GLOCORTICOID REPLACEMENT THERAPY DURING TRANSPHENOIDAL SURGERY: HOW MUCH IS IT REALLY NECESSARY?
Fernando Pimentel, Bernardo Liberman, Arthur Cukiert, Jayme Goldman, Luis Roberto Salgado
The use of glucocorticoid (GC) during and after surgery has been recommended in patients presenting with cortisol deficiency before surgery or in those with chronic hypercortisolism (p.e, cortical adrenal tumor) in whom surgery may cause abrupt onset of adrenal insufficiency. It has been a common clinical practice to use intravenous hydrocortisone (HC) during transesphenoidal surgery (TS) for Cushing’s disease (Tyrrell 78, Fitzgerald 82). GC administration should be based on the magnitude of the stress and the known GC production rate associated with it. However, there have been no studies addressing the GC needs during TS. The GC regimens commonly used during TS have been empirically defined or based on stress data obtained from other types of surgical procedures and vary from 200 to 400 mg of HC during the day of surgery (Williams, Arafah, Tyrrell 78, Fitzgerald 82). In the present study we assessed the need for GC replacement therapy during TS in patients with sellar tumors.
SUBJECTS AND METHODS
Thirty-two patients with sellar tumors were evaluated before and after TS. Sellar tumors were detected by MRI in all patients. Suggestive tumor tissue was visualized and resected by the neurosurgeon during surgery and histological examination and immunohystochemistry confirmed the diagnosis in all patients. Another 6 patients submitted to bilateral adrenalectomy for the treatment of Cushing’s disease were included in order to study the influences of exogenous GC administration on the rate of free cortisol urinary (UFC) excretion in patients without endogenous cortisol production. Informed consent was obtained from patients and the research protocol was approved by the Institutional Review Board.
In 12 patients (group I: age = 36.2 ( 2.2 years; 3 men and 9 women; BMI = 28.6 ( 0.7 kg/m2) no GC was administered during surgery or in the postoperative days. This group included 6 patients with non-secreting adenoma, 5 with acromegaly and one with prolactinoma. All patients had normal cortisol secretion before and after surgery as determined by a serum cortisol peak greater than 18 (g/dL during an insulin tolerance test (ITT) (0.1 – 0.3 U insulin / Kg IV).
Twelve patients (group II: age = 29.6 ( 3.3 years; 6 men and 6 women; BMI = 29.0 ( 1.6 kg/m2) received GC trans- and post-operatively. This group included 10 patients with Cushing’s disease, one with acromegaly and one with craniopharyngioma. They received 400 mg of intravenous HC during the day of surgery (PO-i) (100 mg before and at the end of surgery, and 100 mg b.i.d ) and 200 mg during the first post-operative day (PO-1) (50 mg q.i.d). Eight patients (group III: age = 36.4 ( 4,1 years; 2 men and 6 women; BMI = 26.2 ( 1.2 kg/m2) received a lower dosage of GC during the day of surgery and in the post-operative days. This group included 5 patients with Cushing’s disease, 1 with acromegaly and 2 with non-secreting adenoma. They received 200 mg of intravenous HC during the day of surgery (PO-i) (50 mg before and at the end of surgery, and 50 mg b.i.d) and 100mg of HC during PO-1 (50 mg b.i.d). Groups II and III patients without Cushing’s disease that received GC disclosed a low cortisol reserve during ITT (cortisol peak lower than 18 (g/dL) in a preoperative evaluation. Whole day urinary samples were collected for UFC determination during PO-i and PO-1 from all patients. Six patients (Group IV: age = 40.8 ( 3.0 years; 2 men and 4 women; BMI = 29.1 ( 4.8 kg/m2) that were submitted to bilateral adrenalectomy for treatment of Cushing’s disease received exogenous GC as follows: 400 mg/day of intravenous HC on the first day (PI-i) (100 mg q.i.d) and 200 mg/day on the second day (PI-1) (50 mg q.i.d)), thus resulting in a GC schedule similar to that of group II patients. UFC excretion rate was measured during PI-i and PI-1 in all patients.
Signs and symptoms of peripheral cortisol deprivation were evaluated by means of sequential recording of symptoms and measurements of body weight, body temperature, blood pressure and heart rate. Blood electrolytes, glucose, BUN, creatinine and white and red blood cell analysis were carried out in all patients before and after surgery.
UFC was measured by a radioimmunoassay (RIA) kit after extraction with dichloromethane (Diagnostic Products Corp., Los Angeles, CA). The normal range was 20 to 90 (g/24 hours and inter- and intra-assay variation’s coefficients were 6.9% and 6.0%, respectively.
Data were analyzed by Kruskal-Wallis one way analysis of variance (ANOVA) and the results were reported as mean ( SE. P-Values below 0.05 were considered statistically significant.
No statistical difference was observed in age and body mass index (BMI) between the studied groups.
UFC excretion decrease from PO-i / PI-i through PO-1 / PI-1 in all patients. UFC excretion during PO-i in group II patients (9350 ( 1631 (g/24h) was approximately 14 times greater than that seen in group I patients (679 ( 134 (g/24h) (p<0,05) and 5 times greater than that in group III subjects (1849 ( 298 (g/24h) (p<0,05). Mean UFC during PO-1 observed in group II patients (3233 ( 334 (g/24h) was also approximately 15 times greater than that of patients in group I (217 ( 37 (g/24h) and 4 times greater than that of group III subjects (755 ( 126 (g/24h). No statistical difference was observed between groups I and III. There was no difference between the mean UFC excretion in groups II and IV patients (9350 ( 1631 (g/24h and 10120 ( 1613 (g/24h, respectively) during PO-i / PI-i or PO-1 / PI-1 (3233 ( 334 (g/24h and 3891 ( 782 (g/24h, respectively).
Signs and symptoms of peripheral cortisol deprivation were not observed in any patient during PO-i or PO-1. Eight patients from Group II and 6 patients from group III showed clinical and laboratorial evidence of adrenocortical insufficiency by the third post-operative day (6 patients from group II and 3 from group III had Cushing’s disease). In group II, 2 patients with Cushing’s disease presented transitory diabetes insipidus. No other surgical complication was observed in this series.
A significant increase in blood pressure was observed in all patients during PO-i. This was mainly seen in the first two hours after the end of surgery, but the mean blood pressure measurements obtained in group I during PO-i were not different from those obtained in groups II and III. No difference was observed among groups in serum sodium, potassium and creatinine concentrations.
Adrenal insufficiency after TS is rare but often fatal if not diagnosed and treated (Harris 90 ). Patients who might be at risk to develop post-operative adrenal axis faillure are usually treated empirically with different GC schedules. Sub-physiologic replacement of GC may result in hemodynamic abnormalities such as drop in blood pressure, systemic vascular resistance, left ventricular stroke work index and absent tachycardic response (Udelsman 86 and ref. 9 – Mohler and 10 – Alford do Harris 90). GC excess is associated with well known immunosuppression and increased susceptibility to infection, electrolyte unbalance, reduced carbohydrate tolerance, hypertension and catabolic effects on muscle and wound healing (ref 7-13 do Udelsman, Baxter 90, Ref 9 do Lamberts97, Salem 94). We and others (Arafah, ) used a large dose of GC during and after TS until recently. Data from the present study demonstrated that a dose of 400 mg of HC during the peri-operative day yielded an UFC excretion rate considerably greater (14 times) than that observed in patients who underwent the same surgical stress but did not receive GC replacement. During PO-1 UFC excretion was also greater (15 times) in patients who received HC intravenously than in those who did not. When a lower dose of HC (200 mg/day) was administered during the day of surgery (group III) the mean UFC excretion remained greater (3 times) than that of patients who did not receive HC (Group I), although statistical difference was not achieved.
Several different parenteral GC replacement schedules have been published in order to study the cortisol secretion rate during or immediately after surgery (ver ref. do Kehlet – página 261). Some studies suggested that physiologic cortisol supplementation may be sufficient for patients submitted to major surgical procedures (ref. 14 a 16 do Udelsman 85, ref 30 e 31 do Czerwiec 96). Udelsman et al (1985) demonstrated that in adrenalectomized primates submitted to cholecystectomy the mortality rate in the sub-physiologically GC-treated group was higher than in the controls or in the group receiving full GC replacement. There was no difference between animals receiving full or supra-physiologic GC replacement.
Less than 1% of the excreted cortisol is eliminated as urinary free cortisol (Williams, ). In adults, the adrenal glands produce 75 to 150 mg of HC during the first 24 hours after major surgery (Kehlet 73, ref. 32, 39, 105, 57, 106 e 108 do Salem 94, ver ref doLamberts 97 e outras) and the UFC excretion rate should be approximatly 750 to 1,500 (g / day during this period (about 10 times the normal values in non-stressful situations). Our data are in accordance with this estimate: group I patients mean UFC excretion was 679 (g/24h (range: 185 to 1750 (g/24h) during PO-i. Theorectically, after administration of 200-500 mg of HC, UFC excretion should range from 2,000 to 5,000 (g in 24 hours (1% of the administered HC). Patients in the present study who received 200 mg during PO-i Group III) showed UFC excretion rates which corresponded to that value. However, in subjects receiving 400 mg of HC this percentage was approximately 2%. This increased cortisol excretion rate might be due to the fact that urinary excretion of free cortisol increases when serum cortisol concentration exceeds cortisol binding globulin (CBG) capacity (25g/dL) (Williams, Greenspan 321; Bright 95). This occurs more frequently after bolus injections. Actually, we also observed a great inter-individual variability in UFC excretion as noted previously by others. This may be due to differences in CBG concentration, sex, age and BMI (Bright 95, Barton 93; Lamb 94). We did not observe statistical differences in age and BMI in this series.
Our data strongly suggest that intravenous HC doses of 200 mg during PO-i and 100 mg during PO-1 might be enough to prevent adrenal insufficiency in patients at risk after TS. Kehlet et al. ( ) suggested injecting 25 mg intravenous HC every 4 hours. They also suggested a schedule for continuous infusion consisting of a 100 mg / day regimen until gastrointestinal function allows oral intake of usual GC replacement therapy. Although Czerwiec et al ( 1996 ) had recommended an intravenous HC dosage of 300-400 mg/day for patients undergoing major stress, others (Lamberts 97 e Salem 94) suggested a 25mg/day regimen for minor stress, 50-75 mg/day for moderate stress and 100-150 mg/day for major stress.
We and others (JCEM 97) have observed that ACTH secretion in patients with Cushing’s disease may persist during some post-operative days after total tumor resection, and that prophylactic GC administration during this period might not be necessary for these patients. Thorner et al. (abstr. 97) demonstrated that serum cortisol was still elevated 6 hours after surgery in a great number of patients (12 of 17 patients) who had achieved remission after transsphenoidal surgery for Cushing’s disease and did not receive GC during the procedure. A rapid decrease in blood cortisol levels by 24 to 48 hours after surgery was observed in the same patients studied by the same authors. This endogenous production of cortisol could raise the mean UFC excretion in patients with Cushing’s disease receiving exogenous GC as in group II and III. We have observed no difference between mean UFC excretion during PO-i / PI-i in groups II and IV. These findings suggest that endogenous production of cortisol might play only a small role in determining the rates of post-operative UFC excretion. Moreover, we observed that UFC values from patients with ACTH-secreting pituitary tumors or other sellar tumors in groups II and III were similar during PO-i and PO-1 (Figure 1).
The side effects of high-dose GC administration were also analysed in this study. However, only a few of these effects could be analyzed because there were important differences in the clinical spectrum of the groups. The hyperglycemic effect of GC has been known for a long time (ref 8 and 9 do Wayne 77; ref. Kitabchi do Bowes 91, Malerbi 88, ref – Al-Shoumer do Peacey 97), but we were unable to compare fasting glucose values in group I with groups II and III since there were much more diabetic patients in the latter groups than in the former . The hypertensive effect of GC could not be evaluated because there was a marked increase in blood pressure during PO-i (more significant in the first 2 hours after surgery) in all patients. This immediate post-operative hypertensive effect has been previously described in other surgical procedures and is related to surgical stress (Udelsman 87). No electrolyte unbalance, gastric ulceration or psychiatric disorder was observed in any patient.
Many of the published GC regimens probably represent supra-physiologic steroid replacement therapy. Data obtained in this study strongly suggest that group III regimen (200 mg / day during PO-i and 100 mg / day during PO-1) is probably the best one for patients being submitted to TS for the treatment of sellar tumors.