Abstract
The general compatibility between various polymers (plastic and elastomeric materials) commonly used in medical devices (e.g., syringes) and common solution-based pharmaceutical products was examined. Processes affecting material/solution compatibility that were considered included drug binding and material leaching. Considering binding, material/water equilibrium binding constants (Eb) were determined for 14 organic solutes and the test materials. Correlations between the measured binding constants and the organic solute's octanol/water and hexane/water partition coefficients were obtained. In general, the elastomeric materials behaved similarly and interacted with acidic compounds differently than with neutral compounds. Alternatively, the plastic materials typically used in device components such as syringe barrels were relatively inert in the sense that they only minimally bound even the most lipophilc compounds. Considering leaching, both organic entities and trace elements/metals were extracted from the elastomers over the extraction pH range investigated. The plastic materials generally had lesser quantities of organic leachables than did the elastomers and contained little, if any, extractable trace elements/metals.
LAY ABSTRACT: Polymeric materials are commonly used in medical devices such as syringes. The plastic materials may interact with drug products contained within the device, potentially affecting the quality of the drug products. These interactions may include leaching, which is the migration of entities out of the material and into the drug product, and binding, which is the migration of substances out of the drug product and into the material. This paper examines the magnitude of leaching and binding for several materials that can be used in syringe parts such as the syringe barrel, plunger, and tip cap.
Introduction
Polymers, including plastic and elastomeric materials, are widely used in medical items, such as solution containers, transfusion sets, transfer tubing, and devices (e.g., syringes, vials). The physiochemical nature of these polymers provides medical products with their necessary and desirable performance characteristics. While an important performance characteristic of polymers used in medical application is chemical inertness, interactions between a polymer and a contacted pharmaceutical product are well documented (1⇓⇓–4). Such interactions may include sorption (or binding), the uptake of product components by the polymer, or leaching, the release of polymer components to the product. Because both sorption and leaching can materially affect product safety and efficacy, it is a necessary part of storage/delivery system development to establish a particular system's propensity to interact with the medical product.
Solute (e.g., drug) binding can be modeled using fundamental properties of the drug and the contacting material. Such a thermodynamic model establishes the maximal equilibrium binding of the solute and thus delineates the potential that the product/material couple will fail from an efficacy perspective. Solute binding by a material is essentially a partition-mediated process, and the equilibrium distribution of a solute between a solution and material phase can be expressed as the equilibrium interaction constant, Eb. To establish a material's binding model, an organic solvent system that mimics the binding behavior of the material is defined. For example, the octanol/water system has been proposed as a surrogate for polymers. Octanol/water is chosen due to the ready availability of the octanol/water partition coefficient (Po/w). If such a model is appropriate, then a linear relationship can be established between the polymer/water and octanol/water partition coefficients (4⇓⇓⇓⇓–9):
It has been demonstrated that a coupled model, using dissimilar solvents such as hexane/water and octanol/water, provides a better fit for a wider range of polymers than does a model based on a single solvent system (10⇓⇓–13). Such an expression takes the form:
The slopes and intercept for this relationship can be obtained from a data set generated via model compounds. Specifically, if a compound's log Eb can be determined and the compound's partition coefficients can be obtained (typically available in the chemical literature), Eb values for a set of model compounds can be regressed versus their log (Po/w) and log (Ph/w) values and the model parameters obtained.
In the case of leaching, numerous investigators have documented the accumulation of material-derived substances in solutions contacted by polymeric materials (14⇓⇓⇓⇓⇓⇓⇓⇓–23). Both the identity of the leached substances and their accumulation levels may affect the ultimate utility of the product. While in general the issue of leachables is primarily driven by concerns over product safety (i.e., the safety impact of the leached substances), leachables can and do influence other aspects of product function and quality.
In this study, we examined, in general terms, the compatibility of several polymeric and elastomeric materials with solution-based pharmaceutical products. This assessment included the development of binding models for the evaluated materials as well as the general chemical characterization of material extracts obtained over a wide range of extracting solution pH. The materials examined are commonly utilized in medical devices such as syringes, vials, et cetera.
Materials and Methods
Tested Materials
The polymeric and elastomeric materials used in this study were obtained from ready-to-use devises (or device components) from appropriate commercial vendors. The specific materials tested included four materials representative of materials utilized in syringe barrels (B1 = polypropylene, B2 = Type I glass, B3 = cyclic olefin polymer, and B4 = cyclic olefin copolymer) and six elastomers (numbered as E1 through E6; E3 was Teflon-coated) which could be used as syringe plungers and/or tip caps. For the sake of simplicity, the four materials representative of syringe barrels will be referred to as plastics, even though material B2 is actually glass.
Binding Assessment
Contact Conditions:
The binding assessment involved generating a binding model for each of the test materials. The test materials were contacted with aqueous donor solutions containing known amounts of model organic compounds. These model organic compounds were ones whose theoretical partitioning behavior, as expressed by classical octanol-water and hexane-water partition coefficients, are well known and which thus serve as analytically expedient surrogates for ingredients of pharmaceutical formulations. The materials and donor solutions were equilibrated, and the equilibrium concentration of the models in the donor solution was analytically measured. Differences between the initial prepared and final equilibrium concentrations of the models in the donor solution were utilized to calculate interaction constants. An interaction model was generated for each tested material by correlating the model's partition coefficients with the interaction constants for each material. These correlations were then used to establish the relative binding characteristics of the materials evaluated for any organic compound whose partition coefficients are either known or can be measured.
The donor experiment was performed at a temperature of approximately 40 °C for a period of 7 to 13 days. Equilibrium is typically established under such conditions. The donor experiment was carried out in such a way that the material contact surface area was maximized. Specifically, approximately 8 g of material was contacted with 75 mL of the donor solutions containing approximately 10 mg/L of the model compounds. Duplicate test articles were prepared for each material/donor solution couple. The donor solutions were acidified to ensure that the ionizable solutes were in their protonated (neutral) form. The suite of model compounds used is summarized in Table I. The individual model compounds, as well as all analytical reagents, were obtained commercially (e.g., Aldrich Chemical, Milwaukee, WI) as substances of known, high purity.
Model Compounds and Their Associated Partition Coefficients
Analytical Methods:
The analytical methods employed for donor solution analysis were gradient reversed-phase high-performance liquid chromatography (HPLC) separations coupled with UV detection. Operating conditions for these HPLC methods are summarized in Tables IIA⇓ through IIC. The HPLC analyses were performed with Agilent (Wilmington, DE) model 1100 chromatography systems.
A. Operating Conditions, HPLC Method For Quantifying Group 1 Analytes1
B. Operating Conditions, HPLC Method For Quantifying Group 2 Analytes1
C. Operating Conditions, HPLC Method For Quantifying Group 3 and 4 Analytes1
Quantitation of the model compounds in the donor solution was accomplished as follows. Standards for the target analytes were prepared at two levels, approximately equal to the initial analyte levels in the donor solution and a factor of 10 lower than this (≈10 and 1 mg/L, respectively). For those samples whose analyte response was greater than the mean response of the lowest standard, the analyte concentrations were determined by putting the sample's peak area into a linear calibration curve (standard area versus standard concentration). For those analytes whose response was lower than the mean response of the lowest concentration standard, the analyte concentrations were determined using a single point response factor:
Mathematical Analysis, Equilibrium Interaction Constant, Eb:
The equilibrium interaction constant, Eb, is essentially a partition coefficient describing the equilibrium distribution of a solute between a material phase and a solution phase:
In a binding study as performed herein, Cs is measured as the equilibrium concentration of the analyte remaining in the donor solution. While Cp cannot be measured directly, it can be calculated from the experimental parameters and Ci, the initial analyte concentration in the donor solution as follows. The amount of the solute bound by the material (Am) is calculated from Ci, Cs, and the donor solution volume (Vs) as:
The concentration of solute in the material phase (Cp) is Am divided by the weight of material used (Wm):
While having both Cp and Cs should allow for a direct calculation of Eb, there is the issue of compatible concentration units. Specifically, Cp is a mass concentration [for example mg/kg (ppm)] while Cs is a solution concentration. If the density (D) of the solution can be measured (or assumed to be 1 g/mL), this difference can be reconciled.
The overall equation for calculating Eb is as follows, illustrating units and conversions.
Mathematical Analysis, Interaction Model:
The interaction model takes the form:
The slopes and intercept for this relationship were obtained by performing multivariable linear regression analysis of the data set generated in this study for the model compounds. Specifically, as the compound's log Eb was determined in this study and the compound's partition coefficients were obtained from the chemical literature, Eb values for a set of model compounds were regressed versus the compound's log (Po/w) and log (Ph/w) values and the model (curve fit) parameters obtained.
As the model compounds span nearly four orders of magnitude in terms of their partition coefficients, it is reasonable to anticipate that no single contact condition was most appropriate for all the model compounds. Thus the contact conditions described previously are most appropriately thought of as screening conditions, capable of producing good Eb values for many model compounds but not effective for all the models. If the model compound is only minimally bound by the material under examination, then the concentrations of the control and equilibrated donor solution will be very similar. As the Eb is calculated as the difference between these two numbers, their similarity means that a calculated Eb will have a large associated error (i.e., subtracting two big numbers to get a small number). In such cases the error in Eb may be so large that the Eb is not appropriate for use in developing the binding model. The appropriate course of action in this circumstance is to repeat the interaction experiment, increasing the amount of material contacted per unit volume of donor solution. Doing so will increase the amount of the compound bound by the material and will increase the difference in model compound concentration in the control and tests articles.
The opposite situation is encountered by a strongly bound model compound. The concentration of this compound in the donor solution after equilibration with the material may be so small (because most of it is bound) that it is below the method's quantitation limit. In this circumstance, a more accurate Eb is obtained if the amount of material per unit volume of donor solution is significantly decreased. Adopting this strategy decreases the amount of solute bound by the material, thereby increasing the concentration of the compound remaining in the donor.
As a general rule of thumb, the following conventions are used to assess the accuracy of the calculated Eb:
(A) If the concentration of the compound in the equilibrated donor solution is more than 98% of the concentration of the compound in the control solution, then the Eb calculated is not reliable and the experiment should be repeated with more material (or less donor solution). This requirement is based on the anticipated precision of the HPLC methods used for compound quantitation.
(B) If the concentration of the compound in the equilibrated donor solution is less than 1% of the concentration of the compound in the control solution, then the Eb calculated is not reliable and the experiment should be repeated with less material (or more donor solution). This requirement is based on the sensitivity and linear range of the analytical method. For a compound in the control solution at a level of 10 ppm, this requirement is triggered if the compound's level in the donor is less than 0.1 ppm (100 ppb). In many cases, 100 ppb is near the method's quantitation limit and well outside the standard range.
In general, the most accurate estimation of Eb is obtained if the compound's concentration in the equilibrated donor is between 20% and 80% of the compound's concentration in the control article.
Comparative Plots of Material Binding Characteristics:
While the binding models are very effective in terms of portraying the binding characteristics of individual materials, they are a poor means for comparing one material to another because the x-axis of each binding model is different for each material. However, appropriate comparisons can be obtained if the binding constants for one material are plotted versus those of a second material. If the binding characteristics of the two materials are the same, then such a plot will have a unit slope and a zero intercept. If one material's binding characteristics are more greatly influenced by the chemical nature of the compound being bound, then the plot will have a non-unit slope. If the “intrinsic binding capacity” of one material is different from that of another, the intercept will be non-zero.
Extraction Assessment
Extraction:
Each test material was extracted with three solutions, pH 2 (0.01N HCl), neutral (water, unbuffered) and pH 12 (0.01N NaOH). Approximately 7 g of each test sample was added to 100 mL of extracting solution. Duplicate extracting samples were prepared for each material and each extracting solution. In order to simulate some unspecified duration of anticipated product use (long-term storage at room temperatures), the solutions were autoclaved at ≈121 °C for 1 h. The extractions were performed in Pyrex|Pr glass bottles with Teflon-lined, inert plastic caps. Extraction controls were generated by autoclaving extraction vessels that contained only the extracting solution.
Extract Analysis:
The extracts and controls were analyzed for pH, UV absorbance, total organic carbon (TOC) level (as a measure of total organic leachables) and for selected trace elements and metals. The TOC measurements were performed with an OI Analytical (College Station, TX) model 700 Total Organic Carbon Analyzer. The trace element/metals analyses were performed via inductively coupled plasma atomic emission spectroscopy (ICPAES) using a Varian (Walnut Creek, CA) Liberty model 220 ICP spectrometer. The ICP spectrometer was operated with a V-groove nebulizer and a cyclonic, double-pass glass spray chamber.
Several of the high-pH extracts from the elastomeric materials developed particulate matter. The particulate was collected by filtering portions of the extracts through 0.8 micron mixed cellulose ester filter membranes. The collected material was visually examined via a stereomicroscope to assess its gross morphology and then was removed from the filter with a tungsten probe. A portion of the removed material was mounted on a KBr disk for Fourier transform infrared (FT-IR) analysis, and another portion was mounted on a carbon support with carbon paint for analysis by energy-dispersive x-ray spectroscopy (EDXS).
A Digilab (Randolph, MA) FTS 7000 series infrared spectrometer system equipped with a Universal Microspectrometer Accessory (UMA) 600 was used for the FT-IR analyses. The UMA 600 contained a liquid nitrogen–cooled MCT (mercury cadmium telluride) detector and a KBr beamsplitter. The IR spectra were acquired in transmission using a 2 cm−1 internal aperture, a speed of 20 KHz, a resolution of 8 cm−1, and 64 co-added scans. A Hitachi Hitachi S-3500N variable pressure scanning electron microscope (SEM) was used in high vacuum, yielding a secondary electron image. The accelerating voltage used to acquire x-ray spectra was 30 Kev. A Noran Vantage energy dispersive x-ray spectrometer system was used in conjunction with the SEM. The deadtime was set at ≈30% and x-rays were acquired for 100 s. All elements reported produced k alpha x-rays lines. Calclium was the only element to produce a k beta x-ray line.
Results and Discussion
Binding Models, Elastomeric Materials
The measured log Eb values are summarized in the Table III. It is clear from these data that the elastomers examined in this study are similar in binding characteristics and bind solutes to a much greater extent than do the plastic materials. An example of a typical binding model for an elastomeric material, the model obtained for material E6 is shown in Figure 1. While it is immediately clear that the binding model is a poor fit for the behavior of all the model solutes, it is also the case that the binding data appears to have a bimodal distribution. Upon closer examination, it was observed that the model solutes could be divided into two groups based on whether the solute was an acid or a neutral. Such a bimodal distribution of model solute binding has been reported in the literature (9⇓–11, 13). Independent binding models (for material E6) are characterized by excellent fits, as is illustrated in Figure 2. It is interesting to note that the binding models for the acids closely follows a hexane/water model, while the behavior of the neutrals is reflected in a combined octanol/water and hexane/water model. While this behavior clearly reflects the differing mechanics of the solute/material interaction, it is beyond the scope of this study to interpret these mechanistic implications. Similar behavior was exhibited by all the elastomeric materials. Curve fit parameters for the binding models (acids and neutral solutes) derived from this data are summarized in Table IV.
Typical binding model, elastomer material E6. While the fit for the binding model is poor for all model solutes, the distribution of the data suggests that the there are two unique analyte populations in this data.
Typical binding models, elastomer material E6. Dividing the model solute population into two compound types, acids and neutrals, produces excellent linear relationships between the log of the binding constant and the linear combination of the logs of the two partition coefficients. While the behavior of the acids closely follows a hexane/water model, the behavior of the neutrals is reflected in a combined octanol/water and hexane/water model.
Equilibrium Binding Constants Obtained for Each Model Compound and Each Material
Summary of Binding Models1,2
Ultimately the binding data can be used to rank-order the materials in terms of their ability to bind solutes. In general, the slopes and intercepts of the binding models provide broad insights into the relative binding properties of the various materials studied. In its simplest sense, the slope of the models reflects the material's “sensitivity” to solute properties. Similarly, the intercept reflects the “intrinsic binding power” of the material. Thus two materials with similar slopes but different intercepts (co-linear binding models) have binding properties that are different by the same magnitude for all compounds. An example of this is the models for all the elastomers. As noted from Table IV, the slope1 and slope2 values for these materials are very similar (possibly the same given the experimental uncertainty) but the intercepts are somewhat different. It is the nature of the binding models that the larger (less negative) the intercept, the greater the “intrinsic” binding.
While the binding models themselves may provide a means of comparing the binding properties of several materials, the differences in the binding models for the materials in this study are too small to facilitate such a comparison. Additionally, while the binding models are very effective in terms of portraying the binding characteristics of individual materials, they are a poor means for comparing one material to another because the x-axis of each binding model is unique to that material. Appropriate comparisons can be obtained if the binding constants for one material are plotted versus those of a second material. This is equivalent to defining the following mathematical relationship:
If the binding characteristics of the two materials are the same, then such a plot will have a unit slope and a zero intercept. If one material's binding characteristics are more greatly influenced by the chemical nature of the compound being bound, then the plot will have a non-unit slope. If the “intrinsic binding capacity” of one material is different from that of another, the intercept will be non-zero.
A simultaneous comparison of the binding properties of several materials is facilitated if these properties are “harmonized” relative to a “standard” material. In the case of the elastomers, the E6 material was chosen, more or less arbitrarily, as the “standard” material.
Figure 3 is such a comparison plot, relevant for the interaction characteristics of the elastomers with acidic solutes. Interpretation of such diagrams is relatively straightforward. A material that behaves exactly like the standard material will have a line with an intercept of zero (goes through the origin), a slope of 1, and a coefficient of determination (R2) of 1.0. This relationship is shown in the figures by a line comparing the E6 material to itself. A material with a slope greater than 1 is more “sensitive” to a solute's chemical nature than is the standard material, while the opposite is true if the slope is less than 1. The material with a higher intrinsic binding capacity will have a positive intercept, while a material that has less intrinsic binding will have a negative intercept.
Comparison of binding properties, several elastomers versus material E6, acidic compounds.
As shown in Figure 3, the binding behavior of materials E2, E4, and E6 are very similar. Material E5 is very different in terms of its interaction with acids, possessing a much lower intrinsic binding capacity (but similar binding sensitivity) than the other materials. Material E1 has a somewhat higher intrinsic binding capacity than does material E6. Material E3 acts very differently than the other elastomeric materials, with a different solute sensitivity (slope) but similar intrinsic binding capacity to the other elastomers.
Thus from the perspective of acidic solutes, the elastomer materials can be ranked in the following order (least binding to most binding):
An analysis of the binding data for the neutral solutes reveals that materials E2, E4, E5, and E6 are all very similar with respect to their binding of neutral solutes. While material E1 and material E3 have a similar sensitivity to neutral solute lipophilicity as the other materials (similar slopes), material E3 has a lower intrinsic binding capacity than does material E6 (negative intercept) while material E1 has a higher intrinsic binding capacity (positive intercept) versus material E6. Thus from the perspective of neutral solutes, the elastomer materials can be ranked in the following order (least binding to most binding):
It is noted that materials E2 and E6 are essentially identical in their binding characteristics.
A simpler picture of the relative binding characteristics of the elastomeric materials can be obtained by considering the log Eb values obtained for the most strongly bound acid (BBA) and neutral (DBP) model solute. Such a comparison is shown in Figure 4. From this figure, one obtains a general rank ordering of materials (with respect to their binding characteristics, least binding to most binding) of
which mirrors the conclusions drawn previously from the comparative plots.
Comparison of the elastomer's binding of the most lipophillic neutral (DBP) and acidic (BBA) model solutes.
Binding Models, Plastic Materials
Three of the plastic materials studied exhibited similar and minimal solute binding. In fact, discernible binding for three materials (materials B2, B3, and B4) was only observed for the most lipophilic model solutes. Most of the model solutes exhibited negligible binding with these materials (under the experimental conditions employed), and thus it was not meaningful to generate binding models or comparative plots for these materials.
The fourth plastic material (B1) exhibited a much greater affinity for the model solutes than did the other three plastic materials, although its binding was still considerably less than that of the elastomers. The binding properties of the plastic materials for the most lipophilic model solutes are illustrated in Figure 5. It is clear that a rank ordering of the plastic materials (least binding to most binding) is as follows:
Comparison of the plastic material's binding of the most lipophilic neutral (DBP) and acidic (BBA) model solutes.
Utilization of the Binding Models
To illustrate the utility of the binding models, consider the following example. A drug-containing formulation (3 mL) is packaged in a syringe system that consists of a 1.0 g plastic syringe barrel and a 0.2 g elastomeric plunger. The drug, a neutral compound, has a log Po/w of 1.5 and a log Ph/w of 0.8 (similar to the model solute DMP). In this example, the syringe barrel is material B1 and the elastomeric plunger is made from material E4. The question to be answered is “is this product configuration characterized by an acceptable level of drug binding (less than 10%)?”
The answer to this question is obtained as follows. The relationship between the fractional binding (Fb) and a solute's Eb takes the form
where Wc = the material weight (in kg) and Vs is the solution volume (in liters). The Eb values are obtained by putting the log Po/w values into the material's respective binding models from Table IV, resulting in an Eb = 0.915 for the syringe barrel (material B1) and 0.741 for the plunger (material E4) respectively. Substituting these values into the Fb equation produces
Thus while the elastomer is a suitable component for this application, drug loss to the B1 syringe barrel would be excessive.
Results and Discussion, Potentially Reactive Extractables
pH:
The pH values for the extracts and controls (extraction blanks) are compiled in Table V. For the extracts and controls at the pH extremes, (pH 2 and pH 12), there was no pH change arising from the extraction. This result is desirable, as it means that the pH effect was maintained throughout the extraction. Because the neutral pH samples were not buffered, any difference in their pH (extract versus control) would reflect the leaching of acidic or basic compounds. As shown in Table V, there was little or no difference in pH across material type and between extracts versus controls. This data implies that the level of acids or bases extracted from any of the material types tested was small.
pH Values of Test Articles
Total Organic Carbon (TOC):
The total amount of material-derived (extracted) organic carbon associated with the various tested materials is summarized in Table VI and Figure 6. The following observations are pertinent.
Extracted Total Organic Carbon (TOC, mg/L)1
Profiles of extracted TOC for plastic materials (as potential syringe barrels) and elastomer materials.
Three of the plastic materials have only low amounts of extractable TOC (B2, B3, and B4). Of these, the extracted TOC levels for B3 and B4 are not affected by the pH of the extracting medium. Alternatively, the amount of TOC extracted from material B2 increased significantly at pH 12.
Extracts of material B1 had a much higher TOC than the other plastic materials.
Thus the rank ordering of the plastic materials (least to most extracted TOC) is
All the elastomeric materials had an extracted TOC that was similar at pH 2 and pH neutral but which increased significantly at pH 12. At each extract pH, the levels of extracted TOC from materials E2, E4, E5, and E6 materials were all roughly the same. While material E3 had levels of extracted TOC that were similar to those for these other materials at pH 2 and neutral, this material had a lower extracted TOC level than the others at pH 12.
Material E1 had much higher levels of extracted TOC than did the other elastomers. While its level of extracted TOC also increased at pH 12, the magnitude of the increase was proportionally not as great as the increases observed for the other materials.
Thus the rank ordering of the elastomers (least to most extracted TOC) is
UV Absorbance Characteristics of the Extracts:
The net UV absorbance attributable to extracted compounds [UV absorbance in extract − UV absorbance in controls (blanks)] is summarized in Figures 7 and 8. The UV results for the plastic materials mirror the extracted TOC results shown in Figure 6, with the extract from material B1 exhibiting the highest UV absorbance at all wavelengths. The trend in UV absorbance as a function of wavelength (specifically, the measurable absorbance at wavelengths of 250 and 280 nm) suggest that the B1 material may have conjugated or aromatic extractables. The absorbance values associated with the other plastic materials are very low and thus provide no insight into the nature of the organic compounds extracted from these materials.
UV absorbance profiles for the plastic materials.
UV absorbance profiles for the elastomer materials.
The UV results for the elastomers mirror the extracted TOC results shown in Figure 6, with extract E1 exhibiting the highest UV absorbance at all wavelengths. The UV absorbance of the extracts for all elastomers increases at pH 12. The trend in UV absorbance as a function of wavelength (specifically, the measurable absorbance at wavelengths of 250 and 280 nm), suggests that the elastomer materials may have conjugated or aromatic extractables, especially in the case of material E1.
Trace Elements/Metals:
The trace element/metals analysis by ICPAES included 25 elements. The 12 trace elements that were not present in any sample at levels above their lowest quantity determinable (LQD) are summarized in Table VIIA. Seven additional elements (Al, Cd, Fe, Mn, Na, Pb, and Ti) were present at levels above the LQD in some samples but were not reproducibly present in extracts at levels greater than 2 times their levels in the controls. Thus it is concluded that these 19 trace elements were not extracted from the materials in quantifiable quantities. The concentration of trace elements extracted from the plastic and elastomer materials in quantifiable amounts are summarized in Tables VIIB and VIIC. The following observations are pertinent.
A. Trace Elements Which Were Not Present at Levels Above Their LQD1
B. Concentrations of Trace Elements Extracted from the Plastic Materials1
C. Concentrations of Trace Elements Extracted from the Elastomer Materials1
None of the plastic materials had significant levels of extracted trace metals, with the levels of extracted Ba, Ca, and Zn being typically 0.02 ppm (20 ppb) or lower.
For all elastomers, the levels of extracted trace metals are greatest at pH 2 and least at pH 12.
Differences in the elastomers are reflected in their levels of four extracted elements (Ba, Ca, Mg, Zn). Material E3 has only small extracted amounts (less than 30 ppb) of these metals, and even these low values were obtained only at low pH. Materials E2 and E6 have large (>100 ppb) levels of extractable Mg and only small levels (<100 ppb) of Ba, Ca, and Zn. Material E5 has large levels of extractable Mg and Zn and only small levels of extractable Ba and Ca. Material E4 has the largest level of extractable Mg and smaller levels of extractable Ca.
Material E1 has extremely large (>1 ppm) levels of extractable Ba and Zn and large levels of extractable Ca, especially at an extraction pH of 2. The levels of Mg extracted from this material are low.
In terms of ranking/ordering the materials with respect to their extracted trace elements, all the plastic materials are ranked essentially equivalent. While material E1 is far and away the “worst” elastomer in terms of having the highest extractable metals burden, ranking the other materials is difficult because their extracted metal profiles are different. However, in terms of total extracted cations, the ranking of the elastomer materials is as follows:
Characterization of Particulate Found in pH 12 Extracts of the Elastomeric Materials:
Flocculent-like particulate matter was found suspended in the pH 12 extracts of the all of the elastomers. The particulate material was isolated for identification by FT-IR (which provides functional group information), SEM (which provides images of the surface structure), and EDXS (which establishes the elemental composition of materials). The IR spectra show generic correspondence to stearic acid salts and silicone oil. The absorption peaks at 1556, 1470, and 1418 cm−1, as seen in Figure 9, are found in E4, E5, and E6 materials and are consistent with a typical stearic salt, such as sodium stearate. The absorption peaks at 1554, 1469, and 1417 cm−1, as seen in Figure 10, are found in E1, E2, and E6 materials and are also consistent with stearic acid salt, such as sodium stearate. The absorption peaks at 1258, 1088, 1018, and 796 cm−1 are consistent with polymethylsilicone-type silicone oil. The absorption peaks at 1574, 1542, 1469, 1419, and 1107 cm−1, as seen in Figure 11, are found in extracts of material B3, and the spectrum is consistent with a stearic acid salt, such as calcium stearate. See Table VIII for a summary of the test results for the particulates.
IR spectrum of the flocculent particulate isolated from the pH 12 extract of material E2. The absorption peaks at 1556, 1470 and 1418 cm−1 are found in material isolated from extracts of materials E4, E5 and E6 and are consistent with a typical stearic salt, such as sodium stearate.
IR spectrum of the flocculent particulate isolated from the pH 12 extract of the material E6. The absorption peaks at 1258, 1088, 1018 and 796 cm−1 are consistent with polymethylsilicone type silicone oil.
IR Spectrum of the flocculent particulate isolated from the pH 12 extract of material B3. The absorption peaks at 1574, 1542, 1469, 1419, and 1107 cm−1 are consistent with a stearic acid salt, such as calcium stearate.
Analysis of Particulates Isolated from pH 12 Extracts of the Elastomer Materials
Conclusion
Plastic materials that could be used for syringe barrels and elastomeric materials that could be used for other appropriate parts of syringes, such a plungers and tip caps, have been characterized in terms of their ability to interact with solutions that contact such materials. Interactions that were studied include binding (loss of a substance from the solution via absorption and/or adsorption by the material) and leaching (migration of a substance from the materials into the solution). These interactions were considered, as they can impact the safety of drug product solutions that contact devices, such as syringes, that are made of such materials. The four plastic materials and six elastomeric materials differed, in some cases significantly, in their binding and leaching characteristics. Such differences can be utilized in selecting the proper materials for device applications, where in general the proper materials are those materials whose interactions either are immeasurable or are sufficiently small that there are no undesirable ramifications linked to the interactions.
Conflict of Interest Declaration
The authors declare that they have no competing interests.
- © PDA, Inc. 2012
References
Jump to section
Related Articles
Cited By...
- Science- and risk-based strategy to qualify prefillable autoclavable syringes as primary packaging material
- Low Leachable Container System Consisting of a Polymer-Based Syringe with Chlorinated Isoprene Isobutene Rubber Plunger Stopper
- Materials in Manufacturing and Packaging Systems as Sources of Elemental Impurities in Packaged Drug Products: A Literature Review
- Extractables Characterization for Five Materials of Construction Representative of Packaging Systems Used for Parenteral and Ophthalmic Drug Products