This Study Is To Determine If Inhaled Insulin Is Effective In Treating Type 2 Diabetes Mellitus - Study Results

Study Type:	Interventional
Study Design:	Randomized, Open Label, Parallel Assignment
Condition:	Diabetes Mellitus, Type 2
Interventions:	Drug: Insulin glargine Drug: Inhaled Insulin (Exubera)

Key information relevant to the recruitment process for the overall study, such as dates of the recruitment period and locations
No text entered.

Significant events and approaches for the overall study following participant enrollment, but prior to group assignment
No text entered.

	Inhaled Human Insulin	Insulin Glargine
STARTED	219	194
Received Treatment	212	190
COMPLETED	136	152
NOT COMPLETED	83	42
Adverse Event	7	1
Lack of Efficacy	1	3
Lost to Follow-up	10	4
Withdrawal by Subject	31	7
Unknown	27	23
Randomized but did not receive treatment	7	4

	Inhaled Human Insulin	Insulin Glargine	Total
Number of Participants [units: participants]	212	190	402
Age, Customized [units: participants]
18 to 44 years	27	26	53
45 to 65 years	151	125	276
>=65 years	34	39	73
Gender [units: participants]
Female	92	84	176
Male	120	106	226

Measure Type	Primary
Measure Title	Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the Per Protocol (PP) Population
Measure Description	HbA1c lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Per protocol (PP)=Full Analysis Set (FAS) subjects (i.e., had >=1 study drug dose and >=1 post-baseline measurement) with >=12 weeks treatment and no major protocol violation. LOCF imputed missing data with any post-baseline visit. Number of subjects with HbA1c values at Baseline and Week 26: inhaled human insulin n=154, insulin glargine n=157.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	154	157
Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the Per Protocol (PP) Population [units: percent] Mean ± Standard Deviation	-2.01 ± 1.39	-1.75 ± 1.40

Statistical Analysis 1 for Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the Per Protocol (PP) Population

Groups ^[1]	All groups
Non-Inferiority/Equivalence Test ^[2]	Yes
Method ^[3]	ANCOVA
P Value ^[4]	<0.0001
Mean Difference (Final Values) ^[5]	-0.19
95% Confidence Interval	( -0.38 to 0 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	The null hypothesis was that the inhaled human insulin group was inferior to the insulin glargine group with respect to 26-week change from baseline in HbA1c, while the alternate hypothesis is that the inhaled human insulin group was not inferior to the insulin glargine group with respect to 26-week change from baseline in HbA1c, given the predetermined non-inferiority margin of 0.4%.
[2]	Details of power calculation, definition of non-inferiority margin, and other key parameters:
	The non-inferiority test utilized a one-sided 2.5% level of significance.
[3]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[4]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[5]	Other relevant estimation information:
	A confidence interval (CI) approach was presented with a two-sided 95% CI of the difference between treatment and control.

Statistical Analysis 2 for Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the Per Protocol (PP) Population

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0516
Mean Difference (Final Values) ^[4]	-0.19
95% Confidence Interval	( -0.38 to 0 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	With an established non-inferiority claim, an additional test for superiority was conducted (ie, the margin was 0.0%).
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	The superiority test used a one-sided 2.5% level of significance.

Measure Type	Secondary
Measure Title	Percentage of Subjects Achieving Glycemic Control (HbA1c < 6.5%) at Week 26
Measure Description	Percentage of subjects with glycosylated hemoglobin A1c lab value less than 6.5%.
Time Frame	Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with HbA1c < 6.5% at Week 26: inhaled human insulin n=72, insulin glargine n=42.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Percentage of Subjects Achieving Glycemic Control (HbA1c < 6.5%) at Week 26 [units: percentage of participants]	35.5	22.2

Statistical Analysis 1 for Percentage of Subjects Achieving Glycemic Control (HbA1c < 6.5%) at Week 26

Groups ^[1]	All groups
Method ^[2]	Regression, Logistic
P Value ^[3]	0.0052
Odds Ratio (OR) ^[4]	1.96
95% Confidence Interval	( 1.22 to 3.14 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Percentage of Subjects Achieving Glycemic Control (HbA1c < 7.0%) at Week 26
Measure Description	Percentage of subjects with glycosylated hemoglobin A1c lab value less than 7.0%.
Time Frame	Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with HbA1c < 7.0% at Week 26: inhaled human insulin n=127, insulin glargine n=90.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Percentage of Subjects Achieving Glycemic Control (HbA1c < 7.0%) at Week 26 [units: percentage of participants]	62.6	47.6

Statistical Analysis 1 for Percentage of Subjects Achieving Glycemic Control (HbA1c < 7.0%) at Week 26

Groups ^[1]	All groups
Method ^[2]	Regression, Logistic
P Value ^[3]	0.0022
Odds Ratio (OR) ^[4]	2.01
95% Confidence Interval	( 1.29 to 3.15 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Percentage of Subjects Achieving Glycemic Control (HbA1c < 8.0%) at Week 26
Measure Description	Number of subjects with glycosylated hemoglobin A1c lab value less than 8.0%.
Time Frame	Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with HbA1c < 8.0% at Week 26: inhaled human insulin n=162, insulin glargine n=158.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Percentage of Subjects Achieving Glycemic Control (HbA1c < 8.0%) at Week 26 [units: percentage of participants]	79.8	83.6

Statistical Analysis 1 for Percentage of Subjects Achieving Glycemic Control (HbA1c < 8.0%) at Week 26

Groups ^[1]	All groups
Method ^[2]	Regression, Logistic
P Value ^[3]	0.2480
Odds Ratio (OR) ^[4]	0.71
95% Confidence Interval	( 0.40 to 1.27 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Fasting Plasma Glucose at Week 26
Measure Description	Fasting plasma glucose lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with fasting plasma glucose at Baseline and Week 26: inhaled human insulin n=202, insulin glargine n=189.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Fasting Plasma Glucose at Week 26 [units: mg/dL] Mean ± Standard Deviation	-52.77 ± 69.48	-53.50 ± 62.92

Statistical Analysis 1 for Change From Baseline in Fasting Plasma Glucose at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.6091
Mean Difference (Final Values) ^[4]	2.57
95% Confidence Interval	( -7.31 to 12.46 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26
Measure Description	Postprandial blood glucose lab value (Time 0 min [fasting], Time 30 min, Time 60 min, Time 90 min, Time 120 min, Time 180 min): Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with fasting and postprandial blood glucose measurements determined by standardized meal tolerance tests at Baseline and Week 26: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26 [units: mg/dL] Mean ± Standard Deviation
Time 0 at Week 26 (n=105, 119)	-35.22 ± 51.12	-61.16 ± 54.68
Time 30 at Week 26 (n=95, 107)	-49.91 ± 62.41	-66.02 ± 62.32
Time 60 at Week 26 (n=110, 118)	-56.83 ± 69.81	-64.77 ± 64.47
Time 90 at Week 26 (n=114, 120)	-64.94 ± 75.67	-58.85 ± 64.58
Time 120 at Week 26 (n=113, 124)	-66.59 ± 74.69	-54.89 ± 63.02
Time 180 at Week 26 (n=114, 122)	-66.99 ± 74.19	-51.27 ± 59.92

Statistical Analysis 1 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	<0.0001
Mean Difference (Final Values) ^[4]	32.96
95% Confidence Interval	( 22.66 to 43.25 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 0 (fasting) at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 2 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	<0.0001
Mean Difference (Final Values) ^[4]	28.57
95% Confidence Interval	( 16.55 to 40.59 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 30 at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 3 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0044
Mean Difference (Final Values) ^[4]	18.96
95% Confidence Interval	( 5.96 to 31.96 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 60 at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 4 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.3168
Mean Difference (Final Values) ^[4]	6.69
95% Confidence Interval	( -6.45 to 19.83 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 90 at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 5 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.9222
Mean Difference (Final Values) ^[4]	0.63
95% Confidence Interval	( -12.12 to 13.39 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 120 at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 6 for Change From Baseline in Fasting and Postprandial Blood Glucose as Determined by Standardized Meal Tolerance Tests at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.6950
Mean Difference (Final Values) ^[4]	-2.41
95% Confidence Interval	( -14.51 to 9.69 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Time 180 at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Postprandial Blood Glucose as Measured by 8-Point Profiles at Week 26
Measure Description	Post-prandial=after a meal. 8-point scale: (1 = before breakfast, 2 = 2 hours post breakfast, 3 = before lunch, 4 = 2 hours post lunch, 5 = before dinner, 6 = 2 hours post dinner, 7 = at bedtime, 8 = overnight [between 2 and 4 am]). Postprandial blood glucose lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS = all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with postprandial blood glucose measurements as measured by 8-point profiles at Baseline and Week 26: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Postprandial Blood Glucose as Measured by 8-Point Profiles at Week 26 [units: mg/dL] Mean ± Standard Deviation
Post-Breakfast, Week 26 (n=185,171)	-86.61 ± 85.98	-75.36 ± 79.04
Post-Lunch, Week 26 (n=184, 172)	-86.66 ± 77.49	-58.51 ± 86.93
Post-Dinner, Week 26 (n=187,166)	-81.86 ± 75.94	-59.20 ± 86.21

Statistical Analysis 1 for Change From Baseline in Postprandial Blood Glucose as Measured by 8-Point Profiles at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.1207
Mean Difference (Final Values) ^[4]	-9.61
95% Confidence Interval	( -21.76 to 2.54 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Post-Breakfast, Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 2 for Change From Baseline in Postprandial Blood Glucose as Measured by 8-Point Profiles at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	<0.0001
Mean Difference (Final Values) ^[4]	-29.11
95% Confidence Interval	( -39.77 to -18.44 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Post-Lunch, Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 3 for Change From Baseline in Postprandial Blood Glucose as Measured by 8-Point Profiles at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	<0.0001
Mean Difference (Final Values) ^[4]	-33.76
95% Confidence Interval	( -44.17 to -23.35 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Post-Dinner, Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Lipids at Week 26
Measure Description	Lipid (total cholesterol, high density lipoprotein cholesterol [HDL-c], low density lipoprotein cholesterol [LDL-c], triglycerides) lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with lipids data at Baseline and Week 26: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Lipids at Week 26 [units: mg/dL] Mean ± Standard Deviation
Total cholesterol, Week 26 (n=197, 186)	-2.80 ± 31.97	-6.19 ± 29.95
HDL-c, Week 26 (n=197, 186)	2.97 ± 6.99	0.69 ± 6.07
LDL-c, Week 26 (n=197, 186)	-2.09 ± 25.74	-3.71 ± 23.63
Triglycerides, Week 26 (n=197, 187)	-42.65 ± 87.70	-46.23 ± 147.8

Statistical Analysis 1 for Change From Baseline in Lipids at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0735
Mean Difference (Final Values) ^[4]	5.21
95% Confidence Interval	( -0.50 to 10.91 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Total Cholesterol at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 2 for Change From Baseline in Lipids at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0017
Mean Difference (Final Values) ^[4]	2.09
95% Confidence Interval	( 0.79 to 3.39 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	HDL-c at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 3 for Change From Baseline in Lipids at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.1369
Mean Difference (Final Values) ^[4]	3.43
95% Confidence Interval	( -1.10 to 7.96 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	LDL-c at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 4 for Change From Baseline in Lipids at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.5684
Mean Difference (Final Values) ^[4]	4.26
95% Confidence Interval	( -10.40 to 18.92 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Triglycerides at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Cardiovascular (CV) Biomarkers High Sensitivity C-reactive Protein (Hs-CRP), Leptin, and Spot Urine Microalbumin at Week 26
Measure Description	CV biomarker (hs-CRP, Leptin, and Spot Urine Microalbumin) lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with CV biomarker data (hs-CRP, leptin, and spot urine microalbumin) at Baseline and Week 26: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Cardiovascular (CV) Biomarkers High Sensitivity C-reactive Protein (Hs-CRP), Leptin, and Spot Urine Microalbumin at Week 26 [units: mg/L] Mean ± Standard Deviation
hs-CRP, Week 26 (n=198, 187)	0.61 ± 4.56	0.32 ± 3.69
Leptin, Week 26 (n=193, 186)	3.68 ± 13.50	4.41 ± 10.50
Spot Urine Microalbumin, Week 26 (n=183, 169)	3.47 ± 96.48	3.70 ± 54.53

Statistical Analysis 1 for Change From Baseline in Cardiovascular (CV) Biomarkers High Sensitivity C-reactive Protein (Hs-CRP), Leptin, and Spot Urine Microalbumin at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.5673
Mean Difference (Final Values) ^[4]	0.24
95% Confidence Interval	( -0.58 to 1.06 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	hs-CRP at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 2 for Change From Baseline in Cardiovascular (CV) Biomarkers High Sensitivity C-reactive Protein (Hs-CRP), Leptin, and Spot Urine Microalbumin at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.6099
Mean Difference (Final Values) ^[4]	-0.61
95% Confidence Interval	( -2.97 to 1.75 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Leptin at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 3 for Change From Baseline in Cardiovascular (CV) Biomarkers High Sensitivity C-reactive Protein (Hs-CRP), Leptin, and Spot Urine Microalbumin at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.7684
Mean Difference (Final Values) ^[4]	2.28
95% Confidence Interval	( -12.94 to 17.51 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Spot urine microalbumin at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in CV Biomarkers Adiponectin and Apolipoprotein B (ApoB) at Week 26
Measure Description	CV biomarker (adiponectin and ApoB) lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with CV biomarker data (adiponectin and ApoB) at Baseline and Week 26: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in CV Biomarkers Adiponectin and Apolipoprotein B (ApoB) at Week 26 [units: mg/mL] Mean ± Standard Deviation
Adiponectin, Week 26 (n=195, 184)	0.23 ± 5.58	-0.32 ± 5.15
ApoB, Week 26 (n=198, 187)	-0.06 ± 0.19	-0.06 ± 0.17

Statistical Analysis 1 for Change From Baseline in CV Biomarkers Adiponectin and Apolipoprotein B (ApoB) at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.5005
Mean Difference (Final Values) ^[4]	0.36
95% Confidence Interval	( -0.69 to 1.42 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	Adiponectin at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Statistical Analysis 2 for Change From Baseline in CV Biomarkers Adiponectin and Apolipoprotein B (ApoB) at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.6850
Mean Difference (Final Values) ^[4]	0.01
95% Confidence Interval	( -0.03 to 0.04 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	ApoB at Week 26.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in 24-Hour Continuous Glucose Monitoring System (CGMS) Glucose Values at Week 26
Measure Description	24-Hour CGMS glucose lab value was obtained using the Medtronic MiniMed CGMS. Not all subjects were offered the opportunity to participate in this assessment. Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS-CGMS=all subjects with at least 1 study drug dose, at least 1 post-baseline measurement, and who participated in a 24-hour CGMS substudy. LOCF imputed missing data with any post-baseline visit. Number of subjects who participated in the substudy with 24-hour CGMS values at Baseline and Week 26: inhaled human insulin n=2, insulin glargine n=8.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in 24-Hour Continuous Glucose Monitoring System (CGMS) Glucose Values at Week 26 [units: mg/dL] Mean ± Standard Deviation	-36.53 ± 33.96	-35.63 ± 58.19

Statistical Analysis 1 for Change From Baseline in 24-Hour Continuous Glucose Monitoring System (CGMS) Glucose Values at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.1856
Mean Difference (Final Values) ^[4]	58.70
95% Confidence Interval	( -43.37 to 160.77 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Mean Standard Deviation (SD) of 24-Hour Glucose Values Measured by CGMS at Week 26
Measure Description	SD of 24-Hour CGMS glucose lab value obtained using the Medtronic MiniMed CGMS. Not all subjects were offered the opportunity to participate in this assessment. Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS-CGMS=all subjects with at least 1 study drug dose, at least 1 post-baseline measurement, and who participated in a 24-hour CGMS substudy. LOCF imputed missing data with any post-baseline visit. Number of subjects who participated in the substudy with 24-hour CGMS values at Baseline and Week 26: inhaled human insulin n=2, insulin glargine n=8.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Mean Standard Deviation (SD) of 24-Hour Glucose Values Measured by CGMS at Week 26 [units: mg/dL] Mean ± Standard Deviation	-12.50 ± 2.55	4.46 ± 11.51

Statistical Analysis 1 for Change From Baseline in Mean Standard Deviation (SD) of 24-Hour Glucose Values Measured by CGMS at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.1862
Mean Difference (Final Values) ^[4]	-18.72
95% Confidence Interval	( -51.34 to 13.89 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Number of Subjects With Hypoglycemic Events
Measure Description	A hypoglycemic event was identified by characteristic symptoms or blood glucose levels. An event was severe if the subject was unable to treat him/herself; had at least 1 neurological symptom; or blood glucose of < = 49 mg/dL. Events not meeting all 3 criteria were considered mild-moderate. Overall=mild, moderate, and severe.
Time Frame	Months 1 to 7
Safety Issue	Yes

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population=all subjects who took at least 1 dose of study drug. Number of subjects with at least 1 hypoglycemic event during the course of the study that were evaluable at the specified month: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	212	190
Number of Subjects With Hypoglycemic Events [units: participants]
Overall at Month 1 (n=147, 83)	70	24
Mild/Moderate at Month 1 (n=147, 83)	69	24
Severe at Month 1 (n=147, 83)	0	0
Overall at Month 2 (n=141, 81)	98	38
Mild/Moderate at Month 2 (n=141, 81)	98	38
Severe at Month 2 (n=141, 81)	1	0
Overall at Month 3 (n=135, 81)	84	35
Mild/Moderate at Month 3 (n=135, 81)	83	35
Severe at Month 3 (n=135, 81)	3	0
Overall at Month 4 (n=120, 80)	61	25
Mild/Moderate at Month 4 (n=120, 80)	61	25
Severe at Month 4 (n=120, 80)	0	0
Overall at Month 5 (n=117, 79)	55	30
Mild/Moderate at Month 5 (n=117, 79)	55	30
Severe at Month 5 (n=117, 79)	0	0
Overall at Month 6 (n=109, 77)	42	31
Mild/Moderate at Month 6 (n=109, 77)	42	31
Severe at Month 6 (n=109, 77)	0	0
Overall at Month 7 (n=77, 60)	8	4
Mild/Moderate at Month 7 (n=77, 60)	8	4
Severe at Month 7 (n=77, 60)	0	0

No statistical analysis provided for Number of Subjects With Hypoglycemic Events

Measure Type	Secondary
Measure Title	Number of Total Hypoglycemic Events
Measure Description	A hypoglycemic event was identified by characteristic symptoms or blood glucose levels. Severe=the subject was unable to treat him/herself; had at least 1 neurological symptom; or blood glucose of < = 49 mg/dL. Events not meeting all 3 criteria were considered mild-moderate. Overall=mild, moderate, and severe. Total=events during the study.
Time Frame	Months 1 to 7
Safety Issue	Yes

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population=all subjects who took at least 1 dose of study drug. Number of subjects with at least 1 hypoglycemic event during the course of the study that were evaluable at the specified month: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	212	190
Number of Total Hypoglycemic Events [units: events]
Overall at Month 1 (n=147, 83)	195	41
Mild/Moderate at Month 1 (n=147, 83)	194	41
Severe at Month 1 (n=147, 83)	0	0
Overall at Month 2 (n=141, 81)	365	66
Mild/Moderate at Month 2 (n=141, 81)	364	66
Severe at Month 2 (n=141, 81)	1	0
Overall at Month 3 (n=135, 81)	292	84
Mild/Moderate at Month 3 (n=135, 81)	289	84
Severe at Month 3 (n=135, 81)	3	0
Overall at Month 4 (n=120, 80)	166	51
Mild/Moderate at Month 4 (n=120, 80)	166	51
Severe at Month 4 (n=120, 80)	0	0
Overall at Month 5 (n=117, 79)	157	64
Mild/Moderate at Month 5 (n=117, 79)	157	64
Severe at Month 5 (n=117, 79)	0	0
Overall at Month 6 (n=109, 77)	119	63
Mild/Moderate at Month 6 (n=109, 77)	119	63
Severe at Month 6 (n=109, 77)	0	0
Overall at Month 7 (n=77, 60)	12	5
Mild/Moderate at Month 7 (n=77, 60)	12	5
Severe at Month 7 (n=77, 60)	0	0

No statistical analysis provided for Number of Total Hypoglycemic Events

Measure Type	Secondary
Measure Title	Number of Total Subject Months of Treatment
Measure Description	Number of total subject months of treatment. Subject months = number of days from start of treatment to the last day of active treatment + 1 day lag, including off-drug time)/30.44. Severe=the subject was unable to treat him/herself; had at least 1 neurological symptom; or blood glucose of < = 49 mg/dL. Events not meeting all 3 criteria were considered mild-moderate. Overall=mild, moderate, and severe.
Time Frame	Months 1 to 7
Safety Issue	Yes

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population=all subjects who took at least 1 dose of study drug. Number of subjects with at least 1 hypoglycemic event during the course of the study that were evaluable at the specified month: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	212	190
Number of Total Subject Months of Treatment [units: subject months]
Overall at Month 1 (n=147, 83)	145.8	82.4
Mild/Moderate at Month 1 (n=147, 83)	145.8	82.4
Severe at Month 1 (n=147, 83)	145.8	82.4
Overall at Month 2 (n=141, 81)	138.3	81.0
Mild/Moderate at Month 2 (n=141, 81)	138.3	81.0
Severe at Month 2 (n=141, 81)	138.3	81.0
Overall at Month 3 (n=135, 81)	128.8	80.4
Mild/Moderate at Month 3 (n=135, 81)	128.8	80.4
Severe at Month 3 (n=135, 81)	128.8	80.4
Overall at Month 4 (n=120, 80)	117.7	79.7
Mild/Moderate at Month 4 (n=120, 80)	117.7	79.7
Severe at Month 4 (n=120, 80)	117.7	79.7
Overall at Month 5 (n=117, 79)	112.2	77.9
Mild/Moderate at Month 5 (n=117, 79)	112.2	77.9
Severe at Month 5 (n=117, 79)	112.2	77.9
Overall at Month 6 (n=109, 77)	103.5	75.7
Mild/Moderate at Month 6 (n=109, 77)	103.5	75.7
Severe at Month 6 (n=109, 77)	103.5	75.7
Overall at Month 7 (n=77, 60)	12.8	6.6
Mild/Moderate at Month 7 (n=77, 60)	12.8	6.6
Severe at Month 7 (n=77, 60)	12.8	6.6

No statistical analysis provided for Number of Total Subject Months of Treatment

Measure Type	Secondary
Measure Title	Crude Hypoglycemic Event Rate
Measure Description	crude event rate=(events)/(subject-months). Severe=the subject was unable to treat him/herself; had at least 1 neurological symptom; or blood glucose of < = 49 mg/dL. Events not meeting all 3 criteria were considered mild-moderate. Overall=mild, moderate, and severe.
Time Frame	Months 1 to 7
Safety Issue	Yes

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population=all subjects who took at least 1 dose of study drug. Number of subjects with at least 1 hypoglycemic event during the course of the study that were evaluable at the specified month: n=inhaled human insulin, insulin glargine.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	212	190
Crude Hypoglycemic Event Rate [units: events / subject-months]
Overall at Month 1 (n=147, 83)	1.3	0.5
Mild/Moderate at Month 1 (n=147, 83)	1.3	0.5
Severe at Month 1 (n=147, 83)	0	0
Overall at Month 2 (n=141, 81)	2.6	0.8
Mild/Moderate at Month 2 (n=141, 81)	2.6	0.8
Severe at Month 2 (n=141, 81)	0.7	0
Overall at Month 3 (n=135, 81)	2.3	1.0
Mild/Moderate at Month 3 (n=135, 81)	2.2	1.0
Severe at Month 3 (n=135, 81)	2.3	0
Overall at Month 4 (n=120, 80)	1.4	0.6
Mild/Moderate at Month 4 (n=120, 80)	1.4	0.6
Severe at Month 4 (n=120, 80)	0	0
Overall at Month 5 (n=117, 79)	1.4	0.8
Mild/Moderate at Month 5 (n=117, 79)	1.4	0.8
Severe at Month 5 (n=117, 79)	0	0
Overall at Month 6 (n=109, 77)	1.2	0.8
Mild/Moderate at Month 6 (n=109, 77)	1.2	0.8
Severe at Month 6 (n=109, 77)	0	0
Overall at Month 7 (n=77, 60)	0.9	0.8
Mild/Moderate at Month 7 (n=77, 60)	0.9	0.8
Severe at Month 7 (n=77, 60)	0	0

No statistical analysis provided for Crude Hypoglycemic Event Rate

Measure Type	Secondary
Measure Title	Number of Nocturnal Hypoglycemic Events
Measure Description	A hypoglycemic event was identified by characteristic symptoms or blood glucose levels. Severe=the subject was unable to treat him/herself; had at least 1 neurological symptom; or blood glucose of < = 49 mg/dL. Events not meeting all 3 criteria were considered mild-moderate. Nocturnal hypoglycemia=event occuring from midnight to 5:59 am.
Time Frame	Months 1 to 7
Safety Issue	Yes

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
Safety population=all subjects who took at least 1 dose of study drug.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	212	190
Number of Nocturnal Hypoglycemic Events [units: events]	187	114

No statistical analysis provided for Number of Nocturnal Hypoglycemic Events

Measure Type	Secondary
Measure Title	Change From Baseline in Body Weight at Week 26
Measure Description	Body weight value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with body weight measurements at Baseline and Week 26: inhaled human insulin n=192, insulin glargine n=183.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Body Weight at Week 26 [units: kg] Mean ± Standard Deviation	3.55 ± 4.74	2.33 ± 3.80

Statistical Analysis 1 for Change From Baseline in Body Weight at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0055
Mean Difference (Final Values) ^[4]	1.20
95% Confidence Interval	( 0.35 to 2.04 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Secondary
Measure Title	Change From Baseline in Body Mass Index (BMI) at Week 26
Measure Description	BMI value (kg/m2): Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Number of subjects with BMI measurements at Baseline and Week 26: inhaled human insulin n=192, insulin glargine n=183.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Body Mass Index (BMI) at Week 26 [units: kg/m2] Mean ± Standard Deviation	1.22 ± 1.66	0.80 ± 1.30

Statistical Analysis 1 for Change From Baseline in Body Mass Index (BMI) at Week 26

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0053
Mean Difference (Final Values) ^[4]	0.42
95% Confidence Interval	( 0.12 to 0.71 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	No text entered.
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	No text entered.

Measure Type	Other Pre-specified
Measure Title	Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the FAS
Measure Description	HbA1c lab value: Change = value at Week 26 minus value at Baseline.
Time Frame	Baseline, Week 26
Safety Issue	No

Population Description

Explanation of how the number of participants for analysis was determined. Includes whether analysis was per protocol, intention to treat, or another method. Also provides relevant details such as imputation technique, as appropriate.
FAS=all subjects who took at least 1 dose of study drug and had at least 1 post-baseline measurement. LOCF imputed missing data with any post-baseline visit. Using the FAS yielded supplemental analyses for the primary efficacy endpoint.

Reporting Groups

	Description
Inhaled Human Insulin	Initiation dose of 1 milligram per meal, and individually adjusted doses, per subject's blood glucose, over the 6 month study, in addition to oral agents.
Insulin Glargine	Insulin glargine, label instruction initiation dose (10 units), and individually adjusted doses, per subject's blood glucose, over the six months study, in addition to oral agents.

Measured Values

	Inhaled Human Insulin	Insulin Glargine
Number of Participants Analyzed [units: participants]	203	189
Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the FAS [units: percent] Mean ± Standard Deviation	-1.91 ± 1.41	-1.67 ± 1.38

Statistical Analysis 1 for Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the FAS

Groups ^[1]	All groups
Non-Inferiority/Equivalence Test ^[2]	Yes
Method ^[3]	ANCOVA
P Value ^[4]	<0.0001
Mean Difference (Final Values) ^[5]	-0.27
95% Confidence Interval	( -0.47 to -0.08 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	The null hypothesis was that the inhaled human insulin group was inferior to the insulin glargine group with respect to 26-week change from baseline in HbA1c, while the alternate hypothesis is that the inhaled human insulin group was not inferior to the insulin glargine group with respect to 26-week change from baseline in HbA1c, given the predetermined non-inferiority margin of 0.4%.
[2]	Details of power calculation, definition of non-inferiority margin, and other key parameters:
	The non-inferiority test utilized a one-sided 2.5% level of significance.
[3]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[4]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[5]	Other relevant estimation information:
	A CI approach was presented with a two-sided 95% CI of the difference between treatment and control.

Statistical Analysis 2 for Change From Baseline in Glycosylated Hemoglobin A1c (HbA1c) at Week 26 for the FAS

Groups ^[1]	All groups
Method ^[2]	ANCOVA
P Value ^[3]	0.0062
Mean Difference (Final Values) ^[4]	-0.27
95% Confidence Interval	( -0.47 to -0.08 )

[1]	Additional details about the analysis, such as null hypothesis and power calculation:
	With an established non-inferiority claim, an additional test for superiority was conducted (ie, the margin was 0.0%).
[2]	Other relevant information, such as adjustments or degrees of freedom:
	No text entered.
[3]	Additional information, such as whether or not the p-value is adjusted for multiple comparisons and the a priori threshold for statistical significance:
	No text entered.
[4]	Other relevant estimation information:
	The superiority test used a one-sided 2.5% level of significance.

Responsible Party:	Pfizer, Inc. ( Director, Clinical Trial Disclosure Group )
Study ID Numbers:	A2171095
Study First Received:	January 3, 2007
Results First Received:	July 30, 2009
Last Updated:	September 8, 2009
ClinicalTrials.gov Identifier:	NCT00418522 History of Changes
Health Authority:	United States: Food and Drug Administration