Implications of Continuous Ketone Monitoring in Clinical Practice

Continuous Ketone Monitors: A Potential Airbag for Individuals With Diabetes

Diabetic ketoacidosis (DKA) is a serious acute complication that can be life-threatening for individuals who are insulin-dependent. Despite advancements in treatments and advanced diabetes technologies, epidemiological studies have shown that the incidence of DKA is on the rise. A critical factor in preventing worsening metabolic decompensation is the timely measurement of ketone levels. Yet, it is well-recognized that many individuals living with diabetes often neglect to measure their ketone levels.^1,2

Building on the concept of continuous glucose monitoring (CGM), enzymatic reactions that quantify beta-hydroxybutyrate levels in the interstitial fluid could provide a window into an individual’s physiology. With the potential for passive data collection, the burden of remembering to measure ketone levels could be alleviated. A small study has shown the feasibility of this approach.³

Based on experiences with CGM, a more strategic approach to integrating continuous ketone monitors (CKM) into clinical care can be considered based on the differences in the analytes being measured. In a special issue of Diabetes Obesity and Metabolism, Kong et al highlight some of these factors in their article entitled “Continuous Ketone Monitoring: Exciting Implications for Clinical Practice.”⁴ Rather than the wealth of real-time glucose data, which are crucial for decision-making and incorporation into algorithmically modulated insulin delivery, CKM data would only be meaningful if prespecified thresholds are breached. Although the authors suggest the use of trend arrows, one could speculate that such a feature could be toggled on or off, thereby allowing for personalization of the data received. Individuals with recurrent DKA may already be struggling with healthcare engagement, and more is not always better.

CKM could also provide individuals with type 1 diabetes access to adjunctive insulin agents, like sodium–glucose co-transporters. As regulatory approval has expanded to individuals without diabetes who have chronic kidney disease and/or heart failure, now is the time to expand the arsenal of tools to those with type 1 diabetes. Yet, given the increased risk of DKA associated with these agents, CKM could be the safety net needed.

The authors have provided a mockup of how data from a CKM could be viewed on retrospective reports, aligning these reports with the current color schemes and standards used for CGM. This approach may allow for more rapid adoption by clinicians, as data interpretation may seem less onerous. In addition to the proposed ambulatory ketone profile, which includes 14 days of data overlaid on a 24-hour cycle, assessing a 2-week period that shows the time spent in various “ranges” on an individual day may help clinicians home in on specific days with elevated ketone levels and discuss factors that may have led to ketone detection (an infusion set failure, insulin omission, etc), similar to how current reports help with meal-time excursions.

With any new technology or treatment, the question of cost will be raised. However, to appropriately assess cost, one must consider the utilization of emergency rooms and hospital admissions, including intensive care units. Recurrent rates of DKA are known to occur in certain populations; prioritizing access to CKM for these individuals would undoubtedly be a financially sound decision.

CGM and automated insulin delivery have revolutionized the care of individuals with diabetes. The former allows for the detection of patterns, while the latter alters the trajectory of glucose levels. Years ago, diabetes technologies were compared to self-driving cars. As the authors rightly point out, CKM is similar to an airbag for folks with diabetes; you hope you don’t need it, but it can be lifesaving. Personally, I can’t wait for the day when these technologies are fully integrated into clinical care.

References

1. Albanese-O'Neill A, Wu M, Miller KM, et al. Poor Adherence to Ketone Testing in Patients With Type 1 Diabetes. Diabetes Care. 2017;40(4):e38-e39. https://diabetesjournals.org/care/article/40/4/e38/4098/Poor-Adherence-to-Ketone-Testing-in-Patients-With
2. Hepprich M, Roser P, Stiebitz S,et al. Awareness and Knowledge of Diabetic Ketoacidosis in People With Type 1 Diabetes: A Cross-Sectional, Multicenter Survey. BMJ Open Diabetes Res Care. 2023;11(6):E003662. https://drc.bmj.com/content/11/6/e003662
3. Alva S, Castorino K, Cho H, et al. Feasibility of Continuous Ketone Monitoring in Subcutaneous Tissue Using a Ketone Sensor. J Diabetes Sci Tech. 2021;15(4): 768-774. https://journals.sagepub.com/doi/10.1177/19322968211008185
4. Kong YW, Morrison D, Lu JC, et al. Continuous Ketone Monitoring: Exciting Implications for Clinical Practice. Diabetes Obes Metab. 2024 Sep 24. Doi:10.1111/dom.15921. Online ahead of print. https://dom-pubs.pericles-prod.literatumonline.com/doi/10.1111/dom.15921#

Diabetic ketoacidosis (DKA) represents a potentially life-threatening complication of diabetes, especially in individuals with type 1 diabetes (T1D). Recently, DKA has received increased attention because the use of SGLT2 inhibitors — an important class of diabetes drugs — is associated with an increased risk of DKA. As a result of this risk, the US FDA decided not to approve SGLT2 inhibitors as adjunctive treatments for T1D. However, recognizing the important benefits of using SGLT2 inhibitors in decreasing the risk of hospitalizations for heart failure and slowing the progression of chronic kidney disease, the National Institute of Diabetes and Digestive and Kidney Diseases has invited applications to investigate whether continuous ketone monitoring (CKM) can enable the safer use of SGLT2 inhibitors in patients with T1D.

Kong et al reviewed the potential contribution of CKM in reducing the risk of DKA. They noted that CKM technology may be particularly useful for people at high risk for DKA, including patients with T1D who are receiving treatment with SGLT2 inhibitors. Although intermittent self-monitoring of β-hydroxybutyrate levels failed to prevent an unacceptably high risk of SGLT2 inhibitor-induced ketoacidosis in phase 3 trials of SGLT2 inhibitors in patients with T1D, CKM has the potential to provide rapid insights into interstitial β-hydroxybutyrate levels and trajectories. Nevertheless, the value of CKM relies on (a) accurately predicting DKA risk from interstitial β-hydroxybutyrate levels reported by CKM and (b) the ability to respond to DKA alerts with effective mitigation strategies. Below, we highlight four critical questions that may impact the utility of interstitial β-hydroxybutyrate levels:

• At least three organic acids contribute to the pathogenesis of acidosis in DKA: β-hydroxybutyric, acetoacetic, and lactic acids. Is measuring only β-hydroxybutyrate levels sufficient to assess the risk of DKA without measuring lactate and acetoacetate levels?
• What is the optimal threshold for β-hydroxybutyrate levels to initiate therapeutic interventions to prevent DKA? For example, establishing a threshold of 0.6 mmol/L did not prevent an unacceptably high risk of DKA in phase 3 clinical trials of SGLT2 inhibitors. If the threshold is set at 0.6 mmol/L, would this result in an unacceptably high rate of false-positive results? On the other hand, based on studies involving individuals with obesity, starvation ketosis is characterized by mean β-hydroxybutyrate levels of approximately 4.6 mmol/L but is not associated with clinically significant acidosis. If the threshold is set at 4.6 mmol/L to minimize the number of false-positive test results, would this result in an unacceptably high rate of false-negative results?
• The kidneys excrete ammonium ions to help maintain physiological pH in the face of stress caused by the increased acid production associated with ketogenesis. Will it be necessary to set lower threshold levels of β-hydroxybutyrate in patients with impaired renal function and acid excretion?
• The high rates of DKA observed in phase 3 clinical trials of SGLT2 inhibitors in patients with T1D exemplify the challenges in using ketone monitoring data to develop effective strategies for mitigating the risk of DKA. To what extent was this failure attributable to the limitations of intermittent ketone monitoring, which could be avoided by using CKM? Alternatively, to what extent was this failure attributable to the limitations in the design and implementation of therapeutic interventions aimed at preventing DKA?

Clinical trials can generate extensive CKM databases across diverse patient groups. These databases will help refine the interpretation of interstitial β-hydroxybutyrate levels and also enable the identification of optimal thresholds for initiating effective clinical interventions to mitigate the risk of DKA in vulnerable individuals. “An ounce of prevention is worth a pound of cure!”