A Proposed Methodology for a Risk Assessment-Based Liposome Development Process

The requirements of a liposomal formulation vary depending on the pharmaceutical indication, the target patient population, and the corresponding route of administration. Different preparation methods require various material attributes (MAs) (properties and characteristics of the components) and process parameters (PPs) (settings of the preparation method). The identification of the quality target product profile for a liposome-based formulation, the critical quality attributes of the liposomes, and the possible MAs and PPs that may influence the key characteristics of the vesicles facilitates pharmaceutical research. Researchers can systematise their knowledge by using the quality by design (QbD) approach. The potential factors that influence the quality of the product can be collected and studied through a risk assessment process. In this paper, the requirements of a liposome formulation prepared via the thin-film hydration preparation technique are presented; furthermore, the possible factors that have an impact on the quality of the final product and have to be considered and specified during the development of a liposomal formulation are herein identified and collected. The understanding and the application of these elements of QbD in the pharmaceutical developments help to influence the quality, the achievements, and the success of the formulated product.


Introduction
Liposomes are described as artificially prepared vesicles composed of one or more concentric lipid bilayers that are enclosing one or more aqueous compartments by the European Medicine Agency [1]. Liposomes as drug carrier systems have several advantages [2]. These formulations can be used, among others, to protect active pharmaceutical agents (API), incorporate both lipophilic and hydrophilic drug molecules, and maintain targeted drug delivery [3]. From the beginning until the present day, four different generations of liposomes have been distinguished. The first-generation liposomes (conventional liposomes) are made up of neutral and/or negatively charged phospholipids and cholesterol [4]. These vesicles are taken up by the reticuloendothelial system (RES) (phagocytes) in cases of intravenous administration; thus, their circulation time is short [5]. The second generation consists of long-circulating liposomes, while the third generation is made from surface-modified liposomes that can avoid the defence mechanism of the immune system. The fourth generation is built up from polyethylene glycol (PEG)ylated or the so-called "stealth" liposomes [3,4]. The surface of these vesicles is coated with a hydrophilic polymer, such as polyethylene glycol (PEG), that increases the repulsive forces between the liposomes and thus avoids the protein adsorption and opsonisation associated with the drug substance, the excipients, the intermediates (in-process materials), and the drug product [33]. A CPP is a process parameter that variability has an impact on the CQAs and therefore should be monitored or controlled to ensure that the process produces the targeted quality [33].
(3) Risk assessment (RA): RA is a valuable science-based process that is used to identify and rank the parameters on the basis of their impact on the CQAs of the product. Risk assessment is typically performed as the first step during an early phase of the pharmaceutical development processes and is evaluated again when more information becomes available and higher knowledge is obtained [32,33]. The current experimental knowledge obtained from the former practical studies have to be aligned with information from the relevant literature. To perform a successful RA, first, the research team has to define the precise target product (QTPP) and then has to select the critical factors and estimate the interdependence of the critical factors, ranking them by the severity of their impact. The team members estimate the level of the interactions between the parameters occurring during the formulation process (production settings, materials, etc.). All the elements applied in the RA (QTPP elements, CQAs, CMAs, and CPPs) are defined and selected by the research group; therefore, their knowledge strongly impacts this selection process. Risk is defined as the combination of the probability of the occurrence of harm and the severity of that damage. The RA is a systematic process to evaluate the necessary information for the support of the risk-defining step within the risk management process. It means the identification of hazards and the analysis of risks [31]. The quality risk management tools provide systemic and reproducible methods based on up-to-date knowledge to rate the probability, severity, and sometimes detectability of the risk. These methods can be qualitative or quantitative. Once the risk is expressed quantitatively, a numerical scale is assigned for evaluation [33]. The numeric score of the evaluated risks could arise from the multiplications of the severity and occurrence (or probability) values, or sometimes from the severity, occurrence, and detectability if the same scale was used for the estimation of all of these parameters. The RA software can help in this process, but even during the software-supported assessments, the identification of the risks and the estimations of the severity and the occurrence are the task and responsibility of the research group. The software only makes the calculations and provides the data assessment and visualisation of the final results. These results are the basis of the design of experiments (DoE).
(4) Design space (DS) development: DS is a multidimensional combination and interaction of the input variables (e.g., material attributes) and the process parameters that have been demonstrated to assure quality.
(5) Definition of the control strategy. (6) Life cycle management. For better understanding, the schematic structure of the QbD approach is presented in Figure 1. This paper aimed to collect and evaluate the parameters that influence the manufacturing process of a liposomal pharmaceutical product in order to help the researchers and the professionals in the pharmaceutical industry in the QbD-based new liposome design and development. The authors aim to present a wide range of potential QTPP and CQA elements and their characteristics to highlight the potential decision and target points. It was also intended to give an example of how to use RA to rank the influencing parameters. For this illustration, the thin-film hydration method [35], the most common liposome production process (Table 1), was chosen, as the authors have practical experience and knowledge about this technique from their previous studies [27]. This method was described for the first time and used to prepare the first liposomes by Alec Douglas Bangham and his colleagues in 1965 [35]. Several modified versions of the original technique exist (Table 1), however, the basic steps of the process are mutual [36]: (1) preparation of the lipid film from phospholipids and cholesterol, (2) hydration of the thin film with a hydration medium, and (3) modification of the numbers of layers and the size of vesicles. This paper aimed to collect and evaluate the parameters that influence the manufacturing process of a liposomal pharmaceutical product in order to help the researchers and the professionals in the pharmaceutical industry in the QbD-based new liposome design and development. The authors aim to present a wide range of potential QTPP and CQA elements and their characteristics to highlight the potential decision and target points. It was also intended to give an example of how to use RA to rank the influencing parameters. For this illustration, the thin-film hydration method [35], the most common liposome production process (Table 1), was chosen, as the authors have practical experience and knowledge about this technique from their previous studies [27]. This method was described for the first time and used to prepare the first liposomes by Alec Douglas Bangham and his colleagues in 1965 [35]. Several modified versions of the original technique exist (Table 1), however, the basic steps of the process are mutual [36]: (1) preparation of the lipid film from phospholipids and cholesterol, (2) hydration of the thin film with a hydration medium, and (3) modification of the numbers of layers and the size of vesicles.

Preparation Methods Subtypes Comments
Mechanical dispersion methods probe or bath sonication − the critical parameters vary on the basis of the selected preparation method; therefore, the definition of the production technique has to be the first step of every liposome formulation process − the properties of the liposomes (e.g., number of lamellas, size, and distribution of vesicles) French pressure cells-extrusion freeze-thawed liposomes membrane extrusion lipid film hydration techniques hydration of proliposomes micro emulsification, coalescence of small vesicles dual asymmetric centrifugation heating method, Mozafari method electro-formation Solvent dispersion methods ether injection ethanol injection

Methods
The LeanQbD software (QbD Works LLC, Fremont, CA, USA) was used for the RA procedure. The first element of this procedure was the interdependence rating between the QTPPs and the CQAs, and the CQAs and the CPPs. A three-level scale was used to describe the relation between the parameters: "high" (H), "medium" (M), or "low" (L). In the software, the qualitative three-level scale, used for the estimation, is linked to a selectable numeric scale (0-10, or 0-100), which gives, at the end, the severity scores of the evaluated risk factors on the basis of mathematical calculations. In this study, the 0-10 scale was used. After the categorisation of the interdependence, a risk occurrence rating of the CPPs (or probability rating step) was made, applying the same three-grade scale (H/M/L) for the analysis. As the output of the initial RA evaluation, Pareto diagrams [37] were generated by the software, presenting the numeric data and the ranking of the CQAs and the CPPs according to their potential impact on the aimed final product (QTPP). The Pareto charts not only show the differences of the CMAs and the CPPs by their effect but also help to select the factors of a potential experimental design. Table 2 summarises the potential QTPP elements collected by the authors. Potential CQAs are collected and presented in Table 3.  As the preparation method (Table 1) defines the CPPs of the liposome formulation process, a production technique that provides the target CQAs need to be selected prior to the investigation of CMAs and CPPs. The API can be added to the formulation via passive or active loading techniques [3]. Mechanical dispersion [3,19,38,39], solvent dispersion [3,38,39], and detergent removal [3,38,39] methods belong to the passive loading techniques, in which methods the lipid films are prepared via Pharmaceutics 2020, 12, 1164 7 of 13 different techniques, hydrated to obtain liposomes, and the drug is captured during the manufacturing process [3,39]. In case of active loading, the API is incorporated into the already prepared liposomes via gradient loading techniques using buffers or ammonium sulphate gradients [39]. Besides the conventional preparation methods, there are also numerous approaches that have been recently developed to produce liposomes [39,40]. In this paper, the thin-film hydration method-related factors are presented. The potential CMAs and CPPs of the technique are systemised in a flow chart in Figure 2. The steps of the thin-film hydration method [36] are shown in the middle of the figure, while the related material attributes (MAs) and process parameters (PPs) are presented on the two sides of the chart. The general criticality of the presented factors was investigated in a RA, and the rankings of the elements of CQAs, illustrated with Pareto charts for better understanding, are shown in Figure 3, while CMAs and CPPs are shown in Figure 4. Figure 2. Collection of the properties of the liposome components (material attributes (MAs)) and the preparation method (process parameters (PPs)) that affect the result of the thin-film hydration liposome manufacturing technique.

Results
The general criticality of the presented factors was investigated in a RA, and the rankings of the elements of CQAs, illustrated with Pareto charts for better understanding, are shown in Figure 3, while CMAs and CPPs are shown in Figure 4.   The general criticality of the presented factors was investigated in a RA, and the rankings of the elements of CQAs, illustrated with Pareto charts for better understanding, are shown in Figure 3, while CMAs and CPPs are shown in Figure 4.

Discussion
The QTPP (Table 2) depends mainly on the therapeutic/clinical aims and requirements, as well as the characteristics of the drug substance, and it is always unique. For instance, QTPP may be a nano-sized liposome-containing injection for cancer therapy with a proper dose of drug and drug Pharmaceutics 2020, 12, 1164 9 of 13 release dedicated to the therapeutic needs. Those quality attributes that are critically related to the QTPP are the CQAs. That is the reason why the CQAs are also always unique and depend on the QTPP. The potential CQAs (Table 3) are, e.g., the type of the liposome, its lamellar structure, vesicle size, size distribution, sterility, viscosity, and stability, or the dissolution profile of the formulation. The API encapsulation efficiency is also a critical attribute for the liposomes, in addition to the zeta potential, which refers to the stability of the vesicles. PdI is one further potential CQA for lipid-based nanocarrier systems such as liposomes.
The application of a quality management visualisation tool, such as a fishbone diagram, process mapping, or a flow chart, is always useful for the identification of the CMAs and the CPPs of the aimed liposomal product. In this case, to show the systemic collection and presentation of the potential CMAs and CPPs, we built a flow chart (Figure 2). In the middle of the figure, the steps of the production process, which in this case was the thin-film hydration liposome preparation method, are presented. The left side of the flow chart contains the material attributes (MAs), and the right side shows the process parameters (PPs). These MAs and PPs can affect the result of the thin-film hydration-based liposome manufacturing process. The critical ones have to be selected and labelled as CMAs and CPPs. To make this figure and the tables of QTPP and CQAs, prior knowledge, previous experimental experience, and a thorough literature background survey of the field [31,[41][42][43][44][45][46][47][48] were necessary. Although, the main points of the tables and figures are shreds of evidence from the literature mixed with practical experiences, the systemic collection of all the relevant factors and data in one paper is the novelty of the work. The demonstration of the CMAs and the CPPs parallelly enhances the transparency of their relationships. In the following step, RA can be performed among the elements of the QTPP, the CQAs, and the CMAs and the CPPs. Several tools are suitable for an RA, e.g., the support of an RA software can help to achieve proper and quick implementation. In the presented case, the LeanQbD (QbD Works LLC, Fremont, CA, USA) RA software was applied. The interdependence rating among the elements was made on a three-grade scale, as the interaction is low (L), medium (M), or high (H). This process was made step by step for each pair of factors on the basis of the prior experimental and literary knowledge. The results of the RA are presented in Pareto charts generated by the software (Figures 3 and 4). Figure 3 shows the theoretical ranking of the CQAs of the liposomes according to the initial general RA made by the authors. It may also vary in other cases on the basis of the QTPP. Figure 4 presents the general ranking of the CMAs and the CPPs depending on their severity for the liposomal product. It may vary on the basis of the QTPP and the CQAs. According to the RA, the most influential CMAs, organised in descending order, are the phospholipids, the API content [27], the surface modifiers, the cholesterol content, the ratio between the phospholipids and the cholesterol, the phase transition temperature of the lipids, and the quality of the hydration media and the cryoprotectant, while the CPPs are the working temperature, the duration of the sonication, and the number of filtrations. The effect of the CMAs/CPPs can be accurately investigated if some of the values are set on the same level, while the ones under the scope of the study are changed according to the DoE.
Xu et al. performed a risk analysis study on liposomes prepared using the thin-film hydration technique and loaded with superoxide dismutase via a freeze-thaw cycling technique. They analysed those factors that affect the size, the encapsulation efficiency, and the stability of the liposomes. For this evaluation, they checked the properties of the formulation, the process, the analytical method, and the instrumentation reliability. They found that the "analytical method" and the "instrument reliability" categories can be well-controlled; therefore, the factors of these two categories are not critical. However, the factors of the "analytical method" and the "instrument reliability" are non-negligible for the selection and settings of the characterisation methods. Their findings, namely, the influencing role of the lipid concentration, the cholesterol ratio, and the quality of the phospholipids are consistent with our results [49]. Porfire et al. provided a general overview of the QbD approach for liposomes without defining a production process and described methodologies for liposome characterisation as a control strategy in detail. Their reasonable considerations were built into the tables of this paper with our additions. The facts above draw attention to the low number of studies following the steps of the QbD recommended by the regulatory authorities [31]. Our presented work fits well into this scientific research area; it extends the previous knowledge and gives a detailed overview of the QbD application. The systemised and structured form of the facts and information may help researchers in designing and planning their future studies of liposomes.

Conclusions
This work aimed to collect and systemise all the relevant factors of the liposome formulation development via the QbD technique. The application of the QbD approach is a regulatory requirement in the pharmaceutical submissions, and in these applications, RA is the key step. In this study, the theoretical method was presented, the potential QTPP elements of the liposome-based formulations were determined, and the potential CQAs of the liposomes were also collected. The potential critical material attributes and process parameters that need to be considered during the formulation design of the thin-film hydration liposome preparation method were listed and evaluated. The method of screening was also presented to identify the most critical factors. The phospholipids, the API content, the surface modifiers, the cholesterol content, the ratio between the phospholipids and the cholesterol, the phase transition temperature of the lipophilic phase, and the quality of the hydration media and the cryoprotectant were found to be the CMAs of highest influence. Furthermore, the working temperature, the duration of the sonication, and the number of filtrations were identified to be essential CPPs. The authors believe that the presented concept may help researchers to establish and perform studies on liposomes with less effort and more success.

Conflicts of Interest:
The authors declare no conflict of interest.