8117 Late-night Salivary Cortisol in the Diagnosis of Cushing's Disease Recurrence

Abstract Disclosure: A.E. Alexandre: None. P.C. Elias: None. M. De Castro: None. L.M. Mermejo: None. D.C. Aragon: None. A.C. Moreira: None. Cushing’s Disease (CD) remission following transsphenoidal surgery (TSS) is most often defined by low morning cortisol levels and the requirement of glucocorticoid (GC) replacement therapy. However, CD recurrence is not a well studied event. There is no standard definition of remission criteria and no accurate cutoff for biochemical tests on CD recurrence. Late-night salivary cortisol (LNSC) has been shown to have higher accuracy in confirming CD recurrence than urinary free cortisol and dexamethasone suppression test (DST). However, few studies addressed LNSC performance following remission after TSS. The study aimed to establish cutoff values of LNSC with higher sensitivity (S) and specificity (E) in CD recurrence diagnosis.In a retrospective cohort study of 65 patients with remitted CD after TSS, remission was defined as improvement of sign and symptoms related to CD and a plasma cortisol <5mcg/dL in 30 days following TSS. Eleven patients progressed to permanent adrenal insufficiency and were excluded from the study. All 54 patients had at least four LNSC sampling and a minimal follow-up of 24 months after TSS. LNSC was determined by RIA. A cutoff value of 270 ng/dl for LNSC is used for the diagnosis of CD. Patients were divided in three different groups: Overt Recurrence (OR): elevated LNSC and worsening of CD`s signs and symptoms. Remission (RM): normal LNSC and no clinical signs of CD recurrence. Biochemical Recurrence (BR): at least 2 elevated biochemical parameters (LNSC or DST) but no worsening of signs and symptoms associated with CD. The mean of the 2 LNSC last values when OR was diagnosed and the mean of 2 LNSC last values of follow up for RM patients were used to estimate the accuracy of LNSC for diagnosing recurrence using a ROC curve (AUC) and its 95% confidence interval. Of the 54 patients included, 46 females; 8 males; mean age at diagnosis: 32 years, (5 – 58 yrs); mean time of follow-up: 12.8 years (4 – 31.7 yrs). Thirteen were classified as OR, 18 as RM and 23 as BR; and mean time of GC use after TSS was: 8.7; 34 and 17 yrs, respectively, significantly lower in OR compared to RM (p:0.046). A LNSC cutoff value of 272 mcg/dL had sensitivity of 85% and specificity of 95% (AUC: 0.97; 95% IC = [0.93; 1,00]) for CD recurrence diagnosis. BR patients presented LSNC fluctuation along follow-up.LNSC is a useful tool for CD recurrence screening. Setting LSNC specific assay cutoff values might be needed. LSNC cutoff values were similar for the initial CD diagnosis, as well as for CD recurrence diagnosis. Shorter time of GC replacement may be an important predictor of CD recurrence. Patients with BR should be followed more carefully as they may be at higher risk of recurrence. Multiple LNSC sampling on successive days can be easily performed and may detect subtle changes in cortisol rhythm. Presentation: 6/1/2024