We thank Dr. Molitch (1) for his comments in this issue of Diabetes Care regarding our previous response (2) concerning the issue of insulin stability. Dr. Molitch points out apparent differences among manufacturers in our original letter with respect to in-use recommendations, and he requests actual data on insulin potency and that insulin be provided in smaller containers. We appreciate the opportunity to comment further on these important questions.

First, we will provide storage recommendations for the Lilly insulin formulations and containers available in the U.S. Second, we will explain further why the storage guidelines vary by formulation and type of container. Third, we will provide supporting data and explanations for our recommendations.

In-use recommendations for insulins in vials differ from those for insulin in cartridges or prefilled insulin pens. Recommendations for insulin products in vials are the same regardless of the formulation. However, the in-use recommendations may differ for that formulation when the same insulin is available in a cartridge. The primary reason for the different in-use dating recommendation between vials and cartridges is based on differences in the expected use of these products. For example, insulin in cartridges is expected to undergo more rigorous agitation and exposure to more widely varying temperatures by patients than insulin in vials. The in-use dating guidelines established by the Food and Drug Administration (FDA) for cartridges take into account the smaller volume and fewer total units of insulin compared with vials, as well as the potential exposure of insulin to greater heat and mechanical agitation.

The recommended in-use dating once an insulin vial has been punctured is 28 days for Humalog, Humalog Mix 75/25, Humulin R, Humulin L, Humulin N, Humulin U, Humulin 70/30, and Humulin 50/50. Humalog in cartridges and prefilled pens may also be used for up to 28 days. In contrast, the recommended in-use dating for all other insulins available in the U.S. in prefilled pens varies among formulations, based on a variety of end points in stability studies. For example, Humalog Mix 75/25 and Humulin 70/30 in prefilled pens can be used for up to 10 days, whereas Humulin N is recommended to be used within 14 days.

Dr. Molitch requests specific stability information over time for each insulin formulation manufactured by Eli Lilly, especially those insulins available in cartridges and thus having shorter in-use dating than the same insulin available in vials. Lilly’s in-use recommendations are based on the results of extensive laboratory testing. Because it is difficult to quantify how patients treat insulin, Lilly takes a conservative approach to in-use recommendations. We developed two automated physical stress tests to study the insulin formulations at conditions more extreme than the expected, typical patient usage. The details and results of this extensive testing have been published (3). These tests, developed with input from regulatory agencies, also help to determine adequate product dating during nonrefrigerated storage. The tests include exposing cartridges of insulin to temperature cycling and resuspension tests (TCRTs) in addition to high temperature and extreme agitation tests (HTEATs). Although the article cited (3) reported specifically on testing of Humulin R (solution samples), Humulin NPH, and Humulin 70/30 (suspension samples), these tests are applied consistently to all of our insulin solutions and suspensions, including the analogs.

The conditions of the TCRT include temperature cycling (25–37°C) in an incubator, with resuspensions conducted twice daily on a mechanical device outside the incubator. Cartridges are placed on a resuspension machine that rolls and inverts the cartridges; tests include three sets of 10 rolls and 10 inversions twice daily. At least 50 cartridges per lot are tested. The sample size will vary depending on the test end points. Three end points are assessed throughout the TCRT. These end points include visual assessment by trained operators, total insulin potency as measured by reverse-phase high-performance liquid chromatography, and acid clarification of pooled samples. The last test detects insulin aggregation and confirms the finding of the visual assessment. Insulin potency is no longer measured by in vivo bioassay. Instead, we determine the label claim amount of insulin per unit of volume, defined as potency, using high-performance liquid chromatography.

The conditions of the HTEAT include continuous high temperature (37°C), exposure in an incubator, and periodic daily agitation (30 rpm for 4 h) by rotation inside the incubator. In addition, cartridges are placed in the resuspension device after incubation but before visual assessment is conducted. This resuspension test performs the same rolls and inversions as the TCRT. At least 38 cartridges per lot are tested. Sample size will vary depending on the test end points. Visual assessment by trained operators occurs on days 0, 2, 5, 7, 9, 11, 14, 17, 21, and 28 to detect change in color and clarity for clear solutions and large aggregates (clumps) and/or material adhesion to the cartridge wall (frosting) for suspensions. Acid clarification is performed to confirm the visual changes.

The results of the physical stress tests vary depending on the formulation. Insulins in solution, such as insulin lispro and regular insulin, are not affected by the physical stress tests described above. For insulins in suspension, no visible changes are detected after exposure to the TCRT stress conditions. However, insulin suspensions exposed to the HTEAT stress conditions exhibit visible changes over the 28-day period. Four categories of visual change define the results of the test, including slightly grainy, grainy, clumpy, and severely clumpy. Any macroscopic visual changes detected are further evaluated by microscopic analysis. A micrograph by fluorescent backlighting is taken to detect clumping or frosting, which is then verified by optical microscopy. Insulins in suspension that show aggregated crystals before the end of the 28-day period are then acidified. A lack of dissolution after acidification confirms the presence of insoluble denatured insulin, which may indicate loss of potency.

It is important to note that changes in the visual appearance of insulins in suspension subjected to the HTEAT do not always correlate with reduced potency as measured by reversed-phase high-performance liquid chromatography. Based on the analytical criteria of the study, both the slightly grainy and grainy samples produce acceptable results. Samples designated as clumpy or severely clumped do not pass the criteria due to excessive variability in potency. In summary, the in-use dating recommendations for Lilly insulins are derived from the above analytic methods and results.

Eli Lilly must comply with product standards imposed by the U.S. FDA as well as with standards established by other regulatory agencies, e.g., The U.S. Pharmacopeia (USP), both in the U.S. and abroad. Regulatory agencies interpret the results of the TCRT and the HTEAT to guide nonrefrigerated storage dating considerations. As a result of different interpretations by various regulatory agencies, in-use dating periods for Lilly insulins range from 7 to 28 days, depending on the regulatory agency and formulation.

Like Dr. Molitch, many patients and other health care professionals have recommended that Lilly make a smaller insulin vial. Lilly must consider many factors when determining the size of insulin vials, such as average daily use. Another critical factor is maintaining compliance with existing standards. For example, until 15 May 1995, the USP specifically stated that U-100 insulin products should be packaged in 10-ml size bottles and that U-500 insulin products should be packaged in 20-ml size bottles (2). Developing insulin vials smaller than those currently available would benefit only a small minority of patients who take insulin. Importantly, creating smaller vials would require developing entirely new manufacturing facilities, much testing, and extensive regulatory review, all of which would ultimately increase the cost of insulin.

Lilly understands that labeling recommendations for insulin use may affect how health care professionals prescribe insulin and may have a financial impact on patients. We strive to help patients manage their diabetes by producing products that are safe, potent, and in compliance with regulatory agency requirements. Lilly appreciates the opportunity to provide more information about in-use dating recommendations for insulins. We hope that this response gives health care professionals and patients a better understanding of the reasons for these recommendations.

1.
Molitch ME: How long should insulin be used once a vial is started? Response to Grajower et al. (Letter).
Diabetes Care
27
:
1240
–1241,
2004
2.
Grajower M, Fraser CG, Holcombe JH, Daugherty ML, Harris WC, De Felippis MR, Santiago OM, Clark NG: How long should insulin be used once a vial is started? (Commentary).
Diabetes Care
26
:
2665
–2669,
2003
3.
Shnek D, Hostettler DL, Bell MA, Olinger JM, Frank BH: Physical stress testing of insulin suspensions and solutions.
J Pharm Sci
87
:
1459
–1465,
1998