4.3. Stability Study

The stability study was performed following the guidelines for the practical stability studies of anticancer drugs from a European consensus conference, published by the French Society of Oncology Pharmacy (SFPO) [1].

The guidelines were based on ICH guidelines, particularly ICH Q1A (evaluation for stability data), ICH Q1A(R2) (stability testing of new drug substances and products), ICH Q2A (test on validation of analytical procedures), ICH Q1B (stability testing: photostability testing of new drug substances and products), Q3B (impurities in new drug products), Q5C (stability testing of biotechnological/biological product), European Pharmacopeia (Ph. Eur.), EMA guidelines, and the most relevant literature [3,25,27,29,30,31,32,33].

The chemical stability of azacitidine in aqueous suspension was evaluated using the HPLC–UV system and the chromatographic conditions described later on.

The sample stored in the original container (condition A) was analyzed at time zero and times 4, 8, 12, and 22 h from preparation. Other samples (conditions B and C) were analyzed at time zero and times 4, 8, 12, 24, 36, 48, 54, 60, 64, 68, 72, and 96 h.

At each analyzing time, the Vidaza^® suspensions (25 mg/mL) were placed at room temperature for five minutes, then a 20 µL aliquot of each test sample was diluted with refrigerated (2–8 °C) sterile water for injection to a final concentration of 50 µg/mL to obtain a dilute solution for chromatographic analysis.

To calculate the azacitidine concentration (mg/mL) at each time point, the measured azacitidine % relative area was compared to the initial theoretical azacitidine % relative area (100%) that was correlated to the initial theoretical concentration of 25 mg/mL.

From the obtained azacitidine concentration (mg/mL), the percentage loss of azacitidine at each time point relative to the initial experimental concentration value was calculated.

The experiments were performed on triplicate samples (on three different lots of Vidaza^®). The data were expressed as mean ± standard deviation (S.D.) and reported in a summary table.

The mean loss of azacitidine occurring at 22 h from preparation in three different lots of Vidaza^®, reconstituted and stored refrigerated (2–8 °C) in the original container according to SPC (condition A) was considered as the limit of chemical stability.

The chemical stability limit (as loss of azacitidine) was based on the remaining percentage of the initial experimental concentration value, which was calculated relative to the initial theoretical concentration (25 mg/mL).

HPLC analysis was performed on a Thermo Scientific Dionex Ultimate 3000 HPLC system (Thermo Scientific, Bremen, Germany) equipped with an LPG-3400SD pump, TCC-3000 column oven, and UV VWD-3100 detector.

According to the United States Pharmacopeia’s (USP) pending monograph for azacitidine [34] and certificate of analysis (CoA) of azacitidine reference material supplied by Merck Life Science S.r.l. (Milan, Italy), the HPLC analysis using a reversed-phase high-performance liquid chromatography (RP-HPLC; Ascentis Express C18, 150 mm × 4.6 mm, 2.7 μm; Merck Life Science S.r.l. (Milan, Italy) with a linear A-B gradient (0–4.8 min 0% B, 4.8–12 min 0% to 15% B, 12–15 min 15% B, 15–18 min 15% to 30% B, 18–24 min 30% to 50% B, 24–27 min 50% to 0% B, 27–33 min 0% B) at a flow rate of 0.8 mL/min and a total run time of 33 min was performed. Solvent A consisted of 1.54 g/mL ammonium acetate in water (0.02 M, pH 6.9 ± 0.1) and solvent B consisted of solvent A:methanol:acetonitrile (50:30:20).

UV absorbance was measured at 210 nm. The column temperature was kept at 30 °C. The injection volume was 20 μL.

The Chromeleon data system software (Version 7.2.8) was used for data acquisition and mathematical calculations.

The extensive validation of the analytical method was carried out according to ICH Q2(R1) guidelines [27] (see Supplementary Materials).

A forced degradation study was conducted out on one Vidaza^® preparation (25 mg/mL), to test the specificity and the suitability of the chromatographic method for use as a stability-indicating assay.

As a degradation test is designed to increase the rate of chemical degradation of the drug and determine the nature and chromatographic peaks of all DPs, the sample was exposed at 50 °C/43% RH for 12 h.

The used conditions were such to not obtain a drug degradation of more than 20% in order not induce the formation of DPs completely different from those observed in daily practice.

To identify unknown impurities/degradation products formed during the proposed new storage paradigm of reconstituted azacitidine solutions, whose peaks were detected in the HPLC-UV trace, as to exclude their toxic potential, an ultra-high-performance liquid chromatography high-resolution mass spectrometry (UHPLC–HRMS) of samples stored in the condition C was performed.

Briefly, 2 µl of the 50 µg/mL solution were injected into a Thermo Scientific Dionex Ultimate 3000 UHPLC coupled to a Thermo Ultrahigh-resolution Q Exactive mass spectrometer (Thermo Scientific, Bremen, Germany). The column (Ascentis Express C18, 150 mm × 4.6 mm, 2.7 μm; Merck Life Science S.r.l. (Milan, Italy)), thermostatted at 30 °C, was equilibrated with 0.8 mL/min of 1.54 g/mL ammonium acetate in water (0.02 M, pH 6.9 ± 0.1) (solvent A); after 4.8 min from the sample injection, solvent B (solvent A:methanol:acetonitrile 50:30:20) was linearly increased from 0 to 15% in 7.2 min; B% was then kept constant for 3 min, then brought to 30% in 3 min. From minutes 18 to 24 B% was raised to 50% and brought back to 0% B for the reconditioning step. Each sample required a total run time of 31 min. Centroided MS and MS² spectra were recorded in both positive and negative polarities from 100 to 1500 and 200 to 2000 m/z in full MS/dd-MS² (TOP2) mode, at a resolution of 70,000 and 17,500, respectively. The two most intense ions were selected for MS² nitrogen-promoted collision-induced dissociation (NCE = 30). Precursor dynamic exclusion (15 s) and apex triggering (1 to 6 s) were set. The mass spectrometer was calibrated before the start of the analyses.

To assess the chemical changes to azacitidine structure over time, a Fourier transform infrared spectroscopy (FTIR) analysis was performed. The spectra were obtained with 32 scans in a Bruker Vertex 70 V FT-IR spectrometer (Bruker Optics, Ettlingen, Germany), equipped with a Hyperion microscope attachment.

The sample coated CaF₂ slides were placed under the microscope objective and IR spectra were recorded in transmission mode from 4000 to 650 cm⁻¹ at a spectral resolution of 4 cm⁻¹. The spectra were collected using an attenuated total reflectance (ATR) diamond crystal (KRS-5 lens, Golden Gate model GS10542-K; Specac, Inc., Fort Washington, PA, USA) positioned within the optical bench of the spectrometer.

The sample stored in the original container (condition A) was analyzed at time zero and 22 h from preparation. Other samples (conditions B and C) were analyzed at time zero and times 24, 48, 72, and 96 h. The experiments were performed on three different lots of Vidaza^®.

The determination of the pH value was performed with an electrochemical method using one micro-electrode and a millivoltmeter (pH meter) from Thermo Scientific™ Orion™ Dual Star.

The sample stored in the original container (condition A) was analyzed at time zero and 22 h from preparation. Other samples (conditions B and C) were analyzed at time zero and times 24, 48, 72, and 96 h. The experiments were performed on three different lots of Vidaza^®.

Whenever samples were taken for analysis, vials and syringes were visually checked to assay the change in the initial color or appearance or particulate matter of the suspension.

To evaluate the changes in terms of shape, size, and number of particles, as well as to examine any sign of physical instability, such as aggregation or particle precipitation, microscopic observation and the particle counter were performed.

Microscopic observation was performed following the 2.9.37 current test of European Pharmacopeia [35].

Since in the hospital laboratory, the method light obstruction [36] or turbidimetry [37] based was not available, the size and quantity of particles were evaluated by image-based cytometry [38], which is known to provide comparable data to traditional flow cytometry [39].

The sample stored in the original container (condition A) was analyzed at time zero and 22 h from preparation. Other samples (conditions B and C) were analyzed at time zero and times 24, 48, 72, and 96 h. The experiments were performed on three different lots of Vidaza^®.

The evaluation of crystal morphology was performed on microscopy slides loaded with samples and acquired using a computer (equipped with uEye UI-1460LE-C features a 1/2 inch CMOS sensor with a 2048 × 1536 pixel resolution color sensor) and the program MultiScan v.8.08 Computer Scanning System. The computer was connected to an Olympus BX 40 microscope with a 10× objective (NA 0.25).

The quantity and the size of azacitidine particles were evaluated with Tali^® Image-Based cytometer. For the assay, the Tali^® cellular analysis slides were used. The slide holds the sample in two separate, enclosed chambers. In each chamber, 25 μL of the sample were loaded. The Tali^® Image-Based cytometer captures a series of images (i.e., fields of view) of the sample in the chamber and then analyzes them using algorithms specifically designed to determine total particle counts in a range between 0–60 µm and calculates their concentrations in 1 mL.

Classically, it is considered that many anticancer drugs do not facilitate bacterial growth. Moreover, thanks to the application of good hospital pharmacy manufacturing practice rules, sterile conditions were guaranteed during the manufacturing process, preventing bacterial contamination. Nevertheless, since the maintenance of the sterility in the final container also depends on the nature of the container and the storage conditions, the sterility assay was performed.

The study took place at the Biochem Microbiology Laboratory (Zola Pedrosa, Bologna, Italy).

The sterility assay was performed in triplicate for each storage condition at 22 (condition A) and 96 (conditions B and C) hours from preparation, respectively.

The methodology of the test followed the 2.6.1 current test of the European Pharmacopoeia [40].

The seeding was carried out under a vertical laminar air-flow hood in aseptic conditions and the containers were decontaminated externally with 70° of alcohol.

Then, 1.5 mL of each sample was transferred directly into the thioglycollate medium for the detection of aerobic and anaerobic micro-organisms and into the tryptone soya broth medium for the detection of fungi. Tubes were then incubated for 14 days at 22 ± 2 °C and 32 ± 2 °C, respectively, and observed at 4 and 14 days of incubation. The results were considered satisfactory if no evidence of microbial growth is found. Appropriate negative controls were included.