Design and development of a combined calcium– iron–folic acid prenatal supplement to support implementation of the new World Health Organization recommendations for calcium supplementation during pregnancy Ashley M. Aimone Phillips, Stanley H. Zlotkin, Jo-Anna B. Baxter, Frank Martinuzzi, Tanush Kadria, and Daniel E. Roth Abstract Background. Hypertensive diseases of pregnancy are important causes of maternal and perinatal mortality. Based on meta-analyses of efficacy trials of prenatal calcium supplementation to reduce the risk of hypertensive diseases of pregnancy, the World Health Organization recommends 1.5 to 2.0 g of elemental calcium per day for pregnant women with low dietary calcium intakes (as well as 60 mg of iron and 400 µg of folic acid). However, implementation of this recommendation is challenged by the size and number of calcium tablets required and the need to avoid concurrent ingestion of calcium and iron due to intraintestinal interactions. Objective. We developed a novel micronutrient powder containing microencapsulated pH-sensitive calcium in addition to iron and folic acid, designed to facilitate early intestinal iron release and delayed calcium release. Methods. Two pharmaceutical companies were contracted to develop a prototype, one of which was chosen for clinical testing. Calcium carbonate granules were coated with a trilayer pH-sensitive enteric coating using a fluid-bed spray coater. Iron and folic acid granules were encapsulated with a time-release coating. Iron and calcium dissolution profiles were assessed during exposure to acidic (pH 1.2) and/or basic (pH 5.8) media using a modified USP apparatus 1 (basket) method. Results. At pH 1.2, calcium and iron release was ≤ 10% and > 90% after 120 minutes, respectively. At pH 5.8, > 80% of total calcium was released after 90 minutes.
Ashley M. Aimone Phillips, Stanley H. Zlotkin, Jo-Anna B. Baxter, and Daniel E. Roth are affiliated with the Centre for Global Child Health, Hospital for Sick Children, and the University of Toronto, Toronto, Ontario, Canada; Frank Martinuzzi and Tanush Kadria are affiliated with the Toronto Institute of Pharmaceutical Technology, Toronto. Please direct queries to the corresponding author: Daniel E. Roth, Department of Pediatrics, Hospital for Sick Children, 555 University Ave., Toronto, ON M5G1X8, Canada; e-mail: [email protected].
Conclusions. Based on in vitro criteria, the supplement may be a promising approach for delivering calcium, iron, and folic acid as a single daily dose to pregnant women in settings of low dietary intake of calcium.
Key words: Calcium, encapsulation, hypertension, powder, pregnancy
Introduction Hypertensive diseases of pregnancy, such as preeclampsia, are among the most important causes of maternal and perinatal morbidity and mortality worldwide [1]. The majority of hypertensive diseases of pregnancyrelated deaths occur in low- and middle-income countries where antenatal care is limited, particularly in communities that lack adequate access to emergency obstetric and neonatal intensive care [1]. The prevention of hypertensive diseases of pregnancy is therefore a major public health priority, particularly under resource-limited conditions. In populations where dietary calcium intake is low, randomized, controlled trials and meta-analyses have demonstrated that calcium supplementation during pregnancy significantly decreases the risk of hypertensive diseases of pregnancy [2–5]. A 2010 Cochrane review concluded that supplementation with at least 1 g of calcium reduced the risk of preeclampsia by 55% (95% CI, 0.31 to 0.65), gestational hypertension by 35% (95% CI, 0.53 to 0.81), maternal mortality or morbidity related to hypertensive diseases of pregnancy by 20% (95% CI, 0.65 to 0.97), and preterm birth by 24% (95% CI, 0.60 to 0.97), particularly in populations with low dietary calcium intake [2]. These estimates were updated in a 2012 review by Imdad and Bhutta, which also showed similar reductions in the risk of preeclampsia, maternal mortality or severe morbidity, and preterm birth, as well as an increase in birthweight (mean difference, 85 g; 95% CI, 37 to 133) [6]. Prenatal calcium supplementation has been identified as the
Food and Nutrition Bulletin, vol. 35, no. 2 © 2014, The Nevin Scrimshaw International Nutrition Foundation.
221
222
only intervention to prevent hypertensive diseases of pregnancy that may be feasibly delivered at the community level [3]. Although the policy is not yet widely implemented, the World Health Organization (WHO) recommended in 2011 a routine policy of prenatal supplementation with 1.5 to 2.0 g of elemental calcium per day, starting at around 20 weeks of gestation, for the prevention of preeclampsia in settings where usual dietary calcium intake is low [7]. In 2013, WHO reiterated the same recommendation in a statement focused specifically on calcium supplementation during pregnancy [8]. The WHO statement was further supported by the 2013 Lancet series on Maternal and Child Nutrition, which included calcium supplementation as one of 10 key evidence-based interventions that could significantly improve maternal (and child) health [9], despite very little evidence from effectiveness studies or programmatic settings. Although it is considered to be a simple, low-cost intervention for the prevention of hypertensive diseases of pregnancy, the effectiveness of routine prenatal calcium supplementation is hindered by several challenges: the bulk of supplemental calcium required to deliver the recommended 1.5 to 2 g per day; low adherence to conventional prenatal vitamin and mineral tablet regimens (typically composed of iron and folic acid), which may be further compromised by an increase in the number, size, or frequency of tablets; and intraintestinal inhibition of iron absorption by calcium when the nutrients are concurrently ingested. Calcium is generally assumed to interfere with iron absorption from dietary and supplemental sources in a dose-dependent and dose-saturable manner [10]. However, the mechanism of inhibition remains unclear, and there is ongoing debate regarding the extent to which various forms of supplemental and dietary calcium inhibit iron absorption [11–14]. WHO recommends iron and folic acid supplementation during pregnancy as part of routine antenatal care for the prevention of pregnancy-related anemia and neural tube defects, respectively [15]. The 2011 WHO guidelines on preeclampsia proposed “that concomitant administration of [iron and calcium supplements] should be avoided. Ideally, the two supplements should be administered several hours apart (e.g., morning and evening)” [7]. Furthermore, the 2013 WHO guidelines suggested that the daily dose of calcium should be “divided into three doses (preferably taken at mealtimes)” [8], based on the assumption that fractional calcium absorption declines substantially at doses greater than 500 mg [16]. From a practical standpoint, calcium tablets available on the market do not generally contain more than 600 mg of elemental calcium per unit. Therefore, adherence to current WHO recommendations would require four separate supplements to be consumed daily by pregnant women. We are not aware of published research that has
A. M. A. Phillips et al.
addressed the implications of the WHO recommendation for scale-up of both calcium and iron–folic acid (IFA) supplementation, proposals of feasible alternatives to the temporal separation of the two nutrients in routine antenatal care, or evaluations of potential technologies to enable the simultaneous delivery of iron and calcium in the same supplement. We therefore undertook the development of an innovative microencapsulated prenatal multimicronutrient (calcium, iron, and folic acid) supplement (SUPP) designed to facilitate immediate iron and folic acid release (in the stomach) and delayed calcium release (thereby minimizing intraluminal nutrient–nutrient interactions), and enable a single daily dose of all three nutrients. The objectives of this report are to describe the collaborative process involved in the development of the first prototype and to summarize the results of the in vitro proof-of-principle studies. Our early-stage product development experience provides an example of the challenges of moving an innovative dietary supplement from idea to proof of concept within an academic, investigator-initiated, grant-funded research setting.
Methods Granting process and partnership development
In July 2011, a research group led by investigators at the Hospital for Sick Children (“SickKids”), a public University of Toronto-affiliated teaching hospital in Toronto, Canada, was awarded a seed grant in the first round of the Saving Lives at Birth partnership, a collaborative funding initiative aimed at stimulating innovation of health technologies to reduce maternal and neonatal mortality in rural, resource-poor settings [17]. Cofunding to support the use of stable isotopes was obtained from the International Atomic Energy Agency. In December 2011, SickKids contracted with two private-sector pharmaceutical companies (referred to as company A and company B) to develop a product containing prescribed amounts of powdered calcium, iron, and folic acid, in accordance with the following key design criteria. Dissolution properties: the calcium was to be microencapsulated so as to promote its release in the mid to distal duodenum and thus avoid interactions with iron, which is predominantly absorbed in the proximal duodenum. Physical properties: the encapsulated calcium particle size had to be small enough to avoid disruption of the encapsulate (coating) by chewing, as well as minimize overall powder bulk. Organoleptic properties: the powder was to be tasteless and have a texture and stability suitable for mixing in water and fruit juices and, if feasible, hot beverages (e.g., tea). The contracted companies were not identified to each other, but both were informed that a second bidder was involved. The contracts
223
Development of a prenatal calcium and iron supplement
included time-limited confidentiality clauses to facilitate open exchange of ideas between SickKids and each of the contractors during the period of product development. SickKids and both contractors signed nonexclusive license agreements “to ensure global access to the [project] outputs should the Grantee, or the Grantee’s licensees, collaborators or partners, fail to make them widely available at a reasonable price for the benefit of people in Developing Markets” [18]. This condition would not impinge on intellectual property ownership or rights to exclusive licensure in North America or in other high-income countries. The budgets for the two contracts were negotiated separately; however, the deliverables for both companies were an encapsulated calcium, iron, and folic acid powder prototype and a detailed report of product specifications to be presented to the SickKids team within 6 months. The products were evaluated and compared based on specifications provided by the companies (in vitro dissolution tests, material and particle characteristics, product stability tests, and estimated production costs), as well as qualitative visual inspection and taste tests by SickKids investigators. Production of the supplement prototype
One prototype was selected to proceed to production of a clinical batch for use in initial clinical proof-ofconcept studies. The composition and preparation of the selected prototype are described below. Preparation of base calcium carbonate granules
Using a wet granulation method, very fine calcium carbonate powder (10 µm average particle size, Nutri Granulations, La Mirada, CA, USA) was granulated in a high-shear mixer (GEI Collette Gral 10, Wommelgem, Belgium) with chopper and agitator speeds of 1,600 and 400 rpm, respectively. During mixing, the following ingredients were added to achieve a total weight gain of 34.4% w/w: microcrystalline cellulose, PH101 (8.8% w/w), maltodextrin M200 (17.5% w/w), and sodium starch glycolate (8.1% w/w). The resulting wet massed granules were then dried at 45°C for 30 minutes in a fluid-bed dryer (Uni-Glatt, Glatt, Binzen, Germany) and screened for size using sieves (Ro-TAP, W. S. Tyler, Ohio, USA) with mesh sizes of 30 and 50. Granules were accepted if they passed through 30 mesh and were retained on 50 mesh, which corresponds to a particle size less than 600 µm and greater than 300 µm. The acceptable particle size range was considered optimal for maintaining coating efficiency, while reducing the risk of potential coating damage due to chewing. Application of subcoating and enteric coating to calcium carbonate granules
A 2.0% to 3.0% w/w subcoating, applied to the subset of granules that passed the granulation step, consisted
of a polyvinyl alcohol–polyethylene glycol copolymer (Kollicoat IR, BASF, Mississauga, ON, Canada), talc, and croscarmellose sodium (AC-DI-SOL). The subcoating was applied using a fluid-bed top spray coater (Uni-Glatt, Glatt, Binzen, Germany) with a spray rate of 2.5 mL/min (nozzle size 5), an atomizing pressure of 30 to 45 psi, and an inlet air temperature of 35° to 40°C. The resulting subcoated granules were again screened for size using mesh sizes 30 and 50. Accepted granules from the subcoating step were then coated with three layers of a pH-dependent polymethacrylic acid, methyl methacrylate copolymer enteric coating (Eudragit L 30 D-55, Evonik Industries AG, Darmstadt, Germany), which was formulated to begin dissolving at pH 5.8, and thus target mid-duodenal calcium release. Each coating layer consisted of Eudragit L 30 D-55, triethylcitrate, and talc. The first layer was neutralized to pH 5.5 and applied at the 25% w/w level. The second layer was neutralized to pH 4.5 and applied at the 15% w/w level. The third layer was applied at the same level as the second layer, but not neutralized. Similar to the methods described above, the enteric coatings were applied using a fluid-bed top spray coater (UNI, Glatt, Binzen, Germany) with a spray rate of 5 mL/min (nozzle size 8), an atomizing pressure of 30 psi, and an inlet air temperature of 35° to 40°C. The resulting enteric-coated granules were screened again for size using 30 and 50 mesh sizes. Addition of iron, folic acid, and suspending agents
The iron granules were encapsulated with a timedependent coating (Eudragit RL 30 D, Evonik Industries AG, Darmstadt, Germany), which included folic acid in the matrix and was formulated to target gastric release. The iron coating consisted of Eudragit RL 30 D, triethylcitrate, and talc. Each dose of calcium contained a fixed combination of 60 mg of elemental iron and 400 µg of folic acid to meet WHO recommendations [15]. As a final step, a suspension formulation including sorbitol (Neosorb P60W), xanthan gum, and citric acid anhydrous was added to each individual dose of calcium, iron, and folic acid. Adding these ingredients prevented the powder particles from settling to the bottom of the vessel when mixed with water. The final product was added to plastic vials (approximately 7.0 cm tall by 2.5 cm in diameter) with desiccant-filled caps (DesiCap, Multisorb Technologies, Buffalo, NY, USA). In-house simulation and taste test trials were used to determine the optimal composition of the suspension. For each simulation trial, a single packaged dose of the prototype supplement was shaken by inverting the vial three or four times, then dispensed into a clear glass or plastic drinking cup with approximately 125 mL of clean water and stirred with a plastic or metal spoon until the mixture was homogeneously dispersed and the solution had thickened. The suspension formulation was qualitatively evaluated based on
224
the following criteria: smooth consistency (minimal “clumping”), thickness appropriate for drinking, low residual adherence (the extent to which granules remained on the sides of the cup after decanting), short mixing time, mild to no flavor, odorless, and acceptable tongue feel (minimal “grittiness”). Dissolution testing
The dissolution performance of the enteric-coated calcium granules was assessed by sequentially exposing the coated granules to acidic media (USP 0.1N HCl buffer, pH 1.2) for 120 minutes then basic media (Na-citrate buffer, pH 5.8) for 90 minutes, using a Hanson SR-8 Plus dissolution tester (Hanson Research Corporation, Chatsworth, CA, USA), and following a modified United States Pharmacopeia (USP) basket apparatus methodology for delayed-release (entericcoated) dosage forms (USP 35–NF 30). As per USP standards, the paddle speed was set at 100 rpm, and all tests were performed at 37°C [19]. Modifications included wrapping the baskets tightly with a porous polypropylene material and replacing media throughout the base testing phase to maintain a pH of 5.8. The baskets were wrapped in order to reduce their mesh size and prevent leakage of test granules into the surrounding vessel, thus minimizing subsequent analytical errors (which were noted in preliminary tests). A 25-mL sample was taken from the test media at approximately 10-minute intervals starting at 120 minutes (end of the acid stage). These samples were filtered, and 20 mL of the filtrate was titrated with 0.01M sodium ethylenediaminetetraacetate (EDTA) using a modified USP titration method. The amount of CaCO3 was calculated based on the amount of EDTA required to reach the end point of titration, where 1 mole of CaCO3 (100 g) is equivalent to 1 mole of EDTA in solution form (372.8 g in 1,000 mL of water). The stability of the enteric coating was also assessed by performing dissolution tests after storing the coated granules at room temperature (20° to 25°C/35% to 40% relative humidity) for 4 weeks, and under accelerated conditions (40°C/75% relative humidity) for 12 weeks. The dissolution profile of enteric-coated calcium was considered acceptable if calcium release did not exceed 10% during the acid stage and reached at least 80% within 90 minutes of exposure to basic media. According to industry standards, the dissolution threshold for delayed-release or enteric-coated products should be ≤ 10% during a 2-hour acid stage and 75% to 80% for a 30- to 45-minute basic stage [19]. The percent release tolerance level and duration of a basic stage can be adjusted according to the solubility of the compound being tested. In the case of CaCO3, which has low solubility, the dissolution threshold of the basic phase was maintained at 80%, while the duration of the test was increased to 90 minutes. The encapsulated ferrous fumarate (and folic acid) granules also underwent
A. M. A. Phillips et al.
dissolution testing, although in acidic media only because the coating was formulated to promote gastric release. The dissolution profile of the iron and folic acid granules was considered acceptable if at least 90% iron and folic acid release was achieved within 30 minutes of exposure to acid media. Statistical methods
The results obtained from all dissolution and analytical test runs were summarized and expressed as mean and SD percent release for each specified time point. Dissolution profiles of coated calcium granules stored at room temperature (20° to 25°C/35% to 40% relative humidity) or under accelerated conditions (40°C/75% relative humidity) were compared at 7, 14, and 28 days. Mean percent release values were determined at each time point for the room (reference) and accelerated (test) conditions, then the mean dissolution values from both curves were used to calculate difference (f1) and similarity (f2) factors [20]. The difference factor (equation 1) is a measurement of the relative error between the two curves, and represents the percent (%) difference at each time point. The similarity factor (equation 2) is a logarithmic reciprocal square root transformation of the sum of squared error, and represents the similarity in percent dissolution between the two curves. Equation 1: f1 = {[∑t–1 n | Rt – Tt |]/[∑t–1 n Rt ] ∙ 100} Equation 2: f2 = 50 ∙ log{[1+(1/n) ∑t–1 n (Rt – Tt)2 ]–0.5 ∙ 100} where n is the number of time points, R is the dissolution value of the reference batch at time t, and Tt is the dissolution value of the test batch at time t. In order for curves to be considered similar, the f1 values should be < 15 (acceptability range, 0 to 15) and the f2 values should be > 50 (acceptability range, 50 to 100). All calculations were performed using Microsoft Excel 2007. Sample calculations of f1 and f2 are included in the appendix.
Results Two companies engaged in prototype development under contract from SickKids investigators from December 2011 to July 2012. The companies proposed encapsulated calcium powder supplement prototypes that differed in materials used, particle specifications, palatability, stability, and estimated production costs (table 1). The companies used similar coating approaches but different coating materials. Company A used an acetate phthalate-based polymer, while company B used a polymethacrylic acid, methyl methacrylate copolymer. Company B also applied a
225
Development of a prenatal calcium and iron supplement
subcoating before applying the enteric coating, which may have been partly responsible for observed differences in palatability and stability of the final product. The estimated costs of prototype development and production for the subsequent clinical study also differed substantially between the two companies. Cost was considered to be an important factor when planning for potential future large-scale production. Both companies used similar dissolution testing and in vitro analytical procedures. Despite slight variations in sampling times, the overall dissolution profile for prototype B was more acceptable than that for prototype A, since the release of calcium from prototype B remained below 10% during the acid phase and rose
above 80% during the basic phase, neither of which were demonstrated by prototype A (table 2). In fact, the coating of prototype A became very unstable and demonstrated poor dissolution performance in acidic conditions when retested approximately 2 weeks after application; the resulting calcium granules had unacceptable organoleptic properties (e.g., bitter taste), and were not considered a viable product for use in clinical studies. Company A proposed to refine their prototype upon provision of additional funds (beyond the original contract); however, this was not feasible within the scope of the grant. The SickKids investigators chose to continue to work with company B, which proceeded to finalize
TABLE 1. Enteric-coated calcium prototype characteristics produced by two private-sector pharmaceutical companies Item description
Company A
Particle size of granules Granule formulation Coating method Coating materials
150–250 µm Calcium carbonate, starch Fluid bed Hydroxypropyl methyl cellulose acetate phthalate (Shin-Etsu HP-55)
Coating weight Dissolution test method Analytical method(s) Palatabilitya Stabilityb
Company B
300–600 µm Calcium carbonate, HPMC E6 Fluid bed Subcoating: polyvinyl alcohol–polyethylene glycol (Kollicoat IR) Enteric coating: polymethacrylic acid, methyl methacrylate (Eudragit L 30 D-55) Enteric coating: 37.5% w/w Subcoating: 10% w/w Enteric coating: 25% w/w USP apparatus 1 (basket) with 0.1N Modified USP apparatus 1 (basket) with 0.1N hydroacetic acid and ammonia/ammochloric acid and sodium citrate buffers nium chloride buffers Titration with sodium EDTA Titration with sodium EDTA (titrant) (titrant) Unacceptable Acceptable Unstable at 7 days Stable at 56 days
EDTA, ethylenediaminetetraacetic acid; HPMC, hydroxypropylmethylcellulose a. Based on a qualitative assessment of taste, odor, and mouthfeel in taste tests of dry product. Considered acceptable if tasteless, odorless, and had minimal grittiness on tongue. b. Product was considered to be stable if dissolution test showed < 10% release within 60 minutes of exposure to acid medium after storage at room temperature (20° to 25°C/35% to 40% relative humidity) or accelerated conditions (40°C/75% relative humidity).
TABLE 2. Dissolution profiles of enteric-coated calcium powder prototypes developed by company A and company Ba % calcium release (SD) Dissolution phase and exposure time (min)
Prototype A (37.5% enteric coating weight)
Prototype B (25% enteric coating weight)
Acid phase (pH 1.0–1.2) 30 60 120
0.28 (2.1) 1.12 (1.0) 14.3 (3.7)
NV NV 5.3 (1.2)
Basic phase (pH 5.5–5.8) 10 30 60 120
54.8 (2.2) 72.2 (1.0) 83.5 (5.5) 87.4 (2.6)
18.0 (4.0) 68.5 (12.3) 88.0 (5.9) 95.1 (4.7)
NV, no value (sample was not drawn at specified time point) a. Percent calcium release values represent means (SD) of multiple samples.
226
A. M. A. Phillips et al.
the composition of the prototype supplement, including the incorporation of suspension agents and coated iron–folic acid granules (table 3). The dissolution profiles of the final coated and uncoated calcium carbonate and ferrous fumarate granules (fig. 1) were considered acceptable according to our criteria, indicating that the differential release of calcium and iron was consistently achieved within targeted acid-base environments. According to the similarity and difference factors (f1 and f2 ), calculated at each time sampling point, the dissolution profiles of coated calcium granules stored in accelerated conditions demonstrated some deterioration by 28 days, compared with those stored at room temperature (table 4).
Discussion Short-term contractual engagement of two private-sector pharmaceutical companies by an academic research institution resulted in the successful development of an oral combined calcium–iron–folic acid supplement prototype that met three main predefined acceptance criteria: targeted release of calcium at pH 5.5 to 5.8, and thus at a more basic pH relative to the release of iron and folic acid, based on in vitro dissolution tests; low likelihood of disruption of the enteric coating by chewing, due to sufficiently small particle size and the use of suspension agents to yield a liquid that facilitated the comfortable swallowing of intact particles; and acceptable palatability due to masking of the metallic taste of iron and the chalky texture of calcium. Several challenges were encountered in the development of a usable and palatable powdered supplement. The first was the sheer bulk of material required to deliver 1.5 g of elemental calcium (3.75 g of calcium carbonate) in addition to the multiple excipients (coating constituents as well as suspension agents). The final
TABLE 3. Composition of the final powdered supplement product, including enteric-coated calcium, encapsulated iron and folic acid, and suspending agents Amount (% total weight)
Component Calcium carbonate granules Calcium subcoating Calcium enteric coating Ferrous fumarate/folic acid granules Ferrous fumarate/folic acid coating Suspension agents
TABLE 4. Dissolution profile comparison of enteric-coated calcium carbonate granules stored in accelerated conditions (40°C/75% relative humidity) versus room conditions (20° to 25°C/35% to 40% relative humidity) No. of days stored in accelerated conditions
f1a
f2b
7
4.2
71.7
14
9.0
57.0
28
19.3
40.3
a. Acceptable range of difference factor (f1) is 0–15. b. Acceptable range of similarity factor (f2) is 50–100.
powder formulation could be readily mixed with water or fruit juice, yet at least 1 minute of vigorous stirring was required to produce an acceptable suspension. Multiple rounds of sampling and modification of the concentrations of suspension agents (sorbitol and xanthan gum) were required to arrive at the formulation that optimally suspended all of the calcium and iron–folic acid particles; the resulting fluid had a nectar consistency, with particles that remained visible as well as palpable in the mouth. Our initial intention was to design a product that would deliver the entire 1.5-g WHO-recommended daily intake of elemental calcium
Ferrous fumarate (coated)
Calcium carbonate (uncoated)
100 Ferrous fumarate (uncoated)
% release
80 60 Calcium carbonate (coated)
40 20 0
0
20
40
60
pH 1.2 for 2 h
80
100
34.21 0.67 29.52 1.67 0.12 33.81
120
140
Time (min)
160
180
200
220
240
pH 5.8 for 1.5 h
FIG 1. Dissolution profile of the final encapsulated calcium carbonate and ferrous fumarate granules
Development of a prenatal calcium and iron supplement
in one dose unit. However, we also produced dose units containing 500 mg and 1,000 mg of elemental calcium; these formulations have the advantage of less material per dose, but would have the disadvantage of requiring multiple daily dose regimens. A second challenge was the unexpected degree of technical difficulty involved in microencapsulating calcium carbonate, because of the relatively large granule size, coarse particle texture, and reactivity of the carbonate salt. Both companies found that their initial enteric coats were poorly protective against acid and disintegrated rapidly in ambient conditions, probably due to the coarse texture of the particles, which resulted in a porous coat that permitted moisture to enter and react with the carbonate ions. Company B successfully addressed this problem by improving the granulation process, applying a subcoat to create a more homogeneous particle surface prior to application of the enteric coating, as well as by increasing the thickness of the enteric coating itself. A third challenge was the implementation of a suitable analytical system for dissolution testing. The USP apparatus used in the current project had to be modified because the standard basket mesh dimensions were not suitable for the particle size range that we had specified as acceptable. A fourth challenge was the development of the coated iron–folic acid granule, since we had initially envisioned simply adding unmodified folic acid and off-the-shelf lipid-microencapsulated ferrous fumarate (as used in the original Sprinkles) to the powdered calcium mixture. However, we later recognized that this form of iron would not mix well in an aqueous solution, and also that the lipid coating would quickly disintegrate in any hot liquid, thereby failing to mask the metallic taste of the iron. To address this, company B proposed the use of a time-dependent coating, with planned release of the ferrous fumarate approximately 20 minutes after immersion in water or juice. The small dose of folic acid was readily integrated into the time-release coating. Although this aspect of product development was technically less difficult than designing the calcium coating, it nonetheless entailed unplanned time and financial costs. The final formulation was subsequently used to produce clinical batches for use in two studies that will be reported elsewhere: an acceptability and palatability study among pregnant women in Bangladesh, in which flavored and unflavored formulations of the supplement were compared with conventional and chewable tablets; and a dual stable isotope study of the fractional intestinal calcium absorption in pregnant women, the first step toward establishing biological proof-ofconcept in a clinical setting. The calcium absorption study has been conducted in Dhaka, Bangladesh, in collaboration with the International Centre for Diarrhoeal Diseases Research, Bangladesh (ICDDR,B) and the Texas Children’s Hospital. Based on the results of
227
the in vivo absorption of the enteric-coated calcium, a subsequent step may be a clinical trial to examine potential attenuating effects of enteric-coated calcium on iron absorption. Despite challenges in the development process, the new supplement prototype represents a potential approach for the combined delivery of daily supplemental calcium, iron, and folic acid to pregnant women. This novel delivery method was designed to support the implementation of the recent WHO recommendations for calcium supplementation (1.5 to 2 g per day) among women whose usual dietary calcium intake is low. Adherence to current WHO guidelines regarding prenatal iron and calcium supplementation, using products that are currently available on the market [8], would require a woman to take four separate tablets spaced throughout the day. However, it is well recognized that evidence-based guidelines face a range of barriers to their implementation [21]; we believe that the WHO prenatal calcium recommendation is unlikely to be widely adopted in low-income settings in the absence of a modified delivery vehicle and simplified schedule that combines calcium with other recommended micronutrients in a palatable and easily used format that can be produced and distributed at low cost. Our experience in developing a novel powdered micronutrient supplement may be relevant to other academic researchers engaged in early-stage collaborative efforts to develop innovative health technologies suitable for eventual scaling in low-income settings. Notably, productive public–private contractual partnerships were essential and were executed without compromising adherence to global access principles (including allowance for nonexclusive licensure in developing markets) or opportunities for intellectual property ownership or protection. We found that private-sector partners were as supportive as academic partners of a nonlinear, iterative process of experimentation and strategic risk-taking that is necessary for moving a technology across the so-called “valley of death” that separates conception and proof-ofconcept from commercialization [22]. We were initially concerned that our simultaneous engagement of two private-sector partners in a competitive process was risky. First, we worried that this approach might undermine the academic spirit and scientific aims of our collaborative relationships; however, we found that both companies were wholly comfortable with this approach. Second, we were uncertain whether a seemingly redundant subcontracting plan that is unconventional in the academic setting would be acceptable to the granting agency; however, Grand Challenges Canada (our direct funder for the Saving Lives at Birth partnership) was supportive of this approach from the outset. From the standpoint of achieving our research goals, this decision proved to be a very cost-effective risk mitigation strategy, particularly given that one of
228
the partners did not ultimately produce a prototype that met the design criteria within the agreed-upon budget and timeline. Small-grant or “seed” funding for investigators and entrepreneurs with innovative ideas for potential solutions to address global health problems has recently become a high-profile component of the global health research funding envelope (e.g., Bill & Melinda Gates Foundation Grand Challenge Explorations, Grand Challenges Canada Stars program). Arguably, these funding agencies have spurred researchers to spend more time thinking creatively about the path-to-scale of their research outcomes. Clearly, innovation cannot begin without the spark of an idea, but most ideas burn out early in the process of attempting to convert the concept into an efficacious and commercially viable product [23]. In this paper, we describe the early stages of development of a new once-daily prenatal micronutrient powder combining calcium, iron, and folic acid for the prevention of hypertensive diseases of pregnancy and maternal anemia in resource-poor settings. Successful development of an innovative prototype was enabled by seed funding from a nontraditional source of academic research funding (the Saving Lives at Birth program) and strategic engagement with two private sector partners. Based on a set of pre-established criteria upon which candidate prototypes were evaluated, in vitro and qualitative acceptability assessment suggested that the selected product was a promising approach for the delivery of iron, folic acid, and calcium to pregnant women. Nonetheless, a series of additional criteria (e.g., proof of concept with respect to iron absorption, clinical efficacy, cost-effectiveness, long-term stability) will
A. M. A. Phillips et al.
need to be met to continue to progress along the long road to implementation and scale-up.
Authors’ contributions Daniel E. Roth and Stanley H. Zlotkin designed the research; Frank Martinuzzi and Tanush Kadria produced the supplement and analyzed the data; Ashley M. Aimone Phillips and Jo-Anna B. Baxter wrote the first draft of the paper; all authors reviewed and approved the final manuscript.
Acknowledgments This work was made possible through the generous support of the Saving Lives at Birth partners: Grand Challenges Canada, the US Agency for International Development, the Government of Norway, the Bill & Melinda Gates Foundation, and the World Bank. The manuscript was prepared by the authors and does not necessarily reflect the views of the Saving Lives at Birth partners. The work was also supported in part by additional funding from the International Atomic Energy Agency. We would also like to acknowledge Diego Bassani, Brendon Pezzack, Nandita Perumal, Sohana Shafique, Elaine Gergolas (SickKids); Steve Abrams, Keli Hawthorne (Baylor College of Medicine); Abdullah al Mahmud, Munirul Islam, and Tahmeed Ahmed (ICDDR,B) for their contributions to the design concept and potential delivery applications of the micronutrient product.
References 1. Duley L. The global impact of pre-eclampsia and eclampsia. Semin Perinatol 2009;33:130–7. 2. Hofmeyr GJ, Lawrie TA, Atallah AN, Duley L. Calcium supplementation during pregnancy for preventing hypertensive disorders and related problems. Cochrane Database Syst Rev 2010:CD001059. 3. Imdad A, Jabeen A, Bhutta ZA. Role of calcium supplementation during pregnancy in reducing risk of developing gestational hypertensive disorders: a metaanalysis of studies from developing countries. BMC Public Health 2011;11(suppl 3):S18. 4. Villar J, Abdel-Aleem H, Merialdi M, Mathai M, Ali MM, Zavaleta N, Purwar M, Hofmeyr J, Nguyen TN, Campódonico L, Landoulsi S, Carroli G, Lindheimer M, Group WHOCSftPoPT. World Health Organization randomized trial of calcium supplementation among low calcium intake pregnant women. Am J Obstet Gynecol 2006;194:639–49. 5. Kumar A, Devi SG, Batra S, Singh C, Shukla DK. Calcium supplementation for the prevention of preeclampsia. Int J Gynaecol Obstet 2009;104:32–6. 6. Imdad A, Bhutta ZA. Effects of calcium supplementation
7. 8. 9.
10. 11.
12.
during pregnancy on maternal, fetal and birth outcomes. Paediatr Perinat Epidemiol 2012;26(suppl 1):138–52. World Health Organization. Recommendations for prevention and treatment of pre-eclampsia and eclampsia. Geneva: WHO, 2011. World Health Organization. Guideline: Calcium supplementation in pregnant women. Geneva: WHO, 2013. Bhutta ZA, Das JK, Rizvi A, Gaffey MF, Walker N, Horton S, Webb P, Lartey A, Black RE; Lancet Nutrition Interventions Review Group; Maternal and Child Nutrition Study Group. Evidence-based interventions for improvement of maternal and child nutrition: What can be done and at what cost? Lancet 2013;382:452–77. Hallberg L, Rossander-Hulthen L, Brune M, Gleerup A. Inhibition of haem-iron absorption in man by calcium. Br J Nutr 1992;69:533–40. Dawson-Hughes B, Seligson FH, Hughes VA. Effects of calcium carbonate and hydroxyapatite on zinc and iron retention in postmenopausal women. Am J Clin Nutr 1986;44:83–8. Cook JD, Dassenko SA, Whittaker P. Calcium supplementation: effect on iron absorption. Am J Clin Nutr
229
Development of a prenatal calcium and iron supplement
1991;53:106–11. 13. Hallberg L, Brune M, Erlandsson M, Sandberg AS, Rossander-Hultén L. Calcium: effect of different amounts on nonheme- and heme-iron absorption in humans. Am J Clin Nutr 1991;53:112–9. 14. Gaitán D, Flores S, Saavedra P, Miranda C, Olivares M, Arredondo M, López de Romaña D, Lönnerdal B, Pizarro F. Calcium does not inhibit the absorption of 5 milligrams of nonheme or heme iron at doses less than 800 milligrams in nonpregnant women. J Nutr 2011;141:1652–6. 15. World Health Organization/Food and Agriculture Organization. Vitamin and mineral requirements in human nutrition: report of a joint FAO/WHO Expert Consultation, Bangkok, Thailand, 21–30 September 1998. 2nd ed, 2004. Available at: http://whqlibdoc.who. int/publications/2004/9241546123.pdf. Accessed 9 March 2014. 16. Heaney RP, Weaver CM, Fitzsimmons ML. Influence of calcium load on absorption fraction. J Bone Miner Res 1990;5:1135–8.
17. Grand Challenges Canada. Saving lives at birth: a grand challenge for development. Available at: http://savinglivesatbirth.net/. Accessed 9 March 2014. 18. Grand Challenges Canada Website. Available at: http:// www.grandchallenges.ca/. Accessed 9 March 2014. 19. United States Pharmacopeia/National Formulary. Delayed release tablets. In: Convention USP, ed. United States Pharmacopeia, 35th ed, and National Formulary, 30th ed. Rockville, MD, USA: The United States Pharmacopeial Convention, 2012. 20. Moore JW, Flanner HH. Mathematical comparison of dissolution profiles. Pharm Technol 1996;20:64–74. 21. Haines A, Kuruvilla S, Borchert M. Bridging the implementation gap between knowledge and action for health. Bull World Health Organ 2004;82:724–31; discussion 32. 22. Grand Challenges Canada/Grand Defis Canada. Integrated innovation. Toronto, ON, Canada: Grand Challenges Canada, September 2010. 23. Coller BS, Califf RM. Traversing the valley of death: a guide to assessing prospects for translational success. Science 2009;1:1–5.
Appendix Sample calculation of f1 REFa
TESTb
ABS (Rt – Tt)
SUM (Rt – Tt)
SUM ( Rt)
B/C
D*100
5.9 24.7 52.3 70.8
5.5 18.2 46.2 69.0
0.4 6.5 6.1 1.7
17.7
417.7
0.0
4.2
82.0
82.8
0.9
89.0 93.0
89.2 94.9
0.2 1.9
a. Reference = average percent calcium release from April 19, May 1, and May 15 (2012) dissolution results (total n = 18 vessels). b. Test = average percent calcium release from day 7 accelerated stability dissolution results (total n = 6 vessels).
Sample calculation of f2 REFa
TESTb
(Rt –Tt) ^2
5.9 24.7 52.3 70.8 82.0 89.0 93.0
5.5 18.2 46.2 69.0 82.8 89.2 94.9
0.2 42.4 37.7 2.9 0.7 0.0 3.5
SUM (Rt –Tt) ^2
1 + 1/n*B
C^ –0.5
D*100
LogE
F*50
87.5
13.5
0.3
27.2
1.4
71.7
a. Reference = average percent calcium release from April 19, May 1, and May 15 (2012) dissolution results (total n = 18 vessels). b. Test = average percent calcium release from day 7 accelerated stability dissolution results (total n = 6 vessels).
Ashley M. Aimone Phillips, Stanley H. Zlotkin, Jo-Anna B. Baxter, and Daniel E. Roth are affiliated with the Centre for Global Child Health, Hospital for Sick Children, and the University of Toronto, Toronto, Ontario, Canada; Frank Martinuzzi and Tanush Kadria are affiliated with the Toronto Institute of Pharmaceutical Technology, Toronto. Please direct queries to the corresponding author: Daniel E. Roth, Department of Pediatrics, Hospital for Sick Children, 555 University Ave., Toronto, ON M5G1X8, Canada; e-mail: [email protected].
Conclusions. Based on in vitro criteria, the supplement may be a promising approach for delivering calcium, iron, and folic acid as a single daily dose to pregnant women in settings of low dietary intake of calcium.
Key words: Calcium, encapsulation, hypertension, powder, pregnancy
Introduction Hypertensive diseases of pregnancy, such as preeclampsia, are among the most important causes of maternal and perinatal morbidity and mortality worldwide [1]. The majority of hypertensive diseases of pregnancyrelated deaths occur in low- and middle-income countries where antenatal care is limited, particularly in communities that lack adequate access to emergency obstetric and neonatal intensive care [1]. The prevention of hypertensive diseases of pregnancy is therefore a major public health priority, particularly under resource-limited conditions. In populations where dietary calcium intake is low, randomized, controlled trials and meta-analyses have demonstrated that calcium supplementation during pregnancy significantly decreases the risk of hypertensive diseases of pregnancy [2–5]. A 2010 Cochrane review concluded that supplementation with at least 1 g of calcium reduced the risk of preeclampsia by 55% (95% CI, 0.31 to 0.65), gestational hypertension by 35% (95% CI, 0.53 to 0.81), maternal mortality or morbidity related to hypertensive diseases of pregnancy by 20% (95% CI, 0.65 to 0.97), and preterm birth by 24% (95% CI, 0.60 to 0.97), particularly in populations with low dietary calcium intake [2]. These estimates were updated in a 2012 review by Imdad and Bhutta, which also showed similar reductions in the risk of preeclampsia, maternal mortality or severe morbidity, and preterm birth, as well as an increase in birthweight (mean difference, 85 g; 95% CI, 37 to 133) [6]. Prenatal calcium supplementation has been identified as the
Food and Nutrition Bulletin, vol. 35, no. 2 © 2014, The Nevin Scrimshaw International Nutrition Foundation.
221
222
only intervention to prevent hypertensive diseases of pregnancy that may be feasibly delivered at the community level [3]. Although the policy is not yet widely implemented, the World Health Organization (WHO) recommended in 2011 a routine policy of prenatal supplementation with 1.5 to 2.0 g of elemental calcium per day, starting at around 20 weeks of gestation, for the prevention of preeclampsia in settings where usual dietary calcium intake is low [7]. In 2013, WHO reiterated the same recommendation in a statement focused specifically on calcium supplementation during pregnancy [8]. The WHO statement was further supported by the 2013 Lancet series on Maternal and Child Nutrition, which included calcium supplementation as one of 10 key evidence-based interventions that could significantly improve maternal (and child) health [9], despite very little evidence from effectiveness studies or programmatic settings. Although it is considered to be a simple, low-cost intervention for the prevention of hypertensive diseases of pregnancy, the effectiveness of routine prenatal calcium supplementation is hindered by several challenges: the bulk of supplemental calcium required to deliver the recommended 1.5 to 2 g per day; low adherence to conventional prenatal vitamin and mineral tablet regimens (typically composed of iron and folic acid), which may be further compromised by an increase in the number, size, or frequency of tablets; and intraintestinal inhibition of iron absorption by calcium when the nutrients are concurrently ingested. Calcium is generally assumed to interfere with iron absorption from dietary and supplemental sources in a dose-dependent and dose-saturable manner [10]. However, the mechanism of inhibition remains unclear, and there is ongoing debate regarding the extent to which various forms of supplemental and dietary calcium inhibit iron absorption [11–14]. WHO recommends iron and folic acid supplementation during pregnancy as part of routine antenatal care for the prevention of pregnancy-related anemia and neural tube defects, respectively [15]. The 2011 WHO guidelines on preeclampsia proposed “that concomitant administration of [iron and calcium supplements] should be avoided. Ideally, the two supplements should be administered several hours apart (e.g., morning and evening)” [7]. Furthermore, the 2013 WHO guidelines suggested that the daily dose of calcium should be “divided into three doses (preferably taken at mealtimes)” [8], based on the assumption that fractional calcium absorption declines substantially at doses greater than 500 mg [16]. From a practical standpoint, calcium tablets available on the market do not generally contain more than 600 mg of elemental calcium per unit. Therefore, adherence to current WHO recommendations would require four separate supplements to be consumed daily by pregnant women. We are not aware of published research that has
A. M. A. Phillips et al.
addressed the implications of the WHO recommendation for scale-up of both calcium and iron–folic acid (IFA) supplementation, proposals of feasible alternatives to the temporal separation of the two nutrients in routine antenatal care, or evaluations of potential technologies to enable the simultaneous delivery of iron and calcium in the same supplement. We therefore undertook the development of an innovative microencapsulated prenatal multimicronutrient (calcium, iron, and folic acid) supplement (SUPP) designed to facilitate immediate iron and folic acid release (in the stomach) and delayed calcium release (thereby minimizing intraluminal nutrient–nutrient interactions), and enable a single daily dose of all three nutrients. The objectives of this report are to describe the collaborative process involved in the development of the first prototype and to summarize the results of the in vitro proof-of-principle studies. Our early-stage product development experience provides an example of the challenges of moving an innovative dietary supplement from idea to proof of concept within an academic, investigator-initiated, grant-funded research setting.
Methods Granting process and partnership development
In July 2011, a research group led by investigators at the Hospital for Sick Children (“SickKids”), a public University of Toronto-affiliated teaching hospital in Toronto, Canada, was awarded a seed grant in the first round of the Saving Lives at Birth partnership, a collaborative funding initiative aimed at stimulating innovation of health technologies to reduce maternal and neonatal mortality in rural, resource-poor settings [17]. Cofunding to support the use of stable isotopes was obtained from the International Atomic Energy Agency. In December 2011, SickKids contracted with two private-sector pharmaceutical companies (referred to as company A and company B) to develop a product containing prescribed amounts of powdered calcium, iron, and folic acid, in accordance with the following key design criteria. Dissolution properties: the calcium was to be microencapsulated so as to promote its release in the mid to distal duodenum and thus avoid interactions with iron, which is predominantly absorbed in the proximal duodenum. Physical properties: the encapsulated calcium particle size had to be small enough to avoid disruption of the encapsulate (coating) by chewing, as well as minimize overall powder bulk. Organoleptic properties: the powder was to be tasteless and have a texture and stability suitable for mixing in water and fruit juices and, if feasible, hot beverages (e.g., tea). The contracted companies were not identified to each other, but both were informed that a second bidder was involved. The contracts
223
Development of a prenatal calcium and iron supplement
included time-limited confidentiality clauses to facilitate open exchange of ideas between SickKids and each of the contractors during the period of product development. SickKids and both contractors signed nonexclusive license agreements “to ensure global access to the [project] outputs should the Grantee, or the Grantee’s licensees, collaborators or partners, fail to make them widely available at a reasonable price for the benefit of people in Developing Markets” [18]. This condition would not impinge on intellectual property ownership or rights to exclusive licensure in North America or in other high-income countries. The budgets for the two contracts were negotiated separately; however, the deliverables for both companies were an encapsulated calcium, iron, and folic acid powder prototype and a detailed report of product specifications to be presented to the SickKids team within 6 months. The products were evaluated and compared based on specifications provided by the companies (in vitro dissolution tests, material and particle characteristics, product stability tests, and estimated production costs), as well as qualitative visual inspection and taste tests by SickKids investigators. Production of the supplement prototype
One prototype was selected to proceed to production of a clinical batch for use in initial clinical proof-ofconcept studies. The composition and preparation of the selected prototype are described below. Preparation of base calcium carbonate granules
Using a wet granulation method, very fine calcium carbonate powder (10 µm average particle size, Nutri Granulations, La Mirada, CA, USA) was granulated in a high-shear mixer (GEI Collette Gral 10, Wommelgem, Belgium) with chopper and agitator speeds of 1,600 and 400 rpm, respectively. During mixing, the following ingredients were added to achieve a total weight gain of 34.4% w/w: microcrystalline cellulose, PH101 (8.8% w/w), maltodextrin M200 (17.5% w/w), and sodium starch glycolate (8.1% w/w). The resulting wet massed granules were then dried at 45°C for 30 minutes in a fluid-bed dryer (Uni-Glatt, Glatt, Binzen, Germany) and screened for size using sieves (Ro-TAP, W. S. Tyler, Ohio, USA) with mesh sizes of 30 and 50. Granules were accepted if they passed through 30 mesh and were retained on 50 mesh, which corresponds to a particle size less than 600 µm and greater than 300 µm. The acceptable particle size range was considered optimal for maintaining coating efficiency, while reducing the risk of potential coating damage due to chewing. Application of subcoating and enteric coating to calcium carbonate granules
A 2.0% to 3.0% w/w subcoating, applied to the subset of granules that passed the granulation step, consisted
of a polyvinyl alcohol–polyethylene glycol copolymer (Kollicoat IR, BASF, Mississauga, ON, Canada), talc, and croscarmellose sodium (AC-DI-SOL). The subcoating was applied using a fluid-bed top spray coater (Uni-Glatt, Glatt, Binzen, Germany) with a spray rate of 2.5 mL/min (nozzle size 5), an atomizing pressure of 30 to 45 psi, and an inlet air temperature of 35° to 40°C. The resulting subcoated granules were again screened for size using mesh sizes 30 and 50. Accepted granules from the subcoating step were then coated with three layers of a pH-dependent polymethacrylic acid, methyl methacrylate copolymer enteric coating (Eudragit L 30 D-55, Evonik Industries AG, Darmstadt, Germany), which was formulated to begin dissolving at pH 5.8, and thus target mid-duodenal calcium release. Each coating layer consisted of Eudragit L 30 D-55, triethylcitrate, and talc. The first layer was neutralized to pH 5.5 and applied at the 25% w/w level. The second layer was neutralized to pH 4.5 and applied at the 15% w/w level. The third layer was applied at the same level as the second layer, but not neutralized. Similar to the methods described above, the enteric coatings were applied using a fluid-bed top spray coater (UNI, Glatt, Binzen, Germany) with a spray rate of 5 mL/min (nozzle size 8), an atomizing pressure of 30 psi, and an inlet air temperature of 35° to 40°C. The resulting enteric-coated granules were screened again for size using 30 and 50 mesh sizes. Addition of iron, folic acid, and suspending agents
The iron granules were encapsulated with a timedependent coating (Eudragit RL 30 D, Evonik Industries AG, Darmstadt, Germany), which included folic acid in the matrix and was formulated to target gastric release. The iron coating consisted of Eudragit RL 30 D, triethylcitrate, and talc. Each dose of calcium contained a fixed combination of 60 mg of elemental iron and 400 µg of folic acid to meet WHO recommendations [15]. As a final step, a suspension formulation including sorbitol (Neosorb P60W), xanthan gum, and citric acid anhydrous was added to each individual dose of calcium, iron, and folic acid. Adding these ingredients prevented the powder particles from settling to the bottom of the vessel when mixed with water. The final product was added to plastic vials (approximately 7.0 cm tall by 2.5 cm in diameter) with desiccant-filled caps (DesiCap, Multisorb Technologies, Buffalo, NY, USA). In-house simulation and taste test trials were used to determine the optimal composition of the suspension. For each simulation trial, a single packaged dose of the prototype supplement was shaken by inverting the vial three or four times, then dispensed into a clear glass or plastic drinking cup with approximately 125 mL of clean water and stirred with a plastic or metal spoon until the mixture was homogeneously dispersed and the solution had thickened. The suspension formulation was qualitatively evaluated based on
224
the following criteria: smooth consistency (minimal “clumping”), thickness appropriate for drinking, low residual adherence (the extent to which granules remained on the sides of the cup after decanting), short mixing time, mild to no flavor, odorless, and acceptable tongue feel (minimal “grittiness”). Dissolution testing
The dissolution performance of the enteric-coated calcium granules was assessed by sequentially exposing the coated granules to acidic media (USP 0.1N HCl buffer, pH 1.2) for 120 minutes then basic media (Na-citrate buffer, pH 5.8) for 90 minutes, using a Hanson SR-8 Plus dissolution tester (Hanson Research Corporation, Chatsworth, CA, USA), and following a modified United States Pharmacopeia (USP) basket apparatus methodology for delayed-release (entericcoated) dosage forms (USP 35–NF 30). As per USP standards, the paddle speed was set at 100 rpm, and all tests were performed at 37°C [19]. Modifications included wrapping the baskets tightly with a porous polypropylene material and replacing media throughout the base testing phase to maintain a pH of 5.8. The baskets were wrapped in order to reduce their mesh size and prevent leakage of test granules into the surrounding vessel, thus minimizing subsequent analytical errors (which were noted in preliminary tests). A 25-mL sample was taken from the test media at approximately 10-minute intervals starting at 120 minutes (end of the acid stage). These samples were filtered, and 20 mL of the filtrate was titrated with 0.01M sodium ethylenediaminetetraacetate (EDTA) using a modified USP titration method. The amount of CaCO3 was calculated based on the amount of EDTA required to reach the end point of titration, where 1 mole of CaCO3 (100 g) is equivalent to 1 mole of EDTA in solution form (372.8 g in 1,000 mL of water). The stability of the enteric coating was also assessed by performing dissolution tests after storing the coated granules at room temperature (20° to 25°C/35% to 40% relative humidity) for 4 weeks, and under accelerated conditions (40°C/75% relative humidity) for 12 weeks. The dissolution profile of enteric-coated calcium was considered acceptable if calcium release did not exceed 10% during the acid stage and reached at least 80% within 90 minutes of exposure to basic media. According to industry standards, the dissolution threshold for delayed-release or enteric-coated products should be ≤ 10% during a 2-hour acid stage and 75% to 80% for a 30- to 45-minute basic stage [19]. The percent release tolerance level and duration of a basic stage can be adjusted according to the solubility of the compound being tested. In the case of CaCO3, which has low solubility, the dissolution threshold of the basic phase was maintained at 80%, while the duration of the test was increased to 90 minutes. The encapsulated ferrous fumarate (and folic acid) granules also underwent
A. M. A. Phillips et al.
dissolution testing, although in acidic media only because the coating was formulated to promote gastric release. The dissolution profile of the iron and folic acid granules was considered acceptable if at least 90% iron and folic acid release was achieved within 30 minutes of exposure to acid media. Statistical methods
The results obtained from all dissolution and analytical test runs were summarized and expressed as mean and SD percent release for each specified time point. Dissolution profiles of coated calcium granules stored at room temperature (20° to 25°C/35% to 40% relative humidity) or under accelerated conditions (40°C/75% relative humidity) were compared at 7, 14, and 28 days. Mean percent release values were determined at each time point for the room (reference) and accelerated (test) conditions, then the mean dissolution values from both curves were used to calculate difference (f1) and similarity (f2) factors [20]. The difference factor (equation 1) is a measurement of the relative error between the two curves, and represents the percent (%) difference at each time point. The similarity factor (equation 2) is a logarithmic reciprocal square root transformation of the sum of squared error, and represents the similarity in percent dissolution between the two curves. Equation 1: f1 = {[∑t–1 n | Rt – Tt |]/[∑t–1 n Rt ] ∙ 100} Equation 2: f2 = 50 ∙ log{[1+(1/n) ∑t–1 n (Rt – Tt)2 ]–0.5 ∙ 100} where n is the number of time points, R is the dissolution value of the reference batch at time t, and Tt is the dissolution value of the test batch at time t. In order for curves to be considered similar, the f1 values should be < 15 (acceptability range, 0 to 15) and the f2 values should be > 50 (acceptability range, 50 to 100). All calculations were performed using Microsoft Excel 2007. Sample calculations of f1 and f2 are included in the appendix.
Results Two companies engaged in prototype development under contract from SickKids investigators from December 2011 to July 2012. The companies proposed encapsulated calcium powder supplement prototypes that differed in materials used, particle specifications, palatability, stability, and estimated production costs (table 1). The companies used similar coating approaches but different coating materials. Company A used an acetate phthalate-based polymer, while company B used a polymethacrylic acid, methyl methacrylate copolymer. Company B also applied a
225
Development of a prenatal calcium and iron supplement
subcoating before applying the enteric coating, which may have been partly responsible for observed differences in palatability and stability of the final product. The estimated costs of prototype development and production for the subsequent clinical study also differed substantially between the two companies. Cost was considered to be an important factor when planning for potential future large-scale production. Both companies used similar dissolution testing and in vitro analytical procedures. Despite slight variations in sampling times, the overall dissolution profile for prototype B was more acceptable than that for prototype A, since the release of calcium from prototype B remained below 10% during the acid phase and rose
above 80% during the basic phase, neither of which were demonstrated by prototype A (table 2). In fact, the coating of prototype A became very unstable and demonstrated poor dissolution performance in acidic conditions when retested approximately 2 weeks after application; the resulting calcium granules had unacceptable organoleptic properties (e.g., bitter taste), and were not considered a viable product for use in clinical studies. Company A proposed to refine their prototype upon provision of additional funds (beyond the original contract); however, this was not feasible within the scope of the grant. The SickKids investigators chose to continue to work with company B, which proceeded to finalize
TABLE 1. Enteric-coated calcium prototype characteristics produced by two private-sector pharmaceutical companies Item description
Company A
Particle size of granules Granule formulation Coating method Coating materials
150–250 µm Calcium carbonate, starch Fluid bed Hydroxypropyl methyl cellulose acetate phthalate (Shin-Etsu HP-55)
Coating weight Dissolution test method Analytical method(s) Palatabilitya Stabilityb
Company B
300–600 µm Calcium carbonate, HPMC E6 Fluid bed Subcoating: polyvinyl alcohol–polyethylene glycol (Kollicoat IR) Enteric coating: polymethacrylic acid, methyl methacrylate (Eudragit L 30 D-55) Enteric coating: 37.5% w/w Subcoating: 10% w/w Enteric coating: 25% w/w USP apparatus 1 (basket) with 0.1N Modified USP apparatus 1 (basket) with 0.1N hydroacetic acid and ammonia/ammochloric acid and sodium citrate buffers nium chloride buffers Titration with sodium EDTA Titration with sodium EDTA (titrant) (titrant) Unacceptable Acceptable Unstable at 7 days Stable at 56 days
EDTA, ethylenediaminetetraacetic acid; HPMC, hydroxypropylmethylcellulose a. Based on a qualitative assessment of taste, odor, and mouthfeel in taste tests of dry product. Considered acceptable if tasteless, odorless, and had minimal grittiness on tongue. b. Product was considered to be stable if dissolution test showed < 10% release within 60 minutes of exposure to acid medium after storage at room temperature (20° to 25°C/35% to 40% relative humidity) or accelerated conditions (40°C/75% relative humidity).
TABLE 2. Dissolution profiles of enteric-coated calcium powder prototypes developed by company A and company Ba % calcium release (SD) Dissolution phase and exposure time (min)
Prototype A (37.5% enteric coating weight)
Prototype B (25% enteric coating weight)
Acid phase (pH 1.0–1.2) 30 60 120
0.28 (2.1) 1.12 (1.0) 14.3 (3.7)
NV NV 5.3 (1.2)
Basic phase (pH 5.5–5.8) 10 30 60 120
54.8 (2.2) 72.2 (1.0) 83.5 (5.5) 87.4 (2.6)
18.0 (4.0) 68.5 (12.3) 88.0 (5.9) 95.1 (4.7)
NV, no value (sample was not drawn at specified time point) a. Percent calcium release values represent means (SD) of multiple samples.
226
A. M. A. Phillips et al.
the composition of the prototype supplement, including the incorporation of suspension agents and coated iron–folic acid granules (table 3). The dissolution profiles of the final coated and uncoated calcium carbonate and ferrous fumarate granules (fig. 1) were considered acceptable according to our criteria, indicating that the differential release of calcium and iron was consistently achieved within targeted acid-base environments. According to the similarity and difference factors (f1 and f2 ), calculated at each time sampling point, the dissolution profiles of coated calcium granules stored in accelerated conditions demonstrated some deterioration by 28 days, compared with those stored at room temperature (table 4).
Discussion Short-term contractual engagement of two private-sector pharmaceutical companies by an academic research institution resulted in the successful development of an oral combined calcium–iron–folic acid supplement prototype that met three main predefined acceptance criteria: targeted release of calcium at pH 5.5 to 5.8, and thus at a more basic pH relative to the release of iron and folic acid, based on in vitro dissolution tests; low likelihood of disruption of the enteric coating by chewing, due to sufficiently small particle size and the use of suspension agents to yield a liquid that facilitated the comfortable swallowing of intact particles; and acceptable palatability due to masking of the metallic taste of iron and the chalky texture of calcium. Several challenges were encountered in the development of a usable and palatable powdered supplement. The first was the sheer bulk of material required to deliver 1.5 g of elemental calcium (3.75 g of calcium carbonate) in addition to the multiple excipients (coating constituents as well as suspension agents). The final
TABLE 3. Composition of the final powdered supplement product, including enteric-coated calcium, encapsulated iron and folic acid, and suspending agents Amount (% total weight)
Component Calcium carbonate granules Calcium subcoating Calcium enteric coating Ferrous fumarate/folic acid granules Ferrous fumarate/folic acid coating Suspension agents
TABLE 4. Dissolution profile comparison of enteric-coated calcium carbonate granules stored in accelerated conditions (40°C/75% relative humidity) versus room conditions (20° to 25°C/35% to 40% relative humidity) No. of days stored in accelerated conditions
f1a
f2b
7
4.2
71.7
14
9.0
57.0
28
19.3
40.3
a. Acceptable range of difference factor (f1) is 0–15. b. Acceptable range of similarity factor (f2) is 50–100.
powder formulation could be readily mixed with water or fruit juice, yet at least 1 minute of vigorous stirring was required to produce an acceptable suspension. Multiple rounds of sampling and modification of the concentrations of suspension agents (sorbitol and xanthan gum) were required to arrive at the formulation that optimally suspended all of the calcium and iron–folic acid particles; the resulting fluid had a nectar consistency, with particles that remained visible as well as palpable in the mouth. Our initial intention was to design a product that would deliver the entire 1.5-g WHO-recommended daily intake of elemental calcium
Ferrous fumarate (coated)
Calcium carbonate (uncoated)
100 Ferrous fumarate (uncoated)
% release
80 60 Calcium carbonate (coated)
40 20 0
0
20
40
60
pH 1.2 for 2 h
80
100
34.21 0.67 29.52 1.67 0.12 33.81
120
140
Time (min)
160
180
200
220
240
pH 5.8 for 1.5 h
FIG 1. Dissolution profile of the final encapsulated calcium carbonate and ferrous fumarate granules
Development of a prenatal calcium and iron supplement
in one dose unit. However, we also produced dose units containing 500 mg and 1,000 mg of elemental calcium; these formulations have the advantage of less material per dose, but would have the disadvantage of requiring multiple daily dose regimens. A second challenge was the unexpected degree of technical difficulty involved in microencapsulating calcium carbonate, because of the relatively large granule size, coarse particle texture, and reactivity of the carbonate salt. Both companies found that their initial enteric coats were poorly protective against acid and disintegrated rapidly in ambient conditions, probably due to the coarse texture of the particles, which resulted in a porous coat that permitted moisture to enter and react with the carbonate ions. Company B successfully addressed this problem by improving the granulation process, applying a subcoat to create a more homogeneous particle surface prior to application of the enteric coating, as well as by increasing the thickness of the enteric coating itself. A third challenge was the implementation of a suitable analytical system for dissolution testing. The USP apparatus used in the current project had to be modified because the standard basket mesh dimensions were not suitable for the particle size range that we had specified as acceptable. A fourth challenge was the development of the coated iron–folic acid granule, since we had initially envisioned simply adding unmodified folic acid and off-the-shelf lipid-microencapsulated ferrous fumarate (as used in the original Sprinkles) to the powdered calcium mixture. However, we later recognized that this form of iron would not mix well in an aqueous solution, and also that the lipid coating would quickly disintegrate in any hot liquid, thereby failing to mask the metallic taste of the iron. To address this, company B proposed the use of a time-dependent coating, with planned release of the ferrous fumarate approximately 20 minutes after immersion in water or juice. The small dose of folic acid was readily integrated into the time-release coating. Although this aspect of product development was technically less difficult than designing the calcium coating, it nonetheless entailed unplanned time and financial costs. The final formulation was subsequently used to produce clinical batches for use in two studies that will be reported elsewhere: an acceptability and palatability study among pregnant women in Bangladesh, in which flavored and unflavored formulations of the supplement were compared with conventional and chewable tablets; and a dual stable isotope study of the fractional intestinal calcium absorption in pregnant women, the first step toward establishing biological proof-ofconcept in a clinical setting. The calcium absorption study has been conducted in Dhaka, Bangladesh, in collaboration with the International Centre for Diarrhoeal Diseases Research, Bangladesh (ICDDR,B) and the Texas Children’s Hospital. Based on the results of
227
the in vivo absorption of the enteric-coated calcium, a subsequent step may be a clinical trial to examine potential attenuating effects of enteric-coated calcium on iron absorption. Despite challenges in the development process, the new supplement prototype represents a potential approach for the combined delivery of daily supplemental calcium, iron, and folic acid to pregnant women. This novel delivery method was designed to support the implementation of the recent WHO recommendations for calcium supplementation (1.5 to 2 g per day) among women whose usual dietary calcium intake is low. Adherence to current WHO guidelines regarding prenatal iron and calcium supplementation, using products that are currently available on the market [8], would require a woman to take four separate tablets spaced throughout the day. However, it is well recognized that evidence-based guidelines face a range of barriers to their implementation [21]; we believe that the WHO prenatal calcium recommendation is unlikely to be widely adopted in low-income settings in the absence of a modified delivery vehicle and simplified schedule that combines calcium with other recommended micronutrients in a palatable and easily used format that can be produced and distributed at low cost. Our experience in developing a novel powdered micronutrient supplement may be relevant to other academic researchers engaged in early-stage collaborative efforts to develop innovative health technologies suitable for eventual scaling in low-income settings. Notably, productive public–private contractual partnerships were essential and were executed without compromising adherence to global access principles (including allowance for nonexclusive licensure in developing markets) or opportunities for intellectual property ownership or protection. We found that private-sector partners were as supportive as academic partners of a nonlinear, iterative process of experimentation and strategic risk-taking that is necessary for moving a technology across the so-called “valley of death” that separates conception and proof-ofconcept from commercialization [22]. We were initially concerned that our simultaneous engagement of two private-sector partners in a competitive process was risky. First, we worried that this approach might undermine the academic spirit and scientific aims of our collaborative relationships; however, we found that both companies were wholly comfortable with this approach. Second, we were uncertain whether a seemingly redundant subcontracting plan that is unconventional in the academic setting would be acceptable to the granting agency; however, Grand Challenges Canada (our direct funder for the Saving Lives at Birth partnership) was supportive of this approach from the outset. From the standpoint of achieving our research goals, this decision proved to be a very cost-effective risk mitigation strategy, particularly given that one of
228
the partners did not ultimately produce a prototype that met the design criteria within the agreed-upon budget and timeline. Small-grant or “seed” funding for investigators and entrepreneurs with innovative ideas for potential solutions to address global health problems has recently become a high-profile component of the global health research funding envelope (e.g., Bill & Melinda Gates Foundation Grand Challenge Explorations, Grand Challenges Canada Stars program). Arguably, these funding agencies have spurred researchers to spend more time thinking creatively about the path-to-scale of their research outcomes. Clearly, innovation cannot begin without the spark of an idea, but most ideas burn out early in the process of attempting to convert the concept into an efficacious and commercially viable product [23]. In this paper, we describe the early stages of development of a new once-daily prenatal micronutrient powder combining calcium, iron, and folic acid for the prevention of hypertensive diseases of pregnancy and maternal anemia in resource-poor settings. Successful development of an innovative prototype was enabled by seed funding from a nontraditional source of academic research funding (the Saving Lives at Birth program) and strategic engagement with two private sector partners. Based on a set of pre-established criteria upon which candidate prototypes were evaluated, in vitro and qualitative acceptability assessment suggested that the selected product was a promising approach for the delivery of iron, folic acid, and calcium to pregnant women. Nonetheless, a series of additional criteria (e.g., proof of concept with respect to iron absorption, clinical efficacy, cost-effectiveness, long-term stability) will
A. M. A. Phillips et al.
need to be met to continue to progress along the long road to implementation and scale-up.
Authors’ contributions Daniel E. Roth and Stanley H. Zlotkin designed the research; Frank Martinuzzi and Tanush Kadria produced the supplement and analyzed the data; Ashley M. Aimone Phillips and Jo-Anna B. Baxter wrote the first draft of the paper; all authors reviewed and approved the final manuscript.
Acknowledgments This work was made possible through the generous support of the Saving Lives at Birth partners: Grand Challenges Canada, the US Agency for International Development, the Government of Norway, the Bill & Melinda Gates Foundation, and the World Bank. The manuscript was prepared by the authors and does not necessarily reflect the views of the Saving Lives at Birth partners. The work was also supported in part by additional funding from the International Atomic Energy Agency. We would also like to acknowledge Diego Bassani, Brendon Pezzack, Nandita Perumal, Sohana Shafique, Elaine Gergolas (SickKids); Steve Abrams, Keli Hawthorne (Baylor College of Medicine); Abdullah al Mahmud, Munirul Islam, and Tahmeed Ahmed (ICDDR,B) for their contributions to the design concept and potential delivery applications of the micronutrient product.
References 1. Duley L. The global impact of pre-eclampsia and eclampsia. Semin Perinatol 2009;33:130–7. 2. Hofmeyr GJ, Lawrie TA, Atallah AN, Duley L. Calcium supplementation during pregnancy for preventing hypertensive disorders and related problems. Cochrane Database Syst Rev 2010:CD001059. 3. Imdad A, Jabeen A, Bhutta ZA. Role of calcium supplementation during pregnancy in reducing risk of developing gestational hypertensive disorders: a metaanalysis of studies from developing countries. BMC Public Health 2011;11(suppl 3):S18. 4. Villar J, Abdel-Aleem H, Merialdi M, Mathai M, Ali MM, Zavaleta N, Purwar M, Hofmeyr J, Nguyen TN, Campódonico L, Landoulsi S, Carroli G, Lindheimer M, Group WHOCSftPoPT. World Health Organization randomized trial of calcium supplementation among low calcium intake pregnant women. Am J Obstet Gynecol 2006;194:639–49. 5. Kumar A, Devi SG, Batra S, Singh C, Shukla DK. Calcium supplementation for the prevention of preeclampsia. Int J Gynaecol Obstet 2009;104:32–6. 6. Imdad A, Bhutta ZA. Effects of calcium supplementation
7. 8. 9.
10. 11.
12.
during pregnancy on maternal, fetal and birth outcomes. Paediatr Perinat Epidemiol 2012;26(suppl 1):138–52. World Health Organization. Recommendations for prevention and treatment of pre-eclampsia and eclampsia. Geneva: WHO, 2011. World Health Organization. Guideline: Calcium supplementation in pregnant women. Geneva: WHO, 2013. Bhutta ZA, Das JK, Rizvi A, Gaffey MF, Walker N, Horton S, Webb P, Lartey A, Black RE; Lancet Nutrition Interventions Review Group; Maternal and Child Nutrition Study Group. Evidence-based interventions for improvement of maternal and child nutrition: What can be done and at what cost? Lancet 2013;382:452–77. Hallberg L, Rossander-Hulthen L, Brune M, Gleerup A. Inhibition of haem-iron absorption in man by calcium. Br J Nutr 1992;69:533–40. Dawson-Hughes B, Seligson FH, Hughes VA. Effects of calcium carbonate and hydroxyapatite on zinc and iron retention in postmenopausal women. Am J Clin Nutr 1986;44:83–8. Cook JD, Dassenko SA, Whittaker P. Calcium supplementation: effect on iron absorption. Am J Clin Nutr
229
Development of a prenatal calcium and iron supplement
1991;53:106–11. 13. Hallberg L, Brune M, Erlandsson M, Sandberg AS, Rossander-Hultén L. Calcium: effect of different amounts on nonheme- and heme-iron absorption in humans. Am J Clin Nutr 1991;53:112–9. 14. Gaitán D, Flores S, Saavedra P, Miranda C, Olivares M, Arredondo M, López de Romaña D, Lönnerdal B, Pizarro F. Calcium does not inhibit the absorption of 5 milligrams of nonheme or heme iron at doses less than 800 milligrams in nonpregnant women. J Nutr 2011;141:1652–6. 15. World Health Organization/Food and Agriculture Organization. Vitamin and mineral requirements in human nutrition: report of a joint FAO/WHO Expert Consultation, Bangkok, Thailand, 21–30 September 1998. 2nd ed, 2004. Available at: http://whqlibdoc.who. int/publications/2004/9241546123.pdf. Accessed 9 March 2014. 16. Heaney RP, Weaver CM, Fitzsimmons ML. Influence of calcium load on absorption fraction. J Bone Miner Res 1990;5:1135–8.
17. Grand Challenges Canada. Saving lives at birth: a grand challenge for development. Available at: http://savinglivesatbirth.net/. Accessed 9 March 2014. 18. Grand Challenges Canada Website. Available at: http:// www.grandchallenges.ca/. Accessed 9 March 2014. 19. United States Pharmacopeia/National Formulary. Delayed release tablets. In: Convention USP, ed. United States Pharmacopeia, 35th ed, and National Formulary, 30th ed. Rockville, MD, USA: The United States Pharmacopeial Convention, 2012. 20. Moore JW, Flanner HH. Mathematical comparison of dissolution profiles. Pharm Technol 1996;20:64–74. 21. Haines A, Kuruvilla S, Borchert M. Bridging the implementation gap between knowledge and action for health. Bull World Health Organ 2004;82:724–31; discussion 32. 22. Grand Challenges Canada/Grand Defis Canada. Integrated innovation. Toronto, ON, Canada: Grand Challenges Canada, September 2010. 23. Coller BS, Califf RM. Traversing the valley of death: a guide to assessing prospects for translational success. Science 2009;1:1–5.
Appendix Sample calculation of f1 REFa
TESTb
ABS (Rt – Tt)
SUM (Rt – Tt)
SUM ( Rt)
B/C
D*100
5.9 24.7 52.3 70.8
5.5 18.2 46.2 69.0
0.4 6.5 6.1 1.7
17.7
417.7
0.0
4.2
82.0
82.8
0.9
89.0 93.0
89.2 94.9
0.2 1.9
a. Reference = average percent calcium release from April 19, May 1, and May 15 (2012) dissolution results (total n = 18 vessels). b. Test = average percent calcium release from day 7 accelerated stability dissolution results (total n = 6 vessels).
Sample calculation of f2 REFa
TESTb
(Rt –Tt) ^2
5.9 24.7 52.3 70.8 82.0 89.0 93.0
5.5 18.2 46.2 69.0 82.8 89.2 94.9
0.2 42.4 37.7 2.9 0.7 0.0 3.5
SUM (Rt –Tt) ^2
1 + 1/n*B
C^ –0.5
D*100
LogE
F*50
87.5
13.5
0.3
27.2
1.4
71.7
a. Reference = average percent calcium release from April 19, May 1, and May 15 (2012) dissolution results (total n = 18 vessels). b. Test = average percent calcium release from day 7 accelerated stability dissolution results (total n = 6 vessels).
