Continuous Glucose Monitoring for Detecting Hypoglycemia During Maintenance Therapy for Acute Lymphoblastic Leukemia

Survival rate for pediatric acute lymphoblastic leukemia (ALL) has improved markedly during the 2010s, reaching approximately 93% [1]. Therefore, safe completion of intensive chemotherapy without major complications is increasingly important.

During maintenance therapy (MT), typically involving daily oral 6‐mercaptopurine (6‐MP) and weekly oral methotrexate (MTX), hypoglycemia occurs in 7%–54% of patients. Although symptomatic hypoglycemia is relatively uncommon, severe hypoglycemia may occur, particularly in younger individuals with hepatotoxicity [2].

Continuous glucose monitoring (CGM) allows the identification of symptomatic and asymptomatic hypoglycemia via interstitial glucose levels. Here, we report a case of hypoglycemic seizures during MT that was successfully managed using intermittently scanned CGM (isCGM).

A 5‐year‐old female was diagnosed with B‐cell precursor ALL (intermediate risk) at 2 years and 6 months of age. Chemotherapy was initiated according to the JPLSG ALL‐B12 protocol (JPLSG: Japanese Pediatric Leukemia/Lymphoma Study Group). During induction therapy, she experienced anaphylaxis to L‐asparaginase, necessitating its discontinuation. At the end of the induction therapy, she achieved complete remission, and subsequent chemotherapy was administered as planned without complications. At 4 years and 6 months of age, MT was initiated with 6‐MP (50 mg/m² daily) and MTX (20 mg/m² weekly). Before and during MT, she developed elevated aspartate aminotransferase (35–179 IU/L) and alanine aminotransferase (39–471 IU/L) levels.

Two weeks after starting MT, she experienced hypoglycemia (glucose 68 mg/dL) with vomiting during a morning outpatient visit, followed by recurrent early‐morning vomiting. Seven months after initiating MT, she experienced repeated early‐morning seizures, requiring emergency hospital transport on each occasion. These episodes were considered to be attributable to hypoglycemia and resolved after glucose administration. During one episode, the blood glucose level was 37 mg/dL, with undetectable insulin. During subsequent outpatient evaluations, although seizures were absent, the blood glucose level dropped to 22 mg/dL. Laboratory findings demonstrated low insulin (0.2 μIU/mL) and marked hyperketonemia (total ketone bodies 4667 μmol/L, acetoacetate 448 μmol/L, and 3‐hydroxybutyrate 4119 μmol/L), resulting in the diagnosis of ketotic hypoglycemia. From that day onward, the timing of 6‐MP administration was changed from night to morning based on previous study [3], and we employed isCGM (Freestyle Libre; Abbott Laboratories, Abbott Park, IL, USA) for monitoring hypoglycemia.

CGM demonstrated that hypoglycemia occurred predominantly between 4:00 and 7:00 AM (Figure 1A). We observed that hypoglycemia was more frequent after insufficient evening food intake, indicating that extended overnight fasting was a major contributing factor. Initially, we advised her parents to administer raw cornstarch before bedtime; however, they declined owing to difficulty with ingestion. Therefore, we recommended the parents to allow the patient to consume some snacks, including fruits, before sleeping (Figure 1B). After these interventions, no severe hypoglycemia occurred, and MT was completed safely. Although periods without CGM data were noted, hypoglycemia was observed for approximately 230 min during the total 824 h CGM recording until the end of MT. The lowest glucose level recorded was 53 mg/dL. ALL episodes were asymptomatic and resolved within 15–60 min with oral intake.

FIGURE 1.

Intermittent scanning continuous glucose monitoring (isCGM) profiles before and after bedtime snack instruction. (A) Representative isCGM profiles obtained before instructing the patient to consume snacks before bed. Four non‐consecutive days are shown,with each graph representing one day of isCGM monitoring . Frequent early morning hypoglycemia was observed, whereas no hypoglycemia occurred during the daytime. (B) Representative isCGM profiles obtained after instructing the patient to consume snacks before bed. Two non‐consecutive days are shown, with each graph representing one day of isCGM monitoring. Glucose levels remained stable throughout the day, and early‐morning hypoglycemia was not observed. The x‐axis indicates time of the day (hh:mm), and the y‐axis indicates glucose concentration (mg/dL). Open circles indicate the time points at which isCGM sensor was scanned. The numbers shown below each graph represent the maximum (upper row) and minimum (lower row) glucose values (mg/dL) recorded during each time interval. Red numbers indicate hypoglycemic values (<70 mg/dL).

In this case, isCGM enabled us to identify the timing of hypoglycemia and perform specific interventions based on CGM findings. After the patient started consuming fruits or other foods before bedtime, symptomatic early‐morning hypoglycemia did not occur. Although hypoglycemia was not completely avoided, isCGM helped prevent severe hypoglycemic episodes and supported completion of MT without further development of hypoglycemia‐induced severe complications.

Hepatotoxicity, younger age, extended overnight fasting, and elevated total bilirubin (T‐bil) levels are risk factors for 6‐MP–associated hypoglycemia [2, 4]. In the present case, hypoglycemia first developed approximately two weeks after initiation of maintenance therapy, which is similar to the timeframe reported previously for 6‐MP–associated hypoglycemia [2]. Therefore, a common underlying mechanism related to 6‐MP exposure may have contributed to the development of hypoglycemia in this patient. This patient demonstrated persistently elevated transaminase levels during chemotherapy. During MT, the median T‐bil level was 0.87 mg/dL; however, transaminase levels remained unchanged during the hypoglycemic episode, whereas T‐bil levels increased to 1.65 mg/dL. These findings suggest an association between the transient increase in bilirubin levels and the hypoglycemic episode. In our patient, prolonged maintenance therapy may additionally have contributed to the recurrence of hypoglycemic episodes. Although the mechanism of 6‐MP–associated hypoglycemia remains uncertain, it is suspected that 6‐methylmercaptopurine inhibits the liver conversion of alanine to pyruvate during nighttime fasting [3].

isCGM requires scanning to display glucose levels. Real‐time CGM systems may further help prevent symptomatic hypoglycemia as they automatically provide glucose levels and generate alerts for hypoglycemia.

It should be noted that the bioavailability of oral 6‐MP decreases when administered after meals due to inactivation by xanthine oxidase contained in daily products. However, it has been reported that no difference in relapse rates exists depending on the timing of administration, and treatment outcome can be maintained [5].

In conclusion, this case suggests that CGM may be a useful monitoring tool when 6‐MP–associated hypoglycemia is suspected; bedtime snacks including fruits may prevent early‐morning hypoglycemia during MT.

Author Contributions

K.H., T.S., and E.O. acquired data about the patient. K.H., and T.S., wrote the manuscript. Y.F., E.O., and K.H. had the clinical management of the patient. K.S., S.S., and Y.N. reviewed and critically revised the manuscript. All authors gave approval of the final version.

Consent

Informed consent for publication of the patient's clinical data was obtained from the patient's parents.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.