Dynamic Light Scattering Analysis for the Determination of the Particle Size of Iron-Carbohydrate Complexes

Summary

Dynamic light scattering (DLS) has emerged as a suitable assay for evaluating the particle size and distribution of intravenously administered iron-carbohydrate complexes. However, the protocols lack standardization and need to be modified for each iron-carbohydrate complex analyzed. The present protocol describes the application and special considerations for the analysis of iron sucrose.

Abstract

Intravenously administered iron-carbohydrate nanoparticle complexes are widely used to treat iron deficiency. This class includes several structurally heterogeneous nanoparticle complexes, which exhibit varying sensitivity to the conditions required for the methodologies available to physicochemically characterize these agents. Currently, the critical quality attributes of iron-carbohydrate complexes have not been fully established. Dynamic light scattering (DLS) has emerged as a fundamental method to determine intact particle size and distribution. However, challenges still remain regarding the standardization of methodologies across laboratories, specific modifications required for individual iron-carbohydrate products, and how the size distribution can be best described. Importantly, the diluent and serial dilutions used must be standardized. The wide variance in approaches for sample preparation and data reporting limit the use of DLS for the comparison of iron-carbohydrate agents. Herein, we detail a robust and easily reproducible protocol to measure the size and size distribution of the iron-carbohydrate complex, iron sucrose, using the Z-average and polydispersity index.

Introduction

Iron sucrose (IS) is a colloidal solution comprised of nanoparticles consisting of a complex of a polynuclear iron-oxyhydroxide core and sucrose. IS is widely employed to treat iron deficiency among patients with a wide variety of underlying disease states who do not tolerate oral iron supplementation or for whom oral iron is not effective¹. IS belongs to the drug class of complex drugs as defined by the Food and Drug Administration (FDA), which is a class of drugs with complex chemistry commensurate with biologicals². The regulatory evaluation of complex drug products may require additional orthogonal physicochemical methods and/or preclinical or clinical studies to accurately compare follow-on complex drugs^3,4. This is important because several studies have reported that the use of IS versus a follow-on IS product does not produce the same clinical outcomes. This underscores the criticality of the use of novel and orthogonal characterization techniques that are suitable for detecting differences in the physicochemical properties between IS products^5,6.

The accurate elucidation of the size and size distribution of IS is of clinical importance, as particle size is a major influential factor in the rate and extent of opsonization-the first critical step in the biodistribution of these complex drugs^7,8. Even slight variations in the particle size and particle size distribution have been related to changes in the pharmacokinetic profile of iron-oxide nanoparticle complexes^9,10. A recent study by Brandis et al. showed that particle size measured by DLS was significantly different (14.9 nm ± 0.1 nm vs. 10.1 nm ± 0.1 nm, p < 0.001) when comparing a reference listed drug and a generic sodium ferric gluconate product, respectively¹¹. The consistent batch-to-batch quality, safety, and efficacy of iron-carbohydrate products are entirely dependent on the manufacturing process scale-up, and potential manufacturing drift must be carefully considered⁹. The manufacturing process may result in residual sucrose, and this will vary based on the manufacturer¹². Any modifications in the manufacturing process variables can lead to significant changes in the final complex drug product with regard to the structure, complex stability, and in vivo disposition⁹.

To assess drug consistency and predict the drug's in vivo behavior, contemporary orthogonal analytical methodologies are required to determine the physicochemical properties of complex nanomedicines. However, there is a lack of standardization of methodologies, which can result in a high degree of interlaboratory variation in result reporting¹³. Despite the recognition of these challenges by global regulatory authorities and the scientific community¹⁴, most of the physicochemical characteristics of IS remain poorly defined, and the full complement of critical quality attributes in the context of available regulatory guidance documents have not been defined¹⁵. The draft product-specific guidance documents issued by the FDA for iron-carbohydrate complexes suggest DLS as a procedure to evaluate the size and size distribution of follow-on products^16,17.

Several publications have detailed established DLS protocols to determine IS nanoparticle dimensions^13,18. However, because the sample preparation, procedure conditions, instrumentation, and instrumentation setting parameters are different among the published methods, the DLS results cannot be directly compared in the absence of a standardized method to interpret the results^13,18. The diversity in methodologies and data-reporting approaches limit the appropriate evaluation of these characteristics for comparative purposes¹⁹. Importantly, many of the DLS protocols previously published to evaluate IS do not account for the effect of the diffusion of sucrose in the suspension due to the presence of free sucrose, which has been shown to spuriously elevate the Z-average-calculated hydrodynamic radii of the nanoparticles in colloidal solutions^13,18. The present protocol aims to standardize the methodology for the measurement of the particle size and distribution of IS. The method has been developed and validated for this purpose.

Protocol

1. Operating the machine

1. Starting up the machine and software
   NOTE: Supplemental Figure S1A-D describes the steps for starting up the machine and software.
   1. Switch on the instrument at least 30 min before starting the measurements, and then start the PC.
   2. Double-click on the instrument software icon to start the program.
   3. Enter the username and password in the login window. Ensure that each user has their own account.
   4. Wait for all three black bars at the bottom right corner to light up green, indicating that the unit is ready for operation.
   5. In the case of long periods of inactivity, when the user is automatically logged out, under Security, click on Login, and re-enter the password.
2. Creating a measurement file
   NOTE: A new measurement file is created at the beginning of each measurement day. All the measurements are listed and stored in the measurement file. For details of this procedure, see Supplemental Figure S2A-B.
   1. Create a new measurement file by clicking on File | New | Measurement File. In the opened window, select the storage location, and name the measurement file. Confirm the details by clicking on Save.
   2. To open a file, click on File | Open | Measurement File. Select the measurement file in the opened window, confirm the details by clicking on Open, select the file name, and click on Save.
3. Printing the results
   1. Print or save the results of the system suitability test (SST) (see step 1.5) and the mean value of the sample measurement according to Supplemental Figure S3.
   2. Mark the measurement(s) in the Records View of the Measurement File.
   3. Right-click on Batch Print, and wait for another small window to open.
   4. Select the PSD Intensity report from the choices, and confirm it by clicking on OK.
4. General procedure to start a measurement
   NOTE: The procedure for starting a measurement is described in Supplemental Figure S4A-D. Follow the path outlined below for the required unit parameter file (referred to as SOP):
   1. Select the required SOP from the dropdown list. Since the most recently used SOPs are displayed in the list, if an older SOP is needed, select Browse for SOP, and confirm by clicking on the green arrow. Once the storage location of the SOPs opens, proceed with step 1.4.2.
      NOTE: For details of the important system parameters specific for iron sucrose (e.g., equilibration time of an attenuator), see Table 1.
   2. In the opened window, make the required entries under Sample Name and Notes. Confirm by clicking on OK, and wait for the measurement window to open automatically.
   3. Start the measurement by clicking on the green Start button.
   4. Once an acoustic signal sounds at the end of the measurement, close the measurement window.
5. System suitability test (SST) measurement
   NOTE: Do not touch the lower part of the cuvette (measuring zone). At the beginning and at the end of the sequence, measure the 20 nm particle standard.
   1. Fill ~1 mL of the undiluted particle standard into a polystyrene cuvette, and close with the lid.
      NOTE: The diluted standard prepared in step 1.5.1 can be used for 1 month.
   2. After filling, close the cuvette, and check for air bubbles. Remove air bubbles by tapping the cuvette lightly.
   3. Place the cuvette in the cell holder of the instrument with the arrow mark facing forward, and close the measuring chamber cover.
   4. Load the unit parameter SOP, and enter the following data in the start window:
      Sample Name: SST 20 nm Particle Standard
      ​Then, add a note: Identifier number and expiry date of the standard
   5. Start the measurement.
   6. After the end of the measurement when the acoustic signal sounds, close the measurement window.
   7. Print the report (see point 1.1.3).
      NOTE: The SST is passed if the particle size Z-average corresponds to the value of the certificate of analysis ± 10%.
6. Measurement of the iron sucrose solution
   1. Pipette 0.5 mL of an IS solution with an iron content of 2% m/V into a 25 mL volumetric flask, and fill up to the mark with low-particle water (e.g. freshly deionized and filtered [pore size 0.2 µm]); this solution contains 0.4 mg Fe/mL.
      NOTE: The sample preparation in step 1.6.1. with this specific dilution was established during method development, and this was determined as the optimal dilution for this purpose.
   2. For preliminary cleaning, fill the polystyrene cuvette approximately 3/4 full with the prepared measuring solution, swirl gently, and then empty as completely as possible.
      NOTE: Do not touch the lower part of the cuvette (measuring zone), and avoid air bubbles by not shaking the cuvette.
   3. For the measurement, pipette 1 mL of the measuring solution into the cuvette, and put on a lid.
   4. Check the measuring solution in the cuvette for air bubbles. If there are air bubbles, remove them by lightly tapping the cuvette.
7. Performing the measurement
   1. Place the plastic cuvette with the measuring solution in the device with the arrow mark facing forward, and close the lid.
   2. Load the parameter SOP (see step 1.4.1), and enter the following data in the start window:
      ​Sample Name: Batch number
   3. Start the measurement.
   4. After the end of the measurement, when the acoustic signal sounds, close the measurement window.
   5. Calculate the mean value of the six individual measurements according to Supplemental Figure S5. Mark the individual measurements in the Records View of the Measurement File, right-click on Create Average Result, add the name of the mean value under Sample Name, and confirm by clicking on OK.
   6. Wait for the software to create a new record at the end of the list, and look for the name entered and the averaged result in that record.
   7. Print the report (see step 1.1.3).

Representative Results

The method described was validated according to ICH Q2(R1)²⁰, which involved the measurement of test solutions under varying conditions. The precision was only 0.5% RSD for the Z-average size, while a maximum of 3.5% RSD was calculated for the PDI. The mean results from different analysts and days only differed by 0.4% for the Z-average size and 1.5% for the PDI. Statistics were calculated from 12 measurements performed by two analysts on varying days. Neither changes in the test concentration in the range of 50%-200% nor the storage of the test solutions for up to 5 days in refrigerated conditions had an impact on the final result.

Analyzed parameters
Z-average size
The hydrodynamic diameter is given as the Z-average particle size, and the method for determining this is defined in ISO 22412:2017¹⁷. The Z-average size is a parameter also known as the cumulant mean. The Z-average is the preferred DLS size parameter, as the calculation of the Z-average is mathematically stable, and the Z-average result is insensitive to noise. According to the EMA and FDA, the Z-average size together with the PDI are the recommended values for the characterization of nanomedicines^15,16,21. The Z-average particle size is only comparable with the size measured by other techniques if the sample is monomodal, spherical, or near-spherical in shape, is monodisperse, and is prepared in a suitable dispersant. This is because the Z-average mean particle size is sensitive to even small changes in the sample preparation. The Z-average particle size is a hydrodynamic parameter and is, therefore, only valid for particles in a dispersion or for molecules in solution.

Polydispersity index
This index is a number calculated from a simple two-parameter fit to the correlation data (the cumulant analysis). The polydispersity Index is dimensionless and scaled such that values smaller than 0.05 are rarely seen, except for in highly monodisperse standards. Values greater than 0.7 indicate that the sample has a very broad particle size distribution and is likely not suitable for the DLS technique. Various size distribution algorithms can function with data that fall between these two extremes. The calculations for these parameters are defined in the ISO standard document 22412:2017¹⁷.

Size distribution by intensity/volume/number
Typical size distribution plots (intensity, volume, number) are depicted in Figure 1. The result plots show six independently prepared samples of IS batch 605211 at a concentration of 0.4 mg Fe/mL. For the visualization in Figure 1, the raw data taken from the DLS software were plotted with statistical software without further modification⁹. A size distribution by intensity impacted by a second peak is provided as an example of a bad result in Figure 1A. Figure 2 displays poor-quality data showing an additional signal at 5,000 nm.

Figure 1: Size distribution. (A) intensity, (B) volume, and (C) number¹³. Please click here to view a larger version of this figure.

Figure 2: Representative poor-quality data. Please click here to view a larger version of this figure.

The test solution of IS batch 0371022A (0.4 mg Fe/mL), which was stored for 5 days at room temperature, showed an additional signal at ~5,000 nm, which is indicative of some additional particles (e.g., either dust or precipitation). Accordingly, the PDI originally determined at 0.130 was shifted to 0.184, while the Z-average was still close to the original value (i.e., 11.33) at 11.99 nm (unpublished data).

Precision was tested by two lab technicians on separate days. The mean value of 12 replicates was 11.32 nm with an RSD of 0.4% and 0.125 with an RSD of 1.5% for the Z-average and PDI, respectively, for the two technicians. The acceptance criteria were met (NMT 5% for the Z-average, NMT 20% for the PDI) (unpublished data).

Comparison of analyzable parameters
In addition to calculating the basic parameters-the Z-average and polydispersity-the software of the DLS device also allows the calculation of size distributions that can be weighted according to the intensity of the detector signal or the volume (or number) of scattering particles. The relevance of comparing these parameters is obvious in the results outlined in Table 2. While the size distribution by number differed by up to a factor of 2 from the proposed intensity-based Z-average, only slightly lower values were calculated by the size distribution by volume. It should be noted, however, that intensity-based result reporting may be inaccurate if the iron-carbohydrate complex solutions contain larger particles or aggregates¹³.

Table 1: System parameters for the DLS. Abbreviations: RI = refractive index; DLS = dynamic light scattering¹³. Please click here to download this Table.

Table 2: Examples of how particle size determination by IS is affected by the data reporting approach. This table is adapted from Di Francesco and Borchard¹³. Abbreviations: SD = standard deviation; RSD = relative standard deviation; PDI = polydispersity index; IS = iron-sucrose. Please click here to download this Table.

Supplemental Figure S1: System operating steps. Please click here to download this File.

Supplemental Figure S2: Creating a measurement file. Please click here to download this File.

Supplemental Figure S3: System suitability test. Please click here to download this File.

Supplemental Figure S4: Starting a new measurement. Please click here to download this File.

Supplemental Figure S5: Calculation of measurements. Please click here to download this File.

Discussion

DLS has become a fundamental assay for the determination of the size and size distribution of nanoparticles for applications in drug development and regulatory evaluation. Despite advances in DLS techniques, methodologic challenges still exist regarding the diluent selection and sample preparation, which are especially relevant for iron-carbohydrate complexes in colloidal solutions. Importantly, DLS methods for iron-carbohydrate nanomedicines have not yet been studied extensively in the biological milieu (e.g., the plasma)²². Therefore, there still remains a critical need for the harmonization of best-practice protocols depending on the diluent selection. The selection of the diluent is important, as using purified water versus isotonic saline solutions may affect the colloidal suspension's stability¹⁶.

It should also be noted that iron-carbohydrate complexes should not be diluted below the prescribing information recommendations in an effort to mitigate the challenges of having a dark, opaque solution. Excessive dilutions are not biorelevant and can affect the colloidal suspension's stability via changes in ionic shielding, which can lead to potential precipitation and inaccurate result reporting. Various dilutions and diluents (unpublished data) were tested during the development of this method, and the sample preparation described in step 1.6.1 of the protocol was determined and validated as the optimal dilution for IS. Several modifications must be considered for the DLS analysis of iron-carbohydrate complexes. For example, the preparation of test solutions needs to be performed in the absence of any kind of high-speed agitation. The use of vortex mixers should be avoided, as this induces the creation of iron-sucrose aggregates. For the preparation of test solutions, IS solutions are gently mixed in water with an automatic pipette. Additionally, when running the samples for DLS analysis, the automatic calibration feature should be turned off.

There are several inherent limitations of DLS analysis for iron-carbohydrate nanoparticles. Due to the nature of the light scattering angles and the Z-average output, the hydrodynamic diameters reported are biased toward larger particles in the measuring solution. Thus, the particle size may be overestimated, and the true distribution of the particle size may be underestimated¹³. Reporting result techniques should be considered in the context of how large the iron-carbohydrate complex particles are and the potential for aggregation under the experimental conditions. It should also be taken into account that the results of the intensity-, volume-, and number-weighted size distributions may differ greatly between different DLS units from the same or different manufacturers, as different manufacturers use different algorithms for the calculation. Therefore, ISO22412 only recommends the use of the Z-average and polydispersity, as the algorithm for their calculation is standardized. Regulatory agencies have also recommended Z-average size reporting¹⁶. It should also be noted that minor modifications will be required (e.g., the handling of the software, measuring procedure, and data preparation) when this protocol is applied to other instruments.

Even in light of the challenges associated with DLS, this technique represents a significant advancement on previous analytic methodologies and adds compelling data to the characterization of iron-carbohydrate complexes. It has been endorsed by scientist collaboratives and regulatory agencies^16,18,19,21. Future efforts in applying DLS analysis to iron-carbohydrate complexes should most importantly focus on the global harmonization of protocols for their application to drug development and regulatory evaluation, including ensuring bioequivalence. Overall, the analytical protocol here described aims to standardize the methodology for the measurement of the particle size and distribution of IS and can be a useful tool for the evaluation of the quality of IS.

Materials

Name	Company	Catalog Number	Comments
Equipment
Zetasizer Nano ZS	Malvern	NA	equipped with Zetasizer software 7.12, Helium Neon laser (633 nm, max. 4 mW) and 173° backscattering geometry
Materials
Disposable plastic cuvettes
LLG-Disposable plastic cells	LLG labware	LLG-Küvetten, Makro, PS; Order number 9.406011
low-particle water			(The use of freshly deionized and filtered (pore size 0.2 μm) water is recommended).
Microlitre pipette
Venofer 100 mg/5 mL	Vifor Pharma
Volumetric flask 25 mL
Nanosphere	Thermo	3020A	Particle Standard
Software
Origin Pro v.8.5	Origin Lab Corporation