine diphosphate, and human milk thiamine concentrations to identify infants and young children responding favorably to therapeutic thiamine administration

Authors: HESS, SONJA - University Of California, Davis; SMITH, TARYN - University Of California, Davis; ARNOLD, CHARLES - University Of California, Davis; JONES, KERRY - University Of Cambridge; HAMPEL, DANIELA - University Of California, Davis; HIFFLER, LAURENT - Non ARS Employee; TREHAN, INDI - University Of Washington; FISCHER, PHILIP - Mayo Clinic; MEADOWS, SARAH - University Of Cambridge; PARKINGTON, DAMON - University Of Cambridge; BROWN, KENNETH - University Of California, Davis; SITTHIDETH, DALAPHONE - Lao Tropical And Public Health Institute(LAO TPHI); TAN, XIUPING - University Of California, Davis; KOULMAN, ALBERT - University Of Cambridge; Allen, Lindsay; KOUNNAVONG, SENGCHANH - Lao Tropical And Public Health Institute(LAO TPHI)

Submitted to: Current Developments in Nutrition
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/18/2024
Publication Date: 5/23/2024
Citation: Hess, S.Y., Smith, T.J., Arnold, C.D., Jones, K.S., Hampel, D., Hiffler, L., Trehan, I., Fischer, P.R., Meadows, S.R., Parkington, D., Brown, K., Sitthideth, D., Tan, X., Koulman, A., Allen, L.H., Kounnavong, S. 2024. Assessment of erythrocyte transketolase, whole blood thiamine diphosphate, and human milk thiamine concentrations to identify infants and young children responding favorably to therapeutic thiamine administration: Findings from the Lao Thiamine Study, a prospective cohort study. Current Developments in Nutrition. 8(6). Article 103786. https://doi.org/10.1016/j.cdnut.2024.103786.
DOI: https://doi.org/10.1016/j.cdnut.2024.103786

Interpretive Summary: Little is known about the relationships between thiamine status biomarker [whole blood thiamine diphosphate (ThDP), erythrocyte transketolase activation coefficient (ETKac) and human milk thiamine concentration (MTh)] and thiamine deficiency. Here, we explored correlations between these biomarkers and thiamine responsive disorder (TRD) status, an indicator of clinical response to thiamine administration among young children, and to define a population-specific cut-off of ThDP corresponding to ETKac =1.25 for young children and their mothers. Hospitalized children with symptoms suggestive of thiamine deficiency disorders were treated with parenteral thiamine (100mg daily) for =3 days alongside other treatments and examined in regular intervals. Case reports were reviewed by three pediatricians, who assigned a TRD score (TRD or non-TRD) based on risk factors of thiamine deficiency and response to thiamine. For a community-based comparison group, children were frequency-matched by age, sex and residence. Their mothers were eligible for participation. Maternal and infant venous whole blood thiamine diphosphate (ThDP) was measured as well as the erythrocyte transketolase activation coefficient (ETKac). We found that 55.6% of all children were diagnosed with TRD. Median (IQR) ThDP and ETKac were 66.9 (41.4, 96.9) nmol/L and 1.25 (1.11, 1.48) for the 287 hospital children, and 64.1 (50.0, 85.3) nmol/L and 1.22 (1.12, 1.37) for the community children. ThDP and ETKac were 70.2 (55.1, 88.1) nmol/L and 1.29 (1.18, 1.44) among mothers of hospitalized children, and 73.6 (54.6, 94.6) nmol/L and 1.23 (1.15, 1.30) among community mothers. 89% of breastfeeding mothers of infants <6 months of age had MTh <200 ug/L. ThDP and ETKac were well correlated in hospital (r=-0.84) and community children (r=-0.64), and less strongly in their mothers (hospital: r=-0.51; community: r=-0.19). Among all children, ThDP was strongly predictive of an ETKac cutoff of >1.25 with an AUROC of 0.86; a ThDP cutoff of 64 nmol/L had a sensitivity of 0.79 and specificity of 0.80. Among all women, ThDP predicted the ETKac cutoff of >1.25 with an AUROC of 0.66 with a proposed ThDP cutoff of 74 nmol/L (sensitivity 0.68; specificity 0.61). Our results showed that none of the thiamine biomarkers were clinically useful to identify children with TRD. Whole blood ThDP was predictive of ETKac in young children and their mothers.

Technical Abstract: Background: There is limited information on the relationships between selected biomarkers of thiamine status (whole blood thiamine diphosphate (ThDP), erythrocyte transketolase activation coefficient (ETKac) and human milk thiamine concentration (MTh) and clinical manifestations of thiamine deficiency. The present study aimed to explore correlations between biomarkers of thiamine status and thiamine responsive disorder (TRD) status, a composite indicator of favorable clinical response to thiamine administration among young hospitalized children and to define a population-specific cut-off of ThDP corresponding to ETKac =1.25, separately for young children and their mothers. Methods and Findings: Hospitalized children (aged 21 days – <18 months) were eligible, if they presented with symptoms suggestive of thiamine deficiency disorders. Children were treated with parenteral thiamine (100mg daily) for =3 days alongside other treatments and examined in regular intervals. Per child individual case reports were compiled for review by three pediatricians, who assigned a TRD score (TRD or non-TRD) based on risk factors of thiamine deficiency and response to thiamine administration. For a community-based comparison group, children were frequency-matched by age, sex and residence. Mothers of enrolled children in hospital and community cohort were eligible for participation. From children and women, venous whole blood thiamine diphosphate (ThDP) was determined by HPLC, and erythrocyte transketolase activation coefficient (ETKac) was assessed by UV spectrophotometer in washed erythrocytes. Associations between biomarkers were assessed using Spearman correlations, and biomarker cutoffs corresponding with TRD and ETKac >1.25 were explored using area under the receiver operating characteristic curve (AUROC), with optimal cut-offs selected via the closest-to-(0,1) corner cut-point approach. Thiamine biomarkers were available for 515 children (mean±SD age 4.7±3.5 months; 59.4% male) and 667 women (24.8±6.3 years). Among the hospital children, 55.6% were diagnosed with TRD. Median (IQR) ThDP and ETKac were 66.9 (41.4, 96.9) nmol/L and 1.25 (1.11, 1.48), respectively, among the 287 hospital children, and 64.1 (50.0, 85.3) nmol/L and 1.22 (1.12, 1.37) among community children. ThDP and ETKac were 70.2 (55.1, 88.1) nmol/L and 1.29 (1.18, 1.44) among mothers of hospitalized children, and 73.6 (54.6, 94.6) nmol/L and 1.23 (1.15, 1.30) among community mothers. 89% of breastfeeding mothers of infants <6 months of age had MTh <200 ug/L. Although ThDP and ETKac were significantly different between TRD and non-TRD, they did not have clinical usefulness due to the wide overlapping range in concentration/ratio. ThDP and ETKac were well correlated in hospital (r=-0.84) and community children (r=-0.64), and less strongly in their mothers in the hospital (r=-0.51) and community (r=-0.19) cohorts. Among all children, ThDP was strongly predictive of an ETKac cutoff of >1.25 with an AUROC of 0.86; a ThDP cutoff of 64 nmol/L had a sensitivity of 0.79 and specificity of 0.80. Among all women, ThDP predicted the ETKac cutoff of >1.25 with an AUROC of 0.66 with a proposed ThDP cutoff of 74 nmol/L (sensitivity 0.68; specificity 0.61). Conclusions: None of the thiamine biomarkers were clinically useful to identify children with TRD. Whole blood ThDP was predictive of ETKac in young children and their mothers. Assessment of ThDP by HPLC is simpler than measuring ETKac, so may be more useful for population assessment.