Review Papers Against Their Own Results

 Quick Explanation Copied

 Long Explanation

Core Finding

Di et al. modified the original PAMPA (Kansy et al.) by coating the filter membrane with porcine polar brain lipid (PBL) in dodecane instead of the phosphatidylcholine lipid used for intestinal (PML) models. On a 30-drug development set, CNS+ compounds generally showed Pe >4.0×10⁻⁶ cm/s and CNS− compounds Pe <2.0×10⁻⁶ cm/s; the five misclassified outliers (caffeine, corticosterone, loperamide, astemizole, terfenadine) were all known to undergo active influx, P-gp efflux, or rapid metabolism. An 11-experiment factorial design found 22 mg/mL PBL, 4 µL volume optimal; the validated 14-compound Wyeth set was classified 100% correctly Original PAMPA-BBB methods and validation.

Classification thresholds (CNS+ >4.0; CNS− <2.0) were derived empirically and then validated. Astemizole, loperamide, terfenadine, clonidine (borderline), oxazepam and caffeine fall in or above the uncertain zone.

External Validation & Influence

Independent head-to-head testing across 37 drugs found PAMPA-BBB correlated significantly with in situ brain perfusion while MDR1-MDCKII did not PAMPA-BBB validated against brain perfusion. The assay was subsequently extended to natural-product CNS discovery (1,760 plant extracts) PAMPA-BBB applied to natural products and to brain-penetrant antioxidant screening, where the authors explicitly note its passive-diffusion limitation PAMPA-BBB extension to antioxidant screening.

Critical Blind Spots

• Classification circularity: the development set mixed four classification methods (kinetic rate, steady-state distribution, pharmacology, AUC-based B/P); the authors admit equivalence was assumed, not proven Acknowledged classification assumption.
• Small, proprietary validation set: only 14 Wyeth compounds, all apparently passive-diffusion dominated; no public data.
• Threshold derivation: CNS+/− boundaries were set from the same data used to evaluate the assay — an internal overfitting risk.
• Species/membrane mismatch: porcine brain lipid in a non-bilayer dodecane film cannot reproduce tight-junction geometry or transporter activity; the authors state this openly but quantify neither reproducibility (SD) nor inter-day variability.

Verdict

Reported: a robust, cheap, 96-well passive-permeability screen. Inferred: the method's real value is as a front-end triage filter, exactly as the authors recommend pairing it with Caco-2 P-gp assays — not as a standalone BBB predictor. Uncertain: generalizability to poorly soluble, very lipophilic, or transporter-substrate chemotypes; UV detection may also fail for non-chromophoric or impure samples. Confidence: moderate — strong external replication exists, but the original validation is small and the threshold derivation is partially circular.

 BGPT Paper Review

Study Novelty

First published adaptation of PAMPA specifically for BBB using porcine brain lipid; built on Kansy's 1998 PAMPA concept but introduced a brain-lipid membrane and empirical CNS classification ranges that became an industry standard.

Scientific Quality

Well-designed DOE optimization, QC standards, and honest treatment of outliers raise quality; deductions for circular threshold derivation, mixed classification criteria acknowledged but unresolved, no reported SDs/replicate statistics, and a small proprietary validation set.

Study Generality

Applies broadly across drug-like small molecules for passive permeability, but is deliberately scoped to passive diffusion and excluded transporter/metabolism effects; not a general BBB model.

Study Usefulness

Became a routine, low-cost front-end screen in industry and was later validated against in situ brain perfusion and extended to natural products; the CNS+/CNS− Pe ranges remain widely cited.

Study Reproducibility

Methods, lipid source (Avanti catalog no.), volumes, pH, and software are clearly described and materials are commercially available; however, no replicate SDs or inter-lab data are given and the Wyeth validation compounds are not identified.

Explanatory Depth

Mechanistic reasoning is sound for passive diffusion (outliers mapped to known transporters/metabolism), but the paper does not characterize membrane structure, unstirred water layer effects, or quantitative structure-permeability relationships.