Published in:
01-11-2018 | Original Article
Performance of 111In-labelled PSMA ligand in patients with nodal metastatic prostate cancer: correlation between tracer uptake and histopathology from lymphadenectomy
Published in: European Journal of Nuclear Medicine and Molecular Imaging | Issue 12/2018
Login to get accessAbstract
Purpose
Intraoperative identification of lymph node (LN) metastases (LNM) detected on preoperative PSMA PET/CT may be facilitated by PSMA radioguided surgery with the use of a gamma probe. We evaluated the uptake of 111In-labelled PSMA ligand DKFZ-617 (referred to as 111In-PSMA-617) in unaffected LN and LNM at the level of single LN.
Methods
Six patients with prostate cancer (PCa) with suspicion of LNM on preoperative PSMA PET/CT underwent 111In-PSMA-617-guided lymphadenectomy (LA; four salvage LA and two primary LA). 111In-PSMA-617 (109 ± 5 MBq). was injected Intravenously 48 h prior to surgery Template LAs were performed in small subregions: common, external, obturator and internal iliac vessels, and presacral and retroperitoneal subregions (n = 4). Samples from each subregion were isolated aiming at the level of single LN. Uptake was measured ex situ using a germanium detector. Receiver operating characteristic (ROC) analysis was performed based on 111In-PSMA-617 uptake expressed as standardized uptake values normalized to lean body mass (SUL).
Results
Overall 310 LN (mean 52 ± 19.7) were removed from 74 subregions (mean 12 ± 3.7). Of the 310 LN, 35 turned out to be LNM on histopathology. Separation of the samples from all subregions resulted in 318 single specimens: 182 PCa-negative LN samples with 275 LN, 35 single LNM samples, 3 non-nodal PCa tissue samples and 98 fibrofatty tissue samples. The median SULs of nonaffected LN (0.16) and affected LN (13.2) were significantly different (p < 0.0001). Based on 38 tumour-containing and 182 tumour-free specimens, ROC analysis revealed an area under the curve of 0.976 (95% CI 0.95–1.00, p < 0.0001). Using a SUL cut-off value of 1.136, sensitivity, specificity, positive predictive value, negative predictive value and accuracy in discriminating affected from nonaffected LN were 92.1% (35/38), 98.9% (180/182), 94.6% (35/37), 98.4% (180/183) and 97.7% (215/220), respectively.
Conclusion
Ex situ analysis at the level of single LN showed that 111In-PSMA-617 had excellent ability to discriminate between affected and nonaffected LN in our patients with PCa. This tracer characteristic is a prerequisite for in vivo real-time measurements during surgery.