Introduction

Hyponatremia is the most common electrolyte disorder in hospitalized patients, occurring in 5–30 % of patients [1, 2].

It is commonly defined as a serum sodium level of 135 mEq/l and it can be classified in three different levels: mild (130–134 mEq/l), moderate (125–129 mEq/l), and severe (<125 mEq/l) [3–6].

A variety of risk factors have been reported for hospital-acquired hyponatremia, including older age, diabetes mellitus, chronic kidney disease, surgery, pulmonary infection, diuretic therapy, administration of antibiotics or opioids, and use of hypotonic intravenous fluids [7, 8].

Hyponatremia is associated with a poor outcome in several medical conditions, such as liver cirrhosis, congestive heart failure, and infectious diseases [9–11].

It is also associated with serious complications that have been linked to increased morbidity and mortality [12–14].

Although some of the adverse outcomes associated with hyponatremia can be related by the underlying medical illness, it should be noticed that hyponatremia itself could be associated with adverse outcomes independently of the comorbidities [15].

Gill et al. reviewed that about 14 % of hyponatremia in inpatients was caused by underlying tumor-related conditions [16].

The incidence and prevalence of hyponatremia in cancer patients vary greatly, depending on the cancer type, clinical setting, and serum sodium cutoff point.

The main cause of hyponatremia in cancer patients is the syndrome of inappropriate antidiuretic hormone secretion (SIADH), which is due to ectopic production of antidiuretic hormone (AVP) by extracellular fluid depletion or to renal toxicity induced by chemotherapy, especially platinum-based [17, 18].

SIADH is most commonly found in patients with small cell lung cancer (SCLC; 11–15 %), but it has also been reported in 3 % of patients with head and neck cancer and in other solid tumors and hematological malignancies [19–21].

It is debatable whether the occurrence of hyponatremia may correlate with the stage and the anatomical spread [22–25].

Recently, Hansen and colleagues suggested that low sodium serum values could be associated with the number of metastatic sites and could be a measure of the tumor burden [26].

Recent studies suggested that hyponatremia might be an indicator of poor prognosis, and we previously demonstrated its impact on prognosis in mesothelioma patients [27].

Furthermore, it might have negative effects on hospitalization length and quality of life in non-malignant as well as in malignant diseases. The degree of hyponatremia and its normalization could modify the outcome of hyponatremic cancer patients [17, 28].

The purpose of this study was to evaluate the impact of hyponatremia on the length of stay, with a pharmaco-economical evaluation, and on the prognosis in cancer patients.

Patients and methods

Study population

The study population included all consecutive cancer patients aged 18 years or older who were hospitalized at our Institution from June 2013 to December 2013. We collected data regarding age, sex, staging, histology, chemotherapy, and levels of serum sodium at admission and at discharge.

Statistical analysis

Clinical data were retrospectively collected from medical chart reviews and electronic records. Overall survival (OS) was defined as the interval between the date of hospital admission to last follow-up or death. Survival distribution was estimated by the Kaplan-Meyer method, and differences in probability of surviving were evaluated by log-rank test.

A Cox regression model was applied to the data with a univariate approach and used to assess the prognostic role of the variables. All significance levels were set at a 0.05 value.

Chi-square test and Fisher’s exact probability test were used to evaluate association between categorical variables. Statistical analysis was performed with MedCalc software version 10.4.8 for Windows.

The cost of hospitalization was calculated by summing the direct medical costs, indirect costs, and common business folded [29], supplied by the administrative office of our institution. Quantification considered the costs of the year 2012 (2013 is not already available).

Results

One hundred five patients were included in the analysis. Male to female ratio was 57:48, and median age was 65.8 years (range 26–83 years).

The majority of patients had lung adenocarcinoma (28.6 %), 34 cases (32.4 %) had gastrointestinal cancer, and ten patients (9.5 %) had a diagnosis of sarcoma. The remaining patients examined were affected, in smaller percentages, by other solid tumors. In almost all cases, stage of disease was advanced or metastatic. Table 1 summarizes patients’ characteristics.

Table 1 Patients characteristics
Full size table

A total of 38 % of patients presented at least one episode of mild or severe hyponatremia during hospitalization. Twenty patients (19 %) showed mild or severe hyponatremia at admission, while 20 patients (19 %) developed mild or severe hyponatremia during hospitalization (Table 2).

Table 2 Effect of hyponatremia on the length of hospitalization
Full size table

Median OS from the date of hospitalization was 50 days (range 3–282 days). A statistically significant difference in OS from the date of admission has been shown between eunatremic and hyponatremic patients (p = 0.0255) (Fig. 1). Furthermore, the presence of metastases was related to a worsened median OS (p = 0.0418).

Fig. 1
figure 1

Overall survival in patients divided according to the serum sodium level at the admission (calculated in days from the first day of admission)

Full size image

A statistically significant correlation was found between the length of stay and the detection of hyponatremia (serum sodium <135 mEq/l) both at admission (p = 0.0009) (Fig. 2) and during hospitalization (p = 0.0001) (Fig. 3).

Fig. 2
figure 2

Length of hospitalization according to the value of sodium at admission

Full size image
Fig. 3
figure 3

Length of hospitalization in days based on minum serum sodium level during the hospitalization

Full size image

At multivariate analysis, including sodium level at admission, during hospitalization and at discharge, the length of stay and the presence of metastases, only hyponatremia at admission (hazard ratio (HR) = 2.64, 95 % confidence intervals = 1.72–4.35, p < 0.001), the severity of hyponatremia (HR = 1.35, 95 % confidence intervals (CI) = 1.04–1.88, p = 0.22), and the stage of disease (HR = 1.32, 95 % CI = 1.04–1.76, p = 0.23) resulted as independent factors.

We also conducted a drug economic analysis in order to understand the cost of the prolongation of hospitalization in patients with severe and moderate hyponatremia compared to patients with serum sodium level in normal range. A relevant problem of economic analysis of health interventions is the correct quantification of the costs [30].

The cost of hospitalization in our institute amounts to 381€/day per patient. The cost of hospitalization for a cancer patient, that was calculated to be 7 days on the average (corresponding to the national median length), amounted to €2,672. The cost for the days of hospitalization for a patient with moderate (131–135 mEq/l) and severe (<130 mEq/l) hyponatremia, calculated on the median of days of hospitalization (9 and 21, respectively), amounted to €3,435 in the first case and €8,015 in the second. Comparing the described costs, a patient with moderate or severe hyponatremia costs in rate terms 128 and 299 % more than a normonatremic patient, respectively. Comparing these data with the cost of hospitalization days for patients without hyponatremia in our sample, considering a median of 4 days (€1,527), the difference is 224 and 524 % more, respectively, in moderate and severe hyponatremic patient.

Discussion

Hyponatremia is recognized as the most common electrolyte disorder which occurs in hospitalized patients, and it is reported to be associated with higher morbidity and mortality [1, 2, 31].

Mild hyponatremia, which was thought to be relatively asymptomatic, also showed an excess of mortality compared with patients with a normal serum sodium level [15].

Serum sodium is a readily available, easily obtained, and routinely measured plasma electrolyte; however; hyponatremia is often under-diagnosed and untreated.

About 14 % of hyponatremia in inpatients is due to underlying tumor-related conditions [16]; therefore, it should be suspected and screened in all the patients who are diagnosed and are being treated for malignancies.

The systemic manifestations of many types of tumors and the toxicities of cancer therapy are involved in the pathogenesis of hyponatremia in cancer patients.

Although only few studies have evaluated the impact of hyponatremia on cancer patients outcome, literature data suggest that hyponatremia can be considered a negative prognostic factor in this setting of patients [17, 32].

Low serum sodium levels have recently been associated with poor OS in hepatocellular carcinoma [33], gastric cancer [34], mesothelioma [27], and small cell lung cancer [35].

In localized renal cell carcinoma, serum sodium level below median values has recently been associated with poor disease-free survival and OS after nephrectomy [36].

Cancer-related hyponatremia has been also hypothesized adversely affect the response to treatment [37, 38].

In a Belgian prospective study on hyponatremia in cancer patients, higher death rates were observed in hyponatremic patients compared with the entire examined cancer population. The authors reported that no death was apparently attributable to the electrolyte disorder itself but they suggested that hyponatremia might be a sign of general debility or of the gravity of underlying disease that increase the risk of death [17].

According to literature data, in our study, we observed a statistically significant difference in OS from the date of admission (p = 0.0255) between eunatremic and hyponatremic patients.

Furthermore, the presence of hyponatremia at admission, the severity of hyponatremia, and the stage resulted as independent prognostic factors (p = 0.0088).

Several studies analyzed the prognostic role of hyponatremia at hospital admission [31, 39].

Gill et al. recently found that severe hyponatremia was associated with a significantly higher mortality (27 vs. 9 %, p = 0.009) and a longer duration of hospitalization (16 vs. 13 days, p = 0.005) [16].

Our results showed a statistically significant correlation between the length of stay and the detection of hyponatremia (serum sodium <135 mEq/l) both at admission (p = 0.0009) and during hospitalization (p = 0.0001).

Our findings are consistent with those reported by Doshi and colleagues who, analyzing 3357 hospitalized cancer patients with cancer, showed a higher frequency of hyponatremia in this setting and a strong and independent association between mild and moderate hyponatremia and longer hospital stay (on average 4 to 8 additional days) with an increased 90-day mortality [39].

In a recent retrospective study conducted in a tertiary cancer center on 295 patients who underwent inpatient cancer rehabilitation, Nelson et al. found that a high percentage of patients have hyponatremia and it was associated with a prolonged rehabilitation length of stay (11 and 15 days for patients with mild and severe hyponatremia, respectively, vs. 10 days of eunatremic patients) [40].

There is a growing body of evidence that acting effectively and timely on the normalization of sodium levels could have a positive effect on prognosis of cancer patients.

The rate of serum sodium correction depends on the etiology, pathophysiology, acuity, severity, and clinical presentation of hyponatremia.

Hansen et al. in a retrospective study observed that SCLC patients who did not fully normalized serum sodium had a worse prognosis than hyponatremic patient who did. Moreover, they reported that also the baseline serum sodium levels were prognostic for outcome [26].

In a recent study, Petereit et al. reported that the median survival in lung cancer with hyponatremia was less than that in normonatremic patients and the correction of serum sodium above the level of 138 mEq/l was correlated with an improved survival [28].

In conclusions, our results seem to indicate that the occurrence of hyponatremia at the admission or during the hospitalization might represent a significant factor influencing the outcome. Moderate to severe hyponatremia in cancer patients delays cancer treatments and hospital discharge with a deep impact on the hospital bills.