Validation and Simplification of the Retinol Isotope Dilution Technique in Burkinabe Schoolchildren

Project title Validation and simplification of the retinol isotope dilution technique in Burkinabe schoolchildren

I. Background and rational Biomarkers of vitamin A status that reflect the gold standard, i.e., liver biopsy, are available but undergoing refinement to increase accessibility in population-based applications. Retinol isotope dilution (RID) testing is one such biomarker. In Burkina Faso, we were able to evaluate a 7-d test with a 14-d test with the simple addition of a correction factor to the prediction equation. The values for total body vitamin A stores were similar, but variation increased with increasing liver reserves. The assumption of 3% of body weight as liver weight in school-aged children was confirmed as appropriate in the calculation for total liver vitamin A reserves. Simplifying isotope dilution for population evaluation and building capacity for mass spectrometry analyses are important areas of nutrition development to inform public health programs. This proposal will continue to refine the method for widespread use.

II. The Overall Objective To provide new knowledge on how to use the retinol isotope dilution technique in larger population surveys to assess vitamin A status from deficiency to excess.

II.1. Specifics objectives Several specific objectives have been outlined for this CRP. Through the research that will be conducted in Burkina Faso and on existing data sets, we will focus on the following specific objectives in schoolchildren.

1. To define a vitamin A excess total liver reserve value by applying fasting serum retinyl ester data in schoolchildren with RID test 2. To examine how the existing methodology can be simplified to suit population surveys without the loss of accuracy, using onsite retinol purification techniques 3. To explore how the retinol isotope dilution technique is affected by inflammation at the time of dosing using an existing data set in African preschool children 4. To examine the difference in absorption of 3,4-didehydroretinyl acetate from oil and in encapsulated form in school-age children.

III. Method III.1. Study site and population The proposed study will be carried out in a primary school in the Western part of Burkina Faso, in Orodara, a village at the outer edge of Bobo Dioulasso, where we will target children aged 7 to 12 years. Most inhabitants of the village are middle-income farmers and their families. Burkina Faso is one of the countries involved in the consortium VITAAFRO (Vitamin A in Africa) that is interested in accurately determining VA status to inform existing interventions and policies.

III.2. Study design Two successive cross-sectional studies will be carried out during a school year. The first study at the beginning of the school year, which happen to be the end of the raining season, and which is also out of the mangoes season and the second study, 5 to 6 months later, after the first set of data collection, during the mangoes season these studies will be carried out in a primary school in the Western part of Burkina Faso, in Orodara, a village at the outer edge of Bobo Dioulasso.

III.3. Sampling method In total 158 school children will be recruited in two different ways. At first time, a group of 60 school children will be recruited through a stratified sampling including children from the first to the last primary school year and proportionally to their number per class. To this first group of 60 school children a retinol isotope dilution (RID) test for total body store assessment will be applied to compare the values to fasting serum retinyl ester concentrations at baseline.

The remaining 98 school children will be recruited according to stratified sampling and proportionally to their number per class and divided into two groups of 49 children according to the sex (i.e. 49 females and 49 males). In each participant, two timepoints of blood sampling will be performed. Participant will be will then be randomly assigned to have blood sampling at H0, H2, H3, H4, H5 H6 H8, H10, H24, H72, H168 and H336 as specified below. A randomization list will be generated using tables of random numbers. We have chosen to do block rather than simple randomization because the former approach will allow us to have unbiased comparison groups that will be roughly the same size throughout the trial. By using block randomization, there will be always nearly equal numbers of children in each study group.

To these groups of children, the modified relative dose response (MRDR) test, a qualitative measure of vitamin A status will be used to investigate the absorption of 3,4-didehydroretinyl acetate.

H0 (5), H2 (5), H3 (5), H4 (11), H5 (11) H6 (11) H8 (10), H10 (10) , H24 (5), H72 (5), H168 (10), H336 (10)

III.4. Community awareness and sensitization Before study commencement, we will conduct a series of meetings with key decision makers and community leaders in the study area as part of efforts to raise awareness and mobilize the community for the study. These community outreach and sensitization meetings will inform the community about the existence and purpose of the study as well as the benefits not only to their communities but also to the country and Africa at large. The reasons for choosing the area for each of the human studies and the proposed design will be explained to them.

III.5. Recruitment and screening

As a screening process, a list of the eligible schoolchildren will be made from which the selection will be performed. Children will be eligible to participate if they meet the following screening criteria:

III.5.1. Inclusion criteria:

• School children aged 7-12 years

• Child's guardian or parent is willing to provide informed consent

• Child's family plans to remain resident in the study areas during the study period

III.5.2. Exclusion criteria:

• Clinical evidence of vitamin A deficiency a mis en forme : Anglais (Royaume-Uni)a mis en forme : Anglais (États-Unis)a mis en forme : Anglais (États-Unis)a mis en forme : Anglais (États-Unis)

• Severe malnutrition (weight-for age Z score (WAZ) or height-for-age Z score (HAZ) <-3 SD)

• Serious illnesses including tuberculosis, symptomatic HIV infection

• Child aged <6 years or >12 years III.6. Study variables After screening for eligibility, informed consent will be obtained from the child and their caregiver or parent. We will collect socioeconomic information (e.g. income, education, family assets) and perform anthropometric measurements (weight, height), assessment of hemoglobin status, and prepare thick and thin blood smears for malaria detection. A stool sample will be collected for intestine parasites determination.

III.7. Blood samples processing For the 60 children involved in the total body store analysis, we will target children aged 7 to 12 years and baseline subject characteristics will be obtained. At baseline, a fasting venous blood sample (7 mL) will be drawn on each child. Thereafter, they will be given an oral dose of 1.0 μmol [14,15]-13C2-retinyl acetate, which will be synthesized as published, dissolved in soybean oil, and administered with a 250 microL positive displacement pipet. The dose will immediately be followed by one teaspoon of olive oil to aid absorption. Children will eat a standardized breakfast of white bread and chocolate-flavored hazelnut paste. The follow-up blood will be collected by Orodara's clinic nurses on day 14 after the dose, stored immediately on ice, and transported to the Institut de Recherche en Sciences de la Santé located in Bobo-Dioulasso for processing.

The serum samples of the same group of 60 school children will be aliquoted into three tubes and frozen at -80 degrees C.

All samples will be shipped or hand-carried to University of Wisconsin-Madison (UW) for analysis by gas chromatography-combustion-isotope ratio mass spectrometry (GCCIRMS) as outlined below.

The baseline blood samples will also be analyzed for serum retinyl esters and carotenoids as described below for low sample volumes.

Similar method has been used in Burkina Faso in the past, which revealed low vegetable intake. However, vitamin A-fortified oil has now become widely available in the country since that study was done. This study will add value to the validation of serum retinyl esters as a measure of vitamin A excess.

A pre-purification of retinol from serum will be investigated using solid-phase extraction (SPE) methods with preservative from one aliquot to avoid the extra regulatory burden of shipping human biological samples (serum). We will first determine optimal protocols at the University of Wisconsin - Madison. Pilot studies have suggested that forward-phase SPE using NH2 columns and hexane/ethyl acetate solvents have very high recovery and purity of retinol from serum, and we will investigate butylated hydroxytoluene and alpha-tocopherol as preservatives. If SPE- prepurification is used in the field, we would plan to repurify the retinol by HPLC before introduction into GCCIRMS, so we will determine which preservative will not interfere with HPLC repurification and analysis.

An existing data set in African preschool children aged 3 to 4 years, to explore how the RID technique is affected by inflammation at the time of dosing. It is well-known that serum retinol and retinol binding protein are temporarily decreased by inflammation. This can affect estimates of vitamin A status in populations with high levels of clinical or subclinical infections. However, the potential effect of inflammation on RID has not been evaluated in a large group of children. We have calculated total body vitamin A stores in children from Burkina Faso, Ethiopia, Cameroon, South Africa, and Tanzania. Inflammation status has been assessed by measuring C-reactive protein (CRP) and alpha1-acid-glycoprotein (AGP) obtained from the blood samples at time of dosing. We will look at the association between total liver reserves (TLR) based on the 13C-RID test and continuous values of CRP and AGP, testing whether the slope of the relationship is not equal to zero. We will also compare categorical means for TLR using the following four categories of inflammation: no inflammation, incubation (elevated CRP (>5 mg/L) only), early convalescence (elevated CRP and AGP (>1 g/L)), and late convalescence (elevated AGP only) using an ANOVA test for differences among groups.

For the second cross sectional study involving MRDR test, we will still target children aged 7 to 12 years and baseline subject characteristics will be obtained to evaluate absorption curves using an analog and to determine if there is a difference using capsules in the absorption curves.

This study will investigate the absorption of 3,4-didehydroretinyl acetate as an analogy to labelled retinyl acetate in capsules versus direct oral doses in oil. Using the same awareness and sensitization procedures as noted above, Children will be separated into six groups of 15 and randomly assigned to capsule administration or direct oral dose with different time point of blood samples taken collection, from 0 to 7 hours later. All samples will be analyzed for 3,4-didehydroretinol and 3,4-didehydroretinyl esters. The alcohol form will equate to that which is bound to retinol binding protein (RBP), while the esters will reflect the absorption curves.

All children will receive 7.0 μmole 3, 4-didehydroretinyl acetate dissolved in approximately 200 μL soybean oil. The cutoff for vitamin A deficiency for the MRDR test is 0.060, which is the 3,4-didehydroretinol to retinol ratio at 4 to 6 hours after dosing. Therefore, each child will have this measurement. The area under the ester absorption curve from 2 to 5 hours will reveal the chylomicron distribution from the oral dose. All samples will be analyzed using the standard 250 μL method for the MRDR and an extended UPLC method for 3,4-didehydroretinol and 3,4-didehydroretinyl esters.

The study will have two phases:

An initial phase will be carried out after the mangoes season (November-December) and a second phase of the study will be carried out during the mangoes season (April- May).

III.8. Laboratory analysis

Analysis of samples by gas chromatography-combustion-isotope ratio mass spectrometry (GCCIRMS):

Serum: Serum extraction procedures and instrumentation for GCCIRMS analysis have been previously reported. Briefly, serum us thawed at room temperature. To 0.75 mL (or all available) serum, ethanol (1 x vol) is added to precipitate proteins, C23 beta-apo-carotenol is added as an internal standard, and 3 x 1 mL hexane extractions are performed. Hexanes are pooled, dried under nitrogen, resuspended in 100 μL methanol, frozen at -80⁰C for ~5 min, centrifuged briefly, and injected onto HPLC system 1 for quantification and purification. System 1 is a Gracesmart C18 (5 μm, 4.6 x 250 mm) column with 92:8 acetonitrile:water at 1.0 mL/min. The retinol fraction us collected, dried under nitrogen, and resuspended in 100 μL methanol for injection onto system 2. System 2 is an identical column with 100% methanol at 0.7 mL/min. The retinol fraction is collected, dried under vacuum with a SPD1010 SpeedVac, resuspended in 4-6 μL hexane, and 1-3 μL injected (to inject approximately 75 ng) on the GC. The GC/C/IRMS system was published. Atom % (At%) is directly calculated (Isodat version 2.0) in reference to carbon dioxide, which is calibrated against a sucrose standard (NIST, 8542).

Retinyl ester and carotenoid analysis: Serum carotenoids and retinyl esters will be analyzed as published at the University of Wisconsin-Madison Department of Nutritional Sciences laboratory by ultra-performance liquid chromatography (Waters UPLC®) according to a previously described solvent system, optimized for low sample volumes and desired analytes. Briefly, 150 μL serum, or all available, is aliquoted into borosilicate glass tubes (Fisher Scientific) to which 100 μL C-23 β-apo-carotenol (internal standard optical density in methanol ~1.0 at 325 nm) and 225 μL ethanol are added. Organic compounds are extracted three times with 500 μL hexanes, mixed with a vortex and centrifuged at relative centrifugal force of 1200 units of gravity for 5 min. Pooled hexane layers are dried in conical-bottom glass tubes under a stream of nitrogen. The dried extract is re-suspended in 75 μL of 75:25 methanol:dichloroethane concentration, transferred to UPLC vials (Waters) and 6 μL injected onto the UPLC system.

III.9. Organic synthesis for completion of the project. Synthesis of 13C-retinyl acetate for completion of Objectives 1 and 2. 13C2-retinyl acetate will be synthesized using a procedure that has been published in detail by Tanumihardjo (2001). Briefly, C18-tetraene ketone is synthesized from β-ionone, a commercially available precursor, and carefully purified. Using a modified Whittig-Horner procedure, C-18 tetraene ketone is reacted with the carbanion of 13C2-labeled triethylphosponoacetate to add 2-13C's to the backbone of the retinol moiety. The resulting ethyl ester is reduced to the alcohol and esterified to 13C2-retinyl acetate with acetic anhydride. The synthetic all-trans isomer is purified carefully on 8%-water- deactivated alumina using relatively innocuous, highly volatile, organic solvents, e.g., hexanes and diethyl ether. The resulting preparation is characterized using UV-VIS spectroscopy, TLC, and HPLC. For long-term storage, the 14, 15-13C2-retinyl acetate is stored dissolved in soybean oil at -80oC. We will prepare the final doses in soybean oil (~200 μL/dose) shortly before the study and hand-carry them to each study site.

Synthesis of 3,4-didehydroretinyl acetate for completion of Objective 4: Dr. Tanumihardjo routinely synthesizes 3,4-didehydroretinyl acetate for the Centers for Disease Control, Atlanta, Georgia, to use in their population-based nutrition evaluations. Therefore, we do not foresee an issue in using current rotating stocks that are in storage at -80oC for the proposed vitamin A status evaluation in the infants of the mothers enrolled in the kinetic study. The synthetic procedures used exclusively in the Tanumihardjo laboratory are published. Briefly, retinoic acid is methylated with diazomethane. The resultant methylretinoate is oxidized with manganese dioxide to form 4=oxo methylretinoate. After careful purification and crystallization, the 4=oxo is reduced to 4-OH methylretinoate using sodium borohydride. After crystallization in hexanes at -30oC for several days, 4-OH is dehydrated with p-toluene sulfonic acid and the resultant 3,4-didehydro-methylretinoate is purified and allowed to crystallize overnight. Finally, 3,4-didehydro-methylretinoate is reduced to 3,4-didehydroretinol with lithium aluminum hydride and immediately acetylated to 3,4-didehydroretinyl acetate, purified, dissolved in soybean oil, and stored at -80oC until doses are prepared and customized for individual studies.

III.6. Ethical clearance The protocols and informed consent forms will undergo ethical review in the Institutional Review Boards of IRSS and will be registered at Pan African Clinical Trial Registry (www.pactr.org) database under a unique identification number. We already have experience working with the RID test in children.

IV. Study working plan Time plan, IRSS Burkina Faso Activities Year 1 Year 2 Year 3 Year 4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Protocol finalisation Ethical review and approval SOPs development Study site preparation Study staff selection and training First round of the first cross-sectional study Samples preparation and storage Second round of the first cross-sectional study Samples preparation and storage Samples shipment First round of the second cross-sectional study Samples preparation and storage Second round of the second cross-sectional study Samples preparation and storage Samples shipment Laboratory analysed Data handling Disseminations, local, national, international