Bold, Michael S.

5577 Background: Prostate specific antigen (PSA) has well-recognized limitations as a marker for treatment response and disease progression. A post hoc analysis of the PREVAIL trial reported 24.5% of chemotherapy naïve mCRPC patients on enzalutamide had radiographic progression on conventional imaging with non-rising PSA. In this study, we sought to retrospectively compare PSA levels with C-11 choline positron emission tomography/ computed tomography (PET/CT) images in patients with m-CRPC on 2nd-HT with prior use of chemotherapy. Methods: We identified 123 patients with mCRPC on 2nd-HT following prior use of docetaxel chemotherapy (Abiraterone, n = 106; Enzalutamide, n = 17). Patients underwent serial PSA testing and C-11 choline PET/ CTs every 3–6 months. Disease progression was defined by the increase in blood pool corrected maximum standardized uptake value (SUVmax) of the index lesion on C-11 choline PET/CT scan. Suspicious lesions were confirmed by biopsy and/or conventional imaging. Results: Approximately 43% (n = 53) of patients had radiographic disease progression while on 2nd-HT. At time of radiographic progression, 60.4% of patients showed a parallel rise in PSA (Group-A), while 39.6% showed a paradoxical response; defined as radiographic progression with stable or down-trending PSA (Group-B). Median PSA at time of progression was 3.1 ng/ml for Group-A, and 1.3 ng/ml for Group-B (p-value = 0.0176). Median SUVmax was the same (4.9 Group-A, 4.6 Group-B; p-value = 0.6072). Bone-predominance progression was more significant in Group-B (90%) versus Group-A (65%) (p-value = 0.0309). The median time for radiographic progression was 9.5 months versus 3.9 months for Group-A and Group-B, respectively (Log-Rank = 0.0063). Conclusions: Metabolic imaging is a useful tool that should complement PSA in the evaluation of treatment response and disease progression in mCRPC patients on 2nd-HT, especially considering the paradoxical response observed in our data.

144 Background: Despite having well-recognized limitations, urologists often rely on serial PSA testing as a marker for treatment response or disease progression. To determine if PSA was indeed a reliable marker for treatment response or disease progression, we compared PSA levels against C-11 choline PET/CT in the evaluation of patients with advanced prostate cancer treated with second generation hormonal therapy (2nd-HT). Methods: We retrospectively identified 239 patients who were undergoing treatment with 2nd-HT (enzalutamide or abiraterone) for advanced prostate cancer. While on treatment, patients underwent serial PSA testing and C-11 choline PET/ CTs every 3 – 6 months. Paradoxical response was defined as increasing blood pool-corrected SUVmax of known choline-avid lesions and/or identification of new choline-avid lesions, despite stable or down-trending PSA. Results: Median (IQR) age was 70.4(64.3 – 75.7) years and median (IQR) primary Gleason Score was 8 (7 – 9). In our study, 19% of patients (n = 46/239) who were receiving 2nd-HT exhibited paradoxical response. Median (IQR) PSA and corrected SUVmax at baseline evaluation were 1.3 ng/mL (0.3 – 12.8 ng/mL) and 3.5 (1.8 – 5.8), respectively. Median (IQR) PSA and corrected SUVmax at the time of paradoxical response were 0.4 ng/mL (0.1 – 5.4 ng/mL) and 4.5 (2.8 – 6.8), respectively. The median duration of 2nd-HT treatment prior to detection of paradoxical response was 4.8 months (2.9 – 10.1 months). No significant difference was noted between patients receiving enzalutamide versus abiraterone (p = 0.35). Independent predictors of paradoxical response were prior primary systemic treatment (i.e. hormonal/chemo-hormonal therapy versus local therapy) and patient’s age at time of 2nd-HT initiation on univariate and multivariate analysis. Conclusions: Our retrospective review demonstrated prostate cancer disease progression discordant with PSA down-trending in 19% of patients receiving 2nd-HT. We conclude that in this subset of patients with advanced prostate cancer, PSA may not be a reliable marker of treatment response of disease progression, and routine radiographic evaluation in these patients is warranted.

Abstract Erdheim-Chester disease (ECD) is a rare histiocytosis of the “L” (Langerhans) group with multisystem involvement that can affect the large and medium-sized arteries mimicking vasculitis. Aortic involvement is common but the frequency and outcome of aortic branch vessel abnormalities are less well described. Patients with ECD were retrospectively identified. Images containing information of arterial involvement within 6 months of diagnosis were considered baseline and compared to last follow-up studies. Two physicians independently reviewed the studies to evaluate for presence of abnormalities attributable to ECD. Age and sex-adjusted logistic regression models were used to examine associations between patient characteristics and vessel involvement at baseline. Among a cohort of 64 patients with ECD, 63 had baseline imaging of vascular structures. ECD involvement of at least 1 segment of the aorta was observed in 56%. Abnormalities were also observed in aortic arch branches (26%), visceral branch arteries (40%), iliofemoral arteries (31%), coronary (5%), and pulmonary (3%) arteries. Perinephric fibrosis was strongly associated with the identification of abnormalities in the thoracic aorta (OR 4.92 [1.54, 15.75]; P = .007), abdominal aorta (OR 7.57 [2.28, 25.07]; P = .001) and visceral branch arteries (OR 6.05 [1.52, 24.03]; P = .01) but not pelvic/lower extremity arteries. Complete normalization of arterial abnormalities at follow-up was only observed in 9% or less of arterial segments involved at baseline. Aortic and aortic branch vessel abnormalities are frequently observed in patients with ECD and are often asymptomatic. Partial and/or complete resolution of arterial findings is uncommon.