Interferences With CGM Systems: Practical Relevance?

Introduction

Most systems for continuous glucose monitoring (CGM) are based on a specific electrochemical methodology for measuring glucose in interstitial subcutaneous fluid (ISF). The enzymes needed for the enzymatic reaction are located at the tip of the filaments that are inserted into the subcutaneous tissue using needles. However, depending on the particular enzyme chemistry and the voltage applied to measure the electron flow in relation to the glucose concentration in the ISF around the sensor tip, other substances can also influence the measurements taken by CGM systems, that is, shift the “glucose” data in one direction or the other. These interfering substances can be endogenous substances, as well as drugs or food components circulating in the human body.

Despite the potential impact of interfering substances on CGM accuracy, there are few publications focusing on the interference of drugs and other substances on CGM system measurements.^1-3 In addition, the CGM systems tested in these studies are mostly outdated. Manufacturers are required to perform interference studies during the clinical development process for each new CGM system and product generation to gain regulatory approval. However, the data generated by most of these studies are not published. Although interferences are well known and relevant when it comes to systems for self-measurement of blood glucose (SMBG), once again, publications are sparse. ⁴

In this editorial, the publicly available information about interfering substances and how they influence glucose measurements from various CGM systems, manufactured by different companies, are summarized, with a focus on the clinical situation of people with diabetes (PwD). The subsequent summary is not comprehensive and will not cover all available CGM systems.

Dexcom—Acetaminophen and Hydroxyurea

The CGM systems manufactured by Dexcom use a measurement voltage, where common compounds may get oxidized. In their device design, Dexcom has tried to eliminate interference as much as possible. For example, their CGM systems contain additional diffusion membrane layers in the glucose sensor’s needle tips that can block some interfering compounds based on their molecular mass. In addition, Dexcom has been able to virtually eliminate sensitivity to acetaminophen from one generation of its CGM systems to the next (ie, from G5 to G6) by adding a permselective membrane to block the diffusion of this drug to the electrode surface, thereby minimizing its effect.^5-9 It is worth noting that at least one of the studies on Dexcom’s CGM systems and their interferences was sponsored by Abbott. ⁵

Although sensitivity to acetaminophen has been minimized, Dexcom’s G6 CGM system still shows an interference with hydroxyurea, an antineoplastic drug (brand names: Hydrea, Litalir, Droxia, and Siklos) used primarily in chemotherapy for some forms of cancer, as well as treatments for sickle cell anemia.^10,11 This interference could be detected via postmarket surveillance, a process that can help manufacturers to identify potential interfering substances. According to Dexcom Medical Affairs, potential hydroxyurea interference was identified by an astute clinician who observed and documented discordance between CGM system glucose readings and blood glucose meter values after a patient took hydroxyurea. The clinician queried the medical team at Dexcom about the observed differences. The query led to an investigation, from which Dexcom confirmed that hydroxyurea is an interferent leading to falsely elevated CGM system glucose readings. Following the investigation, Dexcom worked with global regulators to develop a new precaution:

If you are taking hydroxyurea, your G6 readings may be falsely elevated and result in missed hypoglycemia alerts or errors in diabetes management decisions. The level of inaccuracy depends on the amount of hydroxyurea in your body. Use your blood glucose meter for treatment decisions if you are taking hydroxyurea.

The identification of hydroxyurea as an interferent also led to a Class 2 Device recall by the United States Food and Drug Administration (FDA; www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRes/res.cfm?ID=179621). Fortunately, medications like hydroxyurea are taken by a small group of continuous glucose monitor users.

Abbott—Ascorbic Acid

Because of the measurement technology used by Abbott, a lower potential must be applied to the sensor, reducing the risk of oxidizing substances at the sensor electrodes. However, the Abbott CGM system appears to be prone to interference by ascorbic acid. In the US User Guide for a CGM system that is relatively new on the US market (FreeStyle Libre 2, Abbott), an intake limit of 500 mg of ascorbic acid (vitamin C) per day is specified and should not be exceeded. In the German version of the User Guide from 2019, there is a reference to an interference problem with ascorbic acid: “Taking ascorbic acid (vitamin C) while wearing the sensor can lead to false-high sensor glucose values.” The 2019 Guide recommended the usage of “standard doses” of ascorbic acid (page 9), whereas the 2018 Guide lacked similar statements or recommendations. Discussions about interference from ascorbic acid have also occurred on social media sites and forums (eg, for Libre: forum.tudiabetes.org/t/anybody-have-a-cgm-hack-for-taking-vitamin-c/65300). In the United States, a study was performed with the Libre 3 (which is not yet available on the US market, but is available in Europe), which also addresses this topic (www.medtechdive.com/news/abbott-lays-out-plan-for-us-study-of-freestyle-libre-3/581380/).

To understand the background for the recommendations made in the User Guide for the Libre 2, I contacted the manufacturer and was guided to a specific clinical study that evaluated the interference with ascorbic acid for this CGM system. A summary of the data from this study can be viewed on the FDA home page (www.accessdata.fda.gov/cdrh_docs/reviews/K193371.pdf).

The investigation was a prospective, multicenter, single-arm study involving 60 participants across four sites. The study participants were PwD (>18 years of age) taking fairly high doses of ascorbic acid (one gram was given three times at four-hour intervals). Each participant wore two glucose sensors, one on the back of each upper arm, for a period of up to 10 days. The influence of ascorbic acid was determined by the difference (in mg/dL) between measurements with the Libre 2 CGM system and a comparison method. In this study, the comparison method was a laboratory instrument, the Yellow Springs Instrument Life Sciences 2300 STAT PlusTM glucose and lactate analyzer.

The study found a maximum deviation in the measurements recorded approximately two to three hours after each ascorbic acid dose. The maximum average deviation was +9.3 mg/dL after a dose of one gram. The maximum overall mean deviations were 18.4 and 19.7 mg/dL (from baseline) after the second and third 1 g doses, respectively. Based on the results of this study and results from other in-house studies, the expected positive bias from a single dose of 500 mg of ascorbic acid was estimated to be approximately 4.5 mg/dL.

While this information is helpful for the assessment of the interference problem, one may also be interested in the details of the studies, which have not yet been published in a scientific journal. Consequently, some questions remain:

• If the displayed glucose values are 10 (or even 20) mg/dL higher than the true glucose values, then this may have clinical relevance in the lower glucose range.

• What are the “pharmacokinetics” of the increase in ascorbic acid in the blood (and downstream in the ISF) after ingestion of a larger amount of ascorbic acid through food or a substitution? At least some data indicate that this can temporarily lead to circulating increases in ascorbic acid, which can potentially lead to even more pronounced deviations in the measurement results of the CGM system. ¹² There is a paucity of publications that presents blood levels of ascorbic acid over time and in relation to different (oral) doses of ascorbic acid. What were blood and ISF levels of vitamin C after a single dose or repeated doses in people with renal insufficiency across the spectrum of CKD?

• What amounts of ascorbic acid do PwD substitute in everyday life, along with the amount ingested from food, possibly without knowingly doing so?

Ascorbic acid is contained in various foods and it is often not immediately apparent to the user in what quantities. Ascorbic acid is also the most frequently supplemented vitamin (oral supplements are among the most popular sold) and this is supported by data from various studies. For substitution, with most ascorbic acid preparations the recommendation is no more than one gram per day. Such doses are promoted for preventing and treating the common cold, managing stress, and enhancing well-being. ¹² In addition, many users—especially in the winter season—substitute with ascorbic acid to strengthen their immune system. ¹³

With respect to intake of ascorbic acid via the diet, what is the recommended daily allowance (RDA) for ascorbic acid? According to the “US National Institutes for Health” (ods.od.nih.gov/factsheets/VitaminC-HealthProfessional/), the average daily level of intake sufficient to meet the nutrient requirements of nearly all (97%–98%) healthy individuals for an adult is 75 mg per day for women and 90 mg per day for men. The main sources of ascorbic acid for most individuals are fruits/products, soft drinks, vegetables, mushrooms and legumes, and vegetable-based dishes. One glass of orange juice contains 150 mg of ascorbic acid.

The use of ascorbic acid to “protect” against infections via “immune support” may be even more widespread because of the current COVID-19 pandemic. According to the aforementioned report (page 153), the median daily intake of ascorbic acid is 60 mg in men and 70 mg in women. It is worth highlighting that all these numbers are median values. In reality a wide range of behaviors exist. High levels of ascorbic acid can also occur in dialysis patients. The normal kidney clears ascorbic acid when the plasma level exceeds 60 mM, but the effect of dialysis on ascorbic acid is highly variable. ¹⁴ For patients who take large ascorbic acid supplements, the lack of the normal renal clearance mechanism can result in very high plasma levels (>200 mM). ¹⁴

As long as we do not know how much of a dose of ascorbic acid is absorbed and what the interindividual differences in absorption are, it is not easy to predict the levels of this compound circulating in the blood. Other authors have called for standardized clinical methods for measuring plasma ascorbic acid, as the instability of ascorbic acid leads to problems in laboratory analysis.^14-16

In studies that have been performed with systems for SMBG evaluating the impact of high doses of ascorbic acid application in oncology patients or patients with sepsis, some fail with high ascorbic acid levels in blood while others do not.^17-19 However, this may not be relevant for CGM systems as it is not known how high the corresponding levels of ascorbic acid are in the ISF. It is possible that ascorbic acid circulates into the ISF in the same way that glucose does because this vitamin has a similar structure to glucose. It is however known that both ascorbic acid and glucose are able to circulate into the aqueous humor fluid.

If the effects of ascorbic acid on users of the Abbott CGM systems (which all are based on the same measurement technology) are the same for ISF as the effects on the blood of healthy participants, then one can reasonably assume that at least for some users, their routine daily dose of ascorbic acid can lead to an interference problem.

Medtronic

Little is known about interferences with the Medtronic CGM systems; however, according to a statement presented at the annual Meeting of the American Diabetes Association in 2021, their new Guardian 4 CGM sensor is not compatible with acetaminophen. In addition, Medtronic sent out an “Urgent Medical Device Correction” in March 2021 about hydroxyurea interference with their Enlite and Guardian 3 CGM systems (www.medtronicdiabetes.com/res/img/pdfs/Hydroxyurea-Inaccurate-SG-Readings-Patient-Letter.pdf). According to this notice, Medtronic has not received any complaints or reports of patient harm or injuries because of this issue. However, they informed users that hydroxyurea administration results in higher sensor glucose readings compared with blood glucose readings (Table 1).

Table 1.

Conclusions by Medtronic about hydroxyurea interference with their Enlite and Guardian 3 CGM systems.

• Hypoglycemia caused by overadministration of insulin

• Inaccurate graphs or missed alarms and alerts

• Delay to or loss of sensor-enabled insulin suspension (sensor-enabled insulin pump use only)

• SG readings in CareLink reports being substantially higher than BG readings

What you should do

• Do not use continuous glucose monitors while taking hydroxyurea.

• Only use your SMBG meter to monitor glucose levels while taking hydroxyurea.

• Consult with your health care professional if you are taking hydroxyurea and using a continuous glucose monitor.

• When using an insulin pump system, disable the CGM feature by going to the sensor settings. Refer to the user guide that came with your insulin pump for instructions.

• Always check the label of any medication to confirm with your health care professional whether hydroxyurea or hydroxycarbamide is an active ingredient.

Abbreviations: CGM, continuous glucose monitoring; BG, blood glucose; SMBG, self-measurement of blood glucose; SG, sensor glucose.

What to Do?

Manufacturers should make the results of studies performed during the clinical development process of their CGM systems publicly available. In addition, manufacturers should quickly investigate new drugs that gain widespread use among PwD, such as SGLT-2 inhibitors, and their study results and data should be made available. This is the case with SMBG systems.

If interference by a certain substance induces clinically relevant deviations in the glucose measurement results of a given CGM system, then CGM system users and their diabetes treatment team should be made aware. One way to educate continuous glucose monitor users about possible inferences is by informing PwD who participate in training programs for the use of CGM systems. ²⁰ Manufacturers can also inform users about inferences by providing them with resources. For example, users would benefit from a list of drugs that contain acetaminophen or foods that contain ascorbic acid. In principle, PwD can also inform themselves about interfering substances from the information provided by manufacturers in the User Guide for a given CGM system. In reality, it is very likely that most users will not consult their User Guides for information.

If a patient or clinician is suspicious that a medication or supplement may be contributing to erroneous CGM values, then they should contact the manufacturer’s technical support team with that information. Manufacturers of CGM systems should train their global sales teams, technical support teams, and distributors to keep in mind interferences as potential sources of measurement issues with CGM systems. They should also assess potential interferences that are identified through customer complaints and postmarket surveillance. If interferences are identified, then the manufacturers should appropriately communicate the information to users. It might also be worthwhile to evaluate such interferences in more detail via specially designed clinical trials.

Summary

In summary, more attention should be paid to substances that may potentially interfere with glucose measurements by CGM systems. It may be that in everyday life, interferences play a bigger role than many users (and diabetologists) assume. They may be the reason for the infrequent, implausible measurement results reported with CGM systems. PwD and their diabetes team need to be made aware of potential interferences and trained accordingly. As CGM is increasingly utilized in newer environments, such as hospitals and operating rooms, more clinical studies on potential interferences are required for clinicians, regulators, and manufacturers to allow safe and effective use of CGM systems.