Container/Closure Integrity Testing and the Identification of a Suitable Vial/Stopper Combination for Low-Temperature Storage at -80 C

10.5731/pdajpst.2012.00884Access the most recent version at doi: 453-46566, 2012 PDA J Pharm Sci and Tech

Brigitte Zuleger, Uwe Werner, Alexander Kort, et al.

80 °C−Storage at Suitable Vial/Stopper Combination for Low-Temperature

Container/Closure Integrity Testing and the Identification of a

on December 6, 2014journal.pda.orgDownloaded from on December 6, 2014journal.pda.orgDownloaded from

http://journal.pda.org/lookup/doi/10.5731/pdajpst.2012.00884

http://journal.pda.org/


Container/Closure Integrity Testing and the Identification ofa Suitable Vial/Stopper Combination for Low-TemperatureStorage at �80 °CBRIGITTE ZULEGER1, UWE WERNER1, ALEXANDER KORT2, RENE GLOWIENKA2, ENGELBERT WEHNES1,and DEREK DUNCAN3

1Bavarian Nordic GmbH, Robert Roessle Strasse 10, D-13125 Berlin, Germany; 2IDT Biologika GmbH, AmPharmapark, D-06861 Dessau-Rosslau, Germany; and 3Lighthouse Instruments, Science Park 406, NL-1098 XHAmsterdam, The Netherlands ©PDA, Inc. 2012

ABSTRACT: It was recently found that after storage of a live viral vaccine at �80 °C in glass vials closed with rubberstoppers, a phenomenon was revealed which had not been observed before with other viral products stored at �20 °C:overpressure in the vials. As this phenomenon poses a serious safety problem for medical personnel as well as for theproduct itself, an investigation was initiated to identify the root cause of the overpressure.After exclusion of possible root causes (differences in air temperature or atmospheric air pressure during filling andquality control testing, outgassing from the formulation buffer) the remaining hypothesis involved a possiblecontainer closure integrity issue at low temperature. The glass transition temperatures (Tg) of many rubber stopperformulations are in the range �55 to �70 °C. At storage temperatures below Tg, the rubber stopper loses its elasticproperties and there is a risk that the seal integrity of the vial could be compromised. Loss of seal integrity of the vialsnear storage temperatures of �80 °C would result in an ingress of cold dense gas into the vial headspace. Afterremoval of the vials from storage at �80 °C, the rubber stoppers could regain their elastic properties and the vialswould quickly reseal, thereby trapping the ingressed gas, which leads to overpressure in the vial headspace.Nondestructive laser-based headspace analysis was used to investigate the maintenance of container closure integrityas a function of the filling and capping/crimping process, storage and transport conditions, and vial/stopper designs.This analytical method is based on frequency modulation spectroscopy (FMS) and can be used for noninvasiveheadspace measurements of headspace pressure and headspace gas composition. Changes in the vial headspacecomposition and/or pressure are a clear marker for vials that have lost container closure integrity.

KEYWORDS: Storage at �80 °C, Live viral vaccines, Container closure integrity, Headspace analysis, Frequencymodulation spectroscopy, Glass transition temperature

LAY ABSTRACT: After storage of a live viral vaccine at �80 °C in glass vials closed with rubber stoppers,overpressure in some of the vials was observed, posing a serious safety problem for medical personnel as well as forthe product.A working hypothesis to explain this phenomenon involved a possible container closure integrity issue at these lowtemperatures. The glass transition temperatures (Tg) of many rubber stopper formulations are in the range �55 to�70 °C. At storage temperatures below Tg, the rubber stopper loses its elastic properties, resulting in compromisedseal integrity of the vial and ingress of cold dense gas into the vial headspace. Upon thawing, the rubber stoppersregain their elastic properties and the vials quickly reseal, thereby trapping the ingressed gas, which leads tooverpressure in the vial headspace.Nondestructive, laser-based headspace analysis, which is able to detect changes in headspace pressure and gascomposition, was used to investigate the maintenance of container closure integrity. Changes in the vial headspacecomposition and/or pressure are a clear marker for vials that have lost container closure integrity.

* Correspondence author: Brigitte Zuleger, Bavarian Nordic GmbH, Robert Roessle Strasse 10, D-13125 Berlin,Germany. [email protected]. Phone: �49 (30) 9406 3917

doi: 10.5731/pdajpst.2012.00884

453Vol. 66, No. 5, September–October 2012

on December 6, 2014journal.pda.orgDownloaded from


Introduction

A storage temperature of �80 °C for a live viralvaccine (or any other pharmaceutical product) intro-duces many practical challenges, is expensive, andshould be avoided whenever possible. Nethertheless,during clinical testing, a storage temperature of�80 °C is often chosen for live viral vaccines becausestability data are not available to support productstorage at higher temperatures for the required shelflife.

It was recently found that storage of a live viralvaccine at �80 °C in glass vials closed with rubberstoppers revealed a phenomenon which had not beenobserved before with other viral products stored at�20 °C: overpressure in the vials. During quality con-trol (QC) testing, vials stored at �80 °C were thawedto room temperature (RT) and the rubber stopper waspunctured with the needle of a syringe. In approxi-mately 20 – 40% of the vials (percentage depending onthe product batch), the syringe piston moved back-wards upon insertion of the needle. After withdrawalof the needle, vaccine product shot out of the needleinsertion hole indicating a considerable overpressurewithin the vials. For some vials the cap was removedcompletely which allowed the stopper to pop up fromthe vial. As this phenomenon posed a serious safetyproblem for medical personnel as well as a potentialsterility risk for the product itself, an investigation wasinitiated to identify the root cause of the overpressure.

A couple of possible root causes were identified andcould be excluded after thorough investigation. Dif-ferences in air temperature or atmospheric air pressureduring filling and QC testing could result in pressuredifferences of only a few mbar, much lower than theoverpressure actually observed. Identity testing of theformulation buffer confirmed that the specified ingre-dients and the potential for chemical reactions causingoutgassing from the formulation buffer could be ruledout.

After exclusion of these possible root causes, theremaining hypothesis involved a possible containerclosure integrity issue at low temperature. The glasstransition temperatures of many rubber stopper formu-lations are in the range �55 °C �70 °C. At storagetemperatures below the glass transition temperature,the rubber stopper might lose its elastic propertieswith risk that the seal integrity of the vial could becompromised. Loss of seal integrity of the vials near

storage temperatures of �80 °C would result in aningress of cold dense gas into the vial headspace. Afterremoval of the vials from storage at �80 °C, therubber stoppers could regain their elastic propertiesand the vials would quickly reseal, thereby trappingthe ingressed gas. This would lead to overpressure inthe vial headspace due to volume increase of thetrapped cold gas as it warms up.

To further investigate a possible issue with containerclosure integrity at a storage temperature of �80 °C asthe root cause for overpressure in the viral vaccineproduct vials, four sets of experiments were per-formed. Nondestructive headspace analysis was usedto investigate the maintenance of container closureintegrity as a function of the filling process, storageand transport conditions, and vial/stopper packagingcomponents. Asselta et al. (1) previously identifiedlaser-based headspace analysis as an effective meansto identify closure integrity failures at low tempera-tures. This analytical method is based on frequencymodulation spectroscopy (FMS) and can be used fornoninvasive headspace measurements of headspacepressure and headspace gas composition. A number ofprevious publications have described the use of FMSto characterize the headspace in sterile pharmaceuticalproduct vials for various applications (2–5). Of par-ticular interest for this study, changes in the vialheadspace composition and/or pressure are a clearmarker for vials that have lost container closure integ-rity (2). Analytical platforms based on FMS can there-fore be implemented to perform a quantitative physi-cal container closure integrity test. The nondestructivenature of the measurement enables 100% analysis ofproduct as well as the ability to measure a singlesample over multiple time points.

Materials and Methods

The materials listed in Table I were used for theinvestigations.

Buffer Sample and Vial Closure Preparation

In the first set of experiments, the vials were filledeither manually or by use of an automated filling linewith 0.7 mL of different types of buffers, either at RTor at �5 °C. Stoppering and crimping of both manu-ally- and automatically-filled vials was done automat-ically (see description below). For all subsequent setsof experiments, the vials were prepared in a waymimicking the filling of the actual viral vaccine prod-

454 PDA Journal of Pharmaceutical Science and Technology



uct as closely as possible. Formulation buffer or salinewas stirred at RT and filled in portions of 0.7 mL intothe vials by use of an automated filling line. The vialswere crimped using an automated system with cappingand crimping parameters that were set manually. In thecourse of experiments two settings were used: stiff andtight. The stiff setting barely allowed manual move-ment of the cap whereas the tight setting did not allowany movement of the cap after crimping. For the restof this paper, samples crimped using the stiff settingwill be referenced as loosely crimped while samplescrimped using the tight setting will be referenced astightly crimped.

Low-Temperature Storage

For the first set of experiments, filled vials were storedat RT, �5 °C, �20 °C, �80 °C and on dry ice andanalysed after 13 days of storage. This period of timewas chosen for subsequent investigations as the prod-uct vials which initially showed the phenomenon ofoverpressure in QC testing were stored at �80 °C fora period of 13 days. Prior to the measurement, vialswere allowed to equilibrate to RT in normal air for atleast 1 h.

For storage at �80 °C, a deep freezer was used. In thesecond set of experiments, the cold transport processof product vials was mimicked by storing and shippingsample vials on dry ice followed by subsequent head-space analysis of the samples.

The last two sets of experiments used vials shipped atRT and stored on dry ice for defined periods as de-scribed in the Results section. Prior to headspace

oxygen analysis, vials stored on dry ice were allowedto equilibrate to RT in a carbon dioxide-rich atmo-sphere to prevent any gas exchange with normal airduring thawing.

Headspace Analysis

Headspace oxygen measurements were performed us-ing a nondestructive headspace oxygen analyzer(model FMS-760, Lighthouse Instruments, Charlottes-ville, VA, USA). Calibration was performed usingcertified 20% and 0% oxygen standards.

These certified standards were made by backfilling 2Rvials with certified NIST traceable oxygen gas mix-tures and then flame sealing them shut. Because theyare made from 2R vials, the standards can be measuredidentically to 2R vial samples. The results listed inTable II are of repeated measurements of the knownstandards and demonstrate the accuracies and preci-sions of the headspace oxygen measurements for the2R vial configuration.

Five consecutive measurements were made on each ofsix known oxygen standards to verify the performanceof the system (see Table II for an example of head-space oxygen performance data). The headspace oxy-gen in each sample vial was then measured and re-corded.

Headspace pressure measurements were performed us-ing a nondestructive headspace pressure analyzer(model FMS-1400, Lighthouse Instruments, Charlottes-ville, VA, USA). Calibration was performed with cer-tified pressure/moisture standards.

TABLE IVarious Vial Stopper Combinations Using the Components Listed in This Table Were Used in ContainerClosure Studies after Storage at �80 °C

Material Designation Comments

13 mm rubber stoppers SNB1 Stopper Non–Blow back, manufacturer 1Tg of rubber compound: � �60 °C

SNB2 Stopper Non–Blow back, manufacturer 2Tg of rubber compound: � �65 °C

SBB Stopper Blow Back, manufacturer 1Tg of rubber compound: � �60 °C

2R vials (overflowcapacity 4 mL)

VBB Vial Blow Back, manufacturer 3

VNB Vial Non–Blow back, manufacturer 3

13 mm caps Plastic aluminiumflip off cap

Same type of caps was used for all vial/stoppercombinations




These certified standards were made by evacuating 2Rvials to known calibrated pressures and then flamesealing them shut. Because they are made from 2Rvials, the standards can be measured identically to 2Rvial samples. The results listed in Table III are ofrepeated measurements of the known standards anddemonstrate the accuracies and precisions of the head-space pressure measurements for the 2R vial config-uration.

Three consecutive measurements were made on eachof seven known pressure standards to demonstrate theperformance of the system (see Table III for an exam-ple of headspace pressure performance data). Theheadspace pressure in each sample was then measuredand recorded.

The samples were allowed to thaw for at least one hourbefore the headspace analysis. For cases where thesamples were stored in a deep freezer (air environ-ment), the samples were thawed to RT in an airenvironment. For cases where the samples were storedon dry ice (carbon dioxide environment), the sampleswere allowed to thaw in a carbon dioxide–rich atmo-sphere. This was done by removing the samples fromthe storage box with dry ice and placing them into a 25L plastic container. The lid of the plastic container wassealed with a rubber O-ring and had two holes, onehole for a rubber tube connected to a carbon dioxidegas source, the other hole for a rubber tube that servedas an exhaust. When the samples were placed insidethe plastic container, the carbon dioxide gas sourcewas opened and the container was purged continu-ously for 15 min with carbon dioxide at a rate �5SLPM (standard liters per minute). After 15 min ofpurging, the holes of the container lid were sealed withtape and the samples were allowed to thaw in thecarbon dioxide-rich atmosphere for a minimum of 1 h.

Results and Discussion

The objective of the first set of experiments was toconfirm the overpressure phenomenon. After the ini-tial observation of the overpressure phenomenon, anumber of viral vaccine product vials from four dif-ferent batches were investigated for overpressure afterstorage at different temperatures (Table IV).

No vials with overpressure were identified after stor-age at �5 °C and �20 °C. The number of vials with

TABLE IIResults of Five Consecutive Measurements onCertified Standards of Known OxygenConcentration

OxygenStandard Value

(% atm)

Measured Oxygen(% atm)

Mean SD

19.99 20.04 0.10

8.01 8.03 0.11

4.00 4.19 0.04

2.00 2.19 0.12

1.00 0.95 0.08

0.00 0.07 0.07

SD � standard deviation.

TABLE IIIResults of Three Consecutive Measurements onCertified Standards of Known Pressure Levels

PressureStandard Value

(mbar)

Headspace Pressure(mbar)

Mean SD

60.4 57 0.11

121 115 0.32

255 254 0.37

511 519 0.41

655 672 0.40

803 816 0.88

939 943 2.19

SD � standard deviation.

TABLE IVOverpressure Results (>1500 mbar) of ViralVaccine Product Vials Stored at DifferentTemperature Conditions

Product Storage

Number ofVials with

Overpressure

Batch 1 �80 °C 7 out of 10

�80 °C/dry ice 27 out of 74

Batch 2 �80 °C/dry ice 5 out of 25

Batch 3 �5 °C 0 out of 10

�20 °C 0 out of 10

�80 °C 0 out of 35

�80 °C/dry ice 1 out of 25

Batch 4 �80 °C/dry ice 2 out of 98




overpressure after storage at �80 °C or �80 °C/dryice differs remarkably between batches 1 and 2 andbatches 3 and 4. As part of the first set of experimentsit was then decided to investigate the influence of thestorage temperature, filling temperature, and vial de-sign on the potential creation of overpressure vials. Inaddition, three similar compositions of the formulationbuffer from two different vendors were investigatedfor outgassing of buffer components as a possible rootcause for the observed overpressure phenomenon. Atotal of 769 vials were filled either automatically ormanually with 0.7 mL of different types of buffers,either at RT or at �5 °C. Two different types of 2Rvials from the same manufacturer were used, non–blow back (VNB) and blow back neck design (VBB).For all fillings, the same lot of stoppers was used frommanufacturer 1, a non– blow back design (SNB1). Fordetails, see Table V. The buffer-filled vials as well asthe viral vaccine product vials were crimped loosely(see Materials and Methods).

Filled vials were stored at RT, �5 °C, �20 °C,�80 °C and on dry ice for 13 days. Prior to theheadspace pressure measurements, vials were allowedto equilibrate to RT. In contrast to the initially mea-sured viral vaccine product vials, none of the preparedbuffer vials displayed overpressure, even after storageat �80 °C and on dry ice.

To help explain these results, a closer look was takenat the packaging components used for the vials in thefirst set of experiments. For all investigated vials thesame design of stoppers (non– blow back, SNB1) andvials (blow back, VBB) were used. This combinationof non– blow back stoppers and blow back vials hadbeen chosen to limit the number of required primary

packaging components for the studies as these vialsand stoppers are also used for other freeze-dried viralvaccines. The suitabilty of this vial/stopper combina-tion for liquid frozen vaccines stored at �20 °C hadbeen proven for previous liquid products.

The type and lots of vials and stoppers used for thefirst set of experiments are listed in Table VI. FromTable VI it can be seen that for product batches 1 and2, a different lot of stoppers (A) and vials (C) wereused than for product batches 3 and 4 (stopper lot B,vial lot D). This might explain the difference in thenumber of overpressure vials between batches 1 and 2and batches 3 and 4. For the buffer-filled vials, thesame lot of stoppers (B) and vials (D) were used as forbatches 3 and 4. These results indicate that the differ-ent lots of vials and stoppers may be partly responsiblefor the overpressure phenomenon. However, the fact

TABLE VComposition/Production of Buffer-Filled Vials for First Set of Experiments

Stoppers Vials BufferFilling

TemperatureFilling

Procedure

Stoppering/CrimpingProcedure

SNB1 VBB A RT Automatically Automatically

VBB A RT Manually

VNB A RT Manually

VBB B RT Manually

VBB C RT Manually

VBB A 5 °C Manually

SNB1 � stopper non– blow back manufacturer 1, VBB � vial blow back, VNB � vial non– blow back.Buffers A–C represent three similiar compositions of the fomulation buffer.

TABLE VIListing of the Different Types and Lots of Vialsand Stoppers Used for Four Batches of Live ViralVaccine Product and Various Buffer-Filled Vials

Product

Stoppers Vials

Design Lot Design Lot

Batch 1 SNB1 A VBB C

Batch 2 A C

Batch 3 B D

Batch 4 B D

Buffer-filledvials

B VBB � VNB D � E

SNB1 � stopper non– blow back manufacturer 1,VBB � vial blow back, VNB � vial non– blow back.




that none of the buffer vials showed overpressuremotivated further sets of experiments to investigateadditional factors besides the lots of vials and stoppersthat could potentially be responsible for the overpres-sure phenomenon.

In order to verify how vial closure integrity is depen-dent on the vial/stopper combination, a second set ofexperiments was performed. In total, 10 different vial/stopper combinations were tested, all of them crimpedloosely (see Materials and Methods). Two differentdesigns of the 2R vials were chosen from the samemanufacturer, non– blow back (VNB) and blow backdesign (VBB). Different designs of the stoppers wereidentified, from manufacturer 1, non– blow back(SNB1) and blow back design (SBB) and from man-ufacturer 2, non– blow back design (SNB2). Besidesthe different designs, different lots of vials werechecked for size of the inlet diameter (Table VII) anddifferent lots of stoppers were checked for size of theplug diameter (Table VIII). The diameters of the vialsand stoppers are derived from the incoming goodscontrol of the respective lots. The average diametersare calculated from 20 single measurements.

The vial lots (blow back) with the largest and smallestinlet diameter were chosen for investigation, as well asone lot of non– blow back vials. A non– blow backstopper lot from manufacturer 1 with the largest andsmallest plug diameter was identified for subsequentinvestigations, as well as a non– blow back stopperfrom a different manufacturer 2 (SNB2), because thespecified stopper plug diameter from manufacturer 2 isapproximately 0.1 mm larger than the specified stop-per plug diameter from manufacturer 1. In addition, ablow back stopper from manufacturer 1 was chosenfor subsequent investigation. Altogether, 3 differentvials and 4 different stoppers were investigated in 10different vial/stopper combinations (Table IX).

A total of 50 vials of each vial/stopper combinationwere filled with 0.7 mL of saline, and stoppered andcrimped, using an automated filling line. After storagefor 14 days on dry ice, vials were thawed in a CO2

atmosphere and measured for oxygen content. If vialsare stored at dry ice temperature (�78.5 °C), the ini-tial 1 atm headspace pressure of the vials at RT issignificantly reduced due to the cooling of the initialair headspace. If rubber stoppers lose their seal integ-

TABLE VIIInlet Dimensions of Vials from Various Lots

2R Vial Type Lot No.

Diameter of Vial Inlet (mm)

Min Max Average

Blow back vialEN ISO 8362-1, d4

6.80 7.20 N/A

Specification manufacturer 3 6.80 7.20 N/A

VBB 1 (largest blow back vial) 6.85 7.15 7.052 6.86 7.12 7.04

3 6.86 7.05 6.95

4 6.85 7.07 6.94

5 6.85 7.08 6.95

6 (smallest blow back vial) 6.85 7.02 6.907 6.86 7.12 6.98

8 6.86 7.06 6.90

9 6.86 7.03 6.94

10 6.95 7.15 7.04

Non–blow back vialDIN EN ISO 8362-1, d4

6.80 7.20 N/A


VNB 1 (non–blow back vial) 6.88 7.00 6.96

VBB � vial blow back, VNB � vial non– blow back.Lots marked in bold were chosen for investigations.




rity at these low temperatures, the pressure gradientdrives CO2 into the headspace through the leak dis-placing some or all of the original air headspace.

Figure 1 shows the measured headspace oxygen re-sults after 14 days of storage on dry ice. Four of thevial stopper combinations (combinations 1, 2, 3, and9) have produced samples with depleted oxygen levels(�17% oxygen) after storage on dry ice with combi-nations 2, 3, and 9 having especially large numbers ofoxygen-depleted vials; this is summarized in Table IX.The use of the blow back stopper from manufacturer 1in combination with all three different vial types (sets1, 2, and 3) resulted in multiple instances of oxygendepletion. Sets 1 through 3 show 4, 27, and 39 vialswith depleted oxygen levels, respectively. The vial/stopper combination previously used as the standardcombination did not maintain container closure integ-rity at �80 °C (set 9: 31 vials with depleted oxygenlevel), if the combination large vial/small stopper wasused. In contrast, the combination small vial/largestopper (set 7) of the standard combination did notshow any vials with depleted oxygen levels, strength-

ening the hypothesis that design as well as size of vialsand stoppers are important for storage at �80 °C.

After the initial measurements, vials identified as hav-ing depleted levels of headspace oxygen were stored atRT in a normal air environment and measured againafter 4 and 22 days to monitor the time evolution ofthe headspace oxygen content. The time-evolved mea-surements also included measurements of the head-space pressures. The objective of the time-evolvedmeasurements was to correlate depleted oxygen levelsin a vial to the presence of overpressure, and todetermine if leaks in the samples were temporary orpermanent. As an example of the results, Figure 2 andFigure 3 plot the time evolved headspace measure-ments for vial/stopper combination 1. In addition tothe four samples having oxygen levels �17%, threeadditional samples having slightly depleted oxygenlevels (between 17% and 18%) were also monitored.After a 22 day period, vials that were initially identi-fied as having depleted oxygen levels showed a slightincrease in headspace oxygen. Headspace pressuremeasurements confirmed that the oxygen depleted vi-

TABLE VIIIStopper Plug Dimensions of Various Lots of Stoppers

13 mm Stopper Type Lot No.

Diameter of Stopper Plug (mm)

Min Max Average

Non–blow back stopperDIN EN ISO 8362-2, d1

7.35 7.65 N/A



SNB1 1 7.30 7.44 7.37

2 (smallest non–blow back stopper) 7.32 7.40 7.353 7.30 7.42 7.36

4 7.34 7.40 7.37

5 (largest non–blow back stopper) 7.37 7.48 7.416 7.32 7.45 7.38

7 7.30 7.45 7.38

8 7.33 7.42 7.37

SNB2 1 (non–blow back stopperdifferent manufacturer)

7.44 7.55 7.48

Blow back stopperDIN EN ISO 8362-2, d4

7.40 7.80 N/A


SBB 1 (blow back stopper manufacturer 1) 7.70 7.71 7.70

SNB1 � stopper non– blow back manufacturer 1, SNB2 � stopper non– blow back manufacturer 2, SBB � stopperblow back manufacturer.Lots marked in bold were chosen for investigations.




als also contained overpressure. The overpressures inthe headspace were also monitored over time and theresults in Figure 3 show a slight decrease in overpres-sure over this time period.

To help interpret the time-evolved measurements,comparisons can be made to calculations of a head-space leak rate model described in (6). This headspace

leak-rate model has been validated with experimentalstudies and allows the calculation of changing head-space conditions due to a leak defect. Input parametersfor the model are headspace volume, effective leakhole size, and initial headspace conditions. Figure 4demonstrates how quickly an overpressure of 1500mbar in a 2R vial having an empty headspace volumeof 3 mL would come to equilibrium with an atmo-

Figure 1

Headspace oxygen results from measurements of all samples from 10 different vial/stopper combinations in thesecond set of experiments.

TABLE IXOverview of Vial/Stopper Combinations Tested in the Second Set of Experiments, Vials Identified after 14Days Storage on Dry Ice with Depleted Headspace Oxygen Levels

Set No.

Stoppers Vials No. of Vials from Totalof 50 with O2 < 17%Design Size Design Size

1 SBB NA VNB NA 4

2 SBB NA VBB Small 27

3 SBB NA VBB Large 39

4 SNB1 Small VNB NA 0

5 SNB2 NA VBB Large 0

6 SNB2 NA VNB NA 0

7 SNB1 Large VBB Small 0

8 SNB1 Large VNB NA 0

9 SNB1 Small VBB Large 31

10 SNB2 NA VBB Small 0

SNB1 � stopper non– blow back manufacturer 1, SNB2 � stopper non– blow back manufacturer 2, SBB � stopperblow back manufacturer 1, VBB � vial blow back, VNB � vial non– blow back.




sphere environment assuming a very small effectiveleak hole size of 0.2 �m. The rate of pressure changein the graph of Figure 4 can be compared to the rate ofpressure change measured in the samples as plotted inFigure 3. One can conclude that the change in pressureof the samples (less than 100 mbar within 18 days) isminor compared to that calculated by the headspaceleak rate model for a 0.2 �m leak (approximately 400mbar within 18 days). These results indicate that thesamples are sealed and are maintaining the overpres-sure. There is no sizable leak present in the sample

vials during storage at RT which allows equilibrationof the headspace overpressure. The fact that the oxy-gen levels do not appreciably increase over time due toan ingress of air through a leak as predicted by theheadspace leak rate model is a second verification thatthe oxygen-depleted/overpressure vials have sufferedtemporary leaks during storage at low temperature andthen reseal when warming up to RT.

As could be seen in the second set of experiments,container closure integrity at �80 °C depends on the

Figure 2

Time-evolved headspace oxygen results from measurements of oxygen-depleted samples from vial/stoppercombination 1.

Figure 3

Plot of time-evolved headspace pressures measured in oxygen-depleted vials from vial/stopper combination 1.




design as well as on the relative size of the vial neckand the stopper plug. This is a noteworthy observationas it is generally accepted that the most robust seal iscreated by the stopper flange and the top vial finish,provided that the aluminum crimp is applied withsufficient pressure. The non– blow back stopper frommanufacturer 2 (SNB2) was chosen for all future ap-plications, because the stopper plug diameter is ap-proximately 0.1 mm larger compared to the plug di-ameter of the non– blow back stopper frommanufacturer 1, thereby enhancing the chances ofobtaining appropriate container closure integrity at�80 °C due to size of the stoppers. In addition, the Tg

of the rubber stopper from manufacturer 2 is approx-imately 5 °C lower than the Tg of the rubber stopperfrom manufacturer 1.

A third set of experiments was planned and executedto identify the onset of overpressure as a function of

storage time. Three slightly different lots of vials andone lot of the non– blow back stopper from manufac-turer 2 were investigated. The vial stopper combina-tion from set 9 described above in Table IX (largevial/small stopper) was used as a positive control,where the phenomenon of overpressure vials could beexpected. Three hundred vials of each of the fourvial/stopper combinations were filled with 0.7 mLformulation buffer, and stoppered and crimped, usingan automated filling line. Vials were stored on dry icefor 3, 6, 14, and 21 days. After equilibration to RT ina CO2 rich atmosphere, vials were measured for oxy-gen content and overpressure.

No oxygen-depleted samples (samples having �17%oxygen) or samples with overpressure (samples havingheadspace pressure �1150 mbar) were identified afterday 3 and day 6 measurements in all of the fourvial/stopper combinations, including the positive con-trol vials. After 14 and 21 days, respectively, one andthree of the vials of the positive control (a different setof vials was measured at each timepoint) showedreduced oxygen content and overpressure. All othervial/stopper combinations using the stoppers frommanufacturer 2 were without overpressure vials. Asummary of these results is shown in Table X.

The results of the third set of experiments are in strongcontrast to the second set of experiments, where set 9,the same vial/stopper combination used as the positivecontrol in the third set of experiments, showed 31 outof 50 vials with overpressure. The failure rate hasdropped dramatically for this combination in the thirdset of experiments compared to the second. Uponfurther examination, the capping/crimping process

Figure 4

Loss of overpressure for 2R vial, 0.2 �m leak (leakrate model).

TABLE XSummary of Results of Third Set of Experiments

Vial/Stopper Combination

Depleted Oxygen Vials Overpressure Vials

Day 3 Day 6 Day 14 Day 21 Day 3 Day 6 Day 14 Day 21

Set 9 POSITIVE CONTROLSNB1 Small � VBB Large

0 0 1 3 0 0 1 3

Set 6SNB2 � VNB

0 0 0 0 0 0 0 0

Set 6�SNB2 � VNB new lot

0 0 0 0 0 0 0 0

Set 10SNB2 � VBB Small

0 0 0 0 0 0 0 0

SNB1 � stopper non– blow back manufacturer 1, SNB2 � stopper non– blow back manufacturer 2, VBB � vial blowback, VNB � vial non– blow back.




was identified as possibly contributing to the rootcause for the overpressure effect in addition to thedesign and size of vials and stoppers. The vials in thisthird set of experiments were crimped tightly, whereasthe vials of the previous experiments were crimpedloosely (see Materials and Methods). This is again anoteworthy observation. Using a higher crimp pres-sure is not sufficient to maintain a robust seal betweenthe stopper flange and the top vial finish and eliminatecontainer closure integrity failure at �80 °C storage.Only the addition of the extra sealing surfaces from arelatively large stopper plug provides reliable sealintegrity.

To investigate this hypothesis further, a fourth set ofexperiments was performed using the vial/stoppercombination non– blow back vial and non– blow backstopper from manufacturer 2. This vial/stopper com-bination gave the best results in the previous studiesdue to the optimum relative dimensions of a small

inner vial neck diameter and a large stopper plugdiameter. A total of 392 vials were filled with 0.7 mLformulation buffer; half of the vials were “tightlycrimped,” the other half were “loosely crimped.” Thevials were measured for overpressure after 4 and 12days of storage on dry ice. Figures 5 and 6 show theheadspace pressure results.

After 4 days of storage on dry ice, the loosely crimpedset showed three vials having an overpressure. After12 days storage on dry ice, two overpressure vialswere identified in the loosely crimped set (a differentset of vials was measured at each time point). Therewas no vial with overpressure in the sets of tightlycrimped vials. These results are summarized in TableXI and confirm that even an “optimal” vial/stoppercombination from a design and dimensional point ofview needs to be appropriately capped and crimped tomaintain container closure integrity during storage at�80 °C.

Figure 5

Overpressure results for vials stored on dry ice for four days comparing samples capped and crimped with atight setting to samples capped and crimped with a loose setting.

Figure 6

Overpressure results for vials stored on dry ice for 12 days comparing samples capped and crimped with a tightsetting to samples capped and crimped with a loose setting.




Conclusions

The results of this study support the original hy-pothesis that overpressure found in product vials isa result of losing container closure integrity duringcold storage. The storage of stoppered vials at atemperature of �80 °C increases the risk of com-promising container closure integrity due to the factthat this temperature is below the glass transitiontemperature (Tg) of most rubber stopper formula-tions. At storage temperatures below Tg, the rubberstopper can lose its elastic properties and there is arisk that the seal integrity of the vial is compro-mised. This study also demonstrated that stopperedvials that lose closure integrity at low temperatureswill reseal when warming back up to RT. Once thevial reaches a temperature above the Tg of therubber stopper, the stopper regains its elastic prop-erties and reseals the vial closure. Leaks at theselow storage temperatures can therefore be tempo-rary.

Some of the container closure integrity tests describedin the Pharmacopeias and other sources (7–10) are notuseful for identification of container closure integrityfailures during storage at �80 °C. This is becausethese physical container closure integrity tests identifycontainers that are leaking at the time of measurement.Because of the impracticality of analyzing vials whilethey are stored at �80 °C, these measurements forcontainer closure integrity are often performed at RTafter the leaks have resealed. The measurement at RTdoes not reflect the status of the vials during storage at�80 °C and therefore does not detect the temporaryleak.

On the other hand, a nondestructive analytical methodthat quantitatively characterizes the gas conditions ofthe vial headspace provides a way to detect vials that

have been temporarily leaking at low storage temper-atures. The headspace method, based on FMS technol-ogy, enabled studies described in this paper whichidentified the root cause of a headspace overpressurephenomenon in viral vaccine product vials stored at�80 °C. In addition, the studies demonstrated that theheadspace method enabled identification of rubberstopper/vial combinations that maintain closure integ-rity at storage temperatures below the glass transitiontemperature of the rubber stoppers. Critical factorsfor the maintenance of container closure integrityincluded appropriate design of the vial and stopperplug, relative dimensions of the stopper and the vialneck giving a tight fit, as well as an appropriatelytight capping and crimping process. The dimen-sional variability found between different vial andrubber stopper lots as well as different specifica-tions for the 13mm stopper depending on stoppermanufacturer motivates a careful selection of pack-aging components for storage at �80 °C. For futureconsideration, rubber formulations with Tg below�80 °C might also have a positive effect on main-taining container closure integrity at deep frozenstorage conditions.

Acknowledgements

The authors thank all at Bavarian Nordic, IDT Bi-ologika, and Lighthouse Instruments who contributedto solve this problem: identifying a measurement pro-cedure for overpressure in vials, discussing results,designing experiments, and preparing all the materialsfor measurement.

Conflict of Interest Declaration

The authors declare that they have no competing in-terests.

References

1. Asselta, R.; Smith, E.; Sunderland, W.; Trappler,E. Effect of Low Temperatures on Parenteral VialSeal Integrity. Presented at the PDA AnnualMeeting, Las Vegas, NV, 2007.

2. Templeton, A. C.; Han, Y.-H. R.; Mahajan, R.;Chern, R. T.; Reed, R. A. Rapid HeadspaceOxygen Analysis for Pharmaceutical PackagingApplications, Pharm. Technol. 2002, July, 44 –61.

TABLE XISummary of Results Investigating the Influence ofCapping and Crimping to the Loss of ContainerClosure Integrity during Cold Storage

CappingCondition

Number of OverpressureVials out of 98

Day 4 Day 12

Tight 0 0

Loose 3 2




3. Lin, T. P.; Hsu, C. C.; Kabakoff, D. B.; Patapoff,T. W. Application of frequency-modulated spec-troscopy in vacuum seal integrity of lyophilizedbiological products. PDA J. Pharm. Sci. Technol.2004, 58 (2), 106 –115.

4. Mahajan, R.; Templeton, A. C.; Reed, R. A.;Chern, R. T. Frequency modulation spectroscopy:a novel non-destructive approach for measuringmoisture activity in pharmaceutical samples.Pharm. Technol. 2005, 29 (10), 88 –104.

5. Cook, I.; Ward, K. Applications of headspacemoisture analysis for investigating the water dy-namics within a sealed vial containing freeze-dried material. PDA J. Pharm. Sci. Technol. 2011,65 (1), 2–11.

6. Veale, James R. New Inspection Developments.In Practical Aseptic Processing Fill and Finish;Lysfjord, J., Ed.; Davis Healthcare InternationalPublishing/PDA: Bethesda, MD, 2009; pp 305–372.

7. USP 1207 Sterile Product Packaging—IntegrityEvaluation.

8. USP 381 Elastomeric Closures for Injections.

9. Ph. Eur. 3.2.9. Rubber Closures for Aqueous Par-enteral Preparations, for Powders and for Freeze-Dried Powders.

10. PDA Technical Report No. 27. PharmaceuticalPackage Integrity. PDA: Bethesda, MD, 1998.




Authorized User or for the use by or distribution to other Authorized Users·Make a reasonable number of photocopies of a printed article for the individual use of an·Print individual articles from the PDA Journal for the individual use of an Authorized User ·Assemble and distribute links that point to the PDA Journal·Download a single article for the individual use of an Authorized User·Search and view the content of the PDA Journal permitted to do the following:Technology (the PDA Journal) is a PDA Member in good standing. Authorized Users are An Authorized User of the electronic PDA Journal of Pharmaceutical Science and

copyright information or notice contained in the PDA Journal·Delete or remove in any form or format, including on a printed article or photocopy, anytext or graphics·Make any edits or derivative works with respect to any portion of the PDA Journal including any·Alter, modify, repackage or adapt any portion of the PDA Journaldistribution of materials in any form, or any substantially similar commercial purpose·Use or copy the PDA Journal for document delivery, fee-for-service use, or bulk reproduction orJournal or its content·Sell, re-sell, rent, lease, license, sublicense, assign or otherwise transfer the use of the PDAof the PDA Journal ·Use robots or intelligent agents to access, search and/or systematically download any portion·Create a searchable archive of any portion of the PDA JournalJournal·Transmit electronically, via e-mail or any other file transfer protocols, any portion of the PDAor in any form of online publications·Post articles from the PDA Journal on Web sites, either available on the Internet or an Intranet,than an Authorized User· Display or otherwise make any information from the PDA Journal available to anyone otherPDA Journal·Except as mentioned above, allow anyone other than an Authorized User to use or access the Authorized Users are not permitted to do the following:



Documents

Container/Closure Integrity Testing and the Identification of a Suitable Vial/Stopper Combination for Low-Temperature Storage at -80 C