Daily Glucose Profiles after Switching from Injectable to Oral Semaglutide in Patients with Type 2 Diabetes Mellitus

Abstract

Objective This prospective observational study explored the changes in the daily glycemic profile after switching from injectable to oral semaglutide in patients with type 2 diabetes mellitus.

Methods Patients with type 2 diabetes mellitus who were treated with once-weekly 0.5 mg injectable semaglutide and wished to switch to once-daily oral semaglutide participated in this study. Oral semaglutide was initiated at 3 mg and increased to 7 mg a month later, according to the package insert. Before and two months after the switch, participants wore a sensor for continuous glucose monitoring for up to 14 days. We also evaluated the questionnaire-based treatment satisfaction and the preference between the two formulations.

Patients Twenty-three patients participated.

Results Mean glucose levels significantly increased by 9 mg/dL on average, from 132±20 to 141±27 mg/dL (p=0.047), which was equivalent to a change of 0.2% in the estimated hemoglobin A1c (6.5±0.5% to 6.7±0.7%). The inter-individual variability assessed with standard deviation also significantly increased (p=0.004). The change in treatment satisfaction varied considerably among patients, with no specific trend in the overall population. After trying oral semaglutide, 48% of patients responded that they preferred the oral formulation, while 35% preferred the injectable formulation, and 17% had no preference.

Conclusion The mean glucose levels increased by 9 mg/dL on average after switching from once-weekly 0.5 mg injectable semaglutide to once-daily 7 mg oral semaglutide, with an increased inter-individual variability. The change in treatment satisfaction considerably varied among patients.

Introduction

Semaglutide is a glucagon-like peptide 1 (GLP-1) receptor agonist that was originally served as an injectable formation (1). Recently, its oral formation has been developed (2). The bioavailability of orally administered semaglutide is much lower than that of subcutaneously administered semaglutide, so the dosage of the oral formation is set higher (7 mg once daily versus 0.5 mg once weekly for maintenance dose) to guarantee an efficacy equivalent to that of the injectable formation. However, no clinical studies have directly compared the glucose-lowering effect between the oral and injectable formulations at the clinically available dosage.

Another unclear aspect regarding oral versus injectable semaglutide is the difference in treatment satisfaction. Treatment with injectable agents is generally more burdensome and less satisfactory than that with oral agents (3,4). Oral formations are thus expected to be preferred to injectable formations by patients. However, oral semaglutide tablets need to be taken at least 30 minutes before breakfast every day (5), which some patients might find burdensome.

The present prospective observational study explored the changes in the daily glucose profile after switching from injectable to oral semaglutide in patients with type 2 diabetes mellitus who were treated with injectable semaglutide and achieved fairly stable and good glycemic control but wished to switch to the oral formulation in clinical practice. We also investigated the treatment satisfaction after the switch.

Materials and Methods

The current prospective observational study was conducted between December 2021 and September 2022. We registered 23 adult Japanese patients with type 2 diabetes mellitus who were treated with once-weekly 0.5 mg injectable semaglutide and wished to switch the medication to once-daily oral semaglutide. All participants were treated at Shiraiwa Medical Clinic, Kashiwara City, Osaka, Japan. The current study was performed in accordance with the Declaration of Helsinki and was approved by the ethics committees of Osaka University Hospital; the approval number was 21289 (T1). Informed consent was obtained from every participant.

The inclusion criteria were as follows: 1) age ≥20 years old, 2) treated with once-weekly 0.5 mg injectable semaglutide for type 2 diabetes mellitus, and 3) wishing and planning to switch from injectable semaglutide to once-daily oral semaglutide. The exclusion criteria were as follows: 1) semaglutide was scheduled to be stopped due to surgery or other reasons during the study period, 2) other medications were planned to be changed during the study period, 3) patients were not expected to live a regular life with good adherence to diet or exercise therapy, and 4) poor medication adherence was evident. It should be noted that, due to the second to fourth exclusion criteria, patients with poor glycemic control were not included in the present study, and most of the study participants in the present study had relatively stable and fairly good glycemic control.

Before switching to oral semaglutide, participants were asked to wear a FreeStyle Libre Pro™ Flash Glucose Monitoring System sensor (Abbott Japan, Tokyo, Japan) on the back of the upper arm for up to 14 days and to respond to a questionnaire regarding medication satisfaction (see below) (6). After completing the glucose monitoring and the questionnaire, once-weekly injectable semaglutide was discontinued, and oral semaglutide was instead initiated at 3 mg once daily. As advised by the package insert (7), oral semaglutide tablets were taken with up to 120 mL of water, in a fasting state, and at least 30 min before consuming any food, other beverage, or other oral medications. One month later, the dose of oral semaglutide was increased to 7 mg once daily. At two months after switching from injectable to oral semaglutide (i.e., one month after increasing the dose of oral semaglutide to 7 mg once daily), the participants were again asked to wear a continuous glucose monitoring system sensor for up to 14 days and to respond to the questionnaire concerning treatment satisfaction. At two months, the participants were also asked which formulation they preferred, injectable or oral semaglutide.

This study evaluated the clinical outcomes one month after starting treatment with 7 mg oral semaglutide, referring to the pharmacokinetic information of oral semaglutide. From a clinical perspective, 90% of the theoretic steady-state value is often used as a practical definition of steady state (8). Pharmacokinetic studies of oral semaglutide reported the half-life of the drug to be 150-160 h (9,10), and therefore a 1-month administration was expected to be sufficient for the medication to reach >90% of the theoretic steady-state concentration. Indeed, the pharmacokinetic information on repeated administration shown in the package inset of oral semaglutide was derived from data on 28-day dosing (7).

Daily glucose profiles

Daily glucose profiles were measured using the FreeStyle Libre Pro™ Flash Glucose Monitoring System before and two months after switching from injectable to oral semaglutide. According to the international consensus on the continuous glucose monitoring metrics (11), the glucose measurements were summarized as follows: mean glucose, coefficient of variation (CV), time above range (TAR) (>180 mg/dL and >250 mg/dL), time in range (TIR) (70-180 mg/dL), and time below range (TBR) (<70 mg/dL and <54 mg/dL). The mean glucose levels were further mathematically converted to estimated hemoglobin A1c (HbA1c) levels (renamed the “glucose management indicator” in the latest international consensus) using the following recommended formula: 3.31+0.02392×[mean glucose in mg/dL] (11,12). We also calculated the mean amplitude of glucose excursion (MAGE) (13), interquartile range (IQR), and mean of daily difference of blood glucose (MODD) (14) as indices for glycemic variability other than CV. This study evaluated glycemic control using those glucose profiles, rather than directly measured HbA1c levels, simply because HbA1c was expected to reflect glycemic control over two or three months (15), which was longer than the 1 month of 7 mg semaglutide treatment. Directly measured HbA1c levels are provided as Supplementary materials.

Questionnaire concerning medication satisfaction

The participants were asked to respond to a questionnaire concerning medication satisfaction (6) before and two months after switching from injectable to oral semaglutide. The details of the questionnaire were reported previously (6). In brief, the questionnaire contains a total of 10 statements related to treatment satisfaction and asks to what extent a respondent agrees with each statement, based on a 7-point Likert scale ranging from 0, which corresponds to “strongly disagree,” to 6, which corresponds to “strongly agree.” The 10 statements consist of 6 positively and 4 negatively worded ones. A total satisfaction score is calculating by summing the scores from all 10 items, with the rating of the 4 negatively worded items reversed; the possible range of the total score is between 0 (no satisfaction) and 60 (full satisfaction). Details of the questionnaire are provided in Supplementary material 1.

Statistical analyses

Data are given as means and standard deviation (SD) for continuous variables and as percentages for discrete variables, if not otherwise mentioned. A p<0.05 was considered statistically significant. The mean differences in treatment satisfaction, daily glucose profiles, and body weight were examined by the paired t-test. We also tested the difference in SD of those variables using 2,000-time bootstrap resampling. The correlation of those variables with the time interval between taking oral semaglutide and having breakfast was assessed using Spearman's correlation coefficient. Differences in clinical characteristics among more than two groups were tested by a one-way analysis of variance for continuous variables and the chi-squared test for discrete variables, except for the time interval between taking oral semaglutide and having breakfast, which was tested by the Kruskal-Wallis test. Differences in the change in glucose levels among 24-h time points were tested using an analysis of variance for the linear mixed model, where inter-subject variability was treated as a random effect.

All statistical analyses were performed with the R software program, version 4.1.1 (R Development Core Team, Vienna, Austria).

Results

Table 1 summarizes baseline clinical characteristics of the study population. The mean duration of diabetes was 12±9 years, and the mean HbA1c level at baseline was 6.7±0.6%. All participants completed 1-month treatment with 3 mg oral semaglutide and subsequent 1-month treatment with 7 mg oral semaglutide. Two patients responded that they woke up earlier after the switch to oral semaglutide to take the tablets at least 30 minutes before their breakfast (they originally had breakfast within 30 minutes after waking up). The remaining 21 patients originally had breakfast at least 30 minutes after waking up and so did not change their wake-up time or breakfast time. The time interval between taking oral semaglutide and having breakfast was 30 to <60 minutes in 9 patients (39%), 60 to <120 minutes in 7 patients (30%), 120 to <180 minutes in 4 patients (17%), and ≥180 minutes in 3 patients (13%). At two months, all participants completed the questionnaire, but data on daily glucose profiles were unavailable in five patients because a sensor fell off early in two patients (the first and second days), and three patients refused to wear a sensor due to itching and discomfort.

Table 1.

Clinical Characteristics of Study Population.

n		23
Age (years)		59±12
Male sex		14 (61%)
Body mass index (kg/m2)		27.7±4.8
Duration of diabetes (years)		12±9
Hemoglobin A1c (%)		6.7±0.6
Injectable semaglutide use (0.5 mg once weekly)		23 (100%)
Metformin use		12 (52%)
Sodium-glucose cotransporter 2 inhibitor use		5 (22%)
Sulfonylurea use		1 (4%)
Other hypoglycemic agent use		0 (0%)
Dyslipidemia		16 (70%)
Hypertension		16 (70%)
Types of oral medications (per day)		3.2±2.0
Frequency of taking oral medications (per day)		1.7±1.0

Figure illustrates the 24-h glucose levels before and after switching. The change in glucose levels was not significantly affected by time (p>0.999). As summarized in Table 2, the mean glucose levels significantly increased by 9 mg/dL, from 132±20 to 141±27 mg/dL (p=0.047), which was equivalent to a mean change of 0.2% in the estimated HbA1c, from 6.5±0.5% to 6.7±0.7% (p=0.047). In addition, the mean TAR (>250 mg/dL) significantly increased from 0.9±1.6% to 2.4±3.3% (p=0.040). The SDs of those metrics also significantly increased (all p<0.05). Time in range (70-180 mg/dL) was 87.1±6.6% and 81.1±14.3% before and after the switch (p=0.059 for mean and p<0.001 for SD), whereas time above range (>180 mg/dL) was 11.4±7.7% and 17.5±15.2% before and after the switch, respectively (p=0.052 for mean and p=0.002 for SD). The other metrics of daily glucose profiles and body weight showed no significant changes in the mean or SD values (all p>0.05). Changes in directly measured HbA1c levels are provided in Supplementary material 2.

Figure.

24-h glucose levels before and after the switch. Data are presented as the means and standard deviations of glucose levels measured by the FreeStyle Libre Pro^TM Flash Glucose Monitoring System under treatment with once-weekly 0.5 mg injectable semaglutide (gray points and error bars) and once-daily 7 mg oral semaglutide (black points and error bars).

Table 2.

Change of Daily Glucose Profile, Body Weight, and Treatment Satisfaction.

	Before switch	After switch	Mean difference	p value for mean	p value for standard deviation
Daily glucose profile*
Mean glucose (mg/dL)	132±20	141±27	9±18	0.047	0.007
Converted to estimated hemoglobin A1c (%)	6.5±0.5	6.7±0.7	0.2±0.4	0.047	0.007
Coefficient of variation (%)	24.8±4.9	24.8±3.6	0.0±3.4	0.967	0.060
Time in range (70-180 mg/dL) (%)	87.1±6.6	81.1±14.3	-6.0±12.5	0.059	<0.001
Time above range (>180 mg/dL) (%)	11.4±7.7	17.5±15.2	6.0±12.3	0.052	0.002
Time above range (>250 mg/dL) (%)	0.9±1.6	2.4±3.3	1.5±2.9	0.040	0.006
Time below range (<70 mg/dL) (%)	1.5±2.6	1.4±4.4	-0.1±2.9	0.921	0.717
Time below range (<54 mg/dL) (%)	0.1±0.2	0.2±0.6	0.1±0.5	0.533	0.487
Mean amplitude of glucose excursion (mg/dL)	50.4±13.0	53.9±15.2	3.5±11.7	0.220	0.363
Interquartile range (mg/dL)	42.8±14.0	47.2±15.5	4.4±10.3	0.087	0.450
Mean of daily differences (mg/dL)	28.9±10.6	32.7±12.9	3.8±12.0	0.196	0.314
Body weight (kg)	74.3±13.6	74.5±13.9	0.2±1.2	0.413	0.065
Questionnaire for treatment satisfaction (points)
1. Convenience	4.8±1.1	4.4±2.0	-0.4±2.5	0.467	0.010
2. Encouragement of adherence	5.0±1.1	4.6±2.1	-0.4±2.5	0.458	0.025
3. Control over diseases	4.8±1.0	4.3±1.6	-0.5±1.5	0.141	0.027
4. Feeling of healthiness	4.6±1.5	4.4±1.7	-0.2±1.5	0.575	0.087
5. Troublesomeness	2.0±1.4	2.1±1.9	0.2±1.7	0.638	0.013
6. Fear of forgetting to take medicines	1.8±1.7	1.4±1.6	-0.4±2.3	0.433	0.676
7. Suspicion about efficacy	2.0±2.1	2.2±1.8	0.1±1.8	0.734	0.160
8. Undesirably weakened disease awareness	1.6±1.6	1.9±1.7	0.3±1.8	0.431	0.629
9. Satisfaction	4.6±1.2	4.4±1.8	-0.2±1.9	0.672	0.059
10. Hope for receiving the treatment	4.4±1.3	4.4±1.9	0.0±2.0	1.000	0.069
Total satisfaction score	44.9±8.9	43.1±12.2	-1.8±13.0	0.507	0.128

*Data on daily glucose profile were available in 18 patients.

As Table 2 shows, the total satisfaction score changed from 44.9±8.9 points to 43.1±12.2 points, with no specific trend in the overall population (p=0.51). The distribution of the change was -1.8±13.0 points; its standard deviation (i.e. 13.0 points) was relatively large compared with the SD of the absolute value (i.e. 8.9 and 12.2 points before and after the switch), indicating that the change in treatment satisfaction varied considerably among patients. The SD of ratings for some questionnaire items significantly increased after the switch, also suggesting an increased variety of treatment satisfaction after the switch. Indeed, after trying oral semaglutide, 11 patients (48%) responded that they preferred treatment with oral semaglutide to that with injectable semaglutide, whereas 8 patients (35%) responded that they preferred treatment with injectable semaglutide. The remaining 4 patients (17%) had no preference. The preference had no significant association with the baseline clinical parameters or the changes in daily glucose profile, appetite, or gastrointestinal symptoms (Table 3).

Table 3.

Clinical Parameters by Medication Preference.

	Patients preferring injectable semeglutide (n=8)	Patients having no preference (n=4)	Patients preferring oral semeglutide (n=11)	p value
Age (years)	58±12	68±8	56±12	0.72
Male sex	4 (50%)	3 (75%)	7 (64%)	0.68
Body mass index (kg/m2)	27.9±4.5	29.3±7.3	26.9±4.5	0.65
Duration of diabetes (years)	10±8	24±4	10±7	0.94
Hemoglobin A1c (%)	6.7±0.6	6.6±0.3	6.7±0.7	0.74
Dyslipidemia	6 (75%)	4 (100%)	6 (55%)	0.22
Hypertension	5 (62%)	4 (100%)	7 (64%)	0.35
Types of oral medications (per day)	2.5±1.8	4.4±1.8	3.4±2.1	0.41
Frequency of taking oral medications (per day)	1.8±1.3	2.0±0.0	1.6±0.9	0.78
Change of wake-up time
Waking up earlier after the switch	2 (25%)	0 (0%)	0 (0%)	0.13
Time interval to breakfast				0.43
≥30 and <60 minutes	4 (50%)	2 (50%)	3 (27%)
≥60 and <120 minutes	2 (25%)	2 (50%)	3 (27%)
≥120 and <180 minutes	0 (0%)	0 (0%)	4 (36%)
≥180 minutes	2 (25%)	0 (0%)	1 (9%)
Change of daily glucose profile
Mean glucose (mg/dL)	15±11	3±6	7±24	0.45
Estimated hemoglobin A1c (%)	0.4±0.3	0.1±0.1	0.2±0.6	0.45
Coefficient of variation (%)	0.8±2.9	-2.5±0.3	0.2±4.0	0.86
Time in range (70-180 mg/dL) (%)	-10.6±11.5	2.0±4.3	-5.6±4.4	0.53
Time above range (>180 mg/dL) (%)	10.7±11.3	0.1±2.6	4.9±14.4	0.43
Time above range (>250 mg/dL) (%)	2.6±3.0	-0.4±0.6	1.5±3.2	0.56
Time below range (<70 mg/dL) (%)	-0.2±0.5	-2.1±3.0	0.7±3.7	0.53
Time below range (<54 mg/dL) (%)	0.0±0.0	0.0±0.0	0.1±0.7	0.63
Mean amplitude of glucose excursion (mg/dL)	5.9±9.2	-5.0±3.5	4.8±14.3	0.95
Interquartile range (mg/dL)	9.8±8.8	3.3±8.1	1.2±11.3	0.12
Mean of daily differences (mg/dL)	10.3±10.5	-1.4±7.0	1.2±13.3	0.18
(Missing data)	2 (25%)	1 (25%)	2 (18%)	0.92
Change of appetite				0.29
Reduced	2 (25%)	1 (25%)	0 (0%)
Unchanged	4 (50%)	2 (50%)	10 (91%)
Increased	2 (25%)	1 (25%)	1 (9%)
Change of gastrointestinal symptoms
Nausea				1.00
Improved	0 (0%)	0 (0%)	0 (0%)
Unchanged	8 (100%)	4 (100%)	11 (100%)
Deteriorated	0 (0%)	0 (0%)	0 (0%)
Abdominal discomfort				0.083
Improved	0 (0%)	1 (25%)	0 (0%)
Unchanged	8 (100%)	3 (75%)	11 (100%)
Deteriorated	0 (0%)	0 (0%)	0 (0%)
Constipation				0.51
Improved	0 (0%)	1 (25%)	1 (9%)
Unchanged	8 (100%)	3 (75%)	9 (82%)
Deteriorated	0 (0%)	0 (0%)	1 (9%)
Diarrhea				0.56
Improved	0 (0%)	0 (0%)	1 (9%)
Unchanged	7 (88%)	4 (100%)	10 (91%)
Deteriorated	1 (12%)	0 (0%)	0 (0%)

The time interval between taking oral semaglutide and having breakfast was not significantly correlated with any clinical outcomes (Supplementary material 3).

Discussion

The maintenance doses for injectable and oral semaglutide are 0.5 mg once weekly and 7 mg once daily, respectively (7,16). This study investigated the changes in the daily glucose profile after switching from once-weekly 0.5 mg injectable to once-daily 7 mg oral semaglutide in patients with type 2 diabetes mellitus who had been treated with injectable semaglutide and achieved fairly stable and good glycemic control but wished and tried to switch to oral semaglutide. The mean glucose levels significantly increased by 9 mg/dL after the switch, with a significant increase noted in TAR (>250 mg/dL). In addition, the SDs of the mean glucose levels, TIR, and TAR also significantly increased. We further investigated the changes in treatment satisfaction, with no specific trend noted in the overall population, although it varied considerably among patients. After trying oral semaglutide, 48% of patients responded that they preferred oral semaglutide, 35% preferred injectable semaglutide, and 17% had no preference. No clinical parameters were associated with this preference.

A phase 3 trial on injectable semaglutide monotherapy (1) demonstrated that once-weekly 0.5 mg injectable semaglutide had a 1.43% greater reduction in HbA1c (95% confidence interval, 1.26-1.65%) than placebo at 30 weeks. In contrast, a phase 3 trial on oral semaglutide monotherapy (2) reported that once-daily 7 mg oral semaglutide reduced HbA1c by 0.9% (0.6-1.1%) more than placebo at 26 weeks. The HbA1c reduction reported in the trial on oral semaglutide (2) seemed slightly smaller than that demonstrated with injectable semaglutide (1). However, it should be noted that the two trials had non-identical patient backgrounds and different assessment periods. A simple comparison of the two trial results would carry a risk of bias. Similarly, when a phase 3 trial on oral semaglutide as an add-on to insulin±metformin (17) was compared with the corresponding trial on injectable semaglutide (18), the HbA1c reduction reported in the former again appeared smaller than that in the latter [estimated treatment difference from placebo, -0.9% (-1.1 to -0.7%) at 26 weeks vs. -1.35% (-1.61 to -1.10%) at 30 weeks]. However, it should be again noted that a simple comparison of the two trial results would be biased by non-identical patient backgrounds and different assessment periods.

The present study demonstrated that mean glucose levels significantly increased by 9 mg/dL after switching from once-weekly 0.5 mg injectable semaglutide to once-daily 7 mg oral semaglutide, with a significant increase noted in the TAR (>250 mg/dL). These findings were in line with the smaller HbA1c reduction reported in the trials on oral semaglutide than in those on injectable semaglutide (1,2,17,18), The glucose-lowering effect of oral semaglutide thus appears to be inferior to that of the corresponding dosage of injectable semaglutide. The current findings are also consistent with those of a recent report that the mean blood concentration of semaglutide was lower under once-daily oral administration of 7 mg semaglutide than under once-weekly subcutaneous injection of 0.5 mg semaglutide (19).

The increase in the mean glucose levels by 9 mg/dL was equivalent to an increase in the estimated HbA1c levels by 0.2%. In general, the clinically meaningful difference of HbA1c levels is regarded as at least 0.3 to 0.4% (20). The glucose elevation observed in the present study was smaller than this minimal clinically meaningful difference, suggesting that the inferiority of oral to injectable semaglutide with regard to the glucose-lowering effect might not always be clinically important. However, the SDs of the mean glucose levels, estimated HbA1c levels, TIR, and TAR significantly increased after the switch, indicating that the inter-individual variability of glycemic control increased after the switch to oral semaglutide. This increased inter-individual variability can be explained by the pharmacokinetic characteristics of this medication. The inter-individual variability of the blood concentration of semaglutide is reportedly greater for the oral administration of semaglutide than for its subcutaneous injection (19). The increased inter-individual variability of pharmacokinetics would thus lead to an increased inter-individual variability of glycemic control.

In addition to the glucose-lowering effect, treatment satisfaction is another key aspect in the management of type 2 diabetes mellitus. Oral agents are generally superior to injectable agents in terms of treatment satisfaction (3,4), so we suspected that this dogma might hold true for oral versus injectable semaglutide. However, interestingly, the change in treatment satisfaction after the switch demonstrated no specific trend in the overall population (p=0.507), although all study participants had wished to switch from injectable to oral semaglutide. After trying oral semaglutide, 48% of participants responded that they preferred oral semaglutide, 35% preferred injectable semaglutide, and 17% had no preference. While oral semaglutide can relieve the burden of injection, the switch from injectable to oral semaglutide increases the dose frequency from once weekly to once daily, which would attenuate the improvement of treatment satisfaction in some patients (21). Furthermore, oral semaglutide tablets need to be taken at least 30 minutes before breakfast. Some patients might find this inconvenient regimen to be another burden (22). No clinical parameters were significantly associated with a preference between oral and injectable semaglutide, suggesting that asking patients to try both two formulations and select their preferred one would be a practical way to administer this treatment regimen with more satisfaction.

Several limitations associated with the present study warrant mention. First, this was a single-arm, observational study with a limited sample size. Future clinical trials with a larger sample size should validate the current findings. Second, the present study did not provide data on the comparison between once-weekly 0.5 mg injectable semaglutide and once-daily 14 mg oral semaglutide. The mean blood concentration of semaglutide after injection of 0.5 mg was reported to be close to that after oral administration of 14 mg (19). However, according to the package insert of oral semaglutide, the maintenance dose is 7 mg once daily, and the dose may be increased to 14 mg once daily if glycemic control is insufficient. In other words, patients who achieve their target of glycemic control with 7 mg once daily will not take 14 mg once daily in clinical practice. We conducted the present study as an observational, non-interventional study. Within the framework of the observational study, it was impossible to increase the dose to 14 mg once daily for all study participants regardless of their glycemic control under 7 mg once daily. The present study simply presents the data on the comparison between the maintenance dose of injectable semaglutide and that of oral semaglutide. Third, the present study did not adopt diabetes-specific questionnaires. Treatment satisfaction is considerably influenced by the medication regimens patients are prescribed. Patients with type 2 diabetes mellitus often have other chronic diseases and take relevant medications in addition to antidiabetic medications, so for our patients, diabetes was likely merely part of what they were being treated for. We wanted to evaluate whether the overall treatment satisfaction, rather than the diabetes-specific treatment satisfaction, would change after the switch. For this reason, we did not adopt diabetes-specific questionnaires. We believe that the overall treatment satisfaction is of particular clinical importance. However, at the same time, not providing data on diabetes-specific questionnaire results is a limitation of the present study. Fourth, the present study targeted patients who achieved relatively stable and fairly good glycemic control under injection of semaglutide. Patients with poor glycemic control were not included in the present study, which caused selection bias. In patients who do not have good glycemic control to begin with, the degree of deterioration might be much greater than revealed in the present study. Fifth, we did not directly monitor patients' adherence to the proper way to take oral semaglutide. Oral semaglutide needs to be taken early in the morning on an empty stomach and at least 30 minutes before the meal. Not following this protocol would considerably affect glycemic control. Although we confirmed patients' adherence by their self-report, we had no means of directly monitoring their adherence. Sixth, the present study did not assess the pharmacokinetics of the study medications. The association between the change in daily glycemic profiles and the pharmacokinetics change thus remains unknown. Seventh, this study analyzed patients who were originally treated with injectable semaglutide and tried to switch to oral semaglutide. Whether or not similar findings would be observed in patients who were originally treated with oral semaglutide and tried switching to injectable semaglutide is unclear.

In conclusion, this study investigated the changes in the daily glucose profile after switching from once-weekly 0.5 mg injectable to once-daily 7 mg oral semaglutide in patients with type 2 diabetes mellitus who were treated with injectable semaglutide and wished and tried to switch to oral semaglutide. The mean glucose levels significantly increased by 9 mg/dL after the switch, with a significant increase seen in TAR (>250 mg/dL). We also investigated the change in treatment satisfaction, which varied considerably among patients. After trying oral semaglutide, 48% of patients responded that they preferred oral semaglutide, 35% preferred injectable semaglutide, and 17% had no preference. No clinical parameters were associated with the preference.

Author's disclosure of potential Conflicts of Interest (COI).

Mitsuyoshi Takahara: Endowed chair, MSD. Junji Kozawa: Endowed chair, MSD. Iichiro Shimomura: Honoraria, Novo Nordisk Pharma; Scholarship grants, Novo Nordisk Pharma.

Supplementary Material

Supplementary material 1.

Questionnaire for treatmentThe questionnaire.

Supplementary material 2.

Change of directly measured hemoglobin A1c levels after the switch in individual study participants.

Supplementary material 3.

Correlation with time interval to breakfast