DOSING AND SELF-TITRATION

Initiating mealtime insulin

AUTONOMY demonstrated that patient self-titration using Humalog® U-100 KwikPen® can be effective1,2

Thought leaders discuss the AUTONOMY study

These short videos, featuring well-respected leaders from the diabetes community, provide an insightful look at the AUTONOMY study, the study’s implications, and a step-by-step tutorial for the Humalog Self-titration Algorithm.

Full Prescribing Information and Humalog U-100 Patient Information

Self-titration of Humalog U-100 using the Q1D algorithm may help your patients be more engaged in their therapy.1

  • Identify a candidate that you feel comfortable allowing to self-titrate and would like to initiate mealtime insulin starting with one daily injection
  • Show your patient how to utilize tools, such as the logbook, to help them track their blood sugar and make adjustments

Closely monitor blood glucose in all patients treated with insulin. Change insulin regimens cautiously.

For more information, contact your Lilly Sales Representative or call The Lilly Answers Center at 1-800-LillyRx (1-800-545-5979).

Q1D algorithm1

Using the Q1D algorithm, patients self-titrated every day based on premeal or bedtime BG readings, respectively, from the previous day. For example, when adjusting the prebreakfast dose, patients used their prelunch readings from the previous day. The premeal target BG was 85-114 mg/dL. If the patient had a BG reading of ≥115 mg/dL, the patient increased the lispro dose by 1 unit/day until the target was reached. For a BG reading of 56-84 mg/dL, the dose was decreased by 1 unit, and for a reading of <56 mg/dL, the dose was decreased by 2 units.

Q3D algorithm1

Using the Q3D algorithm, patients self-titrated every 3 days based on the median (middle) BG readings from the 3 days before. Accordingly, to adjust the prebreakfast dose, the patient used the median prelunch BG reading from the past 3 days. If the median reading was 85-114 mg/dL, there was no change in lispro dose. For median 115-144 mg/dL, the patient increased the dose by 2 units. For median ≥145 mg/dL, the dose was increased by 4 units. For median 56-84 mg/dL, the dose was decreased by 2 units. For median <56 mg/dL, the dose was decreased by 4 units.

Select Safety Information

Severe, life-threatening, generalized allergy, including anaphylaxis, can occur with Humalog. If hypersensitivity reactions occur, discontinue Humalog and treat per standard of care until signs and symptoms resolve.

AUTONOMY study design

Basal insulin (as insulin glargine [glargine]) was optimized if required during an optional 6-week lead-in period. After the optimization period, glargine was adjusted by study investigators according to fasting blood glucose. Patients performed self-monitoring of blood glucose 4 times daily: before breakfast, lunch, and dinner and at bedtime.

Interventions consisted of adding lispro sequentially by 1, 2, or 3 mealtime injections to their baseline glargine. Initial lispro dose began at breakfast, calculated at 10% of total daily glargine dose.2 Subsequent dosing adjustments were based on the patient self-titration algorithms either every day (Q1D) or every 3 days (Q3D). Once the breakfast dose was optimized based on prelunch blood glucose levels, prelunch prandial dose was started in the same manner, if needed, based on predinner blood glucose readings. Similarly, if needed, predinner prandial dose was started based on 5% of total daily glargine dose and was optimized based on bedtime blood glucose levels.2

Autonomy study design chart

The AUTONOMY study was a 24-week, multicenter, randomized, open-label, parallel trial of adult patients (N=1106) with type 2 diabetes inadequately controlled on basal insulin, conducted in primary care and endocrinology settings, to compare 2 patient self-titration algorithms for initiating and adjusting prandial insulin lispro.

AUTONOMY consisted of 2 identical studies (Study A and Study B) using the same protocol conducted to compare the primary outcome of change in A1C from baseline after 24 weeks between a daily (Q1D) or 3-day (Q3D) algorithm.

In the AUTONOMY study, patients began mealtime insulin with one injection at breakfast1

Is your patient a candidate for self-titration with Humalog U-100 KwikPen?

The patients in the AUTONOMY study1

  • Had type 2 diabetes
  • Had uncontrolled A1C
  • Were on ≥20 units of basal insulin once daily and were on 1 or more oral antidiabetes drugs (OADs)
  • Were aged 18-85
  • Had BMI <45 kg/m2

Overview of the AUTONOMY study

In the AUTONOMY study, mealtime insulin lispro therapy by self-titration was effectively initiated with a single breakfast injection and adjusted in adult patients with type 2 diabetes on basal insulin and OADs using either the Q1D or Q3D algorithm.1

  • Patients began mealtime insulin with one injection at breakfast1
  • Carbohydrate counting and correction factors were not needed to determine dose
  • Under healthcare provider's direction, patients recorded their BG readings and adjusted their prandial insulin based on their readings
  • Patients using the Q1D algorithm experienced a mean decrease in A1C from baseline of 1% by the 24th week of therapy1*
  • With Q1D patients, almost half achieved an A1C ≤7.0% at endpoint1†
  • With the Q1D algorithm, 61.2% of patients required only 1 or 2 mealtime injections at endpoint1,3,4
  • In patients aged 65 and older and the overall study population, there was no statistically significant treatment-by-subgroup difference in A1C from baseline to endpoint1†

All efficacy analyses were based upon the full analysis set (patients in the all-randomized population who took at least 1 dose of insulin lispro).

*Change in A1C in the 24-week AUTONOMY study of Q1D and Q3D algorithms of Study A (N=528) and Study B (N=578). Data were reported as least square mean (LSM) ± standard error (SE) of the mean and 95% CI from a mixed effect model for repeated measures (MMRM) that included baseline A1C, strata, treatment algorithm, visit, and treatment algorithm by visit. Similar results were seen with the Q3D arm of the study.

†The percentages of patients achieving A1C targets at the end of the study (last observation carried forward [LOCF]) were analyzed using a logistic regression model with terms for treatment algorithm and strata.

‡Data are reported as mean (SE) at week 24. N=number of patients in specified treatment and subgroup category. Results are based on patients with non-missing A1C at baseline and at week 24 in the specified treatment and subgroup category. P-values for the 2-way interaction (treatment by subgroup) are from the MMRM model: response=baseline + strata + treatment + visit + treatment by visit + subgroup + treatment by subgroup.

Select Safety Information

Humalog should be given within 15 minutes before or immediately after a meal.

In the AUTONOMY study, patients initiated with one injection at breakfast5

  • Injections were administered using Humalog U-100 KwikPen5
  • Carbohydrate counting was not required to adjust titration1
  • Under healthcare provider direction, patients recorded their BG readings, and then self-adjusted their prandial insulin based on their readings1

Starting dose2

  • The initial breakfast mealtime dose was equivalent to 10% of the total daily glargine dose

Patients assessed and adjusted their own mealtime insulin1

  • Following the Q1D algorithm, patients adjusted the breakfast dose daily based on the previous day’s prelunch BG reading
  • When prelunch BG readings were consistently in the 85-114 mg/dL range, no further breakfast dose adjustments were required
  • yesterday's prelunch blood glucose reading
  • If prelunch blood glucose reading was:
  • today's breakfast dose
  • Adjust today’s breakfast dose:
≥115 mg/dL +1 unit
85-114 mg/dL 0 (no change)
56-84 mg/dL -1 unit
≤55 mg/dL -2 units
Q3D dosing information

Additional doses were added at the healthcare provider’s discretion1,2

  • Once the optimal breakfast injection dose was reached and prelunch blood glucose readings were within the 85-114 mg/dL range, the HCP determined whether or not a lunch dose was needed. If additional control was required (predinner blood glucose readings ≥115 mg/dL), patients made adjustments using the same algorithm that they used for the breakfast dose

For Patients Who Used the Q1D Algorithm, at 24 Weeks1,4:

33.5%

required 1 injection per day

27.7%

required 2 injections per day

38.7%

required 3 injections per day
  • For each meal, the base unit lispro dose was equivalent to2:
    • 10% of the total daily glargine dose at lunch
    • 5% of the total daily glargine dose at dinner
  • Lunch and dinner doses were titrated daily using the Q1D algorithm until the respective BG targets were reached
  • Lunch dose adjustments were based on the previous day’s predinner BG reading
  • Dinner dose adjustments were based on the previous day’s bedtime BG reading
  • 61.2% of patients using the Humalog U-100 KwikPen required only 1 or 2 daily injections of mealtime insulin at study endpoint1,3,4
  • If predinner blood glucose reading was:
  • Adjust today’s lunch dose:
≥115 mg/dL +1 unit
85-114 mg/dL 0 (no change)
56-84 mg/dL -1 unit
≤55 mg/dL -2 units
  • If bedtime blood glucose reading was:
  • Adjust today’s dinner dose:
≥115 mg/dL +1 unit
85-114 mg/dL 0 (no change)
56-84 mg/dL -1 unit
≤55 mg/dL -2 units

Interventions consisted of adding lispro sequentially: first at breakfast, at lunch, and then at dinner.

Select Safety Information

Some medications may alter glucose metabolism, insulin requirements, and the risk for hypoglycemia or hyperglycemia. Signs of hypoglycemia may be reduced or absent in patients taking anti-adrenergic drugs. Particularly close monitoring may be required.

Ask your Lilly Diabetes Sales Representative or call The Lilly Answers Center at 1-800-LillyRx (1-800-545-5979) about the My Mealtime Tracker to assist patients in using the self-titration algorithm.

Efficacy with Q1D self-titration algorithm

Patients experienced a 1.0% mean reduction in A1C from baseline to endpoint with the Q1D self-titration algorithm1

Change in A1C

1% Mean A1C reduction [primary endpoint]

Q1D self-titration algorithm resulted in patients experiencing a 1% mean reduction in A1C from baseline (8.3%).*

*For Q3D, the reduction of A1C from baseline at week 24 was -0.96% for Study A and -0.92% for Study B.

All efficacy analyses were based upon the full analysis set (subjects in the all-randomized population who took at least one dose of insulin lispro). A limitation of this study was the exclusion of subjects with BMIs ≥45 kg/m2.

Change in A1C in the 24-week AUTONOMY study of Q1D and Q3D algorithms, of Study A (N=528) and Study B (N=578). Data were reported as least square mean (LSM) ± SE of the mean and 95% CI from a mixed effect model for repeated measures (MMRM) that included baseline A1C, strata, treatment algorithm, visit, and treatment algorithm by visit.

Similar results were seen with the Q3D arm of the study.

Using the Q1D algorithm, almost half of the patients achieved an A1C of ≤7.0% at 24 weeks1

Patients who achieved A1C ≤7.0% at 24 weeks1,6,7

Patients who achieved A1C ≤7.0% at 24 weeks

Q1D and Q3D algorithms in both studies resulted in similar percentages of patients achieving A1C ≤7%.1

*12 patients had an A1C ≤7% at baseline in Study A and 16 patients in Study B.

The AUTONOMY study was a 24-week, multicenter, randomized, open-label, parallel trial of patients (N=1106) with type 2 diabetes inadequately controlled on basal insulin, conducted in primary care and endocrinology settings, to compare 2 patient self-titration algorithms for initiating and adjusting prandial insulin lispro.

AUTONOMY consisted of 2 identical studies using the same protocol conducted to compare the primary outcome of change in A1C from baseline after 24 weeks between Q1D and Q3D algorithms.

Patients ≥65 years of age who achieved A1C ≤7.0% at 24 weeks1,6,7

Patients ≥65 years of age who achieved A1C ≤7.0% at 24 weeks

Q1D and Q3D algorithms in both studies resulted in similar percentages of patients ≥65 years of age achieving A1C ≤7%.1

*5 patients over 65 years of age had an A1C ≤7% at baseline in Study A and 4 in Study B.

All efficacy analyses were based upon the full analysis set (patients in the all-randomized population who took at least 1 dose of insulin lispro). The percentages of patients achieving A1C targets at the end of the study (LOCF) were analyzed using a logistic regression model with terms for treatment algorithm and strata.

Data are reported as mean (SE) at week 24. N= number of patients in specified treatment and subgroup category. Values are based on patients with non-missing A1C at baseline and at week 24 in the specified treatment and subgroup category. P-values for the 2-way interaction (treatment by subgroup) are from the MMRM model: response=baseline + strata + treatment + visit + treatment by visit + subgroup + treatment by subgroup.

Select Safety Information

Fluid retention and heart failure can occur with concomitant use of TZDs and Humalog. Observe patients for signs and symptoms of heart failure and consider discontinuation to dose reduction of the PPAR-gamma agonist.

Hypoglycemia and weight gain in the AUTONOMY study

Hypoglycemia rates and incidence in the AUTONOMY study7

Hypoglycemia overall

Type of Hypoglycemia (rate and Incidence)

Study A:

Q1D (n=268)

Study B:

Q1D (n=289)
Total hypoglycemia incidence (n [%])

Study A:

231 (86.2)

Study B:

238 (82.4)
Total hypoglycemia rate per 30 days

Study A:

3.15 ± 0.23

Study B:

3.18 ± 0.26
Symptomatic hypoglycemia incidence (n [%])

Study A:

194 (72.4)

Study B:

205 (70.9)
Symptomatic hypoglycemia rate per 30 days

Study A:

1.72 ± 0.16

Study B:

1.50 ± 0.15
Nocturnal hypoglycemia incidence (n [%])

Study A:

169 (63.1)

Study B:

156 (54.0)
Nocturnal hypoglycemia rate per 30 days

Study A:

0.71 ± 0.07

Study B:

0.59 ± 0.07
Severe hypoglycemia incidence (n [%])

Study A:

5 (1.9)

Study B:

7 (2.4)
Severe hypoglycemia rate per 30 days

Study A:

0.00 ± 0.03

Study B:

0.01 ± 0.09

Type of Hypoglycemia (rate and incidence)

Study A:Q1D (n=268)

Study B:Q1D (n=289)

Total hypoglycemia incidence (n [%]) 231 (86.2) 238 (82.4)
Total hypoglycemia rate per 30 days 3.15 ± 0.23 3.18 ± 0.26
Symptomatic hypoglycemia incidence (n [%]) 194 (72.4) 205 (70.9)
Symptomatic hypoglycemia rate per 30 days 1.72 ± 0.16 1.50 ± 0.15
Nocturnal hypoglycemia incidence (n [%]) 169 (63.1) 156 (54.0)
Nocturnal hypoglycemia rate per 30 days 0.71 ± 0.07 0.59 ± 0.07
Severe hypoglycemia incidence (n [%]) 5 (1.9) 7 (2.4)
Severe hypoglycemia rate per 30 days 0.00 ± 0.03 0.01 ± 0.09

Data are n (%), negative binomial mean (NBM) ± SE. Incidence is reported as the number of patients with at least 1 hypoglycemic episode. All safety outcomes were assessed in the all-randomized population (all patients who entered the study, completed the glargine optimization lead-in period [if applicable], and were randomized to 1 of the 2 treatment arms). The rate of total, nocturnal, and severe hypoglycemia per year during the treatment phase was analyzed using LOCF.

  • Similar rates and incidence of hypoglycemia were observed in Q1D and Q3D algorithms in both studies (A and B)

A hypoglycemic episode was defined as any time a patient felt that he or she was experiencing a sign or symptom associated with hypoglycemia, or had a blood glucose level of ≤70 mg/dL even if it was not associated with signs, symptoms, or treatment.

Documented symptomatic hypoglycemia was defined as an event during which typical symptoms of hypoglycemia were accompanied by a measured plasma glucose concentration ≤70 mg/dL.

A nocturnal hypoglycemic episode was defined as any time a patient felt that he or she was experiencing a sign or symptom associated with hypoglycemia, or had a blood glucose level of ≤70 mg/dL even if it was not associated with signs or symptoms and occurred after bedtime and prior to the first meal upon waking (eg, breakfast).

A severe hypoglycemic episode was defined as any time a patient required assistance of another person for treatment and that was accompanied by neurologic (cognitive) impairment.

Hypoglycemia in patients ≥65 years7

Type of Hypoglycemia (rate and incidence)

Study A:

Q1D (n=66)

Study B:

Q1D (n=56)
Total hypoglycemia incidence (n [%])

Study A:

60 (90.9)

Study B:

51 (91.1)
Total hypoglycemia rate per 30 days

Study A:

3.42 ± 0.45

Study B:

4.22 ± 0.68
Symptomatic hypoglycemia incidence (n [%])

Study A:

52 (78.8)

Study B:

39 (69.6)
Symptomatic hypoglycemia rate per 30 days

Study A:

1.71 ± 0.29

Study B:

1.85 ± 0.42
Nocturnal hypoglycemia incidence (n [%])

Study A:

45 (68.2)

Study B:

42 (75.0)
Nocturnal hypoglycemia rate per 30 days

Study A:

0.72 ± 0.12

Study B:

0.99 ± 0.20
Severe hypoglycemia incidence (n [%])

Study A:

3 (4.5)

Study B:

1 (1.8)
Severe hypoglycemia rate per 30 days

Study A:

0.01 ± 0.04

Study B:

0.00 ± 0.03

Type of Hypoglycemia (rate and incidence)

Study A:Q1D (n=66)

Study B:Q1D (n=56)

Total hypoglycemia incidence (n [%]) 60 (90.9) 51 (91.1)
Total hypoglycemia rate per 30 days 3.42 ± 0.45 4.22 ± 0.68
Symptomatic hypoglycemia incidence (n [%]) 52 (78.8) 39 (69.6)
Symptomatic hypoglycemia rate per 30 days 1.71 ± 0.29 1.85 ± 0.42
Nocturnal hypoglycemia incidence (n [%]) 45 (68.2) 42 (75.0)
Nocturnal hypoglycemia rate per 30 days 0.72 ± 0.12 0.99 ± 0.20
Severe hypoglycemia incidence (n [%]) 3 (4.5) 1 (1.8)
Severe hypoglycemia rate per 30 days 0.01 ± 0.04 0.00 ± 0.03

Data are n (%), NBM ± SE.

Incidence is reported as the number of patients with at least one hypoglycemic episode. All safety outcomes were assessed in the all-randomized population (all patients who entered the study, completed the glargine optimization lead-in period [if applicable], and were randomized to 1 of the 2 treatment arms). The rate of total, nocturnal, and severe hypoglycemia per year during the treatment phase was analyzed using LOCF.

  • Similar rates and incidence of hypoglycemia were observed in Q1D and Q3D algorithms in both studies (A and B) in patients ≥65 years of age

Patient weight gain in the AUTONOMY study1

Weight (kg) change from baseline (LSM ± SE) at 24 weeks

Weight (kg) change from baseline (LSM + SE) at 24 weeks
  • Patients following the AUTONOMY Q1D algorithm had mean weight gains of approximately 2.15 to 2.47 kg (4.7 to 5.4 lb) in Study A and Study B, respectively, during the study period1,9
  • In Study A, Q3D participants gained slightly more weight (approximately 1.8 lb) than Q1D participants1

Data were reported as LSM ± SE of the mean based upon an MMRM that included baseline weight, strata, treatment algorithm, visit, and treatment algorithm by visit.

Ask your Lilly Diabetes Sales Representative or call The Lilly Answers Center at 1-800-LillyRx (1-800-545-5979) about the My Mealtime Tracker to assist patients in using the self-titration algorithm.

Select Safety Information

Severe hypoglycemia may be life threatening and can cause seizures or death. Hypoglycemia is the most common adverse effect of Humalog therapy. Patients with renal or hepatic impairment may be at higher risk of hypoglycemia.

Dosing for Humalog® Mix75/25 (75% insulin lispro protamine suspension and 25% insulin lispro injection [rDNA origin])

The following dosing options for patients with type 2 diabetes starting premixed insulin are based on Hirsch et al.10 Dosing guidelines should be individualized for each patient.

Recommended dose

  • Start 10 units predinner
  • Add subsequent prebreakfast dose based on patient's needs, or
  • Start 10 units prebreakfast and predinner

Titration

  • Prebreakfast dose: Adjust based on predinner/evening glucose values
  • Predinner dose: Adjust based on prebreakfast/morning glucose values
  • Do not increase dose if hypoglycemia (<70 mg/dL) or symptoms are present

Titration schedule for once-daily or twice daily insulin regimens9

If most values in the past 3-7 days fall in this range:

Adjust dose by:

<80 mg/dL -2 units
80-109 mg/dL No change
110-139 mg/dL +2 units
140-179 mg/dL +4 units
≥180 mg/dL +6 units

Some Important Dosing Information

Dosage regimens of insulin will vary among patients and should be determined by the healthcare provider familiar with the patient's metabolic needs, eating habits, and other lifestyle variables. In addition, the dose must be individualized for each patient, based on blood glucose monitoring.

Select Safety Information

Hypoglycemia is the most common adverse effect associated with Humalog Mix75/25. Severe hypoglycemia can cause seizures and may be life threatening. The timing of hypoglycemia may differ among various insulin formulations.

Dosing for Humalog® Mix50/50 (50% insulin lispro protamine suspension and 50% insulin lispro injection [rDNA origin])

Transitioning your patients from basal insulin to a premixed insulin

  • To reduce the potential for insulin stacking, start premixed insulin therapy after the typical dosing cycle of the last dose of basal insulin therapy
    • For example, for a patient on a basal insulin analog, do not start premixed insulin therapy until 18 to 24 hours after the last basal insulin dose9
  • Convert the total daily dose of basal insulin as a guide for the initial starting dose of premixed insulin10,11
    • For example, a patient taking 50 units of basal insulin would be moved to a 50-unit total daily dose of premixed insulin
  • Titrate dosage as appropriate based on clinical judgment

Dosage regimens of insulin will vary among patients and should be determined by the healthcare provider familiar with the patient’s metabolic needs, eating habits, and other lifestyle variables. In addition, the dose must be individualized for each patient, based on blood glucose monitoring. Pens and needles are for single-patient use only and should not be shared, even in healthcare facilities, as infection or disease can be spread from one person to another. Do not withdraw insulin from the pen using a syringe.

Helping fit Humalog Mix50/50 into your patient’s life

  • Use average blood glucose (BG) values across 3 to 7 days10
  • A dose change is applied to a mealtime dose preceding the time of the reading10,12
    • For example, if pre-evening glucose is high, then increase insulin dose before breakfast
    • BG reading times should be kept consistent relative to mealtimes due to the time-action profile of a premixed therapy

Select Safety Information

Fluid retention and heart failure can occur with concomitant use of TZDs and Humalog Mix50/50.

The safety and effectiveness of Humalog Mix50/50 in patients less than 18 years of age have not been established.

References

  1. Edelman SV, Liu R, Johnson J, Glass LC. AUTONOMY: the first randomized trial comparing two patient-driven approaches to initiate and titrate prandial insulin lispro in type 2 diabetes. Diabetes Care. 2014;37:2132-2140.
  2. Data on file, Lilly Research Laboratories: HI20140206A.
  3. Data on file, Lilly Research Laboratories: HI20140219B.
  4. Data on file, Lilly Research Laboratories: HI20140225A.
  5. Data on file, Lilly Research Laboratories: HI20140212B.
  6. Data on file, Lilly Research Laboratories: HI20140219C.
  7. Data on file, Lilly Research Laboratories: HI20140219D.
  8. Data on file, Lilly Research Laboratories: HI20150209B.
  9. Data on file, Lilly Research Laboratories: HI20140212A.
  10. Hirsch IB, Bergenstal RM, Parkin CG, Wright E Jr, Buse JB. A real-world approach to insulin therapy in primary care practice. Clin Diabetes. 2005;23(2):78-86.
  11. Tibaldi J. Initiating and intensifying insulin therapy in type 2 diabetes mellitus. Am J Med. 2008;121(suppl 6):S20-S29.
  12. Jain SM, Mao X, Escalante-Pulido M, Vorokhobina N, Lopez I, Ilag LL. Prandial-basal insulin regimens plus oral antihyperglycaemic agents to improve mealtime glycaemia: initiate and progressively advance insulin therapy in type 2 diabetes. Diabetes Obes Metab. 2010;12(11):967-975.