The container is part of the medicine
A perfect tablet in a leaky blister is not a quality product. The container-closure system — the bottle and cap, the blister foil, the vial and stopper — is regulated as part of the medicine because it controls moisture, oxygen, light and microbial ingress. For sterile products especially, a container-closure integrity test proves the seal keeps microbes and gases out across the whole shelf life; a hairline crack can ruin an otherwise flawless injection. The package also has to be compatible: it must not leach into the product or adsorb the drug.
Stability and shelf life
Stability is whether the product still meets its specification over time. The shelf life is the period — printed as the expiry date — during which it is guaranteed to do so in its package. Real-time stability studies store batches at controlled conditions (a common long-term setting is 25 °C / 60% relative humidity) and re-test assay, related substances, dissolution and appearance at intervals, following the ICH stability guidelines. But waiting three years before launching every product would be crippling.
So we also run accelerated stability testing at higher stress (often 40 °C / 75% RH) to provoke degradation quickly. To turn that into a shelf-life estimate, we lean on the Arrhenius equation, which describes how a reaction's rate constant climbs with temperature. Measure the degradation rate at two or more temperatures, fit Arrhenius, and extrapolate down to the storage temperature.
Arrhenius shelf-life estimate (first-order degradation)
Arrhenius: k = A * exp(-Ea / (R*T))
=> ln k = ln A - (Ea/R) * (1/T)
Measured rate constants:
At 50 C (323 K): k1 = 0.0200 / month
At 40 C (313 K): k2 = 0.0100 / month
Activation energy from the two points:
ln(k1/k2) = -(Ea/R) * (1/T1 - 1/T2)
ln(0.0200/0.0100) = 0.6931
(1/323 - 1/313) = -9.89e-5 K^-1
Ea/R = 0.6931 / 9.89e-5 = 7008 K
Ea = 7008 * 8.314 = 58.3 kJ/mol
Extrapolate to storage at 25 C (298 K):
ln k25 = ln k2 - (Ea/R)*(1/298 - 1/313)
(1/298 - 1/313) = 1.608e-4 K^-1
ln k25 = ln(0.0100) - 7008 * 1.608e-4
= -4.6052 - 1.127 = -5.732
k25 = exp(-5.732) = 0.00324 / month
First-order time to lose 10% of potency (t90):
t90 = 0.105 / k25 = 0.105 / 0.00324 = ~32 months (~2.7 years)
So accelerated data point to roughly a 2-year shelf life at 25 C,
to be confirmed by real-time study.It all lands in the dossier
Every test in this track flows into one place: the regulatory dossier a company submits to gain and keep a marketing authorisation. Its quality module pins down the finished-product specification — the agreed list of tests, validated methods and limits — plus the manufacturing process, the container-closure data and the stability study that justifies the shelf life. Once approved, the specification becomes the contract every future batch is tested against.
For a generic product, the dossier leans heavily on this track: matching the originator's specification, demonstrating comparable dissolution, and proving bioequivalence — sometimes via a biowaiver supported by dissolution data alone. That is the quiet triumph of pharmaceutical quality: a stranger's tablet, made on another continent, can be trusted to behave like the brand because the same standards, tests and dossier discipline stand behind both.