High-Voltage Leak Detection of a Parenteral Proteinaceous Solution Product Packaged in Form-Fill-Seal Plastic Laminate Bags. Part 3. Chemical Stability and Visual Appearance of a Protein-Based Aqueous Solution for Injection as a Function of HVLD Exposure

Introduction

Part 1 of this research series documented the development and validation of six high-voltage leak detection (HVLD) method options for leak-testing plastic bags filled with an aqueous solution of the rapid acting insulin analogue, insulin aspart (NovoRapid®/NovoLog®) by Novo Nordisk A/S, Bagsværd, Denmark. Part 1 employed positive controls with laser-drilled holes in the bag film.

Part 2 found that HVLD is able to identify weaknesses or gaps in the bag seals as well. Part 2 also reported that package holes could be found even after allowing 21 days of ambient plus refrigerated storage for protein to clog and block leak paths. As discussed in Part 1 of this series, earlier research performed by Novo Nordisk A/S (not reported) found that product formulation readily clogged leak paths making them undetectable by vacuum decay, which lead to the investigation of HVLD as a container–closure integrity test. No tests were performed to prove that the leaks in the samples tested by HVLD post ambient and refrigerated storage were indeed clogged.

Testing packages for leaks while still cold from refrigerated storage could temporarily elevate HVLD results due to condensation on package surfaces; this effect can be avoided by testing bags dry and at ambient temperatures. Subtle material differences in plastic laminate film lots linked to alternative material sources were found to cause an upward shift in baseline HVLD voltage readings for no-defect packages. This highlighted the importance of testing multiple package lots from multiple sources when establishing HVLD test parameters, and to check HVLD test performance prior to package supplier approval.

This final Part 3 investigated the impact of HVLD exposure on the visual appearance and chemical stability of the packaged product. The product used is the rapid-acting insulin analogue, insulin aspart (NovoRapid®/NovoLog®) by Novo Nordisk A/S, Bagsværd, Denmark. In general, electrical conductivity and capacitance leak tests are considered nondestructive to product contained in HVLD-exposed packages. Published research has shown that in rare instances HVLD exposure has triggered headspace ozone formation in small volume vials causing product oxidation (1). Therefore, prudence dictated that the influence of HVLD exposure on the visual appearance and chemical stability of insulin products be verified prior to adopting HVLD as a nondestructive container closure integrity test.

(Note: In Part 1 of this series, method validation results indicated a small risk of falsely rejecting non-leaking packages exists. One way suggested to address a questionable fail result is to retest the package after instrument performance has been confirmed using a small set of negative controls. Support for such a product retest program would require long-term stability verification of product packages exposed to HVLD at least as many times as the retest program would allow. A retesting program was not adopted by Novo Nordisk A/S; therefore, the stability studies described in Part 3 study did not include samples exposed to HLVD more than twice.)

Part 3 research, including HVLD tests, product stability storage, and visual appearance, was performed by Novo Nordisk A/S of Denmark. The HVLD instrument is a Nikka Densok HDT-1 pinhole inspector (Lakewood, CO) with data acquisition system by TellTech Services Corporation (Mahopac, NY). Dana Guazzo of RxPax, LLC (Bridgewater, NJ) provided consult support and assisted in publication preparation. All work was fully subsidized by Novo Nordisk A/S.