MiniMed 670G System

P160017 · Medtronic Minimed, Inc. · OZP · Sep 28, 2016 · Clinical Chemistry

Device Facts

• Record ID: P160017
• Device Name: MiniMed 670G System
• Applicant: Medtronic Minimed, Inc.
• Product Code: OZP · Clinical Chemistry
• Decision Date: Sep 28, 2016
• Decision: APPR
• Device Class: Class 3
• Attributes: Therapeutic, Expedited Review

Intended Use

The Medtronic MiniMed 670G system is intended for continuous delivery of basal insulin (at user selectable rates) and administration of insulin boluses (in user selectable amounts) for the management of Type 1 diabetes mellitus in persons, fourteen years of age and older, requiring insulin as well as for the continuous monitoring and trending of glucose levels in the fluid under the skin. The MiniMed 670G System includes SmartGuard technology, which can be programmed to automatically adjust delivery of basal insulin based on Continuous Glucose Monitor sensor glucose values, and can suspend delivery of insulin when the sensor glucose value falls below or is predicted to fall below predefined threshold values. The Medtronic MiniMed 670G System consists of the following devices: MiniMed 670G insulin pump, the Guardian Link (3) Transmitter, the Guardian Sensor (3), One-Press Serter, and the Contour NEXT Link 2.4 Glucose Meter. The system requires a prescription. The Guardian Sensor (3) glucose values are not intended to be used directly for making therapy adjustments, but rather to provide an indication of when a finger stick may be required. All therapy adjustments should be based on measurements obtained using a home glucose monitor and not on values provided by the Guardian Sensor (3).

Device Story

MiniMed 670G System is an automated insulin delivery system for Type 1 diabetes management. It integrates an insulin pump, continuous glucose monitor (CGM), and blood glucose meter. The system receives interstitial glucose data from the Guardian Sensor (3) via the Guardian Link (3) Transmitter. The pump uses a proprietary 'SmartGuard' algorithm to automatically adjust basal insulin delivery (Auto Mode) or suspend insulin delivery based on low glucose thresholds (Manual Mode). The user manually calculates and delivers meal boluses. The system is used in home settings by patients under physician guidance. Output is displayed on the pump screen, providing real-time glucose trends and alerts. This automation aims to improve glycemic control and reduce hypoglycemia risk. Data can be downloaded to CareLink software for clinical review.

Clinical Evidence

Pivotal study (G140167) evaluated 123 subjects (14-75 years) over 3.5 months. Primary endpoints were descriptive safety and effectiveness. Results showed mean HbA1c reduction of 0.5% (7.4% to 6.9%). Time in range (70-180 mg/dL) increased from 66.7% to 72.2%. No severe hypoglycemia or DKA reported. Guardian Sensor performance study (G140053) evaluated 89 subjects; mean absolute percent difference (MARD) was 9.64% with 3-4 calibrations/day.

Technological Characteristics

System includes MMT-1780 pump, Guardian Link (3) Transmitter, Guardian Sensor (3), and Contour NEXT Link 2.4 meter. Sensor uses electrochemical sensing principle. Connectivity via RF telemetry. Software includes rule-based and predictive algorithms for insulin suspension and automated basal adjustment. Sterilization of sensor via Electron Beam. Pump is battery-operated micro-infusion pump.

Indications for Use

Indicated for management of Type 1 diabetes mellitus in persons 14 years and older requiring insulin. Contraindicated in children under 7 years old and patients requiring less than 8 units of insulin per day. Adjunctive use only; requires blood glucose meter for therapy adjustments.

Regulatory Classification

Identification

An insulin-only automated insulin dosing system that achieves a target glucose level by automatically increasing or reducing insulin infusion from an insulin pump. Insulin adjustments are based upon specified thresholds of measured glucose levels.

Reference Devices

• Enlite Sensor (P120010)
• MiniMed 530G System (P120010)
• Contour Next Link 2.4 Wireless Blood Glucose Monitoring System (k110894)
• MiniMed 630G pump (P150001)

Submission Summary (Full Text)

{0}

SUMMARY OF SAFETY AND EFFECTIVENESS DATA (SSED)

I. GENERAL INFORMATION

Device Generic Name: Automated Insulin Delivery System

Device Trade Name: MiniMed 670G System

Device Procode: OZP, Artificial pancreas device system, single hormonal control

Applicant’s Name and Address: Medtronic MiniMed, Inc.
18000 Devonshire Street
Northridge, CA 91325

Date(s) of Panel Recommendation: None

Premarket Approval Application (PMA) Number: P160017

Date of FDA Notice of Approval: September 28, 2016

Priority Review: Granted priority review status on July 13, 2016 because the device is a novel technology and availability is in patients’ best interest.

II. INDICATIONS FOR USE

MiniMed 670G System

The Medtronic MiniMed 670G system is intended for continuous delivery of basal insulin (at user selectable rates) and administration of insulin boluses (in user selectable amounts) for the management of Type 1 diabetes mellitus in persons, fourteen years of age and older, requiring insulin as well as for the continuous monitoring and trending of glucose levels in the fluid under the skin. The MiniMed 670G System includes SmartGuard technology, which can be programmed to automatically adjust delivery of basal insulin based on Continuous Glucose Monitor sensor glucose values, and can suspend delivery of insulin when the sensor glucose value falls below or is predicted to fall below predefined threshold values.

The Medtronic MiniMed 670G System consists of the following devices: MiniMed 670G insulin pump, the Guardian Link (3) Transmitter, the Guardian Sensor (3), One-Press Serter, and the Contour NEXT Link 2.4 Glucose Meter. The system requires a prescription.

The Guardian Sensor (3) glucose values are not intended to be used directly for making therapy adjustments, but rather to provide an indication of when a finger stick may be required. All therapy adjustments should be based on measurements obtained using a home glucose monitor and not on values provided by the Guardian Sensor (3).

Guardian Sensor (3)

PMA P160017: FDA Summary of Safety and Effectiveness Data

{1}

The Guardian Sensor (3) is intended for use with the MiniMed 670G system to continuously monitor glucose levels in persons with diabetes. It is intended to be used for detecting trends and tracking patterns in persons aged fourteen years and older, and to be used by the MiniMed 670G system to automatically adjust basal insulin levels. It is indicated for use as an adjunctive device to complement, not replace, information obtained from standard blood glucose monitoring devices. The sensor is intended for single use and requires a prescription. The Guardian Sensor (3) is indicated for 7 days of continuous use.

## One-press Serter

The One-press Serter is used as an aid for inserting the sensor. It is indicated for single-patient use and it is not intended for multiple-patient use.

## Guardian Link (3) Transmitter

The Guardian Link (3) Transmitter is intended for use with the MiniMed 670G System. The Guardian Link (3) Transmitter powers the glucose sensor, collects and calculates sensor data, and wirelessly sends the data to the MiniMed 670G insulin pump. The Transmitter is intended for single-patient multi-use.

## Contour NEXT Link 2.4 Glucose Meter

The Contour Next Link 2.4 Wireless Blood Glucose Monitoring System is an over the counter (OTC) device utilized by persons with diabetes in home settings for the measurement of glucose in whole blood, and is for single patient use only and should not be shared. The Contour Next Link 2.4 wireless blood glucose monitoring system is indicated for use with fresh capillary whole blood samples drawn from the fingertip and palm only. The Contour NEXT Test Strips are intended for self-testing by persons with diabetes for the quantitative measurement of glucose in whole blood samples from 20 to 600 mg/dL. The Contour Next Link 2.4 wireless blood glucose monitoring system is intended to be used to transmit glucose values to the MiniMed 670G pump and facilitate transfer of information to Medtronic CareLink Software through the use of radio frequency communication. The Contour Next Link 2.4 Wireless Blood Glucose Monitoring System is not intended for the diagnosis of, or screening for, diabetes mellitus. It is not intended for use on neonates.

## CONTRAINDICATIONS

A prominent boxed warning is included in the labeling regarding use of the device in subjects under the age of 7 years as follows:

"Medtronic performed an evaluation of the 670G closed loop system and determined that it may not be safe for use in children under the age of 7 because of the way that the system is designed and the daily insulin requirements. Therefore this device should not be used in anyone under the age of 7 years old. This device should also not be used in patients who require less than a total daily insulin dose of 8 units per day because the device requires a minimum of 8 units per day to operate safely."

The following contraindications for this device are also described in the labeling:

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 2

{2}

- Insulin pump therapy is not recommended for people who are unwilling or unable to perform a minimum of four blood glucose tests per day. As insulin pumps use rapid acting insulin only, blood glucose testing is required to help identify rapid glycemic deterioration due to insulin infusion occlusion, infusion site problems, insulin stability issues, user error, or a combination of these.
- Pump therapy is not recommended for people who are unwilling or unable to maintain contact with their healthcare professional.
- Pump therapy is not recommended for people whose vision or hearing does not allow recognition of pump signals and alarms.
- Do not use serter on products other than the Enlite Sensor (P120010) or Guardian Sensor (3). Medtronic cannot guarantee the safety or efficacy of this product if used with other products.
- The reservoir is contraindicated for the infusion of blood or blood products.
- Infusion sets are indicated for subcutaneous use only and not for intravenous (IV) infusion or the infusion of blood or blood products.

## III. WARNINGS AND PRECAUTIONS

The warnings and precautions can be found in the 670G System User Guide labeling.

## IV. DEVICE DESCRIPTION

The MiniMed 670G system is comprised of the following devices:

### MiniMed 670G Pump (MMT-1780)

The MiniMed 670G pump (model MMT-1780) is an ambulatory, battery operated, rate programmable micro infusion pump designed to deliver insulin from a reservoir. The reservoir is driven by a motor to deliver determined basal rate profiles and user selected bolus amounts of insulin into the subcutaneous tissue through an infusion set.

The MiniMed 670G pump is offered in one model (MMT-1780). The pump houses electronics, a pumping mechanism, a user interface, and a medication reservoir within the same physical device. The reservoir is attached to a tube that connects to the user's infusion site on their body. It is intended to deliver insulin through a diffusion mechanism. Model MMT-1780 is compatible with a 3.0 mL reservoir. The pump only displays blood glucose level units in mg/dL and cannot be reconfigured by the user. In addition to its delivery of insulin, the MiniMed 670G pump is designed to receive and display real-time interstitial fluid glucose values via the Guardian Link (3) Transmitter. When used in combination with Guardian Sensor (3), the transmitter sends sensor signals to the MiniMed 670G pump via radiofrequency (RF) telemetry. The 670G Pump has the following features and capabilities:

- Provides predictive sensor glucose alerts when sensor glucose values are high or low (please see 'Manual Mode' section below for details).
- Can receive blood glucose values from the Contour NEXT Link 2.4 Meter to use for sensor calibration.

PMA P160017: FDA Summary of Safety and Effectiveness Data

{3}

- The pump can display Guardian sensor glucose values in real-time, and also store those values (and blood-glucose values from the meter) into its pump memory.
- “SmartGuard” Technology: There are two levels of this technology; the first is available in Manual Mode and the second in Auto Mode.

- The first level of SmartGuard technology is available in Manual Mode:
- This technology automatically suspends insulin when the sensor reaches a preset low limit (referred to as ‘Suspend on Low’)
- This technology automatically suspends insulin when the sensor glucose value is predicted, using a proprietary predictive suspend algorithm, to reach a pre-set low limit, i.e., suspends before the low limit is reached (referred to as ‘Suspend before Low’).
- When a Suspend event occurs, the user can choose to continue to keep insulin suspended, or the user can choose to resume insulin delivery.
- Following a Suspend event, insulin delivery will automatically resume when the sensor glucose levels rise above the pre-set suspend threshold.
- The ‘Suspend on low’ and ‘Suspend before low’ features are optional features available when the system is in Manual Mode.
- This technology (in Manual Mode) provides a bolus calculator called the ‘Bolus Wizard’ that uses your settings to calculate an estimated bolus amount based on the meter blood glucose values and carbohydrates that the user enters. These settings should be set up with the help of a Health Care Practitioner before using the Bolus Wizard. Those settings include:
- Carbohydrate Ratio
- Insulin Sensitivity Factor
- Blood Glucose Target
- Active Insulin Time

- The second level of SmartGuard technology is available in Auto Mode:
- This technology automatically adjusts basal insulin delivery using continuous glucose monitor data, referred to as Auto Mode.
- The Auto Mode feature can automatically increase or decrease the amount of insulin delivered based on sensor values.

During Auto mode operation, the user must manually deliver meal boluses that they calculate using the estimated amount of carbohydrates for meals at the time they are eaten. The user must also use the Auto Mode bolus feature to calculate boluses while in Auto Mode.

The MiniMed 670G Insulin Pump is designed to receive and display real-time glucose values received from the Guardian Link transmitter. Guardian sensor signals are transmitted from the transmitter to the MiniMed 670G Insulin Pump via RF telemetry and converted into glucose concentrations based on calibration values from the Contour Next Link 2.4 meter. Signals are updated and transmitted to the pump every five minutes.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 4

{4}

The real time sensor glucose values, displayed by the MiniMed 670G Insulin Pump, are not intended to be used directly for making therapy adjustments. The user can use the tracking and trending of sensor glucose values to help determine if an unplanned finger stick measurement may be needed. In addition, sensor glucose values should not be used to modify insulin therapy. All manual insulin therapy adjustments should be based on measurements obtained using a blood glucose meter and not based on the sensor glucose value displayed by the MiniMed 670G Insulin Pump.

## Manual Mode

The user can set the pump to temporarily suspend insulin delivery automatically when the sensor glucose level is equal to or less than a selected threshold. The user has the capability to select a ‘Suspend on Low’ threshold within the 50 mg/dL to 90 mg/dL range. When the ‘Suspend on Low’ tool is set to ‘ON’, the system compares the sensor glucose value and the programmed Suspend threshold whenever the sensor glucose value is updated (every five minutes). If the sensor glucose value falls below the set threshold, insulin delivery will be suspended. Once the sensor glucose rises above that set threshold, insulin delivery will resume. The use of the ‘Suspend on Low’ tool is optional and the user can turn the tool ‘ON’ and ‘OFF’.

The pump also includes the new Predictive Low Management tool that allows the user to set a glucose value threshold where the pump will suspend the insulin delivery if the blood glucose value is predicted to reach the selected threshold in the near future (‘Suspend before Low’). The user can set their low sensor glucose threshold value from 50 mg/dL to 90 mg/dL; insulin delivery will suspend when the sensor glucose value is predicted to reach or fall below a level that is 20 mg/dL above the set low limit threshold value within approximately 30 minutes. The user has the ability to program the length of time in which suspension will occur pending the blood glucose threshold 30 minutes before the threshold is reached. Once the sensor glucose rises above that set threshold, insulin delivery will resume. The use of the ‘Suspend on Low’ tool is optional and the user can turn the tool ‘ON’ and ‘OFF’.

When the sensor glucose value is below or predicted to be below the set threshold, an alarm and siren occurs and the pump suspends, and the user may elect to continue the suspend, or cancel the temporary pump suspension of insulin delivery at any time.

If the user does not respond to the ‘suspend on low’ or ‘suspend before low’ alarm or siren, the pump will automatically suspend for up to two hours.

If the user cancels the suspension of insulin delivery, the system will continue to deliver insulin at the programmed basal rate until the next time the sensor glucose value is below the set threshold value. The alarm and siren will then re-sound, and the pump will suspend (unless canceled by the user).

If the user responds to the alarm or siren by electing to accept the insulin suspension, the pump will suspend for at least 30 minutes and up to a maximum of 2 hours; the pump

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 5

{5}

may resume insulin delivery if the sensor glucose value rises above the set low threshold after 30 minutes have passed and before or up to the 2 hours maximum suspension time. The user can manually resume insulin at any time. At the end of the two hour maximum, the pump will resume insulin delivery until the next sensor glucose value is below the set threshold suspend value.

The user can cancel the temporary pump suspension at any time during the two-hour period regardless if the suspension occurred because the user was not able to respond to the initial alarm or he/she accepted the suspension. The pump may also resume insulin delivery on its own based on sensor glucose values or predicted sensor glucose values (if the sensor glucose value indicates that the user has recovered from the set threshold low glucose value). The table below summarizes the suspend features.

Table 1: Suspend on Low and Suspend before Low Feature and Related Resumption Options

| | Suspend Features | |
| --- | --- | --- |
| What Happens | Suspend on Low | Suspend before Low |
| The pump suspends insulin delivery | The predetermined sensor blood glucose threshold is reached. | The predetermined sensor blood glucose threshold is predicted to be reached within 5 to 30 minutes (time period predetermined by user). |
| The user accepts the pump suspension | The pump will suspend for at least 30 minutes and up to a maximum of 2 hours. | The pump will suspend for at least 30 minutes and up to a maximum of 2 hours. |
| The user responds to the insulin suspend and the system suspends insulin for two hours (maximum suspension time) | If insulin was suspended for two hours, then at the end of the two hour maximum, the pump will resume insulin delivery. The pump will not suspend insulin again until after the refractory period* and the next sensor glucose value is below the set low glucose suspend value. | If insulin was suspended for two hours, then at the end of the two hour maximum, the pump will resume insulin delivery. The pump will not suspend insulin again until after the refractory period* and the next sensor glucose value is predicted to go below the set low glucose suspend value. |
| The user cancels suspension | The system will resume insulin delivery at the programmed basal rate until the next time the sensor glucose value is below the set threshold value. | The system will resume insulin delivery at the programmed basal rate until the next time the sensor glucose value is below the set threshold value. |
| The user does not respond to the suspend alert | The pump will automatically suspend for | The pump will automatically suspend for |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{6}

| | up to two hours if the sensor glucose does not detect that the users glucose values are or are predicted to go above the low glucose level | up to two hours if the sensor glucose does not detect that the users glucose values are or are predicted to go above the low glucose level |
| --- | --- | --- |
| The sensor glucose detects a glucose value above the low glucose pre-set level | Pump may resume insulin delivery on its own after 30 minutes based on sensor glucose values glucose values. | Pump may resume insulin delivery on its own after 30 minutes based on sensor glucose values or predicted sensor glucose values. |

* After a Suspend event occurs, there is a period of time when the suspend functionality is unavailable (refractory period). This time will vary depending on whether or not the user responds to the Suspend event. Please see the labeling for the MiniMed 670G System regarding how the suspend functions work. The user can manually suspend insulin delivery at any time.

The MiniMed 670G Insulin Pump is capable of storing 90 days of pump history and glucose sensor data. The pump has a graphical display that the user can use to view the glucose history for the past 3, 6, 12 and 24 hours, high/low glucose alarms and display of retrospective glucose trend information.

Stored pump history and glucose data can be downloaded to a personal computer for review and analysis, to track patterns and improve diabetes management. Data is downloaded from the pump to CareLink therapy management software.

The MiniMed 670G pump is compatible with commercially available Medtronic Paradigm infusion sets and their 3 mL volume reservoirs. It was not necessary to develop new infusion sets or reservoirs for use with this pump.

## Auto Mode

The MiniMed 670G Insulin Pump contains an auto mode feature; this new tool uses an algorithm to automatically adjust basal insulin delivery using continuous glucose monitor data. When in Auto Mode, the pump responds to fluctuations in interstitial fluid glucose levels measured by the continuous glucose monitor; the Auto Mode feature can automatically increase or decrease the amount of basal insulin delivered based on sensor glucose values.

Auto Mode does not administer meal boluses. During Auto mode operation, users must deliver meal boluses by entering the amount of insulin they want to deliver based on the estimated amount of carbohydrates they are eating. Failure to deliver meal boluses in association with meals during Auto mode operation can result in significant post meal hyperglycemia.

The Auto Mode algorithm is designed to adjust the user's basal insulin rates to try to keep them at a target blood glucose level. The standard target glucose setting in Auto Mode is

PMA P160017: FDA Summary of Safety and Effectiveness Data

{7}

120 mg/dL, and the target can also be set temporarily to 150 mg/dL for exercise and other events. In addition, blood glucose readings above 150 mg/dL will prompt the Auto Mode feature to calculate if a correction bolus is needed; if needed, a correction bolus will be recommended to the users, who can choose whether they want to deliver that correction bolus. Users should check their blood glucose levels using a blood glucose meter before administering a correction bolus.

When first using the device, Auto Mode cannot be activated until the system completes a 48 hour warm-up period while the user uses the pump in manual mode to deliver insulin. In addition, before activation, the user must cancel any temporary basal rates, ensure delivery is not suspended, set a carbohydrate ratio, set high and low glucose settings, and enter a blood glucose value obtained by using a blood glucose meter if one has not been entered in the last 12 minutes. Carbohydrate ratios and high and low glucose settings should be discussed and established with the user's healthcare practitioner before use of Auto Mode begins.

There is an additional feature, called the safe basal feature which will be activated in Auto Mode when the system encounters specific issues that the user should address; the user cannot manually enable Safe Basal as it is a mandatory automatic feature that activates when the pump encounters these types of issues. This feature is designed to be a safety net when in Auto Mode. The pump will transition into Safe Basal if the system detects the user is getting too little or too much basal insulin, if the pump detects an issue with the sensor or a discrepancy between blood glucose meter and sensor glucose readings, or if the pump has not received sensor glucose values in over 5 minutes. When the pump encounters an issue, it transitions to safe basal; in safe basal, the pump supplements the user's basal insulin needs by delivering a steady state basal rate, which allows the user time to perform the additional actions prompted by the pump that are required to ensure Auto Mode stays active. Safe Basal does not automatically adjust basal insulin and instead delivers a steady basal rate until the user has addressed the issue that triggered the pump to activate Safe Basal. After 90 minutes in Safe Basal, if the condition(s) that caused the pump to transition into Safe Basal have not been resolved, the pump will automatically exit Auto Mode and enter Manual Mode. If Auto Mode is active, then the 'Suspend on Low' and 'Suspend before Low' features from the 'Manual Mode' are unavailable and inactive. However, while in Auto Mode, basal insulin is adjusted by decreasing or suspending automatically based on continuous glucose monitor sensor values if the user is experiencing low glucose. The table below describes the available features and accessibility within each mode (Auto Mode and Manual Mode). Note that, although Safe Basal is not mentioned in the table below, it is only available in Auto Mode and cannot be activated by the user, as it is a mandatory automatic feature that activates when the pump encounters an issue that the user should address (see description above).

Table 2: Modes and Related Accessibility

| Mode | Description | When is it Active? | Will I receive Alerts? |
| --- | --- | --- | --- |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{8}

| Mode | Description | When is it Active? | Will I receive Alerts? |
| --- | --- | --- | --- |
| Manual Mode: Sensor Augmented Pump | This mode is when the device is functioning as a sensor and pump, but the device is not in Auto Mode and the insulin suspend features are not turned on. | This is the default mode and the user does not have to specifically turn this mode on. | There is a mandatory severe low alarm at 50 mg/dL; The user can also set optional high and low alerts to sound on or before set sensor glucose levels. |
| Manual Mode: Suspend On Low | When this feature is active the device detects that your sensor glucose level has reached a pre-set sensor glucose value and it automatically suspends basal insulin delivery when that value is reached. | The user has to turn this feature on. It is not available when Auto Mode is turned on, and it cannot be turned on if suspend before low is turned on. | There is a mandatory severe low alarm at 50 mg/dL and at the pre-set low level. The user can also set optional high alerts to sound on or before set sensor glucose levels, and an optional alert before low alert. |
| Manual Mode: Suspend Before Low | When this feature is active the device detects when your sensor glucose is predicted to reach a pre-set value and it automatically suspends basal insulin delivery before that value is reached. | The user has to turn this feature on. It is not available when Auto Mode is turned on, and it cannot be turned on if suspend before low is turned on. | There is a mandatory severe low alarm at 50 mg/dL and at the pre-set low level. The user can also set optional high alerts to sound on or before set sensor glucose levels, and an optional alarm before low alert. |
| Auto Mode | When this mode is active, the device can automatically adjust basal insulin by increasing, decreasing, or turning off basal insulin delivery based on sensor glucose levels. | The user has to turn this mode on and certain pre-defined conditions have to be met. | There is a mandatory severe low alarm at 50 mg/dL; The user can also set optional high and low alerts to sound on or before set sensor glucose levels. |

## Guardian Link Transmitter System (MMT-7811)

The Guardian Link Transmitter System consists of the Guardian Link Transmitter (MMT-7811), the Charger (model MMT-7715), and the Tester (model MMT-7736).

The Guardian Link Transmitter interfaces directly with the glucose-sensor assembly. The Guardian Link Transmitter provides power to the glucose sensor, and measures the sensor signal current from the glucose sensor.

The sensor signal current is an electrical current level that is proportional to the glucose level in the user's subcutaneous interstitial fluid. The sensor signal current is converted to a digital signal, which is filtered to reduce noise artifacts. This digital signal is sent to the MiniMed 670G pump every 5 minutes, using radio frequency (RF).

## Guardian Sensor (MMT-7020)

PMA P160017: FDA Summary of Safety and Effectiveness Data

{9}

The Guardian Sensor is a sterile, single-use, single patient glucose sensing component for continuous monitoring of glucose levels in the user's interstitial fluid, when inserted in the user's abdomen for up to seven days. The Sensor is inserted into the subcutaneous tissue using the One-Press Serter and is taped to the user's skin. It connects to the Guardian Link Transmitter, which in turn communicates with the MiniMed 670G Pump.

When making treatment decisions, such as determining insulin dose for meals, the 670G continuous glucose monitor (CGM) values should not be used, as they are not intended to be used to make such treatment decisions. The 670G continuous glucose monitor does not replace a blood glucose meter. Users should always use the values from a blood glucose meter for treatment decisions. Blood glucose values may differ from sensor glucose values. Using the sensor glucose readings for treatment decisions could lead to unwanted high or low blood glucose.

Users should calibrate the Guardian Sensor at least every 12 hours using meter blood glucose values. Calibration is necessary for sensor function, and more frequent calibration can help to increase the accuracy of the sensor. The system requires a minimum of two calibrations per day, and four calibrations per day are recommended. The system is contraindicated for patients unwilling or unable to do frequent blood glucose meter measurements.

If the user obtains blood glucose values using the Contour Next Link 2.4 Meter, the user may transmit blood glucose values via Bluetooth to the 670G pump to be used for sensor calibrations. If the user uses a different FDA cleared blood glucose meter to calibrate the Guardian Sensor, the user must manually input the blood glucose values into the pump to be used for sensor calibration. Additionally, users who use the Contour Next Link 2.4 should calibrate with values obtain using fingersticks; users should not use readings obtained from blood from alternative sites (e.g., palm).

## One-Press Serter

The One-Press serter is a sensor insertion device which aids the user in inserting the Guardian Sensor. The serter was also previously reviewed and approved under P120010/S070. The user must use the One-Press Serter in order to insert the Guardian Sensor.

## Contour Next Link 2.4 Meter (MMT-1352 and MMT-1152) and Test strips

The Contour Next Link 2.4 Meter can be used with the 670G system; the meter wirelessly sends blood glucose values to the insulin pump for sensor calibration via Bluetooth. The meter was also previously cleared under k110894. Specifications and performance requirements were established for the meter and evaluated as part of the class III 670G System. The sponsor verified and validated the specifications and performance requirements of the meter for the 670G System. The sponsor provided blood glucose meter specifications, rationale for requirements for the meter, and impact of error on the sensor, predictive low alerts, threshold glucose suspend, and the predictive low glucose management and hybrid closed loop features in the current submission. The sponsor carried out error impact analysis in order to determine the lot release criteria for

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 10

{10}

the meter test strips. Based on the information provided, the meter specifications meet the clinical needs of the 670G system.

## Additional System Accessories

The following additional accessory devices are compatible with the 670G Insulin Pump:

Table 3: Accessory Devices

| Device | Model |
| --- | --- |
| Reservoirs and Infusion Sets | Model Numbers |
| MiniMed Quick Set Infusion Set | MMT-386, MMT-387, MMT-394, MMT-396, MMT-397, MMT-398, MMT-399 |
| MiniMed Silhouette Infusion Set | MMT-368, MMT-369, MMT-370, MMT-377, MMT-378, MMT-381, MMT-382, MMT-383, MMT-384 |
| MiniMed Mio Infusion Set | MMT-921, MMT-923, MMT-925, MMT-941, MMT-943, MMT-945, MMT-965, MMT-975 |
| MiniMed Sure-T Infusion Set | MMT-862, MMT-864, MMT-866, MMT-874, MMT-876, MMT-886 |
| Paradigm Reservoir | MMT-332A |
| Optional Devices | Model Numbers |
| CareLink USB 2.4 | MMT-7306 |
| CareLink Online (Personal) | MMT-7333 |
| CareLink Pro | MMT-7335 |

## V. ALTERNATIVE PRACTICES AND PROCEDURES

Control of diabetes can be achieved through a combination of various behaviors and methods.

Self-behaviors include healthy eating, taking the clinically indicated medications, and being active. Persons with diabetes may also administer insulin by injection or using other insulin infusion pumps as prescribed by their physician. An insulin pump is an alternative to multiple daily insulin injections (via insulin syringe or an insulin pen). Periodic self-glucose monitoring using home use blood glucose meters provides information regarding variations in glucose levels.

Methods of monitoring glycemic control include periodic measurement of Hemoglobin A1c (HbA1c) which reflects blood glucose control over a three month period. Self-monitoring of blood glucose using glucose meters and test strips provides quantitative measurements of blood glucose at a single point in time for users and their healthcare providers. This helps to monitor the effectiveness of glycemic control, as well as make more immediate treatment modifications.

Currently, cleared or approved insulin infusion pumps may be used for continuous subcutaneous insulin infusion. Additionally, commercially available sensor-augmented insulin infusion pumps or continuous glucose monitoring systems may be used to record

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 11

{11}

continuous interstitial glucose information and provide real-time hypoglycemia and hyperglycemia alerts.

Each alternative method for monitoring glycemic control has its own advantages and disadvantages. A user should fully discuss these alternatives with his/her physician to select the method that best meets expectations and lifestyle.

## VI. MARKETING HISTORY

The 670G System has not been marketed in the United States or any foreign country.

The insulin reservoirs and infusion sets used with the 670G System are the same as those currently used with the MiniMed 530G System (P120010). These devices have not been withdrawn from commercial distribution for any reason, related to either safety or effectiveness.

## VII. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH

Potential device-related serious adverse events include

- Diabetic ketoacidosis (DKA) resulting from high blood glucose due to suspension of insulin delivery or inadequate insulin delivery (which may result from catheter occlusion, hardware or software malfunction, erroneous CGM readings in Auto Mode or suspend mode, or inadequate insulin dosing).
- Severe hypoglycemia resulting from over-delivery of insulin (which can result from hardware or software malfunction, erroneous CGM readings in Auto Mode, or erroneous insulin dosing), which may lead to seizure, unconsciousness, and rarely death.

Potential device related non serious events include:

- Skin irritation or redness
- Infection
- Pain or discomfort
- Bruising
- Edema
- Rash
- Bleeding
- Induration of skin
- Allergic reaction to adhesives

Sensor breakage with fragments retained under the skin is a potential adverse event related to use of the CGM component of the 670G system, but this was not observed during these studies. Based on postmarket experience with similar devices and the results observed in these clinical studies, the occurrence and severity of these events is low.

Infection at the insulin pump infusion set insertion site and sensor insertion site is a potential complication related to insertion of the CGM or the insulin pump infusion set.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 12

{12}

Based on post-market experience with similar devices, and the results observed in these clinical studies, the occurrence and severity of these events are not expected to be different from other approved infusion sets and CGM devices.

Use of insulin pumps are known to carry an increased risk of DKA. However, FDA has received information indicating some patients are willing to accept an increased risk of DKA or ketosis and hyperglycemia (severe hyperglycemia) because of the benefits of pump use (see also Section XII below).

Like other insulin pumps, there is an inherent risk that users of the device who do not use the 670G device as intended could harm themselves. Therefore, the device is for prescription use only and contraindicated for people unwilling or unable to perform a minimum of four fingerstick blood glucose meter tests per day and for people unwilling or unable to maintain contact with their healthcare professional.

As demonstrated under P120010/S046 for the MiniMed 530G System (which has the same ‘suspend on low’ feature, where the insulin delivery will suspend for two hours after the low glucose threshold has been reached), two hour suspension of insulin delivery is unlikely to lead to clinically significant ketosis or ketoacidosis even if the pump inappropriately suspends when the blood sugar is normal or elevated, and should respond to insulin therapy and hydration within a few hours.

There is a theoretical risk of insulin over-delivery due to device malfunction which has a risk of leading to severe hypoglycemia due to malfunction of the 670G System. This event did not occur during the pivotal study or the continuation phase of the pivotal study. If insulin over-delivery were to occur, there are several mechanisms in place designed to help detect and mitigate the risk of impending and/or current hypoglycemia, including the presence of alarms/alerts and the suspension/reduction of insulin delivery.

There is a theoretical risk of insulin under-delivery (due to a hardware or software malfunction) which may lead to severe hyperglycemia or DKA due to malfunction of the 670G system. This event did not occur during the pivotal study or the continuation phase of the pivotal study. If insulin under-delivery were to occur, there are mechanisms in place to help detect impending and/or current hyperglycemia, including the presence of alerts and alarms.

The consequences of falsely high glucose reading on the continuous glucose monitor would be potential over-delivery of insulin via automated insulin delivery and missed low glucose suspensions and alerts/alarms, which have the potential to lead to severe hypoglycemia. The consequences of falsely low glucose reading on the continuous glucose monitor would be potential under-delivery of insulin and missed high glucose alerts, which have the potential to lead to severe hyperglycemia or DKA.

## VIII. SUMMARY OF NONCLINICAL STUDIES

### A. Laboratory Studies

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 13

{13}

Pre-clinical testing was performed on the Guardian Link Transmitter, and the Guardian Sensor. Pre-clinical testing of the MiniMed 630G pump hardware supports safe use of the 670G pump, as the two pumps contain identical hardware. Please see the SSED for P150001 for descriptions of the pre-clinical testing of the MiniMed 630G pump.

The Next Link 2.4 Blood Glucose Monitoring System was previously cleared as a stand-alone blood glucose test system under k110894. For approval as part of the MiniMed 670G System, Next Link 2.4 Blood Glucose Monitoring System analytical performance information was referenced from k110894, and that analytical performance was determined to be adequate to support approval. A description of meter performance can be found in the Decision Summary in the FDA public 510(k) database; http://www.accessdata.fda.gov/cdrh_docs/reviews/K110894.pdf

Pre-clinical testing of the serter was reviewed and approved under P120010/S070.

## Guardian Link (3) Transmitter (MMT-7811)

Twenty-nine Guardian Link transmitters (MMT-7811) were subjected to the following functional and environmental tests to ensure that these devices will continue to function normally when exposed to extreme environmental conditions.

Table 4: Transmitter Functional and Environmental Tests

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Chemical Compatibility | Demonstrate the ability of various components to withstand exposure to the following chemicals for between 45 seconds – 1 minute: U100 Insulin (Humalog or Novolog), a solution of 1 part dish detergent to 9 parts water, 70% Isopropyl Alcohol. | No cracks, crazing, dissolving or discoloration to the transmitter surface. |
| Environmental storage conditions | Transmitters withstand -25 to 55°C, 10-100% relative humidity. | No visible degradation. When connected to a simulated sensor, signal current shall be 53.5 nA ±10%. When leak tested, leak rate shall be <0.40 mbar. |
| Temperature shock test | Transmitters demonstrate reliable performance after 10 cycles from -5 to 45 °C with 5 minute ramp time and 30 minute dwell time at each plateau. | No visible degradation, and signal current when connected to a simulated sensor must be 53.5 nA ±10%. When leak tested, leak rate shall be <0.40 mbar. |
| Operating environment conditions | Transmitters demonstrate the ability to operate with temperature of -5-45 °C, 95% relative humidity, 8.9-15.4 psiA | No visible degradation, and shall have a signal current of 53.5 nA ±10%. |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{14}

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Random vibration test per EN 60601-2-24 | Transmitters demonstrate reliable operation after exposure to 10-100 Hz @ (1 m/s2)2/Hz, 100-200 Hz @ -3dB/octave, and 200-2000 Hz @ 0.5 (m/s2)2/Hz for 30 minutes in each axis. | No visible degradation, and signal current when attached to simulated sensor is 53.5 nA ±10%. |
| Drop test per EN 60601-2-24 | Demonstrate safe operation after six repeated 1 meter drops onto 50 mm thick hardwood – one drop for each of 6 device faces or axes (±X, Y, Z) | No visible degradation, and signal current when attached to simulated sensor is 53.5 nA ±10%. |
| Push test per EN 60601-1 | Demonstrate that transmitters maintain performance after exposure to applied force. | When subjected to a steady force of 56 lbf ±2.25 lbf, the transmitter shows no sign of distortion or damage. |
| Mechanical Shock per IEC 60601-1-11 | Demonstrate reliable transmitter operation after exposure to 150 m/s2 (15g) acceleration, with three shocks per axis in each direction (±X, Y, Z) for a total of 18 shocks. | Transmitters show no visible degradation, and signal current when attached to simulated sensor is 53.5 nA ±10%. |
| Impact Test per IEC 60601-1 | Demonstrate ability of transmitter to withstand an impact from a 500g steel ball with approximate diameter of 50mm dropped from a height of 1.3m | Transmitter must not introduce a safety hazard to the user or other persons in the surroundings. |
| Connector insertion force | Demonstrate force required to connect battery charger and sensor to transmitter is less than 3 pounds. | Insertion force is less than 3 pounds. |
| Connector retention force | Demonstrate force required to retain battery charger to transmitter is greater than 0.5 pounds and less than 3 pounds. | Charger retention force is greater than 0.5 pounds and less than 3 pounds. |
| Connector cycling | Demonstrate that the transmitter can withstand 244 insertion/removal cycles with both the battery charger and the sensor. | Transmitters show no visible degradation, and signal current when attached to simulated sensor is 53.5 nA ±10%. |
| Mold stress relief per EN 60601-1 | Demonstrate that after release of any internal stresses due to the plastic molding process, the transmitter maintains basic safety. | After exposure to 70°C for 7 hours, transmitters are returned to room temperature and tested for basic safety. Transmitter shall not show visible damage or distortion. |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{15}

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Fluid ingress per IPX8 (International Protection) | Demonstrate reliable operation of the transmitter when submerged to a depth of 8ft for 30 minutes. | Visual inspection of transmitter shall show no water ingress. Sensor signal shall be 2.35-2.65 nA in the 2.5 nA range, 24.25-25.75 nA in the 25 nA range, and 145.5-154.5 nA in the 150 nA range. |
| Protection against solid foreign objects per IP48 | Demonstrate that the full diameter of 1.0mm spherical probe cannot pass through any opening of the transmitter. | The full diameter of a 1.0mm spherical probe cannot pass through any opening of the transmitter. |

All protocols, test reports, and acceptance criteria have been reviewed and found to be acceptable. All transmitter devices met all pre-determined acceptance criteria during this testing.

Guardian Sensor (MMT-7020)

Thirty Guardian Sensors (MMT-7020) were subjected to the following functional and environmental testing after sterilization and 365 day accelerated aging:

Table 5: Sensor Functional and Environmental Tests

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Extraction Force | To test the force required to extract the needle after insertion. | The extraction force (along the needle axis) required to extract the needle hub from the sensor base shall not exceed 1 LbF |
| Patch Pull Test | To test the force required to break the bond between the sensor patch and the sensor base | Either an applied force exceeding 2.5 LbF shall be required to separate the adhesive bonds on the patch from the sensor base or the patch material shall fail before any adhesive bond. |
| Sensor Pull Test | To test mechanical integrity of the senor/tube assembly | An applied force exceeding 0.13 LbF shall be required to break the sensor from the sensor base |
| Sensor Connection Force Test | To test the force required to connect the sensor to the compatible transmitter | The sensor shall require a connection force of no more than 4.8 pounds when connecting to the |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{16}

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 17

| | | transmitter/recorder |
| --- | --- | --- |
| Contact Pad Continuity Test | To test the resiliency of the senor contact pads after multiple connect/disconnect cycles | Each contact pad shall have a resistance of less than 10 Ohms after 6 insertion cycles with the transmitter/ recorder connector |
| Latching Force Test | To test that the connection to the transmitter is robust | An applied force exceeding 1 Lbf shall be required to separate the transmitter from the sensor base |
| Water Tightness Test | To test the ability of the connection between the sensor and the transmitter to prevent water from entering the transmitter bore | The connection between the sensor and the transmitter shall be fluid tight and meet the IPX8 requirements of IEC60529 Section 14 (submersion of 8 feet for 30 minutes) |
| SynDaver Insertion Test | To test the overall mechanical functionality of the device in a representative use-case scenario | The sensor shall neither kink nor show signs of damage upon insertion into a representative human tissue; and the needle hub shall shield the needle after removal from the sensor base |
| Accuracy Test | To test that the sensor output is within the system required limits at the extent of the glucose ranges (40 mg/dL and 400 mg/dL) | The sensor shall be capable of achieving a calibration ratio (glucose-to-sensor current ratio) between 1.5 to 15 mg/dL/nA, inclusive, within the operating range of the sensor. 90% of the individual glucose concentration values must fall within the calibration range for both 40 and 400 mg/dL solutions. |
| Linearity Test | To test that sensors show a linear response when glucose levels are driven from 40 mg/dL to 400 mg/dL in a stepwise manner | The sensor shall operate with a measured linearity of at least 0.90 (R²) within the operating range of the sensor |
| Response Time Test | To test that the senor responds adequately to sudden changes in glucose concentrations | The in-vitro signal (current) shall reach 95% of the average stead-state value 15 after a glucose concentration step change. The glucose |

{17}

Expiration dating for the Sensor component of this device has been established and approved at 12 months at storage conditions between 36°F to 81°F. The protocol and data for recommended storage and expiration date were reviewed and found to be acceptable.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 18

{18}

All protocols, test reports and acceptance criteria have been reviewed and found to be acceptable. All sensor devices met all pre-determined acceptance criteria during this testing.

The MiniMed 670G System with all components operating together, including the Guardian Sensor (3), Guardian Link (3) Transmitter, and Contour NEXT LINK 2.4 meter (MMT-1152 and MMT-1352), was subjected to the following functional and environmental tests to ensure that these devices will continue to function normally when exposed to extreme environmental conditions:

Table 6: System Functional and Environmental Tests

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| EMC/EMI Testing per EN 60601-1-2:2007 | Demonstrate ability of the system to operate in environments with EMI which meet the standard of EN 60601-1-2:2007 | Maintain basic safety and essential performance during exposure to EMI – delivery accuracy must be within ±5%. BG Meter commanded bolus amount matches pump displayed delivered amount. |
| Wireless Coexistence | Demonstrate ability of system to withstand expected levels of wireless transmission from other sources | Maintain basic safety and essential performance during exposure to wireless transmission sources – delivery accuracy must be within ±5%. |
| FCC and Avionics | Demonstrate compatibility with FCC regulation | Emitted levels must be per FCC CFR 47 Part 15.247. It is acceptable that the pump pay lose RF communication with the transmitter. In this case the pump will alarm “Lost Sensor” to notify the user. Pump operating mode shall not be affected. BG Meter commanded bolus amount matches pump displayed delivered amount. No interruption of pump alarms, and no change in pump operating mode or programmed settings. |
| X-ray Immunity | Demonstrate reliable operation when exposed to x-ray – 100kV, 100 uA exposure for 2 minutes | Pump history download shows no change in transmitted values. Pump and BG meter maintain association and complete system functionality. BG Meter |

PMA P160017: FDA Summary of Safety and Effectiveness Data

{19}

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| | | commanded bolus amount matches pump displayed delivered amount. No interruption of pump alarms, and no change in pump operating mode or programmed settings. Sensor signal shall be 53.5 nA ±10% when connected to a simulated sensor. |
| RF Performance | Demonstrate reliable system operation when multiple pump/transmitter pairs are operating within close proximity | Less than 10 % packet loss with RF communication between pump and transmitter, and between pump and BG meter. Sensor signal values shall be 53.5 nA ±10%, and there shall be no unexpected lost sensor alerts. BG Meter commanded bolus amount matches pump displayed delivered amount. |
| Electronic article surveillance immunity | Demonstrate that the system operates reliably when exposed to EMI from electronic article surveillance equipment. | Pump delivery accuracy must be within ±5%. Sensor signal values must be 53.5 nA ±10%. BG Meter commanded bolus amount matches pump displayed delivered amount. |
| Cell phone and cordless phone immunity | Demonstrate that the system operates reliably when exposed to EMI specifically in common cell phone spectra (800-960 MHz and 1700-2200 MHz @ 1MHz steps) using WCDMA, WCDMA/3GPP, GSM/EDGE, DECT, IS95, PHS, NADC, PDC, and cordless phone spectra (2400 and 900 MHz) | Pump history download shows no change in transmitted values. Pump and BG meter maintain association and complete system functionality. BG Meter commanded bolus amount matches pump displayed delivered amount. No interruption of pump alarms, and no change in pump operating mode or programmed settings. |

All protocols, test reports, and acceptance criteria have been reviewed and found to be acceptable. All system components met all pre-determined acceptance criteria during this testing.

The results for all of the above validation testing were found to be acceptable. These results support the conclusion that the MiniMed 670G System is safe for its intended use.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 20

{20}

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 21

# Biocompatibility Testing

Biocompatibility testing for the sensor components of the Guardian Sensor (Patch, tube and circuits) was performed in accordance with the recommendations of ISO 10993-1, Biological Evaluation of Medical Devices-Part 1: Evaluation and Testing. The results of these tests are listed in the Table below.

Table 7: Biocompatibility Testing Summary

| Test | Testing Summary |
| --- | --- |
| Cytotoxicity | Pass-Non-cytotoxic |
| Irritation | Pass-Non-irritating |
| Sensitization | Pass-Non-sensitizing |
| Systemic Toxicity (Acute) | Pass-Non-Toxic |
| Repeated (Sub-Acute) Toxicity | Pass-Non-Toxic |
| Genotoxicity: Bacterial Reverse Mutation | Pass-Non-genotoxic |
| Implantation (12 weeks) | Pass |

Material mediated pyrogenicity and hemolysis testing were leveraged from P120010, as the materials that comprise the Guardian sensor are the same as the materials that comprise the Enlite sensor. Please see the SSED for P120010 for additional details on testing for these materials.

# Sterility

The method employed for the sterilization of the Guardian Sensor is Electron Beam Sterilization. The sterilization process used to sterilize the sensor was validated according to the ISO 11137 Sterilization of Health Care Products – Radiation standard. All sterilized components met the standards of ISO 11137 to assure a sterility assurance level (SAL) of $10^{-6}$.

The Guardian Link Transmitter, 670G pump, and Contour NEXT Link 2.4 Wireless Blood Glucose Monitoring System are provided as non-sterile.

# Packaging/Shelf-Life:

The MiniMed 670G System (670G pump, Guardian Link Transmitter, and Guardian Sensor) was tested under conditions of simulated shipping per ASTM D4169, Standard Practice for Performance Testing of Shipping Containers and Systems. Testing included environmental conditioning, manual handling, vehicle stacking, loose load vibration, low pressure testing, vehicle vibration, concentrated impact, and final inspection of samples. The MiniMed 670G Insulin Pump (MMT-1780) has a shelf life of six months based on the internal backup battery, which requires regular

{21}

recharging. The Guardian Link Transmitter is intended to be operable for a period of 12 months.

The shelf-life of the Guardian Sensor (MMT7008) was validated to be up to one year when stored at 36°F to 81°F (+2°C to +27°C) according to the requirements of ISO 11607: Packaging for Terminally Sterilized Medical Devices, ASTM D 4169: Standard Practice for Performance Testing of Shipping Containers and Systems and ASTM F 1929: Standard Test Method for Detecting Leaks in Porous Medical Packaging by Dye Penetration.

The packaging of the Contour NEXT Link Meter 2.4 was subject to and has met the requirements for international shipping and handling using procedures and methods defined in ISTA Procedure 2A, Performance Test Procedure for Individual Packaged Products Weighing 150 Lbs. (68 Kg) or less. The operating temperature of the meter is from 41°F-113°F (5°C to 45°C) and relative humidity range from 10% to 93%.

## Software

Comprehensive verification and validation testing was conducted to confirm that the software used in the MiniMed 670G System meets all specified requirements and that the software will operate reliably and safely under normal or abnormal use conditions.

The software verification and validation were carried out in accordance with the FDA Guidance Document, "General Principles of Software Validation: Final Guidance for Industry and FDA Staff." Software development activities included establishing detailed software requirements, linking requirements with associate verification tests, software code reviews, unit testing, system level testing and defect tracking and dispositioning to ensure the software conforms to user needs and intended uses.

## Human Factors Testing

A human-factors-usability analysis was conducted in accordance with EN62366, Medical Devices – Application of Usability Engineering to Medical Devices.

Task analysis was evaluated by conducting usability studies on selected functions that were determined to have either a high or medium risk to the user. During the usability validation study, 60 representative users performed typical tasks associated with using the pump in both Manual and Auto Modes, and the Guardian continuous glucose monitoring system; four representative user groups were studied (15 per group):

- Pediatric Novice Insulin Pump Users (ages 14-21 years) (5)
- Users in this group were not currently external pump users, had less than 6 months of experience with a Medtronic pump, or were currently using a competitor pump.
- Adult Novice Insulin Pump Users (ages 22 and older) (8)

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 22

{22}

○ Users in this group were not currently external pump users, had less than 6 months of experience with a Medtronic pump, or were currently using a competitor pump.

- Pediatric Experienced Insulin Pump Users (ages 14-21 years)
○ Users in this group currently were using a Medtronic external insulin pump for more than 6 months.

- Adult Experienced Insulin Pump Users (ages 22 and older)
○ Users in this group were using a Medtronic external insulin pump for more than 6 months.

These representative users only included a few truly pump naïve users. This device is available by prescription only, and the healthcare provider will provide training, and will determine who is most appropriate to use the device.

All errors and near errors from participants during the usability study were collected and documented for basic statistical analyses. A root cause of any user error was classified as a device issue (e.g., user did not know how to use a portion of the device such as how to disconnect the tubing from the infusion site), test artifact, or use error. For any use errors, a residual risk analysis was performed to determine the impact of the error that could potentially lead to a residual risk of significant over or under-delivery of insulin.

The usability evaluations performed demonstrated that users understood the instructions provided in the user guide and that they could use the device safely.

## B. Animal Studies

Animal testing was performed on an early version of the 670G System control algorithm in Auto Mode. The animal studies were performed using canines to evaluate the probable safety of the closed loop system (Auto Mode) before entering into human studies. The animal studies showed that the early algorithm was safe and based on these results, Medtronic moved the device into human studies.

## C. Additional Studies

None

## IX. SUMMARY OF PRIMARY CLINICAL STUDIES

Medtronic performed clinical studies to establish a reasonable assurance of safety and effectiveness of the 670G System. A summary of the clinical studies is presented below.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 23

{23}

Table 8: Summary of P160017 Clinical Studies

| Clinical Study | IDE | Patient Population | Study Design/Objective |
| --- | --- | --- | --- |
| Safety Evaluation of the Hybrid Closed Loop (HCL) System in Type 1 Diabetes | G140167 | 14-75 years (Type 1 Diabetes) | Multi-center, single-arm, home and hotel clinical study. The study evaluated the safety of the 670G System and its algorithm with the Guardian Sensor in subjects 14 to 75 years. |
| A Performance Evaluation of the Enlite® 3 Sensor to Support a Full 168 hours (7 Days) of Use | G140053 | 14-75 years (Type 1 or Type 2 diabetes) | Multi-center, prospective, single-sample correlational design without controls. The study demonstrated the measurement performance of the Guardian Sensor over 7 days in subjects 14 to 75 years. |

Pivotal study: Safety Evaluation of the Hybrid Closed Loop (HCL) System in Type 1 Diabetes (G140167):

## A. Study Design

Subjects were treated between June 3, 2015 and March 7, 2016 and included 123 patients. There were 10 investigational sites.

The pivotal study was a multi-center, single-arm home and hotel clinical evaluation in subjects with type 1 diabetes on insulin pump therapy. The sponsor enrolled 126 subjects (ages 14-75 years) at 10 investigational centers. Three subjects did not complete the study (see subject accountability below).

The 123 study subjects wore the 670G pump with the Guardian Link Transmitter, the Guardian Sensor and infusion sets for approximately 3.5 months and participated in three study phases: a two week run-in period, a three month at home use period, and a 5 day/6 night hotel study, which occurred during month 1, month 2, or month 3 of the study. Subjects were instructed to use the device in auto mode for the duration of the 3 month at home study.

## Run-in period

During the two week run-in period, subjects used the study pump (670G) with only the sensor augmented pump function activated (all automated features were off). Prior to wearing study devices, all subjects and their companions were trained on the devices as well as diabetes management principles, such as the treatment of hyperglycemia and hypoglycemia. In addition, there was training regarding the need to have access to oral glucose in case of hypoglycemia. Subjects were instructed to

PMA P160017: FDA Summary of Safety and Effectiveness Data

{24}

monitor blood glucose using self-monitoring of blood glucose (SMBG) 4-6 times a day. As a precaution, subjects were told that they should keep their own insulin pump supplies in case they were asked during the study to revert back to using their own pump. Subjects were also instructed that they should always have insulin and syringes or pens, in case they encountered problems with the study pump (e.g., infusion set occlusion with high glucose).

## At Home Study Period

Following the two week run-in period using the Study Pump (670G), a total of 123 subjects participated in a 3-month at home study period. Prior to entry into Auto Mode, subjects used the pump in Manual Mode during the first 6 days of the study period in order to collect data on insulin utilization and sensor glucose levels. After this 6 day period, the subjects were allowed to enter Auto Mode.

Subjects were required to have a companion with them during the night for the duration of the study period. Companions were instructed to be under the same roof (i.e., within range and able to hear sensor alarms), but not necessarily in the same bedroom as the study subject. Subjects were also required to upload their pump data daily for the first 14 days after entering into Auto Mode to facilitate remote monitoring by the study sponsor.

The required device settings for the study are below unless otherwise stated as optional:

- Manual Mode:
- High sensor glucose alert set at 300 mg/dL
- Low sensor glucose alert set at 70 mg/dL
- Predictive alerts and rate of change alerts are optional
- It was recommended (optional) to have the ‘Suspend before Low’ ON with the low limit setting of 70 mg/dL.

- Auto Mode:
- High sensor glucose alert set at 300 mg/dL
- Low sensor glucose alert set at 70 mg/dL

- The target glucose during Auto Mode
- The target glucose was 120 mg/dL.
- A temporary target of 150 mg/dL could be used in certain scenarios (e.g., exercise).
- Alarms that were mandatory (fixed into the system):
- Sensor glucose at or below 50 mg/dL
- Sensor glucose at or above 300 mg/dL for one hour
- Sensor glucose at or above 250 mg/dL for 3 hours

## Hotel Study Period

A total of 122 of the 123 subjects participated in the Hotel portion of the study (6 days, 5 nights); one subject did not complete the hotel visit and removed the sensor and study devices for the last 30 days of the study due to rheumatoid arthritis and C.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 25

{25}

difficile diarrhea (unrelated to device use and study protocol). There was an average of 4-6 subjects participating in each hotel stay. The hotel stays were to occur during the first, second or third month of study, with a minimum of 20 subjects completing the hotel stay each month. The purpose of the hotel portion of the study was to stress the user with sustained daily exercise and unrestricted eating (see description below).

All subjects underwent daytime and nighttime frequent sample testing for approximately 24 hours during the Hotel study with a laboratory blood glucose method. Overnight frequent sample testing occurred every 30 minutes from approximately 10PM to 7AM. Daytime frequent sample testing was every 60 minutes from approximately 7AM to 10PM. Subjects also participated in daily sustained exercise/activity regimens for a minimum of 4 hours spread throughout the day. With respect to meals, subjects were allowed to eat what they chose.

On Days 1-6 of the Study period, the subjects used the device in Manual Mode to collect data necessary for the system in order to enter into Auto Mode. Entry into Auto Mode occurred on day 7 whether at home or in the hotel. If a subject was scheduled for their hotel visit in month 1, the hotel visit would occur during days 7-14 of the study.

The study was a multi-center, single arm observational at home and hotel clinical study with no controls. There were no statistically powered endpoints in the Auto Mode study (G140167). This was a descriptive study to evaluate the safe use of the 670G System.

1. Clinical Inclusion and Exclusion Criteria
Enrollment in the pivotal study was limited to subjects who met the following inclusion criteria.

1. Subject is age 14-75 years at the time of screening
2. Subject has a clinical diagnosis of type 1 diabetes for 2 years or more as determined via medical record or source documentation by an individual qualified to make medical diagnosis.
3. Subject is willing to participate in a hotel study for the specified duration of hotel stay.
4. Subject must have companion who will sleep in the same dwelling place every night during the study period and should also be able to call the subject daily in the event the subject is traveling. This requirement may be verified by subject report at screening visit.
5. Subject is willing to perform ≥ 4 finger stick blood glucose measurements daily.
6. Subject is willing to perform required sensor calibrations
7. Subject is willing to wear the system continuously throughout the study
8. Subject has a Glycosylated hemoglobin (HbA1c) value less than 10.0% (as processed by Central Lab) at time of screening visit.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 26

{26}

9. Subject has TSH in the normal range OR if the TSH is out of normal reference range the Free T3 is below or within the lab’s reference range and Free T4 is within the normal reference range.

10. Pump therapy for greater than 6 months prior to screening (with or without continuous glucose monitor experience)

11. Subject is willing to upload data from the study pump, must have Internet access and a computer system that meets the requirements for uploading the study pump

12. If subject has celiac disease, it has been adequately treated as determined by the investigator

13. Subject is willing to take one of the following insulins and can financially support the use of either of the 2 insulin preparations throughout the course of the study (i.e., copayments for insulin with insurance or able to pay full amount)

- Humalog® (insulin lispro injection)
- NovoLog® (insulin aspart)

14. Subjects with history of cardiovascular event 1 year or more from the time of screening must have an EKG within 6 months prior to screening or during screening. If subject has an abnormal EKG, participation is allowed if there is clearance from a cardiologist

15. Subjects with the 3 or more cardiovascular risk factors listed below must have an EKG within 6 months prior to screening or during screening. If subject has an abnormal EKG, participation is allowed if there is clearance from a cardiologist

Cardiovascular risk factors include:

- Age &gt;35 years
- Type 1 diabetes of &gt;15 years’ duration
- Presence of any additional risk factor for coronary artery disease
- Presence of microvascular disease (proliferative retinopathy or nephropathy, including microalbuminuria)
- Presence of peripheral vascular disease
- Presence of autonomic neuropathy

16. Subjects with history of cardiovascular event 1 year or more from the time of screening must have a stress test within 6 months prior to screening or during run in period. If subject fails stress test, participation is allowed if there is clearance from a cardiologist

17. Subjects must be able to speak English, and be literate in English

Subjects were not permitted to enroll in the pivotal study if they met any of the following exclusion criteria:

1. Subject has a history of 2 or more episodes of severe hypoglycemia, which resulted in any of the following during the 6 months prior to screening:

a. Medical assistance (i.e., Paramedics, Emergency Room or Hospitalization)

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 27

{27}

b. Coma
c. Seizures

2. Subject is unable to tolerate tape adhesive in the area of sensor placement
3. Subject has any unresolved adverse skin condition in the area of sensor placement (e.g., psoriasis, dermatitis herpetiformis, rash, Staphylococcus infection)
4. Females who are sexually active and able to conceive will be excluded if they are not using an effective method of contraception and do not agree to continue using an effective method of contraception for the duration of the study as determined by investigator.
5. Subject has had any of the following cardiovascular events within 1 year of screening: myocardial infarction, unstable angina, coronary artery bypass surgery, coronary artery stenting, transient ischemic attack, cerebrovascular accident, angina, congestive heart failure, ventricular rhythm disturbances or thromboembolic disease
6. Subject is being treated for hyperthyroidism at time of screening
7. Subject has diagnosis of adrenal insufficiency
8. Subject has had DKA in the 6 months prior to screening visit
9. Subject has taken any oral, injectable, or intravenous steroids within 8 weeks from time of screening visit, or plans to take any oral, injectable or intravenous steroids during the course of the study
10. Subject is actively participating in an investigational study (drug or device) wherein he/she has received treatment from an investigational study drug or investigational study device in the last 2 weeks.
11. Subject has been hospitalized or has visited the ER in the 6 months prior to screening resulting in a primary diagnosis of uncontrolled diabetes.
12. Subject is currently abusing illicit drugs
13. Subject is currently abusing marijuana
14. Subject is currently abusing prescription drugs
15. Subject is currently abusing alcohol
16. Subject is using pramlintide (Symlin), DDP-4 inhibitor, liraglutide (Victoza or other GLP-1 agonists), metformin, canagliflozin (Invokana or other SGLT2 inhibitors) at time of screening
17. Subject has a history of visual impairment which would not allow subject to participate in the study and perform all study procedures safely, as determined by the investigator
18. Subject has elective surgery planned that requires general anesthesia during the course of the study
19. Subject has a sickle cell disease, hemoglobinopathy; or has received red blood cell transfusion or erythropoietin within 3 months prior to time of screening
20. Subject plans to receive red blood cell transfusion or erythropoietin over the course of the study participation
21. Subject diagnosed with current eating disorder such as anorexia or bulimia
22. Subject has been diagnosed with chronic kidney disease that results in chronic anemia

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 28

{28}

23. Subject has a hematocrit that is below the normal reference range of lab used
24. Subject is on dialysis
25. Subject has serum creatinine of &gt; 2 mg/dL

## 2. Follow-up Schedule

Throughout the study period there were a number of scheduled telephone calls. These calls were meant to make sure that the subject was healthy and to remind them about adherence to study requirements, for example uploading the study pump data to CareLink.

During the final study visit, subjects were asked to complete some questionnaires about their experience and also had blood collected for an HbA1c test.

## Hotel Study Phase

The subject received a follow-up phone call from the study doctor’s staff within 6-18 hours from discharge after the visit at the hotel or clinic to address any questions or concerns the subject had and to ask how they were doing.

## 3. Clinical Endpoints

There were no statistically powered endpoints in the pivotal study, nor was there any hypothesis testing. This was a descriptive study to evaluate the safe use of the Auto Mode. The analyses used for the various study phases are as follows:

### At-home Study Phase

- The mean change in HbA1c will be presented from baseline to end of study
- Change of Total Daily Dose (TDD) of insulin and weight from baseline to end of study
- Time spent in Auto Mode versus time spent in Manual Mode
- Time in different range (% of sensor glucose [SG]): SG≤50, 60, 70 mg/dL, 70 mg/dL <sg≤180 (dL,="" 250="" 350="" 50,="" 70="" and="" gL,="" gsc≥180,="" mg="" mg="" ml="" number="" of="" on="" or="" pure="" same="" sg="" sg≥180,="" the="" time="" time="" umc)="" vs="" week,="" x="" xge:="" xuc="" –="">&gt;180, 250 mg/dL, 350 mg/dL
- Number of Events, AUC and Time in the hypoglycemic range: SG≤50, 60, and 70 mg/dL

### Hotel Study Phase

- Time in different ranges (per the comparator method (CM): BG ≤50, 60, 70 mg/dL, 70 mg/dL <sg≤180 (dl,="" 250="" 350="" 50,="" 70="" gsc≥180,="" mg="" ml="" ml="" number="" of="" on="" or="" pure="" same:="" sg≤180,="" the="" time="" umc="" vs.="" www.="" www.dc="" www.dl="" –="">&gt;180, 250 mg/dL, 350 mg/dL
- Number of Events, AUC and Time in the hypoglycemic range: CM ≤50, 60, and 70 mg/dL

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 29</sg≤180></sg≤180>

{29}

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 30

## Safety Endpoints
- Serious adverse events, serious adverse device events
- Unanticipated adverse device effects
- Incidence of severe hypoglycemia
- Incidence of DKA

## B. Accountability of Study Cohort

A total of 126 subjects entered the run-in period, 2 subjects did not meet the run-in criteria and 124 subjects entered the study period; 1 subject was withdrawn during the study period because the subject felt too restricted by the study protocol and the inability to control mildly elevated glucose levels with frequent correction boluses; therefore, 123 subjects completed the home study phase.

The demographics of the study population are typical for studies performed in the Type 1 diabetes population performed in the US.

## C. Study Population Demographics and Baseline Parameters

| Characteristic | Number of Subjects =124 |
| --- | --- |
| Age (Years) | |
| N | 124 |
| Mean (SD) | 37.8 (16.46) |
| Median | 39.0 |
| Min, Max | 14.0, 75.0 |
| Gender N (%) | |
| Female | 69 (55.6%) |
| Male | 55 (44.4%) |
| Race N (%) | |
| Asian | 3 (2.4%) |
| Black/African American | 1 (0.8%) |
| Other | 1 (0.8%) |
| White | 119 (96.0%) |
| Ethnicity N (%) | |
| Hispanic/Latino | 3 (2.4%) |
| Non-Hispanic/Non-Latino | 109 (87.9%) |
| Not reportable per local law or regulation | 12 (9.7%) |
| Diabetes History(Years) | |
| N | 119 |
| Mean (SD) | 21.7 (13.65) |
| Median | 19.0 |
| Min, Max | 2.0, 57.0 |

{30}

| Height (cm) | |
| --- | --- |
| N | 124 |
| Mean (SD) | 171.2 (10.29) |
| Median | 170.3 |
| Min, Max | 147.1, 198.5 |
| Weight (kg) | |
| N | 124 |
| Mean (SD) | 76.9 (17.86) |
| Median | 73.5 |
| Min, Max | 46.7, 175.0 |
| BMI | |
| N | 124 |
| Mean (SD) | 26.2 (5.27) |
| Median | 25.0 |
| Min, Max | 17.0, 46.0 |
| Baseline A1C | |
| N | 124 |
| Mean (SD) | 7.4 (0.91) |
| Median | 7.3 |
| Min, Max | 5.2, 9.7 |

## D. Safety and Effectiveness Results

### 1. Safety Results

The analysis of safety was based on the pivotal study (G140167) and the Guardian sensor performance study (G140053 – see study description below) which included 123 subjects available for the 3.5 month evaluation, and 89 subjects available for two weeks, respectively. Key safety outcomes and adverse effects are reported below for the pivotal study. The continuation phase of the pivotal study (G140167/S004), as well as the confirmatory (post-market) study (G160046) for the 670G system should provide further information regarding the safety of the device (including incidence of harmful events), as well as durability of the results noted during the pivotal study. See below for additional descriptive details for the continuation and confirmatory studies.

### Adverse events that occurred in the pivotal study:

The safety data of the 670G System was assessed by evaluation of the incidence of all serious Adverse Events, Adverse Device Effects (ADEs), Serious Adverse Device Events (SADEs), and Unanticipated Adverse Device Effects (UADEs) experienced by study subjects. Adverse events (AEs) were listed in terms of severity and relationship to the device.

There were 4 reports of serious adverse events.

- 1 report of appendicitis

PMA P160017: FDA Summary of Safety and Effectiveness Data

{31}

- 1 report of bacterial arthritis of right wrist
- 1 report of C. difficile diarrhea
- 1 report of worsening rheumatoid arthritis

There were no reports of unanticipated adverse device effects

There were no reports of DKA

There were no reports of severe hypoglycemia events.

There were 4 procedure-related events
- thrombophlebitis
- pain
- irritation/bruising
- pain at the IV site

There were 24 severe hyperglycemia events reported.

Severe hyperglycemia was defined in the protocol as a glucose concentration &gt;300 mg/dL with blood ketones &gt;0.6mmol/L or accompanied by symptoms of nausea, vomiting or abdominal pain. Table 9 (below) summarizes the severe hyperglycemic events which were reported during the study, and provides a column that identifies any mitigations, where applicable.

The HCL study was not powered to detect if the incidence of adverse events varies by subpopulation.

PMA P160017: FDA Summary of Safety and Effectiveness Data
Page 32

{32}

Table 9: Summary of Severe Hyperglycemic Events

| Event Description | Run-in Phase | Study Phase | Mitigations Implemented |
| --- | --- | --- | --- |
| Not device related | 2 | 5 | N/A- not device related |
| Infusion set issues | 5 | 6 | No changes to the 670G pump or Guardian system as the events were due to infusion set site issues |
| Software or hardware issues resulting in depletion of pump's battery backup | 0 | 5 | A software anomaly that is related to the pump's battery depletion was fixed and applied in the software that is included in the approved device |
| Sensor values triggering the safe basal insulin delivery rate that was not sufficient to maintain normal glucose levels | 0 | 1 | N/A- No changes to the device were made because this is a risk mitigation by design of the 670G System when the user is in Auto Mode and it senses that the sensor is not performing as needed by the system. |

# 2. Effectiveness Results

System Performance: Pivotal Study (G140167)

The data below describe how the device performed during the pivotal study. The study was not designed to determine the effectiveness of the device compared to alternative treatments such as manual daily injections or non-automated insulin pump therapy.

The table below provides an overall summary of the run-in phase and study phase (home and hotel) for all subjects in the study. The data presented in this table

PMA P160017: FDA Summary of Safety and Effectiveness Data

{33}

includes information about subjects' glucose levels, insulin delivered and weight during run-in versus study phases.

Table 10: Percent of Time within Glucose Ranges, Mean Insulin Delivery, and Mean Weight of Subjects during Run-in and Study Phases

| Parameter | Run-In (Two weeks) | Study (Three months) |
| --- | --- | --- |
| Sensor glucose, mean ± SD (median) mg/dL | 150.2±22.7 (150.1) | 150.8±13.7 (149.9) |
| Percent of time with glucose level in range: mean ± SD (95% CI), % | | |
| SG <= 50 mg/dL | 1.0±1.09(0.76, 1.15) | 0.6±0.61(0.49, 0.70) |
| SG <= 60 mg/dL | 2.6±2.32(2.22, 3.04) | 1.5±1.17(1.31, 1.73) |
| SG <= 70 mg/dL | 5.9±4.08(5.13, 6.59) | 3.3±1.97(2.91, 3.61) |
| SG >70 <= 180 mg/dL | 66.7±12.17 (64.58, 68.91) | 72.2±8.83(70.64, 73.78) |
| SG > 180 mg/dL | 27.4±13.69 (24.96, 29.83) | 24.5±9.21(22.90, 26.17) |
| SG > 250 mg/dL | 6.9±7.44(5.61, 8.26) | 5.6±4.45(4.83, 6.41) |
| SG > 300 mg/dL | 2.3±4.19(1.57, 3.06) | 1.7±1.89(1.33, 2.00) |
| SG > 350 mg/dL | 0.7±1.68(0.41, 1.00) | 0.5±0.71(0.36, 0.61) |
| Within-day SD of glucose mean ± SD (median, interquartile range) mg/dL | 50.1±9.9 (48.9, 43.7-56.2) | 46.7±7.3 (45.6, 41.7-50.4) |
| Within-day coeff…