Recurrent parathyroid carcinoma and the need for molecular profiling to aid in prognostication and therapeutics.

Abstract

e17012 Background: Parathyroid carcinoma (PC) is rare, has high morbidity and poor overall survival. Surgery is the mainstay of treatment but recurrence is common. There are no known therapeutic options and efforts to ascertain prognostic factors are restricted by rarity of disease. We report the largest single institution experience and aim to identify prognostic clinicopathologic features associated with aggressiveness of disease. Methods: Retrospective review of PC patients from 1981 to 2014 at a tertiary cancer center was done. Those with a final pathology diagnosis of carcinoma, atypia or neoplasm on were included. Clinicopathological data was analyzed for predictors of disease recurrence. Limited molecular profiling on a subset of patients was done. Cox regression was used for univariate analysis with p ≤ 0.05 considered significant. Results: 87 patients met inclusion criteria. Median follow up was 5 yrs [IQR 3-8]. 36 patients (41%) underwent local resection, while 48 (55%) had en bloc resection. 30(35%) developed recurrent or persistent disease with > 85% requiring ≥ 3 surgeries for palliation or cure. Univariate analysis of clinicopathological data showed only mitotic index (MI) of ≥ 1 was predictive of disease recurrence (PFS) (p = 0.03) with a trend towards decreased overall survival (OS) (p = 0.06). Extent of initial surgery had no bearing on PFS or OS. 8 patients with recurrent disease had molecular profiling. 2 had no discernible mutations, 2 had TP53 mutation. Others mutations identified were CDC73, CTNNB1, KDM5C, NF1 and TLR4. Conclusions: High MI is prognostic of disease recurrence in PC. Radical resection at initial surgery did not improve PFS. Other clinicopathologic features of PC fail to precisely identify risks for recurrent disease. In the subset of patients with limited therapeutic options, genomic profiling is needed to identify prognostic and targetable mutations and drive research aimed at creating novel therapeutics. Variables Values median [IQR] Age (yr) 58 (47-66) Gender n(%) Male 34(39) Female 53(61) Final Pathology n(%) Carcinoma 56(64) Neoplasm 17(20) Atypia 10(12) PTH at diagnosis (pg/ml) 381 [203-1536] Ca+ at diagnosis (mg/dl) 12.9 [12-14.5] 10 year OS 67% 10 year PFS 56%