Improved formulation and lyophilization cycle for rBCG vaccine

Vaccine 29 (2011) 4848–4852

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Improved formulation and lyophilization cycle for rBCG vaccine Tom H. Jin ∗ , Lisa Nguyen, Tianli Qu, Eric Tsao Aeras, 1405 Research Blvd, Suite 300, Rockville, MD 20850, USA

a r t i c l e

i n f o

Article history: Received 17 March 2011 Received in revised form 13 April 2011 Accepted 16 April 2011 Available online 5 May 2011 Keywords: Formulation Freeze drying rBCG Vaccine Tuberculosis

a b s t r a c t To improve the conventional BCG vaccine in cake appearance and integrity, a new formulation with corresponding freeze drying cycle was developed for a recombinant BCG vaccine. The new formulation contains mannitol as a bulking agent, and trehalose, sucrose and sodium glutamate as stabilizers. The formulation and freeze drying cycle were tested with different super cooling rates and secondary drying temperatures, with or without an annealing process. Thermodynamic behavior was characterized using differential scanning calorimetry (DSC). Varying the secondary drying temperature and presence/absence of an annealing step caused marked differences in cake thermodynamic profiles irrespective of different cooling rates. The annealing process allowed efficient crystallization of the mannitol. Failure to crystallize the bulking agent had the potential to depress the Tg  and compromise storage stability in the final lyophile by crystallizing from the solid during storage, even when the secondary drying temperature was as high as 40 ◦ C. The improved formulation and freeze drying cycle resulted in good recovery of 53.2% during lyophilization and a higher survival rate of 61.7% in an accelerated stability study than the conventional BCG formulation and cycle. In summary, full crystallization was necessary for the mannitol bulking formulation. The freeze dried rBCG vials obtained using the formulation and drying cycle developed here met the requirements of BCG vaccine in good cake appearance, high viability post freeze drying and heat stability during storage. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction Tuberculosis continues to be a major global health problem. Between eight and nine million people are infected annually. Currently, two billion people are thought to be infected with Mycobacterium tuberculosis, and over two million people die of the disease each year [1,2]. The first human vaccination using an attenuated strain, named the bacilli of Calmette and Guérin (BCG), was applied in Paris in 1921 to a newborn whose grandmother had pulmonary tuberculosis. The mass vaccination of children was begun in 1928 and was adopted by many countries, after newer and safer production processes were implemented [3,4]. However, the variable success obtained after vaccination and the increasing appearance of multi-drug resistant strains of M. tuberculosis, which is hindering drug-based control measures, mean that new vaccine developments and optimized vaccination strategies for combating this disease are urgently needed. Although a number of live attenuated bacterial vaccines have been developed for a wide range of diseases including anthrax, brucellosis, cholera, dysentery, plague, typhoid and Q fever, the only two currently in widespread use are BCG vaccine and Ty

∗ Corresponding author. Tel.: +1 2404980607; fax: +1 3015472901. E-mail address: [email protected] (T.H. Jin). 0264-410X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2011.04.056

21a oral typhoid vaccine. When produced as a liquid preparation, stability of BCG was a considerable problem. This has been resolved by freeze-drying in a stabilizing medium, the composition of which varies somewhat between manufacturers. BCG vaccines freeze-dried under appropriate conditions meet the WHO requirements for ≥20% survival after 28 days at 37 ◦ C. Where failures do occur, these are usually attributed to excessive residual moisture content and different stains being used. The Japanese strain of BCG (Tokyo 172) was shown to be considerably more resistant to freeze-drying compared with other BCG strains used in various countries [5]. However, the Japanese strain has been reported to have a lower residual virulence and a weaker immunogenicity than other strains. The discovery that sodium glutamate could be used in place of sucrose or lactose to produce heat-stable, freezedried BCG vaccine was an advance in stabilizing the potency of freeze-dried BCG. When sodium glutamate is employed as a main component in the BCG formulation, the dried BCG vaccine is a white powder or “dust-like” substance that can be easily moved by shaking and does not form cake, potentially resulting in product lose on the wall and stopper of the product vial. The goal of this study was to develop a better formulation for a freezedried rBCG vaccine (known as AERAS-422) constructed by Aeras and to be used in clinical trials. The most important qualities of a freeze-dried rBCG vaccine include: good cake appearance, high viability post freeze drying, and heat stability during storage and transportation.

T.H. Jin et al. / Vaccine 29 (2011) 4848–4852

4849

Table 1 Lyophilization cycles with improved rBCG formulationa . Lot no.

Cooling rate (◦ C/min)

Annealing step

Second drying temp. (◦ C)

Crystalline peak

A B C D E F

5 5 5 5 2 2

No No Yes Yes Yes Yes

35 40 40 50 40 37

Yes, large Yes, large No No No Yes, small

Tg  (◦ C) 70 70 90 110 71 68

a To verify whether the annealing process is critical to the formulation containing mannitol as a main component, different parameters, such as annealing step and second drying temperature, were evaluated with similar pre-lyo concentration of rBCG.

2. Materials and methods

At the end of the secondary drying, the vials were sealed under vacuum.

2.1. Chemicals 2.4. Differential scanning calorimetry (DSC) Mannitol, sucrose, and trehalose were bought from Mallinckrodt Baker, Phillipsburg, NJ; sodium glutamate was from EMD, Gibbstown, NJ; and tyloxapol from Pressure Chemical Co., PA. 2.2. Recombinant BCG strain AERAS-422 rBCG (Ag85A, Ag85B, Rv3407, ureC::pfoAG137Q, panCD) is a live attenuated recombinant BCG vaccine being developed by Aeras. It was modeled on the original Danish 1331 strain of BCG vaccine. Ag85A, Ag85B and Rv3407 are overexpressed by AERAS-422, which further encodes an attenuated perfringolysin, PfoAG137Q from C. perfringens str 13, in the ureC locus of the chromosome to enable class I presentation rBCG antigens. A similar rBCG construction strategy was described by Sun et al. [6]. The genetic and phenotypic stability of this mutation has been confirmed in the Accession Cell Bank (ACB), Master Cell Bank (MCB), and in Clinical Trial Material (CTM). 2.3. Freeze-drying formulations and lyophilization cycles After the rBCG bulk material was diluted 30–40 fold with the formulation to be tested, the mixture was filled in 2 mL glass amber vials (0.5 mL of mixture per vial). The formulation for lyophilization was: 0.22 M mannitol, 7.3 mM sucrose, 0.04 M sodium glutamate, 1.8 mM tyloxapol and 13.2 mM trehalose. The filled and half-stopped vials were loaded into a LyoStar II lyophilizer (SP Industrials, Warminster, PA) at a shelf temperature of 5 ◦ C and frozen to −50 ◦ C by reducing the temperature at a rate of 2 or 5 ◦ C/min. The vial contents were kept frozen at a temperature of −50 ◦ C for 2 h. Then an annealing step (−15 ◦ C) was added, if applicable according to the requirements of the process designs (Table 1), as follows: the shelf temperature was increased back to −15 ◦ C at 0.5 ◦ C/min, and kept at this temperature for 2 h; after annealing, the vials were frozen again to −50 ◦ C at a cooling rate of 0.5 ◦ C/min and were kept at −50 ◦ C for another 2 h. Primary drying was then carried out in two temperature steps. In the first step, the shelf temperature was increased at 0.4 ◦ C/min to −30 ◦ C, then at 0.3 ◦ C/min to −10 ◦ C, followed by incubation for 10 h. The second drying step was conducted at a shelf temperature of 35 ◦ C, 37 ◦ C, 40 ◦ C or 50 ◦ C for 6 h under a chamber vacuum pressure of 200 mTorr. The vials were sealed under 300 ± 50 mbar of nitrogen at the end of the cycle. As a control, rBCG was prepared with 0.08 M sodium glutamate and the freeze drying cycle for conventional BCG vaccine using the same lyophilizer. The filled vials were cooled from room temperature to −20 ◦ C at 1 ◦ C/min. After soaking for 30 min, the vials were slowly cooled down to −35 ◦ C at 0.2 ◦ C/min and kept at this temperature for 3 h. Drying, at a pressure of 75 mTorr and ramping rate of 0.1 ◦ C/min, was carried out at −20 ◦ C for 6 h. Secondary drying followed at 37 ◦ C for 7 h at a pressure of 11 mTorr.

The thermodynamic behavior of the powders was determined on a DSC 823e (METTLER TOLEDO, Switzerland). The cover of the crucible containing the powder sample was punched with a small hole before analysis. The sample (around 3–5 mg powder) was cooled from room temperature to −30 ◦ C at 10 ◦ C/min, held at −30 ◦ C for 5 min, then heated from −30 ◦ C to 160 ◦ C or 220 ◦ C with a scanning rate of 10.0 ◦ C/min. The sample cell was purged with a nitrogen gas at 10.0 mL/min. The glass transition temperature (Tg  ) was recognized on the reversing heat flow curve as an endothermic shift of the baseline and determined as the midpoint of this transition by a STARe SW9.01 software (METTLER). 2.5. CFU assay Each rBCG specimen was reconstituted in 0.5 mL saline, and 0.1 mL samples were withdrawn from each vial and diluted in Middlebrook 7H9 broth (BBL, BD, Sparks, MD) using as series of 1:10 dilutions. Then 0.1 mL of each sample with the appropriate number of replicates was plated onto Middlebrook 7H10 plates (BBL, BD, Sparks, MD). The number of serial dilutions could be varied depending upon the estimated CFU/mL range of the specimen. The planted plates were put in Ziplock Bags (three plates per bag) incubated at 36 ± 2 ◦ C in air for three to four weeks until colonies had developed to a visible size. For each dilution series, a dilution plate containing not less than 30 and not more than 300 colonies was counted with a colony Counter (ProtoCol, Synoptics, UK). 2.6. Appearance of freeze dried rBCG vaccine The appearance of rBCG vaccine was visually inspected after freeze drying. The top surface structures of lyophilized cakes were also evaluated by a microscopically with image at ×400 (Axioskop 40, Carl Zeiss, Germany). 2.7. Stability and accelerated study The stability of freeze dried rBCG vaccine was evaluated at a storage temperature of 4 ◦ C. The accelerated study was executed for 4 weeks at 37 ◦ C. The survival value of each lot was determined by comparing between the CFU values of freeze dried vials being stored at 4 ◦ C and 37 ◦ C for 4 weeks. 3. Results 3.1. Freeze drying cycles and thermodynamic properties The lyophilization cycles using the new improved rBCG formulation are listed in Table 1 (Lots A–F). To verify whether the annealing process is critical to the formulation containing mannitol as a main

4850

T.H. Jin et al. / Vaccine 29 (2011) 4848–4852

Fig. 1. DSC thermograms of freeze dried rBCG samples prepared with the different cycles. The samples were cooled from room temperature to −30 ◦ C at 10◦ C/min, held at −30 ◦ C for 5 min, and heated to 160 ◦ C or 220 ◦ C at 10◦ C/min. X-axis is temperature; Y-axis is heat flow.

component, different parameters (cooling rate, with or without annealing step and second drying temperature) were evaluated with similar pre-lyo concentrations of rBCG. Without the annealing step, two different secondary temperatures (35 and 40 ◦ C) were applied at a 5 ◦ C/min of cooling rate. The increased secondary drying temperature (from 35 to 40 ◦ C) did not result in difference in thermograms (curves A and B, Fig. 1) measured by DSC. A large crystalline peak immediately followed the glass transition (Tg  ) activity. The Tg  value for both cycles was the same at 70 ◦ C (Table 1, Lots A and B). However, when the annealing step was added in the lyophilization cycles, it caused marked differences in thermodynamic profiles (curve C vs B, Fig. 1). At a cooling rate of 5 ◦ C/min, the annealing step could totally remove the crystalline peak at the same secondary drying temperature of 40 ◦ C. When drying temperature was further increased to 50 ◦ C, the crystalline peak was also not evident (curve D), but the Tg  s were increased from 90 ◦ C and 110 ◦ C (Table 1, Lot D). The thermograms of glass transition also became broad and melting temperatures were shifted to higher values (Fig. 1). At a cooling rate of 2 ◦ C/min, the annealing step could totally remove the crystalline peak at the secondary drying temperature of 40 ◦ C (curve C vs E, Fig. 1), but not at 37 ◦ C. There was a small crystallization peak left at the lower drying temperature (curve F, Fig. 1). The glass transition temperatures with an annealing step and secondary drying temperature of 37 ◦ C was similar, around 70 ◦ C, to that when no annealing step was used (Table 1, Lot F vs Lots A and B).

lization of rBCG during storage. The candidate cycle of Lot C with the tested formulation had the highest survival of 61.7% after 4 weeks. Lots C and D met the WHO stability requirement of more than 20% in survival [7]. Lot SG, prepared using the conventional formulation and cycle, had both a low recovery rate of 16.0% and low survival value of 5.0%. 3.3. Appearance of rBCG vaccines The appearance of the freeze dried rBCG vaccine prepared using the mannitol formulation was consistent across the different lyophilization cycles, with elegant cake formation in each case (Fig. 2, left). Microscopic examination of the cake surface showed an irregular crystalline structure when no annealing step was used (Lot A, Fig. 3A), while a regular crystalline structure was evident when an annealing step was incorporated into the process (Lot C, Fig. 3B). The freeze dried rBCG vaccine prepared with the conventional cycle and formulation resulted in product with a shrinking “dust-like” appearance that did not form into a cake (Fig. 2, right). 4. Discussion Although freeze-drying is often a preferred method for improving the shelf life of vaccines unstable in aqueous solution, many

3.2. Freeze drying recovery and stability The viability and recovery of rBCG Lots A to F (corresponding to the lyophilization cycles in Table 1) during the freeze drying process, and stability at storage temperatures of 4 and 37 ◦ C are shown in Table 2. As a control, rBCG prepared using a conventional cycle and formulation is listed as Lot SG. Surprisingly, recovery after freeze drying was generally higher when an annealing step was applied in the drying cycle than when no annealing process was employed (Lots C–F vs Lots A and B). Lot C showed the highest recovery of 53.2%, while Lots A and B only had recovery of 11%. Under the accelerated conditions at 37 ◦ C (Table 2), the survival values were lower in the Lots A, B, E and F (ranging from 5.6% to 13.9%) than the Lots C and D (61.7% and 25.9%, respectively). The lower survivals corresponded to the occurrence of a crystallizing peak in the DSC thermograms (Fig. 1). Lack of re-crystallization activity in Lots C and D corresponded to higher rBCG survival, suggesting that the full crystallization of the mannitol formulation is critical for stabi-

Fig. 2. The cake appearance of rBCG vaccines. The left vial was prepared from the improved formulation and freeze-drying cycle; the right vial was from the conventional formulation and cycle.

T.H. Jin et al. / Vaccine 29 (2011) 4848–4852

4851

Table 2 The recovery and stability of freeze-dried rBCG vaccinesa . Lot nob .

Pre-lyo (CFU/mL)

Post-lyo (CFU/mL)

Recoveryc (%)

4W-4 ◦ C (CFU/mL)

4W-37 ◦ C (CFU/mL)

Survivald (%)

A B C D E F SG

4.50E+7 4.74E+7 2.82E+7 2.26E+7 3.75E+7 1.71E+7 2.16E+7

5.26E+6 5.29E+6 1.50E+7 4.66E+6 6.34E+6 2.87E+6 3.46E+6

11.0 11.0 53.2 20.6 19.4 16.8 16.0

4.23E+6 6.80E+6 6.92E+6 4.25E+6 4.81E+6 3.04E+6 2.31E+6

2.36E+5 4.06E+5 4.27E+6 1.10E+6 6.70E+5 2.15E+5 1.16E+5

5.6 6.0 61.7 25.9 13.9 7.0 5.0

a b c d

The results were recorded as CFU/mL for rBCG. Lots A–F were prepared from the improved formulation and cycles (see Table 1); Lot SG was from the conventional formulation and cycle as a control. The recovery of rBCG vaccines were compared between the CFU values of pre- and post-lyophilization. The survival value of each lot was compared between the CFU values of freeze dried vials being stored at 4 ◦ C and 37 ◦ C for 4 weeks.

organisms (bacterium or virus) lose their activity in the freezing and drying process. The efficiency of the BCG vaccine is now considered to depend on the immunizing properties of the strain and the method of manufacture. Any method which ensures that a vaccine contains a high proportion of live, well-dispersed bacilli, free from soluble antigens and stable on storage, is advantageous [1,8–10]. Sodium glutamate or sodium glutamate diluted in one fourth of Sauton medium is the conventional freeze drying formulation for BCG vaccine. The disadvantage of this formulation is cake appearance and integrity after lyophilization: most frozen cakes will collapse and/or shrink during the drying process. The potential powder losses on the stopper and glass wall of the final container during the transportation may result in variable potency test outcomes. The new formulation of our rBCG vaccine developed in this study achieved improvements compared to the conventional glutamate formulation in the following respects: (i) good cake appearance; (ii) high viability maintained during the freezedrying process; (iii) better stability during the accelerated study at 37 ◦ C, indicating longer shelf life under the refrigeration. Generally, the improved rBCG formulation not only has the advantage of better stability, but also provides easier handling (no cake loss during shipping and storage). In the new formulation, in addition to using mannitol as a bulking agent, we employed more stabilizers such as trehalose, sucrose and sodium glutamate. The bacteria of BCG or rBCG tended to clump when they were suspended in solution. The addition of tyloxapol (Triton WR 1339) could reduce the clumping of rBCG organisms in the lyophilization process, thus enhancing the colony forming units in the following assays (data not shown). Sodium glutamate is a salt form of glutamic acid, which has been shown to stabilize the potency of freeze-dried BCG vaccine, but without maintenance of cake integrity in the final product. Disaccharides are amongst the most frequently used excipients, with trehalose and sucrose being

common selections. Sucrose or trehalose are also often added to the formulation to prevent aggregation during lyophilization [11–13]. Trehalose is a nonreducing sugar, with a high Tg  (∼117 ◦ C) coupled with a low tendency to crystallize, making it an attractive model amorphous compound. Moreover, trehalose can be rendered amorphous by different methods including freeze-drying and spray drying. Sucrose, as a cryoprotectant, was found to reduce or eliminate the perturbation caused by freezing [14–16]. Trehalose and sucrose could both be used as stabilizer and cryoprotectant for freeze dried products, with the addition of glutamate as a stabilizer, these excipients may protect bacteria against freezing and drying stresses by different mechanisms. For example, various sugars thermodynamically stabilize organisms in both liquid and frozen solutions by being preferentially excluded from their vicinity [17–19]. According to the popular accepted “water replacement” theory, the excipients protect from drying stress by substituting themselves for water molecules essential to maintaining the organism structure. Mannitol is isomeric, unbranched hexitols, and is known to crystallize readily from the frozen state, making it a widely used bulking agent in lyophilized formulations. It also has excellent cake-forming qualities. The nonreducing sugar alcohol can be dried under relatively harsh conditions, because of its high eutectic temperature of −1.5 ◦ C [20,21]. It is well known that high product temperature yields a fast process, with each 1 ◦ C increase in product temperature decreasing primary time by about 13%. With mannitol being used as a main component in the formulation, the freeze drying cycle could be completed within 24–26 h. Mannitol exits in several polymorphic forms when different cycles are applied: ␣-mannitol, ␤-mannitol, ␦-mannitol, amorphous form, and mannitol hemihydrate (MHH). In some cases mixtures of the different forms are obtained [21–25]. Among the anhydrous polymorphs, ␤-mannitol is the stable form under

Fig. 3. Images of cake surface structures lyophilized under improved and conventional conditions. The top surface structures of lyophilized cakes were evaluated microscopically (×400). A is from Lot A, without an annealing process in drying cycle; B from Lot C, with an annealing step.

4852

T.H. Jin et al. / Vaccine 29 (2011) 4848–4852

ambient conditions. MHH contains water, stoichiometrically incorporated in the crystal lattice, which may not be completely removed unless secondary drying is conducted at ≥40 ◦ C [26,27]. During product storage, MHH may transform to anhydrous mannitol with the release of lattice water. The water will then be available for interaction with the other formulation ingredients. This may lead to loss of live organism activity. Batch-to-batch variations may also result from variable MHH content in the lyophile. In this study, it was found that adding only the high secondary drying temperature (40 ◦ C) did not result in different thermograms compared with drying at 37 ◦ C measured by DSC. The crystalline peak suggested that mannitol did not totally crystallize during the freezing phase. The low survival rate during the accelerated study suggested that MHH formed in the freeze drying cycles. The further crystallizing of mannitol under the release of MHH during the storage would cause the loss of live organism activity and reduce the shelf life of freeze dried rBCG products. However, after the annealing step was applied, variable secondary temperatures caused marked differences in thermodynamic profiles irrespective of different test cooling rates. Under a cooling rate of 2 ◦ C/min, the annealing step could totally remove the crystalline peak at a secondary drying temperature of 40 ◦ C, although it remained a small peak when 37 ◦ C was tested, indicating partial crystallization during the freeze drying process. At a drying temperature of 40 or 50 ◦ C, annealing caused the disappearance of the crystalline peak in both thermal profiles. The high survival rate in the later accelerated study suggested that adding the annealing step plus higher temperature in the secondary drying phase resulted in total crystallization of mannitol and eliminated MHH in the final vials. The crystallization of mannitol could be easily observed under the microscope. In addition, the observed enthalpy of crystallization was more dependent on the annealing step than on the cooling rate. The absence of an annealing step could cause a much larger fraction of the mannitol along with the associated unfrozen water to crystallize from solution, thus potentially depressing the Tg  and compromising storage stability. The presence of an annealing step facilitated both mannitol nucleation and crystal growth, even in the presence of rBCG bacteria and others excipients. DSC is very useful tool in formulation and cycle development of freeze dried products. In unannealed solutions, the onset of crystallization was delayed so that the crystallization exotherm overlapped with the huge eutectic melting endotherm, thus making accurate measurement of crystallization enthalpy difficult. In annealed samples, annealing led to nucleation, resulting in crystallization at a lower temperature. This separated the crystallization exotherm from the eutectic melting endotherm, thereby enabling the accurate measurement of enthalpy. This transformation enables DSC to be utilized as a more sensitive technique for the detection of crystallizing mannitol. In conclusion, the developed formulation and freeze drying cycle for rBCG vaccine not only resulted in elegant cake, but extended its shelf life, potentially allowing transportation and used without refrigeration for short periods of time. This research provides a basis for further clinical trials of rBCG vaccine. Acknowledgements We thank Barbara Shepherd for critical review of the manuscript. This work was supported by the Bill & Melinda Gates Foundation.

References [1] Kamath AT, Fruth U, Brennan MJ, Dobbelaer R, Hubrechts P, Ho MM, et al. New live mycobacterial vaccines: the Geneva consensus on essential steps towards clinical development. Vaccine 2005;23(May (29)):3753–61. [2] World Health Organization (WHO). Global tuberculosis control—epidemiology, strategy, financing. WHO/HTM/TB/2009.411. [3] WHO. WHO vaccine preventable diseases: monitoring system—2000 global summary. WHO/V&B/00.32. [4] Yamamoto S, Yamamoto T. Historical review of BCG vaccine in Japan. Jpn J Infect Dis 2007;60(November (6)):331–6. [5] Corbel MJ. Reasons for instability of bacterial vaccines. Dev Biol Stand 1996;87:113–24. [6] Sun R, Skeiky YA, Izzo A, Dheenadhayalan V, Imam Z, Penn E, et al. Novel recombinant BCG expressing perfringolysin O and the over-expression of key immunodominant antigens; pre-clinical characterization, safety and protection against challenge with Mycobacterium tuberculosis. Vaccine 2009;27(July (33)):4412–23. [7] WHO. Requirements for biological substances. 1988;11:60–92. [8] Gheorghiu M, Lagrange PH, Fillastre C. The stability and immunogenicity of a dispersed-grown freeze-dried Pasteur BCG vaccine. J Biol Stand 1988;16(January (1)):15–26. [9] Gheorghiu M, Lagrange PH. Viability, heat stability and immunogenicity of four BCG vaccines prepared from four different BCG strains. Ann Immunol 1983;134(January–February (1)):125–47. [10] Nascimento IP, Leite LCC. The effect of passaging in liquid media and storage on Mycobacterium bovis—BCG growth capacity and infectivity. FEMS Microbiol Lett 2005;243(February (1)):81–6. [11] Quaak SG, Haanen JB, Beijnen JH, Nuijen B. Naked plasmid DNA formulation: effect of different disaccharides on stability after lyophilisation. AAPS PharmSciTech 2010;11(March (1)):344–50. [12] Kawai K, Hagiwara T, Takai R, Suzuki T. Comparative investigation by two analytical approaches of enthalpy relaxation for glassy glucose, sucrose, maltose and trehalose. Pharm Res 2005;22(March (3)):490–5. [13] Surana R, Pyne A, Suryanarayanan R. Effect of preparation method on physical properties of amorphous trehalose. Pharm Res 2004;21(July (7)):1167–76. [14] Rodrigues JP, Paraguassú-Braga FH, Carvalho L, Abdelhay E, Bouzas LF, Porto LC. Evaluation of trehalose and sucrose as cryoprotectants for hemtopoietic stem cells of umbilical cord blood. Cryobiology 2008;56(April (2)):144–51. [15] Yoshioka S, Miyazaki T, Aso Y, Kawanishi T. Significance of local mobility in aggregation of beta-galactosidase lyophilized with trehalose, sucrose or stachyose. Pharm Res 2007;24(September (9)):1660–7. [16] Breen ED, Curley JG, Overcashier DE, Hsu CC, Shire SJ. Effect of Moisture on the Stability of a lyophilized humanized monoclonal antibody formulation. Pharm Res 2001;18(September (9)):1345–53. [17] Arakawa T, Prestrelski SJ, Kenney WC, Carpenter JF. Factors affecting shortterm and long-term stabilities of proteins. Adv Drug Deliv Rev 2001;46(March (1–3)):307–26. [18] Tsvetkov TD, Tsonev LI, Tsvetkova NM, Koynova RD, Tenchov BG. Effect of trehalose on the phase properties of hydrated and lyophilized dipalmitoylphosphatidylcholine multilayers. Cryobiology 1989;26(April (2)):162–9. [19] Costantino HR, Carrasquillo KG, Cordero RA, Mumenthaler M, Hsu CC, Griebenow K. Effect of excipients of the stability and structure of lyophilized recombinant human growth hormone. J Pharm Sci 1998;87(November (11)):1412–20. [20] Hawe A, Friess W. Physicochemical characterization of the freezing behavior of mannitol–human serum albumin formulations. AAPS PharmSciTech 2006;7(December (4)):94. [21] Kim AI, Akers MJ, Nail SL. The physical state of mannitol after freeze-drying: effects of mannitol concentration, freezing rate, and a noncrystallizing cosolute. J Pharm Sci 1998;87(August (8)):931–5. [22] Liao X, Krishnamurthy R, Suryanarayanan R. Influence of processing conditions on the physical state of mannitol—implications in freeze-drying. Pharm Res 2007;24(February (2)):370–6. [23] Liao X, Krishnamurthy R, Suryanarayanan R. Influence of the active pharmaceutical ingredient concentration on the physical state of mannitol—implications in freeze-drying. Pharm Res 2005;22(November (11)):1978–85. [24] Pyne A, Surana R, Suryanarayanan R. Crystallization of mannitol below Tg  during freeze-drying in binary and ternary aqueous systems. Pharm Res 2002;19(June (6)):901–8. [25] Yu L, Milton N, Groleau EG, Mishra DS, Vansickle RE. Existence of a mannitol hydrate during freeze-drying and practical implications. J Pharm Sci 1999;88(February (2)):196–8. [26] Johnson RE, Kirchhoff CF, Gaud HT. Mannitol–sucrose mixtures—versatile formulations for protein lyophilization. J Pharm Sci 2002;91(April (4)):914–22. [27] Cavatur RK, Vemuri NM, Pyne A, Chrzan Z, Toledo-Velasquez D, Suryanarayanan R. Crystallization behavior of mannitol in frozen aqueous solutions. Pharm Res 2002;19(June (6)):894–900.