Characterization of the freeze sensitivity of a hepatitis B vaccine

Research Paper

IB UT

Characterization of the freeze sensitivity of a hepatitis B vaccine

E.

[Human Vaccines 5:1, 26-32; January 2009]; ©2009 Landes Bioscience

1PATH;

DI ST R

Dexiang Chen,1 Anil Tyagi,2 John Carpenter,2 Shalimar Perkins,1 David Sylvester,1 Mark Guy,1 Debra D. Kristensen1 and LaToya Jones Braun2,* Seattle, Washington USA; 2University of Colorado Denver; School of Pharmacy; Department of Pharmaceutical Science; Aurora, Colorado USA

Key words: vaccine, freeze exposure, cold chain, hepatitis B, aluminum salt adjuvant, freeze-thaw damage

.D

O

NO

T

chain temperature in the specified range is a very challenging task, and temperature excursions, particularly at the periphery of the cold chain, are quite common. Several cold chain monitoring studies have reported that vaccines were inadvertently exposed to freezing temperatures at alarming frequencies.2-11 In some cases, one vaccine shipment was exposed to several episodes of freezing temperatures during cold chain storage and transport.2,4-7 Accidental freeze exposure of vaccines is not limited to resource-constrained settings; a review showed that similar problems are frequently found in developed countries where temperatures are supposedly closely monitored.12 It should be noted that temperature excursions happen and are not usually detected in spite of tools available in cold chain systems. Such tools include the shake test, which is not commonly used due to the inconvenience and difficulty in interpreting the results, and freeze indicators such as Freeze WatchTM which are provided with batches of vaccine and do not allow tracking of individual vials once the vaccine reaches the periphery of the cold chain.13,14 For these reasons, a portion of the vaccines used in humans might have been exposed to freezing conditions as described in cold chain monitoring studies which highlight the risk of vaccine freezing. It should be pointed out that there is no established epidemiological evidence of vaccine failure attributable to freezing although some linkages have been speculated in a few cases.15,16 It is well accepted that vaccines are sometimes frozen inside the cold chain and those affected are impaired as shown by laboratory and in vivo evidence.17-23 However, there is a disagreement among the experts on what portion of the freeze exposures detected in the cold chain monitoring studies have actually resulted in physical freezing. A close examination of the temperature monitoring data revealed that the studies have usually used cut-off temperatures of 0°C or -0.5°C. In more recent studies, where temperatures were monitored with recording devices, deeper freezing temperatures down to -8°C or -10°C were seen.4-6 The durations of exposure to freezing temperatures can vary from days to weeks, and each shipment may experience more than one episode of exposure at different segments of the cold chain.2,4-7 It is important to understand the impact of the temperature excursions under these conditions on the quality and performance of the vaccine. Collecting the field evidence of vaccine freezing and performing comprehensive analysis of vaccines that have been transited through the cold chain is one way to understand the risk of vaccine freezing and its potential impact. However, these studies are very difficult to do logistically. By investigating the freeze kinetics

BI

OS

CI

EN

CE

Recent studies have revealed that vaccines containing aluminum adjuvant are exposed to sub-zero temperatures while in the cold chain more frequently than was previously believed. This raises concerns that these freeze-sensitive vaccines may be damaged and offer inadequate protection. This study was undertaken to characterize the immediate qualitative changes of one such vaccine, hepatitis B, caused by freeze exposure. Hepatitis B vaccine was subjected to freezing temperatures ranging from 0°C to -20°C for up to three episodes with durations ranging from 1 hour to 7 days. The vaccine was analyzed for freezing point, particle size distribution, tertiary structure, and in vitro and in vivo potency. Whether or not hepatitis B vaccine freezes was shown to be dependent on an array of factors including temperature, rate of temperature change, duration of exposure, supercooling effects and vibration. Vaccine exposed to “mild” freezing (-4°C or warmer) temperatures did not freeze and remained qualitatively unaltered. Single or repeated freezing events at temperatures of -10°C or lower were associated with aggregation of the adjuvant-antigen particles, structural damage of the antigen, and reduction of immunogenicity in mice. Damage to the vaccine increased with duration of freezing, lower temperature, and the number of freezing episodes. With vibration, vaccine froze at -6°C after 1 hour and damage occurred. Freezing and freeze damage to vaccines containing aluminum salt adjuvant represent real risks to the effectiveness of immunization and should be prevented by strengthening the cold chain system or, alternatively, development of freeze-stable vaccine formulations.

S

Introduction

20 08

LA

ND E

Many currently used vaccines contain adjuvants based on aluminum salts in order to enhance their immunogenicity. However, vaccines containing aluminum salts are known to be sensitive to damage caused by freezing. As a result, World Health Organization (WHO) guidelines and product labels stipulate that such vaccines should be stored and transported at temperatures between 2°C and 8°C.1 Furthermore, it is advised that vaccines suspected of having been frozen should be discarded. In practice, maintaining the cold *Correspondence to: LaToya Jones Braun; University of Colorado Denver; School of Pharmacy C238-P15; P.O. Box 6511; Aurora, Colorado 80045 USA; Email: Latoya. [email protected] Submitted: 05/06/08; Revised: 06/18/08; Accepted: 06/24/08

©

Previously published online as a Human Vaccines E-publication: http://www.landesbioscience.com/journals/vaccines/article/6494

26

Human Vaccines

2009; Vol. 5 Issue 1

DI ST R

IB UT

E.

Freeze sensitivity of vaccines

T

NO

CE

Results

Single freezing event. Greater than 99% of the particles in the standard vaccine stored at +4°C had a diameter less than 3 μm, with very few large particles (Fig. 2). Following a single freezing event at -12°C, there was a small but statistically significant decrease in the percentage of particle in the 1.5 to 3.0 μm range. Vaccine that was frozen at -14°C or -20°C for 24 hours had a much larger decrease in the same particle size range and statistically significant increase in the populations of larger particles. Samples which were exposed to -10°C for 24 hours exhibited a measurable particle size shift, but it did not reach statistical significance. Samples exposed to freezing temperatures (0°C to -8°C) for 24 hours (without visually discernable freezing) had a particle size distribution indistinguishable from the +4°C control. Therefore, aggregation to form larger particles correlated with whether or not physical freezing had occurred. The degree of particle aggregation was also dependent upon the actual freezing conditions involved. Repeated freeze-thaw events. Vaccine was exposed to -6°C, -10°C, -14°C and -20°C for 24 hours followed by thawing at +4°C for 24 hours. Samples receiving 1, 2 or 3 cycles of the above temperature treatments were analyzed to determine the distribution of particles within various size ranges (Fig. 3A–C). With the exception of samples exposed to -6°C, which did not freeze and did not exhibit a particle size shift, all other treatments resulted in physical freezing of the vaccines and shifts of particle sizes towards the larger aggregates. The degree of shift correlated with both the freezing temperature and the number of temperature treatments. The shift was greatest for the two lowest temperatures. For a given freezing temperature, each additional cycle of treatment resulted in increased number of large-particle aggregates. Duration of freezing at -10°C. Hepatitis B vaccine was exposed to -10°C without disturbance for up to 174 hours. Three vials of vaccine were analyzed at each time point as shown in Figure 4, and the particle size distribution was determined using a Coulter counter. There was no apparent reduction of particles of less than 3 μm when the samples were examined at 5, 6 and 10 hours after the exposure, although vaccine was physically frozen after 10 hours. There was

O

and freeze sensitivity of hepatitis B vaccine in a controlled laboratory environment, we hoped to fill our knowledge gap regarding the risk of vaccine freezing.

.D

Figure 1. Freezing kinetics of hepatitis B vaccine. Freezing status of the vaccine subjected to various treatments in an environmental chamber as shown in the figure was assessed by visual examination. Agitation involved rotation of vaccine vials at a frequency of 1 to 2 rounds per second.

Figure 2. Effect of freezing temperatures on the particle size distribution of hepatitis B vaccine. Vaccine was exposed to specific temperatures as shown in the figure for 24 hours. Data represents the percentage of particles in various size fractions as measured using the Z1 Coulter counter. Particle size greater than 3 μm indicates aggregation and freeze damage. Vaccine was only physically frozen at -10°C and lower temperatures. *indicates p < 0.05, ✓ indicates p < 0.01. Error bars represent SEM.

©

20 08

LA

ND E

S

BI

OS

CI

EN

Freezing point and freezing kinetics of hepatitis B vaccine. The freezing point (the onset melting temperature) of three brands of hepatitis B vaccine ranged from -2.5°C to -2.8°C as determined by the DSC assay. Shanvac vaccine (from Shantha Biotech, India) with a freezing point of -2.8°C was used in the following studies unless stated otherwise. Vaccine in triplicate vials was exposed to temperatures ranging from 0°C to -14°C in 2°C increments without agitation in the environmental chamber. Visual examination of freezing was made 24 hours later by tilting the vials and observing the ability of the contents to flow (Fig. 1). There was no visible freezing at -6°C, partial freezing at -8°C (not shown), but complete freezing at -10°C or lower temperatures. At -10°C, it took 1 hour for the vaccine to partially freeze and 6 hours to completely freeze. Agitation of the vials, intended to represent the motion during transportation, was found to accelerate vaccine freezing. A single gentle tap of the vials exposed to -10°C for 1 hour changed the vaccine from a liquid to a slushy state. Continuous shaking of vaccine vials at 1 to 2 revolutions per second for 30 minutes led to partial freezing at -10°C. At -6°C with agitation, the vaccine was partially frozen after 1 hour and completely frozen after 3 hours, in contrast to the non-agitated vials, where the vaccine did not freeze even after 72 hours of exposure. At -4°C the hepatitis B vaccine did not freeze, even with continuous agitation (Fig. 1). Particle size distribution of hepatitis B vaccine. Greater than 98% of HBsAg in an undamaged vaccine is associated with aluminum adjuvant particles, which is essential for the immunogenicity and efficacy of the vaccine. Therefore it is important to understand the effect of freezing on the integrity of the adjuvant and the antigenadjuvant complex.

www.landesbioscience.com

Human Vaccines

27


S

BI

OS

CI

EN

CE

.D

O

NO

T

DI ST R

IB UT

E.

had the largest effect on the zeta potential. Relative to the control, these samples experienced a significant increase in zeta potential towards zero (Fig. 5). Exposing the vaccine to -10°C for up to 10 hours resulted in particles with zeta potentials similar to the control. With increasing time at -10°C, there was a small trend towards less negative zeta potential values; however, the shift was not statistically significant. Thus, the -20°C treatment for 168 hours was significantly more damaging than the -10°C treatment for up to 174 hours. In vitro antigen potency. The in vivo efficacy of hepatitis B vaccine is dependent upon the correct conformation of the HBsAg, which can be measured in the in vitro potency assay (a conformationdependent ELISA). However, in this study the effects of freezing on the protein as measured by this assay were found to be unpredictable, with the measured antigen potency remaining unchanged, increasing or decreasing after treatment. Repeated freezing and thawing at -20°C (and lower temperatures) invariably caused a significant loss of potency for all three brands of hepatitis B vaccines tested (Fig. 6A). A single freezing event of 24 hours at different temperatures (data not shown) or at -10°C for different durations (Fig. 6B) usually did not result in a loss of in vitro potency. In contrast, the results obtained from samples that had been exposed to -10°C for 10 or more hours were quite variable. In theses cases, the apparent vaccine potency appeared to be equivalent to or greater than the +4°C control. The reasons for this are not clear, but it is possible that these severe freezing conditions had an impact on the conformation of the HBsAg, opening or breaking up particles, thereby making the antigen more accessible to the monoclonal antibodies used to detect HBsAg in the assay. Up to three cycles of freezing and thawing at temperature ranging from -6°C to -14°C did not cause any change or increase in potency (data not shown). In all experiments, exposure of hepatitis B vaccine to sub-zero temperatures that did not result in freezing did not have any effect on the in vitro potency as measured using the AUZYME assay. Under all the freezing conditions tested, the HBsAg remained adsorbed to the adjuvant based on the analysis of the supernatant of vaccine upon removal of the aluminum particles (data not shown). Changes in antigen tertiary structure. The change of the tertiary structure of HBsAg in response to freezing was measured using the fluorescence assay (Fig. 7). There was no change in tertiary structure if the exposure (e.g., -6°C) did not result in freezing. There was also no obvious shift if the freezing temperatures were -10°C or warmer, even if the vaccine was physically frozen up to three times. Upon freezing at -14°C or -20°C, the fluorescence peak of the HBsAg shifted from approximately 334 nm to 336 nm or higher, suggesting that the protein structure had become slightly more exposed to the solvent. The fluorescence peak shift was detected following only one freezing event at temperatures of -14°C or lower. Repeated freezing did not cause any further shift. Immunogenicity in mice. The immunogenicity of hepatitis B vaccine that had been subjected to three 20-hour episodes of freezing at -20°C was compared to that of the untreated control. Immunization with 0.4 to 2.0 μg per dose of control (non-frozen vaccine) resulted in a dose-dependent antibody response. The frozen vaccine at 1 or 2 μg per dose elicited approximately five-fold lower antibody titers than the control vaccine at a comparable dose. The frozen vaccine at 0.4 μg was not immunogenic in this assay (Fig. 8A). Based on the results from the first study, a 1-μg dose was used in a second study to determine the immunogenicity of the vaccines

LA

ND E

Figure 3. Effect of repeated temperature treatments on particle size distribution. Each cycle involved 24-hour exposure to specific temperatures as shown in the figure followed by thawing at +4°C for 24 hours. Data represents percent of particles with size ranges of: (A) 1.5 to 3 μm, (B) 3 to 6 μm and (C) 6 to 30 μm. *indicates p < 0.05, ✓ indicates p < 0.01, + indicates p < 0.001. Error bars represent SEM.

©

20 08

small but not statistically significant loss of particles under 3 μm after 30 hours. By 54 hours, there was a significant increase in the percentage of particles above 3 μm. The shift in the distribution of particles to the larger size ranges increased with prolonged freezing up to 174 hours (Fig. 4). Freezing at -20°C for 168 hours caused more severe particle aggregation than freezing at -10°C. Changes in colloidal electric potential. The changes in the zeta potential as functions of temperature and duration of time exposed to -10°C followed the same pattern as the above-mentioned particle size data. Freezing and storing the vaccine at -20°C for 168 hours 28

Human Vaccines



DI ST R

IB UT

E.

treated under different freezing conditions. Hepatitis B vaccine was exposed to -6°C, -10°C and -20°C for 6 hours or three episodes of 20 hours each followed with a 4-hour thawing at +20°C. The treated vaccine samples were then used to immunize mice. The resulting antibody titers are shown in Figure 8B. Vaccines that were exposed to -6°C for up to three 20-hour episodes showed no reduction in immunogenicity, whereas three 20-hour exposures to -10°C resulted in a 2-fold reduction in average antibody titer compared to the untreated control. Vaccine that was exposed to -10°C for 6 hours (one treatment) had no change in immunogenicity, whereas exposure to -20°C for 6 hours led a 2.6-fold antibody reduction and three 20-hour exposures resulted in a 7.2-fold reduction in immunogenicity.

NO O .D

CE

Hepatitis B vaccine, like any other vaccine containing an aluminum-salt adjuvant, is freeze sensitive. Therefore, it is not surprising that the experiments in this study demonstrated that freezing hepatitis B vaccine alters the physical properties of the adjuvant and antigen, resulting in a reduction in immunogenicity in vivo. However, the data reported here also demonstrate that vaccine freezing and freeze damage is a complex process dependent on many variables including formulation of the vaccine, the temperatures involved, duration of exposure, and whether vibration is involved. The freezing point of hepatitis B vaccine used in this study is approximately -2.8°C. This is lower than the commonly referred freezing temperature of water (0°C) or previous reported findings and is due to the presence of sodium chloride, buffer salt, antigen, adjuvant and preservative. However, it is interesting to note that hepatitis B vaccine did not readily freeze at temperatures as low as -4°C in this study. This is partly due to the presence of the excipients in the formulation and also due to the phenomenon of supercooling, which retards the freezing process. In this study, we were unable to freeze the hepatitis B vaccine at -6°C even after up to 72 hours of exposure if the vaccine was undisturbed. The super-cooling effect disappeared if the vaccine was shaken during exposure to low temperatures. Agitation would be expected during transportation of the vaccine, making vaccine in the field more susceptible to freezing and freeze damage. Throughout the study there were no detectable changes of the antigen and adjuvant if the temperature exposure did not result in physical freezing of the vaccine. The adjuvant and antigen in the hepatitis B vaccine were found to have differential freeze sensitivities. Aluminum adjuvant is more sensitive to freeze damage than HBsAg. This was manifested as particle aggregation of the aluminum adjuvant, detectable by sedimentation assay (data not shown), and more accurately by using the particle sizing assay. Particle aggregation worsened with increased number of freezing cycles. In contrast, the amount of HBsAg as measured using the in vitro potency assay was unaltered if the vaccine had been only partially frozen or even seemingly increased following a single freezing event (which can be explained by increased accessibility of antigenic sites of the partially denatured antigen particles). Only after repeated freezing was there a significant reduction in the in vitro potency of the HBsAg. It appears that the shake test, which is currently recommended for identifying freeze-damaged vaccines, is an appropriate tool because it reflects the degree of adjuvant-particle aggregation. The particle aggregation of adjuvant may be related to the reduction of surface charges as measured in the zeta potential assay.

T

Discussion

©

20 08

LA

ND E

S

BI

OS

CI

EN

Figure 4. Effect of freezing duration on particle size distribution. Vaccine was exposed to -10°C for up to 174 hours. Physical freezing happened at 6 hours after exposure. Data represents percent of particles with size ranges of: (A) 1.5 to 3 μm, (B) 3 to 6 μm and (C) 6 to 30 μm. Vaccines kept at +4°C or -20°C for 168 hours were used as controls. *indicates p < 0.05, ✓ indicates p < 0.01. Error bars represent SEM.

www.landesbioscience.com

Figure 5. Effect of freezing duration on the surface charges of adjuvantantigen particles. The data represents zeta potential measured on the same samples as described in the Figure 4 legend. ✓ indicates p < 0.01. Error bars represent SEM.

For colloids, the more non-zero (either positively or negatively) the charge, the more stable the particles. A single freezing event of -20°C decreased the colloidal stability of the vaccine particles as indicated by zeta potential values and even at the higher freezing temperature of -10°C there was a trend in the zeta potential value towards zero

Human Vaccines

29

DI ST R

IB UT

E.


NO

T

Figure 6. Effect of freezing on the in vitro potency of hepatitis B vaccine. (A) Three brands of hepatitis B vaccine were exposed to three cycles of 24-hour freezing at -20°C, each followed by 24-hour thawing at +4°C. Data represents potency of frozen vaccine compared with the untreated control measured using the AUZYME assay. (B) The data represents in vitro potency on the same samples as described in the Figure 4 legend.

EN

CE

.D

O

shift reflecting a change in the polarity of the environment of the tryptophan residues might also result from the agglomeration of the adjuvant or rearrangement of the phospholipids in the HBsAg particle with or without an accompanying substantial change in the structure of the HBsAg protein. Regardless of which of the explanations is correct, the shift in emission peak reflected a change in the structure of the vaccine that altered the vaccine’s immunogenicity. The in vivo immunogenicity of the vaccine provides an important indication of the vaccine’s integrity. Hepatitis B vaccine immunogenicity was reduced following 6 hours of exposure to -20°C and worsened by repeated freezing at -20°C. However, even after repeated freezing, the antigen remained adsorbed to the surface of the adjuvant particles. Indeed, the vaccine’s immunogenicity was only slightly compromised by a single freezing event, and not completely abrogated even by three freeze-thaw cycles. It should be noted that the assay used detected anti-HBsAg binding antibodies. The relevance to protection is unknown. There is little doubt that lower freezing temperatures (-6°C or colder) are detrimental to hepatitis B vaccine. However, it would be premature to conclude that mild temperature excursions (-4°C or warmer) pose no risks to the vaccine. Although these mild freezing conditions did not have an immediate impact on the quality of the adjuvant or antigen in the hepatitis B vaccine under this investigation, primarily due to super-cooling, we do not know whether exposure to these temperatures has a delayed impact on the quality of the vaccine and subsequent shelf stability. It is crucial to keep in mind that the work reported here primarily examined freeze damage to one brand of hepatitis B vaccine only. Hepatitis B vaccine made by different manufacturers may respond differently to freeze exposure.20 Several gaps still exist in our understanding of the freezing process and its impact on the quality of the vaccine. For these reasons, it is important to follow the established cold chain guidelines for vaccine handling;13,28 this includes the use of temperature-monitoring devices, inspection of vaccines using the “shake test,” discarding vaccines that are frozen or suspected of having been exposed to freezing temperatures, and re-immunization of children who received compromised product. Eliminating the risk of freeze damage to vaccines in the cold chain should be pursued as an ultimate goal.

OS

CI

Figure 7. Effect of freezing on the tertiary structure of the HBsAg. The data represents the peak fluorescence positions of the HBsAg from the samples described in the legend of Figure 3. ✓ indicates p < 0.01, + indicates p < 0.001. Error bars represent SEM.

©

20 08

LA

ND E

S

BI

as the duration in the frozen state was prolonged. It should be noted that aluminum hydroxide adjuvant is usually positively charged; the hepatitis B vaccine in this study had a moderately negative zeta potential probably due to the presence of phosphate buffer in the vaccine. With aluminum’s stronger affinity for phosphate anions than hydroxyl anions,24,25 anion exchange of adjuvant hydroxyl with the phosphate buffer anions26 and the phospholipids of the adsorbed HBsAg27 may have taken place. The fluorescence assay, a more sensitive method for detecting changes to the tertiary structure of the HBsAg, also indicated that freeze damage to the protein occurred only when the vaccine was physically frozen. There was an appreciable change in the fluorescence only when the vaccine was frozen at temperatures of -14°C or lower. Caution must be taken in the interpretation of the fluorescence assay results. The fluorescence emission peak shift observed when the vaccine was frozen at temperatures at or lower than -14°C was significant but small, given that tryptophan residues that are completely exposed have emission peak positions at about 350 nm. The conventional interpretation of emission peak shifts in proteins is that they indicate non-native tertiary structure. However the 30

Human Vaccines


DI ST R

IB UT

E.


O

CE

Materials and Methods

In vitro potency assay. Potency of the vaccine powders was assessed using an immune assay (AUZYME) according to the manufacturer’s (Abbott Laboratories, IL) instruction. The concentration of the HBsAg in the test samples was derived based on a standard curve generated using known concentrations of HBsAg. The antigen-adjuvant association was monitored by measuring the free HBsAg after pelleting the aluminum adjuvant at 1,500 g for 10 minutes using a microcentrifuge. The supernatant was collected, further diluted and assayed using the AUZYME assay as described above. The concentration of HBsAg in the supernatant, which is expressed as the percentage of the whole vaccine determined using aliquot collected before centrifugation, represents the free HBsAg disassociated from the aluminum adjuvant. Fluorescence spectroscopy. To measure conformation changes of HBsAg, vaccine samples (1.5–1.8 ml) were allowed to settle overnight at 4°C in triangular fluorescence cuvettes. Samples were then excited with a wavelength of 280 nm and emission spectra were obtained to monitor the fluorescence of the antigen’s tyrosine and tryptophan residues. All spectra were collected at 25°C and at 1 nm intervals using a QM-4 fluorometer (Photon Technology International, Inc., Birmingham, NJ) with a Peltier temperature-controlled cuvette holder (Quantum Northwest, Shoreline, WA). Emission peak positions were determined by importing the data into Origin 7 software (OriginLab Corp., Northampton, MA), smoothing the spectra using Savitsky-Golay (7 Pt) function and taking the first derivative of the smoothed spectra. The effect of temperature and number of freeze-thaw cycles on the fluorescence emission peak position was compared to the control values using a two-way ANOVA with Bonferroni posttests using Prism5 software (GraphPad Software Inc., La Jolla, CA). Zeta potential. To detect changes in the characteristics of the average surface charge of the particles, the electric potentials of the vaccine particles were measured using a zeta potential analyzer (NICOMP 380/ZLS, Particle Sizing Systems, Santa Barbara, CA). Prior to analysis, each sample was diluted with 1 M KCl. This was transferred to a disposable cuvette and immediately placed in the sample compartment for data collection. Data was collected at 25°C and analyzed using the NICOMP zeta potential software.

.D

It can be achieved by ensuring that vaccines containing aluminum adjuvant are not exposed to freezing temperatures while in the cold chain. Alternatively, developing vaccine formulations that can tolerate sub-zero temperature excursions, and in particular prevent freeze damage, would circumvent many of the problems and uncertainties associated with vaccine freezing that are highlighted in this paper.

NO

T

Figure 8. Effect of freezing on the immuogenicity of hepatitis B vaccine in mice. (A) Data represents the geometric mean antibody titers (ELISA unit) of eight mice immunized with various doses of freeze-thawed vaccine or control vaccine. Freeze-thawed vaccine received three 20-hour exposures at -20°C followed with a 4-hour thawing at +24°C. (B) Data represents the geometric mean antibody titers (ELISA unit) of eight mice immunized with 1 μg of vaccine subjected to specific temperature and cycles of treatment as indicated in the figure.

©

20 08

LA

ND E

S

BI

OS

CI

EN

Vaccines. Hepatitis B vaccine purchased from Shantha Biotech, India, was used for the majority of the experiments. The hepatitis B vaccine was a recombinant yeast product. Each milliliter of vaccine contained 20 μg of hepatitis B surface antigen (HBsAg) adsorbed to approximately 0.5 mg of aluminum hydroxide adjuvant. Freeze exposure. An environmental chamber (LH1.5, Associated Environmental Systems, Ayer, MA) was used to conduct freeze exposure studies with a temperature fluctuation range of ±0.5°C. Actual temperature was confirmed by placing a thermometer inside the chamber. Vaccine vials were not treated with agitation in most of the freeze-exposure studies. In some experiments, vaccine vials were mounted on an agitation shaker (VWR) set at maximal speed of 1 to 2 rounds per second to introduce vibration during freezing. Freezing was monitored initially by holding the vial still and then gently tilting to observe the flow of the content. The status of freezing was scored as “not frozen” (no visible ice and contents flowed freely as water), “partially frozen” (visible ice crystals but the contents still flowed), or “completely frozen” (contents of the vial did not flow). Differential scanning calorimetry (DSC) assay. The freezing point of each sample was determined using a differential scanning calorimeter (Diamond DSC, PerkinElmer, Waltham, MA). Indium was used as the calibration standard. Each sample was prepared by placing 20 μl of sample in an aluminum calorimeter pan. The capped pan was then placed in the chamber and subjected to a thermal cycle. The thermal cycle consisted of the following: (1) heat to 50°C at 10°C/min; (2) hold at 50°C for 1 minute; (3) cool to -40°C at 10°C/min; (4) heat to 50°C at 10°C/min; (5) hold for 1 minute at 50°C; and (6) cool to -40°C at 10°C/min. The melting/ freezing point is determined during the heating phase from -40°C to 50°C. www.landesbioscience.com

Human Vaccines

31


.D

O

NO

T

DI ST R

IB UT

E.

8. Pai HH, Ko YC. Vaccine storage practices in primary care physicians’ offices in Taiwan. Kaohsiung J Med Sci 1999; 15:274-9. 9. Reimer RF, Lewis PR. Vaccine storage in pharmacies on the Central Coast of New South Wales. Aust N Z J Public Health 1998; 22:274-5. 10. Guthridge SL, Miller NC. Cold chain in a hot climate. Aust N Z J Public Health 1996; 20:657-60. 11. Finnegan P, Howell F. Storage and handling of vaccines by family doctors. Ir Med J 1996; 89:64-8. 12. Matthias DM, Robertson J, Garrison MM, Newland S, Nelson C. Freezing temperatures in the vaccine cold chain: a systematic literature review. Vaccine 2007; 25:3980-6. 13. WHO. Guidelines on the international packaging and shipping of vaccines. Geneva: World Health Organization 2005. 14. WHO. PQS Performnce Specification: Irreversible freeze indicator. Geneva: WHO 2006. 15. Chen RT, Hardy IR, Rhodes PH, Tyshchenko DK, Moiseeva AV, Marievsky VF. Ukraine, 1992: first assessment of diphtheria vaccine effectiveness during the recent resurgence of diphtheria in the Former Soviet Union. J Infect Dis 2000; 181:178-83. 16. Edstam JS, Dulmaa N, Nymadawa P, Rinchin A, Khulan J, Kimball AM. Comparison of hepatitis B vaccine coverage and effectiveness among urban and rural Mongolian 2-yearolds. Prev Med 2002; 34:207-14. 17. Boros CA, Hanlon M, Gold MS, Roberton DM. Storage at -3 degrees C for 24 h alters the immunogenicity of pertussis vaccines. Vaccine 2001; 19:3537-42. 18. McLean AA, Shaw R Jr. Hepatitis B vaccine. Ann Intern Med 1982; 97:451. 19. WHO. The effect of freezing on the appearance, potency and toxicity of adsorbed and unadsorbed DPT vaccines. Weekly Epidemiological Record 1980; 55:385-92. 20. WHO. TT vaccine—safer out of the cold chain. EPI Cold Chain Newsletter 1990. 21. Edstam JS, Dulmaa N, Tsendjav O, Dambasuren B, Densmaa B. Exposure of hepatitis B vaccine to freezing temperatures during transport to rural health centers in Mongolia. Prev Med 2004; 39:384-8. 22. Hanjeet K, Lye MS, Sinniah M, Schnur A. Evaluation of cold chain monitoring in Kelantan, Malaysia. Bull World Health Organ 1996; 74:391-7. 23. Davaalkham D, Ojima T, Wiersma S, Lkhagvasuren T, Nymadawa P, Uehara R, Watanabe M, Oki I, Nakamura Y. Administration of hepatitis B vaccine in winter as a significant predictor of the poor effectiveness of vaccination in rural Mongolia: evidence from a nationwide survey. J Epidemiol Community Health 2007; 61:578-84. 24. Hingston FJ, Atkinson RJ, Quirk JP. Specific adsorption of ions. Nature 1967; 215:1459-61. 25. Liu JC, Feldkamp JR, White JL, Hem SL. Adsorption of phosphate by aluminum hydroxycarbonate. J Pharm Sci 1984; 73:1355-8. 26. Morefield GL, Sokolovska A, Jiang D, HogenEsch H, Robinson JP, Hem SL. Role of aluminum-containing adjuvants in antigen internalization by dendritic cells in vitro. Vaccine 2005; 23:1588-95. 27. Iyer S, Robinett RS, HogenEsch H, Hem SL. Mechanism of adsorption of hepatitis B surface antigen by aluminum hydroxide adjuvant. Vaccine 2004; 22:1475-9. 28. WHO. Immunization in practice, Module 3: The Cold Chain. WHO/IVB/04.06 2004. Available at: www.who.int/vaccines-documents/.

Acknowledgements

BI

OS

CI

EN

CE

Statistical significance between the treated samples and the time 0 h control was determined using one-way ANOVA with Dunnet posttests (InStat3—GraphPad Software Inc., La Jolla, CA). Particle size analysis. A Coulter counter (model Z1, Beckman Coulter, Fullerton, CA) was used to count the number of particles of antigen-adjuvant complexes. Particles were monitored in seven different detection ranges; 1.5 to 3, 3 to 6, 6 to 8, 8 to 15, 15 to 20, 20 to 25 and 25 to 30 μm. Briefly, 100 μl of each sample was diluted into 20 ml of ISOTON® II diluent (Beckman Coulter) and assayed for number of particles. Measurements of the diluent without sample served as the background. Statistical analyses of the results were based on the number of independent factors. To compare the effect of freezing temperature or freezing duration on the percentage of particles within each range, a one-way ANOVA with Dunnet posttest was performed using InStat3 software (GraphPad Software Inc., La Jolla, CA). A two-way ANOVA with Bonferroni posttests was performed using Prism5 software (GraphPad Software Inc., La Jolla, CA) to determine whether the number of freeze-thaw cycles at various temperatures had an effect on the percentage of particles within the size ranges examined. In all cases, the 4°C untreated vaccine served as the control. In vivo potency assay. The immunogenicity of the vaccines was determined in eight 5- to 7-week-old BALB/c mice (Spring Valley Laboratory, Woodbine, MD). Mice were dosed (0.2 ml, intraperitoneal injection) on days 1 and 29 of the study. On day 43, 150 to 250 μl of blood was collected from each mouse for determination of antibody concentration to HBsAg. An ELISA was used to quantify the HBsAg-specific antibodies present in the mouse sera. Nunc Maxisorb F8 plates were coated with HBsAg. Diluted test serum, standard sera, or control sera (positive and negative) to HBsAg were added to each well of the coated plate. Following incubation, aspiration and washing, biotin-labeled goat anti-mouse IgG H + L (Sigma, St. Louis, MO) was added to each well. Following incubation and washing, the plates were further incubated with straptavidin-HRP conjugate (Sigma). After stopping color development, the absorbance of controls and specimens was determined using a spectrophotometer at a wavelength of 450 nm. The concentrations of the antibodies in the test sera were calculated by comparing them with standard sera containing a known concentration of antibody to HBsAg.

References

ND E

S

The authors would like to thank Kelly Arthur and Joanie Robertson for their technical assistance in the early stages of this study and Dr. Julian Hickling for editing the manuscript.

©

20 08

LA

1. Milstien JB, Galazka AM, Kartoglu Ü, Zaffran M, World Health Organization. Dept. of Immunization Vaccines and Biologicals. Temperature sensitivity of vaccines. Geneva: World Health Organization 2006. 2. Milhomme P. Cold chain study: danger of freezing vaccines. Can Commun Dis Rep 1993; 19:33-8. 3. Miller NC, Harris MF. Are childhood immunization programmes in Australia at risk? Investigation of the cold chain in the Northern Territory. Bull World Health Organ 1994; 72:401-8. 4. Nelson C, Froes P, Dyck AM, Chavarria J, Boda E, Coca A, Crespo G, Lima H. Monitoring temperatures in the vaccine cold chain in Bolivia. Vaccine 2007; 25:433-7. 5. Nelson CM, Wibisono H, Purwanto H, Mansyur I, Moniaga V, Widjaya A. Hepatitis B vaccine freezing in the Indonesian cold chain: evidence and solutions. Bull World Health Organ 2004; 82:99-105. 6. Techathawat S, Varinsathien P, Rasdjarmrearnsook A, Tharmaphornpilas P. Exposure to heat and freezing in the vaccine cold chain in Thailand. Vaccine 2007; 25:1328-33. 7. Wirkas T, Toikilik S, Miller N, Morgan C, Clements CJ. A vaccine cold chain freezing study in PNG highlights technology needs for hot climate countries. Vaccine 2007; 25:691-7. 32

Human Vaccines
