ASCO 2019: PSMA: Is Seeing Believing?
The prostate-specific membrane antigen physiologically in the gastrointestinal tract acts as a folate hydrolase (releasing folate in the proximal small intestine for nutrition), as a NAALADase in the central nervous system (required for neurotransmitter homeostasis), and it is also present in the renal tubules, salivary glands, and normal prostate. Pathologically, prostate cancer cells overexpress PSMA (100-1000-fold overexpression), and expression also increases in high-grade tumors, castrate-resistant state, and in metastases. But, Dr. Williams notes that PSMA is also expressed in the neovasculature of various benign and malignant conditions.
The clinical trial testing 18F-fluciclovine and 68Ga-PSMA-11 PET/CT in patients with early biochemical recurrence of prostate cancer found that detection rates were significantly lower with 18F-fluciclovine PET/CT than with 68Ga-PSMA-11 PET/CT per-patient (26% vs 56%; p = 0.003) and per-region for pelvic nodes (N) (8% vs 30%; p = 0.003) or any extra-pelvic lesions (M) (0% vs 16%; p = 0.008). Furthermore, agreement for 68Ga-PSMA-11 PET/CT interpretation was significantly higher than for 18F-fluciclovine PET/CT (0.67 vs 0.20, p = 0.002). Dr. Williams agrees that PSMA-PET should be the standard for restaging recurrent prostate cancer, especially at PSA <1.0, however he notes that metabolic imaging may still be useful.
The Osprey cohort B assessed 18F-DCFPyL PET/CT imaging in patients with known or suspected metastatic prostate cancer. This study found that the sensitivity and PPV of 18F-DCFPyL PET/CT as compared to histopathology ranged from 92.9-98.6% (lower bound of 95% CI: 84.0-91.6%) and 81.2-87.8%, respectively. Dr. Williams notes that the PPV for all patients is high (81.9%), but it’s not perfect. Furthermore, he notes that sensitivity is near perfect, but this is not sensitivity as we know it since negative areas obviously aren’t sampled in these patients. This suggests that PSMA will only rarely fail to show a lesion that has positive histology; in Dr. Williams’ opinion, DCFPyL can adequately replace conventional imaging.
The Cornell experience with PSMA-targeted radionuclide therapy showed that high PSMA expression was associated with more frequent prostate-specific antigen (PSA) decline (≥30%: 38.9 vs 16.4% p=0.002; ≥50%: 26.9 vs 7.3% p=0.002). When controlling for dose level, this association remained significant for low (≥30%: 22 vs 5.9% p=0.04) and high doses of radionuclide therapy (≥50%: 36.5 vs 9.5% p=0.02). Finally, 13 (6%) patients with no PSMA uptake (VS=0) had PSA declines. Dr. Williams highlighted that if the target is absent, the chance of >50% PSA decline is only 7.3%, whereas if the target is present, >50% PSA decline is almost 4-fold higher.
So, is seeing believing? Yes, but seeing is not knowing. We have ample documentation of the performance characteristics of a scanning modality that is clearly better than conventional imaging. Furthermore, Dr. Williams feels that we need to move away from “lumpology” and move towards biology: 98% of M0CRPC are PSMA positive, but what does this mean?
Dr. Williams concluded noting that the “tsunami of PSMA imaging will strike” and the majority of patients that have a PSMA PET/CT will have their management changed. He cautions that we should (i) first do no harm, (ii) don’t ignore standard of care, build on it, (iii) if in doubt, observe and rescan, and (iv) don’t create an industry around treating “lumps”.
Presented by: Scott Williams, BSc, MBBS, MD, FRANZCR, Peter MacCallum Cancer Centre, Melbourne, Australia
Written by: Zachary Klaassen, MD, MSc, Assistant Professor of Urology, Georgia Cancer Center, Augusta University/Medical College of Georgia, @zklaassen_md at the 2019 ASCO Annual Meeting #ASCO19, May 31- June 4, 2019, Chicago, IL USA