SAT-455 Analysis of a Group of Prolactinomas Resistant to Treatment with Dopamine Agonists at a University Hospital in Buenos Aires, Argentina

Prolactinomas represent approximately 40-60% of all pituitary tumors. Dopamine agonists (DA) are the first-line treatment, with a 70-90% response. However, up to 10% are resistant (R) to DA. Our objectives were to describe the clinical, biochemical and imaging characteristics of a group of patients with prolactinomas R to DA; and to determine the presence of prognostic factors by performing a comparative analysis with the population of non-R prolactinomas. We analized retrospectively the medical records of patients diagnosed with prolactinoma, with follow-up (f-up) at the same hospital from 1998 to 2017. They were classified as R when prolactin (PRL) was not normalized and/or a reduction ≤ than 50% of the tumor (Tx) volumen, with a cabergoline (CBG) dose ≥2mg/week (mg/w), with, at least 6 months of treatment. Macroadenomas (MA) were defined as Tx ≥10mm and as invasive with a classification of Knosp 3 or 4 by MRI. Of 95 prolactinomas, 14 patients with R to DA were described. The mean age was 32.7 ± 15.6 years (y). We found no statistically significant differences (SSD) in terms of gender, with a higher distribution (53%) and a tendency to age (37.9 vs 26.9, p = 0.09) in the male gender (M). The mean time of f-up was 34.7 ± 28.7 months (m). Majority were MA (87%) and invasive Tx (64%). The initial manifestations were: headache (85.7%), visual field alteration (50%), menstrual cycle alterations (86%), galactorrhea (20%). The mean PRL value was 7419.4 ± 1149.7 ng/ml, being higher in the M group (11652.6 vs 1653.3, p <0.001). 93% of the patients received CBG as the main treatment. The mean dose was 6.8 ± 4.3 mg/w, with a mean time of 34.3 ± 23.1 m. 57% of the patients achieved a reduction ≥ 50% of PRL, 2 normalized it. 4 patients presented ≥ 50% reduction in Tx size. All were MA, invasive, none normalized PRL, in 3/4 surgical resolution was indicated. When comparing with the population of non-R prolactinomas (n = 81), we found a higher proportion of M with a statistical tendency in the R group (p=0.053) and a greater proportion of MA (p=0.04) and invasive (p = 0.03). Higher values ​​of PRL were evident at diagnosis in this group (508.9 vs 7482.5, p <0.0001), higher dose of CBG (1.8 vs 6.6, p <0.001), but without SSD in the time of use of DA, the age at diagnosis or the time of f-up. A value of PRL at diagnosis above 1065 ng/mL has a specificity of 85% for R. Patients in the non-R group had a biochemical response ≥ 50% with the treatment (p<0.001), but not in terms of the structural response (p=0.13) nor in the biochemical response time (p=0.12). Similar to that reported in the literature, we evidenced a frequency of 15% of R in our population. In contrast to classic prolactinomas, a greater proportion of R Tx was found in the M gender, in MA and with invasive characteristics. We can conclude that larger Tx with a higher PRL value at the time of diagnosis and, probably, the M gender are more likely to present R to DA