How do you design an in-use stability study for a multidose or reconstituted product?

1When an in-use study is needed.

EMA’s 2001 note defines the purpose as establishing “a period of time during which a multidose product can be used whilst retaining quality within an accepted specification once the container is opened”, and the dossier includes “either the in-use stability data on which the in-use shelf life is based or a justification why no in-use shelf life is established”.¹ In the US, 21 CFR 211.166(a)(5) requires “Testing of drug products for reconstitution at the time of dispensing (as directed in the labeling) as well as after they are reconstituted”.²

1.1What FDA, the EU, the UK, ICH and WHO have published.

• FDA: no final in-use guidance for human drugs. The only FDA guidance with in-use in its title is CVM GFI #242 (November 2020), for multiple-dose injectable animal drugs.⁷⁸ FDA issued the ICH Q1 draft as its own draft in June 2025,⁹ and its December 2023 ophthalmic draft asks applicants to submit in-use data.¹⁰ The binding rules are 21 CFR 211.166 and 211.137; the final guidance is the 2018 package type guidance, ICH Q1A(R2) and Q5C.²¹¹¹²³⁴
• EU: the 2001 in-use note and the 1998 note on shelf life after first opening set the design and the label wording for human medicines,¹¹³ and preservative efficacy is tested by the Ph. Eur. 5.1.3 method.¹⁴ A veterinary guideline in effect since 30 September 2025 merges the two veterinary notes without changing their science.¹⁵
• UK: regulation 50(5C) of the Human Medicines Regulations 2012 keeps EMA’s scientific quality guidelines in force as they stood at the end of 2020 unless MHRA replaces them, so applications there follow the EU notes.¹⁶
• ICH: the ICH Q1 draft, endorsed at Step 2 on 11 April 2025 and still at Step 3 on 8 November 2026, adds a full in-use section; see the ICH Q1 guideline.¹⁷⁵
• WHO: TRS 1010 Annex 10 (2018) follows the EMA design and adds that “a 30-day in-use period is normally considered acceptable without further supporting data.”⁶

2In-use stability protocol template.

Each row is a protocol section. Replace the middle column with the product’s choices and keep the justification beside each.

Which batch rule applies? Q1A(R2) governs a constituted or diluted product in an ICH filing, with at least three primary batches;³ EMA and WHO govern an opened multidose container.¹⁶ The ICH Q1 draft would make two batches the general rule, with a commitment or justification if no aged batch exists at filing.⁵

2.1Tablets and other solid oral products.

WHO’s example of a multidose product that needs in-use data is a moisture-sensitive or hygroscopic tablet in a large container, such as a polyethylene bottle of 500 tablets.⁶ The ICH Q1 draft lets an applicant justify an open dish study instead.⁵

2.2Choices that decide the result.

• Punctures. FDA’s veterinary guidance, GFI #242, asks for the theoretical maximum number of punctures with “the overall largest needle gauge normally used in practice”, treats all punctures at the start of storage as the worst case, and accepts punctures without withdrawal if the needle touches the product.⁸ The closure should be shown to withstand repeated insertion and withdrawal.⁵⁴
• Preservative. The same guidance wants the minimum preservative assay limit “supported by AET data at the lowest acceptable preservative concentration”, and for self-preserved products asks for AET at the end of the in-use period.⁸

3Example: a preserved multidose vial.

The product, batches and results are invented. A peptide injection is filled at 10 ml into a multidose vial with m-cresol as preservative, stored at 2 to 8 °C, with a dose of 0.25 ml once a day. The proposed label allows 28 days after first use. Batch B1 is 22 months into a 24-month shelf life; batch B2 is 2 months old.

Once a day, on an ordinary laboratory bench, each vial is punctured with a new sterile needle and syringe, 0.25 ml is withdrawn, and the vial goes back to 2 to 8 °C. After 28 withdrawals, 3.0 ml remains, and that final amount is tested. Vials are tested at day 0, day 14 and day 28; further day-0 and day-28 vials go to antimicrobial effectiveness testing by USP <51>, which FDA links to the 28-day period for human multiple-dose injectables,⁸ and by Ph. Eur. 5.1.3.¹⁴

• Every result is within its limit on the last 3.0 ml, and the stopper passes closure integrity testing after 28 punctures.
• m-cresol fell by 0.26 mg/ml. The 2.4 mg/ml lower limit is the lowest concentration at which the product passed AET, so the limit and the microbial claim support each other.
• At 0.0093 mg/ml a day, m-cresol would reach 2.4 mg/ml after about 56 days, but the study supports 28 days, the period tested and the EU’s normal maximum for preserved sterile products.¹³
• The EU label states a maximum of 28 days at 2 to 8 °C once opened; the US label needs no discard statement, because 28 days is FDA’s default for a multiple-dose injectable.¹³¹²

4Example: a freeze-dried antibody diluted for infusion.

The product, batches and results are invented. A monoclonal antibody is supplied as 100 mg of powder, reconstituted with 4.0 ml of water for injections to 25 mg/ml, then diluted in 0.9% sodium chloride to between 0.5 and 4.0 mg/ml. It contains no preservative. As Q1A(R2) asks, the in-use arms run on the three primary batches at release and at 12 months, the last point before filing, and again at the end of shelf life.³

The reconstituted vial is held 24 hours at 2 to 8 °C. Bags at the lowest and highest concentrations, in polyolefin and in PVC, are stored 24 hours at 2 to 8 °C and then, the same bags, 6 hours at 25 °C, the last 2 hours running through an infusion set with a 0.2 µm in-line filter. Protein is measured in the bag and at the end of the set, because contact materials should be shown compatible and sorption addressed.⁵²⁰

• Protein and potency fell together at 0.5 mg/ml in PVC while HMW species stayed level, which points to adsorption to the bag. The label restricts dilution to polyolefin bags.
• The set and filter took another 1% at the lowest concentration, still inside the limit, so the label can allow an in-line filter.
• Microbial arm: polyolefin bags at both concentrations, each inoculated with one of the five pharmacopoeial test strains at 10 to 100 CFU/ml, a level the protocol justifies as touch contamination, follow the same schedule and are counted at 0, 24 and 30 hours. The criterion is the pharmacopoeia’s “no increase”: “not more than 0.5 log10 increase from the initial calculated count”.²¹ No count rose by 24 hours; at 30 hours P. aeruginosa was up 0.3 log10 at 0.5 mg/ml, inside the criterion.
• The EU label states chemical and physical in-use stability for 24 hours at 2 to 8 °C then 6 hours at 25 °C including infusion time, with EMA’s microbiological wording for unpreserved infusions (below).¹³

5How results become the label.

• EU, sterile unpreserved: “Chemical and physical in-use stability has been demonstrated for x hours/days at y °C.” For infusions and injections, storage before use “would normally not be longer than 24 hours at 2 to 8 °C, unless reconstitution / dilution (etc) has taken place in controlled and validated aseptic conditions”.¹³
• EU, sterile preserved: “once opened, the product may be stored for a maximum of z days at t °C”, with z justified case by case and not normally above 28 days.¹³ The period goes on the label, with space for the date of opening if room allows.¹
• US, multiple-dose injectable: FDA generally expects multiple-dose containers to hold 30 ml or less and assumes a 28-day beyond-use date once opened. “If the beyond-use date is other than 28 days for a product in a multiple-dose container, an appropriate discard statement (supported by appropriate data) should be included”, such as “After first use, refrigerate or keep at a temperature not greater than XX for XX days.”¹²

6The study file, the dossier and the inspection.

21 CFR 211.166(a) requires a written stability programme, and that it be followed.² For an in-use study that means:

• Before the first withdrawal: the approved protocol,² and each batch’s number, date of manufacture and size.¹
• During: a log of each withdrawal or puncture and the raw data behind every result.²⁸
• After: a report with the results “summarized and tabulated”, conclusions stated and anomalous results explained.¹
• In the dossier: the data in CTD 3.2.P.8.3, the in-use storage conditions and period in the 3.2.P.8.1 conclusions and the 2.3.P.8 summary, and bag, set and diluent compatibility in 3.2.P.2.6.²⁰

Before approval, FDA investigators compare raw data with the summaries filed, and their programme names finished product stability among the “good candidates for this audit”.²² Have ready the executed log, proof that the aged batch was aged, AET at the lowest preservative limit, and a label period no longer than the period tested.

7What inspectors cite.

Regenerative Processing Plant’s website said its preservative-free eye drops could be “stored at room temperature and used within 90 days” after opening. FDA’s warning letter of 16 August 2024 replied that the in-use period “must be supported by stability data”:²³

“Note that stability studies are needed to support such a statement. Without such studies, the risk for microbial contamination and growth is unknown.”

After complaints of contaminated unpreserved sterile products meant to stay sterile in use, FDA asked Excelvision, a Fareva site, on 25 June 2026 to:¹⁸

“Provide your studies on the growth properties (microbial challenge studies) of each formulation in the event of in-use contamination, based on the maximum number of days of permitted use. In addition, in the event the formulation is deemed suitable following these studies, state whether your firm plans to reduce allowable use periods to no longer than 30 days.”

8How it runs in Seal.

In Seal, an in-use study can run as its own stability protocol beside the product’s long-term study, since several protocols can apply to one product, and each study keeps the protocol version it executed. Removal, receipt, test start and return are separate events, each with its quantity, time, person and source, and a missing result becomes owned work rather than a blank cell.

Neil can plan the next pulls against protocol windows, chamber inventory and laboratory capacity, flagging a unit shortfall before the pull, and can review a stability trend by batch, time point and condition. The study owner assesses the trend, and a trend summary does not establish shelf life.