In Reply: “On Insulin Kinetics Following High-Dose Insulin Therapy, and When to Stop Therapy”

Dear Editor: We appreciate the interest in our case report and the thoughtful comments in the letter from Dr. Gawedski and Dr. Paloucek [1]. We are thankful for the opportunity to further discuss our limitations.

In considering the elimination kinetics of insulin in our case, we agree that renal impairment was contributory. We disagree, however, that 80% of insulin clearance is renal. This appears to be a mis-statement in Rabkin et al. when citing Ferannini et al., who reported that 80% of the extra-hepatic metabolism of insulin under hyperinsulinemia-euglycemia clamp conditions was via the kidneys, accounting for approximately one-third of total metabolism [2, 3]. The remaining sources cited by Rabkin et al. note the kidneys metabolize about 15–40% of intravenous insulin [3–5]. Additionally, we disagree that diminished renal function in our patient undermines our main conclusion—that the half-life of insulin is substantially prolonged after an infusion of 10 U/kg/hour. Our patient’s nadir eGFR was 28 mL/min; it returned to normal (creatinine 1.05 mg/dL, eGFR 79 mL/min) 24 hours later, approximately 10 hours post HDI discontinuation. Although abnormal, this would not fully explain a change in half-life from insulin’s usual 6–9 minutes to the calculated 10.6–18.7 hours [1, 6, 7]. Furthermore, AKI is common in ill patients requiring HDI; a review of 199 cases of beta-blocker/calcium-channel blocker poisonings treated with HDI showed 29% of patients developed an elevated creatinine [8]. Thus our case likely reflects common real-world experience where the exact contribution of renal dysfunction to insulin clearance is unknown.

Additionally, our findings are largely consistent with published insulin overdoses. These data must be considered cautiously as insulin administered by either the subcutaneous or intramuscular routes demonstrates different absorption kinetics than insulin administered via intravenous route; nevertheless, in overdose scenarios half-lives are frequently prolonged, ranging from 1 to 12 hours rather than minutes [9–11].

Unfortunately, it was beyond our capacity to directly measure hepatic and renal clearance of insulin or its volume of distribution. We thus reported half-life, consistent with prior authors [9–11]. In an attempt to further ameliorate this issue, we also presented a table of individual serum insulin concentrations to allow transparency for readers. In discussing the approach to down-titrating insulin, the authors of the letter state that “... a safer general approach would be to titrate as tolerated...” On this point, we respectfully disagree. We are aware of no evidence base supporting an optimal approach to discontinuing HDI. Lacking prospectively derived evidence for HDI discontinuation, we have only expert opinion and a limited number of pharmacokinetics studies on which to base decisions. Available evidence suggests that insulin likely titrates itself due to prolonged elimination kinetics under supraphysiologic dosing [9–11]. Further, although the authors state it would be safer to titrate insulin as needed, they provide no evidentiary support for this claim, failing to acknowledge that prolonging unnecessary insulin infusions may lead to iatrogenic harms including volume overload, hypoglycemia, and hypokalemia, which in a bradycardic patient could precipitate torsades des pointes. Our experience with HDI suggests patients typically do well regardless of the insulin discontinuation strategy employed; as such, we strive to shorten HDI duration as much as possible [8, 12]. We are unaware of any group that titrates insulin as slowly as would be logically indicated by the elimination kinetics under this intense dosing.

Finally, addressing the concern about discontinuing dextrose after 24–48 hours, we agree that we may not have described the origin of this recommendation adequately. We suspect the clinical team continued the dextrose infusion for longer than necessary because of their awareness of the elevated serum insulin levels. In retrospect, as there were no episodes of hypoglycemia, we suspect this was ultimately unnecessary.

Sources of Funding

There are no sources of funding.

Declarations

Conflict of Interest

There are no conflicts of interest.