RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Freeze Drying of L-Arginine/Sucrose-Based Protein Formulations, Part I: Influence of Formulation and Arginine Counter Ion on the Critical Formulation Temperature, Product Performance and Protein Stability 1 ¨ PETER STARTZEL, HENNING GIESELER,1 MARGIT GIESELER,1 AHMAD M. ABDUL-FATTAH,2 MICHAEL ADLER,2 HANNS-CHRISTIAN MAHLER,2 PIERRE GOLDBACH2 1

GILYOS GmbH, Friedrich-Bergius Ring 15, Wuerzburg 97076, Germany F. Hoffmann-La Roche Ltd., Pharmaceutical Development & Supplies, Pharma Technical Development Biologics EU, Basel 4070, Switzerland 2

Received 22 January 2015; revised 30 March 2015; accepted 27 April 2015 Published online 20 May 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jps.24501 ABSTRACT: The objective of this study was to investigate product performance of freeze dried L-arginine/sucrose-based formulations under variation of excipient weight ratios, L-arginine counter ions and formulation pH as a matrix to stabilize a therapeutic monoclonal antibody (MAb) during freeze drying and shelf life. Protein and placebo formulations were lyophilized at aggressive primary drying conditions and key attributes of the freeze dried solids were correlated to their thermal properties and critical formulation temperature. Stability (physical) during processing and long-term storage of the MAb in different formulations was assessed by SE-HPLC. Thermal properties of the mixtures were greatly affected by the type of L-arginine counter ion. High glass transition temperatures were achieved by adding multivalent acids, whereas the temperature values significantly decreased in the presence of chloride ions. All mixtures were stable during freeze drying, but storage stability varied for the different preparations and counter ions. For L-arginine-based formulations, the protein was most stable in the presence of chloride ion, showing no obvious correlation to estimated global mobility of the glass. Besides drying behavior and thermal C 2015 Wiley properties of the freeze dried solids, the counter ion of L-arginine must be considered relevant for protein shelf life stability.  Periodicals, Inc. and the American Pharmacists Association J Pharm Sci 104:2345–2358, 2015 Keywords: protein formulation; freeze drying/lyophilization; amorphous solids; glass transition; solid-state stability; protein aggregation; calorimetry (DSC); L-arginine

INTRODUCTION Freeze drying is commonly used in the pharmaceutical industry to improve protein stability for molecules that are not stable in aqueous solution.1–4 However, some proteins are inactivated during the drying process if not appropriately formulated. Therefore, excipients, for example stabilizers, buffers, and surfactants are added to stabilize the protein during the various stresses exerted by freezing and drying as well as long-term storage.2,5,6 In the solid state, proteins are protected thermodynamically by replacement of water molecules2,7,8 and kinetically by suppression of global motion (also referred to as "-relaxation) within the glass.9,10 "-Relaxation mainly occurs because of the translational and rotational motion of the molecules and therefore strongly affects the diffusion of reactive species, potentially leading to protein aggregation.11 Both concepts have been used to explain protein stability in the lyophilized state,12,13 although in some cases neither concept was found conclusive to the data.14,15 Some of this divergence was attributed to the presence of motions which occur in a shorter time scale and are therefore referred to as fast dynamics or $-relaxations.11,16,17 Global motions are typically characterized by “structural relaxation time” (denoted as J)18–20 as both Correspondence to: Pierre Goldbach (Telephone: +41-61-6885242; Fax: +4161-68-88689; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 104, 2345–2358 (2015)  C 2015 Wiley Periodicals, Inc. and the American Pharmacists Association

degradation (chemical and physical) and structural relaxation require a certain degree of molecular mobility within the glass. Thus, it is suggested that instability-causing processes, such as protein aggregation, are coupled to the structural relaxation time.18 Results are typically compared under application of the parameter $, a stretched exponential constant (J$ ).18 Similar to sugars, various amino acids and their salts (e.g., L-arginine hydrochloride)21,22 also showed lyo- and cryoprotective effects.23 However, effective protection can only be achieved if the amino acid remains in the amorphous state. Addition of various acids was reported to affect the crystallinity and the glass transition temperature of vacuum dried and freeze dried L-arginine.24 Izutsu et al.25 found that mixing L-arginine and a hydroxyl di- or tricarboxylic acid led to amorphous solids with a glass transition temperature of the maximally freeze concentrated solute (Tg  ) significantly higher than those of the individual solute solutions (i.e., the pure acid and L-arginine base). Especially multivalent acids such as citric and L-tartaric acid raised the Tg  as well as the Tg of the freeze dried solids. This was attributed to the formation of an intense interactive network between the involved molecules.25 Increase of the relevant glass transition temperatures (Tg  and Tg ) may allow application of a higher product temperature during primary drying, and decrease molecular mobility in the freeze dried solids which can provide enhanced formulation robustness and may also improve protein stability in the freeze dried state.26 However, the assessment of greater protein stability at higher glass

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

2345

2346

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

transition temperatures of the freeze dried excipient matrix is based on the assumption that stability in the solid state is governed by global mobility.26 However, so far, little information is available on stability of freeze dried therapeutic proteins as a function of the varying physical properties obtained for different types of L-arginine counter ions.25,26 Moreover, Larginine and acid combinations have been used to assist recovery of chemically unfolded proteins or recombinant proteins expressed in inclusion bodies,27–29 due to the ability of L-arginine to suppress aggregation of folding intermediates in the liquid state.30 This unique feature may render L-arginine a versatile excipient for protein freeze drying, as an L-arginine-containing formulation may also stabilize the protein in solution prior to freeze drying and after reconstitution.25,31 In the present study, L-arginine was included into sucrose-based formulations at varying concentrations. Sucrose was used as a second excipient in the light of its frequent application as a lyoprotectant in pharmaceutical freeze drying.2,32 Although analysis of L-arginine and acid combinations showed that the L-arginine counter ion influences the physical properties of the amorphous amino acid, no data have been published on such combinations included into a multiexcipients formulation. However, the presence of additional excipients may significantly alter interactions between molecules which may lead to properties of L-arginine different from those as a single component. Therefore, investigating L-arginine as a part of a typical pharmaceutical, multi-component formulation may be considered essential as a basis for assessment of its stabilizing effect on proteins. The main purpose of this study was to investigate how protein stability in various L-arginine/sucrosebased formulations (including buffer and surfactant) is affected by the type of counter ion during freeze drying and during storage. Long-term stability with regard to protein aggregation was studied for a recombinant humanized monoclonal IgG1 antibody over a maximum period of 180 days at different storage conditions ranging from 5°C to 40°C. Results obtained through this study also provide insight into the effect of the counter ion on drying performance, product appearance and characteristic product attributes (i.e., residual moisture, reconstitution time (RT), and physicochemical properties) for active and placebo formulations.

MATERIALS AND METHODS Materials All chemicals were of analytical grade and used as supplied. Sucrose from Ferro Pfanstiehl Inc. (Waukegan, Illinois) was used during all experiments. L-Arginine base, L-histidine base, and citric acid monohydrate were purchased from Ajinomoto Foods Europe SAS (Hamburg, Germany); polysorbate 20 was obtained from Croda GmbH (Nettetal, Germany). Hydrochloric acid 25% (v/v) and succinic acid (cryst.) were acquired from Merck KGaA (Darmstadt, Germany). Phosphoric acid 85% was purchased from Carl-Roth (Karlsruhe, Germany). Water for injection (WFI) was used for preparation of all formulations. Solutions were sterile filtered through a 0.22 :m membrane filter before use (placebo formulations: PES membrane, MILLEX-GP50; Millipore Corporation, Bedford, Massachusetts; active formulations: PVDF membrane, SteriCupTM ; Millipore Corporation, Bedford, Massachusetts). Twenty milliliters clear glass tubing vials (Fiolax, 20 mm) were St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

¨ purchased from SCHOTT (Mullheim, Germany) and 20 mm D777–1 Lyo Stoppers were acquired from Daikyo Seito Ltd. (Tokyo, Japan). Purified monoclonal antibody (MAb) was supplied as 76.4 mg/mL stock solution in 20 mM histidine buffer (histidine/histidine-hydrochloride) pH 6.0 by F. Hoffmann-La Roche (Basel, Switzerland). The MAb used in this study was a recombinant humanized monoclonal IgG1 antibody with an approximate molecular weight of 149 kDa and an isoelectric point (pI) of 7.8. Methods Preparation of Product Solutions Solutions containing 104 mg/mL (304 mM) L-arginine base were titrated to a pH of 6.0 ± 0.3 with citric, phosphoric, succinic, or hydrochloric acid to investigate the physicochemical characteristics of L-arginine as a single solute in presence of different counter ions. All pH measurements were executed using a calibrated pH-meter (SCHOTT Instruments Lab 870 with Mettler-Toledo InLab Micro Pro pH electrode). For placebo formulations, the calculated mass of excipients was dissolved in WFI as tabulated in Table 1 (F01–F08). The mixing ratio of L-arginine to sucrose was 4:1 and 1:4 (by weight). Both Larginine and L-histidine (buffer) were added in form of the base, and solutions were then titrated to a pH of 6.0 ± 0.3 using the aforementioned acids (cf. Table 1). For investigating the effect of pH (discussed below), several mixtures were adjusted to pH setpoints other than 6.0 (i.e., 5.0 and 7.0) while maintaining accuracy of pH adjustment and the compounding procedure. Polysorbate 20 was added using a 4% (w/v) stock solution (final concentration 0.02%, w/v), and the total solid content of 130 mg/mL in solution (excluding buffer and surfactant) was adjusted by diluting with WFI. Histidine buffer concentration was 20 mM after dilution. Formulations with a higher Table 1. Overview of Formulation Compositions Compositiona Name of Formulation

L-Arginine Base (mg/mL)

Sucrose (mg/mL)

MAb (mg/mL)

Acidb

F01 F02 F03 F04 F05 F06 F07 F08

26 104 26 104 26 104 26 104

104 26 104 26 104 26 104 26

– – – – – – – –

Hydrochloric Hydrochloric Citric Citric Phosphoric Phosphoric Succinic Succinic

F09 F10 F11 F12 F13 F14 F15 F16 F17

16 64 16 64 16 64 16 64 –

64 16 64 16 64 16 64 16 80

50 50 50 50 50 50 50 50 50

Hydrochloric Hydrochloric Citric Citric Phosphoric Phosphoric Succinic Succinic –

a All formulations contained 20 mM L-histidine as a buffer which was added in form of the base (for F17: histidine was added as histidine/histidine–HCl, 20 mM to obtain a pH of 6.0) and 0.02% (w/v) polysorbate 20 as a surfactant. pH of the formulations was adjusted to 6.0 ± 0.3 using the respective acid. b Acid used for pH adjustment.

DOI 10.1002/jps.24501

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

concentration of L-arginine (304 mM) versus sucrose (47 mM), that is, with a mixing ration of 4:1 (by weight) are labeled “L-arginine-rich” mixtures, while formulations with a higher sucrose concentration (187 mM versus 76 mM L-arginine) are referred to as “sucrose-rich” mixtures. For preparing of protein formulations, the protein concentration was adjusted to 50 mg/mL by dilution of the protein stock with a concentrated excipient solution which had been titrated with the corresponding acid to obtain a pH of 6.0 ± 0.3. The total solid concentration of 130 mg/mL in the final solution (excluding buffer and surfactant) for the protein-containing mixtures was maintained by decreasing the excipient concentrations (i.e., sucrose and L-arginine) to 80 mg/mL. The excipient ratios of L-arginine to sucrose were maintained identical to the placebo formulations (cf. Table 1, F09–F17). This consequently led to a molar sucrose to protein ratio of 139:1 in L-arginine-rich systems and 557:1 in sucrose-rich systems. Freeze Drying Microscopy The temperature at the onset of collapse (Toc ) and the full collapse temperature (Tfc ) of product solutions were determined with freeze drying microscopy (FDM), using a Linkam FDCS196 freeze drying stage (Linkam Scientific Instruments, Surrey, UK) and Zeiss Axio Imager.A1 microscope (Carl Zeiss MicroImaging, G¨ottingen, Germany). The applied magnification during the experiments was 200-fold, and a sample volume of approximately 1.5 :L was analyzed. Pictures were captured by the software (Linksys32; Linkam Scientific Instruments, Surrey, UK) in 5-s intervals using a digital camera. All samples were cooled at a constant rate of 1°C/min to −25°C, followed by a fast freezing rate (10°C/min) to −55°C and a 10 min isothermal equilibration. Vacuum (100 mTorr) was applied after 8 min. The heating rate during the heating phase under vacuum was 1°C/min for all analyzed mixtures. SDs of three replicate measurements for onset of collapse (Toc )33 and full collapse temperature (Tfc )33 were not more than 0.2°C. Freeze Drying All samples (placebo and active) were freeze dried using a LyoStar 3TM freeze dryer (SP Scientific, Stone Ridge, New York). Vials were semi-stoppered after filling with a fill volume of 10.5 mL. A total of 36 vials per formulation for placebo mixtures and 28 vials for protein mixtures were loaded in a close hexagonal packing profile. One row of empty dummy vials surrounded the product vials to reduce radiation effects. Product temperature was monitored for one center-positioned vial per formulation via a thin wire thermocouple (36 gauge; OMEGA Engineering, Newport, Connecticut). After completion of the freeze drying cycle, vials were stoppered at 75 mTorr prior to backfill the chamber with dry nitrogen gas. An overview of the freeze drying conditions applied is provided in Table 2. Modulated Differential Scanning Calorimetry Glass transition temperatures of the frozen solutions (Tg  ) and of the freeze dried solids (Tg ) were analyzed using a modulated differential scanning calorimetry (MDSC) Q2000 (TA Instruments Inc., New Castle, Delaware), operating with TA Instruments Universal Analysis software (version 4.5A) and equipped with an autosampler and a refrigerated cooling system (RCS90). For analysis of the liquid samples during this study, 10–20 mg of solution was transferred into a Tzero aluDOI 10.1002/jps.24501

2347

Table 2. Overview of Freeze Drying Conditions Shelf Ramp Temperature Rate (°C) (°C/min) Loading Freezing Primary drying Secondary drying

5.0 −40.0 5.0 25.0

– 1.0 0.5 0.2

Hold Time (min)

Chamber Pressure (mTorr)

60 180 Cycle dependenta 360

– – 75 75

a Primary drying terminated upon reaching a Pirani/Capacitance Manometer differential of ࣘ7 mTorr.

minum hermetic pan. MDSC was performed by freezing the sample (1°C/min) to −60°C followed by reheating to −15°C (1°C/min) and a 60 min isothermal annealing phase. The annealing temperature was selected to be above the Tg  for all formulations (highest Tg  was approximately −26°C), thereby facilitating potential excipient crystallization. After annealing, the samples were cooled back to −80°C and then heated to +5°C at 3°C/min with a superimposed temperature modulation of ±0.636°C every 40 s. For solid samples, powdered product was taken from the center of the cake (2–8 mg) and transferred to the pan and hermetically sealed. MDSC was performed by heating the samples between −5°C and 160°C at 2°C/min using a superimposed temperature modulation of ±0.954°C every 60 s. Glass transition temperatures (Tg and Tg  ) are reported as “midpoint” temperature (1/2 Cp ) based on the reversing heat flow signal. SDs of three replicate glass transition temperature measurements were not more than 0.2°C for liquid and 0.3°C for solid samples. A rough estimation of Tg  of the sucrose/L-arginine systems was obtained by using the Gordon–Taylor equation which assumes ideal miscibility of the two components and additivity of the free volumes.34 Here, the glass transition temperature of a mixture, Tg12 is defined as Tg12 =

w1 Tg1 + Kw2 · Tg2 w1 + Kw2

(1)

where w1 and w2 are the mass fractions of each component and Tg1 and Tg2 are respective glass transition temperatures. The constant K is related to the ratio of the free volumes of the two components and can be calculated using the Simha–Boyer rule.35 Specific K values applied during calculation are given in Table 3. Additional information regarding the use of the Gordon–Taylor equation and the Simha–Boyer rule is described in more detail elsewhere.36,37 X-Ray Powder Diffraction Measurements The physicochemical state of the freeze dried solids was analyzed using X-ray powder diffraction analysis (XRPD). Measurements were executed at a controlled temperature of 25°C using a Philips X’pert MPD (PANalytical, Eindhoven, Netherlands) with Cu K" radiation at 40 kV/40 mA. The freeze dried solids were carefully ground and filled into an aluminum sample holder. The powder was then gently compacted using a cover glass. The sample chamber was purged with dry nitrogen during the measurement, and scans were recorded in the range 22 = 0.5°–40°, with a step size of 0.02°. Resulting diffractograms were analyzed using Microcal Origin. St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

2348

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Table 3. Thermal Characteristics of the Frozen L-Arginine/Sucrose-Based Placebo Formulations Investigated by Freeze Dry Microscopy (Toc , Tfc ) and Modulated DSC (Tg  ) Name of Formulation F01 F02 F03 F04 F05 F06 F07 F08

Toc a (°C)

Tfc a (°C)

Tg  , pH 5.0b (°C)

Tg  , pH 6.0 (°C)

Tg  , pH 7.0 (°C)

Tg  , calc.c (°C)

Td (°C)

−33.4 −40.5 −27.9 −26.3 −29.7 −25.9 −30.4 −32.7

−31.2 −38.0 −26.8 −24.0 −28.3 −24.6 −28.8 −30.7

−32.9 −37.6 −28.4 −25.9 −29.6 −28.8 −30.5 −31.6

−31.6 −38.1 −29.1 −26.1 −28.9 −27.4 −29.6 −32.0

−33.0 −38.0 −27.5 −26.4 −27.6 −26.3 −29.8 −32.7

−33.8 −39.0 −30.9 −26.6 −31.0 −28.0 −32.4 −32.9

−2.3 −1.0 −1.9 −0.5 −2.1 −0.6 −2.8 −0.9

Onset of collapse temperature (Toc ) and full collapse temperature (Tfc ) measured by FDM (n = 1). Glass transition temperature (midpoint) of the maximally freeze concentrated solute, reported from the reversing heat flow signal of the modulated temperature scan (n = 1). pH indicates the formulation pH which was adjusted by adding different amounts of acid. Accuracy of pH adjustment was ±0.3. SD of three replicate glass transition temperature measurements was not more than 0.2°C. c Respective K-values for calculation of Tg  using the Gordon–Taylor equation were based on density of L-arginine-base (1.30 g/cm3 ) and sucrose (1.43 g/cm3 ) as well as on Tg  values of the pure excipients (sucrose, −32.3°C measured under identical conditions) and values provided in Figure 1 (titrated L-arginine base, pH 6.0): KF01/F02 = 0.9, KF03/F04 = 1.3, KF05/F06 = 1.3, KF07/F08 = 1.1. d Difference between calculated and measured values for Tg  (Tg  -calc. – Tg  -measured). a b

Karl Fischer Moisture Determination A Karl Fischer Moisture Analyzer 831 KF Coulometer equipped with a 874 Karl Fischer (KF) Oven Sample Processor (both Metrohm, Filderstadt, Germany) was used for residual moisture analysis. The precision of the measurement was verified using a crystalline water standard (HydranalTM Water Standard KF Oven). The freeze dried samples were carefully comminuted within the vial under dry atmosphere (glove box, RH 0.05) coincided with the likely formation of small amounts of collapsed product that resisted reconstitution longer than the main body of the cake, leading to extended overall RT. Generally, extended RT is a known issue often associated with freeze dried products showing collapse.54,56 RT of the L-arginine-rich systems did not differ significantly from the values obtained for the sucrose-rich samples, except for the phosphate formulation which showed a lower RT (p < 0.05). The glass transition temperature of the freeze dried solids was not only governed by their residual moisture content (cf. Table 4), but was also affected by formulation composition. The highest Tg was determined for mixtures titrated with citric acid, and the results were lower if phosphoric or succinic acid was added to the mixture. In line with observations for the freeze concentrated solutions, Tg varied with the type of counter ion, which may be attributed to the ability of the present L-arginine

Table 4. Analytical Results and Product Appearance of Freeze Dried Placebo Formulations Type of Formulation

26 mg/mL L-Arginine/104 mg/mL Sucrose

Type of L-Arginine Counter Ion Cake appearance, macroscopica

%, No defects %, Minor defects %, Major defects

Average residual moisture (%)b Glass transition temperature (Tg ) (°C) Reconstitution time (s)c

Chloride

Citrate

Phosphate

Succinate

0 31 69 1.8 ± 0.3 31.4 70 ± 20

0 100 0 2.9 ± 0.2 49.4 49 ± 1

34 66 0 2.7 ± 0.2 42.3 50 ± 1

0 100 0 2.0 ± 0.3 44.2 51 ± 2

Type of Formulation

104 mg/mL L-Arginine / 26 mg/mL Sucrose

Type of L-Arginine Counter Ion Cake appearance, macroscopica

%, No defects %, Minor defects %, Major defects

Average residual moisture (%)b Glass transition temperature (Tg ) (°C) Reconstitution time (s)c

Chloride

Citrate

Phosphate

Succinate

0 0 100 3.5 ± 0.3 17.2 49 ± 1

9 63 29 4.0 ± 0.4 65.5 60 ± 20

3 86 11 4.6 ± 0.3 47.5 37 ± 7

3 17 80 5.5 ± 0.9 40.2 49 ± 2

a

Total number of inspected samples/formulation: 35. Mean, n = 4 ± SD. c Mean, n = 3 ± SD. 30 s injection time subtracted from average reconstitution values. b

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

DOI 10.1002/jps.24501

2353

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

and counter ion combination to form an interactive network in the dried amorphous solid. Specific interactions result in an increased Tg of the dry glass, whereas water molecules in the amorphous phase have a plasticizing effect.58,59 These competing effects may further contribute to the divergent correlation between moisture content and Tg , especially if bonds between molecules are strong as is expected for L-arginine citrate or phosphate-based systems. In fact, Tg of these mixtures was higher at high L-arginine concentrations, despite the higher moisture level in the dried cakes (cf. Table 4). On the other hand, higher moisture content was associated with a lower Tg for the L-arginine chloride-based formulation systems in which interactions are weaker. This indicates that thermal characteristics of the L-arginine-based formulations are affected by multiple effects, including the strength and number of intermolecular bonds, and the ability of the respective L-arginine and counter ion combination to retain water within the amorphous glass. Influence of the Protein Fraction on Thermal Characteristics, Drying Behavior, and Product Appearance For manufacturing of stability samples, the investigated placebo mixtures were used as a basis to prepare formulations containing 50 mg/mL of a MAb. Total solid content of the protein-containing systems was maintained at 130 mg/mL by adjusting the respective excipient fraction to 80 mg/mL as provided in Table 1 (F09–F17). The collapse temperature of the various mixtures was higher in presence of the protein (cf. Fig. 5b). This is consistent with a number of reports showing that Tg  as well as the temperature difference between the onset of collapse and full collapse determined in the freeze dry microscope increase as a function of the protein concentration.48,60–62 Lewis et al.61 found no measurable difference in the formulation Tg  for sucrosebased mixtures containing either lysozyme, bovine serum al-

bumin (BSA) or IgG, in which the protein accounted for about 16% (w/w) of the total dry weight, but observed a significant influence on the freeze drying properties at 44% (w/w). In the present mixtures, the protein fraction comprised approximately 39% (w/w) of the cake composition. According to thermocouple data, Tp-TC still exceeded the determined onset of collapse temperatures during primary drying (by about 2°C to 7°C depending on the formulation), comparable to the extent observed for placebo. However, unlike for the protein-free systems, product appearance was less affected by the high product temperature. Visual inspection of the sucrose-rich systems showed only small differences, compared to the placebo cakes, except for a lower number of severely damaged vials among the L-arginine chloride-containing samples (cf. Table 5). However, improvement of cake appearance became obvious for samples containing the high L-arginine concentration. Here, defects were “minor” level for L-arginine citrate and succinate-based products and the cake structure was fully intact if phosphoric acid was added to the system. The physical stability of amorphous protein-containing systems that were freeze dried at or even above their collapse temperature determined by freeze dry microscopy has been reported for various biopharmaceutical formulations.48,60,62,63 Beside the obvious influence of the high-molecular-weight component, recent advances in optical coherence tomography (OCT) provide evidence that collapse temperature measured in the microscope may deviate from collapse temperature in a product vial due to differences in ice nucleation, resulting dry layer resistance and drying rate if the product is dried from a thin film.64,65 In fact, results provided by 3D-imaging of the OCT technique indicated that the effective collapse temperature in a vial may be several degrees higher than found in the microscope. However, even in presence of the protein, about 22% of the vials with the high concentration of L-arginine chloride showed distinct loss of cake integrity and were therefore categorized as “major defects” (cf. Table 5).

Table 5. Analytical Results and Product Appearance of Freeze Dried L-Arginine/Sucrose-Based Protein Formulations Type of Formulation

16 mg/mL L-Arginine/64 mg/mL Sucrose/50 mg/mL MAb

Type of L-Arginine Counter Ion Cake Appearance, macroscopica

%, No defects %, Minor defects %, Major defects

Average residual moisture (%)b Glass transition temperature (Tg ) (°C) Reconstitution time (s)c

Chloride

Citrate

Phosphate

Succinate

0 63 37 1.1 58.2 96 ± 41

7 93 0 1.2 69.9 106 ± 6

48 52 0 1.1 72.7 113 ± 13

0 100 0 1.0 66.4 88 ± 23

Type of Formulation

64 mg/mL L-Arginine/16 mg/mL Sucrose/50 mg/mL MAb

Type of L-Arginine Counter Ion Cake Appearance, macroscopica

Average residual moisture (%)b Glass transition temperature (Tg ) (°C) Reconstitution time (s)c

%, No defects %, Minor defects %, Major defects

Chloride

Citrate

Phosphate

Succinate

0 78 22 1.5 59.6 71 ± 22

67 33 0 1.5 85.6 133 ± 18

100 0 0 1.5 79.2 83 ± 15

4 96 0 1.4 69.8 128 ± 26

a

Total number of inspected samples/formulation: 27. Mean, n = 2. c Mean, n = 3 ± SD. 30 s injection time subtracted from average reconstitution values. b

DOI 10.1002/jps.24501

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

2354

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Influence of the Protein Fraction on Moisture Content, RT, and Glass Transition Temperature Residual moisture of the protein mixtures was lower than for placebo, especially for the high L-arginine concentration. Similar to placebo, the moisture content consistently increased for higher L-arginine concentrations, but did not exceed 1.5%. As expected based on literature data,66–68 the presence of the protein prolonged RT which exceeded 2 min for individual samples (cf. Table 5). Glass transition temperatures of the investigated dried cakes were higher compared with the placebo samples (cf. Table 5). However, the subtle differences between the mixtures showed no clear correlation to the differences in moisture content, but varied as a function of the formulation composition. This may provide further evidence that thermal behavior of the complex formulation systems is governed by a balance between the plasticizing effect of water and molecular interactions in the amorphous phase. Protein Stability during Freeze Drying and Storage Proteins formulated without stabilizer or with an insufficient amount of stabilizing excipients can undergo significant degradation upon freeze drying or shelf life.8,69 Cleland et al.70 and others7,71 showed that stability was related to the ratio of disaccharide stabilizers to protein in the formulation. For a number of pharmaceutical proteins, adequate stability over shelf life was achieved at a molar ratio of approximately 360:1 (lyoprotectant:protein).70,72,73 For the investigated MAb formulations, both sucrose and L-arginine molecules are expected to interact with the protein if they remain in the amorphous state, and hence contribute to its physical stabilization in the lyophilized solid. At the intended protein concentration of 50 mg/mL, the applied total stabilizer concentration of 80 mg/mL (i.e., sucrose and L-arginine) provided a molar ratio of stabilizer to protein which was consistently higher than the molar ratio mentioned by Cleland et al.70 for all compounding ratios and counter ions. However, considering that the molecular weight of L-arginine is about half than that of sucrose, the molar ratio of stabilizer to protein was higher at the high L-arginine

Table 6. In Process Degradation of MAb in all Formulations and Rate Constants for Physical Aggregation k (%A/month0.5 ) During Storage Stability Studies

Name of Formulation F09 F10 F11 F12 F13 F14 F15 F16 F17

Storage Stability 180 Days

Level of Aggregation before FD (% Area)a

Initial Level of Aggregation after FD (% Area)a

b k at 40°C √ ± S.E. (%A/ month)

2.9 2.9 2.7 2.8 2.9 2.8 2.9 2.8 2.6

2.9 2.9 2.6 2.9 2.9 2.9 2.9 2.9 3.0

0.10 ± 0.01 0.08 ± 0.004 0.29 ± 0.06 0.43 ± 0.04 0.31 ± 0.02 0.74 ± 0.06 0.18 ± 0.02 0.27 ± 0.01 ࣘ0.01*

a Reported from integration of the associated peak area relative to the total peak AUC. b S.E. = standard error for k values (%A/month0.5 ) estimated from multiple linear regression (GraphPad PrismTM ). a p > 0.05, for H0 : k = 0 by ANCOVA (analysis of covariance).

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

concentration compared to the sucrose-rich formulations. Assuming that mechanisms associated with the stabilizing effect of both types of molecules are similar,26 a higher L-arginine concentration might be expected to promote protein stability. To clarify these effects, protein aggregation was assayed by SE-HPLC prior to and directly after freeze drying, and after 30 days, 90 days, and 180 days of storage at controlled temperature conditions. Turbidity measurements of reconstituted solution provided no indications for larger aggregates or particulates for the samples stored up to 12 months at the maximum storage temperature of 40°C (data not shown). The difference in the level of protein aggregation before and after the freeze drying process (cf. Table 6) as well as aggregation induced during storage at 5°C and 25°C (data not shown) was found negligibly low. Even at 25°C, the calculated aggregation rate constants were not significantly different from zero (p > 0.05, ANCOVA) indicating stability of the protein over the 180 days storage period. Calculated degradation rates for these conditions were not high enough to reveal formulation-specific differences. Differences in stability were, however, noticed for samples stored at 40°C. Based on the data for the higher storage temperature, protein stability in the freeze dried state varied significantly with formulation composition and the type of L-arginine counter ion. Interestingly, aggregation of the investigated MAb was promoted at the high L-arginine concentration for formulations containing citric, phosphoric, and succinic acid (cf. Table 6). Both sucrose to L-arginine mixing ratios showed increasing aggregation rate constants (p < 0.05, ANCOVA) for the varying counter ions in the order phosphate > citrate > succinate. The lowest aggregation rates were determined for mixtures titrated with hydrochloric acid. In fact, protein stability in the presence of chloride ions was similar to stability observed for a sucrose reference formulation included for comparison (F17). Opposed to the other tested counter ions, the higher L-arginine concentration (i.e., lower sucrose concentration) further decreased the aggregation rate of the investigated MAb at accelerated storage conditions. As discussed earlier, freeze drying at aggressive primary drying conditions led to formation of collapse in some of the cakes. The formation of product collapse not only causes the loss of “pharmaceutical elegance,”3,74 but has been reported to affect the product’s final moisture content and RT. Several studies, however, indicate that product collapse may be inconsequential for protein stability.50,54,75–77 In fact, studies using the concept of foam drying showed that protein stability does not even require an elegantly freeze dried cake.16 Collapse may, however, alter the properties of the freeze dried material. Viscous flow may increase density of the cakes and may eventually result in blockage of pores. This may affect product resistance, sublimation rate and primary drying time.53,54,78 The formation of collapse may also reduce specific surface area and thereby retard secondary drying ultimately leading to elevated moisture content.3 Increased density hinders diffusion of water molecules through the dried matrix, resulting in the same effect. Moisture content may be more critical for protein stability during storage compared to the pure presence of collapsed cake regions. Published data indicate that non-collapsed and collapsed cakes incubated at a comparable moisture level showed comparable physical stability.79 The moisture content of the protein cakes investigated in the present study was almost identical at the same L-arginine to sucrose mixing ratio and may therefore be excluded as the source of counter ion specific differences DOI 10.1002/jps.24501

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

in protein stability (cf. Table 5). There is also some evidence that collapse increases the tendency of amorphous material to crystallize.80,81 This may prevent physical interaction of the stabilizer molecules with the protein and therefore weaken its stabilizing effect.22,53,81–83 Crystallization of amorphous components during storage was not observable for the investigated solids, based on XRPD analysis of additional samples of each composition which were stored at 40°C for up to 240 days (data not shown). Therefore, it may be concluded that observable differences in protein stability can be attributed to the specific ability of the amorphous component system to protect the MAb against denaturation in the glassy state. Estimation of Global Mobility Given the significant effect of the counter ion on glass transition temperature of L-arginine, it appears plausible that molecular mobility (global motion or "-relaxation) varies among the different formulations. To quantify molecular mobility, the relaxation time constant J$ was estimated as described in the methods section. Relaxation time J$ can be regarded as inversely proportional to the molecular mobility for global motion10,16 , and may be correlated to storage stability (assuming that molecular mobility is coupled with the type of motion leading to protein degradation in the solid state).84,85 A plot of the physical aggregation rate constant ln(k) versus the estimated ln(J$ ) (cf. Fig. 6) suggests that stability increases with shorter relaxation time values, that is, increasing molecular mobility. This correlation is somewhat unexpected and an explanation is not obvious at this point. However, the missing correlation between global mobility and protein aggregation may support an integral role of fast $-relaxations for protein stability11,17 but may also be associated with the individual effects of the L-arginine counter ion combinations on protein secondary structure. In investigation of the stability of BSA and IgG, glass-forming organic acid salts (e.g., sodium citrate) revealed a structure stabilizing effect to the freeze dried proteins, indicating the contribution of direct interactions between the salt and the protein molecules.86 Here, hydrogen

2355

bonds to the protein that substitute for water molecules lost upon drying were considered essential. In addition, electrostatic interactions (e.g., ion–ion, ion–dipole) between the salt and basic amino acid residues on the protein surface were suggested to be involved in structural stabilization26,87 as well as number and type of functional groups. For example, the absence of hydroxyl groups was found the likely reason for the limiting structure stabilizing effect of glass forming succinic acid salts.86 In contrast, human growth hormone was destabilized by adding chloride ions.88 Beside potential interactions between the counter ion and the protein, functional groups of the L-arginine molecules may further contribute to the complex interactions in the solid state. Further work is required to deepen insight into mechanisms on a molecular level and their effect of protein structural stability. Finally, the obtained results clearly suggest that both formulation composition and the type of counter ion affect storage stability. Against expectations, the higher glass transition temperatures (Tg and Tg  ) achieved by combination of L-arginine and multivalent acids were associated with a higher level of protein aggregation, whereas the depressing effect of chloride ions on glass transition and onset of collapse temperature correlated with improved protein stability. Therefore, careful optimization of the formulation composition and the freeze drying process conditions is required to benefit from the superior stability achieved if L-arginine is titrated with hydrochloric acid, while compensating for the deficient effect on collapse behavior caused by L-arginine in the presence of chloride ions.

CONCLUSIONS L-Arginine could constitute a promising excipient for formulation of freeze dried proteins. In a sucrose-based formulation design, the type of the L-arginine counter ion greatly affected the collapse behavior of the formulation. Within the narrow pH range dictated by the stability of the therapeutic MAb used in this study, the effect of the acid to L-arginine mixing ratio on Tg  was less pronounced than the effect exerted by variation of the counter ion. The type of L-arginine counter ion also affected residual moisture content and physical properties of the dried solids. However, the high glass transition temperatures achieved by combining L-arginine and multivalent acids (e.g., citric acid and phosphoric acid) did not increase stability of the protein, as aggregation was likely not governed by global mobility of the glasses. Specific mechanisms of degradation may be the reason for the observed differences in stability. The lowest rate of aggregation during storage among the investigated antibody formulations was achieved for the mixtures containing L-arginine chloride, which showed a critically low collapse temperature. Thus, further research is required to assess formulation and process variables in order to possibly combine the beneficial effect of L-arginine chloride-based formulation systems on protein stability with improvement of collapse behavior.

ACKNOWLEDGMENTS Figure 6. Estimated ln (J$ ) versus the rate constant for physical aggregation ln(k) (%A/month0.5 ). Symbols correspond to results obtained at a storage temperature of 40°C. Error bars indicate standard error for aggregation rate constants. DOI 10.1002/jps.24501

The authors would like to acknowledge F. Hoffmann-La Roche for financial support of this project. We would also like to acknowledge Dr. Christian Schmalz at F. Hoffmann-La Roche, Basel, Switzerland for providing the drug substance as well as St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

2356

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

Mrs. Vanessa Haefliger and Mrs. Eva Keller at F. HoffmannLa Roche, Basel, Switzerland for her support with SE-HPLC measurements. Finally, we would like to thank Dr. Fred Lim, Dr. Puneet Sharma, and Dr. Lokesh Kumar at Genentech, San Francisco, California for extremely helpful discussions and critical insight.

REFERENCES 1. Carpenter JF, Chang BS, Garzon-Rodriguez W, Randolph TW. 2002. Rational design of stable lyophilized protein formulations: Theory and practice. Pharm Biotechnol 13:109–133. 2. Carpenter JF, Pikal MJ, Chang BS, Randolph TW. 1997. Rational design of stable lyophilized protein formulations: Some practical advice. Pharm Res 14(8):969–975. 3. Nail SL, Jiang S, Chongprasert S, Knopp SA. 2002. Fundamentals of freeze-drying. Pharm Biotechnol 14:281–360. 4. Tang X, Pikal MJ. 2004. Design of freeze-drying processes for pharmaceuticals: Practical advice. Pharm Res 21(2):191–200. 5. Franks F. 1992. Freeze-drying: From empiricism to predictability. The significance of glass transitions. Dev Biol Stand 74:9–18; discussion 19. 6. Franks F. 1998. Freeze-drying of bioproducts: Putting principles into practice. Eur J Pharm Biopharm 45(3):221–229. 7. Wang B, Tchessalov S, Warne NW, Pikal MJ. 2008. Impact of sucrose level on storage stability of proteins in freeze-dried solids: I. Correlation of protein-sugar interaction with native structure preservation. J Pharm Sci 98(9):3131–3144. 8. Carpenter JF, Prestrelski SJ, Dong A. 1998. Application of infrared spectroscopy to development of stable lyophilized protein formulations. Eur J Pharm Biopharm 45(3):231–238. 9. Franks F, Hatley RHM, Mathias SF. 1991. Material science and the production of shelf-stable biologicals. Bio Pharm 4(38). 10. Pikal MJ. 2004. Mechanisms of protein stabilization during freezedrying and storage: The relative importance of thermodynamic stabilization and glassy state relaxation dynamics. In Freeze drying/lyophilization of pharmaceutical and biological products; Ray L, May JC, editors. New York, London: Informa Healthcare. p 63–108. 11. Cicerone MT, Tellington A, Trost L, Sokolov A. 2003. Substantially improved stability of biological agents in dried form: The role of glassy dynamics in preservation of biopharmaceuticals. BioProcess Int 1:36– 47. 12. Yoshioka S, Aso Y, Kojima S. 1996. Determination of molecular mobility of lyophilized bovine serum albumin and gamma-globulin by solid-state 1H NMR and relation to aggregation-susceptibility. Pharm Res 13(6):926–930. 13. Duddu SP, Zhang G, Dal Monte PR. 1997. The relationship between protein aggregation and molecular mobility below the glass transition temperature of lyophilized formulations containing a monoclonal antibody. Pharm Res 14(5):596–600. 14. Chang LL, Shepherd D, Sun J, Ouellette D, Grant KL, Tang XC, Pikal MJ. 2005. Mechanism of protein stabilization by sugars during freeze-drying and storage: Native structure preservation, specific interaction, and/or immobilization in a glassy matrix? J Pharm Sci 94(7):1427–1444. 15. Davidson P, Sun WQ. 2001. Effect of sucrose/raffinose mass ratios on the stability of co-lyophilized protein during storage above the Tg. Pharm Res 18(4):474–479. 16. Abdul-Fattah AM, Truong-Le V, Yee L, Nguyen L, Kalonia DS, Cicerone MT, Pikal MJ. 2007. Drying-induced variations in physicochemical properties of amorphous pharmaceuticals and their impact on stability (I): Stability of a monoclonal antibody. J Pharm Sci 96(8):1983– 2008. 17. Cicerone MT, Soles CL. 2004. Fast dynamics and stabilization of proteins: Binary glasses of trehalose and glycerol. Biophys J 86(6):3836–3845. St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

18. Pikal MJ, Chang LL, Tang XC. 2004. Evaluation of glassy-state dynamics from the width of the glass transition: Results from theoretical simulation of differential scanning calorimetry and comparisons with experiment. J Pharm Sci 93(4):981–994. 19. Hancock BC, Dalton CR, Pikal MJ, Shamblin SL. 1998. A pragmatic test of a simple calorimetric method for determining the fragility of some amorphous pharmaceutical materials. Pharm Res 15(5):762–767. 20. Liu J, Rigsbee DR, Stotz C, Pikal MJ. 2002. Dynamics of pharmaceutical amorphous solids: The study of enthalpy relaxation by isothermal microcalorimetry. J Pharm Sci 91(8):1853–1862. 21. Seguro K, Tamiya T, Tsuchiya T, Matsumoto JJ. 1990. Cryoprotective effect of sodium glutamate and lysine-HCl on freeze denaturation of lactate dehydrogenase. Cryobiology 27(1):70–79. 22. Izutsu K, Yoshioka S, Takeda Y. 1991. The effect of additives on the stability of freeze-dried ß-galactosidase stored at elevated temperature. Int J Pharm 71(1–2):137–146. 23. Hsu CC, Nguyen HM, Yeung DA, Brooks DA, Koe GS, Bewley TA, Pearlman R. 1995. Surface denaturation at solid-void interface—A possible pathway by which opalescent particulates form during the storage of lyophilized tissue-type plasminogen activator at high temperatures. Pharm Res 12(1):69–77. 24. Mattern M, Winter G, Kohnert U, Lee G. 1999. Formulation of proteins in vacuum-dried glasses. II. Process and storage stability in sugar-free amino acid systems. Pharm Dev Technol 4(2):199–208. 25. Izutsu K, Kadoya S, Yomota C, Kawanishi T, Yonemochi E, Terada K. 2008. Freeze-drying of proteins in glass solids formed by basic amino acids and dicarboxylic acids. Chem Pharm Bull (Tokyo) 57(1):43–48. 26. Izutsu K, Fujimaki Y, Kuwabara A, Aoyagi N. 2005. Effect of counterions on the physical properties of l-arginine in frozen solutions and freeze-dried solids. Int J Pharm 301(1–2):161–169. 27. Buchner J, Rudolph R. 1991. Renaturation, purification and characterization of recombinant Fab-fragments produced in Escherichia coli. Biotechnology (NY) 9(2):157–162. 28. Shiraki K, Motonori K, Shinsuke F, Tadayuki I, Takagi M. 2002. Biophysical effects of amino acids on the prevention of protein aggregation. J Biochem 132(4):591–595. 29. Arakawa T, Tsumoto K. 2003. The effects of arginine on refolding of aggregated proteins: Not facilitate refolding, but suppress aggregation. Biochem Biophys Res Commun 304(1):148–152. 30. Arakawa T, Ejima D, Tsumoto K, Obeyama N, Tanaka Y, Kita Y, Timasheff SN. 2006. Suppression of protein interactions by arginine: A proposed mechanism of the arginine effects. Biophys Chem 127(1– 2):1–8. 31. Tsumoto K, Umetsu M, Kumagai I, Ejima D, Philo JS, Arakawa T. 2004. Role of arginine in protein refolding, solubilization, and purification. Biotechnol Prog 20(5):1301–1308. 32. Wang W. 2000. Lyophilization and development of solid protein pharmaceuticals. Int J Pharm 203(1–2):1–60. 33. Gieseler H. 2012. Quality by design (QbD) in freeze drying. In Encyclopedia of pharmaceutical technology; Swarbrick J, editor. London, UK: Informa Healthcare. 34. Shamblin SL, Taylor LS, Zoegrafi G. 1998. Mixing behavior of colyophilized binary systems. J Pharm Sci 87(6):694–701. 35. Simha R, Boyer RF. 1962. On a general relation involving the glass temperature and coefficients of expansion of polymers. J Chem Phys 37(5):1003–1007. 36. Hancock B, Zoegrafi G. 1994. The relationship between the glass transition temperature and the water content of amorphous pharmaceutical solids. Pharm Res 11(4):471–477. 37. Shamblin SL, Huang EY, Zografi G. 1996. The effects of colyophilized polymeric additives on the glass transition temperature and crystallization of amorphous sucrose. J Therm Anal 47(5):1567– 1579. 38. Pikal MJ, Rigsbee D, Roy ML, Galreath D, Kovach KJ, Wang B, Carpenter JF, Cicerone MT. 2008. Solid state chemistry of proteins: II. The correlation of storage stability of freeze-dried human growth hormone (hGH) with structure and dynamics in the glassy solid. J Pharm Sci 97(12):5106–5121. DOI 10.1002/jps.24501

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

39. Pikal MJ, Rigsbee DR. 1997. The stability of insulin in crystalline and amorphous solids: Observation of greater stability for the amorphous form. Pharm Res 14(10):1379–1387. 40. Hodge IM. 1994. Enthalpy relaxation and recovery in amorphous materials. J Non-Cryst Solids 169(3):211–266. 41. Hodge IM. 1987. Effects of annealing and prior history of enthalpy relaxation in glass polymers. 6. Adam-Gibbs formulation of nonlinearity. Macromolecules 20(11):2897–2908. 42. Shamblin SL, Tang XL, Chang LQ, Hancock BC, Pikal MJ. 1999. Characterization of the time scales of molecular motion in pharmaceutically important glasses. J Phys Chem B 103(20):4113–4121. 43. Yamaki T, Ohdate R, Nakadai E, Yoshihashi Y, Yonemochi E, Terada K, Moriyama H, Izutsu K, Yomota C, Okuda H, Kawanishi T. 2012. Component crystallization and phsical collapse during freeze drying of L-arginine citrate acid mixtures. Chem Pharm Bull (Tokyo) 60(9):1176–1181. 44. Kadoya S, Izutsu K, Yonemochi E, Terada K, Yomota C, Kawanishi T. 2008. Glass-state amorphous solids formed by freeze drying of amines and hydroxyl carboxylic acids: The effect of hydrogen-bonding and electrostatic interactions. Chem Pharm Bull (Tokyo) 56(6):821– 826. 45. Aoki K, Nagano K, Iitaka Y. 1970. The Crystal Structure of LArginine Phosphate Monohydrate. Acta Cryst B (27):11–23. 46. Fukumoto K, Yoshizawa M, Ohno H. 2004. Room Temperature Ionic Liquids from 20 Natural Amino Acids. J Am Chem Soc 127(8):2398– 2399. 47. Terzyan SS, Karapethyan HA, Sukiasyan RB, Petrosyan AM. 2004. L-Arginine nitrates. J Mol Struct 687(1–3):111–117. 48. Meister E, Gieseler H. 2008. Freeze-dry microscopy of protein/sugar mixtures: Drying behavior, interpretation of collapse temperatures and a comparison to corresponding glass transition data. J Pharm Sci 98(9):3072–3087. 49. Pikal MJ. 1990. The collapse temperature in freeze drying: Dependence on measurement methodology and rate of water removal from the glassy phase. Int J Pharm 62:165–186. 50. Schersch K, Betz O, Garidel P, Muehlau S, Bassarab S, Winter G. 2010. Systematic investigation of the effect of lyophilizate collapse on pharmaceutically relevant proteins I: Stability after freeze drying. J Pharm Sci 99(5):2256–2278. 51. Righetti MC, Ajroldi G, Pezzin G. 1992. The glass transition temperature of polymer-diluent systems. Polymer 33(22):4779– 4785. 52. Cavatur RK, Suryanarayanan R. 1998. Characterization of phase transitions during freeze-drying by in situ X-ray powder diffractometry. Pharm Dev Technol 3(4):579–586. 53. Tsourouflis S, Flink JM, Karel M. 1976. Loss of structure in freezedried carbohydrates solutions: Effect of temperature, moisture content and composition. J Sci Food Agric 27(6):509–519. 54. Wang DQ, Hey JM, Nail SL. 2004. Effect of collapse on the stability of freeze-dried recombinant factor VIII and alpha-amylase. J Pharm Sci 93(5):1253–1263. ¨ 55. Schneid SC, Startzel P, Lettner P, Gieseler H. 2011. Robustness testing of pharmaceutical freeze-drying: Inter-relation of process conditions and product quality attributes studied for a vaccine formulation. Pharm Dev Technol 16(6):583–590. 56. Tang X, Nail SL, Pikal MJ. 2006. Evaluation of manometric temperature measurement, a process analytical technology tool for freezedrying: Part I, product temperature measurement. AAPS PharmSciTech 7(1):95–103. 57. Tang XC, Nail SL, Pikal MJ. 2006. Evaluation of manometric temperature measurement (MTM), a process analytical technology tool in freeze drying, part III: Heat and mass transfer measurement. AAPS PharmSciTech 7(4):97. 58. Youmin Y. 2007. Modulation of molecular mobility in sucrose-based amorphous solids detected by phosphorescence of erythrosin B. Graduate School New Brunswick. New Brunswick, New Jersey: The State University of New Jersey.

DOI 10.1002/jps.24501

2357

59. Youmin Y, Ludescher DR. 2008. The effect of salts on molecular mobility in amorphous sucrose monitored by erythrosin B phosphorescence. Carbohyd Res 343(15):2641–2649. 60. Johnson RE, Oldroyd ME, Ahmed SS, Gieseler H, Lewis LM. 2010. Use of manometric temperature mesurement (MTM) to characterize the freeze drying behavior of amorphous protein formulations. J Pharm Sci 99(6):2863–2873. 61. Lewis LM, Johnson RE, Oldroyd ME, Ahmed SS, Joseph L, Saracovan I, Sinha S. 2010. Characterization of freeze-drying behavior of model protein formulations. AAPS Pharm Sci Tech 11(4):1580–1590. 62. Colandene JD, Maldonado LM, Creagh AT, Vrettos JS, Goad KG, Spitznagel TM. 2007. Lyophilization cycle development for a highconcentration monoclonal antibody formulation lacking a crystalline bulking agent. J Pharm Sci 96(6):1598–1608. 63. Fonseca F, Passot S, Cunin O, Marin M. 2004. Collapse temperature of freeze-dried Lactobacillus bulgaricus suspensions and protective media. Biotechnol Prog 20(1):229–238. 64. Greco K, Mujat M, Galbally-Kinney KL, Hammer XD, Ferguson DR, Iftimia N, Mulhall P, Kessler WJ, Pikal MJ. 2013. Accurate prediction of collapse temperature using optical coherence tomography-based freeze drying microscopy. J Pharm Sci 102(6):1773–1785. 65. Mujat M, Greco K, Galbally-Kinney KL, Hammer XD, Ferguson DR, Iftimia N, Mulhall P, Sharma P, Pikal MJ, Kessler WJ. 2012. Optcial coherence tomography-based freeze drying microscopy. Biomed Opt Express 3(1):55–63. 66. Cao W, Krishnan S, Ricci MS, Shih LY, Liu D, Gu JH, Jameel F. 2013. Rational design of lyophilized high concentrated protein formulations – mittigating the challange of slow reconstitution with multidisciplinary strategies. Eur J Pharm Biopharm 85(2):287–293. 67. Lin LFH, Bunnel R. 2013. Overcoming challanges in the reconstitution of high-concentrated protein drug products. Biopharm Int 26(3):28–29. 68. Shire SJ, Shahrokh Z, Liu J. 2004. Challanges in development of high protein concentration formulations. J Pharm Sci 93(6):1390– 1402. 69. Prestrelski SJ, Tedeschi N, Arakawa T, Carpenter JF. 1993. Dehydration-induced conformational transitions in proteins and their inhibition by stabilizers. Biophys J 65(2):661–671. 70. Cleland JL, Lam X, Kendrick B, Yang J, Yang TH, Overcashier D, Brooks D, Hsu C, Carpenter JF. 2001. A specific molar ratio of stabilizer to protein is required for storage stability of a lyophilized monoclonal antibody. J Pharm Sci 90(3):310–321. 71. Sane SU, Wong R, Hsu CC. 2004. Raman spectroscopic characterization of drying-induced structural changes in a therapeutic antibody: Correlating structural changes with long-term stability. J Pharm Sci 93(4):1005–1018. 72. Chang BS, Reeder G, Carpenter JF. 1996. Development of a stable freeze-dried formulation of recombinant human interleukin-1 receptor antagonist. Pharm Res 13(4):243–249. 73. Costantino HR, Langer R, Klibanov AM. 1995. Aggregation of a lyophilized pharmaceutical protein, recombinant human albumin: Effect of moisture and stabilization by excipients. Biotechnology 13(5):493–496. 74. Pikal MJ. 1990. Freeze-drying of proteins. Part I: Process design. BioPharm (3):18–28. 75. Chatterjee K, Shalaev EY, Suryanarayanan R. 2005. Partially crystalline systems in lyophilization: II. Withstanding collapse at high primary drying temperatures and impact on protein activity recovery. J Pharm Sci 94(4):809–820. 76. Jiang S, Nail SL. 1998. Effect of process conditions on recovery of protein activity after freezing and freeze-drying. Eur J Pharm Biopharm 45(3):249–257. 77. Schersch K, Betz O, Garidel P, Muehlau S, Bassarab S, Winter G. 2012. Systematic investigation of the effect of lyophilizate collapse on pharmaceutical relevant proteins, part 2: Stability during storage at elevated temperatures. J Pharm Sci 101(7):2288– 2305.

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

2358

RESEARCH ARTICLE – Pharmaceutics, Drug Delivery and Pharmaceutical Technology

78. Mac Kenzie AP. 1975. Collapse during freeze drying—Qualitative and quantitative aspects. In Freeze drying and advanced food technology; Goldsmith SA, Rey L, Rothmayer WW, editors. London: Academic Press London. pp. 277–307. 79. Lueckel B, Helk B, Bodmer D, Leuenberger H. 1998. Effects of formulation and process variables on the aggregation of freeze-dried interleukin-6 (IL-6) after lyophilization and on storage. Pharm Dev Technol 3(3):337–346. 80. Izutsu K, Yoshioka S, Kojima S. 1994. Physical stability and protein stability of freeze-dried cakes during storage at elevated temperatures. Pharm Res 11(7):995–999. 81. Darcy P, Buckton G. 1997. The influence of heating/drying on the crystallisation of amorphous lactose after structural collapse. Int J Pharm Sci 158(2):157–164. 82. Hora MS, Rana RK, Smith FW. 1992. Lyophilized formulation of recombinant tumor necrosis factor. Pharm Res 9(1):33–36. 83. Izutsu K, Yoshioka S, Terao T. 1993. Decreased protein-stabilizing effects of cryoprotectants due to crystallization. Pharm Res 10(8):1232– 1237.

St¨artzel et al., JOURNAL OF PHARMACEUTICAL SCIENCES 104:2345–2358, 2015

84. Abdul-Fattah AM, Truong-Le V, Yee L, Pan E, Ao Y, Kalonia DS, Pikal MJ. 2007. Drying-induced variations in physico-chemical properties of amorphous pharmaceuticals and their impact on Stability II: Stability of a vaccine. Pharm Res 24(4):715–727. 85. Poirier-Brulez F, Roudant G, Champion D, Tanguy M, Simatos D. 2005. Influence of water content on molecular mobility in starchbased glasses as assessed through structure and secondary relaxation. Biopolymers 81(2):63–73. 86. Izutsu K, Kadoya S, Yomota C, Kawanishi T, Yonemochi E, Terada K. 2009. Stabilization of protein structure in freeze dried amorphous organic acid buffer salts. Chem Pharm Bull (Tokyo) 57(11):1231– 1236. 87. Tian F, Middaugh CR, Offerdahl T, Munson E, Sane S, Rytting JH. 2006. Spectroscopic evaluation of the stabilization of humanized monoclonal antibodies in amino acid formulations. Int J Pharm 335(1– 2):20–31. 88. Pikal MJ, Dellerman KM, Roy ML, Riggin RM. 1991. The effects of formulation variables on the stability of freeze-dried human growth hormone. Pharm Res 8(4):427–436.

DOI 10.1002/jps.24501