Hyponatremia and Other Electrolyte Abnormalities in Patients Receiving Immune Checkpoint Inhibitors

Harish Seethapathy; Nifasha Rusibamayila; Donald F. Chute; Meghan Lee; Ian Strohbehn; Leyre Zubiri; Alexander T. Faje; Kerry L. Reynolds; Kenar D. Jhaveri; Meghan E. Sise


Nephrol Dial Transplant. 2021;36(12):2241-2247. 

In This Article

Materials and Methods

We included all patients who received ICIs at the Massachusetts General Hospital Cancer Center by reviewing oncology infusion records between May 2011 and December 2018. The follow-up period began at the date of the first ICI exposure. The drug classes included in the study were CTLA-4 inhibitors, PD-1 inhibitors, PDL-1 inhibitors and combination therapy (concurrent CTLA-4/PD-1). The cancer type and ICI start date were obtained from oncology infusion records. Clinical data including comorbidities, concomitant medication use and laboratory data were collected using the Partners Healthcare Research Patient Data Registry.[7] Exposure to anticancer agents associated with electrolyte abnormalities (cisplatin, carboplatin, oxaliplatin, vincristine, vinblastine, cyclophosphamide and ifosfamide) within 6 months prior to ICI initiation or any time during the 1-year follow-up was also recorded. Baseline values for electrolytes such as sodium, potassium, calcium and phosphate were obtained from the laboratory results just prior to ICI initiation. Baseline creatinine was calculated by averaging all values obtained in the 3 months prior to ICI start. Excluded patients were those without electrolyte results within 3 months prior to starting ICI, those who did not have at least one follow-up laboratory value after the start date and those on dialysis. Laboratory data were collected for all patients for 12 months or until death, whichever occurred earlier. Comorbidities were defined by using the International Classification of Diseases 9th or 10th Revision codes at the ICI start date as we have previously described.[8] Medications included those active in the electronic health record at the date of beginning ICI. Diuretic use was defined as a prescription for a loop, thiazide or potassium-sparing diuretic.

Hyponatremia was defined as serum sodium ≤134 mEq/dL. The stages of severity were defined using the Common Terminology for Cancer Adverse Events (CTCAE) version 5.0 criteria: Grade 1, 130–134 mEq/L; Grade 2, 125–129 mEq/L; Grade 3, 120–124 mEq/L; and Grade 4, <120 mEq/L.[17] Severe hyponatremia was defined as serum sodium ≤124 mEq/L (Grade 3 or 4). Other electrolyte abnormalities were also graded using CTCAE version 5.0. Corrected calcium was calculated by using the formula 0.8 (four patient albumin levels) + measured calcium. Laboratory errors were defined as any Grade 3 or 4 abnormality that was not confirmed by a repeat test that was also out of the normal reference range. Baseline use of concomitant medications associated with electrolyte disorders was recorded, including the use of selective serotonin reuptake inhibitors (SSRIs), diuretics, proton pump inhibitors (PPIs) and angiotensin-converting enzyme (ACE) inhibitors or angiotensin-receptor blockers (ARBs). In order to identify the etiology of severe (Grade 3 or 4) cases of hyponatremia, a detailed review of the electronic health record was performed. The etiology of severe hyponatremia was categorized as follows: hyponatremia due to a hemodynamic disturbance, including either hypovolemic hyponatremia due to volume depletion or hypervolemic hyponatremia due to cirrhosis or congestive heart failure; hyponatremia from syndrome of inappropriate antidiuretic hormone (SIADH), as defined by a lack of response to intravenous fluids, and elevated urinary sodium and osmolarity; hyponatremia due to endocrinopathy, confirmed by laboratory testing or brain magnetic resonance imaging demonstrating an increase in size/enhancement of the pituitary gland and/or infundibulum; hyponatremia associated with the patient's terminal decline leading to hospice enrollment or death and hyponatremia from other/unknown causes.

Statistical Analysis

Baseline characteristics were described using means and standard deviations (SDs) for continuous variables and counts and percentages for categorical variables. Univariable logistic regression was performed to evaluate for predictors of each electrolyte disorder. Multivariable logistic regression was performed using variables that were statistically significant with a P-value <0.10 in the univariable model. A stepwise selection model was performed to assess collinearity in the multivariable logistic regression models. All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

The Institutional Review Board at Partners Healthcare System approved this study and waived the need for informed consent.