Interference between excipients and other components is a perennial problem in formulation analysis. In one mAb formulation encountered at Covance, Tween co-eluted with one of the product aggregate peaks. It was subsequently found that the Tween absorbance was below detection limits for this particular formulation. Had Tween been present at a higher concentration, the analytic method may have required modification, for example by monitoring protein concentration at wavelengths other than 280 nm, which would have likely led to reduced sensitivity.
One seemingly small alteration in formulation can introduce significant changes in analytic method. Recently, a trimeric cytokine fusion protein submitted by a Covance client, appeared to be unstable above -80°C, which necessitated production and storage as a lyophilized product. This change, in turn, required a switch from a standard biconchoninic acid (BCA) assay (the protein’s extinction coefficient was unknown) to a microBCA assay.
A number of proteins conjugated to polyethylene glycol (PEG) have been tested. Assays for these molecules must include assessment of unconjugated, free PEG within the formulated drug product. Since PEG lacks a UV chromophore, until recently the most common method employed was size-exclusion HPLC with refractive index (RI) detection. The size-exclusion method was necessary since it employs an isocratic gradient that provides a stable signal for the RI detector. The sensitivity of this method is also limited, which restricts its applicability as an impurity assay. In addition, this method is naturally unsuitable where separation is not achieved by SE-HPLC.
Covance is currently evaluating a gradient HPLC method for separating free-PEG and PEG-protein conjugates that incorporates a charged aerosol detection, which has a greater sensitivity and should have greater applicability for PEGylated proteins.
Timing, thoroughness, simplicity, and knowledge are critical components of formulations, particularly with respect to the ongoing analytical work conducted during preclinical and clinical development.
While small molecule drugs rarely fail due to formulation problems, the same cannot be said for biopharmaceuticals. The cytokine fusion protein mentioned above would not have been approved had it not been reformulated in a lyophilized form. Stability failures can occur as late as Phase II, after tens of millions of dollars have been invested, and the prospect of re-formulating is a costly one.
With formulation, the more effort invested up front the better. Establishing analyses appropriate to both formulation and stability-determining assays before entering human studies is ideal since these drugs are unlikely to experience formulation-related issues. The most efficient way to do this is to perform a formulation matrix construction and develop assays to circumvent interference from excipients.
A thorough understanding of one’s biological agent, route of administration, and pharmacokinetics is indispensible for a successful formulation program. Within that framework, the aim is to arrive at the simplest formulation that satisfies requirements of stability, activity/potency, and analytics. This is not always easy to achieve as formulations might contain excipients that are suitable for use in the laboratory but not for use in humans, that interfere with common analytical methods, or that do not support long-term stability studies.
It is no exaggeration to say that for practical purposes formulation carries approximately the same significance for the commercial success of a biopharmaceutical as the drug substance itself.