DIABETIC KETOACIDOSIS ASSOCIATED WITH ALPELISIB TREATMENT OF METASTATIC BREAST CANCER

Abstract

Objective:

Alpelisib-induced diabetic ketoacidosis (DKA) is a rare, but life-threatening, adverse event. There have been only 2 reported cases in the literature. We describe such a case, with emphasis on the importance of screening and achieving adequate glycemic control prior to and after initiation of therapy.

Methods:

A 49-year-old woman, known to have advanced breast cancer, presented with a 3-day history of nausea, vomiting, and diffuse abdominal pain. She had started alpelisib at 300 mg/day 2 months prior to presentation, after failing other options. She was diagnosed with DKA using her clinical and laboratory features, leading to treatment with hydration and intravenous insulin therapy.

Results:

Laboratory data showed high anion gap metabolic acidosis, hyperglycemia, and ketonemia with negative GAD-65 antibodies, leading to the diagnosis of alpelisib-associated DKA. Alpelisib was held, and she was treated with intravenous insulin and hydration. When DKA and hyperglycemia resolved, alpelisib was resumed at a lower dose (200 mg/day) and her blood glucose was managed using a regimen combining insulin and metformin.

Conclusion:

Phosphatidylinositol-3 kinase signaling is important for the metabolic actions of insulin, and alpelisib has been associated with severe hyperglycemia. Metformin is the first-line treatment, however when DKA is the presenting syndrome, insulin needs to be considered. Blood glucose and hemoglobin A_1c should be checked prior to treatment initiation and monitored closely after drug initiation. DKA, albeit rare, must be considered in an acutely ill, alpelisib-treated patients presenting with metabolic acidosis, and if drug discontinuation is not an option, insulin treatment may be required to control glycemia.

INTRODUCTION

Diabetic ketoacidosis (DKA) is a life-threatening condition characterized by ketonemia, acidosis, and hyperglycemia (1). It is typically seen in patients with type 1 diabetes mellitus, however it can occur in type 2 diabetes mellitus as well. Alpelisib is a phosphatidylinositol 3 kinase (PI3K) catalytic subunit α selective inhibitor. It has recently been approved to treat hormone receptor-positive, human epidermal growth factor receptor 2-negative, PI3K-mutated, advanced or metastatic breast cancer (2).

Its approval was based upon a phase 3, randomized, placebo-controlled trial which included 572 patients and demonstrated a median progression-free survival of 11 months (95% confidence interval is 7.5 to 14.5 months) for those who received alpelisib plus fulvestrant in combination, compared with 5.7 months (95% confidence interval is 3.7 to 7.4 months) for those who received placebo plus fulvestrant (hazard ratio 0.65; 95% confidence interval is 0.5 to 0.85 with p <0.001) (2). Adverse events were common, including diarrhea, decreased appetite, rash, anemia, lymphopenia, hepatotoxicity, weight loss, and hyperglycemia (2). In the combination arm, 181 patients (63.7%) reported any hyperglycemia (blood glucose >160 mg/dL), 93 patients (32.7%) had grade 3 hyperglycemia (250 to 500 mg/dL), and 11 (3.9%) had grade 4 hyperglycemia (>500 mg/dL). DKA was reported in 0.7% of patients. At baseline, 56% had pre-diabetes and 4% carried a diagnosis of diabetes based on fasting blood glucose and hemoglobin A_1c (HbA_1c). Patients with uncontrolled hyperglycemia and type 1 diabetes mellitus were excluded.

In the total cohort, 18 patients (6.3%) discontinued alpelisib due to hyperglycemia (2). Interestingly, increases in fasting blood glucose and HbA_1c were more pronounced in patients who had diabetes mellitus or pre-diabetes at baseline. Of the patients with pre-diabetes, 74% experienced clinically significant hyperglycemia during therapy (3). The median time to first occurrence of hyperglycemia was 15 days (range was 5 to 17 days) and the median time to improvement was 8 days after treatment discontinuation (range was 2 to 65 days) (3). In clinical trials, 80% of patients experienced hyperglycemia, and 87% of those required treatment, either by metformin alone, or in combination with other agents (3).

CASE REPORT

A 49-year-old woman, known to have advanced breast cancer with liver metastases, presented with a 3-day history of nausea, vomiting, and diffuse abdominal pain. She had received, and failed, multiple lines of hormonal and chemotherapy treatments. Most recently, molecular profiling performed on a tissue biopsy revealed the presence of a PI3K catalytic subunit α mutation. Thus, the patient was started on alpelisib at 300 mg/day 2 months prior to current presentation.

She was not known to have hyperglycemia prior to drug initiation. Due to coronavirus disease 2019, she had to self-isolate and monitor her free blood glucose at home. Her self-reported glucose levels were within normal range for the first month, and she stopped reporting the next month. She reported an 11-kg weight loss within the month before her admission, however this was attributed to poor appetite. She denied fever, diarrhea, genitourinary, or respiratory symptoms.

On physical exam, she appeared ill. She was afebrile, hypertensive (140/91 mm Hg), and tachycardic (100 beats/minute). She had dry mucous membranes and reduced skin turgor. Cardiovascular, pulmonary, and neurological exams were unremarkable. She had diffuse abdominal tenderness. Laboratory data showed high anion gap metabolic acidosis and hyperglycemia (pH of 7.31, sodium of 133 mEq/L [reference range is 136 to 145 mEq/L], potassium of 4.3 mEq/L [reference range is 3.5 to 5.1 mEq/L], chloride of 97 mEq/L [reference range is 98 to 107 mEq/L], bicarbonate of 4 mEq/L [reference range is 22 to 29 mEq/L], anion gap of 31, creatinine of 0.56 mg/dL [reference range is 0.5 to 0.9 mEq/L], albumin of 4.8 g/dL [reference range is 3.5 to 5.2 mEq/L], and glucose of 387 mg/dL) in the presence of ketonemia and ketonuria, consistent with the diagnosis of DKA. Lactic acid, amylase, and lipase were normal. HbA_1c was 11% (97 mmol/mol) and C peptide was 1.07 μg/L (reference range is 0.8 to 4.2 μg/L). Anti-glutamic acid decarboxylase-65 antibody serology was sent upon admission and found to be negative.

The patient was admitted to the intensive care unit and treated with intravenous insulin at a rate of 0.1 U/kg/hour per DKA protocol. Alpelisib was held. Anion gap and pH levels normalized 3 days after admission and she was transitioned to a basal/bolus subcutaneous insulin regimen (insulin glargine at 10 U/day and insulin aspart at 2 to 4 units subcutaneously 3 times daily).

Her clinical status improved gradually and she resumed feeding. Her average blood glucose ranged between 140 to 160 mg/dL. An attempt was made to resume alpelisib at a lower dose (250 mg/day) and she remained hospitalized for monitoring. Within a day of resuming alpelisib, her blood glucose levels spiked, peaking 4 hours after administration, up to 300 mg/dL. Her insulin requirements increased 3-fold with continued hyperglycemia. Alpelisib was held again and she was discharged home on degludec with insulin aspart combination at 16 U once daily. She was asked to monitor her blood glucose 3 to 4 times/day. When her hyperglycemia resolved, we resumed alpelisib at a lower dose (200 mg/day), started metformin at 1,000 mg/day, and continued the insulin combination, titrating it up to 34 to 36 U daily. The combination of metformin and insulin allowed her blood glucose to remain well controlled (average 120 to 140 mg/dL) while continuing low-dose alpelisib therapy up to the time this report was written.

DISCUSSION

The standard of care for patients with hormone receptor-positive, human epidermal growth factor receptor 2-negative advanced breast cancer is endocrine therapy (aromatase inhibitors or gonadotropin-releasing hormone agonists) with or without CDK 4/6 inhibitors (4). However, acquired resistance to endocrine therapy remains a challenge to oncologists and patients. Alpelisib was recently approved to treat patients with acquired PI3K catalytic subunit α mutations. Insulin activates the PI3K signaling cascade resulting in glucose uptake in peripheral cells. In the liver, insulin stimulates glycogen synthesis and inhibits gluconeogenesis (5). Inhibition of the p110α subunit of PI3K was associated with hyperinsulinemia and hyperglycemia. This side effect is specific to α subunit inhibitors. It was described in selective p110α inhibitors (alpelisib and taselisib) as well as pan-PI3K inhibitors (copanlisib) (6). Hyperglycemia secondary to alpelisib is continuous, with a delayed peak at days 8 and 15. It occurs 4 to 6 hours after each dose, and the glucose nadirs prior to the subsequent dose (7). Fortunately, this effect is reversible (8).

Alpelisib dose modifications and hyperglycemia management follows specific guidelines. For grades 1 and 2, no dose adjustment is required. Initiation or intensification of anti-hyperglycemic treatment may be required. If blood glucose does not fall below 160 mg/dL within 21 days, alpelisib dose should be reduced by 1 level (3). For grade 3, management is similar with the added consideration of an additional drug for 1 to 2 days until hyperglycemia improves, and treatment with intravenous hydration and correction of electrolyte disturbances as indicated. If blood glucose decreases below 160 mg/dL within 3 to 5 days, the drug can be resumed at 1 lower dose level. If blood glucose does not fall below target within 21 days, the drug should be discontinued permanently. Grade 4 should be managed as grade 3. If blood glucose remains above 500 mg/dL, permanent discontinuation of the drug is warranted (3).

Metformin is the first line of treatment (3). It should be initiated at a dose of 500 mg/day and increased gradually. Sodium glucose cotransporter 2 inhibitors have been proposed as the second line of treatment, especially if metformin is not well tolerated due to gastrointestinal side effects, however euglycemic DKA is a potential risk. Sulfonylureas and insulin should be used with caution due to the increased risk of hypoglycemia as the effect of this drug subsides. Glipizide, glyburide, and glimepiride may not be effective given that alpelisib is a CYP2C9 inducer (7). Insulin may be used for 1 to 2 days until hyperglycemia resolves. However, this may not be necessary given the short half-life of alpelisib.

Our patient developed DKA, hence it was decided to continue insulin therapy indefinitely, given the lack of clear guidelines. Metformin was not an option initially given her metabolic acidosis and nausea, however it was added later on at a low dose (500 mg/day) due to severe hyperglycemia and high insulin requirements. Metformin was well tolerated with good control and close monitoring, thus allowing her to continue alpelisib, which was indeed one of her last resort options as she had failed multiple prior therapies.

Management required a multidisciplinary approach to comprehensively address all the patient’s medical issues. A facilitated, multi-pronged communication that included counseling the patient up front, close monitoring throughout the treatment, dose reductions as needed, and proactive anti-hyperglycemic management was able to help this patient.

CONCLUSION

Alpelisib is a promising drug used to treat advanced, resistant breast cancer with PI3K catalytic subunit α mutations. It is, however, associated with significant adverse events, specifically hyperglycemia. On the other hand, alpelisib-associated DKA is rarely reported and its occurrence can add significant morbidity. This underscores the need to check fasting blood glucose and HbA_1c levels prior to treatment initiation, while scrupulously monitoring glycemia during active therapy. Future studies need to help identify patients potentially at risk for occurrence of DKA while on alpelisib, and standardize management strategies, particularly in patients who have failed other available options.

Abbreviations

DKA

diabetic ketoacidosis

HbA_1c

hemoglobin A_1c

PI3K

phosphatidylinositol-3 kinase