Glargine versus regular insulin protocol in feline diabetic ketoacidosis

Abstract Objectives To determine whether basal‐bolus administration of glargine insulin is a safe and effective alternative treatment compared to the standard continuous rate infusion (CRI) protocol. Design Prospective randomized clinical trial. Setting University teaching hospital. Animals Twenty cats diagnosed with diabetic ketoacidosis (DKA). Interventions The cats were block‐randomized to either a CRI protocol using regular insulin (CRI‐group; n = 10) or a basal‐bolus SC and IM glargine protocol (glargine‐group, n = 10). Baseline blood gases, electrolytes, glucose, and β‐hydroxybutyrate (β‐OHB) concentrations were measured at the time of admission and later at predefined intervals until reaching the primary endpoint of the study, defined as a β‐hydroxybutyrate concentration < 2.55 mmol/L. Measurements and main results The main outcome measure was time (h) to resolution of ketonemia. Secondary outcome measures were time until first improvement of hyperglycemia and ketonemia, decrease of glucose to ≤13.9 mmol/L (250 mg/dL), resolution of acidosis, consumption of first meal, and discharge from hospital. Additionally, occurrence of treatment‐associated adverse events and death were compared. Seventeen cats (85%) survived to discharge, with no difference in survival between groups (P = 1.0). Median times to β‐OHB < 2.55 mmol/L were 42 (CRI‐group) and 30 (glargine‐group) hours, respectively (P = 0.114). Median times to first improvement of hyperglycemia (glargine‐group: 2 h; CRI‐group: 6 h; P = 0.018) and until discharge from hospital (glargine‐group: 140 h; CRI‐group: 174 h; P = 0.033) were significantly shorter in the glargine‐group. No significant differences were observed in any other parameter under investigation (P > 0.05). Conclusions Basal‐bolus administration of glargine insulin appears to be an effective and safe alternative to the current standard CRI‐protocol for the management of DKA in cats. The positive outcomes and simplicity make it a viable option for the treatment of feline DKA.


INTRODUCTION
Diabetic ketoacidosis (DKA) is an acute, life-threatening metabolic complication of feline diabetes mellitus (DM), and mortality rates range from 17 to 50%. [1][2][3][4][5][6] Despite this, studies are scarce, and current treatment recommendations are mostly extrapolated from experiences in people and dogs. The 3 cornerstones of therapy are rehydration, correction of electrolyte imbalances, and insulin administration to reduce ketone and glucose concentrations. Published insulin treatment regimens include intermittent IM administration of regular insulin 2,7 or glargine, 8 continuous rate infusion of regular insulin 1,3,5,9 or lispro, 1,6 the combinations of SC glargine with intermittent IM regular insulin 5 or the combination of SC and IM glargine. 8 The relatively small number of patients included in the studies and limited comparisons between different protocols precludes evidence-based recommendations regarding which protocol is superior, and additional research is required. The use of the insulin analog glargine is of special interest as it is considered an insulin of first choice in diabetic cats, 10,11 has a long half-life when injected subcutaneously, 12 and elicits comparable effects to regular insulin when given intravenously 13,14 or intramuscularly 15 in people. In a retrospective study of 15 cats with DKA, SC glargine administration was started immediately after the diagnosis of DKA and combined with IM glargine injections as required. 8 As the survival rate was 100% and hospitalization time was short, this basal-bolus treatment was considered a feasible alternative to traditional treatment regimens. Several studies in ketoacidotic diabetic human patients have shown that IM 17 or even SC 18,19 bolus treatment protocols are safe and an effective treatment option. The objective of this prospective block-randomized study was to compare the basal-bolus glargine regimen with a regular IV insulin protocol and to demonstrate its utility in the treatment of feline DKA.

Case selection
Client-owned cats with naturally occurring DKA admitted to the Uni- blood β-hydroxybutyrate (β-OHB) concentration > 2.55 mmol/L, 22 and a metabolic high anion gap (AG) acidosis (pH < 7.27 and an AG > 20.6 mmol/L), 23 29 rounded to the next half or whole unit. The primary aim in both protocols was to achieve a blood glucose concentration < 13.9 mmol/L (250 mg/dL). At that point, intensive insulin therapy was continued at 10 mL/h, but IV fluids were changed to a 2.5 to 5% glucose-containing solution and adjusted to keep blood glucose concentration between 10 and 13.9 mmol/L (180 and 250 mg/dL). Based on recommendations in people, 19 insulin administration was only decreased (to 5 mL/h) or stopped (at the discretion of the clinician in charge) in cats with glucose < 4.44 mmol/L (80 mg/dL), despite 5% glucose infusions at the calculated infusion rate (see above).

Therapy and monitoring
No glucose was added to the insulin infusion. In cats with inappetence for more than 3 days, a nasoesophageal feeding tube was placed, and the cats were fed 4 times/d with a liquid diet k using standard estimation of resting energy requirements (70 x [body weight in kg] 0.75 ).
When the cats were appropriately hydrated, started to eat spontaneously, and β-OHB was < 2.55 mmol/L, the cats in both groups were changed to an exclusively SC insulin regimen (glargine q 12 h with 0.25-0.5 units/kg [based on estimated ideal body weight and blood glucose concentration]). 27 The primary aim of the study was to compare time to plasma β- The study protocol was approved by the Austrian Ethical and Animal Welfare Committee (BMWF-68.205/0004-II/3b/2014).

Statistical methods
All analyses were performed using statistical software . m Due to the small number of cats in both groups, and because many data were not normally distributed (visual evaluation and Kolmogorov-Smirnov test), all variables are presented as median (range), and data were compared between the groups using the non-parametric Mann-Whitney U-test.
Categorial variables were compared using Fisher's exact test. Kaplan-Meier plots were generated and evaluated with the log-rank test. Cats that did not achieve a specific outcome (eg, discharge from hospital, consumption of first meal, or glucose < 13.

RESULTS
In the study period, 37 cats with diabetes mellitus and clinical suspicion for ketoacidosis were evaluated. Seventeen cats were excluded because of a pH > 7.27 (n = 9), congestive heart failure (n = 4), and stage 4 kidney failure (n = 2). Post hoc exclusions occurred because of initiation of the wrong protocol (n = 1) and variation of the protocol (n = 1). The remaining cats (n = 20) that met all criteria for inclusion were block-randomized to 1 of the 2 groups (10 cats per group). Except for 1 Manx (CRI-group) and 1 Russian Blue cat (glargine-group), all cats were domestic shorthair cats. Demographic characteristics including age, sex, weight, body condition score, and baseline variables were not significantly different between the groups (Table 2). There were 6 and 3 cats with newly diagnosed and untreated DM in the CRI and glarginegroup, respectively (P = 0.37). One cat in each group had been pretreated with prednisolone, 2 cats in the CRI-group and 4 cats in the glargine-group were deemed fractious (P = 0.629). All but 1 cat (score 2) in the CRI-group and 3 cats (all score 2) in the glargine-group had a mentation score of 1 at presentation (P = 0.582).
One cat in the glargine-group with hepatic failure and bacterial cystitis was euthanized due to financial constraints on day 2 before reaching the primary endpoint of the study. Histopathology was not performed.
In the CRI-group, 1 cat died on day 7, likely as a consequence of liver failure (icterus, alanine aminotransferase [ALT], 1000 U/L), and the owners elected euthanasia on day 5 for another cat with worsening mentation (mentation score 3). Postmortem examination was not performed in the first cat, and histopathology in the second cat revealed severe multifocal purulent myocarditis, pancreatitis, insulitis, focal hepatic necrosis, and a thyroid adenoma.
One cat in the CRI-group received 1 bolus, and 1 cat in the glarginegroup received 2 boluses of isotonic crystalloids before initiating CRI.
The cat in the CRI-group additionally received hydroxyethyl starch (2 mL/kg over 5 minutes) for volume resuscitation. Although 2 cats in the glargine-group had a pH < 7.00 (pH = 6.94 and 6.99) at presentation, no cat had a pH < 6.9, which was the criteria for mandatory bicarbonate therapy. Both cats survived to discharge without bicarbonate administration. One cat in the glargine-group received magnesium sulfate. A nasoesophageal feeding tube was placed in 8 cats (CRIgroup, n = 5; glargine group, n = 3; P = 0.65). Two cats in the CRI-group developed furosemide-responsive respiratory distress and hydrothorax, likely caused by overhydration. Eight (CRI-group) and 6 (glarginegroup) cats were treated with antibiotics, respectively (P = 0.629).
No differences were observed in any other parameter under observation ( Table 2) (Table 3). In addition to the first IM insulin injection in the glargine-group, a median (range) of 3 (0-11) additional IM glargine injections were administered.

DISCUSSION
Comparable to the situation in animals, the ideal type of insulin and the optimal route of insulin administration is a matter of debate in human patients with DKA. 18   suggest that this concern is unwarranted. 36,38,39 It has been demonstrated in cats with diabetic ketosis that despite significant decreases of blood glucose concentrations > 120% within the first 72 hours of conventional therapy, sodium, the major determinant of serum tonicity, increases and effective osmolality stays relatively constant, minimizing large osmotic shifts. 39  with metabolic acidosis in cats, 22 and DKA was not diagnosed at values < 3.8 mmol/L in another study. 43 With the resolution of ketonemia to < 2.55 mmol/L, most of the cats in our study began to eat spontaneously and were transitioned to SC glargine alone. The results of blood gas analysis did not influence insulin dose in any cat. and 30  hours, compared to about 11 to 12 hours in people, 46,47 respectively. Further, a decrease of β-OHB by more than 1 mmol/L/h, which suggests adequacy of therapy in people, 44  were found in the second. 9 Neither of the studies compared time until resolution of ketoacidosis or ketonemia.
A topic of debate in human medicine is whether to routinely start SC administration of a long-acting basal insulin (eg, glargine insulin) at the onset of DKA management. The rationale is to provide stable background insulin concentration and to avoid reoccurrence of hyperglycemia during the transition time to SC insulin. 18,48 A common concern raised with SC insulin administration in dehydrated patients is SC insulin accumulation and sudden release after rehydration. 49 A meta-analysis of 4 studies in human patients suggests that this concern is likely unwarranted. The addition of SC glargine insulin to standard protocols using IV infusion of regular insulin significantly decreased the time to resolution of DKA, without increasing the risk of adverse events. 50  Further, like Normosol-R, which is unavailable in many countries, it is calcium free, allowing the addition of phosphorus-containing solutions.
In contrast to human patients where the incidence of hyperchloremia increases markedly within the first 20 hours of infusion therapy, 53 hyperchloremia was rarely observed in our study. The most likely explanation is that 85% of our cats were hypochloremic at presentation, whereas this seems to be less common in ketoacidotic human patients. 54  It is possible that the "refeeding syndrome," which has been associated with hepatic lipidosis and characterized by the development of severe hypophosphatemia, 55 was a causative factor for the development of anemia in some of the cats in the current study.
Several important limitations must be recognized when interpreting the results of our study. The major limitation was that due to stringent criteria for inclusion, the number of cats included was lower than initially planned. A sample size calculation, assuming a power of 80% and a 5% significance level, revealed that a minimum of 37 cats in each treatment group would have been needed to show a 12-hour difference until resolution of ketonemia. In the future, multisite studies are recommended to provide sufficient power. Other limitations of the study are that the patient population described may not be representative for all cats with DKA, that insulin and infusion requirements and final costs were not compared, that bias may have been introduced by missing data, that it was not possible to definitely diagnose all concurrent disorders, that calcium and magnesium were measured merely in selected cases, that not all cats achieved the measured endpoints, and that it is possible that greater familiarity with the CRI-protocol, the standard protocol in our clinic before the study, resulted in more proficient implementation than the glargine-protocol.
In conclusion, although the study was underpowered to detect differences in time until resolution of ketonemia, the results suggest that the basal-bolus administration of insulin glargine is a useful alternative to the current standard CRI-protocol for the management of DKA in cats. It is simple and associated with a shorter time to first improvement of hyperglycemia and decreased hospitalization time.
Additionally, β-OHB measurements using hand-held ketone meters are a useful adjunct for monitoring response to therapy and eliminate the need for frequent blood gas analyses. Further studies are required to evaluate the benefits and disadvantages of IM bolus vs CRI protocols and to compare choices of fluid therapy in management of feline DKA.