Imaging, Pathological, and Clinical Characteristics of Mixed GH/PRL-Producing Pituitary Adenomas in Patients with Acromegaly: A Large Single Center Experience.
Presentation Number: SUN 438
Date of Presentation: April 2nd, 2017
Elena Varlamov*1, Fabienne Langlois2, Joao Prola1, Jung Kim1, Dawn Shao Ting Lim2, Christine G Yedinak1, Justin S Cetas1 and Maria Fleseriu1
1Oregon Health & Science University, Portland, OR, 2Oregon Health & Science University, OR
Background: Unique characteristics of mixed GH/PRL adenomas (MA) have not been extensively studied. Sparsely granulated GH adenomas (SGA) have been recently associated with T2 MRI hyperintensity and more aggressive behavior than densely granulated GH adenomas (DGA) that are hypo- or isointense and have better response to 1st generation somatostatin receptor ligands (1stgen-SRLs).
Objective:To assess T2 MRI, pathological and clinical characteristics of MA in patients (pts) with acromegaly (A), and compare data with DGA and SGA.
Methods:Retrospective IRB-approved data analysis of A pts who had surgery (sx) at OHSU (2006-2016); pts with prior sx, radiation and any preoperative medication for A were excluded. Pathology and imaging were analyzed by dedicated neuropathologist and neuroradiologist. Pathology: DGA, SGA or MA. MA included GH-PRL staining and mammosomatotroph tumors. Silent GH –PRL were excluded. T2 MRI intensity was defined as hypo- (<white matter), hyper- (>grey matter), and isointense (white <signal <grey). Remission was defined as normal IGF-1 3-6 mo after sx; SRL response; normal IGF-1 at 12 mo. Stats: PASW18, independent T test, ANOVA, Bonferroni & Tamhane post hoc analysis.
Results:Analysis of 56 pts: 18 DGA (32%), 13 SGA (23%), and 25 MA (45%). Age and sex were similar for MA, SGA, and DGA, however, post hoc analysis revealed SGA >DGA in females (p=0.05). Tumor size, cavernous sinus invasion, and baseline IGF-1 were similar in all groups. Mean baseline PRL level was higher in MA than DGA or SGA (27.2, 13.9 and 11.2ng/mL; p=0.016). Despite PRL staining, only 28% of MA had elevated PRL. T2 MRI intensity: 64% MA iso-, 75% SGA hyper- (p=0.026), and DGA 29% hypo-, 42% iso-, 29% hyper-. SSTR2A+ staining: 100% MA and DGA, but 64% SGA (p=0.025). Surgical remission rates were similar in all groups. 21 pts with persistent disease treated with SRL: MA and DGA were more responsive than SGA (44%, 78%, 0%; p=0.025).
Discussion: In this treatment naïve cohort of pts with A, ⅔ of MA were T2 MRI isointense, while majority of SGA were hyperintense. MA and DGA responded to 1stgen-SRLs better than SGA. Despite seemingly lower response of MA vs DGA, the difference was not significant likely due to small sample size. While mean PRL level was higher in MA, only ⅓ of pts had elevated PRL.
Conclusion: GH-secreting tumors represent a heterogeneous group and their careful pathological classification is required. Diagnosis of mixed GH/PRL adenoma can be considered in pts with T2 isointense MRI even in the absence of high PRL at baseline. Clinical features and SSTR2A+ status of MAs seem to resemble DGA, however, only half of MA responded to adjuvant 1st gen-SRLs, suggesting that combination therapy might be needed earlier in the course. Further research for both specific and sensitive predictive markers that can be used for individualized treatment options is needed.
Disclosure: MF: Principal Investigator, Novartis Pharmaceuticals, Ad Hoc Consultant, Novartis Pharmaceuticals, Principal Investigator, Chiasma, Ad Hoc Consultant, Chiasma, Principal Investigator, Pfizer Global R&D, Ad Hoc Consultant, Pfizer Global R&D. Nothing to Disclose: EV, FL, JP, JK, DSTL, CGY, JSC