" <;; %" ' Îą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s mentioned, saline and phosphate buffered saline are the most commonly used vehicles for formulating or diluting drugs for direct CNS delivery as well as for flushing the delivery system before and after dose administration. We have discovered that small differences in buffer concentration and pH have a very large impact on in vivo safety and tolerability of the administered solution. A preliminary study in adult cynomulgus monkeys was conducted to evaluate the toxicology and safety pharmacology of repeated IT-lumbar doses of our enzyme. Each animal was implanted with a port-catheter system to facilitate an every-other-week dosing regimen. The device control animals received phosphate buffered saline at pH 7.2. The vehiclecontrol group was dosed with an aqueous solution of 20 mM sodium phosphate, 130 mM NaCl, and 0.005% polysorbate 20 at pH 7.5. This formulation was evaluated as it rendered adequate solubility and stability to the protein. Clinical signs were observed during and immediately after dosing; the incidence was comparable between the control groups (device control and/or vehicle-dosed group) and enzyme-dosed groups, with no evidence of a dose response. Consequently, the study was terminated after the second dose. A representative view of the histopathology is given in Figure 8. These clinical observations, which occurred in all animals including the vehicle-dosed animals, prompted a series of toxicology studies Figure 5: Examples of Clinical and Investigational Intrathecal Formulations balance of CSF composition and intracranial pressure. Thus, without removal of CSF, the dose volume is limited to â‰¤3 mL in humans (and â‰¤1 mL in the adult cynomulgus monkey). Dose volume limitation necessitates high concentration protein formulation (>10 mg/mL) when doses are in the several tens of milligrams. Several factors can impact protein solubility to achieve the desired concentration, including ionic strength, amino acid sequence and other co-solubilising agents. Solution compositions used routinely for CNS administration are isotonic saline (unbuffered) or Elliottâ€™s B solution (artificial CSF) with a composition listed in Figure 3. Isotonic solutions may not render adequate solubility for some proteins. Additionally, Elliottâ€™s B solution contains a very low buffer concentration that may not provide adequate buffering capacity needed to stabilise protein formulations during long-term storage. The artificial CSF solution also contains various salts which are often not compatible with protein formulations. For example, calcium salts may mediate protein precipitation. 18 FORMULATIONS Jun 2010.indd 18 The most common approved CNS bolus formulation composition is saline (150 mM NaCl in water), as shown in Figure 4. Others have been tested (Figure 5), some with poor safety profiles. Proteins typically require controlled pH and specific excipients for their solubility and stability (see Figure 6), so we systematically investigated formulation components that would be suitable for our proteins for direct administration to the CSF. Our initial investigation was targeted to a lysosomal enzyme that required a minimum of # ( )!( ("(& '& *& # ! ! $ T . ! " Figure 6: Most Common Parenteral Protein Formulation Constituents www.ondrugdelivery.com Copyright ÂŠ 2010 Frederick Furness Publishing 2/7/10 09:01:11