Amber Wegner, PharmD

Chronic obstructive pulmonary disease (COPD) affects about 25% of the veteran population and is the third-leading cause of death globally.^1,2 In patients with COPD, cigarette smoking leads to increased respiratory symptoms, a greater annual rate of decline in forced expiratory volume in 1 second (FEV₁), and an increase in COPD mortality rate vs nonsmokers.³ Veterans are at a higher risk of COPD due to increased prevalence of smoking within this population as well as military activities leading to environmental and occupational exposure.⁴

According to the 2024 Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines, the primary treatment goals of COPD therapy are to reduce symptoms and future risk of exacerbations.³ Bronchodilators are recommended for initial COPD pharmacotherapy, including long-acting muscarinic antagonists (LAMAs) and/or long-acting Β₂-agonists (LABAs). In some cases, treatment may include inhaled corticosteroids (ICS). Evidence supports ICS therapy in patients with COPD experiencing hospitalizations for exacerbations, ≥ 2 moderate exacerbations per year, blood eosinophil count ≥ 300 cells/μL or concomitant asthma.³

While the 2024 GOLD guidelines caution against the use of ICS outside of certain patient groups, previous GOLD guidelines recommended the use of ICS more broadly.⁵ Due to these changes, many patients may be using ICS therapy unnecessarily. At the Sioux Falls Veterans Affairs Health Care System (SFVAHCS), ICS overuse was identified as a driver of increased medication burden and potential adverse effects (AEs). To help reduce unnecessary ICS use, a data dashboard was created to identify potential candidates for ICS deprescribing. SFVAHCS clinical pharmacy practitioners are licensed pharmacists who work as independent practitioners with a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics. Pharmacists contacted dashboard patients to de-escalate ICS therapy when appropriate.

The SUNSET trial directly compared the continuation of triple therapy (tiotropium + salmeterol/fluticasone propionate) vs deprescribing to LABA/LAMA (indacaterol/ glycopyrronium) in patients with COPD.⁶ It evaluated whether LABA/LAMA was noninferior to LABA/LAMA/ICS therapy when comparing COPD exacerbations in patients whose COPD exacerbations were infrequent. Participants were randomized to triple therapy continuation or indacaterol/glycopyrronium and followed for 26 weeks. Patients on indacaterol/glycopyrronium did not have a significant difference in exacerbations than patients utilizing triple therapy.

The Implementation of a Targeted ICS De-Escalation in Patients with COPD in the Primary Care Setting trial evaluated the success of pharmacist-led ICS deprescribing in appropriate patients with COPD.⁷ Pharmacists followed GOLD guidelines to recommend ICS deprescribing and have risk vs benefit discussions with certain patients. Patients were considered for ICS deprescribing if they had a history of recurrent pneumonia or had no exacerbations within the previous year and eosinophils < 300 cells/μL (risk-benefit discussion if no eosinophil count available). This study found that 19.6% of patients were unable to tolerate ICS withdrawal and resumed either a standard or reduced dose of ICS therapy.⁷

Current guidelines and evidence recommend deprescribing ICS for appropriate patients. There is no current literature defining the impact of pharmacists on ICS deprescribing within the US Department of Veterans Affairs (VA) system. This study will allow for a quantifiable measure of pharmacists’ impact on reducing AEs associated with unnecessary ICS use.

Methods

This retrospective, single-center study was conducted at the SFVAHCS. Data were collected through manual chart review of SFVAHCS electronic health records. Veterans aged ≥ 18 years with a COPD diagnosis who underwent ICS deprescribing by a SFVAHCS pharmacist between February 2022 and December 2023 were included. Records were examined for 52 weeks prior to ICS withdrawal (baseline) and 52 weeks following withdrawal. Patients were excluded from the study if they had a history of asthma or ICS was used for < 52 weeks before deprescribing. Baseline characteristics were collected, including age, race, sex, current tobacco use, eosinophil count, COPD maintenance therapy, FEV₁/forced vital capacity (FVC) ratio, and mean postbronchodilator FEV₁ improvement.

The primary endpoint was number of COPD exacerbations at 52 weeks before vs after deprescribing. Secondary endpoints included the number of patients restarted on an ICS within 52 weeks of deprescribing, as well as ICS AEs, including pneumonia, oral candidiasis, and throat hoarseness.

Statistical Analysis

The primary endpoint was analyzed using the Wilcoxon signed rank test and secondary endpoints were analyzed using the McNemar exact test. Results with P < .05 were considered statistically significant for both tests.

Results

Seventy-six patients were included. Patients had a mean age of 75 years and 91% identified as White, which is representative of the SFVAHCS patient population (Table 1). Twenty-nine (38%) patients were current tobacco users and 55 patients (72%) used LAMA/LABA therapy (after ICS deprescribing) with an eosinophil count < 300 cells/μL. There was no significant difference in exacerbations before vs after ICS deprescribing (P = .78) (Table 2). There were 7 AEs reported before ICS deprescribing vs 0 following ICS deprescribing (P < .001). Five patients (7%) reported throat hoarseness, 1 (1%) reported pneumonia, and 1 (1%) reported oral candidiasis. Eighteen patients were reinitiated on ICS (24%). ICS reinitiation was most commonly due to patients reporting worsening symptoms (56%) (Table 3).

Discussion

This study sought to determine the impact of pharmacist-led ICS deprescribing on AEs and exacerbations experienced by patients with COPD. COPD exacerbations were not significantly different before vs after ICS deprescribing. The pharmacist-led ICS deprescribing program did not lead to increased COPD exacerbations. Similar to the SUNSET trial, the results of this study showed exacerbations did not significantly increase upon ICS deprescribing; however, this study differed by specifying pharmacist- led intervention.⁶

There was a decrease in ICS-related AEs following ICS deprescribing. Several patients were reinitiated on an ICS. As expected with deprescribing, some patients were not able to tolerate ICS withdrawal or had clinical indications to resume therapy (ie, an exacerbation). Similar results were found in another study where 8.9% of patients were restarted on ICS and 10.9% were de-escalated to a lower dose but were unable to stop completely.⁷ A 2024 systematic review by Georgiou et al found a wide range of patients resumed ICS therapy following withdrawal (21%-74%). Of note, only 2 of the randomized controlled trials and 3 observational studies included this meta-analysis included data on ICS reinitiation. Georgiou et al concluded there was insufficient evidence to determine the proportion of patients reinitiated on ICS but patients were commonly resumed on ICS therapy due to worsening symptoms, experiencing an exacerbation or decline in FEV₁.⁸ Although the rate of ICS reinitiation was unclear in the Georgiou et al meta-analysis, reasons for re-initiation were similar to what was found in our study.⁸

Strengths and Limitations

The retrospective nature of this study and its small sample size of 76 patients limit its findings. The retrospective nature of the study required researchers to rely on proper chart documentation, which is not always accurate or up to date. Lack of documentation of COPD exacerbations or patients who received care outside the VA following initial deprescribing could have biased study results. This patient population is representative of the veteran population in South Dakota but is not representative of the female or non-White patient population which may be more prevalent at other VA Health Care Systems as well as nonveteran patient populations. Additionally, this study was limited to a review of 52 weeks pre- and post-ICS deprescribing which may have impacted results. Patients may have had a COPD exacerbation or were restarted on ICS therapy beyond 52 weeks. Finally, the retrospective nature and small sample size limited the findings for the primary endpoint which could have been improved with a larger sample size and a randomized controlled design.

The comparison of patients with themselves before and after ICS deprescribing reduced the potential for bias seen in retrospective studies. This method did not require a second control group which would potentially introduce confounding factors.

Conclusions

This study found that in a small population of veterans with COPD, pharmacist-led ICS deprescribing did not lead to an increase in COPD exacerbations and decreased the risk of AEs related to ICS therapy. Some patients were reinitiated on ICS therapy; however, reinitiation was rarely attributable to a COPD exacerbation. This study suggests that pharmacists were able to appropriately identify candidates for ICS deprescribing without increasing their risk of exacerbations. By de-escalating ICS therapy, pharmacists decreased medication burden and potential AEs caused by ICS therapy. These findings support expanding the role of clinical pharmacists in COPD management, particularly in identifying candidates for safe ICS deprescribing.

Chronic obstructive pulmonary disease (COPD) affects about 25% of the veteran population and is the third-leading cause of death globally.^1,2 In patients with COPD, cigarette smoking leads to increased respiratory symptoms, a greater annual rate of decline in forced expiratory volume in 1 second (FEV₁), and an increase in COPD mortality rate vs nonsmokers.³ Veterans are at a higher risk of COPD due to increased prevalence of smoking within this population as well as military activities leading to environmental and occupational exposure.⁴

According to the 2024 Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines, the primary treatment goals of COPD therapy are to reduce symptoms and future risk of exacerbations.³ Bronchodilators are recommended for initial COPD pharmacotherapy, including long-acting muscarinic antagonists (LAMAs) and/or long-acting Β₂-agonists (LABAs). In some cases, treatment may include inhaled corticosteroids (ICS). Evidence supports ICS therapy in patients with COPD experiencing hospitalizations for exacerbations, ≥ 2 moderate exacerbations per year, blood eosinophil count ≥ 300 cells/μL or concomitant asthma.³

While the 2024 GOLD guidelines caution against the use of ICS outside of certain patient groups, previous GOLD guidelines recommended the use of ICS more broadly.⁵ Due to these changes, many patients may be using ICS therapy unnecessarily. At the Sioux Falls Veterans Affairs Health Care System (SFVAHCS), ICS overuse was identified as a driver of increased medication burden and potential adverse effects (AEs). To help reduce unnecessary ICS use, a data dashboard was created to identify potential candidates for ICS deprescribing. SFVAHCS clinical pharmacy practitioners are licensed pharmacists who work as independent practitioners with a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics. Pharmacists contacted dashboard patients to de-escalate ICS therapy when appropriate.

The SUNSET trial directly compared the continuation of triple therapy (tiotropium + salmeterol/fluticasone propionate) vs deprescribing to LABA/LAMA (indacaterol/ glycopyrronium) in patients with COPD.⁶ It evaluated whether LABA/LAMA was noninferior to LABA/LAMA/ICS therapy when comparing COPD exacerbations in patients whose COPD exacerbations were infrequent. Participants were randomized to triple therapy continuation or indacaterol/glycopyrronium and followed for 26 weeks. Patients on indacaterol/glycopyrronium did not have a significant difference in exacerbations than patients utilizing triple therapy.

The Implementation of a Targeted ICS De-Escalation in Patients with COPD in the Primary Care Setting trial evaluated the success of pharmacist-led ICS deprescribing in appropriate patients with COPD.⁷ Pharmacists followed GOLD guidelines to recommend ICS deprescribing and have risk vs benefit discussions with certain patients. Patients were considered for ICS deprescribing if they had a history of recurrent pneumonia or had no exacerbations within the previous year and eosinophils < 300 cells/μL (risk-benefit discussion if no eosinophil count available). This study found that 19.6% of patients were unable to tolerate ICS withdrawal and resumed either a standard or reduced dose of ICS therapy.⁷

Current guidelines and evidence recommend deprescribing ICS for appropriate patients. There is no current literature defining the impact of pharmacists on ICS deprescribing within the US Department of Veterans Affairs (VA) system. This study will allow for a quantifiable measure of pharmacists’ impact on reducing AEs associated with unnecessary ICS use.

Methods

This retrospective, single-center study was conducted at the SFVAHCS. Data were collected through manual chart review of SFVAHCS electronic health records. Veterans aged ≥ 18 years with a COPD diagnosis who underwent ICS deprescribing by a SFVAHCS pharmacist between February 2022 and December 2023 were included. Records were examined for 52 weeks prior to ICS withdrawal (baseline) and 52 weeks following withdrawal. Patients were excluded from the study if they had a history of asthma or ICS was used for < 52 weeks before deprescribing. Baseline characteristics were collected, including age, race, sex, current tobacco use, eosinophil count, COPD maintenance therapy, FEV₁/forced vital capacity (FVC) ratio, and mean postbronchodilator FEV₁ improvement.

The primary endpoint was number of COPD exacerbations at 52 weeks before vs after deprescribing. Secondary endpoints included the number of patients restarted on an ICS within 52 weeks of deprescribing, as well as ICS AEs, including pneumonia, oral candidiasis, and throat hoarseness.

Statistical Analysis

The primary endpoint was analyzed using the Wilcoxon signed rank test and secondary endpoints were analyzed using the McNemar exact test. Results with P < .05 were considered statistically significant for both tests.

Results

Seventy-six patients were included. Patients had a mean age of 75 years and 91% identified as White, which is representative of the SFVAHCS patient population (Table 1). Twenty-nine (38%) patients were current tobacco users and 55 patients (72%) used LAMA/LABA therapy (after ICS deprescribing) with an eosinophil count < 300 cells/μL. There was no significant difference in exacerbations before vs after ICS deprescribing (P = .78) (Table 2). There were 7 AEs reported before ICS deprescribing vs 0 following ICS deprescribing (P < .001). Five patients (7%) reported throat hoarseness, 1 (1%) reported pneumonia, and 1 (1%) reported oral candidiasis. Eighteen patients were reinitiated on ICS (24%). ICS reinitiation was most commonly due to patients reporting worsening symptoms (56%) (Table 3).

Discussion

This study sought to determine the impact of pharmacist-led ICS deprescribing on AEs and exacerbations experienced by patients with COPD. COPD exacerbations were not significantly different before vs after ICS deprescribing. The pharmacist-led ICS deprescribing program did not lead to increased COPD exacerbations. Similar to the SUNSET trial, the results of this study showed exacerbations did not significantly increase upon ICS deprescribing; however, this study differed by specifying pharmacist- led intervention.⁶

There was a decrease in ICS-related AEs following ICS deprescribing. Several patients were reinitiated on an ICS. As expected with deprescribing, some patients were not able to tolerate ICS withdrawal or had clinical indications to resume therapy (ie, an exacerbation). Similar results were found in another study where 8.9% of patients were restarted on ICS and 10.9% were de-escalated to a lower dose but were unable to stop completely.⁷ A 2024 systematic review by Georgiou et al found a wide range of patients resumed ICS therapy following withdrawal (21%-74%). Of note, only 2 of the randomized controlled trials and 3 observational studies included this meta-analysis included data on ICS reinitiation. Georgiou et al concluded there was insufficient evidence to determine the proportion of patients reinitiated on ICS but patients were commonly resumed on ICS therapy due to worsening symptoms, experiencing an exacerbation or decline in FEV₁.⁸ Although the rate of ICS reinitiation was unclear in the Georgiou et al meta-analysis, reasons for re-initiation were similar to what was found in our study.⁸

Strengths and Limitations

The retrospective nature of this study and its small sample size of 76 patients limit its findings. The retrospective nature of the study required researchers to rely on proper chart documentation, which is not always accurate or up to date. Lack of documentation of COPD exacerbations or patients who received care outside the VA following initial deprescribing could have biased study results. This patient population is representative of the veteran population in South Dakota but is not representative of the female or non-White patient population which may be more prevalent at other VA Health Care Systems as well as nonveteran patient populations. Additionally, this study was limited to a review of 52 weeks pre- and post-ICS deprescribing which may have impacted results. Patients may have had a COPD exacerbation or were restarted on ICS therapy beyond 52 weeks. Finally, the retrospective nature and small sample size limited the findings for the primary endpoint which could have been improved with a larger sample size and a randomized controlled design.

The comparison of patients with themselves before and after ICS deprescribing reduced the potential for bias seen in retrospective studies. This method did not require a second control group which would potentially introduce confounding factors.

Conclusions

This study found that in a small population of veterans with COPD, pharmacist-led ICS deprescribing did not lead to an increase in COPD exacerbations and decreased the risk of AEs related to ICS therapy. Some patients were reinitiated on ICS therapy; however, reinitiation was rarely attributable to a COPD exacerbation. This study suggests that pharmacists were able to appropriately identify candidates for ICS deprescribing without increasing their risk of exacerbations. By de-escalating ICS therapy, pharmacists decreased medication burden and potential AEs caused by ICS therapy. These findings support expanding the role of clinical pharmacists in COPD management, particularly in identifying candidates for safe ICS deprescribing.

Chronic obstructive pulmonary disease (COPD) affects about 25% of the veteran population and is the third-leading cause of death globally.^1,2 In patients with COPD, cigarette smoking leads to increased respiratory symptoms, a greater annual rate of decline in forced expiratory volume in 1 second (FEV₁), and an increase in COPD mortality rate vs nonsmokers.³ Veterans are at a higher risk of COPD due to increased prevalence of smoking within this population as well as military activities leading to environmental and occupational exposure.⁴

According to the 2024 Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines, the primary treatment goals of COPD therapy are to reduce symptoms and future risk of exacerbations.³ Bronchodilators are recommended for initial COPD pharmacotherapy, including long-acting muscarinic antagonists (LAMAs) and/or long-acting Β₂-agonists (LABAs). In some cases, treatment may include inhaled corticosteroids (ICS). Evidence supports ICS therapy in patients with COPD experiencing hospitalizations for exacerbations, ≥ 2 moderate exacerbations per year, blood eosinophil count ≥ 300 cells/μL or concomitant asthma.³

While the 2024 GOLD guidelines caution against the use of ICS outside of certain patient groups, previous GOLD guidelines recommended the use of ICS more broadly.⁵ Due to these changes, many patients may be using ICS therapy unnecessarily. At the Sioux Falls Veterans Affairs Health Care System (SFVAHCS), ICS overuse was identified as a driver of increased medication burden and potential adverse effects (AEs). To help reduce unnecessary ICS use, a data dashboard was created to identify potential candidates for ICS deprescribing. SFVAHCS clinical pharmacy practitioners are licensed pharmacists who work as independent practitioners with a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics. Pharmacists contacted dashboard patients to de-escalate ICS therapy when appropriate.

The SUNSET trial directly compared the continuation of triple therapy (tiotropium + salmeterol/fluticasone propionate) vs deprescribing to LABA/LAMA (indacaterol/ glycopyrronium) in patients with COPD.⁶ It evaluated whether LABA/LAMA was noninferior to LABA/LAMA/ICS therapy when comparing COPD exacerbations in patients whose COPD exacerbations were infrequent. Participants were randomized to triple therapy continuation or indacaterol/glycopyrronium and followed for 26 weeks. Patients on indacaterol/glycopyrronium did not have a significant difference in exacerbations than patients utilizing triple therapy.

The Implementation of a Targeted ICS De-Escalation in Patients with COPD in the Primary Care Setting trial evaluated the success of pharmacist-led ICS deprescribing in appropriate patients with COPD.⁷ Pharmacists followed GOLD guidelines to recommend ICS deprescribing and have risk vs benefit discussions with certain patients. Patients were considered for ICS deprescribing if they had a history of recurrent pneumonia or had no exacerbations within the previous year and eosinophils < 300 cells/μL (risk-benefit discussion if no eosinophil count available). This study found that 19.6% of patients were unable to tolerate ICS withdrawal and resumed either a standard or reduced dose of ICS therapy.⁷

Current guidelines and evidence recommend deprescribing ICS for appropriate patients. There is no current literature defining the impact of pharmacists on ICS deprescribing within the US Department of Veterans Affairs (VA) system. This study will allow for a quantifiable measure of pharmacists’ impact on reducing AEs associated with unnecessary ICS use.

Methods

This retrospective, single-center study was conducted at the SFVAHCS. Data were collected through manual chart review of SFVAHCS electronic health records. Veterans aged ≥ 18 years with a COPD diagnosis who underwent ICS deprescribing by a SFVAHCS pharmacist between February 2022 and December 2023 were included. Records were examined for 52 weeks prior to ICS withdrawal (baseline) and 52 weeks following withdrawal. Patients were excluded from the study if they had a history of asthma or ICS was used for < 52 weeks before deprescribing. Baseline characteristics were collected, including age, race, sex, current tobacco use, eosinophil count, COPD maintenance therapy, FEV₁/forced vital capacity (FVC) ratio, and mean postbronchodilator FEV₁ improvement.

The primary endpoint was number of COPD exacerbations at 52 weeks before vs after deprescribing. Secondary endpoints included the number of patients restarted on an ICS within 52 weeks of deprescribing, as well as ICS AEs, including pneumonia, oral candidiasis, and throat hoarseness.

Statistical Analysis

The primary endpoint was analyzed using the Wilcoxon signed rank test and secondary endpoints were analyzed using the McNemar exact test. Results with P < .05 were considered statistically significant for both tests.

Results

Seventy-six patients were included. Patients had a mean age of 75 years and 91% identified as White, which is representative of the SFVAHCS patient population (Table 1). Twenty-nine (38%) patients were current tobacco users and 55 patients (72%) used LAMA/LABA therapy (after ICS deprescribing) with an eosinophil count < 300 cells/μL. There was no significant difference in exacerbations before vs after ICS deprescribing (P = .78) (Table 2). There were 7 AEs reported before ICS deprescribing vs 0 following ICS deprescribing (P < .001). Five patients (7%) reported throat hoarseness, 1 (1%) reported pneumonia, and 1 (1%) reported oral candidiasis. Eighteen patients were reinitiated on ICS (24%). ICS reinitiation was most commonly due to patients reporting worsening symptoms (56%) (Table 3).

Discussion

This study sought to determine the impact of pharmacist-led ICS deprescribing on AEs and exacerbations experienced by patients with COPD. COPD exacerbations were not significantly different before vs after ICS deprescribing. The pharmacist-led ICS deprescribing program did not lead to increased COPD exacerbations. Similar to the SUNSET trial, the results of this study showed exacerbations did not significantly increase upon ICS deprescribing; however, this study differed by specifying pharmacist- led intervention.⁶

There was a decrease in ICS-related AEs following ICS deprescribing. Several patients were reinitiated on an ICS. As expected with deprescribing, some patients were not able to tolerate ICS withdrawal or had clinical indications to resume therapy (ie, an exacerbation). Similar results were found in another study where 8.9% of patients were restarted on ICS and 10.9% were de-escalated to a lower dose but were unable to stop completely.⁷ A 2024 systematic review by Georgiou et al found a wide range of patients resumed ICS therapy following withdrawal (21%-74%). Of note, only 2 of the randomized controlled trials and 3 observational studies included this meta-analysis included data on ICS reinitiation. Georgiou et al concluded there was insufficient evidence to determine the proportion of patients reinitiated on ICS but patients were commonly resumed on ICS therapy due to worsening symptoms, experiencing an exacerbation or decline in FEV₁.⁸ Although the rate of ICS reinitiation was unclear in the Georgiou et al meta-analysis, reasons for re-initiation were similar to what was found in our study.⁸

Strengths and Limitations

The retrospective nature of this study and its small sample size of 76 patients limit its findings. The retrospective nature of the study required researchers to rely on proper chart documentation, which is not always accurate or up to date. Lack of documentation of COPD exacerbations or patients who received care outside the VA following initial deprescribing could have biased study results. This patient population is representative of the veteran population in South Dakota but is not representative of the female or non-White patient population which may be more prevalent at other VA Health Care Systems as well as nonveteran patient populations. Additionally, this study was limited to a review of 52 weeks pre- and post-ICS deprescribing which may have impacted results. Patients may have had a COPD exacerbation or were restarted on ICS therapy beyond 52 weeks. Finally, the retrospective nature and small sample size limited the findings for the primary endpoint which could have been improved with a larger sample size and a randomized controlled design.

The comparison of patients with themselves before and after ICS deprescribing reduced the potential for bias seen in retrospective studies. This method did not require a second control group which would potentially introduce confounding factors.

Conclusions

This study found that in a small population of veterans with COPD, pharmacist-led ICS deprescribing did not lead to an increase in COPD exacerbations and decreased the risk of AEs related to ICS therapy. Some patients were reinitiated on ICS therapy; however, reinitiation was rarely attributable to a COPD exacerbation. This study suggests that pharmacists were able to appropriately identify candidates for ICS deprescribing without increasing their risk of exacerbations. By de-escalating ICS therapy, pharmacists decreased medication burden and potential AEs caused by ICS therapy. These findings support expanding the role of clinical pharmacists in COPD management, particularly in identifying candidates for safe ICS deprescribing.

Type 2 diabetes mellitus (T2DM) is a chronic disease becoming more prevalent each year and is the seventh-leading cause of death in the United States.¹ The most common reason for hospitalization for patients with T2DM is uncontrolled glycemic levels.² Nearly 25% of the US Department of Veterans Affairs (VA) patient population has T2DM.³ T2DM is the leading cause of blindness, end-stage renal disease, and amputation for VA patients.⁴

According to the 2023 American Diabetes Association (ADA) guidelines, treatment goals of T2DM include eliminating symptoms, preventing or delaying complications, and attaining glycemic goals. A typical hemoglobin A_1c (HbA_1c) goal range is < 7%, but individual goals can vary up to < 9% due to a multitude of factors, including patient comorbidities and clinical status.⁵

Initial treatment recommendations are nonpharmacologic and include comprehensive lifestyle interventions such as optimizing nutrition, physical activity, and behavioral therapy. When pharmacologic therapy is required, metformin is the preferred first-line treatment for the majority of newly diagnosed patients with T2DM and should be added to continued lifestyle management.⁵ If HbA_1c levels remains above goal, the 2023 ADA guidelines recommend adding a second medication, including but not limited to insulin, a glucagonlike peptide-1 receptor agonist (GLP-1RA), or a sodium-glucose cotransporter 2 inhibitor. Medication choice is largely based on the patient’s concomitant conditions (eg, atherosclerotic cardiovascular disease, heart failure, or chronic kidney disease). The 2023 ADA guidelines suggest initiating insulin therapy when a patient's blood glucose ≥ 300 mg/dL, HbA_1c > 10%, or if the patient has symptoms of hyperglycemia, even at initial diagnosis. Initiating medications to minimize or avoid hypoglycemia is a priority, especially in high-risk individuals.⁵

Clinical evidence shows that GLP-1RAs may provide similar glycemic control to insulin with lower risk of hypoglycemia.⁶ Other reported benefits of GLP-1RAs include weight loss, blood pressure reduction, and improved lipid levels. The most common adverse events (AEs) with GLP-1RAs are gastrointestinal. Including GLP-1RAs in T2DM pharmacotherapy may lower the risk of hypoglycemia, especially in patients at high risk of hypoglycemia.

The 2023 ADA guidelines indicate that it is appropriate to initiate GLP-]1RAs in patients on insulin.⁵ However, while GLP-1RAs do not increase the risk of hypoglycemia independently, combination treatment with GLP-1RAs and insulin can still result in hypoglycemia.⁶ Insulin is the key suspect of this hypoglycemic risk.⁷ Thus, if insulin dosage can be reduced or discontinued, this might reduce the risk of hypoglycemia.

The literature is limited on how the addition of a GLP-1RA to insulin treatment will affect the patient's daily insulin doses, particularly for the veteran population. The goal of this study is to examine this gap in current research by examining semaglutide, which is the current formulary preferred GLP-1RA at the VA.

Semaglutide is subcutaneously initiated at a dose of 0.25 mg once weekly for 4 weeks to reduce gastrointestinal symptoms, then increased to 0.5 mg weekly. Additional increases to a maintenance dose of 1 mg or 2 mg weekly can occur to achieve glycemic goals. The SUSTAIN-FORTE randomized controlled trial sought to determine whether there was a difference in HbA_1c level reduction and significant weight loss with the 2-mg vs 1-mg dose.⁸ Patients in the trial were taking metformin but needed additional medication to control their HbA_1c. They were not using insulin and may or may not have been taking sulfonylureas prior to semaglutide initiation. Semaglutide 2 mg was found to significantly improve HbA_1c control and promote weight loss compared with semaglutide 1 mg, while maintaining a similar safety profile.

Because this study involved patients who required additional HbA_1c control, although semaglutide reduced HbA_1c, not all patients were able to reduce their other diabetes medications, which depended on the baseline HbA_1c level and the level upon completion of semaglutide titration. Dose reductions for the patients’ other T2DM medications were not reported at trial end. SUSTAIN-FORTE established titration up to semaglutide 2 mg as effective for HbA_1c reduction, although it did not study patients also on insulin.⁸

Insulin is associated with hypoglycemic risk, weight gain, and other AEs.^7,8 This study analyzed whether increasing semaglutide could reduce insulin doses and therefore reduce risk of AEs in patients with T2DM.

Methods

A retrospective, single-center, chart review was conducted at VA Sioux Falls Health Care System (VASFHCS). Data were collected through manual review of VASFHCS electronic medical records. Patients aged ≥ 18 years with active prescriptions for at least once-daily insulin who were initiated on 2-mg weekly dose of semaglutide at the VASFHCS clinical pharmacy practitioner medication management clinic between January 1, 2021, and September 1, 2023, were included. VASFHCS clinical pharmacy practitioners have a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics.

The most frequently used prandial insulin at VASFHCS is insulin aspart, and the most frequently used basal insulin is insulin glargine. Patients were retrospectively monitored as they progressed from baseline (the point in time where semaglutide 0.5 mg was initiated) to ≥ 3 months on semaglutide 2-mg therapy. Patients were excluded if they previously used a GLP-1RA or if they were on sliding scale insulin without an exact daily dosage.

The primary endpoint was the percent change in total daily insulin dose from baseline to each dose increase after receiving semaglutide 2 mg for ≥ 3 months. Secondary endpoints included changes in daily prandial insulin dose, daily basal insulin dose, HbA_1c, and number of hypoglycemic events reported. Data collected included age, race, weight, body mass index, total daily prandial insulin dose, total daily basal insulin dose, HbA_1c, and hypoglycemic events reported at the visit when semaglutide was initiated.

Statistical Analysis

The sample size was calculated prior to data collection, and it was determined that for α = .05, 47 patients were needed to achieve 95% power. The primary endpoint was assessed using a paired t test, as were each secondary endpoint. Results with P < .05 were considered statistically significant.

Results

Sixty-two patients were included. The mean HbA_1c level at baseline was 7.7%, the baseline mean prandial and insulin daily doses were 41.5 units and 85.1 units, respectively (Table 1) From baseline to initiation of a semaglutide 1-mg dose, the daily insulin dose changed –5.6% (95% CI, 2.2-14.0; P = .008). From baseline to 2-mg dose initiation daily insulin changed -22.2% (95% CI, 22.0-35.1; P < .001) and for patients receiving semaglutide 2 mg for ≥ 3 months it changed -36.9% (95% CI, 37.4-56.5; P < .001) (Figure).

After receiving the 2-mg dose for ≥ 3 months, the mean daily dose of prandial insulin decreased from 41.5 units to 24.6 units (95% CI, 12.6-21.2; P < .001); mean daily dose of basal insulin decreased from 85.1 units to 52.1 units (95% CI, 23.9-42.0; P < .001); and mean HbA_1c level decreased from 7.7% to 7.1% (95% CI, 0.3-0.8; P < .001). Mean number of hypoglycemic events reported was not statistically significant, changing from 3.6 to 3.2 (95% CI, –0.6 to 0.1; P = .21) (Table 2).

Discussion

This study investigated the effect of subcutaneous semaglutide dose escalation on total daily insulin dose for patients with T2DM. There was a statistically significant decrease in total daily insulin dose from baseline to 1 mg initiation; this decrease continued with further insulin dose reduction seen at the 2-mg dose initiation and additional insulin dose reduction at ≥ 3 months at this dose. It was hypothesized there would be a significant total daily insulin dose reduction at some point, especially when transitioning from the semaglutide 1-mg to the 2-mg dose, based on previous research. ^9,10 The additional reduction in daily insulin dose when continuing on semaglutide 2 mg for ≥ 3 months was an unanticipated but added benefit, showing that if tolerated, maintaining the 2-mg dose will help patients reduce their insulin doses.

In terms of secondary endpoints, there was a statistically significant decrease in mean total daily dose individually for prandial and basal insulin from baseline to ≥ 3 months after semaglutide 2 mg initiation. The change in HbA_1c level was also statistically significant and decreased from baseline, even as insulin doses were reduced. This change in HbA_1c level was expected; previous literature has shown a significant link between improving HbA_1c control when semaglutide doses are increased to 2 mg weekly.¹⁰ Due to having been shown in previous trials, it was expected that HbA_1c levels would decrease even when the insulin doses were being reduced.¹⁰ Insulin dose reduction can potentially be added to the growing evidence of semaglutide benefits. The change in the number of hypoglycemic events was not statistically significant, which was unexpected since previous research show a trend in patients taking GLP-1RAs having fewer hypoglycemic events than those taking insulin.⁶ Further investigation with a larger sample size and prospective trial could determine whether this result is an outlier. In this study, there was no increase in HbA_1c or hypoglycemic events reported with increasing semaglutide doses, which provides further evidence of the safety of semaglutide even at higher doses.

These data suggest that for a patient with T2DM who is already taking insulin, the recommended titration of semaglutide is to start with 0.5 mg and titrate up to a 2-mg subcutaneous weekly dose and to then continue at that dose. As long as the 2-mg dose is tolerated, it will provide patients with the most HbA_1c control and lead to a reduction of their total daily insulin doses according to these results.

Strengths and Limitations

This study compared patient data at different points. This method did not require a second distinct control group, which would potentially introduce confounding factors, such as different baseline characteristics. Another strength is that documentation was available for all patients throughout the study so no one was lost to follow-up. This allowed comprehensive data collection and provided a stronger conclusion given the completeness of the data from baseline to follow-up.

Limitations include the retrospective design and small sample size. In addition, the study design did not allow for randomization. There is no documentation of adherence to medication regimen, which was difficult to determine due to the retrospective nature. Other changes to the patients’ medication regimen were not collected in aggregate and thus, it is possible the total daily insulin dose was impacted by other medication changes. There is also potential for inconsistent documentation of the patients’ true total daily insulin dose in the medical record, thus leading to inaccuracy of recorded data.

Conclusions

A small sample of veterans with T2DM had statistically significant reductions in total daily insulin dose when subcutaneous semaglutide was initiated, as well as after each dose increase. There was also a statistically significant reduction in HbA_1c levels from baseline even as patient insulin doses were reduced. These results support the current practice of using semaglutide to treat T2DM, suggesting it may be safe and effective at reducing HbA_1c levels as the dose is titrated up to 2 mg. There was no statistically significant change in the number of hypoglycemic events reported as semaglutide was titrated up. Thus, when semaglutide is increased to the maximum recommended dose of 2 mg for T2DM, patients may experience a reduction of their total daily dose of insulin and HbA_1c levels. These benefits may reduce the risk of insulin-related AEs while maintaining appropriate glycemic control.

Type 2 diabetes mellitus (T2DM) is a chronic disease becoming more prevalent each year and is the seventh-leading cause of death in the United States.¹ The most common reason for hospitalization for patients with T2DM is uncontrolled glycemic levels.² Nearly 25% of the US Department of Veterans Affairs (VA) patient population has T2DM.³ T2DM is the leading cause of blindness, end-stage renal disease, and amputation for VA patients.⁴

According to the 2023 American Diabetes Association (ADA) guidelines, treatment goals of T2DM include eliminating symptoms, preventing or delaying complications, and attaining glycemic goals. A typical hemoglobin A_1c (HbA_1c) goal range is < 7%, but individual goals can vary up to < 9% due to a multitude of factors, including patient comorbidities and clinical status.⁵

Initial treatment recommendations are nonpharmacologic and include comprehensive lifestyle interventions such as optimizing nutrition, physical activity, and behavioral therapy. When pharmacologic therapy is required, metformin is the preferred first-line treatment for the majority of newly diagnosed patients with T2DM and should be added to continued lifestyle management.⁵ If HbA_1c levels remains above goal, the 2023 ADA guidelines recommend adding a second medication, including but not limited to insulin, a glucagonlike peptide-1 receptor agonist (GLP-1RA), or a sodium-glucose cotransporter 2 inhibitor. Medication choice is largely based on the patient’s concomitant conditions (eg, atherosclerotic cardiovascular disease, heart failure, or chronic kidney disease). The 2023 ADA guidelines suggest initiating insulin therapy when a patient's blood glucose ≥ 300 mg/dL, HbA_1c > 10%, or if the patient has symptoms of hyperglycemia, even at initial diagnosis. Initiating medications to minimize or avoid hypoglycemia is a priority, especially in high-risk individuals.⁵

Clinical evidence shows that GLP-1RAs may provide similar glycemic control to insulin with lower risk of hypoglycemia.⁶ Other reported benefits of GLP-1RAs include weight loss, blood pressure reduction, and improved lipid levels. The most common adverse events (AEs) with GLP-1RAs are gastrointestinal. Including GLP-1RAs in T2DM pharmacotherapy may lower the risk of hypoglycemia, especially in patients at high risk of hypoglycemia.

The 2023 ADA guidelines indicate that it is appropriate to initiate GLP-]1RAs in patients on insulin.⁵ However, while GLP-1RAs do not increase the risk of hypoglycemia independently, combination treatment with GLP-1RAs and insulin can still result in hypoglycemia.⁶ Insulin is the key suspect of this hypoglycemic risk.⁷ Thus, if insulin dosage can be reduced or discontinued, this might reduce the risk of hypoglycemia.

The literature is limited on how the addition of a GLP-1RA to insulin treatment will affect the patient's daily insulin doses, particularly for the veteran population. The goal of this study is to examine this gap in current research by examining semaglutide, which is the current formulary preferred GLP-1RA at the VA.

Semaglutide is subcutaneously initiated at a dose of 0.25 mg once weekly for 4 weeks to reduce gastrointestinal symptoms, then increased to 0.5 mg weekly. Additional increases to a maintenance dose of 1 mg or 2 mg weekly can occur to achieve glycemic goals. The SUSTAIN-FORTE randomized controlled trial sought to determine whether there was a difference in HbA_1c level reduction and significant weight loss with the 2-mg vs 1-mg dose.⁸ Patients in the trial were taking metformin but needed additional medication to control their HbA_1c. They were not using insulin and may or may not have been taking sulfonylureas prior to semaglutide initiation. Semaglutide 2 mg was found to significantly improve HbA_1c control and promote weight loss compared with semaglutide 1 mg, while maintaining a similar safety profile.

Because this study involved patients who required additional HbA_1c control, although semaglutide reduced HbA_1c, not all patients were able to reduce their other diabetes medications, which depended on the baseline HbA_1c level and the level upon completion of semaglutide titration. Dose reductions for the patients’ other T2DM medications were not reported at trial end. SUSTAIN-FORTE established titration up to semaglutide 2 mg as effective for HbA_1c reduction, although it did not study patients also on insulin.⁸

Insulin is associated with hypoglycemic risk, weight gain, and other AEs.^7,8 This study analyzed whether increasing semaglutide could reduce insulin doses and therefore reduce risk of AEs in patients with T2DM.

Methods

A retrospective, single-center, chart review was conducted at VA Sioux Falls Health Care System (VASFHCS). Data were collected through manual review of VASFHCS electronic medical records. Patients aged ≥ 18 years with active prescriptions for at least once-daily insulin who were initiated on 2-mg weekly dose of semaglutide at the VASFHCS clinical pharmacy practitioner medication management clinic between January 1, 2021, and September 1, 2023, were included. VASFHCS clinical pharmacy practitioners have a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics.

The most frequently used prandial insulin at VASFHCS is insulin aspart, and the most frequently used basal insulin is insulin glargine. Patients were retrospectively monitored as they progressed from baseline (the point in time where semaglutide 0.5 mg was initiated) to ≥ 3 months on semaglutide 2-mg therapy. Patients were excluded if they previously used a GLP-1RA or if they were on sliding scale insulin without an exact daily dosage.

The primary endpoint was the percent change in total daily insulin dose from baseline to each dose increase after receiving semaglutide 2 mg for ≥ 3 months. Secondary endpoints included changes in daily prandial insulin dose, daily basal insulin dose, HbA_1c, and number of hypoglycemic events reported. Data collected included age, race, weight, body mass index, total daily prandial insulin dose, total daily basal insulin dose, HbA_1c, and hypoglycemic events reported at the visit when semaglutide was initiated.

Statistical Analysis

The sample size was calculated prior to data collection, and it was determined that for α = .05, 47 patients were needed to achieve 95% power. The primary endpoint was assessed using a paired t test, as were each secondary endpoint. Results with P < .05 were considered statistically significant.

Results

Sixty-two patients were included. The mean HbA_1c level at baseline was 7.7%, the baseline mean prandial and insulin daily doses were 41.5 units and 85.1 units, respectively (Table 1) From baseline to initiation of a semaglutide 1-mg dose, the daily insulin dose changed –5.6% (95% CI, 2.2-14.0; P = .008). From baseline to 2-mg dose initiation daily insulin changed -22.2% (95% CI, 22.0-35.1; P < .001) and for patients receiving semaglutide 2 mg for ≥ 3 months it changed -36.9% (95% CI, 37.4-56.5; P < .001) (Figure).

After receiving the 2-mg dose for ≥ 3 months, the mean daily dose of prandial insulin decreased from 41.5 units to 24.6 units (95% CI, 12.6-21.2; P < .001); mean daily dose of basal insulin decreased from 85.1 units to 52.1 units (95% CI, 23.9-42.0; P < .001); and mean HbA_1c level decreased from 7.7% to 7.1% (95% CI, 0.3-0.8; P < .001). Mean number of hypoglycemic events reported was not statistically significant, changing from 3.6 to 3.2 (95% CI, –0.6 to 0.1; P = .21) (Table 2).

Discussion

This study investigated the effect of subcutaneous semaglutide dose escalation on total daily insulin dose for patients with T2DM. There was a statistically significant decrease in total daily insulin dose from baseline to 1 mg initiation; this decrease continued with further insulin dose reduction seen at the 2-mg dose initiation and additional insulin dose reduction at ≥ 3 months at this dose. It was hypothesized there would be a significant total daily insulin dose reduction at some point, especially when transitioning from the semaglutide 1-mg to the 2-mg dose, based on previous research. ^9,10 The additional reduction in daily insulin dose when continuing on semaglutide 2 mg for ≥ 3 months was an unanticipated but added benefit, showing that if tolerated, maintaining the 2-mg dose will help patients reduce their insulin doses.

In terms of secondary endpoints, there was a statistically significant decrease in mean total daily dose individually for prandial and basal insulin from baseline to ≥ 3 months after semaglutide 2 mg initiation. The change in HbA_1c level was also statistically significant and decreased from baseline, even as insulin doses were reduced. This change in HbA_1c level was expected; previous literature has shown a significant link between improving HbA_1c control when semaglutide doses are increased to 2 mg weekly.¹⁰ Due to having been shown in previous trials, it was expected that HbA_1c levels would decrease even when the insulin doses were being reduced.¹⁰ Insulin dose reduction can potentially be added to the growing evidence of semaglutide benefits. The change in the number of hypoglycemic events was not statistically significant, which was unexpected since previous research show a trend in patients taking GLP-1RAs having fewer hypoglycemic events than those taking insulin.⁶ Further investigation with a larger sample size and prospective trial could determine whether this result is an outlier. In this study, there was no increase in HbA_1c or hypoglycemic events reported with increasing semaglutide doses, which provides further evidence of the safety of semaglutide even at higher doses.

These data suggest that for a patient with T2DM who is already taking insulin, the recommended titration of semaglutide is to start with 0.5 mg and titrate up to a 2-mg subcutaneous weekly dose and to then continue at that dose. As long as the 2-mg dose is tolerated, it will provide patients with the most HbA_1c control and lead to a reduction of their total daily insulin doses according to these results.

Strengths and Limitations

This study compared patient data at different points. This method did not require a second distinct control group, which would potentially introduce confounding factors, such as different baseline characteristics. Another strength is that documentation was available for all patients throughout the study so no one was lost to follow-up. This allowed comprehensive data collection and provided a stronger conclusion given the completeness of the data from baseline to follow-up.

Limitations include the retrospective design and small sample size. In addition, the study design did not allow for randomization. There is no documentation of adherence to medication regimen, which was difficult to determine due to the retrospective nature. Other changes to the patients’ medication regimen were not collected in aggregate and thus, it is possible the total daily insulin dose was impacted by other medication changes. There is also potential for inconsistent documentation of the patients’ true total daily insulin dose in the medical record, thus leading to inaccuracy of recorded data.

Conclusions

A small sample of veterans with T2DM had statistically significant reductions in total daily insulin dose when subcutaneous semaglutide was initiated, as well as after each dose increase. There was also a statistically significant reduction in HbA_1c levels from baseline even as patient insulin doses were reduced. These results support the current practice of using semaglutide to treat T2DM, suggesting it may be safe and effective at reducing HbA_1c levels as the dose is titrated up to 2 mg. There was no statistically significant change in the number of hypoglycemic events reported as semaglutide was titrated up. Thus, when semaglutide is increased to the maximum recommended dose of 2 mg for T2DM, patients may experience a reduction of their total daily dose of insulin and HbA_1c levels. These benefits may reduce the risk of insulin-related AEs while maintaining appropriate glycemic control.

Type 2 diabetes mellitus (T2DM) is a chronic disease becoming more prevalent each year and is the seventh-leading cause of death in the United States.¹ The most common reason for hospitalization for patients with T2DM is uncontrolled glycemic levels.² Nearly 25% of the US Department of Veterans Affairs (VA) patient population has T2DM.³ T2DM is the leading cause of blindness, end-stage renal disease, and amputation for VA patients.⁴

According to the 2023 American Diabetes Association (ADA) guidelines, treatment goals of T2DM include eliminating symptoms, preventing or delaying complications, and attaining glycemic goals. A typical hemoglobin A_1c (HbA_1c) goal range is < 7%, but individual goals can vary up to < 9% due to a multitude of factors, including patient comorbidities and clinical status.⁵

Initial treatment recommendations are nonpharmacologic and include comprehensive lifestyle interventions such as optimizing nutrition, physical activity, and behavioral therapy. When pharmacologic therapy is required, metformin is the preferred first-line treatment for the majority of newly diagnosed patients with T2DM and should be added to continued lifestyle management.⁵ If HbA_1c levels remains above goal, the 2023 ADA guidelines recommend adding a second medication, including but not limited to insulin, a glucagonlike peptide-1 receptor agonist (GLP-1RA), or a sodium-glucose cotransporter 2 inhibitor. Medication choice is largely based on the patient’s concomitant conditions (eg, atherosclerotic cardiovascular disease, heart failure, or chronic kidney disease). The 2023 ADA guidelines suggest initiating insulin therapy when a patient's blood glucose ≥ 300 mg/dL, HbA_1c > 10%, or if the patient has symptoms of hyperglycemia, even at initial diagnosis. Initiating medications to minimize or avoid hypoglycemia is a priority, especially in high-risk individuals.⁵

Clinical evidence shows that GLP-1RAs may provide similar glycemic control to insulin with lower risk of hypoglycemia.⁶ Other reported benefits of GLP-1RAs include weight loss, blood pressure reduction, and improved lipid levels. The most common adverse events (AEs) with GLP-1RAs are gastrointestinal. Including GLP-1RAs in T2DM pharmacotherapy may lower the risk of hypoglycemia, especially in patients at high risk of hypoglycemia.

The 2023 ADA guidelines indicate that it is appropriate to initiate GLP-]1RAs in patients on insulin.⁵ However, while GLP-1RAs do not increase the risk of hypoglycemia independently, combination treatment with GLP-1RAs and insulin can still result in hypoglycemia.⁶ Insulin is the key suspect of this hypoglycemic risk.⁷ Thus, if insulin dosage can be reduced or discontinued, this might reduce the risk of hypoglycemia.

The literature is limited on how the addition of a GLP-1RA to insulin treatment will affect the patient's daily insulin doses, particularly for the veteran population. The goal of this study is to examine this gap in current research by examining semaglutide, which is the current formulary preferred GLP-1RA at the VA.

Semaglutide is subcutaneously initiated at a dose of 0.25 mg once weekly for 4 weeks to reduce gastrointestinal symptoms, then increased to 0.5 mg weekly. Additional increases to a maintenance dose of 1 mg or 2 mg weekly can occur to achieve glycemic goals. The SUSTAIN-FORTE randomized controlled trial sought to determine whether there was a difference in HbA_1c level reduction and significant weight loss with the 2-mg vs 1-mg dose.⁸ Patients in the trial were taking metformin but needed additional medication to control their HbA_1c. They were not using insulin and may or may not have been taking sulfonylureas prior to semaglutide initiation. Semaglutide 2 mg was found to significantly improve HbA_1c control and promote weight loss compared with semaglutide 1 mg, while maintaining a similar safety profile.

Because this study involved patients who required additional HbA_1c control, although semaglutide reduced HbA_1c, not all patients were able to reduce their other diabetes medications, which depended on the baseline HbA_1c level and the level upon completion of semaglutide titration. Dose reductions for the patients’ other T2DM medications were not reported at trial end. SUSTAIN-FORTE established titration up to semaglutide 2 mg as effective for HbA_1c reduction, although it did not study patients also on insulin.⁸

Insulin is associated with hypoglycemic risk, weight gain, and other AEs.^7,8 This study analyzed whether increasing semaglutide could reduce insulin doses and therefore reduce risk of AEs in patients with T2DM.

Methods

A retrospective, single-center, chart review was conducted at VA Sioux Falls Health Care System (VASFHCS). Data were collected through manual review of VASFHCS electronic medical records. Patients aged ≥ 18 years with active prescriptions for at least once-daily insulin who were initiated on 2-mg weekly dose of semaglutide at the VASFHCS clinical pharmacy practitioner medication management clinic between January 1, 2021, and September 1, 2023, were included. VASFHCS clinical pharmacy practitioners have a scope of practice that allows them to initiate, modify, or discontinue medication therapy within medication management clinics.

The most frequently used prandial insulin at VASFHCS is insulin aspart, and the most frequently used basal insulin is insulin glargine. Patients were retrospectively monitored as they progressed from baseline (the point in time where semaglutide 0.5 mg was initiated) to ≥ 3 months on semaglutide 2-mg therapy. Patients were excluded if they previously used a GLP-1RA or if they were on sliding scale insulin without an exact daily dosage.

The primary endpoint was the percent change in total daily insulin dose from baseline to each dose increase after receiving semaglutide 2 mg for ≥ 3 months. Secondary endpoints included changes in daily prandial insulin dose, daily basal insulin dose, HbA_1c, and number of hypoglycemic events reported. Data collected included age, race, weight, body mass index, total daily prandial insulin dose, total daily basal insulin dose, HbA_1c, and hypoglycemic events reported at the visit when semaglutide was initiated.

Statistical Analysis

The sample size was calculated prior to data collection, and it was determined that for α = .05, 47 patients were needed to achieve 95% power. The primary endpoint was assessed using a paired t test, as were each secondary endpoint. Results with P < .05 were considered statistically significant.

Results

Sixty-two patients were included. The mean HbA_1c level at baseline was 7.7%, the baseline mean prandial and insulin daily doses were 41.5 units and 85.1 units, respectively (Table 1) From baseline to initiation of a semaglutide 1-mg dose, the daily insulin dose changed –5.6% (95% CI, 2.2-14.0; P = .008). From baseline to 2-mg dose initiation daily insulin changed -22.2% (95% CI, 22.0-35.1; P < .001) and for patients receiving semaglutide 2 mg for ≥ 3 months it changed -36.9% (95% CI, 37.4-56.5; P < .001) (Figure).

After receiving the 2-mg dose for ≥ 3 months, the mean daily dose of prandial insulin decreased from 41.5 units to 24.6 units (95% CI, 12.6-21.2; P < .001); mean daily dose of basal insulin decreased from 85.1 units to 52.1 units (95% CI, 23.9-42.0; P < .001); and mean HbA_1c level decreased from 7.7% to 7.1% (95% CI, 0.3-0.8; P < .001). Mean number of hypoglycemic events reported was not statistically significant, changing from 3.6 to 3.2 (95% CI, –0.6 to 0.1; P = .21) (Table 2).

Discussion

This study investigated the effect of subcutaneous semaglutide dose escalation on total daily insulin dose for patients with T2DM. There was a statistically significant decrease in total daily insulin dose from baseline to 1 mg initiation; this decrease continued with further insulin dose reduction seen at the 2-mg dose initiation and additional insulin dose reduction at ≥ 3 months at this dose. It was hypothesized there would be a significant total daily insulin dose reduction at some point, especially when transitioning from the semaglutide 1-mg to the 2-mg dose, based on previous research. ^9,10 The additional reduction in daily insulin dose when continuing on semaglutide 2 mg for ≥ 3 months was an unanticipated but added benefit, showing that if tolerated, maintaining the 2-mg dose will help patients reduce their insulin doses.

In terms of secondary endpoints, there was a statistically significant decrease in mean total daily dose individually for prandial and basal insulin from baseline to ≥ 3 months after semaglutide 2 mg initiation. The change in HbA_1c level was also statistically significant and decreased from baseline, even as insulin doses were reduced. This change in HbA_1c level was expected; previous literature has shown a significant link between improving HbA_1c control when semaglutide doses are increased to 2 mg weekly.¹⁰ Due to having been shown in previous trials, it was expected that HbA_1c levels would decrease even when the insulin doses were being reduced.¹⁰ Insulin dose reduction can potentially be added to the growing evidence of semaglutide benefits. The change in the number of hypoglycemic events was not statistically significant, which was unexpected since previous research show a trend in patients taking GLP-1RAs having fewer hypoglycemic events than those taking insulin.⁶ Further investigation with a larger sample size and prospective trial could determine whether this result is an outlier. In this study, there was no increase in HbA_1c or hypoglycemic events reported with increasing semaglutide doses, which provides further evidence of the safety of semaglutide even at higher doses.

These data suggest that for a patient with T2DM who is already taking insulin, the recommended titration of semaglutide is to start with 0.5 mg and titrate up to a 2-mg subcutaneous weekly dose and to then continue at that dose. As long as the 2-mg dose is tolerated, it will provide patients with the most HbA_1c control and lead to a reduction of their total daily insulin doses according to these results.

Strengths and Limitations

This study compared patient data at different points. This method did not require a second distinct control group, which would potentially introduce confounding factors, such as different baseline characteristics. Another strength is that documentation was available for all patients throughout the study so no one was lost to follow-up. This allowed comprehensive data collection and provided a stronger conclusion given the completeness of the data from baseline to follow-up.

Limitations include the retrospective design and small sample size. In addition, the study design did not allow for randomization. There is no documentation of adherence to medication regimen, which was difficult to determine due to the retrospective nature. Other changes to the patients’ medication regimen were not collected in aggregate and thus, it is possible the total daily insulin dose was impacted by other medication changes. There is also potential for inconsistent documentation of the patients’ true total daily insulin dose in the medical record, thus leading to inaccuracy of recorded data.

Conclusions

A small sample of veterans with T2DM had statistically significant reductions in total daily insulin dose when subcutaneous semaglutide was initiated, as well as after each dose increase. There was also a statistically significant reduction in HbA_1c levels from baseline even as patient insulin doses were reduced. These results support the current practice of using semaglutide to treat T2DM, suggesting it may be safe and effective at reducing HbA_1c levels as the dose is titrated up to 2 mg. There was no statistically significant change in the number of hypoglycemic events reported as semaglutide was titrated up. Thus, when semaglutide is increased to the maximum recommended dose of 2 mg for T2DM, patients may experience a reduction of their total daily dose of insulin and HbA_1c levels. These benefits may reduce the risk of insulin-related AEs while maintaining appropriate glycemic control.