Weighing the prognostic role of hyponatremia in hospitalized patients with metastatic solid tumors: the HYPNOSIS study

Abstract

Previous works linked low sodium concentration with mortality risk in cancer. We aimed at weighing the prognostic impact of hyponatremia in all consecutive patients with metastatic solid tumors admitted in a two-years period at our medical oncology department. Patients were included in two cohorts based on serum sodium concentration on admission. A total of 1025 patients were included, of whom 279 (27.2%) were found to be hyponatremic. The highest prevalence of hyponatremia was observed in biliary tract (51%), prostate (45%) and small-cell lung cancer (38.9%). With a median follow-up of 26.9 months, median OS was 2 months and 13.2 months for the hyponatremia versus control cohort, respectively (HR, 2.65; P < 0.001). In the multivariable model, hyponatremia was independently associated with poorer OS (HR, 1.66; P < 0.001). According to the multivariable model, a nomogram system was developed and validated in an external set of patients. We weighed over time the influence of hyponatremia on survival of patients with metastatic solid tumors and pointed out the possibility to exploit serum sodium assessment to design integrated prognostic tools. Our study also highlights the need for a deeper characterization of the biological role of extracellular sodium levels in tumor development and progression.

Figure 1

Kaplan-Meier curves for overall survival. Red lines indicate patients in the control cohort while green lines indicate patients in the hyponatremia cohort. Thick lines indicate raw survival curves while thin lines indicate adjusted survival curves. Patients in the control cohort had higher overall survival compared to patients in the hyponatremia cohort.

Figure 2

Kaplan-Meier curves for overall survival regarding the four most represented histologies. Red lines indicate patients in the control cohort while green lines indicate patients in the hyponatremia cohort. Thick lines indicate raw survival curves while thin lines indicate adjusted survival curves. Patients in the control cohort had higher overall survival compared to patients in the hyponatremia cohort in colorectal cancer (CRC, panel A), non-small cell lung cancer (NSCLC, panel B) and breast cancer (BC, panel C). In gastroesophageal cancer (GEC, panel D), adjusted model indicates that patients in the hyponatremia cohort had a better prognosis than patients in the control cohort.

Figure 3

Nomograms predicting the survival probability within 3 months (panel A) and 36 months (panel B) after the date of hospitalization. Abbreviations. BC: breast cancer, BTC: biliary tract cancer, CRC: colorectal cancer, GEC: gastroesophageal cancer, mos: months, NET: neuroendocrine tumor, NSCLC: non-small cell lung cancer, PaC: pancreatic cancer, PC: prostate cancer, RCC: renal cell carcinoma, SCLC: small-cell lung cancer, tx: therapy.