Beyond Dissolution: Reframing Lipid Functionality

By George ForrestTechnology

Why the next chapter in oral drug delivery is being written by lipids — and not the ones you think.

Lipid-based delivery has been the most credible answer to a brutal pharmaceutical truth: most new drug molecules cannot dissolve well enough to work. Thirty years on, the answer’s ceiling is now visible. The next move is not to optimise the system that depends on dissolution. It is to engineer past it.

6-minute read

The constraint nobody admits

Between 70 and 90 per cent of compounds in the pharmaceutical pipeline are poorly water-soluble (Lipinski, 2000; Di et al., 2012). Around 40 per cent of marketed drugs sit in BCS Class II — high permeability, low solubility — meaning that dissolution, not membrane crossing, is the step that limits whether the dose ever reaches the bloodstream.

Every solid oral product must disintegrate, dissolve in gastrointestinal fluid, and stay dissolved long enough for the intestinal wall to absorb it. Each of those steps is variable. Gastric emptying differs between patients, between meals, between days. Transit time is not a number — it is a distribution. The dose is designed for a target with sub-nanomolar precision; the route by which it gets there is none of those things.

What Neoral actually proved

The single best demonstration that lipid science can defeat part of this problem is cyclosporine.

Sandimmune, the 1983 formulation, was a coarse oil-in-water emulsion. Mean oral bioavailability was about 30 per cent. Absorption was bile-dependent and food-dependent — the difference between dosing in the fasted state and after a high-fat meal could push AUC by a factor of two to three (Mueller et al., 1994; Kovarik et al., 1994). For a drug with a narrow therapeutic window — graft rejection on one side, nephrotoxicity on the other — that swing was not an inconvenience. It was a clinical safety problem.

Neoral, the 1995 microemulsion reformulation, dropped that problem. Relative bioavailability rose to 174–239 per cent of Sandimmune across dose levels (Mueller et al., 1994; Kovarik et al., 1994; Holt et al., 1994). Dose-normalised AUC was 32–63 per cent higher. Droplet size on dispersion was below 0.15 µm. Absorption became bile-independent. Dose linearity, which Sandimmune never had, appeared. The same molecule, with the same target, in the same patient, suddenly behaved.

That is what lipid engineering at its best can do. It is the founding proof of self-emulsifying drug delivery as a serious pharmaceutical discipline. Every later lipid-based product — saquinavir, tipranavir, sirolimus capsules, dutasteride softgels — descended from that moment.

What Norvir proved next

Three years after Neoral, lipid systems hit a wall that no amount of formulation craft could engineer around.

Ritonavir, Abbott’s HIV protease inhibitor, launched in 1996 as a soft-gel capsule. The molecule had one known crystal form. In mid-1998, batches began failing dissolution testing. A second, more stable, far less soluble polymorph — Form II — had appeared (Bauer et al., 2001). It crystallised out of the soft-gel matrix and would not redissolve. The lipid vehicle was no longer in possession of the drug.

Norvir was withdrawn from US distribution. Abbott rebuilt manufacturing — refrigerated gelcap, then melt-extruded amorphous solid dispersion. The direct financial cost was reported at more than $250 million; the clinical cost was the temporary disruption of one of the most important HIV regimens of the decade.

The lesson is not that lipid systems are fragile. The lesson is that they are still inside the dissolution paradigm. A SEDDS holds the drug in a solubilised state ready to disperse in the gut — but it depends on the drug starting in the right form, the patient starting in the right physiological state, lipolysis happening on schedule, and bile arriving on cue. The system optimises within the constraint. It does not remove it.

Optimise the constraint, or engineer past it

Tablets are constrained by dissolution.

Lipid systems optimise within it.

Next-generation platforms eliminate it.

After thirty years inside the SEDDS paradigm, the ceiling of optimisation is visible. The cyclosporine gain — roughly a factor of two in bioavailability, a step-change in dose linearity — is approximately what good lipid engineering can deliver while the dissolution step is still in the system. The Norvir failure is a reminder that even that gain can be reversed by a single change in the API’s crystal landscape.

A structured lipid delivery architecture changes the question being asked. The answer depends on the active. For the compounds the architecture has been proven to solubilise — NSAIDs, potassium cinnamate, and polysorbates in specific ratios with GCC — the active enters the patient already in solution: no solid to disintegrate, no precipitation race, no bile dependency. For other actives carried in the same mix, dissolution still happens in vivo — but from the controlled starting state of the structured matrix rather than an uncontrolled suspension. Either way, the variables that used to belong to the patient’s gut are substantially decided at the point of manufacture.

Lipids, repositioned

In a tablet, lipid excipients compensate — they help a poorly soluble API survive contact with the GI tract well enough to be absorbed. Remedial.

In a SEDDS, lipids carry the API in a metastable state and rely on the gut to do the final emulsification. Facilitative.

In a structured lipid architecture, lipids are not compensating for anything. They are the architecture. They define the colloidal structure that controls dispersion the moment the dose hits aqueous fluid, the rheology that suspends dispersed phases without separate viscosity modifiers, and the substrate that additional excipients integrate into. The same molecules — phospholipids, monoglycerides, fatty acid esters — that were once the supporting cast in a solid product become the load-bearing components of the delivery design.

This is the reframing the dissolution-era literature has been circling for two decades without quite naming. Lipid functionality moves from compensatory chemistry to performance-defining chemistry.

The Ibumix platform is one expression of that shift. It is built around the NSAID class. Ibuprofen and naproxen are the launch chassis — Ibumix and Naproxymix respectively — with the architecture designed to extend across other NSAIDs. Each NSAID is held in solution in a glyceryl caprate (GCC) lipid system that, in aqueous environments, self-assembles into a lyotropic liquid crystal mesophase (Mezzenga et al., 2019; Drummond & Fong, 1999). The NSAID does two things at once. It is the active pharmacological agent. And it is a component of the structured lipid architecture.

GCC itself is well-credentialed for this role. Glyceryl caprate — the monoglyceride of capric acid, a medium-chain fatty acid abundant in coconut and palm kernel oil — is FDA-GRAS and has decades of safe use in food and pharmaceutical excipients. It is a documented intestinal permeability enhancer (Lindmark et al., 1995; Aungst, 2000), which means it does work at the absorption site as well as inside the formulation. It has measured broad-spectrum antimicrobial activity (Thormar et al., 2006), which reduces the burden on separate preservative systems. And as a medium-chain monoglyceride, the lipid component itself is absorbed via the portal vein rather than the lymphatic chylomicron pathway — fast, predictable, and largely independent of bile. The architecture is built on a lipid the regulator already knows.

That architecture does two formulation jobs simultaneously. It solubilises the primary active — the NSAID — in the lipid phase. And it provides the yield-stress matrix that can carry additional insoluble APIs in suspension, without separate viscosity modifiers, without xanthan or cellulose derivatives bolted on after the fact. Two delivery modes — solubilisation and suspension — inside one system. The principle is not unprecedented: polysorbate-based formulation systems have done the same dual job for decades, serving as solubilisers in some products and as suspension stabilisers in others. What is new in Ibumix is the application of that principle to a structured lipid architecture engineered around a specific NSAID.

What this opens up is additive design that works with the lipid architecture rather than against it. Co-solubilisers, permeability enhancers, taste-modifiers, stability adjuvants — each can be selected for how it integrates into the structured phase, not how it survives next to it. The lipid is not the vehicle for the active ingredient. The lipid is the structural and functional design substrate. Several components can be carried inside it — some dissolved, some suspended.

This is what load-bearing chemistry means in practice. One lipid system, doing structural work, solubilisation work, suspension work, and absorption-architecture work simultaneously. It is not an excipient. It is the system.

What changes next

The regulatory surface shrinks. Dissolution testing does not disappear — additional APIs suspended in the structured matrix remain solids that have to disintegrate and dissolve in vivo. But they do so from a controlled starting state rather than a conventional liquid suspension, with all its lot-to-lot variability in shake-up, settling and shelf life. The release profile is narrower, the variability surface is smaller, and the in-vitro / in-vivo correlation is easier to defend. The regulatory burden is minimised, not eliminated.

Clinical profiles for the primary active become design choices. The NSAID enters the patient in solution, so its pharmacokinetics no longer depend on fed-vs-fasted state, gastric emptying, or polymorph stability. For additional suspended APIs, the variability is reduced rather than eliminated — they dissolve from a structured matrix rather than an uncontrolled one. Either way, inter-patient variability — the constraint on the therapeutic index of cyclosporine, tacrolimus, and a long list of others — becomes engineerable rather than inherited.

Late-stage commercial risk falls. The $250 million Norvir cost Abbott was not a one-off accident. It was the predictable consequence of building a high-value product on a delivery format whose key risk — polymorph-driven dissolution failure — was invisible until manufacturing scale-up. A structured lipid architecture does not remove that class of risk entirely. It reduces its surface area: the primary active is already in solution; suspended actives dissolve from a controlled starting state rather than an uncontrolled one.

Beyond the paradigm

The dissolution constraint was not invented by anyone. It is what you get when you choose to package a medicine as a solid object — a choice made in the 1840s for reasons that had to do with manufacturing efficiency and almost nothing to do with biology. Lipid science has done extraordinary work managing the consequences of that choice ever since. Neoral is a monument to it. Norvir is a warning about its ceiling.

The next chapter is not more of the same lipid science inside the same paradigm. It is lipid science put to work in a system that no longer requires it to fight the gut for the dose. That is what a structured lipid architecture makes possible. And it is why the most interesting drug-delivery question of the next decade is not how to dissolve a tablet faster.

It is what happens when dissolution stops being the variable that decides whether the medicine works.

— · —

Ibumix is developing a series of new liquid drug delivery platforms designed to remove the dissolution constraint from oral formulation. The platforms are built on glyceryl caprate (GCC), a medium-chain monoglyceride lipid chemistry, and are protected by a UK patent estate.

— · —

Sources

Part of the Ibumix series on lipid and liquid-crystal delivery. Start with the overview: Lipid and liquid-crystal drug delivery.