Claims:
A method for preparing recombinant Human Insulin crystal comprising the steps of:
a. crystallizing the recombinant Human Insulin in a solution containing a recombinant Human Insulin, an organic solvent, and a mixture of Zinc chloride and salt;
b. adjusting the pH of crystallization solution in the range of 4.8 to 5.2;
c. neat settling of crystallization solution at room temperature followed by neat settling at chilling temperature; and
d. freeze drying of slurry and obtaining recombinant Human Insulin crystal.
The method according to claim 1, wherein the concentration of the recombinant Human Insulin in the crystallization solution is 5.0±0.2 g/L.
The method according to claim 1, wherein the organic solvent is selected from acetonitrile, ethanol, n-propanol and isopropyl alcohol.
The method according to claim 3, wherein the preferable organic solvent is isopropyl alcohol.
The method according to claim 4, wherein concentration of isopropyl alcohol is 19-25 million ppm, preferably 21 million ppm.
The method according to claim 1, wherein the concentration of the zinc chloride is 0.3-0.5 ml per gram of recombinant human insulin.
The method according to claim 1, wherein the salt is selected from sodium chloride, sodium acetate and sodium citrate.
The method according to claim 7, wherein the salt is preferably sodium chloride.
The method according to claim 8, wherein the concentration of sodium chloride in the crystallization solution is 0.5M.
The method according to claim 1, wherein the pH of the crystallization solution is adjusted to 5.0 using 3M acetic acid.
The method according to claim 10, wherein the pH is adjusted within 5 minutes following addition of mixture of zinc chloride and sodium chloride in the crystallization solution.
The method according to claim 1, wherein the crystallization solution is allowed to settle during neat settling for 2.5 to 4 hours at temperature between 21°C to 27°C.
The method according to claim 1, wherein the crystallization solution is further allowed to settle for 10 to 12 hours at temperature between 2°C to 8°C.
The method according to preceding claims, wherein the bed height of freeze-drying tray is 1.1 cm.
The method according to claim 1, comprising the steps of:
preparing a crystallization solution containing 5.0 g/L of recombinant human insulin, 21 million ppm of isopropyl alcohol, 4% zinc chloride, 0.5M of sodium chloride;
adjusting pH value of the crystallization solution to 5.0 using acetic acid;
holding the solution at temperature 24±3°C for 2.5 to 4 hours for neat settling;
cooling the crystallization solution to a temperature of 2-8°C for 10-16 hours; and
freeze drying and thereby obtaining the recombinant human insulin crystal.
The method according to claim 1, comprising the steps of:
diluting HPLC elution pool by water for irrigation and isopropyl alcohol till solution contains 5 g/L of recombinant human insulin and targeting 21 million ppm of isopropanol; thereto adding mixture of 4% zinc chloride and 0.5M sodium chloride at 0.006 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s;
adjusting the pH to 5.0 using 3M acetic acid within 5 minutes at 0.1 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s, wherein, post pH adjustment agitation is continued for 15-20 minutes at 0.21 m/s;
adjusting the temperature of the above crystallization solution to 23±3°C upon stopping agitation to allow neat settling for 2.5 to 4.0 hours; then allow the neat settling at 2-8°C for 10-12 hours;
decanting 85-90% of the supernatant, adding the chilled water to slurry in agitated state at a tip speed of 0.21m/s for up to 5 minutes, then keeping at 2-8°C for 16 hours; and
decanting the supernatant and keeping slurry into the freeze dryer for drying.
The method according to preceding claims, wherein consistent crystal geometry and size of recombinant Human Insulin is obtained between 15µm -30µm.
, Description:FIELD OF INVENTION
The present invention relates to a method for peptide crystallization, particularly preparing recombinant human insulin crystal.

BACKGROUND OF INVENTION
Diabetes is a common endocrine and metabolic disorder. It’s a long-term condition that causes high blood sugar levels. In type 1 Diabetes, the body does not produce Insulin, also referred to as Insulin-dependent diabetes, juvenile diabetes, or early-onset diabetes. In type 2 Diabetes, the body does not produce enough insulin for proper function, or the cells in the body do not react to Insulin (Insulin resistance). Presently patients with type 1 diabetes are treated with regular insulin injections along with a special diet and exercise. Patients with Type 2 diabetes are treated with tablets, exercise and a special diet, but sometimes insulin injections are also required.

‘Insulin therapy’ has always been considered as an important means for treating diabetes and controlling blood sugar level. Purification of recombinant Human Insulin achieved by multiple downstream unit operations that involve a combination of crystallization, enzyme catalysis and chromatography. The requirement of the highest purity of Human Insulin is to ensure the patients do not develop immunogenic or toxic responses to the drug product.

Recombinant Human Insulin crystallization occurs in two phases. The first phase is nucleation, the appearance of a crystalline phase from either a super cooled liquid or a supersaturated solvent. The second phase is crystal growth, which is the increase in the size of particles and leads to a crystalline state. The crystal form of recombinant Human Insulin is a better form, since it has a uniform and steady solid molecular form and small sediment volume, and is easy to separate from the supernatant, the time for centrifugation and freeze-drying is short, and the production efficiency is relatively high. It is thus desirable to prepare recombinant Human Insulin crystals and then apply the crystals to Insulin pharmaceutical preparations.

Commercial Insulin manufacturing processes typically include a crystallization step to convert soluble purified Insulin into solid form, providing increased stability for bulk storage prior to formulation and filling. Classical Insulin crystallization process as disclosed in U.S. Pat. No. 2910014 includes preparation of an acidic solution containing organic acid (acetic or citric), approximately 2 g/L Insulin, and zinc and adjustment of the solution pH to near the isoelectric point of insulin (pH 5.5-6.0), which initiated crystal formation.

It is well known in the art that Insulin may be crystallized in the presence of zinc ions, resulting in a crystalline preparation with significant benefits over amorphous, un-crystallized Insulin with regard to stability, storage, formulation, and/or administration. In the presence of zinc, human insulin self-assembles into stable hexameric structures. Zinc content plays an important role in chemical and physical stability of pharmaceutical insulin formulations.

The use of ZnCl2 for crystallization of recombinant Human Insulin (rHI) is an established and well published technique, however, a similar knowledge is not available for crystallizing HI at a preparative scale. Traditional recombinant Human Insulin downstream purification process involves three crystallization steps termed as crystallization-1, crystallization-2 and crystallization-3, respectively as per the order of the operation. Controlling the level of aggregates, residual zinc and other related impurities during drug preparation is a critical quality requirement to make the final drug product complying with the specifications of the innovator. In the recent manufacturing batches of the biosimilar Insulin process, it was noted that level of high molecular weight protein (HMWP) and other related impurities were on the higher level. Crystallization-3 is a potential step, which is a final step of crystallization, where the chances of formation of aggregates (HMWP) and other related impurities are imminent. Improper settling during the neat settling step was observed during this final crystallization stage(s). This suboptimal performance leads to the lower decantation percentage at neat and wash-1 stage resulting in suboptimal freeze-drying performance.

Therefore, there is a need to control the level of these impurities at the appropriate downstream step.

OBJECT OF INVENTION
An object of the present invention is to overcome the various key process challenges observed during the final crystallization & freeze drying stages of recombinant Human Insulin preparation and accordingly modify the process.

Another object of the present invention of preparative peptide crystallization is to obtain consistent crystal geometry and size of recombinant Human Insulin at preparative scale.

SUMMARY OF INVENTION
In one aspect the present invention provides a method for preparing recombinant Human Insulin crystal comprising the steps of:
crystallizing the recombinant Human Insulin in a crystallization solution containing recombinant Human Insulin, an organic solvent, a zinc compound, a salt, such that, mixture of the zinc compound and salt was added together to the solution;
pH was adjusted to 4.8 to 5.2 ; and
the crystallization solution maintained at ambient temperature before freeze-drying.

In another aspect the present invention provides a method comprising the steps of:
diluting HPLC elution pool by water for irrigation and isopropyl alcohol till solution contains 5 g/L of recombinant human insulin and targeting 21 million ppm of isopropanol; thereto adding mixture of 4% zinc chloride and 0.5M sodium chloride at 0.006 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s;
adjusting the pH to 5.0 using 3M acetic acid within 5 minutes at 0.1 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s, wherein, post pH adjustment agitation is continued for 15-20 minutes at 0.21 m/s;
adjusting the temperature of the above crystallization solution to 23±3°C upon stopping agitation to allow neat settling for 2.5 to 4.0 hours; then allow the neat settling at 2-8°C for 10-12 hours;
decanting about 85-90% of the supernatant, adding the chilled water to slurry in agitated state at a tip speed of 0.21m/s for up to 5 minutes, then keeping at 2-8°C for 16 hours;
decanting the supernatant and keeping slurry into the freeze dryer for drying.

BRIEF DESCRIPTION OF DRAWINGS
Figure 1 represents flow chart of the process of crystallization of present invention.
Figure 2 represents flow chart of the process of crystallization followed prior to present invention.
Figure 3 represents the fishbone diagram of parameters studied in crystallization process of present invention.
Figure 4 represents the observations on reagent addition strategy 1. The red line indicates the bed height.
Figure 5A represents the observations on reagent addition strategy 2.
Figure 5B represents the observations on reagent addition strategy 2.
Figure 6 represents the observations of impact of IPA content in FFC.
Figure 7A represents the observations of impact of rate of pH adjustment.
Figure 7B represents the observations of impact of rate of pH adjustment.
Figure 8 represents RP-HPLC 3 dynamic binding capacity study flow chart.
Figure 9 represents the crystal size 15µm -30µm and 40x images of recombinant human insulin at various concentrations in FFC upon scale-up.
Figure 10A represents the recombinant Human Insulin crystals prepared by traditional process wherein only zinc chloride is used.
Figure 10B represents the recombinant Human Insulin crystals prepared by process of present invention wherein mixture of zinc chloride and sodium chloride is used.
Figure 10C represents the recombinant Human Insulin- drug substance crystals.

DETAILED DESCRIPTION OF INVENTION
Definitions
Unless otherwise defined herein; the scientific and technical terms used in connection with the present invention shall have the meanings that are, commonly understood by those of ordinary skill in the art. Further, unless otherwise required by context, singular terms shall include pluralities and plural terms shall include the singular. The methods and techniques of the present invention are generally performed according to conventional methods well known in the art. The nomenclatures used in connection with, and techniques described herein are those commonly used in the art. The methods and techniques of the present invention are generally performed according to conventional methods well known in the art.

The term ‘ambient temperature’ refers to the air temperature of an environment or object or surrounding an equipment. Room Temperature (RT) is generally defined as the ambient air temperature. In present invention, ambient temperatures or room temperature can range between 20 and 26 °C.

The term ‘crystallization’ refers to the solid-liquid separation and purification technique in which mass transfer occurs from the liquid solution to a pure solid crystalline phase.
The term ‘Insulin’ refers to a hormone secreted by the islets of Langerhans in the pancreas; regulates storage of glycogen in the liver and accelerates oxidation of sugar in cells.

The term ‘recombinant Human Insulin’ and ‘rHI’ refer to a form of insulin made from recombinant DNA that is identical to human insulin.

The present inventors studied the process deviations for multiple batches at final crystallization and freeze-drying step and found that all deviations were due to the poor neat settling during the final crystallization process (also corroborated by the smaller crystal size i.e. <1 µm). After a thorough root cause analysis (elaborated through the examples below), following factors have been identified to be affecting the crystal size and subsequent settling-
IPA content in final FFC (feed for crystallization)
Chronology of reagent addition (NaCl first, followed by ZnCl2)
Insufficient ambient temperature (24±3°C) incubation after final pH (5±0.1) adjustment
Rate of addition of acetic acid (3M) for pH adjustment

Further, freeze drying failure (improper product drying) was primarily due to the cascading effect of the suboptimal settling in preceding crystallization step. This resulted in higher slurry volume post wash decantation that was almost 12% higher than the expected volume. This led to the loading of each freeze-drying tray with higher bed height and lower slurry percentage (5%). It was found that lower ambient temperature resulted in poor settling while higher hold duration would lead to increase in aggregates (HMWP). It was further observed that at lower slurry percentage higher bed height is detrimental for efficient product drying.

Figure 1 describes the process of recombinant Insulin crystallization followed in present invention whereas figure 2 describes the old process of Insulin crystallization followed prior to the present invention. To mitigate the above mentioned factors affecting crystal size and subsequent settling, the observations made at manufacturing scale wherein process parameters as captured in figure 3 were studied.

In order to mitigate the problems of the old process, the present invention provides an improved method for preparing recombinant Human Insulin crystal comprising the steps of crystallizing the recombinant Human Insulin in a crystallization solution containing recombinant Human Insulin, an organic solvent, a zinc compound, a salt, such that, the mixture of zinc compound and the salt are added together in crystallization solution followed by a pH adjustment to 4.8 to 5.2, preferably 5.0.

The present invention, particularly provides an improved crystallization-3 process with FFC targeting 21 million IPA content (ppm)..
The concentration of the recombinant human insulin in the crystallization solution is 5.0±0.2 g/L.
In the present invention, the organic solvent is selected from acetonitrile, ethanol, n-propanol and isopropyl alcohol, preferably isopropyl alcohol. The concentration of isopropyl alcohol is from 19 to 25 million ppm, preferably 21 million ppm.

The zinc compound is selected from zinc chloride, zinc oxide, zinc acetate, zinc bromide and zinc sulfate. The salt is selected from sodium chloride, sodium acetate and sodium citrate. Preferably, the solution contains 4% zinc chloride at a concentration of 0.3-0.5 ml per gram of recombinant Human Insulin and 0.5M sodium chloride, which is added as a solution to the crystallization solution.

The pH of the crystallization solution was adjusted in the range of 4.8 to 5.2 using 3M acetic acid within 5 minutes after addition of mixture of zinc chloride and sodium chloride mixture in the crystallization solution.

In a preferred embodiment, the crystallization solution is held at the ambient temperature of 24±3°C for 2.5 to 4 hours for neat settling, upon which, the chilling of neat settling is achieved at 2-8°C for 10-12 hours. Further, the slurry is held at 2-8°C for another 16 hours. The slurry obtained therein is freeze dried.

Elution pool of HPLC was diluted by WFI and IPA followed by addition of solution of ZnCl2+NaCl mixture. The adjustment of 5.0 pH was achieved within 5 minutes of such addition. It was surprisingly found that the faster rate of pH adjustment and the maintenance of ambient/ room temperature is vital for protein crystallization, settling and consistent crystal size. The steps of neat settling and complete wash are performed at cold temperature (5±3°C), which adds to the process robustness and better control of critical quality attributes at final drug substance stage. The 1.1 cm bed height of loading tray used during freeze-drying step for efficient product drying.

In a particular embodiment, the present invention provides a method comprising steps of:
diluting HPLC elution pool by water for irrigation and isopropyl alcohol till solution contains 5 g/L of recombinant human insulin and targeting 21 million ppm of isopropanol; thereto adding solution mixture of 4% zinc chloride and 0.5M sodium chloride at 0.006 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s;
adjusting the pH to 5.0 using 3M acetic acid within 5 minutes at 0.1 volume per volume per minute under stirring condition at a tip speed of 0.42-0.52 m/s, wherein, post pH adjustment agitation is continued for 15-20 minutes at 0.21 m/s;
adjusting the temperature of the above crystallization solution to 23±3°C upon stopping agitation to allow neat settling for 2.5 to 4.0 hours; then allow the neat settling at 2-8°C for 10-12 hours;
decanting about 85-90% of the supernatant, adding the chilled water to slurry in agitated state at a tip speed of 0.21m/s for up to 5 minutes, then keeping at 2-8°C for 16 hours;
decanting the supernatant and keeping slurry into the freeze dryer for drying.

The present method yields recombinant Human Insulin having a consistent crystal size of about 15µm -30µm. It also reduces the time required for sedimentation at manufacturing scale to about 12 hours compared to traditional process wherein the sedimentation step requires 24-60 hours whereas the present invention achieves the same in 12 hours.

Materials and method
Table 1 elaborates the material and the grade of material used for the altering of crystallization-3 process for crystallizing recombinant human insulin.
Sr. No. Name Grade
ACN (Acetonitrile) CG (Commercial Grade)
CH3COOH (Acetic acid) MCG (Multi Compendial grade)
MgCl2 (Magnesium Chloride) SQ (Special Quality)
Glacial Acetic acid ExcelaR
Tris Buffer AR (Analytical Reagent)
Sodium Sulphate anhydrous MCG
L-Arginine Reagent Grade
ZnCl2 (Zinc Chloride) MCG
Ortho-phosphoric acid HPLC Grade
(High Performance Liquid Chromatography Grade)
Acetonitrile Gradient HPLC Grade
Acetic acid HPLC Grade
TFA (Tri-Fluoroacetic Acid) HPLC Grade
HCl (Hydrochloric Acid) ACS (American Chemical Society) reagent
Acetophenone Reagent Plus
IPA (Isopropyl Alcohol) CG (Commercial Grade)
NaOH (Sodium Hydroxide) SQ (Special Quality)
NaCl (Sodium Chloride) USP (United States Pharmacopeia) grade
H2SO4 (Sulfuric Acid) AR
Ultra clean LDPE (Low Density polyethylene) sheets Medical grade
Ultra clean Tyvek sheets Medical grade
Table 1: Material and its grade

Table 2 elaborates the reagents and method of its preparation used for the present invention of altering of crystallization-3 process for crystallizing recombinant human insulin.
Sr. No Reagent Name Preparation
0.01N HCl Measure 900mL of water and then add 0.86mL of concentrated HCl (11.65N). Make up the final volume to 1L using MilliQ water.
1 1M Acetic acid Measure 750mL of water and then add 57.5 ml of Glacial acetic acid (17.4M). Make up the final volume to 1L using purified water, filter through 0.2 micron PES (polyethersulfone) filter.
2 Diluent (for in-process samples): 95% (0.1% TFA water) + 5% HPLC ACN Measure 949.05mL of water and then add 0.95mL of HPLC grade TFA. Make up the final volume to 1L using HPLC grade Acetonitrile.
3 2M Tris Dissolve 242.28g of Tris base in 900ml purified water and make up the volume to 1L using purified water. Filter through 0.2 PES filter.
4 RP-HPLC 2 Mobile phase A:
100mM Tris (Hydroxy methyl amino methane) buffer + 20mM Magnesium Chloride at pH8.50±0.10 For preparation of 1.0L, add 12.114g of Tris and 4.066 g of MgCl2 in 800 ml of purified water and adjust the pH 8.5±0.05 using acetic acid and make up the volume to 1L using purified water, filter through 1.2 PP (Polypropylene) and 0.2 micron PES (Polyether Sulphone) filter.
5 RP-HPLC 3 Mobile phase A:
25mM Acetic acid For preparation of 1.0L, add 1.43mL of acetic acid to 998.56mL of WFI (Water For Irrigation) and filter through 1.2 PP and 0.2 micron PES filter.
6 25mM Sodium hydroxide Dissolve 1.0g of Sodium hydroxide pellets in 700ml of purified water. Make up the volume to 1L using purified water, filter through 1.2 and 0.2 micron filter.
7 Crystallization-3:
4% Zinc Chloride
Dissolve 40g of Zinc chloride (MCG) in 700ml of WFI water. Adjust the pH to 4.6±0.05 with acetic acid. Make up the volume to 1L with WFI and filter the buffer through 0.2 PES micron filter.
8 Crystallization-3:
3M Acetic acid Measure 750mL of water and then add 172.5 ml of Glacial acetic acid (17.4M). Make up the final volume to 1L using WFI water.
9 Crystallization-3:
0.5M NaCl
Dissolve 29.22g of Sodium chloride in 700ml of WFI. Make up the volume to 1L using WFI, filter through 0.2 micron filter.
Table 2: Reagents and its method of preparation

Table 3 elaborates the analytical method(s) used for the altering of crystallization-3 process for crystallizing recombinant human insulin. The table also elaborates the stage of process during which the particular method been used.
Sr. No. Analytical method Stage
1. Insushort DSP III (Downstream Process III)
(RP-HPLC to Drug substance)
2. HMWP DSP III
(Crystallization-3 to Drug substance)
3. RS
(related substance) DSP III
(Drug substance)
5. LOD
(Loss on drying) Drug substance
6. ROI Drug substance
7. Zinc Drug substance
8. Solvent Analysis Drug substance
Table 3: Analytical methods and stage of process where it has been used

Table 4 elaborates the name and model of the equipment used for the altering of crystallization-3 process for crystallizing recombinant human insulin.
Sr. No. Name Model / Make
Preparative HPLC Shimadzu LC-8A & Akta Explorer100
Analytical HPLC Agilent HPLC- 1200/1100 & Shimadzu LC-2010
Weighing Balance Sartorius & Metter Toledo
pH & conductivity meter Eutech
Cold room Blue star
Deep freezer Vest frost
Magnetic stirrer Shalom Instruments
Preparative column Novasep Process LC50.500.VE100
Lyophilizer Tofflon Lyo-0.4(CIP, SIP) & Telstar Lyobeta
Chiller Werner Finely Pvt Ltd & Julabo Chiller
Fume hood Kewaunee Scientific Corporation
Muffle furnace Servewell instruments Pvt Ltd
Burner Guna Enterprises
LOD Servewell instruments Pvt Ltd
Microscope Olympus CX41
Table 4: Equipment details

Table 5 elaborates the preparative and analytical columns used for the altering of crystallization-3 process for crystallizing recombinant human insulin.
Sr. No. Column Name Model / Make
HMWP SEC 10µm,7.8 ×300Å WATERS HMWP
ACE 250*4.6 mm, 5µm, 300Å C18 Advanced chromatography Technologies
Vydac C18, 5µm, 250*4.6mm Grace Vydac
Kromasil 100 Å C8,
13µm,250*4.6mm Kromasil®
Table 5: List of Preparative and Analytical columns

Table 6 elaborates revised crystallization-3 process parameters and ranges followed in present invention related to the altering of crystallization-3 process for crystallizing recombinant human insulin.
Process parameters Ranges
FFC (feed for crystallization) Concentration 5.0±0.2 g/L
IPA content (ppm*) in FFC 21 million
Molarity of NaCl for FFC volume 40 ± 1 mM
4% ZnCl2+0.5M NaCl addition rate =0.006 vvm
Tip speed during 4% ZnCl2+0.5M NaCl addition 0.42 to 0.52 m/s
Expected consumption of 3.0M acetic acid based on FFC volume 0.3 to 0.4%v/v
Volume of 3.0M acetic acid to be added in first pulse 70 to75%
(of the expected consumption)
Rate of pH adjustment (for first pulse) =0.1 vvm
Tip speed during pH adjustment 0.42 to 0.52 m/s
Agitation time post pH adjustment 15 to 20 min
Tip speed post pH of 5.0±0.1 attained
(15-20 min) 0.21 m/s
Incubation time at 24±3°C post pH agitation completion 2.5 to 4.0 hours
Neat settling temperature post 24±3°C incubation 5±3°C
1st neat supernatant sampling 16 hours
Tip speed during Wash mixing 0.21 m/s
Wash settling temperature 5±3°C
Mixing duration during wash (after adding WFI) Not more than 5 minutes
Wash settling temperature 5±3°C
1st wash supernatant sampling 16 hours
*ppm – refers to “µg” of IPA per “g” of Human Insulin in FFC
Table 6: Revised crystallization process parameter and ranges

Parts per million (ppm) calculations used for measuring small concentrations in a solution. In present invention, it was required to prepare an accurate of amount blank buffer (L) for dilution of RP-HPLC3 Elution Pool (EP) (5.0±0.2 g/L) based on the formula elaborated in table 7 below.

Dilute the RP-HPLC3 EP to 5.0 g/L with blank buffer
Feed for crystallization (L)={[Conc.of RPHPLC3 EP (g/L) × Volume of total RPHPLC3EP (L)]/(5.0 (g/L) )}
Blank buffer volume (L) = Volume of Feed for crystallization (L) – Volume of RP-HPLC3 EP (L)
Formula to calculate IPA (ppm) content in RP-HPLC3 EP
X+Y (ppm) = 21000000 (ppm)
Where X is IPA content in ppm (already available in RP-HPLC3 EP), while Y is required amount of IPA (ppm) to be contributed by blank buffer for dilution of RP-HPLC3 EP.
IPA(%)in RPHPLC3 EP=(%B of the first fraction of EP+%B of the last fraction of EP)/2
X (ppm) = IPA content in ppm (already available in RP-HPLC3 EP)
X={( [IPA in RPHPLC3 EP (%)×Volume of total RPHPLC3 EP (L)×Density of IPA (kg/L)×1000×1000])/(Total product in RPHPLC3 (kg)*100) }
Y (ppm) = 21000000–X (ppm)
Formula to calculate blank buffer IPA strength (%) for RP-HPLC3 EP dilution
Blank buffer IPA strength (%)={[Y(ppm)×Total product in RPHPLC3 EP (kg)]/(Blank buffer volume (L)×Density of IPA stock (kg/L)×1000×1000)}×100
*Density of IPA stock= 0.786 Kg/L
Formula to calculate volume of IPA (L) in blank buffer
Volume of IPA (L)={[Blank buffer IPA strength(%)×Blank buffer volume (L)]/100}
Formula to calculate volume of WFI (L) in blank buffer
Volume of WFI in Blank buffer (L) = Volume of Blank Buffer (L) –Volume of IPA (L)
Table 7: Formulae for IPA content calculations

The present invention is further elaborated with the help of following examples. However, these examples should not be construed to limit the scope of present invention.

Example 1: The impact of reagent addition chronology and temperature of neat settling
Haziness was observed in the FFC (feed for crystallization) upon NaCl addition during the satellite trials conducted at the laboratory. NaCl used to hasten the rate of settling during the crystallization process. Thus, based on this attributed purpose of NaCl, experiments were performed to understand whether the change in chronology of NaCl addition would mitigate the haziness observed in the FFC of the batch.
The reagent addition chronology was studied in two strategies viz.
Strategy 1: Addition of 0.5M NaCl post pH adjustment of the crystallization mixture.
Strategy 2: Addition of 4% ZnCl2+0.5M NaCl mixture to the FFC at 5.0g/L.

The Strategy 1 is Addition of 0.5M NaCl post pH adjustment of the crystallization mixture elaborated as follows in table 8 and 9.
Table 8 elaborates the details of the reagent addition performed in strategy 1. In strategy 1, the RP-HPLC 3 elution pool (EP) of 8.70g/L concentration was diluted to 5.0g/L with WFI. This dilution was followed by addition of ZnCl2, which was immediately followed by pH adjustment to 5.0±0.1. NaCl was added to the mixture and it was held at 24±3ºC until neat settling. Table 9 elaborates the details of the observations on reagent addition of strategy 1.

From the trial experiments T5, T6, T20 & T21 (shown in figure 4) it was observed that reverse chronology (i.e. addition of NaCl at the very last) did not result in a consistent settling. However, the necessity of ambient hold (24±3°C) was evident. This observation of ambient hold (subjective to settling observation) was thus maintained constant for all the successive experiments.

Sr. No. Trial No. Conditions
T5 RP-HPLC 3 EP 8.70g/L concentration. diluted to 5.0 g/L with WFI, the dilution performed to attain 5g/L was inclusive of 40mM NaCl.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 (0.5mL/g).
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, 40mM NaCl was added and the mixture was held at ambient (24±3ºC) temperature until settling occurred.
T6 RP-HPLC 3 EP 8.70g/L concentration. diluted to 5.0 g/L with WFI, the dilution performed to attain 5g/L was inclusive of 40mM NaCl.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 (0.5mL/g).
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed it mix for 5 min.
Post mixing, 40mM NaCl was added and the mixture was held at cold (5±3ºC) temperature until settling occurred.
T20 RP-HPLC 3 EP 11.20g/L conc. diluted to 5.0 g/L with WFI, the dilution performed to attain 5g/L was inclusive of 40mM NaCl.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 (0.5mL/g).
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed it mix for 5 min.
Post mixing, 40mM NaCl was added and the mixture was held at cold (5±3ºC) temperature until settling occurred.
T21 RP-HPLC 3 EP 11.20g/L conc. diluted to 5.0 g/L with WFI, the dilution performed to attain 5g/L was inclusive of 40mM NaCl.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 (0.5mL/g).
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, 40mM NaCl was added and the mixture was held at ambient (24±3ºC) temperature until settling occurred.
Table 8: Details of strategy-1 reagent addition

Sr. No. Trial No. Conditions
(Refer Table 8) Observations
Settling (hr) Settling bed appearance
T5 RP-HPLC 3 EP conc -8.70 g/L
Held at ambient (24±3ºC) temperature until settling occurred. 1.0 Settled bed
T6 RP-HPLC 3 EP conc -8.70 g/L
Held at cold (5±3ºC) temperature until settling occurred. Not Applicable Not settled
T20 RP-HPLC 3 EP conc -11.20 g/L
Held at cold (5±3ºC) temperature until settling occurred. Not Applicable Not settled
T21 RP-HPLC 3 EP conc -11.20 g/L
Held at ambient (24±3ºC) temperature until settling occurred. 1.0 Settled bed
Table 9: observations on reagent addition strategy 1

The Strategy 2 is Addition of 4%ZnCl2+0.5M NaCl mixture to the FFC at 5.0g/L elaborated as follows in table 10 and 11.

Strategy 2 was explored due to the inconsistent results of the strategy 1. Table 10 elaborates the details of the reagent addition performed for strategy 2. In this strategy, experiments were performed by mixing the required amount of 4% ZnCl2 and 0.5M NaCl mixture and in turn added to the prepared FFC at 5.0g/L. Following which crystallization-3 was performed by adjusting the pH with 3.0M Acetic acid to 5.0±0.1. Table 11 elaborates the details of the observations on reagent addition of strategy 1.
From the trial strategy 2 experiments (shown in figure 5A and 5B), it was observed that the % of IPA in crystallization mixture remains same, but the IPA content (ppm) was varied. This resulted in a varied ambient hold temperature duration required to achieve similar settling across all the experiments. To normalize the IPA content in FFC, next set of trials were performed.
Sr. No. Trial No. Conditions
T18 RP-HPLC 3 EP 11.38g/L conc. diluted to 5.0 g/L with RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid).
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
T19 RP-HPLC 3 EP 7.35g/L conc. diluted to 5.0 g/L with WFI.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
T20 RP-HPLC 3 EP 11.38g/L conc. diluted to 5.0 g/L with WFI,
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
T23 RP-HPLC 3 EP 11.38g/L conc. diluted to 8.0g/L with RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0 g/L with WFI.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred
T25 RP-HPLC 3 EP 11.38g/L conc. diluted to 9.0g/L with RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0 g/L with WFI.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
T26 RP-HPLC 3 EP 11.38g/L conc. diluted to 10.0g/L with RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0 g/L with WFI.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
Table 10: Details of strategy-2 reagent addition

Sr. No. Trial No. Conditions
(Refer Table 10) Observations
Settling (h) Crystal size (µm) IPA content (ppm)
T18 RP-HPLC 3 EP conc -11.38 g/L
EP diluted to 5.0g/Lwith RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid). 1.0 > 20 26409600
T19 RP-HPLC 3 EP conc -7.35 g/L
EP diluted to 5.0g/Lwith WFI. 1.5 > 15+ 17965714
T20 RP-HPLC 3 EP conc -11.38 g/L
EP diluted to 5.0g/Lwith WFI. 9.0 > 50 11603515
T23 RP-HPLC 3 EP conc -11.38 g/L
EP diluted to 8.0g/Lwith RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0g/L with WFI. 2.0 > 20 16506000
T25 RP-HPLC 3 EP conc -11.38 g/L.
EP diluted to 9.0g/Lwith RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0g/L with WFI. 3.0 > 20 14672000
T26 RP-HPLC 3 EP conc -11.38 g/L EP diluted to 10.0g/Lwith RP-HPLC 3 Blank Buffer (16% IPA+ 84% 50mM Acetic acid) and further diluted to 5.0g/L with WFI. 2.5 Precipitate 13204800
Table 11: Observations on reagent addition strategy 2

Example 2: Impact and the role of IPA content in FFC
Based on the observations in reagent addition chronology experiments and the observed variation in IPA content (ppm) in FFC; trials were performed by normalizing the IPA content (ppm) in FFC to hasten the settling rate during neat settling.
Table 12 elaborates the RP-HPLC3 EP dilution by targeting IPA content (ppm) in FFC whereas table 13 and figure 6 elaborate the observations of the impact of IPA content in FFC.
The IPA content (ppm) in FFC is vital for achieving desired settling and crystal size. However, the repetition of IPA content target experiments did not result in the similar crystal size. This observation led to study the rate of pH adjustment as one of the key factor that might influence crystal size.

Sr. No Trial No. Conditions
1 T53 RP-HPLC 3 EP 6.58g/L conc. diluted to 5.0 g/L with Blank buffer (2% IPA+ 98% WFI) targeting 20 million IPA content (ppm) in FFC,
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
2 002(16)A RP-HPLC 3 EP 13.13g/L conc. diluted to 5.0 g/L with WFI, (IPA content is not targeted in FFC).
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
3 002(16)B Small portion of 002(16) A crystallization mixture was separated and stimulated with 100% IPA targeting 21 million.
Post IPA stimulation, the mixture was held at ambient temperature for settling
4 002(02) RP-HPLC 3 EP 7.27g/L conc. diluted to 5.0 g/L with Blank buffer (6.5% IPA+ 93.5% WFI) targeting 21 million IPA content (ppm) in FFC,
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1, post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
Table 12: RP-HPLC3 EP dilution by targeting IPA content (ppm) in FFC

Sl.no Trial No Conditions (Refer Table 12) Observations
Settling (h) Settling bed appearance Crystal size (µm)
T53 RP-HPLC 3 EP 6.58g/L conc. diluted to 5.0 g/L with Blank buffer (2% IPA+ 98% WFI) targeting 20 million IPA content (ppm) in FFC. 1 Settled bed >20
002(16)A RP-HPLC 3 EP 13.13g/L conc. diluted to 5.0 g/L with WFI, (IPA content is not targeted in FFC). 3 Not settled Nucleation started
002(16)B Small portion of 002(16) A crystallization mixture was aliquoted and stimulated with 100% IPA targeting 21 million. 3 Settled bed >20
002(02) RP-HPLC 3 EP 7.27g/L conc. diluted to 5.0 g/L with Blank buffer (6.5% IPA+ 93.5% WFI) targeting 21 million IPA content (ppm) in FFC. 3 Settled bed 5
Table 13: Observation of the impact of IPA content in FFC

Example 3: The impact of rate of pH adjustment
Based on the varying crystal sizes (µm) observed despite appropriate addition of reagents (NaCl+ZnCl2) (discussed in example 1) and targeting IPA (ppm) in FFC (discussed in example 2), experiments elaborated in table 14 were performed to understand the impact of rate of pH adjustment.
The IPA content targeted to 21 million in FFC. The addition of 3M acetic acid for attaining the crystallization pH to 5.0±0.1 has to be 0.1 vvm irrespective of the scale of crystallization (refer Table 15).
Sl.no Trial No Conditions
002(09) RP-HPLC 3 EP 7.34g/L conc. at pH 7.30 diluted to 5.0 g/L with Blank buffer (6.2% IPA+ 83.8%WFI) targeting 21 million IPA content (ppm) in FFC.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture at 0.015 vvm of FFC.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 at 0.001 vvm and post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
002(10) RP-HPLC 3 EP at pH 7.30 was further adjusted to pH 7.50 with 2M TRIS.
RP-HPLC 3 EP 7.34g/L conc. at pH 7.50 diluted to 5.0 g/L with Blank buffer (6.2% IPA+ 83.8%WFI) targeting 21 million IPA content (ppm) in FFC.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture at 0.015 vvm of FFC.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 at 0.001 vvm and post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
002(11) RP-HPLC 3 EP at pH 7.30 was further adjusted to pH 7.50 with 2M TRIS.
RP-HPLC 3 EP 7.34g/L conc. at pH 7.50 diluted to 5.0 g/L with Blank buffer (6.2% IPA+ 83.8%WFI) targeting IPA content 21 million in FFC,
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture at 0.015 vvm of FFC.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 by bolus addition and post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
002(13) RP-HPLC 3 EP 7.34g/L conc. at pH 7.30 diluted to 5.0 g/L with Blank buffer (6.2% IPA+83.8%WFI) targeting 21 million IPA content (ppm) in FFC.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture by bolus addition.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 by bolus addition and post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
002(14) RP-HPLC 3 EP at pH 7.30 was further adjusted to pH 7.50 with 2M TRIS.
RP-HPLC 3 EP 7.34g/L conc. at pH 7.50 diluted to 5.0 g/L with Blank buffer (6.2% IPA+ 83.8%WFI) targeting 21 million IPA content (ppm) in FFC.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture by bolus addition and post attaining the pH, the solution was allowed to mix for 5 min.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 by bolus addition and post attaining the pH, the solution was allowed to mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
002(15) RP-HPLC 3 EP at pH 7.30 was further adjusted to pH 7.50 with 2M TRIS.
RP-HPLC 3 EP 7.34g/L conc. at pH 7.50 diluted to 5.0 g/L with Blank buffer (6.2% IPA+ 83.8%WFI) targeting 21 million IPA content (ppm) in FFC.
Post dilution, crystallization-3 was performed by addition of required amount of 4%ZnCl2 + 0.5M NaCl (0.5mL/g+40mM) mixture at 0.015 vvm of FFC.
pH of the resultant solution was adjusted with 3.0M acetic acid to 5.0±0.1 by bolus addition and post attaining the pH, the solution was allowed it mix for 5 min.
Post mixing, the mixture was held at ambient (24±3ºC) temperature until settling occurred.
Table 14: Details for rate of pH adjustment

Sr. no Trial No Conditions
(Refer Table 14) Reagent & Rate of Addition Observations
ZnCl2+NaCl mixture 3M Acetic acid for pH adjustment Settling (hr) Crystal size (µm)
002(09) RP-HPLC3 EP pH 7.30 Slow (0.015 vvm) Slow (0.001 vvm) 1 Micro
crystal
002(10) RP-HPLC3 EP pH 7.50 Slow (0.015 vvm) Slow (0.001 vvm) 1 10
002(11) RP-HPLC3 EP pH 7.50 Slow (0.015 vvm) Bolus 1 >20
002(13) RP-HPLC3 EP pH 7.30 Bolus Bolus 3 >20
002(14) RP-HPLC3 EP pH 7.50 Bolus Bolus 1 >20
002(15) RP-HPLC3 EP pH 7.50 Slow (0.015 vvm) Bolus 1 > 20
Table 15: Matrix of pH adjustment data and observation

Example 4: The impact of RP-HPLC3 product binding capacity on crystallization
To challenge the RP-HPLC3 product binding capacity as well as to understand the impact of RP-HPLC 3 product dynamic binding capacity (DBC) at final crystallization satge, RP-HPLC3 trials were performed at two different DBCs, one at 23g/L and another one at 50g/L to accommodate and propose the wider range of DBC at manufacturing scale unlike the current control limit for RP-HPLC 3 DBC (25 to 40.0g/L). Figure 8 shows the RP-HPLC 3 DBC study flow chart.

RP-HPLC 2 load was procured from the manufacturing facility and RP-HPLC 2 and RP-HPLC 3 steps were performed at pilot lab as per insulin biosimilar process. All the individual fractions (after RPHPLC3) were adjusted to 7.35±0.1 after the elution. After pooling the fractions, pH of RPHPLC3 bulk EP was checked and adjusted to 7.4±0.1. Individual fractions pH essentially were not less than 7.3, as it may trigger protein precipitation.

Table 16 elaborate the RP-HPLC3 Load purity profile wherein table 17 elaborate RP-HPLC3 process performance and quality attributes at different DBC.
Quality attributes of RP-HPLC3 EP, which was generated at different DBC were comparable with control specifications. The same RP-HPLC 3 EP was utilized for crystallization-3 experiments, found no impact on crystallization process.
RP-HPLC3 EP was used for scale up trial.

Sample Conc.
(g/L) Purity
(%) 0.85-0.86 RRT 0.92-0.93 RRT 0.95-0.96 RRT 1.03-1.04 RRT
Load spec. Not more than
5.0 g/L Not more than 98.6% Not more than
0.18% Not more than
0.10% Not more than
0.35% Not more than
0.25%
Load 4.20 98.28 0.17 0.07 1.06* 0.13
*0.95+0.96 RRT was out of specification for RP-HPLC 3 Load.
Table 16: RP-HPLC3 Load purity profile

Sample Loading (g/L) Purity
(%) 0.85-0.86 RRT 0.95-0.96 RRT 1.03-1.04 RRT Conc.
(g/L) Yield (%) EP in CVs %B start %B End
EP spec. NLT 99.50% BLOQ (0.05%) Not more than 0.10% Not more than 0.10%
001(04) EP 50.05 99.75 0.00 0.10* 0.07 12.87 81.62 3.17 15.5 17.4
002(01) EP 23.35 99.85 0.00 0.00 0.09 7.27 80.01 2.57 15.7 17.2
002(03) EP 23.35 99.82 0.00 0.05 0.09 7.49 76.79 2.39 15.4 16.8
002(12) EP 50.42 99.74 0.00 0.10* 0.07 13.15 84.26 3.23 15.8 17.4
* Higher level of 0.95/0.96 RRT was observed due to the higher level of the same at RP-HPLC3 Load stage, though the values are within the specification.
Table 17: RP HPLC3 process performance and quality attributes at different DBC

Example 5: Scale-up trial
From the various earlier trials (referred in example 1-4), it was observed that the reagent addition chronology, IPA content (ppm), pH adjustment rate and ambient hold temperature are vital for crystallization-3 process.

Based on the understanding, gained from various experiments, scale up trials were designed while keeping following process parameters under check (Table 18). Table 19 elaborates the scale up performance with respect to process & quality wherein figure 9 shows the images of crystals obtained upon scale-up.

Process parameters Trial 1 Trial2 Trial
RP-HPLC 3 DBC (g/L) 50 g/L 25 g/L 50 g/L
FFC conc. (g/L) 5.0 ±0.2
Target IPA content (ppm) at FFC stage 21
Addition of 0.5M NaCl +% ZnCl2 mixture 40mM NaCl + 0.5mL/g ZnCl2 @ =0.006 vvm
Mode of pH adjustment (5.0±0.1) Bolus (=0.1vvm)
Agitation time after pH adjustment (min) 15 to 20
Ambient temperature hold during neat settling (h) 3.0 to 4.0
% neat decantation 85-90%
Fold wash after neat decantation 8 fold 10 fold
% Wash decantation = 95-97%
Table 18: Process parameters that were under check during scale-up of crystallization-3

Stage Attributes T003 T004 T006
FFC Conc. (g/L) 4.96 5.01 5.14
FFC (before pH adjustment) pH 7.36 7.60 7.35
Conductivity (mS/cm) 0.39 0.80 0.34
FFC
(After pH adjustment) pH 4.83 5.01 5.02
Conductivity (mS/cm) 3.39 3.67 3.26
Neat Settling Ambient hold (h) 3.0 3.0 2
Temperature 5±3 ºC 5±3 ºC 5±3 ºC
Time 3.5 h 4.0 h 15.5 h
% Decantation 87.70 89.02 85.02
Neat Decanted supernatant Sup conc. (mg/mL) 0.23 0.07 0.10
pH 4.86 5.06 5.03
Conductivity (mS/cm) 3.45 3.56 2.99
Wash settling temperature Temp. 5±3 ºC 5±3 ºC 5±3 ºC
Wash Decanted supernatant pH 4.83 5.02 5.01
Conductivity (mS/cm) 0.58 0.50 0.39
% Decantation 94.86 95.51 96.97
Sup conc. (mg/mL) 0.05 0.03 0.10
Neat Slurry Insushort purity (%) 99.77 99.80 99.74
HMWP (%) 0.072 0.048 0.040
FFL
Feed for Lyophilisation Insushort purity (%) Not Available 99.76 Not Available
HMWP (%) Not Available Not Available 0.045
Table 19: Scale up performance – process & quality

Example 6: The Sub-optimal freeze-drying during Insulin biosimilar batch
Due to the poor settling at neat settling stage during the Crystallization 3 stage of Insulin biosimilar process, lower decantation was performed at each stage (neat and wash). As per the general trend (data obtained upon following old process mentioned in figure 2), volume for freeze-drying is observed to be in the range of 30±2 Litres but it was found to be 34.46 Litres in this batch. As per the BMR limit, maximum volume, which could be loaded on each tray, should be 1.28 Litre (which corresponds to the 1.1 cm bed height) but due to the higher volume after wash decantation, average volumetric distribution of slurry in each tray was 1.436 Litre, which corresponds to the 1.23 cm bed height). In addition, the observed slurry percentage, which was loaded in each tray, was approximately 5%.

To understand the impact of Slurry percentage and bed height on Freeze-drying efficiency, the following trial elaborated in table 20 was designed. Manufacturing slurry (re-dissolved slurry) was procured to understand the impact of bed height & slurry percentage on freeze drying efficiency (Table 21).

From the results, it was quite evident that, at lower slurry percentage (4-5%), increase in bed height can impact on the moisture content (LOD value). Higher the bed height (with lower slurry percentage) in freeze drying tray would result in inefficient drying and sublimation.

Slurry %
Bed height (cm) Total challenge to the condenser
8% 1.2 cm 20% of total capacity
1.4 cm
4% 1.2 cm
1.4 cm
1.6 cm
Table 20: Freeze-drying trial

Slurry % Bed height (cm) LOD value
8% 1.2 cm 4.45%
1.4 cm Didn’t dry
4% 1.2 cm 8.04%
1.4 cm Did not dry
1.6 cm Did not dry
Table 21: Freeze drying process outcome

Summary of Final Drug Substance Quality Attributes (3 scale up experiments)
For a particular crystallization condition (as captured in the ‘condition’ section of each table), observations are captured pertaining to each unique experiment. Table 22 elaborates the summary of final drug substance from revised crystallization process whereas table 23 elaborates the list of critical process parameter and its impact observed during the trails conducted (referred in example 1-6).

Stage Parameter T003 Trial 1 T004
Trial 2 T006
Trial 3 USP monograph Specification
DS ORP% (RS) 0.08 0.07 0.11 Not more than
2.0 %
A21% (RS) 0.04 0.04 0.039 Not more than
2.0 %
HMWP % value 0.13 0.12 0.1 Not more than
1.0 %
Insushort purity (%) 99.54 99.71 99.52 Not Applicable
IPA (ppm) 791.0 1037.53 1432 Not more than
5000 ppm
LOD (%) 1.99 2.68 2.56 Not more than
10.0%
ROI (%) 1.05 1.11 Not Available Not Applicable
Zn (%) 0.38 0.40 0.39 Not more than
1.00 %
Table 22: Summary of Final drug substance from revised crystallization process

Stage Parameter Targeted range Impact of the parameter
Crystallization IPA content (ppm) ppm at FFC stage 21 million Lower IPA would result in inefficient settling of crystals while higher IPA can pose the challenges for its removal in DS
pH adjustment rate first pulse (target pH 5.0±0.1) 0.1vvm vvm directly
(positive correlation) impacts crystal size
Mixing rate & time after final pH adjustment pH 5.0±0.1 0.21m/s mixing rate & 20 min Mixing time Mixing rate and time are crucial to aid crystallization
Ambient Temperature hold duration during neat settling 2.5-4 hours Lower ambient hold would result in poor settling while higher hold duration would lead to increase in aggregates (HMWP)
Freeze drying Bed height Not more than
1.1 cm Higher bed height could result in inefficient sublimation and there by drying.
Table 23: List of critical process parameter and its impact

Conclusion
Based on the above data, it’s evident that performing crystallization-3 process with FFC targeting 21 million IPA content (ppm), followed by addition of ZnCl2+NaCl mixture, & comparatively faster rate of pH adjustment and ambient temperature hold are vital for protein crystallization, settling and consistent crystal size. Performing remaining neat settling and complete wash, both at cold temperature (5±3°C) can certainly add up to the process robustness and better control of critical quality attributes at final drug substance stage.

The process thus reduced time for sedimentation at manufacturing scale. In traditional process, the sedimentation used to take up to 24-60 hours whereas the present invention achieves the same result in 12 hours. The results are shown in Figure 10 which contains three comparative images of human insulin crystals. 10A represent the recombinant Human Insulin crystals prepared by traditional process wherein only zinc chloride is used while 10B represents the recombinant Human Insulin crystals prepared by process of present invention wherein mixture of zinc chloride and sodium chloride is used. The consistent crystal size between 15µm -30µm is noticeable in 10B. 10C represents the recombinant Human Insulin drug substance crystals.