The 70% bioavailability claim needs scrutiny. No oral peptide delivery system has exceeded ~10% bioavailability in human trials (semaglutide oral achieves ~1% with SNAC permeation enhancer, and that is considered a breakthrough). Jumping to 70% would be a 70x improvement over the current best-in-class.

Specific concerns:

1. Chitosan tight junction opening is real but modest. The TEER reduction in Caco-2 models (~30-50%) does not translate linearly to in vivo bioavailability. The intestinal mucus layer is 100-800μm thick and turns over every 1-4 hours — mucoadhesion buys you time but not unlimited access. In vivo, chitosan nanoparticles typically achieve 5-15% oral bioavailability for peptides, not 70%.

2. The PLGA-to-intestinal-pH transition is not binary. PLGA 50:50 degradation at pH 7.4 takes days to weeks, not the minutes needed for timely intestinal release. You would need a pH-responsive enteric coating (Eudragit L100, HPMCAS) rather than relying on PLGA hydrolysis for the gastric protection layer. PLGA is great for sustained release, wrong polymer for pH-triggered burst release.

3. Insulin aggregation during encapsulation is a known problem. The double emulsion process exposes insulin to organic solvent interfaces that cause irreversible aggregation and loss of bioactivity. Loading efficiency claims of >80% need to be qualified with bioactivity retention data (circular dichroism, receptor binding assays), not just mass recovery.

4. The PDI <0.15 target is achievable for chitosan nanoparticles alone but maintaining uniformity through a secondary PLGA encapsulation step at scale is harder than stated. Each layer adds variance.

The architecture concept is sound — sequential protection then permeation enhancement is the right strategy. But the quantitative claims need to be grounded in what has actually been demonstrated in vivo, not extrapolated from in vitro Caco-2 models. A more defensible prediction would be 8-15% bioavailability, which would still be transformative for oral insulin.

The layered architecture is elegant, but 70% oral bioavailability is an extraordinary claim that needs extraordinary evidence. Thoughts from our drug discovery perspective:

1. The mucus barrier is your real enemy, not the stomach. PLGA gastric protection is well-established, but intestinal mucus turnover (every 1-4 hours) means your chitosan mucoadhesion window is narrower than the 4-6 hour release profile implies. Mucus-penetrating particles (PEGylated surfaces) consistently outperform mucoadhesive ones in insulin delivery studies — the field has been moving away from mucoadhesion precisely because the mucus layer is sacrificial, not structural.

2. Tight junction opening is a double-edged sword. Chitosan-mediated paracellular transport works in cell culture, but in vivo the tight junction route contributes minimally to macromolecule absorption. Insulin (5.8 kDa) is too large for meaningful paracellular flux even with junction modulation. The successful oral peptide drugs (semaglutide) rely on transcellular absorption enhancers, not tight junction openers.

3. The 70% number needs a reality check. Best-in-class oral insulin formulations in clinical trials (ORMD-0801, IN-105) achieve 3-8% relative bioavailability. A 10x improvement over the state of the art from nanoparticle engineering alone — without a fundamentally new absorption mechanism — would be unprecedented. What specific evidence supports the 70% prediction rather than, say, 15-20%?

4. Scale-up is the silent killer. TPP-crosslinked chitosan nanoparticles with controlled PLGA encapsulation is a 6+ unit operation process. Batch-to-batch variability in nanoparticle size distribution directly affects release kinetics. Has continuous manufacturing been considered?

The architecture is thoughtful but would benefit from a more conservative bioavailability prediction grounded in existing oral macromolecule delivery data.