Baseline characteristics and risk factors for hyponatraemia in all patients with serum sodium level <135 mmol/L
Among 670 patients who underwent transsphenoidal surgery over 4 years, 162 patients (162/670; 24.2%) developed hyponatraemia (serum sodium <135 mmol/L) postoperatively. Mild hyponatraemia occurred in 90 patients (90/670; 13.4%), 38 patients had moderate hyponatraemia (38/670; 5.7%), and severe hyponatraemia was recorded in 34 patients (34/670; 5.1%).
Patients’ characteristics, tumour histology and surgical approach are shown in Table 1. The median age for hyponatraemic patients was 56 years (IQR 44-68). Younger age was associated with hyponatraemia (p = 0.043). On average, the value of nadir sodium increased by 0.05 mmol/L for each year increase of age (F(1,159) = 4.18, p = .043, R2 = 0.03) (Figure 1).
There was no difference in the incidence of hyponatraemia observed between males and females (p = 0.4). However, female gender was significantly associated with lower nadir level (mean sodium = 124.8 mmol/L) than with male gender (mean sodium = 133 mmol/L) (p < 0.001).
There was no significant difference in the frequency of hyponatraemia between patients with microadenoma and those with macroadenoma (p = 0.6) (Table 1). The frequency of hyponatraemia was similar in patients with pituitary macroadenoma with and without optic nerve (p = 0.7). In addition, there was no difference in the frequency of hyponatraemia between patients with functioning and non-functioning adenomas (p = 0.5). With respect to the neurosurgical technique, the occurrence rate of hyponatraemia was similar between microscopic and endoscopic transsphenoidal surgeries (p = 0.4).
Incidence, time course, and severity of hyponatraemia in all patients with sodium <135 mmol/L
Prior to surgery, ten patients (10/156; 6.4%) had pre-existing hyponatraemia with a level of 132.2 (± 3.5) mmol/L. The median time for serum sodium to decrease below 135 mmol/L was 4 days (IQR 1-6) after tumour resection. The mean (± SD) nadir sodium level for all patients was 128.6 (± 5.2) mmol/L and the median timepoint to exhibit nadir sodium was postoperative day 7 (IQR 2-8). Nadir sodium according to hyponatraemia severity is showed in table 2. Nadir sodium varied according to the aetiology. In SIADH, nadir level was 125.1 (± 5) mmol/L with a median time of onset of 8 days (IQR 6-9 days). In patients with adrenal insufficiency, lowest sodium concentration was 130 (± 2.4) mmol/L, occurring after a median time period of one day (IQR 1-5). In patients with DDAVP overreplacement, nadir sodium was 127 (± 4.4) mmol/L with a median duration of 7 days (IQR 3-9) postoperatively.
We examined the trend of hyponatraemia according to the severity during the first postoperative week. Mild hyponatraemia was more common in the early postoperative period, mostly on day 1 (p = 0.0001), while severe hyponatraemia started from day two and evolved in a delayed pattern most commonly on day 7 (p = 0.0001) (Supplementary Figure). Mean serum sodium during the first seven days following transsphenoidal surgery according to hyponatraemia severity is shown in Figure 2.
Investigations and aetiology for patients with serum sodium ≤132 mmol/L
One hundred and eleven patients had serum sodium levels of 132 mmol/L and below. The commonest cause was SIADH (80/111; 72%), followed by adrenal insufficiency (9/111; 8.1%), overzealous treatment of diabetes insipidus with DDAVP (6/111; 5.4%), and hypotonic hyponatraemia due to hypotonic fluid administration (3/111; 2.7%). There were no documented cases of CSWS. No clear diagnosis was documented in 13 patients (13/111; 11.7%).
All patients had early morning cortisol checked 48 hours postoperatively. In patients with a new onset of secondary adrenal insufficiency, mean morning cortisol was 138.7 (± 84) nmol/L. Thyroid status was assessed in all patients pre-operatively and postoperatively with mean free T4 level of 18 (± 5) pmol/L (reference range 12-22 pmol/L). Notably, 11 patients had low free T4 level prior to surgery which was treated appropriately with levothyroxine replacement.
Treatment of hyponatraemia in patients with serum sodium ≤132 mmol/L
Fluid restriction was imposed on 84 patients of those with serum sodium ≤132 mmol/L; this includes 80 patients with SIADH, 3 patients with fluid overload and 1 patient on desmopressin therapy.
For patients with SIADH, 72 (72/80; 87.8%) had fluid restriction as a monotherapy. Of those, 60 patients had fluid restriction between 500 and 1000 ml daily, 8 patients reduced fluid intake to 1500 ml daily and 4 patients were asked to drink to thirst.
Fluid restriction as a monotherapy achieved a mean increase in serum sodium of 3.3 mmol/L over a 3-day period. Notably, 23 patients (23/72; 31.9%) did not achieve any increase in sodium levels during the first three days of fluid restriction, four patients (4/72; 5.5 %) had a mean sodium increase of 1-2 mmol/L, 8 patients (8/72; 11.1%) had 3-4 mmol/L sodium increase and 37 patients (37/72; 51.3%) had sodium increment of ≥5 mmol/L during the first three days of fluid restriction. The median time to achieve an increase in serum sodium of ≥5 mmol/L was 3 days (IQR 2-6 days) and to achieve normal sodium concentration was 4 days (IQR 2-6). Figure 3. demonstrates sodium levels at baseline and after starting fluid restriction in those received fluid restriction only.
Second-line treatment for SIADH was administered in a total of 8 patients; with hypertonic saline 1.8% being used in 5 patients (5/84; 5.9%), hypertonic saline 2.7% in 1 patient (1/84; 1.1%), sodiumchloride tablets in 1 patient (1/84; 1.1%), and tolvaptan at a dose of 7.5mg in one patient (1/84; 1.1%). The patient who was treated with 2.7% hypertonic saline and the second one who received tolvaptan developed sodium overly rapid correction by more than 10 mmol/L over the first 24 hours.
Patients diagnosed with new onset of secondary adrenal insufficiency (number = 9) received glucocorticoid replacement and achieved a mean sodium increase of 8 mmol/L in the first 3 days post therapy. Patients with DDAVP over replacement (number = 6) were treated by dose down titration, leading to a mean sodium increase of 8 mmol/L during the first 3 days of therapy. No patients were treated with urea or demeclocycline.
Outcome for all patients with hyponatraemia <135 mmol/L
The mean serum sodium level on discharge for the full cohort was 137.3 (± 4.2) mmol/L. For the full cohort (number = 162), hyponatraemia was corrected in 4 days (IQR 2-6 days) and the length of hospital admission was longer for patients with hyponatraemia (median = 8 days [IQR 5-14]) than the patients who remained normonatraemic throughout hospitalisation (median = 5 days [IQR 4-7]) (p < 0.001). Lower nadir sodium was significantly associated with longer inpatient stay. On average, one mmol/L reduction in sodium concentration resulted in an increase of inpatient stay by 0.39 days (F(1,154) = 14.39, p < .001, R2 = 0.09) (Figure 4).
Seven patients out of 162 (7/162; 4.3%) were admitted to the intensive care unit with a median stay of 3 days (IQR 2-4). One patient had a seizure secondary to severe hyponatraemia. No patients developed permanent neurological morbidities and there was no associated mortality.
Among 162 patients who developed hyponatraemia, a subset of eleven patients (11/162; 6.7%) had normal sodium levels during initial hospitalisation; however, they were readmitted post discharge with hyponatraemia during the first 30 days of surgery. Those patients were discharged after a median length of initial hospital stay of 4 days (IQR 4-6). Six patients (6/11; 54.5%) were men and five (5/11; 45.5%). The median time of readmission from surgery was 9 days (IQR 7-10). All patients had severe hyponatraemia due to SIADH with a mean sodium level of 120.1 (± 4.4) mmol/L. Among the 11 patients who were readmitted, 8 patients (8/11; 72.7%) were treated with fluid restriction, two (2/11; 18.1%) were treated with both fluid restriction and 1.8% hypertonic saline in the intensive care unit, and one (1/11; 9%) patient was treated with a combination of fluid restriction and 2.7% hypertonic saline. The median duration to regain normal sodium levels was 4 days (IQR 3-6) and median duration to hospital discharge was 5.5 days (IQR 3-10).