I think you may apply baseline correction procedure and for all endogenous substance you can use baseline correction procedure. 

Measure C peptide and insulin both - endogenous insulin will produce C-peptide along with insulin whereas cloned, purified exogenous insulin will not include C-peptide.

[In the olden days when porcine insulin use was prevalent, species specific insulin differences in sequence could be utilized to make the distinction. These days all insulin is cloned human version. However, in some cases, depending on the version used, a mutation or two that alter half-life of insulin may still offer a similar paradigm for distinction.]

Unfortunately, its not as simple as correcting for baseline. In most cases you can't correct for endogenous insulin. There are ways around it in healthy volunteers but it can't be done post-hoc (i.e., you have to design the study correctly).  The EMA guidance has some good info:

http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2012/12/WC500136392.pdf 

Thank you all for your answers. I will share my experience once study is conducted end of this month.

Rizwan, it is true that a simple insulin measurement before and after exogenous insulin administration is not enough to give you the proportion for endogenous versus exogenous insulin in serum. This is why you have to have an ability to distinguish between the two sources of insulin. One of the reasons for significant difference is due to the very nature how islets normally function - increase in insulin level causes islets to respond by further increasing glucose stimulated insulin output. If you inject exogenous insulin, at or soon after glucose challenge, the level of endogenous insulin will go up significantly higher than it might have, had you not injected insulin. The increase can be as much as 40-50%.

The above can be easily demonstrated if you were to use an immunologically distinct version of insulin, such as the mutant B52-Asp version (with much higher circulating half-life).