Transsphenoidal surgery for malignant pituitary lesions: an analysis of inpatient complications.

BACKGROUND

Fewer than 4% of pituitary tumors are malignant lesions. These tumors predominantly represent metastatic disease from elsewhere. This study evaluates inpatient complications, demographics, and hospitalization characteristics of patients who underwent transsphenoidal surgery (TSS) for malignant pituitary lesions.

METHODS

The Nationwide Inpatient Sample was evaluated for TSS patients from 1998 to 2010. Demographics, hospitalization characteristics, and complications were evaluated among patients with malignant lesions and compared to those with benign tumors.

RESULTS

There were 17,425 inpatient records, 1.0% of which involved malignant pituitary tumors. There was no difference in age between these cohorts (p = 0.378). Patients with malignant tumors had greater length of stay (6.7 days vs 4.5 days, p = 0.003) and higher trending charges ($55,371 vs $40,550 p = 0.091). The most common postoperative complications among patients with malignant lesions included diabetes insipidus (DI) (17.9%), fluid/electrolyte abnormalities (14.0%), neurological complications (5.6%), cerebrospinal fluid (CSF) rhinorrhea (2.2%), and iatrogenic pituitary disorders (2.2%). Patients with malignant lesions had a significantly greater rates of postoperative DI and fluid/electrolyte abnormalities (odds ratio = 2.0 and 1.7, respectively), whereas no statistical difference was noted in the rates of CSF rhinorrhea (p = 0.372).

CONCLUSION

In this analysis of inpatient hospitalizations for TSS patients, malignant pituitary disease was associated with a greater rate of postoperative DI and fluid/electrolyte abnormalities, but no differences in the rates of postoperative CSF rhinorrhea and other complications were found. Patients with malignant pituitary lesions undergoing TSS had significantly longer hospitalizations and higher trending charges than those with benign lesions. This analysis is, however, subject to the limitations of the database.