CPSA Digest 2001

Poster Session Summary

All large and midsize pharma have programs that have ramped in the past 3-5 years. Current goal is to kill compounds only if very early and primarily to build desirable properties into potential leads.

Analyses are still underway to generate better in-silico methods. In-silico will never replace analysis totally, just reduce numbers from 1000's/week to 100's/week. While molecule is still on paper, in-silico can provide solubility, logP, pKa, permeability, potential for metabolism and toxicity.

LC/MS is still primary analytical method but throughput is a concern so analysts are looking to plate reader type assays (fluorescence). Issues remain with sensitivity and specificity outside of MS. Staggered elution zone systems, dual column on-line SPE, using serial or parallel extraction columns are main methods of increasing LCMS throughput.

Chromatography Future
Need higher throughput but without loss of selectivity and sensitivity. Fast gradient elution takes advantage of higher peak capacity and speed so everyone wants to run at higher flow rates.

Monolith columns enable that. Discovery is embracing those columns. However, solvent consumption at ˜6 ml/min is a concern so demand for 1 and 2 mm columns is huge but Merck is having a hard time developing smaller ID's than 4.6 mm. Word was it would be more than one year until the columns are available.

Existing LC systems (flows up to 5 or 10 ml/min) would benefit from higher backpressure spec. (˜8000 psi).

Autosampler speed and carryover still are still primary detractors to HT. Staggered elution zone systems, dual column on-line SPE, using serial or parallel extraction columns are main methods of increasing LCMS throughput.

Plasma Sample Preparation
This discussion centered around two posters which described the on-line extraction of small proteins or peptides from plasma samples. This is an interesting topic in that the standard solid-phase extraction methodology involves removal of the plasma proteins and salts prior to eluting the organic compound of interest. In this instance the organic of interest is a protein and not an easy task. The consensus was that no really good methods exists with the exception of that which was presented during the poster session. There was a suggestion that direct injection of the plasma on-column using MS/MS is a reasonable alternative given the inevitable fact that the column will eventually clog.

Predictive Pharmaceutical Profiling
A discussion that centered around the comparison of in-vitro predictability vs. in-vivo observations was initiated by a poster which presented in-vitro data which could be misinterpreted. This poster suggested that data suggesting CYP450 3A4 oxidation was actually to be attributed to a little known 1A2 oxidation. The question was therefore asked as to whether Predictive Profiles were in fact very predictive of in-vivo observations. In the end, a consensus was reached that while in many cases the actual in-vitro data might not match in-vivo assays the profiles were still a very good means by which a series of compounds could be evaluated relative to another series or within the series being analyzed. In this era of combinatorial and parallel synthesis, these profiles still give the scientist a way to evaluate the potential "drugability" of a series of compounds early in discovery.

Return to Proceedings »