Postoperative complications following transsphenoidal pituitary surgery are generally associated with sodium disturbance, such as DI and delayed hyponatremia [3–12, 27]. However, the precise mechanism of the latter has not been established.
The present study was conducted to determine factors related to delayed hyponatremia following endonasal endoscopic TSS in patients with a nonfunctioning pituitary adenoma. Of 137 patients who underwent that surgical procedure, 31 (22.6%) developed delayed hyponatremia, with equal frequency in females and males, though female gender has been reported to be a risk factor [10]. There were no differences between the delayed hyponatremia and normal natremia groups in the present study in regard to age, BMI, smoking habit, or rate of recurrence.
The associations of comorbidities with occurrence of delayed hyponatremia were also examined. Although a previous report stated that patients with a preexisting renal disorder had a higher likelihood of developing delayed hyponatremia as compared to those without [28], multivariate analysis conducted in the present study revealed that a preexisting renal disorder did not have a significant association, though interestingly, patients with preexisting hypertension had a statistically higher likelihood of avoiding delayed hyponatremia. That latter result suggests that patients with hypertension might have greater renin-angiotensin-aldosterone system hormone levels, resulting in sodium retentive actions [29] and possible influence on sustainable natremia.
Although there was no significant correlation shown by multivariate analysis, univariate analysis results revealed that the anterior-posterior distance of the tumor was correlated with delayed hyponatremia (p = 0.025), suggesting a weak association with compression of the pituitary gland posterior lobe, which might cause inappropriate release of arginine vasopressin (AVP) following surgery. Previous studies have also noted that patients with a large tumor frequently developed delayed hyponatremia as compared to those without [3, 4, 7, 10, 11] and speculated that a larger size tumor is associated with increased attenuation of the functional capacity of pituitary cells during development of a pituitary tumor. It has also been noted that pituitary function recovery in the delayed phase after surgery might cause a surge of AVP release and fluid retention [4].
In the present study, the influence of surgical factors on occurrence of delayed hyponatremia in the present cohort received focus. Based on multivariate analysis results, direct surgical factors, such as operative duration, extent of resection, intraoperative bleeding volume, and intraoperative CSF leakage, had no association with its occurrence. In contrast, the presence of a firm tumor, transient DI, and meningitis, indirect surgical factors, each had a significant relationship with delayed hyponatremia occurrence. These results suggested that a greater level of surgical manipulation of the pituitary stalk could lead to transient DI, while meningitis may be associated with a rebound effect of compensatory treatment for dehydration.
Overall, transient DI appears to be the most common complication after TSS in these cases, with symptomatic hyponatremia the second most common [17, 30]. A previous retrospective study found delayed hyponatremia to be the most common cause for readmission, followed by DI [13]. Other reports have noted occurrence of early transient DI after TSS ranging from 10–60% [2, 31, 32]. In the present study, transient DI occurred in 35.5% of the delayed hyponatremia cases and in 8.5% of cases with normal natremia, while multivariate analysis showed that transient DI was significantly correlated with delayed hyponatremia (OR 6.21, p = 0.001) as was meningitis (OR 12.03, p = 0.006). These results were compatible with previous reports noting that hyponatremia following DI was caused by SIADH due to unregulated released of AVP from denervated posterior pituitary nerve terminals [29, 33].
Postoperative hyperintensity of the pituitary posterior lobe shown by sagittal T1WI was also evaluated to examine the influence of existence of the posterior lobe following surgery on occurrence of delayed hyponatremia. There was no significant difference between the groups regarding hypersignal percentage (p = 0.097). Moreover, measurements of postoperative AVP (Table 4) also indicated no significant differences for AVP values (p = 0.440) and no excess production of AVP, such as SIADH. Thus, prediction of individual patient risk for DI or SIADH remains difficult [34].
As for postoperative blood examination evaluations, multivariate analysis revealed that chloride was significantly lower in the hyponatremia as compared to the normal natremia group (p < 0.001), suggesting that chloride concentration in the renal tubules is tightly coupled with sodium and water transport [29]. Misono proposed that chloride-mediated feedback could play a role in ANP-induced natriuresis in patients with a high level of circulating ANP [35].
We also evaluated whether the degree of hyponatremia was correlated to appearance of symptoms and the results showed that patients with severe hyponatremia presented symptoms with significant frequency (OR = 52, p = 0.002). Such patients should be treated as soon as possible to avoid deterioration and several reports regarding treatment of delayed hyponatremia to prevent severe symptoms have been presented [25, 36–38], with the latter by Deaver et al. noting that mild fluid restriction (to 1.5 liters daily) was an effective approach for preventing readmission for hyponatremia after TSS for a pituitary adenoma. In the present cohort, restrictions of fluid (to 1.0 liters daily) and oral intake of salt (3–6 g/day) were used for patients with delayed hyponatremia to improve symptoms. Finally, though treatment with the oral vasopressin receptor antagonist tolvaptan can be useful for cases with SIADH [9], that drug was not available in Japan until 2020.
This study has some limitations, including inherent bias related to a retrospective review of cases. Also, this was a single-center study, potentially introducing bias to the results. All of the present patients underwent transsphenoidal pituitary surgery with an endoscopic approach, thus, though direct surgical factors did not have an influence on delayed hyponatremia, factors with effects identified in this study were not free from confounding effects caused by differences in surgical techniques. All surgeries analyzed in this study were performed by a single neurosurgeon. Because clinical outcomes after transsphenoidal pituitary surgery can be easily influenced by factors related to the performing physician [39], differences in operator experience and techniques may have had effects on the present results.