Analytical Method Development

Requirement USP General Chapter ⟨711⟩, “Dissolution,” defines four apparatus types in routine use for solid oral dosage forms.

Figure D1.1

D1.1  Apparatus 1–4, and why the choice matters

• Apparatus 1 (basket) — a rotating wire-mesh basket. Well suited to capsules and to tablets that would otherwise float or stick to the vessel wall.
• Apparatus 2 (paddle) — the most common apparatus for conventional immediate-release solid oral products. A sinker is sometimes added for units that would otherwise float, which itself becomes part of the validated method (module D3 returns to this).
• Apparatus 3 (reciprocating cylinder) — moves a sample through a row of vessels, each potentially holding a different medium. Useful for simulating a changing gastrointestinal pH profile, or for gentler agitation than a paddle provides.
• Apparatus 4 (flow-through cell) — continuously replaces the medium rather than recirculating a fixed volume. Favored for poorly soluble or low-dose drugs, or products where maintaining sink conditions in a fixed-volume vessel is difficult.

Requirement USP ⟨724⟩, “Drug Release,” adds Apparatus 5 (paddle over disk) and 6 (cylinder), both generally used for transdermal systems, and Apparatus 7 (reciprocating holder), used for specialized solid orals such as osmotic-pump tablets and other drug-eluting devices.

Practice None of these seven apparatus types is more “correct” in the abstract. The right choice is whichever apparatus produces hydrodynamics that actually represent how the dosage form behaves — a decision made about the specific product, not a default reached for out of habit.

---

Requirement USP General Chapter ⟨1092⟩, “The Dissolution Procedure: Development and Validation,” is an informational chapter guiding apparatus, medium, and agitation selection; the standard validation parameters from Arc A3 (specificity, linearity, accuracy, precision, robustness) applied specifically to dissolution; and practical execution details including deaeration, filtration, and sampling.

D2.1  Discriminating power: the idea that makes this module necessary

Practice A discriminating method reliably detects clinically meaningful batch-to-batch or formulation-to-formulation differences without over-flagging trivial ones that don't matter. A non-discriminating method — typically from too-aggressive agitation or an overly solubilizing medium — dissolves everything to close to 100% quickly, regardless of real underlying differences, masking exactly the variability the method exists to catch. A method dissolving every batch to 100% in ten minutes, whatever its condition, has not passed a quality bar; it has stopped being able to see one.

Practice Biorelevant media — FaSSIF/FeSSIF and fasted/fed gastric analogues that mimic actual gastrointestinal fluid composition (bile salts, lecithin) more closely than simple buffers — are used especially for poorly soluble BCS Class II/IV drugs, for IVIVC development, and to support BCS-based biowaivers.

D2.2  Discriminating power, worked as a real number: the f2 similarity factor

Requirement The most widely used quantitative tool for comparing two dissolution profiles is the f2 similarity factor, recommended in FDA's SUPAC-IR (1995) and SUPAC-MR (1997) guidance documents:

Worked with two test batches compared against the same reference profile:

Figure D2.1

Test batch A — a minor, defensible process adjustment — tracks the reference closely at every time point, giving f2 = 81.7, comfortably above the 50 threshold. Test batch B releases noticeably more slowly at every time point, reaching only 78% at 60 minutes against the reference's 92%, and its f2 = 39.2 correctly falls below the threshold.

Practice This is what “discriminating power” means in practice, made numerical: a method (and the comparison statistic applied to its output) that gives test batch A a passing similarity score and test batch B a failing one has demonstrated it can actually tell the two apart. A non-discriminating method — one whose aggressive conditions dissolve both test batches to a near-identical plateau regardless of their real difference — would blur this exact distinction, and no amount of correctly executed math would recover the information the method itself never captured.

---

Figure D3.1

D3.1  Six ways to get a wrong answer that looks like a normal one

Practice Deaeration — dissolved gas coming out of solution at 37°C can alter buoyancy and hydrodynamics and coat surfaces; differences exceeding 30% in measured dissolution rate between aerated and properly degassed media have been documented for some formulations. Sinker selection is a validation variable in its own right, not an incidental accessory choice made once and forgotten. Vibration sensitivity — from HVAC systems, foot traffic, or a worn stirring motor — can shift results outside acceptable variability without any single obvious cause. Coning is an undissolved mound that forms beneath the paddle at the vessel's hub when agitation is insufficient, shielding particles from the bulk medium and slowing the apparent dissolution rate. Filter adsorption occurs when a membrane material non-specifically adsorbs some of the dissolved drug, lowering the measured concentration — especially significant at low drug concentrations. Mechanical qualification of the apparatus itself — vessel dimensions, shaft wobble, rotation-speed accuracy — is a recurring inspection finding precisely because it's easy to treat as a one-time setup step rather than an ongoing control.

---

D4.1  Advanced Pharmaceutical Technology — one time point, no justification

“In addition, your contract testing laboratory was conducting a single dissolution test point for your drug products without scientific justification that a single time point demonstrates a consistent drug absorption over an extended period of time.”Advanced Pharmaceutical Technology — Warning Letter, March 14, 2025. fda.gov

Practice A single dissolution time point can be scientifically defensible — for a rapidly dissolving immediate-release product, for instance, where module D2's own discussion of the 85%-in-15-minutes convention already implies a full profile comparison may not be necessary. What FDA cited here was the absence of that justification, for an extended-release product where a single time point says nothing about the release behavior across the rest of the dosing interval.

D4.2  Stason Pharmaceuticals, Inc. — an OOS investigated without a root cause

“Your firm failed to thoroughly investigate any unexplained discrepancy or failure of a batch or any of its components to meet any of its specifications.”Stason Pharmaceuticals, Inc. — Warning Letter, July 8, 2020. fda.gov (21 CFR 211.192; temozolomide capsules, lot 18J043B)

An out-of-specification dissolution result at a stability time point was investigated, but the investigation never reached a root cause or produced a preventive corrective action. Practice An OOS investigation that closes without identifying why the failure happened leaves every batch after it exposed to the same undiagnosed risk — the investigation happened, procedurally, but it didn't actually do the thing an investigation is for.

D4.3  Missouri Analytical Laboratories Inc — flagged as the weaker fit

“In addition, your analysts used individualized non-validated (b)(4) spreadsheets to calculate assay, impurity, content uniformity, and dissolution test results for a variety of drug products.”Missouri Analytical Laboratories Inc — Warning Letter, September 30, 2021. fda.gov (21 CFR 211.68(b))