Response to Dr. Deluzio's Letter to the Editor

Dear Editor:

I am writing in response to the letter ¹ of Dr. Tony Deluzio, Senior Medical Director, Medical Affairs-Diabetes, Novo Nordisk, Inc., commenting on my review ² published in a 2011 supplement issue of the journal. (Please note the corrected Figs. 1 and 2 also published in the journal in 2011. ³ )

With regard to Dr. Deluzio's point on the “action profiles of biphasic (premixed) insulins …,” the publications peak on biphasic insulin aspart 70/30 was stated as 1–4 h. ² Apparently, you have information that refutes that, and the peak you quote is 2.4 h. You also point out that insulin detemir is “inaccurately represented where it should follow a similar glucose. …” Finally, you bring to my attention the fact that since this article ² was written for type 2 diabetes mellitus, I should reference type 2 diabetes action profiles reflecting that patient population, instead of the cited type 1 diabetes population.

As to the detemir statement that there were inaccuracies using the study by Klein et al., ⁴ I would share a few thoughts. First, as you are aware, the study of the insulin action profiles for insulins using clamp techniques involves methods that vary and do not always reproduce equally. The reason clamp studies involving patients with type 1 diabetes are preferred as the best way to show the pure insulin action profile is because it is generally accepted that in this model you have less interference of variables that will have great impact on the glucose infusion curves, such as endogenous insulin production and insulin resistance, which are more of an issue when dealing with type 2 diabetes mellitus. ⁵

I appreciated your making me aware of the pharmacodynamics data on type 2 diabetes mellitus patients quoting the articles by Klein et al. ⁴ and Bilz et al. ⁶ ; however, there is another study that looked at this issue. This was published in Diabetes Care by Bolli et al., ⁷ who concluded that insulin glargine exhibited greater metabolic activity when looking at the glucose infusion rate from 0 to 32 h and even when looking at other markers of activity showing consistent decrease in activity with detemir and NPH. These data also correlate well with the data seen in type 1 diabetes mellitus patients. ⁸ In the face of inconsistent results, it is preferable to use type 1 diabetes mellitus clamp models because of lack of standardization of the clamp techniques.

I also mention two tables from the book edited by Leahy and Cefalu. ⁹ (Table 1 is originally Table 2 in the chapter by Wittlin et al., ¹⁰ and Table 2 is originally Table 1 in the chapter by Leahy. ¹¹ ) These charts again show the action profile variabilities that we see in various publications, raising the question, who are we to believe?