Online measurement method of dichloromethane content in aqueous phase during microsphere preparation

Abstract

The present invention discloses an online measurement method for the dichloromethane content in the aqueous phase during microsphere preparation, and belongs to the field of pharmaceutical analysis technology. The method comprises the following steps: designing multiple batches of microsphere preparation experiments, starting microsphere preparation and online acquisition of raw Raman spectra, analyzing the dichloromethane content in the aqueous phase during the curing process by headspace gas chromatography, pairing the raw Raman spectra with the dichloromethane content in the aqueous phase during the curing process one by one, removing outliers and dividing the data into sample sets, preprocessing the Raman spectra to select the best spectral preprocessing method, optimizing the spectral interval, and using the partial least squares method under optimal conditions to establish an online quantitative analysis model for the dichloromethane content in the aqueous phase during microsphere preparation. The present invention is beneficial in that it achieves accurate online measurement and quality control of the DCM content in the aqueous phase during microsphere preparation, effectively improves detection efficiency, and avoids the influence of sample preparation on measurement accuracy.

Description

Technical Field

[0001] The invention relates to a method for measuring dichloromethane content, in particular to an online method for measuring dichloromethane content in an aqueous phase during a microsphere preparation process, and belongs to the technical field of pharmaceutical analysis. Background Art

[0002] Currently, pharmaceutical production typically utilizes traditional production methods, employing offline analytical methods to sample raw materials, intermediates, and final products during the drug production process. These samples are then sent to the laboratory for analysis. This traditional method's measurement data generally lags behind production, preventing real-time monitoring of product quality and, consequently, timely adjustment of process parameters. This consumes significant manpower and material resources, making it difficult to ensure drug uniformity. Furthermore, the preparation process is often invisible, and the conditions within the equipment cavity are not fully understood. The relevant information obtained after the process is limited, making process control difficult and impacting the final quality of the product. Therefore, studying process analytical technology (PAT) in pharmaceutical production is of great value and significance.

[0003] To introduce PAT into pharmaceutical production and quality control, the U.S. Food and Drug Administration (FDA) officially released PAT industry guidance in September 2004. This guidance encourages pharmaceutical companies to monitor critical quality and performance parameters during drug production in real time, enabling timely adjustments to manufacturing processes, preventing or mitigating the risk of inferior products, and ensuring final product quality. Compared to traditional offline analytical methods, PAT eliminates the need for sample preparation and transfer, improving production efficiency. It allows real-time understanding of critical quality attributes during drug production, enabling insights into production processes and enhancing process control.

[0004] PAT provides key technologies for understanding, monitoring, and controlling production, such as near-infrared spectroscopy and Raman spectroscopy. While near-infrared spectroscopy still dominates the pharmaceutical industry, Raman spectroscopy is increasingly being used due to its advantages, including minimal or no sample preparation, rapid analysis, minimal moisture impact, high sensitivity, and simple operation. Raman spectroscopy was invented in 1928 after the Indian scientist C.V. Raman discovered the Raman effect. Raman spectroscopy is a type of scattering spectroscopy based on the Raman effect. The principle is that when light strikes a medium, energy exchange occurs between photons and molecules, causing the photons to change in frequency, energy, and direction, resulting in inelastic scattering. This inelastic scattering phenomenon is known as Raman scattering. In practical applications, Raman spectroscopy offers inherent advantages over near-infrared and infrared spectroscopy due to its specificity, compatibility with aqueous systems, and sampling flexibility.

[0005] Microsphere sustained-release formulations are tiny spherical entities formed by the adsorption or dispersion of a drug within a polymer matrix, typically with a particle size ranging from 1μm to 250μm. Microsphere formulations offer advantages in terms of safety and efficacy. They not only delay drug release in the body, reduce dosing frequency, and improve patient compliance, but also mask the drug's distinctive pungent taste and enhance post-administration stability. However, the unclear correlation between key material properties, critical process parameters, and critical quality attributes of microsphere formulations has severely hampered the advancement of microspheres from the laboratory to the clinic. Over the past 40 years, only a dozen PLGA (poly(lactide-co-glycolide)) microsphere products have been approved by the FDA. To date, research on PAT for microspheres is uncommon. In microsphere preparation, dichloromethane (DCM) is the most widely used and most stable organic solvent for microsphere formulations due to its strong solubility, low boiling point, volatility, and low solubility in water. Furthermore, during the microsphere preparation process, the removal of DCM not only affects the internal structure of the microspheres but also their final morphology and size, severely impacting key quality attributes such as microsphere release. Therefore, the DCM content of the aqueous phase is a key concern during microsphere preparation and requires control. To date, most methods for measuring DCM content in the aqueous phase have relied on offline analysis, which results in hysteresis and cannot provide timely information on the dynamics of the microsphere preparation process, leading to quality issues in the final microsphere product. Summary of the Invention

[0006] To address the deficiencies of the prior art, the present invention aims to provide an online measurement method for the DCM content of the aqueous phase during the microsphere preparation process. Compared with previous offline analysis methods, this online measurement method can effectively improve detection efficiency and avoid the influence of sample preparation on measurement accuracy.

[0007] In order to achieve the above objectives, the present invention adopts the following technical solutions:

[0008] The method for online measurement of dichloromethane content in the aqueous phase during microsphere preparation comprises the following steps:

[0009] (1) Design multiple batches of microsphere preparation experiments by changing a single factor within the range of normal and abnormal formulation process conditions;

[0010] (2) Start preparing microspheres, use an online Raman spectrometer to collect original spectra during the entire curing process, and record the data collection time and Raman spectrum;

[0011] (3) Analyze the dichloromethane content in the aqueous phase during the curing process by headspace gas chromatography;

[0012] (4) Pair the original Raman spectra collected from each batch with the dichloromethane content in the aqueous phase during the curing process determined by gas chromatography, and filter the data to remove outliers;

[0013] (5) The Kennard-Stone method is used to group each batch of data into a calibration set and a validation set;

[0014] (6) Different spectral preprocessing methods were used to preprocess the Raman spectra. The partial least squares method was selected as the modeling method. The best spectral preprocessing method was selected based on the evaluation parameters of the model results.

[0015] (7) The variable importance projection algorithm is used to screen the Raman spectrum bands, compare them with the full spectrum modeling, and select the optimal modeling band;

[0016] (8) Under the conditions of optimal spectral preprocessing method, 4 potential variables, and optimal modeling band, the partial least squares method was used to establish an online quantitative analysis model for the dichloromethane content in the aqueous phase during the microsphere preparation process.

[0017] Preferably, in step (1), the factors include: oil-water ratio, compressed air flow rate and curing temperature.

[0018] Preferably, in step (2), the parameters of the online Raman spectrometer are set as follows: laser intensity of 300 mW, exposure time of 3 s, number of scans of 10 times, spectrum acquisition every 45 s, wavelength range of 100 cm -1 -3200cm -1 .

[0019] Preferably, in step (3), the method for analyzing the dichloromethane content in the aqueous phase during the curing process is:

[0020] (i) Preparation of aqueous phase samples during the curing process: 0.1 mL of the aqueous phase was accurately measured and placed in a 10 mL headspace vial at different time points during the curing process. 0.9 mL of N,N-dimethylacetamide was added, and the vial was sealed and shaken to mix. The time points were consistent with the time points at which the Raman spectra were collected.

[0021] (ii) Prepare a reference solution: Dissolve dichloromethane in N,N-dimethylacetamide to obtain a 0.5 mg / mL reference solution;

[0022] (iii) using gas chromatograph and headspace gas chromatography to collect gas chromatograms of aqueous samples and reference solution during the curing process;

[0023] (iv) Calculate the dichloromethane content in the aqueous phase during the curing process using gas chromatography. The calculation formula is:

[0024]

[0025] Among them, A 供试品 is the peak area of ​​DCM in the test solution, A 对照品 is the average peak area of ​​DCM in the reference solution, C 对照品 is the concentration of DCM in the reference solution, in mg / mL.

[0026] Preferably, in step (4), the data are screened to remove outliers using a combination of principal component analysis and influence diagrams, with the confidence level set at 95%.

[0027] Preferably, in step (5), the samples are divided into a calibration set and a validation set in a ratio of 3:1.

[0028] Preferably, in step (6), the spectral preprocessing method includes: Raw, MC, SNV, SNV+MC, MC+SNV, FD+SG+SNV+MC, SD+SG+SNV+MC, SNV+FD+SG+MC, SNV+SD+SG+MC, FD+SG+MC, SD+SG+MC, FD+SG+SNV, SD+SG+SNV, SNV+FD+SG ​​and SNV+SD+SG.

[0029] The present invention is beneficial in that:

[0030] (1) Accurate online measurement and quality control of the DCM content in the aqueous phase during microsphere preparation were achieved. Compared with previous offline analysis methods, this method effectively improved detection efficiency and avoided the impact of sample preparation on measurement accuracy.

[0031] (2) Using Raman spectroscopy to introduce PAT into the monitoring of DCM content in the aqueous phase during the microsphere preparation process can obtain key information in real time and adjust parameters in time to ensure the final quality of the microsphere product. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a graph showing the change of DCM content in the water phase over time during the curing process;

[0033] Figure 2 is a graph showing the change of DCM content in microspheres over time during the curing process;

[0034] Figure 3 This is the original Raman spectrum of the rotigotine microsphere samples used for modeling (a total of 164 samples from batches 1 to 7, including microspheres during the curing process and final product microspheres);

[0035] Figure 4 This is the result graph of abnormal sample elimination obtained by PCA method;

[0036] Figure 5 This is the abnormal sample elimination result diagram obtained by using the influence diagram method;

[0037] Figure 6 It is the PCA score scatter plot of the sample set partitioned during the microsphere preparation process;

[0038] Figure 7 This is the result of using the VIP algorithm to screen the Raman spectrum bands;

[0039] Figure 8 This is a comparison chart of the predicted value of DCM content in the water phase during the curing process of the external validation set of the full-band PLS model and the reference value;

[0040] Figure 9 is the relationship between the predicted value and reference value of DCM content in the water phase during the curing process of the external validation set of the full-band PLS model;

[0041] Figure 10 This is a graph showing the cumulative release calculation results of the final product, Rotigotine microspheres, prepared in different batches over a period of 40 min to 6 h.

[0042] Figure 11 This is the calculation result of the cumulative release of the final product Rotigotine microspheres prepared in different batches from 40 minutes to 8 days. DETAILED DESCRIPTION

[0043] Rotigotine microspheres were used as a model drug, and the DCM content in the aqueous phase during the microsphere preparation process was used as an evaluation indicator. Sample spectra were collected using a Raman spectrometer, and the DCM content in the aqueous phase was determined by headspace gas chromatography. The partial least squares (PLS) method was used to correlate the spectral data with the gas phase data. After optimizing the spectral parameters, a quantitative analysis model for the DCM content in the aqueous phase was finally established.

[0044] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] 1. Design multiple batches of microsphere preparation experiments under normal and abnormal prescription process conditions

[0046] Since changes in formulation composition and process parameters usually have a certain impact on the quality attributes of microspheres, factors that have a significant impact on the change in DCM content in the aqueous phase during the formulation process, including formulation composition (oil-water ratio) and process parameters (compressed air flow rate, curing temperature), were used to design multiple batches of microsphere preparation experiments within the normal and abnormal formulation and process conditions by changing a single factor. See Table 1 for details.

[0047] Table 1 Parameter settings during microsphere preparation

[0048]

[0049] Rotigotine microspheres were prepared using a single emulsion solvent evaporation method, specifically:

[0050] (1) Add 8 L of 0.5% (w / v) polyvinyl alcohol (PVA) solution to the reactor and control the temperature to 8°C (aqueous phase);

[0051] (2) Take approximately 8.4 g of poly(lactic-co-glycolic acid) (PLGA) 5050 2A, approximately 8.4 g of PLGA 7525 4A, approximately 0.6 g of palmitic acid, and approximately 7.2 g of rotigotine, accurately weigh them, and dissolve them in 120 mL of DCM (oil phase);

[0052] (3) The oil phase was added to the water phase by a peristaltic pump at a homogenization speed of 700 rpm according to a specified oil-water ratio (e.g., 0.5:100, 1:100, 1.5:100) for homogenization and emulsification. After the addition of the oil phase, the homogenization and emulsification were continued for 2 min. After the emulsification was completed, the compressed air flow rate was set (e.g., 20 m / s, 50 m / s, 80 m / s), and the microspheres were solidified at a specified curing temperature (e.g., 20 °C, 30 °C, 40 °C) at 100 rpm for 5 h to volatilize DCM.

[0053] (4) After the solidification is completed, the microspheres with a particle size between 1200 mesh and 80 mesh are collected through 1200 mesh and 80 mesh sieves, washed with deionized water and placed in a -20°C refrigerator for pre-freezing. The obtained sample is then placed in a freeze dryer for freeze drying, and the freeze-dried product is passed through an 80 mesh sieve to obtain the corresponding final product.

[0054] 2. Preparation of Microspheres and Acquisition of Raman Spectra

[0055] The original spectrum was collected by an online Raman spectrometer. Before preparing the microspheres, the probe of the online Raman spectrometer was inserted into the reactor used for preparing the microspheres to prepare the original spectrum. The laser intensity was 300mW, the exposure time was 3s, the number of scans was 10, and the spectrum was collected every 45s. The wavelength range was 100cm -1 -3200cm -1 .

[0056] Microsphere preparation began. Throughout the curing process, data collection time and Raman spectra were recorded. In this embodiment, a total of 189 samples were collected across all batches. Data from batches 1 through 7 were used to establish the model, and data from batch 8 was used for external validation of the model. The sample data collection distribution is shown in Table 2.

[0057] Table 2 Distribution of sampling points during the curing process

[0058]

[0059] The original Raman spectra of the rotigotine microsphere samples used for modeling (a total of 164 samples from batches 1 to 7, including microspheres during curing and final product microspheres) are shown in Figure 3 .

[0060] 3. Analysis of DCM content in samples by headspace gas chromatography

[0061] 1. Prepare blank solution, test samples (microspheres during curing, aqueous phase during curing, and final product microspheres) and reference solution (DCM)

[0062] Blank solution: Accurately measure 1 mL of N, N-dimethylacetamide, place it in a 10 mL headspace vial, and seal it.

[0063] Microspheres during the curing process: Take about 20 mg of microspheres at different time points (consistent with the time point of collecting Raman spectra), accurately weigh them, place them in a penicillin bottle, add 10 mL of N, N-dimethylacetamide to dissolve and shake well, take 1 mL and place it in a 10 mL headspace bottle, seal it, and prepare 2 parallel copies.

[0064] Aqueous phase during the curing process: Take 0.1 mL of the aqueous phase during the curing process at different time points (consistent with the time point of collecting Raman spectra), accurately measure it, place it in a 10 mL headspace bottle, add 0.9 mL of N, N-dimethylacetamide, seal it and shake it well, and prepare two parallel portions.

[0065] Final product microspheres: Take approximately 20 mg of the final product rotigotine microspheres prepared from each batch, accurately weigh them, place them in 10 mL headspace vials, add 1 mL of N, N-dimethylacetamide, seal them, and prepare two parallel batches.

[0066] Reference solution: Take about 50 mg of DCM, accurately weigh it, place it in a 100 mL volumetric flask, dilute it to the mark with N, N-dimethylacetamide, and shake well to obtain a reference solution with a concentration of 0.5 mg / mL.

[0067] 2. Collect gas chromatograms of the test samples (microspheres during curing, aqueous phase during curing, and final microspheres) and the reference substance (DCM)

[0068] Gas chromatogram was collected by gas chromatograph using headspace gas chromatography. The chromatographic conditions were as follows:

[0069] Chromatographic column: DB-624 quartz capillary column (30m×0.53mm×3µm);

[0070] Temperature program: initial column temperature was 40°C, maintained for 6 min, increased to 200°C at 40°C / min, and maintained for 3 min;

[0071] Injection chamber temperature: 200°C;

[0072] Split ratio: 5:1;

[0073] Detector: flame ionization detector (FID), temperature 260°C, hydrogen flow rate 30 mL / min, air flow rate 300 mL / min;

[0074] Carrier gas: nitrogen, flow rate 4.0 mL / min;

[0075] Headspace sampler: equilibrium temperature 90°C, equilibrium time 30 min, quantitative loop temperature 105°C, transfer tube temperature 115°C, injection volume 1 mL;

[0076] GC cycle time: 30 min.

[0077] 3. Calculate the DCM content in the test sample (microspheres during curing, aqueous phase during curing, and final product microspheres) based on gas chromatography

[0078] The calculation formulas for the DCM content in the microspheres during the curing process, the aqueous phase during the curing process, and the final product microspheres are as follows:

[0079] ;

[0080] ;

[0081] ;

[0082] Where: C 对照品 is the concentration of DCM in the reference solution (mg / mL);

[0083] C 供试品 is the concentration of microspheres in the test solution (mg / mL);

[0084] A 供试品 is the peak area of ​​DCM in the test solution;

[0085] A 对照品 is the average peak area of ​​DCM in the reference solution.

[0086] After calculation, the DCM content in the water phase changes with time during the curing process. Figure 1 The DCM content in the microspheres changes with time during the curing process. Figure 2 .

[0087] Depend on Figure 1 and Figure 2 It can be seen that:

[0088] (1) Oil-water ratio: The oil-water ratio has a significant effect on the DCM content in the aqueous phase;

[0089] (2) Curing temperature: At a lower curing temperature, the volatilization rate of DCM in the microspheres is slower, and the DCM content in the microspheres at the end of the reaction is relatively increased; with the increase of curing temperature, the volatilization rate of DCM in the microspheres is significantly accelerated, and the DCM content in the microspheres at the end of the reaction is correspondingly reduced;

[0090] (3) Compressed air flow rate: The compressed air flow rate has little effect on the DCM content in the aqueous phase.

[0091] 4. Establishing a Quantitative Model

[0092] First, the original Raman spectra collected from batches 1 to 7 were paired one by one with the reference data measured by gas chromatography (DCM content in the aqueous phase during the curing process), and the data were screened to remove outliers. Then, different preprocessing method combinations were studied to eliminate the interference of irrelevant information. In order to further simplify and improve the predictive ability of the PLS model, the variable importance projection (VIP) algorithm was used to select the bands of the Raman spectra and compare them with the full spectrum. The model evaluation parameters RMSEC, RMSECV, R 2 C 、R 2 CV and R 2 P Finally, the eighth batch of samples was used as an external independent validation set to verify the predictive ability of the PLS model.

[0093] 1. Eliminate abnormal samples

[0094] The principal components analysis (PCA) method is used to detect abnormal samples. According to the PCA score graph, samples outside the 95% confidence ellipse are identified as abnormal samples. The results of abnormal sample removal are shown in Figure 4 .

[0095] Furthermore, the influence plot method is introduced, using F-residual and Hotelling T 2 Statistics to identify abnormal samples, with a confidence level of 95%, will exceed F-residual and Hotelling T 2 The samples with the threshold are classified as abnormal samples. The results of abnormal sample removal are shown in Figure 5 .

[0096] comprehensive Figure 4 and Figure 5, 7 samples were identified as outliers. These were samples with a high DCM concentration (the amount of DCM in the aqueous phase during the curing process) at an oil-water ratio of 1.5:100 (data from batch 7). Finally, after removing the outliers, 157 samples remained.

[0097] 2. Divide the sample

[0098] The Kennard-Stone method was used to group the data from batches 1 to 7 in Table 2, and the samples were divided into calibration and validation sets in a ratio of 3:1. The Kennard-Stone method in this study was performed in Matlab.

[0099] After grouping the samples using the Kennard-Stone method, we ultimately obtained a calibration set of 118 samples and a validation set of 39 samples. The reference values ​​for the samples in the calibration and validation sets are shown in Table 3.

[0100] Table 3 Reference values ​​of samples in the calibration and validation sets

[0101]

[0102] As shown in Table 3, the DCM concentration range of the validation set falls within the DCM concentration range of the calibration set, indicating that the sample set division is reasonable.

[0103] During the microsphere preparation process, the PCA score scatter plot of the sample set is shown in Figure 6 .

[0104] Depend on Figure 6 It can be seen that the validation set samples are evenly distributed in the calibration set samples, which further shows that the sample set division is reasonable.

[0105] 3. Screening spectral preprocessing methods

[0106] In order to eliminate the error of Raman spectrum acquisition and establish an accurate online Raman spectrum model, different spectrum preprocessing methods (Raw, MC, SNV, SNV+MC, MC+SNV, FD+SG+SNV+MC, SD+SG+SNV+MC, SNV+FD+SG+MC, SNV+SD+SG+MC, FD+SG+MC, SD+SG+MC, FD+SG+SNV, SD+SG+SNV, SNV+FD+SG, SNV+SD+SG) were adopted to preprocess the Raman spectra. The partial least squares method was selected as the modeling method. The parameters (R 2 C 、R 2 CV 、R 2 P , RMSEC, RMSECV and RMSEP) to select the best spectral preprocessing method.

[0107] The model result evaluation parameters of each model are shown in Table 4.

[0108] Table 4 Model result evaluation parameters of each model

[0109]

[0110] As shown in Table 4, for the rotigotine microsphere sample, when the spectral preprocessing method is SNV+FD+SG+MC, the model prediction effect is the best, and its RMSEC, RMSECV, RMSEP, R 2 C 、R 2 CV 、R 2 P The values ​​are 0.5857 mg / ml, 0.7436 mg / ml, 0.4223 mg / ml, 0.9878, 0.9804, and 0.9918 respectively.

[0111] After calculation, the Ratio of Performance Deviation (RPD) value of the optimal model is 10.23, and the model results are good.

[0112] 4. Optimize the spectrum range

[0113] The Variable Importance Projection (VIP) algorithm is used to screen variables for those that contribute most to the potential change in variable X. The Raman spectrum conversion method used in this study uses the VIP algorithm to screen Raman spectrum bands, compare them with full-spectrum modeling, and select the optimal modeling band.

[0114] The results of using VIP algorithm to filter the Raman spectrum band are shown in Figure 7 .

[0115] The comparison results of the prediction accuracy of the models before and after band screening (the original spectra were preprocessed using the SNV+FD+SG+MC method) are shown in Table 5.

[0116] Table 5 Comparison of model evaluation parameters before and after band screening

[0117]

[0118] It can be seen from Table 5 that the modeling effect before screening is better than the modeling effect after screening, that is, the effect of full-spectrum modeling is better than the effect of spectral interval modeling optimized by VIP algorithm.

[0119] Therefore, for the Rotigotine microsphere sample, full spectrum modeling was chosen.

[0120] 5. Build and evaluate the model

[0121] After eliminating abnormal samples, dividing the sample sets, and examining the rationality of spectral interval optimization, SNV+FD+SG+MC was finally selected as the spectral preprocessing method, the number of latent variables was selected as 4, and the partial least squares method was used in the full spectral range to establish an online quantitative analysis model for DCM content in the aqueous phase during the preparation of rotigotine microspheres (full-band PLS model).

[0122] The 25 samples from the 8th batch (not involved in model building) were used as the external validation set to externally validate the full-band PLS model. Regression was performed with Raman spectra as the X variable and the DCM content (mg / ml) in the aqueous phase during the curing process determined by headspace gas chromatography as the Y variable.

[0123] The comparison between the predicted value of DCM content in the water phase during the curing process and the reference value of the full-band PLS model external validation set is shown in Figure 8 , the relationship between the predicted value and the reference value is shown in Figure 9 .

[0124] Depend on Figure 8 and Figure 9 It can be seen that the reference value of DCM content in the aqueous phase during the microsphere preparation process is basically consistent with the DCM content predicted by the full-band PLS model. A certain deviation may be caused by the small number of samples.

[0125] 5. Study of in vitro release

[0126] In order to investigate the effect of DCM content in the aqueous phase on the release of microspheres and thus on the quality of microspheres, the in vitro release of the final product microspheres prepared from the 1st to 7th batches was studied.

[0127] 1. Prepare solution

[0128] Blank solution: acetonitrile-purified water (volume ratio 55:45).

[0129] pH 9.5 ammonium bicarbonate solution: Take 0.79g ammonium bicarbonate, accurately weighed, dissolved in 1000ml water, adjust the pH to 9.5 with ammonia water, and filter through a 0.22μm aqueous microporous filter membrane.

[0130] Release medium: Take about 4.68g of sodium dihydrogen phosphate dihydrate, about 7.02g of sodium chloride, and about 0.33g of Tween 80, accurately weigh them, dissolve them in water and dilute to 3000ml, adjust the pH to 7.4 with 1.75mol / L sodium hydroxide solution, add 6g of sodium lauryl sulfate and 0.6g of sodium azide, dissolve them, and stir evenly.

[0131] Reference substance solution: Take approximately 25 mg of rotigotine reference substance, accurately weigh it, place it in a 100 ml volumetric flask, add an appropriate amount of blank solution, dissolve it by ultrasonication, dilute it to the scale, shake well, and obtain 0.25 mg / ml reference substance stock solution; accurately pipette 5 ml of rotigotine reference substance stock solution into a 100 ml volumetric flask, dilute it to the scale with blank solution, shake well, and obtain 12.5 μg / ml rotigotine reference substance solution I; accurately pipette 4 ml of rotigotine reference substance solution I into a 100 ml volumetric flask, dilute it to the scale with blank solution, shake well, and obtain 0.5 μg / ml rotigotine reference substance solution II.

[0132] Test solution: Approximately 20 mg of the final product, rotigotine microspheres, prepared from batches 1 to 7, were accurately weighed and placed in 250 ml dry, stoppered glass bottles in triplicate. The release medium was preheated to 37°C in an electrically heated constant-temperature water bath. Accurately measure 200 ml of the release medium and add it to the sample bottle. The vial was stoppered and placed in a 37°C electrically heated constant-temperature water bath. At 40 minutes, 2 hours, 6 hours, 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days, and 8 days, 4 ml of the supernatant was removed and replaced with an equal volume of isothermal release medium. Filter 3 ml of the removed solution through a 0.45 μm polytetrafluoroethylene filter membrane, and 1 ml of the filtrate was used as the test solution.

[0133] 2. Chromatographic conditions and calculation methods of drug loading and encapsulation efficiency

[0134] (1) Chromatographic conditions

[0135] The chromatographic conditions are as follows:

[0136] Column: XSelect CSH C18 (4.6 × 50 mm, 3.5 μm);

[0137] Mobile phase: acetonitrile-pH 9.5 ammonium bicarbonate solution (volume ratio 55:45);

[0138] Column temperature: 40°C;

[0139] Flow rate: 1.8 ml / min;

[0140] Detection wavelength: 230 nm;

[0141] Injection volume: 20 μl;

[0142] Running time: 10 minutes.

[0143] (2) Calculation method of drug loading and encapsulation efficiency

[0144] The drug loading and encapsulation efficiency were calculated according to the following formula:

[0145] Drug loading (%) = (drug content in microspheres / microsphere sample weight) × 100%;

[0146] Encapsulation efficiency (%) = (actual drug loading / theoretical drug loading) × 100%.

[0147] 3. Calculate in vitro release

[0148] 25 mg of rotigotine microspheres were accurately weighed into a glass vial, added to 200 ml of pH 7.4 phosphate buffer, and placed in a 37°C electrically heated water bath. 4 ml of the supernatant was removed at 40 minutes, 2 hours, 6 hours, 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days, and 8 days, and an equal volume of buffer was added. Drug concentration was determined by high-performance liquid chromatography using an XSelect CSH C18 (4.6 × 50 mm × 3.5 μm) column with a mobile phase of acetonitrile-pH 9.5 ammonium bicarbonate (55:45, by volume) at a flow rate of 1.8 ml / min, detection wavelength at 230 nm, and an injection volume of 20 μl. All release experiments were performed in triplicate, and all data were processed using Origin.

[0149] The in vitro release of the final product rotigotine microspheres prepared from different batches 1 to 7 was determined according to the above method (n=3). Among them, the 0.5 μg / ml rotigotine reference solution II was used as the reference solution for the test sample at 40 min-6 h, and the 12.5 μg / ml rotigotine reference solution I was used as the reference solution for the test sample at other time points.

[0150] The cumulative release calculation results of the final product rotigotine microspheres prepared in different batches are shown in Table 4. Figure 10 The cumulative release calculation results of the final product Rotigotine microspheres prepared in different batches are shown in Figure 11 .

[0151] Depend on Figure 10 and Figure 11 It can be seen that the final product rotigotine microspheres prepared in the 7th batch (oil-water ratio 1.5:100) showed a higher burst release amount and a faster overall release rate. The release rate in the early and middle stages of release was higher than the preset standards.

[0152] According to USP <1092> The guidelines for dissolution method development and validation stipulate that extended-release dosage forms must have at least three control points for release: early, mid, and late release. Dose dumping should not occur in the early release period. The mid-release time point defines the in vivo release profile of the dosage form, and the final time point indicates that the drug is essentially released. To better control product quality, based on the in vitro release profile, 2 hours is selected for controlling burst release, 3 days is selected for controlling mid-release release, and 8 days is selected for controlling the release endpoint.

[0153] Based on the accumulation of multiple batches of data from clinical batches and stability batches, the preset standards for the release of rotigotine microspheres are tentatively set as follows: the 2-hour cumulative release should not exceed 2%, the 3-day cumulative release should be between 37%-73%, and the 8-day cumulative release should not be less than 85%, in order to evaluate the quality of the microspheres.

[0154] In summary, the present invention provides a new method for online real-time monitoring of the DCM content in the aqueous phase during microsphere preparation, which is beneficial to quality control during the microsphere preparation process, thereby improving the production efficiency of microspheres and reducing the loss of manpower and material resources.

[0155] It should be noted that the above embodiments are merely examples for the purpose of clearly illustrating the present invention and are not intended to limit the embodiments of the present invention. A person skilled in the art would be able to make other variations or modifications based on the above description. It is not possible to enumerate all embodiments here. Any obvious variations or modifications arising from the technical solution of the present invention remain within the scope of protection of the present invention.

Claims

1. An online measurement method for the dichloromethane content in the aqueous phase during microsphere preparation, characterized in that: The following steps are involved: (1) Design multiple batches of microsphere preparation experiments by changing a single factor within the range of normal and abnormal formulation process conditions; (2) Start preparing microspheres, use an online Raman spectrometer to collect original spectra during the entire curing process, and record the data collection time and Raman spectrum; (3) Analyze the dichloromethane content in the aqueous phase during the curing process by headspace gas chromatography; (4) The original Raman spectra collected from each batch were paired one by one with the dichloromethane content in the aqueous phase during the curing process determined by gas chromatography. The principal component analysis method was used to detect abnormal samples. According to the PCA score graph, samples outside the 95% confidence ellipse were identified as abnormal samples. The influence diagram method was further introduced, and the F-residual and Hotelling T 2 Statistics are used to identify abnormal samples, with the confidence level set to 95%, which will exceed the F-residual and Hotelling T 2 The samples with the threshold are classified as abnormal samples. The abnormal sample elimination results of the two methods are combined to determine and remove the final abnormal samples. (5) The Kennard-Stone method is used to group each batch of data into a calibration set and a validation set; (6) Different spectral preprocessing methods were used to preprocess the Raman spectra. The spectral preprocessing methods included: Raw, MC, SNV, SNV+MC, MC+SNV, FD+SG+SNV+MC, SD+SG+SNV+MC, SNV+FD+SG+MC, SNV+SD+SG+MC, FD+SG+MC, SD+SG+MC, FD+SG+SNV, SD+SG+SNV, SNV+FD+SG ​​and SNV+SD+SG. The partial least squares method was selected as the modeling method. The best spectral preprocessing method was selected based on the evaluation parameters of the model results. (7) The variable importance projection algorithm is used to screen the Raman spectrum bands, compare them with the full spectrum modeling, and select the optimal modeling band; (8) Under the conditions of optimal spectral preprocessing method, 4 potential variables, and optimal modeling band, the partial least squares method was used to establish an online quantitative analysis model for the dichloromethane content in the aqueous phase during the microsphere preparation process.

2. The method according to claim 1, characterized in that In step (1), the factors include: oil-water ratio, compressed air flow rate and curing temperature.

3. The method according to claim 1, characterized in that In step (2), the parameters of the online Raman spectrometer are set as follows: laser intensity of 300 mW, exposure time of 3 s, number of scans of 10 times, spectrum acquisition every 45 s, wavelength range of 100 cm -1 -3200cm -1 .

4. The method according to claim 1, wherein In step (3), the method for analyzing the dichloromethane content in the aqueous phase during the curing process is: (i) Preparation of aqueous phase samples during the curing process: 0.1 mL of the aqueous phase was accurately measured and placed in a 10 mL headspace vial at different time points during the curing process. 0.9 mL of N,N-dimethylacetamide was added, and the vial was sealed and shaken to mix. The time points were consistent with the time points at which the Raman spectra were collected. (ii) Prepare a reference solution: Dissolve dichloromethane in N,N-dimethylacetamide to obtain a 0.5 mg / mL reference solution; (iii) Gas chromatograms were collected by gas chromatograph using headspace gas chromatography to collect gas chromatograms of aqueous samples and reference solutions during the curing process; (iv) Calculate the dichloromethane content in the aqueous phase during the curing process using gas chromatography. The calculation formula is: ； Among them, A 供试品 is the peak area of ​​DCM in the test solution, A 对照品 is the average peak area of ​​DCM in the reference solution, C 对照品 is the concentration of DCM in the reference solution, in mg / mL.

5. The method according to claim 1, wherein In step (5), the samples are divided into a calibration set and a validation set in a ratio of 3:1.

Images